[PDF] Language: its structure and use - Download PDF for free (2023)

Download language: its structure and use...

language their structure and use

Language, Its Structure and Use FIFTH EDITION

EDWARD FINEGAN University of Southern California

Australia Brazil Canada Mexico Singapore Spain United Kingdom United States

Language: Its Structure and Use, Fifth Edition Edward Finegan Publisher: Michael Rosenberg Development Lead Editor: Karen Judd Development Editor: Mary Beth Walden Editorial Assistant: Megan Garvey Marketing Director: Kate Edwards Marketing Assistant: Kate Remsberg Marketing Communications Director: Heather Baxley Content Design Manager: Sarah Sherman Senior Art Director : Cate Rickard Barr Senior Print Buyer: Betsy Donaghey

Acquisitions Account Manager, Text: Mardell Glinski Schultz Permissions Researcher: Sue Howard Production Service: Lachina Publishing Services Text Designer: Brian Salisbury Sr. Permissions Account Manager, Images: Sheri Blaney Cover Designer: Gopa & Ted2, Inc Cover Photo: Original art by © Werner Hoeflich, Untitled (Hedge Series) 2003 Printer: West Group

Thomson Higher Education 25 Thomson Place Boston, MA 02210-1202 USA

ALL RIGHTS RESERVED. No part of this copyrighted work contained herein may be reproduced or used in any form or by any means (graphical, electronic or mechanical, including photocopying, recording, recording, web distribution, information storage and retrieval systems or otherwise). . without the written permission of the publisher. Printed in the United States of America 1 2 3 4 5 6 7 09 08 07 Library of Congress Control Number: 2007922714 ISBN-10: 1-4130-3055-6 ISBN-13: 978-1-4130-3055-6

For more information about our products please contact us: Thomson Learning Academic Resource Center 1-800-423-0563 To obtain permission to use material from this text or product, make a request online at http:// www.thomsonrights.com Permissions questions can be emailed[email protected]

Brief Table of Contents Preface xxiii Chapter 1 Languages and Linguistics

P A R T O N E Estrutura da linguagem Kap that lo Kap that lo Kap that lo Kap that lo Kap that lo Kap that lo

2 3 4 5 6 7

Chapter 1 1 Chapter 1 2

Words and their parts: lexicon and morphology 34 The sounds of languages: phonetics 75 Sound systems of language: phonology 105 The structure and function of phrases and sentences: syntax The study of meaning: semantics 172 Universals of language and typology of language 214

PART TWO Language use Chapter 8 Chapter 9 Chapter 1 0

139

247

Information structure and pragmatics 248 Speech and conversational acts 281 Linguistic variation in usage situations: registers and styles 313 Linguistic variation between social groups: dialects 346 Writing 394

PART THREE Language change, language development and language acquisition 419 Chapter 13 Chapter 1 4 Chapter 1 5

Language Change Over Time: Historical Linguistics Historical Developments in English 467 First and Second Language Acquisition 501

420

Glossary 531 Index 551 Language Index 565 Website Index 569 Video Index 571 Imprint 573

This page has been left blank internationally

Contents in detail Foreword

XXIII

Chapter 1

Languages and Linguistics

How many languages are there in the world?

Does the United States have an official language? English Only, English Plus, Multilingualism 4 What is human language? 5 Three faces of a language system Language: mental and social 6

Signs: Arbitrary and Non-Arbitrary 7 Arbitrary Signs 7 Representative Signs 7 Language: A System of Arbitrary Sign Languages as Patterned Structures Discrimination 9 Duality 9 Displacement 9 Productivity 10

Speaking as a standardized use of language

The origin of languages: from Babel to Babble

Languages and dialects 12 What are social dialects? 13 different dialects or different languages? 14 What is a standard grade? 14 Are there good and bad things about using the English language? Language Means of communication Speaking 16 Writing 17 Characters 17

fifteen

sixteen

Do only humans have a language? 19 How animals communicate in their natural environment

VIII

v i i i • Contents in detail Can chimpanzees learn human language? Projeto Nim 21 What is Linguistics? 22 What are the branches of linguistics? Continue

Exercises 26 Especially for Educators and Future Teachers Additional Resources Internet 29 Video 30

Further reading references Further reading references

PREVIOUS CHAPTER 2

linguistic structures

33 words and their parts: lexicon and morphology

Introduction: Words are tangible

What does it mean to know a word?

Lexical categories (parts of words) 35 How to recognize lexical categories 36 Verbs 37 Nouns 37 Adjectives 38 Pronouns 38 Determiners 39 Prepositions and postpositions 39 Adverbs 39 Conjunctions 40 Morphemes are parts of words that have meaning Morphemes can be free or connected 41 The morphemes are derived from in other words 42 inflectional morphemes 43

How are morphemes arranged in words? Morphemes are arranged in sequence 43 Morphemes can be discontinuous 44

• i x portmanteau words contain mixed morphemes morphemes overlap words 45 How do you expand the vocabulary of a language? Some classes of words are open, others closed How to derive new words 46 Compounds 47 Abbreviations 48 Additional formation 49 Functional conversion or modification 49 Semantic modification 50 Word borrowing 50 Word invention 52

46 46

What kinds of morphological systems do languages have? Isolation morphology 53 Agglutination morphology 54 Inflection morphology 54 Pronunciation variants of a morpheme: summary of the allomorphs

Exercises 64 Based on English 64 Based on Languages Other Than English 68 Specific for Educators and Future Teachers 71 Other Resources

Further reading references Further reading references

Chapter 3

The sounds of languages: phonetics

Sounds and Spelling: Not Equal Same Spelling, Different Pronunciation Same Pronunciation, Different Spelling Why and Why Sounds/Spelling Phonetics: The Study of Sounds Phonetic Alphabets 80 The Vocal Tract 80

76 76 77 Deviations

x • Contents in detail Sound description 83 Voice 83 Articulation style 83 Articulation points 84 Consonant sounds 85 Registers 85 Fricatives 86 Africans 87 Obstructives 87 Approximants 87 Nasal sounds 88 Clicks, keys, trills 89 Vowels 89 Tonals and front vowels 90 Diphthongs 91 Summary of other vowel articulation features

98 English Exercises 98 Especially for Educators and Aspiring Teachers More Resources

103

Suggestions for further reading

Chapter 4

102

103

Language Sound Systems: Phonology

Introduction: Sounds in the head 106 Phonemes and allophones 107 Distribution of allophones 109 Phonological rules and their structures 112 Generalization of phonological rules 114 Natural classes of sounds 115 Underlying forms 116 Ordering rules 116 Syllables and restrictions on syllable structure order 120 Sniglets 121

119

105

Contents in detail • x i stress

122

Syllables and accents in phonological processes

122

The interplay of morphology and phonology 123 English plural, possessive and third person singular morphemes 123 English past tense morphemes 124 Underlying phonological form of morphemes in the lexicon 125 Summary

128

Exercises 130 Based on English 130 Based on Languages Other Than English 132 Especially for Educators and Future Teachers 136 Other Resources

137

Further reading references Further reading references

138

Chapter 5 Introduction

137

The Structure and Function of Phrases and Sentences: Syntax 139

140

Constituencies and tree diagrams constituency tree diagrams 141 constituency 142

141

Main clause components: Noun phrases and verb phrases Noun phrases and verb phrases 145 Active and passive phrases 146 Sentence structure rules 147 Expensive noun phrases 147 Prepositional expansive phrases 148 Phrases and verb expansions 148 Sentence structure rules and tree diagrams 149 Grammatical relationships: subject, direct object and other immediate dominance 151 Subject and direct object 151 Grammatical relationships 152 Passive sentences and structural dependency 152

151

144

x i i • Contents in detail Surface structures and underlying structures Syntactic operations 153 Formation of questions 153 Embedded clauses 156 Subordinate clauses 156 Relative clauses 157 COMP nodes 158 Types of syntactic operations

160

Summary of the functions of the syntactic operations

153

160

163

Homework 164 in English 165 Non-English 168 Especially for Educators and Aspiring Teachers 169 Suggested Reading Immersive Reading References

170

171

Chapter 6

The study of meaning: semantics

Introduction 173 What is meaning?

173

Linguistic, social and affective meaning Linguistic meaning 174 Social meaning 175 Affective meaning 176

174

Meaning of words, phrases and statements 177 Meaning of words and phrases 177 Meaning of statements 178 Lexical semantics 179 Semantic fields 180 Hyponymy 181 Part/whole relationships 183 Synonymy 184 Antonyms 185 Inverse 186 Polysemy and homonymy 187

172

S e m o d e tail • x i i i Metaphors 188 Lexical semantics: discovering relationships in the lexicon Function words and categories of tense and modality of meaning 191 Reference 192 Deixis 193 Semantic roles and meaning of sentences

191

197

Semantic functions and summary of grammatical relationships

190

200

204

Exercises 206 Based on English 206 Based on English and Other Languages 209 Special for Educators and Future Teachers 211 Other Resources

211

Further reading references Further reading references

212

Chapter 7

Linguistic universals and linguistic typology

Similarity and diversity between languages Why discover universals? 215 Language Types 216 Universal Semantic Pronouns 218

215

217

Phonological universals 221 Vowel systems 221 Nasal and oral vowels 223 Consonants 223 Syntactic and morphological universals 224 Word order 224 Possessive and possessive noun phrases 226 Prepositions and postpositions 226 Relative clauses 227 General patterns of order 227 Relativization hierarchy 228

214

x i v • Contents in detail Types of language Universals 230 Absolute universals and tendencies universals Implictive and non-implicative universals

230 231

Statements on language universals 231 Source language hypotheses 231 Universals and perception 232 Acquisition and processing explanations 232 Social explanations 233 Summary

236

238 Exercises in English and Other Languages 238 Especially for Educators and Aspiring Teachers 244 Suggestions for Further Reading Immersive Reading References

244

245

PART TWO Using Language 247 Chapter 8 Information Structure and Pragmatics Introduction: Coding information structure

249

Categories of information structure 250 Information given and new information 251 Topics 253 Contrast 254 Specific expressions 255 Reference expressions 257 General and specific expressions 257 Categories of information structure 258 Information structure: intonation, morphology, syntax New information accent 259 Morphemes of information structure 259 Front 260 Left shift 261 Slots and WH slots 262 Passives 264 Word order 266

258

248

Contents in detail • x v The relationship of propositions to discourse: summary of pragmatics

267

270

Exercises 272 Based on English 272 Based on Languages Other Than English 275 Especially for Educators and Future Teachers 278 Other Resources

278

Further reading references Further reading references

279

Chapter 9 Language used

Speech acts and conversation 282

Sentence structure and function of sentences

282

Speech acts 283 Types of speech acts 284 Idioms and illocutions 284 Distinguishing between speech acts 285 Appropriateness conditions and successful utterances

289

Courtesy 291 Respect independence and show attendance at public speaking events

281

292

The organization of the conversation 293 turns and pauses 293 adjacent pairs 296 opening sequences 298 closing sequences 300 conversation patterns 301

292

285

xvi • Corrected content in detail 301 Courtesy: An organizing force in conversation Summary of intercultural communication

302

303

304

305 English-Based Exercises 305 Especially for Educators and Future Teachers Additional Resources

310

Further reading references Further reading references

309

310

311

Chapter 1 0 Language Variations in Usage Situations: Registers and Styles 313 Introduction

314

Language varies within a language community 314 Choice of language in multilingual societies 314 Linguistic repertoires in Brussels, Tehran and Los Angeles Changing varieties within a language 315 Language situations 316 Elements of a language situation

316

Records in monolingual societies

318

Style markers 319 Lexical style and register markers 319 Phonological style and register markers Grammatical register markers 325 Semantic register markers 326

314

322

Similarities and differences between oral and written registers Two registers compared 329 Lexicon and grammar 330 Phonology 333 Register comparison 334

327

Summary in detail • x vi i Summary

335

Exercises 337 Based on English 337 Based on English and Other Languages 342 Especially for Educators and Future Teachers 342 Other Resources

342

Further reading references Further reading references

343

344

Chapter 11

Linguistic variation between social groups: dialects 346

Language or dialect: which one do you speak? 347 Social boundaries and dialects 347 Distinguishing between dialect, pitch and accent How do languages diverge and merge? 348 Linguistic fusion in an Indian village 348 Language/dialect continued 349 National varieties of English 350 American and British national varieties

350

American English Regional Varieties 352 Dialect Mapping 353 Dialect Boundaries 356 Regional American English Dictionary 358 North American English Vowel Fusion Atlas 362 Vowel Shifts 363 ANAE Discoveries 365

362

Ethnic variants of American English 366 African American English 367 Chicano English 369 Ethnic variants and social identification

370

347

x vi i i • Contents in detail Varieties at the socioeconomic level: English, French and Spanish New York City 371 Norwich, England 374 Montreal, Canada 375 Argentina 375 General remarks 376 The linguistic diversity of women and men Why do stigmatized varieties persist? Continue

377

379

381

382 Exercises in English 382 Especially for educators and future teachers

385

Other Resources 387 Internet 387 Video and Audio 388 Suggested Further Reading References on immersive reading

389

390

391

Chapter 1 2 Introduction to Writing

394

395

The historical development of writing 395 The leap from image to writing 396 Writing systems 398 Syllable writing 398 Logographic writing 402 Alphabetical writing and orthography 408 Writing development in newly formed societies Summary

413

Exercises 415 Especially for Educators and Trainee Teachers Other Resources

416

411

371

S ente in D e ta ls • x i x Suggested reading for advanced readers

417

Language change, language development and language acquisition 419

PART THREE

Chapter 13 Language Change Over Time: Historical Linguistics 420 Are living languages always changing?

421

Language families and the Indo-European family Reconstruction of the linguistic past 424 Polynesian and Pacific background 424 Polynesian languages and their history 426 Comparative reconstruction 428 Protopolynesian vocabulary reconstruction Historical linguistics and prehistory 434

422

431

What are the language families of the world? 434 Speaker numbers and languages 434 The Indo-European family 436 The Sino-Tibetan family 442 The Austronesian family 443 The Afroasiatic family 443 The three major language families of Sub-Saharan Africa Other language families of Asia and Europe 447 Native American Languages 449 Aboriginal languages of Australia 451 Papuan Languages 451 Nostratic macro family 451 Contact languages 452 Multilingualism 452 Summary

457

459 Non-English Language Exercises 459 Especially for Educators and Future Teachers 462

446

x x • Content in detail Other resources Video 463

463

Further reading references Further reading references

464

465

Chapter 1 4 Historical development in English A thousand years of change

467

468

Where does English come from? 470 English is a Germanic language 470 Morphology and syntax in the Indo-European periods of English history 473

472

Old English: 700–1100 473 Old English Script 473 Old English Sounds 474 Old English Vocabulary and Morphology 475 Old English Inflections and Word Order 480 Angelic Companions: An Old English Narrative 481 Vocabulary in the Narrative 481 Grammar: Syntax and Morphology in the Narrative 483 Narrative Text Structure 483 Middle English: 1100– 1500 483 The Norman invasion 483 Middle English vocabulary 484 Middle English sounds 484 Middle English inflections 485 Middle English word order 487 Where do men and women go Nus: a Middle English travel fable Vocabulary in the fable 487 Morphology in the fable 489 Syntax in the fable 489 Modern English: 1500–present 490 Early and Late Modern English 490 Phonology: Vowel Shifts in English 490 Morphology of Modern English 491

487

Contents in Detail • x x i Modern English Word Order 492 Modern English Vocabulary 493 Summary

494

exercises

495

Other Resources 498 Internet 498 Video and Audio 499 Suggested Further Reading Advanced Reading

499

500

Chapter 1 5 First and Second Language Acquisition Introduction

502

First language acquisition 503 Principles of language acquisition 504 Adult contributions to language acquisition 505 Stages of language acquisition 507 How do children acquire morphology and grammar? 510 How quickly do children acquire vocabulary? 512 How do children acquire the sounds of language? 512 How do researchers study language acquisition? . when learning a second language 522 Summary

524

Exercises 525 Specific for Educators and Future Teachers Other Resources 527 Internet 527 Videos 528

527

501

x x i i • Set of Details Suggested Reading Immersive Reading References Glossary Index

529

530

531

551

language index

565

Internet-Site-Index Video-Index Credits

573

571

569

528

Preface A Special Word for Students For hundreds, even thousands, of years, philosophers, rhetoricians, and grammarians have analyzed the use of language in their daily lives and the linguistic and social structures that underlie that use. The 19th and 20th centuries proved rich in linguistic discoveries, as first philologists, and then linguists and cognitive scientists, broadened and deepened our understanding of language as a uniquely human trait. Over the past few decades, as space explorers probed our views of the Uranus satellites and microbiologists probed the details of DNA in the Human Genome Project, linguists have also generated an explosion of knowledge about the representation of language in the mind and the interactions between language, use, and social structures of the community. Despite the impressive speed at which researchers have been gaining insights into human language, important questions remain unanswered and many areas remain unexplored or unexplored. There is much more to be discovered about language than is known today, and today's college students, tomorrow's researchers, continue to do much intellectually stimulating and socially useful work. For those of you who wish to contribute to our understanding of the human mind or human social interaction, rest assured that what you know now will dwarf what you and others like you have in your life discover. For you and those who just want to understand what we know about language, this book offers an invitation to dive in and ask your own questions about language and its role in your life and the lives of those around you. One of the helpful proverbs I heard in school was, "One stitch in time saves nine," and another warned, "Look before you jump." On the other hand, one proverb turned out to be wrong: the one that said: "Sticks and stones may break my bones, but words will never hurt me." Most of us learn early in life how powerful a language tool is and how it can be used for good or ill. Language can enlighten and delight us, but it can also do harm, and language is as central to our social interactions as it is to our cognitive activities. They want to learn as much about it as possible. If you read Language: Its Structure and Usage, or LISU for short, you will occasionally see words in bold. When an important concept is discussed for the first time (not necessarily the first time it is mentioned), the relevant term appears in bold to emphasize its importance and alert you to its appearance in the glossary. To learn more about the topics that interest you, see Recommended Additional Reading in each chapter. You will also find lists of videos and internet addresses. For more information, visit the LISU website at http://www.thomsonedu.com/english/finegan.

A word for LISU teacher encompasses more chapters than can be covered in a semester-long course. Typically, Instructors will cover the first six chapters and then repeat the rest as per xxiii

xxiii • Foreword

needs and interests. In this book, the chapter on morphology takes precedence over those dealing with phonetics and phonology. This organization is successful partly because beginners find words more tangible and accessible than sounds, and partly because morphology can be discussed without phonetic symbols, whose strangeness on the portal can be daunting. The sequence of chapters invites teachers to teach morphology before phonology, but to teach phonetics and phonology before morphology, simply defer the section “The Interplay of Morphology and Phonology” (pages 123-127) until you have completed the Morphology chapter to have. Each chapter contains sections on computers and languages, the Internet and other resources, and separate exercises for English and other languages. With the aim of involving the students in everyday situations in which the content of a chapter can play a role, each chapter begins with a few puzzles under the heading "What do you think?". Then, before the exercises at the end of the chapter, there are brief answers to the riddles in What Do You Think? Checked. Some of your students might benefit from pondering the puzzles and reading the Revised section before studying the chapter. I also tried to make LISU more interactive in other ways. The Try It Yourself sections directly apply what has just been explained in the text and encourage students to check their own understanding. The "Educator and Prospective Teacher" exercises may be of particular interest to your intended audience, but most of these exercises will be useful to other students as well.

A Word on Phonetic Transcription Custom in the United States favors a modified version of the International Phonetic Alphabet, and the considerable variation in published sources and on the Internet makes it desirable for students to recognize the need to identify exactly what the represent symbols. To avoid this problem, many linguists prefer the IPA more or less strictly, and from time to time the IPA itself changes. As in all linguistic matters, rule gives way to practice. While I generally prefer IPA symbols, having introduced them in Chapter 3, I sometimes use alternate symbols thereafter and state what the symbols represent. I hope this will better prepare students for real-world practice, including using dictionaries together.

Workbook and Answer Key For the fifth edition of LISU, Paul Frommer and I have prepared a fourth edition of "Languages: A Workbook in Elementary Linguistics". The workbook is useful to help students review, apply, and expand on the basic concepts. The spoken language files that accompany many of the workbook exercises are available on the LISU website: http://www.thomsonedu.com/english/finegan. This book and the workbook have separate answer keys that are only available to publisher faculty.

Foreword • xxiv

Acknowledgments I have drawn on many scholars whose work has provided a basis for the questions raised here. The references in each chapter show the spectrum of science to which I drew, and I am also indebted to many whose work is not cited. I thank readers of previous editions for helpful comments: Michael Adams and his students, John Algeo, Joseph Aoun, Anthony Aristar, Dwight Atkinson, Robin Belvin, Douglas Biber, Betty Birner, Dede Boden, Larry Bouton, Leger Brosnahan, William Brown, Paul Bruthiaux, Ron Butters and his students, Dani Byrd, Steve Chandler, Bernard Comrie, Jeff Connor-Linton, Janet Cowal, Marianne Cooley, Carlo Coppola, Nicole Dehé, John Dienhart, David Dineen, Alessandro Duranti, Paul Fallon, Andreas Fischer, Paul Frommer, John Hagge, Jim Hlavac, John Hedgcock, Kaoru Horie, José Hualde, Larry Hyman, Yamuna Kachru, Christine Kakava, William A. Kretzschmar, Juliet Langman, Peter Lazar, Audrey Li, Ronald Macaulay and his students Joseph L Malone, Erica McClure, David Mortensen, James Nattinger, Michael Newman, John Oller, Ingo Plag, Doug Pulleyblank, Vai Ramanathan, Gregory C. Richter, La Vergne Rosow, Johanna Rubba, Robert Seward, Trevor Shanklin, Harold F. Schiffman, Deborah Schmidt, Barb ara S Peicher, Chad Thompson, Gunnel Tottie, Edward Vajda, Robert R. van Oirsouw, Heidi Waltz, Charlotte Webb, Rebecca Wheeler, Roger Woodard, Anthony Woodbury, Thomas E. Young, and Rüdiger Zimmermann. Thank you for the provided data from Marwan Aoun, Zeina el-Imad Aoun, Dwight Atkinson, Liou Hsien-Chin, Yeon-Hee Choi, Du Tsai-Chwun, Nan-hsin Du, Jin Hong Gang, José Hualde, Yumiko Kiguchi, Yong - Jin Kim, Won-Pyo Lee, Christopher Long, Mohammed Mohammed, Phil Morrow, Masagara Ndinzi, Charles Paus, Minako Seki, Don Stilo and Bob Wu. A very special kudos to Emily Nava for her many suggestions to improve this edition. The photo on page 79 was contributed by Eric Du and that on page 48 by Julian Smalley. William Labov contributed the map on page 365. Thank you for the graphics and Jenny Ladefoged for the photo on page 13. The photo I took on page 505 shows Joanne Smalley from Nottinghamshire, England, talking to her three month old daughters, Anya Smalley Lowe. I received thoughtful and much-appreciated recommendations from the reviewers commissioned for this issue: Steve Chandler at the University of Idaho, Katherine Crosswhite at Rice University, Zygmunt Frajzyngier at the University of Colorado, John B. Gatewood at Lehigh University, Sam Mchombo from the University of California, Berkeley, Timothy J. Pulju from Dartmouth College and Angela Reyes from Hunter College. To Steve Dalphin, editor; Megan Garvey, editorial assistant; Sarah Sherman, director of content projects; and Bonnie Briggle, Project Manager, for shrewd editorial supervision and meticulous production. Publishers are increasingly acknowledging their employees and freelancers, who play a central role in making a textbook what it can be. I welcome this recognition. Special thanks to Mary Beth Walden, who served as Development Editor for this issue; Your patience and insight are greatly appreciated. Thanks to my partner Julian Smalley for countless technical and less tangible contributions.

xxv• Foreword

A word for everyone I appreciate the comments from all readers[email protected]—Edward Finegan Los Angeles

1 Languages and Linguistics WHAT DO YOU THINK? • Two roommates who are arguing about something are discussing the number of languages in the world. One says thousands and the other says there is no way to count them. What you say? • A friend in Los Angeles opens her utility bill and, looking at an insert, says in alarm: “Look at this, in five different languages: Spanish and Chinese, and who knows what! Shouldn't English be the official language of the United States? Is this? • Nan, a ninth grader, is reading a newspaper and looks up and asks what the word note means. You think she knows what phrases like love letter and thank you letter mean, so ask her to read the sentence out loud. After she reads it, you say that bill means "bill," as in "20 dollar bill." Nan asks how hearing the whole sentence helped you. your explanation? • Claire complains that her history teacher corrected the word sneak to sneak into her thesis. Everyone she knows says they ran away and she wonders where the professor got the information! What have you heard people cite as the authority for deciding right or wrong in using the English language, and where can Claire find reliable guidance on stealth and the use of stealth? • At a family picnic, fifteen-year-old Frank teases seven-year-old Seth, "Do you know when your birthday is?" When Seth offers "May 9th," Frank replies, "I didn't ask you when your birthday was, man ! I asked if you knew when it was! What does Seth understand about the question that Frank pretends not to understand?"

2 • Chapter 1 Languages and Linguistics

How many languages are there in the world? Some dictionaries include language names in their entries, and you've probably seen lists containing information about the number of speakers of different languages. When the US Census Bureau compiles its census data each decade, it asks residents what language they speak and publishes that information. Most countries are represented at the United Nations and their ambassadors need to know which languages are spoken in their home countries. With all this information, it should be easy to answer this question: How many languages are there in the world? In fact, listing them is not an easy task. First, it is not always clear whether we should call two language variants dialects of the same language or of different languages. Additionally, languages hitherto unknown to scientists are sometimes discovered in the Amazon, Papua New Guinea, and other remote parts of the world. Some language versions may be limited to spoken languages while others include sign languages. After all, languages die when their last speaker dies, and that happens more often than you might think. Even when the criteria for inclusion in a language list are established, compiling the information may not be easy. For one thing, a given language can have different names, such as B. Hebrew and Ivrit or Irish, Erse, Gaelic and Irish Gaelic. On the other hand, a name can be spelled in different ways. Uyghur, a language primarily spoken in China (but not related to Chinese), was written as Uyghur, Uyghur, Uyghur, Uyghur, Uyghur, Uyghur, Weiwuer, and Wiga; Among its speakers, Uyghur is written in Arabic rather than the Roman alphabet, and is sometimes also rendered in Cyrillic and Chinese characters. Over the course of a century, some languages die and others are born. Some of those who die can later be revived, as the Hebrew did. Similarly, the last speaker of Cornish, a Celtic language, died in 1777, but the language has only recently been revived and is now used by a few thousand speakers in south-west England. Manx, another Celtic language previously spoken in the Isle of Man, has died out as a first language, but some second language speakers are making efforts to revive it. In 1996, in Worcester, Massachusetts, Red Thunder Cloud died, and with it Catawba, a Siouan language. On the other side of the coin, in some places pidgin is spoken as a second language, and when children speak a pidgin as their first language, it becomes a full language called Creole. Creole should be counted among the world languages (although their users still call them pidgins). A reliable source of information, The Ethnologue, lists 6,912 languages. But don't think that it can be an exact number in all situations. Note that we sometimes refer to "Chinese" in this book and that the US Census Bureau allows residents to self-identify as "Chinese" speakers, while Ethnologue does not list Chinese as a language anywhere. Instead, it lists thirteen languages with names such as Hakka Chinese, Mandarin Chinese, Wu Chinese, Xiang Chinese, and Yue Chinese, each of which may have its own dialects. In English-speaking countries, Mandarin Chinese is known as Mandarin and Yue Chinese as Cantonese. Also, a good portion of the 6,912 languages are sign languages. Aside from their channel of expression, most sign languages are like spoken languages and share with them the challenges of how to identify and count them. In the city of Chiangmai, Thailand, Chiangmai Sign Language is well known in the deaf community, but only among older sign writers, while younger signers use another language called Thai Sign Language.

Does the United States have an official language?

• 3

It seems safe to go by conventional wisdom that there are between 6,000 and 7,000 languages in use in the world. Of those thousands, only Arabic, Chinese, English, French, Russian, and Spanish have official status with the United Nations, and French is not among the top 10 most widely spoken languages in terms of number of speakers, while Hindi, Bengali, and Portuguese do . , German and Indonesian-Malay have more speakers than some official UN languages. Also note that for the year 2000, the US Census Bureau lists 30 individual languages used in the United States (and many more, classified under broad labels such as "African languages", "other Indic languages", "other Amerindian languages ' have not been identified' and 'Other Asian languages'). ). Given the 30 languages mentioned (which do not include sign languages and count all variants of Chinese as one language) and given these other broad categories, how could you be sure of estimating the number of languages spoken even in the US? United States? Imagine the challenges of getting an exact number for everyone!

Does the United States have an official language? Many Americans think that English is the official language of the United States. In reality, however, the United States has no official language and never has one. A few states have official languages: Spanish in New Mexico and English and Hawaiian in Hawaii, but not the country as a whole. Some Americans also tend to view the United States as essentially a monolingual nation, albeit with large populations of Spanish speakers in the Southwest, Southeast, and Northeast. In reality, however, nearly 47 million US citizens age 5 and older speak a native language other than English (nearly 18% of this age group). 28 million of them speak Spanish, and more than half of these Spanish speakers report that they also speak English very well. Young people between the ages of 5 and 17 who speak a first language other than English number almost 10 million and the vast majority report that they also speak English very well. In seven densely populated states

Los Angeles, California. Voter information brochures are available in English, Spanish, Tagalog, Korean, Vietnamese, Chinese and Japanese.

4 • Chapter 1 Languages and Linguistics

At least one in four residents aged 5 and over speaks a native language other than English. Arabic, Hindi, Hungarian, Korean, Tagalog, Thai, Urdu, and Vietnamese have speakers in all 50 states, and some Amerindians speak the Native American languages in all 50 states. Navajo, with more than 175,000 speakers, is used in homes in 47 states. According to The Ethnologue, 162 living languages are currently spoken in the United States. They range from Achumawi, Alabama, and French Creole to Hawaiian, Lakota, Maricopa, Uyghur, Vietnamese, and Yatzachi Zapotec. At the moment, the United States is rich in languages, and even ballots can be printed in multiple languages. Yet the linguistic wealth of the United States is neither stable nor reliable. The survival of most Native American languages is under threat, both because speakers tend to be older and because insufficient resources support these ancestral languages, giving way to English among younger Native Americans. After English, no language spoken in the United States comes close to Spanish in terms of the number of speakers. Its 28 million Americans far outnumber the 2 million who speak Chinese and the 1.6 million who speak French. With few exceptions, the children or grandchildren of immigrants can no longer readily speak or understand the language of their grandparents, even in the case of Spanish. Furthermore, despite the wealth of non-English languages in the US, 215 million US citizens aged 5 and older speak English at home, a staggering 80%.

Try it yourself: Using your knowledge of current and past immigration patterns, identify the eight most popular non-English languages spoken by US citizens ages 5 and up: Spanish, Chinese, French, and more. For the record, Polish and Arabic take 9th and 10th place respectively.

English only, English plus, multilingualism Many people in the United States do not see their linguistic diversity as an advantage and seem to prefer English only, at least in public discourse. There are many reasons for this, and there have been a number of laws, ballot initiatives and court decisions in this regard. The English only movement has emerged over the past few decades to push for legislation banning the use of languages other than English in certain circumstances, and is particularly concerned about the use of languages other than English in schools. While much of the concern about the use of languages other than English may be unwarranted, even xenophobic, much of it is also motivated by a legitimate concern about the widespread failure of school-age youth to master standard English. Other Americans, concerned about losing the social, cultural, and political treasure represented by linguistic diversity, want to preserve traditional languages and advocate for bilingual education and "English plus." Bilingual education in the United States is also controversial. For example, in 1998, California voters endorsed a ballot initiative (a citizen-sponsored law) that mandated that "all children in California public schools learn English as quickly and effectively as possible." This "Proposition 227" ended bilingual education, except in certain circumstances. In California, many believe Proposition 227 is the way forward, while others see bilingualism and multilingualism as invaluable resources to cultivate.

What is human language?

What is human language? Modern linguistics has its roots in questions that were first asked millennia ago. As old as speculation on any subject, the study of the nature of language occupied Plato and Aristotle, as well as other Greek and Indian philosophers. In some areas of grammatical analysis, the ancients made contributions that remained useful for 2,000 years and established some of the analytic categories still used today. In the 19th and 20th centuries, the field of linguistics emerged to answer some age-old questions, including: • What is the relationship between signs and what do they mean? • What are the elements of a language and how are they organized into words, sentences and discourses? • What allows us to form and understand sentences that we have never heard before? • How do languages achieve their communicative goals? • What is the origin of language? • How do languages change and develop? • What does it mean to say that two languages are related? • How are languages and dialects related? • What enables a young child to learn a language so well and efficiently? • What makes it so difficult for an adult to learn a language? • Are there right and wrong ways of saying things, and if so, who makes the decisions?

This book provides a modern context for asking and answering these questions.

The Three Faces of a Language System The basic function of any language system is to combine meaning and expression to give verbal expression to thoughts and feelings. A grammar can be viewed as a coin whose two sides are expression and meaning and whose task is to systematically link the two. But language has a third face that is so important to the production and interpretation of statements that it can supersede everything else. This face is the context, and only in a specific context can an utterance convey the meaning intended by the speaker and be properly interpreted by a listener. Imagine a dinner discussion about the cost of living where a guest asks the host, "Is there a state income tax in Connecticut?" Possible answers to this question include "yes", "no" and "I don't know", as the question is likely to be interpreted as a request for information in this context. Now consider an equally direct question asked on the same occasion: "Is there salt on the table?" In this case, a host who sincerely responds with "yes," "no," or "I don't know" and drops the subject would come across as insensitive at best. Is there a state income tax in Connecticut? Is there salt on the table?

The form of the salt question is similar to the form of the income tax question, but the purpose of the questions (their intended meaning) and the expected answers could hardly be more different. At a table, a guest asking for salt naturally expects the host to confirm they want the salt, not the information! Conversely, in a related context, for example, with the host in the kitchen, pepper mill in hand, and asking a guest

• 5

6 • Chapter 1 Languages and Linguistics

who has just come out of the dining room: "Is there salt on the table?" the host should be understood as someone seeking information, even if the form of the question corresponds exactly to that of the guest at the table. In response to the question asked in the dining room, an answer of "yes" or "no" would seem odd. In the kitchen it would be absolutely appropriate. So you can see that speakers cannot interpret an utterance by expression alone. To understand the intended meaning of an expression, listeners must consider it in the light of its context. At the same time, when pronouncing an expression, speakers often rely on the listener's ability to understand and discern their intentions in pronouncing the expression in a given context. Along with meaning and expression, the basis of language use is context, and language can best be viewed as a three-sided figure of expression, meaning, and context, as shown in Figure 1-1.

Figure 1-1 Three faces of language

Norte

MIM

S.I.O.

public relation

GRAMM

EX CONTEXT

Expression includes words, phrases and sentences, including intonation and stress. The meaning refers to the meanings and reference points of these expression elements. Context refers to the social situation in which the phrase is uttered and includes anything previously expressed in that situation. It is also based on knowledge generally shared between the speaker and the listener. What unites expression and meaning is grammar. What unites grammar and interpretation is context. Without paying attention to grammar and context, we cannot understand how language works.

Language: Mental and social language is often viewed as a thought carrier, a system of expression that mediates the transmission of thought from one person to another. In everyday life, language also fulfills important social and emotional functions. Linguists are interested in models of how language is organized in the mind and how the social structures of human communities shape language and reflect those structures in expression and interpretation.

Characters: arbitrary and non-arbitrary

Signals: arbitrary and non-arbitrary In everyday conversations we talk about signs of disruptions in the economy, no signs of a train arriving at a station, a person's vital signs, etc. The signs are indicators of something else. In the examples given, the indicator is inherently related to the thing being indicated. Non-arbitrary signs have a direct, generally causal, relationship to the things they indicate. Smoke is a non-random sign of fire, clouds are a non-random sign of imminent rain.

Random Signals Non-random signals such as clouds and smoke differ significantly from partially or fully random signals. Random characters include traffic lights, ticket markers, wedding rings, and national flags. There is no causal or inherent relationship between arbitrary characters and what they mean or indicate. Arbitrary indicators may be present even if the thing indicated is absent (e.g. a bachelor wearing a wedding ring). Since these are conventional representations, any signals can be changed. If a national transport authority decides to use the color blue as a signal to stop traffic, they could do so. The relationship between words and what they represent is generally arbitrary, and we say that language is a system of arbitrary signs.

Representative Characters Some essentially arbitrary characters are not entirely arbitrary and may be indicative of their meaning. Poison can be denoted by a skull and crossbones, while a symbol can denote the sun, and Roman numerals II and III represent the numbers two and three. Because these signs suggest what they indicate, they are partially iconic. However, there is no inherent connection: the sign can exist without the signified and the signified without the sign. Signs that are basically arbitrary but partially iconic are called representational. Speech examples in English are meow and drip because these words suggest what they mean. The iconic expression can also pop up spontaneously in colloquial speech. Once I called a friend at her house and her four-year-old son answered. He reported that his mother was taking a shower, and when I told him I would call him back in a few minutes, he implied that calling back soon wouldn't be enough. His explanation was this: My mother takes long, very long showers.

By lengthening his pronunciation of the vowel, the boy demonstrated the potential for spontaneous iconicity in human language. By lengthening his vowels, he directly signaled the duration of time, thus iconically emphasizing the most prominent part of its meaning. Representative (or iconic) language is a linguistic expression that in some way imitates or directly suggests its content. Try it yourself: in addition to lengthening the vowel to represent length of time, the boy's expression was iconic in a second sense. Identify this second shape. Then identify another very common example in English where this second form conveys a different meaning of extended length. (Hint: Examine the previous sentence carefully.)

• 7

8 • Chapter 1 Languages and Linguistics

Iconicity can also be expressed in grammar. Remember that there are two ways to organize conditional clauses in English. The condition (the "if" part) can precede or follow the consequence. If you're good, I'll give you some M&Ms. (Condition precedes consequence) I'll get you some M&Ms if you're good. (condition follows consequence)

English allows you to put the condition (if you're good) before or after the consequence (I'll give you some M&Ms). Although contextual factors can affect choice, speakers and writers show a strong preference for condition to precede consequence, a preference found in many other languages as well. The reason has to do with the order of occurrence of real events, which are described by conditional clauses. In our example, the recipient must behave first and then the speaker will deliver the M&Ms. These real world events are ordered in time with the condition before the consequence, and this real world order is reflected in the preferred linguistic order. There is therefore an iconic explanation for preferring condition-precedent-order to reverse order. With the condition before the consequence, the expression iconicly mimics the sequence of events in the real world. Some languages only allow the condition-consequence pattern above; others allow both; but no language seems to confine conditional sentences to a non-iconic order, to a consequence rather than a condition.

Language: A System of Arbitrary Signs Despite occasional iconic features, human language is essentially arbitrary. The form of an expression is generally independent of its meaning, except for associations established by convention. Imagine a parent trying to catch a few minutes of the evening news on TV while cooking dinner. Suddenly, a strong smell of burnt rice fills the TV room. This non-accidental clue will make parents save the dinner. The aroma is created by burning the rice and conveys its message to speakers of all languages. There's nothing conventional about that. Now compare the aroma to the words of a young man who sees the smoke in the kitchen and exclaims, "The rice is on fire!" This phrase will likely get parents going too, but the words are arbitrary. It is a record of English (not rice burning) allowing pronunciation to warn parents. So the statement is an arbitrary character. Other languages express the same meaning differently: Korean by the phrase pap thanda, Swahili by wali inaunguwu, Arabic by yah. Tarik Alruzu etc. The forms of these statements have nothing to do with the rice or the way it is cooked; they are not iconic. Rather, they only deal with the Korean, Swahili, and Arabic language systems. Because the relationship between linguistic signs and what they represent is arbitrary, the meaning of a particular sign may differ from one culture to another. Even words that mimic natural sounds differ between languages. For example, cats don't meow in all languages; In Korean, the word is yaong. As you can see, a central feature of human language is that the association between words and what they mean, between the signifier and the signified, is largely arbitrary. In England, the bakers bake the bread; in France pain; in Russia xleb; in China miànba¯ or in Fiji madrai. Not only do things have different meanings in different languages, even a single language can use multiple characters to represent a simple term. We buy a dozen or twelve bagels at the same price and write 12, XII, TWELVE, twelve or twelve. With more complex content, the range of possible expressions can be unlimited.

Languages as standardized structures

Languages as standardized structures Given the arbitrary relationship between linguistic signs and what they represent, languages must be highly organized systems to function as reliable means of communication. We call the observable patterns that languages follow “rules”. They are not imposed from outside (like traffic rules) and do not dictate how to do anything. Rather, the rules described in this book are based on the observed regularities of language behavior and the underlying language systems that can be inferred from that behavior. They are the rules that even children have subconsciously picked up and used to demonstrate mastery of their native language. A language is a set of elements and a system for combining them into standardized expressions that can be used to perform specific tasks in specific contexts. Statements report news, greet relatives, invite friends to lunch, ask strangers for dates; With language we joke, we scoff, we defend a course of action, we express admiration, we propose marriage, we create fictional worlds, etc. And a language does its job with a finite system that a child can master in a few years. The intelligence that enables speakers to form grammatical phrases like My mom is taking a long shower instead of "Mom is taking a long bath" (or thousands of other possible malformed strings of exactly the same words) is grammatical competence. It enables speakers to produce and understand an infinite number of sentences that they have never heard before. In addition to arbitrariness, four other properties of human language systems are worth mentioning.

Discretion Speakers can identify sound segments in words in their language. English speakers can identify the sounds in cat as the three represented by the letters c, a, and t. The same is true for the percussion sounds, which are recognized as four: the initial consonant group represented by s and p, the final consonant sound represented by ll, and the intermediate vowel. It is a structural feature of language that words are composed of elementary sounds.

Duality Human languages can be analyzed on two levels. At one level it can be seen that they have significant units; thus the table top has the two essential parts table and top. At a lower level, the elements contained in the significant parts have no meaning. The three main sounds make no sense individually; they only form a meaningful unit when combined as above. And precisely because the individual sounds in top have no independent meaning, they can be formed into other combinations with other meanings, such as pot, opt, toped, and popped.

Displacement Human languages are able to represent things and events that are not there but are distant in space or time. We do not confine ourselves to discussing events in the here and now. Instead, we can talk about distant places and events of yesterday or in the past, even events that have not yet happened or will never happen. This is an important feature of human language, so important that you ask whether it is the use of tools or language

• 9

1 0 • Chapter 1 Languages and Linguistics

has contributed most to human development, respected primatologist Jane Goodall said: "The biggest difference between us and chimpanzees and other great apes is that we, and only we, developed this sophisticated spoken language. . . . We can teach our children things that are not present, events in the distant past, [and] we can plan for the distant future!”

Productivity From relatively few elements and rules in a language system, humans can produce and understand an unlimited number of sentences by combining and recombining the same relatively few elements in relatively few patterns. Even week-long issues of, say, Time, Newsweek, and The Economist are unlikely to repeat a single sentence (apart from quoted statements), and the same is true for years of publication. The human capacity for linguistic ingenuity makes repetition of sentences unlikely, and an English speaker is capable of understanding all English sentences in a lifetime of reading. And for many observers, this productivity, this ability to generate and understand an infinite number of sentences by combining and recombining the same few elements and structures, is the great characteristic of human language on which language theory focuses. . .

Language as Patterned Use of Language Knowing the elements of a language and the patterns of how they are assembled into well-formed sentences is still not enough to know how to do the work that speakers can do with their language. This requires not only mastery of grammatical rules, but also competence in the correct use of the sentences generated by those rules. Working with the language used requires, among other things, knowing how to connect sentences appropriately in conversation and how to properly rely on the context to shape and interpret utterances. The ability to use language appropriately is called communicative competence. It allows us to weave statements into narratives, apologies, requests, instructions, prescriptions, sermons, reprimands, jokes, prayers, and anything else we do with language. Fluent speaking requires both communicative competence and grammatical competence. Grammatical competence is the language user's unconscious or implicit knowledge of vocabulary, pronunciation, sentence structure and meaning. Communicative competence is the tacit knowledge underlying the appropriate use of grammatical competence in communicative situations. Because standards for proper language usage vary from one language community to another, even a shared knowledge of grammar in a language like English may not be enough to be fluent in other English-speaking communities. For example, members of one culture may find jokes about other people's mishaps funny while members of another culture find them offensive. In fact, the very concept of telling jokes (Have you heard the one about...?) as opposed to telling "funny stories" doesn't seem to exist in certain societies. Likewise, what is considered rude in one place may be routine interaction in another. Differences in interaction habits explain why even some American visitors to the Big Apple may find New Yorkers brusque or impolite when giving directions, for example, when another New Yorker might interpret the same directions as routinely polite.

The origin of languages: from Babel to Babble

The Origin of Languages: From Babel to the Babbler Many people in all parts of the world share the belief that the origin of language stretches back to the Garden of Eden. Even among people who have little faith in this story, many are convinced that the language came from a paradise where its primitive form was logical and perfectly grammatical. It is a common belief that once pure languages have become contaminated with impurities, illogics and ungrammaticalities over time. As examples of contamination, subscribers to this concerned view cite loanwords such as ok American and disco French, which have spread to many other languages, making them less "pure". Among the alleged illogics, double negatives are a commonly cited example in English. Here's the claim. Just as two negatives make one positive in algebra or logic (no wrong means "it's true"), logically "I don't want none" should mean "I want something," and "he's never done anything right" should mean "he has did something right”. . " Of course not. Suspected misspellings include the personal pronoun me alone between you and me and him and me as the subject in the sentence He and I were friends in the army. The argument offered is that the objects of a preposition must be in the objective case (ie only between you and me) and the subjects of a sentence must be in the common (or subject) case (He and I were friends in the army.) Another supposedly ungrammatical word is the word colado, considered by many to be an improper form of colado We all recognize that millions of speakers around the world use these and other supposedly impure, illogical, and ungrammatical expressions, and the sun still rises on them every morning and sets every night, just as it does for those who use them for keep their language the purest., logical and grammatical.

Try it yourself: If you were trained as a student assistant to search through a vast database of legal opinions written on a project to create a manual on the use of modern legal language, and you find that judges are sneaky as they used to be typed tense of sneak about two thirds of the time and sneak the other third, what would you expect the project manager to say in the handbook about the modern legal usage of the past tense sneak? What if your results were 60/40 or 50/50 instead of 66/33?

In addition to differing views on the origins of languages, people have different ways of explaining why languages differ from one another and why they change. The Old Testament records that before the Tower of Babel was built, all men and women spoke the same language and could understand one another. Ultimately, human pride caused God to mistake their communications for mutually incomprehensible languages. According to this story, language differences between people can be seen as punishment for sinful behavior. Similarly, Muslims believe that Allah spoke to Muhammad in pure and perfect Arabic, which the Qur'an embodies. Rather, the variants of contemporary Arabic spoken in the Persian Gulf, North Africa, and elsewhere are seen as the result of later human frailty and guilt. Professional linguists take a different approach. They see the diversity of languages as the result of natural changes over time, the inevitable product of language remodeling to meet changing social and intellectual needs and to reflect human contact.

• 11

1 2 • Chapter 1 Languages and Linguistics

speak other languages. As groups move to new places and mingle with speakers of other languages, or settle in areas with unfamiliar plants and animals, their language must adapt to the new realities. Encountering people using unknown artifacts and having different points of view, and encountering unknown aspects of nature invite speakers to adapt their language. As a result, languages are evolving differently around the world. However, what is even more surprising than the linguistic differences around the world is the degree of similarity between the languages. The differences are obvious, the similarities more subtle. But the similarities shouldn't come as a surprise, after all, every language has to be tailored to the character and capabilities of the human brain. There is only a relatively narrow range of all conceivable language structure types among the world languages. (Chapter 7 examines the universals of language structure.) As you will have noticed when you first hear a foreign language, there are marked differences between languages. Not only do Japanese and French sound different, but French also differs from its close relatives Spanish, Italian, Romanian, and Portuguese. Different social groups even speak the same language differently, and each social group controls a variety of styles that can be used in different situations; the language of conversation differs from that of political sermons and speeches. Speakers tend to be linguistically diverse, with some language varieties characteristic of user groups (Burmese and Brooklyn) and others characteristic of application situations (law, IT, Manhese). Each linguistic variety marks the speaker's social identity and the situation in which it is used.

Languages and dialects Along with physical appearance and cultural characteristics, language helps define nationality. But even within the borders of a nation, people can speak different languages. Ethnic French-Canadians in Quebec are loyal to the French language, while ethnic Anglo-Saxons are loyal to English. Citizens of Switzerland can speak French, German, Italian or Romansh. Dozens of languages are spoken across India, some confined to villages, others regionally or nationally. Hundreds of languages are heard in Papua New Guinea, and an English-based language called Tok Pisin is used for communication between groups. Whenever speakers of a language are separated by geographic or social distances, significant linguistic differences are likely to occur. Marked differences can be observed between the French varieties spoken in Quebec and Paris and between the Spanish varieties spoken in Madrid and Mexico City. English speakers from Sydney, London, Dublin and Chicago speak remarkably different varieties.

Try it yourself: Identify a feature of your own pronunciation that other people have commented on while traveling outside of your region. How about a vocabulary of yours that others find unusual? Do the same with a roommate or classmate whose speech you notice has a different pronunciation or vocabulary. Some people seem to think that only other people speak a dialect, but that's not the case. Instead, they see themselves as people who speak a language, or even that

Languages and dialects language. The truth is that everyone speaks a dialect. American English, Australian English and British English are national dialects, and all speak a regional dialect as well. Anyone who speaks a dialect of English speaks the English language, and anyone who speaks the English language can do so simply by using one of their dialects.

What are social dialects? Language varieties can vary from region to region within a nation. They can also differ across age groups, ethnic groups, and socioeconomic boundaries. In the United States, communities of white Americans and communities of black Americans may speak differently, even if they live in the same city. Similarly, middle-class and working-class speakers can often be distinguished from one another by their distinctive way of speaking. Ethnic, socio-economic, gender and age group-specific language practices also form dialects. You speak a dialect specific to your nationality, region, gender, socioeconomic status, and other characteristics. And the same goes for everyone else. This fact is amusingly illustrated in the classic film My Fair Lady. Based on George Bernard Shaw's Pygmalion, the musical tells the story of phonetics professor Henry Higgins (played by Rex Harrison) who makes a bet to teach London peddler Eliza Doolittle (Audrey Hepburn) to speak "right". Eliza says: "I'd like to be a lady in a flower shop rather than selling on the corner of Tottenham Court Road. But they won't take me if I can't speak nicer. Determined to win the bet, the teacher threatens, "You'll get your vowels right by the end of the day, or there won't be lunch, dinner, or chocolates." A person's dialect is an important part of who they are, and it is him Changing what some people want to do for themselves and many want to do for someone else is not as irrelevant as changing a sweater or hairstyle. "I'm going to make a duchess out of this little dog with a dragon's tail," prophesies the teacher about Eliza. In real life, of course, things are more complicated.

Image unavailable due to copyright restrictions

• 13

1 4 • Chapter 1 Languages and Linguistics

Other dialects or other languages? The Romance languages evolved from the regional dialects of Latin spoken in different parts of the Roman Empire. These dialects eventually gave rise to Italian, French, Spanish, Portuguese and Romanian, now the separate languages of the independent nations. While these languages may share certain inherited traits of grammar, pronunciation, and vocabulary, the nationalistic pride of Italians, French, Spanish, Portuguese, and Romanian supports the view that they speak distinct languages rather than dialects of a single language. The opposite situation characterizes the Chinese. Not all Chinese dialects are mutually intelligible (e.g. Cantonese and Mandarin speakers cannot understand each other), but the speakers see each other as a common language and emphasize this unity with a common writing system. Whether two varieties are considered dialects of one language or different languages is both a social and linguistic issue, and naming can be influenced by nationalistic and religious attitudes. Hindus in northern India speak Hindi, while Muslims there and in neighboring Pakistan speak Urdu. Opinions differ on how well they get along. Until a few decades ago, Hindi and Urdu formed a single language called Hindustani, and the fact that professional linguists wrote "Hindi-Urdu" grammars reflects the perception that the two variants required only a single grammatical description. Of course, over time, Hindi and Urdu, whose different names proclaim the affiliation of their speakers to different social, political, and religious groups, have become more divergent, as have the Romance languages.

What is a standard variety? No type of English can be called standard. After all, there are different national standards: for British, American, Australian and Canadian English, among others. Also, at least as far as pronunciation is concerned, there can be several standard varieties of a national variety. The simple fact is that there are many variants of Standard English. So what does a standard variety mean? We could identify as a pattern the variety that a group of people uses in their public discourse: newspapers, radio shows, political speeches, lectures at colleges and universities, etc. That is, we could identify as the standard the variety used for certain activities or in certain situations. Alternatively, we might call the standard the variety that has undergone a standardization process during which it has been organized for description in grammars and dictionaries and codified in these reference works. An important point to note about any standard or standardized variety is that it should not differ from other varieties in linguistic character. It's not more logical or grammatical. Also, there's no sense in which you could say it's better linguistically. On the other hand, it is certainly true that for some purposes the use of a standardized distributor is very useful. For example, this book is written in a variety of English that has been standardized, and this fact makes it possible to read it in many parts of the world. Rather than using spellings that reflect my own pronunciation, I use standardized American spellings that differ from British spellings in familiar ways. Typically, the standardized varieties are the local dialects spoken in centers of commerce and government. In these centers, the need arises for a diversity that goes beyond local needs, such as in the dissemination of technical and medical information, in the dissemination of laws, and in the production of newspapers and books. The hubs are also places where dictionary makers and publishers are likely to be located. Meanwhile, Samuel Johnson lived in London

Languages and Dialects wrote his dictionary Noah Webster in New England. If circumstances were different, the varieties represented in their dictionaries might well represent the dialects of other groups. Dictionaries initially serve to describe and anchor a variety of languages that can be used for public discourse in social groups, regions and even countries. Of course, not all situations are typical. With Basque, authorities combined forms from different regions into a single standardized variety to be socially and regionally inclusive. The same happened with Standard Somali.

Is there a good and a bad thing about using the English language? Are there right and wrong ways to say and write things? Which of the two spellings honor and honor is correct? Is the Canadian "schedule" or the American "schedule" correct when pronouncing the schedule? Should the interval between plays be called intermission (as in Broadway) or intermission (as in London's West End)? Americans, Canadians, and Brits may prefer their own expressions or pronunciation, but depending on who you are, where you are, and what you want to achieve, any of these alternatives may be appropriate and correct. If it says "book" in Detroit, it will most likely be identified as Canadian. When you say "skedule" in Toronto, you most likely identify as a Yankee. Similar differences exist within a country. What about vocabulary and grammar questions? Is stealth and stealth okay? Aren't some forms of language (like signs saying Drive Slow instead of Drive Slow) downright wrong? To answer this question, it helps to think of grammar as a description of how language is organized and how it behaves. In this case, ungrammatical English sentences include: Try something different this season. Season experiences something other than this allergy.

These are ungrammatical variants of the grammatical expression This allergy season, try something different than a print ad for an allergy drug. No one who speaks English would normally say or write any of them, and in that sense they are ungrammatical. Another view would consider as ungrammatical any violation of a relatively small set of prescribed "rules" such as these: • • • •

Never end a sentence with a preposition. Never separate an infinitive. Never start a sentence with and or but. It's me is ungrammatical; is i is grammatical.

Recipes like these appeared in the 18th century, and even then they didn't accurately describe the language people used to converse and write. Commentators on this prescriptive tradition have formulated rules for what they consider to be the “appropriate” use of must and will, have condemned phrases such as between you and me, and sought to prohibit the use of ain't. More recently, they've poked fun at it when it's used to mark quoted lines or thoughts (and I say, "Do I really want to do that?"). Partly out of this prescriptive tradition, judgments have spread that some common phrases are ungrammatical, such as He and I would sit around talking all day and He doesn't like to cook or It doesn't matter. These phrases are certainly not standard English, but they are the forms commonly used by millions of English speakers, and in that sense they are perfectly grammatical in some varieties of English.

• fifteen

1 6 • Chapter 1 Languages and Linguistics

One way different English languages differ is in their rules, and different rules result in different structures. It is unreasonable to judge sentences allowed by the rules of one kind as ungrammatical simply because they do not follow the rules of another kind. By this logic, any expression that is allowed in Standard American English but not in British English would be ungrammatical, and vice versa. Since languages essentially rely on arbitrary characters to do their job, there's no reason to pretend that there's only one right way to say something. From a linguistic point of view, there is no reason to favor the structure of one language variety over another. The judgments "illogical" and "impure" are imported from outside the realm of language and represent attitudes toward particular varieties or expressions within particular varieties. They often represent judgments by groups of speakers rather than the speech itself.

Try it yourself: Given the premise that, for example, Japanese and Italian are grammatical, Parisian French and Montréal French are grammatical, and American English and British English are grammatical, you argue that it is equally logical to add Brooklynese, Bostonian, and African consider American. English as grammar. In this book, we apply the word "ungrammatical" in a language variety only to expressions that cannot be pronounced by native speakers of that variety. We limit the term "ungrammatical" to a statement such as Book that reading am I right now (compare I am reading that book now) because it does not appear in the assumption that someone knows English (except as an example of a poorly formed sentence usage in didactic books like this one). We do not consider an expression like just between you and me to be "ungrammatical".

Modes of verbal communication There are three basic modes of verbal communication, corresponding to different modes of perception: oral communication, based on the use of the organs of speech and hearing; Writing, a visual representation; and signature, a visual or tactile representation.

Speech The most widespread medium of verbal communication is the voice, and as such speech is a major mode of human speech, with some advantages over other modes. Since it does not have to be seen, the speaking tube can do its job effectively in the dark and in the light, in front and around the corner. During the evolution of the human species, with hands and eyes busy hunting, fishing, and gathering food, speakers were given the freedom to inform, to question, and to direct, to explain, to promise, to apologize, to haggle, to warn and to flirt. Talking has other benefits as well. For one thing, the human voice is complex and has many channels. It has variable volume, tone, rhythm and speed; it is able to modulate a wide range. In addition to a set of sounds, language uses the organization of these sounds, their sequencing into words and sentences. As with writing and pointing, speaking can benefit from word choice and order. in its natural state,

Of course, linguistic modes of communication, language, evaporate and cannot bridge time, but modern technologies make it possible to preserve language indefinitely.

Writing Long before the invention of writing, humans drew stories on cave walls and used other visual cues to record events. These pictograms were language-independent: a kind of cartoon world in which anyone with knowledge of people's lives but without specific language skills could understand the story depicted. When shown to adult speakers, the stories presented can be told in Tagalog, Japanese, Arabic, Swahili, Vietnamese, Spanish, English, Indonesian or any other language. Pictograms ( ) are understandable in any language because they are direct, non-verbal symbolization, like a silent film or the road signs used internationally to indicate a winding road or the availability of food and shelter. Among the most popular symbols currently in use are the : ) and : ( emoticons, which are common in email correspondence. When the symbols are not associated with the objects they represent, but with the words that relate to the objects, we have a much more sophisticated pictorial system Written representation becomes linguistic when it relies on language for its organization and communicative success Although it is difficult to convey a message through abstractions (such as hunger or danger) with pictograms the task becomes manageable when the graphic signs represent existing words, at the moment when an imaginative mind first realized that the written sign could represent not only the sun itself, but also the word "sun" in its spoken language , the first step in the development of writing was taken.Writing was invented 5000 years ago by ingenious people who once represent spoken words with pictograms. as instead of the objects they normally represent d. Language and writing have different relationships to the world they symbolize. Language directly represents the entities of the world, things like sun, moon, fish, grain, light, and height. Scripture represents the physical world only indirectly. A written sentence, such as Meg caught three fish, is a secondary symbolization in which the written characters represent the spoken words, not the entities and activities themselves. Writing has certain advantages over speaking. Although it usually takes longer to produce than speech, it can be read and understood much faster than speech can be heard and understood. Writing (in correspondence, in books, or on cave walls) lasts longer than unrecorded speech and, when published, has a wider reach. A message on a plaque can be read after its author has left the room; not so for a spoken expression.

Sign Language The third type of verbal communication is sign language, the use of visible gestures to communicate. To accompany their speech, speakers often use gestures and facial expressions to convey meaning in support of oral communication, but signaling can be used as the sole means of performing language work. American Sign Language, the primary sign language of North America, is used by between half a million and two million subscribers, not all of whom are deaf. ASL, as it's called, and similar sign languages use hand signs, facial and body gestures and combine them under a system of grammatical rules to form an infinite number of sentences. The linguistic character of sign languages and the linguistic character of

• 17

1 8 • Chapter 1 Languages and Linguistics

FIGURE 1-2 ASL Characters for SUMMER, UGLY, and DRY

Source: Lucas and Valli 2004

Spoken languages like Japanese and English are similar down to the channel of expression, but these sign languages are fundamentally independent of any spoken language. Hand signs in a language like ASL have three main components: • shape and orientation of the hand (which fingers are open or closed and what the palm looks like) • position of the hand in relation to the body of the sign • movement of the hand.

The signs differ from one another in one or more of these components. As shown in Figure 1-2, the ASL hand signals SUMMER, UGLY, and DRY are identical in hand motion, hand shape, and orientation. In all three, the right hand with the index finger extended, the others closed, is drawn in front of the signer, the index finger closing. All three characters are identical except for the placement of the hands. With SOMMER, the sign is drawn in front of the signer's forehead; for UGLY on the nose; and for DRY on the chin. Note that you cannot guess the meaning of these characters, and most characters would. But some other characters show a resemblance to what they represent. For example, in ASL, the character TREE can be seen as a tree, and the character CAT indicates the whiskers of a cat, as shown in Figure 1-3 ASL characters for TREE and CAT

Source: Lucas and Valli 2004

Do only humans have a language? 1-3 As with spoken languages, sign languages like ASL are fundamentally arbitrary, although some characters are representative, with meanings that can be guessed at. We've focused our discussion of sign language on vocabulary, but it's important to note that sign languages have grammars to combine vocabulary into phrases and sentences, just as spoken languages use grammars to organize phrases and sentences. Spoken languages don't form sentences by stringing words together like beads on a string, and neither do sign languages. Signed and spoken languages observe complex systems of grammatical rules, and ASL rules differ from American English rules in much the same way German rules differ from Spanish, for example. Knowing even a long list of ASL characters and knowing English or French grammar but not ASL grammar would not enable a person to effectively communicate in ASL or understand a conversation between participants using ASL . Given the historical development of languages and dialects, it should come as no surprise that ASL shows regional and ethnic variation, but it may not be so obvious that dialects of sign language need not parallel the dialects of languages spoken in the United States. same places. Spoken American English, for example, is primarily derived from spoken British English, but ASL is not derived from British Sign Language and the two are not mutually intelligible. Thomas Hopkins Gallaudet, founder of America's first school for the deaf, studied French Sign Language in Paris and returned to Connecticut with a French Sign Language teacher, so ASL evolved from a specific combination of FSL and a sign language used at Martha's school . . Vineyard, Massachusetts at the time Gallaudet started school. Another type of signaling used by Helen Keller is fundamentally different from ASL in that it was necessary since she was blind and deaf. Helen Keller's signature was a type of manual spelling system and, unlike ASL, was based on an existing spoken language. Their signaling consisted of spelling words by using their hands and hands to “draw” the shape of written characters used to represent sounds (as letters of the alphabet do). Helen Keller's sign language depended on the prior existence of a spoken language and a written form to represent it. Signaling systems based on letter modeling are two steps removed from the language system acquired by hearing and sighted children. Now ASL must also be based on writing to represent names of people and other words that don't have their own characters. Unlike Helen Keller's manual system, writing names and some other words in ASL is done visually, rather than by hand as Helen Keller required. In this book we focus on the language represented in oral and written communication. It is important to note that writing is a secondary mode of linguistic communication, both historically and evolutionarily. Talking is the main mode. This priority can be challenging for students whose focus and context of language discussion in school was reading and writing.

Do only humans have a language? When you look at animals in groups, it's obvious that they interact, that they communicate with each other. Dogs braid their fangs to express displeasure or aggression; male frogs attract female frogs by croaking. It is natural to wonder how the forms of communication used by animals differ from human language. Sometimes we say that porpoises, chimpanzees, gorillas, dolphins, whales, bees and other animals have language systems similar to humans. Undoubtedly, all animal species have evolved communication systems with which they interact.

• 19

2 0 • Chapter 1 Languages and Linguistics

they can signal things like danger and fear, and we now know quite a bit about how and what bees communicate. In recent decades, chimpanzees have been taught to use signs.

How animals communicate in their natural environment People have long wondered how bees manage to tell each other the exact location of a nectar source and have speculated about a bee 'language'. After years of observation and hypothesis, Karl von Frisch claimed that bees had an elaborate dance system they use to communicate the whereabouts of a stash of honey. Various aspects of the dance of a bee returning to the hive indicate the distance and direction of a nectar source. The quality of the source can be gauged by sniffing the explorer bee. Although some of Frisch's interpretations have been questioned, his careful analysis has shown that the bee dance lacks the kind of creativity characteristic of a child's speech. Bees do not use their communication system to convey more than a limited range of meanings (such as "There is a good nectar source here"), so the analogies between bee dance and children's language are implausible and fundamentally misleading. The same lack of creativity characterizes communication between other animals. In addition to a very limited repertoire of meanings, even intelligent mammals such as dogs lack the mental ability to be communicatively creative. For one thing, much of the communication between animals depends on non-random signals. When gazelles spot potential danger, their flight sends a signal to nearby gazelles that danger lurks, and the communicative function of the action is subordinate to its survival function. Likewise, a dog will signal the possibility of biting by showing its fangs. These actions are non-arbitrary signs that accompany desires and possibilities. Everything that animals express through sound does not seem to reflect a logical sequence of thought, but rather a sequence that accompanies a range of emotional states. The communicative activities of most animals differ from human language in that they do not consist essentially of arbitrary signs.

Can chimpanzees learn a human language? But what about chimpanzees? In nature, they use a limited non-verbal communication system similar to that of other mammals, although generally more sophisticated. As the intelligence of chimpanzees approaches human intelligence among other mammals, researchers have attempted to teach them human language. The first chimpanzee to gain notoriety for her communication skills was Vicki. Raised by psychologists Keith and Catherine Hayes for about seven years, she could only pronounce four words: Mama, Papa, Up and Cup, and she could only pronounce them with significant physical effort. Chimpanzees are simply not equipped with the proper mouth and throat organs to speak. Assuming that chimpanzees lacked the physiological ability to speak, the question remained whether they had the mental ability to learn languages. After seeing a film in which Vicki attempted to vocalize human speech, psychologists Allen and Beatrix Gardner gave a home to a ten-month-old chimpanzee named Washoe and raised him like a human child in every possible way. Eventually, Washoe ate with a fork and spoon, sat down at a table and drank from a cup, and even sort of washed the dishes. He was swaddled and potty trained; She played with dolls and showed them affection. as a human

Can chimpanzees learn a human language? He liked picture books with children his age and enjoyed hearing his human friends tell their stories about the pictures in them. Ingeniously, the Gardners arranged for all communications with Washoe to be in American Sign Language (ASL), which they also used with each other and with members of their investigative team when Washoe was present. The Gardners observed the types of simplified communication that human parents offer children in some societies and used repetition and simplified signaling in Washoe, who learned 4 characters in her first 7 months in this very human environment and 30 more characters in the following 14 months . . After 51 months he had acquired 132 signs describing objects and thoughts and understood three times as many. He used signs to denote specific objects and classes of objects. He used the sign "shoe" to refer to shoes in general; He used the character "flower" for flowers in general and even for smells like the smell of tobacco. He even signed for dogs and trees. He asked questions about the world of objects and events around him. After mastering just 8 characters, he began combining them to form complex expressions: YOU HIDE ME; QUICKLY LEAVE ME HEAR DOG (when a dog barks); MY BABY (referring to her doll); etc. After only 10 months in his foster home he has done dozens of combinations of 3 or more ticks such as ROGER WASHOE TICKLE and YOU TICKLE ME WASHOE. In subsequent work with four other chimpanzees (Moja, Pili, Tatu, and Dar) who arrived at their lab a few days after birth, the Gardners showed that adult, human-reared chimpanzees replicate some of the basic aspects of the language acquisition trait. human children, including the use of characters to refer to natural language categories such as DOG, FLOWER, and SHOE. Remarkably, when these chimpanzees later moved to another lab, a baby chimpanzee named Loulis acquired at least 47 signals that had no source other than the signals from their fellow chimpanzees. In creating Washoe and her chimpanzee companions, the Gardners posited that children acquire human language in a rich social and intellectual environment, and that this richness contributes to the child's cognitive and linguistic life. In other labs, the verbal activities of other famous chimpanzees were not vocal, as in Vicki's, or gestural, as in Washoe's, but visual. Sarah used plastic pieces as word symbols and displayed considerable assembly skills. Lana used a suitably labeled computer terminal to create a series of symbols resembling the plastic used by Sarah. Because of the success of the Gardners and other chimpanzee researchers, some observers, particularly psychologists, have suggested that there may be a continuum between human and non-human communication. But by no means were they all convinced.

Project Nim In fact, some psychologists have expressed skepticism about various projects to teach chimpanzees human language. They believe that the individual words the chimpanzees chose in the different modes were, in some cases, triggered by unintentional cues from the researchers. As a result, they claim, the string sequences produced by chimpanzees are not as productive phrases as those created by human children. Other critics question whether chimpanzees have the ability to use language to comment, ask questions, and express feelings like humans do. In an attempt to give more control over efforts to teach a chimpanzee to speak, a rigorous experiment sought to avoid many of the objections to previous research (although it inevitably brought its own problems). This chimpanzee was named Nim Chimpsky,

• 21

2 2 • Chapter 1 Languages and Linguistics

and in the course of his education, Nim achieved several linguistic achievements, which in part mirrored the achievements of his predecessors. But after five years of working with Nim, psychologist Herbert Terrace concluded that chimpanzees are not capable of learning languages like children. Even with sophisticated training, Nim produced very few long expressions and showed little creativity and spontaneity in his use of characters. Unlike Washoe, Nim only signed at the request of researchers and never initiated any interactions. These characteristics, Terrace argues, clearly distinguish what Nim could learn and what children can do with language. Critics of the Nim Project point out that Terrace employed more than 60 research assistants over the course of five years and believe this fact may have contributed significantly to the limitations of Nim's linguistic achievements. In addition, the keepers were instructed to treat Nim fairly and were forbidden to comfort him even when he cried at night. Critics have questioned how similar Nim's learning environment was to the environment in which a normal human child acquires language, and have argued that Project Nim's research conditions had an adverse effect on the chimpanzee's emotional and linguistic upbringing. So we see that some researchers claim that human and animal language fit into a continuum, while others conclude that chimpanzees cannot learn language like children. To reconcile these views, it is helpful to consider the assessment of respected primatologist Jane Goodall, who identified "speaking" as the greatest difference between humans and chimpanzees, "because we can discuss ideas, we can teach things that do." not exist. We can draw inspiration from and teach one another in the distant past, and we can plan for the distant future. It is above all this discussion of ideas that distinguishes humans from chimpanzees. The consensus among psychologists and linguists seems to be that animal language, including that of chimpanzees, does not share all of the characteristics of human language that we discussed above. While there are some similarities between human language and the signs and other visual languages taught to chimpanzees, some critical features of human language seem to be missing. , especially crowding out and productivity.

What is linguistics? Linguistics can be defined as the systematic study of human language: its structures and uses and the relationship between them, and the development and acquisition of language. The field of linguistics includes both the structure of language (and the grammatical competence on which it is based) and the use of language (and the communicative competence on which it is based). Language is often defined as an arbitrary language system used by humans to communicate with one another. This configuration is as useful as can be, although it minimizes typing and signing. It also downplays an important fact that philosophers have emphasized about language, namely that language is more than communication. It is a social action that requires work to be done. Language is a system that speakers, writers, and signers intentionally exploit. It is used to do things, not just report, describe, or discuss them. "That shirt looks great on you!" it's not just a report (while "Halloween falls on a Tuesday this year" might as well be). Most likely it's a compliment. "Outside!" It's just opinion or conjecture when a fan behind home plate yells this at a baseball game, but the umpire tells them, "Out!" it is a call and as such can end an inning or game. As we have already said, people have been interested in the analysis of language for thousands of years. Plato and Aristotle discussed language in the 4th and 3rd centuries BC. C., and from them we inherited several categories of grammatical analysis. more than a century

What is linguistics?

• 23

Liar,[email protected]I have written a description of Sanskrit, which is one of the finest grammars ever produced for any language. Today, the empirical study of language has gained importance at a time when communication is fundamental to social, intellectual, political, economic, and ethical issues. Now complemented by insights from cognitive and neurosciences, computer science, psychology, sociology, anthropology, philosophy and rhetoric, as well as communications engineering and other sciences, linguistics has become a prominent academic discipline at universities and research centers – worldwide. .

What are the branches of linguistics? Historically, language studies have focused on grammar: speech sound patterns, word structure, sentence formation, and meaning. More recently, attention has also turned to the relationship between expression and meaning on the one hand and context and interpretation on the other. This area is called pragmatics. Some linguists describe specific languages; others examine universal patterns in different languages and try to explain them in cognitive or social terms. Some linguists focus on linguistic variation between language communities or within a single community over time or in usage situations such as conversation and sports speakers. Linguists studying variation look for two types of explanations: cognitive explanations related to limitations in human language processing abilities, and social explanations related to social interaction and the organization of societies. A third group of linguists applies the discipline's insights to real-world problems in education, literacy (reading and writing), and second and foreign language acquisition. clinically understand aspects of Alzheimer's disease and aphasia; in forensics to analyze conversations for evidence of conspiracy, threats, defamation and other legal matters, interpret contracts (from rental agreements to insurance policies to agreements to build airplanes), safety notices (e.g. medical labels and dosage instructions). clarify , and speech recognition and document creation. Some applied linguists deal with language policy issues at national and local level: which languages should be designated for use in schools, courts, ballot boxes, etc.? what kind of writing system to use in a culturally diverse modern nation; what regulation of existing language is needed, such as in the Plain English movement in the United States or in the development and production of standardization tools such as dictionaries and grammar books. As the world shrinks and cultures merge, linguists are also applying their skills to the challenges of cross-cultural communication.

Computers and Linguistics At the end of each chapter in this book is a section that covers some aspects of computers and language that are related to the topics covered in that chapter. You don't need to be computer literate to understand these sections and

AND

benefit from them. Seeing the content of a chapter from a different perspective will help you understand it. The next section serves as an introduction to the parallel sections in later chapters.

2 4 • Chapter 1 Languages and Linguistics

Computer Science and Linguistics What is Computational Linguistics? Computational linguistics aims to test language theories and apply linguistic knowledge to real-world problems using computers. To understand how the use of computers can test theories of language, it is helpful to think of linguistics as an attempt to make explicit what speakers know implicitly about their language. Imagine creating a model of what a child needs to know in order to use your language. Even the simplest model would require a list of items (such as words) and a set of rules to combine them into strings that resemble the child's speech. To the extent that the implicit knowledge possessed by a fluent speaker can be made explicit in the model, researchers can test the accuracy of the model and thus their own understanding of human language on the computer. In other words, a program containing the elements and rules of the model produces strings of words that can be checked to see if they are in fact possible sentences. If a program produced strings like "Caught three fish for Meg" or "Caught three fish for Meg", you would know that your model is wrong. So far it has proved impossible to make fully explicit what even a child knows about their language. Here are two examples of how these programs can be applied. Linguists would like to write programs to synthesize speech from written text. It could then send a printed page to a synthesizer, which would efficiently and naturally read it aloud. Text-to-speech synthesis machines exist, but linguists are far from satisfied with their success. (You can listen to the synthesized speech of your written sentences at the websites listed at the end of Chapter 4.) The other side of speech synthesis is speech recognition. Linguists would like to know enough about language interpretation for computers to translate speech into writing and execute spoken commands. A successful speech recognition program would allow physicians to verbally report their findings during a patient's physical examination and automatically convert the verbal reports into written ones. This task is relatively easy for a human transcriber, but it's not well understood, so we haven't yet been able to allow machines to do it as well as we'd like. Today specific speech and speech recognition and speech synthesis

Tasks can be accomplished, and in later chapters you will see what has been accomplished so far.

Computers and machine-readable text In the 18th century, Samuel Johnson's dictionary provided vivid citations from books to illustrate how words were used. During his own reading, Dr. Johnson phrases whose context made the meaning or use of a word particularly clear. His assistants then transcribed the passages onto sheets of paper, and he organized them into dictionary entries. Essentially the same process was used to compile the Oxford English Dictionary in the 19th century. This project required thousands of readers and lasted half a century. In the 20th century, the creators of Webster's Third New International Dictionary also trawled through a collection of several million citations to discover different meanings of words. Currently, the writing of dictionaries is undergoing dramatic changes due to advances in computers and the availability of machine-readable bodies of text known as corpuses. Corpus linguistics is the gathering of text collections and their use to study language usage. In this context, a corpus is a representative set of texts (corpus is the Latin word for "body"). You know the types of machine-readable text that word processors produce, and the fact that it's machine-readable allows you to search for a specific word or phrase. The first computerized corpus, the Brown University Corpus, comprised 500 texts from American books, newspapers and magazines. The lyrics were chosen to represent 15 genres, including science fiction, romance, news reporting, and academic and scholarly writing. Each text contains 2000 words and the total collection contains one million words. Later, the researchers assembled a parallel corpus of British English called the London-Oslo/Bergen Corpus, or LOB for short. These two corpora are parallel collections of American and British writings that appeared in print in 1961. Recent corpora contain more than 100 million words, and corpora of texts in multiple languages are being compiled. Corpuses are proving to be essential for creating 21st-century dictionaries and in many other ways, including speech recognition and artificial intelligence. ■

Continue

Summary • The total number of spoken and signed languages in the world is between 6,000 and 7,000. • According to 2000 US Census data, 47 million US citizens age 5 and older spoke a native language other than English. • The United States has no official language and has never had one. • Human language is an extremely complex system that children can easily master in a remarkably short time. • Natural processes of language change affect all languages over time, and language change is not language decay. • All languages are equally logical (or equally illogical). • Human language is primarily a system of arbitrary signs, but some linguistic signs are representative. • Grammar is a system of elements and patterns that organize linguistic expression. • The five characteristics of human language systems are arbitrariness, discretion, duality, repression and productivity. • Rather than being a two-sided coin, a language system can best be viewed as a triangle whose sides are meaning and expression and whose base is context. • Verbal communication can function in three modes: spoken, written and signed. • Everyone speaks a dialect, and a language includes all its dialects. • Chimpanzees do not have a proper vocal tract for speaking, but are able to a certain extent to combine multiple signals into a meaningful sequence. • The degree of similarity between the language of chimpanzees and that of very young children is still under investigation, but consensus suggests that they differ fundamentally, at least in movement and productivity. • Computers can be used to test models of language that are supposed to exist in the brain. • In the emerging field of corpus linguistics, large sets of computerized texts, called corpora, are used to explore the use of natural language in all its contexts.

WHAT DO YOU THINK ? R E V I S I T • How many languages? The question cannot be answered exactly, but it is there

there are between 6,000 and 7,000 languages in the world. • Multilingual utility bill. The United States has never had an official language. Customers need to understand utility bills and the only way to communicate with people is in a language they understand. We are all relieved when we come across an important text in a language we do not understand and then find a translation in a language we are familiar with. The United States accepts large numbers of immigrants, and when they crowd into urban or suburban areas, it makes sense for businesses and businesses

• 25

2 6 • Chapter 1 Languages and Linguistics

government agencies to communicate with them in a language they are familiar with. Even voting materials are available in multiple languages depending on what is commonly spoken in a community. • Nan, the newspaper reader. Each word form can have multiple meanings. The note has different meanings in musical note, banknote and metaphorical discordance note. Note can be not only a noun but also a verb. Therefore, context of use can sometimes be essential in determining the correct meaning of a word in an utterance. • Prisoner or poacher? Traditionally, sneaking was the final form of sneaking. But English speakers are increasingly speaking and writing in secret. Languages change and what is “right” for one generation may not be for the next. As usage changes, judgments of good and evil change as well. People often cite "the dictionary" as the authority for one use or another, and some dictionaries call themselves "authoritative." But dictionaries also differ in their right and wrong philosophy of use. Dictionary compilers know that whatever position of authority they hold depends on actual usage: what is being said and written at any given moment. • Seth's birthday. Seth understands that statements must be interpreted in context. Frank states that the statements have only a literal meaning, regardless of the context. Since everyone knows their own birthday, Frank's question would seem silly if taken literally, so Seth has to find an interpretation that makes sense. The "9th May" is information that Seth assumes Frank doesn't have: why else would he ask the question?

Exercises 1-1. Here are a series of questions that can form the basis of your linguistic autobiography. Think about it and write down your answers in bulleted form. (1) When did you first notice that people judge certain linguistic expressions as bad or good, and what do you think the basis of these judgments must have been? (2) When did you first notice that some people judge certain linguistic expressions as grammatically correct or incorrect, and what do you think the basis for your judgment must have been? (3) Since when do you think speaking is more fundamental than writing? (4) Was there a time when you considered writing to be the basis of speaking, and if so, what gave you that impression? (5) What aspects of your current views on language place scripture in a superior position to language? 1-2 Throughout the day, write down all the instances you hear (on radio or TV shows, in lectures, or in conversations with your acquaintances) of different types of representative speech (duration, volume, speed, repetition, intonation, ordering, etc.). (You may find it easier to collect examples from sitcoms or children's shows.)

Exercises 1-3. Below is a list of features that describe verbal communication through speaking, writing, and signing. Decide which types of linguistic communication the feature applies to and give an example to illustrate your claim. Pay particular attention to the different types of spoken, written, and signed communication, as some of these characteristics may apply to some types of communication but not to others. Also consider the impact of modern communications technology on these traits. a. A verbal message is transient, which means it cannot last. B. A voice message can be reviewed after it is created. C. A verbal message has the potential to reach a large audience. i.e. A verbal message can be transmitted over long distances. Y. A linguistic message can depend on the context in which it occurs; the producer can refer to the time and place where the message was produced without fear of misunderstanding. Q. A verbal message is based on the senses of hearing, touch and sight. grams. The ability to produce verbal messages is innate; it doesn't have to be learned consciously. h A verbal message must be carefully planned before it is produced. I. The production of a spoken message can be carried out simultaneously with another activity. 1-4 Consider the following quote from a mid-20th century dictionary (A Proounce Dictionary of American English, by John S. Kenyon and Thomas A. Knott, Springfield, MA: Merriam, 1953, p. VI). As with all reputable dictionaries, the editors have endeavored to base the pronunciation on actual civilized usage. In fact, no other standard has specified the pronunciation. This book can only be taken as a safe guide to pronunciation as much as possible. According to this standard, as is often said, no word is "mispronounced almost universally" because that is contradictory. There is often a great temptation for an editor to prefer what he believes to be the correct pronunciation 'should'; but you have to resist it. a. Argue that publishers should resist the temptation to record their personal pronunciation preferences in a dictionary. Please explain whether your argument also applies to a publisher expressing personal preferences for other aspects of the language, such as spelling or usage. B. He argues that the phrase "mispronounced almost everywhere" is contradictory. C. What do you understand by the term "cultivated use"? How would you determine what usage is "civilized"? How do you imagine a dictionary publisher would determine which uses are "cultivated"? In your opinion, what use should a dictionary describe? Explain your point of view. 1-5 In assignments and exams comparing natural conversation to written variants of English, students sometimes claim that the conversation is riddled with errors like those shown below. Offer an alternative explanation for the claim that they are bugs.

• 27

2 8 • Chapter 1 Languages and Linguistics

I said hello and she said hello. I said Hi Pat, I said she said Hi Chris. 1-6 Consider the following from John Simon (Paradigms Lost, New York: Penguin, 1980, pp. 58-59) on Edwin Newman's book A Civil Tongue: With demonic insight, Newman presents 196 pages of grammatical errors. Clichés, jargon, malapropisms, mixed metaphors, monstrous neologisms, profane ambiguities and parasitic redundancies, interspersed with his own mocking commentary. . . and requests to do better. The examples are mostly open horror, very funny and even more harrowing. . . . Worse than a nation of shopkeepers, we have become a nation of orators or word-busters, and the misuse of language, whether through ignorance or obscurity, leads, as Newman convincingly argues, to a degradation of moral values and standards of life. of life. . a. Simon seems to equate "grammatical errors" with cliches, slang, uncleanliness, etc. Which of these can be described as grammatical errors in the linguistic sense? What is the best way to characterize the others? B. Name two ungrammatical structures that you have heard from non-native speakers of English. Have you heard similar grammatical errors from native speakers? What do you think is the reason for your findings on native speaker errors and non-native speaker errors? C. Newman and Simon's argument about the “abuse of language” leading to a deterioration in moral values and living standards is a common assertion by language watchdogs. What kind of abuse does Simon seem to have in mind when making this statement? Are he and Newman right when they claim that such abuses lead to a degradation of moral values? Could it be the other way around? What part could anyone have in promoting the Newman/Simon claim? (Who are the winners and who are the losers if this view prevails?) d. Do you think genuine grammatical mistakes (like those made by non-native speakers) can lead to moral degradation? Explain your position. 1-7 Writing and gestures are visual forms of verbal communication. What is the relationship between writing and braille (the writing system for blind readers)? Is Braille a linguistic means of communication? How many types of verbal communication are there? 1-8 What are the pros and cons of each mode when choosing between voice modes e.g. B. call a distant friend or send a letter? List some of the circumstances in which each language would be preferred over the other. 1-9 Please list the two strongest reasons you heard for bilingual education programs in schools in your community and the two strongest reasons you heard for monolingual programs in English. How do you rate these arguments?

Another resource

Especially for educators and prospective teachers 1-10. Do you think that for students whose mother tongue matches the language of instruction in school, the primary focus of language arts instruction is reading and writing or speaking and listening? Explain your position. 1-11 Do you think that for the same group of students, the real emphasis in the curriculum is reading and writing or speaking and listening? Explain the basis of your opinion. 1-12 For students whose mother tongue differs from what is taught in school (e.g. students who speak Spanish at home but attend an English language school), would your answers to the two questions above be different? If so, how? 1-13 For students whose native language is a different dialect than the one taught at school, your answers to questions 1-10 and 1-11 would be different (focus on your local situation or the situation in a district where you are likely to work) . If so, how? 1-14 Did your teachers speak the same language as you in your early school years? Same dialect? If not, did they express different attitudes towards his speech and yours? Were there discussions about other language varieties and can you understand which attitudes your teachers promoted towards the language varieties of other students? Can you remember something a certain teacher said about other languages or other dialects? Did you feel comfortable speaking in class? Do you think everyone in your class felt the same as you? Remember a time when a teacher spoke about the importance of language in every child's life and how central it is as an aspect of a person's personal identity? 1-15 At some point during your high school and college years, did anyone give you a glimpse of what they thought of your speech? If so, who were they and what were their attitudes?

Other Internet Sources Extensive information and laboratory research is available on the Internet. In this section of each chapter you will find instructions that can help you understand the chapter and provide a laboratory never before available to linguistics students, even in the best-equipped universities. As you know, web addresses can change unexpectedly, so the dates below may have changed when you try. If they have moved there is sometimes an automatic connection to the new address. You can also check the first website below for updated addresses. There you will also find new addresses that might interest you. LISU website: http://www.thomsonedu.com/english/finegan

For users of this book. Offers up-to-date internet addresses and supplementary material for students and teachers. The field of linguistics: http://www.lsadc.org/info/ling-fields.cfm

For general information, see the Linguistic Society of America website. Here you will find short treatments on language and thinking, computer science and language,

• 29

3 0 • Chapter 1 Languages and Linguistics

Languages, prescriptivism, writing, slips of paper, language and the brain, linguistics and literature, and a dozen other subjects. An animated ASL dictionary: http://commtechlab.msu.edu/sites/aslweb/browser.htm

You can view animated representations of ASL signs. James Crawford Language Guide Website and Emporium: http://ourworld.compuserve.com/homepages/jwcrawford/home.htm

A rich source of information and interpretation of language policies in the United States and various states. You can link to Crawford's Obituary: The Bilingual Education Act, 1968-2002. In his 2000 Census: A Guide for the Perplexed, you'll find notable facts about language trends in the United States as reflected in census data, Crawford's interpretation of those facts, and suggestions for diversity websites. only, English-plus and related topics. Ethnologue website: http://www.ethnologue.com/

Lots of information about language distribution, number of speakers, dialects, etc. Organized by country name and language. Census 2000 Gateway: http://www.census.gov/main/www/cen2000.html This official US Census Bureau website allows you to see what languages are spoken by how many residents in each state and country, with easy readable tables.

Video • The human language series

A series of award-winning videos originally aired on PBS in 1995: Discovering Human Language: "Colorless Green Ideas"; Human Language Acquisition: “Playing the Language Game”; and The evolution of human language: "With and without words". The 55 minute videos are informative and entertaining. Produced by Equinox Films, Inc.

Suggestions for further reading • Jean Aitchison. 1996. The Seeds of Speech: Origin and Evolution of Language (Cambridge: Cambridge University Press). A fundamental treatment of the beginnings of the language. • Douglas Beaver, Susan Conrad and Randi Reppen. 1998. Corpus Linguistics: Examining the Structure and Use of Language (Cambridge: Cambridge University Press). An accessible introduction to corpus linguistics. • David Crystal. 1997. Cambridge Encyclopedia of Language, 2nd ed. (Cambridge: Cambridge University Press). Covers topics in a few pages each with illustrations. • Edward Finegan and John R. Rickford, eds. 2004. Language in the US: Issues for the 21st Century (Cambridge: Cambridge University Press). 26 essays are aimed at a broad audience and cover topics such as multilingualism, Spanish in the Southwest and the

Advanced Reading Northeast, African American English, Asian American Voices, Ebony Controversy, Language and Education, Language of Cyberspace, Rap and Hip-Hop and Slang. • Ray Jackendoff. 1994. Patterns in the Mind: Language and Human Nature (New York: Basic Books). Accessible and fascinating discussion of the cognitive aspects of language structure and language acquisition. • Heinz Kloss. 1999. The American Bilingual Tradition, 2nd ed. (Washington, DC: Center for Applied Linguistics and Delta Systems). An introduction by Reynaldo Macias and Terrence G. Wiley helps put Kloss's classic work into context. • Donna Jo Naples. 2003. Language Problems: A Guide to Everyday Thinking About Language (New York: Oxford University Press). A nice introduction to most of the topics covered in this chapter. •Eduardo Sapir. 1921. Language: An Introduction to the Study of Language (New York: Harvest). This accessible classic continues to inform why it's still in print. • Deborah Tannen. 1986. I didn't mean that! How Conversational Style Builds or Destroys Relationships (New York: Ballantine). This bestselling book provides insight into the sometimes confusing connection between conversational styles and social and romantic relationships. • Simon Winchester. 1998. The Professor and the Madman: A Tale of Murder, Madness, and the Making of the Oxford English Dictionary (New York: HarperCollins). Whether you're interested in a story of murder and insanity or the genesis of OED, this book will grab your attention and prove that lexicographers are no harmless slaves. • Simon Winchester. 2003. The Meaning of Everything: The History of the Oxford English Dictionary (Oxford: Oxford University Press). An engaging introduction to OED creation and lexicography in general.

Further reading Crystal's Dictionary of Linguistics and Phonetics (2003) is a useful reference for a wide variety of terms and concepts. Schiffman (1996) contains useful and interesting chapters on language policy in the United States and California and other parts of the world. For a discussion of the relationship between arbitrary and non-arbitrary signs, see Saussure (1959). Haiman's (1985) articles deal with iconic elements in syntax and intonation. For speaking and writing sources, see “Suggested Reading” in Chapter 12. For standard varieties and attitudes to using English correctly, see Finegan (1998) with an emphasis on British and Finegan (2001) with an emphasis on American, and Milroy and Milroy (1999). ) and Wardhaugh (1999). For information on American Sign Language, see Lucas and Valli (2004), on which our discussion here is based, and for a review of Native American and Australian Sign languages, see Umiker-Sebeok and Sebeok (1978). . For the origins of the language, see Lieberman (1991). For corpus linguistics see McEnery and Wilson (1996), Kennedy (1998) or Meyer (2002); for computers and language, see Barnbrook (1996) or Lawler and Dry (1998).

• 31

3 2 • Chapter 1 Languages and Linguistics

References • Barnbrook, Geoff. 1996. Language and Computers (Edinburgh: Edinburgh University Press). • Crystal, David. 2003. Dictionary of Linguistics and Phonetics, 5th ed. (Oxford: Blackwell). • de Saussure, Fernando. 1959. Course in General Linguistics, trans. from the French by Wade Baskin (New York: Philosophical Library). • Finegan, Eduardo. 1998. "English Grammar and Usage." In S. Romaine, ed. Cambridge History of the English Language, Vol. 4 (Cambridge: Cambridge University Press), pp. 536–88. • Finegan, Eduardo. 2001. "Use." In J. Algeo, ed. Cambridge History of the English Language, Vol. 6 (Cambridge: Cambridge University Press), pp. 358–421. • Gordon, Raymond G., Jr., eds. 2005. Anthropologist: Languages of the World, 15th ed. (Dallas, Texas: SIL International). • Haiman, John, ed. 1985. Iconicity in Syntax (Amsterdam: Benjamins). • Kennedy, Graeme. 1998. An Introduction to Corpus Linguistics (London: Longman). • Lawler, John M. and Helen Aristar Dry, eds. 1998. The use of computers in linguistics: a practical guide (London: Routledge). • Liebermann, Philip. 1991. Uniquely Human: The Evolution of Language, Thought, and Selfless Behavior (Cambridge, MA: Harvard University Press). • Lucas, Ceil and Clayton Valli. 2004. "American Sign Language." In Edward Finegan and John R. Rickford, eds. Language in the US (Cambridge: Cambridge University Press). • McEnery, Tony and Andrew Wilson. 1996. Corpus Linguistics (Edinburgh: Edinburgh University Press). • Meyer, Charles F. 2002. English Corpus Linguistics: An Introduction (Cambridge: Cambridge University Press). • Milroy, James and Lesley Milroy. 1999. Authority in Language: Investigating Language Prescription and Standardization, 3rd ed. (London: Routledge). • Schiffman, Harold F. 1996. Language Culture and Language Politics (London: Routledge). • Umiker-Sebeok, Jean D. and Thomas A. Sebeok, eds. 1978. Aboriginal Sign Languages of the Americas and Australia, 2 vols. (New York: Plenum). • Wardhaugh, Ronald. 1999. Proper English: Myths and Misconceptions About Language (Malden, MA: Blackwell).

part One

Language Structures Ords are at the heart of language, and when you think of languages, you usually think of words. In examining language in this book, words are the focus, and we begin our examination of language structures by looking at words from different perspectives:

C • • • •

o o o o

essential parts of words sounds and syllables that make up words principles that organize words into phrases and sentences semantic relationships that bind words together

In the first part of this book you will see how speech sounds result from a few elements, how a large number of syllables are assembled from a few speech sounds, how syllables combine to form meaningful parts of words, and how languages pack these parts of words and a finite vocabulary into infinite numbers Sentences. You'll also see how systematic principles of language structure help you understand expressions even if you've never heard or read them before. Finally, you examine the semantic relationships that organize groups of words.

Words and their parts: lexicon and morphology

WHAT DO YOU THINK ? • Suppose you are the parent of a three-year-old who asks you if you 'baked' a cake and 'talked' to her friends. How would you describe the pattern your daughter uses to mark the past tense of these verbs? • They agreed to create a list of food and drink that volunteers could contribute to a fundraiser for a college choir that tours internationally. All foods must have a name that English borrowed from another language. One quickly thinks of Japanese sushi, Chinese chop suey, Spanish omelette, French champagne and Indian curry. What other food names can you think of in these languages? • If you had to guess the “top ten words” in printed English, what would they be? Why did you choose this one? • A friend mentions that Washington state is named after a famous person, but most other states have names that don't mean anything in English. "Those are just names!" Where did the names Delaware, Missouri and Illinois come from? What about Virginia and North Carolina?

Lexical categories (parts of speech)

Introduction: Words are tangible The most tangible elements of a language are words. Have you ever heard people say, "There's no such thing as a word" or "What does the word Lollapalooza mean?" Someone doing a crossword might have asked you, "What does a three-letter word mean augmented?" We say that one person likes to use "two-bit" words and another person prefers "four-letter" words. In these cases, people have clear ideas about what a word is. On the other hand, our intuitions are less sure about significant parts smaller than a word. We easily understand that the bookshelf, laptop, and headphones have two important parts, but our intuitions may be less sure when it comes to counter, tennis, femininity, song, and impracticality. This chapter examines words and their essential parts, as well as the principles that govern the composition of words and the functions of words in sentences. You will learn what it means to know a word and how languages expand your vocabulary.

What does it mean to know a word? Think about what a child needs to know to use a word. The kid who asks, "Can you take my shoes off?" knows a lot more about the word shoes than what it refers to. Know the sounds of the shoes and the order in which they occur. He knows that the word can be used in the plural (unlike milk, for example) and that the plural is not irregular like teeth or children, but is formed regularly. They also know how to use the word in a sentence. Using a word requires four types of information:

• their sounds and their order (this is called phonological information, topic of chapters 3 and 4)

• their meaning (semantic information, covered in Chapter 6) • how related words like plural (for nouns) and past tense (for verbs) are formed (morphological information, covered in this chapter)

• its category (e.g. noun or verb) and how to use it in a sentence (syntactical information, discussed here and in Chapter 5)

For children and adults, the use of any word requires information about sounds, meanings, related words, and usage in sentences, and this information needs to be stored in the brain's dictionary (called the mental lexicon, or lexicon for short). There are some parallels between the types of information stored in the lexicon and the types found in a desktop dictionary. Both contain information on pronunciation, meaning, related words and sentence usage. But a dictionary also contains information that is not necessary to talk about, for example the spelling of a word and its historical development (known as etymology). Dictionaries also illustrate how writers or speakers used a word. A mental lexicon does not usually contain etymological, illustrative, or orthographic information.

Lexical categories (parts of speech) The ability to use any word in a sentence requires knowledge of its lexical category. Even young children should know the category of every word they use: what are they?

• 35

3 6 • Chapter 2 Words and their parts: lexicon and morphology

verbs, nouns or adjectives. Of course, the child's knowledge is unconscious, and even a child grammarian would not normally know the names of the categories.

Identifying Lexical Categories There are several ways to identify a word's lexical categories, and to some extent they are based on similar principles that children use to discover the same information. A form focuses on closely related forms of a word. Fork and fork, book and books, truck and truck show parallel patterns of related forms, and words with parallel forms belong to the same category, in this case nouns. Words like old, big, and shiny follow a different pattern. Unlike nouns, viejo, alto y brillante have no cognate -s forms ("viejos", "tall" and "brilliantes"). Instead, the related forms have the endings -er and -est: older/older, taller/taller, lighter/lighter. Old, tall, and shiny are therefore members of another category called adjectives. Finally, words like jump and kick appear with parallel endings, including -ed for jumped and kicked, -ing for jump and kick, and -s for jump and kick. Other words that share this pattern are laugh, play, and come back, all of which fall under the verb category. Another way to identify categories focuses on what words and categories can appear together in sentences. For example, the above nouns and can be preceded by a (or an): a fork/the fork, and plurals in -s can be preceded. Basic adjectives like old, big, and bright can be preceded by much or much, as in very bright. Basic verbs can be preceded by can or will: go laugh. Below are examples of these patterns for the noun, adjective, and verb lexical categories. Bike nouns Tante Lager

Bike Aunt Camps

a bike an aunt a camp

the bike(s) the aunt(s) the camp

older newer redder

very old very young very red

seemed to be playing camp

watch while camping

can watch can play can camp

Adjectives old new red

older younger redder

Verbs look like play camps

seems to be playing camp

go fetch go play go camp

Knowing the typical cognate forms in each lexical category, we can verify that sharper is cognate as the adjective sharp (compare muito afiado, muito afiado), jaquetas as the noun jaqueta, and miss as the verb miss (missing/misses, can miss, can miss, you will lose). To find a word in a dictionary you need to look up its base form, as dictionaries do not have separate entries for words ending in sharper, jackets or miss. Children recognize early on that words belonging to different categories have characteristic endings or forms and characteristic distributions in sentences. (More technically, different categories have different patterns of inflection and category co-occurrence.) Relying on meaning is a third way of identifying lexical categories, although it is not always reliable and is mainly useful to support an initial hypothesis to form a word. Category. From the point of view of meaning, nouns designate (or refer to) people, places, or

Lexical categories (parts of speech) things. So swimmers, Cleveland and trees are nouns. Adjectives name qualities or properties of nouns, such as B. tall and impressive in the expressions tall trees or an impressive swimmer. The cognate forms of large can be identified as stronger and larger, but the cognate forms of impressive do not occur, although strong and impressive can be preceded by much, making them adjectives. Verbs describe actions like jumping and singing.

Verbs English-speaking children know that verbs have a number of related forms (speak, speak, speak, speak) and that the base form of the verb, the one with no ending, can be preceded by can or will. This knowledge is implicit; the child is unaware of this. Subcategories of Verbs In order to use a verb, a child must implicitly know the types of sentence structures that the verb allows. Since children store this knowledge in the mental lexicon, it makes sense to address it here. Consider items 1 through 6, where an asterisk marks the sentence as incorrect. 1st 2nd 3rd 4th 5th 6th

Sarah told a joke. *Sarah laughed at a joke. *Sarah said jokingly. Sarah laughed at a joke. *Sara said. Sarah laughed.

You can see that the verbs said and laughed don't leave the same structures behind. Tell clauses require a noun phrase (noun phrase here means one plus a noun) after the verb, as shown by the misspelled 5. But as 6 shows, not all verbs require a noun phrase after them. In fact, laughter does not allow a noun phrase to follow directly, as 2 shows, but it does allow the phrase to follow in a joke. Some verbs allow but do not require a noun phrase after them, as in 7 and 8:7. The diva played. 8. The diva played the piano.

Sentences 1 to 8 illustrate that words like count, laugh and play belong to the category of verbs, but they do not allow the same sentence structures as they belong to different subcategories. If children did not have accurate (unconscious) information about the sub-categorization of verbs, they would not be able to avoid the pronunciation of sentences like 2, 3 and 5. Verbs followed by a noun phrase are said to be transitive. Those that do not require a noun phrase are called intransitives. In a child's lexicon, each verb is classified as a verb and subcategorized as either transitive or intransitive.

Nouns Nouns are another lexical category. You have already seen that English nouns share certain form characteristics. They share a common set of endings or inflections. The inflection at the end of the forks gives information about the number. Number is the term used to cover singular and plural. Almost all nouns in English have different singular and plural forms, such as B. with the cat(s) and the "normal" plate(s) or the "irregular" tooth(s) and the child(ren). Some exceptions, such as deer and sheep, have the same singular and plural form. Not all languages mark the number in nouns. The Chinese is the one who doesn't do that.

• 37

3 8 • Chapter 2 Words and their parts: lexicon and morphology

Adjectives Many adjectives can be recognized by the pattern of their cognate forms, namely the endings -er and -est, as in big and big. But others, especially adjectives containing more than two syllables, do not allow these endings; *beautifuller and *beautifullest are not well-formed (and therefore asterisked). But beautiful is an adjective. This is demonstrated by co-occurrence patterns as with other adjectives. In particular, beautiful can be preceded by very or too. Like other adjectives, beautiful can come before nouns, as in beautiful flowers. As a third box, the only words that fit under "seems" or "he/she looks" would be adjectives: weird, capable, clear, funny, cute.

Pronouns Pronouns are a category of relatively few words divided into several subcategories. In addition to personal pronouns, there are demonstrative pronouns, interrogative pronouns, relative pronouns and indefinite pronouns. They replace noun phrases. Personal pronouns The most well-known pronouns are the personal pronouns, such as I, I, she, he, she and she. First, personal pronouns are distinguished from one another by representing different parts of a social interaction, e.g. B. a conversation. This aspect of pronouns is called the person: the first person is the speaker or speakers; the second person is the person or persons being spoken to (the addressee); and the third person are the people or things being talked about. First person: speaker: I, I, my, we, we, our second person: addressee: you, your, third person: spoken: she, she, she, he, he, his, it, it, they, her, your

Demonstrative Pronouns Demonstrative pronouns refer to things that are relatively close (this, these) or, conversely, relatively far away (that, those) when the referent can be identified by pointing on or out of the context of a discussion. Examples of demonstrative pronouns are what really bothers Guy, and that's Guys. Interrogative Pronouns Interrogative pronouns are used to ask questions. Who played Emma in Who? So what did you tell Rosie what? and what did you say to rosie? They are interrogative pronouns. In the sentence Whose are these? Whose is an interrogative pronoun (these are demonstrative pronouns). Relative Pronouns Relative pronouns have the same forms as other types of pronouns, but are used differently. Examples in the following sentences are who (in 1), that (in 2), and which (in 3). Other relative pronouns are who and who. Note that a relative pronoun is related to a preceding noun phrase. In the examples, the relative pronoun and associated noun phrase are in italics, with the relative pronoun underlined. 1. Ellen is a board certified gerontologist. 2. The program with the most prizes is "60 Minutes". 3. She is a licensed masseuse which I am not.

Indefinite Pronouns Indefinite is the name given to a set of pronouns whose references are identifiable but unspecified: one, some, someone, all, all, nobody, someone, all, all, nobody, something, anything, everything, nothing.

Lexical categories (parts of speech) pronouns are used independently and not as modifiers of other words. This is pretty clear for words like me and such, but some other word forms can be pronouns or belong to another category. who is in which and this are both pronouns. Whose Red Book is it in? who and this are determiners, as explained in the next section.

Determinators Determinators, a small category, precede nouns (a book, an orchestra, the musicians, this problem, these guys, what a movie, its iPod), although words in some other categories (a great book, an acclaimed orchestra, the best player). Determiners do not have endings like adjectives and verbs. They are divided into several subcategories:

• • • •

artigos definidos e indefinidos: the, a, an demostrativos: this, that, these, this Posesives: my, our, your, her, his, its, their interrogatives: which, what, who

Unlike nouns, adjectives, and verbs, categories whose members cannot be fully enumerated, determiners can be enumerated, as in the subcategories just listed.

Prepositions and Postpositions Prepositions form a class with enough members to be enumerated. They have no ends or other variations and their shape is unchanging. They usually precede a noun phrase, e.g. B. at a concert, on Tuesday or under the table. Prepositions indicate a semantic relationship between other entities. The preposition in The book is on (or under or near) the table indicates the position of the book in relation to the table. Note the underlined prepositions in Tina rode to (or from) Athens (indicating direction in relation to Athens) with (or without) Daniel (indicating accompanying her) to (or next to or beside) her (indicating Daniel's position in relation to Tina). ). Instead of prepositions, Japanese and some other languages have postpositions, which work like prepositions but follow the noun phrase instead of preceding it. Compare the Japanese-English pairs below: Japanese post positions

prepositions in english

Taroo no hasi von Tookyoo e

from taro with chopsticks to Tokyo

The placement of the prepositions that seems natural to speakers of English (and French, Spanish, Russian, and other languages) would seem unnatural to speakers of Japanese, Turkish, Hindi, and many other languages that posit this, rather than prepending this category. . .

Adverbs Adverbs have been called "the most nebulous and confusing of traditional word classes," and that category doesn't fit English very well. For one thing, adverbs cannot be identified by their form alone and generally have no related forms. Many adverbs are

• 39

4 0 • Chapter 2 Words and their parts: lexicon and morphology

derived from adjectives by adding -ly, such as fast, odd, and possibly (from the adjectives fast, odd, and possible), but others, including the most common, bear no distinguishing marks. Also, not all words ending in -ly are adverbs (e.g. masculine and celestial are adjectives). The very common words then, now, here, soon and far can be identified as adverbs simply by their arrangement in sentences, bearing in mind where they occur and with which categories they coexist. The meaning of an adverb can also indicate its category, as adverbs usually indicate when (now, then, often), where (here, there), how (rapidly, suddenly, violently), or to what extent (too a lot of) . . . Some grammarians do not treat the word as an adverb. Grammatically, adverbs perform a number of functions, including modifying verbs, adjectives, other adverbs, and complete sentences. Adverbs modifying verbs (sentences with related adjectives in brackets) He spoke aloud. (She spoke loudly.) She was fast asleep. (She was fast asleep.) She thought quickly. (She was a quick thinker.) They studied hard. (They were diligent students.)

Adverbs modifying verbs He often spoke. She studied here. They will arrive soon. She believes it now.

Modifying adverbs are italicized below and modified adjectives or adverbs are underlined. Adverbs that modify adjectives

Adverbs that modify adverbs

a very tall tree

soon much

a very cold winter

incredible fast

a really great night

really incredibly fast

Sentence-Modifying Adverbs Actually, it was Danielle who said that. I honestly don't know why I'm here. Unfortunately the comment wasn't funny.

Conjunctions There are two main types of conjunctions. Coordinating conjunctions such as and, but and or are used to connect expressions of the same category or state, for example noun phrases (dungeon and dragons, tea or coffee), verbs (sing and dance, travel and fall), adjectives (slow and slow). painful, hot and cold) and clauses (she sang and he danced). Subordinating conjunctions are words like "always," "during," and "because" that join clauses in an uncoordinated role, as in "She was visiting Montreal while she was telling Bates College" or "He said she was ill." (Subordinate clauses are discussed in Chapter 5.)

Morphemes are parts of words that have a meaning. Subordinating conjunctions are often simply referred to as subordinating conjunctions and coordinating conjunctions as conjunctions or coordinating conjunctions.

Morphemes are parts of words that have a meaning. They know that words like girl, question, big, guy, and orange can't be broken down into smaller meaningful units. Orange does not consist of o + gamma or range. None of the uncles are uncle and uncle. But most words have more than one significant part. You will find two items each in Grandma, Bookshelf, Innkeeper, Order, Workshop, Oranges and Uncles. Pretty, supermarket and decadent also have more than one important element. A set of words can be constructed by adding elements to a core element, since the following is built around true: true true true

false true true

really wrong wrong

These words share a common root whose meaning or lexical category has been modified by the addition of elements. The essential elements of a word are morphemes. So True is a morpheme; false and true each contain two morphemes; Falsehood contains five (UN- ⫹ TRUE ⫹ -TH ⫹ -FUL ⫹ -NESS). Truer has TRUE and -ER ("more") elements. The morphemes in True are TRUE and -LY; to false, TRUE and A-; in truth TRUE ⫹ -TH ⫹ -FULL. Most morphemes have a lexical meaning, such as B. Gaze, Kite, and Height. Others represent a part of speech or a semantic term, such as the past tense (the -ed in sight) or the plural (the -s in cometas) or the comparative (the -er in high). Don't be fooled into equating morphemes with syllables. Note that Harvest, Grammar, and River contain two syllables, but only one morpheme each. Gorilla contains three syllables; Connecticut contains four; and hippopotamus five syllables; Each of these words is a single morpheme. Conversely, a single syllable can represent more than one morpheme: the monosyllabic kissing contains two morphemes (KISS ⫹ 'PAST TENSE'); as well as dogs (CÃO ⫹ 'PLURAL') and feet (FOOT ⫹ 'PLURAL'). Men contains three morphemes in a single syllable (MAN ⫹ 'PLURAL' ⫹ 'POSESSIVE').

Morphemes can be free or bound Some morphemes can be single words: TRUE, MOM, ORANGE. Others only work as part of the word: UN-, TELE-, -NESS, and -ER. Morphemes that can stand alone are free morphemes. Those that cannot are bound morphemes.

Try it yourself: Identify all morphemes in these words and whether they are free or bound: Bakery, Baseball, Frontier, Cider, Dusty, Fried, Outlaw, Sentence, Premade, this.

• 41

4 2 • Chapter 2 Words and their parts: lexicon and morphology

Morphemes Derived from Other Words Certain connected morphemes change the category of the word to which they are attached, as do the underlined parts of these words: dubious, established, obscure, startle, and teacher. When added to the noun doubt, -FUL derives the adjective doubtful; -MENT added to the verb establish derives the noun establishment. Dark is an adjective, darken a verb; startle a noun, startle a verb; teach a verb, teach a noun. In English (but not all languages), derived morphemes are added to the end of words (and are called suffixes). We can represent these relationships in the following derivation rules: noun ⫹ -FUL adjective ⫹ -LY verb ⫹ -MENT verb ⫹ -ER adjective ⫹ -EN noun ⫹ -EN

→ → → → → → →

Adjective (doubtful, beautiful) Adverb (beautiful, real) Noun (elevation, amazement) Noun (master, lord, tension) Verb (sweet, ease, harden) Verb (frighten, hurry, baptize)

A similar process uses morphemes added to the beginning of a word (called prefixes). Prefixes in English generally change the meaning of a word but not its lexical category. ⫹ verb ⫹ adjective UN- ⫹ verb SOB- ⫹ verb RE- ⫹ verb EX- ⫹ noun MISUN-

→ → → → → → →

Verb (misspelled, err, err, err, misclassified) Adjective (rude, unfunny, unfair, unfaithful, wrong) Verb (undo, unleash, uncover, unravel, undress) Verb (undercut, underestimate, underestimate, underestimate, underline ) verb (restitution, reformulate, paraphrase, reevaluate) noun (ex-police, ex-nun, ex-husband, ex-convict)

Derivative processes that transform a word into another word that has a related meaning but belongs to a different lexical category are common in the world's languages. Here is an example from Persian. (Note: æ is pronounced like a in English hat and x like ch in German Bach.) dærd 'pain' næm 'moisture' xætær 'danger'

dærdnak 'painful' næmnak 'wet' xætærnak 'dangerous'

The suffix -nak can be attached to certain nouns to derive adjectives. Thus, Persian has the following derivation rule: noun X ⫹ -NAK → adjective 'the property of being or having X'

Another Persian suffix creates abstract nouns from adjectives, as these word pairs show: gærm "warm" pæhn "broad"

gærma 'warmth' pæhna 'width'

This process of derivational morphology can be expressed by this rule: adjective ⫹ -A → noun

Not all words belonging to the lexical category can undergo a particular derivation process. In English, the nouns dude and beauty can have the suffix -ful, but rust and book cannot. Unless the words are marked in the lexicon for specific derived statements

How are morphemes arranged in words? In such cases, the ungrammatical forms *rustful and *bookful would result instead of the grammatical forms rusty and bookish derived from other rules. In Fijian, vaka-, meaning "in the manner of," is a derived morpheme that can be placed before adjectives and nouns to derive adverbs according to these two rules: VAKAVAKA-

⫹ adjective → adverb ⫹ noun → adverb

The following adverbs show the morpheme VAKA-: vaka-viti 'Fiji fashion' (from Viti 'Fiji'), vakatotolo 'quick, quick' (from totolo 'quick, quick'); To illustrate the derivation of a noun, consider vakamaarama "lady" (formed from the prefix vaka- for maarama "lady"). Not all linked morphemes change the lexical category of words. Adding the conjoined morphemes DIS-, RE-, and UN- (disappear, repaint, unfavorable) to a verb changes its meaning but not its lexical category. For example, appear and disappear are verbs, just like paint and repaint; favorable and unfavorable are both adjectives. There is a notable tendency in English to add morphemes that change meaning without changing the lexical category at the beginning of words as prefixes, although this is not universal across languages (and some languages lack prefixes, such as Turkish ). The two types of morphemes we have just examined are derived morphemes. You create new words from existing words in two ways. You can change the meaning of a word: true versus false; paint vs. repaint. Or they can change the lexical category of a word: true is an adjective, true is an adverb, true is a noun.

Inflectional morphemes Another type of bound morpheme is represented in the underlined parts of the words cats, collected, sleeps and stronger. These inflectional morphemes change the form of a word but not its lexical category or core meaning. Inflectional morphemes create different forms of a word to fulfill different roles in a sentence or language. In nouns and pronouns, inflectional morphemes are used to denote semantic concepts such as number or grammatical categories such as gender and case. In verbs they can mark categories like tense or number, while in adjectives they indicate degree. They form the "related forms" we used earlier in this chapter to identify word categories. We will come back to inflectional morphology later in this chapter.

How are morphemes arranged in words? Morphemes are arranged in an order Within a word, morphemes have a strict and systematic linear order; they are not randomly arranged. Affixes Some morphemes, called suffixes, always follow the root to which they are attached, such as "PLURAL" in girls and -MENT in commit: Both *sgirl and *mentcommit are buggy. Prefixes attached to the front of a stem, such as B. wrong, are faded and repainted. (Compare *trueun, *appeardis, and *paintre.)

• 43

4 4 • Chapter 2 Words and their parts: lexicon and morphology

Derived morphemes can be prefixes (unhappy, disappear) or suffixes (lucky, looks). In general, inflectional morphemes are added to the extreme parts of words. Together, the prefixes and suffixes are called affixes. Infixes In addition to affixes, some languages have infixes. An infix is a morpheme inserted into another morpheme. Tagalog (spoken in the Philippines) has infixation. For example, the word gulay meaning "greenish vegetable" can be given the ending -in-, giving rise to the word ginulay meaning "blue-green".

Morphemes can be discontinuous Not all morphological processes can be viewed as a union or concatenation of morphemes by adding a continuous string of sounds (or letters) to a root. In other words, not all morphological processes involve prefixes, suffixes, or infixes. The technical term for discontinuous morphology is not concatenative. Circumfixes Some languages combine a prefix and a suffix into a circumfix, a morpheme that occurs in two parts, one on each side of a root. Samoan has a morpheme FE-/ -A?I, meaning "reciprocal": the verb "fight" is finau, and the verb "fight each other" is fefinauaʔi – FE ⫹ FINAU ⫹ AʔI. Interlocking Morphemes Semitic languages such as Arabic and Hebrew can have interlocking morphemes. For example, Arabic nouns and verbs often have a stem made up of three consonants, such as B.KTB. The Arabic word for "book" is kitaab. Through the interweaving of K-T-B and various other morphemes, Arabic generates many nouns, verbs, and adjectives from this single root. All nouns and verbs in Table 2-1 contain the same KTB stem with different interlocking morphemes.

TABLE 2-1 Derivative morphology in Arabic

livro kaatib school textbook book book

'to write' 'writer' 'office' 'library' 'letter' 'typewriter' 'bookseller'

kataba kaataba ?aktaba ?iktataba takaataba inkataba iktataba

'wrote' 'corresponded to' 'dictated' 'registered' 'letters exchanged with' 'he signed' 'had a copy made'

Incidentally, the English words Muslim, Islam and Salaam all contain the Arabic root SLM with its central meaning "peace, submission".

Root words contain fused morphemes Another phenomenon links multiple morphemes in such a way that the sounds of the word cannot be uniquely assigned to each of its morphemes. The classic example is the

How are morphemes arranged in words? French word du combining the two morphemes DE 'of' and LE 'the'. You can see how difficult it is to assign sounds to one morpheme or another. Some analysts call the mixtures smog (of smoke and fog) acronym words.

Morphemes intersect within words Morphemes are arranged in a highly standardized manner. They have an apparent linear order as well as a layered structure. False is true with a prefix (not a appended with true). Truly consists of a true stem with the suffix -ful (truth is true with -th appended). Untrue would be parsed incorrectly if we said it was composed of false with the suffix -thful . Instead, it is true without a prefix. Now consider uncontrolled. Could it be controlled without a prefix? Or go crazy with the suffix -ably? It is helpful to think of the sequence of control layers of the root morpheme as constructed by a set of derivation rules that are widely used for other words as well: control (verb) verb ⫹ -ABLE → Controllable adjective (adjective) UN-

⫹ Adjective → Adjective

uncontrollable (adjective) adjective ⫹ -LY → adverb uncontrollable (adverb)

The root of uncontrollable is control, which works like the root of -able; controllable functions as the root of the uncontrollable; and uncontrollable functions as the root of the uncontrollable. The structure can be represented with the tree diagram in Figure 2-1 or with square brackets as follows: [[un[[controlVerb]ableAdj]Adj]lyAdv]

Figure 2-1 Hierarchical Structure of Uncontrollable Adverb Adjective Adjective Verb a

check

capable

lie

• 45

4 6 • Chapter 2 Words and their parts: lexicon and morphology

How does a language expand your vocabulary? Languages have three main ways of expanding their vocabulary:

• New words can be formed from existing words and parts of words. • Words can be "borrowed" from another language. • New words can be invented, created from scratch

Some word classes are open, others are closed. In some societies, the need for new nouns, adjectives, and verbs frequently arises, and additions to these categories occur voluntarily. For this reason, nouns, adjectives, and verbs are called open classes. Other categories such as prepositions, pronouns, and modifiers are closed classes, and new words are rarely added to these categories. Century after century, English speakers have added thousands of new words, many borrowing from other languages, and building new ones from elements already available.

Deriving new word affixes Adding morphemes to a word is a common way to create new words. English added the agent suffix -ER to the prepositions up and down to form the nouns upper and downer to refer to phenomena that raise or lower mood. More commonly -ER is added to a verb (V) to make a word meaning "one who forms Vs": cantor "one who sings"; activist 'someone who campaigns'; Designer "the one who designs". English mainly adds morphemes through prefixes or suffixes. Prefixes such as UN-, PRE-, and DIS- change the meaning of words, but usually not their lexical category. The prefix UN- added to the adjectives true, popular, successful, and favorable produces new adjectives with opposite meanings: false, unpopular, unsuccessful, unfavorable. UN- is prefixed to a verb and creates a new verb with the opposite meaning: disconnect, untie, undo, uninstall, undo. DIS- placed before a verb produces a verb with the opposite meaning: disobey, disapprove, disappear, displease, dishonor. PRE- serves as a prefix for several categories: verbs (preplan, prewash, premix, preallot), adjectives (pre-copernican, pre-collegiate, prenatal, pre-surgical), and nouns (pre-ancient, pre-affirmation, pre-placement). ). PRE- has roughly the same meaning in each case and forms a new word of the same lexical category from an existing word. Three newly productive prefixes are CYBER- (cyberspace, cyberpal), BIO- (bioterrorism, biotechnology, biological weapons), and NANO- (nanotube, nanosecond, nanotechnology, nanoworld), none of which change the lexical category of the roots to which they belong refer. belong. attach. Suffixes derived from English are added at the end of a root. Unlike prefixes, derived suffixes usually change the lexical category of the root, for example from a verb to a noun. For example, adding -MENT to a verb makes it a noun: arrangement, agreement, delivery. The suffix -ATION does the same: abandon, organize, implement, observe, reform. Suffixes are widely used in the languages of the world. The Indonesian suffix -KAN turns a noun into a verb, and among the different meanings it can produce are: "make it X" (rajakan "crown" from raja "king") and "put on X" (like in Penjarakan 'to imprison' from penjara 'prison' ⫹ -KAN).

How does a language expand your vocabulary? Reduplication Reduplication is the process of repeating a morpheme or part of a morpheme to create a new word with a different meaning or category. The Mandarin Chinese word sànsànbu "to take a leisurely stroll" is formed by doubling the first syllable of sànbu "to go"; hónghóng 'bright red' is formed by doubling hóng 'red'. Partial reduction only repeats part of the morpheme, while full reduction reduplicates the entire morpheme. In the motu language of Papua New Guinea, mahuta 'sleep' becomes fully mahutamahuta 'sleep all the time' and partially mamahuta 'sleep' (when agreeing to a plural subject). In Turkish, adjectives such as açik 'open', ayri- 'severed' and uzun 'long' are doubled (by prefixing the initial vowel followed by a consonant) as apaçik 'completely open', apayi- 'completely detached' and upuzun 'very long' '.'” Reduction is not a repetition that does not create a new word, but simply repeats the same word, as in very, very, and night. English does not have a productive process like the duplication of Chinese, Motu or Turkish. Reduplication can have different functions in different languages. It can soften or intensify the meaning of a word, as the Chinese, Motu, and Turkic examples just given show. Can tag language categories, such as B. in Indonesian, where certain types of plural nouns are formed by doubling: babibabi 'a multitude of pigs' is a reduced form of babi 'pig'.

Compound English speakers show a willingness to put words together to create new words in a process called composition. Recent connections include Air Kiss, Moonshot, Waterbed, Advance, Color Code, Computer, Dust Bunny, Gutbuster, Plastic Wrap, Speed Dating, Mall and Radiopharmaceutical, as well as V-Chip, Email, Online, Website, website and download. (Note that these compounds emphasize or emphasize the first element more than the second.) To gauge the composition's popularity, consider that a relatively brief article in an issue of the Los Angeles Times contained the following examples. nouns

Substantive

adjectives

Petroleum Engineer Government Documents Government Witness Subcommittee Listens to Aircraft Carrier Training Course

whistleblower troublemaker debt ceiling brain tumor reserve account sea power

payphone call pantry concealment bribery separation

Middle-aged born-again, overzealous, high-level babyface whistleblower

The composition is done in many languages. Mandarin, for example, has several compounds, such as fàn-waˇn "rice bowl", diàn-na aˇo ("electric" ⫹ "brain") "computer", tái-bù "tablecloth", fe¯i-jı¯ ('fly' ⫹ 'machine') 'plane', and he¯i-baˇn ('black' ⫹ 'board') 'table' German is famous for its compositional tendencies. The word telephone has long been the word of choice for what is now commonly referred to as a telephone. A pen is called a ballpoint pen ('ball' ⫹ 'writer'); a glove-glove ('hand' ⫹ 'shoe'); Mayor is mayor ('citizen' ⫹ 'teacher'). The Indonesian explored the composition in a word made known to Westerners through its use as the assumed name of a well-known World War I socialite and spy: matahari, meaning 'sun', comes from mata 'eye' and hari 'day'. . the word for "glasses" is kacamata, a compound of kaca

• 47

4 8 • Chapter 2 Words and their parts: lexicon and morphology

Bear Encounters in Yosemite National Park. A notice in four languages, including German. Can you identify likely connections in the German version? Food makes bears aggressive. Protect the food for all animals in the park.

'glass' and mata 'eye' (similar to English compound glasses but with a different order of elements).

Abbreviations Abbreviations of various kinds are a popular way of multiplying words in a language. Common abbreviations are common: Radials for radial tires, Jet for jet plane, Narc for drug agent, Feds for federal agents, Obituaries for obituaries, Poli-Sci for political science, Indie for independent film, Rec Room for rec room, Offset Time for offset time and application or Applications related to computer application programs. Note that the abbreviations in the full expression do not have to be morphemes: narc is not a morpheme in narcotics. Other types of abbreviations are acronyms, initials, and combinations. Acronyms Abbreviations, where the first letters of a phrase are joined together and pronounced as a single word, are acronyms: UNESCO NATO Radar (Radio Detection and Ranging) WASP NASA Yuppy (Young Urban Professional ⫹-Y) DOS (Security Operating System Disk) Nimby (Not in my backyard) ASCII (American Standard Code for Information Interchange, pronounced "as-kee")

Acronyms Some abbreviations look like acronyms but are pronounced as a series of letters. At U-S-C (University of Southern California) and N-Y-U (New York University), a student's grade point average may be referred to as G-P-A. PC has two meanings: "politically correct" and "personal computer". Because of their pronunciation as a series of letters, they are called initials. Many initials (AI, CD, DNA, DVD, fMRI, MTV,

How does a language expand your vocabulary? NHL, PDA) could not be easily pronounced as ordinary words, while others could but are not: CEO for CEO, ADD for Attention Deficit Disorder, and SUV for Sport Utility Vehicle. Perhaps the most popular acronym in the world is OK.

Try it yourself: What are acronyms, which acronyms? EU, US, FBI, CIA, WMD, GI, SARS, BSE, NYPD, ER, DARPA, MBA, VIP, SEC, DJ, ATM, IT, IM, iPod. Mixtures Mixtures are words formed by combining parts of words. Smog (of smoke and fog), Glob (lump and blob) and Motel (motor and hotel) are older hybrids. Some of the newer ones include fanzine (fan and magazine), punkumentary (punk and documentary), infomercial (information and advertising), and biotech (biology and technology). The modem is well known, although its elements are not (modulator and demodulator). Internet users and netiquette mix the network (short for Internet) with citizens and etiquette. The combination of the existing pollutant mix with the end of the metropolis forms the smogopolis. Mixtures such as Spanglish, Franglais, Chinglish, and Yinglish suggest how much certain languages have borrowed words from each other. The blends also serve as trade names and as names for related products: Amtrak, Eurailpass, Eurorail, Eurotrip, and Flexipass. Most inventions seem to combine two nouns, but the aspirants "people who want to be something other than what they are" and "gimmes" "things they don't deserve" combine other categories.

Try it yourself: Identify the mixed words in these combinations: affluential, andropause, automagically, beefalo, bollywood, botel, brainiac, chunnel, cremains, cybrarian, digicam, emoticon, gasohol, gaydar, guesstimate, hactivist, himbo, assassinabilia , prebuttal, sexpert .

Additional Formation Another type of word formation is illustrated by pronunciation, which some college students may find when they look up the verb that matches the noun pronunciation. By pronouncing it, they "reformed" a new verb. Other later formations include the verbs typewrite, baby-sit, and edit, which are formed from the nouns typewriter, baby-sitter, and editor.

Conversion or functional change In some languages, a word belonging to one category can be converted to another category without changing the form of the word. This is called a feature change. We ask someone to update a report (verb), and then we call the revised report update (noun). We have a staff member send the report via email or fax, both verbs converted from shortened forms of nouns (email, fax). Companies hire (verb) a group of employees, calling them new employees (where hiring is a noun). To promote a product on the market, we market it. A conversion of this type usually results in noun/verb and noun/adjective pairs. Table 2-2 shows that sometimes the same shape

• 49

5 0 • Chapter 2 Words and their parts: lexicon and morphology

It can serve as a noun, verb and adjective. Whenever a form is moved to a new lexical category, it conforms to that category's inflectional morphology: one update, two updates, you're updating the report now, and you updated it last month. TABLE 2-2 Some forms in English that belong to more than one lexical category NOUN

VERB

e-mail bookmark broke outrage market delay plot game page opening illegal medium template ceiling noble brick

E-mail bookmarks bust outrage market lag play action

medium brick red prime model blanket

ADJECTIVE

illegal opening side medium model noble brick ceiling

Semantic modifiers can take on new meanings by narrowing or expanding the scope of their reference. The Oxford English Dictionary recently added meanings to its entries for the nouns coyote, crack, and thumb-nail and for the verb zone (out). Two well-known examples of semantic switching have been popular since the Vietnam War, when hawk was commonly used for supporters of the war and dove for its opponents, expanding the meaning of these words from hawk's combative nature and symbolism. peaceful. Paper Birds Pigeons Today's computer users use a mouse and select web addresses. These new meanings did not replace the previous ones, but expanded the scope of the respective words. This phenomenon, called semantic shift or metaphorical expansion, creates metaphors. Over time, the metaphorical origins of words can fade, as in the meanings of the underlined parts of these sentences: derailment of congressional legislation, a forceful speaker, an aggressive chief of staff, glossing over a billion-dollar farm bill to avoid a veto. . battle.

Borrowed Words "Neither is a borrower nor a lender," Shakespeare advised, but the speakers pay little attention to the language. English was extraordinarily receptive to loan words.

How does a language expand your vocabulary? Words that have accepted words from almost 100 languages in the last 100 years. For most of its history, English borrowed more from French than any other language in the 20th century. French is followed at some distance by Japanese and Spanish, Italian and Latin, as well as Greek, German and Yiddish. In lesser numbers English

How many word formation processes can you identify here? Initialism acronym loan combination invention prefix suffix composition

• 51

5 2 • Chapter 2 Words and their parts: lexicon and morphology

It now includes loanwords from Russian, Chinese, Arabic, Portuguese, and Hindi, as well as several African and some Amerindian languages. In turn, many languages have incorporated English words into their inventory, although some cultures are reluctant to borrow them. The Japanese coined the words beesubooru "baseball", futtobooru "soccer" and booringu "bowling" along with the sports they name, changing them to judo, jiu-jitsu and karate, which joined the English-speaking team. Officially, at least, the French are unwilling to borrow, particularly from English, and have banned the use of words like weekend, pharmacy, brainstorm, and countdown. For using the borrowed term jumbo jet, Air France was heavily fined by the French government, which insisted that Gros Porteur was the correct French name for the jumbo jet. OK, Americanism is used pretty much everywhere now, as are terms like jeans and disco, which accompanied the items they named as they spread across the globe. Recently, the terms Internet, WWW and Web have also gone around the world. As with other languages, most English loanwords were nouns, but some adjectives and some verbs, adverbs, and interjections were borrowed. You can easily identify popular loans like paparazzi, karaoke, and resume as loans. Loanwords related to food and drink include hummus (from Arabic), aioli (from Provence), mai tai (from Tahitian), calzone, focaccia, and pizzetta (from Italian), and burrito, enchilada, fajita, and taco (from Spanish). ). ). ). . Yiddish gave us the more general term nosh. Other popular loanwords include Cantonese wok, glitch German, Italian ciao, Spanish macho, pronto, and mañana, as well as Yiddish chutzpah, clumsy, nebbish, schlep, and tug. Borrowed words sooner or later adapt to the pronunciation patterns and grammar rules of the borrowed language. In Los Angeles, a sign hanging from a restaurant undergoing a kitchen renovation reads "Burritofification in Progress," signaling the opening of a Mexican restaurant by mimicking and mimicking the morphological processes that created the garnish. Over time, credits go through the same processes that affect other words. Nosh was borrowed as a verb that couldn't take an object (I want to eat), but has since taken on a new usage as a verb that can take an object (let's eat some hot dogs). The verb nosh with the suffix -er produces the nosher nosher "who eats", and nosh itself can be used as a noun meaning "a snack". large or elaborate dishes

Inventing Words It's not common to invent words from scratch. The advantages of using familiar elements in forming new words and the ease of borrowing from other languages make the invention of new words rare. The invention added words like granola, zap, and quark to the English vocabulary. Nerd seems from Dr. Seuss and Lewis Carroll's Bandersnatch, Galumph and Snark. Some names like Pyrex, Kodak, and Xerox were invented as trademarks for specific products or companies. But speakers and authors sometimes extend the use of brand names beyond brand names, such that "Xerox" is sometimes used loosely to mean "photocopy," though not necessarily a Xerox-made machine is being used. Likewise, when students “google” something on the internet, they don't necessarily use the Google search engine to do so. Gizmo and Lollapalooza may also have been invented, the origins of which are unknown.

What kinds of morphological systems do languages have?

Jeans and Discs The denim on your favorite pair of jeans is a form of the word genoa, borrowed from Middle English (ie, in Chaucer's day). Jeans is an acronym for Jean Barstian 'Genoa Barstian', which refers to a thick fabric that was once made in Genoa, Italy. The word denim, used for the fabric from which jeans are made, evolved from Serge de Nîmes, a fabric product from the French city of Nîmes. You can wear your favorite pair of jeans to a nightclub. The French word discothèque "library of records" is a combination of two French morphemes, disque meaning "disc" or "record" and the suffix -thèque as in bibliothèque "library". Discothèque appears to have been first used in English in 1954, and ten years later as an abbreviation disco. Disco, the noun, underwent a functional change to disco, the verb, meaning "to dance to disco music", a usage first mentioned in 1979. Both the noun and the verb can be heard around the world in cities whose residents do not speak English. still french.

What kinds of morphological systems do languages have? You have already seen examples of derivational morphology and inflectional morphology in many languages. But not all languages have inflectional morphology, and some have little or no morphology. Others have complex words with different parts, each part representing a morpheme. These three types of morphological systems have been called insulators, trusses, and bends. Some languages are mixed in the types of morphologies they use.

Isolating Morphology Chinese is a language with an isolating morphology in which each word tends to be a single isolated morpheme. An isolating language lacks derivational and inflectional morphology. By using separated words, Chinese expresses certain content that an inflectional language with inflectional affixes can express. For example, while English has a declined possessive pronoun (the boy's hat) and a so-called analytic possessive pronoun (the boy's hat), Chinese only allows the possessive pronouns of the boy's hat. Chinese also lacks tense marks, and no distinction is made in pronouns between gender (he/she), number (she/they), or case (she/they). Where English has six words: he, she, he, she, she and they, Chinese uses only a single word, although it can indicate plurality with a separate word. The following sentence illustrates the typical pattern of a morpheme per word in Chinese: wO gang yào gEi nI nà yì bei tea I'll bring you this cup of tea 'I'll bring you a cup of tea in a moment.'

Even more than Chinese, Vietnamese approaches the one-morpheme-per-word model that characterizes isolated languages. Each word in the following sentence has only one form. You can see that the word tái is translated as "I", "my" and "we". Note that the Vietnamese couples are chúng and toi (the words for "PLURAL" and "I") to say "we". like chinese,

• 53

5 4 • Chapter 2 Words and their parts: lexicon and morphology

Vietnamese lacks tense marks on verbs and case marks on nouns and pronouns, as well as numerical distinctions (although it can indicate plurality with a separate word). khi

I d1n

if I

Haus

b3n

I her

Come home, friend of mine

PLURAL

i b2t d5u do homework for myself

begin

it does

Lesson

"When I got to my friend's house, we started having classes."

However, some languages that tend to downplay inflectional morphology use derivational morphology to profitably expand their word pools. Indonesian, for example, has only two inflectional affixes but uses about two dozen derivational morphemes, some of which we have seen in this chapter.

Union Morphology Another type of morphology is called union. In agglutinative languages, words can have multiple prefixes and suffixes, but they are characteristically different and easily decomposed into their parts, like English or pre-asserted ad but unlike sang (SING ⫹ 'PAST') or men (MAN ⫹ ' PLURAL FORM '). Turk has an agglutinating morphology as shown in this example. (Hashes represent morpheme boundaries within a word.) herkes

I go to university

all yours

University-to-Start-FUTURE-I

believe-GIFT

"Everyone thinks I'm starting at university."

Inflectional morphology Many languages have a large inventory of inflectional morphemes. Finnish, Russian and German maintain sophisticated inflection systems. In contrast, English has eliminated most of its inflections over the centuries, leaving only eight inflections: two for nouns, four for verbs, and two for adjectives, as shown in Table 2-3 on page 55. When new nouns, verbs, and adjectives are added to English, or when a child learns new words, the words are most likely to be inflected as in the examples listed, and all eight English inflectional morphemes are considered productive. Compare this inflection system in English with the examples of the Russian noun ˇzena "wife" and the verb pisat "to write" in Tables 2–4 and 2–5 on page 55. Grammatical Functions of Inflections Consider the following sentences. They contain exactly the same words but express different meanings. 1. The farmer saw the wolf. 2. The wolf saw the farmer.

These sentences illustrate how English uses word order to convey meaning: different word orders convey different scenarios about who did what to whom. When semantic facts, such as who did what to whom, are expressed by word order rather than by inflection, this is not a morphological problem but a syntactic one, and syntax is the subject of Chapter 5.

What kinds of morphological systems do languages have?

TABLE 2-3 Inflectional morphemes of the English LEXICAL CATEGORY

GRAMMATICAL CATEGORY

EXAMPLES

Noun Noun

Possessive Plural

Cars, church cars, children

word for word word for word

Third person past participle present participle

(she) nothing, (she) seems desired, was desired, was shown (or was shown), wanting, showing

Adjective Adjective

Comparatives Superlative

louder sweeter louder sweeter

TABLE 2-4 Russian noun inflections: zˇena “wife” CASE

SINGULAR

PLURAL

Nominative Accusative Genitive Dative Instrumental After some prepositions

woman woman woman woman zenoy zene

busy busy busy busy

TABLE 2-5 Verb inflections in Russian present tense: pisˇat' 'to write' PERSON

SINGULAR

PLURAL

First person Second person Third person

more more more

a little smaller

A comparison with Latin is instructive because Latin had a relatively free word order. Since agricola means "peasant" and lupum "wolf", Latin speakers could have arranged sentence 1 ("The peasant saw the wolf") in two ways (among others): Agricola vidit lupum. Lupum vidit agricola.

• 55

5 6 • Chapter 2 Words and their parts: lexicon and morphology

the farmer saw

Lob

BAUER

LOBO

A wolf

he saw

Agriculture.

LOBO

BAUER

The farmer saw a wolf. FARMER

LOBO

1662663

Latin speakers did not rely on word order to indicate who was looking at whom. Instead, inflections in nouns signaled such information. The next three sentences mean "The farmer saw the wolf"; different word orders do not change this meaning.

The farmer saw the wolf.

saw the farmer

Lupus.

BAUER

LOBO

Lupus

he saw

a farmer

LOBO

BAUER

The farmer was a wolf.

he saw

BAUER

LOBO

1662663

Instead of saying, "The wolf saw the pawn," required different inflections:

The wolf saw the farmer.

The inflectional suffixes -a in agricola and -us in lupus mark them as subjects. The inflections -am and -um in agricolam and lupum make them direct objects. A vague parallel in English can be seen in the inflections of Latin nouns in certain uses of pronouns where pronoun form and word order reinforce each other: Sie lobte dich. (She is the subject, he is the object.) He praised her. (He is the subject, she is the object).

Nouns in English and Latin have inflections for number and case. English nouns have only two cases, called possessive and common. The possessive case (sometimes called the genitive case) is denoted by a suffix (cat, robot). The common case is unmarked (cat, robot) and is used for all grammatical functions other than possession: subject, direct object, indirect object, and object of a preposition. Latin also had a genitive and inflections for various other cases, notably nominative (mainly for subjects), dative (indirect objects), accusative (direct objects and objects of some prepositions), and ablative (objects of some prepositions). Latin generally had five or six singular and plural inflections, although some inflected forms were pronounced the same, as shown in Table 2-6. The set of forms that make up the inflectional variants of a word is called a paradigm, and the paradigms of nouns are called declensions. Latin had several declensions, such as the two for agricola and hortus in Table 2-6. Paradigms for the corresponding English words farmer and garden appear in Table 2-7. You will find that the four written forms in the English paradigms represent only two different pronunciations because Farmer, Farmer and Farmer are pronounced the same, as are Gardens, Garden and Gardens. Spoken English usually has only two forms of a regular noun, but irregular plurals can have four spoken forms and four written forms: man, man's, men, men's; kid, kid, kid, kid.

What kinds of morphological systems do languages have?

TABLE 2-6 Paradigms for two Latin singular nouns

'BAUER'

'GARTEN'

Nominative Accusative Genitive Dative Ablative/Instrumental

farmer farmer farmer farmer farmer

garden garden garden garden

PLURAL

'BAUER'

'GARDENS'

Nominative Accusative Genitive Dative Ablative/Instrumental

Farmers Farmers Farmers Farmers Farmers Farmers

Gardens Gardens Gardens Gardens

Table 2-7 Paradigms for two SINGULAR nouns in English

common possessive

Bauer

garden garden

peasants peasants

Gardens Gardens

PLURAL

common possessive

Some English pronouns have a third form for the objective case. In Table 2-8 you can compare the paradigms for first and third person pronouns. First person pronouns show three different case forms in the singular and three in the plural. Third-person pronouns have distinct masculine, feminine, and neuter forms in the singular, but are gender-neutral in the plural. The second-person pronoun (you) and the neuter pronoun of the third-person singular (it) do not have distinct objective forms, as shown in Table 2-9. Instead, they only have two forms, just like regular nouns. Gender and Consistency In English, gender distinctions in pronouns are based on biological sex: references to males require the masculine pronouns he, his, or him, while references to females require the feminine pronouns she, hers, or her. To refer to something that is neither masculine nor feminine, English speakers use it. In German, French, Spanish, Russian, Old English, and many other languages, nouns do not have biological gender, but grammatical gender. In these languages, certain other categories of words, such as determiners and adjectives, occurring within a noun phrase bear inflections that agree with the noun in gender, number, and case.

• 57

5 8 • Chapter 2 Words and their parts: lexicon and morphology

TABLE 2–8 Paradigms for first and third person pronouns in English FIRST PERSON MASCULINE

FEMALE THIRD PERSON

NEUTRAL

SINGULAR

common possessive goal

i undermine

he be he

she is yours

is this

PLURAL

common possessive goal

we our we

they of them they

TABLE 2–9 Second and Third Person Pronouns Compared to English Nouns SECOND THIRD PRONOUNS

Substantive

you you

Bauer

SINGULAR

common possessive

This is

PLURAL

common possessive

you you

peasants peasants

Unlike the English definite article the (with a single spelling representing the two pronunciations "thuh" and "thee"), the German definite article has forms for three genders and four singular cases, although there are no markings for different genders in the plural , as shown in Table 2-10.

TABLE 2-10 German paradigm of the definite article

MALE

nominative accusative genitive dative

o o o o

SINGULAR FEMALE

o o o o

NEUTRAL

PLURAL ALLE GENRES

or do

o o o o

Variations in pronunciation of a morpheme: allomorphy

• 59

French and Spanish also have variants of the definite article, although none are as diverse as German. French only distinguishes two genders of nouns; marks masculine nouns with the definite article le or the indefinite article a and feminine nouns with the definite article la or the indefinite article une; the plural form of the definite article for both genders is les. Gender is marked in Spanish in both singular and plural. Table 2-11 contains examples in French and Spanish.

TABLE 2-11 Definite articles in French and Spanish with nouns

Male Female

FRENCH

ESPAÑOL

the cat the cats the house the houses

'the cat' 'the cats' 'the house' 'the houses'

There is not always a strict demarcation between agglutinative and inflectional languages, and some languages are difficult to classify. Nonetheless, the distinction between inflection, isolation, and agglutination is useful for characterizing languages in terms of their morphological systems.

Different pronunciations of a morpheme: allomorphy A morpheme may not be pronounced the same in all linguistic contexts. The METAL morpheme is pronounced "a dark metallic silver" in one way and "the metallic sound" in another. It's easy to see that the vowel represented by ⬍a⬎ differs in pronunciation in metal and metal (and in American English, the same thing happens with the sound represented by ⬍t⬎). Note that the final sound in house is [s] and in woman [f], but in houses the CASA morpheme ends in [z] and in women the WIFE morpheme ends in [v]. Alternative pronunciations of a morpheme are called allomorphs, and allomorphic variations are common in some languages, including English. (English spellings generally do not account for allomorphic variations; capture in woman/women, but not metal/metal or house/houses.) Allomorphic variations are not limited to free morphemes such as METAL, WIFE, and HOUSE. It can affect any morpheme, including affixes. For example, the English past tense morpheme [t] is pronounced in pick, [d] in play, and a third tense in twisted. Which pronunciation goes with which verb is governed by rules, and the choice depends on the final sound of the verb stem (a question we shall return to in Chapter 4). Children learn the rules of allomorphic variation early on and of course unconsciously. If you were to teach English speaking children a set of fictional verbs they had never heard before, such as "plick" and "tevin," they would pronounce the past tense by changing [t] to "plick" and [d] to " tewin." ". according to these unconscious rules. The English plural morpheme on nouns also varies. In cups and cuts, the plural morpheme is pronounced [s], but in jabs, pads, and traps it's [z], and in churches, cymbals, and

6 0 • Chapter 2 Words and their parts: lexicon and morphology

Bridges else. These rule-driven pronunciations of the plural morpheme depend systematically on the final sound in the noun stem, and children who learned unheard imaginary nouns added the correct plural allomorph to invented words; You would add [s] to "wuck" and "lutt" and [z] to "mub" and "wug", not the other way around. Of course, not all English nouns form their plural in this "normal" way. In addition to regular or rule-controlled allomorphs, there are nouns that form their plural irregularly. Consider the nouns stag and fish with the plural forms stag and fish, or consider man, ox, and tooth, whose plural allomorphs are irregular and require special study. (Irregular forms often have an explanation in the history of a word, as we discussed in Chapter 14.) Most English verbs form their past tense according to a rule, and new verbs also tend to form “regular” forms of the past tense. But many common verbs don't follow regular patterns and need to be learned, such as B. teach and teach, do and do, speak and speak, sleep and sleep, tell and tell, walk and walk. Children who have not yet mastered the irregular past tenses are likely to produce forms such as "taught," "did," "said," and "went."

Try it yourself: Identify verbs that have an allomorphic irregular past tense: swim, chin up, read, reach, break, bake, melt, say, see, find, repeat, lead, eat, taste, scratch, bleed. Identify the nouns that have an irregular plural allomorph: mouse, house, chicken, egg, boy, girl, woman. Identify nouns that have more than one allomorph: moth, calf, child, woman.

TABLE 2-12 Frequencies of Four Common Words in the Brown Corpus WORD

unusually powerful nursing facility

EVENTS

GENDER

TEXT: %S

52 62 63 63

12 14 14 15

43 56 54 52

TABLE 2-13 Frequencies of five closely spaced words in the Brown Corpus WORD

fiscal anode artery budget dictionary

EVENTS

GENDER

TEXT: %S

51 53 55 75 115

3 7 3 1 5

5 23 5 2 26

Variations in pronunciation of a morpheme: allomorphy

• 61

By using computers to learn words, much information can be derived from a corpus such as the Brown Corpus (described on page 24). The most common and least common verb forms can be identified throughout the corpus or in one of its genres such as science fiction or newspaper editorials. You won't be surprised to learn that three of the four most commonly used words are the, of, and and. Instead, words like oblong, obstinate, radionic, narcosis, and mystification appear only once. Information about how general or specialized a word is can also be evaluated. As you can imagine, the, and, and from appear in all 500 texts in the Brown Corpus, while a proper noun appears frequently in a single text but nowhere else in the corpus. The name Mussorgsky appears seven times, all in the same 2,000-word text. The noun dialysis occurs twelve times, all in a single text; Radiation sterilization takes place six times, all in one text. Compare these extremely specialized uses with a word like as well. In addition, it appears 88 times in 63 texts and in 13 of the 15 genera represented in the corpus. The frequency is not exceptionally high, but the spread is wide. Table 2–12 on page 60 provides examples of the simpler types of information you can derive from the Brown Corpus. Next to each word listed is the total number of times it occurs and the number of genres (out of 15) and lyrics (out of 500) in which it occurs. The four words each occur fewer than 65 times in this million-word corpus, and in any case these occurrences are spread across at least 12 genera. This scatter indicates that the vocabulary is not specialized. You can compare these wide spread words with others that have a more restricted spread. The words listed in Table 2-13 on page 60 occur in less than half of the genera in the corpus. This relatively narrow distribution identifies more specialized words that occur in fewer contexts despite their general frequency. Think of anode, which occurs 75 times, more than any other word listed in Table 2-12. Despite its frequency, the anode appears in only two texts, both of which belong to the same genre. This makes it clear that niche words may not come up very often, but they can be used frequently when dealing with a topic. This is especially true for technical and scientific writing. swindler

AND

Note that the word corpus occurs 15 times in this section but nowhere else in the chapter. In Table 2-13, budget and fiscal are also specialized and appear in less than half of the genres. A corpus also allows researchers to determine which words typically appear next to each other. These collocations are useful for a variety of purposes, such as creating lifelike instructional materials for language learners, and, as you will see in Chapter 7, distinguishing between word meanings using forms rather than lexical category information. In practice, words in a corpus are often "tagged" with information such as the lexical category of a word: nouns carry a noun tag, verbs carry a verb tag, etc. Such tagging makes it possible to examine the properties of the group of words, that bear a specific mark. Manually tagging a large corpus (examining each word and writing the tag into the corpus) would be costly and time consuming. Consequently, researchers have developed ways to automatically mark a corpus. One way is to have a computerized reference dictionary that lists the lexical category of the most common words, or as many words as possible. Words in a corpus can be automatically tagged with the corresponding word in the dictionary. Thus, if the forms of information and dissemination were present in the corpus and lexicon, the nominal designation that accompanies them in the lexicon would carry over to the corpus entries. Likewise, the lexicons and frequents would be tagged as adjectives, the and a as determiners or articles, identify and think as verbs, etc. This process of matching forms in the corpus to tagged forms in a dictionary will not be entirely successful, as some forms have members can be of more than one category (as shown in Table 2-2 on page 50). In this paragraph you will find several words whose forms do not clearly identify them as members of a particular category. Forms, can, use, present and process can be nouns or verbs, for example. Because English has so many forms belonging to more than one category, accurate markup must rely on more complicated procedures than automatic dictionary matching. inside

6 2 • Chapter 2 Words and their parts: lexicon and morphology

Using computers to study the context of words (in actual use), a word form generally belongs to only one category. Thus, markup can be aided by identifying the category of words immediately surrounding a form whose category is ambiguous. Let's take treat as an example: It can be a noun or a verb. Suppose the corpus contained the rather problematic phrase, and suppose that automatic tag dictionary matching already mapped the adjective tag well. Given the choice between an adjective before a noun or a verb, it's safer to assign the noun tag since English adjectives generally come before nouns, not verbs. hence dealing with many problems could reasonably be considered a noun. When you start labeling the words that only belong to one

category, you can use this information to clarify ambiguous cases. Other useful information can be extracted from a tagged corpus, including how often a particular form occurs as a noun or as a verb (if both are possible). To learn about the list of names, you would have to summarize all possible forms (list, list, list, list'); Likewise, for the verb, you need to know its forms (list, list, list, list). Information about which genres (e.g., newspaper reporting, academic writing, or financial news) have common adjectives, nouns, verbs, prepositions, or pronouns compared to other genres can be useful for designing instructional materials and building automatic language recognition systems. ■

Summary • A morpheme is a minimal unit of speech that is assigned a grammatical meaning or function.

• Words can contain a single morpheme (camel, swim) or multiple (library, prediction). • In the mental lexicon, each morpheme contains information about sounds, related words, patterns of phrase co-occurrence, and meaning.

• Free morphemes are those that can appear as independent words: AUTO, HAUS, FOR. • Linked morphemes cannot appear as independent words but must be added to another morpheme: CAR ⫹ -S, LOOK ⫹ -ED, ESTABLISHMENT ⫹ -MENT. • Linking morphemes can be nouns for information like number (e.g. "PLURAL") and case (e.g. "POSSESSIVE") or verbs for information like tense (e.g. "PAST") and person (e.g. "THIRD" PERSON).

• Linked morphemes can derive different words from existing morphemes; B. UN (wrong), DIS- (dislike) and -MENT (compromise).

• Linked morphemes can be affixes (prefixes or suffixes), infixes, or circumfixes. • In words, morphemes have significant linear and hierarchical structures. • The range of morphological processes involved in expanding the vocabulary of a language can include composition, reduplication, fixation and truncation.

• Languages borrow words from other languages and sooner or later subject the borrowed words to their own pronunciation patterns and morphological processes.

What do you think? checked

• Types of morphological systems include inflectional, isolating, and binding systems.

• Isolation systems (eg Vietnamese) usually have one morpheme per word. • Attachment systems (eg Turkish) tend to have different affixes. • The corpus study is useful for showing the distribution of word categories and morphemes and of specific words and morphemes in different text genres, information that can be useful when designing automatic speech recognition systems.

• Collocation is the term used to refer to the co-occurrence of a word with other words. • Words in a corpus can be automatically tagged for lexical categories, although multiple rounds of tagging may be required to tag all words.

WHAT DO YOU THINK ? R E V I S I T E D • I baked a cake. Most English verbs form the past tense by default, but some of the more common ones are formed irregularly. Children learn the rule for forming the past tense of regular verbs relatively early (see Chapter 15) and tend to form all verbs in the same way. Instead of using the irregular past tenses done and tell, this three-year-old forms the past tenses using the rule of thumb as if she were doing and saying her past tenses as if cooked and spelled. • Feeding conditions. Sashimi, wasabi, miso, sake, sukiyaki, teriyaki, ramen, and tofu all come from Japanese; Chinese dim sum, bok choy and wontons; alfalfa, salsa, guacamole, toast and Spanish anchovies; mulligatawny, chutney and samosa from the languages of India; Meat, broth, soup, chops, mackerel, lamb, mustard and the more obvious French crepes and croque monsieur (as well as du jour, au jus and a la mode). • Top ten words. According to Brown Corpus, the top ten words in American printed English are the, be, of, and a, in, he, to (the infinitive marker), have, and to (the preposition), and the next three are it, for, and I Note that the list contains no nouns, adjectives, or adverbs, and the only verbs in the first ten words are be and have. • State names. Carolina is named after King Carlos II; Virginia is named after Queen Elizabeth, known as the Virgin Queen; Missouri is named for a Native American village of the Sioux family; Delaware is named for an Algonquin Native American people; Illinois takes its name from a confederation of Algonquian Indian tribes.

• 63

6 4 • Chapter 2 Words and their parts: lexicon and morphology

Exercises Practice exercises

A. For words with more than one morpheme (sometimes approximately), identify where morpheme boundaries exist. Example: Governments: Government - s 1st man

11. Symbol

21. Conversion

2. tasteless

12. Heads of households

22. cheap

3. search

13. Riot

23. childish

4. not disclosed

14. Herz

24. Unlawfulness

5. indelicate

15. dark horse

25. Supermarkets

6. troublesome

16. bigger

26. Globalization

7. dispassionate

17. Fitters

27. Counter Terrorism

8. respectful

18. indirect

28. Interdependenz

9. Slander

19. superior

29. antidesempego

20. exceeds

30. Resilience

10. rude

B. From the numbered list above, use the number to identify four words that contain a prefix. C. Identify by number four that contain a suffix (each suffix must be distinct from the others). D. Identify four compound words by number. E. Using the number, identify four words that contain an inflectional suffix. f

Identify four words that contain a derived suffix by number.

G. Examine these abbreviations (starting on page 48): Jet for jet plane, Narc for drug agent, Feds for federal agents, Obituaries for obituaries, Political Science for political science, Independent for independent film, and Applications for computer application programs. Which of these abbreviated forms represents a full-form morpheme? Now consider recreation room by recreation room and recovery time by recovery time; Do any of these abbreviations represent a full-form morpheme?

Based on English 2-1.

Identify the category of words in italics in the following sentences. Use the abbreviations N for noun, V for verb, Adj for adjective, Adv for adverb, Prep for preposition, Pro for pronoun. a. People who rarely read in their bedroom may feel abnormal. B. Nobody really knows what normal reading is. C. The audio book market is huge.

2-2.

For the five words in Table 2-2 (page 50) that belong to three lexical categories, give a phrase that illustrates their use in each category. Examples of the mean are given. Is there a difference between mean and median? (Noun)

Exercises A tour guide can receive an average of $75 per day in tips. (verb) He worked hard but only got average grades. (adjective) 2-3.

a. Identify the lexical category of each word in the list below. B. List all morphemes (each word contains more than one here) and indicate whether they are free or bound. C. For each affix, indicate whether it is a derivation or an inflection.

2-4.

he listened

kleiner

unproductive

toys

bookstores

Ohr

upgrade

childish

oberer fixer

paternal

unlikely

improbability

refund

not funny

neater

realignments

calculate

disarm

normal

warned

Disturbance

realigned

without claims

a. The following three sentences contain DEMONSTRATIVE PRONOUNS in capital letters and demonstrative pronouns in italics. Characterize the difference in their use. (Hint: what are the lexical categories of the preceding words?) 1) THIS is the last time I do THIS. 2) I won't be making any of those fancy pizzas this time. 3) I'm sick of THIS; Give me one of those red ones. B. Number each pronoun in the following passage and identify its type (personal, demonstrative, interrogative, relative). In the case of personal pronouns, also indicate the person (first, second, third). And these books? Who are you? They appear to be from the library, so they must be returned. If you want you can put it in a shopping bag and I'll return it if I can get Ashley to give me a lift in her car. It's been in store for a few days. I hope you are ready now.

2-5.

Consider two popular connections. Convenience Food "Food that is convenient to buy, cook or eat" is a compound consisting of a noun and a noun. Natural foods "Food made from natural ingredients, free from chemical preservatives and pesticides" is a compound adjective + noun. Taken as a whole, each compound functions as a noun. Name six compound nouns that contain a noun-noun combination and six that are clearly an adjective + noun combination. (Note that not all adjectives that precede nouns are compounds, and it helps to pay attention to the tonic pattern. In the following sentences, compounds are in italics; say them out loud to see the pattern. Not all house white is the White House. House! Not all blackbirds are blackbirds.)

2-6.

From a passage of about 500 words in a weekly news magazine such as Time, Newsweek, or The Economist, make a list of 20 compounds, labeling the lexical category of each word element in the compound and the compound as a whole. So, given the phone tag, you would identify phone as a noun (or N), the tag as a noun, and the compound phone tag as a noun.

• Sixty-five

6 6 • Chapter 2 Words and their parts: lexicon and morphology

CATEGORY OF:

1. ELEMENT

2. ELEMENT

CONNECTION

telephone etiquette

Norte

Software

adj.

Norte

Bozo-Filter

Norte

2-7.

Cyber- became a popular prefix in the 1990s, mainly added to nouns to form new nouns, as in cyberlove, cyberland, cyberspace, and cybercowboy. List ten words that use the prefix cyber- and identify all instances of cyber- that are preceded by a lexical category other than a noun.

2-8.

Draw trees similar to Figure 2–1 on page 45 for these English words: revacinations disenchantment

2-9.

irrationality recapitalization

unshakable improbability

reassuringly updated

Consider the two analyzes of untruth given below. Explain why you prefer one analysis to the other. a. [[[a[trueAdj]Adj]thN]fulAdj] b. [a[[[trueAdj]thN]fulAdj]Adj]

2-10 The following terms are associated with the use of the computer or the Internet. For each, identify the type of formation (composite, abbreviation, acronym, transformation, etc.) and its lexical category. Example: group chat: connection, client-server noun mouse a flame info pike internetter info superpike mac spamming a twit filter

Cyberspace Cyberspace to exploit a lurker i-way netiquette a remailer a sysop WYSIWYG

FAQ PC IMHO for email, nicer smileys to look at a discussion software

Source code programming language to download newbee code writer or domain name browser beginner for keyboard cyber enthusiast

2-11 a. Graduates from the University of California, Los Angeles call their alma mater "U-C-L-A", but sometimes it is also referred to as "youkla" or "ookla". Which of these three pronunciations would count as acronyms? B. Identify four acronyms and four initials from the following list: DNA, STD, AIDS, SIDS, NBA, HIV, NHS, NHL, UNESCO, UN, UK, BBC, NATO. C. What is interesting about the words CD-ROM and JPEG? 2-12 a. The 26 most common words in printed American English are listed below, as determined by their frequency in a one million word corpus of text (the Brown Corpus). The category of some of these words is already given. For everyone else, enter your category and then answer the following questions. Choose your categories from this list: N (noun), V (verb), Adj (adjective), Prep (preposition), Det (determinants, including article), Pro (pronoun).

practice the

she

to be

com

von

Adverb

was

connection

and

she

is he around

to the

Pro

Infinitive marker

connection

to have

von

Pro

This

Pro

your

since

1) Make a list of the pronouns that are among the 26 most common words in written English: 2) Make a list of prepositions: 3) Make a list of determiners: 4) Make a list of verbs: 5) Make Make a list of adjectives: 6) Make a list of nouns: b. The words listed in both columns are printed so frequently that one in four words in the Brown Corpus is among the first eight words listed (or up to the infinitive marker). In other words, more than 250,000 of the millions of words in the Brown Corpus are the same eight words used over and over again. With that in mind, answer the following questions. 1) Which two lexical categories are conspicuously missing from the list? What explanation can you offer for its oddness? 2) What explanation can you offer for the frequency of prepositions in the Brown Corpus? (Hint: It can be helpful to think about what prepositions do.) 3) What explanation can you offer for the frequency of pronouns compared to nouns? 4) The verbs ser and ter appear in the list. If you knew that word 27 on the list is a verb, which verb would you guess? Why? 5) Of the 21 words whose lexical category you were asked to identify in part a, how many belong to closed word classes and how many to open word classes? 2-13 The following words or phrases are from an article on electronic commerce (Newsweek, July 7, 1997, p. 80). On the line next to each word (or word in italics), write the name of the process by which that word was used in this discussion and extract the terms from this list: composition, fixation, invention, abbreviation, transformation, derivation, semantic change, borrowing, mixture.

• 67

6 8 • Chapter 2 Words and their parts: lexicon and morphology

a. lack of awareness b. information superhighway c. in hyperdrive d. the digital world and the shaky title f. a cutting-edge design g. innovative design h. a virtual shop window i. cyberspace j. that will lubricate the trade k. Ecommerce without Zip l. CDA 'Communications Decency Act'

Based on languages other than English 2-14. Consider the following singular and plural noun pairs for people in Persian. How does Persian form these nouns in the plural? (Note: æ stands for a vowel like the English hat and x stands for a final consonant like the German Bach.) zæn mærd bæradær pesær xahær doxtær

'woman' 'man' 'brother' 'boy' 'sister' 'daughter'

zænan mærdan bæradaeran pesæran xahæran doxtæran

'women' 'men' 'brothers' 'boys' 'sisters' 'daughters'

2-15 Spanish nouns have grammatical gender; for example, car "car" and viento "wind" are masculine; Apple "apple" and sopa "soup" are feminine. Examine the following expressions to discover patterns of agreement between noun, article, and adjective. a. B C D.Y.F

the red car 'the red car' the cold wind 'the short way' the red apple 'the red apple' the cold soup 'the short skirt'

In these expressions, which is the masculine form of the definite Spanish article "el" and which is the feminine form? Given the masculine form of an adjective, give the rule for forming the feminine singular of these adjectives, and then fill in the missing adjective in parts i and j. Gram.h I. j.

the yellow book 'the yellow book' the tall man 'the tall man' the flower _____ 'the yellow flower' the woman _____ 'the tall woman'

Exercises Identify how the adjectives in the following examples differ from the previous ones. Given the masculine form of the following adjectives, set up the rule to form the feminine singular forms and apply the rule to fill in the missing adjectives in the dots s and t. k. the happy child 'the happy child' l. the big cat m. the hot bread the strong horse 'the strong horse' the. the easy quiz p. the lucky girl "the lucky girl" q. the big house 'the big house' r. hot greenhouse person _____ 'strong person' t. the task _____ 'the easy task' 2-16. Consider the following Persian word pairs with their English glosses. Write down the lexical category of the words in column A and give the complete rule for the formation of the words in column B from those in column A to the ship of English sh

dana

'intelligent'

danes

'Wisdom'

Mushrooms; arrows

'Boa'

Member

'Friendliness'

Meister

'closed'

Meister

'Thickness'

Mercado

'big'

Mercado

'Size'

called

'Caramel'

the secret

'Sweet'

2-17 a. Please discuss the Turkish nouns below and provide a list of their constituent morphemes, along with an explanation for each. (Note: stands for a vowel similar to u.) kitap

'Book'

apples

'apples'

The rod

'Talos'

'Horse'

Tempo

'Tabla'

People

'Herren'

aod

'again'

bedroom

'Bedroom'

Table Mountains

'Panel'

Saft

'follow'

horses

'horses'

Counting

'ends'

Elma

'Garbage'

'chica'

fruit

"fruit" (SINGULAR)

B. Based on your analysis, provide the Turkish words for the following English words: books, man, girls, end, fruit (PLURAL). C. Given the Turkish odalarda "in the rooms" and masalarda "at the tables", give the Turkish words meaning "in the books" and "on the horse". 2-18 In the Niutao dialect of the Tuvaluan Polynesian language, some verbs and adjectives have different forms with singular and plural subjects, as in these examples: SINGULAR

PLURAL

Tilt

'transformed'

to struggle

Chat

'Schock'

(Continuation)

• 69

7 0 • Chapter 2 Words and their parts: lexicon and morphology

SINGULAR

PLURAL

Apfel Sit Maasei Takato Crazy Eat

Apple stray, stupid people.

'turn' 'turn' 'bad' 'lay down' 'stupid' 'eat'

a. Describe the morphology rule that derives the plural forms of these verbs and adjectives from the singular forms. B. In the Funaafuti dialect of the same language, the process is slightly different, as shown by the following plural forms of the same verbs and adjectives. (Double consonants indicate that the sound lasts for a long period of time.) How are plural forms formed from singular forms in this dialect? How does this process differ from the Niutao dialect plural formation process described in the first part of this exercise? Valea maffuli feppaki appulu

nnofo maassei takkato kkai

2-19 Using the examples below, determine whether the following languages have isolating, inflectional, or agglutinating morphology, and justify your answer. Samoan water

GIFT

OBJECT

maalamalama aʔu i understand 'I understand the lesson'.

the

Mataʔupu Class

FINNISH tyttö silitti girl SUBJECT SONG. Iron PASTS. The girl ironed the shirts.

camisa payat-OBJEKT-PLURAL

JAPANESE akiko-ga haruko-ni mainiti tegamio Akiko SUBJECT Haruko for daily letter OBJECT 'Akiko writes a letter to Haruko every day.

write kaku

MOHAWK t-en-s-hon-te-rist-a-wenrat-eʔ DUAL-FUTURE-REPETITIVE-PLURAL-REFLECTIVE-metal-cross-PUNCTUAL 'They will cross the railroad tracks.' THAI khruu hây sàmùt nákrian sAam lerm the teacher gives the student three notebooks ARTICLE 'The teacher gave the students three notebooks.'

Exercises 2-20. Examine the following Tok Pisin (New Guinea Pidgin English) sentences to identify the morphemes needed to translate the seven English sentences at the end of this exercise. a. manmeri hello people

Wokabaut long rotted

you go

she

estrada

"People are walking down the street." B. mi harim toktok bilong yupela I

listen to the speech of

you Plural

'I hear your speech (PLURAL).' C. my harim speaks for you i

listen to the speech of

they sing

'I hear your (SINGING) speech.' i.e. I'm not a brat

long time in ole

he is brother of

harim toktok bilong mi

hear about it

Mim

He and his brother listen to my speech. Y mi laikim dispela manmeri long rot I

cut

Those

People

she

estrada

"I like these people who are on the go." F. dispela man bilong mi ol this do not turn on

husband and friend of

Likekim scattered

they liked me

Network

This man and my friend like this speech. Now, based on the meaning of the morphemes you can identify in the Tok Pisin sentences above, translate the following sentences into Tok Pisin: 1) These people like my speech. 2) I walk on the street. 3) I like my friend's speech. 4) I like my brother and these people. 5) These people on the street and my friend like your speech. 6) You (QTY.) and my brother like what these people are saying. 7) These people are listening to my friend and my brother's speech.

Especially for educators and future teachers 2-21. Suppose you are teaching young ESL students to change verbs to their "opposites", for example they appear to disappear. How would you get them to provide so many different prefixes in the English language to turn verbs into other verbs with the opposite meaning? 2-22 a. Suppose you are teaching a high school English class how to decipher the lexical category (part of speech) of the words newest, books, playing, and surprise. Would it be better to present them individually or in sets? Why?

• 71

7 2 • Chapter 2 Words and their parts: lexicon and morphology

B. Now look again at the books and the surprise and try to break them down into two sentences each, using one as a verb and the other as a noun. Any additional comments on how best to present them? What else could you do to provide your students with useful tools for identifying the parts of speech of these words? How did you determine your part of speech? 2-23 Imagine you and your students are looking at a map of the United States and searching for place names and river names. Focusing on one side or the other of the Mississippi, can you think of six names that your students would correctly guess came from Native American languages? If you scan the entire map, which states or state capitals did you identify by someone's name? To another country? How to use the credits in Spanish? French? Dutch? 2-24 How could you encourage your students to identify food names that English borrowed from the languages of the students' ethnic backgrounds? 2-25 How would you determine which words to include in the vocabulary of a textbook for international students learning basic English? 2-26 What would you say to students who thought English had attachments because they heard Eliza Doolittle use the word abso-bloomin'-lutely in that letter in the movie My Fair Lady: "Oh, wouldn't that be wonderful ? Oh, how nice to sit absolutely still! I would never move until spring falls on my windowsill"? Does English have other similar examples?

Other Resources The Internet addresses listed in this section may be helpful for understanding this chapter or for diagnosing language issues. Internet addresses change often, so those given here may be out of date. Wadsworth: http://www.thomsonedu.com/english/finegan

The Language: Its Structure and Use, 5th ed. website provides up-to-date Internet addresses and supplemental materials for students and faculty using this textbook. Merriam-Webster online: http://www.m-w.com

Entry into the world of Merriam-Webster dictionaries. It's worth checking out their definitions, which are available online. Corpus Linguistics Tutorial: http://www.georgetown.edu/cball/corpora/tutorial.html

Catherine Ball maintains a three-hour online corpus linguistics tutorial. If you're interested in corpus linguistics, this page is a good place to start your exploration. Corpus Linguistics: http://www.athel.com/corpus.html

Maintained by Michael Barlow, this site (rated 4 stars by Magellan) is a goldmine of corpus references in many languages, as well as software for exploring corpora and information on many other aspects of corpus linguistics.

Immersive bookmarks for corpus-based linguistics: http://devoted.to/corpora

Maintained by David Lee, this site is another gold mine of information on corpus and corpus-based linguistics, but mostly for English. LTG Helpdesk: http://www.ltg.ed.ac.uk/helpdesk/faq/index.html

It provides a set of Frequently Asked Questions (FAQs) and answers, as well as a large number of links to a variety of language technology projects. Frequently asked questions that will be answered include "I'm looking for a tagged English corpus." "Is there a part of speech tagger for Spanish?" "I'm looking for a list of common words in English, French, Italian, Russian, Polish." You can also find links to online bookmarks that assign part-of-speech tags to the texts you submit.

Suggestions for further reading • Jean Aitchison. 2003. Words in the Mind: An Introduction to the Mental Lexicon, 3rd ed. (Malden, MA: Blackwell). A fun and affordable gift.

• Andrew Carstairs McCarthy. 2002. An Introduction to English Morphology (Edinburgh: Edinburgh University Press). A concise and accessible treatment of English words and their structure, including a chapter on the historical sources of English word formation.

• Ronald W Langacker. 1972. Fundamentals of Language Analysis (New York: Harcourt). An excellent introduction to linguistic analysis and problem solving. Chapter 2 covers morphological analysis with illustrations from various languages, including Native American languages. Helpful sample solutions are provided for some problems.

• Donka Minkova and Robert Stockwell. 2001. English Words: Structure and History (Cambridge: Cambridge University Press). A book rich in examples and easy to use.

• 12,000 Words: Supplement to Webster's Third New International Dictionary. 1987. (Springfield, MA: Merriam). A list of 12,000 new words added to English in the 25 years since Webster's Third New International Dictionary was published in 1961.

• "Among the new words". American Speech. In each quarterly issue of American Speech, you'll find a column that defines the latest additions to the English vocabulary. You will be surprised how many words that are part of your everyday life are new to English. Next time you visit your library's journal room, look for the latest version of Among New Words.

• Merriam-Webster's new book of word stories. 1991. (Springfield, MA: Merriam). This engaging book features word histories for thousands of English words, from killer to zombie, and all sorts of interesting words in between, like Jeep and OK.

Further Reading A good general treatment of morphological processes can be found in Katamba and Stonham (2006). More advanced are Matthews (1991) and Bauer (2003), which contain a short chapter on the morphology of the mind and another on the historical development and disappearance of morphology in languages. Our examples of Turkish reduction come from Underhill (1976). Good treatment of morphology can be found in Shopen (1985), particularly in the chapters

• 73

7 4 • Chapter 2 Words and their parts: lexicon and morphology

by Stephen R. Anderson in "Typological Distinctions in Word Formation" and "Flexional Morphology", by Bernard Comrie in "Causative Verb Formation and Other Verb-Derived Morphology" and by Comrie and Sandra A. Thompson in "Lexical Nominalization". Comrie (1990), from whom several examples are taken in this chapter, provides valuable descriptions of more than 40 major languages, often including a discussion of morphology. The Vietnamese example is from Comrie (1989).

References • Bauer, Laurie. 2003. Introduction to Language Morphology, 2nd ed. (Washington, DC: Georgetown University Press).

• Comrie, Bernard. 1989. Universals of Language and Linguistic Typology, 2nd ed. (Chicago: University of Chicago Press).

• Comrie, Bernard, ed. 1990. The world's major languages. (New York: Oxford University Press).

• Cotton, Francis and John Stonham. 2006. Morphology, 2nd ed. (Basingstoke: Palgrave Macmillan).

• Matthews, PH 1991. Morphology, 2nd ed. (Cambridge: Cambridge University Press). • Shopen, Timothy, eds. 1985. Portions of the Language and Lexicon, Vol. 3 the typology and syntactic description of the language. (Cambridge: Cambridge University Press).

• Unterberg, Robert. 1976. Gramática turca. (Cambridge: MIT Press).

The sounds of languages: phonetics

WHAT DO YOU THINK ? • A reading genius, his third grade niece reports one day that English has five vowels: a, e, i, o, and u. They think English must have a dozen or more different vowels. How would you find out how many vowels English actually has? • His friend Fran says George Bernard Shaw claimed that English spelling was so messy that ghoti could be pronounced fish, and she challenges him to identify words whose pronunciation and spelling may have led Shaw to this seemingly absurd conclusion . What words can you name where ⬍gh⬎ is pronounced "f"? Can you name any names that start with ⬍gh⬎, like Shaw's ghoti? • Your roommate wants to know if English has 26 sounds that correspond to the 26 letters of the alphabet. You know that English has more than 26 sounds, and you point out that there isn't a single letter that represents the initial sound, so English uses two letters. They are then asked to give other examples where two letters represent a single sound. What examples can you name? • A friend quotes put and putt as two English words that are pronounced and spelled differently, but whose spelling does not match the pronunciation, and claims that there are many similar pairs in English and asks him to name just one. You can do it?

7 6 • Chapter 3 The sounds of languages: phonetics

Sounds and spells: not the same As an English reader, you are used to seeing written language as a series of words separated by spaces, with each word being a sequence of letters also separated by spaces. Easily recognizes that words exist as separate units made up of a relatively small number of different sounds. For example, English speakers easily consider the words spit and post as four sounds, while adults have five and set three. A little less obvious is the number of sounds when speaking words, series, letters, and sequences that don't have the same number of letters and sounds. This discrepancy is common in English. Cough has three sounds but is written with five letters; Despite its seven letters, Freight has only four sounds. Seven-letter-through and four-letter-through are alternative spellings for a three-phone word. Fone and Risa each have three sounds represented by five letters. Delicacy, with the same number of sounds and letters, use the letter ⬍c⬎ to represent two sounds, one similar to a k and one similar to an s. It is important because of the close connection between writing and speaking in the minds of educated people to emphasize that in this chapter we are interested in the sounds of spoken language, not the letters of the alphabet that those sounds represent in writing.

Same Spelling, Different Pronunciation Notice the variety of pronunciations represented by the same letter or sequence of letters in different words. Consider the pronunciations of the following words, all partially represented by the letters ⬍ough⬎: hard cough branch over public thoroughfare

"koff" "straight" "arc" "through" "tho" "thurrafare"

While the exact sounds of these words may vary among English speakers, the lesson of the distant relationship between the sounds and the letters remains clear. The spelling ⬍ough⬎ represents at least six pronunciations in English, as shown in Figure 3-1.

FIGURE 3-1 Same spelling, different sounds

"the end"

"uf"

"ai"

"She"

"Ö"

"and"

Sounds and spellings: They are not the same

Same Pronunciation, Different Spelling Other sets of English words are pronounced the same but spelled differently, as school-age children learn as they learn sets of homophones (or homonyms) such as there/your, bear/bare, led/lead, and to /two/ even. Consider the words in Figure 3.2, where nine different spellings represent a single sound, as in the word see. Other spellings could be cited for the sound of the word see, including situ and cee (the name of the letter). Note that the letter ⬍x⬎, as in sexy and foxy, represents the two sounds [k] and [s] represented in folksy.

Figure 3-2 Different spellings, same sounds see/senile/sea/seize/scene/enclosure/roof/cedar/cease/juicy/bright/sexy see

com

to damage

to be

cena

your

wer

"I understand"

Compare the sound and spelling of mujer and mujer and you will see that the difference in the letters ⬍a⬎ and ⬍e⬎ does not represent a pronunciation difference, as the second syllables of these words are pronounced the same. On the other hand, ⬍o⬎—the letter that doesn't change—represents two sounds (in females as ⬍oo⬎ of wood and in females as ⬍i⬎ of victory). The pair Satan (the devil) and Satin (the cloth) illustrate the same point: the ⬍a⬎ of the first syllable represents two sounds, but the spellings ⬍a⬎ and ⬍i⬎ of the second syllable represent the same sound. The same is true for loose and losse, where the only difference in pronunciation is in the final sound ([s] vs. [z]), while the only difference in spelling is in the identically pronounced vowels. Playwright George Bernard Shaw was a strong supporter of spelling reform, and highlighted the problems of mapping English sounds and spellings when he provocatively stated that fish could be written ⬍ghoti⬎: the ⬍gh⬎ as a cough, the ⬍ o⬎ as with women, and the ⬍ti⬎ as in the nation. Despite the efforts of Shaw and other reformers, the spelling of English remained essentially unchanged. You can see very modest success in simplifying in simple notations such as thru, nite, and photo, although even these examples are not widely used for the more traditional forms through, night, and photo.

Why and reasons for sound and spelling discrepancies Here are five reasons for the discrepancy between the pronunciation and the written representation of many English words.

• 77

7 8 • Chapter 3 The sounds of languages: phonetics

1. Written English has several origins with different spellings: • Anglo-Saxon The system that developed in Anglo-Saxon England before the Norman invasion of 1066 gave us spellings like ee for the sound of words like action and seen. • Norman French The system that the Normans, with their French spelling habits, superimposed on the Old English system gave us spellings like queen (for old cwene) and thief (for old theef). • Dutch Caxton, the first English printer who was born in England but lived in Holland for 30 years, gave us spellings like Ghost (replacing Gost) and Ghastly (replacing Gastlic). • Spelling reform During the Renaissance, attempts to reform spelling along etymological lines (ie historically earlier) gave us blame for the earlier det or dette and salmon for the earlier samon. 2. A spelling system established hundreds of years ago is still used to represent a language that is constantly changing its spoken form. For example, the initial ⬍k⬎ in words like Knock, Knot, Know, and Knee has already been pronounced, as has the ⬍gh⬎ in Caballero and Pensamiento. As for the vowels, the pronunciation changes that occurred during the evolution of the writing system and subsequent changes in pronunciation led to discrepancies, as seen in beat vs. great and food vs. foot, where different vowels are represented by the same spelling. 3. English is spoken differently around the world (and in different regions of a nation) despite relatively uniform spelling standards. This uniformity of spelling facilitates international communication, but also increases the differences between English spelling and speech. 4. A certain part of the word can be pronounced differently depending on the surrounding sounds and stress patterns. In electrical engineering, the final ⬍c⬎ represents the sound [k] as in kiss, but in electricity it represents [s] as in fool. In senile the ⬍i⬎ represents the ⬍i⬎ sound in I'll, but in senility it represents the ⬍i⬎ sound in sick. 5. Spoken forms can differ in social situations. The spelling system contains some variations (do not vs. don't and it was vs. 'twas), but there is little tolerance for spellings like going ('going to'), want ('want'), and gotcha. ('have you') and even less or nothing for j'eat ('have you eaten?') and woncha ('haven't you?'). A variable spelling of the same phrase would force readers to determine the pronunciation of the language represented before arriving at the meaning, rather than reading the meaning directly, as adult readers often do.

Advantages of Fixed Spellings Some disadvantages of an inconsistent set of spelling matches are obvious. Although less obvious, the benefits are also significant. Consider Chinese, in which many written characters have little or no relation to sounds but directly symbolize meaning, as do numbers like 3 and 7 and symbols like ⫹ and % for European languages. However, such characters allow groups of people whose spoken language is mutually incomprehensible to communicate well in writing, as is the case between speakers of Cantonese and Mandarin Chinese. At the same time, note that the symbol 7 (or the slight variant 7–) has a unified meaning in European languages, although the word for the term is pronounced and spelled differently: seven in English, sept in French, sette in Italian, seven up German etc. Likewise, the fact that English spelling is somewhat independent of pronunciation isn't a bad thing considering English has exceptionally diverse dialects.

Sounds and spellings: Not the same from New Zealand to Jamaica to India and in places where English is used in official capacities alongside the mother tongue or as a second language for academic and other international activities. Despite differences in pronunciation around the world, a uniformly written word is associated with only one set of meanings. Even in a language with different pronunciations for the same element of meaning, stable spellings can contribute to reading comprehension, such as z]). Independence of writing and speaking It is important to distinguish between the sounds of a language and its written representation. To emphasize phonetic and spelling independence, remember that a given language can be represented by more than one writing system. For example, the language widely known as Hindi-Urdu is written by Hindus living in India in Devanagari, an Indian script derived from Sanskrit but written in Arabic script by Muslims living in Pakistan and parts of India. Sometimes people also adopt a new writing system for their language. In the early 20th century, the Turkish government changed orthography (the technical name for a writing system) to represent Turkish from an Arabic script to one based on the Roman alphabet. Languages sometimes use different hyphens for different purposes. Imagine you are sending an international telegram in a language that uses a different script than the Latin alphabet: for example Japanese, Korean, Greek, Russian, Persian, Thai or Arabic. Instead of using their usual spelling, speakers of these languages use the Latin alphabet to send telegrams internationally. An alternative spelling can also be necessary within a country: In China, each character is assigned a four-digit number, these numbers are sent by telegraph and then "translated" back into Chinese characters. Sometimes a language uses more than one writing system for different aspects of writing. Japanese is based on three spellings: kanji, based on the Chinese character system in which a symbol represents a word regardless of its pronunciation, and two syllables. A syllable is a writing system in which each symbol represents a spoken syllable. All over the world there are discrepancies between spoken and written sounds.

Image unavailable due to copyright restrictions

• 79

8 0 • Chapter 3 The sounds of languages: phonetics

Chapter 12 tells you more about the written representation of languages. Here we focus on the human vocal tract and the sounds it produces; Chapter 4 examines the nature of tonal systems in human speech.

Phonetics: The study of sounds Phonetics is the study of sounds that arise during the production of human languages. It has two main branches.

• Articulatory phonetics focuses on the human vocal tract and describes sounds in terms of their articulation in the vocal tract; It was fundamental to the discipline of linguistics. • Acoustic phonetics uses the tools of physics to study the nature of the sound waves generated in human speech; It is becoming increasingly important in linguistics when attempting to use machines to interpret speech patterns in voice identification and language-initiated mechanical operations.

Our discussion will be confined almost exclusively to articulatory phonetics, the nature of human sounds produced by the vocal tract.

Phonetic Alphabets In order to refer to the sounds of human speech in terms of their articulation, phoneticians have developed descriptive techniques that avoid the difficulties of describing sounds in terms of traditional writing systems. You already know that it is impossible to use standardized written representations to analyze phonetic structure because, even within the same language, some phones correspond to more than one letter and some letters correspond to more than one phone. Also, a single letter can be used to represent different sounds in different languages. So we need an independent system to represent the actual sounds of human languages. In scientific discussion, clarity and consistency are the necessary properties of symbols used to represent sounds. The best resource is a phonetic alphabet, and the most common is the International Phonetic Alphabet (IPA). The IPA provides a unique written representation of every sound in every language. Table 3-1 contains a list of symbols used to represent consonants in English. Show the phonetic symbol for each sound and the words with the relevant parts highlighted. Using an option approved by the International Phonetic Association, we use a standard printed ⬍r⬎ to represent the initial sound for walking, but you should know that the Association maps the symbol [r] to a different sound. (We have included the standard IPA initial symbol in parentheses.) Words illustrate initial, intermediate, and final occurrences of sounds.

The Vocal Tract The processes that the vocal tract uses to produce a variety of sounds are similar to those used in wind instruments and organ pipes, which produce different musical sounds by varying the shape, size, and acoustic character of the cavities through which air flows passes. He has passed.

Phonetics: the study of sounds.

TABLE 3-1 English Consonants by Word Position PHONETIC SYMBOL

INITIAL

MEDIUM

FINALE

pag Seconds Track Grams F v θ ð s z ʃ tʃ re Northern Metro ŋ h j r (ɹ) l w

Count pills till dill kill gills fill fine village then bark stupid zebra ——— jelly mill zero ——— hills yes rent lily go

hijacking petunia seduce sick dagger muscular savior author leather mason shell rash measure kitchen stick fool fool sunny singer ahoy beyond berry fool away

tap tab bat pad lick bag boss grab breathe kiss shoes rush rouge clay fudge broom spoon sing ——— toy deer factory ———

Let Your Source Every speech sound you make is different from every other speech sound because of a unique combination of characteristics in the way you shape your mouth and tongue, moving parts of your vocal tract. Examine the simplified drawing of the vocal tract in Figure 3-3. Here we look at the parts of the vocal tract and how they work together to produce sound. How are speech sounds produced? First, air flows from the lungs through the vocal tract, which shapes it into different speech sounds. The air then leaves the vocal tract through the mouth, nose, or both. Although speakers of all languages have the same vocal tract, no language takes full advantage of opportunities to form different sounds, with marked differences in the sounds produced in different languages. For example, Japanese and Thai do not have the English [v] sound for van, and Japanese does not have the [f] sound for fan. Thai lacks the sounds represented by ⬍g⬎ in Gil, ⬍z⬎ in Zebra, ⬍sh⬎ in Shell, ⬍s⬎ in time signature, and ⬍j⬎ and ⬍dg⬎ in Judge. French, Japanese, and Thai lack the distinct ⬍th⬎ sounds in ether and both.

• 81

8 2 • Chapter 3 The sounds of languages: phonetics

Figure 3-3 The nasal cavity of the vocal tract

Mass of the palate in the nasal cavity (velum)

hard palate alveolar teeth lip crest

uvula

center and blade on the rear tip for k LANGUAGE

root epiglottis

throat wall

Larynx

Larynx Upper surface of the vocal tract

underside of the vocal tract

Adaptado de Ladefoged 2006.

Just as some languages lack the sounds that English has, other languages have sounds that English don't have. You probably know that English lacks the r trill in Spanish and Italian, and that German at the end of words like Bach "stream" and hoch "high" has a sound that isn't in the English sound inventory. Arabic has a similar sound to Bach's German ⬍ch⬎, but in Arabic it can initially appear as a word. A similar (but not identical) word that ultimately occurs in the German word ich occurs in English (for those dialects that pronounce the ⬍h⬎) with the initial human and huge. Still, for English speakers learning German, it can be difficult to pronounce the sound of a word like I do, since English doesn't allow that sound at the end of a word.

The Vocal Cords and Voice Humans do not have organs that are used solely for speech. The organs that produce the sounds evolved primarily to serve the life-sustaining breathing and feeding processes. Language is a secondary function of the human "vocal system" and is sometimes thought of as a parasite on these organs. The vocal folds illustrate the 'parasitic' nature of speech: the main function of these two folds is to prevent food from entering the lungs through the wrong tube. In relation to speech, the vibration of the vocal cords distinguishes voiced from unvoiced sounds. You can tell the difference between voiced and unvoiced consonants by switching between the pronunciations of [f] and [v] or [s] and [z] while covering your ears with your hands. Can you tell if you pronounce the thin words thirty, so what happens to those when [θ] or [ð] are pronounced? Check your conclusions in Table 3-7 on page 95.

Descriptive Sounds

Describing sounds When you explore the inventory of sounds, you use your vocal tract to produce the sounds you describe. Say them out loud, paying attention to the shape of your mouth and the position of your tongue with each sound. This first-hand experience will familiarize you with phonetics benchmarks, make it easier to follow the discussion, and give you the confidence to master articulatory phonetics. As in our initial discussion, we will continue to use square brackets to enclose the symbols that represent sounds. Thus, [t] symbolizes the initial and final sounds in tot, [d] the initial and final sounds in did, and [z] the initial sound in zebra, the middle consonant in busy, and the final sound in hum and dogs. Speech sounds can be identified by their articulatory properties, i.e. where they are located in the mouth and how they are produced. All English consonants can be described using three properties:

• Voice (whether or not the vocal cords vibrate) • Venue of articulation (where airflow is most obstructed) • Form of articulation (the particular way in which airflow is obstructed)

Language Start by distinguishing between [s] (as in bus or sip) and [z] (as in buzz or zip). If you say a long continuous [zzzzz] and alternate it with a long continuous [sssss], you'll find that the position of your tongue in your mouth stays the same, although these sounds are noticeably different. You can feel this difference by touching your larynx (voice box or Adam's apple) while saying [zzzzz sssss zzzzz sssss]. The vibration you feel in your larynx when you say [zzzzz] but not [sssss] is called vocalization; It is the result of air pressure passing through a small opening (called the glottis) between two folds of mucous membrane (the vocal cords) in the larynx. It's like the blade with a slit that kids use to make a vibrating sound when blowing air. When the vocal cords are held together, air forced through them by the lungs causes them to vibrate. It is this very vibration or "voice" that distinguishes [z] from [s] and allows speakers to distinguish between two identical sounds. Using these very similar but different sounds allows us to create words that only differ in a single voice characteristic in a single sound but have vastly different meanings such as bus and buzz, sip and zip, peace and peas, sane and Zane. . Besides [s] and [z], other sounds are characterized by a voiced and unvoiced contrast. Consider [f] and [v], as in fine and vine: both sounds are made by forcing air through a narrow opening between the upper teeth and lower lip; [f] is silent and [v] is uttered. Other unvoiced/voiced pairs are [p] and [b] as in pet and bet and [t] and [d] as in ten and den.

Articulation mode [s] and [z] not only have a voice characteristic, but can also be characterized by their articulation mode. When pronouncing it, air is continuously forced through a narrow opening at a point behind the upper teeth. Compare the pronunciation of [s] and [z] with the sounds [t] and [d]. Unlike [s] and [z], [t] and [d] are not pronounced as continuous

• 83

8 4 • Chapter 3 The sounds of languages: phonetics

The airflow goes through the mouth. Instead, the air comes to a standstill behind and above the upper teeth and is then released (or exploded) in a short burst of air. For this reason, [t] and [d] are called stops, and since the air exits through the mouth (not the nose), they are also called oral stops. Sounds like [s] and [z] that are produced by a continuous stream of air passing through a narrow passage in the vocal tract are called fricatives.

Try it yourself: say the sounds [p], [b], [f] and [v] to determine which are plosives and which are fricatives.

Place of Articulation Of the sounds analyzed so far, [s] and [t] are voiceless, [z] and [d] are voiced. All four are pronounced with the point of greatest closure just behind the upper teeth. Say ten and give out loud, feeling where the tip of your tongue touches your mouth to hear the consonants. Both words start (and end) on the alveolar ridge. Because [t], [d], and [n] all articulate at the alveolar crest, they are called alveolar. [s] and [z] are also articulated on the alveolar ridge, as you will notice when pronouncing the words sin and zen. (Of course, [s] and [z] are fricatives, while [t] and [d] are stops.) There are three main points of articulation for stops in English: alveolar ridge, lips, and soft palate (or velum). As you say pin and bin, you'll notice that the first sound of each word causes air to collect behind both lips and then be released. Hence the point of greatest closure is on the lips, hence [p] and [b] are called bilabial stops (bilabial means “two lips”).

Try it yourself: Compare your pronunciations of [p] and [t]. They both have no voice, so what's the difference between them? Say word pairs like pin and tin or tripe and trite to see examples.

Look at the pronunciation of the first kin sound and you'll see that [k], like [p] in pill and [t] in till, is a voiceless stop, but differs in its place from [p] and [ t] distinguishes the articulation: [k] is pronounced when the tongue touches the palate at the velum (soft palate) and is called velar; It's a voiceless parade. The three unvoiced stops [p], [t], and [k] correspond to three voiced stops: [b] as in bib is a voiced bilabial stop; [d] as in did is a voiced alveolar stop; and [g] as in gig is a voiced velar plosive. English has three pairs of stops, and each pair is pronounced at a specific point of articulation, but one voiced and one unvoiced.

Try it yourself: Identify the pairs of distinct stops on the lips, alveolar ridge and velum.

Consonant Sounds In addition to the lips, alveolar ridge, and velum, English uses other articulators to produce some sounds. The ⬍th⬎ of thin is a fricative pronounced with the tongue between the teeth. It is described as a voiceless interdental fricative and has the Greek letter theta [θ] as its phonetic symbol. [ʃ] (the sound represented by ⬍sh⬎ in Shoot and Wish) and [] (the final beige sound and middle consonant in Bar) are pronounced between the alveolar ridge and the velum (or palate); the sounds produced there are called alveolar-palatal. [ʃ] is a voiceless alveolar-palatal fricative; [ ] is a voiced alveopalatal fricative.

Consonant sounds Consonants are sounds produced by the vocal tract partially or completely blocking the air from the lungs. If you look at the index of English consonants in Table 3–1 on page 81 and pronounce the sounds aloud while focusing on location and mode of articulation, you will see how the rest of the tables represent the distribution of the consonants in your voice. , your place of articulation and your way of articulation. Here we describe these consonants, grouped by their mode of articulation and described in terms of pitch and place of articulation. We focus on consonants in English and mention selected consonants in other languages.

Stops The main stops in English are [p], [b], [t], [d], [k], [g]. By pronouncing words with these sounds (see Table 3–1 on page 81), you can see that [p] and [b] are bilabial stops, [t] and [d] are alveolar stops, and [k] and [g] The sail stops. Stops when air builds up in the vocal tract and suddenly escapes through the mouth. ENGLISH STOPS PLACE OF ARTICULATION WITHOUT VOICE WITH VOICE

BILABIAL

of the alveolus

VELAR

glottis

In addition, many languages have a glottal stop. It is pronounced using the glottis to completely but briefly block the passage of air through the throat. The glottal stop is represented by [ʔ]. In English, glottal stop occurs only as a marginal sound, between the two parts of the exclamation Uh-oh! in American English and Cockney English as a medial consonant of words like butter and bottle, for example. In languages like Hawaiian, the glottal stop is a complete consonant that can distinguish two different words: paʔu "stain" and pau "it's over".

• 85

8 6 • Chapter 3 The sounds of languages: phonetics

Fricatives To pronounce the alveolar fricatives [s] and [z], air is forced through a narrow opening between the tip of the tongue and the alveolar ridge. English has a large number of fricatives, some are articulated before [s] and [z] and some after. Fricatives are characterized by the fact that air is forced through a narrow opening in a continuous stream. When pronouncing the first sound in the words thin, three, and theta and the last sound in teeth and bath, remember that the tip of the tongue is positioned between the upper and lower teeth, where the airflow is most constricted and exerts its most articulation. . The sound of these words, represented by [θ], is a voiceless interdental fricative. The voiced counterpart is the initial sound in the words there and then and the middle consonant sound in both. Note that in English the spelling ⬍th⬎ is used for two different sounds: [θ] as in ether and [ð] as in either or leather.

Try it yourself: say the following words to discover other fricatives and become aware of their common characteristics and different points of articulation: fine/vine; beef/irritated [f] [v] labiodental fricative coxa/thy; ether/quiero [θ] [ð] interdental fricative doline/zinc; alveolar fricative bus/buzz [s] [z] rush/rouge; fishin'/vision [ʃ] [] alveolar-palatal fricatives here; ahoi [h] Glottic fricative

ENGLISH FRICATIVE PLACE OF ARTICULATION

WITHOUT A VOICE WITH A VOICE

LABIODENTAL

INTERDENTAL

of the alveolus

Alveopalatinal

f v

θ ð

s z

glottis

Some languages have other fricatives. For example, Spanish has a voiced bilabial fricative (represented by [β]), as in ⬍b⬎ of the 'fin'-cable. Bringing the two lips together instead of the lower lip and upper front teeth. The West African Ewe language has voiced [β] and unvoiced [φ] bilabial fricatives. Spanish and many other languages have a voiceless velar fricative [x] and a voiced velar fricative [γ], the latter being less common. Pronounce [x] as if you are gently clearing your throat. The sound first appears in the Spanish word joya 'jewel' and in the personal name José (when borrowed into English, José is pronounced with [h], the sound closest to [x] in English). [γ] is represented by ⬍g⬎ in Spanish lago 'lake'. German, Irish, and Mandarin Chinese have a voiceless palatal fricative [ç], as in the German word Reich 'imperium'.

Consonant Sounds You may have noticed that the physical distance in the mouth between the points of articulation of English fricatives is not as great as that of plosives. The bilabial, alveolar, and velar places of articulation of the stop consonants are further apart than the labiodental, interdental, alveolar, and alveolopalatal fricatives. This closer spacing of the fricatives can make them difficult to perceive as different. The differences can be difficult to discern, especially for speakers of languages with fewer fricatives than English or languages whose fricatives are further apart. For example, French lacks the interdental fricatives [θ] and [ð], so French speakers tend to perceive (and pronounce) English words like thin and this as if they were "sin" and "zis". A French fricative familiar to English speakers, although English does not have it, is the voiced uvular r-sound (as in Paris or rue 'street'), pushed further back in the mouth and represented by [ʁ].

Afrikades Two English consonant sounds are more complex to describe than their stops and fricatives. These are the sounds that are first created in Chin and Gin and later in Batch and Badge. If you pronounce these sounds slowly enough, you can see that they are fricative stops, which we will call affricates. When pronouncing an affricate, the breath is established by a full closure of the mouth tract at a specific point in the articulation, then released (similar to a stop) and resumed (like a fricative). The sound in Chin is a combination of the plosive [t] and the fricative [ʃ] and is rendered as [tʃ]. The sound at the beginning and end of the judge is a combination of the stop [d] and the fricative [], represented as [d]. English only has this pair of affricates, and to record their place of articulation they are called palatal-alveolar affricates. Different languages have different affricates. The most common are the alveolar affricates [ts] and [dz], which appear at the beginning of the Italian words zucchero "sugar" and zona "zone," respectively. AFRICAN ENGLISH LOCATION OF THE ALVEO-PALATAL COMMON SOUND SOUND

t d

Obstruents Because they share the phonetic property of restricting airflow through the vocal tract, fricatives, stops, and affricates are called obstruents.

Approximants In English there are four sounds known as approximants because they are produced by two articulators that approach almost like fricatives, but are not close to each other.

• 87

8 8 • Chapter 3 The sounds of languages: phonetics

enough to create friction. The approximants in English are [j], [r] (IPA [ɹ]), [l] and [w]. The sound that begins the word you is the palatal approximant [j]; The word cute begins with the consonant cluster [kj]. Because [r] is pronounced by passing air through the middle part of the mouth, it is called the central approximant. To pronounce [l], air is directed to one or both sides of the tongue to make a lateral approach. To distinguish them from other approximations, [r] and [l] are sometimes called liquids. (In some Asian languages, [r] and [l] are not contrasting sounds, so native speakers of those languages can have difficulty distinguishing them when speaking or hearing them in English. This is a topic we'll come back to until the next chapter but one.) When pronouncing the approximate [w], the lips are rounded, as in savage. In certain dialects, some words have [h] before [w], as in which or if. When [w] is the second element of a consonant cluster (as in chords or rapids), the initial note (in these cases [t] or [k]) is rounded in anticipation of [w]. ENGLISH APPROXIMATE BILABIAL ARTICULATION SITE

of the alveolus

PALATAL

r (ɹ) l

VOICE (MID) VOICE (SIDE)

Nasal consonants are pronounced by lowering the velum, which allows airflow to pass through the nasal cavity instead of the oral cavity. English has three nasal stops: [m] as in mad, drummer, cram; [n] like new, sinner, ten; and a third, symbolized by [ŋ] and pronounced as in the words sing and singer. ENGLISH NOSE BILABIAL ARTICULATION SITE

Metro

of the alveolus

VELAR

Norte

Because of the way it's generally written in English, English speakers may think that [ŋ] is a combination of [n] and [g], but it's actually a single sound. You can test this yourself by comparing your singer and finger pronunciations. Ignore the initial sounds [s] and [f], if your pronunciation of singer and finger is different (not for some English speakers), then you have [ŋ] in singer and [ŋg] in finger (note that if you [ ng ] on your finger you would pronounce it as "finn-ger"). Most speakers of American English have a three-way contrast between boil, sinner, and singer depending on whether the middle consonant is [m], [n], or [ŋ]. By looking at where the tongue touches the top of the mouth when articulating these nasal consonants (and by comparing their place of articulation with other previously identified sounds), you can see that [m] is a bilabial nasal, [ n] is an alveolar nasal, and [ŋ] is a velar nasal. If, while saying [mmmm], you stop the flow of air through your nose by pinching it shut (like a clothespin would), the sound will stop.

The vowels sound sharp, showing that air is flowing through the nose when creating nasal stops. Compare cutting air through your nose while saying [nnnnn] and [sssss] and you'll get an idea of how the nasal and oral cavities work in producing sound. If you cut off the air passing through the nose, there is almost no difference in sound quality for spoken consonants; with a nasal consonant, the effect is completely different. If you have successfully identified the points of articulation of nasal alae and understood why they fit into their spaces on the consonant chart, you may have noticed that in English three sets of consonants are articulated in the same places and that they differ only in their articulation: the Oral stops [p] and [b] and the nasal stop [m] are bilabial; the mouth stops [t] and [d] and the nose stop [n] articulate at the alveolar ridge and are called the alveolar; [k], [g] and [ŋ] are articulated on the velum and are called velars. The English nasal consonants are [m], [n], [ŋ]. Different languages have different noses. French, Spanish, and Italian all have a palatal nose [ ≠], which you can recognize by the French word mignon "sweet" (which in English borrowed from the phrase filet mignon), the Spanish words mañana, señor, and canón (which became from borrowed from English as canyon) and Italian bagno 'bath' and lasagna (also borrowed from English).

Clicks, Rattles, Trills Some languages have consonants that belong to the same classes we discussed, but are quite different from those of European languages. Several South African languages have certain clicks under their stop consonants, which are an integral part of their phonetic system. An example is the lateral click made on the side of the tongue; it occurs in English, for example, when we goad a horse, but it is not part of the solid inventory of the English language; is represented by the IPA icon [ ]. Another clicking sound that occurs in some of these languages can be represented by the tsk-tsk censor in English script. This last click is not a lateral but a dental (IPA []) or (post)alveolar (IPA [!]) click made with the tip of the tongue on the teeth or on the alveolar ridge. Some consonant sounds are not stops, fricatives, affricates, approximants, or nasals. The middle consonant sound in the words butter and metal is commonly pronounced in American English as an alveolar flap, which is a short, high-velocity stop produced by slapping the tongue against the alveolar ridge. We represent these keys by [ɾ] (a sound discussed later in Chapter 4). Spanish, Italian, and Fijian all have an alveolar trill r, as in Spanish corre 'run'. ] (instead of the IPA symbol [r]).

Vowels Vowels are produced by passing air through different shapes of the mouth, with different positions of the tongue and lips, and with relatively unimpeded airflow through narrow passages except the glottis. Some languages only have three different vowels; others have more than a dozen. You may have thought that English only has five vowels, but counting to five reflects writing better than speaking. If you pronounce the following words, you'll find that English has at least a dozen different vowels: peat, pit, pet, pate, pat, put, putt, pool, poke, pot, part, and port.

• 89

9 0 • Chapter 3 The sounds of languages: phonetics

Vowel Height and Frontality Vowels are characterized by the position of the tongue and the relative roundness of the lips. Based in part on auditory perception, we refer to vowels as high or low and front or back. We also take into account whether the lips are rounded (like the pool) or not rounded (like the pill).

Try it yourself: You can get a feel for these descriptions by saying food and food alternately: the first contains a front vowel, the second a back vowel. For the height of the tongue, alternately say feet and fat. If you can't feel the difference between the high and low vowels on this pair of sounds, look in the mirror (or at a classmate saying them); the mouth is more open for the fat vowel than for the foot vowel. The reason? The tongue is less fatty.

Figure 3-4 shows the relationship of English vowels to each other and the approximate positions of the tongue during articulation.

Figure 3.4 The front vowels in English

central

cost

Your

Alt

pretty much

”

under

and

Need

Here are the English words for each of the vowel symbols shown in the image. Note that these words were chosen based on North American English; Pronunciation in British English may differ for some: i  e ε

Mascot, batting pit, a little late, mascot, batting bet, bat

and

vorbei, couch putt, but park (in boston)

your υ is my ɑ

Pool, cleats, foot train, boat dock, purchased marijuana, father

Vowels The symbols [ə] (called schwa) and [] (called circumflex or wedge) represent similar sounds. Both occur in the preceding word [əbv]. We use [ə] to represent a middle central vowel in unstressed syllables, such as B. the second syllable of bus [bsəz] and the second and third syllable of able [kepəbəl]. We also use it before [r] in the same syllable, either stressed as in person [pərsən] and lord [sər] or unstressed as in pertain [pərten] and tender [tεndər]. We use [ ] to represent central middle vowels in other stressed syllables, such as B. suds [sdz] and the first syllable of flooded [flɾəd]. (Some textbooks use [] to represent a middle central vowel with the color r. In systems using the [] notation, person [psən], lord [s], and affiliation [pten] would be transcribed.)

Diphthongs English also has diphthongs represented by pairs of symbols to record the fact that a diphthong is a vowel where speech begins in one place and slides to another. Say it slowly to get an idea of what a diphthong is: [aj] (like in motion); [aw] (how loud); [ɔj] (as in child, toy). (Some books transcribe these diphthongs as [ay] or [a], [au] or [aυ], or [ɔy] or [ɔ].) Slowly pronouncing the words buy, boy, branch. As such, American English dialects have as many as thirteen distinct vowels (plus three diphthongs). There are sixteen distinct vowels and diphthongs in England and certain parts of the United States, including metropolitan New York City. Elsewhere in the United States, there are fewer distinct vowels because there is no distinction between Hawk and Hock vowels.

Other articulation features of vowels To create differences between vowels, languages can explore ways other than pitch and tongue back. Vowels can have stress, rounding, lengthening, nasalization, and tone. Stressed languages make it possible to distinguish between tense and relaxed vowels. These designations represent a set of features that distinguish one group of vowels from another. For example, single vowels do not appear at the end of a stressed syllable and tend to be shorter; They also tend to be more centered than the nearest vowel. The contrast between [i] peat and [] hole is partly a tense/relaxed contrast; likewise for the vowels in bait/bet and in cooed/could. Loose vowels are non-syllable-ending, shorter than tense vowels, and more mouth-centered. Hence English has the single vowels [ ε υ] as in pit, pet, put. The corresponding vowels are [i and u] as in beat, bait, boot. English single vowels [ ] have no corresponding tense vowels. Rounding While in English high front vowels automatically tend not to be rounded (and high back vowels tend to be rounded), some languages have both rounded and unrounded front vowels. French and German have rounded high- and mid-fronted vowels, as well as unrounded ones. French has a high front unrounded vowel [i] in words like dire 'say' and dix 'ten' and a high front rounded vowel [ü] as in rue 'street'; it also has a contrast between the unrounded upper center front [e] (as in fée 'fairy') and the rounded upper center front [ø] (feu 'fire'); and between the unrounded lower front-middle [ε] (serre 'greenhouse') and rounded lower front-middle [] (soeur 'sister'). German has similar contrasts.

• 91

9 2 • Chapter 3 The sounds of languages: phonetics

Length German has two vowels of each type, one long and one short. Long vowels are pronounced longer than short vowels. Long vowels are usually represented in phonetic transcriptions with a special colon after them, or with the double vowel symbol. (In dictionaries and some writing systems, a macron (¯) may be used over the vowel symbol.) For example, in German, alongside the short vowels [i] and [ü], as in bitten "bitten" and must "must", there are words with high front long vowels: unrounded [i] in offer 'wish' and rounded [ü] in mill 'mill'. These examples illustrate how languages can multiply vowel differences by exploiting long and short varieties. English also has vowels of different lengths, although it doesn't use the lengths to form different words (see Chapter 4). To feel differences in vowel duration, pronounce the English words beat, bead, bit. You should be able to hear that the count vowel is longer than the time vowel and both are longer than the bit vowel. Nasalization All types of vowels can be nasalized by pronouncing the vowel while air passes through the nose (as in nasal stops) and exits through the mouth. Nasal vowels are indicated by a tilde (~) over the vowel symbol. French has several nasal vowels parallel to oral vowels: lin [lε˜] 'lino' ment [mɑ˜] '(he) lies' honte [ɔ˜t] 'shame'

Milch [lε] 'Milch' ma [mɑ] 'I' (weiblich) hotte [ɔt] 'box'

Other languages with nasal vowels include Irish, Hindi, and the Delaware, Navajo, and Seneca Native American languages. Pitch In many Asian, African, and North American languages, a vowel can be pronounced at multiple pitches and are perceived as different sounds by native speakers of those languages. Normally, a vowel pronounced in a low pitch contrasts with the same vowel pronounced in a higher pitch. An example of a two-tone language is Hausa, spoken in West Africa. In Hausa, the word for "bamboo" is górà, with a high tone (´) on the first syllable and a low tone (`) on the second syllable. Compare this to the word gòrá, where the tone sequence is reversed and the meaning is "big gourd". Some tonal languages have more complex systems. Beijing Chinese dialect has a high tone (symbolized by ¯); a rising tone (´); a falling-rising tone (ˇ), where the pitch begins to fall and then rises sharply; and a falling tone (`), where the tone drops sharply. There is fourfold pitch contrast between the following vowels, which are distinct words. I (high level)

'1'

step into)

'suitable'

I (falling-rising)

'von'

I (fall)

'Thought'

A particular accent can be used to represent different sounds in different languages. Thus, ´ represents a high tone in Hausa but a rising tone in Chinese; in Hausa ` stands for a low tone, but in Chinese for a falling tone.

Vowel sounds Thai have five tones; the standard Vietnamese six-tone dialect; and the Guangzhou (Canton) dialect with nine distinct Chinese tones. Sound is a ubiquitous and diverse phenomenon. Tables 3–2 through 3–5 are vowel tables illustrating sound patterns for four languages: French, Spanish, German, and Japanese.

TABLE 3-2 French vowels with figurative words UNROUNDED FRONT

ROUND FRONT

Me and

UNROUNDED MIDDLE

ROUND BACK

ORAL

high medium high medium low medium low

you are one

NASAL

low medium low

9 0

is greasy 'grey' is tight 'closed' is fresh 'fresh' 7 sole 'branch'

ü ripe 'mature' ø fast 'fast' young 'young' 8 brown 'brown'

ə way 'way' to through 'through'

u fou 'loco' o mot 'palabra' o forte 'fuerte' 9 fundo 'fundo' 0 faon 'gefunden'

TABLE 3-3 Spanish Vowels with Illustrative Words FRONT UNROUNDED HIGH MID LOW

UNROUNDED MIDDLE

Me and

ROUND BACK

you're a

Witz 'joke' and fe 'belief'

al mar 'ser'

u in 'sul' oder mouth 'mouth'

• 93

9 4 • Chapter 3 The sounds of languages: phonetics

TABLE 3-4 German vowels with illustrative words UNROUNDED FRONT

ROUND FRONT

UNROUNDED MIDDLE

ROUND BACK

I i

uh uh

she

la la

ALT

Long short

MEDIUM HIGH

Long short

and and

PRETTY MUCH

and

LOW MID short

Long short

UNDER

Long short

a a i offer 'I want' i bit 'ask' e wen 'who' e if 'when' ε Käse 'cheese'

ü mill ə honey 'honey' ü muss 'must' a crow 'crow' ø ölig 'oily' a mouse 'mouse' roentgen 'roentgen'

u hen 'chicken' u walnut 'mother' o horno 'oven' o ox 'ox'

TABLE 3-5 Japanese Vowels with Illustrative Words UNROUNDED FRONT HIGH MID LOW

UNROUNDED MIDDLE

BACK NOT ROUNDED

ROUND BACK

Me and

by uno

I am 'now' ε Sensei 'teacher'

the' autumnal'

ɯ buji 'seguro'

o you 'approach-saying'

Tables 3–6 and 3–7 summarize all of the vowels and consonants presented in this chapter.

Swedish vowels

• 95

TABLE 3-6 Vowels discussed in Chapter 3

(All vowels can be nasalized and short or long.)

Time high relaxed medium high medium low medium low

FRONT NOT ROUNDED

ROUND FRONT

I  and

she

UNROUNDED MIDDLE

BACK NOT ROUNDED

ROUND BACK

in them

and

ɔɑ

TABLE 3-7 Consonants discussed in Chapter 3 ARTICULATION PLACE ARTICULATION FORM AND VOICE

BILABIAL

LABIODENTAL

INTERDENTAL

of the alveolus

Alveopalatinal

PALATAL

VELAR

VULARES

glottis

STOPS

without a voice

Metro

Norte

NASAL

≠

Norte

x c

Frightening

without a voice

f b

f v

θ ð

s z

ts dz

t d

AFRICA

DEAF CUSTOMIZED SOUND

zentral sound side sound

r (ɹ) l

OTHERS

curved key curved trill

6 (r) ɾ

9 6 • Chapter 3 The sounds of languages: phonetics

Informatics and phonetics Alphabetic writing systems rely on the notion of discrete sounds, and it has proven useful for linguists to think of speech sounds as discrete. But in reality, for example in a conversation, the sounds are not discrete and do not occur separately. Instead, each sound reproduces the next sound in a word (and in an utterance), and the sounds and words are merged. Imagine a computer that could generate discrete sounds that sound natural when you talk to yourself: it could use the sounds [], [n], [k], [l], [u], [d ], [ə] and generate [d]. ] ]. If you put these sounds together in the sequence [nkludəd], you can expect a sound similar to the word it contains. But there are complications. As normally pronounced, include does not have the same [] sound as sit, and the added [n] in is often pronounced more like the [ŋ] in sing than the [n] in tin. So if natural-sounding words are the goal, their production cannot depend on a simple sequence of discrete individual sounds.

AND

Suppose the computer instead composed the sounds [˜], [ŋ], [k], [l], [u], [d], [ə], and [d]. This combination would sound more natural but would still be rigid. First, the first and second [d] sounds of include are different from each other, and both are different from the [d] sound of dig. Further refinement would not suffice, since the individual sounds would have to meet as in natural language. They could not be separated as phonetically printed or transliterated. As an added complication, consider that a word like photo could be phonetically represented as [foɾo], but the same morpheme in photography would be pronounced [foɾə] and in photographer [fəthɑ]. In other words, without some general principles of pronunciation, a computer could not simply combine the sounds represented in the spelling and produce synthesized speech that sounded remotely like natural speech. We will come back to this topic in the next chapter. ■

Summary • Sounds must be distinguished from letters and other visual representations of speech. • Phonetic alphabets represent consistent and comparable sounds in different languages; Each sound is assigned a unique representation, independent of the usual writing system used to represent a particular language.

• This chapter uses the International Phonetic Alphabet (IPA). • All languages contain consonants and vowels. • Consonants can be produced by impeding the flow of air from the lungs through the vocal tract and through the mouth or nose.

• With fricative consonants, air forced through a narrow opening forms a continuous sound, as in the initial and final sounds of diz [sεz] and fish [fʃ].

• At stop consonants, airflow is completely blocked and then released, as in the initial and final sounds of the words tap and cat.

• Affects are produced by combining a stop and a fricative, as in the final sound of the word pessegueiro or the initial and final sounds of juez.

• As a group, fricatives, stops and affricates are called obstruents. • Approach occurs when one articulator approaches another, but the vocal tract does not narrow tightly enough to produce the audible friction of a consonant. Some examples are the initial sounds of west [wεst], yes [jεs], rest [rεst], lest [lεst].

What do you think? checked

• "Liquid" is a generic term for the sounds [r] and [l]. • Consonant sounds can be described as a combination of articulatory features: vocalization, place of articulation, and mode of articulation. For example: [t] is a voiceless alveolar stop; [v] is a voiced labiodental fricative.

• Vowels are created by positioning the tongue and mouth to create passages of different shapes.

• The airflow for oral vowels is through the mouth; With nasal vowels, the airflow goes through the nose and mouth.

• Vowels are described by relative pitch and frontality. For example: [] is a front low vowel; [u] is a high back vowel.

• Secondary characteristics of vowel production are sometimes specified, such as B. Stress, nasality, elongation, or rounding, as in "long vowel" or "nasal vowel".

• In many languages, vowels (and nasals) can be pronounced at different pitches or pitches. • Languages differ in the number of speech sounds they have. • Although linguists find it useful to conceptualize speech sounds as distinct and distinct from each other, actual speech sounds are actually related and overlap.

WHAT DO YOU THINK ? REVISED • How many vowels? An easy way to find out what vowels there are in English is to take a simple word table like b_t and see how many different vowels you can set up to produce another word: bit, beet, bet, bait, bat, but, boot , boot, bought, bite, attack Still other vowels don't appear in this frame, but they do in a frame like p_t: put and pot. There are already thirteen vowels, far more than the five that the suggest vowels of the alphabet. • Shaw's ghoti. Words in which ⬍gh⬎ is pronounced as [f] include cough, hard and rough. A word that starts with ⬍gh⬎ is spirit, but the pronunciation is not like [f]. No English word that starts with ⬍gh⬎ (there are only a few, like ghetto, Ghanaian, pickle and ghee) is pronounced like [f]. Shaw exaggerated. • Sounds and letters. ⬍gh⬎ for the initial ghost sound; ⬍th⬎ initial sound in fine and ending on the way; ⬍th⬎ initial sound in soon and final sound in soft; ⬍ph⬎ initial in physics or philosophy; ⬍sh⬎ initial sound when shooting and final sound when wishing; ⬍pn⬎ initial sound of pneumonia; ⬍ps⬎ initial sound in psalm; ⬍ch⬎ initial sound in Cheese and chorus; Some of these sounds can be written with one letter. • Place and place. English words that are pronounced and spelled differently but whose spelling difference doesn't match the pronunciation difference include Satin/Satan; bite / bite; illuminated/light; Woman women.

• 97

9 8 • Chapter 3 The sounds of languages: phonetics

Exercises Practice exercises

The words in each pair below differ from each other by a single consonant, and within each pair the different consonant sounds differ in only one or two properties (voice, mode of articulation, or place of articulation). For each pair, give the IPA symbols for the contrasting consonant sounds, and then identify the properties that distinguish the consonants. Examples: i) fat/roaster: f/v (voice); ii) vat/that: v/ð (place of articulation) iii) wren/red: n/d (form of articulation) sin/sing

the/ether

well/drill

arrive/appear

dig/gather

thief / fief

they give

Boss/Beams

Dec/o

step / young

Was in

tion/room

run away / idiot

Escalation/Crime

many / cents

sadie/gloomy

Based on English 3-1.

3-2.

Consult the tables on page 95 or overleaf and give a phonetic description of the following sounds. In the case of consonants, indicate the pitch as well as the place and type of articulation. For vowels, provide pitch, an anterior/posterior dimension, and (if necessary) a tense/lax distinction. Examples: [s]: voiced alveolar fricative; [i] – high vowel of previous tense consonants:

[z] [t] [b] [n] [ŋ] [r] [j] [ʃ] [θ] [ð]

Vocal:

[e] [] [ɔ] [] [υ] [o] [ə] [ɑ] [e] [aj]

A minimal pair is a set of two words that have the same phones in the same order, except that one phone differs: pit [pt] / bit [bt]; bell [bεl] / count [bl]; and [iðər] / ether [iθər]. a. For each of the following pairs of English consonants, give minimal pairs that illustrate their occurrence in initial, middle, and final position. (Examples are given for the first pair.)

[s] [z] [k] [b] [t] [b] [s] [t] [r] [l] [m] [n]

Initial

Middle

Finale

process/zoo

Bus/Buzz

peace/peas

exercises b. For each of these vowel pairs, name a minimal word pair that illustrates the contrast. Example: [u] [] boot/bat. [i] []; [ɔj] [aj]; [u][u]; [] [and] 3-3.

3-4.

In common notation, write the words shown in the transcriptions below. Examples: [pεn] spring; [to smoke] to smoke; [bənnə] banana [lŋgwəd]

[Trick]

[spawnt]

[θBravery]

[Thursday]

[ðs]

[lvd]

[Toque]

[kwkli]

[Sinn]

[Brake]

[εntərprajzŋ]

The following names are phonetic transcriptions of well-known film names. Write their names in standard English spelling. Example: for bɾi you would write "buddy"; para evitar"; para la revista "8 Millas". Kəri trnzemərəka btmn btmn bignz prajd en pratudas brokbk mawntən gıd najt enjt ŋjinn stinnnninnnvinnjin truθ whni ð ðjin stinnnnnnnnnnnnnnnninnv lknen lknin lknen lknen lajn lk lk lk lkne lajn lajn lajin lajin lajn tʃɔklet Factory / tʃɑklet covered with dead crabs mnz tʃεst ða kraɑnekaelz av nɑrnie ða line ða wtʃ ða Wardrobe

3-5.

The following transcript represents a person reading a passage about actor Will Smith (adapted from Newsweek, July 7, 1997). The transcription doesn't show extraneous features like vowel length or consonant struts, and you'll quickly notice that capital letters and punctuation aren't shown either. Write the passage in the common English spelling as indicated on the first and last lines. wl smθ hz ə daɑrk feɾəl flɔ ts ən əbsεʃən əv sɔrts ðə kajnd əv θŋ ðt kən drajv lvd wυnz krezi ən majt ivən frecht ək

Will Smith has a dark and fatal flaw. It's a kind of obsession, something that can drive loved ones crazy.

• 99

1 0 0 • Chapter 3 The sounds of languages: phonetics

prɑməsŋ kerir hi hets bd grmər prənnsieʃən ererz msteks av εni lŋgw響stek sɔrt dé mek hm nts hz gérlfrεnd dhi kträs dede ppket noz ekeneli tə kɔʃən hm Abāwt ğ sriəsnes av hz aflkʃnn sɾŋ dawn ov εr brεkfst Ancho mɔrnŋ n ðεr spnʃ stajel vl ə ð ð ʃ ʃ ʃ Stajel v Spn ksajd pipəl se ðe wärd ɔfən lajk ɔf fən van TS rili pranawnst ɔf tan smθ lıkŋ sp ɑrɾi ên prɑppntpnts gelz afr ə ə ə ə ə ə ə ə ə ə aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa , é o Capitão Correção. 3-6.

The following transcription represents a person reading a passage about love potions (adapted from The Encyclopedia of Things That Never Existed, p. 159). Write the passage in common English spelling. z ðə ni ndekets As the name suggests, ðiz poʃənz ɑrkɑmpawndəd, these potions are composed spəsfəkli, specifically your etrkt ə sbdəkt, to attract a subject that is z rilktənt tə sarəndər, meaning reaction. . . . . . . . . Tə Karnəl Dezazerz ðe poʃən me bi hd ɾə for fram εni lkəm ɔRsan skd ðə prεpəreʃən av Majn tʃendz wtʃəz wtʃəz en sɔrsararz hu ɑr dénrali nɑt ntrasted ɑr ɑr SMTAJMZ əNW TE MNJAFKʃəR ðE POʃəNZ ðE Pərtʃəzerz Onli Prɑblem Me Bi ðt Av PərsWeɾ ðI ɑbdεkt Av H1r Each Dezar Te Swɑlo and Poʃən y Risenpi for ŋkluɾəd

Exercises dnder snamen pulled and ground grape seeds ojstər elk nter and say their frame mel nämal and any suɾable deviation from ðe person stʃ z hz or their nail klpŋz 3-7

3-8.

Write each of the following words as you say them in everyday language. (Don't let the spelling fool you; it might help if someone else pronounces it.) Examples: bed [bεd]; stale [rnsəd]; shnook [ʃnυk] changes

mainly

a lot of

Freund

Professor

Semantics

System

right away

Musician

peanut

Palma

Music

Photographer

lung infection

attitude

Salmo

flammable

Photo

Examine the following list of consonants found in four popular dictionaries (the first three are American and British COD) and compare the dictionary symbols to the IPA symbols. The abbreviation MWCD stands for Merriam-Webster's Collegiate Dictionary, Eleventh Edition; WNWCD for Webster's New World College Dictionary, fourth edition; AHD for The American Heritage Dictionary of the English Language, Third Edition; COD for The Concise Oxford Dictionary of Current English, Tenth Edition. API icon

MWCD

WNWCD

AHD

COD

Page

Sch

tʃ

Norte

von

Norte

Some symbols used by dictionaries are IPA symbols, but not all. North American dictionaries tend to prefer their own symbols, while the British dictionary leans heavily on the IPA. Choose three sounds for which at least one dictionary uses a symbol other than the IPA symbol and discuss why they were chosen. 3-9

Examine the following list of vowels as represented in three dictionaries; Compare dictionary symbols with IPA symbols. (See Exercise 3-8 to identify dictionaries.)

• 101

1 0 2 • Chapter 3 The sounds of languages: phonetics

API-Symbol

Words

MWCD

WNWCD

AHD

COD

peat feet



well, little



mascot bet

wait late

and

e

bed, bat

and

Gramm

and

Soda, Article

and

Putten, Amor

and

pelvis, boots

she

tu-ooo

she

push, put

she

)

she

boat sold

Largo

she

Puerto, o

o¯ o·

umbrella

jug, carafe

and

Cow, pout

au·

buy, fixed



again

boy, trouble

Me: Hello

Hallo

ɔ

Unlike their practice with consonants, the tabular dictionaries differ among themselves and from the IPA in the transcription of vowels. Name three instances of transcription differences in this book and discuss the advantages and disadvantages of the dictionary representation compared to ours. 3-10 George Bernard Shaw's wry claim that English spelling is so messy that ghoti [fʃ] could be pronounced "fish" has been called misleading. This judgment is based on observations such as these: ⬍gh⬎ can appear initially in a few words (e.g. ghost and ghastly), and then it is always pronounced [g]; simply after a vowel in the same syllable (as in tos and duro) ⬍gh⬎ can be pronounced as [f]; therefore ghoti could not be pronounced with an initial [f]. What other generalizations about the English spelling patterns of ⬍gh⬎, ⬍o⬎, and ⬍ti⬎ could be used to argue that Shaw's claim is at least an exaggeration?

Specially for educators and future teachers 3-11. Your ESL class is complaining that the English spelling is messy. It would be easier to read, they say, if the spelling reflected the pronunciation. As examples they give that electricity should be written ⬍elektrisity⬎ or ⬍alektriatee⬎ and electric ⬍elektrikal⬎ or ⬍alektrakal⬎; likewise, cats should be written ⬍kats⬎ and dogs ⬍dogz⬎. To what extent might your students' statement be correct? On the other hand, what arguments could you put forward to support the view that reading is easier when there is little or no variation in the spelling of the ELECTRIC morpheme and the "PLURAL" morpheme, even if the pronunciation is different? In other words, what are the good arguments for keeping the traditional spelling in these cases?

Suggestions for further reading 3-12. Continuing the discussion about the orthographic consistency of the same morpheme in different words, you will find that native speakers from different regions have different pronunciations of the vowels. Some have the same pronunciation for talk and touch (and walk and wok), while others don't pronounce these pairs the same. What would your students' spelling reformers suggest to compensate for these differences in pronunciation in different speaker groups? 3-13 Your ESL course notices that you pronounce words like later, fatter, and metal as if they were spelled with ⬍d⬎ instead of ⬍t⬎; They pronounce them as Dame, Chief and Order. They ask you why you don't pronounce them with the [t] sound of the spelling. What's your explanation?

Additional resources International Phonetic Association: http://www.arts.gla.ac.uk/IPA/ipa.html

Find the latest version of the IPA, including vowels, consonants, diacritics, suprasegmentals, tones, and word accents. You will also find links to websites where you can download IPA fonts for your word processors and information on how to record IPA sounds. Data from UCLA Phonetics Laboratory: http://hctv.humnet.ucla.edu/departments/linguistics/VowelsandConsonants/

Here you will find recordings of all the speech sounds you want to hear. The sounds of the IPA: http://www.phon.ucl.ac.uk/home/wells/cassette.htm

A cassette and a CD with the sounds of the International Phonetic Alphabet are available. For ordering information please use the link on the IPA home page or go directly to this website.

Suggestions for further reading • Michael Ashby and John Maidment. 2005. An Introduction to Phonetics (Cambridge: Cambridge University Press). A basic treatment with separate chapters on voice, point of articulation, mode of articulation, suprasegmentals, speaker and listener and a good introduction to instrumental phonetics.

• David Crystal. 2003. Dictionary of Linguistics and Phonetics, 5th ed. (Oxford: Blackwell). A rich source of information on the meaning of terms.

• Peter B Denes and Elliot N Pinson. 1993. The Language Chain, 2nd ed. (NewYork: Freeman). An accessible account of the physics and biology of spoken language; contains chapters on acoustic phonetics, digital processing of speech sounds, speech synthesis and automatic speech recognition.

• Peter Ladefoged. 2006. A Course in Phonetics, 5th ed. (Boston: Wadsworth). An excellent introduction to the mechanisms of speech production and the variety of sounds in languages.

• 103

1 0 4 • Chapter 3 The sounds of languages: phonetics

• Peter Ladefoged and Ian Maddieson. 1996. The Sounds of the Languages of the World (Malden, MA: Blackwell). An advanced treatment of the articulatory and acoustic phonetics of the various sounds in the languages of the world.

• Ian RA MacKay. 1991. Phonetics: The Science of Speech Production, 2nd ed. (Boston: Allyn and Speck). The most comprehensive elementary treatment of all aspects of phonetics; accessible and with excellent illustrations.

• Ian Maddieson. 1984. Sound Standards (Cambridge: Cambridge University Press). An inventory of sounds in a representative sample of the world's languages; Inventories range from a minimum of 11 to a maximum of 141 sounds.

• Geoffrey K Pullum and William A Ladusaw. 1996. Guide to Phonetic Symbols, 2nd ed. (Chicago: University of Chicago Press). Explains the various symbols used in the International Phonetic Alphabet (IPA) and by other authors in their treatment of phonetics and phonology; Presented like a dictionary with each symbol clearly presented.

• Michael Stubbs. 1980. Language and Literacy: The Sociolinguistics of Reading and Writing (London: Routledge). Contains an excellent discussion of the relationship between sounds and spelling in English and other languages; offers an insight into questions of spelling reform.

4 Speech Tone Systems: Phonology WHAT DO YOU DO? • You are visiting Paris with a cousin who has learned French and is proud to speak the language fluently. He is talented enough to fluently direct a taxi driver to take you back to your hotel and you are amazed at the ease with which he and the driver speak French. Then, as you get out of the cab, the driver asks your cousin if he's Canadian or American. Dejectedly, his cousin later asks him what features of his pronunciation identified him as an English speaker and why he could not remove them. What you say? • At work you get a message that says "Call Jules Biker" but you don't know anyone by that name. You say it out loud, trying to remember, and you realize it must be referring to your friend Jewel Spiker. When you tell Jewel about the misspelling, she can't figure out what explains the secretary's perception of p as b. Can you explain that? • A tech-savvy friend claims that machines are so good at synthesizing speech that you can't tell if it's a real person or not. You are skeptical and determined to grapple with the issue. After half an hour on the Internet, what can you tell your friend about the quality of speech synthesis?

105

1 0 6 • Chapter 4 Speech-Tonal Systems: Phonology

Introduction: Sounds in the head This chapter focuses on the systematic structuring of sounds in languages. It examines which phonetic distinctions are significant enough to signal differences in meaning; the relationship between the pronunciation of sounds and their storage in the mind; and the way sounds are arranged in words. It is useful to look at sound systems from the perspective of children acquiring their native language. Imagine the task of a child listening to what their parents, siblings, and others are saying. From the flood of idioms at the beginning of life, a child has to crack the code of his language and learn to speak his mother tongue. Caregivers in some cultures use slow, careful language sometimes referred to as "baby talk" when addressing children, but this is not done consistently and not all cultures follow this practice. To make matters worse, the utterances heard by children are often incomplete, interrupted or missing in some other way. In Chapter 3, you learned to distinguish between the number of letters in a written word and the number of sounds in the word's pronunciation. We assume that words have a certain number of sounds. But children in the first months of life do not seem to immediately grasp this simple fact. Trying to count the number of words in a few seconds of a conversation or radio broadcast in a language unfamiliar to you will quickly show how difficult this task is, since most of the words usually run together. This applies to all spoken languages and all dialects. If the words were printed without spaces, they would be very difficult to read.

As you can see, separating the individual words would not be easy. However, this is part of the challenge you faced when you started learning your first or main language. In fact, the task is even more difficult than the juxtaposition of words in the printed sentence above suggests: while the letters in the sentence are discrete and separate from each other, the individual sounds of a spoken word mingle in a continuous stream. from noise. . To take our handwriting analogy a step further, imagine trying to locate the beginning and ending points of each letter in a handwritten sample: this would better capture the challenge babies face when it comes to it to decipher the code of various sounds in your language. Consider the following:

In cursive writing, the letters of each word are joined together. Although anyone who knows English and can decipher that letter can count the letters in each word, there is no clear division in their visual representation. Each word is written continuously; The letters are mixed. Cannot recognize the beginning or the end (except the beginning of the first letter and the end of the last letter of each word). The same goes for the language babies hear; there is no separation between the individual sounds in a word, and there is no beginning or end of the individual sounds in the speech stream. And for children the situation is even more difficult, because the words themselves are not separated. However, children learn the words of their language quickly and efficiently, a remarkable feat considering how much they still have to learn in their early years. If you examine a physical "image" of a word captured from a sound spectrogram, you can see that there is no separation between the sounds. An important reason for continuity between sound segments is that the phonetic features of a sound, e.g. B. vocalization and nasalization, do not start and end at the same time. When you say fluff, don't say [l], stop and say [] and when you're done say [n] and then [t]. Instead of,

Introduction: Sounds in the head The individual characteristics of a sound can be continued in the next sound, and the characteristics of a sound can be anticipated by the utterance of a previous sound. For example, the nasalization of [n] in fuzz occurs in the vowel that is partially nasalized. Similarly, the voice characteristic of a given tone can be removed in anticipation of an upcoming unvoiced tone: when you say imp, the final [m] is unvoiced in anticipation of the next unvoiced tone [p]. Figure 4-1 shows a spectrogram that illustrates how the statement Weren't you here yesterday? appears acoustically. There is no separation between sounds within words or between one word and another. Likewise, the acoustic signal picked up by a child's ears is continuous, and part of the child's job is to separate words within sentences and sounds within words. FIGURE 4-1 Speech Spectrogram: Weren't You Here Yesterday?

What does a child need to understand in order to know a word of their language? Well, to know a word is to know its meaning and its sound. Children go through different stages in learning words, and there is some disagreement about how well they do this task. Some children seem to take the phrases and clauses as whole units and then break them down into parts (this is called the gestalt approach). Others are more analytical early on, taking words directly and building sentences and clauses around them as needed. Eventually, all children classify expressions into different units of meaning. Focusing on one key ingredient, we can ask what kind of information a child should learn about the sounds in a word. What does it take to recognize a word? On the one hand, children need to recognize the pronunciation of a word given by different people as the same word. In order to understand speech, it is important to ignore certain characteristics of the voice and peculiarities in volume, speed and pitch. A child must notice the sounds in a word and the order in which they occur; For example, this mala contains the three sounds [b], [] and [d] in that exact order. After all, bad and dab sound the same, but they are not the same word.

Phonemes and Allophones Finally, every child learns that sounds are pronounced differently in different contexts; that is, the "same sound" can have more than one pronunciation. Note that English speakers are generally unaware that the words cop [kɑp] and keep [kip] start with different [k] sounds. Alternatively, you can tell the difference

• 107

1 0 8 • Chapter 4 Speech-Tonal Systems: Phonology

Say the two words. Notice where your tongue touches the roof of your mouth at the beginning of each word, and you'll see that it touches the velum further back on cop than on keep. Here's why: [ɑ] is a back vowel and [i] is a front vowel, and before the back vowel is pronounced in cop, the [k] is pronounced further back in the mouth than the [k] before the front vowel [yo ] hold.

Try it yourself: place your tongue as if to say wait. Once your tongue is in position for the initial sustain sound, say cop instead. You'll find that you have to reposition your tongue to do this; Saying police from the guard position will make it sound weird or weird. The need to reposition the back of the tongue to achieve natural pronunciation shows that the k-like sounds of keep and cop are not identical. As a second example, you should be able to identify the differences in the sounds represented by ⬍p⬎ in Pot and Spot or Poke and Speak. Bring the back of your hand to your mouth as you say these word pairs, and you'll notice a noticeable difference in the breath of air that accompanies the sounds represented by ⬍p⬎. The ⬍p⬎ sound in Pot and Poke is a sucked stop, represented as [ph]. The following sound ⬍s⬎ in the spoken place is not aspirated; we represent it as [p].

Try it yourself: the draw that accompanies the [p] sound on Pot and Poke is powerful enough to put out a lit match in front of your mouth. Be careful! The sound [p] after [s] at its place and radius is not sucked in and does not extinguish the match. You can try saying dot, dot, dot followed by a single jump. If everything is in the right position, the match will burn with the dot, but the frying pan will put it out. In discussing these two p-sounds, we notice that they occur in different places within words. Examine the following list of words to identify the positions where unaspirated [p] and aspirated [ph] appear. pill poker card spill sprint pimple

[phl] [phokər] [phlet] [spl] [sprnt] [espajn]

Note in our list that aspirated [ph] only appears at the beginning of words (pill, poker, plate) and non-aspirated [p] only after [s] (spill, sprint, spine). If it has two phones and neither can occur where the other occurs in a word, we say they are in complementary distribution. In the words listed above, the aspirated [ph] appears only at the beginning and the non-aspirated [p] appears only after [s]. By definition, therefore, [ph] and [p] occur in a complementary distribution; they cannot appear in the same position in a word and therefore cannot distinguish one word from another. We conclude that [ph] and [p] are not distinct sounds in English words. Rather, they form a unit of the English tonal system; they are called allophones of a single phoneme; in this case /p/ allophones of phonemes.

Introduction: Sounds in the head A phoneme is a structural element in the sound system of a language. Allophones are realizations of a single structural element in the phonetic system of a language. The allophones of a given phoneme cannot form distinct words, so we say they are non-contrastive. For native speakers, the allophones are perceived as the same sound despite the physical difference. Since English aspirated [ph] and unaspirated [p] are allophones of the phoneme /p/, there could be no English word pair like [phit] and [pit]. Likewise, the two [k] sounds of cop and keep are allophones of the phoneme /k/ in English and cannot form contrasting words. Note: We now start using forward slashes / / to enclose phonemes and square brackets [ ] to enclose allophones. We will continue this practice, although sometimes we have to choose one representation or the other when one of the two suffices. (The angle brackets ⬍ ⬎ contain letters.) In addition to the aspirated and non-aspirated allophones of /p/, English has a third voiceless bilabial stop that can appear in a word like mop. This third allophone can be at the end of a word if that word is at the end of a sentence: Where's the mop? In this position, the lips can remain closed so that the sound ⬍p⬎ is not made. We represent this allophone as [p]. Now, here's a little complication. In this case we don't have a complementary distribution since both unaspirated [p] and unaspirated [p] can appear in a word at the same position (word-finally). If two sounds can appear in the same position in a word, but do not contrast, ie without producing different words, one speaks of a free variation of these sounds. At the end of a phrase, native English speakers may pronounce lip as [lp] or [lp]. Both unvoiced and unvoiced bilabial stops are allophones of /p/, and thus /p/ has three allophones: aspirated [ph], non-aspirated [p], and unvoiced [p]. To reiterate: the allophones of a phoneme occur in a complementary distribution or in free variation; in no case can a change of meaning be indicated by the different allophones.

Distribution of Allophones It can be useful to think of a phoneme as an abstract element in a language's phonetic system: a basic unit of sound that does not lack a very specific pronunciation, but is pronounced in a specific way depending on where it is in a word occurs . . For example, while the phoneme /p/ would have the skeletal properties of a voiceless bilabial plosive, one allophone could be sucked in, another could not be sucked in, and a third could not be released. The pronunciation of the phoneme /p/ cannot be fully given unless its position in a word (or utterance) is known. Only then can your aspiration and liberation be determined. We have just seen that certain allophones are determined by their place of occurrence. If you look at the word groups in Table 4-1, you'll see that the picture is a bit more complex: TABLE 4-1 Allophones of /p/ in English A-words

[ph]

personal family tree track pilgrimage

parental petunia quirky peninsula

emporio compute rapidograph competitive

[p] Computational Rapid Empathy Competence

• 109

1 1 0 • Chapter 4 Sound Systems of Language: Phonology

stacked In these words, the stress (´) indicates the syllable on which the primary stress falls (as in ridiculous versus ridiculous). All words have /p/ in the syllable-initial position. Words in column A have the main stress on the first syllable, with [ph] as the initial sound. Those in column B also initially sucked in the word [ph], although the main stress is on the second syllable. The words in columns A and B show that /p/ is an originally aspirated word, regardless of whether it occurs in a stressed or unstressed syllable. In column C, aspirated [ph] introduces the second syllable, each of which carries the main stress. Thus the aspirated [ph] not only appears initially in the word, but also internally when it introduces a stressed syllable. The words in column D show that unaspirated [p] occurs internally in the word when unstressed syllables are introduced. In summary, the phoneme /p/ is an initially aspirated word in stressed and unstressed syllables, but is an internally aspirated word only when a stressed syllable begins. Given these observations, we can describe the distribution of /p/ allophones more precisely by considering the stress patterns in a word. We do this in Table 4-2.

TABLE 4-2 Two allophones of English /p/ PHONEME

ALOFONOS

DISTRIBUTION

[ph]

In syllable-initial position, in stressed syllable and in word-initial position

[when]

Elsewhere (as in a consonant cluster after /s/ and word-final position)

/when/

Compare these facts about aspirated [ph] and non-aspirated [p] with the facts about /p/ and /s/. If a child had intended to say the word sat said pat, they would not have noticed one of the significant differences in English pronunciation. This is because /p/ and /s/ are separate phonemes and can distinguish words, as in the following word pairs: A

[pt] chest [st] sitting

[pn] set [sn] call

[lpt] previous [lst] last

[sip] reflux [sis] stops

The words in each pair have different meanings and differ in only one sound. Two words that differ by only one sound form a minimal pair. (Note that the distinction depends on phones, not spelling.) Minimal pairs are valuable for identifying the contrasting sounds, the phonemes, of a language. Each minimal pair above shows that /s/ and /p/ are separate English phonemes and not allophones of a single phoneme. The articulatory descriptions of /s/ and /p/ show that they differ in both location and mode of articulation. VOICE ARTICULATION PLACE ARTICULATION FORM

/s/ voiceless alveolar fricative

/p/ southern bilabial explosive

Introduction: sounds in the head To give another example: /s/ and /b/ differ not only in the place and articulation of articulation, but also in the sonority. VOICE ARTICULATION PLACE ARTICULATION FORM

/s/ voiceless alveolar fricative

/b/ voiced bilabial stop

/s/ is a voiceless alveolar fricative, /b/ a voiced bilabial stop. The fact that /s/ and /b/ contrast (as in the minimal pair sat/bat) shows that they are distinctly different sounds. They belong to separate phonemes. Phonetics (allophones) belonging to a single phoneme share certain phonetic features but differ in at least one other feature. This may be related to voicing (voiced vs. unvoiced), aspiration (aspirated vs. non-aspirated), type of articulation (e.g., stop vs. fricative), or site of articulation (e.g., dental vs. alveolar). When analyzing the phonetic system of a language, it is important to pay special attention to phonetic distributions that have similar phonetic descriptions. Consider this list of English words: [pht] [bt] [thp] [thb] [spt]

knock bat knock tab tab cuspir

[ph] and [b] are bilabial stops; share these two characteristics. On the other hand, [ph] is voiceless and aspirated while [b] is voiced and non-aspirated. Are they separate phonemes or allophones of the same phoneme? We cannot answer this question just by looking at the phonetic descriptions. But since there is a minimal pair above, we know that [ph] and [b] are contrasting. That is, pat/bat shows that [ph] and [b] belong to different phonemes. On the other hand, since there are no examples in English where aspirated [ph] and unaspirated [p] contrast (in fact, they occur in a complementary distribution), [ph] and [p] are allophones of a single phoneme. . From the minimal tap/tab pair above (two words that differ by only one phone) we know that unaspirated [p] is opposed to [b]; so aspirated [ph] and non-aspirated [p], both voiceless bilabial stops, contrast with the voiced bilabial stop [b]. The phonemes /p/ and /b/, as we have seen, contrast in word-initial and word-final positions. But sometimes two tones contrast in some positions but not in all positions. Two sounds are different if they contrast in any position. Consider the position after /s/ as in the word s__at: no two English words like sbat and spit have different meanings. Although /p/ and /b/ are different phonemes in English, the difference between them in position after /s/ is not examined. Now consider the situation in Korean. Like English, Korean has three bilabial stops [ph], [p], and [b], as in the following words: [phul] [pul] [pəp] [mubəp]

'grass' 'fire' 'law' 'lawlessness'

The minimal pair [phul] and [pul] shows that [p] and [ph] are contrasting in Korean. On the other hand, no minimal pair could be found even with a large selection of words

• 111

1 1 2 • Chapter 4 Sound Systems of Language: Phonology

that [p] is in contrast to [b]. This is because [p] and [b] are in complementary distribution in Korean: [b] occurs only between vowels and other sound segments, as in [mubəp], and [p] never occurs in this environment. This shows that [p] and [ph] are separate phonemes in Korean, but that [p] and [b] are allophones of a single phoneme. The diagram in Table 4-3 shows the difference in the phonological systems of English and Korean with respect to these three sounds. The same three sounds occur in both languages, but their systematic relationships in these languages are different. In English, [ph] and [p] are non-contrasting allophones of a single phoneme and therefore cannot signal a difference in meaning. In Korean, [ph] and [p] are contrastively distinct sounds; that is, they are separate phonemes and can distinguish one word from another (as in [phul] and [pul]). We can say that vocalization in English is phonemic (the voiced bilabial stop [b] differs from the unvoiced bilabial stop [p]). However, striving is not phonemic in English (no two English phonemes differ only in striving). In Korean, on the other hand, the voiced bilabial stop [b] is an allophone of /p/: [b] occurs predictably between voiced sounds; therefore [b] and [p] cannot distinguish Korean words.

TABLE 4-3 Comparison of three sounds in English and Korean ENGLISH PHONES

SOUNDS: ENGLISH AND KOREAN

KOREAN PHONEME

[ph]

/ph/

/p/ [p] /p/ /b/

[b]

In summary: in Korean, voice is not contrastive, but striving is; In English, aspiration is not contrastive, but vocalization is.

Phonological Rules and Their Structure You probably know that French has nasal vowels, and you may think that English doesn't. In fact, however, English has nasal vowels. In Table 4-4, the words in column B have nasalized vowels (vowels pronounced through both the nose and mouth), while those in column A have oral vowels (vowels pronounced through the mouth). When you say the words in column B, the air escapes from your lungs through your nostrils; Therefore, when this passage is blocked, the vowel tone changes noticeably.

Try it yourself: Hold your nose as you say the words in each column of Table 4-4. You will notice that there is no noticeable difference for the words in column A; for those in column B, it will make an amazing difference.

Phonological rules and their structure.

TABLE 4-4 Oral and nasal vowels in English A

sit pet light brutto babysitter

Sin Feather Lemon Broom Singer

If you look up nasal vowels in English words, you'll find that they all come before one of the nasal consonants /m n ŋ/. This is another way of saying that the distribution of nasal vowels in English is regular and predictable: a vowel is nasalized before a nasal consonant. Since the distribution in English is predictable, the occurrence of nasal vowels cannot indicate a difference in meaning. (In contrast, in French and other languages where their distribution is unpredictable, nasalization can signal a difference in meaning.) Two sounds whose distribution is predictable relative to each other form allophones of a single phoneme; its distribution can be described by a general rule. Phonological rules have this general form: A → B / C___D

You can read a rule like this as "A becomes B in the area after C and before D" or more simply as "A becomes B after C and before D". A, B, C, and D are generally specified in terms of phonological features, although the rules are presented more informally in this book. In cases where C and D do not need to be specified, one of them is missing. For example, the phonological process of nasalization in English can be represented by the following statement: vowel nasalization rule → nasal / ___ nasal (vowels are nasalized if they precede nasal sounds).

We have already said that in learning a word the child must learn the number of phonemes in the word, what those phonemes are and in what order they occur. As the English cop/keep change for the /k/ allophones shows, and as the poke/spoke change for the /p/ allophones shows, a child must also learn to identify specific allophones of a phoneme depending on the phoneme's position pronounce a word and the character of nearby sounds. This is not done by learning all the sounds in each word as it is pronounced, but by learning what phonemes the words contain and adopting the rules that apply universally, as in the nasalization rule above. The situation of a child acquiring [p] and [b] in Korean is similar to that of an English-speaking child acquiring nasal vowels. Since [p] and [b] never contrast in Korean, they are allophones of a single phoneme in Korean, and only one form is needed to represent them in the lexicon. Of course, Koreans also need to be aware of the phonological rule that establishes the distribution of allophones: [b] between vowels and [p] elsewhere. The alternative of a single lexical representation for [p] and [b] in Korean would require a specific distinction between these sounds in each word.

• 113

1 1 4 • Chapter 4 Speech-Tonal Systems: Phonology

that contains one of them. For example, pep 'law' and mubep 'lawlessness' would have different specifications for [p] and [b]. For English speakers (who do not have a predictable distribution of [p] and [b] since they are separate phonemes), this distinction seems natural and necessary. But having different forms for [p] and [b] in a Korean speaker's lexicon would be tantamount to an English speaker having different representations in the lexicon for the different /k/ sounds of cop and keep, for the different sounds /p/ from would have poke and talk, and through the various sounds /i/ of sit and watch. Instead, each phoneme in a language is represented in the lexicon by a single underlying form. Native speakers internalize the phonological rules that specify the distribution of allophones and automatically apply the rules whenever the phoneme occurs.

Generalization of Phonological Rules So far we have considered phonological rules as if they were formulated to apply to particular sounds; In fact, they are more general. Consider the striving that accompanies the production of the initial /p/ in English words like Pillow and Poke: 1. For /p/: unvoiced bilabial stop

→ aspirated / word first and first in stressed syllables

This rule states that in certain situations a voiceless bilabial plosive is sucked.

Why English Speakers Speak French With a Foreign Accent With a foreign accent, you can infer that the differences between allophones are not stored in the mental lexicon but are governed by regular rules. Imagine a native English speaker who doesn't speak French who was introduced to a neighbor named Pierre by a French-speaking neighbor. English speakers aim for unvoiced primary stops such as /p/, but French speakers do not. Thus the francophonist who introduces Pierre will pronounce his name without aspiration. Although the English speaker hears no aspiration in the pronunciation of Pierre, they tend to pronounce Pierre with aspirate [ph], according to the phonological rules of English (but not French). This indicates that English speakers have a rule of aiming for the initial /p/ (even when pronouncing French nouns). The unconscious application of the phonological rules of your native language to a foreign language is a major factor that contributes to a foreign accent and marks you out as a non-native speaker. On the other hand, if you speak a foreign language, you may not make the necessary distinction. English usually distinguishes between the k-sounds in cop and keep. English speakers do not have to separately learn to copy and maintain which k sound to use. But in some languages, including Basque, Malay, and Vietnamese, these two sounds are opposed; these are different phonemes. The initial sound of cop is represented by [k] and of keep by [c]. In languages like Basque, Malay, and Vietnamese, knowing which velar stop occurs in a word is important, just as English speakers need to know whether /p/ or /t/ occurs, since the [k] and [c] sounds in these languages are not distributed by rule and are not predictable.

Phonological rules and their structure If you examine other English words with stop consonants, you will find that /t/ and /k/ also have an aspiration when they are syllable-initial. Since /p t k/ has parallel distributions of aspirated allophones, English seems to need two additional rules like rule 1: rule 2 for /t/ and rule 3 for /k/. 2. To /t/: voiceless alveolar stop 3. To /k/: voiceless velar stop

→ aspirated / word first and first in stressed syllables

Since these three rules exhaust the list of unvoiced stops in English, they can be combined into a single, more general rule covering /p/, /t/ and /k/ as follows: 4. For /p t k/: unvoiced → aspirated /word initially and initially in stopped stressed syllables

Notice in Rule 4 that the combination of the features "unvoiced" and "stop" leaves the point of articulation indefinite. Hence a phonological rule like 4 applies to all unvoiced registers, regardless of their place of articulation. This is especially true for unvoiced bilabial, alveolar, and velar stops, for /p/, /t/, and /k/. The more general a rule is, the easier it is to state it in phonetic slash notation. There is evidence that internalized phonological rules are not specified in terms of allophones such as [p] and [ph], or in terms of phonemes such as /p/, /t/, and /k/, but rather in terms of specific sound classes , through function sets like "voiceless" and "stop".

Natural phone classes A set of phonemes such as /p t k/ that can be described with fewer resources than would be required to individually describe one of its member phones is called a natural phone class. A natural class contains all (and only) sounds that share a certain set of properties. For example, /p t k/ form the natural class of "voiced stops" in English. /p t k/ has the two features "voiceless" and "stop", and there are no other sounds in English that have these two features. Now consider the set /p t k b d g/. This is the natural class of saves. There are no other stops in English, and all sounds in the ensemble share the "stop" characteristic. The sounds /p t k b d/ do not form a natural class. These five sounds share the "stop" function, but also /g/, which is not included. Any feature we use to describe the set /p t k b d/ would also describe /g/. Likewise, /p t km/ does not represent a natural class, since any feature introduced to specify /m/ would also characterize other phones. Adding the "nasal" feature to the description to include /m/ would mean including /n/ and /ŋ/ since they are also nasal. Note, however, that to specify the quantity /p t km n ŋ/ we need an "either/or" description: "voiceless plosive" or "nasal". Since no combination of functions specifies only these six phones, /pt km n ŋ/ is not a natural class.

• 115

1 1 6 • Chapter 4 Speech-Tonal Systems: Phonology

Underlying forms Thanks to internalized rules that provide the correct allophones for the phonemes of a given word, children can eventually create entries in their lexicons like those in Table 4-5. Such shapes are called underlay shapes or underlay representations, and we represent them between slanted lines, using the same notation that we use for phonemes. The surface form that characterizes the pronunciation of a word results from the application of phonological rules to the underlying forms. In some examples in Table 4-5, the surface form is the same as the underlying form because there are no applicable phonological rules that we discuss in this chapter, and accordingly we indicate "none" in the rule column. TABLE 4-5 Basic and surface forms of six words in English BASE FORM

RULE

SURFACE FORM

WRITTEN

/kler/ /bυk/ /bit/ /p/ /spn/ /pn/

Aspiration None None None Nasalization Aspiration/Nasalization

[khlər] [bυk] [bit] [p] [sp)n] [ph)n]

Coloring book beating rotating pen

Order of Rules An additional phonological rule will illustrate a point about the organization of phonological rules in internalized grammar. Consider the following words: a

AND

write pure string lop lock tap pick up

riding need tunic praise log tab pig

Treat Foul Root Trench wrote punch boots

Delay forest path yard rude cloud booed

If you listen carefully while saying these words out loud, you might notice that the vowels in column B are longer than the corresponding words in column A. In phonetic symbols, we represent long vowels with a colon after them, as in [ a ]. Since English does not have a minimal pair like [pit]/[pit] or [bt]/[bt], vowel length cannot be contrasted in English. Instead, it's predictable; Vowel length can be specified by a phonological rule. If you look beyond the spelling, you'll notice that all of the words in column A end in an unvoiced consonant /p t k/ and all of the words in column B end in a voiced consonant /b d g/. English lengthens vowels when they precede voiced consonants. If we use V to represent the vowel and C to represent the consonant, we can formulate the rule as follows:

Phonological rules and their structure Lengthening rule V → V / ___ C voiced (vowels are lengthened before voiced consonants).

As a result of this rule, the following processes occur in English: ε → ε

/___ /d/ (like in bed vs. bet)

the → the

/___ /g/ (like accent vs. bankruptcy)

also → also

/___ /d/ (wie in Slide vs. Lightweight)

(Note that this rule applies to diphthongs like /aj/.) Vowel length in English is predictable and can be specified by a rule; It is not necessary to learn the length of the vowels for each word individually. In some other languages, vowel length is unpredictable and must be learned word by word. For example, Fijian has a minimal pair oya meaning "he, she" and oyaa meaning "that (thing)". Dredre means "to laugh"; dreedree means "difficult". Vakariri means "to cook"; vakaririi means "fast". Therefore, Fijian vowel length cannot be assigned by a phonological rule. It is contrastive, distinctive, phonetic in this language. Now consider the following pairs of words and note how the pronunciation of each word in column A differs from the pronunciation of the corresponding word in column B. You will notice that the difference in pronunciation is not represented by the spelling ⬍ t ⬎ and ⬍d⬎ . Rather, the difference lies in the length of the vowel. In most dialects of American English, the first vowel of each word in column B is longer than that of the corresponding word in column A. A.

Writer Liter Seat Rooter

ruder seed drill operator pilot

The intermediate consonants ⬍t⬎ and ⬍d⬎ do not represent distinct pronunciations, as Americans tend to clash (or "touch") /t/ and /d/ between the vowels of these words. When pronouncing /t/ or /d/ in words like the above, the tip of the tongue quickly hits the ridge of the jaw. Since the key allophones of /t/ and /d/ are identical (IPA [ɾ]), the difference in pronunciation that may have resulted from the t/d distinction is lost or neutralized. The distinction is not lost with the words write [rajt] and hide [rajd]. Again using V to represent any vowel, the American English flutter rule is: Alveolar stop flutter rule

→

V___V that's all

(/t/, /d/ are perceived as [ɾ] between two vowels, the second of which is unstressed.)

Although the flutter rule neutralizes the t/d distinction in this setting, many Americans pronounce words in column B differently than in column A, and the reason is interesting. By combining the flutter rule with the lengthening rule, American English language pronounce words in column B with a longer vowel, although there is no difference in the pronunciation of the middle consonant. Here's an explanation of how this works.

• 117

1 1 8 • Chapter 4 Sound Systems of Language: Phonology

We specify two rules in English that can work on the same words. Let's explore how they interact to create a pronounceable surface shape. Consider the pair of words writer and gentleman. Suppose the underlying forms in the lexicon are /rajtər/ for writer and /rajdər/ for gentleman. We can represent the derivation of surface shapes as in Table 4-6. (If the shape of a word doesn't meet the requirements of a rule, the rule doesn't apply, so we write DNA.) Taking the underlying shapes and applying the two rules in the order shown (stretch first, then touch) results in the surface shapes [ rajɾər] and [rajɾər]. This is the correct pronunciation of these words for some speakers. Let's call them A dialect speakers.

TABLE 4-6 Derivation of Writer and Rider in A WRITER dialect

CABALLERO

underlying shape

/Writer/

/rajdar/

(Verboten)

stretching rule

ADN

derived form

[instead of]

apply ↓ [rajdər]

(Input-Output)

batting rule

applies ↓ [rajɾər]

(Salida)

surface shape

If we apply the same rules in reverse order (hit first, draw later), the results will be different. Here's why: since the slap sound is voiced, the preceding vowel is lengthened in both words. As Table 4-7 shows, this is exactly what happens with speakers of another variant of English. Call it Dialect B. The two identical surface forms [rajɾər] and [rajɾər] resulting from applying the beat rule before the lengthening rule would not be correct for dialect A. In dialect A

TABLE 4-7 Derivation of scribe and rider in dialect B WRITER

CABALLERO

underlying shape

/Writer/

/rajdar/

(Verboten)

batting rule

applies ↓ [rajɾər]

(Input-Output)

applies ↓ [rajɾər]

(Salida)

Derived form Strain rule Surface form

The syllables and the structure of the syllables (the most common dialect), the scribe and the lord are not pronounced the same. Instead, reiter has a longer vowel than scribe. Even starting with the same underlying forms and the same pair of phonological rules, application in sequence produces correct surface forms for a given dialect; but when applied in the other order, the rules produce incorrect forms. Evidence like this led some researchers to hypothesize that the ordering of rules is part of the organization of phonological rules. Note that the forms resulting from the second derivative (Table 4-7) are incorrect in dialect A, but correct in dialect B. This shows how speakers of different dialects can share underlying shapes and rules, but produce different surface shapes as a result. to rearrange the rules. Dialects with an elongation before the beat have different forms of scribe and lord. Dialects with slaps before the lengthening produce identical forms with a long vowel.

Syllables and Syllable Structure So far we have said little about how sounds are organized in words (although, as you will see, our analyzes assumed some organization). It may seem obvious that sounds in words appear as an abcdef sequence, but that's not the whole story. Phonetics are organized into syllables and syllables are organized into words. Every word consists of one or more syllables, and every syllable consists of one or more sounds. The term "syllable" is not difficult to intuitively grasp, and there is considerable agreement on the number of syllables. But technical definitions have proven difficult. Nevertheless, there is agreement that a syllable is a phonological unit made up of one or more sounds, and that syllables can be divided into two parts: a rhyme and an onset. The rhyme consists of a head and the following consonants. The head is usually a vowel, although certain consonants called sonorants can also function as a head. Sonorants include nasals such as [m] and [n] and liquids such as [r] and [l]. Consider the words button, butter, and bottle, whose second syllable we represent as [əC], which contains a vowel and a consonant. The same words could be represented as [bt%], [bɾ&], and [bɑɾ(), with the [] diacritic under the voiced [n r l] indicating that it is the stem of a syllable. the syllable is the beginning. Any consonant that follows the head as part of the rhyme is called a coda. The graphic below shows the structure of a syllable as just described. beginning of syllable

núcleo de rima

Koda

The only essential element of a syllable is the nucleus. Not all syllables have a beginning and not all rhymes have a coda. This means that a single sound can form a syllable. Because a single syllable can form a word, a word can only have one vowel, but you already knew that about the words a and i. Table 4-8 lists some English words with one, two, three, and four syllables.

• 119

1 2 0 • Chapter 4 Speech-Tonal Systems: Phonology

TABLE 4-8 English words broken down into 1 SYLLABLE syllables

2 SYLLABLES

3 SYLLABLES

4 SYLLABLES

tons [thn]

even

most beautiful [lv-li-əst]

irgendein [ε-ni-bɑ-ɾi]

turn [turn]

gone [all-ðo]

any [ε-ni-wən]

bzw. [re-spεk-tav-li]

by [θru]

consistent [kən-ssts]

Computer [Bass-phju-ɾər]

algebraically

navigate

writer

Syllable [s-la-bel]

Definition [dε-fə-n-ʃən]

Sequence Constraints The possible phonetic sequences in a syllable differ from language to language and are limited within each language. If you examine the following sentence, you will see that English syllables allow for different patterns of consonants (C) and vowels (V). (We use hyphens to separate syllables within a word.)

a previous title

n ə pri vi äs kpʃən VC V CCV-CV-VC CVC-CVC

You can see that English allows several types of syllables: VC, V, CCV, CV and CVC. Some other allowed types can be seen in monosyllabic words like these: past tense

/PST/

CVCC

reina

/quin/

CCVC

transformed

/transformed/

CVCCC

Jets

/jet/

CCCVCCC

Within a polysyllabic English word (e.g. at the end of a syllable other than the end of the word), not as many types of coda are allowed. Not all languages allow such a wide variety of syllable types. The preferred syllable type among world languages is a single consonant followed by a single vowel: CV. Other common types are CVC and a single V. (All three types appear in the example sentence above.) Polynesian languages such as Samoan, Tahitian, and Hawaiian have only CV and V syllables. Japanese also allows syllables of mainly CV and V forms, as well as CVC (but only if the second C is nasal). Korean allows the syllables V, CV, and CVC. Mandarin Chinese allows syllables of the forms V, CV, and (when the second consonant is [n] or [ŋ]) CVC. It is not common in world languages to have initial consonant clusters – CC – as in the English words try, twin and stop, and it is highly unusual to have initial consonant clusters of more than two consonants – CCC – as in yell, run and stress. Even English has a limited range of consonants that can occur as C1 and C2 of a two-consonant (C1C2) initiation cluster, and an extremely narrow range of consonants in each of the C1C2C3 positions of a three-consonant initiation cluster. (It is no coincidence that all

Syllables and syllable structure (three illustrations of the initial CCC begin with /s/). Similarly, the initial clusters of three consonants in English have different restrictions than the clusters forming the coda.

Try it yourself: name three English words that have initial clusters of three consonants. What sound do they all start with? What tones are produced as C2? And the C3? Can you imagine startup clusters that have a different C2 or a different C3? What they are? The rules that characterize the syllable structures allowed in a language are called sequence constraints (or phonotactic constraints) and determine what constitutes a possible syllable. As a result of such limitations, there are - in addition to the words that exist in a language - thousands that could not exist, and thousands upon thousands that could not exist because their syllable structures are not legal consonant sequences and vowels in that language. . The following words would be impossible in Hawaiian and Japanese because they violate the sequence restrictions of those languages: pat (CVC), ruffle (CCVC), and spa (CCV).

Sniglets Comedian Rich Hall has compiled lists of "sniglets" in English, words that don't appear in the dictionary but should. Here are some of these sniglets and their suggested definitions. charp 'the mutant green potato chip found in every bag' elbonics 'the actions of two people heading towards an armrest in a movie theater' glarpo 'the junction of ear and skull where pencils are kept' hozone 'the place where where a sock belongs disappears on all clothing to "splash" the metal barrier on a rotary phone that prevents you from dialing past O.

Sniglets conform to English string restrictions.

Try it yourself: Since the following forms violate English string restrictions, they cannot serve as sniglets: ptlin, brkow, tsmtot, ngang. Add three more to the list. When learning a foreign language whose syllable structure differs from that of the native language, speakers tend to impose sequence restrictions of their native language on foreign words. For example, neither Spanish nor Persian allow initial collations like /st/ and /sp/, so speakers of those languages can pronounce English learning and language words like /εs-tdi/ and /εs-pitʃ/, pronunciations that match the limitations of Spanish and Persian chains. Similarly, Japanese speakers have borrowed the words baseball and strike as beesubooru and sutoraiku, forms obeying Japanese sequencing restrictions.

• 121

1 2 2 • Chapter 4 Sound Systems of Language: Phonology

Stress A common aphorism among American linguists is, "Not all white houses are white houses, and not all blackbirds are blackbirds." The point is that stress patterns can be significant. When pronouncing the phrase each white house, there is a relatively heavy emphasis on both white and house: white house. When it comes to the US President's office, White is given a relatively heavy accent, but House has only a secondary accent: White House. The accent pattern associated with the name of the presidential residence corresponds to that of the word téachèr: white house. The tonic pattern of the same words in the phrase (all) white houses does not. From the fact that stress can vary and the meanings of the two terms are different, it follows that stress in English can be contrastive. Below is a list of several other word pairs. The pairs in column A are distinct words: they form noun phrases comprising an adjective and a noun (and an article); the stress patterns of the pairs in column B follow the pattern of teachèr—they are compound nouns. A

the bulletin board the blue bear the beautiful chair the red neck the jet especially the iced tea the yellow coat (clothes)

a panel a choke an éigchàir a rédnèck a jétstrèam an ícè crèam a yellow jacket (a kind of wasp)

Try it yourself: In the columns above, identify three additional phrases and pairs of compound nouns, such as blue bird and blúebìrd.

English has a variable accent, not a fixed one. The same applies to some other languages, including German. Many others have fixed stress, where stress is assigned to a specific syllable in words. In Polish and Swahili words, the stress is usually on the penultimate syllable (called penultimate syllable). Czech words have an accent on the first syllable. French words usually have a stress on the last syllable.

Syllables and Stress in Phonological Processes We have seen above that certain phonological rules for their formulation depend on the syllable, on the stress, or on both the stress and the syllable. Aspiration of English voiceless stops /p t k/ occurs "word initial and initial in stressed syllables" (page 115). Such a formulation assumes that words are organized into syllables. This in turn means that children need to have some understanding of how words are organized into syllables. The flutter rule that produces [rajɾər] for Writer and [mεɾəl] for Metal is also accent dependent, and by now you can probably imagine that the flutter rule could be formulated in terms of syllables rather than vowel segments, which is the same as we formulated on page 117. Current

Phonology-morphology interaction word models use multiple layers to accommodate phonologically significant levels, including segments, syllables, and stress.

The Interplay of Morphology and Phonology Before leaving the subject of phonology, let's examine the pronunciation of the most productive inflectional suffixes in English and see the surprising regularity in the patterns.

Plural, Possessive, and Third-Person Singular Morphemes in English Regular nouns have different pronunciations of plural morphemes, such as lips [lp ⫹ s], seeds [sid ⫹ z], and fusions [fjuz ⫹ əz]. The surface forms of these different pronunciations of a morpheme are called allomorphs. As the following lists show, the allomorphs of the plural morpheme are determined by the final sound character of the singular form. Allomorphs of English "plural" morpheme [əz] bushes judges peaches buses fuses

[s] cat guide smells births

[z] pen seeds dogs cars lightning

These lists show the distribution pattern of plural allomorphs in English. 1. [əz] occurs in nouns ending in /s z ʃ t兰 d/ (a natural class called sibilants). 2. [s] occurs after all other unvoiced sounds. 3. [z] occurs after all other beeps.

You can think of arguments to posit any of the three allomorphs as the underlying abstract form of the plural morpheme. We'll assume it's /z/. From this underlying form, the three allomorphs must be derived according to general rules that apply to all regular nouns. From an underlying /z/, a rule like the following would derive the allomorphic [əz] that follows sibilants; Note that ⫹ marks a morpheme boundary and # a word boundary. Schwa Insertion Rule A /z/ → [əz] / sibilant ⫹ ___# (Schwa is inserted before a word-ending /z/ after a sibilant-ending morpheme).

To derive the allomorph [s] from the underlying morpheme /z/ that follows the unvoiced sounds would require a rule that partially matches the voiced /z/ to the unvoiced sound of the stem morpheme. Assimilation rule A /z/ → voiceless / unvoiced ⫹ ___# (The last word /z/ is voiceless after a morpheme ending in an unvoiced sound).

• 123

1 2 4 • Chapter 4 Sound Systems of Language: Phonology

In order to derive the correct forms of all regular plural nouns, these two rules must have considerable generality. Table 4.9 illustrates this for the nouns cooperative, judge, and weed. (DNA means that a rule does not apply because a necessary condition is missing; slanting lines // represent underlying forms; square brackets [] represent forms derived from the application of a phonological rule.)

TABLE 4-9 Derivation of plural nouns in English COOPS

PARTIES

weed

underlying shapes

/kup⫹z/

/pis⫹z/

/wid⫹z/

Schwa insert

ADN

derived form

[kup⫹z]

apply ↓ [pee⫹əz]

[wid⫹z]

Assimilation

apply ↓ [kup⫹s]

ADN

[pis⫹əz]

[wid⫹z]

surface shape

You may be surprised to learn that our rules for deriving plural forms from regular nouns in English are widely used. Two other extremely common English inflectional morphemes, namely the possessive marker on nouns (judge, cat, and dog) and the third-person singular marker on verbs (shows, ri, and nothing), have the distribution of their allomorphs the same as that of the distribution. for plural. Possessive morpheme in nouns [s] [z] [əz]

a: boat, cat, cat. . . a: John, arm, dog. . . for: church, judge, fish. . .

Third-person singular morpheme in verbs [s] [z] [əz]

for: jumping, eating, kicking, laughing. . . a: rush, seem, stoop, wish, see. . . to: preach, joke, judge, hum, rush. . .

If we posit /z/ as the underlying phonological form of these morphemes, the same rules that derive the correct allomorphs of the plural morpheme will also derive the correct allomorphs of the possessive morpheme of nouns and the third-person singular morpheme of verbs. (Unlike plural forms, some of which are irregular, all nouns have allomorphic regular possessive morphemes, and all verbs are regular with respect to the third-person singular morpheme except is, has, say, and does.)

English Past Tense Morpheme The inflectional morpheme that marks the past tense of regular verbs in English has three allomorphs:

The interplay of morphology and phonology [t] [d] [əd]

to: wishing, kissing, talking, undressing, preaching. . . to: greet, bathe, play, lie down, fidget, tease, wander, wreck. . . from: want, wade, wait, boo, sow, sow. . .

If we posit /d/ as the underlying phonological form of the past tense morpheme, we only need two simple rules to derive the past tense in all regular verbs. Schwa's insertion rule B /d/ → [əd]/ alveolar stop ⫹ ___# (Schwa is inserted before a word-ending /d/ after a morpheme ending in an alveolar stop.)

Assimilation rule B /d/ → unvoiced / unvoiced ⫹ ___# (The word ending /d/ is perceived as [t] after a morpheme ending in a voiced sound).

The past tense derivations of the verbs wishes, will, and wink are given as examples in Table 4-10. TABLE 4-10 Derivation of past tense verbs in WAVED

DESIRE

MY LOVE

underlying shape

/wev⫹d/

/wʃ⫹d/

/wɑnt⫹d/

Schwa insert

ADN

derived form

[wev⫹d]

[wʃ⫹d]

apply ↓ [wɑnt⫹əd]

Assimilation

ADN

surface shape

[wev⫹d]

apply ↓ [wʃ⫹t]

ADN [wɑnt⫹əd]

The last two sets of rules show striking similarities in the schwa insertion processes and the assimilation processes necessary to produce the correct forms of the plural and possessive forms of nouns, the third-person singular verb forms, and the past tense forms of verbs. (In some derivative-independent theories of phonology like ours, the schwa insertion would derive from a “compelling contour principle”. OCP forbids neighboring phones to be in some respects identical. Scope of this chapter.)

Basic phonological form of morphemes in the lexicon This section examines the phonological form of words as accounted for in speakers' mental lexicons. The form of a word in the lexicon is called its underlying form, and the form in the lexicon may not be the same as the pronounced form. Consonants The same types of phonological processes that occur between a stem and an inflectional suffix also operate between a stem and a derived morpheme (e.g.

• 125

1 2 6 • Chapter 4 Speech-Tonal Systems: Phonology

between DESIRE and FULL). Think of a child who knows the words metal and medal. In American English, the sound that occurs in the middle of both words is the alveolar lobe, not [t] or [d] but [ɾ]. (An alveolar lobe is the sound made when the tip of the tongue quickly smacks against the alveolar ridge: then ladder.) As an American kid hearing metal and medal, you would have put exactly what you heard in your lexicon: /mεɾəl / (with flap) in both cases. But consider what happens when someone says their new car is painted metallic red [məthl ⫹ ək]. Once you realize that metallic is composed of METALL and the derivational suffix -IC (as in atomic, Germanic), the two pronunciations [mεɾəl] and [məthl ⫹ ək] need to be reconciled. The task of a language student is to posit an efficient base form that produces the correct pronunciations when applying the phonological rules of English. Then think about your homework when you hear someone report that their car's medallion is missing from the hood. For medal and medallion one hears [mεɾəl] and [mədljən]. Once it is recognized that the MEDAL morpheme occurs in both words, what underlying form should it postulate in the lexicon? Suppose you recognize METAL as the common element in Metal and Metallic and MEDAL as the common element in Medal and Medallion. Here are the pronunciations that caught your eye: METAL [mεɾəl] metal

MEDAL [means ⫹ ek] metallic

[mεɾəl] medal

[madl ⫹ gen] medallion

You could account for the different pronunciations of the METAL morpheme by positing the form /mεtl/ in the lexicon and applying phonological rules that transform this underlying form into the surface forms that occur. If we focus on the consonants and ignore the vowels for the moment, the underlying form /mεtl/ requires a process that converts /t/ to [ɾ] in the metal word [mεɾəl]. (You'll see why we use . below.) The same operation is required to convert /d/ to [ɾ] in the word medal [mεɾəl]. The flutter rule changes the underlying /t/ and /d/ to [ɾ] when they occur between a stressed and an unstressed vowel. Using a more formal notation, the rule would read: Alveolar Stop Flutter Rule

→

accented singing

——

unbetonter Gesang

Not surprisingly, phonological rules that are postulated to account for a set of facts can also account for other facts. Finally, phonological rules apply to all morphemes and words unless blocked by a specific marker for a particular morpheme. For example, nouns such as tooth and foot, which have irregular plural forms, are marked in the lexicon as not having the regular plural morpheme. If they weren't irregularly marked, one would say teeth and feet, as children do before they learn to exempt these morphemes from these regular processes. Thus, the relationship between the phonological representation of morphemes in the lexicon and their effective linguistic pronunciation is mediated by a set of phonological processes that can be represented in rules of significant generality. Not only metal and medal are affected by the flutter rule, but all words that meet the conditions specified in the rule. These include single morpheme words such as butter, bitter, and meter; words with two morphemes such as Writer, Knight, Raider, Rooter; and thousands more.

The interplay of morphology and phonology

• 127

Vowels Imagine a young man who knows the words photography and photographer (pronounced [foɾəgrf] and [fəthɑgrəfər]). At some point the young man postulates the only entry in the lexicon, FOTOGRAFIE, to represent the core of these two words. A quick reflection suggests that if the underlying form was /fotɑgrf/, that would represent the basic knowledge needed to produce both pronunciations. Given the underlying rendition /fotɑgrf/ and the surface forms [foɾəgrf] and [fəthɑgrəfər], a rule that changes unstressed vowels to [ə] produces the correct vowels. When /ə/ appears in the underlying form, there is no rule that produces the correct surface forms. To generate [ɑ] in [fəthɑgrəfər] from an underlying form with schwas /fətəgrəf ⫹ ər/, we would need a rule that generates [ɑ] from the underlying /ə/. For the word photography, we would need a rule that would generate [o] from underlying /ə/ in the first syllable and [] from underlying /ə/ in the third syllable. This would amount to knowing what vowels exist in the surface pronunciation and encoding that knowledge into the underlying form along with /ə/, but exactly what we assume isn't happening. If we postulate different vowels in the underlying forms instead, a single rule can derive [ə] from any underlying vowel when it occurs in an unstressed syllable. Now we can deduce the pronunciations of these words. We formulate the rule as follows: unstressed vowel

→ [and]

(An unstressed vowel becomes weak.)

This rule does not affect stressed vowels; says that unstressed vowels become schwa [ə]. Of course, a rule based on information about accents requires prior assignment of accents, but the rules for assigning accents in English are beyond our scope. If you wish to continue the subject, some references at the end of this chapter contain treatments of stress rules.

Computers and phonology A few decades ago, researchers thought it would be years before computers could recognize and synthesize speech. (Think of speech recognition as converting speech to printed speech, and speech synthesis as converting printed speech to speech.) The process takes longer than expected, and the reasons are not due to a lack of sophistication in computers or technology. Although children master the phonology of their languages from an early age, adult researchers have yet to discover the extraordinary complexity of the phonological processes that characterize human languages. We have not yet modeled with sufficient accuracy what we do when we produce spoken utterances and how we understand the utterances of others. Natural language occurs in a continuous flow and cannot easily be segmented.

Mind without knowledge of the language concerned. It is not yet well understood how humans break down a continuous stream of spoken language into individual words and recognize the sound segments in those words. The machine's speech synthesis also proved challenging. To understand why, focus on the sequence of sounds that would occur in a simple word like arena. It may seem relatively easy to assemble a machine-generated form of /sænd/: the machine would only have to generate a voiceless alveolar fricative, then the vowel /æ/, then the alveolar nasal /n/, and finally the alveolar stop /d/. Easy enough! Note, however, that the quality of your vowel when pronouncing Sand is quite different from the "same" vowel in Hat. If a speech synthesizer created the vowel in the word arena,

1 2 8 • Chapter 4 Speech-Tonal Systems: Phonology

Computer science and phonology would sound artificial. The vowel sand is nasalized (because it appears before the nasal stop /n/, as described on p. 113). When the vocal tract moves toward this nasal consonant, the preceding vowel takes on nasal characteristics. Therein lies a challenge for speech synthesis: how to combine sounds the way humans do. In ordinary human language there is no division between words, no division between sounds. But the situation is even more complex. A sound is essentially a set of phonetic features. Think of the phonological form of Arena as having not only four /s æ n d/ segments, but also the features given under each segment:

/s/

/æ/

/norte/

/d/

voiceless alveolar fricative

Bass soundfront not rounded

nasal alveolar sonora

healthy alveolar arrest

The phonetic properties of the segments are more complicated than indicated, but the illustration above serves our purposes. Note that the articulation of each phonetic feature in a segment does not begin and end at the same time as the other features. The voicing of /s/ does not end abruptly and the voicing of /æ/ begins exactly in the same millisecond that the fricative character of the consonant ends and the vowel character of /æ/ begins. The mouth and other features of the vocal tract are in constant motion in the production of even the simplest of words. For artificial speech to be natural, much more needs to be understood about the nature of the phonetic realizations of the underlying phonological forms. ■

Summary • Phonology is the study of the sound systems of languages. • A phoneme is a unit in the phonetic system of a language. It is an abstract element, a set of phonological features (e.g. bilabial, stop) with various predictable manifestations (called allophones) in language. • Two words can be slightly different because they have only one distinct pair of phonemes (such as pin/bin or tap/tab). • Each phoneme consists of a set of allophones. Each allophone is the specific rule-driven and therefore predictable realization of the phoneme in a given linguistic environment. • The allophones of a phoneme occur in a complementary distribution or in free variation; they never contrast. The allophones of a single phoneme cannot be the only difference in a minimal pair of words with different meanings. • Two languages can have the same sounds but structure them differently within their systems. Both Korean and English have the three sounds [p], [ph] and [b]. In English, non-aspirated [p] and aspirated [ph] are allophones of one phoneme, while [b] belongs to another phoneme. In Korean, aspirated [ph] and non-aspirated [p] are separate (contrasting) phonemes, while [b] is the allophone of the phoneme /p/ that occurs between voiced sounds. • Every single word in a speaker's lexicon consists of a sequence of phonemes that form the underlying phonological representation of the word. The underlying forms differ from the pronunciations and are usually not directly observable in the language.

What do you think? Revised • Starting from the underlying form of a word, the phonological rules of a language specify the allophonic properties of a phoneme according to its linguistic environment. • A task for children in language acquisition is to discover their phonological rules and to develop efficient and economical basic forms for word units. Given these underlying forms, the phonological rules of a language specify the rule-driven features of the surface form. • Phonological rules can be arranged in relation to each other, with the first applicable rule being applied to the underlying form to produce a derived form, and subsequent rules being successively applied to successive derived forms until the last applicable rule produces a surface form . The shape of the surface is the basis for the pronunciation of a word. Two dialects of a language may contain some of the same rules but apply them in a different order, producing different surface shapes for different pronunciations. • Words are made up of groups of sounds called syllables, not the sounds themselves. • Languages have sequence restrictions on the structure of permissible syllable types and on the occurrence of certain consonants and vowels within syllable types. • Curriculum vitae is the most common syllabic type in the world's languages. English has an extraordinarily wide range of syllable types, including two- and three-consonant clusters. The specific consonants that can appear in each position are restricted. • The accent is contrastive in English: “Not all white houses are the White House”. • Phonological processes (eg, aspiration and flutter in English) may depend on syllable structure and stress, as well as a sequence of sound segments.

WHAT DO YOU THINK ? R E V I S I T E D • French taxi driver. Part of the answer is probably that your cousin is using the English allophone distribution instead of the French one. English tends to aim for the initial /p/, French doesn't. There are other processes that are subtle and operate below the level of consciousness, including vowel and consonant length rules, word stress patterns, and sentence intonation patterns. • Julio Biker. One explanation is that the /p/ in Jewel Spiker's surname is neutralized in the environment after the /s/. In other words, since English in this setting doesn't rely on the /p/-/b/ contrast, the sounds can, and do, be pronounced more similarly. Although the initial /p/ is aspirated in English, the initial /b/ is not aspirated. In the environment after /s/, as in Spiker, /p/ is not aspirated, and the lack of aspiration contributes to the easy confusion of /p/ and /b/ at this point. • Technical friend. On the Internet, one quickly discovers that while speech can easily be synthesized tone by tone, it is not so easy to create a natural connection between sounds in a word, or between words, or to create natural-sounding intonation patterns.

• 129

1 3 0 • Chapter 4 Sound Systems of Language: Phonology

Exercises Practice exercises

A. The following words are given in standard spelling and a phonetic transcription. Examine the allomorphic variation in the pronunciation of the underlined morpheme and give an underlying form from which allomorphic variants can be derived by rule. Pay attention only to the pronunciation presented in the transcription and ignore the spelling. (Nasalization is ignored for this exercise.) Example: Metal [mεɾəl]

metallic [mətlək] underlying form: /mεtl/

Mensch

[Mensch]

mankind [hjumnəɾi]

Mut

[Loan]

brave

Industry

[ndəstri]

industrial

medicine

[your belly]

medicinal [mədsənəl]

B. Transcribe these monosyllabic words and underline the rhyme in your transcription; Then identify the beginning, head, and (if any) coda of each syllable. Example: remainder: [rεst] r, ε, st

erobert [klnʃt] kl, , nʃt

sit down

exhausted

Jets

sin

pumpkin

kneaded

Aroma

I'm sorry

gefüttert

C. The rule on page 117 states that vowels are lengthened before voiced consonants in English. But you may have noticed that all of the examples in columns A and B on this page have consonants that lengthen before they stop. Neither example shows lengthening before fricatives or affricates, which are also consonants. Is the rule given too broad, or can you give pairs like those in columns A and B where the lengthening occurs before a voiced fricative or affricate but not before its unvoiced counterpart? Based on what you find, reevaluate the generality of the stretching rule as indicated.

Based on English 4-1.

4-2.

Consider the following English words for how the sound represented by ⬍t⬎ is pronounced. For each column, indicate the phonetic character of the allophone (how it is pronounced). is sucked off? fluttered? Then, as in this chapter for the /p/ allophones in English, describe the /t/ allophones and specify their distribution. A

more resilient

Stand

later

Petunias

talkative

still

given

Potato

ATM

Story

petal

Returns

Using the monosyllabic words below, give a list of 15 ordered pairs whose stress pattern indicates that they are compounds; H. with emphasis as in the examples. This

The exercises will be useful to mark the accent pattern in the vowels of each element, using ´ for the primary accent and ` for the secondary accent. Examples: tímezòne, showhòrse

4-3.

she was

haz

Quadrant

Cara

cair

for free

gear

thousand

Sombrero

Herz

keep

Horse

Largo

treated

lanza

Tierra

color

life

clear

elegant

porta

reins

walk from

estrada

exhibition

Style

Tabla

beat

the sea

Tempo

principal

Behavior

Weight

again

Zone

It seems that the following words do not exist in English. Some are "sniglet" candidates (they may exist), but others violate English string restrictions and may not exist. Identify the possible sniglets and explain why the others are not allowed. For potential sniglets, provide an appropriate case in the standard case. ptribɑr ppkəss pŋgəkd

4-4.

twntʃ blbjulə sprang

Rezenant lktomŋgjuleʃεn spwənt

a. Make a list of as many words as possible, each representing a different beginning of three consonants. Example: jump in Spread b. Examine the initial clusters you listed in (a) and answer these questions about English: Which consonants can appear first in an initial cluster of three consonants? Which consonants could come second in an initial group of three consonants? Which consonants could come third in an initial group of three consonants? Examine your three lists to decide whether or not they are natural species and list all natural species.

4-5.

Although English makes a contrast between /p/ and /b/ (pill versus bill), it does not take advantage of the contrast in the environment following /s/ (as in spell and twist). So there are no word pairs like /sbn/ and /spn/. When a language exploits a distinction in some settings, but not all, the potential contrast tends to be neutralized where it is not exploited. Consequently, the /p/ of pill differs more from the /b/ of bill than the /p/ of spin (try to distinguish "spin" from "sbin"). First, the /p/ in spin (but not the /p/ in pill) lacks aspiration, as does the /b/ in bill. Therefore, a feature that distinguishes /p/ and /b/ elsewhere is not used after /s/. Below are two sets of words. Those in Column I contain a contrast that English exploits in that environment, but not in the Column II environment. That is, for words in Column II there cannot be contrast based on the difference in sound found in the word pair is shown in the same row in column I. i i seat sit

Ich singe King of the Ring

ii. a little bit

here beer lover

iii. hasse Hut

Hang sang played

4. papi

Hacked star studio

in Channels

skill value hit

• 131

1 3 2 • Chapter 4 Speech-Tonal Systems: Phonology

a. Identify the segment that is likely to result in different phonetic transcriptions and indicate what those transcriptions would look like. B. Characterize the environment (in column II) that supports neutralization. C. Give reasons, based on your knowledge of English phonology (e.g. sequencing limitations), for preferring one of the transcriptions to the other. 4-6

On page 122 we said that you can probably imagine that the flutter rule in English could be rephrased in terms of syllables and their constituents, rather than in terms of vowel segments as formulated on page 117. Formulate the flutter rule in relation to syllables and their components.

4-7.

For each English word below, identify the heading for each syllable and (if applicable) the beginning, rhyme, and coda. Example: for past tense, core: a; beginning :p; rhyme: ast; coda: the second twin became the same love-li-est a-any-body de-fi-ni-tion na-sa-li-za-tion

4-8.

a. The Nasalization Rule (page 113) and Assimilation Rules A and B (pages 123 and 125) cause nearby sounds to become more similar. Which feature is transferred from one sound to another in the nasalization rule? Which property is propagated in assimilation rules? B. One way to characterize the insertion rules of schwa A and B (pages 123 and 125) is to say that they produce different neighboring sounds. Another way to characterize schwa insertion is to say that it separates very similar sounds. What properties do the neighboring sounds have in common before inserting schwa A? And the insertion of Schwa B? C. Given the rules mentioned in (a) and (b), we can see that English has some rules that make neighboring phones more similar and other rules that make neighboring phones more different. Examine these rules carefully and find an explanation for these competing tendencies. Tip: Think about how difficult or easy it would be to pronounce these sequences without the rules; Think about how difficult or easy it can be to understand these sequences without the rules.

Based on languages other than English 4-9.

Fijian has pre-asalized stops in its inventory of phonemes. The prenasalized stop [nd] consists of a pronounced nasal immediately before the stop with which it forms a single sound unit. Consider the following Fijian words pronounced in fast speech: vindi kenda tiko tutu viti dovu dondo

'sprout' 'we' 'stay' 'grandfather' 'Fiji' 'cane' 'reach'

dina dalo vundi manda tina mata mokiti vevendu

'true' 'taro' 'banana tree' 'first' 'stem' 'eye' 'round' (a type of plant)

Tasks Based on this data, determine whether [d], [nd], and [t] are allophones of a single phoneme or form two or three separate phonemes. If you find that two of them (or all) are allophones of a single phoneme, give the rule that describes the distribution of each allophone. If you analyze all three as separate phonemes, justify your answer. (Note: In Fiji, all syllables end in a vowel.) 4-10. Examine the following words from Tongan, a Polynesian language. (Note: In Tongan, all syllables end in a vowel.) Tauhi

'import'

Chino

'Body'

Sissi

'Kranz'

inside

'To the right'

Island

'Island'

AFTER

'applaud'

to whisper

'Garroar'

gordo

'being'

Caliente

'green'

movement

'separate'

answeror

'take a rest'

Mission

'Klang'

a. Based on this data, determine whether [s] and [t] are allophones of a single Tongan phoneme or whether they are separate phonemes. If you think they are allophones of the same phoneme, provide the rule that describes where each allophone occurs. If you conclude that they are different phonemes, justify your answer. B. In each of the following Tongan words, one sound has been replaced with a space. This sound is [s] or [t]. With no more knowledge of Tongan than you suggest from the question above, is it possible to guess which of these two sounds fits in the gap? If so, please set the tone; if not, explain why. ___ ill

'fishing net'

Taxa

'bulk goods'

___ Three

'put something on'

my old me

'ambush'

C. Over the past century, Tongan has borrowed many words from English and adapted them to the phonological structure of its words. kaasete

'Gazeta'

suu

'Zapato'

body

'prato'

Show

'Show'

Gesellschaft

'sociedad'

pata

'Butter'

Salat

'Salat'

criticism

'Zucker'

Solo

'Marshal'

a pig

'Cigarette'

seconds

'second'

Diamond

'Diamond'

How does the phonetic status of [s] and [t] differ in loanwords and native Tongan words? In other words, is it the same situation with these loanwords? Write an integrated statement about the status of [s] and [t] in Tongan. (Note: Your statement must include information about which area of Tongan vocabulary each part of the rule applies to.)

• 133

1 3 4 • Chapter 4 Language Systems: Phonology

4-11 The distribution of the sounds [s] and [z] in colloquial Spanish is illustrated by the following examples in phonetic transcription: izla

'Island'

you go

'Joke'

fuersa

'Force'

alt

'Ski'

Fisch

'pez'

Riezgo

'Risk'

muscles

'Muscle'

tax

'tax'

sin

'which one'

rezva

'underpants'

rasko

'I scratch'

dezde

'from ... to'

take a rest

'remaining'

Open

'Feature'

Same

'even'

beizbɔl

'Baseball'

cost

'Cost'

but

'more'

Are [s] and [z] different Spanish phonemes or allophones of the same phoneme? If they are different phonemes, justify your answer. If they are allophones of the same phoneme, indicate their distribution. 4-12 Consider the following Russian words. Based on this limited list, where does Russian seem to have a contrast between [t] and [d] and where not? (Note: an apostrophe marks a palatalized consonant.) pərxot gz'εtə zapət rat zdan'ijə most

'steamboat' 'newspaper' 'west' 'happy' 'build' 'bridge'

pote t'εlə dergoj d'εlə ʃtat pote

'body' 'sweat' 'honey' 'business' 'condition' 'low'

4-13 In Samoan, the words can have two forms, one called "bad speech" (used in informal speech when addressing colleagues or family) and another called "good speech" (used with superiors or strangers in literary and religious situations). The difference between the two forms can be described by phonological rules. (Note: the Samoan words for "good" and "bad" do not have the same meanings as the English words in this case.) "Bad"

"Boa"

'to'

Others

Secondly

'chica'

Cara

your

'Cara'

I am here

ŋaŋa

'Language'

Counting

kokoŋi

'Preis'

meat

file

'Stay'

Gracias

'is working'

Telephone

fo.o

'Meeting'

a. Describe the phonological difference between the "bad" and "good" forms. Which form is more basic, the "good" form or the "bad" form? (In other words, what can serve as the underlying shape for both shapes?)

exercises b. If possible, fill in the blanks in the table below. If it is impossible to know the form of a missing word, explain why. "Poorly"

"Boa"

Manual

_______

'Vogel'

Freund

_______

'he died'

_______

maŋoo

'Two'

_______

Kilo

'fishing net'

right away

_______

'To the right'

_______

your

'mourn'

4-14 In German, the letter sequence ⬍ch⬎ can represent (among others) one of two sounds: [ç] (an unvoiced palatal fricative) or [x] (an unvoiced velar fricative). Using the following data, determine whether these two phones are distinct phonemes or allophones of a single phoneme. kεlç

Taxes

'Glas'

fçtə

Fichte

'Baum'

knchl

ankle

'Junta'

kɔx

Koch

'Oven'

tsurεçt

rightly

'in good condition'

vυxt

force

'Weight'

çrυrk

The surgeon

'The surgeon'

Estrella

sober

'sober'

bux

mesentery

'Book'

Damm

area

'Range'

until

checkmark

'apostrophe'

bax

Bach

Ribeiro

If [ç] and [x] are different phonemes, justify your answer. If they are allophones of the same phoneme, indicate their distribution. 4-15 On page 120, you learned that Japanese character string restrictions allow syllables of the forms CV, V, and (when the second C is nasal) CVC. Using this information, break down the words given in the Japanese vowel table (Table 3–5 on page 94) into syllables: ima "now"; aki 'autumn'; spark plug "safe"; you 'approach'; Sensei "Teacher". Do the same with the loanwords beesubooru "baseball" and sutoraiku "strike", where ⬍ee⬎ and ⬍oo⬎ represent long vowels, not two vowels. 4-16 In light of our discussions in this chapter and your experience with some of the previous exercises, consider the following quotation from Halle and Clements (1983). The perception of understandable language is . . . only partially determined by the physical signal that reaches our ears. Of equal importance. . . It is the contribution made by the perceiver's knowledge of the language in which the utterance is framed. Acts of perception that rely heavily on the active contribution of the observer's mind are often referred to as illusions, and the

• 135

1 3 6 • Chapter 4 Speech-Tonal Systems: Phonology

the perception of intelligible speech shines. . . to qualify for this description. A central problem of phonetics and phonology is . . provide a scientific characterization of this illusion that is at the heart of all human existence.

Specially for educators and future teachers 4-17. As an exercise for a class of international high school students learning English, she asked them to list English names for games and offered them: skokey, skwinty, twint, stwink, plopo, splopt, sprats, skretsht, stretched, skwickt, spwint, eddy , tprash, stpop, frash, quirt, splast, plsats. Realize that some names are not legitimate because they contain phonetic sequences (not letters) that are not allowed in English. Which ones are impossible and what explanation can you give students as to why they are impossible? 4-18 Using phonological terms from this and the previous chapter, identify two distinctive features of the "foreign accent" for students represented in schools in your community. Try to explain the differences between the way native English speakers and non-native speakers pronounce certain accented words. It might be useful to think about (a) the stock of phones, (b) the phonological rules for the distribution of allophones, (c) the constraints on the phonetic order. 4-19 Recall from Chapter 3 (page 87), and perhaps from your own experience, that French speakers tend to pronounce the English word "thin" as "sin" and "thin" as "zis." From this observation, what can you say about (a) French consonant stock compared to English consonant stock; and (b) whether or not French uses vocal training as a contrastive resource? Finally, what would you say about how a French student would tend to pronounce the English words then and roughly? 4-20 Focusing on the high front vowels, carefully compare the Spanish vowel chart (Table 3–3 on page 93) with the English vowel chart (inside cover and on page 90). Based on these tables and your relevant experience with the IPA symbols, identify which different vowel pairs in English you think will present a challenge for learners of Spanish and explain why. Then name two minimal English word pairs (identical words except vowels) that these students may find difficult to understand and produce. 4-21. When you listen to your Spanish-speaking students speaking to each other, you notice with the English loanwords that their Spanish pronunciation is systematically different from their English. So, for example, the word scanner was borrowed as a scanner and slogans as slogans. What do these pronunciations tell us about phonotactic restrictions in at least some Spanish word-initial consonant clusters? 4-22 One of her students, returning from a summer visit to Berlin, Paris and Madrid, says that listening to local radio from those cities, she couldn't separate the flow of speech into single words: everything seemed to blur. She thinks English is different because English words are separated and easy to identify. What would you tell him about the difference you experienced between your ability to hear words in English and your inability to separate words in German, French and Spanish?

Suggestions for further reading

Other web speaking resources: http://www.tue.nl/ipo/hearing/webspeak.htm#Online

If you are interested in hearing synthesized speech, several websites provide examples. This site is a "jump station" that provides links to speech synthesizers from around the world. Once you've picked one, you can write something you want to hear synthesized. Provided your computer has multimedia capabilities, you can experience state-of-the-art text-to-speech synthesis. AT&T Labs text-to-speech site: http://www.research.att.com/~ttsweb/tts/demo.php

This demo demonstrates the features of the AT&T Natural Voices™ speech synthesizer. Enter up to 300 characters and get an audio file of your input, compatible with your computer and playable with its multimedia capabilities. You can choose between several voices. Speech Links: http://www.speech.cs.cmu.edu/comp.speech/SpeechLinks.html

This is a language technology hyperlink page containing hundreds of links to projects around the world. In addition to links to technical articles (most of which are out of reach for beginners), you will find links to websites dedicated to speech recognition and synthesis. Speech Synthesis Systems Museum: http://www.cs.bham.ac.uk/~jpi/museum.html

These links will take you to a variety of speech synthesis systems, some of which allow testing so you can make your own judgments of promise and naturalness. Ladefoged concatenative speech synthesis: http://hctv.humnet.ucla.edu/departments/linguistics/VowelsandConsonants/vowels/chapter8/chapter8.html

This site features synthesized American, English, and Scottish text-to-speech versions of "The north wind and the sun were arguing which is stronger, when a traveler appeared wrapped in a warm cloak. They agreed that whoever got the traveler to remove his cloak first should be considered stronger than the other."

Suggestions for further reading • Carlos Gussenhoven and Haike Jacobs. 2005. Understanding Phonology, 2nd ed. (London: Hodder Arnold). An excellent continuation of this chapter; rich and universally accessible. • Francisco Katamba. 1989. An Introduction to Phonology (New York: St. Martin's). A comprehensive treatment, sensitive to theoretical and descriptive concerns. • April McMahon. 2002. An Introduction to the Phonology of English (Edinburgh: Edinburgh University Press). A very simple and accessible treatment that includes the phonology of words and phrases. • David Odes. 2005. An Introduction to Phonology (Cambridge: Cambridge University Press). A fundamental treatment, with separate chapters on phonetic transcription, allophonic relationships, underlying representations, abstraction, and psychological reality, and a chapter devoted solely to analysis.

• 137

1 3 8 • Chapter 4 Speech-Tonal Systems: Phonology

Advanced Reading Clark and Yallop (1990) and Carr (1993) are basic textbooks that will be widely accessible to readers who have mastered some of the phonetics and phonology in this chapter. Halle and Clements' (1983) 'Problembuch' covers a wide range of languages and has an excellent introductory chapter that goes beyond what we have covered; It also includes separate chapters on complementary distribution, natural types, phonological rules, and systems of rules. Roca and Johnson (1999) is an excellent workbook with dozens of problems in a variety of languages; is intended to accompany the phonology course by the same authors and can also be used independently. Kaye (1989) is a lively, provocative, and accessible follow-up to this chapter. Hogg and McCully (1987), Bybee (2002) and Goldsmith (1996) offer more specialized treatments.

References • Bybee, Joan. 2002. Phonology and Language Use (Cambridge: Cambridge University Press). • Carr, Philip. 1993. Phonology (New York: St. Martin's). • Clark, John and Colin Yallop. 1995. An Introduction to Phonetics and Phonology, 2nd ed. (Malden, MA: Blackwell). • Goldsmith, John A., ed. 1996. The Handbook of Phonological Theory (Malden, MA: Blackwell). • Hall, rich. 1984. Sniglets (New York: Collier). • Halle, Morris and G.N. Clements. 1983. Book of Problems in Phonology (Cambridge, MA: MIT Press). • Hogg, Richard and C.B. McCully. 1987. Meteral Phonology: A Textbook (Cambridge: Cambridge University Press). • Kaye, Jonathan. 1989. Phonology: A Cognitive View (Hillsdale, NJ: Erlbaum). • Rock, Iggy and Wyn Johnson. 1999. A Course in Phonology (Malden, MA: Blackwell). • Rock, Iggy and Wyn Johnson. 1999. A workbook on phonology (Oxford: Blackwell).

The structure and function of phrases and sentences: syntax.

WHAT DO YOU THINK ? • Your colleague Rudy says he finds it easy for us to produce routine phrases like “What time is it?”. and "Okay, thanks" because we hear them so often. But he asks how we produce and understand sentences that we have never heard before. What do you think? • Her friend Amber reports that she thought about ambiguity while reading Steven Pinker's The Language Instinct. She understands why an ambiguous word like bank can mean "savings bank" or "river bank," but wonders what makes a set of unambiguous words like new drug combinations ambiguous. What's your explanation? • One day, with the Los Angeles Times in hand, reader asks Ron if it is grammatical to say "The extra money can be claimed elsewhere, Romer said" or simply "Romer said the additional money can be claimed elsewhere." it is grammatical. You want to know about the order of subjects, verbs and objects in English. what can you tell him • Nerdy Ned is annoyed that his word processor's grammar checker contradicts almost every passive sentence he writes. Instead of The winning team was ruined by a hodgepodge of friends, the tester recommended A hodgepodge of friends ruined by the winning team. Ned says the checker assumes all passives are bad and wonders what you think. What you say?

139

1 4 0 • Chapter 5 The structure and function of phrases and sentences: syntax

Introduction In this chapter we examine how words and morphemes are organized into phrases and sentences. We also examine the relationships between certain types of sentences, such as B. Declaratives and interrogatives. We examine how a finite grammar can generate an infinite number of sentences and how the "creative" aspects of sentence production and the easy understanding of new sentences are normal parts of everyone's competence. All languages have ways of referring to entities: people, places, things, ideas, events, etc. Expressions used to refer to entities are called noun phrases. The proper nouns Peter and Pennsylvania, the generic names cow and calorie, and the personal pronouns He and She are noun phrases. Plus more complicated terms like your mother, that grimoire, the star of the show, and a Michigan judge. All are reference terms; all are noun phrases. Languages also have ways of saying something about the entities they refer to. All languages have ways of making positive and negative statements. They also allow speakers to ask questions, give directions, etc. Let's illustrate with positive statements. The following sentences refer to an entity and then make a statement about it. reference expression

sermon

Judge Judy A poltergeist Julian

a daughter showed up last night. bought an answering machine.

In the first example, "Judge Judy" is referred to and then something is said about her, which is "she has a daughter."

Try it yourself: The second example above references "a poltergeist" and then delivers a sermon. what is the sermon What is the referent in the third example and what predication is made of it? Syntax is the part of grammar that governs the form of the strings of characters that language users use to make statements, ask questions, give instructions, etc. The study of syntax is concerned with the construction of sentences and their structural and functional relationships to one another. What we functionally call reference phrases are grammatically called noun phrases. From a functional point of view, expressions like having a daughter and buying an answering machine are predicates; grammatically they are verb phrases. As different as languages are in other respects, they all have noun phrases as reference expressions and verb phrases as predicates. A simple sentence contains a single verb (or predicate) and any other expressions that the verb requires as part of its structural features. In Chapter 2, we discussed the subcategories of verbs and said that speakers need to know what types of sentence structure each verb allows. We note that the verbs allow different complements: some require a noun phrase; others do not allow. In the following examples, the verb is in italics: Danny fell. Dimas cooked the hot dogs. A runner from Ohio won Sunday's marathon.

Diagrams of trees and constituencies Britney is buying a new raincoat this fall. His uncle had put the presents in the car. The psychiatrist should have listened to his patient.

Each sentence contains only one verb, although the verb can be a single word (fell, cook, won) or more than one word (go shopping, had piled up, should have listened). From a syntactic point of view, the verb is the central element of a sentence. For one, subcategorization determines which plugins you can have. (We will see later in this chapter that a simple sentence can act as part of another sentence, in which case the simple sentence can be called a clause.)

Constituent and Tree Diagrams When analyzing sentences, it is helpful to recognize that sentences are not made up of words strung like beads on a string, but of organized components. Remember the phrase Harry saw a ghost. It is obviously made up of words, and each word contains at least one morpheme. Since these morphemes have associated phones, we can say that the set of phones (like /g/, /o/, /s/, /t/ in phantom form), of morphemes (VER and 'PAST TENSE'), or of words (Harry and saw). Such an analysis is correct, but misses the point. It resembles the description of a shopping mall made of concrete and electric cables. We mean that a shopping center has retail stores, restaurants, parking lots, cinemas, etc. Then we could go further and describe the composition of these units and their relationship to each other. Each analysis is about identifying the structural units that are relevant for a specific purpose or organizational level. When analyzing sentences, these structural units are called constituents.

Tree Diagrams A useful way to represent syntactic relationships is with tree diagrams. The tree in Figure 5-1 represents the fact that the sentence Harry liked Peeves consists of two parts: the Harry-related phrase and the predicate he liked Peeves. In the tree diagram, S means sentence, N means noun or pronoun, and V means verb. The same tree can represent other sentences, as Harry saw in Figure 5.2. (As we shall see later, the trees in Figures 5-1 and 5-2 are simplified.) FIGURE 5-1

FIGURE 5-2S

Norte

torment

valued

hobbies

torment

This

• 141

1 4 2 • Chapter 5 The structure and function of phrases and sentences: syntax

Constituents We can consider sentences to be composed first of their largest constituents. These large units can be parsed into smaller units that can also be parsed. Linear Order of Constituents It is obvious that the words of any sentence occur in a specific order. The components of a sentence therefore necessarily also have ordered elements. Simply put, sentences are expressed as an ordered sequence of words, as in these examples: A chubby plumber from Pasadena was ice skating in the park. Hillary hated the harp. Xavier is from Xanadu.

Now we ask whether the order in which the words are arranged is fixed. If it is fixed, is it equally fixed in different languages? First, consider the following sentences: The farmer saw the poltergeist. The poltergeist saw the farmer.

Both sentences are well-formed and contain exactly the same words, but mean different things. Due to their identical words, the difference in meaning must be signaled by the different word order. Therefore, word order is an essential part of the structure of English sentences. In this case, by the order of the words, we understand who saw and who was seen. Now consider this: The farmer saw the poltergeist. * Farmer the poltergeist the saw.

While the first string is well formed, the second is not. This also shows that word order is an essential part of English sentence structure. When we rearrange words, we sometimes produce another well-formed sentence with a different meaning, but we can produce a bunch of malformed (or "ungrammatical") words. Sometimes changing the word order can produce another well-formed sentence with the same meaning, as in these example pairs: 1a. He saw a poltergeist in the castle yesterday. B. He saw a poltergeist in the castle yesterday. 2. When preparing the sauce, I forgot to add the basil. B. When preparing the sauce, you forgot to add the basil.

the farmer saw

Sombra

BAUER

ELF

The farmer saw a shadow. The farmer saw the shadow.

16263

Therefore, the order of the words is not absolutely fixed. Not all languages examine word order in the same way that English does. We saw in Chapter 2 that Latin could express "the farmer saw the wolf" in any of several word orders; Here we illustrate it with a similar sentence:

"The farmer saw the poltergeist."

Component and tree diagrams

• 143

Umbria

he saw

SERRA POLTERGEIST

a farmer FARMER

He saw the farmer's shadow. He saw the farmer's shadow.

16263

In these Latin sentences, who did what to whom is not indicated by the word order, but by inflectional suffixes. The same is true for sentences in many other languages, including Russian and German. Hence the following Latin expressions have the same meaning if the same inflection is retained in each noun:

"The poltergeist saw the farmer."

The other three possible orders for arranging these words in an order also indicate the same meaning. Therefore, word order is not the same in all languages. Hierarchical Arrangement of Components As can be seen from the tree diagrams in Figures 5-1 and 5-2, a sentence is more organized than the linear arrangement of its words. To further explore the notion of internal structure, consider the term current information technology, which is ambiguous as it can mean both current information technology and current information technology. is called the constituent structure and can be represented in a tree diagram, as in Figure 5-3 and Figure 5-4.

FIGURE 5-3

Indeed

FIGURE 5-4

education

"current information technology"

Technology

Indeed

education

Technology

"Information Technology That's Current"

As a second example, consider the expression gullible boys and girls. It can mean both "gullible boys and gullible girls" and "gullible girls and boys". In the tree diagram of Figure 5-5 (page 144) you will notice that at the top level there are two branches, that is, two components, while in the tree diagram of Figure 5-6 there are three branches, and therefore , three components. . Also note that in Figure 5.5 boys and girls are components (but gullible boys are not), while in Figure 5.6 gullible children are components (but boys and girls are not). The trees thus capture the two possible constitutive structures (and explain the respective readings) of gullible boys and girls. These figures show that a given string of words can have more than one internal architecture. Figures 5-5 and 5-6 represent the same four words in the same linear order, but with different constituent structures. The ambiguity of meaning in this sequence of words arises from the fact that the sequence has two possible constituent structures.

1 4 4 • Chapter 5 The structure and function of phrases and sentences: syntax

FIGURE 5-5

gullible

FIGURE 5-6

Kinder

girl

"gullible boys and gullible girls"

gullible

Kinder

girl

'gullible boys and (all) girls'

Try it yourself: choose one of these structurally ambiguous expressions and provide a pair of tree diagrams that capture the possible constituent structures: excessive light and glare or modern novel readers. Structural Ambiguity Structural ambiguity can also occur in sentence organization. Examine the following sentence 1. 1. He sold the car to his brother in New York.

Despite the fact that the individual words are not ambiguous, this string of words has more than one possible interpretation, and the ambiguity arises from its two possible constituent structures. Using square brackets instead of a tree diagram, we can represent the ambiguity of 1 in 2 and 3 below. 2. He sold the car [to [his brother in New York]]. 3. He sold the car [to his brother] [in New York].

We can rewrite set 2 to 4 below but not 5 or 6. In contrast we can rewrite set 3 above to 5 or 6 below but not 4:4. It was his brother in New York that he sold Car 5. In New York he sold his car to his brother. 6. In New York he sold his car to his brother.

These examples illustrate that the words in a sentence have an internal organization that is not obvious by simply examining a sequence of words. The linear order of words in a sentence (first, second, etc.) is obvious upon inspection. But only an English speaker can discern the constituent structure of an English sentence and know when a given string of characters has more than one possible internal organization.

Main Clause Constituents: Noun Phrases and Verb Phrases In addition to their apparent linear arrangement, the words in a sentence have a constituent structure that is not obvious but is understood by speakers of the language. consider the

Main Clause Constituents: Noun phrases and verb phrases in Figure 5-7 with their two constituents. More elaborate sentences, like the ones in Figure 5-8, can be analyzed in a similar way.

FIGURE 5-7

FIGURE 5-8

Miguel

cried

1. Alex 2. My uncle 3. The earring nerd

gone, won a bike, spilled the potion

Noun phrases and verb phrases like the ones we studied consist of two main components: the noun phrase (NP) and the verb phrase (VP). (These structures roughly correspond to the functional features of the reference expression and predicate discussed above.) Each NP, in turn, contains a noun (Alex, Uncle, Nerd), and each VP contains a verb (missing, won, spilled). NPs and VPs can be identified by the blanks they fill in a sentence and sometimes by their functions. For example, in Figure 5.8, Alex at 1, My Uncle at 2, and The Nerd with the Earring at 3 serve as reference expressions on which to make a statement. He also disappeared, won a bike and spilled the potions show; they make statements about an NP. NPs can also be identified by substitution methods, such as those contained in the list of alternatives to the two-part structure shown in Figure 5-8. So for Alex, we could replace "My Uncle" or "The Nerd" with the earring. All three are NP because they can appear in the slot, _____ won a bike, or _____ spilled the potion, or _____ disappeared. In sentence 2 below, the Vice President spilled the potion. Unlike the VP of 1, which consists of the single word disappeared, the VP of 2 contains the verb spilled and the SN the potion. Thus a VP can contain an NP. In addition, as example 3, VP can also contain a prepositional phrase (in a competition). NP 1. [Alex] 2. [Genaro] 3. [The nerd with the earrings]

VP [disappeared] [spilled the potion] [won the bike in a competition]

The NPs in the three sets above include Alex, Genaro, the potion, the nerd, the earring, the bike, and a competition. In fact, all you can type in the fields below would be (or at least include) an NP: She liked talking about _____. _____ bothered her without exception.

• 145

1 4 6 • Chapter 5 The structure and function of phrases and sentences: syntax

Inserted anywhere, the following expressions would make a well-formed English sentence and are therefore NP; In each case, the "principal noun" is in italics. Animals the weather your young instructor the thief who stole your bag

his return to his first love his determination to go all the way Walter's race victory an old Cincinnati cyclist

Also note that an NP can be a pronoun: she liked to talk about him/her/it/her/us. Inevitably he/she/she/we harass him/her.

The NP and the pronouns have the same distribution in the sentences; Whenever an NP can occur, a pronoun can take its place. Hence the pronouns are NP. VPs can be identified using similar replacement methods. Consider the sentence Lou cried, where cried represents the PV. Among many others, the following strings can replace crying in the Lou _____ slot. So they fit the frame and are VP (the verb in each VP is in italics):

123

sea

lost the race won an award for his efforts in the tournament

So far we have seen two main parts of a sentence: NP and VP.

Active and passive sentences No matter how many words it contains, an NP acts as a unifying element in a sentence. Even elaborate NPs like the nerd with the earring or what she wanted for her twenty-first birthday are structural entities, as are simple NPs like the lions, she, and Rob. To see more clearly what we mean by a structural unit, a constituent, consider the following sentences, where in each pair the first sentence is active and the second passive: 1a. B. 2. B. 3a. b.

Zelda auctioned off the famous wooden spoon. (ACTIVE) The famous wooden spoon was auctioned off by Zelda. (PASSIVE) The judge fined an elderly plumber from Pasadena. (ACTIVE) An elderly plumber from Pasadena has been fined by the judge. (PASSIVE) The mail truck crushed Karen's bike. (ACTIVE) Karen's bike was crushed by the mail truck. (PASSIVE)

Try it yourself: Using the pattern of the three pairs of sentences above, type in the passive version of Dean Kamen invented the heart stent and the Segway and the active version of The pantry was meant to be stocked by husbands who stay at home with the kids. Speakers implicitly know how a passive sentence relates to an active sentence, but let's try to make explicit what that knowledge should be.

Sentence Construction Rules Based on sentences 1a and 1b above, we can hypothesize this rule: “To change an active sentence into a passive one, replace the first word (Zelda) with the last four (the famous wooden spoon). )". (For present purposes, we ignore the verb was and the preposition for, but a full rule statement would need to specify these features as well.) Our rule produces a well-formed string when applied to sentence 1a; but when applied to 2a it produces the distorted chord given in 2c below, and when applied to 3a it produces the distorted chord given in 3c: 2a. C.3a. c.

The judge fined an old plumber from Pasadena. * Former Pasadena judge's plumber fined. The mail truck crushed Karen's bike. * Truck crushes Karen's bike messenger was nearby.

See for yourself that swapping the first word and last four words of 2a and 3a would give 2c and 3c (assuming the introduction of by and a correct form of the verb BE). Of course, what speakers know about the relationship between active and passive sentences has nothing to do with word count. Instead, active and passive sentences are linked by a structure-dependent operation that depends on the constituent parts rather than the number of words. Refer to the interchangeable parts in the active/passive pairs of sentences 1a and b, 2a and b, and 3a and b. The word sequences in each of the following sets share a structural feature as they work in a similar way: 4. Zelda/The Judge/The Mail Truck 5. The Famous Wooden Spoon/An Old Pasadena Plumber/The Bike Karen

The NPs of sentences 4 and 5 move as units in the syntactic operation relating a passive sentence to an active sentence. When combining active and passive clauses, NSs act as constituents, no matter how many words they contain.

Sentence construction rules Extension of noun phrases Based on the analysis of categories (parts of speech) in Chapter 2, we can now characterize certain types of NP and present them as examples: Noun (N): Karen, oracle, justice, floating determiner (Det) ⫹ Noun : the amulet, a potion, some gnomes, my saucer determinant ⫹ noun ⫹ prepositional clause (PP): the book on the table, a price increase, the market for ideas, the man behind the curtain determinant ⫹ adjective (A) ⫹ noun: an ancient oracle, these hellish halls, the first omen, my flat saucer

To represent these different NP patterns, we use sentence structure rules like the following: 1. 2. 3. 4.

NP NP NP NP

→ N → Is N → Is N PP → Is A N

(NP consists of N) (NP consists of Det ⫹ N) (NP consists of Det ⫹ N ⫹ PP) (NP consists of Det ⫹ A ⫹ N)

• 147

1 4 8 • Chapter 5 The structure and function of phrases and sentences: syntax

These four rules or extensions can be combined into one rule. To do this, we put the optional elements in brackets or those that do not have to appear. Note that N is the only component needed in any NP extension; the rest is optional and must be enclosed in parentheses. The combined rule looks like this: 5. NP → (Det)(A)N(PP)

Rule 5 can be expanded to the four separate rules (1 through 4) that we intend to capture. In addition, however, it includes some extensions that we did not expect. Since Det, A, and PP are optional, we can rewrite NP not only as 1, 2, 3, and 4, but also in other forms: 6. NP → A N 7. NP → Det A N PP

Therefore, Rule 5 proposes additional extensions that we do not intend to capture. If English has well-formed NP structures consisting of AN (as in 6) and Det AN PP (as in 7), plus all other extensions that admit 5, then 5 is valid. Otherwise we would have to revise it to exclude malformed structures. Of course, some English NPs are made up of A and N (ordinary superheroes; natural grace; great imagination), while others are made up of Det A N PP (your sad life on the edge; the white whale on the beach; those fantastic clouds on the darling). An advantage of formalisms like rule 5 in combination is that they often contain unexpected statements that can be verified with other data. In doing so, they prove their own validity.

Prepositional phrase suffix PP means prepositional phrase, like in the car, from Xanadu, New York, to the brother, with the earring, and for the judge. PPs consist of a preposition (PREP) and usually a noun phrase (NP), so the syntax for PP is: PP → PREP NP

If NP is considered optional (as in She walked behind the car/She followed), then the rule would place NP in parentheses.

Expansion of sentences and verb phrases To capture the fact that sentences and clauses have two basic parts, we can formulate the following sentence structure rule: S → NP VP

Each sentence structure rule can produce a tree diagram, and this produces the following tree: S

public notary

Vice President

Sentence Structure Rules Having seen several extensions of NP, we can now move on to the internal structure of VP. The following extensions of our structure for identifying VP show that the structures on the right (those after Lou) are VP; the labels below the VP items indicate their categories. _VP_ 1. Lou called V ____VP____ 2. Lou bought a bike V NP ________VP______ 3. Lou received the bike in May V NP PP

Theorems 1, 2, and 3 above show three ways to expand VP: VP → V VP → V NP VP → V NP PP

V is the only ingredient that appears in all of these rules. On the other hand, NP and PP are optional. Using parentheses for optional elements, the above three expansions can be combined into a single sentence structure rule representing that VP must have V and may have NP, PP, or both: VP → V (NP) (PP)

Just as we discover unexpected options when we combine four NP extensions into one, the combined rule for VP will produce the structure V PP. Note that V PP is not represented between theorems 1, 2, and 3, which formed the basis of the constituent structural rules for VP. We can check the validity of the extension and see that V PP is actually needed to represent the internal structure of VP in sentences like (Finian) played in the yard, (Dana) ran around the track, and (Pat) flew to Ballina. The latter is shown below. _____VP____ Pat

Volon in Ballina V

PAGES

Sentence structure rules and tree diagrams We formulate four sentence structure rules: S → NP VP NP → (Det) (A) N (PP) VP → V (NP) (PP) PP → PREP (NP)

• 149

1 5 0 • Chapter 5 The Structure and Function of Phrases and Sentences: Syntax

These represent the fact that a sentence has an NP and a VP; that an NP has an N; that a VP has a V; and that a PP has a PREP. According to these sentence construction rules, all other possibilities are optional. Our rules can produce the following tree diagram: S

public notary

Vice President

Norte

This is the simplest structure generated by our sentence structure rules, and represents sentences like Lou Gone and Stinks. Now consider the more complicated structure in Figure 5-9, where we provide an example sentence for the structure. Of course, our four sentence construction rules can represent structurally simple or structurally complex sentences.

FIGURE 5-9 PVS

PN was

Norte

PREPARATION

Hall

PAGES

public notary

by Butte

public notary

Has won

and

Norte

forgive

PP PREPARATION

PN was

Norte

just

Try it yourself: create a sentence whose constituent structure is the same as in Figure 5.9. Then, using the four sentence structure rules that produced Figure 5.9, give a different structure and two example sentences to illustrate it.

Grammatical relationships: subject, direct object and others

Grammatical Relationships: Subject, Direct Object, and Others Using syntax rules, we can define the subject and direct object precisely. These two sentence structure rules are important for the definition: S → NP VP VP → V (NP) (PP)

Immediate Dominance We can represent the relevant parts of these sentence structure rules in a tree diagram. In Figure 5-10, the circled NP is directly below the S node, the boxed NP is directly below the VP node, and the VP node is directly below the S node and is immediately dominated by this other node. Thus in Figure 5.10 V is immediately dominated by VP; the circled NP is immediately dominated by S; the boxed NP is immediately dominated by VP; and both VP and NP within a circle are immediately dominated by S.

Subject and Direct Object We can now define subject and direct object using sentence structure rules and the tree diagrams they produce. In English, subject is defined as the NP immediately dominated by S. In our diagram, the circled NP is the subject. The direct object is defined as NP immediately dominated by VP. In Figure 5-10 is the boxed NP. Since NP is an optional element in the VP extension, it follows that not all sentences have a direct object.

FIGURE 5-10 S Subject VP

public notary

direct object v

public notary

Transitive and intransitive Recall from Chapter 2 that a sentence without a direct object contains an intransitive verb. Examples of intransitive verbs are crying, running, laughing, and disappearing, all of which can occur without a direct object. On the contrary, the verbs

• 151

1 5 2 • Chapter 5 The structure and function of phrases and sentences: syntax

verbs that take a direct object are called transitive verbs; Examples are craft, buy, and find, such as B. making a potion, buying a motorcycle and finding a penny. While some verbs can be both transitive and intransitive, as shown in the first three pairs of sentences below, others are only transitive or just plain intransitive: Intransitive

Transitive

Jose won. Taylor sings. Suze studied at Oxford. Michael disappeared. *Michael is scared.

Josh won an award. Taylor sings lullabies. Suze studied economics at Oxford. *Miguel left with the dishes. Michael startled the kittens.

Try it yourself: make a sentence with a verb that can only be intransitive and another with a verb that can only be transitive. Then give a pair of sentences that use the same verb but are transitive in one case and intransitive in the other.

Grammatical Relations Certain structural properties of direct subjects and objects cannot be equated with anything else, including meaning. Subject and direct object are grammatical relationships. Grammatical relationship is the term used to capture the syntactic relationship in a sentence between an SN and the predicate. In other words, grammatical relationships indicate the syntactic role that an NS plays in your sentence. Besides subject and direct object, sentences can have other grammatical relationships, such as B. Indirect, crooked, and possessive objects. English has skewed grammatical relationships for the NPs that are the subject of a preposition (The poltergeist pointed to the tooth) and owner (Josh's car).

Passive Clauses and Structural Dependence Having defined subject and direct object structurally, we can now return to a syntactic relationship discussed earlier. The notions of subject and direct object allow us to reformulate the relationship between active and passive sentences as follows: To transform an active sentence into a passive sentence, we swap the subject SN and the direct object SN.

(As before, you need to predict the preposition by and a form of the verb. See an example:

BEING.)

Surface Structures and Underlying Structures You can see that in a passive sentence, the direct object of the active sentence appears as the subject and the subject of the active sentence appears as the oblique (preceded by the preposition by).

Surface Structures and Underlying Structures We have seen that speakers know more about the structure of a sentence than is revealed by the linear sequence of its words. Not only do we have implicit knowledge of constituent structure, but we often understand more constituents in a sentence than are actually expressed. For example, to understand the meaning of sentences like the following, knowledge of the syntactic rules of English is essential, although this knowledge may be implicit:

1662663

Lisa won an award, but Larry

1. 2. 3. 4. 5. 6. 7.

Not. I do not care. he didn't tell Sarah. He didn't party with her. he wasn't in Paris to buy a tie. he didn't train tigers. won no awards.

Although the list of possible propositions that follow this pattern is endless, the only legitimate interpretation of Proposition 1 is Proposition 7. Propositions 2 through 6 are not possible interpretations of Proposition 1. We understand that Proposition 1 has the implicit conclusion that one prize wins. To explain this, remember that in Chapter 4 we posited underlying forms of sounds and morphemes. Likewise, one way to account for tacit knowledge of sentence structure is to posit underlying syntactic structures. For example, we can represent the meaning of sentence 1 by positing an underlying form similar to sentence 7: Lisa won an award, but Larry didn't win an award. If we were to take such an underlying form, we would posit certain syntactic processes that would preclude the second occurrence of winning a prize, such that Lisa would win a prize but not Larry.

Syntactic Operations We have seen examples of syntactic operations in English, including movement, as in passivation, and deletion, as in . . . but not Larry. Now let's look at the others.

Question Formation English has two main types of questions. Yes/No questions like Was it an open discussion? can be answered with yes or no. Informational questions contain a wh-word such as who, what, or when and require more than a simple yes or no answer. Yes/No Questions Review the following pairs of statements and yes/no questions. 1. Suze deserves a fair wage. Will Suze earn a fair wage? 2. Tony was winning the race when he stumbled. Did Tony win the race when he tripped?

• 153

1 5 4 • Chapter 5 The structure and function of phrases and sentences: syntax

If you compare the form of the sentence with the form of the question, you will see that the auxiliary verb must be placed in front of the subject NP in the yes/no question. (Verbs like will in 1 and was in 2 and did and does in 3 and 4 below are auxiliary verbs and are distinguished from main verbs like win and win. Verbs that can be placed before the subject NP to form questions are (called auxiliary verbs; Auxiliary verbs are also parts of VP that carry the negative element in contractions like can't, should't, and wasn't.) Note that yes/no questions have an auxiliary verb, even if the corresponding statements like those in 3 and 4 not shown: 3. Alvin studied alchemy in college Did Alvin study alchemy in college 4. Inflation always hurts the poor Does inflation always hurt the poor?

Pairs of sentences such as 3 and 4 provide an argument for postulating an auxiliary verb in the underlying structure of each sentence, although not all sentences express an auxiliary verb in the surface structure. But in English, an auxiliary verb must appear in the surface structure of negative sentences (Alvin didn't study alchemy) and questions (Is inflation bad for the poor?). Auxiliary verbs can also be used to express emphasis (But she works out every day!) and other semantic information such as timing (She's going to win) and aspects (You're coming home). (Aspect and tense references are discussed in Chapter 6.) Because an auxiliary is often present in surface structure (and for other reasons not discussed here), an auxiliary constituent is postulated in the underlying sentence structure. Like other parts of the underlying structure, the auxiliary verb is generated by a sentence structure rule. Instead of the previous rule that extends S to NP VP, we can postulate the following rule: S → NP AUX VP

We can represent the structure created by this rule in a tree diagram: S

public notary

FOR HIM

Vice President

The operation that changes the constituent structure of statements 1, 2, 3, and 4 above to the constituent structure of their respective yes/no questions moves AUX to a space before the subject NP, as shown below: S

public notary

FOR HIM

Vice President

FOR HIM

public notary

Vice President

Therefore, we represent the underlying form of the 1-sentences on page 153 as a tree on the left in Figure 5-11. The tree on the right is the constituent structure that results from applying the subject-auxiliary inversion operation.

Surface Structures and Underlying Structures FIGURE 5-11 S NP

Vice President

FOR HIM

Norte

FOR NP

is a

Norte

Wille

To win

Vice President

Norte

PN was

Suze

public notary

Suze

To win

and

AND

Norte

Fairer Lohn

a fair wage

The rule could be written as: NP AUX VP

VP AUX-NP

Information Questions In an information question, the information sought, the item being asked, is represented by a WH word (who, why, when, where, what, what, how), and such questions are sometimes referred to as questions. (Note: in the remainder of this chapter we sometimes ignore the distinction between who and whom in traditional grammar and careful writing and speaking.) Information problems manifest themselves in two ways. Someone repeats the form of the statement, as in these examples: (He boils horse feathers).

what is cooking

(She was looking for Sigmund Freud.)

Who was he looking for?

The linear form of an echo question is identical to that of the statement, except that instead of the question part, a WH word appears: what horse feathers, who for Sigmund Freud. General information questions are more common than echo questions. They have the form shown below, in which an operation called WH move precedes the word WH: 1. What is cooking? (Cooking what?) 2. Who were you looking for today? (Who were you looking for today?)

If you compare these general information questions with the echo questions in parentheses, you can see that two changes have taken place: • The word WH (the part being queried) appears before its clause. • The auxiliary constituent comes before the subject NP.

Note that frequent informational questions leave a "gap" in the structure at the point left vacant by the initial word WH (here denoted by a hyphen). This does not apply to echo questions as the word WH remains in its underlying position.

• 155

1 5 6 • Chapter 5 The structure and function of phrases and sentences: syntax

Embedded sentences We've already seen sentences like Lou. Now let's look at sentences that have built-in sentence-like structures. In the following examples, the phrase in italics is incorporated into another sentence. 1. Suze said that Lou was crying. 2. The fact that Jay won the marathon surprised Sheila.

In sentence 1, the sentence Lou cried is embedded in the structure of the sentence Suze said. Therefore, the phrase Lou cried is structurally equivalent to the word something in the sentence Suze said something. In 2, the clause that Jay won the marathon is built into the structure, Sheila surprised. The clause embedded in 2 (That Jay won the marathon) is structurally equivalent to That in It Surprised Sheila or The News in The News Surprised Sheila.

Subordinate clauses The embedded clause can be introduced by a word that would not appear in that position if the clause were an independent clause. This word, called the subordinate word, serves to mark the beginning of a nested sentence and to help identify its function in the sentence. Point 2 above is an example of such a subordinate. Not all embedded clauses need to be preceded by a subordinate. Compare these pairs of sentences: 1. Suze said that Dan washed the dishes. 2. Suze said that Dan was washing the dishes. 3. We were surprised that she won. 4. *We were surprised.

Note that 1 and 2 are well-formed with or without the that subordinate. But while 3 is well-formed, 4 is not. In the sentences we examined, one clause acts as a grammatical component of the other clause and is called subordinate to it. The subordinate clause is embedded in a main clause. By definition, every subordinate clause is embedded in a main clause and plays a grammatical role in it. (The grammatical functions include the subject and the direct object, which we will see later, but they can also be adverbial, as in She was more successful than she tried harder.) For example, in sentences 1 and 3 below, where functions the brackets highlight the clauses, each embedded clause as a grammatical unit in its main clause and has the same grammatical function in its main clause as the underlined word in the sentence directly below it: 1. Harry said [saw a ghost]. 2. Harry said that. 3. [Josh's fear of witches] upset his wife. 4. He teased his wife.

Tree diagrams can also illustrate the relationship between clauses in a sentence, such as B. Harry said he saw a ghost. In representing such a sentence, we can replace the sentence saw a ghost with the word that, as in Figure 5.12. This tree diagram captures the fact that the embedded clause S2 (Saw a Ghost) functions structurally as part of the array

Surface Structures and Underlying Structures FIGURE 5-12 S1

Norte

torment

saw a ghost

Clause S1 (Harry said—). The embedded clause fills the same space in the parent clause as the word in the Harry clause expresses.

Relative clauses A relative clause is formed when a clause is nested within an SN of another clause to produce structures like the following (relative clauses are in italics): 1. The principal praised [the teacher who let me down]. 2. [The jewelry he borrowed] was fake. 3. Sarah saw [praises French director Kim]'s new film. 4. Sarah saw the new movie [praises French director Kim].

When SNs with the same referent appear in two clauses, a relative clause can be formed by embedding one clause within the other, as in the example below. Identical subscripts (index j) in primes indicate identical referents: I gave your address to my cousinj

My cousin lives in Dublin

I gave your address to my cousin who lives in Dublin.

Relative clauses in English are usually introduced with a relative pronoun, such as who (or who or their), which, or that. As in 4 above, the pronoun can be omitted in certain circumstances. Relative clauses modify nouns, and the noun that modifies the relative clause is called the noun. The main noun is 'repeated' in the embedded sentence, where it is relativized (ie takes the form of a relative pronoun). A relative clause is the part of the noun phrase that contains the main noun. The resulting noun phrase structure can be represented as in Figure 5-13, where the main noun is denoted by N. Note that in this case it is the relativized SN that acts as the subject of its sentence (the SN immediately dominated by S).

• 157

1 5 8 • Chapter 5 The structure and function of phrases and sentences: syntax

FIGURE 5-13 NP

Norte

Vice President

NP V

eng

principal

wer

life

PP PREPARATION

public notary

Dublin

In other sentences, the relativized SN can be the direct object, as in this figure: The jewelry you borrowed was fake.

Here the relative clause he borrowed derives from the underlying clause that he borrowed the jewels, with the jewels being the direct object of the borrowed verb. A relativized NP can also be an oblique as in 1 or a possessor as in 2:1, that's the official I was talking about. (cf. I told you about the cop) 2. This is the cop whose car was totaled. (cf. the police car was destroyed)

In English, a relativized SN can have the following grammatical relationships within its sentence: subject, direct object, oblique, or possessor.

COMP node Let us now discuss the syntactic processes involved in forming relative clauses in English. Examine the following sentences, noting the "gap" in the structure (indicated by a hyphen -): 1. There is the hole I warned you about -.

2. JK Rowling wrote the novel I recommended.

3. The fans who braved the weather paid a price.

We can represent the constituent structure underlying these propositions in a tree diagram, as Figure 5.14 for proposition 2 shows the WH motion operation that we presented above for information questions. Figure 5-14 shows a node named COMP.

Surface Structures and Underlying Structures FIGURE 5-14 S NP N

Vice President v

NP S

PN was

JK Rowling

I wrote

Norte

COMP.

oh romance

I recommended the novel

(to 'complementor'), which we have not previously identified. It is possible to discuss the WH shift operation for relative clauses without using the COMP node (as we did in the information questions), but there is good reason to posit such a node in the underlying structure. COMP serves, among other things, as a kind of substitute and "magnet" for WH components such as that, which, who and other relative pronouns as well as for WH information question components. Because syntactic operations transform one constituent structure into another, we can represent the result of the WH shift applied to Figure 5-14 by the tree shown in Figure 5-15. Thus, the WH shift removes a WH component from S and adds it to the COMP node. We have studied the WH movement in relation to relative clauses, but the

FIGURA 5-15 S NP N

Vice President v

NP S

PN was

JK Rowling

I wrote

Norte

oh romance

COMP.

was

I recommended

• 159

1 6 0 • Chapter 5 The structure and function of phrases and sentences: syntax

the same operation could move any WH component to the COMP node, including question words in information questioning.

Types of syntactic operations Although it is not known how many types of syntactic processes exist in human languages, recent syntactic theories reflect evidence that these operations are considerably more general than our detailed specifications of certain operations suggest. Motion operations are very common in the world's languages, and in a syntax-theoretic model, all transformations are motion rules.

Functions of syntactic operations We have examined various syntactic operations, mainly from a structural perspective, and we have emphasized the fact that syntactic operations depend on structure. For example, regardless of subject and object length or grammatical complexity, active and passive sentences are related to the extent that the object of the active sentence (The judge fined Jaime) is the subject of the associated passive sentence (Jaime was fined occupied by the judge), and the subject of the active clause is a slash (from the judge) in the passive structure. What is the purpose of syntactic operations like those that form passives? We illustrate with examples in English, but comparable analysis applies to syntactic operations in other languages. In English, as we have seen, yes/no questions are formed by shifting the auxiliary verb in certain structural patterns. Thus, a syntactic operation from the underlying structure Jurará produces the structure Jurárá? From a functional point of view, the declarative makes a statement; The interrogative asks a question. Languages must be able to make statements and ask questions. The point is that, as in these English examples, the question form is related to the sentence form and in this case is achieved through the syntactic operation of subject-auxiliary inversion. (There are other ways of asking questions in English, including associating a particular intonation pattern with a declarative structure.) Equally interesting and less obvious are the functions that active and passive structures serve. In general, active and passive sentences mean the same thing. Because if Jaime was fined by the judge, then it must be the case that the judge fined Jaime. Then why should there be two ways of saying the same thing? Why does English syntax offer active and passive versions of a sentence? To find the answer, consider the following passage about a baseball player named Odalis Pérez; was featured in the sports section of the Los Angeles Times. Pérez allowed Barry Bonds an inside single before putting Andrés Galarraga on another inside popout for the final out. Pérez re-engaged Hernández when he left the field, resulting in his dismissal. He was also ejected on June 13 against the Cleveland Indians at Jacobs Field.

Smith is the grammatical subject of the first two sentences, and He (ie Smith) is the grammatical subject of the third sentence. (The he in the second sentence also refers to Perez.) Of the four verbs in italics, the first three are active and

Functions of syntactic operations

• 161

all three have Pérez or he (i.e. Pérez) on the subject: Pérez broke off; Perez reinstated; I was on my way It is clear that the author's focus is on Perez. Let's examine the last sentence as originally written (repeated in 1 below) and as the active sentence (2): 1. He was also ejected on June 13 against the Cleveland Indians at Jacobs Field. 2. He was ejected by a referee on June 13 against the Cleveland Indians at Jacobs Field.

The use of a passive structure in the last sentence of the published passage allows the sportswriter to focus on Perez, since in English the subject tends to be the subject of his sentence, the center of attention (a subject we will discuss in Chapter 2 will come back 8). ). ). Introducing an arbiter into the passage as the grammatical subject of the third sentence would make the arbiter the subject of that sentence and change the focus of Pérez. Also note that in a passive sentence, the subject of the corresponding active sentence can be omitted entirely. Instead of being sent off by a referee, the author could (and did) simply say he was sent off. By using what is known as an agentless passive (one in which the underlying subject is not expressed), the author keeps the passage's focus on Perez.

Try it yourself: Explain why the author of the following passage uses the underlined passive structure in sentence 3 and why this sentence combines an active verb (overrun) and a passive verb (was kicked out). 1. The Dodgers scored just one run despite loading bases twice in the sixth, and Daryle Ward's baserun error in the ninth added further frustration to a team that faced many in Northern California. 2. With Jolbert Cabrera first and an out, Ward hit the hole on the right and sent Cabrera into third place. 3. But Ward stormed the bag and was ejected, quickly dampening Dodgers spirits.

We have also seen that informational questions have alternative forms: those in which the respondent is confronted (What did you find? Who did you meet there?) and those in which you are not confronted (Did you find something? Did you who met?) . Speakers use echo questions when they can't hear something completely or want to express surprise or disbelief at what they're hearing.

Computers and the Study of Syntax In Chapter 2, we saw that computer programs can do a good job of tagging words in sentences with their parts of speech and identifying lexical categories for many English words. Programs that can analyze a variety of lexical categories for their constituent structure are called parsers, and some parsers are impressively successful at this

assign constituent structure to a chain of lexical categories. Tagging the words in a sentence for a lexical category is not the same as identifying the constituent structure. For just as a sequence of words can have more than one constituent structure, it can also have a particular sequence of lexical categories. as we saw

1 6 2 • Chapter 5 The structure and function of phrases and sentences: syntax

Computing and studying the syntax like this. Once the part-of-speech tags have been assigned, a parser uses only a few sentence structure rules to generate a constituent structure bracket or tree diagram for the string. You can visualize the process from bottom to top in a tree structure. For example, the sentence construction rule NP → Det N brackets farmers and wolves as NP, resulting in the bracket NP V NP. The sentence construction rule VP → V NP allows to put the lobes in square brackets as VP, giving NP VP. This in turn is recognized as representing S. Taken together, the tagged string can be parsed as shown in the image at the bottom of this page. For more complicated strings (as is the case with most natural strings), assigning the correct constituent structure may not be as straightforward. Grammar checkers in word processors have relatively simple parsers. Based on these parsers, they sometimes suggest changes to your syntax for the sake of grammatical correctness or stylistic refinement. As you may have noticed, grammar checkers often suggest revisions that indicate they misparsed the sentence. Also, grammar checkers often find that natural sentences are too long to parse, and the best they can do is suggest that the sentence be shortened. ■

above, the noun phrase gullible boys and girls has two possible constituent structures. It follows that the sequence of lexical categories for this sentence can also have two square brackets, or [A [N Conj N]]

[AN] [Conj] [N].

Some sophisticated computer programs can parse tagged sentences and produce a tagged constituent structure or tree diagram. In the case of gullible boys and girls, a parser would produce two candidates for constituent structures. Researchers faced major challenges in designing parsers that could parse a variety of natural English sentences. Many sentences are relatively simple and easy to parse, but not all, and made-up sentences tend to be easier to parse than real sentences found in ordinary conversation. Let's see how a parser would work. First, on a sentence like That rancher saw the wolves, a tagger would assign parts of speech to words like this: That rancher det saw Det's wolves Basically, saw could be a noun or a verb, and that could have multiple possible tags as well, but even moderately skilled ones Taggers will have no trouble determining the correct tags in a sentence.

[S [NP [Das

Ö]

[rancheiro

Norte

]

NP] [VP

[vi

V [NP

[ a

Ö]

[lobos N]

This tagged bracket fully corresponds to the following tree diagram:

SNP

Vice President

public notary

Det v

Norte

Was

Rancher

Norte

Lobos

NP] VP] S]

What do you think? checked

Summary • The rules of sentence formation form the syntax of a language. The study of sentence structure is also referred to as syntax. • All languages have reference expressions and predication expressions. • Syntactically, a reference expression is an NP (noun phrase) and a predicative expression is a VP (verb phrase). • A clause (and a clause) consists of a verb with the required set of NPs. • Speakers of any language can generate an unlimited number of sentences from a finite number of rules for combining sentences. • There are two types of syntactic rules: sentence structure rules and syntactic operations. The latter can be referred to as "transformations". • Sentence structure rules generate underlying constituent structures. • Syntactic operations transform one constituent structure into another constituent structure. • The positioning of the underlying constituent structures captures the surprising regularity of certain relationships between sentences. • The positioning of underlying structures helps explain some elements of meaning and certain syntactic and semantic relationships between sentences. • To explain how speakers relate two structures to each other (how does Martha not believe in poltergeists and Martha not believe in poltergeists?), linguists posit an English rule that transforms the structure underlying the basic declarative sentence into the structure on which the sentence is based. derived question. • It seems that the most important and general syntactic operations involve motion, such as WH motion.

WHAT DO YOU THINK ? R E V I S I S I T WHAT DO YOU THINK? REVISED • Noise and productivity. We hear some expressions so often that it's easy to guess where we know them from. But probably most of the things we say in our interactions we have never heard or read before. Fortunately, there is relatively little underlying structure to what we say and hear. By combining these few structures in different ways, we can easily and accurately form or understand entirely new sentences. • Yellow and ambiguity. Ambiguous words often go unnoticed because the contexts encourage a particular reading and effectively eliminate the alternatives. The statement I'm going to the bank to deposit a check eliminates the likelihood of a river bank. Another type of ambiguity arises when a sequence of words has more than one possible internal meaning.

• 163

1 6 4 • Chapter 5 The structure and function of phrases and sentences: syntax

Organization. New drug combinations may mean "new drug combinations" or "new drug combinations". • Ron reads. You agree that the common word order in English is subject, verb, and object, but you've noticed many exceptions. English allows both "One of the ones I want!" and "I want one like that." So "extra money could be reclaimed elsewhere, Romer said" is similar because the subject Romer is at the end of the sentence. It puts more emphasis on the initial noun phrase "extra money". • Nerd Ned. Ned should consider using a grammar checker that's so simple, it recommends changing every passive sentence you type. Finally, sometimes a passive sentence is just what is needed to keep the focus where the author intended. If Ned wrote about a winning team, you probably want to keep the focus there, and one way to do that is to make the winning team the topic of the sentence.

Exercises Practical exercises

A. In each sentence, identify the subject component and any direct object components. a. Political engagement will achieve this. B. This does not necessarily mean that everyone will have a win-win situation. C. Does one of them ring a bell? i.e. I would have noticed the ideal student of a progressive teacher. Y. So what does the alternator do? Q. The Mission's most prominent figure took a more radical position. Grams. And now the man who wanted to be a woman with all his heart wants to stay a man. B. Each following sentence has undergone one or more syntactic operations. Provide an appropriate underlying form for each, except for those in parentheses, which are provided for context only. a. The program was started by disaffected sophomores. B. The other girls were frustrated with our tactics. C. Erin could eat calamari, but I don't think so. i.e. (I didn't expect that to happen.) Did you? Y. (How are you?) Not bad. It can't be better! Q. Did you know that he failed Chemistry 101? grams. I love apples. I hate pears h (It doesn't have to last.) Or do I?

exercises

Based on English 5-1.

a. Give as many examples of these constituents as you can identify in sentences 1) and 2) below: NP, PP, VP. B. List as many examples of these lexical categories as you can identify in sentences 1) and 2) below: N, PREP, V. 1) A show at a stadium near St. Louis. 2) The trouble started when Axl Rose asked venue security to confiscate a camera he saw near the front of the stage.

5-2.

Provide an illustration for each of the VP extensions provided on page 149. Example: V NP - ate an apple.

5-3.

a. Draw a labeled tree diagram for each sentence. 1) ancient inscriptions 2) in the dark of night 3) brewed a potion 4) borrowed the book recommended by the teacher 5) monstrous members of a terrible kingdom b. Provide a tree diagram for each sentence (ignore italics for now). 1) Witches scare him. 2) Heaven flooded the earth with water. 3) A ghost has the spirit of a dead person. 4) Are there spirits in the physical world? 5) Does she believe ghosts exist? 6) The teacher I described to you won the race. C. For each italicized phrase in the above sentences, determine whether or not it is a constituent and give its name if so.

5-4.

What is the difference in the relationship between Harry and the verb to see in 1) and 2) below? Draw tree diagrams of the underlying structure of the two sentences, showing the difference in structures. 1) Josh advised Harry to see a doctor. 2) Josh promised Harry he would go to the doctor.

5-5.

English has a syntactic operation called dative motion that derives sentence 2) from the underlying structure of sentence 1: 1) I am sending a letter to Hillary. 2) I sent Hillary a letter. The following sentences are also examples of sentences with the dative: He sold his brother a sailing boat. Hal won't tell me Daniel's new phone number. I'm going to buy my cousin new pajamas.

• 165

1 6 6 • Chapter 5 The structure and function of phrases and sentences: syntax

a. Give the three underlying propositions that correspond to the three derived propositions. B. The dative motion applies to prepositional phrases beginning with the preposition to, but cannot be applied to prepositional phrases beginning with most other prepositions: *I'll do your homework for you. (from I'll do your homework with you.) *My neighbor heard the news on the radio. (My neighbor heard the news on the radio.) But the dative does not apply to all sentences that begin with the preposition for. The sentences in 3) below cannot go through a dative motion, as evidenced by the ungrammaticality of the corresponding sentences in 4). 3) Drive a truck to New Orleans. They will forward your complaint to the main office. 4) *Drive a truck in New Orleans. *He will forward the complaint to the main office. Describe the dative in detail. C. Now consider the ungrammatical clauses in 5) below, which are derived via the dative movement of the underlying clauses to the corresponding base clauses in 6). How should you change your description of the dative movement so that it doesn't produce the ungrammatical sentences in 5)? 5) *I gave it to my new neighbor. * I will bring my little sister with me. *They will probably mail it to me. 6) I gave it to my new neighbor. I'll take her to my little sister. They will probably send me to him. 5-6

English has the grammatical relationships of subject, direct object, slash, and possessor. But it is questionable whether the indirect object is a definite grammatical relation, and if so, whether it occurs in sentences like The Witch bot The Boy A Potion or The Witch bot The Boy A Potion. The syntactic properties of the child differ in the two sentences. What syntactic evidence can you offer to argue that the child does not have the same grammatical relationship in each of these sentences? (Note: At least one syntactic operation examined in this chapter does not produce grammatical strings for either sentence.)

5-7.

English has two types of relative clauses. Type 1 has been described in this chapter; Leave the prepositions where they are in the original sentence. This is the man [whom I spoke to last night]. (original sentence: I spoke to the man last night) In type 2, the preposition is shifted to with the word WH at the beginning of the sentence. This is the man [whom I spoke to last night]. Describe the operation of the relative clause that forms type 2 relative clauses, focusing on how it differs from the operation that forms type 1 relative clauses. Base your discussion on the following data:

Exercises This is the man [who left]. (Types 1 and 2) *This is the man [on the left]. (Types 1 and 2) This is the man [that I saw]. (Types 1 and 2) This is the man [that I saw]. (Types 1 and 2) This is the man [that I saw]. (Types 1 and 2) This is the man [that I saw]. (Types 1 and 2) This is the man [to whom I gave the book]. (Type 1) This is the man [I gave the book to]. (Type 1) This is the man [I gave the book to]. (Type 1) This is the man [I gave the book to]. (Type 1) *This is the man [I gave the book to]. (Type 2) This is the man [I gave the book to]. (Type 2) *This is the man [I gave the book to]. (Type 2) *This is the man [I gave the book to]. (Type 2) 5-8.

On page 157 we noted that the relative pronoun can be omitted in certain structures. For example, Ø stands for an omitted relative pronoun in the following sentence: Sally saw a new film by French director Ø Kim loved him. a. For each of the following sentences, identify the grammatical relationship of the relativized NP within its sentence, using S for subject, DO for direct object, and Obl for oblique. 1) I lost the book [that you gave me]. 2) He rented the video [that scared you]. 3) I met the professor [who taught me geometry of solids]. 4) I met the poet [we read about it last week]. 5) I found the video [you lost]. 6) I saw the oak tree [under which you slept]. 7) The new teacher that Lou liked just quit. 8) The new teacher [who liked jazz] just left. 9) I took an apple from the tree you planted. 10) I like the new lyrics [that you complained about]. B. In which sentences could you omit the relative pronoun? C. What would not allow omitting the relative pronoun? i.e. Which grammatical relationships allow to omit the relative pronoun? Y. In which grammatical relationships can the relative pronoun not be omitted? F. Rewrite sentences 4, 6 and 10 and put the preposition in front of the relative pronoun. Gram. Can the relative pronoun be omitted in the paraphrased versions of 4, 6, and 10? h What generalization can you make about when a relative pronoun can be omitted in your sentence?

• 167

1 6 8 • Chapter 5 The structure and function of phrases and sentences: syntax

5-9.

Considering the rules of motion for forming questions, analyze what happened in the derivation of the following sentences to produce the malformed sentence. How would you phrase the auxiliary motion rule to avoid the ungrammatical example? The teacher who will be teaching this class is Lily's aunt. * Could it be that the teacher who will be giving this lesson is Lily's aunt? Is the teacher who teaches this class Lily's aunt?

5-10 Here are six example sentences that a word processor's grammar checker found offensive, along with the comment suggesting a specific correction. In each case, the grammar checker was incorrectly parsed, and the proposed correction would result in a misspelled sentence. For each example, identify the word or constituent structure that the grammar checker parsed incorrectly and explain the basis for the suggested correction. Example: "In this sentence, each embedded sentence acts as a grammatical unit in its parent sentence." Comment: The word each does not match functions. Consider function instead of functions. Explanation: It is likely that the grammar checker incorrectly interpreted the functions as a plural noun (instead of a singular verb of 3. If the parsing were correct, "any embedded clause function" would be well-formed. 1) "It is the order of the words in the sentence , which indicates who is doing what to whom." Commentary: Consider are instead of is. 2) "Do "George Washington" and "the first President of the United States" mean the same thing?" Commentary: Consider presidents instead of presidents, or media instead of average. 3) “When a student volunteers, Disneyland is fun. . . ” Comment: The word “a” does not match among the volunteers. 4) "Linguistic semantics is the study of the systematic ways in which languages structure meaning." Comment: Consider the language or languages instead of languages. 5) "Theorem 2 is true because we know that the word dogs describes entities that are also described by the word animals." Comment: Consider describing instead of describing. 6) "Harold, who has two PhDs, gave me a fascinating insight into Warhol's art last night." Comment: Consider given rather than given.

Based on languages other than English 5-11. Examine the tree diagram for this Fijian phrase: ea-biuta na ŋone vakaloloma na past - leave the poor boy or "The bad man left the poor boy on the bus".

Thomas-Mann

This is bad

is in

and the women the bus

exercises see

Vice President

public notary

give it to her

PAGES

public notary

Norte

AND

Norte

AND

von None vakaloloma

von

attempt

public notary

PREPARATION

daa

Norte

von

after

a. Give the sentence structure rules that produce this constituent structure tree. B. Note that the order of certain phrases in the Fijian sentence differs from English. In terms of the constituent order, what are the main differences between Fijian and English? C. Using the structure of the tree, determine which of the following phrases are components and name each component. based on vakaloloma in tamata and based on tamata ðaa ðaa and based on ea-biuta

ea-biuta e o wakaloma e na e o o o o o wakaloma e os Tomaten ðaa o wakaloma e-biuta e o

Specially for educators and future teachers 5-12. Although we have minimized the difference between who and who in the examples in this chapter, writers and speakers who regularly distinguish between them in relative clauses do the following: 1) It was Lynne who answered. 2) She is an avid golfer and mainly plays in Morro Bay. 3) A nurse who walked her dog off leash was fined $600 4) He then came to meet the woman he wanted to marry. 5) His grandmother, whom he adored, was seriously ill. 6) He married a woman from Devon whom he had met in Australia. 7) Ricardo, with whom we spent a week in Sitges, learned to sail in Baja. 8) He said this to Eddie, who does not deny the lawyer's intuition. 9) It's not just for those who have decided to change their lives. 10) He was a friend of Truman's with whom he had an affair and who encouraged him to write.

• 169

1 7 0 • Chapter 5 The Structure and Function of Phrases and Sentences: Syntax

After examining these sentences, craft a statement that captures the facts about when these authors and speakers use who and who in relative clauses. (Hint: Enclose the relative clause in square brackets and examine the grammatical relationship of the relative pronoun within its sentence.) 5-13. In this chapter we discuss who and who in different ways. On page 155 we said that we would sometimes ignore the traditional distinction between them in many writings and some guarded speech. In a descriptive approach, we rely on forms commonly used by speakers. But in Exercise 5-12, we consider a more traditional analysis associated with some teachers, publishers, and parents, and sometimes referred to as prescriptive grammar, an approach that prescribes certain forms of language as some people think they should be used or have been used. in the past. . Given the common usage criticized by prescriptive grammarians, what position do you think a teacher should take? What do you think students should understand about the role of language prescription in their lives? Do you think this role should be the same for you as a writer and for you as an interlocutor? Should teachers at different levels of education have different approaches to description and prescribing? Explain your position and try to justify it.

Suggestions for further reading • Andrew Carnie. 2002. Syntax: A Generative Introduction (Malden, MA: Blackwell). A challenging introduction to generative theory, with chapters on lexical-functional grammar and head-oriented sentence structure grammar. • Bernard Comrie. 1989. Language Universals and Linguistic Typology: Syntax and Morphology, 2nd ed. (Chicago: University of Chicago Press). A clear discussion of syntactic universals in a variety of languages; The chapters word order, subject, upper and lower case and relative clauses are particularly recommended. • VJ Cook and Mark Newson. 1996. Chomsky's Universal Grammar: An Introduction, 2nd ed. (Malden, MA: Blackwell). A more advanced introduction to formal grammar, with chapters on principles and parameters, X-bar theory, and the minimalist program, going beyond this chapter in accessible steps. • Jim Miller. 2002. An Introduction to English Syntax (Edinburgh: Edinburgh University Press). A basic introduction to English syntax, combining structural and functional considerations; complete and advanced. • Max Morenberg. 2002. Making Grammar, 3rd ed. (New York: Oxford University Press). Focuses on practical understanding of English grammar in traditional terms. • Maggie Tallerman. 1998. Understanding Syntax (New York: Oxford University Press; London: Arnold). A clear introduction to syntax as structure. • Linda Thomas. 1993. Initial Syntax (Malden, MA: Blackwell). A basic introduction to syntax that emphasizes structure but gives some attention to function.

references

Further Reading Comprehensive and accessible reviews of syntax can be found in Aarts (1997) and Radford (1997). The volumes edited by Shopen (1985) contain a wealth of useful material. The publication of the second edition is planned for 2007. Interested readers who have mastered this chapter will likely have access to two excellent chapters from Volume I: "Parts of Speech Systems" and "Passive in World Languages"; Volume II contains valuable chapters on the topics of "complex sentences and complex sentences", "supplement" and "relative clauses". More advanced than any of the readings on the suggested reading list above is Thompson's (1996) treatment of functional grammar. Haegeman (2006) is an excellent textbook showing how to analyze sentence structure.

References • Aarts, Bas. 1997. Syntax and Reasoning in English (New York: St. Martin's). • Hägemann, Liliane. 2006. Syntactic Reasoning: A Guide to Reasoning and Analysis (Malden, MA: Blackwell). • Radford, Andrew. 1997. Syntax: A Minimalist Introduction (Cambridge: Cambridge University Press). • Shopen, Timothy, ed. 1985. Linguistic typology and syntactic description, 3 vols. (Cambridge: Cambridge University Press). • Thompson, Geoff. 2004. Introduction to Functional Grammar, 2nd ed. (London: Arnold; New York: St. Martin).

• 171

6 Meaning Studies: Semantics WHAT DO YOU THINK? • Her colleague Holly, a philosophy student, often asks her friends questions about the language. He recently asked, "Do you think George Washington and the first President of the United States mean the same thing?" What do you say to her? • Your friend Nathan claims that there are no true synonyms. You counter with quick and quick as synonyms for "quick". Nathan wraps things up by pointing out that a fast talker isn't necessarily a fast talker, explaining that they aren't synonymous as you can't always exchange fast and fast. what are you saying now • An uncle, knowing that you study linguistics, asks if there is a term that describes the relationship between word pairs such as uncle and nephew, student and teacher, doctor and patient. "They're not opposites like hot and cold," he says. "But what are they?" What you say? • At a family picnic, he overhears his cousin teasing his four-year-old daughter about a coloring book he took from her. The girl says, "This is mine." His father says, "That's right, it's mine." The girl repeats: "No, it's mine." His father says: "That's what I said: It's mine." "No, it's not," he insists. Then he takes the book and leaves. What is it about the meaning of you and me that makes it possible to poke fun at a four-year-old in this way?

172

introduction

Introduction Semantics is a more familiar term than phonology, morphology, or syntax. Semantics is concerned with meaning, and linguistic semantics is the study of the systematic way languages structure meaning, particularly in words and sentences. In defining linguistic semantics (which we shall simply call "semantics") we must refer to the meaning of words. In everyday interactions, we use the words meaning and meaning in different contexts and for different purposes. For example: The word perplexity means "state of perplexity". Rash has two meanings: "rash" and "skin irritation." Espejo means “mirror” in Spanish. I'm not saying I'm incompetent, just inefficient. The meaning of the cross as a symbol is complex. I wanted to bring you my newspaper, but I left it at home.

What does it mean? Linguists also ascribe different interpretations to the meaning of the word. Because the goal of linguistics is to explain exactly how languages are constructed and used to represent, among other things, situations in the world, it is important to distinguish between different ways of interpreting the meaning of words. A few examples should illustrate why we need to develop a precise way of talking about meaning. Consider these sentences: 1. I went shopping this morning. 2. All dogs are animals.

The truth of sentence 1 depends on whether or not the speaker is telling the truth about visiting the store; nothing in the words of the sentence makes it inherently true. On the other hand, sentence 2 applies because the word dogs describes entities that are also described by the word animals. The truth of 2 does not depend on whether the speaker is telling the truth or not; it just depends on the meaning of the words dogs and animals. Now compare the following pairs of sentences: 3. You are too young to drink. You are not old enough to drink. 4. Matthew spent several years in northern Tibet. Matthew was once in northern Tibet.

3 sentences basically say "the same thing", with the first describing exactly what the second describes. We say that they are synonymous phrases or that they paraphrase each other. In 4 the first sentence implies the second, but not vice versa. If Matthew spent several years in northern Tibet, he must have been there at some point in his life. On the other hand, if Matthew has ever been to northern Tibet, he doesn't necessarily have to have spent several years there.

• 173

1 7 4 • Chapter 6 The study of meaning: semantics

Next consider the following sentences: 5. The single woman is married to a bachelor. 6. My toothbrush is pregnant.

Sentences 5 and 6 are syntactically well-formed, but there is something wrong with their semantics. The meanings of the words in 5 are contradictory: an unmarried woman cannot be married, let alone an unmarried one. Therefore, Proposition 5 is a contradiction. Proposition 6 is not contradictory, but it is semantically anomalous: toothbrushes are not pregnant. To diagnose exactly what is wrong with these sentences, we need to distinguish between contradictory and anomalous sentences. Finally, look at sentences 7 and 8: 7. I saw her bend over. 8. She ate the cake.

Clause 7 can be interpreted in two ways: duck can be a verb referring to bowing quickly (e.g. when walking through a low door), or it can be a noun referring to a type of waterfowl. These word meanings give the expression two different meanings. Since there are two possible readings for 7, it is said to be ambiguous. On the other hand, 8 is not ambiguous but has an imprecise quality, at least when taken out of context. Although we know that the subject of 8 is a woman, we cannot know who it is referring to or what particular cake was eaten, although the expression of the cakes indicates that the speaker has a specific cake in mind. Taken out of context, 8 is vague as certain details are not given; but it is not ambiguous. These observations make it clear that meaning is a complex concept. A sentence can make sense and be true because it states a fact about the world or because the speaker is telling the truth. Two sentences can be related because they mean exactly the same thing or because one implies the other. Finally, when we feel that something is wrong with the meaning of a sentence, it may be because the sentence is contradictory, abnormal, ambiguous, or just plain vague. One of the tasks of semantics is to distinguish between these different ways in which language "means".

Linguistic, social, and affective meaning For our purposes we can first distinguish three types of meaning. Linguistic meaning includes both meaning and reference. Social relevance is what we count when we identify certain social characteristics of speakers and situations from the character of the language used. Affective meaning is the emotional connotation attached to words and statements.

Linguistic meaning Meaning is a very complicated subject, and there is no unified theory on how languages mean. Referential meaning One way to define meaning is to say that the meaning of a word or phrase is the actual person, object, abstract concept, event, or condition to which the word or phrase refers . So the referential sense of Alexis Rathburton would be

The linguistic, social and affective meaning is the person who bears that name. The term Scott's dog refers to the specific domesticated dog owned by Scott. This particular animal can be said to be the referential meaning of Scott's linguistic expression dog, and the dog selected or identified by the expression is its referent. Words are not the only linguistic units that have referential meaning. Phrases also refer to actions, conditions and events in the world. Rahul sleeping on the sofa refers to the fact that a person named Rahul is currently sleeping on a long piece of furniture normally intended for sitting. The referent of the sentence is therefore the state of Rahul in the mobile in question. Meaning Referential meaning is perhaps the easiest to see, but it's not enough to explain how some expressions mean what they mean. On the one hand, not all expressions have referents. Neither a unicorn nor the current King of France have any real point of reference in the real world, but both expressions have meaning. Even leaving aside the social and affective meaning, if the terms had only a referential meaning, the sentences of 9 below would mean exactly the same as those of 10, but no. 9. George Washington was the first President of the United States. George Washington was George Washington. 10. Jacqueline Bouvier married John F Kennedy in 1953. Jacqueline Bouvier married the thirty-fifth President of the United States in 1953.

The 10 sentences do not mean the same thing, and the pair's second sentence seems odd, partly because it would have been impossible to marry the thirty-fifth president in 1953, since the United States did not have a thirty-fifth president. . until 1960. Proper names such as George Washington, Jacqueline Bouvier, and John F. Kennedy constitute a special category, and we can say that the meaning of proper names is the person named, the person to whom the proper name refers. On the other hand, the meaning of expressions such as the first President of the United States and the thirty-fifth President of the United States cannot be reduced to their reference persons. Consider the sentences of 11:11. Al Gore almost became the forty-third President of the United States. Al Gore almost became George W. Bush.

Obviously these terms do not mean the same thing, although the terms George W. Bush and the forty-third President of the United States have the same reference. Therefore, the sentences in 9 do not have an identical meaning. In general, we cannot equate the meaning of an expression with the expression's referent. We say that expressions have "senses", and any theory of how language means must take into account the meaning of the sense.

Social Meaning Linguistic meaning is not the only type of meaning that language users communicate with each other. Consider the following sentences: 1st 2nd 3rd 4th

So I told him, "You can't do anything right." Is he a doctor here? Are you visiting us during the holidays? Enjoy your meal!

• 175

1 7 6 • Chapter 6 The study of meaning: semantics

In addition to representing actions, states, and mental processes, these phrases convey information about the identity of the person who uttered them or the situation in which they were uttered. In Figure 1, the use of the verb say with the first person singular pronoun I indicates something about the speaker's social status. In 2, the it form, in which some other varieties are used, indicates a speaker of an ethnic variety of English (African-American English). In 3, the pronoun y'all identifies a particular regional dialect of American (Southern) English. Finally, the choice of words in 4 indicates that the utterance was made in an informal context. Social status, ethnicity, regional origin and context are all social factors. Therefore, in addition to linguistic meaning, every utterance also conveys social meaning, not only in the sentence as a whole, but also in word choice (y'all and chow) and pronunciation (gonna or nothin').

Affective meaning Besides linguistic and social meaning, there is a third type of meaning. Compare the following examples: 1. Tina, who always boasts of her two PhDs, lectured me all night about Warhol's art. 2. Tina, who has two PhDs, gave me a fascinating insight into Warhol's art last night.

Since these two phrases can be used to represent exactly the same event, we can say that they have a similar referential meaning. On another level, however, the information they convey is different. Sentence 1 gives the impression that the speaker thinks Tina is a pretentious bore. Sentence 2, on the other hand, shows that the speaker finds it interesting. The "attitude" of the speaker in these utterances is therefore different. Word choice is not the only way to communicate feelings and attitudes about expressions and context. Sentences that differ only in emphasis or intonation form a strong contrast. This sequence of words can be interpreted in several ways depending on the intonation: Erin is very smart.

The phrase can, of course, be spoken without stressing a specific word, in which case it is taken literally as a comment acknowledging Erin's intelligence. But when the words really and cleverly are overemphasized, the phrase sarcastic can be interpreted to mean exactly the opposite. Intonation (often accompanied by appropriate facial expressions) can be used as a means of communicating attitudes and feelings, and can substitute for the literal meaning of a sentence. Consider one last example. Suppose Andy Grump, Sara's father, addresses her like this: Sara Grump, how many times have I asked you not to channel?

There would be reasons to look beyond words for the "meaning" of this unusual form of address. Mr. Grump can call his daughter Sara Grump to show his distress, like in this example. By addressing her as Sara Grump instead of the usual Sara, she expresses frustration and annoyance. His choice of name therefore shows that he is upset. Compare the tone of this sentence with a similar one in which he calls her honey.

Meaning of words, phrases, and expressions The level of meaning that conveys the language user's feelings, attitudes, and opinions about specific information or the current context is called affective meaning. Affective meaning is not exclusive to sentences: words like Ai! and live! They obviously have affective meanings, as do words like funny, cute, and nasty. Even the most common words like dad, democracy and old can evoke certain emotions and feelings in us. The difference between synonymous or nearly synonymous word pairs such as vagabond and homeless is essentially a difference at the affective level. In this particular pair, Boom has a negative effect while Homelessness is neutral. Little is known about how affective meaning works, but it is of great importance in all verbal communication. From our discussion so far, you can see that meaning is not a simple concept, but a complex combination of three aspects: • Linguistic meaning, including referential meaning (the real object or concept that is selected or described by an expression) and the meaning of Sense • Social Significance: Information about the social nature of the language user or the context of the utterance. • Affective meaning: how the language user thinks about the current content or context

The linguistic meaning of an expression is often referred to as denotation, as opposed to connotation, which includes both social and affective meaning. This chapter focuses mainly on linguistic meaning, the traditional area of semantics, but occasionally we refer to the three-way distinction. Social significance is explored in Chapters 10 and 11.

Meaning of Words, Sentences and Expressions Meaning of Words and Sentences We have discussed words and sentences as the two units of language that have meaning. Content words—principally nouns, verbs, prepositions, adjectives, and adverbs—have meaning to the extent that they relate to concrete objects and abstract concepts; they are identified as characteristic of certain social, ethnic and regional dialects and particular contexts; and convey information about the feelings and attitudes of language users. Function words like conjunctions and determiners also have meaning, albeit in a slightly different way than content words, as you'll see later in this chapter. Like single words, sentences have social and affective connotations. However, studying the meaning of words is different from studying the meaning of sentences because the units are of different types. For a sentence to convey meaning, we must rely on the meaning of the individual words it contains. How we approach the task of deriving the meaning of sentences from the meaning of words is a complex matter. One obvious hypothesis is that the meaning of a sentence is simply the sum of the meanings of its words. To see that this is not the case, consider the following sentences, in which the individual words (and thus their added meanings) are the same: The lion licked the trainer. The trainer licked the lion.

• 177

1 7 8 • Chapter 6 The study of meaning: semantics

Obviously, the sentences refer to different events and therefore have different linguistic meanings. This is conveyed by the fact that the words are arranged differently in the sentences. So we cannot say that all we have to do to get the meaning of a sentence is to add the meanings of its parts. We must also consider the semantic role assigned to each word. By semantic function we mean things like who did what to whom, to whom, and for whom. In other words, the semantic role of a word is the role played by its referent in the action or state of being described by the sentence. Sentence semantics deals with the semantic roles and relationship between words and constituents within a sentence. Scope of Word Meaning While it is important to distinguish between word meaning and sentence meaning, both interact on many levels, as this sentence shows: He can leave tomorrow when he finishes his thesis.

In this sentence the individual words can, tomorrow and if have meanings: can denote permission or possibility; tomorrow specifies a future time unit starting at midnight; and a condition is displayed. But the impact of these words goes beyond the sentences they appear in and affects the meaning of the entire sentence. In fact, if we replace May with will, the phrase takes on a whole different meaning: he'll leave tomorrow when he's finished his last job.

The term can denote permission or possibility, while the term simply describes a future event. Therefore, it can affect the meaning of the entire sentence. The meaning of the word can is the entire sentence. This also applies to tomorrow and if. These examples illustrate that the meaning of words and the meaning of sentences are closely related.

Try it yourself: limit yourself to the following sentences. Using 1 as a template, provide a sentence that illustrates the scope of Solo for 2 and 3. If the scope of Solo is not clear, provide the alternative "cf". Sentences, each of which is unique. 1. He just wants you to be happy. (cf. He only wants you to be happy; he doesn't care about others.) 2. Only she wants you to win. 3. He just wants to talk to his daughter.

Meaning of Sentences In addition to words and phrases, there is a third entity that carries meaning; However, we may not see it as clearly because we take it for granted in everyday interactions. Consider this statement: I now pronounce you husband and wife.

This phrase can be uttered in many different circumstances: (1) by an official at a ceremony speaking to a couple about to be married in the presence of their family and friends, or (2) by an actor disguised as an official , who speaks to two actors before a congress

Semantics of lexical gations of Hollywood extras staged by a director making a soap opera. First and foremost, I now declare that you, husband and wife, create a marriage for the couple who wish to marry. However, the same statement has no bearing on an actor's marital status on location. Thus the circumstances of the utterance produce different meanings, although the linguistic meaning of the sentence remains unchanged. Therefore, it is necessary to know the circumstances of a statement in order to understand its effect or power. We say that the sentence uttered in the wedding context and the sentence uttered in the film context have the same linguistic meaning, but they are different utterances, each with their own uttered meaning. The difference between the meaning of the sentence and the meaning of the phrase is best illustrated by asking Can you close the window? There are at least two ways a recipient might respond to this question. One would be to say "Yes" (meaning "Yes, I can physically close the window") and do nothing about it. This is the interpretation of the "sage"; not, of course, in the way this question is usually meant. Another form of recipient response would be to get up and close the window. Obviously, these interpretations of the same question are different: the know-it-all interpretation treats the question as a request for information; The second interpretation treats it as a call to action. To describe the difference between these interpretations, we say that they are different statements. The sentence semantics does not deal with the meaning of the expression. (Utterances are the subject of research in another branch of linguistics called pragmatics, which is the subject of Chapters 8 and 9.) One of the premises of sentence semantics is that sentences must be separated from the context in which they are uttered; Words which sentences and expressions are to be distinguished. To experienced users of the language, this premise may seem odd and counterintuitive, since so much of the meaning depends on context. The point is not to dismiss context as unimportant, but to recognize that sentences can have meaning regardless of context, while the meaning of the statement depends crucially on the circumstances of the statement. Semantics is the branch of linguistics that studies the meaning of words and sentences, generally ignoring context. In contrast, pragmatics pays less attention to the relationship of word meaning to sentence meaning and more to the relationship of a statement to its context.

Lexical Semantics The lexicon of a language can be viewed as a compendium of all its words. Words are sometimes called lexical items or lexemes (the ending -ema as in phonemes and morphemes). The branch of semantics that deals with the meaning of words is called lexical semantics. Lexical semantics studies the relationships between the meanings of words. For example, ask about the relationship between the words man and woman on the one hand and human on the other. How are the adjectives big and small related to each other in the same way as the pair dark and light? What is the difference between the meaning of the words like always and never and the meaning of the words how often and rarely? What do language users really mean when they say that a dog is "a kind of" mammal? Lexical semantics investigates such questions. It is the study of how the lexicon is organized and how the meanings of lexical items are related, and its main goal is to model the structure of the lexicon by categorizing the types of relationships between words. Lexical semantics focuses on linguistic meaning.

• 179

1 8 0 • Chapter 6 The study of meaning: semantics

Semantic Fields Consider the following word combinations: 1. cup, mug, wine glass, glass, plastic cup, goblet 2. hammer, cloud, tractor, glasses, leaf, justice

All of the words in Group 1 denote concepts that can be described as "drinking vessels," while the words in Group 2 denote concepts that have nothing in common. The words in Set 1 form a semantic field, a set of words with an identifiable semantic affinity. The next set is also a semantic field whose words all relate to emotional states: angry, sad, happy, exuberant, depressed, anxious.

So we see that words can be divided into groups according to their meaning. In a semantic field, not all lexical items necessarily have the same status. Consider the following sets, which together form the semantic field of color terms (of course there are other terms in the same field): 1. blue, red, yellow, green, black, purple 2. indigo, saffron, royal blue. Aqua Navy Bisque

The colors referred to by words in sentence 1 are "more common" than those described in sentence 2. They are considered less marked members of the semantic field than those in sentence 2. Less marked members of a semantic field are generally easier to identify. and remember what shaped the members the most. Children learn the term blue before learning the terms indigo, royal blue, or aquamarine. Often a less marked word consists of a single morpheme (blue contrasted with royal blue or aqua) as opposed to more heavily marked words. The less marked member of a semantic field cannot be described with the name of another member of the same field, while the more marked members can be described in this way (indigo is a type of blue, but blue is not a type of indigo). ). Terms with fewer tags also tend to be used more often than terms with higher tags; For example, blue is much more common in conversation and writing than indigo or aquamarine. (In the Brown Corpus of a million words in American English, there are 126 examples of blue, but only one for indigo and none for aquamarine.) Blue describes a wider range of colors than either indigo or aquamarine. Finally, the weaker words are not the result of metaphorically using the name of another object or concept, while the stronger words often are; For example, saffron is the color of a spice that gave the color its name.

Try it yourself: rust, silver, orchid and champagne are all part of the color connotation and you can easily see where these color terms came from. Fruits, flowers, gems, and other natural objects are notable sources for terms in this semantic field. Can you identify five additional color terms that come directly from the name of a real-world object of that color?

Using our understanding of the semantic field and markup, we now proceed to identify the types of relationships between words. We will see how the words form a semantic field

Lexical semantics can have different types of relationships with each other and with other words in the lexicon, and we will classify these relationships.

Hyponymy Consider again this series of unlabeled color terms: blue, red, yellow, green, black, purple. What they have in common is that they relate to colors. We say that the terms blue, red, yellow, green, black, and purple are hyponyms of the term color. A hyponym is a subordinate and specific term whose referent is included in the referent of a superordinate term. Blue is a kind of color; Red is a kind of color and so on. They are specific colors and color is the general term for them. We can illustrate the relationship with the following diagram, where the lower terms are the hyponyms (hypo- means "below"). The topmost term, color in this case, is called the parent term (technically the hyperonym). color

(Superior)

blue red yellow green black purple

(Hyponyme)

Another example is the term mammal, which reference includes the references of many other terms. mammal

cow dog

Gatos

(Superior)

monkey human whale

(Hyponyme)

The relationship between each of the lower terms and the upper term is called hyponymy. Hyponymy is not limited to objects like mammals, or abstract concepts like color, or even nouns. Hyponymy can be identified in many other areas of the lexicon. For example, the verb to cook has many hyponyms. oven

(Superior)

toast cook roast grill roast bake microwave

(Hyponyme)

Not all sets of hyponyms have a parent term. For example, uncle and aunt form a lexical field because we can identify a common trait in their meanings. However, there is no term in English that specifically refers to uncles and aunts (i.e. siblings of parents and their spouses). ? uncle

Many

(parent) (hyponyms)

Conversely, some other languages have a higher-order term for the corresponding field. In Spanish, the plural term tíos can include aunts and uncles, and the Spanish equivalents of the terms tio and tia are therefore hyponyms for tíos.

• 181

1 8 2 • Chapter 6 The study of meaning: semantics

Although hyponymy occurs in all languages, the concepts that words have in hyponymous relationships vary from language to language. In Tuvaluan (a Polynesian language), the superordinate term ika (roughly "fish") has as hyponyms not only all terms relating to animals that English speakers would recognize as fish, but also terms for whales and dolphins (that English speakers recognize) . as mammals) and for sea turtles (which are reptiles). Of course, we are dealing here with popular classifications, not scientific classifications. ika

it hurts 'hai'

aaseu 'Mackerel'

Nofu 'Steinfisch'

"Turtle" background

Table 'Dolphin/Whale'

As such, there are differences between languages as to the exact nature of certain hyponymous relationships. In a semantic field, hyponymy can exist on more than one level. A word can have a hyponym and parent term, like the blue in Figure 6-1. Because they refer to different "types" or "shades" of blue, the terms turquoise, aquamarine, and royal blue are hyponyms for blue. Blue, in turn, is a hyponym for color. Thus we have a hierarchy of related terms through hyponymous relationships. Similar hierarchies can be set up almost indefinitely for many semantic fields. In the field of cooking, roast has hyponyms in the terms sauté, sauté, and fry, and is itself a hyponym for cooking. The further down we go in a hierarchy of hyponyms, the more pronounced the terms are: cooking is relatively unlabeled; Eintopf is clearly more pronounced. Midground frying is less pronounced than searing, but more pronounced than boiling.

FIGURE 6-1 Color

Rojo

green

Therefore

violet

Turquoise

blau

gelb

seawater

royal blue

Examples of complex hyponym relationships abound in the field of folk-biological classification, as shown in Figure 6-2 on page 183. Note that the term animal occurs on two levels. English speakers use animal for at least two different references: (1) animals as opposed to plants and rocks, and (2) animals (usually mammals other than humans) as opposed to humans, birds, and insects. It is not uncommon for a word to have different meanings at different levels of a hyponymic hierarchy.

Figure 6-2 Lexical semantics of animals

Vogel

am Board

Error

Corrion Hawk

quail

chicken

there

Mensch

Tier

dog

Gatos

frei

Cobra

pollo

Hyponymy is one of several types of relationships that language users use to organize the lexicon. It is based on the notion of inclusion: if the referent of term A (eg color) includes the referent of term B (eg red), then term B (red) is a hyponym of term A (color). Hyponymy is important in everyday conversation: we use it every time we say "B is some kind of A" (red is some kind of color) and for tasks like using a thesaurus organized by relationships.

Part/Whole Relationships A second important hierarchical relationship between words is found in pairs such as hand and arm or room and house. In each pair, the referent of the first term is part of the referent of the second term. However, a hand is not "a kind of" arm, and thus the relationship between hand and arm is not hyponymous. Instead we call this the part/whole relationship. Part/whole relationships are not just a property of word pairs: hand, elbow, forearm, wrist, and several other words have a part/whole relationship with arm. Other important examples of part/whole relationships are words like second and minute, minute and hour, hour and day, day and week, neither of which could be described without reference to the fact that one is a subdivision of the other. Figure 6.3 illustrates the difference between a part/whole relationship and a hyponymous relationship for the word eye.

FIGURE 6-3 Partial/Full Ratio Face

boca

nose cheek

ojo

(eye 'part of face')

body of hyponymic relation

should

Herz

Lunge

ojo

(eye 'a kind of organ')

• 183

1 8 4 • Chapter 6 The study of meaning: semantics

Synonymy Two words are considered synonymous if they mean the same thing. The terms movie, film, film, and motion picture have the same references in the real world and are often considered synonymous terms. To approach the notion of synonymy more formally, we can say that the term A is equivalent to the term B if every referent of A is a referent of B and vice versa. For example, if every movie is a movie and every movie is a movie, the terms movie and film are synonymous. The "inverse" is important: without it, we would be defining hyponymy. You may be wondering why speakers of a language bother to maintain synonyms when they are just adding redundancy to the lexicon. English has many pairs of synonyms, such as cloudy and fuzzy, help and support, oblique and oblique (the result of English borrowing the second term in each pair from French or Latin). When we claim that two terms are synonymous, we generally base that judgment on the linguistic meaning alone. Although film, cinema, film and cinema share the same linguistic meaning, they differ in social and affective meaning. The film may seem appropriate for classic cinema or auteur cinema; it's a rather intellectual term. Realize that film is mostly used in informal contexts, while film is more traditional or industry related. Therefore, we can consider the terms as synonyms if we indicate that we are only considering the linguistic meaning. On the social and affective level, however, they are not synonymous. In fact, there are very few true synonyms in the lexicon. More often than not, terms that appear synonymous have different social and affective connotations. Even if we limit the meaning to the linguistic meaning, words that appear synonymous at first glance often refer to slightly different conceptual contexts or are used in different situations. The adjectives quick, quick, and swift can be used interchangeably in reference to a person's walking speed, but a fast speaker (a "slippery or tricky person") is distinct from a "fast speaker"; some people live in the fast lane, not the "fast lane"; and swift is the more appropriate term to describe a spirit or a look, while swift is the usual term when referring to a person's pace, particularly metaphorical steps, such as in learning to write or do arithmetic. In these circumstances, is it correct to say that these adjectives are synonymous?

Try it yourself: For each, provide a synonym or near-synonym in the same part of speech. Adjective: keen (sharp), ancient, juvenile, fast, mute, strong, fertile, naked, small, drunk: surround, kidnap, stutter, seek, praise, clothe, shake, disturb, commit, initiate

The fact that there are few true synonyms in a language's lexicon reflects the general tendency of language users to make the best of what is available to them. When two terms have the same reference, the meaning of one is often modified to express differences.

Lexical semantics involves linguistic, social, or affective meaning. Although true synonymy is rare, the term is useful as it helps describe similarities between the meanings of different terms in the lexicon.

Antonymy The word antonymy derives from the Greek root anti- ("opposite") and denotes the opposite in its meaning. Unlike synonymy and hyponymy, antonymy is a binary relationship that can characterize a relationship between only two words at a time. The terms A and B are antonyms if, if A describes a referent, then B cannot describe the same referent, and vice versa. Prototypical antonyms are pairs of adjectives that describe opposite concepts: tall and short, wide and narrow, hot and cold, married and single, alive and dead. However, antonyms are not limited to adjectives. The nouns man and woman are also antonyms, since a person cannot be described with both terms at the same time. Always and never form a pair of antonymous adverbs: they have mutually exclusive references. The verbs love and hate can also be seen as antonyms as they refer to mutually exclusive emotions. Antonymy is therefore a binary relationship between terms with complementary meanings. Intuitively you can see a difference between the big and small pair of antonyms and the single and married pair of antonyms. The first pair denotes relatively subjective ideas. I would agree that blue whales are large mammals and mice are small mammals, but whether German Shepherds are large or small dogs depends on your perspective. A Chihuahua owner will tell you that German shepherds are large, but a Great Dane owner may judge them as small. Adjectives like big and small also have superlative and comparative forms: blue whales are the largest mammals; German Shepherds are larger than Chihuahuas but smaller than Great Danes. Antonym pairs with these properties are called gradable pairs. Unlike the big and the small, the single and the married are mutually exclusive and complementary. A person cannot be single and married at the same time. With respect to marital status, a person cannot be qualified with a term containing non-single or married as a hyponym; thus, single and married are complementary. Also, single and married generally cannot be used in the comparative or superlative sense (it is impossible for anyone to be legally a "singer" than another single person). The pair is an example of non-gradable antonyms (sometimes called complementarity). So there are two types of antonyms: gradable and non-gradable. If the terms A and B are gradeable antonyms, and if A can be used to describe a particular referent, then B cannot be used to describe the same referent, and vice versa. If A and B are non-gradable antonyms, the same condition applies, with an additional condition: if A cannot describe a referent, then that referent must be describable by B, and vice versa. Thus man and woman, married and single, alive and dead can be seen as immutable antonyms, while hot and cold, love and hate are always and never changeable. Usually gradable antonyms will have words that describe intermediate stages: sometimes, rarely, occasionally, often there are gradations between always and never. As you acknowledge, the distinction between qualifiable and unqualifiable antonyms is sometimes blurred by users of the language. In English, for example, it is reasonable to assume that everything that is alive is not dead and that everything that is dead is not alive, and so the adjectives dead and alive form an ungradable pair. However, we have expressions like half dead, almost dead, and more dead than alive. Such expressions indicate

• 185

1 8 6 • Chapter 6 The study of meaning: semantics

in some contexts we see alive and dead as gradeable antonyms. However, the distinction between gradable and non-gradable antonyms is useful because it describes an important distinction between two types of word relationships.

Try it yourself: Give an antonym for each in the same part of speech: Adjective: appetizing (unpleasant), frank, outward, arrogant, superficial, cold, sated, fertile, funny Noun: retrospective (predictive), inward, friendship, failure , freedom, utility, chaos, certainty, fertility Verb: to ignite (extinguish), reveal, remember, dishonor, ignore, appear, expand, purify, fold

Finally, antonyms generally don't have the same status when it comes to awards. For example, if you're asking about the weight of an object, ask What is its weight? and no how bright is it? - unless you already know the object is light. Note also that the noun weight, which describes both relative weight and relative lightness, is associated with heavy rather than light (as in the expressions carrying much weight and throwing weight). From the pair of antonyms heavy and light, heavy is more neutral than light and therefore less pronounced. Similarly, tall is marked as smaller than short, hot as less than cold, and married as less than single (we say marital status, not "single status"). While there are some differences between languages as to which word in a pair is considered less pronounced, there is a surprising consistency from language to language.

Inversion Another important relationship invokes the notion of opposition, although it does so in a different way than antonymy. Consider the relationship between wife and husband. If A is B's husband, then B is A's wife. Thus, wife is the opposite of husband and vice versa. The inversion denotes a reciprocal semantic relationship between pairs of words. Other examples of inverse pairs include terms denoting many other kinship relationships, such as B. grandchildren and grandparents or child and parent; Terms describing professional relationships, such as employer and employee or doctor and patient; and terms denoting relative positions in space or time, such as B. up and down, north and south or before and after. Reversed couples can combine with other types of opposition to form complex relationships. The antonymous father:mother is inversely related to the antonymous son:daughter. In general, conversational pairs denote relationships between objects or between people. Some inverse relationships are a bit more complex. For example, the verb give requires a subject and two objects (She gave him the book). The opposite of giving is receiving, except that the relationship is not an "inversion" of direct subject and object like kissing and kissing would be (Smith kissed Jones vs. Jones was

lexical semantics, kissed by Smith) nor a mutual subject/possessor relationship as husband and wife; rather, the relationship is between the subject and the indirect object. Siddhartha gave Jessie a present. Jessie received a present from Siddhartha.

Other pairs of words with a similar relationship are lend and borrow and buy and sell. Note that Rent is its own inverse in American English. Eve rents an apartment from Adam. Adam rents an apartment from Eve.

If there is a risk of confusion, the preposition out can be added to rent meaning "to lend money". In British English, this rental feeling is described by the verb let (flat to let). In some languages, a single word is used for "buy" and "sell". These facts suggest that conversation is an intuitively recognizable relationship.

Polysemy and Homonymy Two other terms closely related to the basic types of relationships are polysemy and homonymy. Unlike the terms discussed above, polysemy and homonymy refer to similarities rather than differences between meanings. A word is polysemic (or polysemic) if it has two or more related meanings. For example, the word plain can have several related meanings, including: 1. "plain, plain" (plain English) 2. "no decoration" (plain white shirt) 3. "not pretty" (plain jane)

Homographs have the same spelling but different meanings (and pronunciations), such as dove 'a kind of bird' and dove 'past tense of diving', or behavior as a verb and behavior as a noun, with the verb having the main stress on the second syllable and the noun has in the first syllable. Homophones have the same pronunciation but different meanings: sea and lake, so and sew, two and also, smooth and flat, flower and flour, boar and bore, bear and naked or eye, me and yes. Words are homonyms when they have the same written or spoken form but have different meanings. A narrower definition of homonym limits the term to groups of words that are both homographic and homophonic, such as Ribera and Caja de Ahorros, or the adjective still 'still' (still waters) and the adverb still 'still' (still sick). Languages exhibit varying degrees of polysemy and homonymy in their lexicons. A language like Hawaiian, which has a limited set of possible words due to its phonological structure, has much more homonymy than English (see “Sequence Constraints” in Chapter 4 and Figure 7-3 on page 202). A difficulty arises in distinguishing between homonymy and polysemy: how do we know if we have separate lexical items and not a single word with different meanings? Consider it easy. How would we know if the three adjective meanings ('simple', 'no decoration', 'not pretty') are different words that sound the same? Using spelling as a criterion is misleading: many sets of words are different but have the same spelling, for example, the noun sounds "noise" and the noun sounds

• 187

1 8 8 • Chapter 6 The study of meaning: semantics

'water canal' or margin 'financial institution' and margin 'river bank'. However, the problem is relevant to anyone who wants to organize or use the entries in a dictionary (in which different meanings of the same word are grouped under a single entry, but each homonymous form has its own unique entry). There is no easy solution. If there is a clear distinction between polysemy and homonymy, there must be multiple criteria, none of which would be sufficient on their own, and some of which may lead to different results. We have already ruled out spelling as an unreliable criterion. A modestly reliable criterion is the historical origin or etymology of a word. We can assume that there are two words of the form som that correspond to the two meanings given above, since they come from different historical roots. Likewise, the word bank, meaning "financial institution," is a loan word from French, while bank, meaning "river bank," is of Scandinavian origin. The various antonyms and synonyms of a word provide different criteria for distinguishing between polysemy and homonymy. Simple in the sense of "simple, clear" and simple in the sense of "without decoration" share a synonym in simple and an antonym in complex. This fact suggests that they are actually two meanings of the same ambiguous word. No common synonym or antonym can be identified for the two meanings of the sound, as shown in Figure 6-4.

FIGURE 6-4 Word

Feeling "light, clear".

airplane

equivalent

Antonym

simply

Complex

'no decoration' 'water track' sound 'noise'

Finally, we can ask if there is a resemblance between the different meanings of what appear to be the same word. The two meanings of flatness above can be characterized as "without complexity", suggesting that they are related, but there is no such overarching description for the sound "surface of water" and the sound "rushing". polysemic, while the two sound senses reflect lexical elements of the same name. (Of course, other meanings of simple may or may not belong to single words.) While these criteria help distinguish between polysemy and homonymy, they are not infallible. It is often difficult to decide whether a given pair of same and similar word forms are separate homonyms or simply the same polysemic word with different meanings. Although homonymy and polysemy can be distinguished as distinct terms, the boundary between them may not be clear in specific cases.

Metaphors The difficulty in distinguishing between polysemy and homonymy arises in part from the fact that language users often use words metaphorically. A traditional notion of metaphor sees it as expanding the use of a word beyond its primary meaning

Lexical semantics describes referents that share similarities with the word's primary referent. For example, the word eye can be used to describe the hole in the blunt end of a needle, the flower bud of a potato, or the center of a thunderstorm. The similarities between these reference points and the main reference point of the word eye are their rounded shape and their more or less central role or position in a larger form. People often create new metaphors, and once a metaphor is accepted, speakers tend to see the metaphorical meaning as separate from its main meaning, such as foreigners. Language. Therefore, it is difficult to determine whether it is one word with two meanings or two words with different but metaphorically related meanings. Metaphors appear constantly in everyday language and writing because they are a fundamental part of our thinking. The following examples are taken from a typical newspaper front page: Tennis star Serena Williams survives her first matches. The dollar falls sharply. His speech was the trigger for a new popular uprising.

In the first example, the verb breeze obviously doesn't mean literally; It is used to give the impression that Williams has won games effortlessly, much like a breeze can easily blow across a tennis court. Likewise, the underlined words in the other two sentences are to be interpreted as metaphors whose effectiveness depends on our ability to recognize that in some contexts the words are not to be taken literally. (The mechanisms we use to figure out when a word should be interpreted metaphorically are discussed in Chapter 9.) Metaphors are not formed randomly. For example, consider the following metaphors related to the notion of time: I hope to see you again this weekend. Experts do not expect inflation to rise in the near future. He carries old grudges from his youth with him. From time to time we need to look back and see the lessons history has taught us.

A pattern emerges from these examples: In English, we construct time metaphors as if we were physically moving through time into the future. So the future is ahead in the first two examples. Metaphors related to the past use words related to what lies beyond, as in the last two examples. Metaphors that go against this pattern would sound very strange: *See you again this weekend. * Drag out old grudges from his youth.

Another principle that governs the formation of metaphors is: "Ideas are objects that can be felt." So that they can be smelled, felt and heard. Your suggestion smells like fish. I couldn't understand what they were trying to prove. I would like your opinion on whether my plan seems reasonable.

• 189

1 9 0 • Chapter 6 The study of meaning: semantics

Writers and critics often speak of the "cooking" of the writing process. I let my manuscript simmer for six months. Who knows what kind of story he'll make up! His latest book was little more than an unfinished hodgepodge of earlier works.

"In the heart emotions are experienced" is a general tenet on which our metaphors for emotions are based. It is with a heavy heart that I inform you of his death. You must not talk lightly about this tragedy. The rescuers received the heartfelt thanks from the survivors.

The construction of metaphors therefore follows predefined patterns. Most of the metaphors discussed so far are relatively conventionalized, that is, they are common in speech and writing because they are predefined. But language lends itself to creative activity, and language users do not hesitate to create new metaphors. But even when we create our own metaphors, we must follow the principles that govern conventional metaphors. In English, metaphors relating to time obey the convention of "going through time into the future". Some metaphorical patterns are common in the languages of the world. For example, the word "eye" is used metaphorically in many languages to refer to rounded objects, such as the bumps on a potato, and the central portion of an object, such as the center of a thunderstorm. But other principles of metaphor vary from language to language. For example, in many languages, the heart is not the seat of emotions. Polynesian languages like Samoan and Tahitian treat the stomach as the metaphorical seat of emotions. It is likely that some of these principles reflect the worldviews of different cultures. The exact workings of the connection between culture and language are not yet fully understood. Metaphors play an important role in the development of cognitive linguistics, which has shown how important metaphorical thinking is in our language, our thinking and our life in general. Rather than being primarily a poetic device, some researchers view our cognitive system as "fundamentally metaphorical in character" and with profound implications for everyday life. Lakoff and Johnson wrote: How we think metaphorically is important. It can determine questions of war and peace, economic and legal decisions as well as the worldly decisions of everyday life. Is a military attack "rape," "a threat to our security," or "defending a population against terrorism"? The same attack can be designed in any of these ways, with very different military consequences.

Lexical semantics: Finding relationships in the lexicon Hyponymy, part/whole relationships, synonymy, gradable and non-gradable antonymy, conversation, polysemy, homonymy and metaphor: Lexical semantics is primarily concerned with finding relationships in the lexicon of languages. A word's semantic relationships are, in a sense, part of its meaning: the word cold can be defined as an adaptable antonym of hot, because it has the generic term sensation of warmth, and because it is more prominent than hot but less pronounced than cold and freezing. In

Function Words and Categories of Meaning If we know how the meaning of a word interacts with the meaning of other words, we can begin to understand its meaning. Lexical semantics, of course, don't explain the difference in meaning between such disparate words as Gorilla and Doubtful. For lexical semantics to be useful, it must be applied to specific areas of the lexicon where word meanings share common features. Hence the notion of the semantic field becomes useful. When the word gorilla is placed in its proper semantic field, its relationship to the chimpanzee and the great ape can be explored. Similarly, the word doubtful can be contrasted with certain, probable, probable and other words expressing probability or certainty. The different types of relationships described above are the most basic tools of lexical semantics. They are fundamental because one type cannot be characterized by another type. For example, an antonymous relationship between two words cannot be explained by hyponymy, part/whole relationships, synonymy, conversation, or metaphor.

Function words and categories of meaning The lexicon does not consist exclusively of content words such as father, dove, stew and democracy, which refer to objects, actions or abstract concepts. It also includes function words like the conjunctions if, but, and or; the determiners a, these and these; and helpers can, should and want to. The role of these categories is to show grammatical connections.

Tense and modality Many categories of meaning are associated with function words and function morphemes. Linked morphemes can denote different categories of meaning in English, including number (toys vs. toy) and tense (walked vs. walk). In other languages, the same categories are not expressed by connected morphemes, but by separate words. In Tongan, the functional word ʔoku denotes the present, while naʔe denotes the past. 'Well

ʔalu e

gift goes

feinʔeiki to

the woman

Pro

Kolo city

The woman goes into town. naʔe ʔalu e go past

feinʔeiki to

the woman

Pro

Kolo city

The woman wanted to go into town.

Whether the tense is expressed by linked morphemes or separate lexical items is not important to semantics. Importantly, there is a semantic category that affects the meaning of sentences in both Tongan and English. Semantic categories such as tense are mediated by function words and function morphemes, but their scope extends beyond the component in which they occur. The meaning of a tense morpheme affects the whole sentence because the tense of the verb determines the tense of the whole sentence. The tense category (and other similar semantic categories) thus relates to both the meaning of the word and the meaning of the sentence. Modality, or mood, is a category through which speakers can convey (referred to as epistemic modality) or express their attitude towards the truth or reliability of their statements

• 191

1 9 2 • Chapter 6 The study of meaning: semantics

Obligation, permission or suggestion (called deontic modality). The sentences in the following pairs differ in their epistemic modality: 1. He has probably already left town. (Probability) She's already left town. (statement) 2. Harry must have been very tall when he was young. (guess) Harry was very tall when he was young. (Confirmation) 3. You can come to the party. (possibility) They come to the party. (Affirmation)

And those of the following couples differ in their deontic modality: 4. He must come tomorrow. (Command) Come tomorrow. (Confirmation) 5. You can take the dishes. (Permission) They're clearing the dishes. (Explanation)

The two types of modalities are linked, as evidenced by the fact that the same words (e.g. must and may) can denote both types, depending on the context. The modality can be expressed by auxiliary verbs such as may, should, or must (which are called modal auxiliary verbs); through modal verbs such as order, accept and allow; by modal adverbs such as possibly or certainly; and in some languages by suffixes to verbs or nouns. Such affixes are common in Native American languages, some of which may have extremely complex systems of modal and particle affixes.

Reference A noun phrase in an utterance may or may not have a corresponding real-world entity. Referencing refers to the ability of linguistic expressions to refer to entities in the real world. When someone says I read a new James Joyce biography last weekend, the phrases me and a new James Joyce biography are referring to real-world entities. On the other hand, if someone says they would like to find a James Joyce biography short, there exists in the speaker's mind a real-world entity that corresponds to me, but no James Joyce biography short. (A brief biography of James Joyce may exist, but in this sentence the speaker has no real-world entity in mind to which the phrase refers.) In the following examples, note the difference in reference for the various uses of a certain sentence. . In Examples 1, 3 and 5, the underlined sentences are unrelated; we say that they are not referential or that they are not referential. In 2, 4, and 6, the same terms have real-world references; they are referential. 1st 2nd 3rd 4th 5th 6th

Can you recommend a good western for kids? (non-referential) Last night I saw a good western on HBO. (Frame) She would buy a new Ford Bronco if she could find one for sale. (non-referential) He was testing a new Ford Bronco that he liked. (Reference) I'm looking for the best Chinese restaurant in town. (not referential) On Tuesday I ate at the best Chinese restaurant in town. (Notice)

As these examples show, reference is a property, not of words or phrases as such, but of linguistic expressions as they appear in actual language. The same sentence can be referential in one statement and non-referential in another. Also note that the reference cannot be

Function words and semantic categories equated with definition, a subject to which we shall return later. (The reference is explored further in Chapter 8.)

Deixis The word deixis comes from the Greek adjective deiktikos, meaning “to show, to indicate”. Deixis is the marking of the orientation or position of entities and events with respect to specific points of reference. Consider the following sentence addressed to a waiter by a restaurant customer while pointing to menu items: I want this dish, this dish, and this dish.

In order to interpret this statement, the waiter must have information about who I am referring to, at the moment the statement is made, and what the three noun phrases in that dish refer to. We say that I is a deictic expression, as is the present tense of the verb and the three noun phrases on this tablet. Our ability to interpret them allows us to interpret the sentence. Deixis consists of three semantic terms, all of which refer to the orientation or position of events or entities in the real world. Personal deixis is usually expressed by personal pronouns: I versus you versus he or she. Spatial deixis refers to orientation in space: here versus there and this versus that. Temporal deixis refers to orientation in time, such as present versus past. Personal Deixis Many of the statements we make every day are comments or questions about ourselves or our interlocutors. I really should go now. Did you return the video I asked for? In this family we never smoke and rarely drink.

The pronouns I, you, and we, along with she, he, it, and they (and alternative forms) are personal deixis markers. When we use these pronouns, we direct our expressions towards ourselves, our interlocutors and third parties. Personal pronouns are of course not the only means used to indicate personal deixis. The phrase that person in the sentence You may like scary rollercoaster rides, but that person doesn't care can be used to refer to the speaker when the speaker wants to express anger or contempt, for example. Similarly, court etiquette may require you to use the noun phrase Your Honor when addressing a judge: Would Your Honor allow a brief pause? Therefore, personal deixis is not exclusively associated with pronouns, although pronouns are the most common way of expressing personal deixis. In this discussion, we mainly focus on pronouns as personal deixis markers. The most fundamental opposition in personal deixis systems is that between speaker (English I; German I; Persian man; Thai chaˇ n) and addressee (English you; German du; Persian to; Thai thee). This personal opposition is so fundamental that it is reflected in the pronoun systems of all languages. Pronouns referring to the speaker (or a group containing the speaker) are called first-person pronouns, and pronouns referring to the addressee (or a group containing the addressee) become second-person pronouns named person. In addition to the contrast between the first and second person, pronoun systems often have separate forms for the third person, i.e. any entity other than the speaker and the person being spoken to. In English he, she, it and they denote entities in the third person. but

• 193

1 9 4 • Chapter 6 The study of meaning: semantics

Third-person pronouns are not found in all languages. Some languages just don't have special ways of referring to third-person entities. In these languages, third-person entities are named with a demonstrative such as this or that, or are not expressed at all. In Tongan, a verb without an explicit subject is understood to have a third-person subject. oh well

a'u

past

to get

'(He/She/It/It) has arrived.'

Tongan has a third person pronoun form but only uses it for emphasis. oh well

a'u

I one

past

to get

He she

'He/She is the one who has arrived.'

The fact that some languages lack separate third-person pronouns reflects the fact that the third person is less important in personal deixis than the first and second persons. In fact, the third person can be defined as an entity different from the first person and different from the second person. Since it can be described in relation to the other two people, it is generally a less fundamental distinction in language. The singular pronoun system in English can be described as: speakers only

listeners only

your

neither speaker nor listener

he she that

Some languages make finer distinctions in their pronominal systems, while others make finer distinctions (see the “Semantic Universals” section in Chapter 7). However, all languages have separate first and second person pronouns. In addition to the person, personal Deixis systems can mark gender and number differences. In English, the gender distinction is only made in the third person singular: he for male and she for female. In other languages, the gender can also be marked for other people. In Hebrew, the second-person singular pronoun is ata for masculine references, but at for feminine references. The number is given for English pronouns in the first person (I versus we) and in the third person (he/she/it versus they); The second person pronoun is used to refer to singular and plural units. Many languages have separate second-person singular and plural pronouns (French tu and vous; German du and her; Persian to and shoma). Singular and plural are not the only categories of numbers that can be distinguished: some languages have different dual forms to refer to exactly two people, and some languages even distinguish between "some" and "many" references (see table of Fijian pronouns on page 220 of chapter 7). Finally, the personal deixis often reflects the social status of the speakers. In French, the choice of a second-person pronoun depends on the nature of the speaker's relationship with the addressee. When speaker and addressee are of roughly the same social status, the pronoun you is used; To mark or establish social distance or social inequality, a speaker uses the plural pronoun vous instead of tu even when addressing a person. Significantly more complex systems are found in languages such as Japanese, Thai, and Korean. In fact, the use of deictic means to reflect facts about the participants' social relationship is another type of deixis, commonly known as social deixis.

Function Words and Categories of Meaning Personal daixis can thus mark a number of overlapping distinctions: person, gender, number, and social relations. Languages express these distinctions in various combinations, marking some and not marking others. However, the basic distinction between first and second person is found in all languages and seems to be a basic semantic category in all deictic systems. Spatial deixis Spatial deixis is the marking of the orientation or position in space of the referent of a linguistic expression. The most commonly used word categories to express spatial daixis are demonstratives (this, that) and adverbs (here, there). Demonstratives and adverbs of place are not the only categories that have a spatially deictic meaning; The directional verbs ir and vir contain deictic information as well as bring and take. Languages differ in the number and meaning of demonstratives and adverbs of place. The English demonstrative system distinguishes only between this (near - close to the speaker) and that (far - relatively far from the speaker). It is one of the simplest systems found. At the other extreme are languages like Eskimo, which has 30 demonstrative forms. In all languages, however, the demonstrative system treats the speaker as a point of reference. Thus, the speaker is a fundamental reference point for spatial darixis. Many spatial deixis systems have three terms. Three-term systems fall into two categories. In one category are the meanings of the terms "close to the speaker", "slightly far from the speaker", and "far from the speaker". The Spanish demonstrative pronouns este, ese, aquel have these three respective meanings. In another type of three-term demonstrative system, the terms have the meanings "near the speaker", "near the listener" and "far from both speaker and listener". Fijian is an example of such a system. na ŋone oŋgo the child is (near me) 'this child (near me)' na ŋone oŋgori the child is (near you) 'the child (near you)' na ŋone oya the child who (far from you) ) and I ) 'this child (away from you and me)'

In both systems, however, the speaker is taken as either the sole reference point or as one of two reference points. Spatial deixis therefore represents the orientation of actions and states in space and is most commonly conveyed through demonstrative and place adverbs. Languages can have anywhere from 2 to 30 different demonstrative forms, but all demonstrative systems take the speaker as their basic reference point. Temporal Deixis A third type of deixis is temporal deixis - the orientation or position of the reference point of actions and events in time. All languages have words and phrases that are inherently marked for temporal deixis, such as B. before, last year, tomorrow, now and tonight. In many languages, temporal deixis can be marked over time, encoded on the verb with affixes, or expressed in a morpheme in its own right. inside

• 195

1 9 6 • Chapter 6 The study of meaning: semantics

English you must choose between the past tense and the non-paste form of the verbs. I go to school everyday. (no past tense) I went to school every day. (Past)

To express a future tense, English does not have a definite verb inflection (it lacks a future tense), but uses a multi-word verb in the non-past tense. I'm going to school next week. (from non-past to future time)

Tuvaluan is like English: e does not mean past tense while ne is past tense. au e fano ki te fakaala I nonpast go to the banquet 'I will go/go to the banquet'. au ne fano ki I Past go to 'I went to the banquet'.

that of

Wake-Up-Party

In some languages there is a choice between future and non-future (with undifferentiated present and past). In many languages, temporal deixis can only be marked with optional adverbs. This Chinese expression can be interpreted as past, present or future depending on the context: xià yüˇ rain down "It was/is/will rain".

When there is a possibility of ambiguity, a tense adverb ("last night", "now", "next week") is added to the sentence. In non-tense languages, another semantic category called aspect is often needed. The aspect is not directly related to temporal deixis, but refers to the way actions and states are seen: as continuous (he spoke), repetitive (he spoke [every day]), instantaneous (he spoke) etc. . in. Tense is therefore not the only marker of temporal deixis, although it is often exploited by languages as the primary means of marking temporal deixis. The most basic reference point for time is the moment the sentence is spoken. Any event that occurs before this point in time can be marked as past, and any event that occurs after this point in time can be marked as future. The train arrived. (anytime before speaking time) The train is coming. (at the time of the speech) The train will arrive. (at any time after the time of issue)

When the point of reference is at a point in time other than the moment of utterance, we say that time is relative. Relative time is used in many languages when speakers want to compare when two events occurred. After they bought two, they gave me another one. I was sick for a week before I saw you yesterday.

Semantic roles and meaning of sentences Languages sometimes have complex tense convention rules that dictate the form of verbs in relative contexts. Deixis as a semantic term The three types of deixis illustrate how semantic categories permeate language beyond simple word meaning. The deictic orientation of a sentence or clause can be conveyed by bound morphemes such as tense endings, by free morphemes and function words such as pronouns and demonstrative pronouns, or by content words such as here and bring. Deictic meaning is independent of the medium through which it is conveyed. One of the purposes of semantics is to describe which parameters are important or essential to characterize deixis (as well as other semantic categories) in language in general. For example, we observe that the distinction between speaker and addressee is an essential function of the personal deixis system of all languages. Likewise, we note that every spatial Deixis system has at least one reference point, a location near the speaker. A spatial deixis system can also have a secondary reference point near the listener. There is significant overlap between the types of Deixis. For example, personal, spatial, and temporal deixis share a fundamental reference point: the speaker's identity and location in space and time. Many linguistic features can be used to denote more than one type of deixis. The English demonstrative this can be used for personal deixis (that person), spatial deixis (this thing) and temporal deixis (this morning). Personal, spatial, and temporal deixis are clearly closely related terms. Text deixis A type of deixis that we have not yet discussed is the text deixis, which is the orientation of an utterance in relation to other utterances in a series of utterances. For example, consider the following pair of sentences: He started cursing and cursing me. That irritated me even more.

The demonstrative that at the beginning of the second movement does not refer to a spatial or temporal direction, but to something of the above. Label the text deixis. Textdeixis is therefore a tool that allows language users to group statements and show relationships between statements. Because textual deixis is primarily about utterances and their context, it goes beyond the scope of traditionally defined semantics, although its importance should not be underestimated.

Semantic roles and meaning of sentences We find that although sentences, like words, must have meaning in order for speakers of the language to be able to understand one another, the meaning of sentences cannot be determined simply by adding the meaning of each word to the content of the sentence. Judgment This fact was illustrated in the last section, where you saw that connected morphemes and function words can have a meaning that affects the meaning of the whole sentence. We also find that phrases like The coach licked the lion and The lion licked the coach have very different meanings even though they contain the exact same words. Of course, adding the meaning of each word doesn't give the full meaning of a sentence. Such a process will not even distinguish between the two simple illustrative sentences in this paragraph. Defining what makes up the meaning of a sentence needs to consider more than the meaning of individual content words.

• 197

1 9 8 • Chapter 6 The study of meaning: semantics

Consider the following active/passive counterparts, which describe the same situation at the referential meaning level: 1. The lion licked the trainer. 2. The trainer was licked by the lion.

These sets differ in that 2 is a passive structure while 1 is not. Since we are concerned here with meaning, we ask how to explain the synonymy between 1 and 2. Also note the following sentences: 3. David cut the salami with a knife. 4. David used a knife to cut the salami.

Here is a similar situation to the active/passive counterparts of 1 and 2, where the sentences have the same referential meaning. However, we must describe how sentences 3 and 4 mean "the same thing". The situations just presented suggest that the crucial factor in the way the meaning of sentences is constructed is the role that each noun phrase plays in relation to the verb. Therefore we have to introduce the notion of the semantic role of a noun phrase. The semantic role refers to the way in which the referent of the noun phrase contributes to the state, action, or situation described by the phrase. The semantic role of a noun phrase differs from its syntactic role (as subject, object, etc.), as shown by the contrast between sentences 1 and 2: in both 1 and 2 is the way the lion at one action involved, the same; and the way the coach gets involved is the same. Instead, the former, while having the same semantic role in both, has different syntactic roles, as the direct object of the verb in 1 and as the grammatical subject in 2. The semantic role is not an inherent property of a noun clause: a given los sentence noun can in have different semantic roles in different sentences, as in the following: Michael was hurt by a friend. Michael was in love with a friend.

The semantic role is a way of characterizing the meaningful relationship between a noun phrase and the verb of a sentence. Agents and Patients The first semantic roles we need to identify are agent (the initiator responsible for an action) and patient (the entity experiencing a specific state change). In both Set 1 and Set 2, the agent is the lion and the patient is the trainer. The fact that both sentences describe the same situation (and therefore have the same referential meaning) can be explained by the fact that in both sentences each noun phrase has the same semantic role. Experimenters The role of the subject-noun phrases in the following sentences is not that of an agent because Courtney is not actually the initiator responsible for the actions denoted by the verbs: Courtney likes blueberry pancakes. Courtney felt threatened by the lion.

In both sets, Courtney experiences a physical or mental sensation. Courtney's semantic role is Experiencer, defined as someone receiving sensory input. In English, experimenters can be direct subjects or objects, depending on the verb. With-

Semantic roles and meaning of sentences Compare the sentences about Courtney, in which the experimenter is the subject, with the following sentence, in which the experimenter is the direct object: Dwayne sometimes surprises me with his wit.

Tools and causes Now consider the semantic role of the underlined noun phrases in the following sentences: 5. Michael was hit by a rock. 6. Michael was hit by a rock.

The difference between these phrases is that 6 implies someone used a rock to attack Michael, while 5 does not. In Proposition 6 we say that a stone is the instrument or agent through which an actor carries out the action; Note that the definition requires that there is an agent, which agrees with our interpretation of Theorem 6. In Set 5, a stone can only be given the role of an instrument if there was an agent that caused the damage. If the rock that injured Michael were part of a rockfall, a rock would have the semantic role of cause, defined as any natural force causing a change in state. Instruments and causes can be expressed as prepositional phrases (as in the examples above) or subjects. The silver key opened the cellar door. (INSTRUMENT) The snow gave way on the roof. (CAUSE)

That the noun phrase the silver key is actually an instrument and not an agent is supported by the fact that an agent cannot be attached (bound by e), as the following anomalous example shows: *The silver key and Juan opened the door the cellar.

However, an instrument can be conjugated to another instrument and an agent to another agent. A shove and a shove opened the cellar door. John and Chelsey opened the basement door.

Recipient, Beneficiary, Locative, Tenses A noun phrase can be a recipient (someone who receives a physical object), a benefactor (the one for whom an action is performed), a locative (the place of an action or condition), or a temporary (the time the action or condition occurred). I gave Yolanda a puppy. (RECEIVER) Stefan forwarded the message to Yolanda. (BENEFITS) The Midwest is cold in the winter. (LOCATIVE) She left the house the day before yesterday. (IN THE INTERIM)

The goal of this effort is to characterize the possible semantic roles that noun phrases can fulfill in a sentence. Each noun phrase in a sentence is assigned a semantic function, and apart from noun phrase coordination, the same semantic function cannot be assigned to two noun phrases in a sentence. Consequently, a sentence like the following is ruled out as odd or semantically anomalous because it contains two underlined instrumental noun phrases: *This ball broke the window with a hammer.

• 199

2 0 0 • Chapter 6 The study of meaning: semantics

Furthermore, in most cases a single noun phrase can only be given one semantic function. In rare cases, a noun phrase can have two roles; in the sentence Geoff rolled downhill, if Geoff rolled downhill intentionally, he is both agent and patient because he is both the responsible initiator of the action and the entity experiencing the change of state.

Semantic Roles and Grammatical Relationships Semantic roles and grammatical relationships are not the same, and it is important to understand the relationship between them. For example, the subject of a sentence in English can be an agent (as in the underlined noun clause in sentence 1), a patient (as in 2), an instrument (3), a cause (4), an experimenter (5), useful (or receptor ) (6), locative (7) or tense (8), depending on the verb. 1st 2nd 3rd 4th 5th 6th 7th 8th

The caretaker opened the door. (AGENT) The door opened slightly. (PATIENT) Their first album greatly expanded their audience. (INSTRUMENT) Bad weather spoiled the corn crop. (CAUSE) Serge heard his father whisper. (EXPERIMENT) The young artist won the prize. (BENEFIT OR RECIPIENTS) Arizona attracts asthmatics. (LOCATIVE) The next day we made our way to Alice Springs. (IN THE INTERIM)

In certain English constructions, the subject may not play a semantic role, as in the construction "dummy it" where the pronoun it fills subject space but is semantically empty. It was clear the government had him trapped there.

Thus the notion of subject is independent of the notion of semantic role; and we could show the same for direct objects and other grammatical relationships. On the other hand, semantic roles do not appear to be constrained by grammatical relationships. For example, a locative can be expressed as a subject (as in sentence 1 below), a direct object (2), an indirect object (3), or an oblique object (4). 1st 2nd 3rd 4th

The garden will look great in spring. (Subject) William planted cucumbers and tomatoes in the garden. (direct object) Begonias add a cheerful look to the garden. (INDIRECT OBJECT) The door leads to the garden. (ASLANT)

However, there is a relationship between grammatical relationships and semantic roles. Consider the following sentences, all with open as a verb: Michele opened the door with this key. (AGENT) The door opens slightly. (PATIENT) This key opens the door. (INSTRUMENT) The wind opened the door. (CAUSE)

The grammatical subjects of the above sentences are an agent (Michele), a patient (the door), an instrument (that key), and a cause (the wind). Such an extreme variety is not found in all verbs. The verb to calm down can have an instrument or a cause as the subject.

Semantic roles and grammatical relationships This ointment relieves sunburn. (INSTRUMENT) The cool stream soothed my aching feet. (CAUSE)

To reassure an experimenter as the grammatical subject of the verb, we use a passive construction. The herbal tea calmed me down. (EXPERIMENT)

The verb clearly controls the range of variation allowed. Language users know the semantic roles that each verb allows as subject, direct object, and so on. In the mental lexicon there is a tag attached to the verb calm indicating that only instruments and causes are allowed in the subject position, while the tag attached to the verb open allows the subject to be agent, patient, instrument or cause. Semantic roles are universal features of the semantic structure of all languages, but the way they interact with grammatical relationships such as subject and direct object differs from language to language. Equivalent verbs in different languages do not carry similar tags. For example, the tag attached to the English verb like only allows experimenters as test subjects. I like french fries. (EXPERIMENT)

But only patients can be subjects of the corresponding Spanish verb gustar. The french fries

Mim

How. (PATIENT)

For me

cut

Potato Chips

'I like french fries.' (Lit., "French fries are nice to me.")

A similar situation is found for the verbs like and please in many other languages, including Russian. In some languages, the verb "understand" allows its subjects to be experimenters or patients, as in Samoan. The choice depends on the emphasis and focus. water

known a'u i

understand present tense

Wort

lesson object marker

'I understand the lesson.' water

qualified

Wort

I have already

a'u.

Present

I get it

Lesson

Pro

Mim

'I understand the lesson.' (Lit. 'The lesson understands for me.')

Some languages distinguish between agent and experimenter much more carefully than English. For example, the verb can have a subject when the action being described is intended, but it can have a direct object when the action is unintended. In addition to cross-linguistic variation with respect to specific verbs, languages vary in the degree to which different semantic functions can fit into different grammatical spaces in a sentence. In English, subject space can be occupied by noun phrases of any semantic function, depending of course on the verb. Many English verbs allow different semantic roles for subject, direct object, etc. But in many other languages the situation is different. In languages like Russian and German, verbs do not allow for as much variation in semantic roles as verbs in English, and there is a much closer connection between semantic roles and grammatical relationships.

• 201

2 0 2 • Chapter 6 The study of meaning: semantics

Computers, corpora, and computerized semantic corpora are useful for dictionary makers and others to discover language patterns that are not obvious through mere introspection. For example, patterns of collocation (which words belong together) are much easier to understand using a computerized corpus of natural language texts. These patterns can be very useful for emphasizing meaning, including parts of speech and words that occur with a certain frequency. Furthermore, although it may appear that synonymous words can be used in place of others, the corpora can show that it is actually not common for words to be easily interchangeable. For example, small and small, large and large, and fast and fast are often considered synonyms. But as a cursory examination of the Keyword-in-Context Agreements (KWICs) for these pairs shows, they are not directly interchangeable. Table 6-1 shows a selection of KWIC entries for the word small and Table 6-2 on page 203 shows a selection of KWIC entries for small. (Samples were taken from the British National Corpus and agreed with WordSmith.) Note in Table 6-1 that some of the sentences would not tolerate a small substitution.

narrow, for example 2, 3, 5, 6, 9, 10, 11, 15, 16, 17 and 21. Taking 3 as an example, English doesn't allow "anything angry". Of the cases where substitution is possible, some would sound very strange or convey a different connotation, such as B. 1, 4 and 8. In 1, “poor rich kid” and “poor rich kid” have different meanings. As the examples in Table 6-2 show, small is easier to replace with small; one reason is that little, in its adjectival use, has very similar labels and connotations to most uses of small. But if we look again at Table 6-1, we see that the opposite is not true. This is because little is not only an adjective meaning “small”, but also part of an adverb, in the expressions a little loud, a little discouraged and a little open (13, 20, 24) where it is an adjective modifies, and a poco more (12) where it modifies an adverb. However, dictionaries call small and small as synonyms. See the Additional Resources section at the end of this chapter for instructions on example sentences that contain any word or phrase you want to explore. You can learn a lot about the semantics of a particular word or phrase from this list. ■

TABLE 6-1 Close match

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Activities that come together etc. The few hours of poor training to avoid stagnation and I am deeply distressed and neither. Even without the threat to his job, he had a job, he had few options. There may be offers, some of which are not yet registered. But he would have if he had tried. But there are and those of your friends. They have noticed that my success has changed very little between classes. Objectively, he found it necessary to withdraw an energy in his actual work. We talked about there being another Ess, but the fur might be illuminating the whole body instead of just one. However, it is necessary to examine the role of Parliament. IN A SLOW tone with

tiny, tiny, tiny, tiny, tiny, tiny. tiny tiny tiny tiny tiny

Rich kid was served for a second, if anything, spinning. Vienna Dear Miss Angry, for the direction events have taken. Even when applying for a search warrant, for example when activating a maneuver, there may be little or no hope of finding these specific reasons. Current evidence suggests that I was objectively a little more attractive to the conservatively active life where I had p more time, so I bought some chocolate toppings. I want all men to enjoy a little bit of this. There is something else about Parliament's role. ON Rebound, South Africa, once again it was you

Semantic roles and grammatical relationships

• 203

TABLE 6-1 Close match

17 18 19 20 21 22 23 24 25 26 27

and seven balances to spare. In addition to a soft and beautiful vocal body, and an actually ugly body, still moving over her like a move, she closed her eyes, tiredly clenched to his stubborn hands from time to time, and facilitated the original access to the earth, a question. About In this study, the privatization program implied an increase in efficiency. Can you leave us a base with some objectives? We thought we knew an Unge and the dining area was probably for being happy and having a happy face and spreading a smile

tiny tiny tiny tiny tiny tiny tiny

Silicone lube (vacuum grease, timid-looking DOW, and tombstone taken care of. Once the car was down. Then the front zip had been done so fingers could slide in. Progress - with the change of position in the lead or the open improvement at times, but with the pace we know a bit more about what makes children a little smaller.

small small small small small small small small small small small small small small small small small small

For us to analyze (the Scotch, the familiar G crowd waiting for them outside, the kind of unsuccessful business decoration judging by the lack of serious interference now, pinching him in the neck, waist for a quantitative analysis) can be a number of Ammunition, the remains of its mouth contracted in a vicious line. The percentage of additional income that a company earns with a hundred children was filed around the idea - unions and, as a secondary consequence, uninhabited islands, and that the necessary remainder of that nationality outlive the income. perfect as a little velvet and 88% of the suitcases fell into the child and immediately went back to the black and white, that would be enough) so the anteroom through the gallery, the closets, shelves, mirrors and even ch cities and their larger ones Counterparts achieved Is the city a large city and would b enough for a group of local enthusiasts who frequently take steps to promote regions within a country and for group pools and 50 minutes up the big no-name side street should I have compared p? Fuel economy you stick with

TABLE 6-2 Small Agreement

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

Newspaper advert showing individual readers also crying at the sight of the house and the McCloy, who lives here in town, drives a lot The Fuhrer rested in a cursory meeting until the spring of 1941. Even his clothes (although the numbers are, of course, very high) Sick and mortar , and become a gray white face of the captain, the result of which the individual can only keep very uniform products that help both large and old for a pond in the garden, but with the subsequent continuation of a large number of archives, often in near remote reefs, headlands, or the time of the Russian conquest, although only one bought his home and vouched for arbitration, risks were also taken). In the process, he knocked over a teacher's must-have device along with a monitor (a. They ended up dragging the gnomes onto a famous wall rail with hooks for hanging." The most striking contrast between the Mento. Thirty years later, the Mento that the cemeteries since have long been the focus and have a policy of devaluing the yuan by any country-specific factor, even by 5°F/Gas Mark 3), and allow 40 minutes for the Admiralty to return. a can answer “Yes, but these cost savings are

2 0 4 • Chapter 6 The study of meaning: semantics

Summary • Semantics is the study of meaning in language. • Semantics has traditionally focused on linguistic meaning, but languages also convey social meaning and affective meaning. • Words, sentences and expressions can have meaning, and the meaning of the sentence and the meaning of the expression must be distinguished. • The study of the meaning of sentences falls mainly within the realm of semantics. • Within a sentence, words can reach other constituents, since they only reach the bracketed constituent in He only knew [what he had read in the letter]. • Pragmatics is the branch of linguistics that deals with the meaning of statements. • Lexical semantics is the study of meaningful relationships in vocabulary. The types of relationships that exist between word groups are universal, although the specific word groups to which they refer vary from language to language. • Semantic fields are sets of words whose referents belong together based on fundamental semantic features. • Words in a semantic field can be organized in terms of these relationships: hyponymy (a kind of), part/whole (subdivision), synonymy (similar meaning), scalable and non-scalable antonymy (opposite meaning), inverse (mutual meaning) . ). ), polysemy (multiple meanings), homonymy (same written or spoken form), and metaphor (derived meaning). • Semantic terms such as deixis can be expressed by concatenating morphemes (-ed in walk) and function words (he who is in them) as well as content words (morning). • There are several types of deixis: personal (you, me), spatial (here, there) and temporal (now, then). They all require a reference point to be identified. • In relation to the speaker and the moment of utterance, the here and now is very privileged as a point of reference in the three types of deixis. • The meaning of a sentence is not simply the sum of the meaning of its words. • Sentence semantics aims to discover the basic relationships between noun phrases and the verb in a sentence. • Semantic roles (eg agent or instrument) are not inherent properties of noun phrases, but relational concepts. They are independent of the grammatical relationships (e.g. subject or object) of the noun phrase. The verb determines which semantic role can be used in certain grammatical gaps in a sentence. • This chapter describes nine semantic roles: Agent: the initiator responsible for an action Patient: the entity undergoing a particular change in experiential state: the entity receiving sensory information Instrument: the intermediary through which an agent performs an action performs : the natural cause that causes a change of state beneficial: the entity on which an action is performed receiver: the entity that receives a local physical object: the location of an action or transient state: the moment when the action takes place or State • Semantic functions are universal, but languages differ in how specific functions are encoded in the syntax.

What do you think? checked

WHAT DO YOU THINK ? REVIEW • Holly's question. George Washington and the first President of the United States usually refer to the same person, namely the man who lived from 1732 to 1799 and became President of the United States in 1789. George Washington is this man's name, but George Washington does not mean "the first President of the United States." The two terms have the same referent but different meanings. The fact that two terms refer to the same entity does not necessarily imply that they mean the same thing. Think of sentences in which one of these expressions would be appropriate but the other is not, or both could occur but cannot be interchanged. Consider this example: "Some Marylanders like to say that John Hanson was the first President of the United States." Even if this statement were true, it would not be true to say, "Some Marylanders like to say that John Hanson was George Washington. " The fact that one is true but the other does not prove that they do not mean the same thing. • Nathan's idea. Although synonyms may not have the same occurrence patterns as other words, they can still mean the same thing in some contexts. Nathan identified a frame that might fit fast or quick, but not with the same meaning, and this fact seems to provide an argument that the two words are not used exactly the same, but are synonymous. And of course, "fast talker" is an expression that could have come from someone who speaks quickly, but the expression identifies someone who persuades with smooth or deceptive language. • Your uncle. The term that characterizes the semantic relationship between uncle and nephew is "converso". Among other things, hot means “not cold”, but nephew is not the same as “not uncle”. cold where hot and cold are opposites. • Family picnic. Words like mine and yours are "dectic" expressions, and their meaning in this case depends on who pronounces them. Mine means "belongs to the speaker" (or reported speaker), so if your cousin says "it's mine," he claims ownership, and if your daughter uses the same words, she claims ownership. Deictic words must be interpreted in their context.

• 205

2 0 6 • Chapter 6 The study of meaning: semantics

Exercises Practice exercises

A. Provide a word whose reference has the specified semantic relationship to the words below. 1. Mother and father are hyponyms. 2. The knee is the part. 3. Jewelry is the main term. 4. The bike is everything. 5. strong is the antonym. 6. Grandma is the complete opposite. 7. Niece is in arrears. 8. Obese is the synonym. B. From Table 6-1 on pages 202 and 203, give the numbers of the KWIC Concordance Lines where the word little could be explained in each of these senses: 1. "little" 2. "not much" 3. " small amount". from'

Based on English 6-1.

The following sentences are ambiguous. Based on the discussion in this chapter and in Chapter 5, describe the ambiguity. 1) They found peasants repulsive. 2) The car I want to drive is a Lamborghini. 3) Nothing is more alarming than the development of nuclear power plants. 4) Erika doesn't like her husband and neither does Natalie. 5) They said they told him to come to them. 6) Challenging fighters are avoided at all costs. 7) He met his challenger at his home.

6-2.

Identify the differences in linguistic, social, and affective meaning between the words and phrases in each of the following sentences: 1) swindle, deceit, deceit, fraud, ruse, trick 2) delicious, nice, great, elegant, good, pleasant, bad, cool 3) man, dude, dude, athlete, little devil, kid, gentleman, hunk, child 4) eat, devour, nourish, devour, peck, ingest, chew, graze, fill the belly 5) tired, exhausted, tired, tired, listless, exhausted, exhausted, forgiven, exhausted 6) idiot, idiot, nerd, blockhead, idiot, turkey, coward, punk, blockhead, scumbag

Exercises 6-3.

Some of the following term sets form semantic fields. For each set: a. Identifies words that do not belong to the same semantic field as the rest of the sentence. B. Identify the parent term of the remaining semantic field, if any (it can be a word from the set). C. Determine if some terms are less pronounced than others and justify your statement. 1) acquire, buy, collect, hoard, earn, inherit, steal 2) whisper, speak, tell, tell, say, harangu, scribble, instruct, summarize 3) path, path, barn, path, street, highway, avenue, Via, Interstate, Method 4) stench, smell, stench, aroma, bouquet, fragrant, perfume, scent, aroma, olfactory

6-4.

For each semantic relationship given below, provide one or more examples of words whose references have that relationship to the given word, and provide the name of the semantic category used to cover your answer. Example: Fish is the generic term (hyperonym). Answer: Salmon, trout, cod, sole, swordfish, tuna are his hyponyms. 1) Irish Setter, Dalmatian, Cocker Spaniel are the hyponyms. 2) peep, tom, persian, alley are the hyponyms. 3) Dog, cat, fish, parakeet, hamster are hyponyms. 4) Knife, fork, spoon are hyponyms. 5) true is the antonym. 6) imprecise is the antonym. 7) Sister is backward. 8) Teacher is the complete opposite. 9) Partner is backwards. 10) Toe is the part. 11) The menu is the whole. 12) Friend is the synonym. 13) Teacher is the synonym.

6-5.

Consider the following two sequences of dictionary entries, taken (slightly abridged) from The American Heritage Dictionary of the English Language, 3rd ed. (Boston: Houghton Mifflin, 1992): Sequence 1 shellⴢy1 adj. -iⴢer, -iⴢtest. 1. Hoarse or raspy quality: a voice hoarse with emotion. 2 Similar to a shell. B. shells included. [From HUSK] – husk´iⴢly adv. husⴢky2 adj. -iⴢer, -iⴢtest. 1. strongly built; corpulent. 2. Strongly built: Clothing sizes for strong children. – husky n., pl. -is. A hoarse person. [Maybe from HUSK]

• 207

2 0 8 • Chapter 6 The study of meaning: semantics

husⴢky3 n., pl. -kies 1. Often husky or huskie. Breed of dogs developed in Siberia for pulling sleds and with dense multi-colored fur. Also called Siberian Husky. 2. A similar dog of arctic origin. [Probably due to truncation and modification of ESKIMO.] Sequence 2 junⴢior adj. 1. Abbreviation jr., jr., jun., jun., jnr. It is used to distinguish a son from his father when they have the same name. 2. Aimed at youth or below: youth fashion; a youth sports league. 3. Of lesser rank or tenure: a non-commissioned officer; the young senator from Texas. 4. From, for, or forming third-year students of an American school or college: the junior class; junior prom. 5. Scale smaller than normal. -junior n. apr jr., jr., jun., jun., jnr. 1. A person younger than another: a sister four years younger than me. 2. A person inferior in rank or time of participation or service; subordinate. 3. A junior in an American high school or college. 4. A class of dress sizes for girls and slim women. In this sense, she is also called Junior Miss. Using terms introduced in our discussion of lexical semantics, describe in detail how these dictionary entries are organized. Include a discussion of the criteria used to create different entries or subentries for words of the same name. 6-6

In the following sentences, one or more words are used metaphorically. Give a general statement describing the principle underlying each set of metaphors; then add a metaphor to the crowd that follows the principle. Example: I let my manuscript simmer for six months. He came up with a reply that readers will appreciate. There is no simple recipe for writing successful business letters. Parent statement: "The writing process looks like cooking." Additional example: "He's the kind of writer who invents another useless novel every six months." 1) Viewers peppered him with counter-arguments. His opponents tore up his arguments. My reasoning left them out of ammunition. The others will never be able to destroy this argument. His question betrayed a defensiveness. 2) This heat is suffocating. The sun hits these poor workers hard. The clouds seem to be rising. The northern part of the state is under heavy snowfall. The cool breeze drove away the oppressive heat.

6-7.

Determine whether the words in each of the following groups are polysemous, homonymous, or metaphorically related. State the criteria you used to reach your conclusion. You can use a dictionary.

Exercises 1) run down (the stairs); trample (an enemy); below (a list of names) 2) the seat (of the pants); the government website); the (driver's) seat (of a car) 3) an ear (for music); an ear of corn); an ear (as an organ of hearing) 4) for throwing (a baseball); pitch black); the tone (of someone's voice) 5) spell (one word); (under) a spell; a spell (dry) 6) vision (the ability to see); (a man with) vision; vision (as hallucination) 7) the butt (of a gun); the butt (in a joke); beat (like a ram) 6-8.

Identify the semantic role of each underlined noun phrase in these sentences: 1) In October I looked from the wooden bridge at the small river behind our university. 2) I forgot everything I learned in elementary school. 3) The Grand Tetons rise majestically over the valley. 4) Snow completely buried my car during the last storm. 5) Fifty kilos of cocaine were seized by the DEA. 6) Natalie won a trip worth a thousand dollars. 7) The hurricane destroyed the island. 8) Your joke never ceases to amaze me.

6-9.

a. Examine Table 6-1 on pages 202 and 203 to determine which words occur frequently with little, either before or after. B. List all immediate ingredients, few of which are an element in the examples in Table 6-1; Based on that, tell what type of sentence it works wrong in, e.g. B. in an adjective or adverbial clause. Example 12: a little more - adverbial phrase; 20: a little discouraged - adjective phrase

Based on English and 6-10 other languages. A "tag" is attached to each verb in the lexicon, indicating what semantic role each noun argument can be assigned. For example, the verb asssar can have an agent as a subject (as in clause 1), a patient (as in clause 2), a cause (3), or an instrument (4). But in the subject position it does not allow locatives (5) or tenses (6). 1) Matthews Baked Scones. 2) The cake bakes. 3) The sun crisps my lilies. 4) This oven makes wonderful cakes. 5) *The kitchen bakes well. 6) *Tomorrow it will bake well. a. From the given sentences, determine what semantic functions these verbs allow as a subject: feel, provide, absorb, thaw, enjoy. 1) His hands were loose and sweaty. He could sense the presence of an intruder in the apartment. This room feels damp.

• 209

2 1 0 • Chapter 6 The study of meaning: semantics

They felt everyone under the covers to see what was there. This semester looks very different from last semester. 2) Gas lamps provided light for the outdoor picnic. These fields provide enough wheat to feed a city. Who delivered these buns? The accident worried me a lot. His books provide many illustrations of this phenomenon. In such cases, the Articles of Association provide for the dissolution of the Management Board. 3) The students have absorbed so much material that they can no longer understand it. This type of sponge does not absorb water well. The United States took over the Republic of Texas in 1845. My working hours eat up all my free time. The ground absorbs the rain. 4) If Antarctica suddenly thaws, sea levels would rise dramatically. Chicken doesn't defrost well in just two hours. The crowd thawed after Kent's arrival. Kent's arrival thawed the party. The heat from the sun will melt the ice in the cooler. Ice melts at 0 degrees Celsius. The peace treaty will unfreeze relations between the United States and China. 5) This wine tastes like vinegar. He tasted all the appetizers at the party. I can taste the capers in the sauce. B. Languages can differ in terms of the semantic roles that certain verbs can take. The following are semantically well-formed French sentences with the verb goûter 'I like': Il n'a

I have never tried caviar.

He never

tasted the caviar

"He never tasted caviar." It gives me a bitter taste.

Try a

bitter taste inside

Café.

this coffee

"I have a bitter taste in this coffee." On the other hand, the following sentence is not well constructed: *Les cuisses de grenouille goûtent bon. He

leg

toad

I like

Boa

Frog legs taste good. What is the difference between English gusto and French goûter in terms of the range of semantic roles they allow as a subject?

Another resource

Specially for educators and future teachers 6-11. Your high school ESL class is asking if Riverside Bank and Savings Bank are the same word or different words. You realize that both are nouns and are spelled and pronounced the same. By attempting to identify synonyms and antonyms (as in Figure 6-4 on page 188), you construct an argument designed to convince your students that these are different words and not the same word with different meanings. To show the contrast, identify another pair of word forms that represent different meanings of the same word and rebuild the argument by identifying synonyms and antonyms. 6-12 In his freshman year, he told his high school English class that the subject of a sentence is the "doer" of the plot, and Devon got the most points for the example. If you ask for other examples, a student says Disneyland is fun, and you immediately see a problem: Disneyland is the subject of the sentence, but not the actor of an action. What do you say to correct your explanation of the roles played by subjects in sentences? 6-13. Draw characterizations of one or at most two sentences at a time to help your students remember the difference between a grammatical relationship (e.g. subject or object) and a semantic role (e.g. agent or medium). 6-14 Name three pairs of terms in which the referent of the two terms is the same but the meaning is different. Example: assemble. McKinley and the tallest peak in the United States. 6-15 When writing manuals, authors are sometimes asked to be careful where they just put the word in a sentence. You may recommend placing it immediately before the component within its scope (manuals may spell it to come directly before the words it modifies). In the following sentences, enclose the member within the scope of just in parentheses and insert a caret where just could be placed so that it is directly before the structure in its scope. (Note: These phrases are taken from the British National Corpus.) Example: That leaves [a logical explanation]. (⫽This leaves only a logical explanation.) a. It was only a matter of time. B. She just needed to rest. C. I only saw a small part of it. i.e. Opportunities were only half used. Y. Newspaper advertisements usually only offer one product or a small range of products. F. Cassie knew only one such stone.

Other British National Corpus resources: http://www.natcorp.ox.ac.uk/

Here you can get up to 50 example sentences, randomly selected from the 100 million word resources of the British National Corpus.

• 211

2 1 2 • Chapter 6 The study of meaning: semantics Roget's Internet Thesaurus: http://www.thesaurus.reference.com/

On this page you will find access to an online thesaurus. It allows you to examine the relationships between words, especially those in hyponymous relationships.

Suggestions for further reading • Stephen R. Anderson and Edward L. Keenan. 1985. "Deixis", in Timothy Shopen, ed., Language Typology and Syntactic Description, vol. 3 (Cambridge: Cambridge University Press), pp. 259-308. A relatively brief and comprehensive treatment of Deixis. • Sandra Chung and Alan Timberlake. 1985. "Tense, Aspect, and Mode", in Timothy Shopen, ed., Language Typology and Syntactic Description, vol. 3 (Cambridge: Cambridge University Press), pp. 202-258. Provides a brief discussion of temporal and related terms. • George Lakoff and Mark Johnson. 1980. Metaphors We Live By (Chicago: University of Chicago Press). This remarkable, accessible, and classic work on metaphor has been reprinted with an added "Afterword, 2003," from which our quote on page 190 is taken; Some ideas presented in the original book are expanded and partially corrected in the reprint. • George A Miller. 1996. The Science of Words (Indianapolis: W. H. Freeman). An accessible and award-winning treatment of the psychology of lexical meaning. • Sebastian Lobner. 2002. Understanding Semantics (London: Arnold; New York: Oxford University Press). Published in the Understanding Language Series, this is a complete and comprehensive introduction to semantics in general. It goes beyond the current chapter by covering the meaning of sentences more comprehensively, addressing cognition, translation, and formal semantics.

Further Reading The main reference work on semantics is Lyons (1977), which provides a wealth of information and critical discussion. Lyons (1996) and Saeed (2003) are simpler and more accessible. Lexical semantics is discussed in Lehrer (1974), who focuses on semantic universals (discussed in Chapter 7 of this book), and in Wierzbicka (1985), who mainly deals with the meaning of the term "type of". Cruse (1986) is a good summary of lexical semantics. Several articles by Holland and Quinn (1987) examine connotations and cultural elements in the organization of semantic fields. The ideas presented in Lakoff and Johnson (1980) are further developed in Lakoff (1987). Other valuable work on this and related topics is Lakoff and Johnson (1989) and Fauconnier and Turner (2002). The deixis is discussed in detail in chapter 2 of Levinson (1983). See Palmer (1986) for a full discussion of mood and modality. Corporation-based approaches to lexicography can be found in Sinclair (1991).

References • Cruse, DA 1986. Lexical Semantics (Cambridge: Cambridge University Press). • Fauconnier, Gilles and Mark Turner. 2002. The Way We Think: Conceptual Mixture and the Hidden Complexities of the Mind (Nueva York: Basic Books).

References • Holland, Dorothy and Naomi Quinn, eds. 1987. Cultural Patterns in Language and Thought (Cambridge: Cambridge University Press). • Lakoff, George. 1987. Women, Fire, and Dangerous Things: Which Categories Reveal the Mind (Chicago: University of Chicago Press). • Lakoff, George and Mark Johnson. 1989. More Than Reason of Brilliance: A Field Guide to Poetic Metaphor (Chicago: University of Chicago Press). • Teacher, Adrienne. 1974. Semantic Fields and Lexical Structure (Amsterdam: North Holland). • Levinson, Stephen C. 1983. Pragmatics (Cambridge: Cambridge University Press). • Lyon, John. 1977. Semantics, 2 vols. (Cambridge: Cambridge University Press). • Lyon, John. 1996. Linguistic Semantics: An Introduction (Cambridge: Cambridge University Press). • Palmer, F.R. 1986. Mood and modality (Cambridge: Cambridge University Press). • Saeed, John I. 2003. Semantics, 2nd ed. (London: Blackwell). • Sinclair, John. 1991. Corpus, Concordance, Collocation (Oxford: Oxford University Press). • Wierzbika, Anna. 1985. Lexicography and Concept Analysis (Ann Arbor, MI: Karoma).

• 213

7 Universals of Language and Linguistic Typology WHAT DO YOU THINK? • One day your third grade niece comes home from school and announces that her teacher said English has 13 vowels. She asks if all languages have 13 vowels. At first you want to say that English only has 5 vowels, but then you realize you don't have to. what do you say to her • Several cousins are visiting you in Chicago for your twenty-first birthday and you notice that Laura from Texas says "du" when addressing more than one person, while Rudy from New York City sometimes says "du" under the same circumstances . You wonder if their dialects are flawed (because they don't sound "standard") or possibly better than yours (because they make a useful difference with your Chicago English). You're also wondering if it's normal for languages to have equivalents of y'all and youse, or like Standard English only has a singular and plural form. What is your conclusion? • A classmate in a Japanese class notices that the order of words in Japanese is strange. He says that Japanese put verbs at the end of the sentence instead of after the subject: "where they belong". It confirms that the logical order is subject-verb-object, as in English. A Japanese exchange student you taught found the complete opposite, that the Japanese subject-object-verb order was more logical. He is convinced that logic is not the problem, that both orders are equally logical. What argument can you use to convince your colleague that word order is not a matter of logic and that Japanese and English are equally logical (or illogical) in this regard?

214

Similarity and diversity between languages.

Similarity and diversity between languages The different languages of the world are structured according to many different patterns of phonology, morphology, syntax and semantics. Some languages have large pools of phonemes; others have very few. In some languages, including French, Italian and English, the basic sentence structure is SVO, i.e. H. the subject comes before the verb and the verb before the direct object, as in these examples: French English Italian English

Affairs)

Verb

object (the)

Haussmann Haussmann Keplero Kepler

refurbished redesigned modified modified

the space. Square. Copernican theory. Copernican theory.

In languages like Japanese and Persian, both the subject and the direct object come before the verb, in a SOV pattern: Persian Japanese

Affairs)

object (the)

Verb

I'm hebi ga This snake Ali Ali

inu the dog ketabhara the books

totes Korosit. mibrd. loaded.

This snake killed the dog. 'There he carries the books.'

Given this variation, you may be wondering if the languages of the world share some characteristics. It turns out that there are basic principles that govern the structure of all languages. These universals of language determine what is possible and what is impossible in the structure of language. For example, while some languages have both voiced and unvoiced stops (h and p; d and t) and others only unvoiced stops (p and t), no language has yet been found that has voiced stops but no unvoiced stops. This observation can be translated into a rule expressing what is possible in the structure of a language (i.e. a language can have voiced and unvoiced stops or only unvoiced stops in its phonetic inventory) and a law excluding a combination of phonemes. do not occur in any of the world's languages (ie voiced pauses without silent pauses).

Why Discover Universals? Universals of language are statements about what is possible and impossible in languages. From a purely practical point of view, such principles make sense because if we can assume that they apply to all languages, there is no need to repeat them in the description of each language. Thus, the study of linguistic universals underscores the underlying unity of the vast diversity of languages found in the world. Language universals are also important to our understanding of the brain and the principles that govern interpersonal communication across cultures. Only the human species developed the ability to speak in the course of evolution and thus differed from all other animals, including other higher mammals. Humans have not developed a single language that everyone speaks and understands, but more than 6,000 different languages, each complex and sophisticated. If basic principles govern all languages, it is likely that they are the result of the cognitive and social skills that enabled humans to develop the ability to speak in the first place. When learning the language

• 215

2 1 6 • Chapter 7 Language universals and language typology

universal, we begin to understand what it is about the human brain and the social organization of everyday life that enables people to communicate through language. The study of language universals offers insight into the cognitive and social foundations of human language, about which so little is known. Researchers must be careful when positing linguistic universals because relatively few languages in the world are adequately described. In addition, much more is known about European languages and major non-Western languages (such as Chinese, Japanese, Hindi, and Arabic) than about the much more numerous languages of Africa, Asia, America, and Oceania. . In Papua New Guinea alone (an area the size of the states of Washington and Oregon combined) more than 700 languages are spoken, although grammatical descriptions are only available for a few dozen; very little or, in some cases, nothing is known about the others. Linguists proposing world languages need to ensure that the principles proposed are applicable to more than the known European languages. Linguistic universals should apply to the languages of the world in general, whether those languages are spoken by a few dozen people in a village in the highlands of Papua New Guinea or by millions of people in Europe, Africa, or Asia. Because little or nothing is known about the structure of hundreds of languages, universal principles can only be proposed as tentative hypotheses based on the languages for which descriptions are available. Fortunately, many linguists today study lesser-known languages. For the most part, first-time grammars confirm rather than refute the proposed linguistic universals. Caution is also required when drawing conclusions about world languages. These universal principles help explain why language is species-specific, but there is a big step between discovering a universalism and explaining it in terms of human cognitive or social abilities. Most often, explanations of language universals as symptoms of cognitive or social factors are based on logical arguments rather than solid scientific evidence. Of course, the fact that explanations can only be tentative doesn't mean they shouldn't be suggested, but it does mean that linguists have to be careful and remember that languages play many roles simultaneously.

Types of Languages A prerequisite for the study of universals is a thorough understanding of the diversity found among the languages of the world. Linguistic typology focuses on classifying languages according to their structural features. (Typology means the study of types, or the classification of objects into types.) Examples of typological classifications are "languages that have both voiced and unvoiced stops in their phonemic stocks" (such as English, French, and Japanese) and "languages." which only have unvoiced stops. "voice" (like Mandarin Chinese, Korean, and Tahitian). Since no language has voiced stops without unvoiced stops, this type does not exist. Of course, the languages in each category will be different depending on the criteria of, for example, if we have a Typology of languages according to whether or not they have nasal vowels in their phonetic inventory, English, Japanese, Mandarin Chinese, Korean and Tahitian fall into the category of languages having no vowels In contrast, standard French has four nasal vowels (some French dialects have only three): /7/ as in faim /f7/ 'hungry'; /8/ as in brun /bʁ8/ 'brown'; /0/ as in manger /m0e/ 'to eat' and /9 / as in maison /mεz9)/ 'house' Thus, it falls into the category of languages that have nasal vowels, along with Hindi, Tibetan, and Yoruba (language widely spoken in Nigeria.) Of course, linguists can categorize

Universal semantic egories only according to certain criteria; The languages of the world are so diverse in so many different ways that there is not even a general typological classification of languages within a single level of language structure such as phonology. The typological categories have no necessary correspondence with language groups descended from the same mother tongue; In fact, typological categories extend across language families. In the example just given, English, Japanese, and Tahitian are not related languages; However, they fall into the same language type in terms of the presence or absence of nasal vowels. On the other hand, French and English are related but divided into different types. Language types are in principle independent of language families, but members of the same family often share certain typological features because of their common heritage. Consequently, linguists include as many unrelated languages as possible in their proposed language types to ensure that similarities between languages in any one category are not the result of familial relationships. This chapter examines both the diversity of the world's languages and the unity that underlies that diversity. Discovering linguistic universals and classifying languages into types are related and complementary tasks. In order to discover universal principles, we first need to know how different languages are structurally. We don't want to postulate a universal language based on a limited set of languages, only to find that the proposed universal language doesn't work for a language type we don't consider. A universal should work for all language types and all languages. Likewise, the way we classify languages and describe different types of structures is largely determined by the search for universals. For example, it would be possible to form a typological category by grouping all languages that have the sound /o/ in their phonetic inventory. But such a typology tells us nothing about a universal principle underlying the structure of these languages; In fact, their structures may have little in common, apart from the fact that /o/ is an element in their phonetic inventory. Instead, a typology of languages based on the presence or absence of nasal vowels reveals interesting patterns. It turns out that no language in the world has only nasal vowels. All languages must have oral vowels, whether or not they have nasal vowels. This suggests that oral vowels are in some sense more "basic" or indispensable than nasal vowels, which could be of great interest to our understanding of language structure. Therefore, this typology is useful as it helped reveal a universal language. Whether a particular typological classification is interesting or useful depends on whether it helps to discover universal principles in the structure of languages. The following sections present examples of universal languages and language types from semantics, phonology, syntax and morphology. For each example, look at the interaction of typologies with universals and note how the different types of language universals are declared. Some examples are revisited at the end of the chapter, where we examine the cognitive and social explanations that have been proposed to account for linguistic universals and linguistic types.

Semantic Universals Semantic universals determine the composition of the vocabulary of all languages. That there are semantic universals may come as a surprise at first glance. whatever you have

• 217

2 1 8 • Chapter 7 Language universals and language typology

Anyone who has studied a foreign language knows how different the vocabulary of two languages can be. Some ideas that can be conveniently expressed by a single word in one language may require a full sentence in another language. For example, the English word “privacy” has no simple French equivalent. (That's not to say the French don't have a sense of privacy!) Similarly, English lacks an equivalent for the Hawaiian word aloha, which roughly translates to 'love', 'compassion', 'pity', 'hospitality'. friendship' and is also used as a general greeting and farewell. However, despite these linguistic differences, there are some fundamental areas of the vocabulary of any language that are subject to universal rules. These areas include color terms, body part terms, animal names, and sensory verbs. Semantic universals usually deal with the less distinctive members of semantic fields (see Chapter 6), which in this context are called base concepts. As an example, consider the following terms that refer to shades of blue: turquoise, royal blue, and blue. Blue is a simpler term than the others: turquoise derives from the name of a gemstone of the same color, while royal blue refers to a shade of blue. Thus, the word blue is more fundamental than any of the other words, albeit for different reasons: unlike turquoise, blue primarily refers to a color, not an object; Unlike royal blue, blue is a simple, unchanging term. The combination of these qualities makes blue a less pronounced, more basic color concept than the others. Basic concepts have three properties: 1. Basic concepts are morphologically simple. 2. Basic terms have a less specialized meaning than other terms. 3. The basic terms have not recently been borrowed from another language.

Semantic universals deal with terms like blue rather than terms like turquoise and royal blue.

Pronouns Although pronoun systems can vary greatly from language to language, every language's pronoun system follows the same universal principles. First, all known languages without exception have pronouns at least for the speaker and the addressee: the first person (I, I) and the second person (you). But there are big differences between the languages of the world in the number of distinctions made by pronouns. The following table represents the English pronominal system (we limit ourselves to subject pronouns). English pronouns FIRST PERSON SECOND PERSON THIRD PERSON

SINGULAR

PLURAL

I you he she it

we you them

In this chart, the columns represent number: the first column singular, the second column plural. List of person rows: The first row shows first-person pronouns, the second row shows second-person pronouns, and the third row shows third-person pronouns. American Standard English uses the same form for the second person singular and the plural pronoun (you).

Semantic Universals The pronoun systems of other languages have different patterns. Spoken Spanish has different singular and plural forms in each person; In the example below, the two plural forms are the masculine and feminine pronouns. (Spanish also has "polite" pronominal forms, but we ignore them here.) Castilian Pronouns SINGULAR FIRST PERSON SECOND PERSON THIRD PERSON

I you he she

PLURAL M

we you them

Some languages make finer distinctions in number. Ancient Sanskrit speakers distinguished between two persons and more than two persons. The form for two people is called dual, and the form for more than two people is called plural. (In the table below, the three words for the third person are the masculine, feminine, and neuter forms.) Sanskrit Pronouns FIRST PERSON SECOND PERSON THIRD PERSON

SINGULAR

DOUBLE

PLURAL

aham tvam sas, tatuaje, sa

avam u¯vam tau, of, from

vayam yayam te, tani, tas

Other languages have a single pronoun that refers to the speaker and the addressee (and sometimes other people) simultaneously, and a separate pronoun that refers to the speaker along with other people but excludes the addressee. The first is called the first-person inclusive pronoun, and the second is called the first-person exclusive pronoun. In English, both terms are encoded in nodes. In contrast, Tok Pisin has separate inclusive and exclusive pronouns. Tok Pisin Pronouns FIRST PERSON EXCLUSIVE FIRST PERSON INCLUDED SECOND PERSON THIRD PERSON

SINGULAR

PLURAL

Mypela Yumi all of you

you in

Tok Pisin is an English-based Creole (see Chapter 13), with most of its vocabulary derived from English. The English pronouns and other words that Tok Pisin speakers used to form their pronoun system are easy to recognize: mi is from me, yu from you, probably from him, yumi from you-me, ol from everyone, and that plural suffix - from probable partners. Fiji has one of the largest pronoun systems of any language. It has a singular form for each pronoun, a dual form for two people, a separate "judgment" form referring to about three people, and a plural form referring to more than three people (in actual usage refer experimental pronouns refer to a few persons and the plural refers to a multitude). Also, like Tok Pisin, Fijian has first-person dual, evaluative, and plural, separate inclusive and exclusive forms.

• 219

2 2 0 • Chapter 7 Language universals and language typology

Fijian pronouns

FIRST PERSON EXCLUSIVE FIRST PERSON INCLUDED SECOND PERSON THIRD PERSON

SINGULAR

DOUBLE

TRIES

PLURAL

keirau kedaru kemudrau rau

keitou kedatou kemudou iratou

keimami keda kemunii ira

power to learn

There are many variations between the English and Fijian extremes. Some languages have separate double pronouns while others do not; Some systems distinguish between inclusive and exclusive pronouns, others do not. However, all languages of the world have distinct first and second person pronouns, and most languages have third person pronouns, first person inclusive pronouns, and first person exclusive pronouns. A four-person system (including first person and first, second, and third person) is by far the most common. Therefore, the four-person pronoun system is a bit more basic than the two- or three-person pronoun system. In that sense, English is atypical. Variations in pronoun systems obey a set of universal rules. In order to discover these universals, we need to create a typology of pronoun systems. Some types of pronoun systems in world languages are systems with singular and plural forms - for example English, Spanish systems with singular, dual and plural forms - for example Sanskrit systems with singular, dual, experimental and plural forms - for example , Fijian Systems without inclusive/exclusive distinction in the 1st person plural, e.g. B. English, Spanish systems with inclusive/exclusive distinction in the 1st person plural, e.g. B. Tok Pisin, Fiji

Some types of pronominal systems that do not occur Systems without first and second person pronouns Systems with singular and dual forms but without plural forms Systems with singular, dual and experimental forms but without plural forms Systems that make inclusive/exclusive distinction but not in the first person (a logical impossibility)

Based on what we find and don't find in our typology, we posit some universal rules. Some Universal Rules 1. 2. 3. 4.

All languages have at least first and second person pronouns. If a language has singular and dual forms, it will also have plural forms. If a language has singular, dual, and experiential forms, it will also have plural forms. When a language makes an inclusive/exclusive distinction in its pronoun system, it will do so in the first person.

Please note that the reverse of these rules does not apply. For example, the converse of Universal Rule 2 would state that if a language had separate plural forms, it would also have separate dual forms. But English, too, refutes this generalization: it has separate plural forms but no dual forms. Therefore, the implications are only one-way.

Phonological universals It is important to note that semantic typologies and universals are not measures of complexity in language or culture. All we can conclude from these differences is that some categories are more pronounced in some cultures than in others. Comparing the two examples of semantic universals discussed in this section, we also see that the English pronoun system is one of the most restricted in the world, despite the fact that English has very extensive academic and colored lexicons, to name just two areas . Thus, different areas of the lexicon show different degrees of elaboration in different languages. This is not to say that some languages are "richer" or "better" or "more developed" than others.

Phonological Systems with Universal Vowels Another level of language structure at which we can identify universal rules and classify languages into useful typological categories is phonology. In Chapter 3 we discussed the fact that languages can have very different phonetic inventories. Figure 7-1 depicts the standard American English vowel system, classified by place of articulation. Compare this to Figure 7-2, which shows the vowel system of Parisian Standard French (a conservative dialect that retains certain opposites lost in many other French dialects). The symbol /ü/ represents a high front rounded vowel as in the word /ʁü/ rue 'street'; /ø/ is an upper middle rounded vowel as in /fø/ feu 'fire'; // is a lower middle rounded vowel as in /bʁ/ beurre 'butter'; and /7/, /8/, /9/ and /0/ are nasal vowels. Finally, Figure 7-3 on page 222 compares the vowel systems of Quechua (spoken in Peru and Ecuador) and Hawaiian.

Figure 7.1 Vowels in American English

your υ is my ɑ

 identical

and and

and

Figure 7-2 Oral and nasal vowels in Parisian French

you oral

Nasal

and and

and

he is

9 0

• 221

2 2 2 • Chapter 7 Language universals and language typology

Figure 7.3 Quechua and Hawaiian vowels

or for the Quechua

Me and

you are hawaiian

The first thing these four examples show is that different languages can have very different vowel sets: English has many vowels in its inventory that French doesn't have, and vice versa. Second, the number of vowels in a language can vary significantly. Quechua only has 3 different vowels; Along with the vowel systems of Greenlandic Eskimo and Moroccan Arabic, the Quechua vowel system is one of the smallest in the world. Hawaiian has 5 vowels, a very common number among world languages. At the other end of the spectrum, English has 13 vowels and French has 15, including the four nasal vowels. However, behind this diversity we find universal patterns. If we were to map the vowel inventories of all known languages, we would confirm that languages in general have vowel systems that fall between the two extremes of Quechua and French. Therefore, every language has at least 3 vowel phonemes. Some have 4 vowels, such as Malagasy, the language of Madagascar (whose vowels are /i ε ə υ/), and Native American Kwakiutl (with /ia ə υ/). Some have 5 vowels, such as Hawaiian, Mandarin Chinese, and, as shown in Tables 3-3 (page 93) and 3-5 (page 94), Spanish and Japanese. Others, like Persian and Malay, have 6 vowels; and so on through 15. Comparing the graphs, we find that all languages have in their vowel inventory a high unrounded front vowel (/i/ or //), a low vowel (/a/), and a high back vowel more rounded Vowel (/u/ or /υ/) or unrounded vowel (/ɯ/). These vowels have allophones in some languages, especially languages with few vowels. For example, in Greenlandic Eskimo, /i/ has the allophones [i], [e], [ε], and [ə], depending on the surrounding consonants; but there are no minimal pairs that depend on these variants. Minor variations also exist, but these variations don't really contradict the universal rule, which can be formulated as follows: All languages have a front high unrounded vowel, a low vowel, and a back high rounded or unrounded vowel in their inventory of phonemes. Note that this first universal rule describes what makes up the minimal type and what is contained in all other types. The second universal rule is stated: Of the languages that have four or more vowels, all have similar vowels like /i to u/ (as indicated by the first universal rule) plus a central high vowel /i–/ (as in Russian vi-'usted') or an unrounded middle front vowel /e/ or /ε/. The third universal rule that we can discover in our vowel tables is: languages with a five-vowel system contain a leading unrounded vowel. For example, in the Hawaiian five-vowel system, /e/ has the allophones [ε] and [e]. Other languages with inventories of five vowels are Japanese (whose inventory is /i ε a ɔ ɯ/) and Zulu (/i ε a ɔ u/). Most five-vowel languages have a mid-back rounded vowel (or /ɔ/ or /o/) in their inventory, such as Japanese, Hawaiian, and Zulu. Some languages with a five-vowel system lack a rounded back vowel, although a similar sound is often included, such as B. Mandarin, whose constituents (/i ü to ë u/) contain the unrounded back vowel /ë/.

Phonological universals We can thus conclude that five-vowel stock languages usually (but not always) have a rounded vowel. This observation also applies to languages with more than five vowels. The fourth universal rule is therefore: languages with five or more vowels in their stocks tend to have a rounded vowel phoneme in the middle. This rule is worded differently than the first three rules because it is not absolute. But it's a useful observation because it describes a significant trend across languages. Languages with six-vowel inventories, such as Malayalam (spoken in south-western India) contain /ɔ/ in their inventory and // or /e/. Malayalam has in its inventory the three "obligatory" vowels /i through u/; the vowels /e/ and /ɔ/ predicted by the second and third universal rules; Y //. These universal rules can be summarized as shown in Figure 7-4. FIGURE 7-4 Summary of Universal Vowel Rules LANGUAGE TYPE

Example of the number of vowels

1 you and me

ε ɔ

or and

3 Quechua

4 Malagasy

5 Hawaiian

6 strategies

Nasal and oral vowels Other universal rules governing the vowel stocks of world languages can be discovered, but we shall mention only two more. The first states that if a language has nasal vowels, the number of nasal vowels never exceeds the number of oral vowels. We find examples of languages with fewer nasal vowels than oral vowels: Standard French, for example, has four nasal vowels and eleven oral vowels. We also find examples of languages with equal numbers of nasal and oral vowels: Punjabi (a North Indian language) has ten each. But there are no languages with more nasal vowels than oral vowels. The second universal rule of interest is not a rule in the usual sense, but a description of the most common vowel system: a five-vowel system consisting of a front high unrounded vowel (/i/ or //), a front vowel middle unrounded vowel (/e/ or /ε/), a low back vowel (/a/), a back middle rounded vowel (/o/ or /ɔ/), and a back high rounded vowel (/u/ or /u/ ). Hawaiian is an example of this system, as you can see from the symmetry in the Hawaiian Vowel Chart (page 222). Each vowel is placed as far away from the others as possible, minimizing the chance of confusing two vowels. This five-vowel system therefore has an ideal quality, which we shall return to later in this chapter.

Consonant vowel systems are not the only area of phonology where universal rules apply. The consonant inventories of the world languages also show many universal properties.

• 223

2 2 4 • Chapter 7 Language universals and language typology

A few examples are presented here, although not in great detail, since they are not different in nature from the universals of the vowel systems. Recall (from Chapter 3) that the sounds /p t k/ are voiceless plosives. Every language has at least one of these silent stops as a phoneme. While some languages lack affricates or trills, unvoiced stops are found in all languages. In fact, most languages have all three of these sounds, even languages with small consonant inventories. Niuean (Polynesian language), for example, has only three stops, three nasals, three fricatives and one approximative, making a total of ten consonants (as opposed to twenty-four in American English). However, all three stops are /p t k/. Put in universal form, this generalization reads: Most languages have all three registers /p t k/ in their consonant inventory. This universality suggests that in some sense these three consonants are more fundamental than others. It is clear from our discussion that this universal is not an absolute rule. Hawaiian (a language related to Niuean) only has /p/ and /k/. (Hence English words with the /t/ sound are borrowed into Hawaiian with /k/, like kikiki 'ticket'.) This universal is therefore a tendency, not a statement of what is and is not found among the languages of the world. Another important universal regarding stops has already been mentioned. Remember that the difference between the two sets of registers /p t k/ and /b d g/ is that the first set is unvoiced and the second set is voiced. All six sounds have phonetic status in English, French, Spanish, Quechua, and many other languages. However, in some languages we only find unvoiced stops, such as Hawaiian (and all other Polynesian languages), Korean, and Mandarin Chinese. So far we have identified two types of languages: languages with voiced and unvoiced stops and languages with only unvoiced stops. As mentioned, every language has at least one silent stop in its inventory; Consequently, there are no languages that have both voiced and unvoiced stops, and there are no languages that do not have both voiced and unvoiced stops. From this typology, the following universal rule can be derived: no language has voiced stops without unvoiced stops. Note that of the register register universals examined so far, only one rule (and it is only a trend) says anything about which registers are included in the language registers. But there are other universals that address this issue. Here we give just one example: if a language has no stop, there is a strong tendency for that language to include in its inventory a fricative with the same place of articulation as the missing stop. For example, Standard Fijian, Amharic (the main language of Ethiopia), and Standard Arabic do not have the phoneme /p/, which is a labial stop. As the universal rule predicts, all of these languages have a fricative /f/ or /v/ whose place of articulation is similar to that of /p/. The fricative "fills" the missing stop. This rule is also just a trend as there are languages that break it. Hawaiian, which lacks /t/, has none of the corresponding fricatives /ð/, /θ/, /z/, or /s/. But most languages follow the rule.

Syntactic and Morphological Universals Word Order Speakers of English and other European languages often assume that the normal way of forming a sentence is to put the subject of the sentence first, then the verb,

Syntactic and morphological universals, and then the direct object (if any). In fact, in English, the expression Mary saw John following this order is well-formed, while variations such as John Mary saw and Mary saw John are not well-formed. However, the normal order of words in a sentence differs significantly from language to language. Consider the following Japanese sentence, where the subject is a girl named Akiko, the verb butta is "to hit (over)", and the direct object is a boy named Taro. Akiko

Georgia

bread from

Akiko Asunto Taro

Object

throw punches

Akiko atiu Taro.

In Japanese, therefore, normal word order is subject first, then direct object, and last verb. If we were to change this order (to make Japanese syntax compatible with English, for example), the result would be ungrammatical. Now let's look at Tongan, where the verb should come first, the subject second, and the direct object last. In the following sentence, the verb taai is "to hit", the subject is a person named Hina, and the direct object is a person named Vaka. naʔe taaʔi ʔe Last blow

gris a

Wake up

Subject Hina Object Vaka

'Hina hit Vaka.'

Of course, not all English sentences follow the order of subject-verb-direct object or SVO. To emphasize certain noun phrases, English speakers sometimes place direct objects at the beginning of a sentence, as in Sewing I hate it, but I'll do it for you. For questions like Who(m) did you see? the direct object who(m) comes first. Similar variations in word order are found in most languages of the world. However, phrases like sew and examples of which are derived from simpler phrases are also less common than phrases that follow the SVO order. So, although some English constructions do not follow this order, let's say that the SVO order in English is "simple" and that English is an SVO language. Examples of SVO languages are Romance languages (such as French, Spanish, and Italian), Thai, Vietnamese, and Indonesian. Japanese is an SOV language, along with the Native American languages Turkish, Persian, Burmese, Hindi and Navajo, Hopi, and Luiseño. Tongan is a VSO language, along with most other Polynesian languages, some dialects of Arabic, Welsh, and several Native American languages such as Salish, Squamish, Chinook, Jacaltec, and Zapotec. Besides VSO, SVO, and SOV, there are three other logical ways to combine verbs, subjects, and direct objects. Surprisingly, however, very few languages have VOS, OVS, or OSV as their basic word order. Few languages are VOS, the most well-known being Malagasy and Fijian. Below is a simple sentence in Fijian showing that the direct object comes before the subject. and the

taja

na ŋone na yalewa

past

beat

the boy the girl

"The girl hit the boy."

OVS and OSV are the basic word order of only a handful of languages in the Amazon Basin, including Hixkaryana (OVS) and Nadëb (OSV). By far the most common word orders among world languages are SVO, SOV and, to a lesser extent, VSO.

• 225

2 2 6 • Chapter 7 Language universals and language typology

Try it yourself: What characterizes the order of S and O in the languages SVO, SOV and VSO (the most common) and distinguishes them from the languages VOS, OVS and OSV (the rare ones)? In each of the three common configurations, S precedes O; in unusual configurations S follows O. So we can make a general statement: in the basic word order of world languages there is an overwhelming tendency for the subject of a sentence to precede the direct object. Universals are much more than just syntax. Two extreme cases are languages in which the verb appears first in the clause (so-called verb-final languages and illustrated by Tongan) and languages in which the verb appears last (so-called verb-final languages and illustrated by Japanese). . For the sake of simplicity, we have excluded the VOS and OSV languages from our discussion, although they basically follow the same rules as the VSO and SOV languages, respectively.

Possessive and possessed noun phrases If we look at the order of other syntactic components in verb-initial and final languages, we find surprisingly regular and interesting patterns. First, in most verb-final languages, such as B. Japanese, possessive noun phrases precede possessive noun phrases. taroo no imooto Taro's sister 'Taro's sister'

In languages with initial verbs, the reverse order is more common; in the following example from Tonga, the possessed entity is expressed first, the possessor last. ko e tuongaʔane ʔo vaka el

sister

by Waka

'Vakas Schwester'

We therefore establish the following rule: There is a strong tendency for possessed noun phrases to precede possessed noun phrases in verb-finals and to follow possessed noun phrases in verb-initials.

Prepositions and Postpositions Many languages use prepositions to express position or direction. As the word suggests, prepositions come before modified noun phrases (NPs). For example, in Tongan, the prepositions ki denoting direction and ʔi denoting place precede the NP they modify. Ki Tonga

ʔno

Tongan

in Tonga

Other languages have postpositions instead of prepositions. Postpositions perform the same functions as prepositions but follow the NP, as in this Japanese example.

syntactic and morphological universals takeyoo

Tokio

Pro

'to Tokyo'

Predominantly verb-initial languages have prepositions and verb-final languages postpositions. The third rule can be formulated as follows: There is a strong tendency for verb-initial languages to have prepositions and verb-final languages to have postpositions.

Relative clauses Depending on the language, relative clauses come before or after main nouns. In relative clause constructions in English (the book that Judith wrote), the relative clause (that Judith wrote) follows its head (the book). The same goes for Tonga. ko and tohi o

[naʔe faʔu

last book

Hina]

write subject as

'the book that Hina wrote'

In Japanese, however, the relative clause comes before its head. [Hiroo

Georgia

fence] love

Hiro

object

written book

'the book that Hiro wrote'

The vast majority of verb-initial languages place the relative clauses after the main noun, and the vast majority of verb-final languages place the relative clauses before the main noun. Thus we can observe the following generality: There is a strong tendency in verb-initial languages to place relative clauses after the main noun and in verb-final languages to place relative clauses before the main noun.

General Ordering Patterns We have found that Verb Initial (VSO) languages place possessors after possessed nouns, relative clauses after main nouns, and have prepositions. Verb-final (SOV) languages, on the other hand, put possessors before possessed nouns, put relative clauses before main nouns, and have postpositions. In all of these contexts, a pattern emerges. Note that possessors and relative clauses modify nouns; The noun is a more important element of a noun phrase than either of the modifiers. Similarly, noun phrases modify prepositions or postpositions; similarly, although not intuitively obvious, the most important element of a prepositional phrase is the preposition itself, not the noun phrase; It is the preposition that makes it a prepositional phrase. Finally, the direct object in a verb phrase modifies the verb. Given these observations, we can make a generalization about the order of constituents in different types of languages: in verb-initial languages, the modifying element follows the modified element, while in verb-final languages, the modifying element precedes the modified element . . This pattern is shown in Table 7-1 on page 228. Of course, this generalization is based on trends rather than absolute rules. At each level of the table, some languages violate the mappings. Persian is for example

• 227

2 2 8 • Chapter 7 Language universals and language typology

TABLE 7-1 Summary of Constituent Orders EARLY VERBAL LANGUAGES (EXAMPLE: TONGAN)

VERBAL LANGUAGES (EXAMPLE: JAPANESE)

Modified - Modifying verb - Direct object possessed - Preposition of possessor - Noun phrase main noun - Relative clause

Modifier - modified direct object - possessive verb - possessed noun phrase - relative postpositional clause - main noun

an SOV language such as Japanese and therefore must have the characteristics listed in the right column of the table. But in Persian, possessors follow possessed nouns, prepositions are used, and relative clauses follow principal nouns, all of which are properties of verbal languages. However, such counterexamples to correlations are rare. Note that our discussion didn't mention anything about spoken languages (SVO) like English. These languages do not seem to follow a consistent pattern. English, for example, puts relative clauses after main nouns and has prepositions (both properties of verb-initial languages). Regarding the order of possessors and possessed nouns, English has both standards (man's poor and man's poor). In contrast, Mandarin Chinese, another verb-medial language, has characteristics of verb-final languages. Word order universals exemplify the level at which linguists attempt to describe the universal properties of language. Table 7-1 suggests that in the structure of virtually all verb-final and verb-initial languages, the same organizing principle is at work at the noun phrase, prepositional phrase, and whole sentence level. This is remarkable because it applies to many languages whose speakers have never come into contact with one another. It is therefore likely that a cognitive process common to all humans underlies this organizing principle.

Hierarchy of Relativization Another area of syntactic structure where striking universal principles are found is in the structure of relative clauses. English can relativize the subject of a relative clause, direct object, indirect object, obliques and possessive noun phrases (see Chapter 5). The following examples in English illustrate these different possibilities. the teacher [who was speaking at the meeting] (subject) the teacher [whom I was speaking of - to you] (direct object) the teacher [whom I was telling the story to -] (indirect object) the teacher [whom I was telling the story to - ] (oblique) the teacher [whose book I read] (owner)

Other languages do not allow all these possibilities. Some languages only allow relativization in some of these categories, but not in others. For example, a relative clause in Malagasy is grammatical only if the relativized noun phrase is the subject of the relative clause.

Syntactic and morphological universals ny

student [who

the student

wer

Women]

Women

'the student who saw the woman'

In Malagasy, there is no way to directly add a relative clause whose direct object has been relativized ('the student whom the woman saw'), or the indirect object ('the student whom the woman gave a book'), or an oblique clause translate an ('the student from whom the woman heard the news') or an owner ('the student whose book the woman read'). When Malagasy speakers need to convey what these relative constructions represent in English, they need to passivate the relative clause so that the noun phrase to be relativized becomes the grammatical subject of the relative clause ("the student who was seen by the woman"). . Alternatively, they can express their idea in two sentences, ie instead of 'the woman saw the student who failed the exam' they could say 'the woman saw the student and the same student failed the exam'. They have relative clauses, in which direct subjects or objects can be relativized, but indirect, oblique, or possessive objects cannot. An example of such a language is Kinyarwanda, spoken in East Africa. Other languages, like Basque, have relative clauses in which the subject, direct object, and indirect object can be relativized, but not obliquely or possessively. Another kind of language adds forward slashes to the list of relatable categories; This is the case with Catalan, which is spoken in north-eastern Spain. After all, languages like English allow for all possibilities. Table 7-2 recapitulates the types of relative clause systems found in the languages of the world; the plus sign indicates a part of speech that can be put into perspective, the minus sign one that cannot be put into perspective. Note that a plus sign does not imply anything about the characters to its right; they can be more or less. A plus sign implies that all categories to the left of it can be relativized. It is a notable fact that there are no languages in which, say, an oblique can be relativized ("the man [from whom I heard the story]"), but neither are there subjects, direct objects, and indirect objects. In fact, the formation of relative clauses in all languages is sensitive to a hierarchy of grammatical relationships: Hierarchy of relative clauses subject ⬍ direct object ⬍ indirect object ⬍ oblique ⬍ possessor

TABLE 7-2 Relativization Hierarchy: Types of Relative Clause Systems LANGUAGE TYPE

OBJECT

DIRECT OBJECT

INDIRECT OBJECT

ASLANT

OWNER

1 2 3 4 5

⫹ ⫹ ⫹ ⫹ ⫹

– ⫹ ⫹ ⫹ ⫹

– – ⫹ ⫹ ⫹

– – – ⫹ ⫹

– – – – ⫹

EXAMPLE

Malagasy Kinyarwanda Basque Catalan English

• 229

2 3 0 • Chapter 7 Language universals and language typology

The hierarchy predicts that if a language allows relativization of a particular category in the hierarchy, then that language's grammar also allows relativization of all positions to the left of it. For example, possessors in English can be relativized ("the woman [whose book I'm reading]"). The hierarchy predicts that in English all positions to the left of the possessor can be relativized (i.e. oblique, indirect object, direct object and subject). The hierarchy also predicts that Basque, which allows relativization of indirect objects, will allow relativization of direct objects and subjects; Basque does not allow the categories to the right of the indirect object to be relativized in the hierarchy (oblique or possessive). The hierarchy is thus a concise description of the types of relative clause formation patterns found in the languages of the world.

Types of Language Universals In this section, we build on the universals discussed in the previous sections to classify the different types of universals. It should be clear by now that linguistic universals are not all created equal. Some have no exceptions. Others apply to most languages, but not all. It is important to distinguish between these two types of universals, since the first type seems to be the result of an absolute limitation of language in general, while the other is the result of a bias.

Absolute Universals and Universal Tendencies The first two types of universals are distinguished by the fact that they can be stated as absolute rules. The typology of vowel systems presented above shows that the minimum number of vowels in a language is three: /i to u/. The two universal rules proposed by the typology are the following: 1. All languages have at least three vowels. 2. If a language has only three vowels, these vowels are /i through u/.

From the descriptions of all the languages studied so far, it is clear that these two rules have no exceptions. The two rules are therefore examples of absolute universals, universal rules that have no exceptions. Other examples of absolute universals are: if a language has a set of dual pronouns, it must have a set of plural pronouns; If a language has voiced stops, it must have unvoiced stops. Unlike absolute universals, many universal rules have some exceptions. A good example is the rule that if a language has a gap in its stop inventory, it is likely to have a fricative with the same place of articulation as the missing stop. This rule applies to most languages that have gaps in their inventory, but not all. Such rules are called universal tendencies. (One possible explanation for universal biases is that they represent the union of partially competing universal rules.) Of course, researchers must be careful when deciding that a given rule is absolute. Until a few years ago it would have been easy to assume that there was no language with OVS or OSV as base word order (since none has been described) and that there was a universal absolute stating that "no language has OVS or OSV". for basic words." Word order". However, we now know of a few OVS and OSV languages, all of which are spoken in the Amazon region.

Explanations for the Universal Languages Bowl. Thus, the rule that was declared absolute generality appeared absolute only because no one had found a language that violated it.

Implicit and Nonimplicative Universals Notwithstanding the contrast between absolute and biased universals, we can make another important distinction: between implicative and nonimplicative universals. Some universal rules take the form of conditional implication, as in the following examples: • When a language has five vowels, it usually has the vowel /o/ or /ɔ/. • If a language is verb-final, then possessors in that language will likely precede possessed noun phrases.

All rules of the form "if the condition P is satisfied, then the conclusion Q is satisfied" are called implicative universals. Other universals can be declared unconditionally: all languages have at least three vowels. These universals are called non-implicative universals. So there are four kinds of universals. Types of Universals Absolute Implicational Universal If a language has property X, it must have property Y. Implicational Bias If a language has property X, it will likely have property Y property property X. Bias non-implicative Most languages have property X.

Explanations of linguistic universals It is remarkable that all languages of the world, due to their extreme structural diversity, fall into clearly defined types and are subject to universal rules. So it's reasonable to ask why there are universal rules. The subject is extremely complex and no one has come up with a definitive explanation for anything universal. But for many universals we can hypothesize, or at least conjecture, about the reasons for their existence.

Original Language Hypothesis The first explanation for linguistic universals that might come to mind is that all the languages of the world historically descended from the same original language. However, this hypothesis is difficult to support. First, the archaeological evidence strongly suggests that the ability to speak evolved among our ancestors in different parts of the world at about the same time, and it is difficult to imagine that different groups of speakers were out of contact with one another

• 231

2 3 2 • Chapter 7 Language universals and language typology

together they would have developed exactly the same language. Second, even if we ignore the archaeological evidence, it is impossible to prove or disprove the existence of an original language because we have no evidence for or against it. So the original language hypothesis is not a very good explanation; it is so hypothetical that it does not adequately serve the explanatory function.

Universals and Perception A more likely hypothesis to explain language universals is that they are symptoms of how all people perceive the world and engage in verbal interactions. In the following sections some of these explanations are applied to the universals set out earlier in this chapter. When looking at vowel systems, you may have noticed that the three vowels found in all languages, /i through u/, are far apart on a vowel chart. The two vowels /i/ and /u/ differ in frontality and general roundness, and /a/ differs from the other two in frontality and tone. From these observations it is easy to infer why these three vowels are the most basic vowels in all languages: there is no other group of three vowels that differ more dramatically from one another.

Acquisition and Processing Explanations Some world languages have psychological explanations with no physiological basis. Proposed explanations for word-order universals, for example, are based on the idea that the more regular the structure of a language, the easier it is for children to acquire. Therefore, the fact that verb-initial languages have prepositions and place adjectives after nouns, possessors after possessed nouns, and relative clauses after principal nouns can be summarized by the following rule: In verb-initial languages, the modifier follows the modified element. . Languages that strictly follow this rule show great regularity from one construction to the next; a single ordering principle governs the order of verbs and direct objects, adpositions and noun phrases, nouns and adjectives, possessive and possessive nouns, relative clauses and nouns. It seems that such a language would be easier to acquire than a language with two or more ordering principles underlying different areas of syntax. The fact that so many languages in the world follow one or another general pattern of order (modified-modified) or another (modified-modified) with such regularity thus reflects the proposed general trend for language structure to be as regular as possible and to acquire so make it easy as possible. Psychological explanations have also been proposed to account for the relative hierarchy of sentence formation. Relative clauses in which the head acts as the subject of the relative clause ('a mulher [que partiu]') It is easier to learn and understand which relative clauses in which the head acts as the direct object of the relative clause ('o homem [the I saw]'). Young children usually acquire the first type before beginning the second type. Also, it takes people less time to understand the meaning of relative clauses in subjects than in direct objects. In turn, direct object relative clauses are easier to understand than indirect object relative clauses, and so on in the hierarchy: subject ⬍ direct object ⬍ indirect object ⬍ oblique ⬍ possessor. So there is a psychological explanation for cross-linguistic patterns in the typology of relative clause formation: A language only allows a "difficult" type of relative clause if all "easier" types are also allowed in the language.

Explanations of the universals of language

Social Explanations Finally, remember that language is a cognitive and social phenomenon. Some universals of language have a basis in cognition; others reflect the fact that language is a social tool. The universals of pronoun systems can be explained in terms of usage of the language. For example, why do all languages have first and second person singular pronouns? Remember that the most basic type of verbal interaction is face-to-face. Other contexts in which language is used to communicate (written, over the phone, over the air, etc.) are relatively recent inventions compared to the ability to hold a conversation; They occur less frequently and perhaps less naturally than face-to-face interactions. In a face-to-face interaction, it is important to be able to relate efficiently and concisely to the speaker and the addressee, the two most important entities involved in the interaction. A discussion between two people who don't need you and me, or who need to call each other by name, would be noticeably less efficient. Obviously, the first and second person singular pronouns are essential for normal efficiency in social interaction. It is not surprising, then, that all languages have first- and second-person singular forms of pronouns, even though they may have a gap in other parts of their pronoun system. So the generality that all languages have first and second person pronouns is socially motivated. Also, as noted above, the most common pronoun system has separate first, second, and third person forms, and separates the first person inclusive ("you and I, and maybe other people") and exclusive ("other people"). . and I, but not you'). Why would this system be so widespread and somehow more fundamental than other systems? Pronoun systems can be viewed as a matrix, with each place in the matrix identified by whether the speaker and addressee are included in the pronoun reference, as in Table 7-3.

TABLE 7-3 Matrix of pronoun systems

RECIPIENT INCLUDED RECIPIENT EXCLUDED

SPEAKERS INCLUDED

CLAIMERS eliminated

—— first person plural including first person singular first person plural excluding

second person singular second person plural third person singular third person plural

Given that the speaker and the addressee are the most important elements in face-to-face interactions, it should come as no surprise that the inclusion or exclusion of the speaker and the addressee is the determining factor in defining any space in the Matrix . The most basic (and common) type of pronoun system is therefore the more balanced matrix, in which each space is filled with a separate form. Thus, language universals can emerge from the way people perceive the world around them, learn and process language, and organize their social interactions. Behind the search for universals is the desire to learn more about these areas of knowledge and social life.

• 233

2 3 4 • Chapter 7 Language universals and language typology

Computers and the Study of Language For more than half a century, universal researchers have attempted to develop devices that can translate between languages. However, this goal has, to a limited extent, eluded even the best attempts. As anyone who has ever visited a foreign country knows, a word-for-word translation is not enough. On the one hand, as we have seen, languages differ in their word order, and on the other hand, metaphors in one language may not translate to the corresponding metaphor in another language. Many other reasons also contribute to the failure of word-for-word translation, making even computerized bilingual dictionaries inadequate for each translated language. Consider two translation models. In the first model, one set of rules or procedures is used for the translation from language A to language B and a second set for the translation in the other direction, i. H. from B to A. The rules would have to be fully explicit and a set of methods in each direction since the translation is not symmetric. If a machine translation (MT) device were set up for up to six languages, 6×5 (i.e. 30) sets of procedures would be required to translate each language into the other five. Such a model, called a translational transmission system, can be represented as in the first figure below. Now consider an alternative model in which the basic semantic elements of any language can be abstracted and then encoded in other languages. In this model, each language would require one procedure to decode it into abstract semantic elements (thereby forming an abstract semantic representation), together with a second procedure to convert abstract semantic representations into the lexicon, syntax, and (for spoken texts ) to encode. Phonology of any language language.-target. Similar

One model is called a cross-language translation system and might look something like the second figure below. The interlingual translation model would require twelve procedures: a decoding procedure and an encoding procedure for each of the six languages. This model is much simpler than the transfer translation model, which requires 30 procedures. Unfortunately, it is not clear to what extent sentences can be decomposed into the kinds of abstract semantic representations that would be necessary for a cross-language model, particularly to make the intermediate representation language-neutral. The difficulties related to translation in one of the models relate to what one language encodes what another cannot. For example, as you saw in the table of their pronouns on page 220, Fiji has four different second-person pronouns, while English has only one. This would make it very easy to translate any Fijian second-person pronoun into English, as long as the abstract semantic representation of the Fijian pronouns includes the “second-person” element. English - that is, you. But what about the back translation? Would it be just as easy? For an English sentence containing the pronoun you, no machine could determine the underlying semantic representation from the form itself, except for the second person. In other words, since English does not encode the potential distinction between second-person singular, binary, experimental, and plural numbers, an MT device would not be able to decide which Fijian pronoun to choose if only by the Form of the English pronoun would depend on you. We can represent the first part of the problem in the following scheme where it would be easy to translate from Fijian to English.

Download English translation

French

Italian

fiyiano

Japanese

Irish

English

French

Italian

fiyiano

Japanese

Irish

Explanations of the universals of language

• 235

Interlingual translation from English

French

Italian

fiyiano

Japanese

Irish

abstract semantic representations of sentences

Fijian English as kemudrau kemudou kemunii

French

Italian

english you

But translating you from English to Fijian would be quite a challenge. Although the text or context can make clear how many recipients the pronoun you represent is, except in rare cases (e.g. you two, you three), it would be difficult or impossible for a TM program to decipher this information . As we just saw, it would be easy to switch from Fijian second-person pronouns to plain English second-person pronouns, but switching from English to Fijian is next to impossible. Curiously, the situation is reversed for third-person singular pronouns. In this case, Fiji does not distinguish between masculine, feminine and neuter singular pronouns, but English does; As you saw in the table on page 220, Fijian has only the koya pronoun, which corresponds to the three English pronouns he, she, and it. Fijian English? he koya? is he around Using the Fijian pronoun alone, no device could decide which English pronoun would be the correct translation. Again, context can make this clear, but it would be difficult for an MT device to decipher this information. Having highlighted several ways in which machine or machine translation would be difficult or impossible, it is also important to note that considerable progress has been made in the development of MT devices. The problems discussed can be minimized by restricting the translation machinery to two languages and very specialized language domains.

fiyiano

Japanese

Irish

course in these languages. For example, if you are translating only medical texts or only technical documents from English to another language, you will be dealing with a limited subset of vocabulary and structures. Analysis of certain types of text may show that certain lexical or grammatical options rarely (or never) appear in them. By creating a list of words that appear in medical journals, for example, most of the informal vocabulary can be eliminated from consideration. Coming back to the pronoun issue discussed above, it is almost certain that English-language medical journals would not use the full range of possible semantic distinctions represented in Fijian. Therefore, medical document translation projects might omit some potential Fijian pronouns for convenience. The need for translation has greatly increased in recent decades with the establishment of the European Union, and significant financial resources have been dedicated to researching machine translation. In the search for better machine translation methods, corpora containing more than one language have been created, and their analysis will yield discoveries that will be useful in the design of machine translation devices. Some multilingual corpora contain the same text types but not identical texts. One such corpus is the Aarhus corpus of Danish, French and English law. Their texts are not translations of one another, but constitute a reservoir of information about the language of legislation in these three languages. Other corpora contain texts translated from one another, as in the case of the Canadian Hansard Corpus, which contains parliamentary acts in translations from French into English and from English into French. When a corpus contains translations, it is possible to create a "parallel aligned corpus". It is a corpus containing texts in different languages, some of which have been automatically coordinated

2 3 6 • Chapter 7 Language universals and language typology

Computers and the exploration of linguistic universals that sections correspond paragraph by paragraph or sentence by sentence. Researchers can use these corpora to study the mathematical properties of vocabulary and syntax between languages and language pairs. Based on knowledge of such properties, machine translation

could partially avoid the complexity of any of the models described on pages 234 and 235. Instead, certain mathematical property languages would help determine the probable translation, regardless of how the human mind processes languages and performs translations. ■

Summary • Underlying the great diversity of the world's languages are universal principles that play a role at all levels of language structure: phonology, morphology, syntax and semantics. • The study of typology attempts to catalog languages by types, while the study of universals attempts to formulate the universal principles themselves. • In lexical semantics, the composition of pronominal systems in which interlinguistic variations are found is dictated by various universal rules for number and person distinctions. • Vowel systems and registers of the sounds are two examples of universals that play a role in phonology. • In syntax and morphology, there are universals that regulate the basic order of components in sentences and phrases. • In the syntax, the relativization hierarchy is a striking example of a universal principle. • The outstanding feature of all universals is that the most common patterns are the most regular and harmonious. • Four types of universal rules can be distinguished, depending on whether they have exceptions (absolute versus biased) and their logical form (implicative versus non-implicative): Absolute implicit universals: Languages with property X must have property Y. Implicit biases: Languages with property X are likely to have property Y. Non-implicative absolute universals: All languages have property X. Non-implicative tendencies: Most languages have property X. • Ultimate The goal of studying linguistic universals is to provide explanations for such universal principles. • The universals of language can have physiological, psychological or social explanations. Physiological: Universals often indicate how we perceive the world around us. Therefore, languages tend to emphasize categories that are physiologically and perceptually salient, such as B. Vowels. Psychological: Structural simplicity and consistency make languages easier to acquire and assimilate. Therefore, many universals predict that the simplest and most consistent systems will be preferred. Social: Distinctions on the expression side of language reflect important social distinctions on the content side.

What do you think? checked

WHAT DO YOU THINK ? R E V I S I T E D • Your third grade niece. Many people tend to think of languages in terms of their written form rather than their spoken form. This is why so many people report that English has 5 vowels (a, e, i, o and u). But Figure 7-1 (p. 221) shows that your niece is right. In addition to the 13 vowels, English has 3 diphthongs, as in the words buy, toy, and cow. Languages differ in the number of vowels they have. Quechua, Greenlandic Eskimo, and Moroccan Arabic have 3 vowels; Hawaiian has 5; English 13; Parisian French 15. • Visiting cousins in Chicago. You may have noticed that many English speakers will tell you if they are addressing one or more people, but other speakers use a different form for more than one addressee; they say you, you, or you (you in western Pennsylvania and parts of the Ohio Valley). You or some of your friends may say you when addressing more than one person. As shown in the pronoun tables on pages 218–220, it is not uncommon for languages to have different singular and plural forms for second-person pronouns, sometimes even distinguishing two addressees from more than two. Fiji, for example, has four different forms: one for one person (singular), one for two persons (dual), one for three (intent) and one for more than three (plural). It should be noted that English has different configurations in its dialects, with some varieties, but not all, distinguishing between second-person singular and second-person plural pronouns. They seem to be used more and more by speakers whose stem does not have a distinct second-person plural form. • Introductory Japanese classmate. It seems inevitable to think that language itself is natural and logical, and to suspect that other languages are weird or illogical when they deviate from it. Perhaps nowhere is this more true than in word order. Like other aspects of language, however, word order is not a matter of logic. While some languages have an SVO order, others have an SOV or some other order. German uses SVO in main clauses and SOV in subordinate clauses, while English uses SVO for main clauses (I bought it) and subordinate clauses (because I wanted it). Although the basic word order in English is SVO, other word orders are possible, e.g. B. in sentences like Peas I like. The OSV order of Peas I like is just as logical (or illogical) as the SVO order of I like peas. However, as this chapter shows, some word order patterns in a language often go together, inviting hypotheses as to why they do so.

• 237

2 3 8 • Chapter 7 Language universals and language typology

Exercises Practice exercises

The statements in sentences A and B contain universal statements, some of which are true and some of which are false. Identify which set contains absolute universals and which set contains implied universals, and for each universal state, whether it is true or false. Then say for the set C whether each statement is true or false. A

I. Every language has vowels and consonants. ii. Every language has subject-verb-object as its basic word order. iii. All languages have the same number of personal pronouns. 4. Each language distinguishes between singular and plural “you” in its pronouns. v. Every language distinguishes between “we” inclusive and exclusive in its pronouns.

I. If a language has only three vowels, they can all be back vowels. ii. If a language has only three vowels, they can all be back vowels. iii. If a language has only three vowels, they can all be high vowels. 4. If a language has stop consonants, they can all be expressed.

I. Linguistic universals probably exist because all languages have historically descended from an original language. ii. There is no limit to the ways in which human languages can differ from one another. iii. Some language universals have a basis in human cognition. 4. Some world languages are based on human social interaction.

Based on English and Other Languages 7-1.

Rate how common or unusual the following features of standard English are compared to other languages discussed in this chapter. Justify your judgment in each case. 1) a 13-vowel system 2) no nasal vowels (phonemically different) 3) subject-verb-object word order 4) adjectives before main nouns 5) relative clauses after main nouns 6) no dual pronoun forms 7) no experimental pronoun forms 8) no other pronouns as the second person plural 9) no distinction between inclusive and exclusive pronouns

7-2.

Determine whether each of the following statements is an absolute implicit universal bias, an absolute nonimplicit universal bias, an implicit universal bias, or a nonimplicit universal bias.

Exercises 1) The consonant lists of all languages contain at least two different stops, which differ in the place of articulation. 2) Languages always have fewer nasal consonants than verbal stops. 3) In all languages, the number of front vowels of different pitch is greater than or equal to the number of back vowels of different pitch. 4) Most VSO languages have prepositions, not postpositions. 5) Diminished particles and affixes tend to have high vowels. 6) If a language has separate terms for "foot" and "leg", it must also have separate terms for "hand" and "arm". 7) The future tense is used in many languages to express hypothetical events, and the past tense is often used to express non-hypothetical events. 8) Languages that have relatively free word order tend to be inflected for case. 9) Many languages with verbs at the beginning put relative clauses after the relative clause heading. 7-3.

In English, conditions can be expressed in two ways: the condition first and the condition afterwards, as in (1) below, or the condition after and the condition first, as in (2). However, in several languages only the first pattern is grammatical. In Mandarin Chinese, the conditional clause must come first, as in (3); when it comes second, as in (4), the resulting string is ungrammatical. No language allows only patterns (2): second conditional clause, first conditional clause. 1) If you cry, I turn off the TV. 2) I turn off the TV when you cry. 3) rúguO wo dìdi he jiU wo jiù hen shengqì When my younger brother drinks wine, I get very angry "When my younger brother drinks wine, I get very angry". 4) *wo hen shengqì rúguO wo dìdi he jiU I get very angry when my younger brother drinks wine a. From this information, formulate descriptions of an absolute implied universal, an absolute nonimplicative universal, and a tendency universe, all relating to conditional clauses. B. Suggest an explanation for the universal patterns of order you formulated in (a). (Hint: Remember the order in which the actions specified by condition and condition must be performed.)

7-4.

The composition of the vowel inventories of world languages is predicted using the hierarchy shown in Figure 7-4 (p. 223). The hierarchy provides for the compilation of an inventory of vowels consisting of six phonemes. Complete the next step in the hierarchy by determining the composition of the seven vowel inventories. Use the following information about the composition of seven-vowel inventories from three languages that you should assume are representative of possible seven-vowel inventories. Sundanese Burmese Washkuk

ieεaɔou isεaɔuə ieεaɔu

• 239

2 4 0 • Chapter 7 Language universals and language typology

7-5.

Consider the following typology of pronoun systems found in the languages of the world. The first column of each sentence represents singular pronouns; the second column, double pronouns; and the third column, plural pronouns. A language example is also provided for each type (incl. = inclusive, excl. ⫽ exclusive). 8-pronoun systems 1) I we-2 you you-2 he/she 2) I you you-2 he/she she-2 3) I we-2-incl. you/er Systems of 9 pronouns 1) I we-2 you you-2 he/she she-2 2) I we-2-incl. we-2-excl. you he/she 3) I we-2-incl. we-2-excl. you you-2 he/she

we you they we we you they we-incl. you they we you they we-incl. us excl. you they we

Greenland Eskimo

Arabica

Southern Paiute (North America)

Lapland (arctic Scandinavia)

Maya (Mittelamerican)

Lower Kanauri (India)

you them

10 pronoun systems 1) I we-2-incl. we-2-excl. you you-2 he/she she-2 2) I we-2-incl. we-2-excl. you you-2 he/she she-2

you they we-incl. us excl. you them

11 pronoun systems 1) I we-2-incl. we-2-excl. you you-2 he/she she-2

us-incl. us excl. you them

Coos (North America)

kanauri (if)

hawaiisch

Exercises 2) I you

we-2-incl. we-2 excl. you 2

us-incl. us excl. your

is he around

she-2

she

• 241

Sheep (West Africa)

he and she a. Based on this data, which you can assume to be representative, formulate a set of absolute universal principles that describe the composition of the 8-, 9-, 10-, and 11-pronoun systems. Make your principles as general as possible. B. Of these systems, the most common is the 11 Type 1 pronoun system, exemplified by Hawaiian, followed by the 9 Type 1 pronoun system, exemplified by Lapp. Formulate a set of universal trends that describe the predominance of examples of these two systems. 7-6

Logically, the possible basic subject, verb, and direct object order combinations are SOV, SVO, VSO, VOS, OVS, and OSV. We have seen that there is wide variation in the percentage of languages that show each combination as a basic word order. Linguists have recognized this fact for several decades, but there is little consensus on the exact distribution of these basic word-order variations in the languages of the world. These are the findings of five researchers who conducted multilingual analyzes of the distribution of basic word order possibilities. (The figures are from Tomlin 1986.)

INVESTIGATOR

LANGUAGE EXAMPLE SOV

AFTER

Greenberg Ultán Ruhlen Mallinson/Blake Tomlin

30 75 427 100 402

43 34,6 35,6 35 41,8

37 44 51,5 41 44,8

PERCENTAGES VSO VO

20 18,6 10,5 9 9,2

0 2,6 2,1 2 3,0

OVS

ETC

NOT QUALIFIED

0 0 0 1 1.2

0 0 0,2 1 0

0 0 0 11 0

a. How do these researchers' data agree and disagree? Describes in detail. B. What are the possible causes of the discrepancies in the results? C. What lesson can typologists draw from this comparison? 7-7

Relative clauses can be formed in different ways. In English, we “replace” the relativized element with a relative pronoun that connects the relative clause to its heading (Type 3). Other languages do not have unambiguous relative pronouns, instead replacing the relativized element with a personal pronoun (type 2). For example, in Gilbertese (spoken in the central Pacific), the position of the relativized element in the relative clause is marked with a personal pronoun. Type 2 I'm fine

Typ 2

say coke

she

at home

"the coconut [that fell into the house]"

te anen

[yo nori-a]

the coconut that i saw the coconut [that i saw]

2 4 2 • Chapter 7 Language universals and language typology

In other languages, such as B. Finnish, relative clauses are formed simply by removing the relativized element from the relative clause; no relative pronoun or personal pronoun is added to the relative construction (type 1). type 1

Typ 1

[to dance]

Sohn

[what I saw]

Sohn

had danced

Chico

I have seen

Chico

'the boy [who had danced]'

'the boy [he had seen]'

Some languages have several types of relative clauses. Mandarin Chinese has types 1 and 2. (In Mandarin, the relative clause comes before its head and is separated from the head by the particle of.) Type 1

Typ 2

[mai pingguO de] comprar manzanas

called

[who is still alive

particle man

meiguó

von]

called

Sister is particle man in America

'the man [who bought apples]'

'the man [whose sister is in America]'

In Mandarin Chinese, type 1 is only used to relativize a subject or direct object, while type 2 can be used to relativize a direct object, indirect object, slash, or possessor, as indicated in the table below . Whenever two types of relative clauses are found in a language, the pattern is the same: if we go down the relativization hierarchy (from subject to direct object, to indirect object, to oblique to possessor), one type may end, but the other type may be finished. assumes These are the patterns for some languages: Relativized grammatical relation SUBJECT

DIRECT OBJECT

INDIRECT OBJECT

ASLANT

OWNER

Aoban (South Pacific) Type 1 Type 2

⫹ -

– ⫹

Dutch Type 1 Type 2

⫹ -

– ⫹

Japanese Type 1 Type 2

⫹ -

⫹ ⫹

Kera (Central Africa) Type 1 Type 2

⫹ -

– ⫹

Mandarin Chinese Type 1 Type 2

⫹ -

⫹ ⫹

– ⫹

– ⫹ (continued)

exercises

• 243

OBJECT

DIRECT OBJECT

INDIRECT OBJECT

ASLANT

OWNER

Roviana (South Pacific) Type 1 Type 2

⫹ -

– ⫹

Tagalog (Philippines) type 1 type

⫹ ⫹

– –

Catalan (Spain) Type 1 Type 2

⫹ -

– ⫹

– –

What cross-language generalizations can you draw from this data about the distribution of relative clause types in each language? How can we extend the universal rules associated with hierarchy to describe these patterns? 7-8

Below is a quote from the Officium Liner Notes, produced by ECM Records. After that, the sentence is repeated in sections with translations of the German and French show notes. Comparable sections are typographically marked. After checking the English sentence and the three translations, answer the following questions. The oldest tracks on this record (if we can use words like "new" and "old" in this context) are the songs whose provenance we don't know. the oldest pieces

in this recording

the oldest pieces

this recording

The oldest parts of these places.

the oldest

disk

appearing in this record

(if one can use words like "new" and "old" in this context) - if one can speak of "new" and "old" in this context - if one can speak of "new" and "old" in this context

Lage

somehow

von

(if the new conditions are like that

are the corners, are the corners whose are

corners

it is

is old

speak

"old" are suitable for this

is old

Dura

not for us

we are the source

the corners, their origin

for us it is

known

may I

Context)

to this connection

whose origin we do not know. the origin is not known to us.

its origin

are the songs of which

neu

It is.

unknown unknown

a. Which of the languages have prepositions and which have postpositions?

2 4 4 • Chapter 7 Language universals and language typology

B. Each of the translations contains three clauses which are equivalents of i. the oldest pieces on this record are the ii cantos. whether words like “new” and “old” can be used in this context iii. whose origin we do not know. Does any of the languages in the main clause use a different word order than SVO? In subordinate clauses? If other word orders are represented, identify them. C. Which languages have adjectives that precede their main nouns? Which ones have adjectives after main nouns? i.e. Neither German nor French use a prepositional phrase to express what English expresses to us. What do they do instead and how is meaning conveyed without a preposition?

Especially for educators and prospective teachers 7-9.

This chapter describes the “universal languages”. Do you think that everything that has been said in this chapter about linguistic universals applies to all varieties of all languages, including all dialects of a language? What about non-standard dialects? Explain your position.

7-10 Explain at a level appropriate to the students you are teaching or preparing for what a universal language is and how there can be such diversity in the languages of the world when such universals exist.

Suggestions for further reading • Bernard Comrie. 1989. Language Universals and Linguistic Typology: Syntax and Morphology, 2nd ed. (Chicago: University of Chicago Press). Affordable and easy. • William Croft. 2003. Typology and Universals, 2nd ed. (Cambridge: Cambridge University Press). More advanced and complete than Comrie (1989); particularly good at explaining different types of universals. • Song by Jae Jung. 2001. Linguistic Typology (Harlow, Essex: Pearson). It is also accessible and focuses on morphological and syntactic typologies. • Lindsay J Whaley. 1997. An Introduction to Typology: The Unity and Diversity of Language (Thousand Oaks, CA: Sage). This book is the most basic of the four listed.

Further reading Mallinson and Blake (1981) is a good introduction to typology. Shopen (1985) is a collection of excellent essays by distinguished researchers in selected areas of syntactic typology and is also useful for the range of morphological and syntactic variations found in the world's languages. The first volume deals with sentence structure, the second with complex constructions and the third with grammatical categories and the lexicon. Some of the most influential works on language.

Reference Universals was edited by Greenberg, who published a four-volume compendium of detailed studies of universals in specific areas of language structure (1978); The chapters in these volumes provided data for some of the exercises in that chapter. Brown (1984) is an interesting study of universal words for plants and animals. Teacher (1974) is a good summary of research on semantic universals. Tomlin (1986) examines the basic word order of the world's languages. The hierarchy of relativization was discovered by Edward L. Keenan and Bernard Comrie, and Comrie's Chapter 7 (1989) provides a clear discussion of the subject. Butterworth, Comrie and Dahl (1984) is a collection of articles on the theoretical explanation of world languages.

References • Brown, Cecil H. 1984. Language and Creatures: Consistency in Popular Classification and Nomenclature (New Brunswick, NJ: Rutgers University Press). • Butterworth, Brian, Bernard Comrie and Osten Dahl, eds. 1984. Explaining Language Universals (Berlin: Mouton). • Greenberg, Joseph H., editor. 1978. Universals of Human Language, 4 vols. (Stanford: Stanford University Press). • Teacher, Adrienne. 1974. Semantic Fields and Lexical Structure (Amsterdam: North Holland). • Mallinson, George and Barry J Blake. 1981. Typology of Language (Amsterdam: North Holland). • Shopen, Timothy, ed. 1985. Linguistic typology and syntactic description, 3 vols. (Cambridge: Cambridge University Press). • Tomlin, Russell S. 1986. Basic Word Order: Operating Principles (London: Croom Helm).

• 245

This page has been left blank internationally

The second part

Language Use In Part One, you examined the structure of words, phrases, and sentences. In the second part, you examine how you use these structures in shared social interactions. You will see that languages offer alternative ways of expressing the same thing and you will see what these alternative ways do socially and communicatively. Language exists only to be used, and our use of language distinguishes humans from all other animals. It is the use of language that makes us unique human beings. By using language we achieve things and can achieve deep social and intellectual satisfaction. The forms of language you use reflect your social identity and reflect the character of the situation in which you are communicating. Part Two examines dialects - patterns of language variation in different social groups - and charts patterns of language variation in communicative situations. Here you will also study writing systems and the relationships between written and spoken expression.

247

8 Information Structure and Pragmatics WHAT DO YOU THINK? • An international student you teach asks about the function of definite and indefinite articles in English. You declare that the definite article (o) refers to a specific person, place, or thing: the Golden Gate Bridge, the Mayor. In contrast, you say, the indefinite article (a or an) is used to refer to any person, place, or thing: a cook, a park, an apple. The student says she pays attention to what people say and her answer doesn't match what she heard. She recalls yourself recommending "a movie" you've seen and meaning "Chicago," which is a specific movie. You admit she's right. Is there a better explanation for the use of definite and indefinite articles? • During an ESL class discussion about active and passive sentences, a Taiwanese student asks why English has these two ways of saying exactly the same thing. For example, give the following sentences: The Chicago White Sox won the 2005 World Series. (active) The 2005 World Series was won by the Chicago White Sox. (passive) Luckily the bell saved you and you can think of your answer overnight. What statement will you give at the next class reunion? • A commercial colleague would like to know when objects can be placed before subjects in English. He noticed a television commentator saying of the mayor of New York: I like him! And he wants to know what you think. will we go

248

Introduction: Structure of the coding information

Introduction: Encoding of Information Structure Syntax and semantics are not the only regulators of sentence structure. A sentence can be grammatically and semantically well-formed, but still have problems when used in a specific context. Take a look at the following two versions of a local report. (Version 1 sentences are numbered because we will refer to them later.) Version 1 (1) At 3 am. Last Sunday, the Santa Clara Fire Department evacuated two apartment buildings on the corner of Country Club Drive and Fifth Avenue. (2) An oil leak from a furnace was discovered in the basement of one of the buildings. (3) Firefighters sprayed chemical foam on the oil for several hours. (4) By 8 a.m. the situation was under control. (5) Any risk of explosion or fire has been avoided and the leaking furnace has been sealed. (6) The occupants of the two apartment buildings were given temporary housing in the Country Club High School gymnasium. (7) They retook their homes at 5 p.m. Version 2 The Santa Clara Fire Department evacuated two apartment buildings on the corner of Country Club Drive and Fifth Avenue at 3 a.m. last Sunday. In the basement of one of the buildings, someone discovered an oil leak in the furnace. What firefighters sprayed on the oil for several hours was a chemical foam. It was around 8 a.m. that the situation was under control. What someone prevented was any risk of explosion or fire, and as for the leaking firebox, it was sealed. What the residents of the two apartment buildings in the Country Club High School gymnasium received was temporary housing. At 5 p.m. they repossessed their homes. Virtually the same words are used in both versions, and every sentence in both versions is grammatically and semantically well-formed. However, there is something fundamentally wrong with version 2. It contradicts our expectations of how information should be presented in a text. It somehow emphasizes the wrong elements or emphasizes the right elements at the wrong time. Although grammatical, the structures of version 2 seem to be inadequate. The problem with version 2 is the way different pieces of information are labeled for their relative importance. In any sentence it is important to mark elements as more or less important or necessary. Speakers and writers are responsible for bringing certain elements to the fore and pushing others to the background, just as a painter uses color, shape, and position to emphasize some details and emphasize others. In linguistic texts, this emphasis and de-emphasis is called information structure. Unlike syntax and semantics, which are sentence-based aspects of language, information structure requires consideration of discourse: sequences of sentences rather than isolated sentences. Taken out of context, there's nothing wrong with the first sentence of Version 2: As for the Santa Clara Fire Department, at 3 a.m. they evacuated two apartment buildings on the corner of Country Club Drive and Fifth Avenue. last Sunday. However, when you open a report, it looks weird and inappropriate. When we talk about the structure of information, we must consider the context of the discourse, that is, the

• 249

2 5 0 • Chapter 8 Information structure and pragmatics

Environment in which a sentence occurs and, above all, what precedes it. We can describe a discourse as a series of oral or written utterances that "go together" in a particular situation. A dinner conversation, a newspaper column, a personal letter, a radio interview, and a court summons are all examples of speeches. We could even say that a statement like Oh look! (e.g. an utterance to call attention to a beautiful sunset), while not a sequence of utterances, it is discourse in that it takes place in a situational context that helps create an appropriate determine information structure. To mark the structure of information in a sentence, speakers rely on syntactic operations to provide alternative ways of sentence formation. For example, the following sentences are alternative ways of saying the same thing. 1st 2nd 3rd 4th 5th 6th 7th 8th

Firefighters found a leak in the basement. Firefighters found a leak in the basement. Firefighters found a leak in the basement. It was the fireman who discovered a leak in the basement. What the fireman discovered in the basement was a leak. It was a leak that the fireman found in the basement. What the fire department discovered was a leak in the basement. Firefighters found a leak in the basement.

Try it yourself: Add to the above eight sentences two sentences that say the same thing, with different statements that contain the same information as sentence 1 (and all others), no more and no less. It is this range of alternatives that we explore to mark the structure of information. You may be wondering which question each of the sentences above corresponds to. This chapter describes how to determine this. Pragmatics is the branch of linguistics that deals with the structure of information. In Chapter 9 we will discuss other aspects of language use that fall under pragmatics.

Information Structure Categories In order to describe the differences between alternative expressions for the same thing, we need to identify the basic information structure categories. These categories should apply to all languages (although the way each category is used may differ). With these categories we want to explain how discourses are structured in any language. Ultimately, these explanations can suggest hypotheses about how different components of the human mind (like memory, attention, and logic) function and interact. Therefore, categories of information structures, like other aspects of linguistics, should be as independent as possible from specific languages. There is an important difference between the types of syntactic constructions in specific languages and the categories of information structure. The range of available syntactic constructs differs significantly from language to language. For example, some languages have a passive construction (it was tricked by a thief), but others do

Information structure categories No. Because information structure categories are language-independent, they cannot be defined in relation to specific structures. However, there is a close relationship between pragmatics and syntax. In all languages, an important function of syntax is to encode pragmatic information. What differs from language to language is the mapping of pragmatic structure to syntax.

Given Information and New Information One category of information structure is the distinction between given information and new information. The information provided is information of which the recipient is currently aware; New information is information that has just been introduced into the discourse. Consider the following two-round interaction: Alice: Who ate the pizza? Dana: Erin ate the pizza.

In Dana's answer, the noun phrase Erin represents new information because it is introduced into the language there; Instead, the pizza gets information in the answer, as it's presumably on Alice's mind, having just inserted it into the speech from the previous round. (We'll see shortly that the information given is often expressed in a condensed form, such as how Erin ate or Erin did.) The information given need not be introduced into a discourse by a second speaker. In the following sequence of sentences, uttered by a single speaker, the underlined element represents information that was given because it was just introduced in the previous sentence and can therefore be considered to be in the addressee's head. A man called while you were on your break. He said he'll call back later.

For another example see Santa Clara Fire Department Newspaper Version 1 on page 249. Note in (1) that the noun phrase two apartment buildings is new information and in (2) that Oil and a Furnace is new information has already been mentioned and cannot be assumed to exist in the mind of the reader. Also note that in (2) the buildings are reported after the mention of two apartment buildings in (1). In (3), the oil receives information due to the fact that the oil is mentioned earlier in (2). Similarly, reference is made to the leaking furnace in (5), which is said because a furnace was previously mentioned in (2), along with the fact that it was leaking. Next we will see that the difference between new information and given information is related to the use of definite and indefinite articles in sentences like oven and oven. The information does not have to be explicitly stated in order to receive information. Information is sometimes taken for granted because it is closely related to something introduced in the speech. For example, when introducing a noun phrase into speech, all subparts of the referent can be treated as given information. Kent finally got my car back last night. The gas tank was almost empty. My mom went on a Caribbean cruise last year, she loved the food.

In the first sentence, my car is new information, but since a car usually has a gas tank, mentioning my car for the gas tank is enough to provide the information. Also the

• 251

2 5 2 • Chapter 8 Information structure and pragmatics

in the second example, the food receives information; Mention of a Caribbean cruise is enough to enable the reader or recipient to recall all things normally associated with a cruise, including the meals.

Try it yourself: Consider version 1 of the newspaper report on page 249. In sentence (2) consider a furnace component in the basement of one of the buildings. Sentence (1) mentions two dwelling houses, but before (2) no furnace or basement is mentioned. Evaluate whether an oven and a cellar in (2) constitute new or supplied information and justify your assessment.

Since personal conversations and most other forms of speech have at least implied speakers and addressees, participants always use first-person speakers and pronouns, like I and addressee, and second-person pronouns, like you. Receive information. They do not have to be brought into the discourse as new information. Expressing new information Noun phrases that represent new information are often more heavily stressed than those that contain given information, and are usually expressed in more elaborate ways, e.g. B. with a full noun phrase instead of a pronoun and sometimes with a relative clause or other modifiers. The following is typical of how new information is introduced into a speech. As I entered the office, I saw a tall man with an old-fashioned hat.

Expressing the information provided The information provided is usually expressed in a shorter or abbreviated form. Typical reductants for encoding given information include pronouns and unstressed noun phrases. Sometimes the information provided is simply omitted from a sentence. In the next interaction, the information from Adam's question (i.e. it's coming up) is completely omitted from Bella's answer, which only expresses new information. Adam: Who's at the door? Belle: The postman.

Try it yourself: Look up version 1 of the newspaper article on page 249 for the abbreviation features used for certain information. E.g. in (4) instead of Until 8 o'clock. Last Sunday, the report omits last Sunday because, as already stated in (1), information is given. Identify two additional instances of given information that were omitted in Version 1. Next, identify a pronoun used to encode given information and indicate which full noun phrase the pronoun represents.

The contrast between the information provided and the new is important in characterizing the function of various constructions in English and other languages, as you will see in the next section.

Categories of information structures

Subjects The subject of a sentence is its focus: what it is about, its starting point. The notion of subject is opposed to the notion of comment, which is the element of the sentence that says something about the subject. Often the information provided is the element of the sentence we are saying something about; is the subject The new information represents what we are saying about the subject; is the comment. So if Erin ate the pizza in answer to the question What did Erin do? offered, the subject would be Erin (the information given) and the comment ate the pizza (the new information). Sometimes the subject of a sentence can be phrased like these examples: Speaking of Erin, she ate the pizza. Erin ate the pizza.

The subject does not always receive information. In the second sentence of the following sequence, the noun phrase his little sister does not receive any information (it is new information), but it is the subject. Erin ate the pizza. The little sister preferred ice cream.

Notice that in the sentence her little sister, the word her links the new information to the given information presented by Erin. In a conversation, we usually first define a topic with a preliminary remark or question and only then comment on it. Sheila:

Remember that guy I said was bothering me?

Eammon: Sim. Sheila:

Well, today he fell off his bike in front of the whole class.

He's the topic in Sheila's second round and he fell off his bike in front of the whole class today, is the comment. The information given can sometimes serve as a comment, as in the underlined dot in the following sequence: Hal didn't believe anything the charlatan said. As for Sara, she believed everything.

Therefore, the default/new contrast differs from the topic/comment contrast. It is difficult to define exactly what a topic is. While subject is the element of a sentence that acts as the focus of attention, a sentence like Oh look! uttered to draw attention to a blinding sunset has an unspoken subject (the setting sun or the sky). Thus the subject is not necessarily a property of the sentence; can be a property of the discourse context. Topics are less central to English grammar than to the grammar of other languages. In fact, the only construction that clearly marks topics in English is the relatively unusual construction for a sentence like this: As for Colin, he's had enough and gone to bed.

In English, marking the subject of a sentence is much less important than marking the subject. The markup issue is much more important in some other languages. Korean has function words whose sole purpose is to mark a noun phrase as the subject. The same applies to Japanese, as we will see below. There are no special features in Chinese and some other languages

• 253

2 5 4 • Chapter 8 Information structure and pragmatics

Function words attached to topical noun phrases but marked for word order. Subject-marked noun phrases are very common in Korean, Japanese, and Chinese. Therefore, despite the difficulty of defining it, the topic is an important concept and must be distinguished from other categories of information structure.

Contrast A noun phrase is said to be contrastive when it is in opposition to another noun phrase in speech. For example, here Sara in Beth's answer is contrasted with Matt in Alan's question. Alan: Did Matt see the ghost? Beth: No, it was Sara.

Compare Beth's answer to another possible one where the noun phrase would not be contrastive: Yes, it did. The contrast is also marked in sentences expressing the reduction of a choice from several candidates to one. In such sentences, the noun phrase relating to the candidate so selected is marked in contrast. Of all those present, only Sara knew what was going on.

Compare this sentence with the following one, in which Sara is not contrastive. Gerard knew what was going on, and so did Sara.

There is a simple contrast test: if a noun phrase can be followed instead of , it is contrastive. Speaker A: Did Matt see the ghost? Speaker B: No, Sara saw the ghost instead of Matt.

A single phrase can have multiple contrasting noun phrases. On the next exchange, Sara is contrasted with mate, and a whole bunch of ghosts are contrasted with one ghost. Aaron:

Matt saw a ghost?

Bella:

Yes, Matt saw a ghost, but Sara saw a whole bunch of ghosts.

The entity with which a noun phrase is contrasted can be understood from the discursive context or from the situational context. In the example below, Sara might be contrasted when the phrase is part of a conversation about how the speakers don't like going to Maine in the winter. Sara likes to go to Maine in the winter.

Next, in an exchange between an employee and one of several managers, the noun phrase I in the manager's response can be contrasted with other managers, which is not expressed but is understood from the situational context. Employees:

Can I check out earlier today?

Manager:

I do not care.

Information Structure Categories With a strong emphasis on I, the implication of the manager's response is, "I'm fine, but I don't know about other managers." . In English, contrastive noun phrases can be marked in a number of ways, most commonly by pronouncing the contrastive noun phrase with heavy stress. You may be smart, but he's popular.

Other ways of marking contrastivity are explored in the next section.

Definite expressions Speakers mark a noun phrase as definite when they assume that the addressee can identify their referent. Otherwise, the noun phrase is marked as indefinite. In the example below, the noun phrase that defines the neighbor in Bundy's answer assumes that Andrea can tell which neighbor Bundy is talking about. Andrea: Who's at the door? bundle:

is the neighbor

Bundy's answer is appropriate if she and Andrea have only one neighbor, or have reason to expect a specific neighbor. If they have multiple neighbors and Bundy can't assume Andrea can tell which neighbor is at the door, the answer to Andrea's question would be undefined: She's a neighbor. Pronouns and proper names are generally defined. Pronouns like you and we usually refer to specific people who are identifiable in the context of the speech. And a speaker who refers to someone by name (e.g. Laura or Tony Blair) assumes that the addressee will be able to determine the references of those proper names. Nevertheless, there are exceptions. Employees of an authority could say to themselves: I have a Susie Schmidt here who has not paid her taxes since 1997.

The use of the indefinite article marks Susie Schmidt as indefinite. In other words, employees can do this because neither the speaker nor the recipient knows the specific person calling themselves Susie Schmidt. The definition in English and many other languages is characterized by the choice of articles (definite or versus indefinite a) or by demonstrative pronouns (this and that, both definite). Indefinite noun phrases in English are marked by a or an (Furnace, Apartment Building) or by the absence of an article (Oil, Fire, Apartment Buildings). While the definite article can be used with singular and plural nouns (el edificio, los edificios), the indefinite article can only be used with singular nouns (un edificio, *a edificios). Although plural nouns do not accept indefinite articles, they can still be indefinite.

Try it yourself: To see how English expresses vagueness with plural noun phrases, examine the first words of sentences (3) and (6) in version 1 of the Santa Clara Fire Department Newspaper Report on page 249.

• 255

2 5 6 • Chapter 8 Information structure and pragmatics

The choice of article is not always a way to mark the definition. Some languages only have one article. Fiji only has one article, well, and that is definitive. To denote the vagueness, Fijians use the phrase e dua, meaning "there is one". 1. na tuuraŋa article knight 'the knight'

(Are defined)

2. e dua na tuuraŋa there is an article Gentleman 'a gentleman'

(indefinite)

In contrast, Hindi has only one indefinite article ek, and a noun phrase without an article is interpreted as definite. 1. ma) kitaab d·hu˜u˜r h rahii thii I book search -ing past tense 'I was looking for the book.'

(Are defined)

2. ma) ek kitaab d hu˜u˜r h rahii thii I a book search -ing past tense 'I was looking for a book.'

(indefinite)

Many languages do not have articles and must rely on other means to mark the definition, if it is marked at all. Mandarin Chinese depends on word order. When the subject comes before the verb, as in 1 below, it is defined; when it follows the verb, as in 2, it is indefinite. 1. huOche lái le train arrives New situation 'The train has arrived.'

(Are defined)

2. lái huOche le arrivalr train New situation 'A train has arrived.'

(indefinite)

There are other systems too. In Rotuman, spoken in the South Pacific, most nouns have two forms, definite and indefinite. Futi vaka rite defined

Undefined 'the banana' 'the canoe' 'the scion'

caja fut rjot

'a banana' 'a canoe' 'a sprout'

The indefinite form can be derived from the definite form by a set of phonological rules. defined vs. definite datum is distinguished from datum because a noun phrase can be undefined and new, definite and new, undefined and given, or defined and given, with the first and last combinations occurring most frequently. Next, a speech is indefinite and new, and the speaker is defined and given. Last night we went to the Hayden Planetarium for a lecture and the speaker passed out.

You can also define a noun phrase that refers to new information. The following sequence defining the plumber is acceptable whether or not the speaker introduced a specific identifiable plumber in the previous speech.

Information Structure Categories Kitchen faucet drips; Better call the plumber.

In certain circumstances, a noun phrase can be indefinite and given, as with the underlined noun phrase in this example: I had a hamburger for lunch, a hamburger that was the worst I've ever eaten.

Definition and wrapper are clearly different categories of information structure.

Try it yourself: In version 1 of the newspaper article on page 249, identify at least one noun phrase in each of the following categories: (a) indefinite and new; (b) definitive and new; (c) indefinite and definite; (d) defined and given.

Reference phrases A noun phrase is referential if it refers to a specific entity. In the first example below, the blue-eyed Italian expression does not refer to a specific person and is therefore non-referential. Instead, in the second example, the same sentence has a referent and is referential. Kate wants to marry a blue-eyed Italian but she hasn't met anyone yet.

(not referential)

Kate wants to marry a blue-eyed Italian; His name is Mario.

(Notice)

Taken out of context, Kate wants to marry a blue-eyed Italian is ambiguous because nothing in the sentence indicates whether a specific Italian is meant or not. In everyday language, given the clarifying power of context, sentences of this kind are rarely ambiguous. Because referentiality and definition are not the same; a noun phrase can be both referential and definitive – where is the key to the vault? referential and indefinite – He rented a new Ford Bronco. non-referential and defined - What's the smartest thing to do right now? non-referential and indefinite: you need to buy a new car.

While pronouns and proper nouns are generally referential, certain pronouns like you, he, she, and um are generally non-referential. If you own a house in this country, you have to pay taxes. The governor is widely suspected of having ties to the insurance industry. They're forecasting storms tonight. The person just doesn't know what to do in such circumstances.

Since none of these pronouns refer to a specific entity, they are not referential.

General and Specific Expressions A noun phrase can be generic or specific, depending on whether it refers to a category or specific members of a category. In the first example below is the bombardier beetle

• 257

2 5 8 • Chapter 8 Information structure and pragmatics

generic because it refers to the set of all bombardier beetles; but in the second, which could have been pronounced while visiting a zoo, La giraffe refers to a particular animal and is therefore specific. The book is about the bomb beetle that spits a deadly concoction in its enemy's face. The giraffe slowly leaned forward and carefully took a carrot from my hand.

In the first sentence, the bombardier beetle is generic and unambiguous, while a deadly mix is generic and undefined. In the second sentence, a giraffe is specific and definite, and a carrot is specific and indefinite. Thus, the generic/specific contrast differs from the definite/undetermined contrast.

Check it out for yourself: The following sentence appears in a US Supreme Court ruling on driving a truck on Iowa's interstate highways. This leaves a third noun phrase. Is it specific or generic? Which of the noun phrases are definite and which are indefinite? Are all noun phrases referential? In fact, the state points to just three ways in which the 55-foot single is undeniably superior.

Information structure categories Information structure is not only characterized in noun phrases. Other parts of speech, especially verbs, can represent given or new information and also contrast it. In the following exchange, the underlined verb stands for new information marked in high contrast. Jerry comes by from time to time, but Sara goes camping every vacation.

Similarly, prepositions can sometimes be marked for information structure, as in this example of contrastive marking. I said the book was on the table, not under it!

In this chapter we focus almost exclusively on the designation of information structure in noun phrases, partly because the role of other components in the discourse structure is not yet well understood.

Information Structure: Intonation, Morphology, Syntax Languages differ in the amount of pragmatic information they encode and how they encode it. In many languages, intonation is used to mark contrasts. While stress in English is an important tool for identifying information structure, it is less important for information structure in languages such as French and Chinese. Other languages, like

Information structure: intonation, morphology, Japanese syntax, words have functions whose sole purpose is to indicate pragmatic categories. Still others, including English, rely on syntactic structures such as passives to convey pragmatic information. Therefore, different languages use different strategies to encode pragmatic information. What follows is a selection of these strategies.

Emphasis on new information In English and some other languages, intonation is an important way of emphasizing information. In general, noun phrases that represent new information are stressed more heavily than those that represent given information, and they are pronounced slightly higher than the rest of the sentence. This is called new information stress. Aaron:

Whose footprints are on the couch?

Bianca: These are Lou's footprints.

English speakers also use stress to emphasize contrasts. 1. Aaron: Are those your fingerprints on the couch? Bianca: Those aren't mine, those are Lou's. 2. Hal was told he needed two more years to graduate, but Sara got the okay.

Phonetically, reinformation stress and contrastive stress are similar, but they differ functionally. English uses stress in complex ways, much more so than languages like French and Chinese.

Information Structure Morphemes Some languages have grammatical morphemes whose sole function is to mark categories of information structure. In Japanese, the functional word wa, which comes after noun phrases, marks dates or contrast. When a noun phrase is neither given nor contrasted, it is marked with another function word (usually ga for subjects and or for direct objects). That wa is a marker of given information is illustrated by the following exchange: Kenn:

onibus basuga

kimasuka

Subject comes question

'Is the bus coming?' Kimiko: Basu bus

von

Kimasu

given

comes-is

'The bus is coming.'

In Kenn's question, basu couldn't be tagged wa unless he and Kimiko talked about the bus in the earlier speech. But in Kimiko's answer, basu is given information and should be marked wa. The Japanese wa also denotes contrasting information, as in the following sentence: basu wa bus

Kimasu

demostración de takushi wa

Contrast comes-is

Taxi

Kimasen

Contrast is coming, isn't it?

'The bus is coming. But the taxi doesn't come.'

• 259

2 6 0 • Chapter 8 Information structure and pragmatics

Here basu wa need not represent given information, as wa may simply mark the fact that the noun phrase to which it is attached contrasts with another noun phrase also marked wa (takushi 'taxi'). Many other languages use function words to denote categories of information structures. This is the most transparent way of identifying the structure of the information. Grammatical morphemes like Japanese wa do not affect the general form of a sentence. Rather, they simply state what element of a sentence is given, what the contrast is, etc.

Fronting Fronting is one of the various syntactic operations used to characterize the structure of information. Fronting works in many languages, although its exact function varies from language to language. Form sentence 1 in English from the basic structure of sentence 2, which has the same meaning. 1. He may be a boy, but he is certainly not a boy. 2. He may be a boy, but he's certainly not just a boy.

In English, a function of fronting is to mark the given condition, and a fronting noun phrase must represent given information. avi:

I heard you really like mushrooms.

Berto:

Mushrooms I would kill for.

A noun phrase may be prefixed if its referent is part of a group previously mentioned in the speech, even if the referent itself was not mentioned. In the example below, Barnyard is a family film hyponym mentioned in the question to Courteney Cox Arquette immediately before the original noun phrase. the result is pragmatically acceptable. P:

Did you intentionally shoot two family films in a row?

Supported communication:

No, I did Barnyard when I was breastfeeding.

Frontal noun phrases often contrast in English. All:

do you like madonna

They say:

I like her early songs, but Madonna herself doesn't drive me crazy.

Fronting noun phrases do not always have the same function in other languages as they do in English. In Mandarin Chinese, leading noun phrases are often used to represent the subject of the sentence. 1. zèi bEn shu¯ pízi hen hAo kan the cover of this classifier's book is very beautiful 'This book, the cover is beautiful.' ) I saw many paintings in this exhibition.'

The interesting thing about Chinese frontal noun phrases is that they don't necessarily play a semantic role in the rest of the sentence. For example, in the following sentence

Information structure: intonation, morphology, syntax mógu¯ 'mushrooms' cannot be a patient because the sentence already has a patient: zhèi ge do¯ngxi 'something like that'. However, the sentence is grammatically and pragmatically acceptable. mogu¯

Yes indeed

seta yo

xIhuan

very similar

who zhèi ge eats this

Dongxi

classify thing

"Mushrooms, that's what I like to eat."

Also, leading noun phrases in Chinese do not need to be contrasted, although they often do in English. The comparison between English and Chinese fronting makes an important point clear: a grammatical process, such as a move operation, can have comparable syntactic properties in two languages, but their pragmatic functions can differ significantly.

Shift Left Shift left is an operation that derives sentences like 1 from the same underlying structures as basic sentences like 2. 1. Holly, I can't stand you. 2. I can't stand Holly.

Although front left shift is syntactically similar, there are some differences between the two. In particular, a forward noun phrase leaves no pronoun in the sentence, while a left-shifted noun phrase does. Holly, I can't take it. (front) Holly, I can't stand you. (move left)

Unlike frontal noun phrases, a left-shifted noun phrase is separated from the rest of the sentence by a very short pause, represented in writing by a comma. The left shift is similar in nature and function to the right shift, which shifts noun phrases to the right of a sentence. I can't stand you, holly.

In this discussion, we will focus on the shift to the left. Scrolling to the left is mainly used to reintroduce information provided that hasn't been mentioned in a while. In the long example below, the speaker lists and comments on several people. Hal, mentioned at the beginning of the speech, is reintroduced in the last sentence. Since nothing was said about him in the previous two sentences, the speaker reintroduces Hal as a left-shifted noun phrase. I stayed in touch with many of my classmates. I still see Hal who was my best friend in high school. And then there's Jim, my college roommate, and Stan and Sara, who I met my sophomore year at Ohio State. I really like Jim, Stan and Sara. But Hal, I can't stand you now. In addition to reintroducing the information provided, the shift to the left is contrastive. In this example, Hal is in stark contrast to Jim, Stan, and Sara. Because of its dual function, left shift is often used when speakers are going through lists and annotating each list item. Some languages use left shift

• 261

2 6 2 • Chapter 8 Information structure and pragmatics

more often than English. In colloquial French, left-shifted noun phrases are significantly more common than corresponding base clauses. Mon

Brothers,

s'en va

eng

brother goes

through in Mongolia

Mongolia

'My brother, he is going to Mongolia.'

A shift to the right, illustrated by the following theorem, is also common. I already know it

no,

and,

what he

Would you like it.

I know

Mim

what he

Would you like it

'I, I don't know what he wants.'

The left shift in colloquial French has a different function than the corresponding operation in English. In French, a left-shifted noun phrase represents a subject. Left-shifted noun phrases are particularly common when introducing a new topic of speech (as in the first example below) or when the speaker wants to change the topic of speech (as in the second example). 1. [asks a stranger on the street for directions] où est-elle? Pardon, la gare, excuse me, which train station is that. 'Excuse me, where is the train station?' 2. Pierre: Moi, j'aime bien les croissants. I really like croissants 'I really like croissants'. frais, c'est bon Marie: Oui, mais le pain yes, but fresh bread is good 'Yes, but fresh bread is good too.'

To. continue

The pragmatic role of the left shift is therefore much larger in French than in English.

It clefts and WH clefts slit transformations are used in English and many other languages to mark the structure of information. In the examples below, set 1 is an it-split set, set 2 is a WH-split set, and set 3 is the base set, matching 1 and 2. 1. It was Nick that Stan saw at the party. 2. Who Stan saw at the party was Nick. 3. Stan saw Nick at the party.

(Nut) (WH-Nut)

Sentences with spaces are of the form It is/was. . . the one where between the first and second parts of the construction is the noun phrase, the prepositional phrase (It was in March the last time he visited) or the adverb (He recently learned to sing). WH splitting constructs can take the form WH word. . . is/was/will be, where the word WH is usually what. In slotted WH constructions, the slotted noun phrase, slotted prepositional phrase, or slotted adverb is placed after the verb to be, and the rest of the sentence is placed between the two parts of the construction. There are also other variants of WH split sentences, as in these examples:

Information structure: intonation, morphology, syntax It was Stan who saw Nick at the party. Nick is the one Stan saw at the party.

Aside from is and was, some other modes of being can also appear in the cracks. The constructs it-slit and WH-slit are used to identify the donation. In a split construction, the split sentence presents new information and the rest of the sentence receives information. Thus, the information question at 1 below can be answered as 2, where the answer to the question (i.e. the new information) is shared, but not 3 because the shared item is not the requested new information. 1. Who did Stan see at the party? 2. It was Nick that Stan saw at the party. 3. * It was Stan who saw Nick at the party.

That the following phrase in a split sentence contains specific information is illustrated by the fact that it can refer to something just mentioned in the previous sentence. In the example below, the second sentence is a splinter construction, where the following elements from the previous sentence are simply repeated in the discourse. Alice told me Stan saw someone he knew from high school at the party. Turns out it was Nick Stan saw at the party.

The next element in a shared set uniquely represents given information. WH-slit constructs are similar to it-slit constructs. In sentences with WH slots, the new information comes after the verb to be, and the rest of the sentence is between the word WH and the verb to be. 1. What did Stan see at the party? 2. What Stan saw was Nick dancing salsa.

Question 1 cannot be answered with any of the following clauses, since neither in 3 nor in 4 is the nominal clause the new information. 3. *The person who saw Nick dancing salsa was Stan. 4. * Where Stan saw Nick was at the party.

The remainder of a WH split sentence marks the information given, as in an it split sentence. The following pair of sentences, in which the information given is underlined, makes this fact clear. I really enjoyed his latest novel. What I particularly liked was the character development.

Both the it slot and the WH slot emphasize which element is new information and which element receives information. Also, both constructions can mark contrast. Consider the following two sequences. In 1 (whose second sentence is an it-column construction) and 2 (whose second sentence is a WH-column) the new information can easily be understood as contrastive. Possible implied information is given in square brackets after each example. 1. Alice said Stan saw someone he knew from high school at the party. Turns out it was Nick Stan saw at the party. [ . . . not Larry as you might have thought.]

• 263

2 6 4 • Chapter 8 Information structure and pragmatics

2. I really enjoyed your latest novel. What I particularly liked was the way the characters' personalities were developed. [I liked the character development more than the writing style.]

You may be wondering why English should have two constructs with the same function. Languages often use different structures for different purposes, and indeed there is a subtle difference in the use of these two constructs. A cleft construct can be used to mark information that the listener or reader is not necessarily thinking of. However, a slotted WH construction requires the listener or reader to reflect on the information provided. Therefore, it is possible to start a narrative with a splitting construct in it, but not with a splitting construct in WH. The first sentence below is a split construction and would provide an acceptable opening for a historical narrative; but the second set is a WH slot construction and wouldn't normally be a good start. To gain independence from Britain, the colonists started the revolution. * What the colonists began to gain with the revolution was their independence from Britain.

The first sentence is an acceptable opening as it does not necessarily assume that the reader will have the information given (the colonists started the revolution) in mind when the narrative begins. The second sentence assumes that the information provided ([what] the colonists began to gain with the revolution) is in the reader's mind and is therefore not a good opening sentence. The difference between the constructs it-slit and WH-slit shows that the information provided is not an absolute term. There can be different types of donations: information that the recipient knows but is not necessarily thinking about at the time, and information that the recipient knows but is thinking about.

Passive As in other languages that have a passive construction, the choice between an active sentence and its passive equivalent in English can be used to mark the structure of the information. Compare the following sentences: 1. Bureaucrats could easily store and retrieve data about citizens. (active) 2. Bureaucrats can easily store and retrieve data about citizens. (passive) 3. Data about citizens can be easily stored and retrieved. (passive)

Of these three sets, which can all represent the same situation, set 1 is active while the other two are passive structures. In Fig. 2, the agent (bureaucrats) is expressed, and the structure is called the passive construction of the agent. But since no agent is expressed in 3, it is said to be passive without an agent. Agentless liabilities and agent liabilities are used for specific purposes. An agentless passive serves well when the agent is particularly unimportant to the action or state represented by the sentence, such as when the agent is a generic entity whose identity is irrelevant to the point of the sentence. A new shopping center is being built near the airport. These laws, harmful as they are, are rarely enforced.

Information structure: intonation, morphology, syntax In the first sentence, the broker is probably a real estate developer; in the second sentence police authorities. In each case, the precise identity of the agent concerned is either known or irrelevant to the situation represented by the sentence. In spoken language, agentless passive clauses are often equivalent to active clauses with the indefinite, non-referential pronoun sie, as in these examples: They are building a new mall near the airport. They issue new Christmas stamps every year.

A passive construction of the agent is used when the given information is another noun phrase of the agent. Imagine a story that begins like this: The World Health Organization held its annual meeting in Geneva last week.

This sentence establishes the annual meeting as information for the remainder of the report. If the following sentence uses the noun phrase the meeting, this sentence is likely to appear in the subject position since it represents given information. If the noun phrase the encounter in the following sentence does not have the semantic role of agent, it is likely that the sentence will be expressed as a passive construction so that the encounter can be the grammatical subject. The meeting was organized by health administrators from 50 countries.

This generalization is not absolute, and there is nothing fundamentally wrong with a sequence in which the second sentence is active rather than the passive sentence predicted by the generalization, as shown below: The World Health Organization held its annual meeting in Geneva last week. Health administrators from 50 countries organized the meeting.

But the equivalent sequence with a second passive sentence flows better and may be easier to understand: The World Health Organization held its annual meeting in Geneva last week. The meeting was organized by health administrators from 50 countries.

In English, the choice of a passive sentence over its active counterpart is governed by information structure. Agentless passives are used in particular when the agent is well known or not particularly important (as in that very sentence). Passive agents (or passive as they are sometimes called) are used when a noun phrase other than the phrase agent is more prominent as the information provided than the agent itself.

Try it yourself: In the following passive sentences, identify the two non-agents and say which is the likely agent. For the passive agent, specify the agent. 1. The governors of the states were appointed by the president. 2. Most people would be scared if their capital was attacked. 3. Japan Post was affected by the stock market crash.

• 265

2 6 6 • Chapter 8 Information structure and pragmatics

Not all languages have a passive construction. The Chinese and Samoans, for example, do not. These languages have other ways of expressing what English speakers express with the passive voice. If the agent of a sentence is not important in Samoan, it is simply not expressed; the sentence remains an active structure. water

ʔhad the

Correction in present tense

you

the young man

The young woman is insulted. (Lit.: "He scolds the young woman").

order of words

1. gato sobaku está perseguindo cachorro 2. sobaku presleduet koska 3. presleduet koska sobaku 4. presleduet sobaku koska 5. koska sobaku presleduet

154626453

Many languages use the sequential order of noun phrases to mark differences in information structure. English cannot use all word order features for this purpose, as it uses word order to mark direct subjects and objects (see Chapter 5). In the sentence The cat chases the dog, the order of the words indicates who hunts and who is hunted. If we reverse the two noun clauses, the semantics of the clause change (who is the agent and who is the patient): The dog chases the cat. However, in a language like Russian, we can encode noun phrases without changing the semantics. All following sentences mean the same thing. (Note that s is pronounced like sh in Boot.)

The cat chases the dog.

In each of these sentences, we know who is doing what to whom because the inflections of the noun vary. The -u ending of sobaku 'dog' denotes it as a direct object (if it were the subject it would be sobaka), and the ending -a of koska 'cat' denotes it as a subject (if it were the direct object it would be kosku). The differences between these versions of the same sentence lie in their information structure. More precisely, the order of the words in Russian marks the donation. The question of information Sˇto koska presleduet? 'What is the cat chasing?' can only be answered as follows: koska presleduet sobaku cat

hunt dog

The cat chases the dog.

On the other hand, the question Sˇto presleduet sobaku? "What's chasing the dog?" should be answered as follows: sobaku presleduet koska dog

chase cat

The cat chases the dog.

The relation of sentences to language: pragmatics So, the first sentence in Russian is not the subject, but the given information, and the last one is new information. In response to the question What is the cat chasing? the dog is new information and appears at the end of the sentence in Russian. On the contrary, with the question What chases the dog? the cat is new information and takes the last place in the sentence. The word order is used in Russian, as in many other languages, to denote donation. Similar explanations could be offered for the other variants of the Russian expression we have cited, but we will not develop them here. (See Exercise 8-9 on page 277.) Typically, in languages that exploit word order to encode pragmatic information, syntactic constructions such as passives, It-splitting, and WH-splitting do not exist (or are rare). Russian has a grammatical construction reminiscent of the English passive voice, but it is rarely used. The reason is simple: given the rich inflectional system to mark grammatical relationships, word order is left free to mark information structure and there is no need to use complex structures like passives to mark the gift. Passives are useful in languages that exploit word order for other purposes and thus cannot manipulate it to indicate pragmatic information.

The Relation of Sentences to Discourse: Pragmatics We describe some of the basic concepts necessary to describe how information is structured in discourse, and we analyze a number of constructs related to information structure. From the discussion in this and previous chapters it should be clear that the syntactic structure of any language is determined by two factors. On the one hand, the syntax must encode the semantic structure: the syntactic structure of a sentence must enable language users to recognize who is doing what with whom: the agent of a sentence, the patient, and other semantic roles. On the other hand, the syntax must encode the structure of the information: which element of a noun phrase receives information, what is new information that can be easily identified by the addressee, which is not, etc. Schematically, the relationship is as follows:

Therefore, syntax is used to convey two types of information: semantic information and pragmatic information.

• 267

2 6 8 • Chapter 8 Information structure and pragmatics

Computers and pragmatics Finally, a full understanding of pragmatics will be important for speech recognition and, to a lesser extent, for speech synthesis. However, pragmatics in computational linguistics has not been researched as well as the morphological, lexical, phonological, grammatical, and even semantic features of texts. One reason for the relative neglect of pragmatics is that modeling the world and discourse knowledge that speakers rely on to produce and understand spoken and written texts is more challenging than modeling structural aspects of language such as morphology and syntax. Another reason is that the linguistic features by which some pragmatic categories are realized are not always expressed in texts in a way that computers can easily track. As we have seen, speakers base many aspects of the utterance on their beliefs about what the addressees know and what they have in mind. This applies, for example, to the identification of noun phrases as definite or indefinite, to the choice of active or passive, and to the identification of contrasts through intonation. Although these three features have some representation

Resentments in a text, others, such as dates and new information, have little or no textual implementation and would be extremely difficult or even impossible for a computer to identify. If you've ever used a grammar checker, you know that even the most rudimentary passive verbs are easily identified (identification forms of the verb BE together with, but not necessarily together with, a past participle, as is necessary, are realized, and not explored, which all appear in the first two paragraphs of this section). What existing verifiers cannot do is distinguish between passives that actually serve a pragmatic function like updating and those that don't. As a result, an author using a grammar checker may find that it marks all passive verbs and insists on rephrasing them as active. If an author were to rewrite all passives into active ones, the rewritten sentences would remain grammatical, but the changes could damage the pragmatic structure of the text. (Exercise 8-7 on page 275 asks you to think about reviewing liabilities in a short text.) Computer programs can only identify pragmatic categories if they are identified in the text. So

FIGURE 8-1 Average number of provided and new reference terms in three text types

Source: D. Biber, S. Conrad, R. Reppen, 1998.

The relation of propositions to discourse: pragmatics

The Japanese function word wa can be identified automatically, just like a passive word in English. Likewise, most noun phrases in English can be automatically identified as definite or indefinite. But other categories, such as subject, dates, and referentiality, cannot be automatically identified. If researchers wanted to use such categories, their texts would have to be manually tagged to reflect those categories. To do this, a program would mark each potential item, such as all reference phrases, as "given" and then present a human editor with a menu of alternatives to tentatively mark, just as a spell checker would offer alternatives. Once these categories were tagged on the reference phrases in a corpus, researchers could explore related topics, relying on the speed and precision of the computer. Suppose a corpus contains reference terms that have been manually marked as delivered or new. It would be easy to calculate the number of returned and new references to any group of text in the corpus, such as conversations or news reports. It turns out that different types of text differ significantly in the average number of delivered and new references. Figure 8-1 shows two sets of relationships: those between the provided information and the new one of three types.

of text and between the three types of text. For example, the conversation has three times as many first name phases as the new ones. Academic prose, on the other hand, has reported about half of the nominal sentences as new. While conversation and news have about the same number of noun phrases per 200 words of text, the proportion of existing and new noun phrases is reversed in this type of text. Speakers use noun phrases, which are mostly given. Journalism introduces new speakers about twice as often as it refers to the information provided. Many noun phrases in conversation are first- and second-person pronouns, which always carry information. (By contrast, the use of first- and second-person pronouns in news reporting is practically limited to quoted speech.) As Figure 8.1 shows, a second type of information can be gleaned from the number of referential utterances: given and new. different types of text. For example, the number of new noun phrases in conversation is relatively small compared to academic prose or newspaper reporting. Not surprisingly, news reports contain more than twice as many new references as conversations, as does academic prose. To take another example, when an index has been assigned to the third-person pronouns in a corpus

Figure 8-2 Mean distance between pronouns and their antecedents, measured in number of between-reference phrases

Data source: D. Biber, S. Conrad, R. Reppen, 1998.

• 269

2 7 0 • Chapter 8 Information structure and pragmatics

Computers and Pragmatics whose reference matches an earlier noun phrase, then computer programs can track the distance between them and their predecessors. For this purpose, the number of intervening noun phrases is a useful measure, and different types of text are found to have significantly different distances between third-person pronouns and their antecedents. As Figure 8-2 on page 269 shows, conversation has fewer than half the number of intermediate reference expressions (i.e., noun phrases) that we find in academic prose, and

News. In part, this reflects the fact that the conversation is unforeseen and the speakers accommodate the fact that their addressees must follow referents identified only by third-person pronouns (he-who? that-what?). Properly labeled corpora can teach a great deal about the character of different types of text. This information is critical for speech recognition, machine translation, and other real-world language applications where computers and natural language meet. ■

Summary • Pragmatics deals with the coding of information structure: the relative importance of different elements in a sentence, mainly noun phrases. It deals with the relation of sentences to their discursive environment. • Relational categories include giving (whether the information is new or already existing in the context of the discourse), subject (the focus), and contrast (whether the information is contrasted with another). • Non-relational concepts include definition (whether the referent of a noun phrase is identifiable) and referentiality (whether a noun phrase has a referent). • Some syntactic operations serve to mark certain sentence elements for pragmatic categories. In English, fronting, left displacement, it cleft, WH cleft, and passivization emphasize certain noun phrases as subject sentences or as information provided or new information. • Contrast is marked by the accent of the sentence and is a secondary function of certain operations, such as B. Fronting. • Many languages use word order or grammatical morphemes to mark the structure of information. • The functions of a particular transformation device or information structure may differ from one language to another because each language favors certain strategies over others. • Syntax encodes two types of information: semantic information (the semantic function of a noun phrase) and pragmatic information (the relative importance of noun phrases in a discourse).

What do you think? checked

WHAT DO YOU THINK ? R E V I S I T • The international student. Conventional wisdom holds that definite articles are used to refer to specific people, places, and things, as in the Eiffel Tower and film. But it is inaccurate to say that the indefinite articles are used to refer to a person, place, or thing (ie, not a specific one). This is a feature of an indefinite article (I'm looking for a gift in the $50 range for Tony, any suggestions?). Usually, however, the use of an indefinite article indicates that the speaker believes the addressee does not yet have the referent in mind (I bought Tony a present after receiving a suggestion from Barry). In other words, an indefinite noun phrase can indicate that an entity is new to discourse. Once the entity is in the recipient's mind, speakers use certain noun phrases (the present tense, the signal, this) to signal that the entity is receiving information and the speaker believes the recipient can identify it. In a conversation (or newspaper article or other speech) it is common for an initial mention to be vague and subsequent mentions to be unambiguous. • Discussion in ESL lessons. In many ways, the active and passive versions of the phrase about the Chicago White Sox mean the same thing. For example, if one version is true, the other must be true; if one is wrong, the other must be wrong. Furthermore, both describe the same situation in the real world. Still, the two phrases would not be used in the same circumstances as they focus on different noun phrases. The active version (the Chicago White Sox won the 2005 World Series) says something about the Chicago White Sox. The sentence is about the White Sox. The passive version (the 2005 World Series won by the Chicago White Sox) is about the 2005 World Series. So the sentences mean the same thing but present different perspectives. Consider the two queries below and decide whether the active and passive Version above can equally be used as an answer: Tell me something cool about the Chicago White Sox. Tell me about the recent history of the World Series. Speakers (and writers) tend to be efficient, and while two phrases can mean the same thing, they can't necessarily be used interchangeably in all language contexts. Different versions serve different purposes and situations. • Your classmates. English generally has subjects before verbs and verbs before objects (I like popcorn!), but to express contrast, an object can sometimes be moved to the first position before the subject.

• 271

2 7 2 • Chapter 8 Information structure and pragmatics

English-based exercises 8-1.

In an article titled "Ellen's 'Heart Trouble': A Friend's Report," actress Kathy Najimy writes about her first meeting with Ellen DeGeneres. After reviewing the passage (sentence numbers added), answer the following questions: (1) I met Ellen three or four years ago when I was a guest on her show. (2) She was funny, smart, and charming, and I was touched by her vulnerability and what she was going through in terms of her sexuality. (3) We sat in the trailer and talked about what was going on with her personally and politically. (4) It was interesting for me because I know a lot of gay people and a lot of famous people, but I've never met someone who is famous and gay and has a hard time knowing what to do about it. [Los Angeles Times, "Calendar," December 21, 1997, p. 79]. In 1, is it the pronoun I gave or some new information? Explain the basis for your answer. B. Is new information given in 2? Explain the basis for your answer. C. Is 3 data or new information? Explain the basis for your answer. i.e. Is 4 data or new information and to which element does it relate? Y. Identify the noun phrase that is the subject in sentences 2, 3, and 4. f. What conclusions can you draw from your answers to b a and above on the topics of this paragraph? Gram. Based on the examples you just looked at and the other pronouns in the passage, would you say that personal pronouns generally represent given information or new information? h Is given in cubits or new information? Specified or unspecified? to explain. I. In 1, what program is your program related to? Is your show final or indefinite? j. List all indefinite noun phrases in the passage. (Remember that pronouns are noun phrases.) k. Prior to its mention in 3, the trailer wasn't mentioned, so how do you explain it's final?

8-2.

Examine the following section and answer the following questions: (1) Beginning in 1999, the Rose Bowl will no longer have the option of playing the top teams from the Big Ten and Pac-10 football conferences first. (2) Instead, the "Bowl Alliance," which aims to pair the country's top two teams for the national title each year, will decide who goes where. (3) The alliance, a collaboration between six of the country's top football conferences, has a seven-year deal with ABC Sports, which will broadcast the title game. (4) This agreement guarantees each conference champion and Notre Dame a seat in one of four bowl games: Rose, Orange, Fiesta, or Sugar, with the national championship game rotating between the four locations annually. [Adapted from a promotional insert in the Los Angeles Times "Calendar", December 21, 1997, S. SUBWAY]

Exercises A. Identify two noun phrases that are referential and two that are not. B. Identify any contrasting noun phrases in the passage. C. At 4, this agreement is reported and marked as final. Which noun phrase in 3 has the same reference as this agreement but is not defined? Explain why the first of these noun phrases is undefined and the second is definite, even though they have identical references. i.e. From the passage, identify a noun phrase in each of the following categories: (i) referential and definite (ii) referential and indefinite (iii) non-referential and definite (iv) non-referential and indefinite 8-3.

Consider the following text as a complete story in a newspaper. Analyze each sentence in its context and tell what is odd about it in terms of information structure. (1) The Santa Clara Fire Department evacuated two apartment buildings on the corner of Country Club Drive and Fifth Avenue at 3 a.m. last Sunday. (2) Nancy Jenkins discovered a stove in the basement of a building on Country Club Drive and Fifth Avenue, and the stove was leaking oil. (3) Firefighters sprayed chemical foam on the oil for several hours. (4) It was around 8 a.m. last Sunday that the situation was under control. (5) What had been avoided was any danger of explosion or fire, and as for a leaking firebox, it was sealed. (6) What the residents of the two apartment buildings on Country Club Drive and Fifth Avenue in the gymnasium of Country Club High School received was temporary housing. (7) The occupants of the two apartment buildings on the corner of Country Club Drive and Fifth Avenue confiscated their homes at 5:00 p.m. m. last Sunday.

8-4.

Choose a short article (about a newspaper column) or an excerpt from a front-page article in a newspaper. Identify any sentences that have undergone a syntactic operation (such as passivation or splitting). In each case, explain the most likely reason for using a transformed set instead of the equivalent basis set.

8-5.

In certain dialects of English, a syntactic operation moves a noun phrase to the beginning of its clause. Sentence (1) is derived from the same basic structure as principle (2): 1) He wants a bottle of sparkling wine and caviar. 2) He would like a bottle of champagne and caviar. The operation is called the "Yiddish movement" because it is characteristic of the English dialect spoken by native Yiddish speakers. The Yiddish movement is syntactically similar to fronting but differs in its pragmatic function. Here are three pragmatic contexts in which the Yiddish movement is appropriate. Using this data, briefly describe the pragmatic role of the Yiddish movement. 1) Speaker A: Speaker B:

What does he want? A bottle of champagne and caviar he wants!

• 273

2 7 4 • Chapter 8 Information structure and pragmatics

2) Speaker A: Speaker B:

How is your daughter? This worries me so much!

3) Speaker A: Speaker B:

are you willing to help me I wouldn't lift a finger for you!

In particular, compare the following interactions. In the first case, the answer may come through the Yiddish movement; in the second it can not.

8-6.

4) Speaker A: Speaker B:

Who will Deborah marry? A villain Deborah is getting married!

5) Speaker A: Speaker B:

Who will marry Deborah? * Deborah is getting married!

Below is an excerpt taken and slightly adapted from a US Supreme Court case (Kassel v. Consolidated Freightways Corp.); some noun phrases have been underlined and sentences numbered. Read it and answer the following questions. (1) Iowa does not seriously challenge any of these findings. (2) In fact, the state only indicates in three ways that the 55-foot single may be superior: singles take less time to pass through and clear intersections; greater distances can be withdrawn; and they are slightly less vulnerable to the razor. (3) The first two of these features are of limited relevance on modern motorways. (4) As the District Court noted, the negligible difference in the time required to pass and cross intersections is negligible on four-lane divided freeways because overtaking requires no crossing of the oncoming lane, and freeways require few, if any, intersections to have . (5) Concern about backup capacity is also negligible as backup is rarely needed on a highway. (6) In any event, no evidence suggests a difference in carrying capacity between the 60-foot doubles that Iowa allows and the 65-foot doubles that it prohibits. (7) Similarly, 65-foot doubles are less likely to buckle than 60-foot doubles, although doubles tend to buckle slightly more than singles. a. Using the new and provided information, explain how you know in sentence 2 that the state refers to Iowa B. Identify an underlined noun phrase in each of these categories: (i) generic and definite; (ii) specific; (iii) final; (iv) generic and indefinite; (v) final and referential; (vi) indefinite and referential; (vii) given and defined; (viii) given and indefinite; (ix) perpetual and new. C. Given the discussion of the relationship between indefinite and new on the one hand and definite and given on the other hand, explain why these indefinite noun phrases in sentence 7 represent given information: double, single, double 65 ft. , 60 - single ft. d. Identify the noun phrase that is the subject of sentences 1, 3, 4 and 5. e. The first sentence uses the passive voice. The equivalent of active voice is that Iowa doesn't seriously challenge any of these findings. Which noun phrase is addressed in the passive version? Given the topic you just identified, what is the paragraph before this passage likely about?

Exercises 8-7.

Examine the following passage and note the underlined passive verbs. Then give a pragmatic reason that might have motivated the authors to use each of the underlined passives. Then write an active verb for each phrase or sentence whose passive voice you cannot justify. Finally, look at your revised passage and assess whether the text is pragmatically more effective than the original. explain your answer A consequence of the ideological position that individuals are the foundation of society is that these individuals should be treated as equals. As will be discussed below under the topic of facial systems, this egalitarianism of utilitarian discourse does not apply to all people, but only to "those who can be improved by free and equal discussion" (Mill 1990:271-2). That is, this egalitarianism applies only to members of the utilitarian discourse system. [Adapted from Ron Scollon and Suzanne Wong Scollon, Intercultural Communication (Malden, MA: Blackwell, 1995, p. 110)]

Based on languages other than English As with Russian, Spanish word order is used to encode information structure. The parts of a sentence can be arranged in different ways, as shown in the following English examples, all of which can describe the same event. (S = subject; V = verb; O = direct object) Consuelo sent the package. (SVO) Consuelo sent

the package

Send the package to Consuelo. (VSO) be

comfort the package

package sent

comfort. (YOU)

the comfort package

The package

the broadcast

the package you sent

Kompfort. (OVS) Komfort

16666266663

8-8.

'Consolation sent the package.'

Consider the following conversations, focusing on the order of the parts in the responses. 1) Q: What is Consuelo? What

Consuelo

"What did Consuelo do?" A: Consolation prepared

Sangria

Consuelo made the sangria 'Consuelo made the sangria.' 2) Q: Who ate my sandwich? who

ate

my sandwich

"Who ate my sandwich?"

• 275

2 7 6 • Chapter 8 Information structure and pragmatics

For her

Consuelo aß das Sandwich.

your sandwich

ate

Consuelo

'Consuelo ate her sandwich.' 3) Q: Who did Consuelo give this gift to? Per

who comforted this gift

"Who did Consuelo give this gift to?" A: This gift was given by

comfort for

it is

madre

Consuelo gave this gift to her mother 'Consuelo gave this gift to her mother.' 4) Q: What happened? What

It occurred

"What happened?" A: Consuelo died. he died

Consuelo

'Consolation died.' 5) Q: Consuelo received the

forgive?

Consuelo received the award 'Did Consuelo win the award?' A: No, it doesn't.

award he received

the award

Paquita.

received Paquita

'No, Paquita won the award.' 6) Q: Did Consuelo receive this letter? He got

console this letter

"Did Consuelo receive this letter?" A: No, Consuelo did not receive anything

this package.

Consuelo received this package

'No, Consuelo received this package.' 7) Q: Did Consuelo receive the award? Consuelo received the award 'Did Consuelo win the award?' A: Yes, him

award he received

if the price

Consuelo.

received Consuelo

"Yes, Consuelo won the award." Using this data, describe how word order is used to mark information structure in Spanish sentences (but not in questions). In particular, indicate which information structure categories are characterized by which word order option. Make your rule statement as general as possible. B. Notice that in certain sentences the pronoun lo 'it' comes before the verb. What is the syntactic rule that dictates when it should and shouldn't appear? What English rule does the presence of the pronoun in these sentences remind you of?

Exercises 8-9.

Given the role of word order in Russian, ask an informational question (in English) for which sentences 5 and 6 on p. 266 would be pragmatically acceptable Russian answers.

8-10 Examine the following statements in Japanese, made while two friends were waiting at a bus stop. Explain why Yumiko Basu uses Ga to refer to the bus while Kimiko uses Basu Wa. Yumiko: Bus basu ga subject

Viene Kimasu

'The bus is coming.'

Kimiko: Basu wa Bus gegeben

konde-imas crowded-é

The bus is full.

8-11 Tongan has an operation that embeds the direct object (object) in the verb, forming a verb-noun conjunction. Generate a sentence like (a) from the basic structure of principle (b): a) naʔa

to you

past I

Pia

(embedded object)

drink beer

I drunk beer. (literally: 'I drank beer.') b) naʔa

ʔa

to you

past I

Pia

(object not embedded)

Object to drink beer

'I had a beer.' Here are three more examples of built-in object constructions (loosely translated to emphasize the meaning of the Tongan expression): ʔoku

nau

present she

He wants to eat

ika

hungry for

pez

(embedded object)

You are hungry for fish. no

Breast

sio

to build

past us

I understand

Film

(embedded object)

"We went to the cinema." 'Well

This is

present she

Do you

hiva-kakala

(embedded object)

compose a love song

He composes a love song. An embedded direct object cannot be followed by a restrictive relative clause, but an unembedded direct object can. Compare: *naʔa

to you

past I

Pia

[no

you come]

(embedded object)

Drinking beer in the past tense that they give me

'I had beer [they gave me]' naʔa

to you

past I

ʔa

Pia

[no

nau omai] (Non-embedded object)

drink object to beer past give me

'I had a/a beer [I was given]' Assuming that restrictive relative clauses have the same function in Tongan as in English, describe the pragmatic function of object embedding in Tongan.

• 277

2 7 8 • Chapter 8 Information structure and pragmatics

Specially for educators and future teachers 8-12. To refresh his students' memories of what constitutes given and new information, and how the state of information affects whether a noun phrase is marked definite or indefinite, he offers this brief account: I was at a wedding last weekend. The bride and groom were my friends. Then give your students these opening lines (below) of a news article about Annika Sorenstam (adapted from Los Angeles Times, March 20, 2003). It prompts you to fill in the blanks with a/an or the accordingly. Explain to a colleague how you would teach your students how to partially recover definite or indefinite articles using an explanation you think would be appropriate for your students. The world's best female golfer stands in ___ grill, on ___ black rubber mats, ___ pots and pans fly in formation on hooks overhead. Annika Sorenstam feels right at home on her ___ Cooking Tour ___ at Lake Nona Resort, where her mind is far from her world of pro golf. In the kitchen ___ you don't think of Pak, Webb or Inkster. These are crab cakes, stuffed mushrooms and tiramisu. That's why Executive Chef Gary Hoffman recently supplied 60 steaks for ___ a dinner in the ___ Club Dining Room. Sorenstam also knows how to carve ___ flowers from ___ melon slices, the easiest way to peel potatoes, and how to cook rice pilaf. Here is ___ story you love to tell. One night ___ a club member was enjoying his meal at dinner. "He said, 'Bring the cook,'" Sorenstam said. "Then Gary came out and the ___ member said, 'No, not that chef, that other one.' So I left." 8-13 In a dictionary suitable for your current or future students (or a desktop dictionary that you use yourself), check the entries for a and die to see which of the different definitions of each word best suits your understanding fits as explained in this chapter. Then assess whether these definitions adequately represent the facts about the indefinite and definite articles as you understand them. If not, write a corrected definition at a level appropriate for your students.

Andere Ressourcen AT&T Text-to-Speech-Demo: http://www.research.att.com/~ttsweb/tts/demo.php

In a previous chapter, you may have visited this page to see a speech synthesis demo and were impressed by the synthesizer's ability to produce the consonant and vowel sounds of the sentence you are transmitting. It is worth returning to the site to submit sentences illustrating some features of the information structure, such as: B. left shift or contrast. Judge for yourself how well this speech synthesis engine captures the intonation that conveys the pragmatic information in sentences.

references

Suggested Reading • Geoffrey N. Leech. 1983. Principles of Pragmatics (London: Longman). An accessible introduction to pragmatics. • Kenneth R. Rose and Gabriele Kasper, eds. 2001. Pragmatics in Language Teaching (Cambridge: Cambridge University Press). Especially for language teachers and those learning a second language, this series of essays by different authors is useful for highlighting often overlooked issues in language learning and teaching. • Jorge Jule. 1996. Pragmatics (Oxford: Oxford University Press). A concise and accessible introduction to pragmatics, appearing in a series designed to introduce students to various subfields of linguistics. In addition to the material covered in the current LISU chapter, The Christmas Book also covers the material covered in Chapters 6 and 9.

Further Reading For reviews of the topics discussed in this chapter, see Brown and Yule (1983), Lambrecht (1994), Foley and Van Valin (1985), Givón (1979a), Chafe (1976), McCarthy (1991) and Georgakoplou and Goutsos (2004). See Chafe (1994) for a thorough discussion of the issues discussed in this chapter. Works in Givón (1979b) and Li (1976) examine the interaction of syntax and pragmatics in different languages, while Chafe (1970) examines this interaction in English. Data and related issues are discussed in Prince (1979), Finality in Lyons (1999). The discussion of the slit-it and slit-WH constructions in this chapter draws on Prince (1978) and the discussion of the front and the Yiddish movement on Prince (1981). Lambrecht (1981) discusses left and right shifting in spoken French. English passive constructions are examined in Thompson (1987). See Comrie (1979) for a brief discussion of the function of Russian word order. For an overview of research on sentence intonation and stress and their pragmatic functions, see Bolinger (1986). In the classic by Halliday and Hasan (1976) other means of characterizing pragmatic structures in English are discussed. Schiffrin et al. (2001), a comprehensive treatment of discourse, is a useful resource for teachers.

References • Bolinger, Dwight L. 1986. Intonation and Its Parts: Melody in Spoken English (Stanford: Stanford University Press). • Brown, Gillian and George Yule. 1983. Discourse Analysis (Cambridge: Cambridge University Press). • Chafe, Wallace L. 1970. The meaning and structure of language (Chicago: University of Chicago Press). • Chafe, Wallace L. 1976. “Given, Contrastive, Definitive, Subjects, Themes, and Viewpoint,” in Li (1976), pp. 25–55. • Chafe, Wallace L. 1994. Speech, Consciousness, and Time: The Flow and Displacement of Conscious Experience in Speaking and Writing (Chicago: University of Chicago Press).

• 279

2 8 0 • Chapter 8 Information structure and pragmatics

• Comrie, Bernard. 1979. "Russian", in Timothy Shopen, ed., Languages and Their Status (Cambridge, MA: Winthrop), pp. 91–151. • Foley, William and Robert Van Valin, Jr. 1985. “Information Packaging in the Clause”, in Timothy Shopen, eds., Language Typology and Syntactic Description (Cambridge: Cambridge University Press), 3, pp. 282–384. • Georgakoplou, Alexandra and Dionysis Goutsos. 2004. Discourse Analysis: An Introduction, 2nd ed. (Edinburgh: Edinburgh University Press). • Givon, Talmy. 1979a. For Understanding Grammar (New York: Academic). • Givón, Talmy, ed. 1979b. Syntax and Semantics 12: Discourse and Syntax (New York: Academic). • Halliday, M.A.K. and Ruqaiya Hasan. 1976. Cohesion in English (London: Longman). • Johnstone, Barbara. 2002. Discourse Analysis (Malden, MA: Blackwell). • Lambrecht, Knud. 1981. Subject, Antisubject, and Verbal Agreement in Non-Standard French (Amsterdam: Benjamins). • Lambrecht, Knud. 1994. Information Structure and Sentence Form: Theme, Focus, and the Mental Representation of Referees in Discourse (Cambridge: Cambridge University Press). • Li, Charles N., editor. 1976. Subject and Subject (New York: Academic). • Lyon, Christopher. 1999. Definition (Cambridge: Cambridge University Press). • McCarthy, Michael. 1991. Discourse Analysis for Teachers (Cambridge: Cambridge University Press). • Prince, Ellen F. 1978. “A comparison of 54:883–906.

cracks w

and language cracks”, language

• Prince, Ellen F. 1979. „On the Given/New Distinction“, Papiere vom Fünfzehnten Regionaltreffen der Chicago Linguistic Society (Chicago: Chicago Linguistic Society), S. 267–78. • Prince, Ellen F. 1981. „Topicalization, Focus Movement, and Yiddish Movement: A Pragmatic Differentiation“, Proceedings of the Seventh Annual Meeting of the Berkeley Linguistic Society (Berkeley: Berkeley Linguistic Society), S. 249–64. • Schiffrin, Deborah, Deborah Tannen und Heidi E. Hamilton. 2001. Das Handbuch der Diskursanalyse (Malden, MA: Blackwell). • Thompson, Sandra A. 1987. „The Passive in English: A Discourse Perspective“, in Robert Channon und Linda Shockey, Hrsg. 497–511.

9 Speech Acts and Conversation WHAT DO YOU THINK? • Her friend Isabella wonders aloud why the words “I now pronounce you husband and wife” constitute a legal marriage between two people at a wedding but not when uttered in a play. First, you dismiss his question as stupid. "No, really," she says. "What is the difference?" You think the answer is obvious, but you try. What's the best explanation you can give? • Kimberly complains that last week her boyfriend, Tyler, promised to take her next time he went skiing. So he went skiing this weekend and didn't invite her. When asked what Tyler said when he made his promise, Kimberly reports that he said, "I will. I'm coming. Honestly, I will. You ask if Tyler said "I promise" and Kimberly says "No." You tell him that you think Tyler is an idiot, but also gently points out that you're not sure if he made a promise because he didn't use that word. "Come on," he says, "a promise is a promise." How do you explain your interpretation to her? • Your younger brother Brandon complains that Daniel, a French college friend, takes forever to get straight to the point when he calls you and apologizes endlessly! Brandon wonders what's wrong with Daniel and why he can't get down to business. You know that people from different cultures behave differently on the phone. What explanation do you offer Brandon for Daniel's behavior on the phone?

281

2 8 2 • Chapter 9 Speech acts and conversation

Common Language People use language primarily to do things: ask a favor, promise something, report news, give directions, say greetings, look up information, invite someone to dinner, and hundreds of other verbal activities. . actions of daily life. Sometimes the things we do with the language have serious consequences: proposing marriage, going to court, taking an oath, telling the truth, firing an employee, etc. These speech acts are part of speech events such as conversations, lectures, student-teacher conferences, Newsreels, weddings and court cases. Aside from births, deaths, fires, robberies, hurricanes, car accidents, and the like that are not speech acts, much of what is reported on the newspaper pages is speech acts: arrests, predictions, denials, promises, accusations. warnings and so on. The word and sentence structure has been discussed in earlier chapters of this book. Now let's look at what we do with these structures and how our statements work. Knowing a language does not simply mean knowing how to encode a message and transmit it to a second person, who decodes it to understand what is meant. If language use were simply about coding and decoding messages, i.e. grammatical competence, each sentence would have a fixed interpretation regardless of its context of use. However, this is not the case, as the following scenarios illustrate. 1. You are pulled over by a police officer who surprises you by telling you that you have just run over a stop sign. "I didn't see the stop sign," you say. 2. A friend gave you directions to her apartment, including directions to turn left at the first stop sign after the intersection of Oak and Broad. You're about 30 minutes late and you say, "I didn't see the stop sign." 3. You are driving with an aunt who is in a hurry to get to church. He brakes and runs over a stop sign, knowing full well that there is hardly any traffic at this intersection on a Sunday morning. As she enters the intersection, she sees a car approaching and slams on the brakes, surprising her aunt. "I didn't see the stop sign," you say.

For the officer, your statement ("I didn't see the stop sign") is an explanation for not stopping and a subtle plea not to be cited for the violation. For the friend, his testimony is an apology for his tardiness and a reproach that he was not at fault, either intentionally or entirely. He says the same phrase to his aunt (incorrectly in this case) as an apology for scaring her. She acknowledges his apology intent and says, "Okay. But be careful. The linguistic meaning of the phrase "I don't see the stop sign" is the same in all three cases, but the pronunciation in these different contexts serves different purposes and conveys different messages.

The structure of sentences and the function of sentences Traditional grammar books say that declarative sentences make statements (It's raining), imperative sentences give instructions (Close the door), and interrogative sentences ask questions (What time is it? ). This analysis is oversimplified, even misleading. Look at the sentence, can you close the window? Taken literally, its question structure requires a

Speech acts challenge the recipient's ability to close a particular window. If a roommate were to ask this question while a college marching band was practicing nearby, they probably wouldn't interpret it as a question about their ability (and therefore require a verbal response), but rather as a request to close the window. (A request in the form of a question is indicated in speech by the absence of a loud voice, and sometimes in writing by the absence of a question mark - answer promptly). On the other hand, the imperative structure Tell me your name again would not normally be used as an instruction to do something, but as a request for information. Let's take another case: Suppose there's a knock at the door and Megan tells Alex that she'd like to know who's at the door. If Megan believed Alex knew the answer, this statement could be pronounced as a request for information. Often, however, it was a polite request for Alex to open the door. Finally, interrogative sentences can sometimes be used to make statements, as in Suze's answer to Eric's question. Eric:

Is it very easy to get along with Amy?

Suze:

Do chickens have teeth?

Suze's question conveys an emphatically negative response to Eric's question. Two things, then, are clear: (1) people often use declarative, interrogative, and imperative clauses for purposes other than making statements, asking questions, or giving orders; and (2) a fundamental element in interpreting a statement is the context in which it is uttered. Recall the three faces of language usage presented in Chapter 1 (page 6), showing context as the base of a triangle connecting meaning and expression. You realize that a sentence is a structured sequence of words that have a specific meaning. Rather, an utterance is a sentence said, written, or signed in a specific context by someone with a specific intent, where the "speaker" is trying to make an impact on the addressee. So as a question: Can you close the window? has the meaning of a request for information ("Can you close the window?"), but as a contextualized expression it would most often be a call to action ("Please close the window"). Drawing the appropriate conclusions from the conversation is an essential part of interpreting the utterances. To understand expressions one must have the ability to "read between the lines" and the skills one uses to use and interpret the sentences formed through grammatical competence are part of one's communicative competence.

Speech Acts In addition to what we accomplish through physical acts like cooking, eating, biking, gardening, or driving the bus, we accomplish a lot every day through verbal acts. In face-to-face conversations, phone calls, letters of application, scribbled notes to a roommate, and countless other language events, we engage in verbal actions of all kinds. In fact, language is our primary means of greeting, congratulating, and insulting, begging, or begging on a typical day flirt, seek and provide information, and perform hundreds of other tasks. Actions performed through language are called speech acts, and a surprising number of news reports are speech act reports.

• 283

2 8 4 • Chapter 9 Speech acts and conversation

Try it for yourself: Decide which of these headlines report speaking acts and which report physical acts: Sony unveils ambitious plan for musical fight beers in borders FCC judge Skid Row sweeps free smuggling suspects

Gap wins a contract with CSC and hires the founder of Etoys Winds Wreak Havoc POW who was rescued from his captors

Types of Speech Acts Among the different types of speech acts, six have received particular attention: 1. Representatives represent a state of affairs: statements, propositions, assertions, hypotheses, descriptions, suggestions. Representatives can generally be characterized as either true or false. 2. Commissioners commit the speaker to a course of action: promises, promises, threats, vows. 3. Directives are intended to prompt the recipient to act: orders, requests, challenges, invitations, requests, challenges. 4. The utterances bring about the facts they name: blessings, appointments, dismissals, baptisms, arrests, marriages, lawsuits. 5. Expressions indicate the psychological state or attitude of the speaker: greetings, apologies, congratulations, condolences, thanks. 6. The judgments make evaluations or judgements: classification, evaluation, evaluation, tolerance. Because some judgments (such as calling out a baseball player) combine the characteristics of pleas and representatives, they are sometimes referred to as representative actions.

Idioms and illocutions Every speech act has several main components, two of which interest us directly here: the utterance itself and the speaker's intention to make it. First, each utterance is represented by a sentence with a grammatical structure and a linguistic meaning; This is the sentence Second, speakers have some intention in making a statement, and what they are trying to achieve is called illocution. (A third component of a speech act, which we will not elaborate on, is the effect of the act on the listener; this is perlocution, or "absorption".) Consider the statement: Can you close the window? Like all utterances, it can be considered an idiom and an illocution. The voiceover is a yes/no question about the recipient's ability to close a particular window; as such, the convention would require a yes or no answer. Suppose the intention of the speaker (the illocution) is to ask the addressee to close the window; thus the Convention would allow the addressee to recognize the structural problem as a call for action and to act or not to do so. In discussions of speech acts, it is common for the illocutionary act itself to be referred to as a speech act; hence promises, promises, threats, invitations, etc. are all speech acts.

speech acts

Distinguishing between speech acts How do people distinguish between different types of speech acts? How do we know if a sentence like yours is tense? Is it a yes/no question (do you have time [to help me]?) or a request for information about the time of day? To state the question more technically, given that an utterance can serve many functions, how do addressees know the illocutionary power of a speaker's utterance? The answer, of course, is “context”. But how do people correctly interpret the context? We begin our analysis by distinguishing between two broad types of speech acts. Compare the following two statements: 1. I now declare you husband and wife. 2. It's going to be a very windy day.

In the appropriate context, the first statement establishes a new relationship between two people; it is a statement that defines a marriage. The second statement is a simple statement or statement of fact. As any weather forecast will attest, it has no bearing on the weather. As you have already seen, statements such as sentence 2 make statements or express opinions and are representative. Expressions like sentence 1 change the state of affairs and are propositions; They impressively illustrate how the language used is a form of action. Children are subjected to fantastical statements such as "Abracadabra, I'll turn you into a frog!" They eventually learn that real-world objects are more unruly than fairy-tale objects, but all speakers acknowledge verbal power over certain aspects of life, particularly in relation to social relationships. With the declaration that I am now declaring you husband and wife, the nature of the social relationship between two people can change fundamentally. Likewise the statement You are under arrest! it can have social liberty implications, just like the cold case. A referee can change a baseball game with a statement as simple as "Sure!" or hit three! Normally, to be effective, declarations of this type must be made by a specially designated person. If you get a call from an unassigned person, tell a fan in the stands, get out! it would be a judgment, not a finding. In fact, a statement from a particular referee will reverse the contrary decision of an entire stadium of fans.

Eligibility Conditions and Successful Claims The validity of a claim depends on established conventions. I now declare that you, husband and wife, may unite two persons in marriage, but only if several conditions are met: the setting must be a marriage ceremony and the declaration must be made in due time; the speaker must be destined to marry others (a minister, rabbi, justice of the peace) and must intend to marry them; the two people must be legally entitled to marry; and they must intend to become spouses. Finally, of course, the words themselves must be spoken. If a condition is not met, the utterance of the words is ineffective as a performative speech act, one whose words produce the action. Made in a Hollywood film played by an actor playing a pastor and directed at two actors playing characters about to be married, the declaration may help secure an Oscar, but it will not affect marriage. The conventions governing the conditions under which an utterance serves as a particular speech act, such as B. marriage, promise, arrest, invitation, appropriate conventions were mentioned.

• 285

2 8 6 • Chapter 9 Speech acts and conversation

Property relations of the philosopher John Searle and can be divided into four categories. 1. The condition of propositional content requires only that the propositional words are conventionally associated with the intended speech act and convey the content of the act. The idiom must contain conventionally acceptable words to perform the specific speech act: Is it raining? Now I pronounce you husband and wife, you're under arrest I promise. . . , swear . . . . 2. The preparatory condition presupposes a conventionally recognized context in which the speech act is placed. A wedding must be a real (albeit informal) wedding ceremony where two people intend to exchange vows in the presence of a witness. 3. The condition of sincerity requires that the speaker make the statement sincerely. At a wedding, the speaker must intend the wedding words to bring about a marriage; Otherwise, the sincerity condition is violated and the speech act fails. 4. The essential condition requires that all parties involved intend the outcome; For example, in a marriage ceremony, when the participants speak the words I declare to you now to husband and wife, they must have intent to create a marriage bond.

Successful Promises Now look at the Commissioner, I promise to help you with your math tonight. For such a statement to be successful, it must be recognizable as a promise; In addition, the preparatory, sincere, and essential conditions must be met. In the propositional content condition, the speaker must use the conventional term promise to express intention to help the addressee. The preparatory condition requires that speaker and listener be sane and responsible, that the speaker believes they can help with math, and that the recipient wants help. For example, the preparatory condition would be violated if the speaker knows that he or she cannot be present or cannot perform the calculations alone, or if the participants are reading a script that contains the utterance. If the speaker knew that the listener did not want help, the promise would not succeed. For the sincerity condition to be valid, the speaker must sincerely intend to help the addressee. This condition would be violated (and the promise formula abused) if the speaker did not intend it. Finally, the essential condition of a promise is that the speaker intends, through pronunciation, to undertake to provide assistance to the listener. These four fitness conditions define a promise of success. Successive prompts and other speech acts Suitability conditions are useful to describe not only utterances and commitments, but all other types of speech acts. In a typical request (please pass me the salt), the content of the statement must identify the action requested by the listener (pass me the salt) and its form must conform to the conventionally accepted form for making requests. The preparatory condition involves the speaker's belief that the recipient is able to dispense the salt and would not have dared if he had not asked her to dispense it. The sincerity condition requires that the speaker genuinely wants the listener to pass the salt. Finally, the essential condition is that the speaker intends the statement to get the hearer to pass him the salt.

the cooperative principle

The Cooperative Principle The principles governing the interpretation of statements are varied and complex, and differ somewhat from one culture to another. Even within a single culture, they are so complex that one might wonder how language manages to communicate so well. The principles we examine in this section are by no means universal, sound as they may appear to Western readers; As you will see later, what seems like common sense for one group may not be common sense for all groups. Despite the occasional misinterpretation, in most situations people manage to understand the phrases essentially as they are meant. The reason for this is that interlocutors, often without reason to expect otherwise, trust that they and their interlocutors are following the same conventions of interpretation. Listeners simply assume that the speakers followed the conventions of interpretation in constructing their utterances. Speakers, on the other hand, must make a double assumption: not only that the listeners themselves are guided by the conventions, but also that the listeners trust that the speakers have respected those conventions in constructing their utterances. There is an unspoken agreement that people will work together to communicate with each other, and speakers rely on this collaboration for efficient conversation. The cooperative principle, as formulated by the philosopher H. Paul Grice, is as follows: Make your contribution to the conversation as needed, at the stage at which it occurs, for the accepted purpose or direction of the conversational exchange in which you are engaged . This cooperation agreement covers four areas of communication, each of which can be described as a general maxim or principle.

Maxim of Quantity First, speakers are expected to provide as much information as is necessary for their interlocutors to understand what they are saying, but not to provide more information than is necessary. If you ask someone you know if they have pets and they say I have two cats, that's the maximum number you can assume they have no other pets. The conversational implication of such an answer is "I have two (and only two) cats (and no other pets)". Notice that I have two cats, even though the speaker has six cats, or six cats, two dogs, and a llama. But if he had other pets, you would have reason to feel betrayed. While your answer wasn't that far wrong, it would have violated your culturally defined expectation that no relevant information would be withheld. In most (but not all) Western cultures, listeners expect speakers to obey this maxim, and just as importantly, speakers know that listeners believe they are doing so. It is this tacit cooperation that creates conversational implicatures. As another example, suppose you ask a man who is painting his house what color he chose for the living room and he replies: The walls will be cream to contrast with the black sofa and Regency chairs that I inherited from my grandmother. God bless his soul, he passed away last year after a long marriage to my grandfather who never appreciated his love

• 287

2 8 8 • Chapter 9 Speech acts and conversation

Performing Arts. So the trim will be peachy except near the door, which Amber says should be salmon colored so it doesn't clash with the black and red Picasso print I brought back from Spain when I was vacationing there. Let's see, I think that was 2002. Or was it 2001? I actually forget. hell again! Time flies fast doesn't it? And the stairs leading to the bedrooms will be pale yellow. By providing too much information, much more than was requested or expected, the man is just as uncooperative as the woman who withholds information about her pets. The quantity principle dictates that, under normal circumstances, speakers say enough, giving no less and no more information than is necessary for the purpose of the communication: to be sufficiently informative. Society stigmatizes individuals who habitually violate the maxim of quantity; those who provide too much information are described as "never shut up" or "always tell everyone their life story," while those who routinely do not provide enough information are described as moody, reserved, or uncommunicative.

Maxim of Relevance The second maxim instructs speakers to organize their statements so that they are relevant to the current context: be relevant at the time of the statement. The following interaction illustrates a violation of this maxim. Zane: How's the weather outside? Zora: There's a great movie on HBO Thursday night.

Literally, Zora's expression doesn't seem to have anything to do with what Zane just said; if that were so, it would violate the principle of relevance. Because of the principle of relevance, when someone makes a seemingly irrelevant statement, it is often difficult for the audience to understand how it could be relevant (perhaps as a joke or as an indication of displeasure with the direction of the conversation). Chronic violations of this maxim are characteristic of schizophrenics, whose sense of "context" differs radically from that of other people.

Maxim of manners Third, people follow a variety of rules that are summarized under the maxim of manners. Summarized in the guideline order and clarity, this maxim requires speakers and writers to avoid ambiguities and ambiguities and to bring order to their statements. In the following example, the order maxim is violated. A birthday cake should have frosting; use unbleached flour and sugar in the cake; bake for an hour; Preheat oven to 325 degrees; and beat in three fresh eggs. This recipe is odd for the simple reason that English speakers tend to follow a chronological order of events when describing a process like baking. The order is not only determined by the order of events: in every language there are rules that dictate a "natural" order of details in a description. Since more general details often precede more specific ones in American English, the result seems odd when a speaker breaks this rule.

Violations of the cooperative principle There are five shopping centers in my hometown. It is the county seat. My father and mother were born there. My hometown is a Midwestern city of 105,000 people in the center of the Corn Belt. I grew up there until I was 13 years old. As a third example, consider the statement that Ted died and was struck by lightning. If it was the lightning that killed Ted, then good manners had been violated here. Although in logical and union clauses whose temporal reference is irrelevant (ie She studied chemistry and She studied biology is logically equivalent to She studied biology and she studied chemistry), the maxim dictates that a statement like You had a baby and married something other has conversational meanings. Implications that they married and had a child. The maxim of manner in this case suggests that the sequence of expressions reflects the sequence of events or is irrelevant to a correct interpretation. Of course, English and other languages offer ways to avoid misinterpretation: they had a baby before they were married; first they had a son and then they married; they married after having a child; etc.

Quality maxim The fourth general principle that governs the rules of language interpretation is the quality maxim: be true. Speakers and writers are expected to say only what they believe to be true and have evidence for what they say. Again, the downside is that the speakers are aware of this expectation; They know listeners expect them to live up to the epitome of quality. Without the quality maxim, the other maxims are of little value or interest. Short or long, relevant or irrelevant, orderly or messy, all lies are false. However, the maxim of quality applies above all to utterances and some other representative speech acts. The expressive and the directive can hardly be judged true or false in the same sense. It makes sense to think further about the maxim of quality. On the one hand, it is this maxim that obliges the interlocutor to tell the truth and to have their statements documented. Ironically, however, it is this maxim that also makes lying possible. Without the quality maxim, speakers would have no reason to expect listeners to believe what they are saying, and without the assumption that speakers assume someone is telling the truth, it would be impossible to lie. Lying requires the speaker to be expected to tell the truth.

Violations of the cooperative principle It's no secret that people sometimes violate the maxims of the cooperative principle. Certainly not all speakers are perfectly honest; others have failed to realize that efficiency is the desired Western norm in conversational interaction. Even more interesting is that speakers are sometimes forced by cultural norms or other external factors to violate a maxim. For example, regardless of your aesthetic judgement, you may feel embarrassed to say what a beautiful picture! for a host who is openly proud of a newly acquired work of art. The need to adhere to social conventions of politeness sometimes invites people to violate the maxims of the cooperative principle.

• 289

2 9 0 • Chapter 9 Speech acts and conversation

Indirect Speech Acts As mentioned above, interrogative structures can be used to make polite calls to action, imperative structures can be used to ask for information, etc. These uses of a meaningful structure to accomplish another task often play a role in mutual interaction, as in this exchange between colleagues who stayed in the office after dark. Kayla:

Is it the boss?

Ryan:

The light is on in your office.

Kayla:

Gracias.

Ryan's answer contains no obvious reference to the information Kayla is looking for. In theory, this seems to violate the maxim of relevance. However, Kayla is happy with the answer. Kayla acknowledges that the literal interpretation of Ryan's answer violates the maxim of relevance, but assumes that Ryan is relevant as a cooperative speaker and pursues an indirect interpretation. To help her, she learns some facts about her boss' habits: that he works in his own office, that he doesn't work in the dark, and that he's not in the habit of leaving the lights on when he leaves the house. during the day. Based on this information, Kayla deduces an interpretation of Ryan's statement: Ryan believes the boss is inside. For example, an indirect speech act could be based on an obvious violation of the quality maxim. When we describe a friend as someone who never loses a dime, we don't mean it literally; we exaggerate. By exaggerating the information, it may appear that we are not respecting the highest quality. But listeners generally appreciate that the statement is not meant to be taken literally and adjust their interpretation accordingly. In the same way, we can exclaim in front of the Sears Tower in Chicago: It's an awfully small building! This statement seems to violate the maxim of quality insofar as we make an obviously wrong assessment. But speakers readily recognize the irony of such expressions and regard them as indirect speech acts intended to convey an opposite meaning. Characteristics of indirect speech acts Four characteristics of indirect speech acts can be identified from these examples: 1. Indirect speech acts violate at least one maxim of the principle of cooperation. 2. The literal meaning of the phrase of an indirect speech act differs from its intended meaning. 3. Listeners and readers identify indirect speech acts by noting that an utterance has feature 1 (violation of a maxim) and assuming that the speaker follows the cooperative principle. 4. Once listeners and readers have identified an indirect speech act, they use context and world knowledge to identify its intended meaning.

Thus, to interpret indirect speech acts, listeners use maxims to resolve the discrepancy between the literal meaning of the statement and an interpretation appropriate to the context in which it is uttered.

courtesy

Try it yourself: In this short exchange, what would you call the speech act in A? in B? Does D represent a direct or indirect speech act? And the A? A

Anna: Who ate the bread I baked yesterday?

Juan:

Anna: Yes.

Juan:

With the raisins? Did you ask Raul?

Indirect speech acts and shared knowledge A prerequisite for a successful indirect speech act is that interactors exchange sufficient information about the context of the interaction, about each other and their society, and about the world in general. When Jacob asks Emma, have you finished your sociology paper? and she answers Rome is in Spain? Jacob will probably recognize the answer as an indirect speech act. But whether you can interpret it or not depends on your knowledge of geography. Using and understanding indirect speech acts requires familiarity with language and society. To give an example from another culture, when speakers of the Polynesian Tuvaluan language want to comment on the fact that a particular person is in the habit of talking about themselves, they might say: koo tagi te tuli ki tena igoa “The plover is singing his own name. ." The expression derives from the fact that the plover's cry sounds like a high-pitched "tuuuuuliiiiiii," from which Tuvaluan speakers coined the word tuli to refer to the bird itself. Thus the expression became an indirect form of criticism, the quality of praising oneself. To interpret the phrase as an indirect speech act, one must be familiar not only with the plover's cry and its resemblance to the bird's name, but also with the fact that Tuvaluans see people who speak about themselves as a bird who "sings his own name". Of course, a significant amount of information about language, culture, and the environment is required to interpret indirect speech acts.

Politeness Indirect speech acts appear to be a complicated mode of communication. Not only do you have to identify them, you also have to go through a complex thought process to interpret them. One might think that it would be more efficient to communicate directly. But the fact of the matter is that indirect speech acts are about asking and answering questions, criticizing others, etc. Sometimes they add humor, and sometimes they show politeness. Emma's indirect response (Is Rome in Spain?) to Jacob's question suggests, “Don't be ridiculous; of course I'm not done yet.' Ask how to close the window? are perceived as more polite and less intrusive and aggressive than a command like close window! One message conveyed by indirect speech acts is "I'm being polite to you." Indirect speech acts are therefore an effective means of communication: they can convey two or more messages at the same time.

• 291

2 9 2 • Chapter 9 Speech acts and conversation

Respect Independence and Show Commitment Courtesy has two basic aspects. The first is based on the fact that humans respect the privacy, independence, and physical space of others. We avoid intruding on other people's lives, we try not to be overly curious about their activities, and we are careful not to impose our presence on them. We respect your independence and do not impose (some call this negative courtesy). On the other hand, when we show people that we enjoy their company, feel comfortable with them, know something about their personality, or care about their well-being, we are showing engagement (what some refer to as positive politeness). While everyone expects independence and commitment, the former requires that we leave people alone, while the latter requires that we do the opposite. Fortunately, these conflicting needs often arise in different contexts. When we lock ourselves in a room or walk alone on the beach, we assert our right to independence. When we go to a party, take someone out to dinner, or call friends to see how they are doing, we are showing commitment. Both are forms of politeness. In conversation, partners exchange messages about their needs for independence and their desire for involvement, and they also acknowledge each other's needs for both types of courtesy. The expectation that others will not ask embarrassing questions about our private lives stems from the need for independence. On the other hand, when you tell a friend about a personal problem and expect sympathy, look for involvement. If you apologize before asking a stranger the time, you acknowledge the stranger's right not to be intrusive. When we express the hope of meeting a listener at a later date (Let's meet soon!), we recognize the listener's need for engagement and companionship.

Lecture Events News broadcasts, public speaking, class lectures, religious sermons, and conversations are lecture events in which members of a community interact on one or more topics with a specific purpose and awareness of the social relationships between the interlocutors. "Speech" events do not have to include a speech: Personal letters, shopping lists, office notes, birthday cards, and newspaper articles also count as speaking events. Conversation represents the matrix in which native languages are acquired and stands out as the most common, natural and representative verbal interaction. A person can spend a lifetime without writing a letter, composing a poem, or debating public policy, but only in rare cases does someone not engage in frequent conversations with friends and associates. Conversation is an everyday language event. We do it for fun (to gossip, pass the time, strengthen social bonds) and for work (getting help with studying, renting an apartment, ordering a meal at a restaurant). Regardless of its purpose, conversation is our most basic verbal interaction. While lovers in movies can have frank conversations back-to-back, the conversation usually involves people facing each other and taking turns. They do not speak at the same time or prolong the conversation. In some societies, even with multiple speakers in a single conversation, there are only tenths of a second between turns and very little overlap in speech. At the start of a conversation, people perform certain rituals, shake hands, or shake hands.

The organization of the conversation over time. Likewise, at the end of a conversation, people don't just turn around and walk away; They make sure everyone in the group has finished what they wanted to say, and only then say something like "I have to walk" or "Take care." Throughout the interaction, the interlocutors maintain a certain order: they take turns, do not interrupt each other too often, and follow other highly structured but unspoken patterns of conversation. These guidelines can be thought of as rules of conduct that govern speaker behavior. While it's tempting to imagine a relaxed conversation as essentially free of rules or constraints, the fact is that many rules are at work, and unconsciously recognizing those rules helps identify certain interactions as conversations.

The Organization of Conversation If it is surprising that informal conversation is organized by rules, it is because, as with most public speaking events, more attention is paid to content than to organization; the organization of discussions is a matter of course for us. A conversation can be seen as a series of speech acts: greetings, questions, congratulations, comments, invitations, requests, refusals, accusations, denials, promises, farewells. For these speech acts to work, a certain amount of organization is essential: we take turns speaking, answering questions, marking the beginning and end of a conversation and making corrections if necessary. To get this job done quickly, callers can give each other traffic alerts. Okay, now it's your turn to talk. I just asked one question; Now you must react, and do it now. If you have anything else to add before closing this thread, please do so now as I'm leaving in a minute.

However, such guidance would be ineffective and distract from the content. In unusual circumstances, speakers will invoke the rules (Would you please stop interrupting? or Well, say something!), but invoking the rules underscores the fact that they have been violated and can be perceived as impolite. Conversations are often organized in secret, and the organizing principles provide an unobtrusive interactive structure. The architecture of the covert conversation should achieve the following: arrange turns so that more than one person has an opportunity to speak and turns are orderly in turn; enable those involved to anticipate what will happen next and, if there is a choice, how the choice will be decided; provide a way to troubleshoot failures and errors when they occur.

Turns and pauses Participants must tacitly agree who should speak and when. We normally take turns keeping our word, and we do so without open negotiation. A useful way to discover alternation conventions is to observe what happens when they fail. When a participant fails to speak when asked to take their turn, the other speakers often pause and then someone else begins speaking. In this example, Emily repeats her question, assuming Sarah didn't hear or understand her the first time.

• 293

2 9 4 • Chapter 9 Speech acts and conversation

Emilio:

But don't they like it? [break]

Emilio:

[louder] So you didn't like it?

Sara:

Oh no. But then he said no.

It also violates take-off conventions when two people attempt to speak at the same time. In the following example, the beginning and end of the overlap are marked with square brackets. Speaker 1:

After John's party we went to Ed's house.

Speaker 2:

So you-

so you-you-[

Speaker 3:

]

What time did you get there?

When such competition arises in a casual conversation, the speaker may quickly blurt out the word or turn up the volume and continue speaking. Both silence and double-talk are serious problems in conversation, and the rules for taking turns are designed to minimize them. Different cultures have varying degrees of tolerance for inter-phrasal silences, overlap in speech, and competition between speakers. For example, in Inuit culture and some other Native American cultures, people sit comfortably in silence. On the other hand, in French and Argentinian cultures, multiple interlocutors tend to speak at the same time and interrupt each other more often than Americans typically do. Tolerant as they are of silence and overlap, people across cultures seem to regulate rounds of conversation in essentially similar ways: speakers signal when they wish to end their round by selecting the next speaker or leaving the option open; the next speaker takes the floor and begins to speak. These simple principles, which seem second nature to us, govern conversational turns very efficiently. Turn signals Speakers use verbal and non-verbal signals to signal that their turn is about to end. Since phrases often end in a complete sentence, the completion of a sentence can signal the end of a phrase. A sentence ending with the tag question (no? right?) explicitly invites a speaker to speak. Speaker A: It's pretty windy today, isn't it? Speaker B: Of course!

The end of a phrase can also be signaled by sharply raising or lowering the pitch of the voice, or by drawing the last syllable of the last word of the phrase. In very informal conversations, a common signal is the phrase or something like that. Announcer 1: So he was pretending like he got hit by a truck or something. Speaker 2: Really?

Other phrases that can signal the end of a curve are know, something like don't know (or don't know), and um at the end. As you know, some of them can also alternate, allowing the speaker to keep the floor while he thinks about what to say next. Another way to signal the end of a round is to pause and not try to speak again.

The organization of the conversation Daniel:

David:

I really don't think I should have said that at the meeting, let alone in front of the whole committee. It was really insensitive. [pause] Yes, I agree.

Of course, speakers often need to stop in the middle of a phrase to think about what to say next, to emphasize a point, or to catch their breath. To indicate that a speaker has finished once, the pause must be long enough, but "long enough" varies from culture to culture. Both nonverbal and verbal cues can signal the end of a round. Although the main role of gestures in speaking is to support and emphasize what is being said, continuing our hand gestures lets our interlocutors know that we have more to say. After resting our hands, our interlocutors can infer that we give up the word. In a more subtle way, staring can help control lift and spin. In mainstream American society, speakers generally do not look at their interlocutor; Instead, his gaze darts quickly and almost imperceptibly between his listener and another point in the room. On the other hand, since listeners tend to fix their gaze on the speaker, a speaker who reaches the end of a line may simply stare back at a conversation partner, signaling that it is their turn to listen and it is their turn to speak . In cultures where listeners look away while speakers stare, a speaker who wants to stop simply looks away. While visual gaze plays a supporting role in taking turns, the success of phone calls makes it clear that staring is not essential when taking turns. Taking the floor In multi-party conversations, the speaker can use the floor to choose who will speak next, or the next speaker can choose themselves. First, the speaker can signal the choice by addressing the next speaker by name (What have you been up to lately, Helen?) or by addressing the next selected speaker. If the speaker does not choose the next speaker, anyone can take the floor, often starting the round at a fast pace to block other potential speakers. If the spokesperson does not choose the next speaker, a contest may ensue, as in the example below where overlaps are indicated in square brackets. Speaker 1: Who will be at Jake's party on Saturday night? [pause] Speaker 2: Todd stops - [] Speaker 3: I don't know - [pause] Speaker 2: Todd told me - [

]

Announcer 3: I don't know who it is... [brief pause] Announcer 2: [to Announcer 3] Come on in! Speaker 3: I don't know who will be there, but I do know that there will be some activity. Interlocutor 2: Yes, that's what I wanted to say. Todd told me there would be a lot of people there.

Friendly participants strive to solve this competition quickly and smoothly.

• 295

2 9 6 • Chapter 9 Speech acts and conversation

The social inequality between the interlocutors (boss and employee, father and son, doctor and patient) is often reflected in the frequency and objection of the participants. In American work environments, supervisors often start conversations by asking a question and letting subordinates report back. Subordinates keep their word longer than superiors; Subordinates act while superiors act as spectators. In some cultures, superiors talk while subordinates listen.

Adjacency Pairs A useful mechanism in organizing secret conversations is that certain turns have specific following turns associated with them. Questions that request information require answers. The response to a greeting is usually also a greeting, to an invitation an acceptance or rejection, etc. Certain loop sequences go together, as in these adjacency pairs. Requests for information and provision of information Adam: Where is the milk I bought this morning? Betty:

At the counter.

Invitation and Acceptance Alex:

I invited some friends over for dinner on Saturday and I'd love for you to come.

Berto:

To the right!

Evaluation and Disagreement Angel: I don't think Nick would play a trick on you. British:

Well, you obviously don't know Nick very well.

These adjacency pairs include two loops, one directly following the other. In a question/answer neighborhood pair, the question is the first part and the answer is the second part. Here are more examples of adjacency pairs. Ask a favor and grant a guest: May I use your phone? Host:

To the right.

Apologies and acceptance Eli:

Sorry to bother you so late at night.

David:

No, everything is OK. And since?

Quotation and identification marks:

Invoice!

Invoice:

Sim?

Structural Features of Adjacency Pairs Three features of adjacency pairs can be observed. 1. They are contiguous. The two parts of an adjacency pair are connected and spoken by different speakers. A speaker who makes a statement before responding to a

The conversation organization question that was asked sounds odd (and can be frustrating) because adjacent pairs are structured to follow each other: Adam: Where's the milk I bought this morning? Betty:

They said on the radio that the weather would get better at noon. It's on the counter.

2. You are neat. The two parts of an adjacency pair are ordered. Except on TV shows like Jeopardy, the answer to a question must not come before the question. Typically, an invitation cannot be accepted until it has been offered, and an apology cannot be accepted until it has been delivered (except sarcastically). 3. They are compatible. The first and the second part of an adjacency pair are paired accordingly. Proper correspondence avoids awkward exchanges like the following: Kimi: Would you like more coffee? Sascha:

It's okay, you don't bother me one bit!

Insertion Sequences Sometimes the requirement that the two parts of an adjacency pair be contiguous is violated in a socially acceptable way. Adam: Where's the milk I bought this morning? Betty:

Skimmed milk?

Adam: Ja Betty:

At the counter.

In this example, in order for Betty to give an accurate answer to Adam's question, she must first know the answer to another question and therefore start an insertion sequence: an adjacency peer that interrupts the original and puts it "on hold". The interaction therefore consists of an adjacency peer embedded in another, as in the following telephone conversation. S Speaker 1: May I speak to Mr. Higgins? main d Announcer 2: May I ask who is calling? Adjacency Insertion Sequence d Announcer 1: Arthur Wilcox. Pair S Speaker 2: Please wait. Preferred and Non-Preferred Responses Certain types of adjacency pairs are characterized by a preference for a particular type of second party. For example, requests, reviews, and invites have preferred and non-preferred responses. Compare the following interactions where the first exchange has a preferred second party and the second exchange has a non-preferred party. Frank:

I really liked that movie last night. You managed?

Frank: Yes, it was very good. Frank:

I really enjoyed the film last night. You managed?

Frank: No, I didn't like it, but I can imagine that you liked parts of it.

The second preferred part also accepts an evaluation. Fiona:

I think Ralph is a very good writer.

Kieran: I think so too.

• 297

2 9 8 • Chapter 9 Speech acts and conversation

Fiona:

I think Ralph is a very good writer.

Kieran: Well, his drawings are interesting, but otherwise I don't think he writes well.

Non-preferred second parts are usually preceded by a pause and begin with a hesitant part like ok or huh. Preferred second parts tend to follow the first uninterruptedly and consist of structurally simple statements. Michelle: Do you want to have lunch with me tomorrow? Miguel:

To the right!

Michelle: Would you like to meet me for lunch tomorrow? Miguel:

Oh well. . . tomorrow is the 24th right? I told Lori I would have lunch with her tomorrow. And it's her birthday, so I can't cancel. How about Wednesday?

Furthermore, even non-preferred sequels often begin with a token approval or acceptance, or with an expression of thanks or apology, and characteristically include an explanation. Wade:

Can I use your telephone?

Frank: Oh, I'm sorry, but I have an important long-distance call. could you wait a bit

Try it yourself: In an apology, a preferred second party is an acceptance, while a non-preferred second party is a denial. For each of these speech acts, name a second preferred and a non-preferred part: request for information; invitation to a party; Regards; Accusation; congratulatory offer; Evaluation.

Opening Sequences Conversations are opened in a socially acceptable manner. Before the first conversation of the day, the interlocutors often shake hands, like when two office workers meet in the morning. Jeff:

Good morning Stan!

You are: Hello. How are you'? Jeff:

Oh I can't complain I guess. Ready for this afternoon's meeting?

Stan: Well, I don't have much of a choice!

Greetings illustrate opening sequences, phrases that make it easy for people to engage in conversation. They convey the message "I want to talk to you". Greetings are usually reserved for acquaintances who have not seen each other for a long time, or as opening sequences for lengthy conversations between strangers. Some situations do not require a greeting, for example when a stranger comes up to you on the street and asks you what the time is: Excuse me sir, do you know what time it is? The expression Excuse me, sir serves as a context-appropriate opening sequence. So greetings aren't the only type of opening sequence.

Conversational Organization Few conversations don't begin with some sort of opening sequence, even as often as this one: Eric:

He assumes.

Juan:

Was?

Erich:

I broke a tooth.

Speakers also use opening sequences to announce that they are about to invade their interlocutor's personal space. Here, two friends converse with a stranger on a park bench; during a pause in the conversation, the stranger intervenes: stranger:

Sorry, didn't want to interrupt, but I couldn't help but hear you were talking about Dayton, Ohio. I'm from Dayton. [The conversation between the three people continues.]

Unsurprisingly, in situations like this, the opening sequences take the form of an apology. Finally, the opening sequences can serve as a representation of one's own voice, so that the caller can recognize who is speaking, especially at the beginning of a telephone conversation. The phone just rang in Alfred's apartment. alfred: hello? Helena:

Hallo!

Alfredo: Hello Helena! How are you'?

In the second round, Helen reveals her voice for Alfred to recognize. In the third round, Alfred shows his acknowledgment while giving the second part of the greeting neighborhood pair started in the previous round. Opening Sequences in Other Cultures In many cultures, the appropriate opening sequence for a situation in which two people find themselves after not seeing each other for a while is a question about the health of the person, as in the American greeting How are you? These queries are essentially stereotyped and not meant to be literal. In fact, most speakers respond with a conventional optimistic formula (I'm fine or fine, thanks) even when they're feeling down. In other cultures, the traditional greeting may take a different form. Traditionally, Mandarin Chinese ask N I chi guo fàn le ma? Have you ever eaten rice? When two people meet on a street in Tonga, they ask Ko hoʔo ʔalu ki fe? 'Where are you going?' These greetings are phrased like How are you? In formal contexts or when there are differences in social status between participants, many cultures require a long and stereotypical opening sequence. When a person visits a village in Fiji, a very ceremonial introduction is made before any other interaction takes place. This event is about speeches that are subject to a complex set of rules about what needs to be said, when and by whom. This ceremony serves the same purpose as the opening sequences in other cultures. Functions of Opening Sequences A final aspect of opening sequences in which cultural differences can be found is the relative importance of their various functions. In telephone conversations in the United States, opening sequences serve primarily to identify speakers and to attract the caller's attention. In France, the opening sequences of telephone conversations often include an apology for invading someone's privacy.

• 299

3 0 0 • Chapter 9 Speech acts and conversation

named person:

Hallo?

Person calling:

Hi? Sorry to bother you. May I speak to MarieFrance? ("Hello? Sorry to bother you. May I speak to MarieFrance?")

In an American telephone conversation, this opening sequence is unusual. Therefore, the role played by the opening sequence in a phone call is different in two relatively similar cultures. As a result, the French may find Americans pushy and rude on the phone, while Americans are stunned by French apologies, which they find unhelpful and overly ceremonial.

Closing sequences Conversations must also end correctly. A conversation can only end when the participants have said everything they wanted to say. Also, a conversation should end before participants feel uncomfortable with having nothing more to say. As a result, speakers are carefully negotiating the timing of the closure, trying to appear as if they don't want to rush or delay anything. These objectives are reflected in the characteristics of the final sequence. First, a closing sequence contains a conclusion to the last topic of conversation. In the conclusions, the speakers usually agree on a later date or express the hope for such a meeting. These arrangements can be real, as in the first example, or stereotyped, as in the second. Carlos: Dana:

It's ok, nice to see you again. I think you're at Kathy's party tonight. Yes, see you there.

Elizabeth: See you later! Faruk: See you later!

The first step in an ending sequence helps ensure that no one else has anything to say. This is done by simply acting short moves like ok or ok. Typically, these pre-closing sequences are accompanied by a series of pauses between and within turns that slow the exchange and prepare for the end of the interaction. In the example below, Dana takes the opportunity to bring up a final topic, after which Carl initiates another closing sequence. Carl: Dana: Carl: Dana: Carl: Dana: Carl: Dana: Carl: Dana:

It's ok, nice to see you again. I think you're at Kathy's party tonight. Yes, see you there. IT'S OKAY. I heard there will be a lot of people there. Apparently he invited half the town. It should be fun. Yes thats OK. IT'S OKAY. i see you there Later!

The Organization of the Conversation Sometimes, after a concluding conversation, the speakers refer to the original motivation of the conversation. In a courtesy call to inquire about a person's medical condition, the caller will sometimes refer to this fact after the preliminary interview. Person calling:

Well, I just wanted to see how you're doing after your surgery.

named person:

That was very kind of you.

If the purpose of a conversation was to seek a favor, this brief exchange could take place: Alex:

Well, listen, I really appreciate you doing this for me.

To fill:

Forget. I like to help.

Finally, the conversations end with a parting word: goodbye, see you later, see you later. One amazing thing about fasteners is their deceptive simplicity. In fact, they are complex. Participants are very careful not to give the impression that they want to leave or stay and try to ensure that everything on each participant's unwritten agenda has been touched upon. As informal and abbreviated as they are, the closing sequences are bustling with activity.

Conversational Routines Both opening and closing are more routine than the central parts of conversations. The main parts are relatively less predictable; Although people are trained from childhood not to ask certain types of questions, they are also trained to start and end conversations properly. Because of the routine nature of opens and closes, conversations can start quickly and, just as important, end quickly.

Repairs A repair occurs in a conversation when a participant feels the need to correct himself or another speaker, to correct a previous utterance, or simply to repeat something, as in the following examples where a hyphen indicates an abrupt break. 1st speaker: 2nd speaker:

I was going to Mary's house, uh, Sue's house. And I went to the doctor to get a new... er... new, whatever it's called, a new prescription because I ran out of the old one.

3. Alex: Catalina:

Aren't these daffodils beautiful? They're beautiful, but they're daffodils.

4. Winston: David: Winston:

Todd came to visit us during spring break. What? I said Todd was here for spring break.

For 2, the cause of the problem is that the speaker cannot find a word. At 4, David initiates a redress for not hearing or understanding Winston's testimony. Therefore, speakers make up for a variety of reasons. Initiating a repair indicates that someone has misunderstood or misunderstood a term, that information is incorrect, or that someone is having trouble finding a solution.

• 301

3 0 2 • K apitel 9 Sprechhandlungen und Konversation

Word. To solve a fix, someone must repeat the misunderstood or misunderstood phrase, correct the inaccurate information, or specify the word. To initiate a repair we can ask a question, as in 4; Repeat part of the repair instruction, as in example 5 below; stop speaking abruptly, as in Example 6; or use particles and phrases like uh i mean i mean like in example 1. 5. Narrator:

I'm sure... I'm absolutely sure it was him that I saw hanging around last night.

6. Nelson:

And this means - [pause] The carburetor? Yes, right, the carburetor.

Johannes: Nelson:

Repairs can be initiated and resolved by the person who spoke the words being repaired, or by another speaker. Therefore, there are four possibilities: self-initiated and self-healing repairs; third-party repairs and self-repairs; own and third-party repairs; and repairs initiated and repaired by others. Of these possibilities, speakers show a strong preference for self-initiated self-repairs that are less detrimental to conversation and social relationships between speakers. In general, speakers wait for obvious signs of communicative distress before correcting another person's utterance. The least preferred pattern is for repairs started and repaired by others. Individuals who have the habit of starting and repairing utterances for others are referred to as bad speakers or know-it-alls. These preference patterns, found in many cultures, reflect a general but unspoken rule that all participants in a face-to-face conversation have the opportunity to say for themselves what they want. Speakers only assist others in initiating and resolving repairs when no other options are available.

Politeness: An organizational force in conversation Violating the principles of rotation by interrupting or not taking turns is considered impolite. Turning away from the interlocutor at the end of a conversation without going through a closing sequence is also stigmatized in polite conventions. Other aspects of politeness are more subtle but still play an important role in structuring the conversation. There are secret ways we instill respect for independence and participation. If we expect speakers to allow us to initiate and resolve a repair ourselves, we expect them to respect our right to contribute to the conversation without third-party interference; That said, we ask that you respect our independence. We also recognize the need for independence from another person when, rather than abruptly ending a conversation, we initiate an exchange before the conclusion and give our interlocutors an opportunity to say something more before the end. Conversely, when we begin a conversation with a greeting, we express concern for the health and well-being of our interlocutor, thereby acknowledging the other's need for participation. Many of the principles of conversational architecture can be explained with politeness and recognizing the politeness needs of others.

Intercultural communication

• 303

Intercultural communication When people from different cultures have different norms about the type of education required in a particular context, problems can easily arise. We describe how callers in France start phone calls with an apology; Such apologies are rarely part of the opening sequence of an American telephone conversation. Of course, people in both cultures view telephone conversations differently: Americans generally view the act of calling as a polite sign of participation, while the French view it more as a potential intrusion. Because of this variability, people from different cultures often misinterpret each other's signals. In Athabascan Indian conversation, a pause of even about a second and a half does not necessarily indicate the end of a curve, and Athabascans often pause within a curve for the same length of time. In contrast, most European Americans consider a pause of more than a second sufficient to signal the end of a turn (although there can be social differences). When Athabascan Indians and Europeans interact, the latter often misinterpret Athabascan mid-curve pauses as end-of-curve signals and feel free to claim the word. From the Athabascan perspective, the assertion of the word by European Americans represents a disruption on this point. When the same situation arises repeatedly in the interactions between the two groups, negative stereotypes emerge. Athabascans view European Americans as rude, aggressive, and uncontrollably talkative, while European Americans view Athabascans as uncooperative, short-tempered, and unable to maintain coherent conversation. Unknowingly, European-American teachers bring these stereotypes into the classroom and may judge Athabaskan students as distant or unintelligent because the teachers' unspoken cultural expectations are for students to speak, interact, and respond quickly. While these are typically the actions of children in mainstream European-American culture, Athabaskan children, who respect the norms of their own culture, tend not to behave this way. Although most people are unaware of these subtle cross-cultural differences, they can have profound social consequences.

Computers, Speech Acts, and Conversation As we saw in the previous chapter of this section, pragmatics in computational linguistics and corpus research is not yet fully explored. The modeling of politeness, variety, and other phenomena discussed in this chapter has yet to be modeled before many of the applications of computer technology to language can be mastered. The construction of written language corpora has been faster than the assembly of spoken language corpora, and the reasons are obvious. In recent years in particular, machine-readable texts originally published as books, magazines, and newspapers have become widespread. Also, scanners can effectively transform many shapes

AND

Materials from the past in machine-readable text. Another thing is creating electronic representations of transcribed spoken language. First of all, it has to be recorded, for example in audio or video. It then needs to be transcribed, a challenging and expensive task that depends in part on the quality of the recording and the level of ambient noise in the original environment. Nonetheless, one of the first machine-readable corpora was a transcribed version of spoken English. Referred to as the London/Lund Corpus, it formed a basis for extensive research. A significant part of the British National Corpus has more recently been based on language. Around 100 volunteers were deployed across the UK to transport tapes.

3 0 4 • Chapter 9 Speech acts and conversation

Computers, speech acts and conversation recorders as part of ordinary multi-day activities, with a notebook noting the conditions of recorded conversations and other exchanges, e.g. B. Participants and their relationships to each other, the physical environment of the recorded speech, etc. The recordings were then transcribed in standard English spelling. in the

The transcripts are currently being used to examine the character of the conversation. We have reported the findings of the British National Corpus in earlier chapters of this book (see, for example, the conversation category in Figures 8-1 and 8-2 on page 269) and will report further findings in later chapters. ■

Summary • Affirmations do things like confirm, promise, ask and greet. Actions performed through language are called speech acts. • That the language commonly used to take action is made clearer by statements such as “You're fired” or “Case dismissed!”! All speech acts, whether uttered or not, can be described in terms of four adequacy conditions that identify aspects or prerequisites for a successful speech act: the content, the preparatory condition, the sincerity condition, and the essential condition. • Under most normal circumstances, language users are bound by an implicit agreement that they abide by and expect others to abide by. This “principle of cooperation” consists of four maxims: quantity, quality, relevance and manner. • Occasionally a speaker disregards a maxim to signal that the literal interpretation of the utterance is not the intended one. • To encode and decode the intended meaning of indirect speech acts, people use conversational implicature patterns based on knowledge of their language, their society, and the world around them. • Indirect speech acts convey more than one message and are often used out of politeness or humor. • Respecting other people's privacy shows independence, while showing interest and sympathy shows courtesy of participation. • A lecture event is a social activity where language plays an important role. • Speech events are structured and the corresponding verbal and non-verbal behavioral characteristics of given speech events can be systematically described. • Conversations are organized according to certain normative principles. • Rounds are governed by a set of rules. • Neighboring couples are structured by local organizing principles, and many have preferred and non-preferred partners. • Organizational principles shape the opening and closing of discussions. • The repair organization can be described as a set of rules that rank different repair patterns in terms of preference. Self-initiated and self-performing repairs are preferred.

Exercises • At the root of many organizing principles in conversation is the need to show courtesy in independence and courtesy in participation with others. • Culture-specific norms determine when and where courtesy of independence and courtesy of participation are appropriate. • Since the organization of polite conversational behavior differs from culture to culture, misunderstandings between cultures often arise.

WHAT DO YOU THINK ? REVISED • Question from Isabella. Every speech act is surrounded by conditions of appropriateness. Among those associated with the act of two people declaring marriage through the utterance of certain words is the intention of the two people to get married at that moment. In the case of a play, the actors only want to portray a wedding ceremony, not a wedding. This means that an essential prerequisite for a declaration of marriage is missing and the declaration is ineffective as a declaration of marriage. • Kimberly's Complaint. To make a promise, you must use the word promise, as in "I promise to do the dishes if you make dinner." Without the word there is no promise. (Of course, if someone asks "Will you promise me?" and the answer is "yes," the answer would constitute a promise.) Explicit promises and intentions don't require the word promise, so Tyler may have meant well, but he did didn't make a promise, a promise. • Brandon's Complaint. Brandon's comment probably has less to do with Daniel as an individual than with his French social practices. The French consider calling someone curious and may apologize for the call and may take longer to get to the point than Americans expect. Americans, on the other hand, view calling a friend as engagement, generally positive, and requiring no apology.

Exercises based on English 9-1.

Make a list of the headlines on the first two pages of a newspaper. Indicate which of the headlines report physical acts and which report verbal acts.

9-2.

Look at a typical lecture from one of your courses and identify the characteristics that define it as a lecture (as opposed to a keynote speech, workshop, seminar, or lab meeting). Identify the distinctive features of the areas listed below. To what extent is there room for variability in the way a conference is conducted (depending on, for example, the personality of the participants)? When is a conference no longer a conference?

• 305

3 0 6 • Chapter 9 Speech acts and conversation

a) Setting (physical environment, clothing, social identity of participants etc.) b) Non-verbal behavior of participants (body movement, posture and position in relation to others etc.) c) Verbal behavior of participants (alternation, opening, closing, pairing of the Participants, etc.) d) Topic (what should be discussed? how far may one deviate? etc.) 9-3.

Record the first minute of a radio interview. Write down what was said in the first minute in as much detail as possible (for example, indicate who is speaking, when there are pauses, and any delays). Label each phrase according to its illocutionary power (greeting, question, praise, etc.). Then describe in detail the strategies used in opening the radio interview. Illustrate your description with specific examples from your transcript.

9-4.

Capture the first minute of an evening radio or television news report. Write down what was said that minute in as much detail as possible. Then answer the following questions and cite specific figures from your transcript. a. What impact do radio or television presenters want to achieve first? B. How is this achieved? Describe at least two strategies using concrete illustrations. C. Suppose you played your recording to friends without identifying what was recorded. Exactly what features would help them recognize it as an evening news taping? In addition to the content, list three specific telltale traits. i.e. Which of the messages are physical action messages and which are speech act messages?

9-5.

Observe the following interaction between two people working at adjacent desks. Amy:

¿Zach?

Zach:

Sim?

Amy:

do you have a rule

Amy's first time is an opening sequence. What does this indicate and what does Zach's answer indicate? Why didn't Amy just open with Do you have a rule? 9-6

Next time you're on the phone with a friend, pay attention to the ins and outs of the call and take notes as soon as you hang up. Identify several ways in which a phone conversation differs from an in-person conversation. Try to recreate specific linguistic examples from your phone conversation to illustrate your points.

9-7.

Consider the following code snippets, each containing a fix. For each iteration, determine whether the repair was (a) self-initiated and repaired, (b) self-initiated and repaired, (c) self-initiated and repaired, or (d) initiated by others and self-repaired. self repaired, or (d) initiated by others and repaired by another. a. January: James:

What is the sales tax in this state? Five cents on the dollar.

Exercises Patricia: Five cents on the dollar? You mean six cents on the dollar. Jaime: B.A-N-A:

There's a party at the Rod tonight. I want to go?

Sam:

With Rod? Rod is out of town!

I mean Ricks.

C. Pedro: Franco:

And then he comes over and tells me he's quitting his accounting, er, economics degree. Yes, he told me that the next morning too.

d. Always:

Your dog has been ill for the last month and you cannot attend the wedding because you have to take care of it.

Alice:

Well, actually, your dog has been sick for at least two months. So it's nothing new.

me. Oneself:

Do you remember the names of all your children? The eldest is Daniel, the girl is Priscila, then there is another girl - what was her name again?

Director:

Susie I think.

Sam:

Yes, Susie, that's all.

F. Ellie:

How much do you charge for car insurance?

ted:

Two thousand dollars a year, but there is a three hundred dollar deductible. Three hundred or one hundred... I don't remember.

Elli:

It's probably a hundred, right?

ted:

yes i think you are right A hundred sounds good.

grams. Will be:

She's been cooking all day for this dinner.

In fact, he's been cooking for three days.

h will:

9-8.

Oh yes, six cents on the dollar.

There wasn't much he could do for her. He needed five thousand dollars to pay his tuition and I didn't have it.

David:

I thought it was four thousand.

Wille:

Yes, four thousand, but still not that much.

Consider the following snippets, all of the previous structure for starting a conversation. Describe in detail the structure and function of each pre-structure using the terms branch (or junction), signal, adjacency pair, first part, second part and floor claim. a. Larry:

He assumes.

Lauren: Or what? Larry: b. Tom:

Pat is coming tomorrow. [reads newspaper] I can't believe it!

Fred:

Was?

Tomás:

Congress passed another new immigration law.

C. Varilla:

[laughs while reading a book]

What are you laughing about?

Ruta:

This story is so awesome!

• 307

3 0 8 • Chapter 9 Speech acts and conversation

9-9.

Consider the following excerpt from a conversation between three friends. 1) Cindy: Have you heard from Jill lately? He hasn't written or called in years. 2) Larry:

Yes, he sent me a postcard from England.

3) Splitter:

From England?

4) Run:

Ah, maybe he was from France, I don't remember.

5) Cindy: What is he doing... 6) Barb:

No, I know she must have come from France because she would stay there all year.

7) Cindy: What is she doing in France? 8) Larry:

Why are you asking about her?

9) Cindy: I don't know, I was just thinking about her. 10) Larry:

He's on some kind of exchange program. Learn French or something.

11) Cindy: Sounds pretty good to me. 12) Larry:

Yes, I don't know. She said she was fed up with Europe and wanted to go home.

a. How many turns does each speaker in the above conversation have? B. Identify one example of each from the conversation above: Blinker, Challenged Word, Preferred Response, Rejected Response, Repair, Cause of Problem, Introduction, and Solution. C. Identify an adjacent pair in the conversation by naming the first and second participants. 9-10 Conversations in novels and dramas, as well as those reenacted in films or on stage, are often different from ordinary everyday conversations. The following is an excerpt from a conversation in Isak Dinesen's autobiographical novel Out of Africa (New York: Random House, 1937). "Do you know anything about accounting?" Asked. "No. Absolutely nothing," he said, "it was always very difficult for me to add two numbers." "Do you know anything about cattle?" I continued "Cows?" I ask. "No no. I'm scared of cows.” “Can you drive a tractor?” I asked. A faint glimmer of hope appeared on his face. 'No,' he said, 'but I think I can learn that.' 'Not on my tractor,' I said, 'but then tell me, Emmanuelson, what did you do? Emmanuelson straightened up. "What am I?" he exclaimed "Why, I'm an actor." I thought, thank God, it's beyond my ability to help this lost man in any practical way; general human conversation. “Are you an actor?” I said, “That's good. And what were your favorite roles when you were on stage?

Exercises "Oh, I'm a tragic actor," said Emmanuelson, "my favorite roles were Armand in 'La Dame aux Camelias' and Oswald in 'Ghosts.' Using this example, analyze the differences between the organization of written conversations and the organization of real conversations. Why are there these differences? 9-11 The excerpt of conversation transcribed below contains two adjacency pairs. Provide the change letter or letter(s) for each of these categories: (1) insertion sequence, (2) first part of first adjacency pair, (3) second part of first adjacency pair, (4) first part of second adjacency pair, (5) second part of the second adjacency pair. He then relates each of these speech acts with an exemplary phrase: (6) clarification, (7) refusal, (8) suggestion, (9) request for clarification. A. Eric: Do you want to watch Civil War tonight? B. Nan: The Ken Burns series? C. Eric: He did it with his brother. D.Nan: I'm sorry. I have an audit tomorrow.

Specially for educators and future teachers 9-12. Below are two sets of conversational rounds among college students that are mostly spoken within a few months of graduating from high school. In statements (1) to (4), the underlined word as is used in at least two different ways. Analyze (1)–(4) and characterize the two forms. In statements (5) to (7), the word all is highlighted. Characterize your role in these changes. Next, imagine having a discussion with your students where you and they review a transcript. One student reports that another of his teachers makes fun of using likes and another student says that his father also mocks him. In your discussion with them, what points would you make to indicate that these relatively new uses of like and everything work in a similar way to more traditional conversational expressions? With what words to like and compare everything? Discuss why these new uses are sometimes mocked. 1) Adam: I don't want to break up with her like that. . . this time. Brent: Yeah, don't break up this time. Break up like Thanksgiving or something. 2) Ben: So he called and said I can't believe you did that, I can't believe you did that. I'm like—, 3) Ali: I was like, I was like why, why you know. . .

• 309

3 1 0 • Chapter 9 Speech acts and conversation

4) Rod: I was so happy and stuff. 5) Jose: Wait, why is he even complaining about you? 6) Jaime: I'm sitting here trying to read. 7) Danny: I was happy and stuff. 9-13. Your students are likely familiar with a version of the old adage that goes, "Sticks and stones may break my bones, but words will never hurt me." At a level appropriate for your students, create a lesson plan that relates this adage investigates the power of speech acts. You could begin the lesson by asking if any of your students have ever felt hurt by what others have said to or about them. 9-14 Invite students from different cultural backgrounds to share their experiences of making and receiving calls from members of other cultural groups. Also, discuss your attitudes towards telemarketing calls and how your attitudes reflect cultural values.

Additional Sources • John J. Gumperz, TC Jupp, and C. Roberts. 1979. Crosstalk: A Study of CrossCultural Communication (London: National Center for Industrial Language Training and BBC) An hour-long video illustrating and analyzing communication breakdowns between East Indian immigrants and bankers, librarians and other institutional figures in London; a moving demonstration of the painful difficulties that can arise from differing conversational norms across cultural boundaries.

Suggested Reading • Diane Blakemore. 1992. Understanding Expressions: An Introduction to Pragmatics (Oxford: Blackwell). A basic introduction; a natural continuation of the content of this chapter. •Peter Grundy. 2000. Making Pragmatics, 2nd ed. (New York: Oxford University Press/London: Arnold). A clear and basic introduction in an effective conversational and interactive style, containing a chapter on project work in pragmatics and others on politeness, speech acts and deixis (the latter we covered in a previous chapter); mainly British examples. • Jacob L. May. 2001. Pragmatics: An Introduction, 2nd ed. (Oxford: Blackwell). A comprehensive and more advanced treatment that includes chapters on speech acts, cross-cultural pragmatics and conversation analysis, and literary pragmatics. • Deborah Tannen. 1990. They Just Don't Get It: Women and Men in Conversation (New York: Ballantine). Accessible and hugely popular bestseller exploring the misunderstandings between the sexes; it also treats Grice's maxims very simply.

References • Deborah Tannen. 1994. Gender and Discourse (New York: Oxford University Press). An accessible treatment of Tannen's You Just Don't Understanding background. • Ronald Wardhaugh. 1985. How Conversation Works (New York: Blackwell). A well focused, essential and accessible book.

Further Reading The analysis of speech acts has been primarily an enterprise of philosophers. Austin (2005) is a series of 12 readable lectures outlining the nature of locutionary and illocutionary (as well as perlocutionary) actions. Grice (1975, 1989) formulates the cooperative principle and lists the maxims of conversation discussed. Searle (1976) discusses the classification of speech acts and their syntax, while Searle (1975) discusses the structure of indirect speech acts. Alongside these primary sources, Levinson (1983) remains invaluable. Inductive studies by conversation analysts, which differ fundamentally from philosophical traditions in their methodological approach, are difficult to read: Sacks et al. Turn-taking systematically analyzed for the first time. (1974), closures by Schegloff and Sacks (1973) and repairs by Schegloff et al. (1977). More accessible to student readers are these books on the analysis of conversation and the use of language in informal contexts: Levinson (1983) and Wardhaugh (2006), Chapters 11 and 12. The theoretical basis for the study of language events is presented in Goffman (1986 ; original edition 1974) and Hymes (1974). Goffman (1981) presents engaging and entertaining analyzes of various speaking events, including lectures and radio broadcasts. Goodwin (1981) describes how language and gestures are integrated into conversation. The organization of conversation in the workplace is examined in Boden (1988). The characterization of communication between subordinates and superiors as spectator/performer or performer/spectator was proposed by Bateson (2000), a re-edition of a classic text that lays the philosophical foundations for the study of human communication. Intersocial and intercultural differences in the organization of conversation are discussed in Gumperz (1982a, 1982b), BlumKulka et al. (1989), Trosborg (1995) and Scollon and Scollon (1981, 1995), some examples of which appear in this chapter. Godard (1977) is an interesting study of French-American differences in telephone behavior, and this and other aspects of French interaction are discussed in Chapter 10 of Ager (1990). Brown and Levinson (1987) and the chapters by Levinson (1983) and Wardhaugh (2006) deal with politeness. Drew and Heritage (1993) is a collection of essays discussing interactions in institutional contexts.

References • Ager, Dennis E. 1990. Sociolinguistics and Contemporary French (Cambridge: Cambridge University Press). • Austin, John. 2005. How to Do Things with Words, 2nd ed. (Cambridge: Harvard University Press). • Bateson, Gregory. 2000. Steps Towards an Ecology of Mind: Collected Essays in Anthropology, Psychiatry, Evolution, and Epistemology (Chicago: University of Chicago Press). • Blum-Kulka, Shoshana, Juliane House, and Gabriele Kasper, eds. 1989. Cross-Cultural Pragmatics: Excuses and Excuses (Norwood, NJ: Ablex).

• 311

3 1 2 • Chapter 9 Speech acts and conversation

• Boden, Deirdre. 1988. The Business of Talk: Organizations in Action (Cambridge: Polity). • Brown, Penelope, and Stephen C. Levinson. 1987. Courtesy: Some Universals in Linguistic Use (Cambridge: Cambridge University Press). • Drew, Paul and John Heritage, eds. 1993. Lecture at Work (Cambridge: Cambridge University Press). • Godard, Daniele. 1977. “Same Environment, Different Rules: Initiating Telephone Calls in France and the United States,” Language in Society 6:209–19. • Goffmann, Erving. 1986. Structural Analysis: An Essay on the Organization of Experience [Repr. ed., with a foreword by Bennett Berger] (Boston: Northeastern University Press). • Goffmann, Erving. 1981. Conversational Forms (Philadelphia: University of Pennsylvania Press). • Goodwin, Carlos. 1981. Conversational Organization: Speaker-Hearer Interaction (New York: Academic). • Grice, H.Paul. 1975. "Logic and Conversation," in Peter Cole and Jerry L. Morgan, eds., Syntax and Semantics 3: Speech Acts (New York: Academic), pp. 41–58. • Gumperz, John J. 1982a. Discourse Strategies (Cambridge: Cambridge University Press). • Gumperz, John J., ed. 1982b. Language and Social Identity (Cambridge: Cambridge University Press). • Hymes, Dell. 1974. Foundations of Sociolinguistics (Philadelphia: University of Pennsylvania Press). • Levinson, Stephen C. 1983. Pragmatics (Cambridge: Cambridge University Press). • Sacks, Harvey, Emanuel A. Schegloff, and Gail Jefferson. 1974. "A Simpler Systematic for Organizing Turn-Taking in Conversation", Language 50: 696–735. • Schegloff, Emanuel A., Gail Jefferson, and Harvey Sacks. 1977. “The preference for self-correction in organizing repair in conversation,” Language 53: 361–82. • Schegloff, Emanuel A. and Harvey Sacks. 1973. "Opening Closures", Semiotics 7:289-327. • Scollon, Ron and Suzanne B.K. scollon 1981. Narrative, Literacy, and Face in Interethnic Communication (Norwood, NJ: Ablex). • Scollon, Ron and Suzanne Wong Scollon. 1995. Intercultural Communication (Oxford: Blackwell). • Searle, John R. 1975. “Indirect Speech Acts”, in Peter Cole and Jerry L. Morgan, eds. Syntax and Semantics 3: Speech Acts (New York: Academic), pp. 59–82. • Searle, John R. 1976. “A Classification of Illusory Actions,” Language in Society 5:1–23. • Trosborg, Anna. 1995. Interlanguage Pragmatics: Requests, Complaints, and Apologies (Berlin: Mouton de Gruyter). • Wardhaugh, Ronald. 2006. Introduction to Sociolinguistics, 5th ed. (Malden, MA: Blackwell).

Linguistic variation in situations of use: registers and styles

WHAT DO YOU THINK ? • Elementary school student Stefanie asks why her teachers don't like colloquialisms and slang. He also wonders what the difference between them is. what can you tell her • His classmate Michael comments that he is surprised that this book uses contractions and asks if contractions are used in textbooks, are not, and should be avoided. They find they create a casual, relaxed tone and are appropriate here. What justification can you offer for your preference? • Davin, an English major, comments that the dialogue of a P.D. James you are reading is "stunning and totally natural". He scoffs because he recently corrected a transcript of a statement he made in connection with a car accident, and his transcribed answers to questions sounded muddled with false starts and um and um. Their answers were nothing like fictional dialogue. And neither do the lawyer's questions! What do you tell Davin about what's totally natural in fictional dialogue? • Hunched over a cookbook, your Uncle James wails, “What English is that! Toast the pine nuts in a medium-sized pan. Remove and add 1 tbsp. oil and garlic. Cook for 4 minutes and drain off the remaining liquid. Salt and pepper the trout cave and fill with the spinach mixture. Brush the trout with the remaining oil. What happened to words like from, to, they, the and all the other Anglo-Saxon glues in the language? What you say?

313

3 1 4 • Chapter 1 0 Language variation in situations of use: registers and styles

Introduction You are familiar with the term dialects and know that they are variants of languages spoken by different social groups. In the United States, people recognize dialects called "Brooklynés" and "Bostonese" and sometimes speak of a "Southern accent" or a "Minnesota accent". Cockney is another well-known dialect. Dialects are the subject of Chapter 11. In this chapter we are concerned with language varieties that are characteristic of social situations rather than social groups. We talk about jargon and legal jargon and other linguistic varieties that are typical of certain situations. The linguistic varieties characteristic of particular social situations are called styles or registers. In different circumstances, everyone varies the forms of speech. For example, we might call some people Michelle or Michael; another dr Lavender or Mr. Olson; still others, Your Honor or Mr. President; to some we say sir or ma'am or miss. When using the address term friend, you certainly aren't using it indiscriminately for everyone you're in contact with. In some communities, different social situations require completely different languages; In other communities, different social situations require alternative variants of the same language.

Language varies within a discourse Community language choice in multilingual societies One might assume that in multilingual countries such as Switzerland, Belgium and India, different groups of people speak different languages. In most cases, however, each language is also systematically assigned to specific social situations. In multilingual language communities, the choice of language is not arbitrary. Rather, a particular environment, such as school or government, may favor one language while other languages are appropriate in other speaking situations. While there can be more or less equivalent expressions in two languages, the social meaning associated with the use of one language is often different from that associated with the use of the other. As a result, speakers must pay attention to the social significance of language choice, however unconscious that choice may be.

Language repertoires in Brussels, Tehran and Los Angeles The use of selected varieties of two languages among government officials in the Belgian capital illustrates the nature of language choice in a European community. Government officials in Brussels who are of Flemish origin do not always speak Dutch to each other, even if they all speak Dutch very well and equally well. Not only are there times when they speak French to each other instead of Dutch, but there are also times when they speak standard Dutch and times when they use one or another regional variant of Dutch with each other. In fact, some of them also use different variants of French among themselves, with one variant particularly associated with government officialdom, another corresponding to the non-technical conversational French of the highly educated and educated circles in Belgium, and yet another not just a "more colloquial French". more or

The language varies within a colloquial French-speaking community of Flemings. In short, these different variants of Dutch and French form the linguistic repertoire of certain social networks in Brussels. (Fishman [1972], pp. 47–48.) The variety of languages used by Brusselsians is due to the environment in which the conference takes place, the subject, the social relations between the participants and some other characteristics of the conference .Situation . . . In general, the use of Dutch is associated with informal and intimate interaction, while French has more official or "intellectual" connotations. Given these associations, the choice of French or Dutch carries an associated social meaning in addition to its referential meaning. We use the term linguistic repertoire for the set of linguistic varieties that are evident in the speech and writing patterns of a language community. As in Brussels, the linguistic repertoire of each language community can consist of several languages and include several varieties of each language. In the mid-1970s, there was considerable multilingualism in Tehran, the capital of Iran. Christian families spoke Armenian or Syriac at home and in church, Persian at school, all three in different situations when playing or shopping, and Azerbaijani Turkish in bazaar shops. Muslim men from northwest Iran who worked as laborers in the sprawling capital spoke a variety of Persian with their supervisors on construction sites, but switched to a variety of Turkish and a local Iranian dialect with their colleagues when visiting their homes. cities on vacation; they also listened to daily radio broadcasts in High Persian and heard excerpts from the Koran recited in Arabic. It was not uncommon for people from any social position to know up to four or five languages and use them in different situations. The situation is almost the same there today. Elsewhere, the Korean-speaking community in Los Angeles supports bilingual institutions of many kinds: banks, churches, businesses, and a wide range of services from pool halls and video rental stores to hotels, construction companies, and law firms. At some banks in the Koreatown area of Los Angeles, all of the cashiers are bilingual, frequently switching between Korean and English during the business day. Of course, when cashiers switch customers, they switch between Korean and English as needed.

Try it yourself: Think of situations in your community where people switch from one language to another during a conversation depending on who they are talking to, the topic or some other aspect of the social situation.

Changes in varieties within a language If we look at the situation in Europe, apart from the changes between languages, we also see examples of internal language changes. The people of Brussels switch not only between French and Dutch, but also between the varieties of French and between the varieties of Dutch. In Hemnes, a town in northern Norway, residents speak two very different varieties of Norwegian. Ranamål is a local dialect and is used to identify the speakers of this region. Bokmål, one of two forms of Standard Norwegian (the other is Nynorsk), is used in

• 315

3 1 6 • Chapter 1 0 Language variations in situations of use: registers and styles

Hemnes for education, religion, government transactions and the media. All members of the community control Ranamål and Bokmål and are always welcome to speak one or the other. There are differences in pronunciation, morphology, vocabulary, and syntax, and speakers do not mix the two variants in their speech. Here is an illustration of a simple phrase meaning "Where are you from?". ke du e ifra

(rob target)

Where do you come from?

(Bokmål)

While Bokmål is the expected variety in certain well-defined situations, the residents of Hemnes do not accept its use among themselves outside of those situations. In situations where Ranamål is commonly used, using Bokmål would indicate social distancing and even disregard for the community spirit. In Hemnes the usage of Bokmål with other places is snakkfint or snakk jalat 'to make an air'. it conveys the importance of authority, expertise, and courtesy to strangers who are clearly detached from their personal lives” (Blom and Gumperz [1972], pp. 433–434). Therefore, it is important to consider the social situation both when choosing varieties of the same language and when changing languages.

Language situations As we have seen in Hemnes, Los Angeles, Brussels and Tehran, language change can be triggered by a change in a number of situational factors, including the context and purpose of the communication, the person addressed, the social relationships between the interlocutors and the topic.

Elements of a speech situation If we define a speech situation as the union of essential situational factors such as purpose, subject, and social relationships, then any speech situation in a bilingual community generally admits of the use of only one of the two languages. . . Table 10-1 illustrates this concept for a bilingual community in Los Angeles.

TABLE 10-1 Linguistic repertoire

LAGE

A B C D E

LIST OF SPEAKERS

place, place

TOPIC TYPE

ESPAÑOL

intim intim intim intim intim

home school home school

Non-Academics Non-Academics Non-Academics Academics

ENGLISH

X X X

Oral situations As you can see, in situation A a Spanish variant is appropriate, but in situation C an English variant is appropriate. Only in the relatively rare case of Situation E can a person have a real choice between Spanish and English without drawing attention to the language chosen. In situation E, the choice is made possible by the conflict between privacy (which usually requires Spanish, as in situations A or B) and an academic subject (for which English is usually preferred). Table 10-2 shows some aspects of a speech situation that may require a change in language diversity.

TABLE 10-2 Elements of a Speech Situation PURPOSE

THE SETTINGS

ATTENDEES

purpose of the activity

Theme placement mode

Speaker Addressee Social roles of speaker and addressee Character of the audience

In terms of purpose, the type of activity matters, as does its goal. Shop, preach, tell a story? Are you entertaining, conveying information, reinforcing a social relationship? Say hello to a friend or invite an aunt to dinner? Activity may affect your choice of language. As for the environment, you can switch from one language to another if the topic changes from a topic of local interest, for example to one of national interest, or from a personal matter to a topic about your college or university. Location can also influence language choice, as you may use one language in an academic setting but another in a religious setting or at home in related situations. Your mood, whether you speak or write, can also influence the form of language you choose. In relation to the participants, the identity of the speaker will influence the choice of language as much as the identity of the addressee. Speakers usually adapt their facial expressions to the age of the person being spoken to. In some societies, the older the person, the higher the social position; young people should address their elders with more respect than their peers. In French, the second-person singular pronoun “you” has two forms: tu is used to address a social equal or as an expression of intimacy, while vous is reserved for a person of higher social status or to mark distance when addressing more than of a person, regardless of their status). A younger person addressing an older person may be expected to use vous rather than tu unless the older person is a close relative. Since tu is the grammatical singular form and vous is the plural, French shows one way in which morphology can vary depending on the recipient's age or social status. Persian also shares many of the same patterns as many other European languages. Not only the social identity of the speaker and the addressee is relevant, but also their role in the respective speech situation. A judge, for example, often speaks one variant at home where she is mother, wife, neighbor, and another as judge in the courtroom. A father who works as a teacher and has his son as a student can speak different varieties at home and at school, even if the subject and addressee are the same.

• 317

3 1 8 • Chapter 1 0 Language variation in situations of use: Registers and styles

The various aspects of the speech situation come together in a particular choice of linguistic diversity. In any situation, whether general, like at home or at church, or specific, like having a coffee chat with a close friend about politics, only one variation is usually appropriate. In fact, people become so accustomed to speaking a particular language in a particular environment that they may have difficulty communicating in another language in that environment, no matter how familiar the other language is in other environments. (Exceptions to this generalization are professional translators, bilingual educators, and certain business people who regularly interact with members of their own and other cultures.) As a result, switching between language variants is very common around the world and is known as code switching.

Records in Monolingual Societies The recognition that in multilingual societies there exist language configurations and situations in which one language or another is appropriate has a direct parallel in monolingual language communities, where varieties of a single language make up the entire linguistic repertoire. Consider the difference between the full forms of careful speech and the characteristic rapid speech abbreviations and reductions that occur in relaxed face-to-face communication: not just everyday contractions like will't and I'll, but abbreviated phrases like Jeetyet? [ditjεt] and Wajjasay? [wɑdəse] for 'Have you eaten yet?' and what did you say?' To cite another example, you know that you don't typically use the same terms for specific body parts when you're talking to friends and when you're talking to a doctor. You can use the clavicle at home and the clavicle with a doctor, while both can be used with friends depending on other aspects of the language situation. Decisions made for certain other body parts would be more noticeably different. The distribution of alternative terms for the same referent may seem arbitrary and without any communicative benefit. With body parts, for example, all terms can be known (and used) by all parts in equivalent situations. A doctor talking to his own doctor can use the clavicle, but he can use the clavicle to his family and friends. When addressing a doctor, non-medical people can use appropriate terms to discuss a medical situation. Since all terms would be understood equally well and could communicate referential meaning equally well, choosing a socially appropriate variant is cognitively useless. You may wonder why the linguistic expression differs in different speaking situations. The answer is that different forms of the same content can indicate its affective relationship to salient aspects of the situation (attitude, addressee, subject, etc.). It can be assumed that the centuries-long variation of a language satisfies a fundamental need of human interaction.

Try it yourself: name a conversational situation (e.g. a conference with a teacher, dinner with your grandparents or a job interview) and list six words that you would probably talk about in that situation, but for which the expression you would use to a close friend would be different from what you think is appropriate in the speaking situation. Next to each term that corresponds to the speaking situation, write the term that you would use for that speaker with a close friend.

Stylistic features Just as a multilingual language repertoire assigns different linguistic variants to different speech situations, this also applies to a monolingual repertoire. For all speakers, both monolingual and multilingual, there are distinct differences in the forms of language used for different activities, audiences, subjects, and attitudes. These forms make up the styles or registers of a linguistic repertoire. By choosing between varieties, situational variation is created and reflected. From an early age, everyone learns to master different language variants in order to use them in different speaking situations. No one is limited to a single variety in a single language. These language varieties can belong to one language or to several. Which speech situations (which purposes, settings, participants) lead to a different variety depends on social norms. In one society, the presence of in-laws may call for a different variety (as in Dyirbal and many other Australian Aboriginal societies). In other societies, the presence of children or members of the opposite sex can be decisive. In Western societies, adults have a number of words that they do not pronounce in the presence of children (and children try not to pronounce them in the presence of adults). There are also differences related to the mode of writing or speaking the language. You are familiar with the term “colloquial” as a term for informal speech.

Style markers Because languages differ in vocabulary, phonology, grammar, and semantics, styles or registers can differ at all levels. There can also be different patterns of interaction in different speech situations; For example, taking turns in conversation is different from taking turns in courtrooms and classrooms. In addition, social rules regulate non-verbal behavior, such as physical proximity, face-to-face position, standing up and sitting down, which are also associated with register variations. Both interaction patterns and body language are beyond the scope of this book and will only be mentioned in passing. If you find characteristics of a style at one level of grammar, you can expect to find corresponding characteristics at other levels as well. For example, describing legal language requires attention to its distinctive vocabulary, sentence structure, semantics, and even phonology.

Lexical style markers and registration records vary along specific social dimensions. For example, people often speak (and write) in markedly different ways in formal and informal situations. Formality and informality can be viewed as opposite poles of a situational continuum along which forms of expression can be organized. The four words pickled, drugged, intoxicated, and intoxicated can all mean the same thing, but you can rank them by their formality, and you would probably agree that in the order given, they could be ranked from least to most formal. In one context, the reference to intoxication might require the word drunk, while in another a more appropriate expression might be drunk or under the influence. Bombarded and annoying are terms used especially by the youngest in very informal situations. A thesaurus lists more than 125 expressions for "drunk". It goes without saying that they are not situationally equivalent and are not interchangeable.

• 319

3 2 0 • Chapter 1 0 Language variation in situations of use: Registers and styles

Not all words that can be described as "intoxicated" are appropriate for use on all occasions when an indication of intoxication is intended. The choice of words can indicate very different attitudes towards the state, the recipients, the person being described, etc. It can also indicate the speech situation in which the term is used, e.g. B. intimate or distant, formal or informal, serious or humorous. Different expressions for "intoxication" have different connotations depending on the usage situations they are associated with. These associated situations of use add another dimension of meaning than referential meaning. Imagine the following dialogue between a judge and a defendant in a court case: Judge:

I see the police say you got drunk last night and drove an old car in the middle of the street. This right?

Defendant: Your Honor, if I may address this baseless allegation, I would like to inform you that I was not drunk or under the influence of any alcoholic beverage; Just for the record, I didn't drink anything last night.

First, the judge's language seems out of whack: words like cop, wasted, and clunker seem inappropriate, even bizarre, for a judge in a courtroom. The defendant's response also seems out of place, especially after the judge's extremely informal speech. Even if the judge had used high-level language, the defendant's language might appear too formal. It also seems strange that the defendant uses the colloquial word alcohol in a sentence that includes the slang words drunk, intoxicated, drinking and pleading. Compare the judges' language above with the following, which is more appropriate to the speaking situation. Judge: You are accused of driving a blue 1992 Ford under the influence of alcohol. How do you declare?

You can see that even within the same language, registers or styles are chosen for specific usage situations. Salutation The salutation for the same person can vary from situation to situation. The Queen of England is referred to as Her Majesty (or Lady), although her husband is believed to use a more intimate term when speaking to her in private. In court, judges are referred to as Your Honor or Judge, although their friends and neighbors may call them Judy or Lance. He addresses each of us differently depending on the situation: by the first name (Pat); Surname (Smith); surname with title prefix (Dr. Smith, Mrs. Jones); the second person pronoun (you); terms showing respect (sir, ma'am); and various informal generic terms (dude, dude). At the other end of the scale are disrespectful terms like asshole or bastard. Slang The register used in situations of extreme informality is called slang. Jargon is probably the record whose name is best known, and it is often said that the use of jargon can indicate rebellious undertones or a conscious distancing of its users from certain dominant values. Therefore, slang is particularly popular among teenagers and students. However, its use is by no means limited to these groups, as the slang originates from specialized groups of all kinds, from doctors and hackers to cops and stockbrokers.

Style markers Some jargon changes as quickly as clothing trends. Still, slang dictionaries exist, and their existence suggests that some slang expressions have a longer lifespan.

College Slang: Top 20 in Slang and Sociability, Connie Eble reports on the slang most commonly used by students at the University of North Carolina between 1972 and 1993. Which ones have you used or heard of? Cute Chill Out Slide Hit Bag Killer Jam Reach Waste Clueless Diss Pig Evil Crash Cheesy Hook (Up) Trip (Out) Stupid Buzz/Catch a Buzz tool

'excellent, excellent' 'relax' 'simple course' 'slop, no show' 'slop, no show' 'excellent, exciting' 'make music, dance, party' 'find a partner for sex or romance' 'drunk' 'unconscious ''lay down, criticize' 'greedily eaten' 'good, excellent' 'go to bed' 'unattractive, boring' 'find a partner for sex or romance' 'have a strange experience' 'socially inept person' 'experience a mild intoxication' totally acceptable”

These examples of Gíria dictionary overlay are illustrative: Wow, Sinos e Assobios, Cubra sua Bunda, Droga de Designer, Idiota, Emoticon, Chute Ass, Mallie, Netiquette, Pocket Pool, Puzzle Palace, Spam, Boobs and Pimples and what should always be! Slang has a legitimate place in the linguistic repertoire of language communities. As with any record, its effectiveness depends fundamentally on the circumstances in which it is used. In an appropriate situation, anyone of any age and social position can use slang. Just as casual attire can extend your greeting from informal circumstances to slightly more formal circumstances, slang generally moves up the social ladder and becomes acceptable in more formal circumstances. The words mob and pants are among many that were colloquial earlier in its history but can now be used in circumstances that are not very informal. As words become established in more formal circumstances, they lose their slang status and are replaced by new slang. (Although such climbing of the social ladder is common, some slang expressions persist in informal circumstances. Chaucer used Bones, meaning 'given', in the 14th century and remains slang today.) specialized interests when involved in activities related to those interests, including talking about them, is called slang.

• 321

3 2 2 • Chapter 1 0 Language variation in situations of use: registers and styles

Slang is the specialized vocabulary associated with professions like medicine, finance, and engineering, and activities like sports, music, and computers. Unlike slang, slang is not limited to situations of extreme informality and does not carry rebellious connotations. Slang is another term associated with "professional" language or activities, but slang is more indicative of the language of clandestine or criminal activity. Because slang consists of terms that are used in a different sense elsewhere, or perhaps used nowhere else, conversations, including newspaper and magazine articles dealing with specific topics that indicate slang, can be confusing to the uninitiated. Here is a quote from a newspaper report about a baseball game. If you're familiar with the sport, the meaning of the sentence will be clear, but if not, you probably won't understand it, despite its simple grammatical structure: "The drive has the bases loaded in the lower part of the sixth in two bases." on balls and one penalty, but Suppan beat José Valentín and sent Chávez into midfield to end the inning. Hit, Run, Out, Home Run, Left Fielder, Baseman, Starter, Close, Substitute, Runner, Reliever, Warning Track and Pocket, as well as the Verbs Tie, Double, Pitched, Single and Home Run. For baseball fans jargon is easy to understand, for others it can be completely obscure.

Try it yourself: In the illustrative sentence from the previous paragraph, identify six additional nouns and three additional verbs with specialized baseball-specific meanings. You should be able to identify baseball slang based on your familiarity with the sport or because your unfamiliarity makes the slang opaque.

Phonological style markers and registration registers are marked not only for word choice, but also for grammatical features, morphology, and other levels of grammar. For spoken registers, this includes phonology. In a study of speech patterns in New York City, which we will discuss in detail in the next chapter, considerable phonological variation was discovered among all groups of speakers in different usage situations. Figure 10-1 shows frequencies for pronouncing -ing as /ŋ/ in three language situations. We use -ing to represent the pronunciation of the suffix in words like talk, run, eat, and look. The speaking situations in this case consist of three types of interactions in the course of a sociolinguistic interview in the homes of four groups of respondents (designated LC, WC, LMC and UMC). The style of the interview, with its intertwined questions and answers, can be described as "cautious" discourse. Respondents read a defined passage aloud, considering the "reading" style of speaking more carefully than the interview style. Towards the end of the interview, to encourage relaxed language, the interviewer would ask respondents if they had ever been close to death, and this tactic usually resulted in loose and sloppy variation, referred to here as "casual" language. In their informal speech, respondents from LC (LC is an acronym for lower class, a socioeconomic classification based on a combination of income, education, and type of employment) uttered the suffix -ing as /ŋ/ 20% of the time ( die other 80% than /n/). In his cautious speech, /ŋ/ occurrences increased to 47% (while /n/ decreased to 53%).

Style markers FIGURE 10-1 Percentage of pronouncing -ing as /IN/ in three language situations in four social groups in New York City 99

96 100

89 79

78 69 47

Bad

LMC

UMC

Social Groups: Lower Class, Working Class, Lower Middle Class, Informal Upper Middle Class

Careful

file

Fonte two cubes: Labov 1966.

When reading a passage aloud, LC respondents pronounced /ŋ/ 78% of the time (and /n/ only 22%). This represents a dramatic increase in the pronunciation of /ŋ/ as the speech situation becomes more formal. Exactly the same general pattern applies to the other three social groups. Each of them uses more pronunciations of /ŋ/ in careful speech than in casual speech, and more in reading style than in careful speech. We can generalize this finding by saying that in this language community /ŋ/ indicates formality and more frequent pronunciations of /ŋ/ indicate more formality. In another study, college students in Los Angeles collected data showing that both men and women used the pronunciation /ŋ/ more often in discussions than in jokes. Again, we can think of arguing as a less relaxed or more cautious registration than joking. The frequencies are given in Figure 10-2. Although males and females use this phonological variable differently (a topic we will return to in Chapter 11), both sexes use it equally to indicate different situations of use. A study in Norwich, England found similar patterns of variation between data sets. Among five different social groups, the middle-middle class (the highest-ranking group in the study) always used /ŋ/ in the formal register of the reading style, while lower-working-class residents never used it in its more informal language. Thus, while the five social groups used both pronunciations in their speech, the difference at the extremes of socioeconomic level and situational formality was 100%. As the frequencies in Figure 10-3 show, the pattern in Norwich is the same as in New York City: each social group uses /ŋ/ more in reading style and less in casual speech, with a medium percentage for cautious speaking. . It is clear that in this variable, three widely separated English-speaking communities use /ŋ/ to denote situations of greater or lesser formality. Note that it is not the absolute percentage that indexes the situations, but the relative percentage with respect to other situations. The data show that this linguistic situation marker is a continuous variable capable of indicating subtle differences in the level of formality in different language situations.

• 323

3 2 4 • Chapter 1 0 Language variations in situations of use: registers and styles

FIGURE 10-2 Percentage of -ing pronounced as /IN/ in two language situations by men and women in Los Angeles 76

Women

male

It's a prank

what is discussed

Fonte dos dados: B. Wald y T. Shopen, „A Researcher’s Guide to the Sociolinguistic Variable (ING)“ en Shopen y Williams (1981), p. 247.

As another example of phonological variation (or the corresponding spelling), we examine the distribution of common contractions such as can't, won't, and I'll in different usage situations, from telephone conversations between personal friends and between people they don't know. They are not known to write in newspapers (press) and scientific journals. Even with a remedy as direct as contractions, speakers in different speech situations exhibit different forms. The counts in Figure 10-4 are based on a corpus of spoken and written British English and represent the average CONFIGURE 10-3 percentage of pronunciation of -ing pronounced as /IN/ in three language situations in five social groups in Norwich, England

97 100

90 72

56 34 26

2 THE TYPE

12 CMM

13 YOU

LMC

CMM

Social groups: lower working class, middle working class, upper working class, lower middle class, middle middle class Informal data source: Trudgill 2000.

Careful

file

The style draws bullet points for 1000 words. Notice that as you move from a phone call with friends to phone calls to strangers, to interviews, to radio broadcasts, etc., the formality gradually increases. The increase in formality is accompanied by a decrease in the frequency of contractions.

FIGURE 10-4 Number of contractions per 1000 words in different registers

Data source: Beaver 1988.

Recording grammar markers Usage situations are also marked by syntactic variables. As an example, consider the occurrence of prepositions at the end of a clause or sentence. You may remember from your school days that some teachers disapproved of prepositions at the end of sentences. Instead of This is the teacher I was talking about, they recommended This is the teacher I was talking about. Well, it's no secret that, despite warnings to avoid them, English abounds with sentence-ending prepositions. What is less well known is that they do not occur with the same frequency in all language situations. Using the same body of text used for contractions, Figure 10.5 shows the number of sentence-ending prepositions per 1,000 prepositions for nearly a dozen spoken and written records. This figure doesn't show the same continuous slope from less formal to more formal that we've seen with contractions. Instead, there is an important distinction between speaking and writing non-fiction, with fiction writing (including fictional dialogue) having values in between. In spoken recordings, an average of between 33 and 56 prepositions per 1,000 appear in the final sentence position. However, in the non-fiction books, the closing prepositions are fewer

• 325

3 2 6 • Chapter 1 0 Language variations in situations of use: registers and styles

than in all spoken registers. Therefore, there is a notable difference between speaking and writing regarding the final prepositions of the sentence.

Figure 10-5 Number of sentence-ending prepositions per 1000 prepositions in different registers

Data source: Beaver 1988.

As a second example of grammatical variation in different usage situations, consider this short passage from legalese where a record is identified by name. At the request of the Borrower, the Lender may, in its sole discretion, make future advances to the Borrower prior to the Trustee returning the Property in full to the Borrower. Said future prepayments with interest thereon will be guaranteed by this Escrow Term if evidenced by Notes declaring that said Notes are guaranteed by this Instrument. This passage clarifies some characteristic syntactic features of the legal term: 1. Frequent use of passive structures: must be insured, are insured 2. Preference for the repetition of nouns over pronouns: Kreditor/Creditor's Discretion, Promissory Notes/Said Promissory Notes, Advances / Such Future Advances 3. Omission of Certain Indefinite and Definite Articles: At the request of the Borrower, Borrower, Creditor, Creditor, by the Trustee

Semantic Record Markers A given word often has different meanings in different records. Look at the word notes. Promissory notes within the meaning of the above legal passage are promissory notes or promissory notes. inside

Similarities and differences between what is spoken and what is written It registers its daily meaning, but the notes refer to brief and informal written communications on each subject. Words that have a meaning in ordinary everyday usage but have a different meaning in the legal registration include those given here. Expressions that have a different meaning in legal jargon should continue to be used in an alienating way to save the part

Action of the audience Actions taken Process notes

Judgment pilot movement reasonable man respect

Not only lawyers, but also some of their clients can assign special meanings to words. Criminal jargon contains many words and phrases that are commonly used but have a different meaning when used in the context of criminal behavior. The following two lists are for illustrative purposes. Criminal general lingo mafia hot close-up bite

Sing a sign to Rat Bat

Bug Bird Cage Slammer Joint ('prisión')

Marijuana weed very high

Low Speed Slider Manifold Link ('Weed Cigarette')

The jargon of the drug world Crack Coke Snow Rock Dime

Each of these expressions has a meaning in everyday situations, but a different meaning in the underworld.

Check it out for yourself: while some of the following expressions have no colloquial meaning, each also has a colloquial meaning in extremely informal social situations. Give a colloquial meaning for each term: the nouns Skinny, Main Squeeze, Hunk, Jerk, Nerd, Wimp, Spaz, and Tube; the verbs vege out, party, wig out, and nuke; the adjectives great, great and clueless; Instructions come alive and provide a clue.

Similarities and Differences Between Spoken and Written Records Although writing is sometimes referred to as simply written language (visual language as opposed to audible language), writing and speaking often serve different purposes and have different linguistic characteristics. The conversation, of course, is not a written record, but it can be depicted in novels and screenplays. Legal contracts are also not generally spoken of. Imagine the words and syntax of a handwritten last will

• 327

3 2 8 • Chapter 1 0 Language variation in situations of use: registers and styles

or it will differ from one made by a tester speaking on a videotape. Or consider the linguistic differences between a note pinned to a refrigerator door and the same basic message spoken to someone face-to-face. You will quickly realize that speaking and writing are not mirror images of each other. 1. Oral communication can use intonation and intonation to convey information. Face-to-face communication can also use gestures, postures, and physical proximity between participants. In writing, the only channels available are words and syntax, supplemented by typography and punctuation. When speaking, communication is possible on several channels at the same time. We can criticize a person's personality in a seemingly objective way while using tone of voice or body language to express how much we admire the person, or vice versa. When writing, you need to communicate much more lexically and syntactically, although there are ways to achieve ironic and sarcastic undertones that allow recipients to read "between the lines." 2. Speaking and writing differ in the scope of planning. Most written records give you time to edit and review them. On the other hand, during a conversation, pausing to find the right word can test the patience of your interlocutor and risk losing the word. The difference in the time available to plan and edit the written records creates distinctive syntactic patterns that are difficult to achieve under the immediate processing limitations imposed on spontaneous language. Written records tend to have a more specific and varied vocabulary, in part because authors have time to choose their words carefully and even consult a thesaurus. Of course, not all written records are more artificial than all spoken records. Academic conferences and job interviews reflect some of the characteristics of planned writing. On the other hand, some typefaces are created with relatively little planning, and the language of a letter scrawled a few minutes before it's mailed tends to be fairly spoken. 3. Speaker and recipient often face each other, but writer and reader do not. In face-to-face interactions, the immediacy of the interlocutors and the interaction contexts allow them to relate to themselves (I think you see) and their own opinions and be more personal in their interaction. Rather, writing contexts limit the degree to which written expression can be personal. But be careful not to generalize too much. For example, think of a personal letter and a friendly face-to-face conversation. People may feel entitled to be equally likable in both contexts. Thus an impersonal attitude is a characteristic of only a few written registers, just as a personal attitude is a characteristic of only a few spoken registers. 4. Written records tend to be less dependent on the interaction context than spoken records. Writing is more context independent. In spoken registers, spatial deixis (like the demonstrative pronouns this and that) and temporal (like today and next Tuesday) expressions can be understood in relation to the here and now of the statement. In contrast, the lack of a common writing environment makes such expressions rather opaque or confusing. In written text without a date, which day would you refer to today? And what would it refer to if found in a printed document? Like other distinctions between registers, reliance on deictic expressions is not an absolute difference between speech and writing. In telephone conversations, for example, this cannot be said (refer to something in the speaker's environment) without risking opacity. Instead, you can leave a written note on the kitchen table that says: Please do not eat! as long as the reference point of it is obvious

Two records compared to what is written next to the bill; an author of a textbook can certainly refer to this page or sentence.

The spoken and written registers differ in many ways. But when we examine the differences, we don't find an absolute dichotomy between them. For example, some words may occur only in speech or only in written form, although certain words may occur more frequently in one form or another. Written records tend to be more formal, informative, and less personal. Along a continuum of "personal/impersonal," the manner of writing in legal documents is on the impersonal end, while casual conversations tend toward the personal end. But personal letters can come close to a conversation in their linguistic character. Writing and speaking therefore do not form a simple dichotomy, and in order to describe their differences we need to look at which written register and which spoken register are being considered. For all languages, the usage situation is the most influential factor in determining the linguistic form.

Comparison of Two Registers To illustrate the nature of register variation, let us examine two short passages from an English text. The first passage is immediately recognizable as legal language. Although critics have pointed out that legal language can be considered a foreign language because it is so different from ordinary writing and speaking, it is just one of many registers of English. It may be more obscure than other registers for those not used to it, but it is not a foreign language. This passage comes from a Lord by an act of trust. A trust deed is a written agreement that places title to the property in the hands of a trustee to guarantee that money borrowed as security against the property will be paid; a pilot is merely an addition to the base document. Binding to a Deed of Trust Notwithstanding anything to the contrary in the Deed of Trust, this is the case

line

Phrase

agreed that the loan secured by this trust agreement pursuant to

to, and will be construed and governed by, the laws of the United States

States and the rules and regulations they make,

including federal laws, rules and regulations governing federal savings

and credit unions. If a paragraph, clause or provision of this

Deed of trust or promissory note or other obligations guaranteed by this deed

Trust is constructed or interpreted by a court of competent jurisdiction

jurisdiction is invalid or unenforceable, such a decision will be made

affect only the paragraph, clause or provision so construed

or interpreted and will not affect the remaining paragraphs,

Clauses and provisions of this trust deed or promissory note or

other obligations guaranteed by this Trust Term.

• 329

3 3 0 • Chapter 1 0 Language variation in situations of use: registers and styles

The second passage is from a personal interview with former President Harry Truman by biographer Merle Miller (Plain Speaking [New York: Berkley Books, 1974], p. 242). An interview with Harry Truman

line

Phrase

Q. What do you think is the biggest mistake you made as President?

A. That damn idiot from Texas that I first appointed Attorney General

and then taken to the Supreme Court.

I don't know what happened to me.

He didn't do well as Attorney General and on the Supreme Court

To block . . . It doesn't seem possible, but he was even worse.

He didn't make a right decision that I can imagine.

So if you're asking me what my biggest mistake was, it's this one.

Placement of Tom Clark on the United States Supreme Court.

I thought maybe I'd get better when I got on the pitch

but of course that didn't happen.

I told you when we were talking about this other guy.

After a certain age, it's pointless to think that people will do this

change a lot

At first glance it is striking how strikingly different these passages are. The Deed of Trust is 138 words long and consists of just two sentences. In contrast, Truman's 135-word interview unfolds in eight sentences. The average sentence length is 69 words for the deed of trust, 17 for the interview. (In transcribing Truman's words, the interviewer formed nine sentences; in numbering, we use the letters a and b here to indicate a combination of two of the interviewer's sentences into single sentences, so as not to exaggerate the number of single sentences.) You find it instructive to examine the passages carefully to identify other linguistic features that contribute to register differences.

Try it for yourself: Before reading the following review, jot down as many observations as you can about the differences in vocabulary and grammar between the trust deed and the Truman interview.

Lexicon and Grammar One easily discernible difference between the passages is in the vocabulary. The Deed of Trust contains certain words and phrases that may seem strange when appearing in the interview. Also, the Truman language contains certain earthy words that may seem inappropriate in a legal document. You will also see that there are clear differences between passages in the placement of words with other words, as well as in preferred lexical categories and syntax. Such features, taken together rather than in isolation, help mark passages as if they were

Two registers Comparison of certain types of text, specific language variants, adapted to specific speech situations, specific registers. Vocabulary In contrast to the short, everyday words of the interview, the act of confidentiality uses more unusual words notoriously characteristic of legal language. His vocabulary is more "Latin", the words are longer: promulgated, interpreted, regulated, regulation, obligations, decision, jurisdiction, provisions, void, unenforceable, compliant, guaranteed. Note also the very legal placement of competent jurisdiction, where 'competent' does not have the ordinary meaning of 'capable' but has the legal meaning of 'having proper authority over the matter to be decided'. another legal meaning. Besides competent, other words in the passage have specific legal meanings: deed, trust, obligation, decision, provisions, and note (as well as gentleman not appearing in the passage itself). Nouns and Pronouns For comparable amounts of text, the Deed of Trust has a total of 40 nouns, while the Interview has only 17. On the other hand, the Interview has many more pronouns than the Deed of Trust. He frequently uses first- and second-person pronouns (twelve times in all): I, I, and we eight times, and you four times. (The possessive adjunct my also occurs once.) In contrast, there are no first or second person pronouns in the trust deed. The interview also contains common third-person pronouns: Truman uses it five times in reference to Tom Clark. Rather, entire noun phrases are repeated in the trust deed: the trust deed appears six times, the coordinated noun phrase Rules and Regulations twice, and the coordinated paragraph, clause or provision three times (once singular and twice). Singular). in plural). A longer than average noun phrase component is repeated and contains a repetition of the fund contract: this fund contract or notice or other obligations secured by this fund contract. There are other differences in pronominal usage as well. Truman uses the demonstrative that as a "sentence" pronoun, referring not to a noun phrase but to a whole clause, since it was not (line 11). That's all (line 8) that can be traced back to my biggest mistake at or before the indictment of Tom Clark in the US Supreme Court. Given the need for a deed of trust to be very specific, and given that the function of prepositional phrases is to express certain semantic functions, e.g. B. Agent (by a court), instrument (by this term), place (in the term) - The frequency of prepositions in the deed of trust is not surprising. Records whose purpose is primarily informational tend to have a much higher proportion of prepositions than other types of records precisely because prepositions provide a framework for semantic information. Note that the interview contains only one occurrence of prepositional clauses used sequentially (in the United States Supreme Court), but the deed of trust has seven, including this triad: in the reverse deed of trust. Also, the interview includes an example of a preposition at the end of the sentence (He didn't make a right decision you can imagine), a feature that doesn't appear in legal passages and is very rare in formal writings of any kind (as in Figure 10-5 on page 326). Verbs Considering that the sentence should be interpreted and regulated as containing two verbs, the number of verb groups in the trust deed is nine, about one-third the number of the trust.

• 331

3 3 2 • Chapter 1 0 Language variation in situations of use: Registers and styles

Interview with Truman. Therefore, the interview is very verbal. As for specific verbs, Truman uses "think," "know," and "seem," and his interviewer uses "consider." These "private" verbs represent the internal states of a speaker or writer. They are appropriate in the interview and come up too often in the conversation even though they were out of place on the trust deed. Truman also uses proverbs of various kinds (proverbs replace other verbs, just as pronouns replace nouns): do and happen, which can replace many verbs; Put and Get, which are more limited but still have a variety of uses. In conversations where there is pressure to come up with words quickly, proverbs appear frequently, in part because they save the time it would take to find a more explicit verb. In this brief passage, got appears twice, and Truman uses put and wear (Supreme Court) instead of, say, named for. The verb to be, the most common verb in English, appears seven times as a main verb, while in the Deed of Trust it appears four times as an auxiliary verb (is shared, is done, is interpret and is interpret), but only once as a verb. a main verb (to be empty). Some verbs in the deed of trust relate to the subject of discussion and thus to the passage record: agree, interpret, govern, enact, interpret, and influence. Regardless of the topic, but characteristic of legal language is the use of should as an auxiliary verb. Although it appears in other records, its use in legal language is exceptionally common. It must appear as an auxiliary clause in both sentences of the trust deed. The interview relates to the years of the Truman presidency, which is reflected by the predominance of past tense verbs. Of his 25 verb groups, 14 are in the past tense, while the 8 verbs in the present tense generally refer to the ongoing interaction between Truman and the interviewer, or to Truman's own train of thought during the interview: What do you think when you ask? I don't know, I can think. The only verb that refers to the future tense uses the construction they are going to instead of should or will. Negation In the interview, four of the five negative morphemes appear as the negative adverb not (attached to the verb as a contraction). The fifth is the unchanged adverb (damn) good. Instead, the Trust Deed contains elements of negation in adjectives or prepositions through processes of derived (invalid, unenforceable) or compound (yet) morphology; there is an isolated non (which occurs in relation to the future tense, not, in contrast to a positive future tense). A distinctive difference between speech and writing is the much higher frequency of negation in spoken registers, where the vast majority of negative elements are separated into words as no (commonly perceived as -n't) rather than incorporated as invalids. Legalese Adverbs is famous for its use of compound adverbs like the same and below. Our passage contains only one example of this. In fact, apart from a single not, the legal passage contains only the two adverbs just and so. Truman's adverbs are different. You use them for reference time (first, then) and as billboards to indicate your position in relation to what you're saying, e.g. B. Natural and maybe. Passive What is striking about the trust agreement is the frequent use of passive verbs (wake up, do, interpret and drive, interpret or interpret). Passive constructions degrade an agent subject to the object of a preposition, allowing the agent to be omitted (Lightning fall on the house/The house fell on Lightning/The house fall). In legal, both agentless passives (those without the phrase by) and passives with by are common. In stark contrast to the act of confidentiality, Truman and his interviewer use only active verbs.

Two-Plate Comparison Questions By using the form of a direct question (when he asks me what my biggest mistake was) rather than an indirect question (when he asks what my biggest mistake was), Truman contributes to an informal impression. And while it may seem too obvious to mention, the interview includes a question (as is natural in interviews) which is not only a word structure not found in this Deed of Trust but would be unusual in such legal documents. Reduced Relative Clauses Another feature of legal language is the frequency of reduced relative clauses, in which the relative pronoun and a form of the verb to be do not appear where they should. These examples show the omitted words in parentheses. Loans (i.e.) guaranteed rules and regulations (which) are promulgated in accordance with the paragraph, clause or provision (which) so interpreted

Conjunction Truman's interview shows frequent coordinating conjunctions such as and, but, and then and so, which are mainly used to connect clauses, as in lines 5, 6 and 8. These conjunctions are absent from the legal passage, with the exception of and . which is not used to link clauses, but to link verbs, nouns, or adjectives. Another typical feature of legal usage is the combination of triple phrases "X, Y and Z" or "X, Y and Z". In legal records, X, Y, Z can be members of almost any category but are usually noun phrases, adjectives, or verbs; they are usually members of the same lexical or phrase category. Here are some examples of this pattern: Laws, Rules, and Regulations (nouns) Paragraph, Clause, or Provision (nouns) Deed of Trust or Promissory Note or other obligation (noun terms) Void, void, or unenforceable (adjectives)

Sometimes variations within Constituents X, Y and Z result in similar but not entirely parallel structures, as in these examples: 1. Manufactured in accordance with and construed and governed by 2. United States laws and rules and regulations

In 1 there are two verb structures conjugated by e, but the second verb phrase itself contains two conjugated verbs (interpreted and governed). In Figure 2 we cannot more accurately describe the structure as "X, Y and Z" but as "X and Y" where Y is a composite M and N; i.e. "X and (M and N)".

Phonology Since only one of the two passages originated linguistically, we cannot make any direct phonological comparisons. We do not have a phonetic transcription, but we can infer from the transcribed text that Truman had frequent phonological abbreviations. Instead of full forms like don't, even in this small sample there are eight contractions: don't, does't, is not, has not, he's, hed, that's and it's. On line 1, the only place in the deed of trust where a comparable form is allowed to appear, it happens, it doesn't. If we compared two forms of spoken English and had correct transcriptions, we could say more about the phonological similarities and differences.

• 333

3 3 4 • Chapter 1 0 Language variations in situations of use: registers and styles

Comparing Records In comparing and contrasting the two passages, neither feature indicates which records they illustrate. Rather, various features that appear in combination characterize the first passage as legal language and the second as an interview. Truman's style is so informal that it suggests a conversation rather than a formal interview; this may be partly because the interviewer spent several months morning and night with Truman. Undoubtedly, as the days went by, the interview became more and more of a conversation among friends. You have now seen that the properties of language differ from one speech situation to another. Sometimes there is more than one feature in a given dataset than in another, and occasionally a feature occurs exclusively or almost exclusively in one dataset. Sometimes the same form appears in more than one register, but with different meanings or uses.

Computers and the Study of Record Variation In the fields of artificial intelligence, expert systems, and many high-tech fields that are extremely important today, the role of data sets is crucial. The reasons are complex, but you can get an idea of some simply by considering the different syntax and vocabulary patterns in records that each system should be able to handle, e.g. B. Information in the form of headlines, medical, legal, etc. or talkative. . Think of it this way: When your corpus contains nothing but journalistic writing, but you can't differentiate between different types of journalistic writing (news, personals, editorials and editorial letters, advertisements, cartoons, sports commentary, business analysis, stock markets). and weather reports, etc.) would have to be far more complicated than a series of individual systems designed to process multiple sets of data sequentially. The importance of computers in the study of registry and registry variation can hardly be overstated. Corpus compilers have always been aware of the importance of classifying text into records. (In effect, this means labeling each text as belonging to a specific dataset.) Because many registry studies have been quantitative, large-scale corpora help ensure reliability and validity, although corpus design is extremely important in establishing validity. infront of us

He saw that the English Brown and LOB corpora were each about 1,000,000 words. These are not large corpora by today's standards. Although not the largest corpus in the world, the British National Corpus contains 100,106,008 words. According to the BNC website, The Corpus occupies about 1.5 gigabytes of storage space, which is the equivalent of more than a thousand high-capacity floppy disks. To put these numbers in perspective, the average paperback is about 250 pages per inch thick; Assuming 400 words per page, we estimate that the entire corpus, printed in fine print on fine paper, would take up about ten feet of shelf space. Reading the entire corpus aloud at a breakneck speed of 150 words per minute, eight hours a day, 365 days a year would take just over four years. Some of the research reported in this chapter, with an emphasis on quantitative corpus assessments, was computational. Aside from the task of creating them physically on paper, the data in several tables and figures were generated without computers, such as Figure 10-2 on page 324, which shows the frequency of -ing, pronounced /ŋ/, in males indicates women. Los Angeles women. But for other data were computers

Continue

necessary, at least in practical terms. With a tagged corpus, it would be fairly easy to identify some features. In the direct category we can include nouns, prepositions, demonstrative pronouns, private verbs, etc. Depending on the extent of marking, other categories could be identified, such as B. Verbs in the past tense, but if the corpus were not tense-marked, an algorithm would have to be specified to tell the computer what to look for. Algorithms would also be needed to create structures such as

• 335

Sentence pronouns and final prepositions of sentences. Some algorithms would be particularly difficult to design. With that in mind, you might want to think about the kind of algorithm that would tell a computer to recognize these omissions, as in She said she selected instead of She said she selected. After all, it's one thing to write an algorithm that identifies a feature that is present, but identifying a feature that isn't is much more difficult. ■

Summary • Three main elements determine every speaking situation: environment, purpose and participants. • The subject and location are part of the framework. • The nature of the activity and the goals are part of the purpose. • On the part of the participants, it is not only the people themselves who influence the form of the language, but also the roles they play in the speaking situation. • Because we wear different clothes for different occasions and activities, we often don't speak in the same way in court, at dinner, and on the soccer field. • In multilingual communities, different speaking situations sometimes require different languages and sometimes different varieties of the same language. • Registers are characteristic linguistic varieties of certain language situations. Records are also sometimes called styles. • The set of varieties used in a language community in different language situations is called its linguistic repertoire or its verbal repertoire. • The linguistic repertoire of a monolingual community contains many registers that differ absolutely or generally relatively in their linguistic characteristics. • Each register is characterized by a set of linguistic features, not a single feature. • The sum of these features (lexical, phonological, grammatical and semantic) together with the characteristic patterns of language use in a given situation determines a register. • Since all registers or stylistic variants within a language are based on the same grammatical system, the different uses of this system to identify different registers are relative. • Writing differs from speech in several fundamental ways, but the linguistic differences between the two modes are not absolute.

3 3 6 • Chapter 1 0 Language variation in situations of use: registers and styles

WHAT DO YOU THINK ? R E V I S I T D • Stefanie and the jargon. Probably not all of Stefanie's teachers like slang and colloquialisms, and certainly not all of them in all situations. But teachers understand that jargon is characteristic of extremely informal situations and can view lectures or essays as relatively formal situations. Colloquial expressions, by definition, characterize spoken language. Because language-related homework focuses primarily on reading and writing, teachers may classify certain phrases as "colloquial" when they find them in student essays. Teachers may also feel that students are already familiar with slang and colloquialisms and will need to master more formal registers in school. Language appropriate for conversation may not be appropriate for essay writing. • Michael and contractions. Contractions are an abbreviation that usually represent words in writing as they are often spoken in informal situations. Written contractions therefore mimic the relaxed tone of conversation. For example, when used in friendly letters, they reflect the informality of the conversation. By extension, textbooks can achieve a more relaxed and conversational tone through the use of contractions, since contractions not only reflect a conversational tone, but help create one. Part of this book has an interactive emphasis, asking readers to answer questions ("What do you think?") and find things out ("Try it for yourself"). Contractions attempt to create a more interactive style and engage the readers with the content of the text. • Davin and the fictional dialogue. Few people read the transcript of an actual speech, let alone a transcript of an ordinary conversation. Given the spontaneous and improvised nature of conversations, speakers often need to look up words and organize their syntax to convey what they intend. Sometimes they hit syntactic dead ends and have to go back. For Davin, "natural" can simply mean dialogue that doesn't sound stiff or contain colloquialism or slang. If you were reading a real conversation (with the kind of hesitations and reboots and uhm and uhs in a statement) you would surely get impatient. To avoid such impatience, novelists deliberately avoid making their dialogue completely natural. • Recipe from Uncle James. Because recipes were often passed from one chef to another, scribbled on, for example, used index cards or envelopes, the family cooks probably omitted unnecessary words and used a kind of telegraph language. It is therefore not surprising that the omitted words are these

exercises

which can be easily delivered: "Toast the pine nuts in a medium-sized pan. Remove them and add 1 tbsp. of oil and garlic. Cook for 4 minutes and drain off the remaining liquid. Sprinkle some salt and pepper in the trout cave and fill it with the spinach mixture. Brush the trout with the remaining oil. It's an old-fashioned recipe.

English-Based Exercises 10-1. Consider the following expressions. Please turn off the lights when leaving. Please turn off the lights when leaving. The content of the policy is basically the same in both terms, but the societal meanings differ significantly. Identify the features that highlight the differences between the two policies; Then discuss what impression each might make and under what circumstances each might be appropriate. 10-2 a. Name five pairs of terms for body parts or bodily functions, e.g. B. collarbone/collarbone, which would differentiate a conversation you are having with a doctor from a conversation with a friend on the same topic. B. Arrange the words in each sentence below in order of formality: 1) teacher, teacher, trainer, mentor, educator 2) guru, mullah, teacher, coach, trainer, don c. Are any of the words in (1) or (2) so informal that they are colloquial? to explain. 10-3 In Slang and Sociability, Connie Eble reports the top 40 slang words used by students at the University of North Carolina between 1972 and 1993. The 20 most important are listed in the box on page 321. The top 20 are listed below. some with definitions. Provide concise definitions to others when you are familiar with them; if no, what do you mean by that? eat (verb)

Caliente

addicted

'poor, lazy' loose

Granola

destroyed

Couch potato/~girl/ couch potato

do nothing)

no!

Word (above) 'I agree'

wing (verb)

Great

Partner

book 'get out, quick'

the fountains

there

Bagger 'member of the fraternity'

Fuchs/Fuchs

Mark 'Error'

Sorority Sue/Sue/Suzi

• 337

3 3 8 • Chapter 1 0 Language variation in situations of use: registers and styles

10-4 Record 30-45 seconds of a radio report and a TV report (use the same news report if possible). After transcribing the passages, compare them to see how the medium affects your choice of forms of speech. 10-5 Here is the immediate sequence of the Truman passage quoted in this chapter; Sets are numbered for reference only. Q. (1) How do you explain why he is such a bad judge? A. (2) The director is . . . Well, it's not so much that he's a bad man. (3) It's just that he's such a stupid son of a bitch. (4) He's the dumbest man I've ever met. (5) And this is often the case. (6) Stupidity is the worst thing when it comes to high office, and this is especially true when it comes to the United States Supreme Court. (7) As I said before, I will never know what happened to me when I made this appointment and I am sorry that I did. [Speaking Plain, p. 242]. a. Is it clear what is meant in this case (sentence 5) and is it particularly true (sentence 6)? If so, what type of component are you referring to in these cases? B. What is the name of the language source you examined in question a above? C. Identify all occurrences of be as the main verb. How many? i.e. What is the function of the good in sentence 2? Y. If possible, give a noun phrase that has the same referent as the pronoun it in sentences 2, 3, 6 (twice), and 7. Explain the cases where a noun phrase cannot be identified with the same referent can. . 10-6 a. Look up the definition of slang in a good desktop dictionary and, using the definition, list as many slang terms and phrases as possible for two terms in (1) and (2) below, respectively. B. What makes the terms represented in (1) more susceptible to jargon and expressions than those in (2)? C. As far as you can cite the jargon of the points in (2), do they have positive or negative connotations? i.e. Does the dictionary definition of slang help explain the different distribution of the slang terms in (1) and (2) and the connotations associated with the slang terms in (2)? If yes, please explain how. If not, check the dictionary definition to take into account what you discovered about slang connotations. 10-7 Some of the most common English words (the, of and a, to, it, is, that) appear in both the Deed of Trust and the Interview, as well as almost all English-language records. But one entry where these words are relatively rare is "heading". a. Identify two other records where you can see that these words are used relatively infrequently. B. Choose a sample from one of the two indexes you identified or from newspaper headlines and identify the lexical categories that come to mind.

Exercise more often than in conversation. Notice which lexical categories are relatively rare, if any. C. Hypothesize why the distribution is as you found it. i.e. Of course examine line 11 of the Truman interview at p. 330. At one level it could be analyzed as a prepositional phrase consisting of the preposition of and the noun course. If you consider it compound, what lexical category would it belong to? (Hint: Replace compound words with single words and decide which category the substitutes fall into.) and. On line 9 of the trust deed, decide which lexical category it belongs to in terms of its distribution relative to other word classes. Using the same criteria, what is the lexical category of so in line 10? What is that about line 8 of Truman's interview? F. Name the determinants in the trust deed and those in the Truman interview. Include the specific part of speech for each determiner in your list (eg, article, statement). Grams. The Deed of Trust has one instance of this (line 2) and the Truman Interview has six: lines 2 (twice), 7, 8, 11, and 12. Identify the part of speech for each of these seven instances. h Gives two arguments to categorize however (trust, line 1) as a preposition. I. Remembering that compounds (however) are not always written as a single word, identify another example of a compound preposition in the deed of trust. j. The Deed of Trust contains several compounds (e.g. the preposition "despite all" and the pronoun "any" on line 1 and the compound noun "United States" [composed of an adjective and a noun] on lines 3 and 4). Identify all of the compounds in Truman's interview and note their lexical categories. What are the similarities and differences between the categories compiled in the trust deed and the interview? k. Examine the occurrences of to in the Deed of Trust (lines 1, 3, and 9) and in the Truman interview (line 13). Which, if any, is a preposition? what are the others I. Assuming that the passages are typical of your registers, what generalizations can you make about the registers in terms of your exploration of specific word classes? 10-8 Examine the three letters below. The first is a letter of recommendation for a student seeking admission to a master's program in linguistics, the second is a letter to a journal, and the third is a personal letter from a woman to a friend in another state. Identify the distinctive characteristics of each type of letter.

Letter of Recommendation (182 words) I met Mr. John Smith as an undergraduate on three of my courses at State University, and based on knowing him, I recommend that he should definitely be admitted to graduate school. John was my student in the Linguistics 100, where he did exceptionally well and wrote some really excellent work. Based on this paper, I

• 339

3 4 0 • Chapter 1 0 Language variation in situations of use: Registers and styles

I encouraged him to major in linguistics, and then I was lucky enough to have him in two more classes. In one of these (Historical Linguistics) he led the class and apparently worked with more insight than the other seventeen enrolled students. In the other course (Introduction to Phonology) he did not do as well, perhaps because he was under financial pressure and forced to work twenty hours a week with a full course load. In all three courses, John worked very hard and did much more than was necessary. I recommend John Smith unreservedly. You know what you want to achieve and have a strong motivation to be successful in graduate school.

Editorial (91 words) Your Afghanistan story was wrong when you claimed that the 1973 Russian-backed coup was bloodless. As a Peace Corps volunteer in Afghanistan, I saw the dead bodies and the blood and dodged the bullets. It is estimated that between 1,000 and 1,500 were killed, but it is difficult to get an accurate count when a tank reaches the home of the Shah's supporters and fires repeatedly from 30 feet, or when entire families disappear in the middle of the night.

Personal letter (142 words) What's up? Not much happens here. I'm at work right now and it's been really slow this week. We haven't done anything. I hate it when it's so slow. The week never seems to end. Hey how are you? Did you get the photos and the letter I sent you? We haven't heard from you for a long time. Mom has her birthday present ready to send to you and Dan, but she doesn't know when she will send it. how are the kids Does Dan like kindergarten? Well, Al went to school. I miss him much. He went to the LLTI on Monday. It's an upstate business school. You only have to go two years, and it will need air conditioning and cooling, and then it will need heating. 10-9 a. Look again at what was said in the discussion on p. 11 about jurisdiction (as in competent jurisdiction). 331. Next, try to specify the legal meaning of the following words, which are also used in the Deed of Trust with special meanings: deed, trust, obligation, decision, dispositions, and promissory note. List all the words that were used with special meanings in Truman's interview and specify the meaning. B. Give another example of a reduced relative clause in the trust deed in addition to the three on p. 333. c. List all the examples of reduced relative clauses in the Truman passages on p. 330 and in the order given in Exercise 10-5 above. 10-10 Below are personal ads (slightly adjusted) from a weekly Los Angeles newspaper. Examine your linguistic peculiarities and answer the following questions.

Exercises 1) Aquarius SWM, 33 years old, strong build, blue eyes. You: marriageable bilingual Latina woman, 23-30, ok kids. 2) Busty, bright and beautiful businesswoman, 40 years old (she looks 30). Looking for a possibly younger, taller, handsome and thoughtful SWM that respects individuality. Someone living the impossible dream, financially secure, good conversationalist, for relationship, n/s. 3) SWM, 28 years old, attractive student, works for major US airlines, likes to travel. Seeking Woman, 23-32, approachable and intelligent for world class romance and possibly marriage. 4) vegetarian English. SWM, 31 years old. Sincere, sensitive, original, thoughtful, atypical, funny, shy, playful, lovingly professional. I'm looking for a warm, funny and outgoing WF under 29 to share my life with. 5) Slim, young, GWM, very straight appearance, masculine, athletic, healthy, clean-shaven, decent. Seeking similar good looking WM under 25 for monogamous relationship. 6) SBM very romantic, 24 years old, studying. I am looking for a rich, healthy and beautiful woman for friendship and maybe romance. Fakers and pranksters don't have to apply. 7) Hispanic DF, small but full of life, likes sports, dancing, travelling, looking for someone with the same interests, 30+, no matter race. 8) Evolutionary, positive thinking, spiritual, loving, honest, handsome, healthy, confident, 36, 6', 160#, blue eyes, humble, uncompromising, professional. Looking for mate, soul mate, marriage, family. a. Which lexical categories are more common in advertising compared to conversation? Which ones are particularly rare? B. Identify eight distinctive linguistic features of personal ads. These can be features of syntax, morphology, vocabulary, abbreviation conventions, etc. C. List the verbs in all ads, identifying their grammatical person (1st, 2nd, 3rd) and number (singular, plural) whenever possible. (Hint: Include the pronoun that would serve as the subject of each verb to determine person and number.) d. Choose one of the ads and try to write it entirely in colloquial English, simply by providing additional words. Retains the word order and word forms of the original ad. Y. What do you intend to indicate that the ads present an abbreviated or abbreviated form of conversational English? If you think the ads aren't abbreviations for English conversational sentences, what explanation can you offer for the shape of their sentences? Q. What personal ad language features do you find so conventional that prior knowledge of the data set is required to understand them? 10-11 Examine a recent issue of your campus newspaper and identify any entries it contains (eg, editorials and movie reviews). Choose a passage from a register and list eight linguistic features which, due to their high frequency, contribute to the characterization of this register; Give an example of each feature of your passage.

• 341

3 4 2 • Ch a p te r 1 0 Language variation in situations of use: registers and styles

10-12 Prescriptions, obituaries, classifieds, picture ads, telegrams, birthday cards, loan applications, course descriptions in college catalogues, directions for using medication and essay questions are just a few of the unique records you can use on a regular basis. From one of these datasets, choose a small sample of text and give a list of its distinguishing features, with an example of each feature from your sample.

Based on English and other languages 10-13. Identify multiple instances of language features that vary between records in a foreign language you're learning. (Some features may be mentioned in your foreign language textbook, others by your teacher.) Identify at least one phonological feature, one grammatical feature, and several vocabulary words that vary depending on the usage situation. For each function, indicate a situation in which it is appropriate and a situation in which it is not. (Tip: Consider situational differences such as writing vs speaking, formal vs informal, speaking quickly vs speaking cautiously, interaction between you and, say, a teacher vs you and a close friend.)

Especially for educators and prospective teachers from 10 to 14 years. Examine a book in a foreign language and look for clues from the author that the specific language varies from situation to situation. This evidence may focus on formality versus informality, differences between language and writing, dealing with recipients of different social status, slang or slang, or any other linguistic variation that depends on the situation of use. How do you rate the book's clarity on the importance of such differences when it comes to sounding or writing like a native speaker? 10-15 Look at the opening material in your dictionary (or whatever you recommend students do) and find the discussion of how you deal with jargon. (You may need to look under "usage" or "tags" for discussion.) Compare what the dictionary says about slang at the beginning with the definition given in the main part of the dictionary entry list. Finally, look up six of the most common slangs your students (or colleagues) use and see if the dictionary notes the meaning of the slang you have in mind and labels it slang. Based on this exercise, do you think this particular dictionary is a useful source of slang information for you? For your students? Would students typically use a dictionary to gather information about slang? Who would normally use a dictionary to determine the meaning of slang? To what extent should a dictionary try to accommodate slang terms and slang meanings?

Other British National Corpus resources: http://www.natcorp.ox.ac.uk/

The British National Corpus home page allows you to submit queries to the BNC itself and obtain sample sentences containing the phrase queried.

Advanced Reading

Suggestions for further reading • Allan Bell. 1991. The Language of Media (Cambridge, MA: Blackwell). The most accessible in-depth analysis of a single data set. • Vijay K Bhatia. 1993. Gender Analysis: Language Use in Professional Settings (London: Longman). A qualitative approach to record keeping, one step further than this book. • Robert L. Chapman, editor. 1995. American Slang Dictionary, 3rd ed. (New York: HarperCollins). A good slang dictionary; it also discusses the nature and sources of the jargon. We take slang examples for illustration in this chapter from the dust jackets of this volume. • David Crystal and Derek Davy. 1969. Inquiry into English Style (London: Longman). Contains accessible chapters on slang, religion, newspaper reports and legal documents. • Connie Eble. 1996. Slang and sociability: group language among college students (Chapel Hill: University of North Carolina Press). Very informative with a glossary of over 1000 slang terms. • J.E. Lighter, ed. 1997—Random House Historical Dictionary of American Slang (New York: Random House). This is an important work of interest to historians of American English and American slang, and to anyone interested in the history of certain slang terms. Two volumes were published (about the letter O); the final volume will be published shortly by Oxford University Press. • Timothy Shopen and Joseph M. Williams, eds. 1981. English Style and Variables (Cambridge, MA: Winthrop). Essays on discursive, literary and other styles.

Advanced Reading Brown and Fraser (1979) examine the elements of language situations that can affect language. Joos (1962) is a popular treatment of the notion of language style. Fishman (1972) describes the changeover in Brussels, while Blom and Gumperz (1972) describe the changeover between Bokmål and Ranamål. Biber (1988) is a quantitative study of variation in a corpus of spoken and written English, while Biber (1995) analyzes textual variations in Korean, Somali, and other languages. O'Donnell and Todd (1991) discuss English in the media, in advertising, in literature, and in the classroom. For discussions of other written records, see Ghadessy (1988). The chapters of Biber and Finegan (1994) describe recordings of sports coaches, personal courtship, and table talk, as well as variations on recordings in Somali and Korean. Andersen (1990) describes the use of registers in children. Finegan (1992) discusses the development of fiction, essays, and letters over several centuries, along with attitudes toward standardization during this formative period. Lambert and Tucker (1976) report several socio-psychological studies of forms of treatment, primarily in Canadian French, Puerto Rican Spanish, and Colombian Spanish. Useful and insightful discussions of French registers can be found in Sanders (1993) and George (1993), while Burke (1988) extensively illustrates French colloquial usage and slang. Barbour and Stevenson (1990) contain two chapters discussing aspects of situational variation in German, and Clyne (1999) also discusses situational variation. For a more advanced discussion of registration, see Leckie-Tarry (1995). Also advanced, Duranti and Goodwin (1992)

• 343

3 4 4 • Chapter 1 0 Language variation in situations of use: registers and styles

offers descriptive and theoretical perspectives on the importance of context. Eckert and Rickford (2001) reflect on anthropological approaches to style, the traditional sociolinguistic concept of style as attention to language, the important question of audience design, and functionally motivated situational variations. Accessible chapters on American slang (by Connie Eble), rap and hip-hop (by H. Samy Alim), the language of cyberspace (by Denise E. Murray), and doctor-patient conversations (by Cynthia Hagstrom) appear in Finegan and rickford . (2004).

References • Andersen, Elaine S. 1990. Talking with Style (London: Routledge). • Barbour, Steven and Patrick Stevenson. 1990. Variation in German: A Critical Approach to German Sociolinguistics (Cambridge: Cambridge University Press). • Beaver, Douglas. 1988. Variação entre Fala e Escrita (Cambridge: Cambridge University Press). • Beaver, Douglas. 1995. Dimensions of Register Variation: A Cross-Language Comparison (Cambridge: Cambridge University Press). • Biber, Douglas and Edward Finegan, eds. 1994. Sociolinguistic Perspectives on Recording (New York: Oxford University Press). • Blom, Jan-Petter, and John J. Gumperz. 1972. “Social Meaning in Language Structure,” in John J. Gumperz and Dell Hymes, eds., Directions in Sociolinguistics (New York: Holt), pp. 407–34. • Brown, Penelope and Colin Fraser. 1979. “Speech as a Marker of Situation”, in Klaus Scherer & Howard Giles, eds., Social Markers in Speech (Cambridge: Cambridge University Press), pp. 33–62. • Burke, David. 1988. Street French: How to Speak and Understand French Gyria (New York: John Wiley). • Clyne, Michael G. 1999. A Lingua Alemã em Uma Europa em Mudança (Cambridge: Cambridge University Press). • Duranti, Alessandro and Charles Goodwin, eds. 1992. Rethinking Context: Language as an Interactive Phenomenon (Cambridge: Cambridge University Press). • Eckert, Penelope and John R. Rickford, eds. 2001. Style and Sociolinguistic Variation (Cambridge: Cambridge University Press). • Finegan, Eduardo. 1992. “Style and Standardization in England: 1700–1900,” in Tim William Machan & Charles T. Scott, eds., English in Its Social Contexts: Essays in Historical Sociolinguistics (New York: Oxford University Press), pp. 102– 30 • Finegan, Edward and John R. Rickford, eds. 2004. Language in the US: Perspectives for Century 21 (Cambridge: Cambridge University Press). • George, Ken. 1993. “Alternative French”, in Carol Sanders, ed., French Today: Language in Its Social Context (Cambridge: Cambridge University Press), pp. 155–70. • Ghadessy, Mohsen, eds. 1988. Notes on Written English: Situational Factors and Linguistic Features (London: Pinter). • Joos, Martin. 1962. The Five Clocks (New York: Harcourt).

References • Labov, William. 1966. The Social Stratification of English in New York City (Washington, DC: Center for Applied Linguistics). • Lambert, E. Wallace and Richard G. Tucker. 1976. You, You, You: A Socio-Psychological Study of Address Patterns (Rowley, MA: Newbury House). • Leckie-Tarry, Helen. 1995. Language and Context: A Functional Linguistic Theory of Record (London: Pintner). • O'Donnell, W.R. and Loreto Todd. 1991. Variety in Contemporary English, 2nd ed. (London: HarperCollins). • Sander, Carol. 1993. “Sociosituational Variation”, in Carol Sanders, ed., French Today: Language in Its Social Context (Cambridge: Cambridge University Press), pp. 27–54. • Trudgil, Peter. 2000. Sociolinguistics: An Introduction to Language and Society, 4th ed. (New York: Penguin).

• 345

Linguistic variation between social groups: dialects

WHAT DO YOU THINK ? • Her nine-year-old niece, Nina, who is returning from summer camp, reports that one of the caregivers was “talking very funny”: she called the TV TV, Trucks, Trucks, and Cookies Cookies. What do you tell Nina about who "talks weird" and who doesn't? • Daniel, a friend of his who teaches in Chicago, says that one day after a teacher who had grown up in Alabama filled in for him, his students said the successor spoke with a distinctly Southern accent. But the sub claimed he had no accent. The students wondered how the sub could guess that he didn't have an accent. What explanation would you have offered them? • In a discussion about whether teachers in the US should know about Ebonics, his colleague Justin claims that Ebonics is “just bad English” and that teachers don't have to learn it. What arguments can you make that if Ebonics is "broken", then all variants of English are "broken" from the perspective of all other variants? • In the canteen, you and your classmates discuss how differently students speak. Sammy says they say the same thing. what do you say to her

346

Language or dialect: which one do you speak?

Language or dialect: which one do you speak? It is an obvious fact that people from different nations tend to use different languages: Spanish in Spain, Portuguese in Portugal, Japanese in Japan, Somali in Somalia, etc. Besides looks and cultural traits, language is part of what distinguishes one nation from another. Of course, different languages are not only spoken across national borders. In the Canadian province of Quebec, ethnic Franco-Canadians maintain a strong loyalty to the French language, while ethnic Anglo-Saxons maintain a loyalty to English. Dozens of languages are spoken in India, some restricted to small areas, others spoken regionally or nationally. There is considerable international variation among the speakers of each widely spoken language, as in the case of Australian, American, British, Indian and Irish English, among others. Distinct differences can be seen between the French variants spoken in Montreal and Paris, and between the Spanish variants spoken in Spain, Mexico and several Central and South American countries. Furthermore, even casual observers know that residents of different parts of a country speak regional variants of the same language. When Americans speak of a "Boston accent," a "Southern accent," or "Brooklyn," they reveal that their perception of American English varies from place to place. These language region markers identify people as belonging to a particular social group, even if that group is as loosely integrated as most American regional groups. In countries where regional affiliation may have social correlates of ethnicity, religion, or clan, regional differences can be important markers of social affiliation. Like the existence of different languages, the existence of regional varieties of a language suggests that people who speak to each other tend to speak like one another. It's also reasonable to assume that people who differ from other groups tend to make that distinction in their speech. A language can be viewed as a collection of historically related dialects that are similar in vocabulary and structure. Monolingual dialects characterize social groups whose members choose to be speakers of the same language.

Social Boundaries and Dialects Language varies from region to region and also across ethnic, socioeconomic and gender boundaries. American English speakers know that white Americans and black Americans tend to speak differently, even if they live in the same city. Likewise, middle-class speakers can often be distinguished from working-class speakers. Women and men also differ in language use. Across the world, along with regional dialects, there are ethnic, social class, and gender varieties. These form what some call social dialects, although the word dialects is usually restricted to regional variants.

Distinction between dialect, register and accent Dialect and register The term dialect denotes the linguistic diversity that is characteristic of a particular region or social group. Partly through their dialect we recognize a person's regional, ethnic, social and gender affiliation. The term dialect has to do with language users, with groups of speakers. Furthermore, as we saw in the previous chapter,

• 347

3 4 8 • Chapter 1 1 Linguistic differences between social groups: dialects

All dialects vary according to the situation in which they are used, creating what we called registers in the previous chapter: linguistic varieties corresponding to the situations of use. In this chapter we deal with dialects: varieties of language that are characteristic of certain social groups. Idioms, dialects and registers are varieties of languages. This means that there is no linguistic distinction between a language and a dialect. Every dialect is a language and every language is realized in its dialects. From a linguistic point of view, what is called language and what is called dialect are indistinguishable. Dialect and Accent Dialect refers to a language variety in its entirety, including vocabulary, grammar, pronunciation, pragmatics, and all other aspects of the language system. The terms language and diversity also refer to a complete language system. On the contrary, the word accent only refers to the pronunciation. When we speak of a "Southern accent" or "Boston accent," we are referring to the distinctive pronunciation of the Southern or Boston dialect.

How do languages diverge and merge? How is it that certain linguistic varieties that were once similar become different over time, while other varieties remain very similar? There is no easy answer to this question, but the more people interact, the more similar their language remains or becomes. The less contact there is between social groups, the more likely it is that their language varieties will develop distinct characteristics. Geographical separation and social distancing encourage different ways of speaking. Spoken around 6,000 years ago, the Proto-Indo-European language gave rise to most of today's European languages and many languages of Central Asia and the Indian subcontinent. Not only the Romance languages developed from Proto-Indo-European, but also the Germanic, Celtic, Greek, Baltic, Slavic and Indo-Iranian languages. If you consider that only about 200 generations lived and died in these 6,000 years, you can imagine how quickly a multitude of languages can develop from a single main language. Just as physical distance can foster dialect distinctions, social distance can help create and maintain distinct dialects. In part, middle-class dialects differ from working-class dialects because of a relative lack of sustained interactive contacts across class lines in American society. African American English continues to differ from other variants of American English, in part because of the social distance between whites and African Americans in the United States. A dialect connects its users through recognition of common linguistic features, and speakers' ability to use and understand a dialect marks them as "insiders" and allows them to identify (and exclude) "outsiders". But, as we shall see, it is not necessarily the case that varieties differ from one another in any orderly manner. It may be that two varieties share vocabulary but differ in pronunciation, or share much of their phonology but differ in some other respects. All language variants are constantly changing and evolving.

Linguistic merging in an Indigenous town Just as physical and social distance allows speakers of one variety to distinguish themselves from speakers of other varieties, close contact encourages frequent communication

How do languages diverge and merge? linguistic similarity. Because varieties of the same language spoken by people in close social contact tend to become similar, different languages spoken in a community also tend to merge. Kupwar is a village in India on the border between two main language families: the Indo-European family (which includes the North Indian languages) and the unrelated Dravidian family (the South Indian languages). The 3,000 residents of Kupwar are divided into three groups and routinely use three languages in their daily activities. Jains speak Kannada (a Dravidian language); Muslims speak Urdu (an Indo-European language closely related to Hindi); and the untouchables speak Marathi (the regional Indo-European language around Kupwar and the main literary language of the region). These groups have lived in the village for centuries and most of the men are bilingual or multilingual. Over time, as people switched between at least two of these languages, the varieties used in Kupwar became more and more similar. In fact, the grammatical structures of the village varieties are now so similar that a literal translation between the languages is possible because the word order and other structural features of the three languages are now practically identical. This merging is remarkable because the varieties of these languages used elsewhere are very different. But even in Kupwar, where the three grammars were merged, the vocabulary of each language remained quite distinct. On the one hand, the need for communication between different groups favored grammatical convergence. On the other hand, the social division necessary to maintain religious and caste differences has encouraged the continuation of separate vocabularies. As things stand today, communication between the groups is relatively easy, with group affiliation and identity remaining clear. This is the linguistic equivalent of having your cake and eating it too. In the example sentence below, word order and morphology are relatively consistent across the three Kupwar variants, but the vocabulary identifies which language is being spoken. Fusion of languages in Kupwar URDU MARATHI KANNADA

pala jəra kaat ke le pala jəra kap un ghe tapla jəra khod i təgond vegetables have made a small cut 'I cut some vegetables and brought them in'.

I have

a a b vem

I passed

yes this is mine

Remarkably, the three grammars were fused by combining grammatical elements from each language, while social differences were (and are) preserved through differences in vocabulary.

Language/dialect continued In contrast to the situation in Kupwar, the Romance languages, which include Spanish, French, Italian and Portuguese, developed distinct national variants of the colloquial Latin spoken in their regions in Roman times. While the language varieties spoken in Kupwar have converged, the language varieties derived from Latin have diverged over the centuries. The reasons are the same in both cases. First, people use language to mark their social identity. Second, people who talk to each other tend to

• 349

3 5 0 • Chapter 1 1 Linguistic differences between social groups: dialects

speak like the other One consequence of the second principle is that people who do not speak to each other tend to be linguistically different. Today you can see the languages of Europe separately and neatly on a map. In reality, they are not so sharply distinguishable. Rather, there is a continuum of variations and languages "merging" into one another. The national border between France and Italy also serves as a dividing line between the French and Italian-speaking countries. But the French spoken within the French border shares characteristics with the Italian spoken outside. From Paris to the Italian border there is a continuum along which French landraces become increasingly "Italian". Likewise, the Italian varieties are becoming more “French” from Rome to the French border. Swedes from the far south use their local dialects to communicate better with Danish speakers in neighboring Denmark than with their compatriots in Sweden's far north. The same situation exists for residents along the border between Germany and the Netherlands. German speakers can communicate better with local Dutch speakers using their own local varieties than with speakers of southern German dialects. Examples of continuous geographic dialects can be found across Europe. Although the standard Italian, French, Spanish, Catalan and Portuguese varieties are not mutually intelligible, the local varieties form a continuum from Portugal to Spain, halfway through Belgium and then France to the southern tip of Italy. There is also a Scandinavian dialect continuum, a West Germanic dialect continuum, and a South Slavic and North Slavic dialect continuum. In the case of Kupwar, if there were no external reference varieties against which to compare the varieties spoken in the village, we might be inclined to say that the varieties spoken there were dialects of one language. However, Kupwar residents find it socially valuable to continue to speak 'different' languages despite increasing grammatical similarity. What counts most in deciding on language variety names and whether such names represent dialects of a single language or separate languages are the opinions of their speakers.

National Varieties of English In this section we briefly examine some national varieties of English, with a focus on American and British English.

American and British National Varieties The major varieties of English around the world are often divided into British and American types. British English is the basis for the varieties spoken in England, Ireland, Wales, Scotland, Australia, New Zealand, India, Pakistan, Malaysia, Singapore and South Africa. American (or North American) primarily includes English people from Canada and the United States. Despite the proposed groupings, certain features of Canadian English are closer to British English, while certain features of Irish English are closer to North American English. And there are many differences between, say, Standard British English and Standard Indian English. But we can still make a number of generalizations about the British and American varieties as long as we recognize that neither group is entirely homogeneous.

Spelling Variants of National English There are well-known spelling differences between British and American English. Some are systematic, others are restricted to a specific word. American red, white, and blue are colors in Britain, and many other words that end in -o in American English end in -our in British English. Some idiosyncratic spellings include British tires and curbs versus American tires and curbs. Interestingly, Canadians often use British rather than American spelling practices, reflecting their close historical association with Britain. For the most part, these spelling differences do not reflect spoken differences. Below are some common American ~ British spelling matches. American

British

north american

British

Work, left pro, defense spelled, burned, analysis spilled, org center, try play, abbreviation checked, act aborted, skillful

Profession, licensed professional, spelled defense, burned, analysis buried, organization center, audition, abbreviation checked, action aborted, skillful

tire brake program pajama check to zar catalogue

pajama tires braking program consult the tsar ton catalogue

Pronunciation Differences in the pronunciation of vowels and consonants, and in the stress and intonation of words combine to form American and British accents. Speakers of both variants pronounce the vowel of words in the class cat, fat, mat with //. For similar words ending in a fricative, such as fast, path, and half, American English has //, while some British variants have /ɑ/, the stressed parent vowel. Americans pronounce the vowels in the class new, tune, and duty with /u/ as if they were spelled "noo," "toon," and "dooty." Variant British English often pronounces them with /ju/ as they are spelled "nyew", "tyune" and "dyuty", a pronunciation also heard among some older Americans. As for the consonants, perhaps the most notable difference has to do with the intervocalic /t/. When /t/ occurs between a stressed and an unstressed vowel, Americans and Canadians often pronounce it as a flap [ɾ]. As a result, the word Sitter is pronounced [sɾər], and Last and Ladder are pronounced the same. In contrast, speakers of some British varieties pronounce the intervocalic /t/ as [t]. As another example, most American varieties have a word-terminating retroflex /r/ in words like car and near, and also precede a consonant as in cart and bart, while some British varieties, including British English Standard, do not have it. In relation to this postvocalic /r/, speakers of Irish and Scottish English follow the American pattern, while speakers of New York City, Boston, and parts of the South Coast dialects follow the British pattern. Among the differences in word stress, British English tends to stress the first syllable of garage, filet, and ballet, while American English stresses the second syllable. The same applies to patois, massage, rubble, beret and other borrowings from French. Stress patterns differ for certain polysyllabic words, such as lab, desk, and sink, with American English retaining a secondary stress on the penultimate syllable.

Try it yourself: Use the IPA symbols on the front and back of this text to transcribe the word laboratory to represent British and American four-syllable pronunciations.

• 351

3 5 2 • Chapter 1 1 Linguistic differences between social groups: dialects

Syntax and Grammar Some noun phrases denoting places in time or space receive an article in American English but not in British English. American

British

the next day in the hospital for college

Some collective nouns (referring to groups of people or institutions) are treated as plural in British English but generally as singular in American varieties. An American watching a football game might say that Cornell has a two-way advantage, while a British observer might say that Manchester has a two-way advantage. Americans trust more than common sense. So when a sportswriter or TV station talks about the Anaheim Angels baseball team, they might say that Anaheim won again or that the Angels won again. In both British and American English, a noun such as policeman takes a plural verb, as in The police are trying to help the neighbors. Another illustration of the grammatical differences between the two varieties is the use of the verb in conjunction with auxiliary verbs. When asked if you've done your homework, American English allows Yes, I'm done, while British English allows this and Yes, I'm done. When asked if flight times to Los Angeles vary, a British Airways flight attendant may respond: It may vary. Vocabulary There are also vocabulary differences between American and British English, as shown below. American

British

north american

British

TV Roof lift (of a car) Gas/petrol Biscuit liner Tea towel/face towel

elevator tv bonnet cookies gas can glue flannel

Second floor Flashlight Trunk (of a car) Dessert cart Rotating area/rotating outlet

First Floor Torch Boots Pudding Truck Interval Roundabout Exit

Try it yourself: In some cases, a word used in Britain is little known in the United States. In other cases, the most common British term is not the most common American term. Please provide the common US English equivalent for each of the following terms: fortnight, holiday, freeway, detour, road works, intersections, rear, rental car, parking lot, windshield, wrench.

Regional Varieties of American English Beginning in the late 1940's, research into vocabulary patterns in the Eastern United States suggested Northern, Central, and Southern dialects. Midland was divided into North Midland and South Midland varieties. Boston and the New York metro area were

Regional variants of American English that are considered distinct variants of the northern dialect. Midwestern states such as Illinois, Indiana, and Ohio, previously considered representative of "North America", were considered primarily in the North Midland dialect, with a narrow band of the northern dialect in their most distant counties. Belonging to the South Midland variety in its southern counties. Recent research suggests refinements to this scheme, such as those illustrated by the geographic patterns in Figure 11-1.

Figure 11-1 Major US dialectic regions, by vocabulary

Which: Carver 1987.

Dialect Mapping To arrive at a map like the one shown in Figure 11-1, dialectologists study usage patterns. Investigators can rely on vocabulary, pronunciation, or grammar. Typically, a researcher visits a city with a lengthy questionnaire and asks residents what they call certain things or how they express certain meanings. Figure 11-1 is based on the regional vocabulary. Note that this interpretation of the data divides the United States into two major dialects (Northern and Southern), each further divided into upper and lower sections, with Western seen as an extension of the dialect. The map in Figure 11-1 is based on field research done for the Dictionary of American Regional English (DARE) in the 1960s and 1970s. We'll see this project a little later. Several language atlas projects were conducted prior to DARE, which were part of a project called the Language Atlas of the United States and Canada. Data collection in several regions has been completed and the results have been published, but parts of the project are still incomplete. Nonetheless, the data collected provide a useful overview of regional differences. take something

• 353

3 5 4 • Chapter 1 1 Linguistic differences between social groups: dialects

For example, when Atlas researchers asked respondents the common term for the large transparent-winged insect that often swims above water, local terms came to the fore. Figure 11-2 shows darning needle as the most common term in New England, New York State, the New York metropolitan area (which includes northern and eastern New Jersey and Long Island), and northern New York Pennsylvania. Elsewhere, other terms prevailed: Mosquito Hawk off the coast of North Carolina and Virginia, Snake Doctor in inland Virginia, and Snake Feeder along the northern Ohio River in West Virginia, heading toward Ohio, western Pennsylvania and the upper reaches of the Ohio Valley Pittsburgh.

Figure 11-2 Words for "dragonfly" in the eastern states

Quelle: Kurath 1949.

Regional variants of American English You can see from Figure 11.2 that not all terms for "dragonfly" are well distributed. In some areas there was only one form, in others more than one. The larger O's on the map of New England indicate that darning needle was the only regional term found there. You can see in Figures 11-3 and 11-4 that Mosquito Hawk was pretty much the only regional response in parts of southeast Texas and parts of central Texas, as well as in Louisiana and Florida, and much of the United States. Southern Alabama, Mississippi and Georgia. But the snake doctor was the preferred form in west, north, and northwest Texas, western half of Tennessee, northern Alabama and Mississippi, and parts of northwest Georgia. Snakeeaters were occasionally found in Oklahoma along the Canadian and Arkansas Rivers (which are not identified in our figure but can be identified in Oklahoma near the filled triangles in Figure 11.4). Both the hawk midge and the snake doctor have been used in the southern half of Arkansas (in Figures 11.3 and 11.4). The darning needle, so popular in New York and New England, was a rare occurrence to register on these southern charts. Some respondents were unaware of local terms and reported using only Libelle. (Note: If you live in or are from an area shown on the maps but are unfamiliar with the terms shown there, remember that the data was often collected from rural areas and included both "vernacular" and " vernacular" represent .cultured." Some of the interviews were made decades ago and the usage of the word may have changed by now).

Figure 11.3 Words for “Dragonfly” in Texas, Arkansas, Louisiana, Oklahoma

Arkansas

Texas

New Mexico

dragonfly mosquito hawk snake doctor snake feeder

Which: Atwood 1962.

• 355

3 5 6 • Chapter 1 1 Linguistic differences between social groups: dialects

Figure 11-4 Words for "Dragonfly" in the Gulf States

Tennessee

Arkansas

Georgia

Alabama

Texas

Florida

dragonfly snake doctor snake feeder mosquito hawk

Those: Pederson 1986.

Determining Isoglosses After a map has been marked with symbols for different features, lines called isoglosses can often be drawn at the boundaries of the different shapes. For example, in Figure 11-5 on page 357, the four isoglosses running through the north-central states of Ohio, Indiana, and Illinois represent the northernmost boundaries of greasy /grizi/, pronounced /z/, for fodder snake term for "dragonfly" and two more features. Figure 11-6 on page 358 shows seven isoglosses in the Upper Midwest. Three of them mark the southernmost boundaries of the northern features: Mood pronounced [hjumər] (/hj/ is represented as /hy/ in the map legend); Boulevard refers to the strip of grass between the curb and sidewalk; and enter (fresh), meaning "to give birth," usually referring to a cow. The other four isoglosses mark the northernmost limits of Midland features: the word on is pronounced with a rounded vowel ( /ɔ/ or /ɒ/ , where /ɒ/ is pronounced like /ɑ/ but with lip rounding) instead of an unrounded vowel vowel /ɑ/; the term Caterwampus, meaning "crooked" or "wrong"; the term roasted ears for "corn on the cob"; and light bread for "white bread".

Dialect Boundaries Imagine each isogloss map stacked on top of each other on a slide. The result would be a map similar to Figure 11.6 showing how far

Regional variants of American English isoglosses on different maps are "packed". The geographic limit for using a particular word (e.g. Caterwampus) is usually roughly the same as the limit for other terms or pronunciations. Where isoglosses cluster, dialectologists draw dialectal boundaries. Hence a dialect boundary is simply the locus of a set of isoglosses. The map in Figure 11-1 on page 323 is a composite of dozens of maps similar to those in Figures 11-5 and 11-6.

FIGURE 11-5 Four isoglosses in the north central states (northern borders)

big snake eater sook (calls the cows) sugar tree

Wisconsin

ILLINOIS

Virginia Occidental

Kentucky

Tennessee

Source: Markwardt 1957.

• 357

3 5 8 • Chapter 1 1 Linguistic differences between social groups: dialects

Figure 11.6 Seven isoglosses in the Upper Midwest

Borders North South / Hy / In the mood of the boulevard enters (fresh)

NORTH DAKOTA

Minnesota

Midland northern borders /W-Å/ on roasted caterwampus ears on light bread

Dakota del Sur

THIS IS

Source: Allen 1973.

Language patterns are influenced in part by geographic and physical boundaries that facilitate or hinder communication, and in part by the migratory routes one follows to settle in a place. Among the isoglosses in Figure 11-5, that of /grisi/ versus /grizi/ essentially follows a line (now approximated by Interstate 70) that was the main path for pioneer migration during the post-colonial era. In the western United States, the dialect situation is more complex than in the older areas of the East, South, and Midwest. The west attracted settlers who spoke dialects from different parts of the country. California continues to receive immigrants from other parts of the country and the world.

Dictionary of American Regional English The Dictionary of American Regional English provides more information than ever about regional words and phrases in the United States.

Regional Varieties of American English Based on answers to more than 1,800 questions asked by field researchers who toured 1,002 communities across the country, the maps used to present DARE's findings do not represent geographic space like the most maps, but the population density. Therefore, the largest states on a DARE map are those with the largest populations. As a result, the DARE maps represent states in an unknown way, as shown by the comparison in Figure 11-7. Figure 11-8 on page 360 shows the distribution of the terms Mosquito Hawk and Skeeter Hawk on a DARE chart and on a traditional chart. Along with occasional occurrences in California and New Mexico, you can see the distribution of these terms in FIGURE 11-7. Comparison of the DARE map and the conventional map with state names.

Fonte: Dictionary of Regional American English, I, 1985.

• 359

3 6 0 • Chapter 1 1 Linguistic variation between social groups: dialects

Figure 11-8 Distribution of Mosquito Hawk and Skeeter Hawk on DARE chart and conventional chart

Fonte: Dictionary of Regional American English, I, 1985.

Regional Varieties of American English FIGURE 11-9 Cruller distribution on a DARE chart

Fonte: Dictionary of Regional American English, I, 1985.

Gulf States and as far east as the East Coast, and occurs occasionally in Minnesota, Wisconsin, Michigan, and a few other states. The word cruller, used for "a crooked donut," has a very different usage, as shown in Figure 11.9. Cruller is used in the Northeast, New England, New York, New Jersey, Pennsylvania, etc., and some Great Lake states and California, but is not found in Alaska, Washington, Oregon, Nevada, New Mexico, and Hawaii. As a result of various regional dialect projects, most notably DARE, a complex picture of American English dialects emerges, as shown in Figure 11-1 (page 353). The darker the shading of a dialect area in this figure, the greater the number of vocabulary items that distinguish it from other dialect areas. As you can see, the further West you go, the less special vocabulary you will find. Judging by the vocabulary, the boundaries of American dialects are better established in the eastern states than in the younger western states. Based on the DARE vocabulary results, the United States appears to have primarily northern and southern dialects, each divided into upper and lower regions, as shown in Figure 11-1. The Upper North contains dialects of New England, the Upper Midwest,

• 361

3 6 2 • Chapter 1 1 Linguistic differences between social groups: dialects

and the Northwest, with some less pronounced dialect boundaries in the Midwest and Northern California. The Southwest is also a dialect area, and Southern California has some distinctive features. The south is divided into Alto Sur and Bajo Sur, each with subsidiaries.

The Atlas of North American English In the 1990s, extensive research was conducted into pronunciation in urban areas of the United States and Canada. The Atlas of North American English or ANAE is not related to the Linguistic Atlas of North American English and Canada or dare. In terms of its goals, methods and results, it is an independent project. ANAE was created using data from a telephone survey of North American city centers in a project called Telsur. Based on telephone conversations with respondents who identified themselves as being born or raised in the language community in which they were contacted, Telsur combined impressionistic pronunciation judgments with rigorous acoustic analysis of recorded conversations. Telsur and ANAE focused on vowels, specifically several vowel pronunciations that are known to change.

Vowel amalgamations Notable changes in North American pronunciation include previously separate vowel amalgamations: /ɑ/ and /ɔ/ in words like cot and catch, and // and /ε/ in words like pin and pen. To distinguish these last two elements, many speakers call the first a pin or safety pin and the second an ink stick. Fusion Cot ~ Caught The traditional pronunciations of cot and catch differ in that the former has the head /ɑ/ and the latter /ɔ/. Because /ɑ/ is a lower back vowel and /ɔ/ is a lower middle back vowel, the merging is often referred to as a lumbar merging. These are word pairs like don and dawn, wok and walk, and hock and hawk. For many American English speakers who do not mix these vowels, /ɑ/ and /ɔ/ are distinct phonemes, and for these speakers all of these word pairs are minimal pairs (discussed in Chapter 4). The merging of these two phonemes reduces the number of vowels in the English stock, which can lead to many pairs of homophones. Pin ~ Pen merger Another merger involves vowels in word pairs, such as pin ~ pen, him ~ hem, lint ~ lent, and cinder ~ sender. For many speakers these vowels remain distinct as [] and [ε], but for many others they are homophones and cannot be distinguished in speech. This merging is sometimes referred to as an IN~EN merging. Conditional and unconditional mergers Merging of /ɑ/ and /ɔ/ is not limited to specific phonological attitudes within a word, but occurs everywhere. This unconditional merging affects all words containing the sounds, causing the loss of vowel contrast. In contrast, the vowels // and /ε/ merge only when they precede the nasal /n/ or /m/, but nowhere else. Therefore, speakers who pronounce pin and pen identically do not mix pit and pet ([pt] vs. [pεt]), lit and let, whipped and cry, etc., because these words do not fit the specified phonological environment. necessary for the merger.

The Atlas of American English We can summarize the discussion of mergers in the box below. Surname

Vocal

illness

examples

Cradle ~ Melting Pen Captured ~ Feather Melting

/ɑ/ ~ /ɔ/ // ~ /ε/

unconditional precedent /n/ or /m/

cradle ~ captive hock ~ hawk pin ~ feather, gray ~ sender

Vowel Changes Other important changes in American English involve changing the pronunciation of vowels from one place in the mouth to another. The effect is that strangers hear a word pronounced with a specific vowel as if it had a different vowel. For example, the word cod can be heard as cad. You know that the vowels can be plotted on a chart like the one shown in Figure 3-4 on page 90. In addition to the individual vowels in the figure, English has three diphthongs: /aj/ (my, line), /ɔj/ (toy, coin), and /aw/ (cow, town). (In this book, we generally represent other English vowels as single or monophthous vowels. Therefore, we represent the underlying vowel of made as /e/, flowed as /o/, and food as /u / represent. These vowels are often pronounced as diphthongs and are represented as diphthongs in some other books that indicate their underlying forms, such as /ey/, /ow/, and /uw/). Northern cities change major northern cities including Syracuse, Rochester and Buffalo in New York, Cleveland and Akron in Ohio, Detroit in Michigan, Chicago and Rockford in Illinois, and Milwaukee and Madison in Wisconsin: a series of vowel changes that are notable is within its scope. They represent the compensation of the northern cities and can be represented as in Figure 11-10. This change includes Canada as a change in northern cities

Source: Adapted from Labov 1996.

• 363

3 6 4 • Chapter 1 1 Linguistic differences between social groups: dialects

as well as US cities.It has various aspects including the following dates. (The numbers in parentheses refer to the numbered layers in Figure 11.10; for the sake of simplicity we have not included layer number 6 in our list.) 1. // rises and precedes [iə] – one and bad can even like the stressed vowel in the idea sound: [miən], [biəd] (1 in the figure) 2. /ɑ/ precedes [] – cod sounds like cod (2 in the figure) 3. /ɔ/ is lower and before [ɑ]: Squawk sounds like cod (3 in figure) 4. /ε/ is lower and centered on [ ] : Ked sounds like cud (4 in figure) 5. // is backwards in [ɔ] – sounds cud like quack (5 in the figure)

Southern Shift A different series of vowel shifts takes place in the south. They form the south fault and can be represented as in Figure 11-11. Southern Shift has several aspects, including the five listed below, using italicized words as examples. (The parentheses refer to the numbered turns in Figure 11.11; for simplicity, we have not included turns 3, 7, or 8 in our list.) 1. 2. 3. 4.

/aj/ is monotonized to [a] - hidden sounds like [had] or [had] (1 in figure) /e/ is lowered and centered on [aj] - slade sounds like slide (2 in figure) / o / is in front – code and boot sound like [kεod] and [bεot] (6 in picture) //, /ε/, // are raised and in front – kid sounds like keyed, ted like tid, pat like pet (4 pictured) 5. /u/ is frontal – great sounds like “kewl” (5 pictured)

FIGURE 11-11 South offset

Source: Adapted from Labov 1996.

Atlas of North American English

ANAE Results Based on 439 telephone interviewees for whom acoustic analysis was performed, the Atlas of North American English (ANAE) provides a map of the United States and Canada where new dialect boundaries are proposed. You can get a clear picture of these results on the ANAE website. Meanwhile, the map in Figure 11.12 shows the major dialect regions of North America based on Telsur pronunciation data. On the map you can see that, apart from Canada, there are four main pronunciation regions of the USA: West, North, Midland and South. To the north are the Inland North and Western New England dialects, and to the south are the Texas South and Inland South dialects. You'll also notice names for dialects called Mid-Atlantic, New York City (NYC), Eastern New England (ENE), Western Pennsylvania, and others. Below is a table, adapted from the ANAE website, giving some salient features of the pronunciation of some dialects. As shown in Figure 11.12, we give features of each region as a whole and sometimes of dialects within the region. NORTH

Inland New England Northwest

Less facade of /o/ than other areas Change in northern cities Less advanced change in northern cities

Figure 11.12 Urban dialect areas in the United States, by pronunciation

Source: Adapted from Labov 1996.

• 365

3 6 6 • Chapter 1 1 Linguistic differences between social groups: SOUTHERN dialects

Interior South Texas SOUTH MIDLAND WESTERN CANADA

Atlantic provinces

Monophthongization of /aj/ South Shift South Shift Transient Back Fusion Fronting of /o/ Low Back Fusion (cot ~ catch) Fronting of /u/ stronger than /o/ Low Back Fusion (cot ~ catch) No Back Fusion short

Additional data appears on the map but is not accessible enough for further discussion in this black and white image. Among the information you can glean from the color maps on the website is the fact that St. Louis falls within the Midland region but still has the Northern Cities Offset. You may also notice that almost all of Florida is outside of the southern region. This is because Florida does not participate in the south drift, although it does show the front of /u/ (step 5 in Figure 11-11) but not /o/ (step 6 in Figure 11-11). In Canada, the Atlantic provinces have a conservative pronunciation and do not (yet) participate in the changes characteristic of other parts of Canada, such as B. lumbar fusion (cot ~ catch).

Ethnic Varieties of American English Just as oceans and mountains can separate people and eventually lead to different language patterns, so too do social boundaries encourage different ways of speaking. Perhaps the most notable social variants of American English are the ethnic variants. Ethnicity is sometimes racial and sometimes not. For example, differences in the language of Jewish and Italian New Yorkers have been noted, and the variety of English influenced by Yiddish speakers who settled in America is sometimes referred to as "Yinglish". But the social division that leads to ethnic language variants is particularly noticeable in the distinctive language patterns of urban African Americans. In Philadelphia and other cities, the language of African American residents is becoming increasingly different from that of white residents. This distinction between social groups is also reflected in the language patterns characteristic of other ethnic groups. Spanish-speaking immigrants in Los Angeles, New York, Chicago, Miami, and elsewhere have learned English as a second language, with their English marked by a foreign accent. The children and grandchildren of these immigrants acquire English as their first language (and many are bilingual), but the native variety of English spoken by many Hispanics identifies them as being of Hispanic ancestry or growing up in neighborhoods with children of Hispanic descent. The following discussion identifies certain characteristics of African American and Chicano English. Both are true variants of American English like any other regional or social variant. Both have complete grammatical systems that overlap heavily with other variants of English. And like Standard American English, both have a range of registers. While both varieties share many traits with other varieties of American English, they also share certain distinctive traits and a number of common traits that together set them apart from all others. Like all other social variants, these two have rules that determine what is well-formed and what is ill-formed. The rules govern the structures and use of all dialects and no dialect

The ethnic variants of American English exist without phonological, morphological, and syntactic rules. All world languages described in Chapter 7 also apply to African American English and Chicano English.

African American English Not all African Americans are fluent in African American English, and not all African Americans who speak English are African American. After all, people grow up with the diversity of languages that surround them. In an ethnically diverse city like Los Angeles, you might meet English-speaking African-American teenagers whose foreign-born parents speak Chinese or Vietnamese. The diversity of English spoken by these Asian American teenagers reflects the different ways in which their friends speak and the neighborhoods in which they learned English. To emphasize an obvious but often misunderstood fact, the acquisition of a particular language or dialect is as independent of skin color as it is of height or weight. The history of African American English is not fully understood, and there are competing theories about its origins and further development. But there is no disagreement about its structure and functioning. It has distinctive phonological, morphological, and syntactic features, as well as its own vocabulary. Like all other social groups, African American English speakers share different styles of interaction. In this section, we examine some phonological and syntactic features of African-American English, but not lexical or interactional features. Phonological features We examined four pronunciation features of African American English (AAE). 1. Simplification of Consonant Clusters In AAE, consonant clusters are often simplified. Typical examples are found in the words desk, pronounced "des" [dεs], past pronounced "pass" [ps], and wild pronounced "wile" [wajl]. Consonant cluster simplification also occurs in all other variants of American English. Among Standard English speakers, the consonant clusters ⬍sk⬎ in ask and ⬍ld⬎ in wild are also often simplified, such as asking in "asthem" [sðəm] and saying "tole" [tol]. However, consonant cluster simplification occurs more frequently and to a greater extent in African-American English than in other varieties. 2. Omission of ending consonants In AAE, ending consonants such as /d/ can be omitted in words such as sidewords and loanwords. AAE speakers often exclude some end stops, pronouncing lado as a sigh and lent as lent. This exclusion rule is systematically influenced by the phonological and grammatical environment: a. Whether a word-final consonant represents a separate morpheme (as in past tense marked pursued and tested) or does not represent a separate morpheme but is part of the word stem (as in lateral and fast). Final [d] is conserved much more often when it is a separate morpheme. B. Regardless of whether the last word ends on a heavily stressed syllable (try) or a weakly stressed syllable (rapid), note that the second syllable of rapid is not as stressed as the first syllable. Heavily stressed syllables tend to retain endings more than weakly stressed syllables. C. Whether the stop is followed by a vowel (as in Angular Lateral and Treaty) or by a consonant (as in Strong Treaty and Calle Lateral). A following vowel helps preserve the stop; In fact, this seems to be the most important factor in deciding whether to rule out a definitive cap.

• 367

3 6 8 • Chapter 1 1 Linguistic differences between social groups: dialects

3. Interdental ~ labiodental substitution Other phonological features are less common. For some AAE speakers, the th of words like both, with, and Bethlehem may not be perceived as the voiceless interdental fricative /θ/, but rather as the voiceless labiodental fricative /f/ that produces [bof] or [wf]. for example. Likewise, the voiced interdental fricative /ð/ in words like soften or bathe and bruder or mother can be realized with the voiced labiodental fricative /v/, giving rise to [smuv], [bev], [brvə] and [mvə]. Also note the lack of the word ending /r/ in brother and mother, a feature AAE shares with English in New York City, eastern New England, and parts of the South Coast. 4. Aunt and ask Two other pronunciations of AAE are often seen. The first is that the initial vowel of tia and tia is pronounced /ɑ/, a pronunciation also characteristic of eastern New England but not most other American dialects that have //. The second is the pronunciation of ask as [ks] instead of [sk]. This pronunciation is not unique to AAE, but it is a trait that has been stereotyped and stigmatized.

When examining ongoing changes in American English pronunciation, researchers were surprised to find that African Americans living in cities affected by the Northern Cities Change do not appear to be participating. This is a clue that leads some observers to conclude that AAE and American Standard English differ rather than resemble each other. Grammatical Features We examined four grammatical features of African American English. 1. Elimination of the Copula Below, compare the use of the copula, the verb BE, in African American English and Standard American English. Sentences 1 and 2 illustrate that AAE allows the present tense to be excluded precisely where Standard English allows for a contracted form of the copula. AFRICAN AMERICANS

AMERICAN STANDARD

1. This is my bike. 2. Cold coffee. 3. The coffee is cold there.

this is my bike The coffee is cold. The coffee is (always) cold there.

2. Being As shown in example 3 above, AAE speakers express recurring or habitual actions with the form be. For speakers of other varieties, AAE may be equivalent to American Standard English is. In fact, in sentences like 3, it is equivalent to a verb expressing a habitual or continuous state of affairs. As African-American linguist Geneva Smitherman wrote of sentences like 2 and 3: "If you're the cook and the coffee is cold, you may be the only talk of the day, but if the coffee is cold, then you won't be there." " . unemployed!" Thus, the verb be (or its variant bees) is used to indicate a continuous, repeated, or habitual action. The following examples further illustrate this trait. AFRICAN AMERICAN

AMERICAN STANDARD

Do they play all day? Yes, the boys play a lot. I see her when I'm on my way to school.

Do they play all day? Yes, the boys move a lot. I see her when I'm on my way to school.

Ethnic Varieties of American English 3. Existential it Another characteristic of African American English is the use of the phrase where Standard American English uses it, as a New Orleans resident reported after Hurricane Katrina: There is nothing left. Here are two more examples of Existential: AFRICAN AMERICAN

AMERICAN STANDARD

Is there a Miss Jones in this office? She was a wonderful wife and that's not good for her.

Is there a Miss Jones in this office? She was a wonderful woman and there is nothing too good for her.

4. Negative Match The following examples of what is technically called a negative match, better known as double negative or multiple negative, provide a final illustration of the distinction of this ethnic variant: AFRICAN AMERICAN

AMERICAN STANDARD

Nobody helps me with my work. He's not going anywhere.

Sentences in African American English contain more than one word marked for negative. Several negative constructions are well-formed in AAE, as in many other variants of American English and more generally in earlier periods of English. The fact that these constructions are not well-formed in current Standard English does not affect their grammaticality or appropriateness in other varieties.

Chicano English Another important group of ethnic dialects of American English is called Latino English or Hispanic English. The most well-known variant is Chicano-English, spoken by many people of Mexican descent in the major urban centers of the United States and in rural areas of the Southwest. As with African-American English and all other variants of English, certain characteristics of Chicano English are shared with other variants, including other variants of Hispanic English as spoken in the Cuban community of Miami and the Puerto Rican community of New York City. . Chicano English includes many registers for use in different situations. Undoubtedly, some characteristics result from the persistence of Spanish as one of the language variants of the Hispanic American community, but Chicano-English has become a different variant of American English and cannot be considered English with a foreign accent. Many children acquire it as their first language and it is the mother tongue of hundreds of thousands of adults. As such, it is a stable variant of American English, with distinctive grammar and pronunciation patterns. Phonological features A well-known phonological feature of Chicano-English is the replacement of "ch" [tʃ] with "sh" [ʃ], as in the pronunciation of she as [tʃi] instead of [ʃi] and shoes as [tʃuz] ( homophone with choose) instead of [ʃuz] and specifically as [εspεtʃəli]. This trait is so distinctive that it has become a stereotype. There is also the substitution of "sh" for "ch", as in "preash" [priʃ] for predicate and "shek" [ʃεk] for verify [tʃεk], although this feature does not seem to be stereotyped. Other phonological features of Chicano-English include the simplification of consonant clusters, as in [s] for it's, "kine" for kind,

• 369

3 7 0 • Chapter 1 1 Linguistic differences between social groups: dialects

"ole" for old, "bes" for better, "un-erstan" [nərstn] for understanding. Much of this can be represented in the phrase It's kind of hard, which is pronounced [s kɑnə hɑr]. Another important feature of the Chicano-English phonological system is the deafening /z/, especially at the end of words. Because of the widespread occurrence of /z/ in English inflectional morphology (in plural nouns, possessive nouns, and present 3rd person singular verbs such as going), this prominent feature is also stereotyped. Chicano English pronunciation is also characterized by the fact that the standard fricatives represented by th in the spelling are replaced by plosives: [t] for [θ] and [d] for [ð], as in [th k] for thick and [dεn ] for then. Another notable feature is the pronunciation of verbal -ing as "een" [in] instead of /n/ ([ən]) or /ŋ/. Other words in -ng like sing and long end with a velar nasal combination /ŋ/ and a velar plosive /g/; therefore singing is pronounced [sŋg], not [sŋ], and long is [lɔŋg] instead of [lɔŋ]. Another distinctive feature of Chicano English is its use of certain intonation patterns that may seem uncertain or hesitant to speakers of other dialects of American English. Like AAE speakers, speakers of Spanish variants of English living in cities affected by the Northern Cities Shift do not appear to participate in these changes, at least not to the same extent as other groups. Grammatical features Chicano-English also has distinctive syntactic patterns. Often omits the past tense marker on verbs ending in the alveoli /t/, /d/, or /n/, producing "wan" for desire and "wait" for waiting. At least also in Los Angeles. . . or sometimes it is heard instead of someone. . . or, as in I'll buy one, or Terry will. Another feature is the use of dialect-specific prepositions, such as B. out from for away from, as in They party, to get out of their problems. As with many other variants, Chicano English allows multiple negations, as in You owe me nothing and We Little People Receive Nothing.

Ethnic Varieties and Social Identification It is important to reiterate that some commonalities between Chicano English and African American English are characteristics of other varieties of American English. In some cases, such as B. the simplification of consonant clusters, these features are common in mainstream varieties, including Standard English. In other cases, such as B. the negative agreement, they are not characteristic of American Standard English, but are shared with other non-standard varieties. What makes a variety distinctive is not a single characteristic, but a set of characteristics, some of which may also be present in other varieties. Ethnic dialects are an important part of social identity, and traits that are recognized as characteristics of particular social groups can be used to promote or reinforce belonging to that identity. When speaking, an African American man or woman who wishes to emphasize their social identity as African American may emphasize or exaggerate characteristics of African American English. The same is true for speakers of Hispanic varieties of English who wish to emphasize their Hispanic identity. News correspondents on English-language radio and television programs generally speak without strong social group accents. However, to emphasize their ethnic identity, some correspondents use an accented ethnic pronunciation of their own names at the end of a report. A reporter named María Hinojosa self-identifies as mah-REE-ah ee-noh-HOH-sah, with a trill /6/ in REE. Geraldo Rivera pronounces his first name heh-RAHL-doh. Such ethnically influenced pronunciations underscore a reporter's pride in his or her ethnic identity.

Socioeconomic Level Varieties: English, French, and Spanish

Try it yourself: Consider these Spanish name pronunciations: "deh-lahCROOS" for de la Cruz; "FWEHN-tehs" for sources; "GAHR-sah" for Garza and "ehr-NAHN-dehs" for Hernandez. Say these names as loudly as you think they would be pronounced without ethnic pronunciation. Compare these pronunciations with those in quotation marks and identify two features in Spanish pronunciation that are characteristic of Chicano English. Identify two other characteristics that we did not discuss but that you think may reflect characteristics of Hispanic English.

Socioeconomic Varieties: English, French, and Spanish Less impressive than the regional and ethnic variants, but just as important, are the striking language patterns that characterize different socioeconomic groups. Here we describe some speech patterns of English spoken in New York City and Norwich, England, as well as French from Montreal and Spanish from Argentina.

New Yorkers sometimes pronounce /r/ and sometimes drop it in words such as car and beer, wagon and room (where /r/ follows a vowel in the same syllable and appears at the end of the word or precedes another consonant). The presence or absence of this /r/ does not change the referential meaning of a word. A "Cah Pahked" in a red zone is punished just as surely as a similarly parked car. And if you live in New York or "New Yoahk", you have the same mayor (or "Maya"). However, the occurrence of /r/ in these words is anything but random and meaningless. Linguist William Labov hypothesized that the pronunciation of /r/ in New York City depends on social class, and that two groups of social rank New Yorkers differ in their pronunciation of /r/. Based on some preliminary observations, he predicted that members of higher socioeconomic groups would pronounce /r/ more often than speakers of lower socioeconomic groups. To test his hypothesis, Labov examined the discourse of clerks in three different social classes of Manhattan department stores: Saks Fifth Avenue, an expensive upper-middle-class department store; Macy's, a mid-range, mid-priced store; and S. Klein, a discount store frequented primarily by working-class New Yorkers. He questioned supervisors, clerks, and clerks about the whereabouts of merchandise he knew was on display on the fourth floor of the store. In response to a question like "Where can I find the lightbulbs?" got an answer from the fourth floor. He then pretended not to understand the answer and said, "Excuse me?", provoking a repeated and more cautious testimony from the fourth floor. Thus, each employee had the opportunity to pronounce the post-vowel /r/ four times (twice in the fourth and first) in a natural and realistic setting where the speech itself was not the focus. Employees at Saks, the top-rated store, say /r/ more often than employees at S. Klein, the lowest-rated store. At Macy's, the middle store, the staff would pronounce a middle number. Figure 11-13 on page 372 shows the results of

• 371

3 7 2 • Chapter 1 1 Linguistic differences between social groups: dialects

FIGURE 11-13 General stratification of /r/ by store in New York City 70

percentage of respondents

Some /r/ All /r/

40 30 30 20 20

bags

macys

4 S. Small

125

10 0 N= Sources: Labov 1966.

Labov research. The darker sections represent the percentage of employees who pronounced /r/ four times; The lighter sections over the darker areas represent the percentage who said it once, twice, or three times (but not four times). Employees who do not pronounce /r/ are not directly represented on the bar chart. As can be seen, 30% of Saks employees all pronounced /r/ and another 32% partially pronounced /r/. At Macy's, 20% said /r/ four times and another 31% said /r/ multiple times. At S. Klein, only 4% of employees pronounced all /r/s, a further 17% pronounced one, two or three /r/s. Labov's hypothesis about the social stratification of the postvowel /r/ seemed surprisingly confirmed. Can you suggest other possible explanations for these results, as factors other than socioeconomic status might have influenced the results, as Labov acknowledged? For example, if you've spoken to more men than women in a store, or more clerks than clerks, or more African Americans than whites, the difference in pronunciation of /r/ could be due to differences in gender, occupation, or ethnicity. . . It found that there were more salespeople in white stores than any other group, and analyzing their pronunciation separately from everyone else's would eliminate the possibility of the results being skewed by gender, occupation, or ethnicity. Figure 11-14 shows a general distribution pattern similar to that of the entire respondent sample. The white clerks at Saks pronounced /r/ more than the women at Macy's, who in turn pronounced it more than the women at S. Klein. Therefore, Labov was able to rule out the possibility that his results reflected ethnic, gender, or professional differences in the business. After the department store study, Labov conducted a different type of investigation. Equipped with detailed sociological descriptions of individual residents of Manhattan's Lower East Side, he spent several hours with each of the approximately 100 respondents there and recorded their conversations. As we discussed in Chapter 10, his interview techniques caused respondents to use speech samples characteristic of different speech situations or pitches. Here are six variables he examined:

Variants of Socioeconomic Level: English, French, and Spanish FIGURE 11-14 Stratification of /r/ by store in New York City Caucasian saleswoman 70 34

percentage of respondents

31 50

Some /r/ All /r/

40 30

33 28

10 0 N=

bags

macys

4 S. Small

Sixty-five

Sources: Labov 1966.

• postvocalic /r/ • th in words like thirty, through and with (New Yorkers sometimes say thirty with /θ/ and sometimes with /t/) • th in words like this, they and breathe (the infamous “dis”, “ dat words", "dem" and "dose", with variants /d/ and /ð/) • Alternative pronunciation of -ING words like Running and Talk, with variants /ŋ/ and / n/ ( a Often referred to as “drop the g,” you know from Chapter 3 that the change is between velar /ŋ/ and alveolar /n/—only in the spelling is there a “g” for fall.) • Pronunciation of vowels into a class from the word brown, soft, catch • Pronunciation of the vowel in the word class bad, care, sagging

In interviews, Labov spoke to women and men, fathers and children, African Americans and whites, Jews and Italians, a cross-section of Lower East Side residents. Based on extensive background information, he assigned each respondent to a socioeconomic class based on a combination of three factors: • the respondent's education • the respondent's family income • the occupation of the main breadwinner in the household

Using these criteria, he classified people into one of four socioeconomic status categories, which he called lower class, working class, lower middle class, and upper middle class. As expected, and as Figure 11-15 on page 374 shows, Upper Middle Class (UMC) respondents showed more /ŋ/ than Lower Middle Class (LMC) respondents, who in turn showed more than Working Class (WC) respondents, who used it more often than lower class (LC) respondents. Each group also pronounced more /ŋ/ as the attention paid to the language in different styles increased. Through different graded pitches - casual

• 373

3 7 4 • Chapter 1 1 Linguistic differences between social groups: dialects

Figure 11-15 Percentage of the suffix -ing pronounced /IN/ by four socioeconomic groups in New York City 99

100

100 95

90 78

70 60

casual reading

50 40 30 20

100

WC LMC Socioeconomic Status Groups

UMC

Fonte two cubes: Labov 1966.

Style, interview style, and reading style: Respondents from all socioeconomic groups increased the percentage of pronouncing /ŋ/. The interview style is not shown here. Labov found that all six variables were socially stratified. Each group at the socioeconomic level had distinctive pronunciation patterns, and the percentage of pronunciation variants was ranked the same as the groups themselves. The upper-middle class spoke more /θ/ for th (as in thing), more /ð/ for th (as in then), more /ŋ/ (as in running) and more /r/ (as in car). Lower-class respondents expressed fewer of these variants, while lower-middle-class and working-class respondents fell somewhere in the middle, with lower-middle class expressing more than working-class. These regular patterns of variation suggest that even subtle differences in social class can be reflected in language use. The vowels were layered in a similar way. New Yorkers have different pronunciations of the first vowel in coffee, ranging from high back vowel [u] to middle back vowel [ɔ] to low back vowel [ɑ]. The vowel of the bad class words also varies, from relaxed low front [] to tense high front [iə] with a slip, as we saw in our discussion of Northern Cities in Transit. In New York City, higher socioeconomic groups preferred low vowels in both cases.

Norwich, England To see if the kind of linguistic differentiation found in New York City exists elsewhere, British linguist Peter Trudgill studied the speech patterns of Norwegian residents.

SES varieties: English, French and Spanish, meaning England, and found strikingly similar results in syntactic and phonological variation. Trudgill divided respondents into five groups: middle middle class (MMC), lower middle class (LMC), upper working class (UWC), middle working class (MWC), and lower working class (LWC). Figure 11.16 illustrates the distribution of the final /ŋ/ on the suffix -ing among these colloquial and reading style groups.

FIGURE 11-16 Percentage of suffix -ing pronounced as /IN/ by five socioeconomic groups in Norwich, England 100

100

72 occasional reading

40 20 0

0 TYP

Socioeconomic Level Groups UWC LMC MWC

CMM

Data source: Trudgill 2000.

Comparing the data from New York City (Figure 11.15 on page 350) and Norwich (Figure 11.16) shows that the distribution patterns of socioeconomic status are similar in the two cities. Each successively higher socioeconomic status group pronounces /ŋ/ more often than the group immediately below.

Montreal, Canada In Montreal, French speakers vary the pronunciation of pronouns and definite articles. Except in the word le, /l/ is sometimes pronounced and sometimes omitted in personal pronouns such as il 'he' and elle 'she' and articles (and pronouns) such as les 'the (plural)' and la 'the (feminine)'. ).' (See Table 2-11, page 59.) Using two occupational groups, skilled and manual workers, manual workers consistently dropped /l/ more often than skilled workers, as shown for four of these words in Figure 11-17 on page 376.

Spanish speakers in Argentina show similar patterns of phonological variation. To cite an example in Argentina, speakers sometimes drop /s/ before pauses (like in English, /s/ is a common word in Spanish, occurring in plural nouns and in various verb forms). In a study of six Argentine occupational groups, the percentage of exclusion of /s/ was highest in the lowest-status occupations and lowest in the highest-status occupations, as shown in Figure 11-18 on page 376.

• 375

3 7 6 • Chapter 1 1 Linguistic differences between social groups: dialects

FIGURE 11-17 Percentage deletion of /l/ in Montreal French for two occupations 100

100 90

professional workers

50 38

40 30

100

People)

is he around

The pronoun)

there (pronoun)

Data source: Sankoff and Cedergren 1971.

FIGURE 11-18 Percent deletion of /s/ before the break in Argentine Spanish for six occupational groups 70

60 50

40 30 20

sixteen

IV (high)

100

I'm going down)

third party

Classified Occupations

Data source: Terrell 1981.

General remarks Based on the evidence from this and other studies, parallel distribution patterns for phonological variables are to be expected if comparable social structures are found.

Linguistic variants of females and males found. There are also morphological and syntactic variations, although evidence for variation at these grammatical levels is sparse. What is true of variations in English, French, and Spanish is likely to be true of similarly structured communities of other languages, although here again the evidence is sparse.

The linguistic diversity of women and men You already know that in many language communities women and men do not speak identically. In the United States, certain words are more associated with women than men and can therefore "sound" feminine. Adjectives such as adorable, cute, and sweet can have feminine connotations, as can words that describe precise shades of color, such as lilac and teal. In some languages, the differences between the way women and men speak are more dramatic than in English. In informal situations among Japanese speakers, even the first-person pronoun is different for women (atasi) and men (boku). In French, je is the first-person pronoun for both males and females, but since adjectives are marked for gender agreement, Je suis heureux "I am happy" identifies a male speaker, while Je suis heureuse identifies a male speaker. Reports of marked differences between generic varieties have been reported for Chukchee (spoken in Siberia) and Thai. In polite Thai conversations between men and women of the same position, women say dicˇ hàn, while men say phoˇm for the first person singular pronoun "I". Answers like "thank you" and "sorry." The polite formula used by men is khráp, while women use khá or khâ. Because these polite particles are common in day-to-day interaction, the differences in speech between men and women in Thai can seem very clear, even though few words are so different. There are also more subtle differences between men's and women's speech, the kind of quantitative differences we've seen between other social groups. In Montreal, for example, where professionals are less likely to omit /l/ from articles and pronouns than workers are, men and women also differ in their pronunciation of the same words. Figure 11.19 on page 378 shows that males delete /l/ more often than females for il (personal, as in il chante "he sings"), for elle, and for the pronouns les and la. Patterns are also emerging in New York City and Norwich where females are less likely to drop sounds than males. In these cities, when higher socioeconomic classes behave more linguistically than lower classes, women tend to behave more like higher socioeconomic groups than men. In English, along with vocabulary differences, more subtle linguistic differences between the sexes can go unnoticed. One study looked at the pronunciation of the suffix -ing in words like run and talk. In a semi-rural New England town, the speech samples of a dozen boys and a dozen girls between the ages of 3 and 10 showed that even among these young children, all but three used the pronunciations alveolar [n] and velar [ŋ ] for verbal -ing. Interestingly, twice as many girls as boys showed a preference for the /ŋ/ forms, as shown below.

• 377

3 7 8 • Chapter 1 1 Linguistic differences between social groups: dialects

FIGURE 11-19 Percent deletion of /l/ in Montreal French for females and males 100 90

94 84

80 70

67 60

women Men

50 41

20 10 0

People)

is he around

The pronoun)

there (pronoun)

Data source: Sankoff and Cedergren 1971.

-ing pronunciation by 12 boys and 12 girls in a New England town GIRLS BOYS

PREFERENCE FOR /ŋ/

NO PREFERENCE FOR /ŋ/

10 5

2 7

The finding that girls and boys differ in this way may come as a surprise, given that girls and boys in this New England town (as is often the case in Western societies) are often in personal contact. A separation of communication channels, previously suggested as a motivating factor for distinguishing dialectal language patterns, does not seem to explain this case. Then what is the explanation? One hypothesis is the trait of "hardness" associated with working-class lifestyle combined with the trait of "masculinity" associated with "in" forms. In other words, an association between masculinity and "dropping the g" can overcome the associations with prestige and higher socioeconomic status that otherwise accompany the -ing variant (with the g). This analysis suggests that language differences between men and women have little to do with biological sex and much to do with socially constructed gender roles.

Masculinity and the Resistance Factor There is evidence for the prestige of faster spoken pronunciations and not "lower g" pronunciations. Here are two facts. (1) English speakers who use both variants (virtually all of us) are more likely to “pronounce the g” in more formal situations. (2) Social groups with a higher socioeconomic level more than pronounce it

Why do stigmatized varieties persist?

lower status groups. Interestingly, girls and women use the pronunciation of -ing more often than boys and men. One explanation could be that women are more status conscious than men; Sociologists have noted that this is the case in other areas as well, so it wouldn't be surprising. But linguists suggest another reason. Think of it as the "endurance factor." Both boys and men can associate pronunciations like walking and talking with working-class 'toughness', and this association seems to outweigh any association with prestige. One could say that the preference for the less prestigious pronunciation denotes masculinity. Now you might object that using the term 'masculinity' to explain the language behavior of boys and men raises the question. After all, what's the point of calling a pronunciation "masculine" just because men use it more often than women? Well, masculinity and femininity are not the same as masculinity and femininity. Gender differences (male and female) are biological, and language differences do not reflect biology. Rather, they reflect the sociocultural phenomenon of gender: what it means to be a man or a woman. They are aware of gender differences marked by clothing, hair length, body jewelry, and jewelry use. ("Wear earrings, for Christ's sake," says the mother of Emma Thompson's character in The Winter Guest after she cuts her hair. "Let people know you're a woman!"). Don't be surprised that the language also reflects the important social identity of gender roles. It will be interesting to see how the ongoing efforts to equalize gender roles in Western societies can soften the differences between male and female pronunciation and other language patterns!

Why do stigmatized varieties persist? You may be wondering why speakers don't abandon their stigmatized variants for more prestigious ones. The explanation seems to lie in the fact that a person's identity—as a woman or a man, as an American or as an Australian, as a member of a particular ethnic or socioeconomic group—is linked to the language patterns of the group to which he belongs. heard . Changing the way you speak signals changes in who you are or how you want to be perceived. For a New Yorker transplanted to California, speaking like a Californian is giving up a certain New York identity. Stopping speaking African American English means giving up a specific identity as African American. Giving up working-class language patterns acquired in childhood means embracing a new identity. In short, adopting new speech patterns means reforming and presenting yourself anew. Language is one of the main symbols of our social identity, and we have seen how it can be remarkably attuned to that identity. If you identify with “non-native” regional, socioeconomic, or ethnic groups and wish to have sufficient contact with them, your speech will be similar to theirs. We can illustrate this with a revealing study of linguistic and social identity on Martha's Vineyard, an island off the coast of Massachusetts. There the vowels /aj/ and /aw/ have two main variants, with the first element of each diphthong alternating between [a] and the more central vowel [ə]. Words like night and why can be pronounced [aj] or [əj]; Words like "shout" and "how" can be pronounced with [aw] or the more focused diphthong [əw]. These variants are not typical features of the dialect; they do not reflect gender, ethnicity, or socioeconomic status. Rather, the vowel centering represents identity with the traditional values of the island and its life. the islanders

• 379

3 8 0 • Chapter 1 1 Linguistic differences between social groups: dialects

They have more vocal focus than residents in sections geared toward summer visitors. Adolescents who intended to leave the island and live on the mainland showed lower vowel centralization, while higher vowel centralization was demonstrated by an adolescent who moved to the mainland but returned to Martha's Vineyard. Thus, centered diphthongs can be seen as a rejection of continental values and a positive view of island life values. The symbolic value of a person's linguistic diversity should not be underestimated. When evaluating oral arguments in Britain, speakers of regional variants rated the quality of an argument higher when presented with a standard accent, but found the same argument more persuasive when presented with a regional accent. It's easy for speakers from a higher socioeconomic background to ask speakers from a lower socioeconomic background, "Why don't they start speaking like us?" The answer may be simple: Their social identities are different and they don't necessarily share the values of the highest socioeconomic group. To get a feel for the topic, think of gendered dialects. Although there have been recent neutrality movements, it is perfectly acceptable for women to speak like women and men to speak like men. Imagine men asking women to speak like them in order to get ahead in the "man's world". Or imagine a female CEO urging her truckers to speak more like women in order to get ahead in the "women's world." These are clearly unacceptable scenarios. Similarly, most regional varieties are accorded approximate equality. Imagine a Bostonian who moves to Atlanta, only to be told by his boss to "lose" his New English accent in order to be successful. The officer can conclude that the Boston origin, not the accent, was in question. Language is a central factor in a person's identity. Asking people to ask you to change your habitual speech patterns isn't like asking them to wear sweaters of different styles or colors. They challenge it to adopt a new identity and to retain the values associated with that identity, i.e. the identity of speakers of a different dialect. One of the reasons non-standard varieties resist the lure of education so strongly is that colloquial varieties are deeply intertwined with the social identities and values of their speakers.

Computational and Dialect Studies Despite the vast amount of quantitative and qualitative data contained in our discussion of dialects, it should come as no surprise that dialectologists use computers to achieve their goals. Researchers are digitizing the types of data that were manually plotted in the past, as in some of the maps in this chapter. For example, researchers at the Linguistic Atlas of the Mid- and South-Atlantic States (LAMSAS) used a program called MapInfo to plot the longitude and latitude coordinates of the homes of the 1,162 LAMSAS informants, which will allow for maps of various sizes and degrees. of details to represent the features extracted from the informants. You

GRAMM

We have also seen the use of computers in generating non-traditional maps for dialectology in Figures 11-7 on page 359 and 11-8 on page 360. On the other hand, the work represented in the Telsur project and the Atlas of North American English depends critically on the use of computers for the acoustic analysis of vowel sounds. In addition to the diverse tasks that computers have been used to perform for map-related activities, the types of resources that corpora provide to researchers interested in linguistic variations are beginning to revolutionize the study of dialects. A large project called the International Corpus of English aims to collect texts totaling ca

Continue

Millions of words of spoken and written English from the 1990s from each of the 20 world centers representing the English spoken in the Caribbean, Fiji, Ghana, Hong Kong, India, Kenya, Nigeria, the Philippines and Singapore to name a few regions. The texts in these corpora are tagged and annotated, making their use in dialect comparisons extremely valuable. In our discussions of the differences between dialects, we have seen that vocabulary and pronunciation vary. In the real world, an understanding of the vowel

• 381

The deviation proved useful in keeping an innocent person out of jail. Computers were used to help analyze vowel features in recordings that contained illegal speech acts, in this case threatening acts. Someone had called a major airline with a serious threat of violence, and workers who heard the call thought they heard the voice of a disgruntled former employee. A computer analysis of the quality of the caller's vowels revealed that his dialect was not the same as the former employee's. ■

Summary • When physically or socially separated, people who share speech speak differently. With enough time and separation, different languages can emerge. • On the other hand, the language of people speaking as members of the same community can evolve in unison and in some situations even tend to merge. • There is no linguistic basis for distinguishing between a dialect and a language. Every language consists of dialects and every dialect is a language in terms of linguistic principles and linguistic universals. • There are linguistic differences between social groups within each language community. • Linguistic forms can vary greatly from one social group to another, and social groups can be defined as regional in many other ways. • A social group may have ethnicity or socioeconomic status as a basis for membership. • Women and men can also be viewed as members of different social groups, so-called gender groups. • The combination of these different group distinctions gives a complex picture of the composition of society. Within a given ethnic group we find socioeconomic classes whose members are either male or female. Differences in speech underpin these social identities. • Whatever the social group, its linguistic diversity usually has characteristics that distinguish it from the linguistic varieties of other social groups. • Linguistic features that characterize social varieties can also serve as markers (or symbols) of social identity. • One way of emphasizing belonging to an ethnic group is to emphasize or even exaggerate the characteristics of that ethnic linguistic diversity. • If a woman wants to appear particularly feminine, she can choose to display characteristics associated with feminine language and avoid "masculine-sounding" expressions. • People can use socially tagged language features for their own ends. • Everyone speaks with a pronunciation characteristic of social identity. Nobody can speak without an accent, although we are very aware of the accents of others and think that members of our own social group do not have an accent.

3 8 2 • Chapter 1 1 Linguistic differences between social groups: dialects

WHAT DO YOU THINK ? R E V I S I T E D • Nina also speaks strangely. Nina's camp attendant appears to speak British English. As funny as the counselor's dialect may have sounded to Nina, her dialect could have sounded as "funny" to the counselor. The linguistic diversity of each group differs – more or less – from the other groups. Of course, there's nothing "funny" about anyone's language, except maybe that it's different, and patterns that differ from ours can seem odd simply because they're different from what we're used to. • Substitute for Daniel in Alabama. Probably nobody in the world doesn't have an accent; no one speaks a linguistic variety whose pronunciation does not reflect their social identity. (English speakers recognize British, American, Canadian, and Australian accents, among others.) Even if you could get rid of your native accent, you'd have to replace it with another, since you can't speak a language without an accent. (See Ex. 11-10b.) • Justin and Ebonics. In 1997, at the height of the Ebonics controversy, much commentary in newspapers and on the radio indicated a widespread perception that a dialect can be judged good or bad by how closely it resembles the standard variety of the same language. All language variants are more or less different from each other. French is not grammatical when judged by the rules of Spanish or Japanese. American English is ungrammatical when judged by the rules of British English. Ebonics is ungrammatical when judged by standard English rules. And Standard English is ungrammatical when judged by Ebonics rules or Standard French. • Women talk, men talk. The extent to which the language of men and women differs is not the same from one cultural group to another. This also applies to the variation between social groups within a given culture. The survey found some differences, such as B. Different words for the same object or some forms that are more commonly used by men or women. Remembering how boys and girls are raised differently, Sammy shouldn't be surprised that men and women speak a little differently too.

English Based Exercises 11-1. Distinguish accent and dialect. Distinguish dialect and language. What is meant by "linguistic diversity"? Does it make sense to say of linguistic diversity that "it's not a language, it's just a dialect"?

Exercises 11-2. Examine an issue of a newspaper or magazine published in Britain (e.g. The Times, The Economist, Punch, The Spectator, The Listener) and list as many examples of differences between American and British English as you can in one or several can notice two sides. . Add examples of vocabulary, syntax, spelling, and punctuation. 11-3. Which of the following words do you know? Make two lists, one of words you usually use and the other of words you don't use but have heard from others. With which regional or national group do you associate words that you have heard from others but do not use? Compare your judgments with those of your classmates. dragon-fly

Darning Needle, Mosquito Buzzard, Spindle, Snake Eater, Snake Doctor

pancakes

Fried dumplings, warm cakes, flannel cakes, dumpling cakes

Quark

Quark, cottage cheese, cottage cheese, Dutch cheese, curd cheese

green beans

green beans, green beans

Wurm

Night Crawler, Angelworm, Angleworm, Earthworm, Red Worm, Mudworm

Firefly

Firefly, Firebug

baby carriage

Prams, baby trainers, baby taxis, strollers

11-4. The following questions (some slightly adapted) are taken from the questionnaire used to collect data for DARE. Answer each question yourself and then compare your answers with those of your classmates. Do you and your colleagues agree on the regions in which certain variants are used? (DARE provides maps to answer these questions.) a. How do you talk about streets with numbers or letters? For example, if someone asks you how to get to (enter the name of the city nearby), you might say, "Go to ___". B. What names are used here for: 1) the part of the house below the ground floor? 2) The kind of sandwich on a big long loaf, is that a meal in itself? 3) a small creek not big enough to be a river? 4) a round dough cake, fried in fat, with a hole in the middle? 5) a piece of cloth that the woman folds over her head and ties under her chin? 6) Is the common earthworm used as bait? 7) Vehicles for babies or toddlers, the kind that can be put down? 8) a skin tag that someone sucked hard on and brought blood to the surface? 9) a horseshoe-shaped chicken breastbone? 10) That spot on your elbow that gives you a weird feeling when you hit something? 11) very young frogs if they still have a tail but no legs? 11-5. What was Labov's hypothesis about the prevalence of /r/ in New York department stores? Are there three socially rated businesses in your city or community that could be studied in a similar way? What two or three phonological features do you expect to be socially differentiated in your stores? Write a question for each attribute

• 383

3 8 4 • Chapter 1 1 Linguistic differences between social groups: dialects

Find the data you need to test your hypothesis. (Of course, ask the question for the type of business you have in mind.) Would you ask respondents to repeat their answers, as Labov did? Explain why or why not. 11-6. "William Labov. . . he once said of the use of black English, "The goal of most black Americans is to gain full control of the standard language without endangering their own culture." . I wonder if perhaps the good doctor also considers the goals of those black Americans who fully understand standard English but are occasionally annoyed by this fluttering, colorful dialect of black English. Case study: me. So wrote a 21-year-old African-American college student in Newsweek (December 27, 1982, p. 7). She cites several characteristics of African American English as described in this chapter. a. Look up the meaning of patois in your desktop dictionary and remember the connotations when referring to specific language variants. Are these positive or negative connotations? What does your use of the phrase "grammar to the wind" say about a student's attitude toward African-American English grammar? B. What characteristics of African American English do you think the student is referring to when they call you the "Diatoah Grammar of the Winds"? C. What would be the implications for communication if any kind of language were truly "ungrammarized"? Give two reasons why African American English cannot properly be described as a "grammar for the winds" dialect. i.e. What do you think is the reason for students' attitudes towards African American English? What could you explain about the linguistic patterns in each variety and about the status of certain varieties in relation to their linguistic characteristics? 11-7. Describe two ways you noticed that women and men differed in greetings, threats, swear words, and promises. What do you think explains these differences? Do you think these differences are increasing or decreasing? Explain the basis for your answers. 11-8. Among many functions of the word, as in English, it is used by certain speakers to mark the beginning of a direct quote. Here are two examples of this type of quote: "And then she said, 'I don't want to go.'" "Then he said, 'But you promised!'" To complete this exercise, you need natural language data. Your friend Collect 20 examples of natural quotes from at least five people (including some younger people and some older people). Write the examples exactly as they were spoken, being careful not to draw attention to or observe your acquaintance's language. Based on five-year intervals (15-19, 20-24, etc.), note the approximate age of each speaker you want to monitor (regardless of whether they use quotation marks or not). a. Some researchers call this feature "to be like" because their data suggest that this use of "like" often occurs with the verb to be, as in the examples above. Please explain whether or not your data supports the use of the alternative name.

exercises b. Identify the tense (past or non-past) of the verbs preceding the quote as in your data. Identify the tense (present, past, or future) to which the verbs refer. Please note that time and weather are not the same phenomena. C. In the two examples above, the verb form has been shortened to 's. What percentage of your examples show a similar contraction? i.e. In the two examples above, the subject of being in the quoted sentences is a pronoun (he, she). What lexical categories are the topics in your examples? Y. Identify which age groups are using this feature and which are not. Based on your admittedly limited evidence, hypothesize whether use of this feature is age-related. F. Compare your findings on use and age with those of some colleagues and reconsider your hypothesis in light of the pooled data. grams. Do you think that younger users will continue to use quotes as they get older (which would make it an example of ongoing language switching) or that they will stop using it past a certain age (which would make it a feature for age rating? ?) )? Explain your point of view. h Have you observed instances in your data that show the use of like beyond the citation, apart from its use as a preposition (He looks like his father), as a subordinate conjunction (Winston tastes as good as a cigarette), or as a verb (She likes asparagus)? If so, analyze these uses and try to characterize them; What name(s) could serve you? I. What other phrases have you heard that work like double quotes?

Especially for educators and future teachers 11-9. a. The following is the opening statement of a presentation given by a university professor to a group of Southern professors at a professional meeting. (Imagine, in a thick Southern accent, the college professor was born in the South and clearly wanted to play with those affiliations.) "Years ago, during my first week in Wisconsin, a colleague asked me, 'You mean, you stopped teaching English? The narrator was a Canadian with what I thought was a very odd accent. A short time later, a lady attending a lecture course asked me, with all the sweetness and kindness of which she was capable, if I would allow her to teach me to speak properly. If I had had your zeal, patience, and kindness, I might have made the offer first, as I found your speech very unsatisfactory." Provide answers to these questions, the most common of which the teacher asked her audience: 1) Who should teach whom to speak "well"? 2) Is there a pronunciation standard for American English, and if so, which one? 3) Should education aim to make everyone sound like everyone else? 4) Is it possible for training to make everyone sound like everyone else? 5) If the training was successful, how would they all sound? 6) Assuming uniformity could be achieved, how long might it last?

• 385

3 8 6 • Chapter 1 1 Linguistic differences between social groups: dialects

B. The same university professor reported the following comments from a Southern professor and a Southern doctor: Doctor: [ mn av ma pəjʃəns θn ma spitʃ] kɔwz æz ən εdəketd pəsən a don av kɔəs hæv ə səðən æksnt ] 1) After reading the comments out loud, write them out using the pattern in orthographic. 2) Give the standard spelling for these words pronounced in the same dialect: i) [mɔwnn] ii) [kaəd] iii) [kent] iv) [hεp] v) [spikn] vi) [ həjd] vii ) [mn] viii) [bnft] c. Compare Figures 11-11 and the description of the southern alteration (p. 364) with the transcripts of the southern teacher's and physician's comments. For each of the following Southern Change features, cite two words from the comments or from the word list in (2) that illustrate it: (i) monophthongization of /aj/; (ii) /ε/ decidedly higher and borderline; (iii) // distinctly higher and frontal. (For all examples, please provide the words in standard case and the transcribed version.) d. In the transcriptions, name a word pair that indicates whether the amalgamation pen ~ pen is characteristic of that dialect. Quote a pair of words with the /r/ omission after the vowels. Name a pair with the /l/ ellipsis after the vowels. (Adapted from Jane Appleby, "Is Southern English Good English?" in David L. Shores and Carol P. Hines, eds., Papers in Language Variation [Tuscaloosa: University of Alabama Press, 1977], p. 225.) 11 - 10. Consider the content of what the doctor and teacher reported in Exercise 11-9 and reflect on and answer the following questions: a. Does the doctor think northerners have an accent? B. Does he or she believe that education eliminates or should eliminate a regional accent? C. Do you think the doctor is satisfied that the patients think their doctor is from the North? i.e. The teac