Language in the Real World

Language in the Real World: An introduction to linguistics challenges the traditional approaches taken to linguistics to provide an innovative introduction to the subject. By first examining the real world applications of core areas of linguistics and then addressing the theory behind these applications, this text offers an inductive, illustrative, and interactive overview for students. Key areas covered include animal communication, phonology, language variation, gender and power, lexicography, translation, forensic linguistics, language acquisition, American Sign Language, and language disorders. Each chapter, written by an expert in the field, is introduced by boxed notes listing the key points covered and features an author’s note to readers that situates the chapter in its real world context. Activities and pointers for further study and reading are also integrated into the chapters and an end of text glossary is provided to aid study. Professors and students will benefit from the interactive companion website that includes a student section featuring comments and hints on the chapter exercises within the book, a series of flash cards to test knowledge, and further reading and links to key resources. Material for professors includes essay and multi-choice questions based on each chapter and additional general discussion topics. Language in the Real World shows that linguistics can be appreciated, studied, and enjoyed by actively engaging real world applications of linguistic knowledge and principles and will be essential reading for students with an interest in language. Susan J. Behrens is Professor of Speech–Language Pathology and Audiology at Marymount Manhattan College. Judith A. Parker is Professor of English and Linguistics at the University of Mary Washington.

Language in the Real World
An introduction to linguistics

Susan J. Behrens and Judith A. Parker

First published 2010 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN Simultaneously published in the USA and Canada by Routledge 270 Madison Avenue, New York, NY 10016 Routledge is an imprint of the Taylor & Francis Group, an informa business This edition published in the Taylor & Francis e-Library, 2010. To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to Selection and editorial matter © 2010 Susan J. Behrens and Judith A. Parker Individual chapters © 2010 the contributors All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data Language in the real world / [edited by] Susan J. Behrens and Judith A. Parker. – 1st ed. p. cm. 1. Applied linguistics. I. Behrens, Susan J., 1959– II. Parker, Judith A. P129.L365 2010 418—dc22 2009031835 ISBN 0-203-83981-1 Master e-book ISBN

ISBN10: 0–415–77467–5 (hbk) ISBN10: 0–415–77468–3 (pbk) ISBN13: 978–0–415–77467–3 (hbk) ISBN13: 978–0–415–77468–0 (pbk)


Alternate Contents Acknowledgments Contributors Introduction I: Language, Education, and Cultural Change 1 Language Variation: Students and Teachers Reflect on Accents and Dialects Susan J. Behrens and Rebecca L. Sperling 2 Speech Communities: Language as a Mediator of Messages and Perceptions Miriam Eisenstein Ebsworth 3 Teaching Pronunciation: Using Phonology in the ESL and Foreign Language Classroom Joanna Labov II: Literature, Translation, and Computers 4 Lexicography: What Dictionaries Reveal about Language and Dictionary Makers Paul D. Fallon 5 Text Translation: Approaching Otherness Mary Boldt and Esperanza Roncero 6 Machine Translation: The Challenge of Ambiguity Nan Decker 7 The N-Word, the F-Word, and All that Jazz: Race, Sex, and Transgressive Language in Contemporary American Literature and Popular Culture Carmen Gillespie III: Language, Power, and Identity 8 Language, Power, and Sexual Assault: Women’s Voices on Rape and Social Change Judith A. Parker and Deborah Mahlstedt 9 Gender, Language, and Power: Surname or Sirname? Diana Boxer 10 Linguistics as a Forensic Science: The Case of Author Identification Carole E. Chaski

vii xi xii 1 9 11



65 67 89 107


137 139 164 180



IV: Forms of Language and Communication 11 First Language Acquisition: Developing Native Linguistic Competence Janine Graziano-King and Helen Smith Cairns 12 ASL: A Visual Language Miako Villanueva, Deanna Twain, and Laura Leigh Wood 13 Animal Communication: The “Language” of Honey Bees Wyatt A. Mangum V: Language and Communication Science 14 Communication Disorders: A Personal Perspective Ann D. Jablon 15 Analyzing Narratives: An Example of Cross-Cultural Research Methods Cecile L. Stein 16 Neurolinguistics and Psycholinguistics: Contributions to Understanding Healthy Aging and Dementia Yael Neumann, Linda Carozza, and Anastasia Georgiou 17 Language of Children with Autism: The Two Worlds Underlying Verbal Communication Marion Blank and Mary Beth Cull Glossary Index

205 207 226 255

275 277 297



349 367

Alternate Contents
This Alternate Contents guides readers to chapters that cover the traditional areas of linguistic theory and subdisciplines.

Discourse Analysis 2 Speech Communities: Language as a Mediator of Messages and Perceptions Miriam Eisenstein Ebsworth 6 Machine Translation: The Challenge of Ambiguity Nan Decker 8 Language, Power, and Sexual Assault: Women’s Voices on Rape and Social Change Judith A. Parker and Deborah Mahlstedt 9 Gender, Language, and Power: Surname or Sirname? Diana Boxer 10 Linguistics as a Forensic Science: The Case of Author Identification Carole E. Chaski 15 Analyzing Narratives: An Example of Cross-Cultural Research Methods Cecile L. Stein 16 Neurolinguistics and Psycholinguistics: Contributions to Understanding Healthy Aging and Dementia Yael Neumann, Linda Carozza, and Anastasia Georgiou 17 Language of Children with Autism: The Two Worlds Underlying Verbal Communication Marion Blank and Mary Beth Cull 27 107

139 164 180 297



Lexicon/Semantics 2 Speech Communities: Language as a Mediator of Messages and Perceptions Miriam Eisenstein Ebsworth 4 Lexicography: What Dictionaries Reveal about Language and Dictionary Makers Paul D. Fallon 5 Text Translation: Approaching Otherness Mary Boldt and Esperanza Roncero 6 Machine Translation: The Challenge of Ambiguity Nan Decker 27

67 89 107


Alternate Contents

7 The N-Word, the F-Word, and All that Jazz: Race, Sex, and Transgressive Language in Contemporary American Literature and Popular Culture Carmen Gillespie 8 Language, Power, and Sexual Assault: Women’s Voices on Rape and Social Change Judith A. Parker and Deborah Mahlstedt 10 Linguistics as a Forensic Science: The Case of Author Identification Carole E. Chaski 11 First Language Acquisition: Developing Native Linguistic Competence Janine Graziano-King and Helen Smith Cairns 12 ASL: A Visual Language Miako Villanueva, Deanna Twain, and Laura Leigh Wood 13 Animal Communication: The “Language” of Honey Bees Wyatt A. Mangum 16 Neurolinguistics and Psycholinguistics: Contributions to Understanding Healthy Aging and Dementia Yael Neumann, Linda Carozza, and Anastasia Georgiou


139 180 207 226 255


Morphology 4 Lexicography: What Dictionaries Reveal about Language and Dictionary Makers Paul D. Fallon 11 First Language Acquisition: Developing Native Linguistic Competence Janine Graziano-King and Helen Smith Cairns 12 ASL: A Visual Language Miako Villanueva, Deanna Twain, and Laura Leigh Wood 14 Communication Disorders: A Personal Perspective Ann D. Jablon 17 Language of Children with Autism: The Two Worlds Underlying Verbal Communication Marion Blank and Mary Beth Cull 67 207 226 277


Neurolinguistics/Psycholinguistics 8 Language, Power, and Sexual Assault: Women’s Voices on Rape and Social Change Judith A. Parker and Deborah Mahlstedt 14 Communication Disorders: A Personal Perspective Ann D. Jablon 15 Analyzing Narratives: An Example of Cross-Cultural Research Methods Cecile L. Stein 139 277 297

Alternate Contents


16 Neurolinguistics and Psycholinguistics: Contributions to Understanding Healthy Aging and Dementia Yael Neumann, Linda Carozza, and Anastasia Georgiou 17 Language of Children with Autism: The Two Worlds Underlying Verbal Communication Marion Blank and Mary Beth Cull



Pragmatics 2 Speech Communities: Language as a Mediator of Messages and Perceptions Miriam Eisenstein Ebsworth 14 Communication Disorders: A Personal Perspective Ann D. Jablon 17 Language of Children with Autism: The Two Worlds Underlying Verbal Communication Marion Blank and Mary Beth Cull 27 277


Phonetics/Phonology 1 Language Variation: Students and Teachers Reflect on Accents and Dialects Susan J. Behrens and Rebecca L. Sperling 2 Speech Communities: Language as a Mediator of Messages and Perceptions Miriam Eisenstein Ebsworth 3 Teaching Pronunciation: Using Phonology in the ESL and Foreign Language Classroom Joanna Labov 4 Lexicography: What Dictionaries Reveal about Language and Dictionary Makers Paul D. Fallon 5 Text Translation: Approaching Otherness Mary Boldt and Esperanza Roncero 11 First Language Acquisition: Developing Native Linguistic Competence Janine Graziano-King and Helen Smith Cairns 14 Communication Disorders: A Personal Perspective Ann D. Jablon 16 Neurolinguistics and Psycholinguistics: Contributions to Understanding Healthy Aging and Dementia Yael Neumann, Linda Carozza, and Anastasia Georgiou 17 Language of Children with Autism: The Two Worlds Underlying Verbal Communication Marion Blank and Mary Beth Cull 11



67 89 207 277




Alternate Contents

Sociolinguistics 1 Language Variation: Students and Teachers Reflect on Accents and Dialects Susan J. Behrens and Rebecca L. Sperling 2 Speech Communities: Language as a Mediator of Messages and Perceptions Miriam Eisenstein Ebsworth 8 Language, Power, and Sexual Assault: Women’s Voices on Rape and Social Change Judith A. Parker and Deborah Mahlstedt 9 Gender, Language, and Power: Surname or Sirname? Diana Boxer 11


139 164

Syntax 4 Lexicography: What Dictionaries Reveal about Language and Dictionary Makers Paul D. Fallon 6 Machine Translation: The Challenge of Ambiguity Nan Decker 8 Language, Power, and Sexual Assault: Women’s Voices on Rape and Social Change Judith A. Parker and Deborah Mahlstedt 10 Linguistics as a Forensic Science: The Case of Author Identification Carole E. Chaski 11 First Language Acquisition: Developing Native Linguistic Competence Janine Graziano-King and Helen Smith Cairns 12 ASL: A Visual Language Miako Villanueva, Deanna Twain, and Laura Leigh Wood 14 Communication Disorders: A Personal Perspective Ann D. Jablon 17 Language of Children with Autism: The Two Worlds Underlying Verbal Communication Marion Blank and Mary Beth Cull 67 107

139 180 207 226 277



The publishers wish to thank the following for permission to use copyright material: Definition of “Red” Copyright © 2006 by Houghton Mifflin Harcourt Publishing Company. Adapted and reproduced by permission from The American Heritage Dictionary of the English Language, Fourth Edition. Shonna L. Trinch, “Managing euphemism and transcending taboos: Negotiating the meaning of sexual assault in Latinas’ narratives of domestic violence,” in Interdisciplinary Journal for the Study of Discourse, 21:4, Berlin, New York: Mouton de Gruyter, 2001, pp. 567–610, Table 1: “Spectrum of directness for Latina women’s reference to sexual violence”, pp. 582–583. “Red” from the New Oxford American Dictionary (2005) and “Explode” from Oxford English Dictionary (1989), edited by Simpson J. and Weiner E. By permission of Oxford University Press. Definitions from The World Book Dictionary, © 2007 World Book, Inc. By permission of the publisher. All rights reserved. Definition of “amity” Random House Dictionary of the English Language, Unabridged (1987) 2nd ed. New York, Random House. With kind permission from Studio Editions. Signed instruction spread from Clerc’s school throughout the country. Lucas and Hogue 2004, used with permission of C. Lucas and R. Hogue. The British Sign Language (BSL) and American Sign Language (ASL) manual alphabets. Gary, N. E. (1992). “Activities and behavior of honey bees,” in The Hive and the Honey Bee, ed. Joe M. Graham. Dadant and Sons, Hamilton, ILL. “Replot of von Frisch” (1967) Reprinted by permission of the publisher from The Dance Language and Orientation of Bees by Karl von Frisch, translated by Leigh E. Chadwick, pp. 64,296, Cambridge, Mass: The Belknap Press of Harvard University Press, Copyright © 1967, 1993 by the President and Fellows of Harvard College. Stein, C. (2004) Narratives of Bhutanese and rural American 7-year-old children: Issues of story grammar and culture. Narrative Inquiry, 14(2), pp. 369–394. Produced with kind permission by John Benjamins Publishing Company, Amsterdam/Philadelphia. Every effort has been made to trace the copyright holders but if any have been inadvertently overlooked the publishers will be pleased to make the necessary arrangements at the first opportunity.

Susan J. Behrens, Ph. D., Professor of Communication Sciences and Disorders, Marymount Manhattan College, New York, NY. Marion Blank, Ph.D., Co-director, Developmental Neuropsychiatry Program, Columbia University, New York, NY. Mary Boldt, Ph.D., Associate Professor of German, York College of Pennsylvania, York, PA. Diana Boxer, Ph.D., Professor of Linguistics, University of Florida at Gainesville, Gainesville, FL. Linda Carozza, Ph.D., CCC-SLP, Assistant Professor of Speech, Communication Sciences & Theatre, St. John’s College, Staten Island, NY. Helen Smith Cairns, Ph.D., Professor Emerita, Linguistics and Speech and Hearing Sciences, the Graduate Center, and Linguistics and Communication Disorders, Queens College, the City University of New York, New York, NY. Carole E. Chaski, Ph.D., Institute for Linguistic Evidence, Inc., ALIAS Technology, LLC, Georgetown, DE. Mary Beth Cull, Developmental Neuropsychiatry Program, Columbia University, New York, NY. Nan Decker, Ph.D., senior linguist, Logovista US, Inc., Belmont, MA. Miriam Eisenstein Ebsworth, Ph.D., Director of Doctoral Programs in Multilingual, Multicultural Studies, New York University, New York, NY. Paul D. Fallon, Ph.D., Associate Professor of Linguistics, University of Mary Washington, Frederickburg, VA. Anastasia Georgiou, M.S., CCC-SLP, Jersey City, NJ. Carmen Gillespie, Ph.D., Associate Professor of English, Bucknell University, Lewisburg, PA. Janine Graziano-King, Ph.D., Associate Professor, English, Kingsborough Community College, the City University of New York, Brooklyn, NY. Ann D. Jablon, Ph.D., CCC-SLP, Professor of Communication Sciences and Disorders, Marymount Manhattan College, New York, NY. Joanna Labov, Ph.D., Clinical Assistant Professor of TESOL, Steinhardt School of Culture, Education, and Human Development, New York University, New York, NY. Deborah Mahlstedt, Ph.D., Professor of Psychology and Women’s Studies, West Chester University, West Chester, PA. Wyatt A. Mangum, Ph. D., Visiting Assistant Professor of Mathematics, University of Mary Washington, Fredericksburg, VA Yael Neumann, Ph.D., CCC-SLP, Assistant Professor of Linguistics and Communication Disorders, Queens College, City University of New York Queens, NY.



Judith A. Parker, Ph.D., Professor of English and Linguistics, University of Mary Washington, Fredericksburg, VA. Esperanza Roncero, Ph.D., Associate Professor of Spanish, Hartwick College, Oneonta, NY. Rebecca L. Sperling, Ph. D., Associate Professor of Social Work/Sociology, Marymount Manhattan College, New York, NY. Cecile L. Stein, Ph.D., Adjunct Assistant Professor of Communication Sciences and Disorders, Marymount Manhattan College, New York, NY. Deanna Twain, Adjunct Professor of Communication Sciences and Disorders, Marymount Manhattan College, New York, NY. Miako Villanueva, Ph.D., Instructor, Department of Linguistics, Gallaudet University, Washington, DC. Laura Leigh Wood, MA, St. Louis, MO.

SB: To Tony, of course JP: For Gladys B. Parker, my beautiful mother


Language in the Real World: An introduction to linguistics is a different kind of book. It approaches linguistics by examining how the various branches of the discipline are put to use in the real world. This book offers an introduction to the traditional areas of linguistics and linguistic analysis in an inductive, illustrative, and interactive way by examining and telling stories about real world applications. This text conveys the immediacy and enjoyment of reading about and studying linguistics and conducting research into language. The book’s tone is one of discovery as readers become more linguistically aware as they work through the chapters. Most traditional linguistics texts first lead students through linguistic theory, with applications relegated to the final few chapters. Many of these texts reflect a belief that theory must precede application and inform the work of beginning linguistics study. Other texts that focus on the application of linguistics seem narrow in scope, often directed to the specific needs of a different population: teachers of English as a Second or Foreign Language. Our text challenges these beliefs and practices. From our experiences in the classroom, and commitment to student-centered, authentic tools for learning, we have developed a text that takes a more inductive approach to uncovering linguistic phenomena. Readers do indeed encounter linguistic theory, but they do so in ways that embed theory in everyday phenomena. In some courses, this textbook will stand on its own, satisfying the needs to engage students in metalinguistic study and the examination of linguistics. The book will serve well, and be especially exciting to, students new to linguistics, with its richly varied chapters covering introductory and more advanced levels; its helpful interactive exercises; and the resources both in the book and on the book’s companion website. Linguistics professors might also choose to use one of the many fine traditionally oriented texts on the market in conjunction with our text to help students connect real world events and issues with the many facets of language study. One of the strengths of this book is the variety of approaches to the study of language. Each part has a broad introductory chapter followed by chapters that narrow in focus. While broader chapters provide an overview to readers, other chapters present a particular aspect of a field of linguistics. No text covers everything; our chapters are carefully chosen and arranged to give readers many flavors of what the title promises: language in the real world. In addition, each chapter is written by a scholar or group of scholars who share their experiences of linguistics. While this format offers readers a variety of voices across chapters, we (the editors) have worked to ensure a consistency of level and style to make the reader’s experience of the book smooth.

Who This Book Is For
Our book aims to reach two main types of readers: students studying the discipline of linguistics and those in related fields that are informed by language issues. Readers might encounter this book because they have chosen Linguistics as a major or minor. Then



again, they might be engrossed in a subject that intersects with the material in this text—language in its real world applications—and yet might not have taken a traditional linguistics course. Readers might be English majors, Modern Language majors, Philosophy majors, or Performing Arts majors who are interested in language in its spoken and written forms; or studying Psychology, Sociology, or Anthropology: the sciences of our species. All these areas are informed by language and constructed through the application of language. As a discipline that touches on so many other fields of study, and has so many manifestations in human interaction, linguistics should be introduced to students as a real world phenomenon. Specifically, this text will benefit students studying the following areas: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Linguistics General Education/Liberal Arts and Sciences Composition Communication Arts and Theory Education English language English literature Foreign languages Philosophy Gender Studies Performing Arts Social sciences Psychology Communication Sciences and Disorders

We, the editors, each have over 20 years of teaching experience. We have seen again and again that most students get excited about linguistics when they can relate it to daily encounters and when they explore linguistic knowledge and principles through real world applications. Our philosophy is that linguistics can be appreciated, studied, and enjoyed by actively engaging in real world applications of linguistic knowledge and principles.

What Readers Learn from This Book
The lessons readers will be engaged in when they work with this book are two-fold. There is the discovery of linguistics, a discipline devoted to uncovering how language is structured and used to communicate meaning. But this book goes beyond what is already available in linguistics texts; its other lesson is that language is everywhere, and we all benefit from a keener awareness of language and its application. And while a reader might be immersed primarily in one discipline, he or she will benefit from a more comprehensive look at language applications in many areas. This text covers up-to-date issues of language and its applications, as seen from the viewpoint of our many contributors. To take just a few examples, readers will learn about how linguistics applies to crime work in forensics; how the novels of Toni Morrison offer a linguistic view of our culture; what autism is and how the notion of the autistic spectrum evolved; what is beneficial and frustrating about machine translation software; and how linguistic analysis reveals the power dynamics of telling sexual assault narratives in different contexts.



How to Read This Book
Organization of the Text
The length of the text is designed for a single semester’s course of study. The order of chapters takes the reader thematically through the various applications of linguistics; furthermore, each part leads off with a chapter that is broader in scope than its companions, with the subsequent chapters in that section focusing in more detail on another aspect of the theme of that section. Given this design, readers can get an overview of related areas in linguistics by choosing to read initial chapters in each part, and then going back to explore the rest of the part in more depth. Since our book is designed to inform readers in various areas of study, we supplement our table of contents with one that lists the traditional areas of linguistics (phonology, morphology, semantics, syntax, and discourse) and the chapters that cover these areas in some depth. In addition, our chapters use examples from many languages other than English, including American Sign Language (ASL). The parts are organized in such a way that readers first encounter a real world phenomenon they all share: being in school, especially in a new century of globalization and diverse communities (Part I: “Language, Education, and Cultural Change”). The book next presents another section of real world language use familiar to students: books! In this case, dictionaries, novels, and how translators work (Part II: “Literature, Translation, and Computers”). From there, the textbook goes into more specialized aspects of language use. In Part III (“Language, Power, and Identity”), we read about language analysis and what it reveals about the language user in different social contexts. Part IV (“Forms of Language Communication”) turns to what different communication forms look like, in young children, users of a signed language, and even in honey bees. We end with Part V (“Language and Communication Science”), the most specialized section of the text, exploring communication science as it assesses language development, disorders and differences. Yet even in Part V, readers from many disciplines will benefit from examining language in its clinical applications.

Organization of the Parts
Part I, “Language, Education, and Cultural Change,” takes readers into the classroom and examines attitudes towards language diversity by college students and teachers (Chapter 1). Emotions that arise from language differences across cultures are next explored (Chapter 2). Part I ends with a look into the philosophy and phonological training of a teacher of pronunciation to both English as a Second Language (ESL) students and native language learners (Chapter 3). Part II, “Literature, Translation, and Computers,” starts with an exploration of something very familiar to college students, the dictionary, and uncovers new connections between such a reference tool and human language, specifically how dictionaries preserve and document linguistic trends (Chapter 4). We then look at how translators work, both the human and the machine type (Chapter 5 and Chapter 6). Part II finishes with a discussion of potentially inflammatory, transgressive language in literature, focusing on Toni Morrison (Chapter 7). In Part III, “Language, Power, and Identity,” we turn to how linguistics is used to explore ourselves and our identity, with chapters on the language and benefits of telling sexual assault narratives (Chapter 8); a look at naming practices specific to gender (Chapter 9); and a discussion of the legal applications of linguistics, in a chapter on linguistics as a forensic science (Chapter 10).



Part IV, entitled “Forms of Language and Communication,” looks at how we acquire our native language (Chapter 11). Next we consider American Sign Language (Chapter 12). Part IV ends with a discussion of the latest findings from an expert on bees and probes our notions of what language really is (Chapter 13). Part V, “Language and Communication Science,” turns to the clinical side of linguistics, with a look at how speech–language therapy puts the field of linguistics to work for people with language disorders (Chapter 14). Part V then looks at the role of children’s narratives and storytelling in our lives via a cross-cultural application of a diagnostic paradigm (Chapter 15); an in-depth look at how linguistic research can be employed to study the aging brain (Chapter 16); and discussions of language traits in children on the autism spectrum (Chapter 17). Key words appear in bold in each chapter and are defined in a glossary at the back of the book. The glossary, in turn, uses a co-indexing system to alert readers to those chapters that discuss a given term in some depth.

The Organization of Each Chapter
Each chapter begins with a short list delineating the key points to be discussed. Then follows a quotation from an outside source that can only be fully appreciated by working through the chapter. Next is a note from the contributor to the reader, explaining how he or she became involved in the area being introduced and why linguistics has been a compelling force in this work. In-chapter exercises throughout are interactive and take the reader further in his or her engagement with the material. Some of these exercises can be worked out in class, while others call for library or Internet research, or even fieldwork, and are more suitable as assignments outside of class. In addition, many of the exercises have companion comments that can be found on the book’s companion website. Sometimes serving as an answer key, or a debriefing narrative, these comments allow the reader to process the exercise, but without interrupting the flow of the text. Finally, each chapter includes a concise list of resources cited and additional sources for further exploration, such as websites and contact information for relevant organizations.

How to Read the Book
The book can be approached in several ways, depending on the readers’ interests, purposes, and level of familiarity with the material. The first chapter of each part can be read for an overview of the various areas of linguistic application, and then readers can go back and read each part in full. Readers can approach parts in an order different from how they appear in the book, depending on the students’ discipline. Readers in both linguistics and other disciplines, ultimately, benefit from working through the entire book. The text tells a full story of linguistics and language in our world, and while readers can customize the book by rearranging the order of chapters and parts, we hope that all readers engage in the book as a whole. To do so is to better understand the ways language is around us and develop keener critical thinking skills about those applications. In other words, while the different parts have been carefully constructed by theme, the chapters have also been written to hold up independently and work with a different arrangement of companion chapters. The overall effect, however, is a cumulative one. While students in different fields of study may choose to rearrange the table of contents to suit them, the book is a unified product. Those in the following disciplines might want to enter the book in the following ways:



Linguistics While most linguistics courses tend to work with more traditionally oriented texts, our experience with many students over the years is that a companion book of applications goes far to contextualize the theory students are learning. For linguistics courses offered in departments that do not have a major in the discipline, our text would function equally well, or perhaps better, as a primary text. Students studying linguistics would find all sections relevant and can read in the suggested order or in an order more customized to the organization of the course. An alternate table of contents is included specifically for linguistics courses, so readers can select chapters as they relate to the traditional areas of linguistic study: phonology, morphology, semantics, syntax, and discourse, as well as subdisciplines of linguistics. Graduate students in linguistics would also encounter a full, rich array of career possibilities in linguistics. Pondering what one “does” with a degree in linguistics? Here are many portraits of linguistics working daily. General Education/Liberal Arts/Composition Readers who are interested in a more general view of language, perhaps for a General Education or composition course, would encounter language in its broadest applications and find many exercises throughout the book that ask for written critiques of the issues being covered. These readers may choose to read these chapters first, to get an overview of language in real world situations. In this case, readers would first encounter students and teachers reflecting on attitudes towards language variation (Chapter 1); then move to a discussion of dictionaries (Chapter 4); a linguistic analysis of narratives in a much different context (Chapter 8); how language is acquired in childhood (Chapter 11); and the diagnosis and treatment of speech and language disorders (Chapter 14). Communication Arts and Theory Those studying Comm Arts and theories of communication would benefit from an exploration into other aspects of communication, such as sciences and disorders. These readers might want to start with Part V. From there, Comm Arts students could delve into the many other ways that linguistics sees communication and interpersonal interactions by going back to the suggested order in the Contents. Education Readers studying Education would certainly need to understand language in all its facets. These readers might start with Part I, and read Chapters 1–3, as they relate to teaching and cultural differences. They could then go to Part IV and read about language in different forms: native acquisition of linguistic knowledge, American Sign Language, and how honey bees communicate as a contrasting view on human language. Continue on to Part V, for teachers of all age groups need to understand the nature of communication, its norms, range of differences, and impairments. English English majors will already love language. This book can broaden a student’s thinking about written and spoken language. We suggest first reading Part II, covering dictionaries (Chapter 4); how translators build bridges between languages (Chapter 5); issues of parsing syntax in machine translation (Chapter 6); and the use of transgressive language in popular culture and literature, specifically Toni Morrison’s novels Beloved and Jazz (Chapter 7). From there, read the remaining chapters in order to see the full picture of linguistics at work.



Foreign Languages Similarly, students studying foreign languages will benefit from a scientific look at language and its uses. These readers could start with Part II and then proceed in the order in which the sections are arranged. Philosophy Readers studying Philosophy and specializing in the Philosophy of Language, Logic, or Artificial Intelligence will find much to learn from this text. They could read the text in the given order or look at our Alternate Contents and explore the subdisciplines of linguistics. Gender Studies Many programs in Women’s Studies and Gender Studies have been established since the 1980s. Students in these programs need to understand the linguistic dynamics of gender. These readers could start with Parts I and III, then return to the given order of the book. Performing Arts Many people in the performing arts enjoy and benefit from learning more about voice, speech, communication, and language. For those in these fields, the book as a whole tells an integrated story of language as a tool of expression. Part V will provide an intensive focus on voice, language, and communication. Psychology and Social Sciences Readers in these fields are already studying cognition, behavior, systems, and norms. Language is intrinsic to all these issues. Such readers might start with Parts I and III, sections that cover a broad range of topics about language as a social marker and as a marker of identity; then proceed with the book as it is arranged. Communication Sciences and Disorders Those in the fields of communication, speech, language, and hearing sciences also deal with issues that would be informed by our text. Start with Part V (Chapters 14–17) and read about speech–language pathology, psycholinguistic research into the aging brain, and about Autism Spectrum Disorders. From there, go back to Part I and explore the full range of language applications.
Service-learning and community partnerships are taking hold in education. Some of our chapters, and exercises within chapters, supply faculty planning service-learning units with ideas for interaction with community groups.

Our Contributors
Each chapter is written specifically for this text by an expert in his or her field, all professionals with many years of teaching or related technical experience. There are 26 of us, from over a dozen institutions. Some contributors have specific training in linguistics, while others came to love and work with language via other routes. We are specialists in Sign Language, neurological underpinnings of language, language as a legal tool, diversity, narratives, speech pathology, autism, language development and teaching, dictionaries, translation, and even honey bees! In other words, we come from many places in the real world. We also represent colleges and universities from many parts of the United States. We work in different fields informed by the science and art of language. Fittingly, throughout the book, multiple languages and cultures are represented as readers explore the



many facets that make up our real worlds, our common experiences, as well as our unique experiences. Some of us continue to work primarily in the classroom; others are now in private practice or researching fulltime, but we are all familiar with the college classroom and have demonstrated an affinity for communicating complex material clearly and with enthusiasm to undergraduate students. This text is the next best thing to our visiting your classroom and engaging you in discussion about our work, about linguistics, about language as it lives all around us.

The Companion Website
A book such as this one calls out for an interactive, up-to-date website. The real world does not slow down. While our text is up to date as of its publication, the companion website allows readers to work beyond the printed page in each chapter. Professors and students will benefit from the interactive companion website that includes a student section featuring comments and hints on the chapter exercises within the book, a series of flash cards to test knowledge, and further reading and links to key resources. Material for professors includes essay and multiple choice questions based on each chapter and additional general discussion topics.

Our Thanks
As editors, we would like to extend our gratitude to the following people for their support and efforts on behalf of this book: our valued contributors, all of whom were true professionals and true to the book through the long process of seeing it come to life; our Routledge editors in England: Louisa Semlyen, who started the ball rolling with an enthusiastic response to our prospectus; and Nadia Seemungal, who took up the day-today dealings with us once the organization was in place and became our troubleshooter, sounding board, and supporter; Ursula Mallows, Samantha Vale Noya, and Eloise Cook in England for attention to detail; Russell George at Routledge for work on the companion website; Ivy Ip in New York for supporting the project in our home town; all anonymous reviewers at Routledge, for your very detailed and constructive comments and respect for our work; Ann Marie Tevlin Peterson, a valuable reader and shaper in the early stages of this book; our professors at Queens College and Brown University for helping us get to this point; our families, for putting up with LRW talk over breakfast and dinner (as well as extended editor-to-editor calls between NY and VA): thank you, Gladys, Geoff, Habiba, and Tony; our colleagues at University of Mary Washington and Marymount Manhattan College for similarly listening to our tales and tolerating us on days we were in a “book head”; Radmila Dym at Marymount; Tanya Budilovskaya for valuable help with references; Katharine Thomas for her work on the website and Erika Troseth for organizing our glossary; anonymous reviewers; Antonio Barrenechea, Christofer Foss, Warren Rochelle, Patricia Towle, and Stacey Schlau; and of course our students: This book is for you. We wanted to get you the best book possible, and we realized a few years back that, with a deep breath and several-year commitment, we could do that. We want to share our love of language and linguistics, and we want to show you how it matters, how it is not just a theoretical subject you study to fulfill a requirement and then move on. We both have devoted most of our adult years to linguistics, and even now, see ourselves, other linguists, and the Linguistic Society of America moving beyond academic walls to talk, write, and teach about the value of linguistics and understanding how language works in our world. Hopefully, for many of you reading this right now, as your exploration begins, it will similarly take you to amazing places.

Language, Education, and Cultural Change
We are not all the same. Our customs, lifestyles, attitudes, and norms, all differ in large and small ways. This variety includes language use. We might all speak the same language, but differences exist at many levels of daily life, in school, between different cultures, and across nationalities. Furthermore, these differences, especially the ones we don’t particularly pay attention to, can have wide-ranging ramifications for us. In Part I, we start out by examining the attitudes about language variation held by teachers and students. The real world includes the classroom, and the language behavior and expectations we bring into that classroom. Chapter 1, by Susan J. Behrens and Rebecca L. Sperling, introduces the idea of accent and dialect. They use this information about linguistics to illuminate what happens in the classroom when speakers of the standard and non-standard dialects of English make assumptions and limit access to voice. Miriam Eisenstein Ebsworth, in Chapter 2, continues the conversation about differences, this time regarding pragmatic norms, and she widens the exploration to communities and cultures in contact, sometimes living side by side in the same city. We finish Part I with a view into the world of a pronunciation teacher. Joanna Labov discusses the training involved in helping students learn the phonology of another language, and why it is important to us all to understand such a process.


Language Variation Students and Teachers Reflect on Accents and Dialects

Susan J. Behrens and Rebecca L. Sperling
This Chapter Explores:
Standard Language Accents Dialects Linguistic Markedness African American Vernacular English/Ebonics Slang Style Jargon

So, if you want to really hurt me, talk badly about my language. Ethnic identity is twin skin to linguistic identity—I am my language. Attacks on one’s form of expression with the intent to censor are a violation of the First Amendment. Gloria Anzaldúa, 1999

We are professors at the same college but in different departments, who collaborate to teach about issues of diversity. Rebecca Sperling (RS), a social worker, designed and teaches a course called Valuing Difference, using categories such as ethnicity, culture, class, race, gender, sexual orientation, age, ability, nationality, religion, and language to reflect on social mechanisms that both construct and replicate socio-economic and political power, and oppression. Susan Behrens (SB), a linguist, teaches Language and Culture, an examination of the intersections between societal constructs and language forms. We began our collaboration more than ten years ago, and, with our students, continue to examine the material presented in this chapter. In our classrooms, we encourage students to reflect on their use of language, attitudes about accents, dialects, the use of supposedly ungrammatical forms, and language usage in the world around us. We ask our students to challenge their years of exposure to the prescriptive approach to language study most commonly embraced in educational settings. And we collect words of our students, through freewritings and journal entries, tracking changes in attitudes. Here we try to give you a sense of our classroom work. As you read through the chapter, we ask you to refer to and work through the interactive, fieldwork-like exercises we include, and keep track of attitudes you notice (others’ and your own) about the issues we raise.


Behrens and Sperling SB: I became interested in the field of language diversity while in high school. A New Yorker, I moved to central Connecticut and became the target of mockery by classmates who disliked my pronunciation of various words, or claimed not to understand me. In fact, these experiences were partly responsible for my choice of linguistics as a major in college. RS: Though several other aspects of diversity were quite important to me when I was young, I did not pay much attention to language until years later, when I first began teaching about diversity. I don’t think this is unusual for those of us who unquestioningly embrace the practice of speaking “proper” English, and do it with some success, as I did. After all, I reaped many rewards for writing and speaking “well.” It was not until I began thinking about the concepts of “power” and “privilege” that I could even entertain the notion that, throughout my life, I had eagerly conformed to a way of writing and speaking that somehow had been determined by someone to be the “best,” at the expense of others.

It is a basic principle of linguistics that language changes. This is true when groups of people are separated. Speakers may be separated across time, resulting in historic language changes. They may be separated geographically, leading to localized language differences; groups may also be separated socially. And language, while largely a biological faculty, is shaped by our surrounding speech communities. Hence, we “sound” most like our family, friends, and neighbors, locating us in a culture, place, and time. Another linguistic principle is that varieties of language are equally effective communication systems. All natural dialects are regular and logical. However, just as with factors such as race, class, sexual orientation, gender, and age, societies create a hierarchy of acceptability. Societies do this for language forms as well. There are real world consequences to having such hierarchies where elevated value is placed on one form that is then believed to be more intrinsically “normal” or “correct” than other forms. Those people associated with the normal category are rewarded financially, socially, and emotionally; and those who fall outside the norm are disenfranchised. Conformity to the so-called normal has a pay-off, and those who conform develop a vested interest in protecting and perpetuating the hierarchy. The most highly valued language form is called the standard language. For English, it would be termed Standard English (SE) or Standard American English (SAE). A main enforcer of this form is the educational system. We are aware of this clash between the view of linguists that all native language forms are valid, and the aspect of our job as teachers that reinforces the value of the norm. When we grade papers, coach oral presentations, and even choose which student to call on and listen to in the classroom, who to give credence to, we are acknowledging users of this standard language over users of other language forms. This approach to language is called prescriptive. By extension, we decide who graduates successfully from an educational institution and who does not; and who gains power in society, and who does not have an equal opportunity to succeed. In fact, we’re embracing and modeling adherence to the hierarchy.

Exercise 1

Discuss the competing issues raised by the notion that (1) all language forms are equal, and (2) teachers place value on one language form over others.

Language Variation


The Classroom
Linguists and educators such as Shirley Brice Heath (1983) and Jack Richards and Charles Lockhart (1994) have documented the types of language behavior teachers tend to use in the classroom: teacher-talk. For example, teachers mainly ask questions. Further, we usually ask questions without true information gaps, in that we already know the answers. Studies on teacher-talk recognize the gatekeeping function of teachers, who choose which students to call on and validate, how long a student’s turn should be, and what is considered the “correct” answer or an “appropriate” question to ask of the teacher. As a microcosm of how beliefs and practices about language operate, the classroom is an appropriate setting for research into the issues of language attitudes and their impact on education and learning. We have conducted research for the last 10 years, obtaining student reactions to accent and dialect variation and their responses once exposed to linguistic principles that challenge their attitudes. We gather student comments before and after we discuss the linguistic views of language variety. In the beginning, we ask our students, through free-writing exercises and journal assignments, to write candidly in response to the following prompts: • Tell about a time your language was noticed. • Describe what “good English” is to you. • What verbal behavior is annoying to you? Some of the verbatim quotations from our students tell us something distinct about how our students see themselves and others in the classroom and the world around them through their language attitudes.

Student Quotations (verbatim)
“Coming from the South and speaking with a heavy Southern accent, I would feel different. I am a little doll that people can play with. My friend has begun to ‘play’ with me and try to get me to say certain things for her amusement. At first, I enjoyed the attention; who wouldn’t? I have begun to feel like a minority. Many are constantly amused by the way I talk as if it is funny. This is me. I just wish that people would take me more seriously and listen to the things I have to say rather than interrupt my every thought with, ‘Oh, wow, I could just listen to your cute voice all day.’ Notice there was a comment on my voice, not my thought.” “While I always knew accents were not indicators of intelligence, I somehow still judge people based on their speech patterns.” “For many years in grade school and high school we are taught that there is a proper or correct way to speak.” “In New York, people have stopped me while I am talking to tell me that I have a ‘cute’ accent, or they notice me saying ‘ya’ll.’ I had a teacher once say that she thought it sounded uneducated to say ‘ya’ll.’ ” “I mix up words because of Italian grammatical forms. People make fun of me. Plus, every day I get picked on because of my accent. I’m insecure about the way I speak, and I know I always will be.”


Behrens and Sperling “I think people with Spanish accents are so cute. Sometimes it’s really frustrating when I can’t understand someone because their accent is so heavy.” “Some kids (in DC) don’t really use good grammar. Some people call it Ebonics. I don’t like that word because it sounds stupid, it’s like trying to cover up the fact that some children are not getting a proper education.” “People say to me, ‘Wow, you have such a different accent.’ It’s like my accent has become who I am here.” “I never knew I had an accent! Because I sounded like ‘everyone’ else around me. Then I moved to college.” “I do not think that there should be a right or wrong, but there is a line that should be made when involved in education. I had a professor here who spoke very poor SAE (Standard American English) and wrote it even worse. It would take half of my test time to decode the questions he was saying. This was quite frustrating because that language barrier existed and my grade was on the line. I did realize that I was being somewhat selfish.” “I know that in a way (my accent) does give away that I am not ‘fully American.’ ” “We should not compromise the English language.”

Now you try it.
Ask five people to identify five linguistic behaviors they dislike. For example, are you bothered by someone pronouncing the /t/ in “often”? Now ask them to identify linguistic behavior they admire. Such language usage might include correct use of “whom” or distinguishing between “eager” and “anxious.” Ask them to explain the reasons they like or dislike the examples they raise.

Exercise 2

Language vs. Language User
We notice that some of our students respond to the prompts in Exercise 2 not with statements such as “I can’t stand it when a speaker uses double negatives” but rather with something like “I hate people who use double negatives.” Each illustration expressed in the latter fashion suggests that there are ramifications to our societal attitudes about different language forms. The difference between an expression of distaste for a language event and a negative evaluation of the speaker of that linguistic form is a distinction that is very easy to overlook, but it is a very significant difference. As illustrated in the opening quotation by Gloria Anzaldúa, people experience feedback about their language use as commentary on aspects of their personal identity, as a validation or critique of their cultural and/or familiar heritage, class, ethnicity, etc. The ease with which people merge their opinions of language with attitudes about the language user, and the marginalization or privilege afforded that user as a result, is a strong reason to uncover people’s feelings about language usage, and to make the distinctions explicit. At the same time, as people comment on the language of others, they position themselves as language users as well. While language biases have consequences within the contexts of each individual’s immediate social exchanges, perhaps more importantly, they have collective, long-term

Language Variation


social consequences. Sociologists use the notions of “private” and “public” issues to make thematic connections that explain how the lives of individuals are shaped by the social and historic contexts in which they live. Self-esteem, family and peer cultural acceptance, and social class mobility, for example, are all intricately entwined with how some forms of language are socially valued while others are not. In fact, renowned sociologist C. Wright Mills (1959) has suggested that any of us who wish to gain insight into the circumstances of our own lives must develop a “sociological imagination” in order to see our individual beliefs and actions, in some part, as reflections of the social and historical contexts in which we operate. When we can imagine our own behaviors as a sample of larger social and historical macro-dynamics, we gain some ability through such recognition and understanding to change things. Linguistic dynamics in the classroom, where unconscious language attitudes can easily interfere with learning, can be made conscious.

Standard and Non-Standard Accents: Regional and Social
Talking about language variation leads to the question, “Variation from what?” How does a particular language differ from speaker to speaker? One way is by a speaker’s pronunciation, called accent. Accent is the system of a speaker’s phonology: how that person pronounces his or her phonemes, the consonants and vowels of a language, and how those phonemes interact with one another. Accent variations are not random: there is a system to all accents, a regularity that linguists document. This approach to language is termed descriptive, a cornerstone of linguistics. Linguist William Labov, in the 1960s, sought to document the variation within one geographical locale (reported in Labov, 2006). He initiated a real-world-type methodology of obtaining accent data from speakers of different socio-economic groups. He approached clerks in three different department stores in New York City and, using phonetic notation, noticed the way people pronounced the phrase “fourth floor” in response to a carefully worded query for a product. (The /r/ phoneme after a vowel was the target linguistic variable he was interested in.) He had the speakers repeat the phrase by pretending that he did not catch the answer, thus obtaining a casual (informal, less monitored) and then more careful (stylistically formal) utterance. He chose three department stores that reflected three different socio-economic strata of the city: working class, middle class, and upper class. What he discovered was that the higher percentage of post-vocalic /r/ utterances was associated with the higher socio-economic speakers. The speakers at the store whose target shopper was working class had the fewest post-vocalic /r/ instances. Labov thus documented a positive correlation between the more standard forms and the higher socio-economic demographic. Is there a right vs. wrong way to pronounce a word? The “normal” accent in Labov’s study, which would include the pronunciation of the /r/ phoneme, is called the standard accent and is termed unmarked by linguists. In other words, it is a variant that we have been conditioned to treat as normal or neutral and not recognize the privileged status it represents. Not pronouncing post-vocalic /r/ phonemes would be linguistically marked, that is, more noticeable because it is a deviation from the norm. The unmarked forms are valued more in mainstream aspects of society. Indeed, using an unmarked accent gives the impression of being more formal. Since the academic world values and prescribes standard forms, sounding standard also gives listeners the impression of the speaker sounding more educated (note the circular reasoning here). The reverse, of course, can be true: that speakers with many marked forms sound both less formal and less educated. However, the connotations of using marked forms can go further than that. As we


Behrens and Sperling

said earlier, talking about someone’s language use is a code for talking about the speaker. We might say such a speaker is not only lacking in education (didn’t she pay attention in school?) but, more seriously, also not able to be educated; that is, unintelligent.
Find recordings of speakers with different accents, or tape people you know who have different accents reading a passage from a book. You can find accented English samples at Play these audio samples for people in your daily life and ask them to • guess what the speaker looks like • guess what the speaker does for a living. Ask why they answered the way they did. Do you notice patterns in the listeners’ answers? What might this exercise indicate to you about the language attitudes of your speech community?

Exercise 3

Standard and Non-Standard Dialects
Language users do not simply differ from one another by their pronunciation of words. They also differ by grammar, verb endings, slang, and idiom use. A certain speaker from New York City might say “idear.” She might also say she “waits on line for coffee every morning,” that she “don’t like to wait long,” and that “me and my family are going to make a party soon for a friend tying the knot.” This speaker has a non-standard dialect. How is dialect different from accent? Many use the terms interchangeably, but linguistically an accent is variation at the pronunciation level while a dialect is variation manifested at more than one language level. The examples above showed our speaker using a preposition that varies from the prescribed norm (wait on line, not in line); a non-standard conjugation (she don’t); objective case pronoun in the subject position (me instead of I); a non-standard verb usage (make a party); and a idiomatic expression (tie the knot). While the example above illustrates a regional dialect, other demographics are associated with our language use. Our age, gender, ethnic group, class, and any non-English language influences in the home will also affect our dialects. Language users tend to have access to a range of dialects, and we are pretty good about making language choices that allow us to fit in with varying social circumstances. We also have a pretty good ability to violate the norms when we want to for specific effect.
Go back to the accented speakers you found in Exercise 3. Did they also display grammatical and vocabulary choices that differ from your own? What aspects of their speech could be labelled “accent” and what was “dialect”?

Exercise 4

Let’s return to the idea of an (ostensibly) unmarked, normal, socially unnoticed way of using our language. Linguists call this variety the standard dialect; the world of education calls it the standard language. This is the variety of a language that is valued by most mainstream aspects of society, so much so that many would not call it a variety or a dialect at all; it is commonly considered the language, and varieties and dialects are those forms that deviate from the standard. Linguists, however, consider all varieties of a language as dialects of that language family and the standard language as one of these dialects.

Language Variation


Linguists also consider all dialects to be communicatively effective, regular, and logical. Despite this, the standard language is the power dialect, the preferred dialect, chosen for use in print, on the air, and in the classroom. Speaking a dialect that closely approximates the standard dialect is usually associated with middle class and upper class socio-economic groups. So, simply through the act of speaking, we convey our social class and concomitant status to those who recognize the earmarks. Bourdieu, a French sociologist, discusses the concept of linguistic behavior having a market value (Bourdieu and Thompson, 1991). Our language choices denote qualities that are valued and can be exchanged for social prestige. The standard language has a high market value, especially in formal situations, school and the workplace among them. According to Bourdieu, for successful performance in mainstream aspects of society, standard language forms constitute our society’s valid linguistic currency. Given this, those who question the arbitrary nature of the standard risk fiscal compromise. For many people, the main factor that influences the acquisition of the standard dialect is the amount of standardized education to which they are exposed. In fact, educational systems are the main conveyor of the message that standard = right and good, and non-standard = inferior and wrong. Outside of a linguistics book or class, we rarely encounter the messages highlighted in this chapter describing the social construction of privilege via the institutionalization of preferred language forms. As a result, it is often hard (even for those of us who understand this clearly) to believe that the standard language is simply a dialect, one of many fully functioning dialects of English. There is an educational bias towards the standard forms of a language, largely because those controlling systems of education decide on the rules. The dialect of those inhabiting academic institutions has been elevated and codified to be the proper form of the language. Those who come from a middle or upper-middle class background use a home language already close to, or at, the standard. The transition to school dialect is much easier for them (because it is tailor made), compared to those accustomed to using non-standard forms at home and having to convert to the standard at school. Those whose first language is not English, and those from working class environments where a non-standard dialect form of English is often spoken, have to work harder to be successful in educational institutions in the US because they have two simultaneous tasks, one to learn content and the other to change their form of expression to the standard.

African American Vernacular English
In the 1990s, linguistics as a discipline appeared in the news over issues of what language forms students should, could, and must not use in the classroom. In 1996, the Oakland, California School Board put forth a resolution that the home language of the majority of their African American students was not Standard English (SE) but instead Ebonics, a language form stemming from the languages of the West and Niger-Congo areas of Africa. Further, the school system proposed to instruct their teachers in the language differences between the two forms to help the teachers better educate the students and lift test scores for the African American students in the school district. Their resolution was met with a social outcry. Students were quoted as being appalled that their school claimed they didn’t speak English. Parents were outraged that teachers would be trained to teach in Ebonics and withhold Standard English from their children. All of this was a distorted view of the situation, but it highlighted the strong connection people make between mastery of the standard language and educational and economic benefits. In response to the outcry, the Oakland School Board reworded its original proposal to emphasize the goal of proficiency in Standard English and de-emphasize the view that Ebonics was a language separate from English.


Behrens and Sperling

Theresa Perry, in a very sane book co-authored with Lisa Delpit and entitled The Real Ebonics Debate, writes:
The members of the Oakland school board had it right in their initial resolution when they affirmed the importance of fluency in Black Language and Standard English. They knew that fluency in the standard code can never be the singular goal if, and this is a big if, our schools are to participate in the creation of the next generation of AfricanAmerican scholars, preachers, dramatists, writers, blues men and women—AfricanAmerican leaders. (Perry, 1998, 15)

Further, the Linguistic Society of America (LSA), a national association of linguists, issued a position statement that applauded the Oakland move and affirmed the need for teachers to meet students where the students are linguistically. Ebonics was not a new term in 1996. The term was coined in 1973 by psychologist Robert L. Williams; linguist John Baugh had already written extensively, before Oakland, about language forms whose influences can be traced back to slavery, calling these languages Ebonic (cf. Baugh, 2000). Linguists had already documented a variety of English spoken widely in urban working class areas, usually—but not only—by African American users, termed alternatively African American Vernacular English (AAVE), Black English, or African American English (AAE). William Labov wrote about the language form’s logic and systematic nature in 1970. There is disagreement among linguists as to whether this variety of English is a dialect or constitutes a separate language. The original Oakland resolution took the latter view, fueling the fire of the ensuing criticism. But the distinction between two dialects of the same language and two forms that should be labeled separate languages is a social and political matter more than a linguistic one. Mutually comprehensible language forms such as Danish and Norwegian are considered separate languages, while the “dialects” of Chinese are so different as to be mutually incomprehensible. Yet they share a writing system and a national identity, and those criteria go into their label of “dialects.” Just as the term “Ebonics” did not first arise with Oakland, the idea of teachers meeting non-standard English users’ linguistic needs did not first surface with this case in California. Almost two decades before Oakland, the schools in Ann Arbor, Michigan, were challenged to build a bridge between the language form of their African American students and the language of the classroom, Standard English. In this 1979 case, however, the plaintiffs were the African American families of the school children, bringing charges against the school board that teachers were not knowledgeable about the children’s home variety of the language and their culture. The judge overseeing the case ruled that the schools were remiss in not overcoming the barriers between the children’s home language variety and the Standard English of the classroom. His proposed plan for the school board, however, “fell far short of the mark” in breaking down these barriers, according to scholar Geneva Smitherman (1998: 169). The publicity of the Oakland resolution seemed to put the term Ebonics, and the issue of tolerance for non-standard language forms, in the national media and conversation for the first time. Yet linguist Wayne O’Neil, writing in The Real Ebonics Debate, says, “Language prejudice remains a ‘legitimate’ prejudice; that is, one can generally say the most appalling things about people’s speech without fear of correction or contradiction” (1998: 42). The Ebonics story helps us make our point that the boundaries of the educational system reinforce the hierarchy of racial, ethnic, and class structures.

Language Variation


Exercise 5

Read the original Oakland School Board resolution at shoebox/oakland-ebonics.html (DeVoto 1997) and the amended resolution at Now find three different media reports on the Oakland resolution. Compare the overall media coverage of the Oakland resolutions and look for quotes from linguists in the news. Now read the position statement of the Linguistic Society of America at–57.html#1. Draw up a side-by-side comparison chart of the points the media and the LSA are making.

Bilingual Education
In 1998, another linguistic item appeared in the news: California millionaire Ron Unz, not an educator, linguist, or even a parent (but instead a failed candidate for governor), was spearheading a move to block government funding going to bilingual education. The largely native Spanish-speaking school population of California had been benefiting from a law that provided education in a child’s first language. Unz rallied support for a more stringent policy that moved students into English-only classrooms sooner by portraying these children as living in a non-English ghetto, kept from the rewards garnered by Standard English users. His bill was passed, and a similar bill was subsequently passed in Arizona in 2000. These two examples highlight the tremendous support Standard English has in many areas of the education world. Be it pitted against a dialect of English (Ebonics) or of a non-English language (Spanish or other native languages), Standard English should prevail.

Writing Practices in College
The gatekeeping role of the teacher demands that he or she require Standard English forms in students’ writing. Grading rubrics used in college writing courses will most likely factor “grammatical sentence structure” into an evaluation of a paper’s grade. Such a rubric is offered and discussed by Pfeifer and Ferree, in a journal article aimed at teachers entitled “Tired of ‘reeding’ bad papers” (2006). Their suggested rubric contains several criteria on a 1–5 scale, 5 being the highest. Under Criterion 3, Sentence Fluency, 1 out of 5 points is awarded “for lack of sound sentence structure,” while a full 5 points are given when “sentences are clear and the paper is easy to read” (p. 140). Non-standard grammar is equated with unclear and unreadable sentences. Criterion 4 is Proof Reading. A paper will receive 3 out of 5 points when “a few problems with grammar and punctuation cause the reader to stumble or pause now and again.” A mere 1 out of 5 points is awarded when the paper is “replete with grammatical errors and misuse of punctuation” (p. 141). A teacher, then, is unable to read or find clear thoughts in a paper with “ungrammatical” sentences. (Sample rubrics can be found at The ramifications of this reduction in grade are vast in terms of progressing in the educational system. Consider these three: • A student who has not mastered Standard English will not, by these guidelines, ever receive an A on a paper. • Scholarship awards will be limited. Access to higher levels of education will be compromised.


Behrens and Sperling

• The student’s chances of support by a mentor are lessened. It is ironic that we spend so much time shaping our linguistic selves to conform to the standard language when in fact Standard English is highly irregular. It is also everchanging.

Standard English: Irregular and in Flux
Despite being the privileged dialect, Standard English is pretty messy. Remember having to memorize all those irregular plurals and verb tenses? The most irregular words in languages tend to be the oldest and most common words. For English, consider the verb to be. We have irregularity in the present tense, the past, and little phonetic similarity between the infinitive be and forms like is, are, am, was, were. At least there is some family resemblance with been and being. Speakers of dialects that use a more regular system, such as the unconjugated I be, you be, she be, tend to garner negative social reactions, i.e. are negatively marked. Just as their language is negatively marked, so are they. Another irregularity in Standard English can be found in the pronoun system. The adjectival possessives are mine, yours, his, hers, its, ours, and theirs. Notice mine is the only form without a final “s.” Dialects that smooth out the irregularity and use mines generate another stigmatized form. One more example of the irregular nature of Standard English involves reflexive pronouns. To create reflexive pronouns, the rule appears to be: take the possessive pronoun and add “self” or “selves.” So we get “myself/ourselves,” “yourself/yourselves,” “herself,” and “itself.” Following this pattern, we should also get “hisself” and “theirselves.” While these forms are found in non-standard dialects, they are negatively marked; the Standard English forms break the pattern and use the objective case pronouns “him” and “them” as the roots of the reflexives. In other words, what we label as standard is not always regular. Some standard language users will take pride in having mastered all the irregularities, and might express annoyance at speakers who, supposedly, have not made the effort. We could imagine that teachers, who got ahead partly by mastering Standard English, fall into this category.
Consider the following sentences: 1 2 3 4 5 6 7 8 Everyone wants their life to be easy. None of these books are available on-line. I promise to never waste money again. Who did you give that gift to? Keep this secret between you and I. Either Nancy or myself will handle the problem later. I am taking two less courses this term. We will vote for whomever lowers taxes.

Exercise 6

Now, in small groups, judge the sentences above for structure and clarity. Ask three friends or family members, and three professors at your college to judge the sentences. Ask them to discuss the grammatical structure, clarity, and logic of each sentence. Do you see any differences in the answers given by the two groups of interviewees? Now consult style guides from the 1960s (or earlier) to the present to investigate any change in the standard usage of these aspects of English.

Language Variation


These sentences are all non-standard in some way. Yet they might sound fine to you. This is because the standard language is always in flux. For example, it used to be grammatically “correct” to use the masculine singular pronoun to refer to “everyone”: Everyone wants his life to be easy. In the 1970s, that usage started to be called sexist. Now some stylebooks require the non-sexist “his or her,” while many speakers use the plural pronoun “their.”

Social Indicators and Markers
Some deviations from a standard form are acceptable, even neutral, in societal terms and do not raise social eyebrows: choosing between the two pronunciations of the word “either,” for example, may or may not have social implications for a speaker; the same is true of choosing one of the two pronunciations of the word “often.” Your choice might convey some information about your social identity, but nothing that is highly marked. These varieties are called social indicators. Choosing among other varieties, however, might have largely negative associations, and these linguistic variables are called social markers. We say the pronunciation “aks” rather than “ask” is socially marked, as is the use of “don’t” for the third person singular (“she don’t”), double negatives, and what are called double modals (“I might could do that,” a form found among some speakers in the southern US). There are a few social markers that have positive connotations: using “whom,” for example. The social status of the user influences the acceptance of a usage. The phrase “between you and I” is one that can be heard frequently in educated circles. If we wanted to play “gotcha,” we could say that the use of the pronoun “I” in the objective position in this prepositional phrase is non-standard; however, this construction seems to carry overt prestige because (1) educated people tend to use it, and (2) “I” is perceived as more correct than “me.”

Sometimes the linguistic insecurity of a speaker is the force behind the overuse of a supposedly prestige form. For example, non-standard use of the reflexive pronoun “myself” is evident in the speech of those in power. Its use could stem from confusion about the difference between “I” and “me”; it could also be that “myself” sounds more important. “If you have any questions, ask Dr. Smith or myself” is a non-standard usage, but one that is often heard, especially in academic arenas. Along the same lines, here is an example of overuse of “whom”: we saw a t-shirt proclaiming, “I am for whomever beats Harvard.” The “whomever” usage is non-standard in this sentence since the pronoun is the subject of the predicate ‘beats Harvard.’ Such overuse of supposedly correct words, pronunciation, or structure is called hypercorrection. If you don’t quite know the way “whom” should be used, but believe that it is more prestigious than “who,” you might indeed overuse it. If hypercorrection stems from linguistic insecurity, which speakers tend to be the most insecure? Remember that W. Labov found that the middle class speakers he examined increased post-vocalic /r/ pronunciation more frequently than the other populations in conditions where they were made more aware of their own speech. In other words, they were hypercorrecting, using a greater number of post-vowel /r/ phonemes in their more formal, conscious pronunciations of words than did the upper class speakers, and importantly, more than one would expect from their other pronunciations. (This is a slightly different meaning of the term hypercorrection as reflected in the “whomever” usage, where the result was a non-standard form.)


Behrens and Sperling

We are all inconsistent in our language attitudes. A teenager might be corrected for saying that he or she “just graduated Bryant High School” (where did the preposition go? Don’t we “graduate from” a school?), but a business manager could say that this year’s budget “will impact the new project.” Would the businessman “get away” with using “to graduate” in the transitive sense? Would the teenager be admired for using “impact” in the corporate sense? Both verbs are being used as transitive verbs, in that these usages allow a direct object to immediately follow the verb. Both usages are relatively new, for these verbs were originally intransitive verbs and couldn’t be followed by a direct object. These usages, however, do not generally have the same social impact on listeners because of the age and social context of the user. Let’s look at another example of where one speaker can get away with non-standard usage and another cannot: the non-standard usage of double negatives. A student in a freshman composition course would never get away with writing the sentence “The character in the play did not do nothing to influence the plot.” A politician, however, could easily use the double negative in “It is not without regret that I announce my retirement.” Some double negatives are acceptable, it seems. You might argue that these two phenomena are different: the student’s double negative signals a negative meaning (and hence is illogical, following the logic that two negatives equal a positive), while the politician’s use of two negative words in the same sentence results in an intended (and logically) positive meaning. Both sentences, however, contain two negative words that suit the purpose of the speaker. Since the linguistic phenomenon is held constant in these two examples, then, the influential variable must be the language user.

Other Ways We Vary Our Language—and Why
While the chapter so far has focused on society’s privileging of standard language forms, in reality we all vary our language, and we all should. We talk and write differently depending on such factors as our regional influences, our level of education, and even the language use of our family and peers. And we continue to make choices about our language each time we use speech—or writing or signing or instant messaging— based on the people with whom we are communicating, the purpose of the exchange, and sometimes even the time of day. Style is a linguistic parameter we saw manipulated by Labov. Style is the level of formality we decide upon for our communication. Style encompasses a continuum of choices, from informal and relaxed to levels of formality that terminate in a frozen style, much like a politician or keynote speaker addressing a large audience. Style consists of choices at all levels of language: what vocabulary we select; our use or non-use of grammar-bearing morphemes such as possessive /s/ or third person singular /s/ or past tense vs. past participle (“I have went” vs. “I have gone”); syntactic decisions such as single vs. double negative; and use of simple, compound, or complex sentence types. In the informal style, we might say, “Hey” when greeting someone; formally, we might say, “How are you today?” Vocabulary associated with the informal style is called slang. Slang also tends to be associated with use by younger members of the population. In interactions with peers, which would mainly occur in intimate settings, informal style is deemed to be appropriate. Slang works to both solidify group identity and exclude outsiders. If you “get” the slang, you belong; if you don’t, you’re an outsider.

Language Variation


Exercise 7

Look up terms in a slang dictionary, such as the McGraw-Hill Dictionary of American Slang and Colloquial Expressions (4th edn.) (Spears, 2005) or Discuss in a small group the terms you are familiar with, and try to define the unfamiliar terms. Compare your answers.

Exercise 8

Correspond with a student at another school. Collect slang and other words that the student believes are specific to his or her school alone.

Jargon, like slang, works to include and exclude. Jargon is the vocabulary associated with a field of study or occupation. Hence the terms “legalese” and “medical jargon,” vocabulary that might be hard to decipher if we are not in the field. In fact, this book is conveying linguistic jargon right now!
Guess what field these terms are from and try to define them. • • • • • • • • • • • black hat hacker (vs. white hat hacker) deckle gutter juvenilia kiting latching milemarker paradiddle rubric stet western blot

Exercise 9

So far, we have discussed how we vary our language. But why do we vary? Many factors play a role in our language choices. Sociolinguist Dell Hymes developed an informative mnemonic device for remembering the eight basic social factors that influence our language choices. When we are about to speak, write, or sign, we calculate (at some quasi-automatic level) the following variables of the situation, the first letter of each making up the acronym SPEAKING.
S = the Setting: time and place of the communication event P = the Participants: participants could also include those not directly involved in the communication but present at the time, and even non-human or inanimate entities (one’s pet, a family portrait) E = the End result of the event: what the language user wants to result from the event A = the Act order: the order of turns or other linguistic units, e.g. who goes first, what is said first, etc. K = the Key or emotional tone of the event I = the Instrument(s) used: if the communication is speech, signing, writing, e-mail: and if it is being conducted in the standard dialect, one’s native language, second language, etc.


Behrens and Sperling

N = the Norms of the community as to how such communication usually happens: is this communication event adhering to or breaking the norms? G = the Genre of the communication event, the type of event it is: is this event a lecture, discussion, story-telling event, joke-telling, gossip? Source: Adapted from Hymes, 1974.

It is pretty amazing to think that, given all these variables, we can so quickly make socially appropriate decisions about our language choices, or purposefully choose to violate the social norms. Everything we do, linguistically, sends a message about ourselves to the world.

Let’s return to the classroom of SB and RS. Once we have worked together with the concepts presented in this chapter, we ask our students to free-write in their journals about their language attitudes. Some quotations are given in the box below.

Student Quotations (verbatim)
“I used to think some people’s different speech is funny. But now my attitude totally changed. It’s not funny. The different aspects of it is what makes it interesting.” “Language is a system of advantage and disadvantage that can disenfranchise people, so why don’t we study this in mainstream education the way we do racism?” “I am proud that my accent shows who I am and where I am from.” “I do not want to be the person from nowhere.” “For the first time, I actually thought about what impact language has on others’ lives.” “This topic can change attitudes about language in society because, by educating people on how and why different people talk the way they do, people may be more tolerant of others who speak differently and not judge them on the way they sound.” “A professor I had last year had an accent . . . and often it would be hard to understand him. However, I made it my responsibility to listen as carefully as possible to what he was saying.”

Some of our students agree with Anzaldúa: that they are their language and that to criticize a language form is to take a stand against the speaker. Some students begin to see that a standard language form perpetuates the continuation of a hierarchy, another way to judge groups of people having nothing to do with communicative effectiveness. Others continue to express the concerns they raised at the start of the exercise: that a loss of a standard will lead to a loss of communication. We hope that the concepts and issues we have raised in this chapter compel you to reflect on and continue to develop more complex understandings of social, political, and economic dynamics embedded in language use and language norms, e.g. cultural identity, financial mobility, mechanisms of inclusion and exclusion. As informed participants, better able to understand the consequences of language norms and attitudes (for ourselves

Language Variation


and others), you have the ability to make conscious and purposeful language choices, as both speaker and listener. We hope this is your experience.
What do you think? Take our Language Sensitivity Survey: Free-write on any changes you notice about your views towards language diversity after reading this chapter and completing the exercises. There may be no substantial changes. There might be some confusion for you about what you think. Just notice your thinking, especially as the semester continues and you talk to others about this class.

Exercise 10

4Teachers.Org. (2009) RubiStar [Internet]. Available from: index.php?screen=NewRubric&section_id=5#05 [Accessed June 10, 2009]. Anzaldúa, G. (1999) Borderlands/La frontera. San Francisco, Aunt Lute. Aristar, A. R. (1997) LSA resolution on Ebonics [Internet], The Linguist List. Available from:–57.html#1 [Accessed June 10, 2009]. Baugh, J. (2000) Beyond Ebonics: Linguistic pride and racial prejudice. Oxford, Oxford University Press. Bourdieu, P. and Thompson, J. B. (1991) Language and symbolic power. Cambridge, Harvard University Press. DeVoto, J. A. (1997) Oakland school board Ebonics resolution [Internet]. Available from: [Accessed June 10, 2009]. Fiske, R. H. (2009) The Vocabula Review [Internet]. Available from: http://www.vocabula. com/ [Accessed June 10, 2009]. Grammar Blog. (2009) Grammar Blog [Internet]. Available from: http://www.grammar [Accessed June 10, 2009]. Heath, S. B. (1983) Ways with words: Language, life, and work in communities and classrooms. Cambridge, Cambridge University Press. Hymes, D. (1974) Foundations in sociolinguistics: An ethnographic approach. Philadelphia, University of Pennsylvania. Labov, W. (1970) The study of nonstandard English. 8th edn. Illinois, National Council of Teachers of English. Labov, W. (2006) The social stratification of English in New York City. 2nd edn. Cambridge, Cambridge University Press. The Linguist List. (2009) Resolution of the board of education adopting the report and recommendations of the African-American task force; A policy statement and directing the superintendent of schools to devise a program to improve the English language acquisition and application skills of African-American students [Internet]. Available from: [Accessed June 10, 2009]. Mills, C. W. (1959) The sociological imagination. Oxford, Oxford University Press. O’Neil, W. (1998) If Ebonics isn’t a language, then tell me, what is? In: Perry, T. and Delpit, L. The real Ebonics debate. Boston, Beacon Press. Perry, T. (1998) ‘I’on know why they be trippin’’. In: Perry, T. and Delpit, L. The real Ebonics debate. Boston, Beacon Press. Perry, T. and Delpit, L. (1998) The real Ebonics debate. Boston, Beacon Press. Pfeifer, H. L. and Ferree, C. W. (2006) Tired of “reeding” bad papers? Teaching research and writing skills to criminal justice students. Journal of Criminal Justice Education 17(1), pp. 121–142.


Behrens and Sperling Richards, J. C. and Lockhart, C. (1994) Reflective teaching in second language classrooms. Cambridge, Cambridge University Press. Smitherman, G. (1998) “What go round come round.” In: Perry, T. and Delpit, L. The real Ebonics debate. Boston, Beacon Press. Spears, R. A. (2005) McGraw-Hill dictionary of American slang and colloquial expressions. 4th edn. New York, McGraw-Hill. Urban Dictionary. (2009) Urban Dictionary [Internet]. Available from: [Accessed June 10, 2009]. Weinberger, S. H. (2009) The speech accent archive [Internet]. Available from: http:// [Accessed June 10, 2009].

Other Resources
Lippi-Green, R. (1997) English with an accent: Language, ideology, and discrimination in the United States. London, Routledge. Wolfram, W., Adger, C. T., and Christian, D. (1999) Dialects in schools and communities. New Jersey, Lawrence Erlbaum.

American tongues. (1986) Directed by Louis Alvarez and Andrew Kolker. New York, Center for New American Media [video:VHS]. Do you speak American? (2005) New York, PBS/MacNeil-Lehrer Productions [video:VHS].

Center for Applied Linguistics. (2009) Dialects: African American English [Internet]. Available from: Ideonautics Corporation. (2007). Ideonautics [Internet]. Available from: http:// [Accessed June 10, 2009]. Meier, P. (1997) International Dialects of English Archive (IDEA) [Internet]. Available from:


Speech Communities Language as a Mediator of Messages and Perceptions

Miriam Eisenstein Ebsworth
This Chapter Explores:
Cross-linguistic Communication Discourse Styles Narratives Communicative Competence Reactions to Language Speech Acts Pragmatic Failure and Stereotyping Critical Incidents Sociolinguistic Sensitivity

I don’t get no respect! Rodney Dangerfield, 1980; comedian—born Jacob Cohen, 1921–2004

I was born in the United States into a Yiddish-speaking household. I learned the working class English of New York City as my second language. Part of my family in Argentina and Puerto Rico speaks Spanish, another of my languages. Starting in kindergarten, I attended a bilingual school where we learned in English in the morning and in Hebrew in the afternoon. Later on, I studied French in high school and college. For over 25 years, I have taught a broad range of students from kindergarten through graduate school, representing many speech communities. Currently I teach graduate students who are involved in second language pedagogy and research. The sociolinguistic diversity in my life has made me want to understand how members of different communities use language to understand each other. It has also made me realize that language can be a crucial link in how people feel about each other.

First impressions are important. We spend a lot of time choosing just the right clothing when going to particular events. Sometimes we want to look professional so that others will take us seriously. Other times we want to appear relaxed, or alluring to a potential love interest. Language also forms an impression. We react to others based on the way they speak as well as the way they look. The branch of linguistics that looks at the connection between language use and social factors is called sociolinguistics. An exchange can range from a chat on the telephone, a service encounter, a conversation



between teachers and students, or any talk relevant to business, friendship or family. After just a few moments, a listener can develop an impression, a feeling about what somebody might be like. In this chapter we will consider a range of topics that reveal connections between how language is used and its role in mediating communication between individuals and communities. We will also see that particular forms of language can affect not only our ability to send and retrieve messages accurately, but also our feelings about ourselves and others. From first impressions to choices of when to speak or be silent, we are constantly going beyond the boundaries of language and addressing issues of culture, traditions of expression, and sociolinguistic norms and values.

First Impressions and Contrasting Discourse Styles
A friend of mine who grew up with me in Brooklyn, New York—we’ll call her Sheila— had a typical New York working class accent (Labov, 2006). Like the speech of Archie Bunker in the classic 1970s TV comedy All in the Family, and of gangsters in The Sopranos, the New York accent is often depicted as associated with comic or rough characters, typically from working class backgrounds. My friend Sheila, however, was unaware there was anything unusual about her language use. She moved to California in the middle of high school. She wrote to me a short while later, saying she was upset that nobody seemed to listen to her ideas. Instead, people commented on her New York accent. Some people thought it was funny; others thought it sounded tough or sexy. The New Yorkese accent actually covers a broader geographic area that includes parts of New Jersey and environs (Eisenstein and Berkowitz 1981). It reflects several distinctive pronunciations of vowels and consonants (the vowel in “cab” is similar to the first vowel in “air”); the consonant sounds normally made with the tongue on the alveolar ridge are made with the tongue more forward towards the teeth, which would affect the pronunciation of the consonant in words like “time” (sometimes called a dentalized ‘t’). Sheila was unhappy in her new community and felt that she didn’t fully belong, in part because of her distinctive language use. But there were more differences than just pronunciation that affected how others responded to her. New Yorkese also has conversational norms. Deborah Tannen (1984) studied the conversation of friends and acquaintances, most of whom were New Yorkers and Californians, who met at a Thanksgiving dinner. She placed a tape recorder on the table with the turkey and recorded the conversation. Later, she transcribed what people had said and contacted them for their impressions. Among the differences in the speech of the two groups were distinctive rules for turn taking, expressing interest, and choosing topics appropriate for discussion with acquaintances. The New Yorkers at Tannen’s dinner made brief comments as others were speaking to show interest and enthusiasm (1993). Such comments are referred to by linguists as backchannels (Gass and Selinker, 2001). In addition, just as someone was about to finish what s/he was saying, a New Yorker started talking, eliminating the possibility of a silence in between turns. Tannen (1993) refers to this as cooperative overlapping. However, the Californians experienced the New Yorkers’ overlaps as interruptions. In contrast, when New Yorkers were speaking, the Californian listeners were comparatively silent, showing respect for what the other person was saying. In fact, they waited for a second, to be sure the New Yorkers had finished speaking before taking their turns. This behavior, intended to show consideration, was greeted with discomfort by New Yorkers. While Tannen (1993) acknowledged the uniqueness of each individual’s style, there were commonalities among the groups. She described the New York style as “high involvement” and the style of non-New Yorkers as “high-considerateness.” Notice that

Speech Communities


these positive terms encode the good intentions of each group of speakers rather than the way listeners from other communities may understand and interpret what they say. It is also the case that New Yorkers and Californians from different socio-economic groups and ethnicities may subscribe to the norms represented here to different degrees. Of course, individuals from different regions have equally distinctive sociolinguistic norms. An additional source of misunderstanding arose from a difference in the topics each group found appropriate for a “getting to know you” conversation. Californians tended to stick to impersonal, non-controversial topics such as film and fashion, while New Yorkers opted for more personal subjects. Despite the good intentions of both groups of people, their contrasting discourse styles did not allow them to communicate their feelings successfully. Beyond what this research shows about the two particular groups, it demonstrates that even speakers of the same language can misunderstand each other. Indeed, our language variety and the norms and conventions we use to express particular intentions to others can have a major effect on how others feel about us.

Exercise 1

Think of some different people you have met for the first time. What were the circumstances? What was your relation to the other person? Was his/her language like yours, or was it distinctive in some way? What did each of you say? How did you feel about the person? Now, imagine that you’re going to meet somebody in a social situation for the first time. What adjustments will you make in your language? What intention do you want to convey? Try to answer these questions for several different situations and different kinds of people.

Typically, people are under the impression that when somebody wants to be polite, the rules are essentially the same, irrespective of the individual’s language, language variety, culture, or subculture of origin. Individuals who have not had access to sociolinguistic contrasts or formal language study often have impressions and beliefs about how language is used. These impressions draw on understandings that are part of folk linguistics. From a folk linguistics perspective (Preston, 1996), it is thought that one need only translate from one language or variety into the new language or dialect to convey a friendly and respectful tone. However, as we will see, many aspects of polite—or impolite—language are specific to language varieties and subcultures.

Narratives and Monologues: Listeners Are Needed Too!
While Tannen’s work focuses on conversational interaction, even storytelling can pose problems. There are norms for how a story is structured, what is explicitly said, and what is not said or implied. There may even be contrasts in what the role of the listener to the story is supposed to be. The result is that even an event like storytelling, in which primarily one person has the floor, can affect our experiences and feelings. Coates (1996) notes that in Western culture usually one person speaks at a time when telling a story. For example, white middle class American listeners may show interest through “nods, minimal responses, laughter, and comments to express interest, sympathy, or surprise” (Polanyi, 1985). In contrast, in the African American community, the audience is often expected to respond while a story or talk unfolds (Foster, 2002; Smitherman, 1977).



How much you say is also linked to your speech community. Holmes (2003) gives the example of Maori narrators from New Zealand who “often leave meaning implicit.” Thus, Maori stories contrast with the relatively more explicit narratives of the European culture. Holmes (2003) points out that when a listener is from the European majority, the Maori style conflicts with expectations that a speaker will be explicit about the points to be made. This distinction cannot only interfere with communication; it can give both storyteller and listener a false impression of the other. Another distinction in communicative style is associated not with speech but with silence, an important aspect of language. We saw how even small differences in speech vs. silence resulted in negative feelings between New Yorkers and non-New Yorkers. This issue is not only a local one but also a global one. Mesthrie et al. (2000) explain that in particular communities there are norms for when it is appropriate to speak and when to remain silent. In fact, silence is not uncommon in a range of aboriginal cultures to show one is considering a serious topic, to show respect, and to convey a willingness to let another interlocutor complete his or her turn (Eades, 2004). Linguists often study the role of silence and pauses in particular languages and dialects since the use of pauses in particular places is an important part of the message sent and the impression made by a speaker. Language research may focus on “the temporal dimensions of speech” (Griffiths, 1991: 345), including speech rate, silent pause phenomena, and hesitations. Mora (1995) studied a class with non-native students and found that silence played an important role in the classroom. He found that differing orientations to silence were at the root of some communication difficulties and negative feelings between teacher and students. Silence can also result from a mismatch in student and teacher role expectations. While American culture encourages learners to express opinions even when they may differ from those of the teacher, other cultures may discourage such behavior. Sato (1982) notes a distinction between Asian and non-Asian participation in student–teacher interactions. Asian students, according to Sato (1982), felt less free to participate actively in an ESL course than their non-Asian peers. Ogbu and Matute-Bianchi (1986) also report that Punjabi children were inhibited from expressing their own opinions in discussions with adults. Thus, these relatively silent students may have something important to say but may be hesitant to express their ideas. Teachers could misconstrue this behavior and receive the false impression that quiet students do not have information to offer.

Exercise 2

Watch a drama, soap opera, or reality show on television or on the Internet. Record 15 minutes of it if possible. Notice what happens as characters take turns in the conversation. Pay special attention to the pauses and overlaps in the conversation. How do the overlaps and pauses make you feel? What do you think the characters are trying to convey to each other? Is there a difference in the conversations of the actors compared to the contestants on the reality shows? Now pay attention to the turn taking rules and silences in the conversation of friends and family members. Ask yourself and/or your friends, when do silences feel comfortable to you? When do they feel uncomfortable?

Communicative Competence and Feelings
We have seen that silence can be intended to show respect for the speaker; however, it can also be misinterpreted either as a refusal to answer or a lack of knowledge or

Speech Communities


engagement on the part of the listener. These contrasts highlight the challenge in adequately expressing our intentions and feelings to others. How do we use language to express those intentions and feelings? What is universal and what is language-specific? Clearly, there are more pitfalls to cross-linguistic communication than most people realize. We now understand that successful communication involves communicative competence. This construct, introduced by Hymes (1972) and expanded by Canale and Swain (1980) and Bachman (2000), includes all the information about a language a person needs to use it in a socially effective way and to understand the speech and writing of others, in terms not just of content but of implied meanings as well. Kasper and Rose (2002) remind us that successful communication involves the collaboration of other interlocutors as well as the knowledge of the speaker. We can react to differences in languages and language varieties on the following levels: pronunciation, grammar, vocabulary, idiomatic expressions, rules of conversation and writing, gender, and more. Language features, including lexicon (a formal term for vocabulary), grammar (the rules that govern how words are used in sentences), phonology (the sounds of the language and how to combine them), and semantics (deriving meaning from language), interact synergistically to send messages to others and help us retrieve the messages others are trying to send to us. Prosody, the melody of our speech, also conveys both meaning and emotion. Prosody includes alterations in our pacing and pausing, pitch, and amplitude. Our language can also be tied to our emotions more directly. The particular tone, rhythm, and stress we assign to the way we say things convey how we are feeling to others. Often, we’ll say to people we know, “You sound upset,” “You sound angry,” or “You sound tired.” We are not necessarily reacting to the actual content of what they are saying. Rather, we are responding to prosodic elements in their speech that convey an emotional tone.
First, imagine you are waiting for somebody you know who has just walked in. First, say the following sentence in a neutral way. I see you’re here. Then, imagine that you are very angry at that person and want to convey that to him/ her. Say the same sentence in an angry way. I SEE YOU’RE HERE! What were the differences in the way you said the two sentences? Was one louder than the other? Did you stress different words? Did your pitch change in any way? Now try the same sentence to show you’re very happy to see that person. And try it once more to show romantic interest. As a variation, try doing this in pairs and have one person be the speaker and one the listener.

Exercise 3

For those of you who speak more than one language, how would you convey different reactions in your second language? Would you know how to express these ideas and what prosodic features to use? Despite the fact that emotions are universal, the way we use prosodic features to convey them changes from language to language and from variety to variety. When speakers use the emotional tone of their first language or dialect to speak with somebody from another community, the emotions they wish to convey can be misunderstood. Rintell (1989, 1990) studied how native and non-native speakers recognized emotional tone in the language of others. She found that even relatively advanced second language learners had difficulty recognizing the emotion contained in the speech of



others. If somebody intends to sound friendly or helpful and the listener receives a different message, the resulting challenges to relationships are not difficult to imagine. Our ability to make friends in a new community may rest exactly on being able to read the emotional tone in the language of others and sound the way we intend to them. If people seem distant and unfriendly, even when they are not, this can make us feel bad and create unnecessary barriers to forming and maintaining relationships. Once again, language is the link between our feelings about others and ourselves. Indeed, when two individuals or groups are operating under differing rules of communicative competence, problems can occur, often tapping the feelings of speakers and hearers, readers and writers. Non-verbal elements of discourse such as gestures, touch, or the use of space are also tied to language and culture and add an important element to the messages we send and receive (Philippot et al., 1999). When Americans meet someone for the first time, they often stand about a handshake length away from the other person. Latinos may be more comfortable standing a bit closer, an arm’s length away. Individuals from Asian cultures may position themselves relative to each other so that there is room to bow. When relationships become more personal, these parameters may shift, but when distances in cross-cultural encounters don’t match, people may feel uncomfortable and not even know why (Miller, 1988).

Researching Reactions to Language
A useful technique for discovering how people feel about speakers of particular languages and varieties was developed by Lambert, Hodgson, Gardner, and Fillenbaum in 1960. Here is how their matched guise methodology works. A listener is instructed to try to get a sense of what speakers might be like, just from hearing them speak, perhaps on tape. Actually, the listener is being exposed to one person telling a story or doing a reading two different ways, either in two different languages or alternate varieties of the same language. The tapes are matched because the same person is speaking, providing a control for differences in voice and personality that might influence results. The guise is the fact that most listeners assume they are hearing two different individuals rather than one. Often, listeners are asked to rate the speaker on each tape quantitatively on semantic differential scales with opposite adjectives such as friendly–unfriendly, educated–uneducated, or tall–short (Osgood, 1964). Eisenstein (1986) used the matched guise technique to show that users of alternative non-prestige dialects can be stereotyped. Reactions come from both native speakers and second language learners. In a series of studies of native and non-native reactions to familiar and unfamiliar varieties of English using matched guise and follow-up interviews, She found that not only did native speakers attribute negative characteristics to speakers of New Yorkese and Black English (also known as African American Vernacular English or Ebonics), but English language learners living in the community acquired the judgments of their native neighbors towards these individuals. In this study, she asked college students to listen to paired sentences and choose the more prestigious of the two when they were not identical, respond to matched guise tapes on a series of semantic differential scales, and choose a picture the listener thought best represented the speaker. Follow-up interviews explored participants’ understandings of why they chose particular pictures, whether they had encountered other such speakers before, and under what circumstances, and the nature of the relationship they might like to have with the speaker. Not only did the interviews confirm the quantitative data, but they also indicated that listener attitudes derived from a range of sources, including television, film, exposure to the attitudes of others, and media personalities.

Speech Communities


Exercise 4

Identify some personalities in the media and in film, TV, or radio who have distinctive speech styles and accents. Who sounds educated or uneducated? Who sounds friendly or snobbish? Who speaks the way you’d like to speak? What speech pattern would you want to avoid? Notice what elements of each person’s speech you find distinctive. Can you associate these distinctions with the feelings you get when you listen?

Pragmatics and Speech Acts
One of the least well known but most important aspects of language that is linked to our feelings about ourselves and others is associated with speech acts, an area of pragmatics. Pragmatics, defined broadly, includes a range of areas such as presupposition, issues of discourse markers and coherence, appropriateness, and politeness (Koike, 1989; Searle, 1969). A speech act links an interlocutor’s intention to a particular utterance, an intention that may go beyond the literal meaning of what is actually said or written (Horn and Ward, 2006; Mey, 2001). Some examples of speech acts include: suggestions, greetings, apologies, compliments, invitations, acceptances, criticism, and even insults. One must have an excellent command of a language variety to insult somebody successfully. Sometimes a speech act says exactly what it means. For example, I promise to try harder from now on. This semantic formula consists of the verb “promise” plus a description of what the promise entails. While the speaker might or might not be sincere in making the promise, most listeners will recognize the “official” intention being stated. This is a direct speech act; its intention is overtly stated. But every language and variety has particular “semantic formulas,” fixed expressions to express particular intentions, that don’t literally state what the speaker means. For example, in the common expression “Can I leave the room?” everybody understands that the speaker is asking for permission. But sometimes the listener plays with the language and says something like, “You can leave but you may not.” This contrasts the intention of the speech act with what it literally means. Can I plus action is a conventionally indirect speech act. It uses a semantic formula that is common in the language. Often, speakers don’t even realize that they are using indirect speech acts. Sometimes speakers use language in a new way to send a message indirectly. When this happens they are using an unconventionally indirect speech act (Barron, 2003). For example, Blum-Kulka and Olshtain (1984) describe situations in which people refuse indirectly. In one example, a woman wishes to say no indirectly to an offer of a date with a man. A conventionally indirect formula might be, “I’m busy.” This is such a common excuse that if a woman is really busy and actually wants to show interest in the other person, she might add, “Honestly, I’m really busy doing X. But I’d love to get together some other time.” Blum-Kulka and Olshtain (1984) report the memorable unconventionally indirect response, “I’m sorry. I’m a nun!”

Exercise 5

Think of a situation that would require an apology or get you out of an unwanted invitation, as above. Now see if you can come up with both direct and indirect formulas to convey the apology or refusal to others. Think of another common speech act and a typical context for its use. Come up with some different ways the


Ebsworth other person might respond to the apology, assuming the apology was accepted. What might be said if the apology were not accepted, at least at first? Again, imagine direct and indirect ways feelings might be expressed.

Pragmatic Failure and Stereotyping
Eisenstein and Bodman (1986, 1993) studied intercultural expressions of gratitude, and Eisenstein, Bodman, and Carpenter (1995) researched cross-cultural greetings. They found that in both cases, these speech acts were expressed in different situations and in different ways by speakers of different languages. Not everybody said “thank you” in the same situations, and people used different semantic formulas to do so. The nature and length of greetings also changed from one sociolinguistic group to another. For example, the “greeting on the run,” common in New York English, was found to be insulting to non-native speakers from several backgrounds. Also, the need to greet somebody in a service encounter was much more important in the Southern US than in the North. The direct “getting down to business” approach common in New York City was perceived as rude in other places (Beebe, 1997). Thomas (1983, 1995) explains that when people from different speech communities misunderstand each other’s intentions, the result is pragmatic failure. This failure cannot only stem from a distinction in transferring what is typically done or said in particular situations in one’s speech community to another, sociopragmatic failure; it can also happen when a non-native speaker’s command of the second language just isn’t strong enough for the message to get across, pragmalinguistic failure. Of course, both types of failure can happen at the same time. When contact with individuals from different communities repeatedly results in pragmatic failure, participants in the conversation could erroneously conclude that members of the other community are not polite or have other negative qualities. House (2003) notes that the perception of politeness is central to the analysis of misunderstanding as it is a basic social guideline for human interaction. She finds, “Interfering emotions tend to be exacerbated in the process of cross-cultural communication, because it involves interactants who habitually use different communicative styles” (2003: 27). Misunderstandings, House suggests, can become emotionally charged when people from different cultures are involved. Potential misunderstandings include a range of areas that are vulnerable to pragmatic failure, e.g. the introduction of a taboo topic or the difference in relative status between participants (Turner and Hiraga, 2003). In some countries it is acceptable to inquire about people’s ages or salary levels in order to learn more about them and show interest. However, this would not be recommended in an American context. An international student who had been invited to a welcome tea being given by the dean and her husband tried to show appreciation by offering the following compliment, “Hey, that’s a really cool dress!” The dean, a refined woman in her fifties, did not find the comment appropriate.

Responses to Critical Incidents
Beyond conversational norms and contrasts in how particular speech acts and semantic formulas may represent speakers’ intentions, the strategies used to solve the challenges of everyday situations put pragmatics to use in crucial ways. A situation in which it is difficult to know exactly what to say and do in order to influence a successful outcome is called a critical incident. Such situations are particularly challenging when the people involved come from different sociolinguistic and cultural backgrounds.

Speech Communities


The work of Ebsworth and Ebsworth (1997, 2000) illustrates how the way we deal with critical incidents can affect the feelings of the people involved. The researchers found that alternative approaches to critical incidents for Americans and multicultural Puerto Ricans speaking English resulted in differing understandings and responses. First, through natural observation and personal interviews, the researchers identified five critical incidents that had created problems for participants in cross-cultural situations in contexts including business, social, and academic areas. Having elicited typical responses from members of each group, they had a large number of participants judge the alternative answers on semantic differential scales (the same kind used in Lambert’s research; Lambert et al., 1960). Consider the circumstances and reactions of the two critical incidents highlighted in the boxes below.

Critical Incident A
A stranger pulls up in a car and asks you about a neighbor. Although you know the person, you are not sure how much information to reveal. Typical response from Puerto Rican subjects in study: chat with the person for a while and try to find out more. If a participant decided not to share much, s/he might simply say, “I’m sorry. I don’t know the person you’re looking for.” Typical response from American subjects in study: say, “It’s nothing personal, but I don’t give out information to strangers.”

While other responses were suggested, these two were interpreted differently by the two groups in the study, with the first evaluated more positively by Puerto Ricans and the second evaluated more positively by Americans. Interviews with bicultural experts revealed that the major distinctions in judgments involved the difference between how to treat a stranger and the distinction between a real lie and a white lie. The Puerto Rican informants said they felt they owed attention to a stranger, as long as s/he was polite and acting appropriately. If they decided not to reveal the information about the neighbor, they would tell a white lie, indicating they did not know the person, rather than implying the stranger might not be trustworthy. They found the Americans’ “nothing personal” strategy to be unacceptable. By contrast, the Americans thought the “nothing personal” strategy was a reasonable solution that would not cause the stranger to take offense as it was a general policy, not directed at anyone in particular. They felt it was not necessary to be dishonest in this situation. Thus, the research revealed that underlying the alternative pragmatic strategies of the two groups were subtle differences in social norms and values that affected behavior. The research also offered clues into how such incidents could affect how group members felt about the opposite group. Consider the incident below.

Critical Incident B
You are late for a dinner party being given in your honor when a friend from out of town shows up unexpectedly at your door. Typical response from Puerto Rican subjects in study: invite the friend in and spend substantial time with him or her, and then show up really late for the dinner, explaining what happened on arrival.


Ebsworth Typical response from American subjects in study: quickly explain the situation and apologize for running off after just a few moments of conversation.

Interviews showed that for some Puerto Ricans, the American solution of being left abruptly by the person going to the dinner would make them feel hurt and insulted. In contrast, for Americans, the host of the dinner party would feel the guest was very inconsiderate to be so late without calling ahead, and could end the friendship as a result of the affront.

Exercise 6

Come up with a few critical incidents from your own experiences, situations where you had to make decisions about what to say. Think of how they developed and how you felt about the other person when they were over. Now share some of these incidents with others and see what language and strategies they might have used in addressing them. What do acceptable and successful or unacceptable and unsuccessful language and strategies tell you about what is important in your culture? Try to find somebody who speaks another language or dialect. (If you don’t know somebody personally, you might consider trying the Internet to find an e-pal to work with.) See how that person would approach the critical incidents. How do you feel about the solution given? How does that person feel about yours? Are any differences in cultural norms and values revealed by this exchange?

The Process of Acquiring a Second Language or Variety
What about the feelings of individuals who are themselves speakers of a second language and/or alternative varieties towards these varieties and their speakers? How does their awareness motivate the acquisition of the additional language or variety? Verdi (2000) investigated the journeys of four working class women as they became acquainted with the mainstream language and culture of the academy. The women explored their feelings of frustration and challenge and ultimately pride, as they discovered the dissonance between their own language and culture and that of English speakers who understood standard academic language and mainstream cultural norms and values. Annas noted the many differences confronted by working class individuals who attend college; these include dichotomies of language, dress, and even how anger is expressed (1993: 171). One woman in Verdi’s (2000) study related that her attempt to use a more standard language resulted in rejection by her original peer group, revealing the need to become bicultural and bidialectal in order to maintain original associations and community membership. Dykman (1999) studied the texts of three authors who acquired English as a second language and wrote about their experiences. Focusing on their early language learning experiences, she interviewed each one and found that all had strong feelings about their original and acquired languages and cultures. All three authors thought English sounded harsh at first, compared to their native Russian, Polish, and Spanish. Dykman notes that Rodriguez (1982), Hoffman (1989), and Lvovich’s (1997) sense of alienation seemed to coexist with their need to communicate (Dykman, 1999: 83). Being forced to speak a foreign language was particularly painful, since it entailed the possibility of being misunderstood and making a fool of oneself in public. The language learners experienced feelings of nostalgia and frustration as they tried to express themselves in the second

Speech Communities


language. Dykman concurs with Schumann’s (1997) position that “it is emotion that underlies most, if not all cognition, and that variable success in second language learning is, in large measure, emotionally driven” (Dykman, 1999: 95). Pavlenko (2003, 2005) explores many dimensions of the language and emotion connection for second language learners. As a female graduate student in a second language environment, she found that it was difficult to present her ideas in ways that were taken seriously in class. “Oftentimes my questions and statement were ignored, and I was beginning to wonder about their validity” (Pavlenko, 2003: 184).
Either from your own experience or from a story you know, think of an instance in which you or a character must confront a new language, culture, or subculture. What are the challenges in learning how to communicate and understand others? How might the initial mismatch between the first language and culture and the new one affect the feelings of the individuals involved?

Exercise 7

Early Development of Sociolinguistic Sensitivity
Our awareness of language variation and its association with particular communities starts when we’re very young. Rosenthal (1977) studied young children using “Magic Boxes,” cardboard boxes with painted red faces and blue ears and noses, each of which contained a tape recorder that spoke to the children. By presenting tape recordings of speakers using Black English or Standard English, Rosenthal (1977) discovered that even very young children ascribed particular kinds of speech to members of different language communities when asked to match a picture with a tape-recorded voice. By the age of five years, these young children also associated different varieties with socioeconomic status, as revealed by their responses to two magic boxes. When the researcher asked children whether they wanted to give a gift to the speaker or take a gift from the speaker, the children made comments like: Kenneth, the Black English speaker, needs it more “cause Kenneth doesn’t have nothing”; or Steve, the Standard English speaker, “is better off” (Rosenthal, 1977: 58–59). Some of the children revealed emotional responses to one or the other speaker, clearly mediated by the variety of English spoken. One child commented, “I like him [Steve] cause he sound nice; I don’t like him [Kenneth].” Cazden (2001) and Michaels (1984) also found that how children structure their stories is tied to their sociolinguistic community and can affect their feelings about school and themselves. Holmes (2003) explains, “A story is an account of an event, structured in a particular way, which in western culture basically entails having a beginning, a middle, and an end.” In her landmark study, Michaels (1984) described how African American children who used a topic-associating story structure found it hard to tell their stories during “sharing time,” also known elsewhere as “show and tell.” This oral narrative in which children bring something from home to tell their fellow students about is considered an important bridge to written discourse. However, African American children whose teachers were unfamiliar with the discourse structure of African American English stories were puzzled by children’s organization of ideas and constantly interrupted them, urging them to stay on topic. Many of the children expressed feelings of extreme frustration as a result of this experience. Melzi (2002) studied the co-construction of stories by American and Latina mothers of Central American origin. She found that the elicitation styles of the mothers differed, with “European American” mothers (from families with European roots) actively



keeping the children focused on narrative structure, centered on a single topic. In contrast, the Latina mothers emphasized “maintaining a conversation that allowed children to share experiences.” Choi (1992) found that, compared to Canadian mothers, whose behavior was similar to that of Melzi’s (2002) European American group, Korean mothers restricted the degree to which young children could shift the topic or add their own information to a conversation. Rather, the child’s role was more to confirm the information the mother had presented.
Observe a caretaker and young child co-constructing a story about a past event. To what extent does the caretaker focus on keeping the child’s story within a particular structure? To what extent is the child able to follow her/his own thought patterns in constructing the conversation? If you aren’t able to observe first hand, ask somebody among your family or friends who cares for a young child to reflect on what he or she does.

Exercise 8

Approaching Language and Feelings
We have seen that language and feelings are intertwined and can pose a range of problems from between-group understanding to personal acceptance. What are solutions from personal, educational, social, and policy perspectives? We can see that non-native speakers and speakers of alternative varieties must face many challenges in the feelings they evoke in others. Often such negative experience translates into a loss of self-esteem (Cummins, 1989), as well as difficulties in finding acceptance and understanding in the mainstream culture. Certainly, there are examples of individuals who have overcome such challenges. In fact, teaching of the heritage language has been found to increase feelings of self-esteem among minority students (Wright and Taylor, 1995). In Verdi’s (2000) study, all four working class women succeeded in becoming university academics, despite the limitations created by the dissonance of their vernacular languages and subcultures. Each woman was ultimately able to successfully navigate both worlds and understand the language, norms, and values of the various communities she belonged to. One of the participants in the Tannen (1993) dinner, Brad from Los Angeles, who found the New Yorkers’ conversational style difficult, took a trip to Alaska, only to discover that much longer silences were common in verbal interaction there than he had experienced in California. He returned with the insight that the Alaskans viewed his speech style in much the same way as he had viewed the New Yorkers’. Ultimately, the key to having good feelings about ourselves and others lies in understanding and education. The more we comprehend the different ways that language can be used to express ideas and feelings, the more likely we are to accept others for the good people they are. We can go beyond any unintentional violations of our own sociolinguistic norms and work on understanding the messages they are trying to send. If we can succeed in doing this, cross-community and cross-cultural understanding will be increased. And Rodney Dangerfield may finally get the respect (albeit posthumously) that had eluded him for so long.

Thanks to Justin Bennett and Ilka Henits for help with this chapter.

Speech Communities


Annas, P. (1993) Pass the cake: The politics of gender, class, and text in the academic workplace. In: Tokarczyk, M. and Fay, E. A. eds. Working class women in the academy. Amherst, University of Massachusetts Press, pp. 165–178. Bachman, L. (2000) Modern language testing at the turn of the century. Language Testing, 17(1), pp. 1–42. Barron, A. (2003) Acquisition in interlanguage pragmatics. Philadelphia, John Benjamins. Beebe, L. (1997) Rude awakenings: Ways of responding to rudeness. In: Bouton, L. F. ed. Pragmatics and Language Learning Monograph Series, no. 8. Illinois, University of Illinois at Urbana-Champaign, pp. 1–36. Blum-Kulka, S. and Olshtain, E. (1984) Requests and apologies: A cross-cultural study of speech act realization patterns. Applied Linguistics, 5(3), pp. 196–213. Canale, M. and Swain, M. (1980) Theoretical bases of communicative approaches to second language teaching and testing. Applied Linguistics, 1(1), pp. 1–47. Cazden, C. (2001) Classroom discourse: The language of teaching and learning. 2nd edn. Westport, Heinemann. Choi, S. H. (1992) Communicative socialization processes: Korea and Canada. In: Iwasaki, S., Kashima, Y., and Leung, K. eds. Innovations in cross-cultural psychology. Amsterdam, Swets and Zeitlinger, pp. 103–122. Coates, J. (1996) Women talk: Conversation between women friends. Oxford, WileyBlackwell. Cummins, J. (1989) Empowering minority students. Sacramento, California Association for Bilingual Education. Dangerfield, R. (1980) No respect. MCA Special Products [sound recording: CD]. Dykman, E. (1999) Exploring second language acquisition and acculturation through autobiographical texts: A qualitative study of second language learners/authors. Ph.D. thesis (unpublished), New York University. Eades. D. (2004) Understanding aboriginal English in the legal system: A critical sociolinguistics approach. Applied Linguistics, 25(4), pp. 491–512. Ebsworth, M. E. and Ebsworth, T. (1997) Island Puerto Rican and continental American responses to critical incidents: An intercultural study of language, norms and values. Journal of Research on Language and Social Interaction, 30(3), pp. 193–233. Ebsworth, M. E. and Ebsworth, T. (2000) The pragmatics and perceptions of multicultural Puerto Ricans. International Journal of the Sociology of Language, 142(3), pp. 119–155. Eisenstein, M. (1986) Target language variation and second language acquisition: Learning English in New York City. World Englishes, 5(1), pp. 31–46. Eisenstein, M. and Berkowitz, D. (1981) The effect of phonological variation on adult learner comprehension. Studies in Second Language Acquisition, 4(1), pp. 75–80. Eisenstein, M. and Bodman, J. (1986) I very appreciate: Expressions of gratitude by native and non-native speakers of American English. Applied Linguistics, 7(2), pp. 167–185. Eisenstein, M. and Bodman, J. (1993) Expressing gratitude in American English. In: Kasper, G. and Blum-Kulka, S. eds. Interlanguage pragmatics. New York, Oxford University Press, pp. 64–81. Eisenstein, M., Bodman, J., and Carpenter, M. (1995) Cross-cultural realization of greetings in American English. In: Gass, S. M. and Neu, J. eds. Speech acts across cultures. Berlin: Mouton de Gruyter, pp. 89–108. Foster, M. (2002). Using call and response to facilitate language mastery and literacy acquisition among African American students. Washington, D.C., Center for Applied Linguistics. Gass, S. and Selinker, L. (2001) Second language acquisition: An introductory course. 2nd edn. Mahwah, Lawrence Erlbaum.


Ebsworth Griffiths, R. (1991) Pausological research in an L2 context: A rationale and review of selected studies. Applied Linguistics, 12(4), pp. 345–364. Hoffman, E. (1989) Lost in translation: A life in a new language. New York, E. P. Dutton. Holmes, J. (2003) I couldn’t follow her story: Gender and ethnic differences in New Zealand narratives. In: House, J., Kasper, G., and Ross, S. eds. Misunderstanding in social life. London, Longman, pp. 173–198. Horn, L. R. and Ward, G. L. (2006) The handbook of pragmatics. Massachusetts, WileyBlackwell. House, J. (2003) Misunderstanding in intercultural university encounters. In: House, J., Kasper, G., and Ross, S. eds. Misunderstanding in social life. London, Longman, pp. 22–56. House, J., Kasper, G., and Ross, S. (2003) Misunderstanding in social life. London, Longman. Hymes, D. (1972) On communicative competence. In: Pride, J. B. and Holmes, J. eds. Sociolinguistics. Harmondsworth, Penguin, pp. 269–293. Kasper, G. and Rose, K. (2002) Pragmatic development in a second language. Language Learning, 52 (Supplement 1), pp. 1–339. Koike, D. A. (1989) Pragmatic competence and adult L2 acquisition: Speech acts in interlanguage. Modern Language Journal, 73(3), pp. 279–289. Labov, W. (2006) The social stratification of English in New York City. 2nd edn. New York, Cambridge University Press. Lambert, W. E., Hodgson, R. C., Gardner, R. C., and Fillenbaum, S. (1960) Evaluational reactions to spoken languages. Journal of Abnormal and Social Psychology, 60(1), pp. 44–51. Lvovich, N. (1997) The multilingual self: An inquiry into language learning. Mahwah, Lawrence Erlbaum. Melzi, G. (2002) Cultural variations in the construction of personal narratives: Central American and European American mothers’ elicitation style. Discourse Processes, 30(2), pp. 153–177. Mesthrie, R., Swann, J., Deumert, A. and Leap, W. (2000) Introducing sociolinguistics. Edinburgh, Edinburgh University Press. Mey, J. L. (2001) Pragmatics: An introduction. 2nd edn. Oxford, Wiley-Blackwell. Michaels, S. (1984) Listening and responding: Hearing the logic in children’s classroom narratives. Theory into Practice, 23(3), pp. 218–224. Miller, P. W. (1988) Nonverbal communication: What research says to the teacher. 3rd edn. West Haven, NEA Professional Library. Mora, R. (1995) Silence, interruptions, and discourse domains: The opportunities to speak. Applied Language Learning, 6 (1–2), pp. 27–39. Ogbu, J. U. and Matute-Bianchi, M. E. (1986) Understanding sociocultural factors in education: Knowledge, identity, and adjustment in schooling. In: Evaluation, Dissemination, and Assessment Center, California State University ed. Beyond language: Social and cultural factors in schooling language minority students. Sacramento, Bilingual Education Office, California State Department of Education, pp. 73–142. Osgood, C. E. (1964) Semantic differential technique in the comparative study of cultures. American Anthropologist, 66(3), pp. 171–200. Pavlenko, A. (2003) The privilege of writing as an immigrant woman. In Casanave, C. and Vandrick, S. eds. Writing for scholarly publication: Behind the scenes in language and multicultural education. Mahwah, Lawrence Erlbaum, pp. 177–196. Pavlenko, A. (2005) Emotions and multilingualism. Cambridge, Cambridge University Press. Philippot, P., Feldman, R. S., and Coates, E. J. (1999) The social context of nonverbal behavior. Cambridge, Cambridge University Press. Polanyi, L. (1985) Telling the American story: A structural and cultural analysis of conversational storytelling. Norwood, Ablex.

Speech Communities


Preston, D. (1996) Whaddayaknow?: The modes of folk linguistic awareness. Language Awareness, 5(1), pp. 40–74. Rintell, E. (1989) That reminds me of a story: The use of language to express emotion by second-language learners and native speakers. In: Eisenstein, M. ed. The Dynamic Interlanguage. New York, Springer, pp. 237–256. Rintell, E. (1990) That’s incredible: Stories of emotion told by second language learners and native speakers. In: Scarcella, R., Andersen, E., and Krashen, S. eds. Developing communicative competence in a second language. Boston, Heinle and Heinle, pp. 75–94. Rodriguez, R. (1982) Hunger of memory: The education of Richard Rodriguez. Boston, David R. Godine. Rosenthal, M. (1977) The magic boxes: Children and black English. CAL/ERIC-CLL Series on Languages and Linguistics, 43. Illinois, University of Illinois Press. Sato, C. (1982) Ethnic styles in classroom discourse. In: Hines, M. and Rutherford, W. eds. On TESOL ’81: Selected papers from the fifteenth annual conference of teachers of English to speakers of other languages. Washington, D.C., TESOL, pp. 11–24. Searle, J. (1969) Speech acts: An essay in the philosophy of language. Cambridge, Cambridge University Press. Schumann, J. (1997) The neurobiology of affect in language. Oxford, Wiley-Blackwell. Smitherman, G. (1977) Talkin and testifiyin: The language of black America. Boston, Houghton Mifflin. Tannen, D. (1984) Conversational style: Analyzing talk among friends. Norwood, Ablex. Tannen, D. (1993) Gender and conversational interaction. New York, Oxford University Press. Thomas, J. (1983) Cross-cultural pragmatic failure. Applied Linguistics, 4(2), pp. 91–112. Thomas, J. (1995) Meaning in interaction: An introduction to pragmatics. London: Longman. Turner, J. and Hiraga, M. (2003) Misunderstanding teaching and learning. In: House, J., Kasper, B., and Ross, S. eds. Misunderstanding in social life: Discourse approaches to problematic talk. London, Pearson Education, pp. 154–172. Verdi, G. (2000) Navigating language and cultures: An ethnographic study of four workingclass women academics. Ph.D. thesis (unpublished), New York University. Wright, S. C. and Taylor, D. M. (1995) Identity and the language of the classroom: Investigating the impact of heritage versus second language instruction on personal and collective self-esteem. Journal of Educational Psychology, 87(2), pp. 241–252.

Other Resources
Adger, C. T., Christian, D., and Taylor, O. eds. (1999) Making the connection: Language and academic achievement among African American students. McHenry, IL, CAL, and Delta. Brown, S. and Attardo, S. (2000) Understanding language structure, interaction, and variation: An introduction to applied linguistics and sociolinguistics for non-specialists. 2nd edn. Ann Arbor, University of Michigan Press. Chambers, J. K. (1995) Sociolinguistic theory: Linguistic variation and its social significance. Cambridge, Blackwell Publishing. Champion, T. (2003) Understanding storytelling among African American children: A journey from Africa to America. Mahwah, Lawrence Erlbaum. Conrad, S. and Biber, D. (2001) Variation in English: Multi-dimensional studies. UK, Longman/Pearson Education. Garcia, O. and Fishman, J. (1997) The multilingual apple: Language in New York City. Berlin, Walter de Gruyter. Gass, S. and Lefkowitz, N. (1995) Varieties of English. Ann Arbor, University of Michigan Press.


Ebsworth Kanno, Y. (2003) Negotiating bilingual and bicultural identities: Japanese returnees betwixt two worlds. Mahwah, Lawrence Erlbaum. Lantolf, J. (2000) Sociocultural theory and second language learning. Oxford, Oxford University Press. Lippi-Green, R. (1997) English with an accent: Language, ideology, and discrimination in the United States. London, Routledge. McCabe, A. and Bliss, L. (2003) Patterns of narrative discourse: A multicultural life span approach. Boston, Allyn and Bacon. Markee, N. (2000) Conversation analysis. Mahwah, Lawrence Erlbaum. Michaels, S. (1980) Sharing time: Oral preparation for literacy. Proceedings of the Ethnography in Education Research Forum, March 21–23, 1980, University of Pennsylvania. Montgomery, M. (2008) An introduction to language and society. 3rd edn. Oxford, Routledge. Tannen, D. (2006). You’re wearing that? Understanding mothers and daughters in conversation. New York, Random House. White, F. and Peacock, J. (2008) Ancestral language acquisition among Native Americans: A study of a Haida language class. Lewiston, Edwin Mellen Press. Wolfram, W., Adger, C. T., and Christian, D. (1999) Dialects in schools and communities. Mahwah, Lawrence Erlbaum. Young, R. and He, A. (1998) Talking and testing: Discourse approaches to the assessment of oral proficiency. Amsterdam, John Benjamins.

Eisenstein, M. (1983) Native reactions to non-native speech: A review of empirical research. Studies in Second Language Acquisition, 5(2), pp. 160–176. Eisenstein, M. and Ebsworth, T. (1993) Sensitizing learners to sociocultural aspects of L2: A critical incident activity. Idiom, 23(2), 1–6.


Teaching Pronunciation Using Phonology in the ESL and Foreign Language Classroom

Joanna Labov
This Chapter Explores:
Phonemes Minimal Pairs Phonological Rules Speech Organs Teaching Consonants and Vowels Miscommunication and Pronunciation Errors

“I put my head on my head.”
Non-native speaker of English

As is true of many readers of this textbook, I have always loved learning languages and the beautiful sounds of languages. As a child I studied French and Chinese before moving on to study Spanish, Italian, and German as I became older. My exciting career preparing teachers to teach English as a Second Language (ESL) flowed naturally from my years teaching ESL and experiences teaching Spanish as a foreign language. I began teaching ESL in 1983 and right away started specializing in the teaching of pronunciation. I remember one experience I had teaching English to my Mexican neighbor in my home when I lived in California. I was helping this woman to improve her pronunciation of the English consonants /v/ and /b/ by showing her how they differ physically. When she said the words very and berry, the words sounded the same and she did not know how to make them sound different. I showed her how to use her upper teeth and her lower lip in order to produce /v/ and two lips when pronouncing /b/. This was a good start but I realized that in order to improve my student’s pronunciation of other English sounds such as /k/ and /g/ in which the lips are not used, I would need to know more about phonetics (the study of speech sounds) and phonology (the relationship between speech sounds).

A knowledge of phonology and phonetics is useful for people who want to understand the challenges that second language learners face in attempting to communicate in



their second language. It is essential for teachers who want to improve their students’ pronunciation and ability to communicate in their second language. Specialists in pronunciation training who have a background in phonology and phonetics have contributed to what many language teachers know about how to improve their students’ pronunciation (Derwing and Munro, 1997; Field, 2005). However, there are many pronunciation teachers currently teaching in the world who have little or no training in phonology. Pronunciation teachers undergo many types of preparation before teaching second language learners; not all types include training in general linguistics, phonology, or phonetics. Some teachers may have earned a master’s degree to teach their subject language to adult learners in the US and overseas; others, a certification to teach in grades K-12 in public schools in the United States; and there are those who receive only a brief training session in a private institute or learn on the job with no training. The goal of this chapter is to describe how phonology and phonetics can be used to analyze the pronunciation of second language learners. It will introduce the knowledge, skills, and tools that pronunciation teachers use to examine second language learners’ pronunciation analytically. As you read this chapter, keep in mind the people you know who speak a second language with a strong accent. Ask yourself how a pronunciation teacher would use the skills you will learn here to improve the learner’s pronunciation and what pronunciation problems he or she would focus on. Another benefit of learning more about second language learners’ accents is to understand in what ways they differ from the ways native speakers speak. Although native speakers may not realize it, they also have an accent when they speak their first language. Second language learners enroll in pronunciation classes because they want to improve their ability to be understood when they speak. They may find that they are able to read and write at an advanced level but are sometimes unable to communicate via speech. The following example illustrates this case. A native French speaker used English to order a sandwich in a restaurant and was not understood by the person behind the counter. She became frustrated because the sandwich she received was not the one that she had ordered. The woman ended up paying for the sandwich and throwing it in the trash. She left the restaurant as an unsatisfied customer and wasted her money. A pronunciation teacher informed with a knowledge of phonology and phonetics could analyze the speaker’s pronunciation of English to determine which features prevented her ability to communicate at the lunch place. In this chapter, exercises will be provided that are designed for you to further your understanding of the phonetics and phonology concepts covered. Sections marked APPLICATION are designed to show you how a pronunciation teacher might use the concepts covered to teach pronunciation. There will be four real life examples of miscommunication between native and non-native speakers that occurred as a result of inaccuracies in the non-native speakers’ pronunciation. After reading each miscommunication, you will be given several questions to consider and discuss regarding the possible reasons for the miscommunications. Comments will be provided on the book’s companion website which will explain the reasons for each miscommunication.

Language learners need to learn that the way words are spelled may not always be a good representation of the way they are pronounced. This means that they need to learn phonetic symbols, which are symbols that represent each sound, or phoneme, of a language. Phonetics, the study of the physical sounds of human speech, describes the articulation of speech sounds and where they are made physically by a speaker, e.g. the

Teaching Pronunciation


lips, tongue, open or closed jaw, mouth, and/or nose. Phonetics is useful for pronunciation teachers because it provides a description of speech sounds and a focus on speech. We are influenced by the written word as a result of living in a literate society. When people think of words, they often think of the letters that form the words but not the sounds that make up the words. Pronunciation teachers generally focus on teaching the sounds of the language and not how the words are spelled. I use the following exercise in Figure 3.1 to demonstrate to my ESL students that a focus on spelling will not be useful in improving their English pronunciation. Early in the semester I write eight words on the blackboard and review them with the class. As a group, the students read each word aloud, one word at a time. After the students read each word they identify and pronounce the vowel in the word. As the students read each word, I write the symbol /u/ under it so that they can see that all the vowel sounds are the same.

Different Letters Represent the Same Sound to FIGURE 3.1 two too through threw clue shoe Sioux

Sound and Letter Relationships

Source: Avery and Ehrlich (1992: 3).

Next, I erase the /u/ words and write the five words in Figure 3.2 on the blackboard. The students read each word aloud to discover that the letter “a” is pronounced differently in each word.

The Same Letter Represents Different Sounds cake FIGURE 3.2 mat call any sofa

Letter and Sound Relationships

Source: Avery and Ehrlich (1992: 3).

After the students read the word “cake” I write the symbols /kek/ on the blackboard next to the word “cake” to illustrate that its pronunciation is very different from its spelling. (Symbols in slash marks represent a word’s phonemes.) I ask Spanish speakers in the class to read the first word in the second line as if they were reading a Spanish word. When they read the word “cake” it sounds like “ka-ke” because the letters “a” and “e” are pronounced with the Spanish vowels /a/ and /e/. Unlike English, Spanish has a one-to-one relationship between its letters and sounds, so every letter is pronounced. These two activities deepen the students’ understanding of the difference between sounds and spelling as well as creating an appreciation for this new way of looking at English.



Exercise 1

Make a list of words that shows the different ways that the sound /i/ (as in “ski”) is spelled in English.

Exercises 1 and 2 will further your understanding of the differences between the ways the letter “i” is pronounced and how variably the vowel phoneme /i/ is spelled in English.

Exercise 2

Make a list of words that shows the ways the letter “i” is used to represent different sounds in English.

Exercise 3 will help you distinguish between the spelling of English words and the ways that the words are pronounced.

Exercise 3

Count the number of letters and sounds in the words below: Five, sofa, this, cream, movies, plants, Spanish, computers, television

My students come to understand that we can’t rely on the English spelling system as a tool to improve their pronunciation and why we need to use the International Phonetic Alphabet (IPA) in class. The IPA is a standardized notation system used to transcribe sounds spoken in languages around the world. (See the Handbook of the International Phonetic Association (IPA, 1999), and Pullum and Ladusaw (1996), for a guide to the IPA symbols.) Some IPA symbols are the same as letters, for example /p/, /b/, /t/, /d/, /s/, and /z/. Other symbols are different from the letters, such as the symbols /θ/ and /ð/, which represent the “th” sound in the words “thin” and “the” respectively. The symbol /θ/ also represents Castilian Spaniards’ pronunciation of the Spanish letter “c” as the “th” sound in the word “Barcelona”. Figure 3.3 shows the International Phonetic Alphabet, which consists of symbols used to represent the pronunciation of consonants and vowels in spoken languages. You can view the IPA online by checking out the International Phonetic Association’s website: (International Phonetic Association, 2009). This useful website will enable you to type symbols that are not found on your computer keyboard: (Wikipedia, 2009). (Note that later figures—Fig. 3.12 and Fig. 3.15—differ somewhat in IPA symbols. Even with an “international” system there exists variation.)

Teaching Pronunciation



The International Phonetic Alphabet/general American English

The following application describes how I use IPA symbols in my classroom.

APPLICATION: Making the IPA Useful for Students
How do I use the IPA symbols in the classroom? I write all my notes on the blackboard for the students using IPA. After a while, they become very used to the symbols and they are able to understand how the symbols show different sounds, even when they don’t have those sounds in their own languages. I also use the symbols to show how


Labov English spelling is not very directly related to pronunciation and to encourage them to pay attention to sound rather than the way that words are spelled. My students learn IPA quickly because we use it every day in the class. I write welcome messages for them on the blackboard such as “Welcome” /wεlkəm/ and “Hello! My name is Joanna Labov /hεlo! mai nem z d o nə ləbov/. The students learn to read and write questions using IPA notation such as “What is your name?” “How are you?” “How are you feeling today?” I ask them to read sentences written in IPA notation on the blackboard and to transcribe sentences into IPA. The students are pleased to see that they can learn such an interesting and useful transcription system in a short time. Some of my Chinese students already know how to read and write a notation system similar to IPA called K&K because they learned it in China.

The IPA is an indispensable tool for teaching pronunciation and analyzing students’ pronunciation errors with them in class. Some students pronounce words as they are written in their textbooks because they have never heard them pronounced (called “reading mistakes”). For example, they may pronounce the word “syllable” as /s l′eble/ instead of /′s ləbəl/, which is the way it is usually pronounced. Teachers can write the IPA symbols for both pronunciations on the blackboard in order to show their students how their pronunciation of the word “syllable” differs from the usual pronunciation. IPA is also useful for students who cannot perceive a difference between two similar words. One of my students thought the vowels in the first syllable of the words “woman” and “women” were the same, i.e. /w mən/. Therefore, I wrote the IPA symbols for “woman” /wυmən/ and “women” /w mən/ on the blackboard so that the student could see that the words differ in terms of their pronunciation. Exercise 4 will provide you with practice transcribing words using the International Phonetic Alphabet.

Exercise 4

Record one of your classmates saying the following words and transcribe them using IPA. After you finish your transcription, ask your classmate to review its accuracy. Use the symbols in the IPA chart for your transcription of these words: 1 2 3 4 5 Hello, my name is How are you? Hi. I am fine. Good. I am glad. Catch you later.

Speech Organs
People may not be aware of which parts of their bodies they use to produce consonants and vowels due to the automaticity of the speaking process. The organs that we use to produce speech are referred to by the term speech organs. It is important for pronunciation teachers to know which organs are used to produce speech to improve their students’ pronunciation. They can also directly teach their students about their speech organs and how they are used to produce speech. By doing so, teacher and students will have a shared knowledge base and vocabulary about speech production to work from. The speech organs consist of the speaker’s vocal cavity and nasal tract. The vocal tract

Teaching Pronunciation



Sagittal View of the Vocal and Nasal Tracts

Source: Adapted from Tserdanetis and Wong, 2004: 51.

includes the lips, tongue, alveolar ridge, the velum, the uvula, and the larynx. (The parts of the vocal tract that are places of movement or contact, such as the tongue and alveolar ridge, are called articulators.) The nasal tract includes the nostrils and the space within the nasal cavities. See Figure 3.4 for a view of the vocal and nasal tracts. Consonants are produced with relatively more constriction in the vocal tract compared to vowels; the vocal tract is relatively more open during vowel production, allowing a freer flow of air. The following exercise will allow you to feel which speech organs you use when you pronounce a variety of consonants. It can also be used to familiarize ESL students and foreign language students with their speech organs.
Hold a mirror to your face and produce the consonants /b/, /p/, and /m/. Look at your face when you say the consonants several times. Which speech organs do you use to produce these consonants? Now say the consonants /f/, /v/, and then /s/, /z/, and /k/, /g/. Copy the diagram in Figure 3.3 and mark where you think you are producing these sounds.

Exercise 5

Phonology is the study of the relationships and patterns of the sounds of language. A knowledge of phonology is useful for pronunciation teachers because it enables them to



understand the sound patterns of the language they teach. In addition, the teachers’ understanding of the sound patterns of their students’ first languages will help them to understand the reasons for their students’ pronunciation errors. Phonemes are language specific since what is a phoneme in one language is not necessarily a phoneme in another. For example, the English phoneme /I/ (the vowel in the word “sit”) is not a phoneme of Italian. An important concept in phonology is the difference between a phoneme and a grapheme. The term phoneme refers to the smallest unit of sound that is distinguished from other sounds in a language (Ladefoged, 2006). The term grapheme refers to the same idea with regard to letters. The symbol “A” and the symbol “a” are the same graphemes. The word “grapheme” has eight graphemes, whereas it has six phonemes because the grapheme ph is used to represent the /f/ phoneme. Phone means sound and graph means writing, as seen in the words telephone and telegraph. The combination of phone + eme refers to a unit of sound since eme means unit. A unit of sound is any sound that is used by speakers to contrast one word with another, e.g. /p/ contrasts the word pat with hat.

You may be surprised to know that English has three types of /p/ sounds even though it has only one letter “p” and one phoneme /p/. If you ask an English speaker how many / p/ sounds there are in English he or she will probably say that there is one /p/ sound. This abstract conception of the sound /p/ is called a phoneme and is perceived as a unit of sound by English speakers. The slash mark notation / / is used to represent phonemes and the bracket notation [ ] is used to represent the actual sounds or phones people produce when they speak. The phoneme /p/ is a stop consonant, a type of sound produced by first closing off airflow, and then allowing the built up pressure to be released in a burst. In order to understand the English phoneme /p/ one needs to know not only that it has at least three types of sounds (variants), but also where they are distributed in words. Each variant is called an allophone and has a predictable distribution in words. The following exercise provides you an opportunity to notice how much air is produced when English speakers produce /p/ in the beginning of words.

Exercise 6

Put a tissue in front of your mouth and say the following words several times: “puff ”, “spin”, and “lip”. Notice how much the tissue moves as you say each word. Which word produces the most movement of the tissue? Is this the same word as was discussed in the chapter? Ask three of your friends or family members to complete this exercise. Which of the three words produces the most movement of the tissue for the people that you asked to do this exercise? Try the words “puff” and “buff” to see differences in the amount of movement in the tissue.

One can predict that this puff of air, called aspiration, is always found in /p/ at the beginning of a word and not in the second or final position. The three [p] allophones are located in the following position of words: (1) the beginning of the word “puff”; (2) the second consonant of the word “spin”; and (3) the final consonant of the word “lip.” The

Teaching Pronunciation



[ph] aspiration of air “puff” [p] “no aspiration of air “spill” [p ]no release of air “lip”


English Phoneme /p/ and Its Three Allophones

Source: Adapted from Stewart and Vaillette, 2001: 72.

/p/ on the left side of Figure 3.5 is a phoneme and the three [p] symbols on the right side of the figure represent the three allophones of /p/. Notice the allophones are distinguished by superscript marks called diacritics. The three allophones of /p/ are not used in all locations in words spoken by native English speakers. Therefore the allophones do not compete with each other for the same position in words nor serve a contrastive function resulting in a difference in meaning between two words. The relationship among the three allophones is one where they complement one another (make complete, whole) and their distribution in words is called complementary distribution. Each of the three allophones of /p/ is needed to make complete the concept of the English phoneme /p/. This is true for the other voiceless stop consonants as well, /t/ and /k/, which are produced without vocal fold vibration. The tissue exercise is helpful for learners to experience the aspiration that occurs with different consonants in differing environments. The following application describes how I use the tissue activity in Exercise 6 in my ESL classroom.

APPLICATION: Using Tissues to Teach Aspiration
I hold a tissue in front of my lips and say the word “puff” so that my students can see the tissue move. The students can see that the tissue moves more when I say the words “puff” than “spin.” I distribute tissues to my students and ask them to put the tissues in front of their mouths as they say the words “puff” and “spin.” This experiential method of teaching pronunciation helps the students to learn firsthand: (1) the differences between aspirated and nonaspirated English consonants and (2) that aspiration occurs in the beginning of English words but not in second and final positions. Now try the same activity with the words “two” and “stew”; and “key” and “ski.” This activity is helpful for French and Italian learners of English, who often do not produce voiceless stops with aspiration in the beginning of a word. French and Italian do not have aspiration in syllable-initial position (Swan and Smith, 2001).

The concept of allophones demonstrates the importance of phonetics for pronunciation teachers and students. Native speakers do not produce allophones incorrectly because when they learned the phonology of their first language they learned which allophones are produced and in which positions of a word they are allowed. This can be seen by native English speakers’ consistent use of aspiration in the initial position of words for /p/ such as the word “pie.” However, non-native English speakers often produce allophones inaccurately, which leads to miscommunications such as when the word “pie” is produced with no aspiration and thus understood as “bye.” Learners need to understand which sounds of the language they are learning contrast with each other. They often study a language for years but do not learn how to



differentiate two phonemes by saying them or perceiving the difference between them. They may not be able to produce two sounds or be able to perceive the difference between them. I found this to be the case when I interviewed 25 German students who had studied English for an average of eight years in Germany but were not able to produce a difference between the vowels / / and /ε/ as in the words “pat” and “pet” (Labov, 2000). The IPA symbols / / and /ε/ are used in this chapter to represent the vowels usually referred to by the nontechnical terms “Short A and short E.” These speakers learned English / / and /ε/ as one phoneme, ε, and did not understand psycholinguistically that they are two English phonemes. Figure 3.6 illustrates English speakers’ knowledge of the phonemes / / and /ε/.


The English Phonemes / / and /ε/

Figure 3.7 illustrates the English /æ/ and /ε/ phonemes of the selected native German speakers studied (Labov, 2000). The native German speakers’ pronunciations of the vowels / / and /ε/ are often heard by native English speakers as one vowel, e.g. “sat” and “set” both as “set” (Labov, 2000).


The English / / and /ε/ Productions of Selected Native German Speakers

The box below provides a miscommunication between a native German speaker and a native English speaker about the events that occurred during a shopping trip.

A Shopping Trip
One day I was talking with my German friend Claudia in English about what she had done during the afternoon. She told me that she had gone shopping and that she had a new acquisition. Then she said to me, “I put my head on my head,” which at first confused me because how can she put her head on her head? What was Claudia trying to tell me? 1 Consult Figures 3.6 and 3.7. Would Claudia’s statement be confusing for people who have a background in phonetics and phonology? Why or why not? 2 What is the most polite response one can give to someone who says to you, “I put my head on my head”?

Pronunciation teachers use minimal pairs as a tool to teach their students how to produce and perceive two phonemes of their second language. A minimal pair is a pair

Teaching Pronunciation


of words that differ by only one sound, which is thus considered a “minimal” difference. The sound that differentiates the two words in terms of meaning is the phoneme. The differentiation of two words by phonemes can be seen by the phonemes /h/ and /s/ that differentiate the word “hat” from “sat.” It can also be seen by the vowels /υ/ and / / that differentiate the word “woman” from “women.” Figure 3.8 provides the spelling and IPA symbols for the words “hat” and “sat” to illustrate two words that are a minimal pair. One can see that the pronunciation of the words “hat” and “sat” differs only in terms of the initial consonant because the vowel / / occurs in both words.

Spelling hat sat FIGURE 3.8 Minimal Pair Words Hat and Sat

IPA /h t/ /s t/

Figure 3.9 provides the spelling and IPA symbols for the words “soap” and “rope” in order for you to consider whether they are a minimal pair. Since the concept of a minimal pair assumes a difference in one sound (not a letter), the two words in Figure 3.9 are a minimal pair. We can see that /s/ contrasts with /r/ in the words soap and rope. We also see that the symbol /o/ is used to represent the phoneme /o/ in the words “soap” and “rope,” which have the same sound even though the spelling is different. Therefore, the only differences in terms of pronunciation between the two words are /s/ and /r/.

Spelling soap rope FIGURE 3.9 The Words Soap and Rope

IPA /sop/ /rop/

The following exercise will help you to distinguish between pairs of words that are minimal pairs and those words that are not.

Exercise 7

Circle the words “yes” or “no” after each pair of words to indicate whether the words constitute a minimal pair. Minimal Pair 1 fair 2 low 3 feet 4 pit bear grow feat grit yes no yes no

yes no yes no yes no

5 moon boon



The following application is a suggested activity for language teachers to improve their students’ ability to perceive and produce minimal pairs.

APPLICATION: Using Minimal Pairs to Teach Sound Contrasts
I read aloud a list of minimal pairs and two repetitions of the same word to my students. After I say a word pair my students say “one word” or “two words” depending upon how many words they hear. They circle the words “same” or “different” on a paper to indicate whether they heard the same word repeated. This exercise can be modified by having students raise one finger if they think they hear the same word repeated twice or two fingers if they hear two different words. I ask selected volunteers to read the list and the rest of the class raise their fingers to indicate if they hear minimal pairs or the same word repeated.

The box below provides an example of a miscommunication between a native Greek speaker and a native English speaker that occurred about the type of restaurant the Greek speaker had gone to for dinner.

Dining in Peace
One day a friend of mine and I were walking along the beach with her Greek boyfriend, named Kallius. I remember Kallius told me that he and my friend ate dinner at a die restaurant the night before and how delicious the food tasted. Why would he tell me that he ate at a die restaurant? Who would want to eat at a die restaurant? 1 Would you be confused if Kallius were to tell you that he ate at a die restaurant? What would you say to Kallius in reply? 2 What role do you think the homophones “die” and “dye” play, if any, in my misunderstanding of Kallius’ statement?

Phonological Rules
Phonological rules determine the specific arrangement of phonemes in a language that makes the language distinctive. The following two phonological rules illustrate different types of relationships between English phonemes: (1) syllable structure constraints, and (2) assimilation between similar sounds.

Syllable Structures
English has a complex syllable structure system, as can be seen in Figure 3.10, which provides selected examples of syllable structures. The figure illustrates the consonant combinations that occur in English in the beginning of syllables. English has strict rules regulating which sounds can occur in the beginning of syllables. For example, monosyllabic words that begin with three consonants can only occur with the combinations /spr/, /str/, and /skr/, e.g. “spring,” “string,” and “scream.” By completing Exercise 8 you will draw upon your knowledge of what consonant combinations are acceptable in English.

Teaching Pronunciation


Spelling spring click go Key: C = consonant V = vowel FIGURE 3.10

IPA /spr ŋ/ /kl k/ /go/


Examples of English Syllable Structures

Exercise 8

Provide as many consonant clusters that occur in the beginning of words as you can. Underline the consonants you are focusing on and write example words. Provide five examples of words that have two consonants initially. Example: /kl/ (clean) Provide six examples of words that have one consonant initially. Example: /g/ ( go)

Learners of English need to understand which sound combinations occur in the initial positions of syllables as well as in the final positions. This is true for learners of all foreign languages, which have their own distinctive combinations of phonemes in initial and final positions of syllables. Japanese learners of English often insert a vowel between two English consonants because Japanese has a simpler syllable structure than English, i.e. CV. Thus, the word “match” is often pronounced as “mach-ee” /matʃi/, which turns a CVC structure into a CVCV structure (Swan and Smith, 2001). The learner’s insertion of the vowel “ee” /i/ is called epenthesis. Epenthesis also refers to the insertion of a sound in the middle of a word, for example adding /p/ between the /m/ and /s/ of the word “hamster” so that it sounds like “hampster.” Korean learners of English also use epenthesis when they produce English syllables that have two consonants in a row, e.g. “desk” is pronounced as desk-ee /dεski/. The branch of phonology which addresses the combinations and arrangements of sounds of a language is called phonotactics. The following application demonstrates how ESL teachers can use their knowledge of the syllable structure constraints of their students’ first language to help them acquire the English consonant cluster /sp/ in the beginning of words and syllables.

APPLICATION: Teaching Native Spanish Speakers to Produce /sp/
When I teach ESL to native Spanish speakers I use dialogues that contain many occurrences of the consonant cluster /sp/. Spanish speakers tend to pronounce the consonant cluster /sp/ as /əsp/ because the consonant cluster /sp/ does not occur


Labov in syllable-initial position in Spanish. Thus, they often pronounce the word “speak” as /əspik/. In contrast, learners of Spanish do not have problems saying “español” because the vowel-consonant combination /əs/ occurs in English, e.g. essential. The Spanish learners circle words in texts that have the consonant cluster “sp” in order to focus their attention on this consonant cluster. They practice saying these words in class for me to listen to and provide feedback.

Language learners need to learn the phonological rule assimilation, which is a rule that changes a phoneme to one that is more similar to a neighboring phoneme. Assimilation occurs because it is easier for speakers to produce two phonemes that are similar to each other than two phonemes that are different. Ask yourself the following question about pluralization in English: Are there differences in the plural markers of the words “cats” and “dogs”? What is the rule that creates the differences in pronunciation between the singular and plural? The answers can be seen in the pluralization of the words “cat” and “dog” in Figure 3.11.

Spelling cats dogs FIGURE 3.11 The Words “Cats” and “Dogs”

Phonemes /k ts/ /dɔ z/

One can see that the plural of the word “cat” is pronounced /k ts/, but the plural of the word “dog” is pronounced /d gz/. The suffix -s, which represents plurals, is sometimes realized as the phoneme /s/ and sometimes as the phoneme /z/. This assimilation rule results in two consonants next to each other with similar voicing features, i.e. / t/ and /s/. The word “cats” is pronounced with the voiceless consonant /s/, which matches the preceding voiceless consonant /t/ in the word “cat.” Voiceless sounds are produced with no vibrations of the vocal folds, the two flaps of membrane and muscle that are in our throats at the level of the Adam’s Apple. However, the plural marking of the word “dogs” is pronounced as /z/, which matches the voicing feature of the preceding consonant /g/. The phonemes /g/ and /z/ are voiced consonants. (Voiced sounds are produced by vibrations of the vocal folds.) ESL students whose first languages do not have word-final voiced consonants, e.g. Chinese and German speakers, often have difficulties producing voiced consonants at the ends of words. Exercise 9 is an analytical exercise designed to help you discover the assimilation rules involved in the past tense of the English words “buzzed” and “wished.”

Exercise 9

What types of assimilation rules account for the past tense of the words “buzzed” and “wished”? Provide three words that sound like “buzzed” and “wished” and come up with the reason why the past tense of these words is pronounced differently.

Teaching Pronunciation


The assimilation of consonants that are similar to each other is a universal phonological tendency. For example, in Spanish the word “conversación” (conversation) is pronounced as “combersación,” where the /n/ in the first syllable changes to /m/, which results in two consonants produced with both lips touching each other, i.e. /m/ and /b/. The type of assimilation rule demonstrated here by Spanish speakers is called Place of Articulation Assimilation Rule because it involves changes in the speech organs where the articulation of sound occurs, i.e. the speaker’s lips. In French, the final consonant of the verb “mettre” (to put) is not pronounced with a voiced /r/. Instead, the /r/ becomes voiceless because it follows /t/, a voiceless consonant. The assimilation rule that is used by the French speakers in this example is called a Voicing Assimilation Rule because it involves the assimilation of the voicing feature of the /r/ to the nonvoicing feature of the preceding /t/. The following activity is suggested for ESL teachers to help their students learn to link consonants between words together in order to speak English informally.

APPLICATION: Teaching Students to Speak English Informally
I teach my students to speak informally by providing them with an opportunity to practice linking similar consonants together. I give them dialogues to practice reading that contain expressions such as “How are you doing?” and “What are you doing?” These expressions when spoken informally sound like “How ya doin’?” / hau jə duən/ and “Wha’ cha doin?” / w tʃə duən/ respectively.

Teaching Consonants
Description of Consonants
Good pronunciation teachers need to understand how different consonant sounds are made and which sounds are likely to be confused by learners. This can only be done if the teachers understand the key features that make consonants different from each other. An important way of describing consonants is to note the places in the vocal and nasal tracts where they are articulated: lips, tongue, teeth, hard palate, velum, nasal cavity, and vocal folds. This type of description, called place of articulation, describes which speech organs are used to produce a speech sound. The Spanish speaker mentioned earlier who pronounced /b/ when saying the word “very” as in “a very” produced a bilabial consonant (two lips) instead of /v/, a labiodental consonant (lower lip and upper tongue). The speaker needs to learn which speech organs should be used in order to produce /v/ accurately. Consonants are also described by how we shape the airstream that is expelled from our mouths when we produce speech. This type of description is called manner of articulation because it describes how the airstream is shaped in order for speech sounds to be articulated. Some Spanish speakers pronounce “very” not with a /b/ but with a soft and scraping sound due to a partial obstruction of air by the lips; this is called a bilabial fricative. However, the English /b/ is a stop consonant, which means that the airstream expelled from the speaker’s mouth during speech should be stopped completely by her lips. The speaker needs to learn how to manipulate the airstream that comes out of her mouth so that her production of /b/ vs. /v/ is comprehensible. Figure 3.12 provides a chart that describes English consonants in terms of their manner of articulation, place of articulation, and voicing feature (voiced or nonvoiced).




Place of Articulation and Manner of Articulation Chart

Source: Adapted from Stewart and Vaillette, 2004: 46.

Teaching the /l/-/r/ Contrast to ESL Students
Many Chinese, Japanese, Korean, and Thai learners of English (Swan and Smith, 2001; Avery and Ehrlich, 1992) have difficulties producing and perceiving a difference between the English consonants /l/ and /r/ because these languages do not have both phonemes /l/ and /r/. The box below provides a misunderstanding that occurred as a result of a nonnative English speaker’s production of a consonant that was perceived as /r/ instead of /l/. Am I wrong?
One day I was teaching pronunciation to my ESL class at a university language institute. One of my students, Qiao, a Chinese speaker, was having problems pronouncing /l/ and /r/. Many of his /l/s sounded like /r/s and his /r/s sounded like /l/s. Qiao was not able to produce a consistent distinction between the two consonants although he tried earnestly to do so. I remember working with Qiao by helping him to read a dialogue that contained many /l/ and /r/ words. After many attempts, Qiao was able to read the dialogue perfectly by pronouncing every /l/ and /r/ word correctly. Imagine my surprise, when at the end of class Qiao got up to leave, put his hand on the classroom door and told me, “So wrong!” What did I do wrong? What happened with Qiao?

Figure 3.13 illustrates the location of the tongue in the production of the consonants /l/ and /r/.

Teaching Pronunciation



Articulation of the Consonants /l/ and /r/

Source: Avery and Ehrlich, 1992: 22.

Technique to Teach /l/ and /r/
The following is a tactile exercise which will enable you to feel the differences in articulation between the consonants /l/ and /r/ in syllable-initial position. It is also an exercise that ESL and foreign language teachers can use to help their students feel which speech organs are used to produce /l/ and /r/.

Exercise 10

Liquid Consonant Exercise Place a toothbrush, tongue depressor, or chopstick in your mouth horizontally while you say the four sentences within this Exercise, which contain /l/ and /r/ words in contrasting position. By doing so, you will feel the contrast between the words that require the articulation of the tip of the tongue for /l/ and the pronunciation of /r/ words, which does not. I introduce this exercise to my ESL classes during the middle of the semester after the students have gotten to know each other and feel at ease with each other. Four Sentences with /l/ and /r/ Alterations in Contrasting Positions 1 2 3 4 I collect long sentences. I correct wrong sentences. I collect wrong sentences. I correct long sentences. (tongue tip, tongue tip) (no tongue tip, no tongue tip) (tongue tip, no tongue tip) (no tongue tip, tongue tip)

Source: Adapted from Suzuki, 2000, with permission.

Which of the four sentences is more effective in making you feel how you produce /l/ and /r/? Do you feel that Sentence 3 is more effective than Sentence 4? Is Sentence 4 more effective than Sentence 3?



The following application demonstrates how I teach my students to complete the exercise above by using toothbrushes.

APPLICATION: Using Toothbrushes to Teach /l/ and /r/
I model for my students how to complete the exercise shown in Exercise 10 and then I perform it with them. By doing so, I remove any uncertainty they may have about completing this tactile exercise. After students pronounce the first two sentences, in which they produce two of the same consonants consecutively, we say as a class the third and fourth sentences, which have words alternating with the consonants /l/ and /r/. The sentences “I collect wrong sentences” and “I correct long sentences” require the students to alternate their tongues back and forth between the tip and back of their tongues. I ask the students which of the two sentences is more effective in showing them how to produce /l/ and /r/. They always have strong opinions, which they voice vigorously, about which sentence is more effective for them. This experiential activity produces the “Aha!” moment, whereas showing the students Figure 3.13 produces an intellectual understanding of the differences between /l/ and /r/.

Vowels are characterized according to two major dimensions: (1) the height of the jaw and tongue, and (2) how far front and back they are produced in the mouth. Figure 3.14 shows 12 vowels and three diphthongs in general American English; diphthongs are characterized as two vowels produced together, i.e. /a/ as in the word “I”, /aυ/ as in the word “house,” and /ɔ/ as in the word “toy”. The English vowel system of 15 vowels is more complex than in many other languages, such as those languages which have a five vowel system consisting of the vowels /a/, /e/, /i/, /o/, and /u/. Languages which have a five vowel system include Greek, Japanese, Spanish, Swahili, and Zulu. The five vowel system is the most common vowel system of the languages of the world (Ashby and Maidment, 2005). One vowel pair that is challenging for ESL learners from many language backgrounds to produce is the contrasting /i/ and //, especially for speakers of Arabic, Catalan, Chinese, Dutch, Farsi, French, Greek, Italian, Polish, Portuguese, Spanish, Swahili, and Vietnamese (Swan and Smith, 2001; Avery and Ehrlich, 1992). These languages have the vowel /i/ but not the vowel //, which means that the English learners have to learn a new phoneme to contrast with /i/. The vowels /i/ and //, which are high vowels, differ in terms of the tenseness of the mouth muscles and the degree of opening of the mouth. The vowel /i/, a tense vowel, is produced with a tightening of the lips, the muscles surrounding the mouth, the tongue and a slight opening of the mouth. The vowel //, a lax vowel, is produced with a relaxing of the lips, the muscles surrounding the mouth, the mouth and a somewhat larger opening of the mouth. Figure 3.14 provides a vowel chart of the English vowel system displayed inside the mouth. It shows the location of the vowels as they are produced in the mouth and the IPA symbols for each vowel. You can see that there are five vowels produced in the front of the mouth, two vowels in the middle of the mouth, and five vowels in the back of the mouth. (Note that this Figure doesn’t show the three diphthongs or the r-colored vowels / / and / /.)

Teaching Pronunciation



The Vowel Chart in Relation to the Inside of the Mouth

Source: Avery and Ehrlich, 1992: 33.

APPLICATION: Teaching ESL Speakers to Produce /I/ and /i/.
I give a mirror to each student so that he or she can practice producing the minimal pairs below while looking in the mirror. The students practice relaxing their mouths when producing the words that have the vowel /I/. Minimal pairs: e.g. /sIt/ and /sit/, /f It/ and /fit/, /pIt/ and /pit/. 1 2 3 4 I want to sit with you. I want a seat with you. Will these fit your feet? Give the pit to Pete.

The students also listen to an English speaker produce the minimal pairs and sentences so that they can learn to hear the difference between /I/ and /i/.

A knowledge of phonetics and phonology provides pronunciation teachers with an understanding of how speech sounds are produced and the patterns that exist within languages. It is also important for language teachers to know how vowels and consonants are produced during speech. These concepts are essential for language teachers to know in order to improve learners’ pronunciation of their second language.

Ashby, M. and Maidment, J. (2005) Introducing phonetic science. New York, Cambridge University Press. Avery, P. and Ehrlich, S. (1992) Teaching American English pronunciation. 2nd edn. New York, Oxford University Press. Derwing, T. and Munro, M. (1997) Accent, intelligibility, and comprehensibility. Studies in Second Language Acquisition, 19(1), pp. 1–16.


Labov Field, J. (2005). Intelligibility and the listener: The role of lexical stress. TESOL Quarterly, 39(3), pp. 399–423. International Phonetic Association. (1999) Handbook of the International Phonetic Association: A guide to the use of the International Phonetic Alphabet. New York, Cambridge University Press. International Phonetic Association. (2009) International Phonetic Alphabet [Internet]. Available from: [Accessed June 11, 2009]. Labov, J. L. (2000) The roles of distinctive and redundant features in the production of the English / / and /ε/ contrast by L1 German speakers of English. Ph.D. thesis [unpublished], University of Pennsylvania. Ladefoged, P. (2006) A course in phonetics. 5th edn. Boston, Thomson. Pullum, G. K. and Ladusaw, W. A. (1996) Phonetic symbol guide. Chicago, University of Chicago Press. Stewart, T. W. and Vaillette, N. (2001) Language files: Materials for an introduction to language and linguistics. 8th edn. Ohio, Ohio State University Press. Suzuki, M. (2000) Metacognitive strategies differentiating the phonological features of /l/ and /r/. Proceedings of the PENN-TESOL East Conference, October 20–23, 2000, University of Pennsylvania. Swan, M. and Smith, B. (2001) Learner English: A teacher’s guide to interference and other problems. New York, Cambridge University Press [sound recording: CD]. Tserdanetis, G. and Wong, W. Y. P. (2004). Language files: Materials for an introduction to language and linguistics. 9th edn. Ohio, Ohio State University Press. Wikipedia. (2009) Pronunciation respelling for English [Internet]. Available from: http:// [Accessed June 11, 2009].

Other Resources
Annual Review of Applied Linguistics Applied Linguistics ELT Journal Journal of Speech, Language, and Hearing Journal of Verbal Language and Verbal Behavior Language and Communication Language and Speech Language Learning Language Teaching Modern Language Journal Second Language Research Studies in Second Language Acquisition TESOL Quarterly

Selected Pronunciation Related Books and CD-Roms
Celce-Murcia, M., Brinton, D. M., and Goodwin, J. M. (1996) Teaching pronunciation: A reference for teachers of English to speakers of other languages. New York, Cambridge University Press. Clarity Language Consultants. (2009) Connected Speech. Clarity [sound recording: CD]. An interactive multimedia computer program for teaching pronunciation and effective communication skills.

Teaching Pronunciation


Hewings, M. (2004) Pronunciation practice activities: A resource book for teaching English pronunciation. New York, Cambridge University Press. Pennington, M. C. (1996) Phonology in English language teaching: An international approach. New York, Longman.

Selected Pronunciation Textbooks
Dale, P. and Poms, L. (1994) English pronunciation for international students. Englewood Cliffs, Prentice Hall. Gilbert, J. B. (1993) Clear speech: Pronunciation and listening comprehension in North American English. New York, Cambridge University Press. Grant, L. (2001) Well said: Pronunciation for clear communication. 2nd edn. Boston, Heinle and Heinle. Hagen, S. A. and Grogan, P. E. (1992) Sound advantage: A pronunciation book. Upper Saddle River, Prentice Hall. Hahn, L. D. and Dickerson, W. B. (1999) Speechcraft: Discourse pronunciation for advanced learners. Ann Arbor, University of Michigan Press. Kozyrev, J. (2004) Sound bites: Pronunciation activities. Florence, Cengage Learning. Lane, L. (1997) Basics in pronunciation: Intermediate practice for clear communication. New York, Longman. Orion, G. F. (1997) Pronouncing American English: Sounds, stress and intonation. Boston, Heinle and Heinle. Reed, M. and Michaud, C. (2005) Sound concepts: An integrated pronunciation course. New York, McGraw Hill.

Crystal, D. (2003) A dictionary of linguistics and phonetics. 5th edn. Cambridge, WileyBlackwell. Trask, R. L. (1996) A dictionary of phonetics and phonology. New York, Routledge.

Relevant Organizations
Teachers of English to Speakers of Other Languages American Association for Applied Linguistics

Online Databases
The Education Resources Information Center (ERIC). (2009) Education Resources Information Center (ERIC) [Internet]. Available from: [Accessed June 11, 2009]. This database contains abstracts of articles in education found in more than 600 journals dating from 1966 to the present. Linguistics and Language Behavior Abstracts (LLBA). You can access this database by searching online for LLBA. This extensive database allows you to search for relevant articles about linguistics and language behavior.

Department of Linguistics, University of California. (2009) A slow-motion animation of the vocal folds vibrating during speech [Internet]. Available from: http://www.linguistics.


Labov [Accessed June 11, 2009]. UCLA phonetics department’s weblink for a real time view of vocal folds in action. Ladefoged, P. (2006) Vowels and consonants [Internet]. Available from: http:// [Accessed June 11, 2009]. Peter Ladefoged’s IPA website. Meier, P. (1997) International Dialects of English Archive (IDEA) [Internet]. Available from: [Accessed June 11, 2009]. The International Dialects of English Archive (IDEA) website. Szynalsk, T. P. (2009) Type IPA phonetic symbols online [Internet]. Available from: http:// [Accessed June 11, 2009]. This website will enable you to type symbols that are not found on your computer keyboard for IPA symbols as well as orthography in the following languages: Czech, French, German, Hungarian, Italian, Polish, Portuguese, Romanian, Spanish, Swedish, and Turkish.

Literature, Translation, and Computers
A fundamental source of linguistic knowledge we have been using our entire student lives is the dictionary. Chapter 4 is written by a linguistics professor who loves dictionaries. Paul D. Fallon brings to us a fresh look at an old friend. Chapters 5 and 6 delve into translation. Translators Mary Boldt and Esperanza Roncero show us in Chapter 5 why translating from one language to another is not simple, and they make us think about how our worlds widen when we can see connections among languages. Nan Decker, in Chapter 6, offers us a view into the world of machine translation. When a software program does the work of a translator, what are the results? We finish this part with a chapter by Carmen Gillespie, a Toni Morrison scholar and professor of creative writing and literature (Chapter 7). She explores the use of what she calls transgressive language in popular culture and in contemporary literature, specifically concentrating on Morrison’s novels Beloved and Jazz.

Pronunciation Meaning Usage

Lexicography What Dictionaries Reveal about Language and Dictionary Makers

Paul D. Fallon
This Chapter Explores:
Prescriptivism and Descriptivism Language Variation and Change Bias in Definitions

Dictionaries are like watches: the worst is better than none, and the best cannot be expected to go quite true. Samuel Johnson, 21 August 1784, letter to Francesco Sastres . . . for a desert island, one would choose a good dictionary rather than the greatest literary masterpiece imaginable, for, in relation to its readers, a dictionary is absolutely passive and may legitimately be read in an infinite number of ways. W. H. Auden, 1948, “Prologue: Reading,” in The Dyer’s Hand and Other Essays

I have loved dictionaries since I was a child. I fondly remember one from school with the aardvark illustrated on the first page, and one at home that had a table of mysterious alphabets on the inside back endpaper that I consulted to create codes for the neighborhood club of secret agents. Throughout school, I used dictionaries as you probably do—to consult for spelling, meaning, or pronunciation—though I also loved to browse through them for their maps, illustrations, usage notes, and etymologies. Once I received my training in linguistics, I came to appreciate them even more as a description of language. Now, I suffer from the expensive habit of collecting dictionaries, revelling in comparing their infinite nuanced differences, constantly changing over time.



“Look it up!” How many times have we been told to find a spelling or meaning by looking up a word in a dictionary? Yet do we really know what a dictionary is and what it contains? It is in many ways a user’s manual for much of what you know about language. A dictionary is a model of our language. It is a key document that contains highly detailed information about our mental lexicon (vocabulary) and our grammar—what we know when we know language. In addition, it encodes a substantial portion of our grammar, including how sounds combine to form words and how words form sentences. It is a distillation of the properties of our language, and as such is an incredible intellectual achievement. In this chapter, we will explore how dictionaries encode our mental grammar and provide a wealth of information about language in many different ways. In examining dictionaries, we will also look at the role they play in society and some of the biases their makers can display. As for that word in this chapter’s title, lexicography, you can guess that it means something about writing and words. If you look the word up in the dictionary, you’ll read that it is the study of dictionaries and their contents. But dictionaries are much more than mere lists of words. And those who compile and study dictionaries, lexicographers, are more than mere compilers of words.

Organization of Dictionaries
Although most dictionaries are organized alphabetically, many are not. Sometimes, in what is called a reverse dictionary, you can look up meanings by general category, and then you can choose the exact word you were looking for. For example, let’s say you forget the name of the dessert you ate in a restaurant. By looking up “dessert” in the reverse dictionary, you come across a table of various desserts. If you recognize the term, you’ve found it; otherwise, you can read through the definitions to find that your “dish of cream or custard with a caramelized sugar top” was a “crème brûlée.” If you know there’s a word for a hatred of people, by looking up “hatred,” you can find the term “misanthropy” in a reverse dictionary. A visual dictionary is another way of finding a word you want. In a visual dictionary, there are many illustrations with the various parts or categories labeled. For example, say you want to know the name of the little groove above the lips and below the nose. By looking up the entry in “face,” you’ll find that the word you want is “philtrum.” If you’re writing the lyrics to a song or a poem, you might want to find words organized by sound, so you’ll need a rhyming dictionary. For example, if you want to rhyme the word “popularity,” you can find all the words that have the same pattern by looking under three-syllable rhymes; then, following the pronunciation key, you can find lots of rhyming words such as “vulgarity,” “familiarity,” “disparity,” “regularity,” and so on. There are other dictionaries for special needs. There are several different dictionaries for pronunciation, which contain only the pronunciation but no definitions. These can help you settle a dispute about pronunciation, but most don’t reflect the full range of pronunciations, nor do they contain much information about which pronunciations are truly stigmatized, such as “ask” being pronounced as “axe,” or whether it is standard English to make the words “special” and “spatial” sound alike. Other dictionaries are in reverse alphabetical order. This is especially handy if you are a linguist wanting to find all the words that end in -ism, for example. Finally, I have a dictionary that lists words by the fifty-year span in which they entered the language (e.g. 1650–1700), and then groups words by subcategories such as food, entertainment,

What Dictionaries Reveal


sports, etc. This is useful if you want to trace the growth of vocabulary, or if you are writing a creative work of historical fiction and wish to avoid anachronisms (words from another time period). The usual method of organizing a dictionary, however, is by alphabetical order. The advantage of listing words by this method is that the alphabet is fixed and the order is well known (though dictionaries from the eighteenth century did not yet distinguish the letters i from j, nor u from v). Users can thus usually find words they know how to spell fairly easily, but how do you look up a word you’ve heard but are not sure how to spell? A good dictionary will usually have a table of sound-spelling correspondences, often listed in order of frequency. Let’s say you wanted to check the spelling of “scourge.” You could look up the table for the vowel sound pronounced like the vowel in “dirt” (symbolized in dictionaries as (ûr). There you’d find the spelling <ur>, but you wouldn’t find the desired word spelled incorrectly as *“scurge” (where the asterisk means the word or phrase is ungrammatical in some way)—but compare “urge.” You would then find variants <ear>, <er>, <err>, <eur>, <ir>, <or> before coming upon the correct <our> for “scourge.” This is what makes this word so tough to spell—it’s irregular. Initial silent letters certainly won’t help you find your word. For example, if you hear someone say a word that means “pertaining to mythological deities or spirits under the earth” and sounds like “thonic,” you’d probably never think to look it up starting with two silent letters: “chthonic” (unless you could relate it to the word “autochthonous,” meaning “indigenous”, in which the ch of Greek origin is pronounced like a k). Fortunately, this is a word seen in print far more commonly than it is heard in speech. Finally, I should mention that alphabetical order does not make sense for all languages. First, some languages don’t use an alphabet, but use a script based on words. Chinese characters, for example, represent a word or word element, but have no systematic correlation to pronunciation. Therefore, Chinese dictionaries are organized by the number of strokes in each character. Chinese school children need to spend a considerable amount of time learning the order of strokes and what to count as a stroke to both write the characters and to look them up in the dictionary. Even languages with a sound-based script such as Arabic have reason not to always follow an “alphabetical” order. Arabic and Hebrew do not typically write the language with the vowels; they use a consonant-only alphabet. (For precision in holy scriptures, and for foreign learners, diacritics—small, modifying symbols—are sometimes included for the vowels.) This is because these languages have a root and pattern system of language. Most Arabic verb roots, for example, consist of three consonant (triliteral) roots: k-t-b “write,” q-t-l “kill,” d-r-s “study.” These roots then follow different patterns for various forms of the verb, along with a different pattern for the vowels. For example, katab- is the uninflected past form of “wrote,” the causative form is kattab- “cause to write,” the reciprocal is kaatab- “correspond; write to each other.” The passive perfective root is kutib- “was written.” Because many related words such as “office” maktabu, “book” kitaabun, and “Koran school” kuttaabun are derived from the root, the dictionary lists words by root. One advantage to this system is that all related words are grouped together, and a range of meanings for each root can be grasped at once. In addition, several related words can be learned by a student at once. Some even claim such an arrangement allows readers to look up words faster.
If you could start from scratch, how would you organize a dictionary? Alphabetically? By sound? By thematic meanings? How might computers let us approach words in a variety of ways?

Exercise 1



Although most dictionaries are organized alphabetically, we can use this type of dictionary to deduce sound patterns in our language. Speech sounds, phonemes, pattern in a language in a specific way. These patterns are a language’s phonology. For example, if we find a few words beginning with the letter g or k before n, as in gnat or gnostic or knee, we’ll see that the initial consonant is never pronounced. It is a violation of the possible sound sequences of English to pronounce a k or g sound before a nasal sound like n. The words gnat and knee reflect the fact that a consonant that was once pronounced (and thus included in the spelling) now is no longer pronounced. Not all silent letters are useless relics, however. We can learn something about the mental representation of sound by examining related forms. Take the word “agnostic.” When we put the negative prefix a- in front of the word, the g closes the first syllable and thus is not word-initial before the nasal. It is therefore pronounceable in English in agnostic. A similar phenomenon happens when we compare the pronunciation of (silent) n in the word “hymn”; it is sounded when vowel-initial suffixes are added, as in “hymnal.” Compare a similar phenomenon in word pairs like “autumn/autumnal” and “damn/damnation.” These sounds, which vary according to their position in the word, or the type of root or suffix they combine with, are known as alternations and are a major topic in phonology.
Say the following pairs of words out loud. What do you notice about them? sign design malign benign paradigm phlegm signal designation malignant benignant paradigmatic phlegmatic

Exercise 2

Another aspect to pronunciation can be found by examining the pronunciation key to any dictionary. Because English is written using the Roman alphabet, our language has had to adapt a limited number of letters to fit all the sounds of the language. For example, because Latin did not have a th-sound like in the word “think,” we need to use two letters to symbolize that one sound. In addition, English developed a distinction between two types of th-sounds. Compare the initial consonants of the words “thigh” and “they,” the medial consonants of “ether” and “either” and the final sounds in “breath” and “breathe.” You’ll notice that in the second of each pair (but not the first), there is some vibration at your throat and at the placement of the tongue. Dictionary pronunciation keys need to reflect these two sounds since they are used to create differences in meaning: such sounds, phonemes, are the contrastive sound categories of language. Many dictionary keys will try to use a system that is similar to English spelling conventions, so they will use similar patterns, sometimes modified typographically. The New Oxford American Dictionary (2005), for example, uses the symbols /th/ and /tm/.
Can you think of other sounds that need two letters in spelling?

Exercise 3

In English we are really hindered by the Latin alphabet in representing our vowel sounds. Latin had five different vowel qualities, a, e, i, o, u, pronounced as in French or

What Dictionaries Reveal


Spanish, or roughly as in the English words “spa,” “day,” “me,” “go,” “too.” All dialects of English, however, have many more vowels than Latin had—at least 14. To be able to represent these different vowel sounds, pronunciation keys need to add diacritic marks or invent new symbols. The letter a can represent a variety of sounds in English. Compare how a is pronounced in “hat,” “hate,” “hare,” “father,” “any,” “wander,” “tall,” and “alone.” Dictionary keys will represent these vowel sounds as roughly (a, a, â, ä, e, o, ¯ ô, ə). The exact details of each dictionary key differ from publisher to publisher, and so the basic key is sometimes reprinted on each two-page spread, and a fuller key is usually given inside the front cover or endpapers. In the second edition of the Oxford English Dictionary (1989), the pronunciation key uses the symbols of the International Phonetic Alphabet (IPA). This is an international standard, something like the metric system, first proposed in 1886 by what is now known as the International Phonetic Association (also abbreviated IPA). This set of symbols can be used to describe any speech sound from any language in the world. In fact, as new sounds are discovered and documented, new symbols are admitted to the official list. This happened most recently in 2005, with the acceptance of a new symbol for the labiodental flap, / /, a sound found in over 20 languages, mostly in central Africa. The IPA has strict criteria for the use of its symbols, including that they be similar in value to similar letters, but foremost among them is that each distinctive sound should have its own separate symbol. Therefore the IPA does not use two symbols to represent one sound, unlike most dictionary keys. Table 4.1 shows several English words where a combination of letters is used to represent a single phoneme. Note also that there are two phonemes that correspond to the spelling “th”: the initial sound in “thin,” which does not involve vocal fold vibration, and the “th” of “then,” that is produced with the vocal folds vibrating (see IPA chart on p. 47 and at the end of the book).
TABLE 4.1 Word chart budge thin then shin treasure bang Two Letters ch dg th th sh su ng One Phoneme
tʃ d

θ ð ʃ ŋ

In fact, the IPA is not just a set of letter-like symbols. Rather, the IPA is something like chemical abbreviations {H (hydrogen), or O (oxygen)}, which stand for information about atomic weight and composition; phonetic symbols represent a wealth of knowledge about the production of sound, codified in an organized manner just like the periodic table in chemistry. Thus the phonetic symbol [p], which represents a p-sound, is also shorthand for a “voiceless bilabial stop,” the phonetic description of a sound made without vibration of the vocal folds while airflow is halted by the complete closure of both lips. Because of its precision and because it is a universal standard, the IPA is used by many learner’s dictionaries to represent the pronunciation of English words. (A phonetic symbol in square brackets indicates the physical sound, while the symbol in forward-slashes indicates its status as a contrastive phonological sound. Dictionaries often use back-slashes for their pronunciations.)



Exercise 4

Find two college or unabridged dictionaries made by different companies and compare their pronunciation keys. By reading the front matter, you may be able to relate the dictionary symbols to the international standard. When you are finished, you should have a complete inventory of the sounds of English. Compare your answers with someone else in the class. How consistent are the dictionaries?

Recent technological innovations may threaten the pronunciation key in coming years. The inclusion of CD-ROMs with many versions of dictionaries allows the user to consult a dictionary on a computer, or, more recently, on a cell phone or PDA. Most CDs, as well as a few Internet-based dictionaries, include one pronunciation so that the user can hear the word, and not have to decipher the phonetic symbols (although some also include the audio pronunciation using written key symbols, making it easier to learn the symbols since you hear the sound as well). Non-native speakers can thus have access to a native speaker model, and native speakers are unencumbered by stress marks and diacritics. There are, however, some disadvantages. For one, equally acceptable variant pronunciations for some words are almost never given, e.g. as in economics. Further, the words with pronunciations will virtually never include the majority of words in a college or unabridged dictionary. Finally, it’s usually faster to read the pronunciation than wait for the sound file to load.

Exercise 5

Look up in the dictionary the words apricot, pecan, crayon, and caramel. Does the dictionary reflect your own pronunciation of all of these words? What about catch? Does it rhyme with latch or fetch? Some variants are listed for words such as aunt, pronounced both to rhyme with ant or gaunt. Some dictionaries give notes about pronunciations such as ask, when pronounced like axe, as “chiefly substandard.”

A regular pronunciation key aims to capture the important sound contrasts in the language. However, if you speak a dialect that has merged some vowels, such as making “cot” and “caught” sound alike (homophones), then the key may make more distinctions than you do, and, following the key words, you may never realize the difference. Webster’s Third New International Dictionary (1961) contains a highly detailed pronunciation key that takes into account a variety of different dialects. For example, the dictionary pronunciation is listed as \aa\ and refers to “bad, bag, fan as often pronounced in an area having New York City and Washington, D.C., on its perimeter.” These dialects have developed a new vowel that sounds a little like “ee-a”—phonetically [i ]. Studying the key, and the accompanying article in the front matter, can reveal a wealth of information on different accents from around the country.

Exercise 6

Try to think of some variant pronunciations and see if the dictionary records these. For example, do you drop the r in words like park and car? Do you add an r in words like Cuba or wash? Do you pronounce the t in often? What about the l in salmon? Do you feel the dictionary should record all these pronunciations? How would you feel if the way you say a word were not in the dictionary?

What Dictionaries Reveal


Syntax is the aspect of our grammar that describes how words can make up phrases, clauses, and sentences. The dictionary usually contains a brief part of speech label that reflects what linguists call the lexical category—for example, n. for noun, adj. for adjective, and so on. These labels are not always accurate linguistically but represent traditional parts of speech. The word to used as a marker of the infinitive, as in I want to read now is still listed as a preposition in all major dictionaries, as in the sentence I went to the store. Although it had its origin as a preposition in the English spoken hundreds of years ago, it might now be best analyzed as subordinator of verb phrases (Huddleston and Pullum, 2002). With verbs, more sophisticated dictionaries inform the user about some of the collocations (words that go with the entry word). Transitive verbs are those that take an object noun phrase, while intransitive verbs cannot take an object. For example, a verb like “manipulate” is transitive—it must take some kind of object. You can’t say *“They manipulated.” (The asterisk means “not occurring” in the language.) Instead, you need to say something like “They manipulated the stock market.” A verb like “vanish,” on the other hand, is intransitive since it cannot take an object. You can’t say *“Lee vanished the money”; instead, you would say “Lee vanished,” “Lee vanished with the money,” or “The money vanished.” Sometimes, verbs can be both transitive and intransitive. You can “read a book” or just “read.” Often, verbs have slightly different meanings depending on whether they are transitive or intransitive. For example, one meaning of “report” as a transitive verb is “to give an account of, often at regular intervals.” Intransitively, it can mean “to make a report.” Sometimes verbs have very different meanings depending on transitivity. The verb “obtain” in the transitive sense means “to acquire,” while in the intransitive sense, it means “to be customary.”

Exercise 7

Browse through the dictionary at random. Try to find a verb that has both transitive and intransitive meanings. How similar are the meanings? Try to find a word with very different meanings depending on whether the verb is transitive or intransitive. Does slang always reflect the standard grammatical pattern? Look in the dictionary under “represent.” Is it transitive or intransitive? How is it used in a phrase like “The Bronx represents!”? Does it have the standard dictionary meaning here?

Learner’s dictionaries often have much more detailed grammatical information to help learners of the target language. They note not just transitivity, but whether verbs can take two objects (to give somebody something), or whether the verb takes a subordinate clause (to doubt whether they will succeed). Some recent research has suggested that learner’s dictionaries are more helpful to native speaking college students than college dictionaries, showing how words are used in their standard patterns. English language learners often have difficulty with phrasal verbs—verbs that contain another word or a few words to narrow the meaning. Compare “give away” (“bestow”), “give back” (“return”), “give in” (“abandon an argument”), “give off” (“emit”), and “give up” (“surrender”). These words are difficult for dictionary writers and for nonnative speakers, but they are a crucial characteristic of modern English. Think about the difference between “turn up the street” and “turn up the heat.” The former uses “up” as a preposition; the latter uses the phrasal verb “to turn up.” For nouns, learner’s dictionaries also give us detailed grammatical information. For example, if you look up “flour” or “information” in the dictionary, you’ll find that



they are nouns. And you know nouns can be made plural, but there’s nothing in unabridged dictionaries to let us know that certain nouns cannot be made plural. Certain nouns, called mass or noncount nouns, cannot be made plural. We cannot say *“Add three flours”; instead we need to say that you “add three cups of flour.” Likewise, we don’t gather *“informations,” but “pieces of information.” Nouns that can be pluralized, such as “book” and “student,” are count nouns. Learner’s dictionaries encode many of these basic grammatical distinctions, which are often overlooked in the standard dictionaries.
What is the difference in meaning between the singular and plural of words like the following: people fish wine sugar peoples fishes wines sugars

Exercise 8

Does your dictionary distinguish these meanings in the same way that you do?

In the very act of choosing a word to look up, you show your knowledge of the language. For example, if you saw the word “absconding” in the sentence “The witness saw the teller absconding with the cash,” you wouldn’t look it up with the -ing at the end. Your knowledge of English is such that you know to remove the verbal -ing suffix and look up the word under “abscond.” What you would be doing, then, is figuring out a word’s lemma: the base form for a collection of related word forms. The lemma “eat,” to give another example, would thus encompass the word forms “ate,” “eating,” and “eaten.” If you wanted to look up “ate” in the dictionary, you would look for “eat.” And what you would be looking for is the dictionary’s headword: the listing of an entry, usually in boldface. Headwords in the dictionary, then, usually reflect a word group’s lemma. While other, especially irregular, forms of words are listed, it is rare that they would get their own headwords. The dictionary is a great place to learn about morphological information. Morphology is the study of word formation and the creation of words from roots and affixes. All these parts of a word, adding meaning to the full word, are called morphemes, units of meaning. The root is the main part of the word, usually able to stand alone as its own word; the affixes are the units of meaning that attach to the root. In English, we have prefixes, which attach to the front of the root, and suffixes, which attach to the end of the root. For example, what is the plural of “phenomenon”? What is the past participle of “strive”—“strived” or “striven” (or both)? Sometimes the choice of the past participle will vary depending on how the past participle is used. For example, when used as an adjective, the past participle can take one form (like molten lava), while in the perfect verb form, it takes another (The ice has melted ). Sometimes, there is a difference in meaning in a verb like “hang.” Pictures are hung on the wall, while criminals used to be hanged. Good dictionaries will also list major prefixes and suffixes. To save space, forms that predictably change the meaning of a base word are simply listed on the bottom of the page. For example, the prefix mis- means roughly “wrongly, incorrectly, mistakenly” or some negative meaning. So words like “misact” can be listed without a definition (though

What Dictionaries Reveal


“misbehave” has its own entry, perhaps because it’s more common). In a word like “miscarry,” the meaning is less clear from the sum of its parts, since it’s the verb meaning “to suffer a miscarriage,” not to “carry wrongly,” and so it deserves a separate entry. Dictionaries can often help us distinguish prefixes we might have thought were the same thing. Dictionaries will often do this by adding a superscript number to distinguish two homonyms (words that sound the same and are spelled the same but which differ in meaning). For example, did you know that there are two distinct unprefixes? One is a prefix that means “not,” as in “unhappy,” “unable,” “unambivalent,” etc. The other un- has a meaning of reversing or removing, as in the words “undo,” “unzip,” “untie,” and “unfasten.” We can also distinguish these prefixes by the words they attach to. The negative un- attaches mostly to adjectives (as in “unhappy”), and the reversative un- attaches to verbs (as in “untie”), reinforcing the notion that they are two homonymous prefixes. If you look carefully through the dictionary, you can also find evidence of language change in morphology. Your dictionary might list -eth as “an archaic verb form of the third person singular.” You might find under “shoe” the plural form “shoon” (especially in British dialects). That plural with -en or -n is found regularly only in “oxen,” and in “children,” and with a specialized meaning in “brethren.” Verb forms are more likely to show irregular morphology. The word “holp” is a nonstandard form of the past tense “helped,” sometimes used in the Southern Midland and South of the U.S. It actually reflects the original past form before the standard dialects regularized it by grouping it with the verbs that take -ed in the past. Often, you will see choices—this form or that one—which indicate modern variation and change in progress. Most often, irregular verbs will become regular by taking an -ed form in the past tense. Dozens of verbs present us with such choices. Ask yourself which forms you use in a sentence like : awoke/awaked; besought/beseeched; knelt/kneeled; leapt/leaped; “Yesterday, Lee lit/lighted; strove/strived” and so on for many more.
Exercise 9 The past participle is the form of the verb used after an auxiliary verb like have, as in Robin has given me her pencil. What is the past participle of the following verbs: mow, prove, show? Does the dictionary list the forms that you chose?

A dictionary will often alert us to ongoing disputes by language users as to which forms are considered correct. Historically, for example, the word “data” comes from a Latin plural meaning “things given”; the singular is “datum.” Therefore, some conservative writers insist that “data” take a plural verb, as in “These data are inconclusive.” A majority of speakers treat “data” as a collective word like “information,” and would say “The data is inconclusive.” The original singular, “datum”, has also come to take a regular plural “datums” when it refers to points used as reference in surveying or geology. There is very little the dictionary can do to change how people use language.

One of the main reasons you’ll look a word up in a dictionary is for semantic information to learn the word’s meaning, and again, the dictionary can provide at least one model for how we organize words in the mind, relating them by meaning and grouping various senses together. Writing a dictionary definition is harder than you might think. Many lexicographers follow precise guidelines on how to define. But this sometimes leaves us frustrated when we find definitions such as the following:



luminosity—‘the condition or quality of being luminous’ contemptible—‘worthy of contempt’ absorbency—‘the quality or state of being absorbent’

These definitions are logically sound and literally true, but they are also tautologous— they use the same words or roots in the definition as are found in the headword. I bet if you knew the meaning of “luminous,” you could probably figure out the meaning of “luminosity.” With tautologous definitions, the user needs to look up an additional word. Some dictionaries, however, will amplify the traditional tautologous definition or avoid it all together. Thus “accuracy” is defined in one dictionary as “the quality of being accurate,” while in another, it is defined as “the quality or state of being correct or precise.” Other traditional definitions first name the class, and then what distinguishes the class from other things within the class. A “skink,” for example, is “any of a family (Scincidae) of widely distributed lizards having an elongated, shiny body, smooth scales, and short legs.” (From Webster’s New World College Dictionary) Different dictionaries serve the needs of very different populations. An unabridged dictionary needs to be authoritative, but it also serves the general population, not specialists, who will consult more specialized dictionaries of law or physics, for example. Compare the definitions of a word like “red”. Here is the first definition from The American Heritage Dictionary, 4th edition (2000): “the hue of the long-wave end of the visible spectrum, evoked in the human observer by radiant energy with wavelengths of approximately 630 to 750 nanometers.” Here is “red” in The New Oxford American Dictionary (2005): “of a color at the end of the spectrum next to orange and opposite violet, as of blood, fire, or rubies.” The World Book Dictionary (1994) defines “red” as ‘the color of blood or of a ruby. Red has the longest light wave in the color spectrum.’ The background assumptions and level of abstraction can vary considerably. Some dictionaries differ dramatically in the organization of each entry. The Oxford English Dictionary (OED), for example, is an excellent tool to trace the historical development of the language, but may not be your first choice if you want to find a current meaning quickly. The OED was first called A New English Dictionary on Historical Principles and thus lists the oldest attested meaning first, and then allows you to trace the word’s development in meaning through the centuries, with detailed contextual quotes often starting in Old and Middle English. Thus, if you look up “explode,” you’ll first encounter the obsolete definition originally used in English in 1621:
To clap and hoot (a player, play, etc.) off the stage; hence gen. to drive away with expressions of disapprobation; to cry down; to banish ignominiously.

Not until you get to the fifth sense do you encounter anything like the common meaning today:
To “go off” with a loud noise. Of gas, gunpowder, etc.

Other dictionaries try to organize meaning with the most frequent meaning first, figuring that’s what most people would want to know. Thus, if you’re reading the financial news and want to know what “amortize” means, the OED contains two obsolete definitions (marked by daggers), and a third definition before you get to the most current and common meaning, listed fourth. The New Oxford American Dictionary (2005), published by Oxford University Press, is like the OED, but it is a much more user-friendly and contemporary general-purpose dictionary; it lists the common meaning first: “reduce or extinguish (a debt) by money regularly put aside: loan fees can be amortized over the life of the mortgage.” Many definitions are supplemented with sample

What Dictionaries Reveal


sentences that provide a greater context and illustrate the syntactic properties of the word. Within a definition, meaning is grouped by senses and subsenses, but the exact details are subject to a judgment call on the part of the editors. Some very common words have incredibly complex uses in the language, and it can take months for a lexicographer to detail all the senses of such words. Words that have several related meanings are polysemous. Polysemy often arises from extending a word’s meaning in a metaphoric, less literal way. For example, the word “broadcast” used to mean only “to scatter seed broadly.” Now it is used for the transmission of radio or television signals through the air-waves. Metaphorically, the broadcasting of information is like casting seeds. Note another word that is used in both ways: “disseminate,” which means “to spread seed”, or “to give out information.”
Exercise 10 In the Oxford English Dictionary (1989) or in Webster’s Third New International Dictionary (1961), look up one of the following verbs: go, run, set. How many definitions did you find? How many subdefinitions? Do you agree with the way the editors organized these entries? How easy is it to find a particular meaning in these entries?

One aspect of semantics deals with homonyms—words that are spelled the same and pronounced the same, but which often have dramatically different meanings and different origins. For example, “bear” = “the ursine mammal exemplified by polar bear, grizzly bear, etc.” vs. “bear” = “to carry.” It is sometimes difficult for editors to distinguish homonyms from related senses of the word. If words share the same etymology, they will sometimes be included in the same headword, even if the meaning is widely divergent.
Exercise 11 How many senses of foot can you think of? Compare your answers with the definitions given in a large dictionary. Would you group the different senses in quite the same way as the editors? Would you want to separate the sense used in poetry from any of the others?

Some other important knowledge we have about words is the relations between them. For example, when asked to name word opposites (antonyms), everyone can name the opposite of “hot” as “cold.” We can also name words that have substantial meaning overlap (synonyms) such as “happy,” “joyful,” “glad,” and “elated.” A good dictionary will list such words. And a great dictionary will help you distinguish synonyms or related words through usage notes and sample sentences.
Exercise 12 Look up the word “imply”. Does your dictionary distinguish between “imply” and “infer”? The distinction between these words is important to many linguistically conservative writers. What guidance, if any, does your dictionary provide for synonyms such as “impertinent,” “insolent,” “intrusive,” “meddlesome,” “obtrusive”? Compare the New Oxford American Dictionary (2005) with its gray boxes of “The Right Word” or “Usage” notes. Compare also The American Heritage Dictionary’s (2000) synonyms and usage notes. How can these features improve your use of language as a writer?



Computers use bodies of data known as corpora (singular, corpus) to help define words. Corpora are gathered from several sources, such as magazines, newspapers, the Internet, and others, carefully coded to reflect the type of source—sports magazine, literary review, cutting-edge review of technology, popular music magazine, and so on. By searching through a corpus, which can run into the billions of words, lexicographers have found, for example, that the word “cause,” which means, roughly, “to make happen” is used about 90 percent of the time with negative events. For example, “The storm caused a flood.” It is important to capture these primarily negative connotations in a formal definition, both for accuracy’s sake and to alert readers to potential awkwardness in sentences like “This policy caused great prosperity.” This use of computers allows lexicographers to search through vast quantities of data with the ease of a few clicks, and to more accurately define the words around us.

Word Origins
Many dictionaries supply us with the word origin or etymology of a word. Some are succinct and often contain cryptic symbols and abbreviations. For example, the Random House Dictionary (1987) contains the following etymology for amity:
[1400–50; late ME amit(i)e < MF amitie, OF amiste(t) < VL *amicitat-, s. of amıcitas, ¯ ¯ ¯ deriv. of L amıcus. See ami, amiable, -ity] ¯

The inside front cover contains the etymology key in which the abbreviation ME stands for Middle English, < stands for “descended from, borrowed from,” MF and OF are Middle and Old French, VL is Vulgar Latin, the asterisk is a reconstructed form with no documentary evidence, the s. stands for ‘stem,’ deriv. stands for ‘derivative’ and L stands for Latin. The user is then cross-referenced to some words and an affix related to the word in question. In contrast, a dictionary like The World Book Dictionary (1994), which is designed for “all members of the family” and for “students of various ages,” contains the following:
[< Middle French amitié < Old French amistie < Vulgar Latin amıcitas < Latin amıcus ¯ ¯ ¯ friend]

One need not consult a key of abbreviations to understand this information. The American Heritage Dictionary (2000) etymologies also eschew abbreviations, but often contain a reference to see a form in an appendix. For example, if you look up “fable,” you’ll see that it ultimately comes from Latin farı ‘to speak.’ Then “See bha-2 in ¯¯ ¯ Appendix I.” If you look in that Appendix, you’ll see a list of Indo-European roots, roots linguists have reconstructed using a technique known as the comparative method, in which linguists compare various linguistic forms from related languages, use their knowledge of likely sound and meaning changes, and project backwards in time to reconstruct the undocumented language. The reconstruction of Indo-European represents our best guess at the language spoken in the fifth millennium b.c. by the Indo-Europeans, a people who spread their dialects, which evolved into different languages, from Iceland to India. Thus the following families of Indo-European are all distantly related: Irish and Welsh (Celtic), English and German (Germanic), French and Spanish, and Latin before them (Italic), Russian and Polish (Slavic), Lithuanian (Baltic), Pashto and Farsi/Persian (Iranian), Hindi and Urdu (Indic), and languages with their own branches such as Greek, Armenian, and Albanian, and languages no longer spoken (Hittite, Tocharian).

What Dictionaries Reveal


Many dictionaries include a family tree of these languages. But The American Heritage Dictionary (2000) allows us to explore the many related roots that descend from this primeval tongue and which directly or indirectly are now in the English language. Under ¯¯ bha-2, we find that Latin farı has given us words having to do with speech in some way, ¯ such as “fable,” “ineffable,” “infant,” “preface,” and “famous.” Greek has given us related words like “prophet,” “aphasia,” “euphemism,” “phonetic,” and “blaspheme.” Germanic has given us words like “bann,” “banish,” “bandit,” and “boon.” It takes professional training to be able to fully understand and justify all of the claims of word relatedness, which includes a knowledge of plausible sound and meaning changes. However, any language lover should be able to enjoy this etymological feast of words in such a never-ending banquet.

Exercise 13

Using The American Heritage Dictionary (2000), trace the origins of a word like father back to its Indo-European roots. How many words in English have descended from this root? How about for the verb bear?

Tracing the different origins of words can give us a better idea of language history and the associations we have with words. Compare all these words, which derive from a root meaning “heart”: the native English “hearty,” the Latinate “cordial,” and the clinical “cardiac” from Greek. There are different connotations and levels of formality in words like these. Likewise, compare the Greek origin of “cathedral,” meaning “seat” (in this case, a church with the seat of a bishop), with the Old French origin for “chair,” with the more recent borrowing of a type of reclining chair, a “chaise longue.” In addition to complex etymologies, The American Heritage Dictionary (2000) contains many brief word histories that shed light on language and culture. The entry for “tattoo” (meaning the ink design on the skin) tells us that the word was first introduced in English from Captain James Cook, the English captain who sailed around the world from 1768 to 1771. When he and his crew encountered Polynesian societies such as the Tahitian and Samoan societies, they introduced the word and the practice, until recently still associated primarily with sailors, to English speakers worldwide. Although its etymology is interesting, and the word itself comes from the Greek root for “truth,” we should never think that the “true meaning” of a word is its historical one. Meanings are constantly changing, redefined by each generation. While currently the word “nice” has positive connotations, it comes originally from a Latin word meaning “ignorant,” and was used to mean “foolish” in English for many years. Although upon reflection we could connect the roots for “decimal” and “decimate,” only a very small percentage of speakers use the verb “decimate” in its literal and historic sense of “kill every tenth person as a punishment.” It has now come to mean “destroy or kill” not just a fraction but a large part of a group or an area. Would you trust someone who tried to argue that “meat” still means any food (as it still does in compounds like “mincemeat” or “nutmeat”)? At what point can we agree that “girl” means “young female human,” and not just “young person” as it used to centuries ago? How many years does it take before an innovative meaning becomes the standard one? For example, in recent years, especially since the 1960s, the word “gay,” which used to mean “happy,” has been used as a euphemism for homosexuals, especially men. This now is the normal current use of the term, and the “happy” meaning of “gay” has receded like the tide.



Encyclopedic Information
College dictionaries include much encyclopedic information—information about the world, as well as information about the words of language. Such information allows us to answer questions like “Where is Myanmar?” and “What does the acronym BATF stand for?” It is sometimes difficult to tell where one’s lexicon ends and where one’s encyclopedia begins. Some dictionaries blur the distinction by including much encyclopedic information. The Century Dictionary and Cyclopedia (1897) is well named. Dictionaries such as the Random House Dictionary (1987, now Webster’s Random House Dictionary) and The American Heritage Dictionary (2000) include much encyclopedic information between their covers, listing geographic and biographical terms, making the dictionary a general reference resource. Other dictionaries, such as The World Book Dictionary (1994), restrict the dictionary mostly to regular vocabulary since the reader will presumably also have the World Book Encyclopedia (2003) at hand. The Webster’s Third New International Dictionary (1961) had to cut out all the encyclopedic information in order to fit into one volume and include over 100,000 new terms. The company spun off the other information into its line of independent geographic and biographical dictionaries, but maintains the terms, integrated with a much smaller vocabulary, in its college dictionaries. Lexicographers sometimes wrestle with how much encyclopedic information to include in a definition. For example, is the fact that beavers gnaw trees to make dams part of a definition of what a beaver is, or is it simply part of our knowledge of the world? How much culturally specific information should be included in a definition of “prom” as a formal dance, as opposed to a thorough description of a prom that includes rented limousines and corsages?

Role of the Dictionary in Society
A Brief History of Some Dictionaries
The dictionary today is very different from some of the first dictionaries. These were bilingual word lists, usually from Latin to English. Later, when English began to be used more broadly in intellectual circles, and many Latin words were freely used in English, dictionaries came to be glossaries of hard words that helped the growing bourgeoisie. Thus in 1604, Robert Cawdrey published the first monolingual dictionary, his Table Alphabeticall (Cawdrey, 1970), a list of about 3,000 less common words, usually with brief, single-line definitions. Here are a few examples from the letter C:
calamatie calcinate calefie calygraphie calliditie trouble, affliction to make salt make warme, heate, or chafe (g) fayre writing [the (g) means from Greek] craftiness, or deceit

Cawdrey began a lexicographic tradition of plagiarism and piracy that continues to this day. Each lexicographer builds on the tradition of preceding works. Cawdrey took words and definitions from earlier Latin-English dictionaries with little change. Cawdrey’s successors such as John Bullokar in his An English Expositor (1616; Bullokar, 1971) lifted words and definitions from Cawdrey and other sources and published a dictionary of about 6,000 words. Throughout the seventeenth century, dictionaries included larger numbers of words,

What Dictionaries Reveal


reflecting both competition and a swelling of learned vocabulary from the growing classes of literate merchants. By 1706, a dictionary with both common and “hard words” contained 38,000 words. In 1721 Nathan Bailey added etymologies to the 40,000 words in his An Universal Etymological English Dictionary. The creation of a language academy in France, and especially its 1694 publication of a prestigious dictionary of the French language, led many Englishmen to realize the inadequacy of their own state of language. Many literary figures such as Daniel Defoe and Jonathan Swift called for an English language academy, which they thought would regulate the language by establishing authoritative standards and preventing further language change. In 1747, Samuel Johnson rose to the challenge of creating a Dictionary of the English Language, published in 1755, in which he wrote highly colorful definitions of 40,000 words, many of fine literary quality, which he supplemented with illustrative sentences from English literature, the first time this was done in a monolingual English dictionary. The Preface to his dictionary is an outstanding literary achievement in its own right, and still worth reading today. Noah Webster’s The American Spelling Book (1783), widely used in schools and widely imitated by competitors, led him to champion American copyright laws. Its success in the schools inspired him to create his first dictionary in 1806, his A Compendious Dictionary of the English Language, with 37,000 words, very briefly defined, but which contained many Americanisms for the first time—words such as “skunk.” This dictionary contained characteristically American, as opposed to British, spellings in words such as “color” and “honor” without the u, and “theater” and “center” ending in -er instead of the British -re. It is an unremarkable dictionary, with definitions usually only one line a column wide. It also contained an appendix with lists of currencies, weights, measures, post offices, exports, and a chronology of the world, which lists the Creation in 4004 b.c. Webster’s 1828 American Dictionary of the English Language was far more ambitious, with 70,000 words, much scientific vocabulary, and better-defined and distinguished word senses. Webster spent ten years studying languages in order to improve his etymologies, though he ignored the growing scholarship in German from early linguists such as Jacob Grimm, and was wrong in virtually every one of his Semitic etymologies, e.g. saying that “abash” came from Hebrew and Chaldean instead of Old French. From 1830 to 1864, there was a “war of the dictionaries” between Noah Webster and his successors, and rival lexicographer Joseph Worcester. There was a lexicographic arms race, with each dictionary outdoing its predecessors in number of words, use of illustrations, and other features. Ultimately, the race was won by the G. & C. Merriam Company, now known as Merriam-Webster, which continued to expand its line of dictionaries until it produced in 1934 a dictionary with 600,000 entries, the largest ever. In the United States, however, the success of the Merriam company, coupled with expiration of copyright on earlier editions of Webster’s dictionary, and several lawsuits, enabled any company to claim that their dictionary was a Webster’s dictionary. In fact, the name Webster is increasingly used by a variety of companies since the name has come to be used synonymously with dictionary. The World Publishing Company originally issued the Webster’s New World Dictionary in 1951. A defunct Funk and Wagnalls Comprehensive Standard International Dictionary published by Ferguson in the 1950s (and revised in 1973) was reissued in 1984 by another company as Webster Comprehensive Dictionary. The Random House Dictionary (1987) is now retitled the Random House Webster’s. Even Microsoft’s Encarta World Dictionary was the Encarta Webster’s Dictionary (Soukhanov, 2004) in its second edition. (Given all the resources of the Internet, Encarta was discontinued at the end of 2009.)



Contemporary Issues
Many people believe that the dictionary should be a gatekeeper to our language. In this view, it should keep out the unpleasant or faddish words, and preserve the original and correct meanings. This approach is known as a prescriptive one because it tells users what they should do. Samuel Johnson began his dictionary, as he wrote in his Plan, with the intent “to preserve the purity and ascertain the meaning of our English idiom” (when “ascertain” meant “to make certain; to fix; to establish”). However, he notes in his Preface to the completed dictionary that he could not justify his original purpose.
When we see men grow old and die at a certain time one after another, from century to century, we laugh at the elixir that promises to prolong life to a thousand years; and with equal justice may the lexicographer be derided, who being able to produce no example of a nation that has preserved their words and phrases from mutability, shall imagine that his dictionary can embalm his language, and secure it from corruption and decay, that it is in his power to change sublunary nature, or clear the world at once from folly, vanity, and affectation. With this hope, however, academies have been instituted, to guard the avenues of their languages, to retain fugitives, and repulse intruders; but their vigilance and activity have hitherto been in vain; sounds are too volatile and subtile for legal restraints; to enchain syllables, and to lash the wind, are equally the undertakings of pride, unwilling to measure its desires by its strengths . . .

Almost all dictionary makers, however, view their role as recorders—as descriptive. Read the introductory section to almost any dictionary and you will find statements similar to the following, from Webster’s New International Dictionary (1909): “It is the function of a dictionary to state the meanings in which words are in fact used, not to give expression to the editors’ opinions as to what their meanings should be.” The G. & C. Merriam Company continued this descriptive view in the second edition of the New International Dictionary (1934), backing up their definitions with 1,665,000 citations—printed examples of the word used in a variety of contexts and printed sources. For their third edition in 1961 (still the current edition), they had gathered an additional 4.5 million citations, in addition to those gathered in the Oxford English Dictionary (1989) and other sources, including linguistic atlas studies. Little would the publishers realize the intensely negative reception it would receive, largely as a result of changing or dropping some of the usage labels such as “colloquial” or “illiterate.” The New York Times claimed that Webster’s had “surrendered to the permissive school,” which “serves to reinforce the notion that good English is whatever is popular” (Sledd and Ebbitt, 1962: 100). Dozens of critics expected the dictionary to do more to instruct the user about “good” and “proper” English, to preserve distinctions fewer and fewer people made, for example, between the meanings of “disinterested” and “uninterested” (Sledd and Ebbitt, 1962: 110). Lexicographer Sidney Landau observed that “such a fastidious attitude serves to mark the critics as belonging to a high social class.” He then quotes linguist Dwight Bolinger, who noted that the subtle distinctions between words such as “lie” and “lay” are “fragile and impractical, and the price of maintaining them is too high. But that is exactly what makes it so useful as a social password: without the advantage of a proper background or proper schooling, you fail” (Landau, 2001: 75). The American Heritage Company attempted to buy out G. & C. Merriam in order to suppress the third and reissue the second edition while working on a future edition of its own. When this attempt failed, it began working on its own dictionary that would

What Dictionaries Reveal


“faithfully record our language” but also give guidance to its users. A unique feature of the American Heritage dictionaries (now published by Houghton Mifflin) is their usage panel, which in each edition consists of a variety of writers, critics, and academic figures who are asked to vote on hundreds of usage questions. The original panel of 105 people consisted mostly of older (average age 64) white men (90 percent), though more recent panels, headed by linguist Geoffrey Nunberg, have been younger and more balanced, but sometimes more conservative. For example, while 90 percent of the 1969 panel disapproved of structures like “quite unique” (on the grounds that “unique” is an absolute adjective and thus cannot be modified), 80 percent of the panel from 2000 disapproved of it. In contrast, although 44 percent of the 1969 panel accepted the sentential use of “hopefully,” as in “Hopefully, we shall complete our work in June,” only 27 percent accepted this usage in the panel’s 1986 survey.
Some people think definitions should quote only the best writers and usage. Johnson quoted Milton, while Shakespeare is quoted extensively by the OED. Webster’s Third New International Dictionary took flak for including a few quotes from Ethel Merman and Yogi Berra. What sources do you think a dictionary should use for its sample sentences? Should lexicographers make them up to clearly and concisely illustrate the words, or should they prefer famous literary figures?

Exercise 14

A dictionary records new words and our changing culture. Dictionaries that came out in the 1980s boasted that they included words like AIDS, CD-ROM, etc. Dictionaries from the twenty-first century brag about including trendy words or phrases like “blog” and “low-carb,” and the media love stories about new words like “podcast” being admitted into a dictionary. Typically, lexicographers want to find a certain number of written citations of words from a number of different sources used over a period of years before they will enter a word in the dictionary. A delicate area for lexicographers is how to handle trademarked terms that have entered the common language. For example, many people call any facial tissue a Kleenex. A dictionary may define the product name generically, but will then note that it is a trademark. Can you “xerox” on a Canon copier?
In a dictionary, look up a few trademarked product names that you use generically. Think of everyday products you consume: food, drink, pharmacy items, and so on. Does the dictionary reflect your use? Does it distinguish between the word’s use in upper vs. lower case?

Exercise 15

As the number of Americans who attend college has steadily increased, especially since World War II, so too has the competition for college dictionaries. The marketing departments of most dictionary publishers boast about the number of entries in their dictionaries. Presumably, if you are considering buying two dictionaries for a similar price, and one advertises 150,000 entries, while another 163,000, you’d probably think you were getting a better deal by purchasing the one with more entries. But how are entries counted? The publishers and editors never say. Some dictionaries, such as Webster’s New World College Dictionary (2004), list the nouns and verbs under one headword. Others, such as Merriam–Webster, list them as separate headwords. Most dictionaries simply list, without defining, clearly related words such as “fluently” under



the headword “fluent.” Does this increase the total number of words? Sure, they are separate words, but are they entries? Sometimes dictionary marketers finesse the issue by listing the number of “definitions,” though as we’ve seen a single word like “foot” can have multiple definitions. Citing the number of definitions in a dictionary is a good way to increase their sales.

Dictionaries and Bias
When dictionaries were first written by individuals, it was quite easy to detect the bias and personality of their authors. Samuel Johnson famously defined “oats” as “A grain, which in England is generally given to horses, but in Scotland supports the people.” He defined a “lexicographer” as “A writer of dictionaries; a harmless drudge, that busies himself in tracing the original, and detailing the signification of words.” His definition of “excise” was “a hateful tax levied upon commodities, and adjuged not by the common judges of property, but wretches hired by those to whom excise is paid.” It would be difficult to find such bias (as well as style and humor) in today’s dictionaries. However, here is a counter-example: Noah Webster’s 1828 American Dictionary of the English Language, despite its flaws, small vocabulary (70,000 entries), and being 180 years out of date, is being reprinted and marketed by the Foundation for American Christian Education precisely because of the biases of its author. Webster frequently used sample sentences from the Bible, in part because the Bible was used as a literacy tool, and dictionaries held nearly equal prestige in many homes. And precisely because of the use of Biblical quotations, some segments of society would prefer that we use Webster’s definitions of words like “sin,” and traditional definitions of “marriage.” A dictionary must tread quite delicately between recording uses of actual language and appearing to condone them. This is particularly difficult in referring to different races, religions, and ethnicities. Look up the definition of “Jew” and what do you find? In addition to a fairly straightforward definition meaning an adherent of Judaism, you will find in some dictionaries notes that its use is highly offensive as an attributive adjective, as in “Jew lawyer” (vs. the more acceptable “Jewish lawyer”—if the religion of the lawyer must be described). A few years ago, some African Americans were shocked to see the word “nigger” defined in the dictionary as “Negro-usu[ally] taken to be offensive; a member (as an East Indian, a Filipino, an Egyptian) of any very dark skinned race—usu. taken to be offensive.” They lobbied the dictionary publishers to omit the word from future editions since they objected to the definition’s denotation of African Americans and people of color. They wanted to be able to argue that the n-word was solely a negative epithet without any reference to color of skin. But if the word did refer in meaning to people of color, then they thought it would be possible (if not permissible) for them to be called by that word. The publisher did not remove the word, but agreed to work to improve its usage note. Here is one very detailed note on the word from The New Oxford American Dictionary (2005):
The word nigger was used as an adjective denoting a black person as early as the 17th century and has long had strong offensive connotations. Today it remains one of the most racially offensive words in the language. Also referred to as ‘the n-word,’ nigger is sometimes used by black people in reference to other black people in a jocular or disparaging manner, or in some variant.

This note succinctly captures much of the controversy behind this emotionally charged word. It further explains its use within the group (correctly implying that non-African Americans are never able to use this word in this way). Finally, it draws a historical

What Dictionaries Reveal


parallel to another word that was used as an epithet of gays, but, in a move of linguistic judo, has been appropriated as a term of self-reference, a badge of honor. In the same vein, some women use the powerful negative connotations of the word “bitch” in a positive way when talking among women. Should dictionaries include other taboo words (inappropriate terms) like swears? Have you ever looked up some of the “four letter words”? Some school districts insist that publishers print special dictionaries with taboo words expurgated. Yet, sometimes because of the informal nature of taboo, the precise meanings can vary widely. To “fuck up” means “to do something badly or ineptly,” while to “fuck off” means (rudely) “go away.” The sensitive nature of the word was tested during the 2003 Golden Globe awards, when Bono of U2 said, “This is really, really, fuckin’ brilliant.” Because this was said in an over-the-air broadcast before 10:00 p.m., the Federal Communications Commission (FCC) investigated the complaints it received. The FCC originally ruled that the word “may be crude and offensive, but, in the context presented here, did not describe sexual or excretory organs or activities” and was therefore not indecent and thus did not merit fining stations for airing the word. Later, however, the word, whether used in the participial, exclamatory sense or the sexual sense, was deemed by the head of the FCC to be obscene, though no fine was issued.
Exercise 16 Look up a few taboo and/or offensive words in your dictionary. Do their definitions fit your senses of these words. Does your dictionary record the meaning of shit in the sense of Grab your shit and let’s go!? Should it? What guidance should the dictionary provide users for these emotionally charged words?

We have examined dictionaries from a variety of viewpoints. As Johnson (1747) said in the quotation at the beginning of the chapter, the best dictionaries are rarely if ever true to all of our understandings of all words in every context. However, they are incredible documents of our language and our knowledge of language. We can consult dictionaries to check pronunciation or learn about pronunciation differences. We can learn about language change in progress by examining word histories or choices in different pronunciations, word forms, and meanings. We can refine our own understanding of words and their many meanings by consulting definitions and sample sentences. We can delight in unusual etymologies, or fret about the new words or uses that have made it into the dictionary or challenge the definitions of some words. We can decide for ourselves whether to follow the usage notes and suggestions that different dictionaries make. After reading this chapter, I hope you’ll reflect more carefully on the dictionary. Do you own one? Do you instead use online dictionaries? What are the advantages of each? Depending on your needs, you might just find uses for several dictionaries, each of which, as Auden says, may be read in an infinite number of ways.

The American Heritage Dictionary of the English Language. (2000) 4th edn. Boston, Houghton Mifflin. Bailey, N. (1721) An Universal Etymological English Dictionary. Oxford, Oxford University Press.


Fallon Bullokar, J. (1971) An English expositor. Hildesheim, Georg Olms Verlag. Cawdrey, R. (1970) A table alphabeticall: London 1604. New York, Da Capo Press. The Century Dictionary and Cyclopedia. (1897) New York, Century Company. Funk and Wagnalls Comprehensive Standard International Dictionary. (1973) New York, Funk and Wagnalls Company. Huddleston, R. and Pullum, G. K. (2002) The Cambridge grammar of the English language. Cambridge, Cambridge University Press. Johnson, S. Dictionary of the English language. (1747) Princeton, Princeton University Press. Landau, S. I. (2001) Dictionaries: The art and craft of lexicography. 2nd edn. Cambridge, Cambridge University Press. An outstanding account of how dictionaries are made, written by a leading practitioner. The New Oxford American Dictionary. (2005) 2nd edn. Oxford, Oxford University Press. Oxford English Dictionary. (1989) 2nd edn. New York, Oxford. Random House Dictionary of the English Language, unabridged (1987) 2nd edn. New York, Random House. Sledd, J. and Ebbitt, W. R. (1962) Dictionaries and that Dictionary: A casebook on the aims of lexicographers and the targets of reviewers. Chicago: Scott, Foresman. A collection of essays on and criticism of Merriam–Webster’s Third New International Dictionary. Soukhanov, A. (2004) Encarta Webster’s dictionary of the English language. 2nd edn. New York, Bloomsbury. Webster, N. (1783) An American Spelling Book. New York, Hudson and Goodwin. Webster, N. (1806) A Compendious Dictionary of the English Language. Hartford, Sidney’s Press. Webster, N. (1828) American Dictionary of the English Language. Chesapeake Foundation for American Christian Education. Webster’s New International Dictionary of the English Language. (1909) Springfield, Merriam. Webster’s New World College Dictionary. (2004) 4th ed. New York, Webster’s New World Publishing. Webster’s New World Dictionary. (1951) Cleveland, World Publishing Company. Webster’s Third New International Dictionary, unabridged. (1961) Springfield, G. and C. Merriam–Webster. The World Book Dictionary. (1994) Chicago, World Book. World Book Encyclopedia. (2003) Chicago, World Book.

Other Resources
Aitchison, J. (2003) Words in the mind: An introduction to the mental lexicon. 3rd edn. Malden, MA, Wiley-Blackwell. Atkins, B. T. S. and Rundell, M. (2008) The Oxford guide to practical lexicography. Oxford, Oxford University Press. Creswell, T. H. (1974) Usage in dictionaries and dictionaries of usage. Chicago, University of Chicago, Department of English Language and Literature. Fontenelle, T. (2008) Practical lexicography: A reader. Oxford, Oxford University Press. Green, J. (1996) Chasing the sun: Dictionary makers and the dictionaries they made. New York, Henry Holt. A highly readable account about dictionaries through the ages. Hitchings, H. (2005) Defining the world: The extraordinary story of Dr. Johnson’s dictionary. New York, Macmillan. An entertaining and highly readable account of Johnson’s 1755 dictionary, a landmark lexicographic and literary achievement.

What Dictionaries Reveal


Kister, K. (1992) Kister’s best dictionaries for adults and young people: A comparative guide. Phoenix, Oryx Press. Now very outdated, but an invaluable guide to selecting the right type of dictionary, with very insightful criticism of many dictionaries. Available in many public and university libraries. Morton, H. C. (1995) The story of Webster’s Third: Philip Gove’s controversial dictionary and its critics. Cambridge, Cambridge University Press. An account of the planning and creation of Merriam–Webster’s Third New International Dictionary, and the unexpectedly heavy criticism it faced upon publication. Murray, E. K. M. (1977) Caught in the web of words: James A. H. Murray and the Oxford English Dictionary. New Haven, Yale University Press. A detailed biography of the leading OED editor, written by his granddaughter. Rollins, R. M. (1980) The long journey of Noah Webster. Pennsylvania, University of Pennsylvania Press. A biography of the first and most famous American lexicographer. Willinsky, J. (1994) Empire of words: The reign of the OED. Princeton, Princeton University Press. A critical examination of the linguistic authority wielded by the OED. Winchester, S. (1998) The professor and the madman: A tale of murder, insanity, and the making of the Oxford English dictionary. New York, HarperCollins. This national bestseller examines the story behind the thousands of contributions to the OED from a man in an insane asylum. Winchester, S. (2003) The meaning of everything: The story of the Oxford English dictionary. Oxford, Oxford University Press.

Neubach, A. and Cohen, A. D. (1988) Processing strategies and problems encountered in the use of dictionaries. Journal of the Dictionary Society of North America, 10, pp. 1–19.

The Century Dictionary. (2009) The Century dictionary [Internet]. Available from: [Accessed June 12, 2009]. One of the most detailed and beautiful dictionaries ever made is available for free online with thousands of high-resolution scanned pages and decent search functions. Originally published in 1889 by the Century Company, it was edited by American linguist William Dwight Whitney. This version is of the revised 1911 edition, which contained over 8,500 pages and thousands of woodcuts. Although outdated, it is still useful. This website is great for the visually impaired since it allows magnification up to four times. Cornerstone Baptist Temple. (2008) Webster’s 1828 dictionary [Internet]. Available from: [Accessed June 12, 2009]. A church-sponsored website that allows you to search an online version of Webster’s 1828 dictionary. Dictionary. (2009) Dictionary [Internet] Available from: [Accessed June 12, 2009]. A collection of a number of online dictionaries. Love to Know Corporation. (2009) Your dictionary [Internet]. Available from: [Accessed June 12, 2009]. Dictionary based on Webster’s New World. Includes a collection of other dictionaries, language articles, and games. Merriam–Webster. (2009a) Merriam–Webster online [Internet] Available from: http:// [Accessed June 12, 2009]. Free access to the words in the Merriam–Webster Collegiate Dictionary, with roughly 225,000 definitions. A subscription is required to access the over 476,000 entries in the unabridged version.


Fallon Merriam–Webster. (2009b) Visual dictionary online [Internet] Available from: [Accessed June 12, 2009]. A free site with over 6,000 images and graphics that will let you find the word for what you want to name. Oxford University Press. (2009) Oxford English dictionary [Internet]. Available from: [Accessed June 12, 2009]. By subscription. Better to access the site from a public library or at a college or university. Great for studying the historical development of the language. Princeton University. (2006) WordNet: A lexical database for the English language [Internet]. Available from: [Accessed June 12, 2009]. A free lexical database that groups words by synonyms. It allows one to study a variety of shades of meaning and word collocations. ThinkMap. (2009) Visual thesaurus [Internet]. Available from: [Accessed June 12, 2009]. A subscription-based web service that creates word maps illustrating related meanings. Urban Dictionary. (2009) Urban dictionary [Internet]. Available from: [Accessed June 12, 2009]. Users can define their own slang words and vote on others’ creations. Wiktionary. (2009) Wiktionary, the free dictionary [Internet]. Available from: en.wiktionary .org/wiki/Wiktionary:Main_Page [Accessed June 12, 2009]. This collaborative free-content multilingual dictionary is the dictionary companion to Wikipedia. As of this writing, it contained about 137,000 definitions, some with British and American pronunciation in IPA. Many were taken from public domain dictionaries through the use of bots.


Text Translation Approaching Otherness

Mary Boldt and Esperanza Roncero
This Chapter Explores:
Semantics Regional Variation Code-switching Spanglish Verbs and Tenses Register Subtitling

Translation is not a double of the original text . . . it is a literary genre in its own right, different from the other ones, with its own rules and goals. This is for the simple reason that a translation is not the original piece, but a path towards it. José Ortega y Gasset, 1937

As language professors at two different colleges, we have translated texts in such fields as literature, culture, medicine, social justice, and law. Both of us have taught seminars on translation, where we have seen that translation exercises are necessarily an act of reading in depth. Translation is a marvelous tool for any reader, but especially for students and professors wishing to improve their skills in entering into texts. In this chapter, we will attempt to show translation as a microscope that allows us to immerse ourselves in some of the innumerable elements that fuse together to form a text, and to show how they mirror the depths of the human mind and imagination, and of language itself. MB: One of my first encounters with another language was a book called See It and Say It in Spanish. I had pestered my father into buying it with the idea that, by looking at the pictures and memorizing the Spanish words written beneath them, I would be able to teach myself to speak fluent Spanish. Years later, I realized that languages are much more than lists of isolated words and pictures that can be translated back and forth on a one-to-one basis. Each language is a rich, complex system with its own beauty, reflecting its own special culture. ER: When I went back to Madrid, Spain, for the first time after living in the United States for almost four years, I felt the accent from Madrid as soon as I descended from the airplane. Even more shocking to me was that the experience of arriving in my own country felt as if I had landed in a completely new place. As I moved through the city, it appeared to me like a play or a movie performed in a language that I knew very well, Spanish. Then, from the river


Boldt and Roncero of people, a dear old friend from my childhood greeted me with enthusiasm, and when I searched for words to greet her, my mind went totally blank. I was so used to greeting people in English that I could not come up with the words, the tone, the intonation, the emphasis needed to express my delight at seeing her, and my eagerness at telling her all about my new life abroad. How could I possibly have lost my ability to think and speak in Spanish? Had English erased the Spanish language from my brain? It made no sense.

Actually, Esperanza Roncero’s (ER) experience in Madrid made perfect sense. After four years, her whole verbal thinking process was in English and had created a separate world from that of Spanish. ER realized that translation was the only means of bringing her two worlds closer. There would be multiple occasions on which she might not be able to make something understandable, but understanding is not necessarily the only goal of translation. Perhaps, at times, it is enough to make others aware of things they do not understand. Literary theorist Homi Bhabha uses the term “translation” “in the etymological sense of being carried across from one place to another. He uses translation metaphorically to describe the condition of the contemporary world, a world in which millions migrate and change their location every day” (Bassnett, 2002: 6). Deciphering meaning in a world filled with motion and change is central to translation. But can meaning stay constant as it leaps from one culture to another? Underpinned by the implications of work by the famous linguist Noam Chomsky, arguments have been made in favor of one single, underlying meaning that transcends cultures. On the other side, Wilhelm von Humboldt and, later, Edward Sapir (Sapir and Mandelbaum, 1949) and Benjamin Lee Whorf (Whorf and Campbell, 1956), argued that language and meaning are relative, varying with the culture in which they are situated. Translators’ approaches will be informed by the degree to which they choose to emphasize the commonality or the diversity of meanings and cultures. As we consider meaning’s relationship to people of different cultures, translators must ask ourselves at the outset what is most important: faithfulness to the original, source author and text, or comfort and ease of reading for the intended, target reader? Translation theorist Ernst-August Gutt (2000) argues that a translation should resemble the original text “only in those respects that can be expected to make it adequately relevant to the receptor [‘target’] language audience.” Furthermore, “the translation should be clear and natural in expression in the sense that it should not be unnecessarily difficult to understand” (Gutt, quoted in Venuti and Baker, 2000: 378). In other words, translators should not be blindly faithful to the original author and text to the point of overtaxing, and perhaps alienating, their readers in the target language. Philosopher Friedrich Schleiermacher (1812) and others would counter that good translations should indeed make target readers work: the texts in the original language (or “source texts”) come from a different culture, and the translated texts should contain the flavor and exoticism of that culture. If this means moving readers out of their comfort zone, so be it. This underlying tension between source- and target-based translation has been framed under numerous names, including accuracy vs. style, faithfulness vs. transparency, formal vs. dynamic (or functional) equivalence (Nida and Taber, 1969), semantic vs. communicative (Peter Newmark, 1991), and documentary vs. instrumental translation (Christian Nord, 1997). As we examine the choices that translators must make, we will see that each choice is influenced to some degree by this tension between author and reader, between source text and translated text, and between original culture and target culture.

Text Translation


Approaching the Source Text
Initially, translators might seem to have an easy job: after all, livre, Buch, and libro all translate as “book.” Their blood pressure will rise, however, in situations like the one described by translator William Weaver:
An Italian friend of mine called me up; she was practically in tears. She was translating an American short story. I think it was by Tillie Olsen—and in it there was a sentence more or less like this: ‘After school, Jimmy walked through the backyard, opened the screen door, came into the kitchen, and fixed himself a peanut butter sandwich.’ She was in despair, because in Italy there are no backyards, no screen doors, and no peanut butter sandwiches. (Covi et al., 1987, p. 88)

What can translators do in a case like this? Translate the sentence word for word, making the text correct but incomprehensible for the reader at the other end of a cultural divide? Add a footnote explaining the concepts of backyards, screen doors, and peanut butter sandwiches? Insert an explanatory phrase right into the text, as if it were part of the original? Replace the original references with more characteristically Italian landscapes and foods? Or choose a more general term? In no case can translators be completely faithful to the original text while making it accessible to readers with no knowledge of the American terms. Examples like Weaver’s highlight the elusive nature of equivalence, that is, the idea that a translated text can convey the meaning, style, intention, and impact of the original. A translator’s philosophy as to how much equivalence is achievable or even preferable will integrally shape the experience of target text readers.

Translating Meaning
One set of decisions confronting translators occurs at the semantic level, that is, with regard to meaning. If every word had just one meaning in every language, translation would be a simple, mechanical matter, reducible to a formula. But that is not the case. For example, if your college Spanish class makes plans with community leaders in Peru to visit at a given time and they are not on hand to greet you, it is not for lack of interest. For the community members, it is normal to keep doing their chores until someone comes to inform them of the arrival of the group. At that time, they will slowly proceed to the place where the students await. In this instance, each group lives with a totally different understanding of time: one group sets its appointments by the watch, while the other organizes them around their chores. Even among speakers of the same language, expressions of time can be challenging. A quick examination of the American Heritage Dictionary reveals five definitions of the word “hour.” Listeners must mentally sort through the total list of possible meanings, or semantic field, of that term to determine which meaning is intended in a particular context. In the statement “I’ll meet you back here in an hour,” the speaker is apparently referring to a 60-minute period of time, but the precise meaning depends on the speaker’s habits. Other categories of meaning can be equally ambiguous. Take, for example, the word “hamburger.” At first glance, the meaning seems straightforward: the American Heritage Dictionary contains just three definitions of “hamburger,” all revolving around the idea of ground meat. Further reflection, though, reveals the semantic field as broader than it initially appears. While the speaker might be referring to a Wendy’s hamburger, the


Boldt and Roncero

listener might be anticipating a Big Mac. The speaker might be talking of a patty on a wholewheat bun, while the listener might be salivating at the thought of a sesame-seed bun. The ground meat in question is probably ground beef, but it could be Aunt Sally’s special combination of ground beef and ground pork, or, in certain countries, ground horsemeat. And do we really need both halves of the bun, or can a hamburger still be a hamburger with no top half of the bun, or even with no bun at all? Differences in the images conjured up by words become even more numerous from culture to culture. The term “cheese sandwich” can conjure up images of a bright orange square between two soft, white, cottony slices of bread for certain North American readers. How do translators of a German novel react when confronted with the term Käsebrot? Do they simply translate it as “cheese sandwich,” or do they somehow indicate that a German cheese sandwich may be made with very thin, very dense squares of dark, possibly seed-filled bread, or perhaps with a chewy roll? And can translators convey the fact that it is unlikely that a German sandwich would contain American cheese? Similarly, seemingly simple words like “bread” and “salad” evoke vastly different images depending on the cultures of the author and reader. So far, we have spoken of differences in literal meanings, or denotations, of words confronting translators in a text. Traditionally, grammarians have spoken of a second level of meaning connected with a word, in addition to denotations. Connotations are the suggested, implied meanings of a given term. They go beyond literal dictionary associations, evoking the whole range of emotions, memories, and related information that a word might contain. For example, an American confronted with the term “apple pie” would very likely understand its literal meaning: a pastry made of apples, spices, and sugar, inside a crust (or under a crumb topping). Just as likely, though, is the set of associations that the term would also call forth: perhaps memories of the pie as comfort food concocted by a beloved relative; and probably an acquaintance with the idea of apple pie as an American icon of solid, likable tradition, an idea that is institutionalized in the saying “as American as apple pie.” Translators encountering the French tarte aux pommes could translate it as “apple pie.” However, not only is a tarte aux pommes different in appearance and taste than an American pie, but it also differs in its connotations.
a. At Hartwick College, it is a great honor to get a “JCH”—that is, a full-tuition scholarship named for the College’s founder, John Christopher Hartwick, and awarded each year to the top ten students in the third-year class. Since “JCH” is only understood through the lens of a Hartwick insider’s knowledge, though, translating the term would be an unenviable task. Make a list of at least 10 terms that would be understood by other students at your institution (or by your co-workers, or in your family) but not by “outsiders.” As you list the terms, list also their denotations and connotations. b. Try to come up with a working definition of the words “late” and “night” by interviewing 10 fellow students and other informants. If possible, find informants from different cultures.

Exercise 1

Exercise 2

Explain the difference between the following words and create situations in which only one of the pair or trio would be preferable: happy—glad dark—obscure

Text Translation to come back—to return to enter—to come in she receives—she gets lots of books—tons of books once more—again to recall—to remember sorcerer—wizard scream—cry—wail


Non-literal Meanings
Non-literal meanings, or figures of speech, such as similes and metaphors, present a special challenge to translators. They depend heavily on a shared body of knowledge and experience between author and reader, one that may or may not exist between two different languages and cultures. Similes compare two objects using like or as. An American may be “drunk as a skunk” in English but “drunk as a donkey” (saoul comme une bourrique) in French. Not only are skunks and donkeys very different animals, but the rhyme of the English phrase disappears in translation. One of the great tools of a writer wishing to make an impact is the metaphor. Like a simile, a metaphor compares two objects. Unlike the simile, the metaphor makes the audacious claim that the objects are not merely like one another, but actually the same. We would not be exactly flattered if someone called us “a snake in the grass.” In such a case, the speaker would be grouping together two objects, us and a snake, with two very dissimilar qualities: [human, walking upright on two feet, full of capacity for good (we hope!)] vs. [reptile, slithering along the ground, full of venom and negative associations stretching as far back as Adam and Eve]. The implication would be that one of the semantic traits, or qualities of meaning, associated with snakes—in this case, the venom—is actually characteristic of us, too. We are venom—but the speaker is not quite saying so directly. When we try to carry this powerful tool from one language to another, we are bound to run into challenges. Different languages have developed different sets of metaphors, and the shock value of comparing two apparently dissimilar objects can be diluted between languages. One category of metaphors involves terms of endearment. A mother might conceivably call her young daughter “Pumpkin.” The French literal translation of this metaphor would be citrouille or potiron, terms that would leave a French reader baffled, and wondering why the child is being called a soup vegetable. An American reader, meanwhile, might be subconsciously making the leap from “pumpkin” to “pumpkin pie,” with its connotation of sweetness. On the other hand, more than one English speaker has puzzled over the metaphor mon petit chou as a term of endearment. Since the primary definition of chou is “cabbage,” the French speaker would seem to be calling his or her loved one “My little cabbage.” A glance further down the list of dictionary definitions, however, reveals that a chou is also a cream puff pastry—a more likely association. Metaphors can vary fascinatingly in scope. Students at Marburg University in Germany conducted interviews that illustrated some universal tendencies among metaphors in different languages, but also some cultural variants. When expressing the sense of doing something difficult, speakers’ metaphors varied considerably in terms of dominant image:
Farsi: Japanese: to try to take two melons with one hand (to do something impossible) to be locked from eight sides (to be in a desperate situation)


Boldt and Roncero

Russian: Greek:

this does not go into the gate (a matter is too big to handle) to walk uphill (to have many problems) (Callies and Zimmerman, 2002: 47)

In each case, translators would have to decide whether to keep the more “exotic”sounding original metaphor or to switch to phrasing that would not interfere with comprehension.
Make a list of as many similes in English as you can think of (“as quiet as a mouse,” “as happy as a lark/clam,” etc.). Then find a speaker of another language (or do it yourself, if you speak more than one language) and compile a list of comparable similes in the other language. Where the images differ, comment on how the difference would alter a reader’s impression.

Exercise 3

Cultural Referents
The above examples all revolve around the fact that speakers of different languages come from different backgrounds, with different sets of shared values and experiences. What happens when the target language has no equivalent at all for a term in the source text? Cate Hutton (1994) describes such a situation in a report detailing her year as an American Library Association Fellow at the Library of Tibet. One of her duties was to teach a course developing a common vocabulary in library science. Her task was complicated by the fact that many common library science terms had never before existed in Tibetan. She notes that, at the time of her stay in Tibet, the term gar-chak was used interchangeably for “index,” “table of contents,” “catalog,” and “bibliography.” Lotfallah Karimi (2006) speaks of a similar translation conundrum involving the Farsi term mashhadi, denoting someone who has returned from a pilgrimage to Imam Reza’s shrine in the Iranian city of Mashhad. What Karimi terms a “religiously loaded” expression forces translators to decide whether to stress the literal meanings or connotations; how much to insert themselves into the text as visible mediators of meaning and connotation; and whether to opt for a “foreign” flavor or optimal comfort level for readers of the target text.
Students who come from abroad to study in the United States leave behind a world that they know very well. Many of them will find that certain familiar foods, items of clothing, furniture, and traditions do not exist in the United States. Ask international students to talk about these differences and make a list of words or expressions, including holidays, food, and clothing, that don’t have an equivalent in your own culture. Discuss ways of expressing those ideas in English. For example, huipil is a piece of clothing traditionally used by Mayan women both in Chiapas and in Guatemala. It is a kind of blouse that can have many colors, and it is profusely and carefully embroidered. Each community has its own distinctive huipil.

Exercise 4

Cognates, which are words of similar form in two different languages, can seem like a gift from heaven to harried translators. There is no need to hunt down a meaning—it’s

Text Translation


already there, gift-wrapped in this readily recognizable word. Sometimes, though, if a translation seems too easy to be true, it’s worth taking a second look. In Lancaster County, Pennsylvania, many shop names reflect the region’s Pennsylvania German heritage. One shop, selling gift items, bears the seemingly innocent, Germanic name of “Gifthaus.” A bit of investigation into the meanings of Gift and Haus in German, though, reveals that the shop’s name actually translates as “Poison House”! We assume that’s not what the shop owner intended to convey. Translator Catherine Porter (1987) calls cognates “invaluable, and insidious.” She points out that, even if cognate terms are close in meaning between the target and source languages, they may contain different connotations. She notes, as a translator of French into English, that English style often requires the use of short Germanic words instead of the longer Latin-based words that characterize French (Rose, 1987: 41–42).

Exercise 5

Using a French–English dictionary, translate the following French terms: action, addition, avarice, déception, défiance, effervescent, indolent, insupportable, and trombone. Write down the meaning in English, and, where applicable, which English synonym could be used in its place. Then do the same thing using a Spanish–English dictionary for the following words: eventualmente, librería, adición, excelente, embarazada, asistir, éxito, últimamente, largo, molestar, vaso, rápidamente.

Multiple Meanings
In the film Decoration Day, an African American soldier single-handedly takes out a German infantry unit during World War II in the bloody Battle of the Bulge. A white American officer then appears on the scene and asks, “You made all this?” Half dazed, half proud, the soldier says, “Yes.” Immediately, the officer shoots the soldier, leaving him for dead. The soldier survives, though, and is marked for years by this physically and emotionally devastating attack, believing he was shot by a fellow American acting out of racism. Eventually, research conducted by a childhood friend of the wounded soldier reveals that the shooting was not the act of a racist American officer after all, but of an enemy combatant in disguise. German soldiers did impersonate Americans in this battle, and this particular German soldier in disguise gave himself away with a slip of the tongue, in using “made” instead of the more native “did.” It is no wonder that the German character in Decoration Day was confused about the verb “to make.” Consider the flexibility of this extremely common word: we can make a birthday card; make money; make hay; make good time while traveling from Virginia to Florida; make trouble; make a bed; make someone happy; make a face; make out with another person; make out well in business; make someone out to be a liar; make the baseball team; make believe; make love; make good on a promise; make do; make eyes at someone; make peace; and make someone’s day. A translator would have to be an expert on English language idioms to translate these uses of “to make” correctly, by referring to the context of the phrase in question. It is often the most common words that cause difficulties (Nevin, 2006). Roughly thirteen columns of the HarperCollins German– English Unabridged Dictionary are devoted to the verb “to get”; ten to the verb “to go”; six and a half to the verb to come; and about four to the verbs “to do” and “to work.” Not only do the words have multiple meanings, but also many of those meanings may differ only by shades, necessitating a real feel for nuance in the original and target languages.


Boldt and Roncero

Although we readily throw around terms like Hispanic or French, it is easy to forget that these are shorthand terms referring to a wide variety of people and backgrounds. Pauline Courchesne (2007), a Franco-American speaker of French from Holyoke, Massachusetts, recounts an episode in which, while in France, she used the term from her region for pineapple, which is pomme de pin (literally, “apple of pine”). Unaware that the word for pineapple in France is ananas, she wrote “pommes de pin” on a customs declaration form for a box of pineapples that she was sending to Czechoslovakia. The French postal worker gave her a strange look but passed the box through, and the package eventually arrived in Czechoslovakia. It wasn’t until later that she learned that, in the eyes of the French postal worker, she had sent a box of pine cones to Czechoslovakia! Just as there are regional variations within a language, language variations, or dialects, can occur in other domains as well. In the cases where they link members of (and exclude outsiders from) social, ethnic, professional, religious, and other groups, they are called sociolects. When language characteristics of a particular group are lexical, or vocabulary based, and often specialized, we call those terms jargon. The term argot is similar in meaning to jargon but can carry the implication of being a secret language created to hide shady dealings. All of these variations demand translators who can cross language boundaries and convey words as a source of identity as well as of meaning.

Exercise 6

Check the respective British and American meanings of the following terms: biscuit, bonnet, boot, braces, chips, coach, football, jelly, jumper, pudding, torch, and trainer.

Another situation translators may encounter is the phenomenon of code-switching, or going back and forth between more than one language. Speakers who are competent in two or more languages may begin a sentence in one language and then finish it in another, or just sprinkle isolated words or phrases from a second language into their conversation. They may also superimpose a translation from one language onto a word or structure in another language, termed a calque; a calque is a word or phrase translated literally into another language. Code-switching may serve as a way to communicate quickly without the need to grope for a word or phrase in the original language, or it can function as a means of building a rapport with a mutual speaker of two (or three or four!) languages. Similarly, it can—even intentionally—be a mechanism for excluding or marginalizing monolingual listeners. Whatever its purpose in a given situation, codeswitching is impossible to translate from source text to target text in all its meaning, nuance, and force. Code-switching occurs frequently in the adventures of Agatha Christie’s Belgian detective Hercule Poirot, who lives in England but is a native speaker of French. One of the decisions translators such as Louis Postif and Robert Nobret had to make was how or whether to include Poirot’s code-switching, which gave him an air of exoticism in English, in their French translations. In Murder in Mesopotamia, Poirot says at one point, “I make the spring—like the panther—and, mon Dieu! The consternation!” Postif simply retains mon Dieu (“my God”) and the use of the definite article “the” in “the consternation” (la in French, with none of the jarring effect of its use in English), translating the

Text Translation


phrase as “je bondis comme la panthère . . . et, mon Dieu! Je sème la consternation autour de moi” (Christie, 1981). (Literally, “I leap like the panther . . . and, my God! I spread the consternation around me.”) Nobret, on the other hand, drops the mon Dieu but does nothing to show that Poirot switched languages in the original text: “je bondis comme un tigre, et . . . ah là là! . . . consternation générale” (Christie, 1998). (Literally, “I leap like a tiger, and . . . oh! general consternation.”) Translation theorist André Lefevere argues, “Obviously the writer of the original put [foreign words and phrases] there for a reason . . . To ‘regularize’ them, to translate them as if they were not foreign words in the original, may . . . detract from the complexity of the original” (Lefevere, 1992, p. 29).
Find two or three bilingual informants (include yourself, if you happen to be bilingual) and ask them which language they use at the grocery store; while having dinner with their family; and at the doctor’s office in the company of a friend or relative who speaks the same languages. Then, if they haven’t already answered this question, ask them which language they use to count, pray, express anger, and dream. Language is used to convey meaning, but its role goes deeper than that: it can express aspects of ourselves at the deepest level. To choose not to translate code-switching can gloss over an important statement about how a speaker perceives and organizes reality.

Exercise 7

One prominent example of code-switching in the United States is a blend of Spanish and English called (alternately) Spanglish, espanglish, and espaninglish. This phenomenon of blending two languages is not new or unique. Spanish itself, for example, is a blend of Latin and the indigenous languages of the Iberian Peninsula during the invasion of the Roman Empire. Spanglish can be encountered on television, and in music, advertisements, posters, buses, on the streets, and in countless other situations. In the sentence, “Tiempo is money,” we find not only a perfect example of the blending, or hybridizing, of two languages that create another, but also the hybridizing of two cultures, often leaving speakers of Spanglish with a culturally ambiguous identity. Translators would perhaps not have any problem translating the above sentence as Tiempo es dinero, or “Time is money,” but it would be very difficult to translate the cultural significance and the meaning behind this code-switching. To complicate this situation further, there is not one Spanglish, but several. It is clear that the language form spoken by Cuban Americans is different from that of Mexican Americans and Nuyoricans (Puerto Ricans living in New York City). Spanglish is not a unified tongue, but a series of variations, and although Spanglish speakers of different backgrounds can clearly communicate among themselves, they also need to translate to each other the variations of their particular Spanglish. All of the variations reflect a new cultural identity that is neither that of the English-speaking world nor that of the Spanish-speaking world, but rather that of Latinos in North America. Speakers of Spanglish sometimes feel embarrassed because of the frequent denigration of Spanglish as a faulty combination of English and Spanish, but they also use the term with a sense of defiance and pride in the new culture to which Spanglish gives voice. This cultural ambiguity is very well expressed in the work of the Chicana (Mexican American) writer Gloria Anzaldúa in her work Borderlands/La Frontera (2007), where she uses a unique Spanglish comprised of two variations of English and six of Spanish. In this text, the reader is introduced to the difficult reality of an identity residing in the cultural


Boldt and Roncero

ambiguity of the borderland. The borderland, in this case, is constituted by Anzaldúa’s condition of being a Chicana, a woman in a Chicano and Latino culture, a Chicana in white American society, and a lesbian in a straight world.
Read some Hispanic/Latino literature that uses Spanglish, or listen to some news broadcasts and shows directed at Hispanics and Latinos. Look for occasional switches into English such as these examples: “mi boss me pagó con cash por haber trabajado overtime” (my boss paid me with cash because I worked overtime); “Me tomo ahora mi break” (“I am taking now my break”). Look also for loan words such as “dealer,” “high school,” “biles” (“bills”), “librería” (“library”). And finally, look for calques, such as “rascacielo,” which is a calque (literal translation) of the English “skyscraper” (rasca means “it scrapes” and cielo means “sky”), “alto el fuego,” a calque of “cease-fire” (alto means “cease” and fuego means “fire”), and “lavaplatos,” which is a calque of “dishwasher” (lava means “wash” and platos means “dishes”). Write them down and then, with the help of a native speaker from one or more Spanish-speaking cultures, try to determine if the terms are a code-switch, a loan word, or a calque.

Exercise 8

Verb Structures and Tenses
We have been talking primarily about words and phrases as conveyors of meaning. However, the structure of those words and phrases can also function as a powerful, if subtle, conveyor of meaning and nuance. Consider the complex web of forms that can be woven by a single verb. As translators create a whole new structure in which to house the meaning of the original text, verb forms may undergo significant changes. For example, subjects in English may take the form of gerunds, or verbs acting as nouns. Gerunds take the form of a present participle (that is, they end in “-ing”). In Spanish, on the other hand, verbs can also be used as subjects but are structured as infinitives (the “to do” forms of verbs): so “Smoking is bad for your health” uses a gerund as the subject, and Fumar es malo para la salud uses the infinitive “to smoke.” Questions of verb structure are frequent in literary translation as well as in other texts. They present interesting challenges, and sometimes difficult choices, for translators. For example, the following fragment of Diamela Eltit’s Chilean novel, Lumpérica (1991),
atrapar cualquier cosa con la mano, sonreír, reír sin tope, estar sentada con descuido, en un banco de piedra, escondida de perfil en un árbol, bajarse las faldas en un viento repentino. (Eltit, 1991, p. 110)

was translated into English by Ronald Christ as:
catching any object whatsoever in her hand, smiling, laughing uncontrollably, being seated negligently on a stone bench, hidden in profile by a tree, settling her skirts after a sudden gust. (Eltit, 1997, p. 109)

The verbs in italics in the Spanish original were written in the infinitive form, while Ronald Christ translated them as present participles.

Text Translation


The infinitive form of a verb does not carry a sense of time. An infinitive is like a coat in a closet, hanging there, doing nothing. It does not even care if it is winter or summer. It is there, suspended in time and outside of time. However, the present participle does contain information about time. It is an action that is happening continuously in the past, present, or future. Ronald Christ chose to translate infinitives as present participles so readers would understand better the meaning of the fragment, but, in doing so, he had to change the sense of time in which those actions were taking place in Spanish.
Taking into consideration that the tense of a verb refers to different times in the real world, try to explain how the following sentences differ from each other. First identify the tense of the verb and then the time to which the verb refers in the real world, and finally identify the differences among the sentences. Elena comes tomorrow afternoon. Elena will come tomorrow afternoon. Elena is coming tomorrow afternoon. If you know other languages, create a list of sentences where the verb tense and the reference to real time differ. Explain how meaning is affected, and in which situations speakers use the respective forms.

Exercise 9

Intimacy and Respect
Languages can vary widely with regard to levels of formality. Contemporary Standard English, for instance, has just one way of saying “you” (although we also may hear “you all,” “y’all,” and “youse,” among other variants). Standard French, on the other hand, has two ways of saying “you,” and German has three ways. In Spanish, the singular “you” can be expressed in two ways: formally, as usted, and informally as tú. The plural “you” has another two forms: ustedes (formal) and vosotros (informal). The use of each of these forms varies greatly from region to region. (We are not talking here about Argentina, where we find still other forms of address.) A related challenge for translators is contained in verbs such as the French tutoyer, meaning to address someone with the informal tu. In addition to second-person forms of address, languages contain countless other nuances that convey intimacy and respect (and lack thereof). Do we translate the German Herr Professor Doktor literally as “Mr. Professor Doctor”? Or do we avoid translating it altogether, since the literal translation sounds so stilted? And what about the folksy, friendly way of addressing strangers as “hon” and “luv,” terms that have no romantic overtones and that can be heard on the streets of such varied locales as Baltimore and London? As translators address these nuances, our decisions are of crucial importance in avoiding hurt feelings and sometimes even international incidents.
a. It can be difficult for English speakers to imagine the levels of formality built into ways of saying “you” in other languages. As a very rough substitute, try the following experiment: approach someone with whom you are normally on a first-name basis and address him or her as Mr., Ms., Miss, or Mrs. . Also, if you think you can get away with it, address someone you normally address by a title (such as your professor) by his or her first name. Quickly explain why you were using such an unusual form of address, and then ask how the person felt when you used this odd

Exercise 10


Boldt and Roncero form. Try this out with several friends and acquaintances to see if you notice a pattern. b. Talk with several friends, classmates, or other acquaintances who are native speakers of languages whose “you” form varies according to degree of intimacy or formality. Ask your informants at what age they first remember being addressed with the formal “you” form, as opposed to the informal form (which is typically used for children). Find out what effect this change had on your informants’ sense of identity. Ask also which groups of people they address with the formal and informal forms of “you” in their native language, and why. (There are reasons of age, profession, respect, relationship, and other factors to be taken into account.) Ask the older informants how their own use of the formal form changed once people started using this form with them, and also when they accepted being addressed by the formal form. Ask if they can recall one or more examples of being addressed informally when they had expected to be addressed formally, and vice versa, and how they felt about it.

Studies in Sound
Sometimes word choice is influenced by phonetics, that is, by the way a language sounds. One of the contexts in which translators face phonetic considerations is when working with poetry. Two English-language poems about snakes help illustrate this point. Emily Dickinson’s A Narrow Fellow in the Grass (Poetry Foundation, 2009) and D. H. Lawrence’s Snake (1921) both contain a number of sibilant sounds reminiscent of a snake’s hissing. Phrases like “His notice sudden is” and “trailed his yellow-brown slackness soft-bellied down” would be hard to reproduce with exactly the same combination of imagery, associations, meter, and hissing sounds in another language. Translators must also decide whether or not to attempt to focus on the music of poetry by translating rhyme or to concentrate on the message of the source text. Eloquent arguments have been made for both choices, which accounts for the fact that readers may select from a range of poetry and prose translations of such great works of literature as the Iliad, the Odyssey, Dante’s Inferno, and Goethe’s Faust.

Beyond the Text: Who Is the Translator?
Amid all the questions confronting translators, there is not one monolithic, cookiecutter response. Translators of literature may have different priorities than translators of legal documents—or of feature films, as we will see in the next section. It can be helpful, though, to consider some of the qualities that set one translator apart from another. First, translators must possess outstanding skills in both the source and target languages and cultures. It is strongly recommended and common practice that translations be into translators’ mother tongues. A good dictionary is a necessary tool for translators, but just as important are encyclopedias, grammar books, books on regional and local usages of the target language, and history and culture books. Because translators must be able to read the original text in depth, we must have a profound knowledge of the language of the source text, as well as a good understanding of the context from which it came. When translators encounter a text, we do not necessarily have the luxury of inquiring of the author about the meaning of expressions or words. Instead, we might need to start what potentially could be a long search, very much in the manner of a detective.

Text Translation


Like detectives, translators do not operate in a vacuum. As we straddle the line between author and reader, source language and target language, our decisions will be informed to some degree by factors other than language or culture. Sometimes those factors are social or political. A character in a Mark Twain essay once said, “You tell me [where] a man gets his corn pone, [and] I’ll tell you what his ’pinions is” (Neider, 2000). In other words, opinions and decisions—and translations—can be influenced by people’s sense of their own empowerment or helplessness. Are we translating a text into, or out of, the language of a colonizer? Are we constrained by any real or perceived censorship? How shocking are we willing to make our language? How visible do we want to be, or should we be? Which texts are we even willing to translate? Our responses to these questions will be informed by our backgrounds and value systems.
You have just written a best-selling novel that is slated for translation into a language other than English. Write the text of an interview with the candidate applying for the job of translating your text, making sure that your questions reflect the qualities you find most important. As a helpful example, read the text of an interview with Geoffrey Brock, the successor to William Weaver as translator to the celebrated Italian author Umberto Eco: .html (Eco, 2005).

Exercise 11

Exercise 12

Compare three different versions of Psalm 23 from three different translations of the Bible. (This popular psalm was originally written in Hebrew.) Note the differences of wording and structure and the possible effects of those variations upon readers. Then try to compose a version for readers who have never seen a sheep before. The translation of sacred texts from their source language has had a profound impact on the shaping of languages and cultures throughout the world. Additionally, it has been a catalyst in the field of translation studies.

Specialized Fields of Translation
A few common fields of translation include commercial, literary, legal, medical, sacred, scientific, and technical. Translators of texts in any of these domains must first become experts in these fields in their own language, and develop a specialized vocabulary. Agatha Christie’s Murder in Mesopotamia requires a mastery of such terms as “school tablets” (cuneiform exercises in clay), “querns” (flat stones against which grain was ground), and “celts” (stone chisels). Each field, too, comes with its own particular audience and set of expectations. Translators have greater artistic leeway, for example, when translating a novel than when translating a contract or a user’s manual. While any of these fields of translation could be the subject of in-depth study, we will look briefly at one specialized field.

Sometimes films are dubbed from one language into another, but frequently they are subtitled. We are indebted to subtitlers for making films from other languages and cultures accessible to us. Still, who, when watching a foreign film, has not listened to a seemingly endless stream of words in another language, only to see them translated in a


Boldt and Roncero

clipped, two- or three-word phrase? We can sometimes be quick to critique, but it is helpful to remember that subtitlers of films and television shows operate under a different set of constraints than, for example, translators of a novel. Most strikingly, subtitlers are limited by the number of words they may use. In order to allow viewers to read quickly enough to follow the story line, and to avoid becoming too great a distraction from the images on screen, subtitles are generally limited to 32–40 keyboard spaces per line; no more than two lines may appear on the screen at any time. Subtitles generally remain on screen for anywhere from two to seven seconds (Hatim and Mason, 2000, p. 430). Under these circumstances, subtitlers must be concise while at the same time attempting to reproduce the spontaneity and informal register that generally mark oral texts. Anne-Laure Sallé (2000) offers an intriguing example of her line of thought when translating an episode of the American sitcom Frasier into French. In the original version, there is a reference to “that prune Danish that Dad dropped.” In Sallé’s French version, it becomes “ce sablé [‘shortbread cookie’] tombé [‘fallen’].” She notes:
Many words referring to food are untranslatable. In those instances, the best thing to do is to replace the source language by a more general word in the target language. A prune Danish is a kind of Viennese bun, a ‘viennoiserie’ in French. However, this word took up too much space . . . gâteau [“cake”] was still too long, but sablé fit in . . . What was important to convey in the context was that food was still on the floor.

Exercise 13

Choose a scene from a film that has enjoyed worldwide popularity, such The Lord of the Rings. Watch the film, making a list of possible challenges that you think subtitlers may face: obscure wording, obscenities, rapidfire speech, culture-specific references, etc. Then: a. Interview speakers of other languages who have seen the film subtitled in their language, and check online and print reviews to compare your list of anticipated challenges to their critiques of the film’s subtitling. b. If you speak the target language and can get your hands on a DVD of the film with a foreign language subtitle option, compare the English soundtrack with the subtitles, noting differences such as length of utterance versus length of subtitle, and questionable word choice.

As ER says to her students, a translation is never finished. We just decide to call it finished when we are fairly satisfied. Does this sound familiar? Yes! It sounds very much like the relation that artists and creative people have to their work. This is because translation is in itself an art. A catchy Italian phrase, traduttore, traditore, equates a translator with a traitor, but a good translation actually can improve the original text. In translating one world into the other, the major challenge turns out to be not language per se, but how to be demanding yet respectful of the social and cultural realities breathing within two different languages. Upon her return to Spain, ER had two choices about which language to use: either to translate so her experience would be easily understandable to her Spanish listeners, or to put her emphasis on the experience itself. The first choice would have been simpler, but would have sacrificed the reality of the original language while

Text Translation


perpetuating the cultural and social constructs of the Spanish one. The second one was more laborious and time consuming, in its quest to bring to view a social and cultural reality and identity other than that known by her specific group of Spanish speakers. ER chose the second option, reminiscent of a line by translation theorist Antoine Berman: “Good translation . . . stages an ‘opening, a dialogue, a cross-breeding, a discentering’ and therefore forces the domestic language and culture to register the foreignness of the foreign text” (Venuti, 1998, p. 81). Translation’s beauty and difficulty is precisely to perform the otherness of a language as it walks its path towards the original text, and thus to expose our own otherness in a world of difference.

Anzaldúa, G. (2007) Borderlands/La frontera: The new mestiza. 3rd edn. San Francisco, Aunt Lute Books. Bassnett, S. (2002) Translation studies. 3rd edn. New York, Routledge. Callies, M. and Zimmerman, R. (2002) Cross-cultural metaphors: Investigating domain mappings across cultures. Germany, Philipps-Universität Marburg. Available from: http:// [Accessed August 27, 2007]. Christie, A. (1981) Meurtre en Mésopotamie. Translated from the French by L. Postif. Paris, Librairie des Champs-Élysées. Christie, A. (1998) Meurtre en Mésopotamie. Translated from the French by R. Nobret. Paris: Librairie des Champs-Élysées. Courchesne, P. (2007) Franco-American language and identity. e-mail (July 29, 2007). Covi, G., Rose, M. G., and Weaver, W. (1987) A conversation on translation with William Weaver. In Rose, M. G. ed. Translation perspectives III: Selected papers, 1885–86. Binghamton, National Resource Center for Translation and Interpretation, pp. 84–91. Eco, U. (2005) Geoffrey Brock talks “The Mysterious Flame of Queen Loana” [Internet]. Available from: [Accessed June 13, 2009]. Eltit, D. (1991) Lumpérica. Chile: Grupo Editorial Planeta. Eltit, D. (1997) E. Luminata. Translated from the Spanish by R. J. Christ. Santa Fe: Lumen. Gutt, E. (2000) Translation as interlingual interpretive use. In Venuti, L. and Baker, M. eds. The translation studies reader. New York, Routledge, pp. 376–396. Hatim, B. and Mason, I. (2000) Politeness in screen translating. In Venuti, L. and Baker, M. eds. The translation studies reader. New York, Routledge, pp. 430–445. Hutton, C. (1994) 1993–1994 ALA Library Fellows Program Report, unpublished report. Karimi, L. (2006) Equivalence in translation [Internet]. Available from: http://www. [Accessed June 23, 2008]. Lawrence, D. H. (1921) Snake. In Lawrence, D. H. Snake and other poems, ed. Blaisdell, B. Massachusetts, Courier Dover Publications, pp. 44–45. Lefevere, A. (1992) Translating literature: Practice and theory in a comparative literature context. 2nd edn. New York, Modern Language Association of America. Neider, C. (2000) The complete essays of Mark Twain. New York, Da Capo Press. Nevin, H. (2006) Basic words can be the most difficult to translate [Internet]. Available from: [Accessed June 13, 2008]. Newmark, P. (1991) About translation. Clevedon, UK, Multilingual Matters, Series 74. Nida, E. A. and Taber, Charles R. (1969) The theory and practice of translation, Leiden: E. J. Brill. Nord, C. (1997) Translation as a purposeful activity: Functionalist approaches explained. Manchester: St. Jerome.


Boldt and Roncero Ortega y Gasset, J. (1937) Miseria y esplendor de la traducción. In Paz, O. ed. Obras Completas. 2nd edn. Madrid, Alianza Editorial. Peak. (2009) Globulated [Internet]. Available from: [Accessed June 13, 2009]. Poetry Foundation. (2009) A narrow fellow in the grass, Emily Dickinson [Internet]. Available from: [Accessed June 11, 2009]. Porter, C. (1987) Translating French feminism: Luce Irigary’s Ce sexe qui n’en est pas un. In Rose, M. G. ed. Translation Perspectives III: Selected Papers, 1885–86. Binghamton, National Resource Center for Translation and Interpretation. Rose, M. G. (1987) Translation perspectives III: Selected papers, 1885–86. Binghamton, National Resource Center for Translation and Interpretation. Sallé, A. (2000) Subtitling: ambiguity and complexity. Unpublished Master’s thesis. Sapir, E. and Mandelbaum, D. G. (1949) Selected writings of Edward Sapir in language, culture, and personality. Berkeley, University of California Press. Schleiermacher, F. (1812) On the different methods of translating. In Lefevere, A. ed. Translating literature: The German tradition from Luther to Rosenzweig. Amsterdam: Rodopi, pp. 66–92. Translation Directory. (2009) Portal for language professionals and their clients. [Internet]. Available from: [Accessed June 13, 2009]. Twain, M. (1901) Corn-pone opinions. In Neider, C. ed. The complete essays of Mark Twain. New York, Da Capo Press, pp. 583–586. Venuti, L. (1998) The scandals of translation: Towards an ethics of difference. New York, Routledge. Venuti, L. and Baker, M. (2000) The translation studies reader. New York, Routledge. Whorf, B. L. and Carroll, J. B. (1956) Language, thought, and reality: Selected writings of Benjamin Lee Whorf. Cambridge: MIT Press.

Other Resources
Baker, M. (1992) In other words: A coursebook on translation. New York, Routledge. Baker, M. and Malmkjær, K. (1998) Routledge encyclopedia of translation studies. New York, Routledge. Bassnett, S. and Lefevere, A. (1990) Translation, history and culture. New York, Pinter Publishers. Benjamin, W. (1923; Zohn, trans. 1968) The task of the translator: An introduction to the translation of Baudelaire’s Tableaux Parisiens. Translated from the French by H. Zohn. In Venuti, L. ed. The translation studies reader. 2nd edn. New York, Routledge, pp. 75–83. Bly, R. (1991) The eight stages of translation: With a selection of poems and translations. St. Paul, Ally Press. Borges, J. L. and Di Giovanni, N. T. (1973) Borges on writing. London, Dutton Borges. Burgin, R. and Borges, J. L. (1998) Jorge Luis Borges: Conversations. Jackson, University of Mississippi Press. Cronin, M. (2008) Translation goes to the movies. New York, Taylor and Francis. Eco, U. (2001) Experiences in translation. Toronto, University of Toronto Press. Karamitroglou, F. (2000) Towards a methodology for the investigation of norms in audiovisual translation: The choice between subtitling and revoicing in Greece. Amsterdam, Rodopi.

Text Translation


Kristal, E. (2002) Invisible work: Borges and translation. Nashville, Vanderbilt University Press. Mueller-Vollmer, K. and Irmscher, M. (1998) Translating literatures, translating cultures: New vistas and approaches in literary studies. Berlin, Erich Schmidt Verlag GmbH and Company. Paz, O. (1981) Traducción: Literatura y literalidad. Barcelona: Tusquets. Sallis, J. (2002) On translation. Bloomington, Indiana University Press. Schulte, R. and Biguenet, J. (1992) Theories of translation. Chicago, University of Chicago Press. Simon, S. (1996) Gender in translation: Cultural identity and the politics of transmission. New York, Routledge. Toury, G. (1978) The nature and role of norms in literary translation. In Venuti, L. ed. The translation studies reader. 2nd edn. New York, Routledge. Tymoczko, M. and Gentzler, E. (2002) Translation and power. Ahmerst, University of Massachusetts Press. Venuti, L. (2003) The translator’s invisibility: A history of translation. New York, Routledge. Vermeer, H. J. (1996) Skopos and commission in translational action. In Venuti, L. ed. The translation studies reader. New York, Routledge, pp. 219–238. Waisman, S. G. (2005) Borges and translation: The irreverance of the periphery. Lewisburg, Bucknell University Press. Zentella, A. C. (1997) Spanish in New York. In Garcia, O. and Fishman, J. A. eds. The multilingual apple: Languages in New York City. 2nd edn. New York, Mouton de Gruyter, pp. 167–202.

Abe, K., Kesper, N., and Warich, M. (2002) Domain mappings: General results. In Callies, M. and Zimmerman, R. eds. Cross-cultural metaphors: Investigating domain mappings across cultures. Germany, Philipps-Universität Marburg, pp. 29–40. Available from: Echevarría, R. G. (1998) Kay possa?! Is ‘Spanglish’ a language? New York Times, March 28, p. 7.

Alvarez, A. (2000) Basic computer Spanglish pitfalls [Internet]. Available from: http:// [Accessed June 13, 2008]. American Literary Translators Association. (2009) American Literary Translators Association (ALTA) [Internet]. Available from: [Accessed June 13, 2009]. El Castellano. (2008) La página del idioma español [Internet] Available from: www.el [Accessed June 13, 2009]. International Association for Translation and Intercultural Studies. (2003) International Association for Translation and Intercultural Studies (IATIS) [Internet]. Available from: [Accessed June 13, 2009]. Leonardi, V. (2000) Equivalence in translation: Between myth and reality [Internet]. Available from: [Accessed June 13, 2009]. Llorente, M. A. (1998) La página de la lengua española [Internet]. Available from: [Accessed June 13, 2009]. Information about dictionaries, grammar, culture, literature, and translation. Nida, E. A. (1994) The sociolinguistics of translating canonical religious texts [Internet]. Available from: [Accessed June 13, 2009].


Boldt and Roncero O’Hagan, M. (2003) Middle Earth poses challenges to Japanese subtitling [Internet]. Available from: [Accessed June 13, 2009]. Punto y Coma. (2009) Indice numérico [Internet]. Available from: translation/bulletins/puntoycoma/numeros.html [Accessed June 13, 2009]. RedIris (1998) Traducción en España [Internet] Available from: traduccion [Accessed June 13, 2009]. Shi, A. (2005) Accommodation in translation [Internet]. Available from: http:// [Accessed June 13, 2009]. SIL International. (2009) Translation theory and practice [Internet]. Available from: [Accessed June 13, 2009]. Translation Directory. (2009) Portal for language professionals and their clients [Internet]. Available from: [Accessed June 13, 2009]. University of Portsmouth: Paul Joyce German Course. (2009) Lost in translation [Internet]. Available from: [Accessed June 13, 2009]. The University of Texas at Dallas. (2009) Center for translation studies [Internet]. Available from: [Accessed June 13, 2008]. Wycliffe. (2006) Wycliffe Bible translators [Internet]. Available from: [Accessed June 13, 2009].


Machine Translation The Challenge of Ambiguity

Nan Decker
This Chapter Explores:
An Overview of Machine Translation Analyzing the Input Sentence: Finding Sentence Boundaries Finding Word Boundaries Multiple Meanings of Words Using Semantic Information Parsing the Sentence Converting the Input Sentence to the Target Language: Problems Translating Words Problems Converting Sentence Structures from Source Language to Target Language

The British left waffles on the Falkland Islands.

I have worked in the language industry for more than twenty years, much of that time helping to develop LogoVista EtoJ, machine translation software that translates English text to Japanese text (LogoVista, 2009). My contribution has been on the English side of things, helping to build an enormous, richly annotated dictionary of English and augmenting the very robust grammar of English used by its parser. This work has given me a tremendous appreciation for the vastness of human language and the mind that makes sense of it.

Introduction: Overview of the Machine Translation Process
Machine translation (MT) is the translation of text or speech from one language to another by the use of a computer program. Attempts to translate one language to another automatically have been ongoing since the advent of the computer age. Most other software applications under development for such a long time—word processing programs, say, or inventory tracking programs—operate more or less flawlessly and predictably. Such is not the case with MT, as anyone who has experimented with a machine translation program knows. When MT was first conceived of in the late 1940s and 1950s, it was thought that, with time, systems would be developed that could operate fully automatically with no human intervention. This turned out to be much too optimistic a view of what was possible. Similarly, fears that MT software would put human translators out of business turned out to be unfounded. Human translators now use MT programs to increase their



productivity. In computer-assisted translation, a translator can highlight a word in a text and be shown a list of possible translations from an online dictionary. An MT program might produce a rough draft that a translator can clean up more quickly. Or a person who pulls up a web page in an unfamiliar language may at least get a sense of what the page is about, even if many of the details are wrong. MT is most helpful processing texts that deal with very narrow subject matter (the weather, say, or scheduling meetings), formulaic sentence structure and a limited vocabulary. In general, the more complex and less standard the input language, the less accurate the output translation will be. This means that poetry, newspaper headlines, literary criticism, and other abstruse academic writing are not good candidates for MT. And of course, the reality of “garbage in, garbage out” holds here too—if there are spelling or grammatical errors in the source text, they will adversely affect the translation. First, some terms: the human language (also referred to as a natural language, as opposed to, say, a programming language) we wish to translate from is called the source language. The language we want to translate the source language into is called the target language. So when a product translates English into Japanese, English is the source language and Japanese is the target language. There is more than one approach to machine translation. The approach I have used in my work is rule-based translation. It requires intensive human labor to produce linguistic rules and data that are used in the translation process: online dictionaries containing very detailed information about the behavior of words in the source language; a robust grammar of the source language; and conversion rules that move parts of the source sentence around to reflect the correct order of these parts in the target language. It might seem inevitable that machine translation be rule based, but that is not the case. In recent years other approaches have been developed which do not require the codification of human knowledge about the source and target languages. For example, the translation-by-example approach works by gathering very large bodies of text (called corpora, the plural form of corpus) of the source language with corresponding human translations of the same corpora in the target language, aligned so that one knows which parts of the target text correspond to sections of the source text. So, for example, the English phrase “with just over eight weeks remaining until Election Day” would be stored with its humanly produced Spanish translation in a database. This source-target pair can then be used when the source phrase is encountered in a new text to be translated. Allowances can be made for inexact or “fuzzy matching” of text, so that “with just over one month remaining until Election Day” is also dealt with correctly. Another MT method that has received a great deal of attention in recent years is statistical machine translation, which uses mathematical formulas derived from enormous bilingual corpora. Both translation-by-example and statistical MT avoid the great amount of time and expense involved in codifying so many facts about human language in dictionaries and grammars. The goal of this chapter is to give you an understanding of why MT is so difficult. The chapter will describe a rule-based system, because this is the tradition I have worked in and because it gives a clear idea of the linguistic phenomena that cause a machine so much trouble with the translation task. The basic components of a rule-based MT program are: (1) a machine-readable dictionary; (2) a grammar that contains rules for recognizing well-formed sentences in the source language; (3) a computer program called a parser, which applies the grammar rules to the input sentence and produces a parse tree of its structure, performing something like the lost art of sentence diagramming that junior high school students used to learn decades ago; (4) conversion rules to convert the parse tree of the source

Machine Translation


language to the grammatical structure of the target language; and (5) experts, which are used to rank the possible parse trees of a sentence so that the most likely one can be presented to the user. Using all these tools, the basic process in rule-based MT is the following. The MT program must read each sentence of the source text and for each sentence do the following: 1 2 3 4 5 Look up each word in the online dictionary so that all information stored in the dictionary about each word is available. Parse the sentence: analyze the syntactic structure of an input sentence in the source text, labeling the subject, verb, object, and so forth, and make a diagram of the structure, called a parse tree. Rearrange portions of the parse tree so that it conforms to the format of the target language. Replace the source language words and phrases in the tree with target language words and phrases. Output the sentence in the target language.

So, given the sentence: “The dog bit him,” a human might look for the phrases in the sentence (its constituents) and make the following labeled bracketing:
[S [NP The dog] [VP [V bit] [NP him] ] ]

This bracketing indicates that the sentence S consists of a noun phrase (NP) subject “the dog,” followed by a verb phrase (VP) which consists of the verb “bit” and the direct object NP “him.” A parser might represent this labeled bracketing as the following parse tree (although it doesn’t look very much like a tree):

ART = article as in “the” and PRN = pronoun as in “him.” Next, this tree structure might need to be rearranged to conform to the grammar rules of the target language. Italian, for example, puts object pronouns (“him” in this sentence) before the verb (“The dog him bit”), not after it. So for Italian, we would need to move the object pronoun so that the tree looks like this:



And finally, we would have to consult the dictionary to find the Italian equivalents of the input words and insert them in the tree:

The output of this sentence would be “Il cane lo ha morso.” In the following sections, we will examine these steps in greater detail in order to help you appreciate the complexity of MT.

Analyzing the Input Sentence: The Problem of Ambiguity
One of the fundamental problems in MT is ambiguity: words, phrases, sentences, and even punctuation can have more than one interpretation. Writers and readers of a text share an enormous amount of information about how the world works, what kinds of things people would want to convey, and why. A computer program has access to very little of this kind of information—only the information that a human is able to encode for it—and so it has to make lots of guesses about meaning. And these guesses often turn out to be wrong. The next few subsections describe the problems ambiguity causes in determining what a sentence is, what a word is, and how a sentence is parsed.

Finding Sentence Boundaries
Rule-based MT works on a sentence-by-sentence basis. So the first task in the MT process is to locate sentence boundaries in the source text, to determine where one sentence ends and the next one begins. Most descriptions of rule-based MT don’t acknowledge this task as the thorny problem it really is. Capitalization can signal where a sentence might begin and punctuation (a period, question mark, or exclamation point) should tell you where a sentence ends, but complications inevitably arise. Take the periods that occur after abbreviations. “St.” as an abbreviation for “Street” could conceivably occur at the end of a sentence, and the period in it could also indicate where the sentence ends. But “St.” as an abbreviation for “Saint” would probably not occur at the end of a sentence.

Exercise 1

Consider the following text: I think it is Evergreen Dr. Smith has considered carefully. As you read this text, how do you parse it? That is, what phrases do you think make up this sentence and how are they related to each other? Is this text one sentence or two? Why? Would you have parsed it differently if it had read “Mike” instead of “Smith”?

Machine Translation


Finding Word Boundaries
Once sentence boundaries are found, the text must be segmented into words. English inserts spaces between words, but not all languages do so. Chinese and Japanese, for example, present sentences as a string of characters with no spaces in between. The following are some Japanese sentences that can be segmented in more than one way:

which can be segmented as:
(a) Sensei-ga tyoo-no yooni mau. (“The teacher is dancing like a butterfly.”)

(b) Sensei, gatyoo-no yooni mau. (“The teacher is dancing like a goose.”)

Another example is:

which can be segmented as:
(a) Boku-wa daibu tukat-ta. (“I spent a lot (of money).”)

(b) Boku-wa daibutu kat-ta. (“I bought a huge statue of Buddha.”)

We can see the same problem arising in English if we remove the spaces between words.

This string could be segmented in the following ways:
(a) (b) (c) (d) Therapists can send it. Therapist scans end it. The rapists cans end it. The rapist scan send it.

Not all of these are meaningful or grammatical sentences, and that is precisely the problem for MT. MT is made more difficult by the fact that language is constantly changing. In the past decade or so, no doubt in part due to the influence of text messaging and instant messaging, the trend in English has been to loosen the rules of punctuation and capitalization. It’s become quite the fashion for business and product names to start with a lowercase letter (e.g. iPod, eBay) or even to eschew upper case letters altogether. The old clues to sentence and word boundaries have become less reliable.



Looking up Words in the Dictionary: Dealing with Lexical Ambiguity
The lexicon of a language is the words that make it up, and each word is a lexical item. Words that look the same but have different meanings are called homonyms and are the source of lexical ambiguity. Homonyms can be different parts of speech (“record” as a verb vs. “record” as a noun) or one part of speech with different meanings (“switch” as a noun to mean a change, a light switch, or a thin stick). In order to distinguish among homonyms of a word, the MT dictionary must contain far more information about each word than any print dictionary would. A robust MT system might include the following information about a noun: whether it can be a count noun, which is preceded by “a,” or a mass noun, which can appear alone (She built a fire vs. She replied with fire); whether it can appear as a subject, an object or both (book vs. him); whether it takes a singular or plural verb (boy takes singular, fish takes either); whether it can appear as an attributive noun (in front of another noun, as in “home decorating,” “cookie sale”) or not (*“tires department”);1 whether it is a common noun or a proper noun (boy vs. Fred); and whether it is typically followed by a certain preposition (an attack on his views). All this information guides the parser in finding the correct tree structure. For example, knowing that tires cannot be an attributive noun leads the parser to choose (a) over (b) as the intended parse of the following sentence: Even the standard tires can cost too much.
(a) [Even the standard tires] [can cost too much] (b) * [Even the standard tires can] [cost too much]

The Importance of Multiword Phrases
A good rule-based MT system also needs to have many phrases in its dictionary. It is impossible to achieve a decent translation if the source sentence is considered on a word-by-word basis. This is because when the words making up a phrase each have multiple meanings (and therefore translations), the possibility of an incorrect translation of the entire phrase is great. So, for example, take the sentence: “The troops retreated under a hail of fire.” If our dictionary has two translations for “hail” and six translations for “fire,” there are twelve possible outputs for “hail of fire.” One combination would lead to the embarrassing interpretation: The troops retreated under hard frozen raindrops of passion. “Hail of fire” is a rather common phrase: googling2 it produces many thousands of hits. It is better to include it in the dictionary with its own translation. An idiom is a phrase whose meaning can’t be derived from the meanings of the words that make it up. Idioms also need to be included in the dictionary so that they are not translated word by word; consider the idioms “to keep one’s eye on the ball”; “the real McCoy”; “the buck stops here”; “in hot water”; “to carry the day.” Word-by-word translation of idioms often gives nonsensical, humorous or embarrassing results.

Complete Descriptions of Words Reduce Ambiguity
An MT dictionary needs to describe the behavior of words more fully than print dictionaries do. Lexicographers categorize words by their part of speech: noun, verb, adjective, adverb, article, conjunction, preposition. A print dictionary might tell you that “late” can be an adjective with several meanings, among them “tardy” and “deceased.” We need to know more to assure the correct translations for “Her husband is late” and “her late husband was a writer.” In the first sentence, the adjective appears after the verb

Machine Translation


“to be.” In the second sentence, it appears before the noun “husband” as a modifier. In most dialects of English, the “deceased” meaning can’t appear after the verb “to be.” This information needs to be recorded in the dictionary. And consider these meanings of the adjective “mad,” which depend upon what follows “mad”:
I’m mad about your brother. (I like him a lot—“mad” is followed by a prepositional phrase that starts with “about”) I am mad at your brother. (I’m angry at him—followed by a prepositional phrase that starts with “at”). Your brother was mad to react the way he did. (he was foolish—followed by a toinfinitive clause)

Verbs also need to be fully described in the dictionary. Verbs describe actions or states. Intransitive verbs take no object: He died; she sneezed. Transitive verbs take an object: We ate cookies; I visited France. English uses many phrasal verbs, which consist of a verb followed by particle words like “out,” “in” or “over,” where other languages might use a single word. The meaning of the intransitive verb “turn,” for example, depends on the preposition that follows it:
She turned against her old friend. The car turned down the road. The crowd turned on the mayor in anger. The coach turns into a pumpkin at midnight. Please turn to page 16.

It is not enough to simply enter “turn” in the dictionary as an intransitive verb. Separate entries are needed for each verb-particle pair.

Using Domain Information
Another feature that improves translation is the tagging of homonyms for the field of study or knowledge (the semantic domain) they usually appear in. This increases the chance that the correct meaning will be chosen. For example, current as a noun can mean a flow of electricity or a flow of water. These two meanings would be translated differently in most languages. The “flow of electricity” version of “current” might be tagged for electronics, engineering, and other technical domains so that in a technical text it would be chosen instead of the “flow of water” meaning. When MT is used for a specialized field, say, translating auto repair manuals or scientific reports, additional technical dictionaries are necessary to translate terms specific to that domain. Specialized terms tend to have one meaning only.

Parsing the Input Sentence
Remember that we are still working with our source language sentence, putting it into a form that can be used to create a translation in the target language. Once the parser has extracted a sentence from the source text, it applies a set of grammatical rules in order to determine how its words can be combined into phrases and what the grammatical relationships between these phrases might be. Consider our sample sentence from above: “The dog bit him.” To create the parse tree for this, the parser would need to make use of the following grammar rules:



S → NP + VP NP → ART + N NP → PRN VP → VT + NP

The first rule tells you that a sentence S consists of a noun phrase (NP) plus a verb phrase (VP). This rule would account for many of the input sentences a parser would encounter, since English is an S–V–O language (sentences generally appear in Subject– Verb–Object order).3 The second rule tells you that an NP consists of an article (ART, for example “the”) and a noun (N). The third rule tells you that an NP can consist just of a pronoun (PRN, for example “him”). The fourth rule tells you that a VP consists of a transitive verb (VT) and an NP. If these rules are to apply correctly, every word encountered in an input sentence must be located in the dictionary with all of its parts of speech indicated. The parser will look through our sample sentence and find an NP (“the dog”), another NP (“him”) and a VP “bit him.” It will combine the first NP with the VP to make S, a successful parse. There is usually more than one good parse to a sentence—a moderately complex input sentence might result in thousands of parse trees. Now consider the following headline which appeared after the brief 1982 Falklands War, in which Britain and Argentina fought over ownership of the Falkland Islands:
British Left Waffles on Falkland Islands. (Geocities, 2003)

Headlines are written in a telegraphic style that omits some words, leaving them vulnerable to (often humorous) misinterpretation. So let’s put it in non-telegraphic style:
The British left waffles on the Falkland Islands.

This sentence is still ambiguous—it has at least two interpretations. We can illustrate two ways a parser might analyze this sentence by two different labeled bracketings:
(a) [NP The British ] [VP [VT left ] [NP waffles] [PP on the Falkland Islands]] [VP [VI waffles ] [PP on the Falkland Islands]] (b) [NP The British left ]

We need to add a few more rules to our grammar above to get these two bracketings:
PP → Prep + NP (a prepositional phrase can consist of a preposition plus an NP) VP → VT + NP + PP (a verb phrase can consist of a transitive verb + NP + PP) VP → VI + PP (a verb phrase can consist of an intransitive verb + PP)

The dictionary will find that “left ” can be both a noun (a liberal faction, a left-hand turn), a transitive verb (“she left these cookies for you”), and an intransitive verb (“she left yesterday”), and that “waffles” can be both a noun (“we ate waffles”) and an intransitive verb (“She waffles and never makes a decision”), which can appear with or without a following “on” (“She waffled on the issue”). The parser will then apply its grammar rules to the possible paths through the sentence, constructing phrases and combining them into larger phrases, until it has grouped them all into a sentence. If it tries to find a path through the sentence with both “left” and “waffles” as nouns, it will fail to find a complete sentence because a sentence

Machine Translation


needs a verb. If, on the other hand, it assigns the verb part of speech to both “left” and “waffles,” it will fail to find a rule that combines two verbs this way. Which interpretation will an MT system use for its translation? One way an MT system deals with ambiguities such as the one in our Falklands sentence is by assigning probabilities to dictionary entries. So, if the lexicographer decides that “waffles” is more often used as a verb than a noun, the dictionary entry for “waffles” V will be given a higher probability assignment than the entry for “waffles” N. Similarly, a decision about whether “left” is more often used as a noun or verb will have to be made and codified as a probability. How do humans decide which interpretation is correct? Humans use context and their expectations about the way the world works to clear up confusion. When this sentence is spoken, prosody helps clear up the confusion: the stress and intonation of the first four words will lead to one interpretation or the other: “The BRITISH left waffles” favors the verb interpretation of “left” and the noun interpretation of “waffles,” while “The British LEFT waffles” favors the noun interpretation of “left” and the verb interpretation of “waffles.” But these auditory cues are not available in written text. When we read, we need to rely on surrounding context for disambiguation. This means considering previous and following sentences, something which most MT systems are not very good at. Humans rely on world knowledge and semantic information to make decisions about words as they listen or read. If we read this sentence in a newspaper, our knowledge about what kinds of things the news talks about would probably lead us to prefer interpretation (b) over interpretation (a). Consider some other sentences that are syntactically ambiguous:
The report that he left was incorrect. Interpretation (a): He left us an incorrect report. Interpretation (b): The news that he had departed was incorrect. Visiting relatives can be boring. Interpretation (a): When you visit relatives, you can experience boredom. (“visiting” is a verb) Interpretation (b): Relatives who are visiting you can be boring. (“visiting” is an adjective) The chicken is ready to eat. Interpretation (a): The chicken is hungry and wants to eat. Interpretation (b): The chicken is finished cooking and we can eat it.

There is nothing in the semantics of these sentences given in isolation to make us choose one interpretation over the other.
Exercise 2 In linguistics, a garden path sentence is one that causes human readers/listeners confusion. They start out parsing the sentence one way (going down one garden path), until additional words in the sentence cause them to reconsider or back up, and parse it a different way. Consider the following well-known garden path sentences (Fun With Words, 2009). Can you pinpoint where you need to back up and re-parse? Can you characterize the cause of the problem? The horse raced past the barn fell. Every woman that admires a man that paints likes Monet. The cotton clothing is made of grows in Mississippi. The old man the boat.



Using Semantic Information for Disambiguation
The ambiguity in some sentences can be cleared up by applying experts to the translation process. An expert is a subroutine of the MT program that embodies knowledge about one particular aspect of language. One indispensable expert is the semantics expert. Consider these two sentences:
(a) Fred is bright enough. (b) The lamp isn’t bright enough.

The adjective “bright” can mean, among other things, “smart” or “emitting light.” Humans know that the “smart” meaning applies to (a) and the “emitting light” meaning applies to (b) because they know that people (but not appliances) are smart and that appliances (but not people) emit light. This information is built into an MT system by means of the semantics expert. Each noun in the dictionary must be tagged with semantic features which describe the category it fits into. “Fred” might have the semantic feature +human. “Lamp” might have the semantic feature +appliance. Likewise, each adjective in the dictionary must have a list of the semantic features it prefers for the nouns it modifies. This list is called the adjective’s selectional restrictions. So the “smart” meaning of “bright” would have a list of semantic features it prefers: +human, +idea, etc. The “emitting light” meaning would have a list of semantic features it prefers: +appliance, +star, etc. The job of the semantics expert is to determine how well the selectional restrictions on the adjective match the semantic features on the nouns. Each verb in the dictionary also has selectional restrictions, which specify the semantic features it prefers on the nouns that appear as its subject and object. Let’s look at our Falklands sentences again, with the semantic features the nouns might take and the selectional restrictions the verbs might take:
(a) [NP The British ] +human (b) [NP The British left ] +human [VP [VT left ] [NP waffles] [PP on the Falkland Islands ] ] Subj: +human +food +place Obj: +physical_object [VP [VI waffles] [PP on the Falkland Islands]] Subject: +human +place Object of Prep: +idea, +plan

Exercise 3

Can you tell which labeled bracketing the parser would prefer? Why? What is the subject of each sentence? The verb? What are the selectional restrictions on the verbs (what kind of subjects and objects do they prefer?) What are the semantic features on the nouns?

The semantics expert can also help determine which meaning of a phrasal verb is intended. Take this pair of sentences:
(a) Mary turned down the street. (b) Mary turned down the heat.

On the surface, these sentences look like they have the same structure. But in fact these are two different verbs.

Machine Translation


(a) [NP Mary] [VI turned ] [PP down the street ] (b) [NP Mary ] [VT turned down] [ NP the heat]

In (a) “turn” is an intransitive verb followed by a prepositional phrase “down the street.” “Down” is a preposition here. Look at these two sentences:
(a1) Down the street Mary turned. (a2) * Mary turned the street down.

(a1) is a fine sentence. That’s because the whole prepositional phrase (the preposition “down” plus its object NP “the street”) has been moved to the front of a sentence. But (a2) is not grammatical. That’s because the preposition “down” has been moved rightward away from its object “the street.” You can’t break up a prepositional phrase this way in English. In (b) “turn down” is a transitive phrasal verb which takes a noun phrase object “the heat.” Here, “down” is a particle (a little word that appears after the main word in a phrasal verb), and particles can move rightward. Consider these sentences:
(b1) * Down the heat Mary turned. (b2) Mary turned the heat down.

(b1) is ungrammatical because the particle “down” has been moved along with the object “the heat.” However “down the heat” is not a phrase in this sentence. But (b2) is fine because only the particle has been moved rightward. The verb “turn” in (a) above takes subjects that are +human and +vehicle and the preposition “down” takes noun phrase objects that are +road. The verb “turn down” in (b) above takes +human subjects (but not vehicles) and objects that refer to appliances and systems (e.g. heat, lights, volume).
Consider the following sentences. (a) John ran up a big bill. (b) John ran up a big hill. What is the verb for each sentence? Determine whether the verb in each sentence is intransitive or transitive. Does the word “up” function the same way in each sentence? Hint: in one sentence “up” is a particle and in the other it’s a preposition. Try moving one or more words in the sentence to figure out its structure. Can semantics help to distinguish these two meanings?

Exercise 4

Converting the Parsed Input Sentence to the Target Language
So far we have focused on difficulties in determining the structure and meaning of the source text. But once we have analyzed the source text, we are only halfway through our task. The second part of the task is using the analysis of the source sentence to create a translation in the target language. This part involves assigning target translations to the words in the source sentence and converting the source structure to a target structure. Both these steps present challenges.



Challenges Translating Words
One problem that exists at the word level is the fact that cultures vary in their everyday practices, and languages divide up the world differently. There is not a one-to-one mapping between the words in one language and the words in another. So, for example, take the problem of translating the following from English to Italian:
She is working in the garden.

The English word “garden” can refer to either a vegetable garden or a flower garden. In Italian, there are two separate words, orto and giardino, respectively. In an Italianto-English dictionary, we could provide the compound “vegetable garden” as a translation for orto and “flower garden” for giardino but going in the other direction is more problematic. How would we translate “garden” into Italian? We would need to provide two translations in the English-to-Italian dictionary and then either arbitrarily give one translation a greater probability so that it appeared as first choice, or present both in brackets, assuming that a human translator will make the decision during a post-editing step. Another example of lexical mismatch is the difference between English and Japanese in the use of transitive verbs indicating “to wear,” as in “She wore a red dress.” English uses a single verb for all kinds of clothing; Japanese uses different verbs depending on the article of clothing worn. The verb ki-ru is used for “wear a dress,” kabur-u is used for “wear a hat,” hak-u is used for “wear shoes,” etc. The dictionary needs to have several entries for “wear”, one for each category of clothing. Each entry needs distinct selectional restrictions on the object of the verb and a distinct Japanese translation. But the following example is trickier:
Susan bought a new red dress. The next day she wore it to a party.

Determining the referent to “it” so that the proper translation of “wore” can be inserted requires looking back to the previous sentence in the text, something that most MT systems have not attempted. Lexical holes, where the target language has no word at all for a word in the source language, are another conversion challenge. Lexical holes are frequent in language pairs that represent very different cultures. Differences in the way people make their living, their religions, their sports, their kinship systems, and their food will all result in lexical holes. Here, the lexicographer will have to use a loan word, a word borrowed from another language, or make up a word or phrase that translates the source word as best as possible (a neologism). Terms like “sticker shock,” “friend with benefits,” “meme,” “smoothie,” and “presidential primary” are common English terms that might have no equivalent term in many other languages. The term klûnen in the Frysian language (spoken in parts of the Netherlands and Germany) has no equivalent in English—it describes the action of walking on skates over boards or carpeted stretches of road around obstacles or patches of weak ice (Diekema, 2003, p. 22).

Structural Challenges
Beyond the difficulty of choosing the right translation of a single word, differences in syntax between source and target languages also cause problems. Mismatches between languages can be structural as well as lexical. For example, English requires an explicit subject in declarative sentences. Italian does not:

Machine Translation


Italian: English, literally:

Ha comprato quella casa. Has bought that house.

The Italian auxiliary verb “ha” above indicates the third person singular, which could mean that the understood subject is “he,” “she,” or “it.” Context tells you which is meant, but as stated earlier, MT programs cannot extrapolate from context. The solution is to pick one pronoun arbitrarily, require the user to pick one in post-editing, or provide a feature in the user interface that allows you to pick one pronoun form as the default. A human editor will have to clean up mistakes later. Or the source language might use one type of phrase where the target uses another. Take this English–French pair:
English: She ran into the room. French: Elle entra dans la salle en courant. (literally, She entered in the room while running.) (Arnold et al., 1993)

English uses a manner verb (“run”) plus a directional prepositional phrase (“into the room”) to describe entering a room by running. French uses a directional verb (entrer) plus an adverbial phrase that describes the manner of movement (en courant). Somehow this information has to be built into the conversion part of the MT program. Or consider this English–Italian pair:
English: My arms hurt. Italian: Mi fanno male le braccia. (literally, Me make-hurt the arms.)

Italian expresses this with the direct object (“mi”) in sentence-initial position and the subject (“le braccia”) in sentence-final position. The verb (“fare”) is transitive. English expresses the same thing with an intransitive verb (“hurt”). Note also that English uses a possessive adjective (“my”) before parts of the body but Italian uses the definite article “the,” another difference that must be stipulated in the conversion step. English expresses the idea of liking something as “I like X” (Subject–Verb–Object), whereas Italian and Spanish express this idea as “To me is pleasing X” (Indirect Object– Verb–Subject). So the English sentence “I like soccer” is translated like this:
Italian: Spanish: Mi piace il calcio. (literally, To me is pleasing soccer.) Me gusta el futbol. (literally, To me is pleasing soccer.)

The conversion step of MT relies on a set of conversion rules that look for structural configurations in the parse tree and move, delete, or add structure to conform to the syntax of the target language. In the case of translating the Italian above to English, the program would have to identify piace as the verb to be translated in English as “like.” The indirect object mi (to me) must be converted to the subject pronoun “I.” And the subject el futbol in the Spanish sentence must appear as the object of “like.”

I hope that the preceding discussion has given the reader an appreciation for the ambiguity and complexity of human language that make automatic translation so difficult. The amount of lexical, syntactic, semantic and world knowledge that go into understanding even simple sentences in the source language, compounded by the difficulty of converting that information into the structures of the target language, present a daunting challenge to automation.



There have been attempts to improve MT performance by controlling the input to MT programs, for example by instructing writers of industrial manuals to use a sublanguage with simple sentence structures and a controlled vocabulary. When the semantic domain of a text is very limited, controlling input is possible. In Canada, for example, machine translation has been used for decades to produce English–French weather reports with great success. But when the domain of discourse is unlimited, say in newspapers or blogs concerning everyday life, restriction is not possible. The Internet gives you access to a number of free MT programs. Type the keywords “online,” “translation,” and “free” in your search engine to find them. Usually these free tools allow you to translate only a small amount of text at a time, but they are interesting to play around with. Try typing in some of the sentences in this chapter. Choose English as your source language and, if you speak a foreign language, choose that as your target language. Google has a good online translator available by clicking on Language Tools in iGoogle. Their FAQ page says they use a statistical approach to MT, which of course makes sense given their access to enormous amounts of text. Try pasting the first paragraph of a story from your favorite online newspaper into the translation window and see how it does. Newspaper language seems to work especially well with this statistical approach. It is also interesting to translate from source language to target language and then back again to source language. This gives you an idea of how well the translation works when you don’t speak a foreign language. For example, when you type in “My arms hurt” in iGoogle’s translation window and ask to translate from English to Italian, you get “Mie braccia male.” When you translate “Mie braccia male” from Italian to English, you get “My arms evil.” Try our sentence “The British left waffles on the Falklands.” English to Spanish and back to English gives “The British left wafers on the Falklands.” Using this bidirectional approach, compare the output of different language pairs. Pairs like English–Spanish, English–Italian, and Spanish–Italian often work relatively well because the languages in the pair are well studied, relatively close in structure, and have a long MT history. Language pairs that are linguistically very different, that haven’t had much MT development, or for which there are not very large aligned corpora don’t fare as well. English to French to English, for example: “She turned down the job and then she turned down the alley” gives “She turned down the job and then she rejected the alley.” English to Bulgarian to English gives “It turned down for a job and then she turned to determine the alley.” The next time machine translation output makes you smile, hoot with laughter, or shake your head in consternation, stop and try to understand the factors that led to the confusion. In a shrinking world, where written text in hundreds of languages is available with a few keystrokes, an MT program that can convey even the topic of a foreign web page is useful. And I hope that by now if you translate a page and the result is comprehensible, if broken, you have the knowledge to appreciate the enormous amount of work and resources that went into the task.

1 In the field of linguistics, an asterisk (*) before a sentence or phrase means that it is ungrammatical or unacceptable in some other way and would not be used by an expert speaker or writer of the language. 2 “Googling” is a good example of a word that didn’t exist ten years ago but now is common as a noun and verb. Times of political, economic, medical, and technological change create new vocabulary that must be added to the MT dictionary. Language is constantly changing, nowhere faster than in its vocabulary.

Machine Translation


3 English is a fixed word order language: most sentences appear in Subject–Verb–Object order. If we change this order, we need to mark the change in some way. For example, in passive sentences like “Mary was bitten by him,” the auxiliary verb “was,” past participle “bitten,” and change from “he” to “him” mark the fact that the subject and object have switched positions. S–V–O is the most common word order across the world’s languages but all other possible orders exist. Japanese, for example, is S–O–V and Hawaian is V–S– O. Some languages, like Finnish, have free word order, meaning that subjects, objects, and verbs may appear in any order; case markers (endings added to nouns) are used to make clear which word is functioning as the subject and which the object.

Arnold, D. J., Balkan, L., Meijer, S., Humphreys, R. L., and Sadler, L. (1993) Machine translation: an introductory guide. London, Blackwells–NCC. Available from: http:// [Accessed June 12, 2009]. Diekema, A. R. (2003) Translation events in cross-language information retrieval: Lexical ambiguity, lexical holes, vocabulary mismatch, and correct translations. Syracuse, Syracuse University Press. Fun With Words. (2009) Ambiguity and garden path sentences [Internet]. Available from: [Accessed October 31, 2008]. Geocities. (2003) Newspaper headlines [Internet]. Available from: http://www.geocities .com/Principessa07/Headlines.htm [Accessed October 31, 2008]. LogoVista. (2009) LogoVista Translation [Internet]. Available from: http://www.logovista [Accessed June 12, 2009).

Other Resources
For the reader who would like more detail about MT, check out the online book Machine translation: an introductory guide (Arnold et al., 1993) and also Raskin, V. (1997) Linguistics and machine translation. In Dallin D. Oaks ed. Linguistics at work: A reader of applications. Boston, Heinle and Heinle, pp. 634–661. Two of the best English grammar books are each over 1,700 pages long and commensurately expensive. But they give an overview of English, descriptively written, that is breathtaking in its completeness and will help you appreciate the scope of the MT challenge: Quirk, R., Greenbaum, S., Leech, G., and Svartvik, J. (1985) A comprehensive grammar of the English language. London, Longman. Huddleston, R. D. and Pullum, G. K. (2002) The Cambridge grammar of the English language. Cambridge, Cambridge University Press.

If you are interested in exploring the semantic relationships among words, check out the WordNet program, developed at Princeton University. You can download WordNet at and read an overview of it in Wikipedia.


Decker Mark Davies at Brigham Young University has developed a wonderful resource, the Corpus of Contemporary American English. It contains over 385 million words of English from different genres and is continually growing. Its excellent user interface allows you to explore how words and phrases are used in real text. It’s free at And finally, a comprehensive online resource for those interested in finding out what’s going on in the field of linguistics, from conferences to academic programs, publications, online resources, and tools is at


The N-Word, the F-Word, and All that Jazz Race, Sex, and Transgressive Language in Contemporary American Literature and Popular Culture

Carmen Gillespie
This Chapter Explores:
Transgressive Language Expletives Pejorative Language Racialized Language

We die. That may be the meaning of our lives. But we do language. That may be the measure of our lives. Toni Morrison, Nobel Acceptance Speech, 1993; in Morrison, 1994

I am a professor of English and creative writing. Most often I teach courses that explore literary texts. Language usage and meaning, particularly as it is connected with issues of race, identity, class, and gender, are central to the ways in which I approach these classes and examine the works of literature. I am interested in the ways that language works to reinforce our sense of personal identity, and this chapter is an attempt to illustrate the pervasive power of language in identity constructions in popular culture and literature. This chapter also examines the ways that artists, particularly Toni Morrison, endeavor to subvert traditional definitions of language in a quest to dismantle the social hierarchies that proliferate as the result of language designations that are static and bound to preconceived, often erroneous, concepts of race, gender, and other social categories.

This chapter explores the following questions: What does it mean to identify oneself (or to have one’s self identified) racially? What happens to identity when a person thinks about him or herself without race as a marker? How does racial categorization affect



interpersonal relationships? What is the impact of disparities in social/economic/ educational opportunities? What are some of the ways artists confront these questions? Is there a creative way to redefine language so that people are empowered rather than confined by its effects? Using examples from popular culture and literature, this chapter explores the ways that language reinforces the boundaries created by the human tendency to create social hierarchies rooted in racial difference. The explicit use of obscene language in this chapter—language that has been defined alternately as profane, obscene, and/or transgressive—is an exploration of the connections between and among language, race, and social hierarchy. The discussion of this category of language briefly outlines the historical and contemporary uses of such language and attempts to unmask the significance of taboo language and its associations with race in both popular culture and literature.
Exercise 1 In a small group consider the following questions. Characterize and take notes on your discussion and be able to share the outline of your conversation with the class. How do you define obscene language? Do you use language that is defined as obscene? Why or why not? Why do you think that certain words are considered obscene and others are not? How is obscene language used in American culture? By whom and to what effect is obscene language used? What would happen if these words were forbidden?

During the summer of 2007, an animated video called “Read a Book” (YouTube, 2009) premiered on the BET (Black Entertainment Television) cable show The 5ive. The video is an animation of a song by Bomani “D’Mite” Armah, an African American musician (Armah, 2009). The song and accompanying video serve as useful examples of the ways in which the languages of US popular cultures are encoded with racial markers. The song and video “Read a Book,” according to Armah (YouTube, 2009), are a satirical commentary on the current state of hip hop music and of African American communities. The release of the video/song on BET sparked protest from a wide constituency, including activist Jesse Jackson (Black Voices, 2009). The protestors maintained that, despite Armah’s (2009) assertions that the video/song was intended as satirical, the language and images contained in the video were inherently racist. For example, in his conversation about the video on his radio show “Keep Hope Alive,” Jesse Jackson maintained that “language should lift us [African American communities] up not bring us down” (Black Voices, 2009). Jackson’s statement imbues language with the power to enact transformation in the social conditions and positions of African America and implies that the language of the video/song has the power to affect the status of African American communities. A closer examination of the video/song demonstrates the power of the associations between language and race Jackson and others found objectionable. Of particular interest for the purposes of this discussion is the beginning of the video/ song when the narrator states: “I usually write songs with hooks and concepts and shit./ Well fuck that shit. I am about to go blacker. / Read a book. Read a book. Read a motherfucking book” (YouTube, 2009). The central claim of the song/video is that African Americans should be reading instead of indulging in other problematic and self-destructive behaviors. There is an attempt to convey this message by using “authentic” black vernacular, so defined by the song’s use of particularized and stereotypical

The N-Word, the F-Word, and All that Jazz


language. The narrator’s proclamations throughout the song reinforce the notion that there are words, specifically certain expletives, that are explicitly associated with popular definitions of blackness and that the use of words like fuck and nigger are the equivalent of enacting or performing blackness. The video/song in its seeming desire to be “real” (i.e. authentically black) is replete with its insistent repetition of the words fuck and nigger. The refrain “Read a book nigger/ a fucking book nigger/ read a book nigger/ read a motherfucking book,” coupled with the stereotypical images in the video of African Americans as unkempt, lazy, and irresponsible, is but one example of the ways that the language of US popular culture becomes layered with multiple significances, many of them with specific racial connotations. In other words, American popular culture employs language that encodes particular racialized associative links.
Go to YouTube and type in “Read a Book.” Watch the video by Bomani “D’Mite” Armah (YouTube, 2009) and then respond in writing to the characterization of African Americans that you find in the video. What language is used to describe black people? Bring your observations to class and be prepared to support your findings with specific evidence in a group discussion.

Exercise 2

The N-Word and the F-Word: A Brief History
The words nigger and fuck are two of the most explosive and catalytic words in the English language. Because of their inherent volatility, analytical conversations about the words and their social, cultural, and linguistic functions are too often silenced or avoided. An investigation of the histories of, and connections between, these two words can serve as a useful discussion about the racial coding of language and is one of the purposes of this chapter. In his book Nigger: The Strange Career of a Troublesome Word (2002), Randall Kennedy traces the United States’ history of the word nigger from its earliest uses— recording the arrival of the first Africans in Virginia in 1619 and the word’s frequent employ in early legal documents—through its applications in the twenty-first century. According to Kennedy (2002), the word originally was not intended as pejorative, but descriptive, as a way of indicating geographical origin and phenotypical differentiation. As Africans in America gradually lost equal legal status from the colonial period, through emancipation, and into the Civil Rights movement of the 1960s, the word eroded into a term of denigration and derision. In the contemporary era, Kennedy (2002) asserts that there have been various attempts to diffuse the power, hurt, and historical weight of the word through reappropriation—reclaiming and redefining the word. During the Black Arts movement of the 1960s, for example, some African American writers used the word nigger deliberately in their writing in an attempt to repossess it as a positive term—one indicating solidarity and a non-assimilationist political sensibility. A similar, although less calculated and focused phenomenon occurred with the rise of hip hop culture and rap music, a world within which the word nigger is often and casually employed as a form of fraternal address. This usage is even claimed and utilized by some white consumers and white purveyors of hip hop’s cultural products. The word nigger has strong ties to another culturally proscribed word, fuck. Before engaging in a discussion of those connections, it is essential to unpack some of the complex semantic and symbolic significances of the words. In an article entitled “Four-Letter Threats to Authority,” David L. Paletz and William F. Harris assert that:


Gillespie Using the four-letter word ‘fuck,’ the following scheme of distinction appears: . . . a) the word fuck was described in a letter to the Duke University Publications Board during its 1971 obscenity controversy as ‘the most beautiful word in the English language,’ for it (realistically) depicted the intimate physical act of love . . . b) the epithet ‘fuck you’ may be taken as quite a severe negative exhortation . . . c) the word comes to mean more, something beyond the physical act of the incidental verbal harassment. (Paletz and Harris, 1975: 973)

As these definitions reveal, the word fuck may receive such scorn in English and may be thought of as perhaps the worst word in the language (as well as in other languages) as a result of its inherently paradoxical significances. The word simultaneously represents the ultimate human act of intimacy and creation, and denotes dismissal, derision, and perhaps even destruction.
Exercise 3 Consider the definitions of the word fuck as presented by Paletz and Harris (1975). Write a brief paragraph on your reaction to the various definitions of the word as outlined in the excerpt, and then write your own definition based on your understanding and experiences. Also respond to the assertion that the power of the word lies in its conflation of the experience of sexuality with violence and denigration. Share these reflections with the class.

Although the associative/symbolic universe of the word fuck is clearly not limited to its connections with African Americans, the specific histories of African Americans provide a rich mine whose revelations affirm Paletz and Harris’ (1975) assertions regarding the symbolic power and unexpected cultural resonances of the word in the American context. When Africans first began to arrive in the Western Hemisphere as indentured servants and, eventually, as slaves, one of the tensions created by their presence was the irreconcilable distinction between public and legal claims justifying slavery on the grounds that blacks were subhuman and the reality of the voluntary and mostly involuntary sexuality that occurred on a massive scale when and wherever blacks were enslaved. As such, the historical realities of the African American situation and cultural status enable the word fuck to serve as a signifier of those myriad historical complexities—in other words, as the representation of the concept. (In theoretical terms, the word signifier generally refers to the symbolic representation of a concept or idea. For example, a common brand of computer and a red fruit are both signified by the sign apple.) The word fuck evokes both the physical violations that were an inherent part of slavery and the consequences of the traditional position of African Americans at the bottom of the social hierarchy. As Hortense J. Spillers has maintained in her ground-breaking article “Mama’s Baby, Papa’s Maybe: An American Grammar Book,”
In effect, under conditions of captivity, the offspring of the female does not ‘belong’ to the Mother, nor is s/he related to the ‘owner,’ though the latter ‘possesses’ it, and in the African-American instance often fathered it, and, as often, without any benefit of patrimony . . . the child does become the man/woman on the boundary, the one whose human and familial status, by the very nature of the case, had yet to be defined. (Spillers, 1987: 74)

In other words, the African American situation, as a result of its social and legal ambivalence, became the embodiment of both of the definitional poles of the word fuck, in its reproductive and dehumanizing connotations.

The N-Word, the F-Word, and All that Jazz


From this theoretical linking of the word fuck with the historical particularity of the African American situation emanates an immense associative/symbolic constellation that can be found in both popular culture and literature. There was a striking example of the power of these racialized linguistic associations during the summer of 2006 when studio executives were promoting the cult film Snakes on a Plane prior to its release. Actor Samuel L. Jackson was cast in the film in a role familiar to him and to Hollywood films generally—as the “bad nigger”—a stereotype he most famously enacted in the film Pulp Fiction and for which he was nominated for an Academy Award. During the promotion for the film Snakes on a Plane, the Internet buzz generated a false rumor that one of Jackson’s lines would be “I have had it with these motherfucking snakes on this motherfucking plane.” Because of the rumor and the expectation that Jackson would say the line, studio executives actually reshot a key scene in the movie in order to have Jackson say the line. This event is intriguing in that it speaks to the understandings of some members of the general public about the language associated with the stereotype of the “bad nigger” and demonstrates the popular culture connection between that stereotype and the associations and uses of the word fuck (NPR, 2009). In his article entitled “The Dozens: An African Heritage Theory,” Amuzie Chimezie (1976) theorizes that the term motherfucker is a linguistic survival for African Americans from West Africa. Chimezie (1976) maintains that the definition of the word as the worst insult a person can possibly utter maintains similar usage in West African and in African American cultures. Whether the semantic link between the word fuck and disparaging characterizations of African Americans results from the historical realities that place them squarely within the definitional polarities of the word, or the association emerges from an African survival or perhaps from a combination of both, there is a profound subtextual relationship between the two words in the vernacular of the United States—one that has deep historical roots and continues its insidious flowering in the present.

Beyond Censorship
In light of the problematic and volatile histories of these two words, is the solution, as many have advocated, banning the use of them altogether? I would argue that censorship is not the answer and that there are many more creative and potentially successful solutions to undermining the destructive and eviscerating impact of these words in isolation and in their problematic connotative partnerships. What would be the result if there were a way to simultaneously reveal the racialized and racist definitional histories of the words nigger and fuck and to dismantle their agency and power?
What is your response to the question above? In a small group, attempt to answer the question. As a group, ponder a way to creatively expose and defuse these powerful words. Write down some possibilities and be prepared to share them with the class.

Exercise 4

Generative approaches to this dilemma lie in artistic creation. There are artists who use their media with the intention of exploring such innovative and progressive possibilities. The two I will discuss here are Whoopi Goldberg, as a brief example from popular culture, followed by a more extensive discussion of Toni Morrison. In the spring of 2007, Whoopi Goldberg aired a new comedy special entitled The Word According to Whoopi. At the end of the special, Goldberg warns her live audience



that she is about to tell a “nigger joke” and proceeds to relay three. After the last one, Goldberg asks her audience to think about the ways in which the joke does or does not resonate with them. She asks whether the audience finds the joke amusing and, if so, why. In other words, Goldberg turns her comedy routine into a teachable moment. Rather than avoiding the pregnant, poisoned word, she acknowledges both its existence and its possible power and asks the audience to understand and to claim its meaning in their own way on their own terms, with no silencing or damning judgment upon their honest assessment; such a moment is rare in popular culture, but it is necessary in order to encourage people to reflect upon the intricacies and nuances of language and language use in concert with its context and in light of its prospective consequences.

Toni Morrison: Doing Language
This chapter begins with a quotation from American Nobel Laureate in Literature Toni Morrison. The quotation reads, “We die. That may be the meaning of our lives. But we do language. That may be the measure of our lives.” This quotation, along with the rest of Toni Morrison’s Nobel lecture, and, I would assert, Morrison’s entire literary canon, shares with the efforts of Goldberg a desire and determination to reveal an omnipresent, often determinative power of language. Throughout the speech, Morrison establishes that she thinks of language partly as a system, partly as a living thing
[over] which one has control, but mostly as agency—as an act with consequences. . . . Sexist language, racist language, theistic language—all are typical of the policing languages of mastery, and cannot, do not permit new knowledge or encourage the mutual exchange of ideas. (Morrison, 1994)

Exercise 5

Think about Morrison’s statement. Does language have agency? Is using language the equivalent of acting? On a sheet of paper, write down four examples of the way that using language becomes an act. Be sure to describe the consequences and the effects of the language. Be prepared to use your examples in a classroom discussion of this issue.

Morrison builds upon these assertions with her writing, using her novels to demonstrate the generative, reproductive, and empowering possibilities of language. She reveals these opportunities by exposing language in all of its connotative complexity and then by challenging her readers to escape from the boundaries that imposed definitions inevitably present to psychological, spiritual, and social development. She makes available to her readers the opportunity to reconceptualize language, and, therefore, to rethink their understandings and definitions of their subjective and objective worlds. Such an opportunity opens the door for new models for social interactions—for configurations that are not artificially stabilized by language connotations that are rooted in oppression and hierarchy. As the keynote speaker at Princeton University’s 1995 Race Matters conference, Toni Morrison addressed the specific ways in which race, racial constructions, and the realities of racism inform her work. Morrison expressed her desire to explore the possibilities of encouraging the emergence of a world in which race has significance in terms of the cultural and individual specifics derived from historical experience, but where race does not imply or designate hierarchy—levels of authority, inclusion, or exclusion.

The N-Word, the F-Word, and All that Jazz


She proposed that, as an author, the avenue upon which she approaches this daunting task is through the potentially liberating possibilities of language. It is the agency of language, its ability to construct and inform human reality and, in this instance, to create and sustain oppressive social phenomena like racism, that Morrison is particularly concerned with in her literature. In order to illustrate her struggle to articulate, to write in a language that possesses agency and resistance, Morrison discussed a moment that happened in the final stages of the publication of her novel Beloved (1987).
There was a moment of some significance to me that followed the publication of Beloved. It concerns the complex struggle and frustration inherent in creating figuratively logical narrative language that insists on race-specificity without race prerogative. Someone read the last sentence of Beloved as it was originally written. In fact, it was the penultimate sentence if one thinks of the last word in the book (the resurrection of the title, the character, and the epigraph) as the very last sentence. In any case the phrase “Certainly no clamor for a kiss,” which appears in the printed book, is not the one with which I had originally closed the book. My friend was startled by the change. (Morrison, 1997: 5–6)

Morrison elaborates on the word change and, of course, never reveals what the original word was that ends her Pulitzer Prize winning novel. She did, however, describe the word and the reasons why, in her estimation, it was the perfect word choice to end the novel.
My efforts were to carve away the accretions of deceit, blindness, ignorance, paralysis, and sheer malevolence embedded in raced language so that other kinds of perception were not only available but were inevitable. That is the work I thought my original last word accomplished; then I became convinced that it did not and now I am sorry that I made the change. (Morrison, 1997: 7)

Morrison illustrates the potentially transformative power of language through her focus on this unnamed word, a word that she believes has the ability to lay bare language in such a way that the constraint of racial constructions ceases to provide the foundations of its meanings—a word that forces those who encounter it to rethink assumptions and, ideally, to create new possibilities for meaning and understanding. I believe that the word Morrison alludes to is the key to unlocking and apprehending the aesthetic and political energies of her work and also presents a template for a creative examination of the issue of racially connotative disempowering language. As Morrison herself articulates, investigations into the nuances of this word have the ability to move the examiner to “emphasiz[e] racial specificity minus racist hierarchy” (1997: 8). Curiously, no scholar to date has elaborated on the implications of Morrison’s sphinxlike riddle, but, in light of Morrison’s insistence that the use of the word provides an opportunity to utilize the agency of language in order to alter the social order it supports, such scholarship is critical, not only for a full apprehension of Morrison’s canon, but also, and perhaps more significantly, for the revisioning of the American racial home and of the primary role of language in the construction and function of that home. Morrison says that the word that she edited out of the original, pre-published version of Beloved (1987) is “racially charged and figuratively coherent . . . disjunctive, a sore thumb, a jarring note combining . . . two linguistically incompatible functions—except when signaling racial exoticism” (Morrison, 1997: 8). The original word, which would indeed have stuck out like a sore thumb, I believe, is the word fuck.



It is the “two linguistically incompatible functions” of the word fuck that allow it to reverberate so arterially through Beloved (Morrison, 1997: 8). These are the two meanings of the word previously discussed that racialize the term. The word both is a vernacular reference to sex and to the concomitant possibility of the creation of life and also means to ruin or damage someone. When applied as a tool for iterating and comprehending African American histories, the word can illumine the violence of the rape of women during slavery and also the abject subordination the black race has so often experienced in American racial hierarchies; yet the word also contains the urgent need for connection and human survival and creation that can transcend the construct of race altogether. Somewhere in the fertile terrain between these oppositions lies the possibility for new understandings that are both racially specific and unbound by racial hierarchy. Unlocking Morrison’s mysterious word led me to questions about the ways that sexuality generally, and the word fuck in particular, function in Morrison’s larger canon as signifiers—potential moments of semantic revelation, moments when the polarities embodied in the word fuck are destabilized by Morrison’s deliberate use of it as a way of engendering rich, honest language that refuses to reinscribe oppressions. Here I will briefly explore the ways that Toni Morrison democratizes language by opening the possibility for all to reinterpret its meanings without the constraints of easy oppositionality, particularly with respect to race. Through her employ of the risky, taboo, and evocative word fuck, and more generally her employ of sexuality as a terrain of inherent contradictions, Morrison in her fiction provides a way “not only [to discover] the safety and freedom outside the race house, but to suggest contemporary searches and yearnings for social space that is psychically and physically safe” (Morrison, 1997: 10). Although here I will not go into a full application of this theory to Morrison’s Beloved (1987), a brief sketch of the early moments in the novel reveals the rich possibilities for understanding the novel as a guide to linguistic redefinition. Within the first dozen pages of the novel, the reader encounters two vastly different sexual experiences of the novel’s central protagonist, Sethe Garner—the first a flashback where she remembers trading sex with the town engraver in exchange for his inscribing the letters B-e-l-o-v-e-d on her dead daughter’s tombstone; the other occurs when Paul D, the novel’s other major protagonist, arrives and he and Sethe have Paul D’s much imagined sexual encounter, an intimacy that lacks the perfection of the imagined ideal, but that begins the process, for both of them, of liberation from the tyranny of the narratives of the past and freedom from the limitations of memory. These events, which happen within the first few pages of the novel, contrast the two significances of the word fuck and also reveal the word’s racial overtones. The reason Sethe has to barter her body in order to memorialize her dead baby is inseparable from her race—the infant’s death is motivated by her unwavering willingness to do anything in order to protect her child from the hideous uncertainties of race-based slavery. Even as a free African American woman, her history as an ex-slave leaves Sethe without financial resources. Her only recourse for monumentalizing her daughter lies in the commodification of her dark, female flesh, an act which reasserts the racial associations of the word fuck. In similar fashion, the intimacy shared between Paul D and Sethe upon his arrival at Sethe’s home in Cincinnati, 124 Bluestone Road, is also tainted with the legacy of slavery and the fight against dehumanization that those deprived of personhood have to negotiate. Paul D’s decades-long sexual fantasies about Sethe are rooted in his enslavement and the deprivation of his basic freedoms. While Paul D is enslaved at Sweet Home, the same farm as Sethe, he is one of five male slaves their owner, Garner, calls the Sweet Home men. Garner is fond of making a joke with his neighbors about his “nigger men”

The N-Word, the F-Word, and All that Jazz


(Morrison, 1987: 10). Garner asserts that his slaves are men, because he believes that he embodies the characteristics of true masculinity and, as such, possesses the authority to define his male slaves as men without being personally threatened by the label. Garner replies to his neighbors’ objection to his characterization of his slaves as men with the retort “if you’re a man yourself, you’ll want your niggers to be men too” (Morrison, 1987: 10). Nearly invariably, the neighbors reply, “I wouldn’t have no nigger men round my wife,” to which Garner replies “Neither would I” (p. 11). Garner’s response is perceived as an insult because the implication is that, as a man, Garner has control over his wife’s sexuality and that the man making the comment does not. Garner believes that he embodies masculine control and therefore is unthreatened when he gives his slaves freedoms his neighbors perceive as dangerous and taboo. It is important to note that Garner’s contextually oxymoronic phrase, nigger men, does not, in his mind, imbue his slaves with humanity; rather, the assertion is Garner’s way of publicizing his ability to control and dominate his universe. With this passage, Morrison once again uncovers the hierarchies, oppressions, and racisms that are embedded in language. As the only young woman at the farm where she is enslaved, ironically named Sweet Home, Sethe becomes the object of the Sweet Home men’s desire. These men are denied agency and the fundamental human ability to fulfill their basic physical and psychological needs. As a result, they end up, as Morrison writes, “fucking cows” rather than sharing intimacy with another human being, which reinforces Paul D’s doubts, not only about his manhood, but about his very humanity (Morrison, 1987: 11). Through her use of both the words nigger and fuck, and her exploration of sexuality within a racialized context, Morrison demonstrates the dehumanizing impact of linking race and sexuality in language. She does not use the words nigger or fuck gratuitously; neither does she avoid them. Rather, Morrison uses these words in her novels as occasions for exploration of the historical roots of these signifiers and as an opportunity for the reader to unearth his or her own internalized definitions of these words. As such, her use of these expletives creates the prospect for the reinvention of language—the rebuilding of the microcosmic psychological and macrocosmic social home with linguistic building blocks that are more organic, fluid, and, as a result, less susceptible to (re)creating structures rooted in stasis, calcification, and hierarchy.

All that Jazz
If, as I have suggested previously, Morrison’s novel Beloved (1987) originally ended with the phrase “No clamor for a fuck,” then the next novel in her canon, Jazz (1993), may indeed pick up where Beloved leaves off. Indeed, the word jazz has been defined in both its noun and verb form as a euphemism for the word fuck. According to the Oxford English Dictionary, one 1927 definition from the Journal of Abnormal and Social Psychology states that “The word jazz . . . Used both as a verb and as a noun to denote the sex act . . . has long been common vulgarity among Negroes in the South, and it is very likely from this usage that the term ‘jazz music’ was derived” (Johnson, 1927). Although this definition contains its own problematic insinuations, this and other entries affirm the origin of the word jazz as frequently synonymous with the word fuck. This definition is particularly pertinent to Morrison’s employ of the word in her novels, particularly in her novel Jazz (1993). Toni Morrison’s sixth novel, Jazz, is—like the music it is named after—a study of the complex blending and melding that becomes the United States. The United States is many things, but it is often conceptualized as a coming together of opposites, of the synthesis of urban and rural, black and white, rich and poor, male and female, and young and old. All of these polarities combine in a



unique formulation that creates the sound, the look, and the character of the country known as the United States—in other words, jazz. In Jazz, Toni Morrison examines this definition of America by creating characters that can provide access to the experience of what it means to live in a space that is defined by the idea of opposition. Morrison describes these particularly American coming-togethers in Jazz as an attempted union (a fuck, if you will) that is primarily characterized by cycles of yearning, movement, desire, and loss that affect the entire country as well as the individuals who call it home. The novel moves in its journey through the small orbits of the individual characters, to ever-larger circles of concern. This movement metaphorically replicates the inscription of music on a record—the “page” upon which all music, but especially jazz, would have been imprinted in years past. The reader of the novel Jazz becomes, then, like the needle on a record player, gently caressing the surface of the words in order to discern the meaning of the text as it moves slowly towards the center— all the while revealing more and more pieces, notes of the entire score of the novel. The novel’s female protagonist, Violet Trace, has a conflict between her past and her future. While she is in the fourth decade of her life, this tension is compounded by her relocation from the South to the North. Violet imagines herself as a fundamentally different woman while she is living in the country than she becomes in the city. When she lived in the country, Violet envisioned herself as someone strong and capable. During the later years of life in her new home in the city, Violet seems to undergo a kind of breakdown; the fracture of self that seems to correlate in part with her move north begins to manifest in her actions. The source of Violet’s rootless self can be found in the abandonment she experiences as a child as the result of her father’s involuntary absence and her mother’s suicide. As a woman, these losses manifest in Violet as a fissure or crack. Violet is inbetween the spaces, like the dark, ever-moving lanes between the lines on the surface of a record. As the novel moves ever closer to the center of the story, it must make the journey back to the South, back home, in order to understand the pieces of the whole. At the narrative heart of the novel is the story of the interaction of two core figures in the novel, Golden Gray and Wild. This interaction is not only the source of the crises of Jazz’s main characters, but is the fundamental contradiction and conflict the novel slowly brings the reader to consider—the coming together of blacks and whites in the formation of the United States. Jazz music is often spoken of as quintessentially American since it is an art form that seems to have originated on this continent. Jazz is also the fusion or mixing of different musical traditions—particularly European and African. Morrison’s Jazz takes on the questions and problems raised by that fusion and represents them with the characters Golden Gray and Wild. Although Golden Gray appears white, he is of mixed racial heritage—the product of an illicit, forbidden, and taboo, yet historically common, sexual union between a black man and a white slave mistress. When Golden Gray discovers this reality, it sends him into a crisis, and he angrily seeks revenge on his black father. Golden Gray, particularly in light of the way he dresses and acts, can be seen as a representation of America’s founding fathers, Jefferson, Washington, and others, who, although believing themselves to be white, in fact were descended from the same African ancestor as the rest of humanity. These founding fathers were also slave holders and during the formation of the country denied African Americans their basic human rights and yet were often sexually involved with their slaves. The female core character in Jazz, Wild, is the personification of a composite of beliefs about Africans and African Americans. She is unkempt. She is fertile. She is preverbal and illiterate. She is perceived as dangerous. She is motherless. She, as her name suggests, is the embodiment of wildness. All of these ideas have, at one time or another, been held

The N-Word, the F-Word, and All that Jazz


as common currency about black people, particularly about black women. Significantly, Wild’s description bears a striking resemblance to the character Beloved in Morrison’s novel of the same name. Since Jazz is the second in the series of three novels Morrison has called a trilogy, it follows that the character might exist in both books. In Beloved, the titular character comes to represent all of the souls lost during slavery—the soul “clamor[ing] for a fuck.” So in Jazz, that character becomes reproductive and gives birth. As the cumulative spirit of loss, Beloved/Wild embodies the impact of slavery and the omnipresence of that loss. Wild echoes and replicates for others her own experiences. She abandons, yet haunts. The encounter between Golden Gray and Wild becomes the clash between narratives and between representations of the other—the oppositions contained in the word fuck. Although the two characters are seemingly polarized, they are strangely drawn to each other. Like the violent coming together of blacks and whites in the United States throughout the brutal enterprise of slavery, Wild and Golden Gray’s encounter creates a hierarchy in which Golden Gray feels superior, even though what is most noticeable in his behavior towards Wild is his inhumanity. Golden Gray values the concept of honor over the actual expression of humanity. He demonstrates this preference by the revulsion he shows towards Wild as well as by the attention and care he lavishes on his horse rather than Wild, the injured and laboring woman. This intriguing narrative comes and goes in riffs. The reader discovers that the baby that Wild gives birth to, yet will not nurse, is the infant Joe, the novel’s primary male protagonist. The reader learns that Golden Gray’s father, Hunter’s Hunter, becomes Joe’s surrogate father. The information that sends Golden Gray on his journey to find his father also connects the narrative to Violet. Yet as there is no resolution to the conflict between the races in the United States, there is no conclusion to the story of the encounter between Golden Gray and Wild. The two figures enter into memory and mythology, and haunt both Joe and Violet as the couple move from the South to the city. The unresolved story of Wild and Golden Gray is the deeply rooted and buried tale of the United States—a story whose ambiguity and uncertainty continues to affect and haunt all Americans. The word jazz never appears in Toni Morrison’s novel Jazz, and the word fuck only occurs once, yet the novel is fundamentally an explication of the encounter between blacks and whites in the United States—a relationship, like the word fuck, that is traditionally defined by arbitrary polarities yet is in reality a union that is at the narrative heart of the country’s existence—jazz. As previously established, the word jazz has, since its origins, been associated with sexuality, temptation, and taboo. As such, it follows that Morrison’s novel should be rooted in illicit passions. Sexuality is at the center of the novel, as are other forms of desire and longing. The final passages of the novel reveal the deeply interactive structure of the story when the narrator identifies herself as the book itself and professes that, like a record, there is no life without the active and compassionate participation—intercourse, if you will—between the book and the “listening” reader.
The word nigger has a long and difficult history in the American context as well as a complex contemporary life. When you hear this word, what feelings does it evoke for you? How do you feel about its usage? Is it a word that can and should be reclaimed? Why or why not? In a one-page reaction, write an unedited response to these questions and share as a part of a larger discussion about Morrison’s use of the word and your subjective experiences of it.

Exercise 6



Throughout the history of the United States, the common language of its people has possessed a vocabulary that, like most languages, inscribes, often subtextually, social hierarchies. In the US, many of these linguistic markers of the social order are racialized. Of particular interest for the purposes of this chapter are the words fuck and nigger and their function in popular cultures as placeholders and as indicators of the pervasive and persistent marriage between language and racial subordination. This chapter attempts to explore possibilities for reinventing these explosive words in ways that not only diffuse their power and ability to stratify, but plow and unearth a fertile ground within which a new way of making meaning can begin to germinate. As representations of the possibilities inherent in this sowing, this chapter examines briefly the work of Whoopi Goldberg and, more extensively, Toni Morrison, as examples of artists whose work models the potentialities of reimagining the world by tackling the foundation of its creation—language and communication. Toni Morrison’s exploration of language in her novels is rooted in (re)inventing constructions of meaning and interaction that do not rest upon the hierarchical semantic skeletons of tradition. As the quotation from her Nobel lecture affirms, she is committed to doing language. Through her deliberate effort to collapse meaning through exploration of the connotations and parameters of sexuality, Morrison uses the racialized and taboo word fuck and, by not avoiding this social taboo, engenders the possibility of imagining other, less deterministic, oppressive, and segregated understandings of human existence. Goldberg and Morrison’s examples are instructive and create for an open-minded, engaged, and thoughtful audience the possibility of reinventing identity, interactions with others, and even the structures of human society through the exploration and contemplation of subjective understandings and, most importantly, use of language. Note that the words fuck and nigger do not appear in quotation marks in this chapter.

Armah, B. (2009) The hustle [Internet]. Available from: [Accessed June 14, 2009]. Bonami “D’Mite” Armah’s website. Black Voices (2009) Read a book artist responds to Jesse Jackson [Internet]. Available from: [Accessed June 14, 2009]. Jesse Jackson quotation about the power of language to affect the status of African American communities. Chimezie, A. (1976) The dozens: An African heritage theory. Journal of Black Studies, 6(4), pp. 401–420. Johnson, G. B. (1927) Double meaning in the popular negro blues. Journal of Abnormal and Social Psychology, 22(1), pp. 534–536. Kennedy, R. (2002) Nigger: The strange career of a troublesome word. New York, Vintage Books. Morrison, T. (1987) Beloved. New York, Plume. Morrison, T. (1993) Jazz. New York, Plume. Morrison, T. (1994) Lecture and speech of acceptance, upon the award of the Nobel Prize for literature, delivered in Stockholm on the seventh of December, nineteen hundred and ninety-three. New York, Random House. Morrison, T. (1997) Home. In Lubiano, W. ed. The house that race built. New York, Random House, pp. 3–12.

The N-Word, the F-Word, and All that Jazz


Nobel Prize (2009) Toni Morrison Nobel lecture [Internet]. Available from: [Accessed June 14, 2009]. Toni Morrison’s 1993 Nobel Lecture. NPR (2009) Samuel L. Jackson, playing the ‘champ’ [Internet]. Available from: http:// [Accessed June 14, 2009]. Samuel L. Jackson’s NPR interview. Paletz, D. L. and Harris, W. F. (1975) Four-letter threats to authority. Journal of Politics, 37(4), pp. 955–979. Spillers, H. J. (1987) Mama’s baby, papa’s maybe: An American grammar book. Diacritics, 17(2), pp. 64–81. The Word According to Whoopi, 2007. [TV program] Dir. Paul Miller. Perf. Whoopi Goldberg. Bravo Networks, April 5, 2007. YouTube (2009) Bomani “D’Mite” Armah: Read a book [Internet]. Available from: http:// [Accessed June 14, 2009]. “Read a Book” video.

Other Resources
Brodie, F. M. (1975) Thomas Jefferson: An intimate history. 4th edn. New York, Bantam Books. Gordon-Reed, A. (1997) Thomas Jefferson and Sally Hemings: An American controversy. Virginia, University of Virginia Press. Hughes, G. (1998) Swearing: A social history of foul language, oaths, and profanity in English. New York, Penguin. Lewis, J. and Onuf, P. S. (1999) Sally Hemings and Thomas Jefferson: History, memory, and civic culture. 2nd edn. Virginia, University of Virginia Press. Salzman, J., Smith, D. L., and West, C. (1996) Encyclopedia of African-American culture and history. New York, Macmillan Library Reference. Sheidlower, J. (1999) The F-word. New York, Random House Reference. Smith, P. (1976) Jefferson: A revealing biography. New York, American Heritage. Sylvester, T. L. and Knight, J. (1999) Slavery throughout history reference library. 1st edn. Farmington Hills, UXL.

Articles in Journals and Media
Ellis, J. J. (1998) When a saint becomes a sinner. U.S. News and World Report, November 9, pp. 11–12. Lord, L. (1999) The Tom-and-Sally miniseries (cont.): Rallying around the founding father. U.S. News and World Report, January 10, pp. 15–16. Marshall, E. (1999) Which Jefferson was the father? Science, January 8, pp. 153–155. Miers, P. (1999) Language and the structure of desire. MLN, 114(5), pp. 1078–1091. Randolph, L.B. (1999) The Thomas Jefferson/Sally Hemings Controversy: Did Jefferson also father children by Sally Hemings’ sister? Ebony, February 1, pp. 21–22. Slotkin, A. R. (1994) Two new “obscenities”: The acceptability of taboo words in the media. American Speech, 69(2), pp. 220–224.

Language, Power, and Identity
Part III takes us through three chapters that question how language interfaces with our sense of self, our identities. Judith A. Parker and Deborah Mahlstedt, a linguist and a social psychologist, explore how talking about sexual assault experiences is mediated in different contexts by social norms and power structures (Chapter 8). Diana Boxer (Chapter 9) ponders the importance of naming practices in marriage, specifically the trading of women’s maiden names for their husbands’ surnames (sirnames?). In Chapter 10, forensic linguist Carole E. Chaski introduces us to the world of linguistics as a forensic science. Those readers familiar with forensics as part of police and legal work will be convinced that analyzing linguistic data using scientifically defensible methods can be just as informative.


Language, Power, and Sexual Assault Women’s Voices on Rape and Social Change

Judith A. Parker and Deborah Mahlstedt
This Chapter Explores:
Language of Sexual Assault Experiences: Lexicon, Syntax, Semantics, Discourse Survivor Discourse Power Relations in Context Courtroom Discourse Empowerment and Social Change

At that time I was victimized. A survivor is not the end of the road. You become a survivor. But survivor is not the end of the road. A survivor is what you are until something better comes along and then you move to something better. But we don’t have a name for that. Young and Maguire, 2003 Connecting (smaller linguistic) routines to larger societal discourses requires that we think about how small acts ramp up into big ones. Above all, it requires thinking about how a single individual’s verbal move could get picked up by others and eventually make it into public discourse. Eckert and McConnell-Ginet, 2003

JP: As a professor, I enjoy teaching students in courses on psycholinguistics and speech processing, linguistics and text, women’s studies, and an advanced seminar in speech, narrative, and emotion. I’m committed to interdisciplinary approaches to research and teaching. Fresh ideas spring up when applying compelling questions and disciplinary practices from one area of study to another—and then, within this broadened framework, you work relentlessly to really understand it in your own terms. As a feminist, I have been committed to teaching and contributing to feminist research and scholarship, principles and values; and while I love the beauty and work of making theories, I also have a strong commitment to practical applications and to social justice, to practices that might make the world a better place. That’s what motivated me to try to understand and make the connections I could among linguistics, psychology, and women’s studies. Experiential and activist


Parker and Mahlstedt theater work brings animating, practical, and insight-bearing tools for exploring social problems into our many communities. I hope you’ll find a way to weave your own interests along with the threads that come together in this chapter. DM: I became interested in the intersection of progressive education, social psychology, and social justice issues. That interest led recently to a senior seminar I developed, the Psychology of Social Change, a class that is also informed by my commitment to feminist pedagogy and research. I also teach small group psychology, psychology of women, human sexuality, and small group facilitation. Since I believe that emotion is an important part of any learning process, my teaching is experientially based. My interest in language came out of teaching psychology of women and ways of doing research that focused on interaction. Talk is integral to human interaction! My primary area of research has been violence against women. With a male student I developed an action-research project and peer-education dating violence prevention program for college men. Two decades later the Fraternity AntiViolence Education Project is still running; we have also produced a video called Men’s Work. The question that motivates my inquiry, teaching, and activism is: What motivates people, especially as members of a dominant group, to work to end social injustice, and, specifically, what motivates a member of a fraternity to work to stop dating violence? N.B. Our primary focus in this chapter is on women’s use of language about male sexual violence against women. This is not to diminish the reality of sexual violence committed against girls, boys, and men as well as in same-sex interactions, which all clearly deserve attention.

It’s a complex issue, how women who have been sexually assaulted deal with language as they come to grips with their experiences of sexual violence. Negotiating how to refer to oneself as victim or survivor or to use both or other words is one facet of using language to describe one’s experience of sexual assault. The words a woman chooses can convey to herself and others a sense of personal empowerment: they might even be considered a form of social activism. Naming has social, cultural, and political consequences that we will explore in this chapter. At the end of the chapter you’ll find useful information and relevant resources to explore these issues further. In this chapter, you will begin to: • Develop an understanding of the language women use when talking about sexual assault. You will encounter this language in several genres: interviews, narratives, reports, and courtroom testimony. • Develop an understanding of how language is used in the larger social and cultural contexts in which sexual assault experiences are told. You will learn about the different factors that influence our use of language, such as gender, race, and hierarchies of power. • Explore how language can be a source of personal empowerment for sexual assault survivors and also challenge prevailing norms that tolerate (indeed promote) violence against women and other marginalized groups. • Apply linguistic methods to analyzing language, in order to understand how social inequities are produced and perpetuated through language and society. These topics continue to be central concerns for counselors, lawyers, academic researchers, rape crisis intervention advocates, social activists, and, of course, survivors

Language, Power, and Sexual Assault


themselves. This chapter offers an introduction to research on the language and linguistic strategies of talking about sexual assault. We take an interdisciplinary approach that draws on knowledge and applications developed in linguistics, psychology, philosophy, women’s studies, and communication studies. Let’s begin with an exchange between two close friends:
Exercise 1 Scenario: “To tell or not to tell?” Alice is talking with her best friend Jesse about something that happened to a girl down the hall and is surprised by Jesse’s reaction. Should she say anything at all about her own experience? Alice: She’s saying that the guy she was with . . . you know, the one who’s always joking around . . . did something to her last night . . . she was crying. Jesse: Like what? You mean forced himself on her? No way, he’s sweet. Alice: She said she went to his room and . . . Jesse: (friend jumps in) and did what? Tell me she sat on his bed. Alice: Well . . . I don’t know . . . she just said they were hanging out . . . laughing and talking . . . when he started kissing her. She thought he was joking around but . . . all of a sudden he’s getting all worked up . . . she said she told him to stop it. What are you doing? He said: Oh c’mon, don’t give me that . . . and just went ahead. Jesse: Well . . . sounds like he thought it was okay . . . did she try to push him away or hit him? It’s really hard to say when you go that far and then want to stop. Some girls just don’t get it. Alice: Don’t get what? She didn’t start it and did tell him to stop . . . Alice is surprised and upset by her friend’s reaction. She feels she could never tell her best friend that she too had been raped in a very similar situation. Assume that Alice has been sexually assaulted but has never told anyone and therefore this conversation is making her uneasy. She wants to tell her friend, but at the same time she’s reluctant to tell. What do you think Alice will say next? Write out the next verbal exchange between Alice and Jesse. N.B. One resource available to someone who needs support or information about sexual assault is the local or national rape crisis hotline telephone number. These telephone numbers tend to be available in the phone book or online.

Talking about sexual assault experiences takes place in a variety of contexts: with a friend or a counselor, a researcher conducting interviews, or in a courtroom answering questions from a prosecutor or defense attorney. Talking with someone supportive might be very different from talking with someone who is not. We will examine how the constraints, conventions, and social forces present in the immediate situation and the broader levels of society influence what people say about sexual assault. This analysis of language and sexual assault will begin at the level of choosing words.

The Lexicon of Sexual Assault
The lexicon, broadly speaking, is the component of grammar that consists of words and their smaller units that convey meaning. For example, the lexical item sexual assault is composed of two words but three units of meaning: sex-, -ual, and assault.


Parker and Mahlstedt

Choosing Words: Reference, Denotation and Connotation
What words are available to a person who has experienced a violent crime? Which words will she choose to describe her experience of sexual assault? Will she refer to herself as a victim of sexual assault or as a survivor of sexual assault, both, or neither? How does she integrate the meaning of available words into her identity? And how might her use of language change over time, with different people, and in different situations? Linguists examine the meaning of words in relation to other words, in relation to the sentences in which they occur, and in relation to the context of the real world in which people speak and listen to each other. For example, the words survivor and victim might refer to the same person, but the connotation of each word conveys different meaning, and therefore has significance for persons who have experienced sexual assault. In terms of semantics, survivor and victim have the same denotation when they indicate or denote the same referent (same person in this case), but they each connote or convey different social or affective significance. Recently, to make the lexicon more inclusive of sexual violence experienced by people of different genders and relationships, scholars and researchers sometimes substitute phrases like intimate partner violence for domestic violence and male violence against women. Each of these phrases has a different denotation, that is, their referents are different. The semantic relationship of the phrases domestic violence (DV) and male violence against women (and likewise the semantic relationship between intimate partner violence and male violence against women) differs from the semantic relationship of survivor and victim in terms of denotation and connotation. The most important difference between DV and IPV is that they bear different social and emotional meanings, yes, but the two phrases do not refer to exactly the same individual or group of people. For example, the phrase domestic violence is related to battering and to wife beating: all refer to male violence against women, but each with different semantic connotations and each with different syntactic structures implicating or not implicating a (male) perpetrator of the violent crime. But the gender of the perpetrator of DV or IPV remains unidentified. Thus the group of people denoted here is broader; it could include perpetrators who are males, females, children, adults, and others associated with home and family. These changes in terminology present us with a set of questions: how can we be more inclusive without becoming so general that we lose critical specific features of the phrase? If we want to denote male violence against women, then we need to specify that when we choose our words. If we don’t care to acknowledge or focus on the predominance of males as perpetrators of violence in domestic violence and intimate partner violence, then we can use the more general phrasing that doesn’t refer specifically to men as the perpetrators. Why would it be desirable to not name men as the perpetrators of violence against women? Who determines the terms used? When is it appropriate to use the more general, inclusive term? When is it problematic? If the focus of a study is actually male violence against women, why use the phrase “intimate partner violence” throughout the article? It is important not to mask the reality that most intimate partner violence is perpetrated by males against females and that certain social groups (e.g. women, people of color, LGBTQA [lesbian, gay, bisexual, transgender, questioning, and allies]) are subject to greater violence from the dominant group.

Words: Naming and the Reluctance to Name Rape, Date Rape, and Acquaintance Rape
If there’s no name for it, it’s as if the phenomenon does not exist.

Language, Power, and Sexual Assault


Exercise 2

Imagine a time when the terms “date rape” and “acquaintance rape” did not exist. What would you have called the experience of being raped by someone you were dating or knew as a friend?

In the United States, terms associated with sexual assault, such as survivor, date rape, and acquaintance rape, came into existence in the 1970s (Warshaw, 1988). Before that time, even to identify oneself as a victim of date rape was legally difficult or impossible. If a man you knew raped you, it was not considered a crime. How would a woman refer to what happened before that time? She might say, “He took advantage of me” or “He went too far.” Since then, researchers have examined the ways women use language to describe what happened to them and how they refer to themselves after a sexual assault experience. Linda A. Wood and Heather Rennie (1994) studied how women formulate what rape is and their sense of self as victim and/or survivor following rape. Analyzing the interviews of eight women who had been sexually assaulted, they found a range of responses. The reluctance to acknowledge sexual assault as date rape is illustrated in the following words of the participant named Ann:
“Something that I’m still, calling this date rape as you said it’s rape, I see rape as more violent, but maybe . . .” (Wood and Rennie, 1994)

Ann questions the use of the term date rape because she’s always understood rape to be “more violent.” This notion that rape must include excessive violence is a common yet inaccurate social construction of rape. Similarly, in a study by Stacy Young and Katheryn Maguire (2003), most of the ten women participants referred to their experiences as “having sex,” “it,” “what happened to me,” “the incident,” or some other term. Indeed, both studies found that only some women used the terms “sexual assault” or “rape.” Why? Wood and Rennie (1994) explain that it was more difficult to label the experience as a rape if the woman knew her rapist. To know the man who raped you contradicts the images most people have of a “real” rapist and a “real” rape, that is, a disturbed male stranger who violently sexually attacks a woman walking alone in an isolated, poorly lit area. This misconception persists even though research indicates that most often a woman is raped by a man she knows (Warshaw, 1988; Koss et al., 1987). While Wood and Rennie (1994) suggest that it may be a way for women to mitigate the impact of their experience of sexual assault, Young and Maguire (2003) suggest that naming the experience rape or sexual assault could be a phase in a sense-making process for some women. In addition, Wood and Rennie (1994) note that “[t]he naming of what happened is worked out and negotiated in interactions with other people. It is not constructed in a social vacuum.” Keith Bletzer and Mary Koss (2004) report on narratives of experiences of sexually assaulted women from three different social groups in the Southwest United States: Cheyenne (n = 25), Mexicana (n = 13), and Anglo (n = 24). They conducted interviews with each of the sixty-two women in the study. The researchers found differences among these three groups of women in the lexicon they used to describe their experiences of sexual assault. The Mexicana participants in this study, for instance, tended to use language that expressed that something was being done to them:


Parker and Mahlstedt “He made me his by force” (que me hizo suya a la fuerza) “against my will” (contra mi voluntad ) “He made me” (me obligó)

Bletzer and Koss explain that this characterization that something was being done to them serves to indicate that the sexual assault was nonconsensual. Seventy-two percent of the Cheyenne women used action words in their descriptions of sexual assault experiences, including “grab,” “thrown,” “push,” and “pull.”
“He grabbed me, then started taking my pants off, and then he done his thing.” “He just grabbed me, started kissing on me, and then he sort of forced himself on me, had his ejaculation and got off.” (Bletzer and Koss, 2004)

These words describe the coercive nature of the sexual assault and therefore preclude the interpretation that there was any mutuality characteristic of a willingness to have sex. The Cheyenne women participants’ life stories and accounts also included explicit articulation of their own volition, such as “I told him I did not feel like it” or “I told him I would tell my brothers, therefore nothing happened with that experience.”

Using Words to Acknowledge (or Not) Sexual Assault: Deixis, Hedging, and Euphemism
Sometimes words and phrases can also be used to avoid saying what is difficult: they might serve to either acknowledge or not the enormity of what happened to us and our readiness to cope with it. Speakers might use a variety of linguistic resources to manage their (un)willingness to acknowledge their sexual assault experience or their reluctance about saying difficult or socially transgressive taboo words. Three of these resources are deixis, hedging, and euphemism. These discursive strategies are the focus of much research that demonstrates the complex, rich nature and many purposes served when women speak about sexual assault in context (e.g. Wood and Rennie, 1994). The word deixis (pronounced dike-sis) comes from Greek and refers to language whose interpretation requires real-world information. For example, the personal possessive pronouns “my” and “your” refer to the speaker and addressee, respectively. The meaning of “today” depends on when the word “today” is uttered. In false deixis, there is no explicit referent for the pronoun “it” in the sentence “It’s just something that happened.” Another common linguistic strategy speakers use is hedging. Here a speaker modulates the impact of her own utterances by using a word or phrase that is called a hedge. By using such terms as kind of, sort of, maybe, I think and others, a speaker can moderate the strength or weakness of an utterance, and can affect the likelihood that an utterance will or won’t be taken as certainly true.
Example 1: “And you know, I saw the ad [for the study] two weeks ago, and I read it, and I’m going, ‘Well, let’s go and see this person [interviewer]. Let’s find out whether it was an actual rape.’ Something that I’m still, calling this date rape as you said it’s rape, I see rape as more violent, but maybe . . .” (Ann, in Wood and Rennie, 1994)

Language, Power, and Sexual Assault


In Example 1, the phrase “actual rape” suggests that the experience had the appearance of a rape, and that’s one reason that Ann calls it a rape. But she distinguishes between “date rape” and “rape,” though she hedges or moderates her statement by using the word “maybe” that linguistically weakens the effect of her statement.
Example 2: “. . . and I tried to even approach Dan, after that, and he wouldn’t speak to me, you know. I wanted to talk to him about, you know, like ‘Am I imagining things?’, and in a way, I’m glad in a way that he didn’t talk to me ’cause I think it forced me to draw, you know, like to realize it.” (Mary, in Wood and Rennie, 1994)

In Example 2, besides the self-reflective comment “Am I imagining things?”, we see more than one type of hedge, including the use of “in a way,” and the verb of cognition in “I think.” These hedges mitigate the degree of certitude of the statement. In Example 2, the speaker never explicitly names what happened to her as a sexual assault or rape or attack. Example 2 shows hedging as well as false deixis. In this type of linguistic strategy, we’d say there’s a “dummy ‘it’.” (Penelope, 1990).

Exercise 3

Analyze the following sentences to test your understanding of hedging and false deixis: “Sometimes now I can’t believe that it happened to me, ’cause it’s so long and I’ve healed. And I know it happened, but I’m so, I’m so separated from it now . . . I feel better and I’ve healed over it and I’ve come to terms with it.”

Euphemism and Taboo
Another linguistic strategy is to use a euphemism to substitute pleasant, acceptable language in place of stating something that is unpleasant and socially unacceptable, and therefore unmentionable (taboo) in one’s speech community. Shonna Trinch (2001) studied the language of Spanish-speaking Latinas who applied for protective orders in a legal setting. These women narrated their experiences of domestic violence to interpreters who would later develop the legal documents (e.g. affidavits) to secure the protective orders. When narrating these violent experiences, which included instances of marital rape, 80 percent of these women chose a euphemism in place of a term for rape or sexual assault. One woman used the word “violar” (“to rape”) in telling about her violent sexual experiences. In her other references to being raped, she used euphemistic language more characteristic of romantic or erotic experience:
Este, después que trató de hacerme el amor, él me tocaba todo el cuerpo. ‘Then, after he tried to make love to me, he touched me all over my body.’ (Trinch, 2001: 584)

Trinch identifies six gradients forming categories of women’s use of language about their experiences of sexual violence in Table 8.1. Some of the women who want to obtain a protective order use a lexicon and style of storytelling that violate norms of the institutional discourse of a legal report, which needs to be precise and definitional. Trinch (2001) suggests that, as some of these women tell their stories, they are pulled between their sociolinguistic competence to avoid taboos and their sociolinguistic competence of contextual appropriateness.

Table 8.1: Spectrum of directness for Latina women’s references to sexual violence Group C Less direct (legal, euphemistic and unspecified) Group D Less indirect (not legal, euphemistic and ambiguous) Group E Indirect (not legal, euphemistic and vague) Group F Inexplicit, nearly nonexistent (possibly would not have been reported if interviewer had not asked) #3

Group A Direct and explicit (legal, not euphemistic, and unambiguous)

Group B Direct (not legal, euphemistic, yet unambiguous)


He raped me





He raped me,

#7 #9

So he tried to rape me,

If I don’t um, if I don’t go sleep with him, uh, he, he, twists, my arm,

He was pulling my clothes off. He was tearing them off . . .


Try to, try to rape me again #12 And see, I had a stepfather, right? And when I was seven years old, or six, he uh, he sexually, um, molested me, you know? Right? #13 Because one of my girls got molested at school.

#21 Y trató de, de violarme tres veces ‘And he tried to, to rape me three times.’

#10 Hace, la última vez que tuvo relaciones sexuales conmigo, fue a fuerzas, . . . ‘It was, the last time that he had sexual relations with me, it was by force . . .’ luego quiere sexo por, por dónde no se debe, y yo no soy una persona que a mí me ( ) ‘And then he wants sex where sex is not supposed to be had, and I am not the kind of person that ( )’

Y, y, él trató de, de abusarme sexualmente en el carro, . . . ‘And, and, he tried to, to sexually abuse me in the car, . . .’

No words were spoken about the sexual assault in the interview. Victim did write about it on the intake sheet, though.

#17 No, Me, ah, me hizo que tuviera relaciones con él ‘No, he uh, he made me have relations with him . . .’

#11 It happened Saturday, the night before that #16 Sexual violence he forced narrative was ‘himself on me. elicited by #14 ‘forced himself.’ interviewer. Near [This utterance the end of the comes from an interview the topic intertextual link is re-introduced by (see Bauman the interpreter. and Briggs 1990) to the victim’s #19 Sexual violence narrative was conversation elicited by the with the service interviewer provider prior to tape-recording.]

#18 Y, entonces, este, y me forzó a tener relaciones con él ‘And, then, um and he forced me to have relations with him, . . .’

#20 Yo tengo, yo le tengo miedo de él, eh, este, entonces, entonces, trató tres veces de intentarlo, de, de tener relaciones conmigo, . . . ‘I am afraid, I am afraid of him, um, then, then, he tried three times to try it, to, to have relations with me. . . .’

#22 Sexual violence #15 Y me quiso narrative was morder, elicited by ‘And he tried to interviewer bite me, . . .’

Source: Trinch, 2001: 582.


Parker and Mahlstedt

Words: Survivor and Victim
Historically, feminist activists have often rejected the position of victimhood by preferring to use the word “survivor.” Survivor or victim can refer to a component of a woman’s sense of identity after the experience of sexual assault. So, after a possibly traumatizing experience, a person is confronted with rather paltry choices for describing her sense of self in relation to this experience:
“Victim, I hate with a passion. I hate it. I think that I felt powerless enough, and the last thing I’d ever want to think of myself as is a victim.” (Barb, in Wood and Rennie, 1994)

Wood and Rennie (1994) point out that this person is making a distinction between being victimized and being a victim; she’s a person who’s experienced an injustice, but by using a verb (victimized) instead of a noun (victim), she doesn’t have to take on a new and possibly enduring identity as a victim. A woman whose sister had been murdered after being raped adds another dimension: “My sister was murdered after she was sexually assaulted. The word ‘survivor’ doesn’t recognize her.” This too might change how we understand and use the word “survivor.”

Syntax and Semantics in the Language of Sexual Assault
We’re moving on now from word-level to sentence-level meaning. The study of syntax centers on the structure of the sentence, in which words are grouped into units called constituents. Words are organized into larger constituents called phrases, which then are combined into larger units called clauses, which then are joined together to create sentences and larger groupings of language called texts or discourse (oral and written). Semantics is the study of meaning at the word level, sentence level, and at the level of using language in real-world interactions. Clearly, sentence meaning comes from more than the words alone; we also derive meaning from the structure of a sentence. The sentence “The man sexually assaulted the woman” has the same words as “The woman sexually assaulted the man,” but the meanings of the two sentences are different because the word order is different. Analyzing the effects of language in news reports about rape and sexual assault, we will ask, “What is the influence of syntax on semantics?” In a classic psycholinguistic study, Nancy Henley, Michelle Miller, and Jo Anne Beazley (1995) set out to understand the influence of syntax and semantics on female and male readers’ perceptions of sexual violence described in written news reports. Here are three test sentences they developed based on sentences from news reports of incidents of violence against women:
(a) In the U.S. a man rapes a woman every 6 minutes. (b) In the U.S. a woman is raped by a man every 6 minutes. (c) In the U.S. a woman is raped every 6 minutes.

In this data set, sentence (a) is in the active voice, in which the subject “man” is the agent (or doer) of the action. Sentence (b) is in the passive voice. That is, the syntax of this sentence includes a “by-phrase” (in sentence (b) the by-phrase is “by a man . . .”) that tells us who the agent of the action is. In both sentences (a) in active voice and (b) in passive voice, the agent is “a man,” and the patient (the person acted upon) is “a woman.” Sentence (c) is another type of passive sentence, in which the agent is omitted.

Language, Power, and Sexual Assault


Therefore, this kind of passive syntactic construction is called an agentless passive or truncated passive. As you notice, this kind of passive construction does not have a “by-phrase” and it does not communicate who the agent is; all we can assume is that someone is raping a woman every six minutes. Employing four different questionnaires of attitudes about violence against women, Henley et al. (1995) investigated the effect of the use of the passive voice on participants’ assessment of harm to the victim/survivor, responsibility of the male perpetrator, and their attitudes of acceptance or rejection of sexual violence in relationships. The researchers found that male readers attributed less harm to the victim/survivor (who is female) and less responsibility to the perpetrator (who is male). In addition, while female readers did not have the same response as male readers to the dimensions of harm and responsibility, female and male readers both showed more acceptance of rape, woman battering, and misconceptions about rape when they read text written in the passive compared with the active voice. Penelope Eckert and Sally McConnell-Ginet (2003) note that a particular syntactic construction does not automatically or always result in a particular social meaning. However, syntactic choices do convey information in different ways, such as emphasizing different roles in a sentence or different information. When there are alternate linguistic structures we can use to express an idea, the choices we make can, with or without our awareness, convey different information. If, for example, using an active syntactic construction rather than an agentless passive construction makes a difference in how a reader will interpret the seriousness of an act (such as a sexual assault), we might well want to make an informed choice about the words and syntax we choose to use. The psycholinguistic research by Henley et al. (1995) raises the concern that the use of passive syntax might contribute to increasing apathy toward male violence against women.

Exercise 4

(1) Discuss how the different syntactic structures in the two sentences below might influence the meaning you assign these sentences: (a) The man raped the woman repeatedly. (b) The woman was raped repeatedly. (2) Consider ethical issues related to the use of agentless passives vs. the active voice in college or university news reports on sexual assault.

Having looked at the lexicon, syntax, and semantics of sexual assault experiences, let’s move on to look at how language is used in additional social and cultural contexts. More specifically, we will focus on the different factors that influence our use of language.

Survivor Discourse: Silence, Voice, and Activism
In linguistics, the term discourse has many different meanings (Jaworski and Coupland, 2006; van Dijk, 1997a). Discourse can be examined at the level of describing the language people use; and it can be examined at the level of much broader societal influences that shape the language people can and cannot use in given situations. The phrase survivor discourse includes, but also refers to more than, the actual language or speech used by those who have been sexually assaulted or who talk about sexual assault. It encompasses the interactions among people using language, as well as the role of


Parker and Mahlstedt

context. Survivor discourse also is up against the forceful influence of strongly entrenched normative belief systems, such as sexism and patriarchy. One of the most immediate factors influencing talking about sexual assault is safety: Is it safe to talk about sexual assault experiences? Most would agree it is not safe, while others would say that the voices of sexual assault survivors are actively silenced. In a classic article about survivor discourse, Linda Alcoff and Laura Gray (1993) refer to the silencing of the speech of survivors as “excluded speech.” Survivors’ speech is constrained by various rules, sometimes completely disallowed, at other times and in different places categorized as mad or untrue or rendered inconceivable. A woman subjected to sexual assault by someone she knows may be met with disbelief, accused of lying or imagining it. She may also be stigmatized as provoking her sexual assault, as a “slut” or as the “bitch” making it up to seek revenge.
Identify three pejorative words that stigmatize women’s sexual activity. How are these words used to silence women? Gather your data by talking to your friends and using media like email, text messaging, newspapers, talk shows, and other sources.

Exercise 5

Gwendolyn Etter-Lewis (1991, 1993) uses the construct “suppressed discourse” to refer to the “internal censorship of one’s own speech”. In her linguistic analyses of professional African American women’s narratives, Etter-Lewis (1993) explains that suppressed discourse can be an indicator of “chronic oppression” as experienced by these African American women, and indeed the language of sexual assault survivors can be examined in this way also. According to Alcoff and Gray (1993) the enforced silence of those victimized by violence has necessitated the development of both safe places to begin the healing process and collective public actions to end the social injustice of sexual and racial violence. These spaces now include local and national support groups for survivors, feminist therapy, and organized public actions such as “Take Back the Night”, the “White Ribbon Campaign” (male allies who work to end violence against women), The Vagina Monologues, and LGBTQA initiatives to stop violence against members of the LGBTQA community. Alcoff and Gray explain that during the second wave of feminism in the 1970s and 1980s, a prevailing idea was that “having a voice” or “breaking the silence” would empower women in their activism within and against patriarchy. From working collectively in community organizing efforts and talking with each other in various political and consciousness-raising groups, women felt empowered. Many women recognized that speaking out about their personal victimization had significant social and political consequences for more than the one individual and her experience.
Identify four organizations in your community and/or campus that address social injustice and offer support to members of marginalized groups who might be silenced, including women, children, diverse ethnic and racial groups, people living in poverty, LGBTQA people, and people with disabilities.

Exercise 6

Alcoff and Gray (1993) suggest, however, that speaking out about sexual assault in an exploitative context can serve to maintain the status quo and contribute to the subordination of the group speaking out for change. For instance, certain talk shows

Language, Power, and Sexual Assault


seek to attract viewers with sensationalized self-disclosure about private and often taboo experiences. But who benefits? Probably not the survivors. If breaking the silence is to be effective as a strategy for personal empowerment and social change, it’s important to identify the obstacles and risks along with the available resources inherent in the very structure of context itself. The speech of those sexually assaulted is still likely to elicit negative responses and consequences. Women of oppressed races and social classes who have been sexually assaulted are far less likely to receive comparable media, community, and police attention; the message implicitly shouted by the systematic neglect is “don’t expect justice” (Irving, 2008). This does not mean that people should remain silent, but rather that the risks are different for members of different social groups and the strategies of resistance, therefore, require a different line of thinking. Knowledge about the context of speaking is critical.

Context, Power, and Difference
What is context? In different types of talk we hear people arguing, joking, and insulting. These different types of talk are organized into speech events: making an argument, telling a joke, or testifying as a witness in court. A speech event is embedded in a situation, which provides meaning and form: how it begins, is structured, and how it ends (Eckert and McConnell-Ginet, 2003). According to Teun van Dijk (1997), “. . . not all properties of a social situation are part of the context of discourse.” Here he differentiates between the general notion of a social situation and the formal, specific concept of context as used in linguistics. Van Dijk notes, “. . . context may involve such parameters as participants, their roles and purposes, as well as properties of the setting such as time and space.” Authority and power can also be active features of context. For instance, typical dimensions of power relations include hierarchical relationships: doctor/patient, teacher/student, parent/child. A teacher has access to certain speech acts, such as commands and directives. Part of the role of doctor, teacher, parent—or, for that matter, judge or police officer—means having linguistic power, that is, having access to specific types of commands. Power is a central and complex aspect of context; it’s really about the processes and dynamics of exerting influence and control. How is this relevant to language and sexual assault? In what daily situations can we observe the working of power relations, including that of inequality, through language? We might consider a conversation between two individuals, talk among members of a sexual assault support group, the process of applying for a restraining order, or answering an attorney’s questions in court. For now, in our analysis we will focus on one situation, examining different levels of power. Let’s return to Alice and Jesse at the beginning of this chapter. What are some of the power issues in the very question “To tell or not to tell?” Consider carefully both the possible benefits and empowerment, and the possible drawbacks that could result from talking with a close friend about your experience of being sexually assaulted. What words might Alice use to talk with Jesse about her own experience of sexual assault, if she chooses to? How can she exercise some control over how this unfolds? Will she tell Jesse she’s had a “bad experience” or describe the sexual assault in another way? Will she identify herself to Jesse as a survivor or victim, or use neither word? What might her concerns be? These issues are multifaceted. For example, will she evoke positive and negative stereotypical images of feminism, of her school’s women’s studies classes, or other images that you can imagine? Will she be ostracized by her peers and classmates for these associations? Will her residence hall advisor support her or not? If she goes to a university official responsible for student life issues, will the administrator downplay


Parker and Mahlstedt

the incident of sexual assault or encourage her to take further steps to deal with it? Where are her sources of further empowerment? In this introductory discussion of power and sexual assault, we have raised questions about different levels of power that operate simultaneously and affect Alice and Jesse’s conversation. These sites of power exist at the level of the two individuals (Alice and Jesse), groups of Alice’s peers, the institutional response from one university official, and the normative values invoked through stereotypical beliefs and images related to sexual assault.

Exercise 7

Here we alter some details to the scenario above: (1) How might being Latina American affect whether or not Alice decides to tell anyone about her sexual assault? How might being Anglo American affect whether or not Alice decides to tell anyone about her sexual assault? (2) Now change some elements of the context: (a) If alcohol or drugs were involved, how would that affect Alice’s decision to tell or not to tell? What reactions might her friend Jesse have? What reactions might a university official have? (b) If the setting was a fraternity party or a room in a residence hall, how might that affect Alice’s decision to tell or not to tell? What reactions might her friend Jesse have? What reactions might her brother have?

What power dynamics are involved in the scenarios in Exercise 7? For example, often people blame the woman who has been assaulted. In this instance, what are the power dynamics of the phenomenon of victim-blaming? Why do so many women who have been sexually assaulted blame themselves for a crime committed against them rather than blaming the perpetrator? Where do these beliefs about who is responsible for sexual assault come from? Why do certain people’s beliefs carry more weight than other people’s beliefs? Part of the answer is that, under patriarchy, in the U.S. certain social groups (especially white males) have disproportionate amounts of power to assert their attitudes and beliefs over other social groups (Fonow, Richardson, and Wemmerus, 1992). Why is that? We will now look at theoretical frameworks that examine language, power, and inequality in context. In the large body of linguistic research on power relations, feminist theory and critical discourse analysis (CDA) are prominent. Each of these frameworks proposes that power relations exist at the levels of the individual, social category (e.g. gender, race, dis/ ability), and institutional structures (e.g. government, judicial system, economy). Analyses of power relations in context can involve identifying sources of power, locating the sites where power is negotiated and exercised, how power is experienced internally as well as externally, and how power enforces and maintains the status quo. In addition, social change is a central component of both feminist theory and critical discourse analysis.

Feminism and Feminist Linguistics Feminist theory offers ways to conceptualize, analyze, and eliminate power inequities that maintain social injustices associated with gender, race, social class, ethnicity, age, disability, sexual orientation, and nation. Such social groups derive structural power from their status in our primary institutions: judicial, legislative, religious, economic,

Language, Power, and Sexual Assault


and educational. Social groups that have majority status—white males, affluent classes, heterosexuals, speakers of Standard American English—have the structural power to make the decisions that determine for all others law, public policy, and allocation of resources. In the U.S., the structural power afforded members of these dominant social groups grants them access to greater economic (material), social, political, and linguistic resources than are available to those belonging to other social strata. These resources are automatic benefits or unearned privilege of dominant group membership (McIntosh, 1988). What does this mean for a woman who has been sexually assaulted? Structural power allows and supports the beliefs or rape myths that it is a woman’s fault that she was raped: she was drinking, she was wearing provocative clothing, she is sexually active, she belongs to a certain ethnic group or race, and so on. These beliefs, which stigmatize women, are enforced by structural power. Martin Schwartz and Walter DeKeresedy (1997) claim that males benefit from a social system that does not hold men accountable for date rape; in their words, there is an “absence of deterrents” for sexual assault. Dianne Herman (1984) and Peggy Sanday (1990) argue that rapesupportive culture functions to control women through sexual violence and the threat of sexual violence by men, through intimidation and/or force. Carole Sheffield (1984) coins the phrase “sexual terrorism” to describe the threat of violation that women face on a daily basis. Structural power is one dimension of power relations in context. It is not, however, monolithic. Patricia Hill Collins (1999) focuses on the intersection of multiple forms of oppression based on gender, race, social class, citizenship, ethnicity, disability, sexual orientation, and nation. This matrix of domination characteristic of intersectionality theory expands the locus of power from a static “power over” structural model to a dynamic view of shifting, multiple, interrelated systems of power and oppression formed by the intersection of social categories such as gender, race, and social class. This theory of power relations thus proposes that social groups and individual members of social groups have the potential to intervene, resist, and transform power relations that oppress and discriminate against them. That is, the intersection of an individual’s social identities allows for and creates spaces for resistance where power becomes negotiable (Collins, 1999, 2005; Crenshaw, 1991). Intersectionality also points to the need for sexual assault education and support services appropriate for the distinct needs of diverse groups of women. Later in this chapter, we will apply these ideas about power relations to language used in courtroom proceedings concerning a sexual assault case. Feminist ideas about power relations apply directly to language use as well. In their examination of power, language, and miscommunication, Nancy Henley and Cheris Kramarae note that:
Hierarchies determine whose version of the communication situation will prevail; whose speech style will be seen as normal; who will be required to learn the communication style, and interpret the meaning, of the other; whose language style will be seen as deviant, irrational and inferior; and who will be required to imitate the other’s style in order to fit into the society. (Henley and Kramarae, 1991)

When applied to sexual assault, Henley and Kramarae go on to explain that men’s greater social power privileges their account of a date rape situation. Typically the woman was viewed as responsible for date rape because she did not say “No” clearly or assertively enough (Corcoran, 1992). The date rape was then explained away as an unfortunate result of miscommunication. Indeed, this commonly held view led sexual assault activists to develop slogans such as “What part of ‘No’ don’t you understand?” or


Parker and Mahlstedt

“No means No.” Henley and Kramarae warn that “Miscommunication is often a way to stress difference while ignoring hierarchy.” That is, men’s greater social power affords them the right to pay less attention to women’s protests. Feminist linguists since the 1970s have also examined the relationships among language and gender inequity (Cameron, 1998; Lakoff, 1975; Lakoff and Bucholtz, 2004; Eckert and McConnell-Ginet, 2003; Ehrlich, 2001, 2004; Wodak, 1997). Early on, feminists focused on the consequences of sexist language in titles (e.g. Mr./Mrs.), pronouns (he/she), and pejorative language (e.g. mister/mistress; and bitch). Discussing linguistic inequities between women and men, as well as identifying characteristics of men and women’s language, Robin Lakoff (1975) queries the possibility of remedying a social injustice by changing linguistic usage: “Does one correct a social inequity by changing linguistic disparities?” This question of the real-world effects of sexist language usage was prominent in the 1970s and 1980s. In subsequent decades many feminist linguists refocused their attention on how the uses of language and discourse support structures of inequality based on such social categories as gender and sexualities, race, and social class. Thus Kate Clark (1998) examines the influence of the disproportionately high news reporting of stranger rapes, which misrepresents the reality that most rapes are committed by someone the woman knows. Also with regard to sexual assault, linguists working on the language of law have, at one level, examined the effects of legal definitions of rape and sexual assault in courtroom proceedings. Feminist linguistic analyses of sexual assault cases also address the larger context of the institutional and cultural forces governing the practice of law (Matoesian, 1993; Ehrlich, 2001, 2004).

Critical Discourse Analysis Within the field of linguistics, critical discourse analysis (CDA) is a subset of discourse analysis that describes and theorizes the dynamics of language, power, and inequality. CDA has an explicit commitment to social change, exposing how talk and interaction operate in the interest of dominant social groups and institutions (see Fairclough, 1989; Fowler, 1991; van Dijk, 1988 on the discourse of news media). Multimodal studies represent another area where CDA analyzes various media, including text, visual images, and sound. Here too linguists can reveal how social hierarchies are encoded in these modalities (Kress and Van Leeuwen, 1996). Next we examine ways in which linguistic and feminist concepts have been applied to analyze courtroom discourse about sexual assault and rape.

Words in Public Spaces: Speaking about Rape in the Courtroom
Linguists have examined the linguistic strategies and language itself in courtroom discourse about rape and sexual assault because the courtroom is a prominent site of public discourse about violence against women (e.g. Matoesian, 1993; Ehrlich, 2001). According to Susan Ehrlich, language, so important in constructing social reality, is especially significant in a rape trial: courtroom talk (e.g. testimony) might constitute the only evidence that is examined during a trial. In his book Reproducing Rape: Domination through Talk in the Courtroom, Gregory Matoesian (1993) delineates the power structure manifested in the roles, rules, and procedures of the courtroom. For example, the role of attorney in the courtroom has certain prescribed, explicit tasks and responsibilities, such as determining and thus controlling the questions asked of a witness on the stand. The power to control questioning gives the attorney asymmetrically more structural power than the witness. The role of witness also has prescribed tasks, such as answering questions.

Language, Power, and Sexual Assault


While power is differentially distributed, in this example, witnesses can exert some power through the language and linguistic strategies they employ. In addition, as Matoesian states, discourse strategies and conversational styles must be examined to reveal the power dynamics as they are accomplished through the use of language. While these linguistic structures—roles, question/answer structure—are prescribed to develop and manipulate the direction of discourse, Matoesian is clear that these structures in themselves do not necessarily create dominance or subordination. Following Alcoff and Gray (1993), we will provide a critical analysis in this instance of the language of courtroom transcripts to identify some of the obstacles and risks of speaking in the courtroom. In addition, this analysis will examine possible benefits and opportunities that arise for a sexual assault complainant, i.e. the survivor, while giving testimony. Even within the context of “patriarchal and legal domination” (Matoesian, 1993), the witness still can resist domination by drawing upon a number of mechanisms of talk in crafting her or his account. Both Matoesian (1993) and Ehrlich (2001) point out that the difference in power for women and men in the law courts of patriarchy exists at the same time that patriarchal conventions and institutions maintain the false assumption that women and men are conceived as equals before the law. In practice, as discussed by Matoesian and Ehrlich, gender assumptions and beliefs serve to advantage males and disadvantage females. At this point we will examine power dynamics of language as deployed in a specific context: the courtroom. Matoesian and Ehrlich focus on specific language and linguistic strategies the attorneys use in question-and-answer exchanges with complainants and defendants. For example, in question and answer sequences, the defense attorney uses the complainant’s words to construct an image of the complainant that fits the prevailing negative stereotype of the promiscuous female. Perhaps an even greater concern about the legal process, however, is in the recognition that the professional legal players are not only expert rhetoricians, but carry out their “covert attacks against the [victim] . . . through the interactional manipulation of words, utterances, and sequences” (Matoesian, 1993). Further, and most important, legal discourse practice that might appear neutral on the surface activates patriarchal conceptualizations of rape (Matoesian). As MacKinnon (1989) has noted, “The crime of rape is defined and adjudicated from the male standpoint, presuming that forced sex is sex and that consent to a man is freely given by a woman” (MacKinnon 1989, p. 180, cited in Matoesian, 1993). Through feminist methods and critical discourse analysis, Ehrlich (2001) examines courtroom and university tribunal proceedings of sexual assault and rape cases, demonstrating how the discourse and processes used in these settings contribute to the creation and reproduction of social and institutional inequities. In her linguistic analyses, Ehrlich observes that in the actual testimony of the two female complainants, the women attribute a high degree of agency to the defendant (alleged perpetrator of sexual assault) for “sexual acts of aggression.” Here is an example from Connie’s testimony (all names are pseudonyms; CD = Connie): CD: He grabbed my hair from the back of my neck and sort of wrapped it around his hand and pushed me down in between his legs and told me I could put it in my mouth or he would put it in. [CT = court trial] (Ehrlich, 2001) Analyzing the male defendant’s testimony, Ehrlich records how Matt uses language to represent himself as not responsible for the “unlawful sexual acts of aggression” for


Parker and Mahlstedt

which he is standing trial. When asked by the university lawyer in the tribunal to comment on Connie’s claim that Matt “grabbed her by her hair”, Matt denies the university lawyer’s formulation of Connie’s claim and instead he says, “Yeah I was caressing her hair,” using the “language of love” instead of “the language of violence” (Ehrlich, 2001). In the testimony that follows the defense’s strategy aims to establish that the sexual activity (i.e. kissing) was mutual and thus any sexual interaction that follows was also mutual. (Here, MA = Matt; additional initials represent persons in legal roles, e.g. attorneys): OD: MA: OD: MA: OD: MA: Okay. You’re the one who started to kiss her first; is that correct? That’s incorrect. You started kissing her? That’s incorrect. Who started kissing who? As I was saying, we were lying in the bed and before coming in, she was lying— and I lied next to her and we laid there for about two minutes and then we both came and kissed each other mutually. (CT = court trial) (Ehrlich, 2001)

Ehrlich reports that the defendant describes his role in the events under questioning by assigning responsibility and agency to the complainant: “. . . mitigated the force of his agency or the negative nature of his agentive acts.” In the next example, Ehrlich presents excerpts of the courtroom transcription of the complainant Connie’s response while being cross-examined by the attorney for the defense (of the accused perpetrator) and identifies how the female complainant uses linguistic resources to resist her questioner during question-answer interactions in the courtroom. At one point, Connie testifies that Matt has insisted that she have sex with him, if not today another day; she finally says yes to stop his harassment, after which he threatens that she had better not back out on her “promise.” He falls asleep and she waits for him to leave. The question-answer process continues: Q: What did you think would happen if you didn’t do what he told you? CD: I don’t know. I don’t think I really gave much thought to what would happen if I didn’t, except I was afraid not to. Q: What were you afraid of or what were you afraid would happen if you did not do the oral sex that he told you to? CD: I was afraid that he would hurt me. Q: When you said that you felt that you had no choice, what did you mean by that? CD: I felt that I had to do it or get hurt. (Ehrlich, 2001) Ehrlich discusses Connie’s behavior as an act of agency in a dangerous context in which she has very limited options to avoid sexual assault and further physical injury. Connie’s motivation of fear and her response of self-protection are manifest in the language she uses to describe her fear of being hurt if she did not comply with Matt’s aggression. Also, Connie’s language can be, and in this instance was, being used against her. How? The legal contexts frame her very act of resistance, that is, her submitting to coercive sexual aggression to avoid perceived danger, as her consenting to the unwanted sexual activity.

Language, Power, and Sexual Assault


Exercise 8

In date rape and stranger rape, the fear Will he (or they) hurt me or kill me? is prominent when a woman is being sexually assaulted. In Connie’s situation she explains that her fear motivated her to comply with Matt’s sexual demands. Do you think that the woman’s fear of being physically harmed, perhaps killed, should be considered in these proceedings?

It is important to note that the interactional work (i.e. cross-examination) between the defense lawyer and the complainant alone would not work to discredit Connie. Matoesian (1993) explains that the linguistic strategies used by a defense attorney to misrepresent the female, in this instance as consenting, engage patriarchal ideology which as an interpretive framework reinforces and validates the belief that if you’re raped by someone you know, it’s consensual or that “normal” young men do not rape women. If the dominant discourse of patriarchal ideology did not exist as an interpretive framework, the defense attorney’s construction of the revenge-seeking “bitch” or the promiscuous female might not have the same impact. Drawing on this ideology brings forth more than the image of the avenging bitch or the promiscuous female; it is designed to also conjure up related patriarchal beliefs such as the idea that the female is responsible for the male’s sexual arousal and therefore must satisfy him. In this sense the dominant discourse of patriarchal ideology “manufactures” consent where it did not exist. According to Ehrlich (2001), to counteract this prevailing patriarchal discourse in the courtroom it is necessary to establish an alternate interpretive framework that explains that “women’s and men’s differential expressions and interpretations, to the extent that they exist, are shaped by social inequities.”
How does Matt use language in his attempt to diminish his own responsibility for the actions he’s accused of committing?

Exercise 9

The issues of agency and resistance in the courtroom are further sharpened and extended when we turn to a linguistic analysis by Elaine Richardson (2006) of proceedings of the perjury trial against rap artist Kimberly Jones and Jones’ assistant Monique Dopwell. In the context of hip hop culture, Richardson (2006) argues that rappers have “language, literacy, and rhetorical practices” aimed at challenging oppression. As she states, “Although Kimberly Jones’ case is theoretically about perjury, it is also about racism, sexual exploitation, and gender oppression in the larger society and in Hiphop” (Richardson, 2006). Richardson argues that one of the prosecution’s strategies in its perjury case against Kimberly Jones was to construct a sexist, racist stereotype of her as a “Black Jezebel,” that is, as a “lying, ambitious gold-digger and an immoral, promiscuous operator.” To confront and correct the prosecutor’s characterization of Jones, Kimberly Jones’ lawyer engages his client in a series of question-answer interactions that create a space for her to tell her life story, from her upbringing through her later development as the performing artist rapper Lil’ Kim. It is Jones’ language and storytelling that communicate to the jury her humanity. With its focus on matters important to the storyteller, in the courtroom storytelling gains a public hearing for the storyteller’s voice and affords some degree of selfrepresentation and control. According to Richardson (2006), “Storytelling remains one


Parker and Mahlstedt

of the most powerful language and literacy practices that Black women use to convey their special knowledge.” And, in Kimberly Jones’ trial, telling her life story is an act of resistance.

Language, Empowerment, and Social Change
. . . we want to emphasize that change comes in subtle ways. At any historical moment, both the gender order and linguistic conventions exercise a profound constraint on our thoughts and actions, predisposing us to follow patterns set down over generations and throughout our own development. Change comes with the interruptions of such patterns, and while sometimes that interruption may be sudden, it comes more commonly through infinitesimally small events that may or may not be intentional . . . Connecting (smaller linguistic) routines to larger societal discourses requires that we think about how small acts ramp up into big ones. Above all, it requires thinking about how a single individual’s verbal move could get picked up by others and eventually make it into public discourse. (Eckert and McConnell-Ginet, 2003)

Eckert and McConnell-Ginet make the connection between language usage and social change. Social change begins in personal and in collective expressions of resistance. Even in a conversation between two people, each person can contribute to a new understanding, while larger societal forces at the same time play a significant role. As Deborah Cameron (1998) says, “feminist activity . . . has sensitized language users to the non-neutral nature of linguistic representation.” One consequence of feminist efforts in linguistics is that now many people are more likely to question our own and others’ use of language. Feminist linguists and practitioners of critical discourse analysis share the focus on language use, power, and inequality. In both approaches, linguistic analysis can lead to real-world applications, such as changes in how news reporters and editors might choose language to describe sexual assault. If an individual student reporter resists the normative use of agentless passives to describe a rape for her own school newspaper, she sets a precedent that could influence other reporters and lead to a change in her newspaper’s policy for reporting incidents of sexual assault. We will now explore examples of how individuals, activist groups, and organizations initiate social change by challenging prevailing norms of violence against women through their various means of expression and strategic use of voice and language.

Empowerment, Health, and Healing: The Benefits of Telling Stories
Elaine Richardson (2006) argues that Kimberly Jones’ telling her life story in court is an act of resistance against the negative stereotype of the Black Jezebel that the prosecution invokes to erase the person Kimberly Jones. Kim’s life story reveals her humanity to the jury, who learn that she worked for H&R Block, a bank, Bloomingdale’s, had a difficult family situation as her mother and father were separated, and went on to develop a career as a renowned rap artist. Telling stories can be humanizing for those telling the stories also. Much psychological research indicates that telling or writing of difficult or traumatic experiences can lead to improved health (e.g. Pennebaker and Beall, 1986). Researchers have found that survivors of sexual assault can benefit from making repeated accounts or narratives of their sexual assault experiences and that hearing their own accounts can also benefit survivors (e.g. Foa and Rothbaum, 2001; Booker, 2002).

Language, Power, and Sexual Assault


In her book Counseling to End Violence against Women: A Subversive Model, Mollie Whalen (1996) writes about healing and empowerment in the context of different types of feminist therapies. One aspect of feminist therapy focuses on individual capabilities: for a person to gain a sense of her own existing capacities or power. Empowerment involves one’s own personal power, aspects of self-determination, and the ability to identify resources, including people who become part of their social support system. Although this is indeed an important element of healing and empowerment, a feminist perspective offers an additional goal for therapy: clients critically examine and learn ways to challenge oppression within patriarchal societal structures. Engaging in social change is thus a necessary component of a client’s process of healing in feminist therapy (Brown, 2006). Women will take their own paths during the process of healing and empowerment. For many survivors, reaching out to others beyond the silences to break free of the shame, fear, and isolation is in part a desire to build community with other women (and, for some, with male allies). Charlotte Pierce-Baker offers these words:
The way out is to tell: speak the acts perpetrated upon us, speak the atrocities, speak the injustices, speak the personal violations of the soul. Someone will listen, someone will believe our stories, someone will join us. And until there are more who will bear witness to our truths as black women, we will do it for one another. For now, that is enough. (Pierce-Baker, 1998)

Public Activism to End Sexual Assault and Battering
Social change begins in personal and in collective expressions of resistance. Speaking out collectively in public spaces is a way to achieve specific goals: raise consciousness about violence against women, develop community resources and services for survivors and their families, and challenge the structures that maintain and enforce prevailing oppressive beliefs. Collective actions of public resistance have appeared in many forms. Begun in 1977, “Take Back the Night” is one of the oldest grassroots public demonstrations dedicated to increasing awareness of sexual assault. In a direct expression of resistance to male violence against women, women march through the streets at night in a symbolic reclaiming of the freedom to fearlessly walk outside at night. This is followed by a candlelight vigil where survivors of sexual violence publicly share their experience. Eve Ensler (2000) developed The Vagina Monologues out of interviews with women, some centering on sexual assault. Annually, women join their voices in public readings of the monologues to raise awareness, break silences, and support grassroots efforts and organizations to end violence against women. The Clothesline Project is an ongoing, grassroots public art project that began in 1990. Survivors of domestic violence as well as allies honoring the memory of murdered friends and family members represent their experiences of abuse on symbolically colored t-shirts that are then publicly displayed on a clothesline to foster awareness of woman abuse. By these actions, participants resist social norms that pressure women into not speaking about their experiences of violence. (See The rhetoric, art, and new insights coming out of survivor standpoints continue to expand public awareness of the actual nature of woman abuse in a powerful expression of social resistance (Droogsma, 2008). Rape prevention work with men emerged in the 1980s. Beginning in 1988, the Fraternity Anti-Violence Education Project began a one-year training program based on feminist principles and pedagogy for fraternity men. Undergraduate male peers co-facilitate


Parker and Mahlstedt

the seminar, creating an all-male space for men to examine their own sexism, as well as how to identify sexist behavior and the causes of violence against women. The project is experientially based, and over the period of one year the men, in a conscious way, learn how men can relate to men and women outside the traditional roles of hypermasculinity (Mahlstedt and Corcoran, 1999; Mahlstedt and Doyon, 1999/2006). Spurred on initially by one man’s massacre of fourteen women in a college classroom at L’École Polytechnique in 1989 in Montreal, Quebec, the White Ribbon Campaign was organized and run by Canadian men who held week-long programming, such as workshops and speakouts, to educate men about violence against women and to motivate men to take action. The White Ribbon Campaign now takes place beyond Canada and is a regular feature of programming on many university campuses in Canada and the United States.
What information and support services does your institution or local community provide for sexual assault and rape survivors? For example, does any local organization sponsor ongoing support groups for survivors of sexual assault, rape, incest, and/or battery?

Exercise 10

Activist or street theater can generate energy and awareness of social conditions. Speaking, writing, and using theater methods in a neighborhood rape crisis center can activate different members of that community. Using methods and techniques of Theatre of the Oppressed (TO), members of a rape crisis center or domestic violence shelter can explore together current, oppressive situations and conditions (e.g. limited resources, poor communication, racism, and heterosexism at domestic violence shelters). According to Augusto Boal, director, writer, and former member of the Brazilian parliament, “The goal of the Theatre of the Oppressed is not then to create calm, equilibrium, but rather to create disequilibrium which prepares the way for action” (Boal, 1995). With practitioners and TO centers around the world, this form of activist theater enables self-selecting audience members to participate as spect-actors (“I see, I act”) to identify, develop, and explore issues and oppressive circumstances by replacing one of the existing actors (or spect-actors) in a scene and trying out a new strategy. Violence against women is a social injustice we can do something about. Language and linguistics play an important role in meeting this challenge. Linguistics can be an active force for social change when it is applied to reveal the often invisible workings of power through language.

To provide insight into the diversity of researchers and scholars, we use first and last names of authors on first citation outside parentheses.

Alcoff, L. M. and Gray, L. (1993) Survivor discourse: Transgression or recuperation? Signs: Journal of Women in Culture and Society, 18(2), pp. 260–290. Bletzer, K. V. and Koss, M. P. (2004) Narrative constructions of sexual violence as told by female rape survivors in three populations of the Southwestern United States: Scripts of coercion, scripts of consent. Medical Anthropology: Cross-Cultural Studies in Health and Illness, 23(2), pp. 113–156.

Language, Power, and Sexual Assault


Boal, A. (1995) Rainbow of desire. London, Routledge. Booker, M. (2002) Stories of violence: Use of testimony in a support group for Latin American battered women. In Collins, L. H., Dunlap, M. R., and Chrisler, J. C. eds. Charting a new course for feminist psychology. London, Praeger, pp. 307–320. Brown, L. S. (2006) Still subversive after all these years: The relevance of feminist therapy in the age of evidence-based practice. Psychology of Women Quarterly, 30(1), pp. 15–24. Cameron, D. (1992) Feminism and linguistic theory. 2nd edn. New York, Palgrave Macmillan. Cameron, D. (1998) Gender, language, and discourse: A review essay. Signs, 23(4), pp. 945–973. Clark, K. (1998) The linguistics of blame: representations of women in the Sun’s reporting of crimes of sexual violence. In Cameron, D. ed. The feminist critique of language: a reader. 2nd edn. London, Routledge, pp. 183–97. Collins, P. H. (1999) Moving beyond gender: Intersectionality and scientific knowledge. In Ferree, M. M., Lorber, J. and Hess, B. B. eds. Revisioning gender. London, Sage Publications, pp. 261–284. Collins, P. H. (2005) Black sexual politics: African Americans, gender, and the new racism. New York, Routledge. Corcoran, C. B. (1992) From victim control to social change: A feminist perspective on campus rape and prevention programs. In Chrisler, J. and Hower, D. eds. New directions in feminist psychology. New York: Springer, pp. 120–140. Crenshaw, K. W. (1991) Mapping the margins: intersectionality, identity politics, and violence against women of color. Stanford Law Review, 43(6), pp. 1241–1299. Crenshaw, K. W. (1995) Mapping the margins: Intersectionality, identity politics, and violence against women of color. In Crenshaw, K. W. ed. Critical race theory: The key writings that formed the movement. New York, New Press, pp. 357–383. DeFrancisco, V. (1997) Gender, power and practice: Or, putting your money [and your research] where your mouth is. In Wodak, R. ed. Gender and discourse. Thousand Oaks, Sage Publications, pp. 37–56. Droogsma, R. A. (2008) “He might of cracked my spirit, but he never broke it”: A feminist standpoint analysis of woman abuse survivors’ messages in The Clothesline Project. Paper presented at National Women’s Studies Association, June 2008, Cincinnati, OH. Eckert, P. and McConnell-Ginet, S. (2003) Language and gender. Cambridge, Cambridge University Press. Ehrlich, S. L. (2001) Representing rape: Language and sexual consent. New York, Routledge. Ehrlich, S. L. (2004) Linguistic discrimination and violence against women: Discursive practices and material effects. In Lakoff, R. T. and Bucholtz, M. eds. Language and woman’s place: Text and commentaries. Oxford, Oxford University Press. Ensler, Eve. (2000) The vagina monologues. New York, Dramatists Play Service, Inc. Etter-Lewis, G. (1991) Standing up and speaking out: African American women’s narrative legacy. Discourse & Society, 2(4), pp. 425–437. Etter-Lewis, G. (1993) My soul is my own: Oral narratives of African American women in the professions. New York, Routledge. Fairclough, N. (1989) Language and power. London, Longman. Foa, E. B. and Rothbaum, B. O. (2001) Treating the trauma of rape: Cognitive-behavioral therapy for PTSD. New York, Guilford Press. Fonow, M. M., Richardson, L. and Wemmerus, V. A. (1992) Feminist rape education: Does it work? Gender and Society, 6(1), pp. 108–121. Fowler, R. (1991) Language in the news: Discourse and ideology in the press. London, Routledge. Henley, N. M. and Kramarae, C. (1991) Gender, power, and miscommunication. In Coupland,


Parker and Mahlstedt N., Giles, H., and Wiemann, J. eds. Miscommunication and problematic talk. Newbury Park, Sage, pp.18–43. Henley, N. M., Miller, M. and Beazley, J. A. (1995) Syntax, semantics, and sexual violence: Agency and the passive voice. Journal of Language and Social Psychology, 14(1–2), pp. 60–84. Herman, D. (1984) Rape culture. In Freeman, J. ed. Women: A feminist perspective. Palo Alto, CA, Mayfield Publishing. Irving, T. (2008) Decoding Black women: Policing practices and rape prosecution on the streets of Philadelphia. National Women’s Studies Association Journal, 20(2), pp. 100–120. Jaworski, A. and Coupland, N. (2006) The discourse reader. 2nd edn. London, Routledge. Koss, M., Gidycz, C. and Wisniewski, N. (1987) The scope of rape: Incidence and prevalence of sexual aggression and victimization in a national sample of higher education students. Journal of Consulting and Clinical Psychology, 55(2), pp. 162–170. Kress, G. and Van Leeuwen, T. (1996) Reading Images: The grammar of visual design. London, Routledge. Lakoff, R. (1975) Language and woman’s place. New York, Harper & Row. Lakoff, R. and Bucholtz, M. ed. (2004) Language and woman’s place. Rev. and expanded edn. Oxford, Oxford University Press. Lazar, M. M. (2007) Feminist critical discourse analysis: Articulating a feminist discourse praxis. Critical Discourse Studies, 4(2), pp. 141–164. McIntosh, P. (1988) White privilege and male privilege: a personal account of coming to see correspondences through work in women’s studies. Working paper no. 189. Wellesley College Center for Research on Women, Wellesley, MA. MacKinnon, C. (1989) Toward a feminist theory of the state. Cambridge: Harvard University Press. Mahlstedt, D. and Corcoran, C. B. (1999) Preventing dating violence. In Crawford, M., Davis, S. and Sebrechts, J. eds. Coming into her own. San Francisco, Jossey-Bass, pp. 311–327. Mahlstedt, D. and Doyon, D. (1999, 2006) Men’s work [video: Insight Media]. Matoesian, G. (1993) Reproducing rape: Domination through talk in the courtroom. Chicago, University of Chicago Press. Parker, J. A. and Corcoran, C. B. (2006) Acoustic and linguistic assessment of emotional stress in oral accounts of sexual assault experiences. Presentation at the 3rd Annual Conference on Innovations in Trauma Research Methods, November, 2006, Hollywood, CA. Penelope, J. (1990) Speaking freely: Unlearning the lies of the fathers’ tongues. California, Pergamon Press. Pennebaker, J. W. and Beall, S. K. (1986) Confronting a traumatic event: Toward an understanding of inhibition and disease. Journal of Abnormal Psychology, 95(3), pp. 274–281. Pierce-Baker, C. (1998) Surviving the silence: Black women’s stories of rape. New York, W. W. Norton and Company. Richardson, E. B. (2006) Hiphop literacies. New York, Routledge. Sanday, P. R. (1990) Fraternity gang rape: Sex, brotherhood, and privilege on campus. New York, NYU Press. Schwartz, M. D. and DeKeresedy, W. S. (1997) Sexual assault on the college campus: The role of male peer support. Thousand Oaks, Sage Publications. Sheffield, C. J. (1984) Sexual terrorism. In: Freeman, J. ed. Women: A feminist perspective. Palo Alto, CA, Mayfield Publishing. Trinch, S. L. (2001) Managing euphemism and transcending taboos: Negotiating the meaning of sexual assault in Latinas’ narratives of domestic violence. Text, 21(4), pp. 567–610. van Dijk, T. A. (1997a) Discourse as social interaction. Thousand Oaks, Sage Publications.

Language, Power, and Sexual Assault


van Dijk, T. A. (1997b) Discourse as structure and process. Thousand Oaks, Sage Publications. van Dijk, T. A. (1988) News as discourse. Hillsdale, NJ, Lawrence Erlbaum Associates. Warshaw, R. (1988) I never called it rape: The Ms. report on recognizing, fighting, and surviving date and acquaintance rape. New York, Harper & Row. Whalen, M. (1996) Counseling to end violence against women: A subversive model. Thousand Oaks, Sage Publications. Wodak, R. (1997) Gender and discourse. Thousand Oaks, Sage Publications. Wood, L. A. and Rennie, H. (1994) Formulating rape: The discursive construction of victims and villains. Discourse & Society, 5(1), pp. 125–148. Young, S. L. and Maguire, K. C. (2003) Talking about sexual violence. Women and Language, 26(2), pp. 40–52.

Other Resources
Books and Articles
Berman, L. (1998) Speaking through the silence: Narratives, social conventions, and power in Java. New York, Oxford University Press. Foucault, M. (1970/1972) The discourse on language. In Foucault, M. ed. The archaeology of knowledge. New York, Pantheon Books. Lakoff, G. and Johnson, M. (1980) Metaphors we live by. Chicago, University of Chicago Press.

Discourse & Society Narrative Inquiry Journal of Narrative and Life History Journal of Language and Social Psychology

Centers for Disease Control and Prevention (CDC): The Clothesline Project: International Society for Traumatic Stress Studies (ISTSS): Linguistic Society of America (LSA): LinguistList (international listserv): RAINN (Rape, Abuse & Incest National Network): U.S. largest anti-sexual assault organization. Founded by Tori Amos. National Sexual Assault Network: 1–800– 655-HOPE, “Free. Confidential. 24/7.”


Gender, Language, and Power Surname or Sirname?

Diana Boxer
This Chapter Explores:
Surnames and Social Labeling Practices Surnames, Status and Power in Society A Brief History of Women and Surnames Recent Trends Waves of Feminism in the U.S. A Quantitative Survey of U.S. Women and Surname Choices Women’s Narratives about Surnames, Marriage, and Children Surnames and Identity Surnames in Russia and the U.S. Surnames across Cultures

It was my mother’s advice [to keep my surname], and now I realize that she was right. Husbands come and go, but my name is always with me. Anonymous informant

This chapter assesses the current state of affairs in women’s naming choices in the U.S. and brings to light some cross-cultural comparisons. I became interested in the subject over the years through thinking about what women around me were doing about the surname issue when faced with marriage or partnership. Even though I am married, I never thought to take my husband’s surname. I have an independent identity as an academic woman; moreover, changing my name always meant ceding to a perceived more “powerful” member of a pair. That just would not do. Naming choices, like all linguistic choices, affect one’s personal as well as professional identity; however, this is only the tip of the iceberg. There is a deeper issue of societal power underlying surnames. Indeed, there is much more involved in the surname question than merely to change one’s name or not to change. I collaborated on a large project on this topic with Russian linguist Dr. Elena Gritsenko, who visited the University of Florida a few years ago as a Fulbright fellow. As linguists, we bring to the issue of surnames and women a new perspective on the social meaning of labeling practices. As scholars in gender and language, we are interested in women and surnames as reflections of social identity and power. The cross-cultural data in this chapter reflect our collaboration. Our research was funded by the Rockefeller Foundation.

Gender, Language, and Power


The above quotation comes from the many stories I collected on the issue of women and surnames in marriage or partnership. Another young woman disclosed something to me that really struck a chord during our interview. She and her fiancé had discussed her intent to keep her birth surname after marriage, and she thought the matter was settled. However, the night before the wedding he confronted her. He told her, in no uncertain terms, that if she did not take his surname she could forget about walking down the aisle. In other words, she was faced with the choice of taking his sirname (a term I devised) or giving up the relationship. An examination of what people choose to call themselves is interesting not merely from the perspective of “identity” but even more so from the perspective of male hegemony: “Naming conventions, like the rest of language, have been shaped to meet the interests of society, and in patriarchal societies the shapers have been men” (Miller and Swift, 1976: 15). While that article dates back more than thirty years, it remains a fact that our society has certainly been shaped by male leaders. From the perspective of sociolinguistics, narratives by women about their own surnames are telling in terms of the speech acts either directly or indirectly referenced by women themselves (speech act refers to an expression associated with a particular intent by the speaker). Women’s naming narratives are replete with such speech acts/behaviors as “insult,” “regret,” “insist,” and “demand.” These directly affect women’s lack of autonomy in this domain, providing examples of how surname choices result from external (or internal) encouragement or discouragement. In an era in which the social constructions of gender, ethnicity, race, and sexual orientation have become critical foci for examination in the social sciences generally (e.g. Le Page and Tabouret-Keller, 1985; Wiley, 1994) and in gender and language specifically (e.g. Kotthoff, 2000; Pujolar, 2000; Scott, 2000), this emphasis on identity is not surprising. Who we are and how we fit into the social and political structure of where we live is critical for our sense of well-being, belonging, and generally participating as a citizen of a society. Thus, identity as a social science issue will no doubt continue to be important. Likewise, from the perspective of critical discourse analysis (the study of language and power), the study of identity and identity politics is not unrelated to the hierarchical power structure of societies. Naming choices, especially when they reflect one’s social identity as a member of a gendered, ethnic, or racial group, have important repercussions for where groups fit into the extant hierarchy. Indeed, women’s stories about changing or keeping their surnames reveal something about how far we have come in gaining a sense of independent identity and power in our lives and in our communities. Names as social labeling practices give us a clue about underlying patterns and shifts in expected roles for women in society. Because of this, it is useful to look across time, from the recent past to present trends. The study of names is called onomastics, and indeed there is a whole segment of scholarly endeavor interested in how names reflect societal values. Naming practices are social markers, just as are address terms like using first names or titles with last names. In the Romance languages, for example, we must choose a formal or informal address term, e.g. in French, tu is the address term for informal “you” and vous is used for the formal “you.” These terms of address reveal levels of deference to our interlocutor as well as our demeanor regarding their status in relation to ours. Just as terms of address provide sociolinguistic information about interlocutors’ relative status and power, the same is true for surnames. The taking of husbands’ surnames remains an entrenched and persistent tradition in U.S. society. Moreover, in my grandmothers’ time married women were identified as the



wife of someone through formal address terms that used the title Mrs. plus the husband’s first and last names. This is not as common a practice now as it was as recently as the 1950s. Nonetheless, recent indicators show that this practice is on the upswing again.1 The fact that women not only relinquish their surname identity but even their given name identity tells us quite a bit about who holds power in a society.

Exercise 1

Make a list of people you know who have and haven’t changed their last names after marriage or formal partnerships. Look at wedding announcements in the local paper. What patterns are you finding? Compare answers.

A study conducted at Harvard’s economics department by Claudia Goldin and her former student Maria Shim (2001) found that U.S. professional women are increasingly taking their husbands’ surnames. This is a change from the 1980s. While just two to three decades ago many young women tended to want to keep their birth surname, the tide has changed. Why is this so? Goldin and Shim (2001) concluded that the decline in surname retention may be due to economic and social gains women have made over the past few decades. They reason that today’s generation of women has gained the ability to have professions and thus economic mobility. They no longer believe social labels to be anything more than symbolic. My own daughter, a physician, took her husband’s name when she married. Her attitude was, “What’s the big deal, mom? We no longer have anything to prove.” My reply was that her generation of women having nothing to prove derives from the struggles of women of my generation. But are surnames merely symbolic? Gender and language studies show us over and over again that societal traditions and gendered power relations are difficult to overcome. Indeed, temporary resistances often see reversals.

Surname or Sirname? A Brief History
The fact is that until the early 1800s in the U.S., all of a woman’s possessions became her husband’s when she married. Women themselves were seen as property, just as slaves were in that era. Women who were slaves, of course, had no right to possessions. Every state had laws whereby a husband owned all property and whereby women were property and could not own property. Both slaves and women took on the surname of the head of the household. In the middle of the nineteenth century, a series of “Married Women’s Property Acts” in several states gave women the right to retain property held before marriage. These were called Couverture laws, and varied by state. Slave women, however, were not given this right. The right to retain one’s birth surname was not considered an issue even for non-slave women. One of the first women to take up this agenda was the early women’s rights advocate Elizabeth Cady Stanton (Mrs. Henry Stanton), who kept her birth surname as a middle name when she married Henry Stanton. This was unusual for the time. Discussing the common convention of married women going by Mrs. followed by the husband’s first name and surname, she stated:
I have very serious objections . . . to being called Henry . . . the custom . . . is founded on the principle that white men are the lords of all. I cannot acknowledge the principle as just; therefore, I cannot bear the name of another. (Kupper, 1990: 10)

Gender, Language, and Power


Cady Stanton was apparently objecting here to the use of her husband’s given first name, not his surname. That women were expected to automatically take their husbands’ surnames began to be challenged in the middle of the nineteenth century. Lucy Stone, early abolitionist and an important figure in the advocacy of social causes, was the first American woman to retain her birth surname after marriage. For a while after that, women who elected to retain their birth names after marriage came to be called “Lucy Stoners.” Unfortunately, as with many expressions ascribed to women or groups of women, that designation has taken on a negative connotation. Perceptions of Lucy Stoners nowadays are not always positive.

Recent Trends
In the past twenty years, new data on the topic have emerged. These indicate pendulum swings in only a short period of time. For example, a 1985 survey of U.S. college students (Intons-Peterson and Crawford, 1985) found that both sexes strongly identified with their birth surnames. Only a dozen years later Murray (1997) found that things seemed to have changed. That study surveyed over ten thousand Mid-Westerners in twelve states covering four age groups and three social classes. It found mixed assessments of female surname retainers. Women who kept their surname were judged by both sexes as independent, feminist, young, self-confident, and well educated, but also unattractive. Male respondents thought them more likely to work outside the home and less likely to enjoy cooking, less likely to attend church, and less likely to make good wives. The results showed stereotypical perceptions of women who made the choice to keep their birth surname after marriage. Not all of these qualities are positive. As these views show, being career oriented precludes being family oriented. Women cannot be both, it seems. Moreover, being career oriented, for many people, entails being non-religious, not good at cooking, and, alas, not good wives! Women’s surname choices reflect changing societal expectations for the roles that we should play. But do these reflect an increased independent identity and societal power for women? The historical shifts and their underlying causes are worth deeper exploration. As a linguist interested in gender and language, I was left with the nagging notion that there is much more involved in the surname question than meets the eye.

Waves of Feminism in the U.S.
Surname choices over the years reflect societal trends toward women’s rights. First wave feminism, begun in the nineteenth century and extending into the early twentieth century, focused primarily on legal rights for women such as suffrage. Second wave feminism began in the early 1960s and extended to the late 1980s. This period was concerned with political and social rights for women that went beyond the legal domain. That is to say, second wave feminists were concerned with the oppression of women through power structures of society that are inherent in daily life. During this second wave, women’s consciousness was heightened about issues of their economic and social status in society. This period saw an increase in surname retention by women. Third wave feminism has existed since the early 1990s, and focuses less on political and social equality and more on taking account of the diverse characteristics of women of all races, ethnicities, and sexual orientations. For third wave feminism, the strides made by prior generations of feminists are taken as a given by most women. First wave feminism seemed to have little impact on names as social labels in the



U.S. In the 1970s, second wave feminism led to widespread shifts in behavior, including the choice made by many women to keep their birth surnames. Now, in the midst of third wave feminism, we see that women are making more traditional naming choices. This is somewhat contradictory to other recent social movements that are currently challenging and sometimes changing civil law regarding sexualities and marriage practices. Nowadays, U.S. women have the legal right to retain their surnames when they marry. In fact, in some states it is the default to keep one’s name unless explicitly stated otherwise. The issue of surname retention is no longer a legal struggle, and has not been for decades. Civil laws regarding marriage practices are now the foci of struggle. Third wave feminists can take the social labeling practice of surnames as a given legal right. Other issues have become more timely and critical. However, lest we become too complacent about naming, we must still reflect on what a return to taking husbands’ surnames really reflects in gender equality. While surnames are symbolic, they have to do with power structures that are difficult to topple. With this in mind, let us examine the topic more closely.

The Initial Study
Given the historical pendulum swings discussed above, it seemed high time to dig more deeply in order to find out more on the surname issue and what these choices mean for women in their communities. I started with my own students. During a class discussion on social labeling practices in a course on Gender and Language, I asked the question, “What will you do about keeping or changing your surname when and if you marry?” Bear in mind that these students, totaling 21 (18 women and 3 men), were taking the course not as a requirement, but out of genuine interest in the subject.2 Among the 18 female students, only one of whom was married and had just taken her new husband’s surname, 11 said that they would take their husband’s surname. Three of the women said they would hyphenate, and 3 said they would use both names. Only one graduate student specializing in Gender and Language said that she would keep her birth surname. The majority of the students reflected strong sentiments about wanting to take their future husbands’ surname, as the following quotation illustrates:
Ever since I have been a little girl, I’ve been dreaming of the day I change my name to my future husband’s name, and I still feel that way. To me it symbolizes the two of us becoming one family, and a commitment to that man. No one ever starts a marriage with the anticipation that it will end in divorce, so by taking on his name it symbolizes this commitment.

There were only three male students in the class, and one was from Korea. He indicated that Korean women keep their birth surnames upon marriage. The other two males, both gay, said that if they marry or form some sort of union they plan to either blend their names or hyphenate. A New York Times Magazine article (Denizet-Lewis, 2008) mentions two examples where gay men were also involved in surname choices. In one case the man took his partner’s last name and in the other the men hyphenated their two surnames. There is a relative paucity of data on this subject regarding same-sex couples. As we see from the two gay men in my Gender and Language class, unlike the young women involved in heterosexual relationships, these young men indicated that their surname choices would not entail dropping either birth name.

Gender, Language, and Power


Exercise 2

Think about your own inclination to retain, keep, use two names, or hyphenate your surname if/when you marry. What would you expect of your spouse? Ask some of your friends to tell you their thoughts. Write down their responses and compile their stories. What patterns do you find?

All of the students in that class, without exception, said that their mothers used their fathers’ surnames, sometimes with their birth surname as a middle name. Many students said that despite the fact that their mothers were divorced, they retained the surname of their ex-husbands. Some of the students’ views indicate their desire to continue this naming tradition. As a whole, their comments indicated conflicting social expectations. For example, Jessica, single and 19 years old, talked about not wanting to lose her surname but at the same time wanting to show solidarity with her future husband. Elizabeth, single and 20, said, “I’m not sure I like the fact that women completely lose their identity and men do not, but I would like to share the same last name with my husband. I don’t think my boyfriend or society is ready for him to take my last name.” Kira, a 21-year-old bisexual woman, clarified the difficulty of the naming issue for her, depending upon sex of partner. She stated that she would want both her partner and herself to take on a new name along with their old ones. She also said, “If I marry a male, this may be a little more difficult to sell.” How rare is it for men in heterosexual unions to take a hyphenated name or to consider taking a wife’s surname? While women state that taking their partner’s surname is symbolic of their “union,” few men in heterosexual partnerships have such sentiments. Why is this still so? Why do even young women who hold feminist viewpoints look forward to taking a husband’s or partner’s surname? Why do women do the “merging?” The students’ responses stunned me. These were thoughtful young women who had been influenced by the advances of second wave feminism. One might expect young women to eagerly embrace the idea of keeping their names, but they did not. Why? There exists a conflict between the concept of surname as symbolic of family union and women’s struggles for independence and equality. Even now, becoming part of a family entails, for women at least, the desire for affiliation and symmetry. Sociolinguist Deborah Tannen, in her work in Gender and Conversational Interaction (1993), was one of the first to discuss the complexities of female/male discourse. She shows that concluding whether power or solidarity (or both) is being expressed is highly dependent on the context of the discourse. In other words, the power/solidarity dichotomy is not always clear cut. One must carefully analyze what is being said, by whom, and in what kind of situation. Nonetheless, solidarity in general is consistent with what we know about women’s ways of talking, doing, and knowing. Sociolinguistic research has consistently shown women to be interested in establishing rapport or solidarity with others. Indeed, throughout the last thirty-five years, since a surge in research on gender and language, studies have found that while men’s ways of talk tend to express hierarchy (see, for example, Maltz and Borker, 1983; Tannen, 1990), women’s talk concentrates on the expression of symmetry. If this is true, it seems clear that women do the merging, since it is consistent with symmetrical structures. But because women are the ones who merge, male hierarchy wins out. It brings up yet again the issue of traditions that condone and perpetuate hegemonic structures, and the conflict with feminist discourse over the past forty years (Cameron, 1998).



As for women always doing the “merging,” many studies have borne this out. In the early 1980s, for example, Pamela Fishman tape-recorded women and their partners in conversation at home. Her findings showed that the women did most or all of what she called “conversational shitwork.” They worked hard to get their male partners to talk to them, using such linguistic strategies as tag questions (e.g. “That was an exciting movie, wasn’t it?”) in order to force the men to take a substantive turn in the conversation (Fishman, 1983). A follow-up study almost a decade later (De Francisco, 1991) had similar findings. After polling my students, I wanted to widen the research. In 2004, I interviewed a dozen highly successful adult professional women to get their views on the issue of merging, affiliating, and surnames. These were women with an average age of midforties who all had advanced degrees (M.A., Ph.D., or equivalent). Such women are the ones we would expect to be the most enlightened in terms of reaping the benefits of second wave feminism. Among them, three had changed their surname when they married and were satisfied with the choice. One chose to hyphenate, resulting in a very difficult, multi-syllabic name (her husband never thought to hyphenate). Two had changed back from their ex-spouse’s surname to either their birth surname or another name chosen strategically. Six, or half of these adult professional women, had kept their birth surnames.

The Larger Study
I then embarked on a larger project that was both quantitative and qualitative in nature. A quantitative study uses numerical data to conclude findings and requires large amounts of data. A qualitative study uses data that are not measured by numbers but rather obtained from what people say, such as their stories or texts. Both approaches together offer depth and breadth to a research project. The data presented here are gleaned from 134 women surveyed in the U.S. and 103 in Russia. In both countries the data reflect women’s responses in various regions of the countries and among varied age groups and educational levels. The samples are not intended to be representative of either context; they are meant to provide a backdrop for women’s stories and comments. In both the U.S. and Russia, narratives were derived from quasi-ethnographic oral interviews (12 U.S.; 7 Russia) and extensive written comments on the questionnaires. These highlight some of the salient issues for women in choosing surnames upon marriage or partnership.

Quantitative Data on Women and Surnames in the U.S.
Table 9.1 reports the findings.
Exercise 3 Before you read the data in Table 9.1, cover up the numbers and try to predict the results. What is your prediction of surname choice for respondents by age, religion, educational level, and region of the U.S.?

Unlike the original dozen professional women interviewed, where half had retained their surname, the quantitative data showed different numbers. Of 134 married women who responded to a short questionnaire, 108 had changed their surname and only 24 had retained it. There was only one who had returned to her birth surname, and one used a hyphenated name. A very interesting finding was that 17 of the 20 divorced

Gender, Language, and Power TABLE 9.1: U.S. Data, 2004 (Key: NR = not recorded) U.S. Total Husband’s Name Birth Name Hyphenated



1—Marital Status Married Divorced Widowed Single Separated Partnered Total 2—Age 40–55 20s 30s 55–65 65–75 over 75 NR Total 3—Religion Christian Jewish Muslim Buddhist Mormon Hindu Other None NR Total

134 20 6 8 3 3 174 63 47 34 14 8 2 6 174 122 6 2 1 5 1 3 34 174

108 17 6 2 133

24 1 8 1 3 37


1 2



52 30 26 14 8 2 1 133

8 17 8



4 37


1 3

98 5 1 5 1 23 133

21 1 1 1 1 2 10 37 1 1



4—Education at the Time of Marriage High school 54 44 B.A. 89 71 M.A. 22 11 Professional degree 4 2 NR 5 5 Total 174 133 5—Region Northeast South Midwest West Foreign NR Total 38 92 18 8 15 3 174 35 75 10 3 7 3 133

8 17 11 1 37

2 1 1 1 3

2 14 8 5 8 37


1 2





women had kept their former husband’s surname after divorce. This is consistent with what the preliminary data revealed about the majority of the mothers of the 21 students in the Gender and Language course. So, regardless of the age group (from twenties to over 75), the vast majority of women had taken their husbands’ surnames. This finding directly reflects the strength of existing traditions in taking on a man’s family name.

Exercise 4

Calculate the percentages of women who retained their surname in all of the age groups above. Are there any striking differences?

Qualitative Data: Stories about Surnames
Most of the women in this study indicated their desire to share one family name as the main reason for changing. The desire for affiliation with the new family was especially important for women who married young. Anna, married for thirty-two years, had this to say:
Taking on my husband’s last name was an outward sign of our union. It served to make me feel that I was ‘really married’ (I was 17) and that we were forming a brand new family together, and that I had a new identity as his wife, by law. Much time and education later, I came to feel that I had also lost a little of myself and my heritage when I changed to my husband’s name when we married.

Surname decisions are determined by language as much as by culture. Notably, women themselves often use language-related terms, such as “sign,” “symbol,” “meaning,” to account for their naming choices (“It is a sign of unity and solidarity”; “It is a symbol that we have created a new family”). Some women mentioned that one surname made their union “official” and “more complete”; some emphasized emotional satisfaction in their choices (“I enjoy having my husband’s last name”; “I was pleased to take my husband’s name. He is the ‘head’ of our household”).

Surnames and Children
The stories repeatedly indicate that the basic understanding of roles of women remains: as wives and mothers and in relation to other family members. In fact, the most important part of family unity for the women who participated in my study was the issue of surnames for children. The overwhelming number of respondents and interviewees discussed this problem in their stories. Connie, a woman in her thirties married for three years, talked about the possible confusion caused if she had a different surname from her children’s. Many others expressed sentiments consistent with the hassles of more than one surname in any family with children. Several women said that even though they didn’t change their surname when they married, they plan to do so when they have children. Donna had the following to say:
I haven’t changed my name since marriage, but my plan is to do so before children are born. Both of us plan to change our names, to varying degrees. Our plan is for one of our last names to become a middle name for both of us, the other to become the last name for both of us.

Gender, Language, and Power


While many recently married women said that they are considering giving their children their birth surname, the reality is that few families do this. Their decision consistently favors choosing the man’s surname for children. Indeed, when women do retain their birth surname, it results in their having a different surname from that of their children. Family unity was a constant thread in stories among women who had their husbands’ surnames, even after divorce, particularly when children were involved. Enid, recently divorced, said that she did not change back to her birth surname because of the children. She felt consistency with their names to be of primary importance. This is a condition of many divorced women in U.S. society now. The “hassle,” it seems, is to change back and to take a surname that differs from the rest of their nuclear family. While this sentiment was expressed repeatedly, it is noteworthy that many women in U.S. society continue to adopt a new husband’s surname upon remarriage, even though it changes their surname to one that differs from that of their children. Fran, just turned 40, related that she married at age 15 and a few months later had a baby boy. She had taken her husband’s name. Even when they divorced after ten years, she kept her husband’s name, since her son used it. Several years later she remarried. At that point, Fran said, “There seemed no need to keep my first husband’s name, so I took this new surname.” They divorced shortly afterward, but later were remarried. The second time Fran didn’t take his name, since it had been so hard to change back to her birth surname after the divorce. She said, “Who knows? Maybe I had a secret feeling it wouldn’t work again.” This naming story brings up several issues. First, Fran stated that she retained her first husband’s name after divorce, since her child used this name. She then took her new husband’s name when she remarried, even though her son would now have a different surname. She went to the trouble of changing back to her birth surname after the second divorce. Despite all of these changes, she indicated in the interview that if she married again she would probably again take her new husband’s surname! Power is directly reflected in surname changing in marriage. In divorce the man cannot influence the woman’s choice—she is free to return as a result of a process of differentiating herself from him, or to continue to use the surname of her children. That is, she may elect to stay with the surname that identifies her as a mother rather than an independent woman. In a new marriage, men’s power to influence this decision is at work once again.

Exercise 5

(a) Find out what is involved in legally changing a surname in your state. What is the procedure and what is the cost? Is there a legal procedure when a person is getting married? Is there another legal procedure when getting divorced and returning to a birth surname? Is the procedure identical for women and for men? (b) In what aspects of your life would your legal name make a difference?

Identity and Power
Among the large number of interviewees who did choose to adopt their husband’s surname, identity was repeatedly discussed as a pivotal issue in their stories. Some were reluctant or regretful about having to give up symbols of their ethnic identity (“I wish I had kept my mother’s maiden name. I’m from an Italian background. I identify with that name more than my husband’s name or my father’s”). Stories reflecting eagerness



to shed ethnic identities, revealing what the women perceived as identity stigmas were fewer. Recall that in the in-depth interviews with highly educated, professional women in the U.S., half were retainers. The general sentiment was that they had never considered making a change. A recently married academic in her thirties stated that, despite having retained her surname, her friends and relatives often address her by her husband’s surname in formal contexts. The stories are mixed. Those who took their husbands’ names and are satisfied with the decision mentioned: • satisfaction in affiliation: “I have become part of one more loving family”; “One surname is like a symbol of our union”; “Changing my surname gave me a feeling that we have become closer to each other.” • shaping the new identity: “It must have influenced my self-concept . . . in ways unknown to me it shaped the me I am today”; “A new name was a sign of my new status (married) . . . it helped me and others (my parents, friends) to realize the change and marked the beginning of a new phase in my life.” Women who took their husband’s names and are not satisfied referred to: • sense of lost identity: “To some extent, the new surname changed my personality and my thoughts. They became more like my husband’s”; “It seems to me sometimes that I have lost a part of myself”; “I miss my maiden name and have been missing it all my married life—34 years.” • Dissatisfaction with inconvenience: “It added confusion to my life . . . people do not ‘recognize’ me under my new surname”; “I feel embarrassed by my new name”; “I associate my new surname with my husband’s relatives, whom I dislike”; “It added problems—change of a passport, other documents at work.” All women who retained or returned to their birth surnames indicated that they are satisfied with their choice. Recall that there were a few women who chose to legally change their surname from their father’s to their mother’s or grandmother’s birth surname. These women indicated that they either had no relationship or a bad relationship with their father, and thus they would never consciously choose to bear his surname. An underlying issue is what it means to be a retainer or returner when retaining or returning is to one’s father’s surname. I have several anecdotes of respondents who, having a newly raised consciousness about women’s issues, have taken on the surname of a female foremother or a historic or mythical woman. For example, one woman mentioned in discussions about this topic indicated that she had legally changed her last name to Cerridwen, the Celtic goddess of the moon.
Exercise 6 Choose one type of media (e.g. TV, radio, Internet) and listen for how women refer to their own surnames. Do you find a lot of double/hyphenated names?

Surnames in Russia vs. U.S.
The large research project I discussed above continued after the initial U.S. data collection. The larger study across two societies, the U.S. and Russia (Boxer and Gritsenko, 2005), found many similarities in reasons for such choices articulated by women from both countries. However, the narratives reflected different discourses. American women spoke about (not) being owned, (not) losing their identity, and about having an equal

Gender, Language, and Power


partnership. These responses reflect traces of the prevailing feminist discourse in the U.S. of the past several decades. The Russian discourse was more about personal relationships and patriarchal cultural traditions that persist due to the relative absence of feminist discourse in Russia during Soviet times. Before the October Revolution of 1917, Russian women were required to take their husband’s name and retain it all throughout their married lives and even after divorce, should the marriage be legally dissolved (Legal Code, Decrees of Peter the Great and the Holy Synod, supreme governing body of the Russian Orthodox Church). Later in the century, with the advent of the Soviet Union, the law provided for women and men to have the choice of taking either spouse’s surname or keeping their birth surnames. The option of taking a hyphenated name was not provided in the Soviet Union until the current Civil Code of the Russian Federation (passed in December 1995, in effect since March 1996). Legally, then, the opportunity for women to keep their birth surname, hyphenate both names, or for both spouses to take the wife’s surname has been available for many years. In reality, however, choices have been governed more by “the law of tradition.” The Russian narratives directly reflect women’s lack of autonomy, providing examples of how surname choices result from external (or internal) encouragement or discouragement.

The Russian Qualitative Data
Table 9.2 provides data from Russian respondents. The Russian respondents who took their husbands’ surnames after marriage did so due to a long-established tradition. Over 50 percent of women of all age groups stressed this as the main reason for their choice. The word “tradition” was not always mentioned, but usually was implied. These excerpts are all translations from Russian:
“It was always done this way.” “Such is the tradition.” “This is the way to do it here” (in Russia).

A portion of the respondents (about 12 percent) said that they did not even think about whether to change or keep their surname, the decision coming “naturally”:
“I don’t remember actually making my choice, it came as part of the ritual of getting married and was taken for granted by both my husband and me.”

Some answers clearly indicated that respondents did not feel they had the option of keeping their surname. These answers were given by both older women (e.g. “They did not ask you then whether you want to keep your maiden surname”) and younger women (e.g. “I made my choice obeying the tradition”). One respondent who answered “Yes” to the question of whether she was satisfied with the choice of surname for herself added, “but if I could, I would have kept my own.” This and similar answers indicate that what seems to reflect a free and individually determined choice may in fact reflect gender in-group identification and beliefs about the capabilities of women as a social group (cf. Nosek et al., 2002).

Surnames across Cultures
It is well known that there are vast cross-cultural variations on the theme of women, marriage, and surnames. Name changes for women at marriage have been found to be


Boxer TABLE 9.2 Russian Quantitative Data (NR = not recorded) Russian Total 1—Marital Status Married Divorced Widowed Single Separated Partnered Total 2—Age 40–55 20s 30s 55–65 65–75 over 75 NR Total 3—Religion Russian Orthodox Other None NR Total Husband’s Name Birth Name Hyphenated Returned

92 6 5

81 3 2

8 3

3 3

103 35 28 28 8 4




30 25 23 6 2

2 3 3 1 2

3 2 1

103 73 5 25 103




64 4 18 86


2 1 3 6

4 11

4—Education at the Time of Marriage High school 17 15 B.A. (equivalent) 41 38 M.A. (equivalent) 37 29 Professional degree 8 4 NR Total 103 86

1 2 5 3 11

1 1 3 1 6

more prevalent in technologically more complex societies. However, some changes have occurred toward the norm of taking husbands’ surnames even in less industrialized countries. Following are some examples.

In Sweden, the attitude toward women who retain their surname has become more accepting, even among males (Trost, 1991). In Norway women seldom take their husbands’ names any more, though it was legally required from 1923 until 1961. Moreover, since a law enacted in Norway in 1980, children automatically receive the mother’s surname unless the couple informs the authorities otherwise (Romaine, 1999).

Gender, Language, and Power


Even more clear cut is the Civil Code of Quebec, which states that in marriage both spouses retain their last name and exercise their civil rights under those names (Duggan et al., 1993).

Hispanic Cultures
Different languages have different ways of representing hierarchy in marriage names, showing women as the affiliating parties. Typically, in Spanish-speaking communities, married women are referred to as “Señora de (husband’s surname).” This is probably the most telling example of the action of affiliating. De here means of, indicating possession and suggesting—at least from the point of view of linguistic form—that a woman is the property of her husband (P. F. de Mola, personal communication, July 2004). Across the world, in legal terms western societies are becoming more “liberal” about surname retention after marriage. Nonetheless, western naming traditions continue to influence indigenous cultures. In some parts of the world where surname change was not previously assumed, it has become an issue when marriages are officially registered.

In rural areas of Pakistan women still do not change their names after marriage unless they face the need to obtain an official document or enter into some other type of communication with government authorities (Z. B. Sarwar, personal communication, July 2004). When this occurs, a woman’s “full” name is required. In Pakistan this will be her first name plus the first name of her husband; for an unmarried woman it will be her first name and the first name of her father or the nearest male relative. As Sarwar pointed out, in Pakistani society women from westernized urban families take their husbands’ names far more often than women in remote villages.

Originally the concept of a surname did not exist in this West African country. Women and men had their own names that were not changed after marriage. Surnames themselves and the tradition for women to take their husbands’ surnames after marriage are the result of western influence and Christian traditions (A. A. Ampofo and A. Darkwah, personal communication, July 2004). In fact, a husband’s surname and title “Mrs.” has become a class marker. In rural areas, indigenous women are more likely to keep their own names and not change them upon marriage. This is similar to naming patterns in Pakistan (where rural women do not change surnames) but a clear reversal of the situation in the U.S. The commonality to all these societies is that men rarely think of changing their names.

So, where do we go from here? We’ve seen reflections on the issue by women of varied ages and educational levels across societies. The results of these studies on surnames and women inform not only the field of linguistics, but also sociology, anthropology, political science, women’s studies, and psychological research on language and gender. Clearly, gender cannot be abstracted from other aspects of social identity. In making naming choices, women do not act only as women, but also as mothers or future



mothers, professionals, representatives of certain ethnic groups, regions, religions, and other social categories. Gender identity issues are always difficult to separate from women’s complex forms of participation in their communities. The meanings behind naming choices may vary. Surname options often indicate compliance with existing power hierarchies that reflect loyalty to tradition. Since so many respondents indicated that changing their surnames was symbolic of their “union” with their husbands, there exists a conflict between the desire for familial solidarity and the desire to resist traditional male power structures. For many women, the act of name change is a conscious one of building a new married identity. What is below the level of consciousness is what these conscious acts reflect.

1 Wedding invitations are now increasingly referring to women by their husbands’ surname as well as his given name. One that I just received listed the maternal grandmother by her late husband’s first and last names, despite the fact that he died some thirty years ago. In other words, this woman did not even have the honor of having her own first name listed. She was “Mrs. John Smith.” The mother of the bride appeared on the invitation as “Mrs. James Jones,” even though her husband is not the father of the bride. Thus, the bride’s mother’s identity was completely hidden. The only people who got true name billing on the invitation were the men: the groom and the bride’s father. 2 Readers of this textbook may have more traditional inclinations regarding surname changing upon marriage. However, the class that sparked this research was a self-selected group that was particularly tuned into gender and feminist issues.

Boxer, D. and Gritsenko, E. (2005) Women and surnames across cultures: Reconstituting identity in marriage. Women and Language, 28(2), pp. 1–11. Cameron, D. (1998) The feminist critique of language: A reader. 2nd edn. London, Routledge. De Francisco, V. L. (1991) The sounds of silence: How men silence women in marital relations. Discourse and Society, 2(4), pp. 413–423. Denizet-Lewis, B. (2008) Young gay rites. New York Times, April 27, p. 19. Duggan, D. A., Cota, A. A., and Dion, K. L. (1993) Taking thy husband’s name: What might it mean? Names, 41(2), pp. 87–102. Fishman, P. M. (1983) Interaction: The work women do. In Thorne, B., Kramarae, C., and Henley, N. eds. Language, gender, and society. Cambridge, Newbury House, pp. 89–101. Goldin, C. and Shim, M. (2001) Making a name. NBER Working Paper No. W8474. Cambridge, Harvard University Department of Economics and National Bureau of Economic Research. Available at: id=283346 [Accessed June 14, 2009]. Intons-Peterson, M. J. and Crawford, J. (1985) The meanings of marital surnames. Sex Roles, 12(11/12), pp. 1163–1171. Kotthoff, H. (2000) Gender and joking: On the complexities of women’s image politics in humorous narratives. Journal of Pragmatics, 32(1), pp. 55–80. Kupper, S. J. (1990) Surnames for women: A decision-making guide. Jefferson, McFarland. Le Page, R. B. and Tabouret-Keller, A. (1985) Acts of identity: Creole-based approaches to language and ethnicity. Cambridge, Cambridge University Press. Maltz, D. N. and Borker, R. A. (1983) A cultural approach to male–female miscommunication.

Gender, Language, and Power


In Gumperz, J. J. ed. Language and social identity. New York, Cambridge University Press, pp. 195–216. Miller, C. and Swift, K. (1976) Words and women. Garden City, Anchor Press. Murray, T. E. (1997) Attitudes toward married women’s surnames: Evidence from the American Midwest. Names, 45(3), pp. 163–183. Nosek, B. A., Banaji, M. R., and Greenwald, A. G. (2002) Math = male, me = female, therefore math ≠ me. Journal of Personality and Social Psychology, 83(1), pp. 44–59. Pujolar, J. (2000) Gender, heteroglossia and power: A sociolinguistic study of youth power. Berlin, Walter de Gruyter. Romaine, S. (1999) Communicating gender. Mahwah, Lawrence Erlbaum. Scott, K. D. (2000) Crossing cultural borders: ‘Girl’ and ‘look’ as markers of identity in black women’s language use. Discourse and Society, 11(2), pp. 237–248. Tannen, D. (1990) You just don’t understand: Women and men in conversation. New York, William Morrow. Tannen, D. (1993) Gender and conversational interaction. New York, Oxford University Press. Trost, J. (1991) What’s in a surname? Family Reports, 19, pp. 1–24. Wiley, N. (1994) The semiotic self. Chicago, University of Chicago Press.

Other Resources
Professional Organizations
Organization for the Study of Communication, Language, and Gender: http:// Journal, Women and Language. International Gender and Language Association (IGALA): organisations/igala/Index.html. Journal, Gender and Language. American Name Society: Journal, Names.

Carole E. Chaski
This Chapter Explores:

Linguistics as a Forensic Science The Case of Author Identification

How to Study Language and Law Language in the Law Forensic Linguistics with a Focus on Author Identification When the Law Needs Forensic Linguistics Crimes, Civil Disputes, and Security Issues Case 1: The Mother, the Daughter, and the Daughter–Mother Documents Case 2: The Teacher’s Love Letters Case 3: The Hacked Profile on

How Forensic Linguistics Applies Linguistic Theory and Method Linguistic Tools for Forensic Problems Different Methods for Author Identification How Forensic Linguistics Functions as a Science Validation Testing Possibility of Linguistics as a Forensic Science Legal Standards for Admissible Evidence How Linguistics Can or Cannot Meet These Standards

It is the responsibility of intellectuals to speak the truth and expose lies. Noam Chomsky

Linguistics first attracted me because I wanted to find an objective and repeatable method for analyzing language. I had majored in English and Ancient Greek, so I was familiar with literary methods, and then earned a Master’s in the Psychology of Reading, so I was also familiar with pedagogic methods. When I finally found psycholinguistics and computational linguistics, I knew I had found an objective and repeatable method—a science of language at last. The first experiment I designed in psycholinguistics investigated how syntactic and semantic paraphrases of text affected professors’ grading: whether stylistic shifts in sentence structure or vocabulary would predict grades (yes, they did). My first project in computational linguistics was modifying a parser so that it could process really natural natural language, the kind of language that people actually produce, with “grammatical errors,” misspellings, abbreviations, etc. (yes, with lots of modifications). Both of these projects influenced my approach to forensics. I was teaching summer-school linguistics to blear-eyed

Linguistics as a Forensic Science


engineers and technical writers at North Carolina State University, wondering if and praying that there was something useful I could do with linguistics in the real world. The telephone rang. It was Detective W. Allison Blackman of the Raleigh Major Crimes Unit, asking me if there was any way to determine, just from the language itself, who had written a suicide note left on a home computer. There was no ink, no paper, no handwriting, just the language, with three potential authors. Did I think I could do something as a linguist? (Now that was an answer to a prayer, an invitation to use linguistics in the real world.)

Language and law pervade human experience. Just about any human experience can be examined by studying either the linguistic or the legal aspects of it. One might even venture to say that language and law are fundamental to what it means to be a human being. Both law and language provide vast opportunities, but this chapter can only introduce the reader to one small prospect; to explore the case of author identification. In order to put author identification in context, this chapter presents (1) three ways of studying language and law, (2) scenarios in which the law needs linguistics as a forensic tool, with three particular cases of author identification highlighted, (3) examples of how forensic linguistics applies linguistic theory and methods, (4) procedures for validating forensic linguistics methods, exemplified by authorship identification, and (5) the possibility of linguistics as a forensic science, given current legal standards. As we will see, any method used needs to stand up to the criteria of validity and reliability. Validity means that one’s findings are genuine: what you think they represent they actually do represent; reliability means that the results would turn out the same with repeated testing.

Studying Language and Law
Language and law intersect in two ways: how language is used in legal processes, and how language serves as evidence in legal processes. By focusing on language in law, the linguist is typically an observer of legal processes in the same way that linguists observe child language in home or educational settings. In the second focus on language patterns as evidence, the linguist is a direct participant, as a forensic scientist—a forensic linguist—in the investigative or legal processes. This distinction is maintained by the International Association of Forensic Linguists (IAFL), whose website states that the IAFL
is an organization which primarily consists of linguists whose work involves them in the law. Narrowly defined, this means linguistic evidence in court (authorship attribution, disputed confessions, etc.), but in fact the association aims to bring together those working on any aspects of language and the law. (International Association of Forensic Linguists, 2009)

This distinction is also maintained by Roger Shuy, who has
classified this work that forensic linguists do into two types: work that is done without becoming involved in specific litigation, which I call outsider work, and work that is carried out within individual law cases, which I refer to as insider work. Both types of work are important. Both advance knowledge. Both advance the field of linguistics into


Chaski another significant area of life. But working from the outside can be the more comfortable, more academic way. (Shuy, 2000)

A third approach, one which has been embraced by The Association for Linguistic Evidence (TALE), is for linguists to produce research on some law-related linguistic issue and then to provide services based on their research on that language-related legal issue. TALE is
a professional association for researcher-practitioners, the legal profession, and law enforcement/investigators whose purpose is the development and use of validated methods for using language analysis as reliable and potentially admissible evidence in criminal, civil and security cases. (Institute for Linguistic Evidence, 2008)

In this third approach, the research done outside the courtroom relates directly to the expert testimony inside the courtroom. In fact, this approach allows for a symbiotic relationship between language and law, where answers to questions researched in linguistics can become helpful to lawyers, and issues raised in the law can become research questions for linguists. One overarching research question for linguists is how language is used to codify, interpret, and adjudicate laws. As part of this research agenda, linguists have studied the following: • • • • • • • • • • • legalese; the comprehensibility of jury instructions; judicial language; the plain language movement; police interviewing tactics; multilingualism in trials; standards for court interpreting; language and power in courtroom strategies; perjury; language policy; linguistic rights.

Levi (1994) and the International Association of Forensic Linguists website ( provide extensive references to work in these and many other topics in language and law. Many of these topics could be or already have been the subject of expert testimony. For instance, linguists have studied the comprehensibility of insurance policies, pension plans, and jury instructions as a research question, and linguists have also brought expert opinion to bear on this issue in pre-trial investigations and sometimes in trial. As forensic linguistics matures into a forensic science, the same linguists who are producing litigation-independent research will serve as the linguists who bring this research-based knowledge to bear on litigated issues.

It is obvious to any non-lawyer that legal language differs markedly from both spoken language and other genres such as newspaper journalism, academic textbooks, fiction, and even scientific articles. The term legalese identifies legal language. The term legalese first appeared in print in 1914, according to the Merriam–Webster’s Collegiate Dictionary,

Linguistics as a Forensic Science


10th edition (2002); but the recognition that legal language is different goes back to early English records. Mellinkoff (1963) showed that English legal language has a different vocabulary, while Tiersma’s (1999) historical description of language and law traces that vocabulary to the Roman conquest of Britain and Norman invasion of Britain in 1066. English legalese is extremely Latinate, whether from the Roman or the French influence. One of the most interesting facts about Early English law is that it was conducted in French, rather than English. Baugh (1935: 182–183) notes in his classic history of the English language how legal proceedings, first recorded in 1310, were only sporadically conducted in English until finally the Statute of Pleading enacted in 1362 legislated that legal proceedings would be conducted in English and recorded in Latin.
Exercise 1 Write a reasoned opinion essay on the following topic. Include two opposing viewpoints with support for each. Before you begin, conduct Internet searches to determine the status of English as the official language in your state and the use of court interpreters. Should laws in the United States be available in languages other than English?

Given the prevalence of legalese, how do non-lawyers fare in understanding legal language? Some characteristics of legalese are syntactic complexity and semantic qualification. Syntactically complex language has many clauses with items referring to items in other clauses, while semantic qualifiers such as “if,” “unless,” “notwithstanding,” “in the event that,” “the foregoing,” “hereof,” “herein” can make the meaning very precise, but difficult to comprehend since information is densely packed. Meanwhile, statements including multiple negation or words with negative content such as “without,” “avoid,” “neglect,” and words prefixed with mis- or un- can become especially confusing. Therefore, Charrow and Charrow (1979) initiated a long line of subsequent research focusing on whether jurors comprehend legal language and how jury instructions could be made more comprehensible. As experts in the law, and experts in legalese, do judges process and use language in consistent and particular ways? Solan (1993) is one of the earliest works on language and law, focusing on the language of judges. Solan (1993) shows, for instance, how judges handle certain types of ambiguity (where a word, phrase, or sentence can have more than one meaning), how the meaning of words can determine constitutional rights, and how judges talk to juries in instructions. The “plain language” movement prescribes how legal language should be used to communicate to lawyers and non-lawyers as plainly as possible. One linguist-lawyer, Peter Tiersma, has drafted such linguistic guidelines, available through the National Center for State Courts at
Exercise 2 Perform an Internet search for the laws of your state. Typically state codes are available through your state’s dot gov website. Select a portion of the code and translate, as best you can, into non-legalese “plain” English. Compare your answers. If your state laws appear to be comprehensible to you (as a result of the plain language movement), visit Adam Freedman’s website http://, which lists examples of legalese. Using examples you find most interesting, discuss the relationship between the legalese and the translation into plain English.



The second way that language and law intersect—and the focus of this chapter—is that some legally significant situations contain evidence which itself is linguistic. This is called forensic linguistics. In forensic linguistics, the linguist applies standard linguistic analytical methods to answer forensically significant questions that have arisen in a criminal or civil case.

The Need for Forensic Linguistics
In both criminal and civil cases, language itself can serve as evidence.
Before reading further, compile a list of all the ways you can think of that language can be evidence in a trial. Compare answers.

Exercise 3

There are some text types that instantly come to mind as potential language evidence, such as ransom notes, bank robbery notes, threatening letters, and suicide notes. Some crimes are verbal by definition, speech itself that performs criminal behavior, such as slander and perjury. Roger Shuy (1993, 1998, 2002) has written about his work on bribery, threatening, perjury, conspiracy, trademark disputes as well as Miranda rights, and false confession. Shuy is especially well known for his discourse analysis of consent in conversations taped during surveillance of drug deals and other sting operations. Hannes Kniffka (2007) has documented his work on defamation, libel, slander, extortion, and blackmail. Kniffka has brilliantly shown how dialectal features and patterns of linguistic interference can narrow a list of suspects. In my practice since 1992, linguistic evidence has appeared in cases involving: • • • • • • • • • • • • • • • • • • • • homicides; suicides; kidnapping; bombing resulting in death and destruction of property; solicitation to murder; terroristic threatening; sexual assault and sexual exploitation of minors; hate crime; obstruction of justice; theft; habeas corpus; copyright infringement; patent infringement; trademark infringement; defamation; employment disputes; custody disputes; inheritance disputes; plagiarism; discrimination.

Text types which we would usually not associate with criminal activity—blog posts, corporate emails, love letters, and diaries—have also served as linguistic evidence. In several cases involving blog posts, authorship identification connected a person to a

Linguistics as a Forensic Science


specific computer keyboard. In several white collar crime cases, authorship identification demonstrated that a person had authored a document revealing knowledge known only to the perpetrator. In several cases involving sexual crimes, authorship identification provided evidence about texts which revealed crucial relationship information. Three actual cases from my casefiles show the diversity of text serving as linguistic evidence.

Case 1: An Inheritance Dispute
In a civil case in Virginia, I was asked to determine the authorship of several letters allegedly authored by an aging mother to her four children. In these letters, the mother revoked her earlier and well-known decisions regarding her estate and bequeathed all of her property to one daughter, effectively leaving the other children with no legacy. The property included holdings in both Lebanon and the United States, with acreage attached to a waterfront family home in Tidewater, Virginia. Three siblings did not believe that their mother had authored the letters that effectively disinherited them, while the one sibling to whom all the family property was to be bequeathed held to the claim that their mother had indeed authored the questioned letters. The mother had lived with this one sibling prior to her death. During this time period, the daughter had sometimes drafted documents that her mother had copied into her own handwriting and sent to the other children. This fact was proven because drafts in the daughter’s handwriting of the mother’s later letters were found among the documents produced after the three siblings disputed the authenticity of the mother’s letters. Because the daughter had authored documents for her mother, which her mother had then copied by her own hand, there was the potential that this scenario had also been played out with the questionable inheritance letters. The issue, then, was whether the mother had actually authored the letters or not. There were three potential authors of the questioned letters: the daughter, the mother, or the daughter-as-drafter-with-mother-as-handwriter. Without forensic linguistics, this issue was essentially a matter of “she said, they said,” with one sibling claiming that the mother had written the inheritance letters, and the other siblings claiming that she had not.

Case 2: Endangerment of a Minor
In a criminal case in North Carolina, I was asked by a defense attorney to determine the authorship of love letters that had been left on a school computer. The defendant, a female teacher, was accused of having a sexualized relationship with a male student, during which the love letters were composed. The male student had admitted to writing some love letters and alleged that the teacher had written the computer-generated ones in question. The question, then, was whether the teacher had authored the love letters or not. There were two potential authors: the teacher, who denied writing them, and the student, who had already admitted to writing love letters to the teacher. Without forensic linguistics, this issue was certainly a matter of “she said, he said.”

Case 3: The Hacked Profile on
In a civil case in Georgia, I was asked by a plaintiff’s attorney to determine if a profile and a threatening letter shared common authorship in a case requesting injunctive relief so that the defendant would no longer contact the plaintiff. The



plaintiff had posted a profile which was allegedly later hacked and changed. Some of the changes included posting the plaintiff’s name, place of employment, and supposed desire to meet dangerous men. The plaintiff also received a long threatening letter. The defendant originally denied authoring the threat letter and later admitted that she had authored it, but she denied hacking the profile. Potential authors of the hacked profile were the plaintiff, the plaintiff’s ex-husband, and the defendant. Without forensic linguistics, there was no way to determine who was at the keyboard changing the profile into a dangerous post. It is obvious that language can be crucial evidence in many different kinds of civil and criminal cases. It is not surprising then that, as Levi (1994) shows, almost all of linguistics has been applied to forensically significant questions.

Forensic Linguistics as Applied Linguistics
Essentially, forensic linguistics applies linguistic theory and method to forensic issues. Linguistics embraces both the qualitative, i.e. non-numerical, case study approach, and the quantitative, i.e. numerical, statistical and experimental approach. For instance, in studying how children acquire their first language, typically intense case studies of individual children are then followed by experiments with children of the same age to refine the observations in the case studies. In forensic linguistics, the quantitative approach is better suited to meeting the legal requirements for evidence than the qualitative approach. The case of author identification illustrates this fact; the quantitative, computational approach to author identification has been admitted as legal evidence without restrictions, while the qualitative, case study approach has been excluded or severely restricted as evidence. Just as sociology and psychology rely on data collection and statistical analysis, linguistics also relies on experimental design, data collection and statistical analysis. The experimental practice in linguistics is especially important when linguistics is applied to forensic issues because each case is in essence an experiment with particular controls determined by the particular case, but whose general design already fits into some empirical paradigm in linguistics. The most interesting aspect of practicing forensic linguistics is determining which empirical paradigms in linguistics can be applied to any individual case. The linguist must first be applying a method that is already established in linguistics, has been used empirically to make predictions about language structure or language behavior, and can be implemented experimentally. Table 10.1, adapted from Chaski (2006), shows the wide range of linguistic fields that can be applied to answer forensically relevant questions. In fact, most theoretically grounded forensic linguistic methods will draw upon several subfields in linguistics. Linguistics in the real world requires a multi-faceted approach when drawing upon language to uncover the particular linguistic variables that in combination can answer a forensic question with a high degree of accuracy. Linguists identify and analyze languages by their structures: how a language is built. Language is a combination of sound (or hand movement) and meaning. Sound sequences evoke meanings for native speakers of the language being spoken or read. The sound part of language is known as phonetics, phonology, and phonotactics, with the meaning part of language including morphology, semantics, syntax, and discourse analysis. Phonetics (the consonants and vowels used) and phonology (how the phonetic segments combine) describe the sounds in a language. Morphology (how words are formed from roots, prefixes, and suffixes), lexicography (the work of compiling dictionaries and studying words), syntax (how words combine to form phrases, and phrases combine to form sentences), semantics (what words and sentences mean), and

Linguistics as a Forensic Science TABLE 10.1: Linguistics Required to Answer Some Forensic Questions Forensic Questions Who spoke this message or statement? Who authored this message or document? Is this document actually from the time from which it appears to be? Is this statement or document defamatory?
Source: Adapted from Chaski, 2006.


Linguistics Phonetics, phonology, morphology, dialectology, sociolinguistics, computational linguistics Syntax, morphology, dialectology, corpus linguistics, psycholinguistics, computational linguistics Lexicography, language change, dialectology, sociolinguistics Lexicography, psycholinguistic experimentation, syntax, semantics, discourse analysis

discourse analysis (what combinations of sentences mean in particular contexts) describe the ways a language conveys meaning. Psycholinguistics explains (through psychological experimentation) the way the human mind works in creating and understanding the linguistic structures in any language, and the kind of structures that aren’t cognitively possible in any language. Within a language, there can be structural differences that do not prevent understanding even though people notice them; dialectology records these language-internal structural differences in relation to where the different speech communities occur geographically. Sociolinguistics focuses on the social demographics (e.g. class, gender) associated with dialects. Corpus linguistics measures a speech community’s performance of language structures, while computational linguistics creates computer software— using everything from phonetics to quantitative linguistics—which analyzes language and can perform some linguistic task, such as answering forensic questions about a document.

Language Identification
The problem of language identification has become especially important due to antiterrorist investigations. Phonetics, phonology, and morphology are immediate clues for identifying a language, specifically through the method of phonotactics. Phonotactics refers to the possible and impossible combinations or sequencing of sounds in particular languages. For instance, *kt at the beginning of a word is not possible in English (* denotes non-occurring), but it is in Modern Greek’s puristic or Katharevusa dialect. Just by discovering the sound patterns that are possible in a language, the linguist can narrow down which languages the speaker might be speaking, and in some cases, actually identify the language or dialect within the language. Computational linguistics offers software which can quickly analyze the phonotactics of a text or statement, and corpus linguistics and statistics offer mathematical techniques for determining language identity based on the phonotactic pattern counts.

Regional Association
The question of regional association relies heavily on dialectology. Dialectology studies the geographical distribution of language patterns. Mostly, dialectologists have focused



on phonetic, phonological, morphological, and lexical patterns that are common in some geographical area but not in others. For instance, “gum band” is a lexical feature of the Pittsburgh, Pennsylvania area, while most American dialects use the term “rubber band.” Again, using what has already been discovered in linguistics, a forensic linguist should be making use of resources such as Cassidy and Hall’s multi-volume Dictionary of American Regional English (2002), which focuses on lexical dialectology, and Labov, Ash and Boberg’s Atlas of North American English (2006), which focuses on phonetic dialectology. If enough dialect features are actually in the text, a forensic linguist could identify the dialect of a speaker/author. Whether the dialectal features point to the native dialect of a speaker is another question, however, since adults may borrow words and phrases from dialects they encounter as adults. Therefore, the linguistic subfields of sociolinguistics and language acquisition (how children and adults learn languages) would also come into play in answering a question about regional association.

Dialectology and Identification
Dialectal differences are easily recognized so that non-linguists and non-scientists might be tempted to misuse this information. Let us imagine, for instance, that a ransom note contains a statement:
Don’t you tell no cops about our little business or he dies.

This sentence contains a double negative, as underlined. Double negatives are considered “bad grammar” from a prescriptive or pedagogical perspective. From a descriptive perspective, which is guided by observational reports of language and language use rather than the introduction of judgments or rules, speakers of the prestige dialect of English do not use double negatives (unless for emotional effect), while speakers of socially stigmatized dialects of English do. Now suppose that known samples of a suspect’s writing are discovered and they contain double negatives, e.g.
I didn’t know nothing about a hold-up.

If, without scientific training, I am hypothesizing that “grammatical errors” are a reliable way to identify authorship, I might stop right here: the fact that double negatives occur in both the ransom note and the suspect’s writing might be enough for me to conclude that the suspect must be the kidnapper. They match! To be sure, I might then look around to see if there are any other grammatical errors in common, and indeed I might find them, since dialects that allow double negatives also allow for subject–verb disagreement (you was saying instead of you were saying; you gots nothing on me instead of you got nothing on me), and other non-standard but nonetheless predictable patterns. From the non-linguist’s perspective, such a clustering of features might seem very important and weighty as evidence, but from a linguist’s perspective, such a clustering of features is simply what is expected about dialects. So from the linguist’s perspective, a clustering of features, especially features that are considered poor grammar, does nothing more than identify the dialect, and dialects are always shared by groups of people. Showing that two documents exhibit features belonging to a group does not relate those two documents to any individual in that group. Dialectal information may be very helpful in screening potential authors of a questioned document, but it would be overstating the case to claim that dialectal information can be used to make identifications as an author or a speaker. A well-trained linguist knows the limits of his or her method in linguistics, and knows that applying the

Linguistics as a Forensic Science


method to new data does not suddenly make the method able to do what it could not previously do.

Speaker Identification
The problem of speaker identification is central to many criminal and security investigations. Hollien (2002) shows how computational, machine-based approaches to speaker identification work, reporting in particular on his own SAUSI: Semi-Automatic Speaker Identification System and Rodman’s computational work. Rodman has continued to develop his computational method (Rodman et al., 2002). Campbell and his research group at Lincoln Laboratories have also produced highly accurate work in speaker identification (Reynolds et al., 2003). Rose (2002) in Australia has especially focused his work on statistical approaches for interpreting machine-based speaker identification results. In current work in speaker identification, phonetics, dialectology, computational linguistics, and statistics are being combined to produce highly accurate methods.

Text Similarity
The problem of textual similarity arises in criminal, civil, and security cases when the intention or text type of a document is important. For instance, a suicide note may be real or faked; without regard to authorship of the note, the question arises of how similar a particular suicide note is to real suicide notes or phony suicide notes: what does a real suicide note typically contain? Discourse analysis, lexicography, syntax, and semantics can all help us determine the characteristics of text type. Computational linguistics offers two methods that can be readily applied to text similarity. The first method is quantifying linguistic features: computer software can be very fast at finding and counting textual patterns. The second method is parsing: a parsing program (also called a parser) analyzes and tags the syntactic, semantic, and discursive structures, which are the grammatical rules through which the text is produced. These structural tags can then be counted. Once linguistic features are counted, then statistical analysis can begin. Text similarity is essentially a statistical problem—how close is close?—based on the quantity of common linguistic features. Thus, many different aspects of language are used as variables for determining the text type of a questioned document.

Author Identification
Author identification is also a statistical problem based on the quantity of structurally tagged patterns. This quantitative, syntactic approach to authorship identification began with Mendenhall (1887), and was first applied to a forensic issue by Svartik (1968). Continuing this line of research, I have developed a syntax-based authorship identification method called ALIAS_SynAID, which stands for Automated Linguistic Identification and Assessment System_Syntactic Author Identification, and I have implemented it in computer software. As a method, the ALIAS system draws upon sociolinguistic data gathering, psycholinguistic experimental design, phonotactics, syntax, semantics, pragmatics, discourse analysis, corpus linguistics, computational linguistics, and statistics. The authorship identification component of ALIAS_SynAID is based on statistically comparing specific theoretically motivated linguistic features of the unknown or questioned document to known or authenticated documents of at least two potential authors. A full description of the method is available online at www.freepatentsonline. com; a more succinct description is available online at (Chaski, 2005). ALIAS_SynAID automates the process of syntactic tagging, which is labeling each word



for its syntactic function, and categorizing those labels into phrase type. Any linguist trained in current syntactic theory can perform tagging by hand, but this can be laborious and tiring, so automating the tagging provides better speed and consistency. ALIAS_SynAID also incorporates statistical analysis to handle more than twenty variables for each document. The statistical procedure determines which of these twenty variables serve to differentiate different authors. Human identification, in general, typically relies on a small set of features that are usually biologically determined, such as eye color, hair, facial shape, height, weight, sex, and body type. We do not use all of these features in each identification we make; we only use those identifying features that actually serve to differentiate. For instance, if I need to identify two brothers, sex is not determinative, but eye color, height, and facial shape might be the three features that enable me to accurately identify one brother and not misidentify the other brother. In the same way, ALIAS_SynAID’s statistical procedure begins with a set of features that are linguistically determined and then selects those features that accurately differentiate authors from each other. Since it is not possible to apply the ALIAS_SynAID method by simply looking at or reading documents, I cannot offer you in this chapter an exercise to implement a syntactic analysis for authorship, but the following description provides a basic understanding of the method. The specific linguistically determined features in my method derive from syntax, punctuation, and word length. The syntactic features are based on patterns for each grammatical category such as noun, verb, preposition, etc. In the syntax of any language, each grammatical category combines with other grammatical categories to create phrases which are then combined to create sentences. When a grammatical category combines with other words to make a phrase which has the same function as the grammatical category, this grammatical category is called a headword. So in the phrase “the books” two grammatical categories determiner (the) and noun (books)—combine, and the whole phrase “the books” could be replaced by “books” but not “the.” Thus the head of the phrase is the noun (not the determiner). The phrasal combinations can be relatively simple or complex. For instance, a noun phrase can be just a two-word combination, like “the book” or “the chapter,” which are combinations of a determiner, “the” or “a,” and a noun, “book” or “chapter.” Alternatively, a noun phrase can be a combination of this two-word phrase with a prepositional phrase or a relative clause such as “the chapter in the book which you are now reading.” The phrasal combinations for each grammatical category are classified by their complexity and counted. Additionally, punctuation marks signal the edges of syntactic combinations and are counted. Finally the average word length is calculated. ALIAS_SynAID thus outputs a numerical description of each document for each author, and for each questioned document. Known-author documents and questioned documents are analyzed in exactly the same way. Suppose that the following text were input into ALIAS_SynAID:
Please don’t leave me. I will do anything you ask to make our marriage work . . . I don’t want to get a divorce. I want us to live the rest of our lives together.

First, SynAID breaks the text into sentences:
Please don’t leave me. I will do anything you ask to make our marriage work . . . I don’t want to get a divorce. I want us to live the rest of our lives together.

Linguistics as a Forensic Science


Second, SynAID tags each word in each sentence for its syntactic function.
Please don’t leave me. Interjection aux-verb negation base-verb pronoun I will do anything you ask to make our marriage work . . . Pronoun modal-verb base-verb noun pronoun finite-verb infinitival-to infinitive-verb possessive noun base-verb I don’t want to get a divorce. Pronoun aux-verb negation base-verb infinitival-to infinitive-verb indefinitedeterminer noun I want us to live the rest of our lives together. Pronoun finite-verb pronoun infinitival-to infinitive-verb definite-determiner noun preposition possessive noun modifier

Third, these tags are categorized by the markedness of the phrases around each grammatical headword, such as noun and verb. One interesting fact about language is that contrasts in language (such as the words [young/old], the speech sounds [p/b], or the phrases [how are you?/howdy!]) are never equal; there is always one member in the pair that is more frequent than the contrasting member, or more complicated than the other, or more general in meaning than the other. This asymmetry or unequal quality in linguistic contrast is known as markedness. Generally, simple phrases are unmarked, while complex phrases are marked.
Unmarked Noun Phrases Marked Noun Phrases Our marriage Anything you ask A divorce The rest of our lives Unmarked Verb Phrases I want Marked Verb Phrases Don’t leave Will do To make work To live

The marked and unmarked phrases of each type are counted within each sentence, and then these counts are summed together so that the entire document is represented by numerical output from ALIAS_SynAID. Fourth, this numerical output is then analyzed statistically using discriminant function analysis, a statistical method that builds a model of how well two different authors’ documents can be differentiated from each other. This model provides an accuracy rate, so if all the documents from each author can be differentiated from the other author, the accuracy rate is 100 percent. The accuracy rate allows us to determine if the model is reliable, i.e. can the statistical model reliably distinguish between the two sets of known documents? If the model is reliable (if the accuracy rate is high), then this statistical model is used to predict the author class of the unknown document. The most important feature of this statistical approach to author identification is that it takes into account the variation that is within the known documents of each potential author (intra-author variation) as well as the variation that is between the known documents of the different potential authors (inter-author variation). If the model is not reliable, then there will be a low accuracy rate because the variation inside each author will overlap too much with the variation between the authors. Let’s return to the three cases I introduced earlier.



Case 1: Resolution of the Inheritance Dispute
Using ALIAS_SynAID to analyze, categorize, and count syntactic patterns with the numerical output subjected to statistical analysis, I tested the daughter’s known documents, the mother’s known documents, and the daughter–mother known documents, as well as the questioned inheritance letters. The daughter’s and the mother’s known documents were clearly and accurately separated into two classes by the statistical procedure. This test simply showed the reliability of the method: it accurately distinguished the mother’s from the daughter’s known writing. (In research, your results should be able to be replicated by further study; hence, the criterion of reliability is crucial to this work.) The daughter–mother known documents were clearly separated from the mother’s documents, but not from the daughter’s documents. The daughter–mother known documents and the daughter’s documents formed a class unto themselves. Again, this test showed the reliability of the method, since in this scenario the daughter first authored the documents that the mother copied, so it made sense that these documents would be more like the daughter than the mother, and actually be classified to the daughter rather than the mother. The real question could then be answered: were the inheritance letters in a class with the mother’s documents or not? The inheritance letters were not classified as the mother’s documents. Instead, the inheritance letters were assigned, by the statistical procedure, to the documents authored by the daughter independently or as a draft for the mother. The forensic linguistic results supported the three siblings’ contention that their mother had not independently authored the inheritance letters disinheriting them. At this point the case settled, with an undisclosed division of property.

Case 2: The Teacher’s Trial
In addition to the questioned love letters, the attorney provided known writing samples from the teacher and the admitted love letters from the student. Again, using ALIAS_SynAID, I tested the teacher’s known documents and the student’s known documents to determine how distinguishable each set was. The teacher and the student were clearly differentiable from each other in their linguistic features. I then tested the questioned love letters. The love letters were each classified to the teacher, not to the student. The evidence was not offered by the defense, since by constitutional law in the United States, citizens do not have to incriminate themselves, but the defense offered no rebuttal to the prosecution’s presentation of the letters in trial. The teacher was convicted of taking indecent liberties with a child.

Case 3: The Hacked Profile
In addition to the questioned profile, the attorney provided known writing samples from the plaintiff, the plaintiff’s ex-husband, and the long threatening letter admittedly authored by the defendant. Due to the length of the threat letter, I was able to split it into six segments so that the defendant had approximately the same number of writing samples as the plaintiff and ex-husband. Again, using ALIAS_SynAID, I tested the plaintiff’s known documents and the defendant’s known documents to determine how distinguishable each set was, and classify the hacked profile. The plaintiff and the defendant were clearly differentiable from each other in their linguistic features, and the hacked profile was classified to the defendant. In each test of paired authors, the hacked profile consistently was classified to the defendant. At this point, the case proceeded to trial.

Linguistics as a Forensic Science


New Computational Tools for Author Identification
Other computational tools available for author identification are described in GuillénNieto et al. (2008) and Juola (2008). Both Guillén-Nieto and her colleagues at the University of Alicante and Juola and his colleagues at Duquesne University are developing software to implement different methods for authorship identification. Guillén-Nieto et al. (2008) appear to be focusing on fairly simple features such as sentence length, word length, and character distributions, within a forensic context. Juola’s (2008) work is gathering many statistically sophisticated, character-, word-, and syntax-based approaches into one software package. At this time, neither research group has reported results of validation testing for their software, but such validation results are eagerly awaited. It is especially encouraging that these projects are being undertaken.

Forensic Stylistics
An alternative to computational author identification, forensic stylistics first emerged in the 1940s, originally among forensic handwriting examiners, and was later adopted by literary critics and linguists (hence it is also called, by its proponents, text analysis, discourse analysis, sociolinguistics, or psycholinguistics) (McMenamin, 1993, 2002). Forensic stylistics is a qualitative method in practice, because the analyst decides which linguistic features matter from case to case. By far the best explication of the forensic stylistics method is McMenamin (2002). In the forensic stylistics method, the analyst selects, through his or her personal evaluation, the linguistic features that serve as stylemarkers, idiosyncratic similarities or differences, such as spelling, grammatical error, word choice, and formatting. The stylemarker clusters are claimed to be unique to individuals. If the same set of stylemarkers is found in two documents, then the analyst concludes the two documents were authored by the same person. McMenamin (2002) provides a list of hundreds of features that might be considered. On the same set of documents, one analyst could select features A, F, G, H, and L, while another analyst could select features B, C, D, and E, and both analysts would be producing a legitimate forensic stylistics analysis. The forensic stylistics method could be greatly improved with a standard protocol for checking features consistently in the questioned and known documents, but this idea has not been adopted by forensic stylistics proponents, who argue that they never know what is going to be an important feature until they find it. Given this approach, the method must rely completely on the intuitive evaluation of the practitioner, which is difficult if not impossible to replicate. Even if the practitioner has the best of intentions, which Professor McMenamin no doubt has, the method itself does not constrain the confirmation bias—seeing what we want to see—that all humans are prone to. Nor does the method require disconfirming evidence to be weighed in an objective way; the examiner can simply show only similarities or only differences, and is not required by the method to take into account both similarities and differences. Any method that allows anyone to ignore contrary evidence feeds into confirmation bias. At this point, the quantitative, computational method, such as ALIAS_SynAID, and the qualitative, intuitive method, such as forensic stylistics, might appear similar because both start with a potential set of features, out of which a subset is used to identify authors. But the methods differ in fundamental ways. First, the set of potential features in ALIAS_SynAID is based on syntactic theory, which enables linguists to precisely and accurately describe the world’s languages. The set of potential features in forensic stylistics is based on handwriting examination and prescriptive grammar (the pedagogic



grammar valued by educational systems). Second, the set of potential features in ALIAS_SynAID is small, well defined and used for all analyzed documents in the same way, while the set of potential features in forensic stylistics is large, open-ended, ill defined, because the set is not based in linguistic theory nor empirically tested, and differs from case to case and analyst to analyst. The subjective assessment that features that co-occur in two different documents prove common authorship of the two documents has been shown empirically to be usually wrong. Third, ALIAS_SynAID does not claim any unique set of stylemarkers or a unique syntax for each writer, because in order to communicate in a particular language, all of us have to access dialect patterns shared by others in our dialect, the same set of punctuation marks, and the same set of syntactic patterns to make noun phrases, verb phrases, and sentences. ALIAS_SynAID doesn’t claim uniqueness is even viable, but it has demonstrated that the marked or unmarked way in which the syntactic patterns are used can differentiate author one from author two. Fourth, and most important, the subset of identifying features in ALIAS_SynAID is determined through statistical testing, while the identifying features in forensic stylistics are determined through the examiner’s intuition. The statistical testing provides a replicable and objective way to determine which linguistic patterns are the ones enabling us to differentiate author one from author two. The statistical testing guards against confirmation bias, because the computer is counting and determining the weight of the evidence from the counts. Further, the statistical testing takes into account that one author’s documents can vary in the distribution of features, i.e. intra-author variation. Other linguists have expressed their concern that forensic stylistics is not quantitative, so that there is no objective measure of the features that are apparently important. One major issue is that without objective measurement, there is no way to know if the inter-author variation is larger than the intra-author variation. If the intra-author variation is not attended to, then we really cannot tell if a pattern truly belongs to an author or not. The world-renowned linguist David Crystal (who has authored, among his 100 books, two books about English style as well as two encyclopedias published by Cambridge University Press) rejected forensic stylistics as linguistics in a review of McMenamin’s Forensic Stylistics (1993). Crystal’s (1995) review was published in Language, the prestigious journal of the Linguistic Society of America, and is available at Crystal asks
what norms are used as the baseline for the judgments? When McMenamin says, concerning the use of the percent sign and ampersand, that ‘what . . . they have in common is their occasional use. Their use is not frequent or abnormal’, or ‘parentheses . . . are used very frequently’, or ‘the semicolon . . . occurs very frequently,’ how are we to interpret these remarks? Is this linguistic SCIENCE? (Crystal, 1995, p. 384)

If we do not know the rate at which parentheses are used by author A and author B, then how can we know if the rate at which parentheses are used in the questioned document is more like author A or author B? Another well-respected linguist, Geoffrey Nunberg, has rejected forensic stylistics in case reports (personal communication, 2003). Although Roger Shuy (1998) spoke favorably of forensic stylistics, in a 2008 Language Log post he has also expressed doubts about British psychologist Tim Grant’s work applying forensic stylistics to text messages, echoing Crystal’s points regarding McMenamin’s work:

Linguistics as a Forensic Science even if there are a few language features that seem to stand out (like the use of “my” and “myself” mentioned in the BBC article), without a large corpus of data I can’t tell whether the writers used these variably. And if they did use them variably, what percentage of variation of use in a given feature is necessary in order for anyone to be able to claim that the writer has reached it? (Shuy, 2008)


The full text of Shuy’s log is available at
The forensic stylistics method looks for stylemarkers, the similarities and differences between texts, in terms of formatting, spelling, punctuation, and grammatical errors (from a prescriptive point of view such as learning in school not to end a sentence with a preposition or use double negatives). (a) Do you think that different writers never make the same spelling mistakes? (b) Do you think that different writers always use different punctuation marks? (c) Do you think that two different writers could never share the combined pattern of making the same spelling mistakes and using different punctuation marks?

Exercise 4

Exercise 5

Gather about ten emails from at least ten of your friends. For each person, make a list of the formatting, punctuation, spelling and grammar in each email. (a) For each person, is each email totally consistent with the other emails from that person? (b) For all ten persons, is each person’s style totally different from the others’? (c) Did you actually find ten totally different styles or did you find overlapping styles? (d) Did you find any single email from one person that is stylistically just like an email from a different person? (e) How much confusion between authors would take place in your data? How many different authors would you identify based on your data? (f) How much linguistic training did you need to perform this experiment? (g) What linguistic information did you rely on to perform this experiment?

Validating Methods for Forensic Linguistics as a Science
Just as this chapter was undergoing final revisions, the National Academies of Sciences (NAS) (2009) released its review of forensic science in the United States. A dominant theme throughout this review is that forensic methods require validation testing in order to be used with reliability and fairness throughout the nation. Underpinning this theme is the recognition that qualitative, experience-based or intuition-based methods are prone to confirmation bias.
A body of research is required to establish the limits and measures of performance and to address the impact of sources of variability and potential bias. Such research is sorely needed, but it seems to be lacking in most of the forensic disciplines that rely on subjective assessments of matching characteristics. These disciplines need to develop


Chaski rigorous protocols to guide these subjective interpretations and pursue equally rigorous research and evaluation programs. The development of such research programs can benefit significantly from other areas, notably from the large body of research on the evaluation of observer performance in diagnostic medicine and from the findings of cognitive psychology on the potential for bias and error in human observers. (NAS, 2009, p. 23)

The primary difference between non-linguist methods and linguistics as a forensic science is the scientific approach. Any scientific approach is characterized by hypothesistesting, experimentation, replicability of method and of results, and predictive power. Further, any scientific approach guards against confirmation bias by seeking disconfirming evidence by quantifying and automating any tiresome task so that response bias is mitigiated. In forensic linguistics, the scientific method requires hypothesis testing to find out if the particular idea about language use actually works in an analysis. Validation testing is essential to the development of linguistic methods as well as standards for the use of the methods. Validation testing is a series of experiments that test a method’s ability to obtain the correct answer on a database for which we know the correct answers. Using the previous example of dialectal “grammatical errors,” consider the steps required to test the hypothesis that grammatical errors identify authors. Suppose I need to test the hypothesis to find out if any individuals share the same kinds of grammatical errors. So, without being involved in any specific legal case, but just for the sake of scientific method and validity, I need to take the following steps: 1 2 3 4 5 6 Create a database of documents which are authenticated for their authors. Create an objective and repeatable method for identifying all the grammatical errors in each document. Count and classify the grammatical errors for each author. Compare the grammatical errors across all the authors. Determine if any authors have the same grammatical errors. If any authors have the same grammatical errors, I can figure out how many times two authors would be confused based on grammatical errors: this gives me an error rate for the technique. This error rate then tells me how much predictive power the technique has: if this method is only accurate 50 percent of the time, then the grammatical errors method is not a good way to predict authorship of an unknown document, because I might as well flip a coin. If no authors have any grammatical errors (because they have run their documents through an automatic grammar checker or because they are Standard English writers writing fairly formal English), then the dataset cannot tell me anything about the accuracy of this method except that it will not separate “good” writers from each other. If authors do exhibit grammatical errors but ones peculiar to each author, I can now use other databases to repeat this experiment and perhaps replicate these results.



This kind of validation testing has been demonstrated by Chaski (1997, 2001, 2005, 2007), DeVel, Anderson, Corney, and Mohay (2001), Koppel and Schler (2003), St. Vincent and Hamilton (n.d./post-2001), and Gamon (2004), all of whom have independently tested hypotheses about author identification in English. Other languages in which objective, quantitative, and computational experiments on author identification have been conducted include Greek by Stamatatos, Fakotakis, and Kokkinakis (2000, 2001); Turkish by Diri and Amasyali (2003); and Dutch by Baayen, van Halteran, Neijt, and Tweedie (2002). This body of scholarship shows a general consensus that

Linguistics as a Forensic Science


solutions to the author identification problem rely on the objective measurement of syntactic categories and punctuation subjected to strong statistical analysis. Most syntax-based solutions are currently achieving at least 80 percent accuracy on fairly brief 200–1,000 word texts such as are common in forensic settings. Empirical evidence showing that the forensic stylistics method has rather low accuracy is also readily available (Chaski, 2001; Koppel and Schler, 2003; St. Vincent and Hamilton, n.d./post-2001). In Koppel and Schler’s (2003) experiments, the highest accuracy for author identification using 99 forensic stylistics features was 67.2 percent. In St. Vincent and Hamilton’s experiments, the highest accuracy for author identification was 51.46 percent. Notably lacking among forensic stylistics proponents is any actual research that could provide an accuracy rate for forensic stylistics or evidence about the uniqueness of stylemarker clusters, a fact that McMenamin concedes (2002: 207).

The Possibility of Linguistics as a Forensic Science
Within the law, there are two main standards for evidence for expert witnesses such as forensic linguists: the Frye and Daubert standards. One of these two standards (or variations of them) is found in every state court, while the Daubert standard is found in the Federal courts. These two standards differ in how scientific methods obtain credibility before the court and how the judge decides if the scientific evidence is or is not allowed into the trial as evidence. In 1923, the United States Supreme Court set forth what is known as the Frye standard for scientific evidence. The Frye standard was intended to solve the problem of the point at which a scientific method or technique becomes proven: in other words, when does a scientific discovery cease being hypothetical and turn into fact? The Court ruled that the best indicator for this shift in epistemological status is the general acceptance of the method or technique by the relevant scientific community. The general acceptance standard certainly makes sense because it puts the epistemological question in the hands of the scientists. Scientists are doing science in order to explicate rule-governed behavior, to show that certain conditions produce certain consistent results, to identify an object and predict its behavior based on that identity. Scientists should know whether a particular method (1) gets the same results when it is replicated by different groups, or (2) gets the same results when it is tested on different data sets, or (3) explains a set of observable facts which a previous method could not explain, or (4) is able to predict a consequence from a constellation of determinative conditions. In other words, scientists should be the ones who can tell whether a particular method is functioning like science or not. But over the seventy years between 1923 and 1993, the general acceptance standard eventually revealed the weakness of science as a very real human weakness, i.e. that general acceptance could be claimed for just about any method or procedure as long as the scientific expert said it existed based on a handpicked scientific community. Within the traditional forensic sciences such as fingerprint and handwriting examination, this issue of general acceptance within a small and select community of practitioners has become a legal bombshell. Neither fingerprint nor handwriting examination has an academic discipline to which it belongs because both of these forensic methods essentially developed out of police investigative techniques rather than standard sciences. Further, since statisticians began in recent years to look at handwriting and fingerprint data, the assumption of uniqueness has been seriously questioned and undermined. Therefore, the examiner’s ability to individuate has also been seriously questioned, and claims about being able to individuate phenomena which may in fact not be unique are



obviously difficult to maintain. Both handwriting and fingerprint examination have been attacked in recent years as unreliable and without a real scientific community because a more empirical paradigm for scientific evidence has come into the judicial arena, led by the development of DNA identification. In 1993, the United States Supreme Court put forth another paradigm for determining if scientific evidence should be allowed into court testimony. This paradigm, known as the Daubert standard, includes general acceptance but focuses on empirical reliability. This standard takes the approach that scientific methods must actually be proven to be accurate in their predictions before a scientific community accepts them, and so, hand in hand with general acceptance, real scientific methods will be able to show how reliably they perform on real data. In fact, even more important than how reliably they do perform is how often the methods produce errors. The error rate is especially important in a legal context because it tells the judge and jury whether a scientific method is wrong a lot of the time, or not so often. Among traditional forensic sciences, the Daubert standard has stimulated research into the basic assumptions. For instance, the individuality of handwriting is now being tested empirically. The development of software based on optical character recognition algorithms has shown that the individuality assumption is probably not tenable, but more importantly such software is being validated on different datasets to demonstrate that, even if handwriting is not unique, it can reliably be distinguished from different writers at very high accuracy rates and very low error rates (Chaski and Walch, 2009; Srihari, et al. 2002). The same kind of research is currently ongoing in fingerprint and other pattern evidence. Using language as evidence began even before the Frye era with notions that spelling errors, grammatical errors, and other ideas from prescriptive grammar might reliably indicate the authorship of a will. During the Frye era, when general acceptance from a scientific community became important, handwriting examiners continued to include prescriptive grammar in their arsenal, and several academic linguists who became involved in the judicial system as experts in forensic linguistics in the late 1970s and early 1980s adopted the handwriting examiners’ approach. The academic linguists had the advantage of belonging to an actual academic discipline, holding doctoral degrees in linguistics or a language-related discipline, and publishing in peer-reviewed journals. Academic linguists could thus appear quite impressive as experts, according to the Frye standard. In fact, appearing as an expert forensic linguist during the Frye era could be quite lucrative. One forensic linguist candidly stated in a recent interview with Lawyers USA (Dahl, 2008) that forensic linguistics twenty years ago was “just a sideline where we made a few bucks on the side helping out some lawyers and getting to do something a little different.” Also during this period, forensic linguists could testify using untested methods which had never been used in standard linguistics, a loophole which is easy to pass through given that so few judges or jurors even know what linguistics is. But during this same time period, some linguistic experts were testifying on subjects in which their entire research careers were relevant and truly germane to the issue on hand, such as Labov’s (1982) work on the relationship of Black English to educational policy. Since the Daubert standard and related rulings have focused judicial attention on the fact that scientific procedure is by definition repeatable, objective, and produces reliable results, the bar for forensic linguistics has been raised. Forensic linguistics, like other forensic sciences, is expected to demonstrate that its methods are accurate at predicting, classifying, identifying, and possibly even individuating. While the Daubert standard has certainly not totally prevented academic linguists from pursuing a sideline or hobby at the expense of a defendant’s money, time, or even life, or testifying on issues which they

Linguistics as a Forensic Science


have never researched or published in, the Daubert standard has at least awakened linguists to the reality that working in the legal arena means understanding legal standards and conforming to scientific standards. Howald’s (2006, 2008) work analyzes legal standards for linguistic evidence.
Visit the website, where you will find a very important and much-cited ruling in forensic linguistics, the case of United States v. Van Wyk 83 F.Supp.2d 515 (d. New Jersey, 2000). After reading this case, what do you think is the best course for the discipline of forensic linguistics to take?

Exercise 6

Case 3, Continued: The Hacked Profile Trial
As the case was preparing for trial, the defendant’s counsel chose not to depose me even though my client and I attempted to arrange a deposition in person, via video or telephone. During the trial, the defendant’s counsel asked the Court to exclude my testimony based on Georgia’s statute incorporating the Daubert standard. The Court overruled this motion and my testimony was admitted without any restrictions on my ability to state a conclusion about the authorship of the hacked profile. During my testimony, I was asked specifically about software development in computational linguistics, the results of validation testing for ALIAS_SynAID as well as the results of using ALIAS_SynAID to analyze the data in this case. At the conclusion of the trial, the Court gave a directed verdict which provided injunctive relief to the plaintiff, in that the defendant is not to make any contact with the plaintiff in the future. Legally acceptable science is challenging to many sciences, including linguistics. Linguistics is a very broad discipline because language is involved in almost every aspect of human life. Since linguistics reflects the pervasiveness of language, there are many branches of linguistics and methods in linguistics that may not be able to meet the challenge of legally acceptable science. Some branches of linguistics are simply more developed as science than others, and some branches of linguistics align themselves with non-scientific methods. For instance, theoretical linguistics (including phonology, morphology, syntax, and semantics) focuses on the development of a theory of universal grammar which predicts what is or is not possible in human language, and how differences in human languages can be predicted. Each time a new language is examined in light of this linguistic theory of universal grammar, the theory is either supported or not. If the theory is shown to be unsupportable, then the theory is changed. Psycholinguistics takes the theoretical constructs from linguistic theory and tests the psychological reality of these constructs using methods from experimental psychology. Thus, both descriptive taxonomic grammars of particular languages and psycholinguistic experimentation provide a way for linguistic theory to be validated. The theory either correctly predicts what is found in newly studied languages or it does not; the theory either correctly predicts particular constraints on human language processing or it does not. Some fields in linguistics, notably psycholinguistics, neurolinguistics, and computational linguistics, are closer in their everyday work habits to legally acceptable science than other fields in linguistics, such as translation or language teaching. Shuy (2008) discusses how the kind of discourse analysis he does may not be amenable to experimental, i.e. validation, testing, and thus not admissible under Daubert; while Tiersma and Solan (2002) argue that discourse analysis may not be welcome in courts for other reasons.



The truth is that not all of linguistics is viable as a forensic science, and some of linguistics will become viable as a forensic science only when such standard scientific procedures as validation testing and the discovery of error rates are consistently pursued, whenever they can be pursued within the current research paradigms in linguistics. Meanwhile, the courtroom standards challenge us to improve linguistics as a science in the classroom and laboratory. Through embracing wherever possible a model of linguistic science in which we pay attention to our error rates, our theory’s ability to predict accurately, our method’s limitations in terms of data quality and quantity, in the future, linguistics can truly become a mature forensic science.

Comments on Exercises
Exercise 1: English-only laws or laws which make English the official language of the United States of America appear to be brought up in cycles. Recent stresses on the immigration policies are causing English-only laws to be legislated again. You can find a good overview of recent news items by searching the New York Times archives. Exercise 2: If you were fortunate enough to find laws which were plainly stated, read your (or your parents’) insurance policy. Exercise 4: Stylemarkers are very easy to understand and find, so much so that one linguist told me he uses them just because a jury can understand them, but the accuracy rate for forensic stylistics appears to be about 60%. Would you want to be convicted of a crime based on a method which gets it right about 60% of the time? Exercise 5: Congratulations, you have conducted a linguistic experiment! Exercise 6: Since the van Wyk ruling, other Courts have applied the van Wyk ruling so that forensic stylistics expert can not actually give an expert opinion in other court or forensic stylistics has been excluded completely. Although the Daubert standard still includes general acceptance as one criterion, and this criterion can be abused by hiring experts as hired guns to simply state that they do not know about or accept a method which is being presented, empirical reliability and validation testing is the strongest mark of mature forensic science. It is my goal that forensic linguistics hits the mark.

Baayen, H., van Halteran, H., Neijt, A., and Tweedie, F. (2002) An experiment in authorship attribution. Journées internationales d’analyse statistique des données textuelles, 6, pp. 1–7. Baugh, A. C. (1935) History of the English language. New York, D. Appleton–Century–Crofts, Inc. Charrow, R. P. and Charrow, V. R. (1979) Making legal language understandable: A psycholinguistic study of jury instructions. Columbia Law Review, 79(7), pp. 1306–1374. Chaski, C. E. (1997) Who wrote it? Steps toward a science of authorship identification. National Institute of Justice Journal, 233, pp. 15–22. Available from: [Accessed June 27, 2009]. Chaski, C. E. (2001) Empirical evaluations of language-based author identification techniques. Forensic Linguistics: International Journal of Speech, Language and Law, 8(1), pp. 1–64. Available from: [Accessed June 27, 2009]. Chaski, C. E. (2005) Who’s at the keyboard? Authorship attribution in digital evidence investigations. International Journal of Digital Evidence, 4(1), pp. 1–13. Available from: [Accessed June 27, 2009].

Linguistics as a Forensic Science


Chaski, C. E. (2006) Forensic linguistics. McGraw-Hill. Chaski, C. E. (2007) The keyboard dilemma and author identification. In Shinoi, S. and Craiger, P. eds. Advances in digital forensics III. New York, Springer. Available from: [Accessed June 27, 2009]. Chaski, C. E. and Walch, M. A. (2009) Validation testing for FLASH ID on the Chaski writer sample database. Proceedings of the American Academy of Forensic Sciences Annual Meeting, February 16–21, 2009, Denver, USA. Chomsky, N. (2006) Noam Chomsky debates with Fryar Calhoun, E. B. Murray, and Arthur Dorfman [Internet]. Available from: [Accessed June 28, 2009]. Crystal, D. (1995) Review of forensic stylistics by Gerald R. McMenamin. Language, 71(2), pp. 381–385. Dahl, D. (2008) Forensic linguists make a science of syntax [Internet]. Available from: http:// [Accessed June 27, 2009]. DeVel, O., Anderson, A., Corney, M., and Mohay, G. (2001) Mining email content for author identification forensics. Association for Computing Machinery—Special Interest Group on Management of Data “Sigmod Record” Web Edition, 30(4). Available from: http:// [Accessed June 28, 2009]. Dictionary of American Regional English. (2002) Cambridge, Harvard University Press. Diri, B. and Amasyali, M. F. (2003) Automatic author detection for Turkish texts. 13th International Conference on Artificial Neural Networks and Neural Information Processing (ICANN/ICONIP), October 3–6, 2006, Hong Kong. Available from: mygetfile.php?id=265 [Accessed June 27, 2009]. Freedman, A. (2007) The party of the first part: The curious world of legalese. New York, Henry Holt and Co. Related website at: [Accessed June 27, 2009]. Gamon, M. (2004) Linguistic correlates of style: Authorship classification with deep linguistic analysis features [Internet]. Available from: [Accessed June 27, 2009]. Guillén-Nieto, V. G., Sierra, C. V., Juan, M. P., Barco, P. M., and Cueto, A. S. (2008) Exploring state-of-the-art software for forensic authorship identification. International Journal of English Studies, 8(1), pp. 1–28. Hollien, H. F. (2002) Forensic voice identification. New York, Academic Press. Howald, B. S. (2006) Comparative and non-comparative forensic linguistic analysis techniques: Methodologies for negotiating the interface of linguistics and evidentiary jurisprudence in the American judiciary. University of Detroit Mercy Law Review, 285(3), pp 288–289. Howald, B. S. (2008) Authorship attribution under the rules of evidence: Empirical approaches in a layperson’s legal system. International Journal of Speech, Language, and the Law, 15(2), pp. 219–248. Institute for Linguistic Evidence. (2008) International Association of Forensic Linguists. (2009) [Accessed June 27, 2009]. Juola, P. (2008) The JGAAP wiki [Internet]. Available from: http://www/ ~fa05ryan/wiki/index.php/Main_Page [Accessed June 29, 2009]. Kniffka, H. (2007) Working in language and law: A German perspective. Basingstoke, Palgrave Macmillan. Koppel, M. and Schler, J. (2003) Exploiting stylistic idiosyncrasies for authorship attribution. Proceedings of IJCAI’03 workshop on computational approaches to style analysis and synthesis, August 10, 2003, Acapulco, Mexico. Available from: [Accessed June 29, 2009].


Chaski Labov, W. (1982) Objectivity and commitment in linguistic science: The case of the black English trial in Ann Arbor. Language in Society, 11(2), pp. 165–201. Labov, W., Ash, S., and Boberg, S. (2006) The atlas of North American English: Phonetics, phonology and sound change. Berlin, Walter de Gruyter. Levi, J. N. (1994) Language and law: A bibliographic guide to social science research in the USA. Chicago, American Bar Association. McMenamin, G. R. (1993) Forensic stylistics. Amsterdam, Elsevier. McMenamin, G. R. (2002) Forensic linguistics: Advances in forensic stylistics. Boca Raton, CRC Press. Mellinkoff, D. (1963) The Language of the law. Toronto: Little, Brown and Co. Mendenhall, T. C. (1887) The characteristic curves of composition. Science, 9, pp. 237–249. Merriam–Webster’s Collegiate Dictionary. (2002) 10th edn. Springfield, Merriam– Webster, Inc. National Academies of Sciences (NAS) (2009) Strengthening forensic science in the United States: A path forward. Washington, National Academies Press. Party of the First Part (2009) The legalese hall of shame [Internet]. Available from: http:// [Accessed June 27, 2009]. Reynolds, D., Andrews, W., Campbell, J., Navratil, J., Peskin, B., Adami, A., Jin, Q., Klusacek, D., Abramson, J., Mihaescu, R., Godfrey, J., Jones, D., and Xiang, B. (2003) The SuperSID project: Exploiting high-level Information for high-accuracy speaker recognition. Center for Language and Speech Processing Workshop. Available from: groups/supersid/icassp03_overview.pdf [Accessed June 28, 2009]. Rodman, R., McAllister, D., Bitzer, D., Cepeda, L., and Abbitt, P. (2002) Forensic speaker identification based on spectral moments. Forensic Linguistics: International Journal of Speech, Language and Law, 9(1), pp. 22–43. Rose, P. (2002) Forensic speaker identification. New York, CRC Press. St. Vincent, S. and Hamilton, T. (n.d./post-2001) Author identification with simple statistical methods [Internet]. Swarthmore, Swarthmore College, Department of Computer Science. Available from: [Accessed June 29, 2009]. Shuy, R. W. (1993) Language crimes: The use and abuse of language evidence in the courtroom. Malden, Wiley-Blackwell. Shuy, R. W. (1998) The language of confession, interrogation, and deception. Thousand Oaks, Sage Publications. Shuy, R. W. (2000) Breaking into language and law: The trials of the insider-linguist. In Alatis, J. E., Hamilton, H. E., and Tan, A. eds. Linguistics, language, and the professions: Education, journalism, law, medicine, and technology. Georgetown University round table on languages and linguistics. Washington, Georgetown University, pp. 67–80. Shuy, R. W. (2002) Linguistic battles in trademark disputes. Basingstoke, Palgrave Macmillan. Shuy, R. (2007) Baboons and Daubert [Internet]. Available from: ~myl/languagelog/archives/005006.html [Accessed June 28, 2009]. Shuy, R. (2008) Author identification in the news [Internet]. Available from: http:// [Accessed June 28, 2009]. Solan, L. M. (1993) The language of judges. Chicago, University of Chicago Press. Srihari, S. N., Cha, S. H., Arora, H., and Lee, S. (2002) Individuality of handwriting. Journal of Forensic Sciences, 47(4), pp. 1–17. Stamatatos, E., Fakotakis, N., and Kokkinakis, G. (2000) Automatic text categorization in terms of genre and author. Computational Linguistics, 26(4), pp. 471–495. Stamatatos, E., Fakotakis, N., and Kokkinakis, G. (2001) Computer-based authorship attribution without lexical measures. Computers and the Humanities, 35(2), pp. 193–214.

Linguistics as a Forensic Science


Svartik, J. (1968) The Evans statements: A case for forensic linguistics. Göteborg, University of Gothenburg Press. Tiersma, P. M. (1999) Legal language. Chicago, University of Chicago Press. Tiersma, P. M. (2006) Communicating with juries: How to draft more understandable jury instructions. Williamsburg, National Center for State Courts. Available from: http:// [Accessed June 28, 2009]. Tiersma, P. M. and Solan, L. M. (2002) The linguist on the witness stand: Forensic linguistics in American courts. Language, 78(2), pp. 221–239.

Other Resources
Associations and Websites
American Academy of Forensic Sciences,, Forensic Linguistics Society, International Association of Forensic Linguists, International Language and Law Association, Institute for Linguistic Evidence, National Center for State Courts, National Clearinghouse for Science Technology and the Law, The Association for Linguistic Evidence,

Dialectal Sources
Dictionary of American regional English. (2002) Cambridge, Harvard University Press, Volume I A–C, 1985, Volume II D–H, 1991, Volume III I–O, 1996, Volume IV P–Sk, 2002, Volume V Sl–Z, 2009.

Forensics Literature
Solan, L. M., Tiersma, P. M., and Conley, J. M. (2005) Speaking of crime: The language of criminal justice. Chicago, University of Chicago Press.

Forms of Language and Communication
Part IV looks at language in several incarnations: the emerging language in childhood, a signed language, and the communication system of a different species. Chapter 11, by Janine Graziano-King and Helen Smith Cairns, explains how linguistic competence develops in childhood. First language acquisition is an amazing feat for so young a child, and this chapter takes us along for the journey. Miako Villanueva, Deanna Twain, and Laura Leigh Wood, the authors of Chapter 12, are users of American Sign Language (ASL). We learn what ASL is linguistically from ASL scholars and sign interpreters. We finish this section with a chapter by a honey bee expert (Chapter 13). When he is not out in the fields, tending to his hives, Wyatt A. Mangum is teaching us how these creatures convey information via an elaborate and symbolic dance. The chapter challenges readers to think more broadly about language, for the real world comprises many types of communication.

This Chapter Explores:

First Language Acquisition Developing Native Linguistic Competence

Janine Graziano-King and Helen Smith Cairns

Stages in Early Child Language Linguistic Competence Linguistic Performance How a First Language Is Acquired

Acquisition of Phonology, Lexicon, Morphology, Syntax Metalinguistic Skill Critical Period

There is—so to speak—in every child a painstaking teacher, so skillful that he obtains identical results in all children in all parts of the world. The only language men ever speak perfectly is the one they learn in babyhood, when no one can teach them anything! Not only this, but if at a later age the child has to learn another language, no expert help will enable him to speak it with the same perfection as he does his first. Maria Montessori, 1967

Maria Montessori was an educator and psychologist who devoted her life to the study and education of young children. She believed that the child’s first six years of life were the most significant. She developed the Montessori philosophy of education, which downplays the role of instruction and, instead, views the role of the teacher as creating opportunities for each child to develop and fulfill his or her highest potential. In the quotation above, Montessori is referring not specifically to the language of men, but to the language of adults. We begin with this quotation because it, like the title of the book in which it appears—The Absorbent Mind—captures the features of first language acquisition that we emphasize in this chapter. That is, that first language acquisition is natural, automatic, and effortless. JGK: My interest in language learning grew out of my own struggle to learn a second language in high school and college and the struggles of my adult ESL students. However, when my son was born, I became more interested in how a first language is learned—particularly in


Graziano-King and Smith Cairns the processes that underlie lexical learning—and I decided to research first language acquisition, studying under the guidance of my co-author. HC: When I had my first child in 1959, the study of first language acquisition had barely begun. Not only did I know nothing about language development, no one knew very much. In the mid-1960s, when I was a graduate student and had my last child, there had been an explosion of work in this field. My experience of the youngest child’s language development was entirely different from the first because I knew how to listen to him, what to pay attention to, and something about the acquisition process. It is truly amazing how much we have learned in the last forty-plus years about how children acquire language, yet how much there is we still don’t know. As you read this chapter, we encourage you to work on the exercises as they come up because they will help you understand, through experience, some points about language that are important to understanding first language acquisition.

If you are reading this book, then you have not only learned how to read, but you have learned how to read English. Most likely, you also know how to speak English. How did you learn English? If English was the first language you learned, you probably do not remember much about learning it. For a child, the process of learning a first language— English or any other language—is essentially effortless and unconscious. Researchers in the field of language acquisition, who are called acquisitionists, question how so complex a task as first language acquisition (or FLA) can be achieved by children with such apparent ease.

How Is FLA Studied?
To investigate the process of FLA, acquisitionists look at what children do as a window into their knowledge of the language they are learning and how they are learning it. Linguistic knowledge—what you know when you know a language—is often referred to as linguistic competence and differs from what you might consciously know about language. For example, most English-speaking people know the word child—they know how it sounds and how it looks in written form, they understand what it means, they know its plural form is children, and they can use it appropriately in a sentence. Their knowledge of the word child is part of their linguistic competence, regardless of whether or not they know that it is a noun. They also know what can and “can’t be” sentences in English, an important point that we will return to later. Test your linguistic knowledge of some English sentences by completing Exercise 1 below.

Exercise 1

Linguistic competence—sentences in English A. Working by yourself, identify which sentence in each of the following pairs of sentences sounds better to you. Put an asterisk (*) before the one in each pair that sounds worse. * is a symbol used by linguists for non-occurring forms. 1 (a) (b) 2 (a) (b) The girl thinks that the boy admires himself. The girl thinks that the boy admires herself. What did John eat ice cream and? What did John eat ice cream with?

First Language Acquisition 3 (a) (b) 4 (a) (b) 5 (a) (b) 6 (a) (b) 7 (a) (b) 8 (a) (b) 9 (a) (b) 10 (a) (b) John ate yesterday. John devoured yesterday. There’s the girl I talked to yesterday. There’s the girl I talked to her yesterday. Mary slept the baby. Mary diapered the baby. The people are sue the doctor. The people are suing the doctor. The horse is jumping the fence over. The horse is jumping over the fence. Mary bought the dress green. Mary bought the green dress. Bill is eating pizza. Bill have eating pizza. He is watching the football game. Is watching the football game.


B. In small groups, compare your judgments to those of other students. Your answers should be pretty comparable to your peers, even if you cannot explain why one sentence sounds better than the other.

To find out how children learn language, early researchers observed children, kept diaries of their utterances—that is, what children said—and looked for evidence of the processes by which acquisition took place. Today, researchers doing such observational, naturalistic methods of investigation often audio or videotape children as they speak in such everyday settings as eating meals and playing with friends. Researchers often share these data in their original form, or transcribed into written form, through a searchable, online, computer database known as the CHILDES database (MacWhinney, 1991). Other researchers have developed a number of experimental methods to figure out what children know about language. These include: imitation tasks (children repeat what they hear); act out tasks (children “act out” what they hear using small toys); judgment tasks (children judge words or, as you did in Exercise 1 above, sentences); and production tasks (children are prompted to produce language)—each designed to tap into children’s linguistic competence in a slightly different way. The way that children respond in all these tasks is with their linguistic performance: how they use their competence to produce and comprehend language. (See Table 11.1 for a summary of research techniques.) A particular challenge is learning about what infants perceive, given their limited mobility and language abilities, and researchers have come up with some clever methods for working with infants. For example, the high-amplitude sucking (HAS) technique uses a rubber nipple attached to a device that records the pressure on the nipple as babies suck. This technique gives researchers information about babies’ ability to distinguish sounds. Babies suck harder when they hear new sounds; then, as they become accustomed (habituated) to the sound, sucking pressure decreases. If a new sound is introduced and babies suck harder in response to it, researchers have evidence that babies can distinguish the new sound from the previous one. Another technique for conducting research with babies is the head-turn technique, which takes advantage of babies’ ability to turn their heads, and trains them to do so in response to sound changes.


Graziano-King and Smith Cairns

Table 11.1
Research Methodologies in FLA Research Diary Studies Databases Imitation Tasks Act Out Tasks Judgment Tasks Production Tasks High-Amplitude Sucking Technique Head-Turning Technique

Research in FLA is robust, but two things are worth noting about the field. First, although there were some very early child language studies (e.g. Darwin, 1877), the field did not really explode until the 1960s. So, in comparison to other scientific fields, FLA research is relatively young. Many questions that have been raised by researchers have not yet been answered, or the answers themselves have raised additional questions. So it is important to recognize that more work needs to be done and that research is ongoing. Second, as in other scientific fields, researchers do not agree on everything; for example, they may consider some kinds of data to be more important than others, interpret the same data differently, or have different underlying assumptions. As a result, they may draw different conclusions about how language is learned. While there is general agreement that FLA is achieved through interaction between the developing child and the speech of the language community, the main difference among acquisitionists regarding their views of FLA concerns the relative contributions of the child (nature) and the environment (nurture) to the task. We will return to this issue at the end of the chapter, but we believe that it is the contribution of the child (nature) that is primary in an account of FLA, though we recognize that not all linguists share this perspective.

The Course of FLA
Next, we will consider how language acquisition unfolds. We will consider four major aspects of language—phonology, the lexicon, morphology, and syntax—and we will consider them separately. However, it is important to remember that all aspects of language and language use are developing at the same time. For instance, in the first six months of life, even though babies can only make soft coos and gurgles, they achieve eye contact and like to play interactive, turn-taking “games” with adults (Stern et al., 1975)—the beginning of interactive communication.

Language Development from Birth to Age 4 Learning Sounds: Phonological Development Children must learn the phonology of their native language. This refers to the speech sounds of the language, as well as its intonational patterns, or rhythm. Infants seem to be specially attuned to the sounds of language. Studies have shown that newborn babies prefer to listen to their mother’s voice than to voices of other people. They also prefer the language of their community and other languages with the same rhythm (Mehler et al., 1988). These findings suggest that some speech is available to infants in the womb.

First Language Acquisition


An important aspect of all human languages is that each has an inventory of phonemes. These are the speech sounds that are perceived as distinct sounds. While all languages have phonemes, all languages do not share the same inventory of phonemes. For instance, in English, /l/ and /r/ (which are produced in relatively similar ways) are phonemes and so English speakers hear lip and rip as different words. But a language such as Korean does not have /l/ and /r/ as phonemes, so speakers of Korean do not hear a difference between lip and rip. On the other hand, Korean distinguishes between the phonemes /ph/ and /p/, while English does not. (The first is aspirated—followed by a puff of air—as in the English word pin; the second is unaspirated—no puff of air—as in the English word spin.) So while Korean speakers hear [phul] (grass) and [pul] (fire) as distinct words, English speakers do not. (Slash marks denote phonemes; brackets show how the phoneme is produced by a speaker.) However, this inability to perceive the difference between similar sounds that are not phonemes in the native language is not the case for infants. When they are about six months old, babies can perceive distinctions between all pairs of speech sounds, even those that are not phonemes in the language they are acquiring. However, by the time they are about a year old, this special perception has disappeared, and, like adults, they perceive as distinct only those speech sounds that are phonemes in their language (Werker and Lalonde, 1988). Of course, if these children need to learn another language, they can recover the perceptual distinctions needed for the new language. It is as though all infants are born prepared to acquire the phonemic system of any language in the world. Then experience with their own native language allows them to narrow the field down to just what they need for their language. We think this is a metaphor for all language development. All infants are born with the capacity to acquire any of the world’s languages, and experience with the language of their environment allows them to develop that particular language. As noted above, the earliest sounds babies make are soft coos and gurgles. At about six months of age “babbling” begins and babies produce syllable-like sounds: “ba,” “da,” “ga,” etc. As babbling becomes more complex, the syllables become connected, so babies produce, for instance, “bababa” and later “baga.” Stretches of babbling also take on the rhythm and intonation of sentences in the language the children are acquiring, making babies sound as though they are making statements or asking questions. As infants develop, there is a great change in the speech organs. At birth the larynx (vocal folds and accompanying structures) is high and close to the mouth, and the tongue is proportionally much larger than that of the adult, so the vocal tract is very small. During the first year, the larynx moves down and the control mechanisms in the brain that are responsible for articulatory timing and precision develop (Lieberman, 2006). This is why early words contain only “simple” speech sounds, like /b, d, g, k,/ called stop consonants, and vowels; the production of these speech sounds, which have either complete closure or complete opening of the vocal tract, does not require a great deal of vocal tract control. Only later will children begin to use more “complex” sounds, like sh and s, where the mouth must be only slightly open and air must be forced through the teeth and the narrow passage at the palate and alveolar ridge, respectively. For a similar reason, in their early words children do not produce clusters, like the sk sounds at the beginning of school, instead pronouncing it “kool.” This is an example of cluster reduction, one strategy that children use to tackle difficult sounds, and such strategies go on for several years. As infants age, their vocal control mechanisms mature, as well as the neural systems of the brain that are responsible for controlling speech. Consequently, their speech becomes gradually more adult-like (Nittrouer et al., 1989).


Graziano-King and Smith Cairns

Learning Words: Lexical Development At the same time that children are learning the phonology of their language, they are also learning its words. As we discuss word learning—and all language learning for that matter—it is important to recognize that much of what we say about what children do varies a good deal from one child to the next. So when we say “Children do this” or “Children do that,” we are speaking in general terms and recognize that the behaviors of individual children vary.
Identifying words in running speech Ask someone you know who speaks a language that you do not know to say a few sentences at their normal speaking pace. Try to break the sentence into words. Check with your speaker to see if you were successful. How successful were you? How easy or difficult was this to do? How do you think your experience compares to that of children learning their first language?

Exercise 2

As you can see from Exercise 2, when you don’t know a language, word boundaries— where words begin and end—are not obvious. And yet children manage to “cut up” or segment the continuous speech stream in order to begin acquiring words. Experiments (e.g. Mandel, Jusczyk, and Pisoni, 1995) have shown that babies as young as five months old can pick out familiar words from running speech. Research suggests that babies can recognize the recurring patterns of words they hear repeatedly, or that they may use cues such as stress and rhythm (e.g. Thiessen and Saffran, 2003). Adults who know a language have vocabularies of roughly 50,000–100,000 words, with receptive vocabularies (the words they understand) being larger than their productive vocabularies (the words they are able to use in speech and writing). Linguists refer to speakers’ vocabularies, or mental dictionaries, as their lexicons. So, as children are learning words, they are building a lexicon that contains information about the words they learn, including information about how words sound, what they mean, and how they are used—in sentences and in social contexts. It is important to recognize that children (and adults) build their lexicons in two ways: they learn more words and they learn more about the words they know.

Learning More Words Children’s first words generally appear between the ages of 10 and 15 months (Benedict, 1979; Bates, et al., 1994). Usually, a baby’s first word is not an “adult” word, but a string of sounds that the baby uses again and again to refer to the same thing, such as saying “baba” for “bottle” (Huttenlocher and Smiley, 1987). Researchers (e.g. Bates et al., 1994; Marchman and Bates, 1994) have examined the vocabularies of babies and discovered that the first fifty words a child learns are mostly nouns. At around the fifty-word mark, most children experience what is called the vocabulary spurt (Dromi, 1987), during which they acquire new words very quickly. By the age of 6, a child’s vocabulary is estimated to contain between 8,000 and 14,000 words. Beginning with one word at 12 months, this means that the child must acquire an average of four to eight new words every day during the preschool years. Learning More About Words Fast mapping is what allows children to learn so many words in so short a period of time. Through fast mapping, children learn a word with minimal exposure to it—often

First Language Acquisition


after only one or two encounters with it (Carey, 1978). Clearly learning words so quickly gives children only minimal information about words; more exposures to words in more contexts help children expand and refine their knowledge of words. Two interesting examples of the need for children to refine word knowledge are overextension and underextension of word meaning. When children overextend a word’s meaning, they give the word a scope of meaning that is too wide—for example referring to all four-legged animals as dogs. The opposite occurs with underextension. In this case, the scope of meaning is not wide enough. One of our children thought that being a “friend” only meant “someone who does a puzzle with you”; another, in his own bedroom, cried that he wanted to “go home,” thinking “home” referred only to the living room and not the entire apartment—probably because the entrance to the apartment was the living room, and when entering, we would say to him, “We’re home!”

Learning How to Build Words: Morphological Development As children learn words, they unconsciously begin to analyze them into smaller units of meaning. Linguists define a morpheme as the smallest unit of language that has meaning. The study of how words are formed from morphemes is called morphology. Some words, called simple words or monomorphemic words, are made up of only one morpheme; these are free morphemes. For example, dog and house are each free morphemes or simple words. (In fact, they can come together to make a compound word—doghouse.) Other morphemes are not free, i.e. by themselves they are not words; these are called bound morphemes. While the word dog is a single morpheme because it cannot be analyzed into smaller meaningful parts, the word dogs is made up of two morphemes: the free morpheme dog and the bound morpheme -s, a suffix (i.e. a bound morpheme that attaches to the end of words) which has the meaning of “more than one.” Note that prefixes, which attach to the front of words, are also bound. As children unconsciously analyze words, they identify morphemes and patterns for combining morphemes and use these to create words they have not yet heard. For example, in English the past tense of regular verbs is formed by adding a suffix, so we say washed, kissed, etc. Other, irregular verbs do not take -ed, so we say, for example, ate, went, etc. When children learning English learn the regular pattern for past tense formation (i.e. add suffix) they overregularize this pattern, producing words such as eated and goed (e.g. Marcus et al., 1992; Marcus, 1995). These overregularization errors—errors children make when they over-apply a pattern that they have recognized—serve as evidence that children are not simply relying on what they hear in their environment; they are applying the patterns of their language to produce forms they have never heard.
Exercise 3 Combining free and bound morphemes A. Do you think you know how to make a noun plural in English? Do you think you add the suffix -s? That is what most speakers of English would say, and it may be true when we write (though sometimes we add -es in writing), but it is not true when we speak. On your own or in small groups, say the plural of the following; for each, stress the last sound you make: 1 2 3 4 5 dog lark cat bird grub


Graziano-King and Smith Cairns 6 7 8 9 10 pup giraffe dove wug blick

B. What did you notice about the last sound you made when you pluralized dog, bird, grub, dove, and wug? C. What did you notice about the last sound you made when you pluralized lark, cat, pup, giraffe, and blick?

In Exercise 3, you were able to apply a morphophonemic rule of English to the plural morpheme when it attached to two made-up words: wug and blick. A morphophonemic rule determines how a bound morpheme is pronounced. These same words were used in a famous early FLA experiment (Berko, 1958) to demonstrate that young children acquire the morphophonemic patterns or rules of their language. Using drawings that were given made-up, or nonce, names, children were prompted to produce items such as plural and possessive nouns and third-person singular present tense verbs (where the -s can sound like s (cats, cat’s, pats), z (boys, boy’s, runs), or iz (foxes, fox’s, watches)) and past tense verbs (where the -ed can sound like t (parked), d (hugged), or id (glided)). For example, children were shown a drawing of a bird-like creature and were told, “This is a wug.” (Wug is a nonce word, so children could not have heard it before.) They were then shown a picture with two such birdlike creatures and were told, “Now there is another one. There are two of them. There are two ————,” and children were prompted to complete the final sentence. The children, just as you did for the plural, produced the correct forms of the nonsense words, although they had, of course, never heard them. The ability to do this means that children do not need to hear a word before they can produce it; they can extract the rules of the language and produce forms, in this case suffixed words, that they never heard before.

Learning How to Build Phrases and Sentences: Syntactic Development Possibly the most interesting aspect of all human languages is the fact that each has its own syntax. This refers to the way sentences are formed in the language. A fundamental feature of all languages is that their sentences must conform to certain syntactic patterns. We don’t simply say strings of individual words; they must be organized into a particular structure to carry the meaning we intend. The most basic aspect of syntax is word order. English standard word order is Subject-Verb-Object (SVO). The dog chased the cat conveys the meaning that the dog did something to the cat. The cat chased the dog means something else entirely. Other languages may have different word order; for example, word order in Japanese and Hindi is SOV and in many Native American languages it is VSO. There are many other more complicated structures involved in syntax. For instance, to create a Yes/No question in English, we move the auxiliary verb (sometimes called the helping verb) to the front (in sentences without auxiliary verbs, we put do before the subject or, if the main verb is a form of the verb to be, we move that). So to make a question of The kitten is playing with string we move the auxiliary is to the front for Is the kitten playing with string? This is called subject/auxiliary inversion. Languages also provide syntactic forms for sentences to be embedded to make longer, more complex sentences. Thus, we can say You think [the kitten is playing with string], or I know [you think [the kitten is playing with string]]. These are only two examples of the complex syntactic rules the developing child must acquire.

First Language Acquisition


While children are learning their first fifty words, their “sentences” consist of single words; this is called the holophrastic stage. So, if a baby says “baba” while reaching for a bottle, baba might mean “Give me the bottle,” or, in the absence of a bottle, might mean “Where is the bottle?” or “I want a bottle.” When the lexicon is around fifty words, the words begin to be combined into two-word sentences in which the two words are given the intonation pattern of a sentence. Even at this relatively early stage, children’s sentences conform to the standard word order of the language. A child learning English will use sentences such as See doggie (VO) or Daddy go (SV), both consistent with the English SVO pattern. Experiments have shown that even before children learning English demonstrate their knowledge of word order by producing word combinations, they show receptive knowledge of word order (Hirsh-Pasek et al., 1996).

Exercise 4

Child sentences The speech of children younger than 3 who are learning English has some very predictable features. Working in small groups, indicate which sentence you would expect to hear earlier for each set of sentences below and discuss the reason for your choice. 1 (a) (b) (c) (a) (b) (a) (b) (a) (b) (a) (b) (a) (b) Puppy play. Puppy playing. The puppy is playing. Mommy goes. Mommy go. Baby in highchair. The baby is in the highchair. I show you again. I will show you again. Sun not shining. No sun shining. Why is the dog barking? Why the dog is barking?

2 3 4 5 6

After the two-word stage, children’s sentences become longer, but are still short and incomplete. However, children do not produce words in just any order; their sentences seem like “edited” versions of adult sentences, suggesting that children are using some sort of syntax to produce their sentences—just not yet exactly the syntax used by the adults around them. Some examples from a 23-month-old girl are: “Mommy like it,” “Daddy push in swing,” and “Where go Mom?” While all children produce such sentences, the information that is omitted varies depending upon the language they are acquiring. English-speaking children tend to leave out suffixes and function words, such as the articles the, a, and an. Instead, they include those words that communicate the most meaning. The inclusion of bound morphemes and function words comes over time. So, even though they both contain the same number of words, a sentence such as Mommy likes it (which contains three words and four morphemes) is more advanced than Mommy like it (which also contains three words but only three morphemes). For this reason, children’s sentences beyond the two-word stage are not measured by the number of words, but by the number of morphemes they contain. The average length of their sentences, measured in morphemes, is an indicator of the maturity of a child’s


Graziano-King and Smith Cairns

language in the early years. This measure is called MLU for mean length of utterance, and it gradually increases after the child moves away from the two-word stage. Child language shares many common features. For example, early questions go through similar stages for all children learning English. At first, children simply form questions with rising pitch at the end of the utterance and do not use subject/auxiliary inversion (which we discussed earlier) for Yes/No questions; later they do. Questions beginning with who, what, when, where, why, and how, called wh-questions, must also undergo subject/auxiliary inversion. At first, children form wh-questions without moving the auxiliary, so they produce questions like Why the kitten is chasing a toy mouse? Gradually, they learn that it is necessary to use subject/auxiliary inversion in both types of question (Weinberg, 1990). Another common feature of English child language is the development of negative sentences. In early negative sentences, the negative word comes at the beginning of the sentence—for example, No Daddy want cake. In the next stage, the negative word comes between the subject and verb—what is known as internal negation (Daddy no want cake). Finally, when children learn about auxiliary verbs, they produce the almost adultlike form (Daddy don’t want cake). The completely adult Daddy doesn’t want cake is later still. (Klima and Bellugi, 1966, were the first to describe the stages of acquisition of negation.) Like the overregularization of bound morphemes mentioned above (the examples eated and goed), these early sentence patterns show that children are producing forms they could never have heard spoken by the adults in their environment. As the sentences children produce become increasingly more adult-like, so does their interpretation of sentences. Adults know that in a sentence such as Grover scratched himself, himself must refer to Grover. Experiments have shown, however, that until children are about 4 years of age, they misinterpret this sentence to mean that Grover scratched either himself or any other male character (Cairns et al., 1994). Similarly, they know that in the sentence Grover told Cookie Monster to jump over the fence, it is Cookie Monster that will jump, but they also think that Cookie Monster will do the jumping in the sentence Grover kissed Cookie Monster after jumping over the fence (Cairns et al., 1994). The fact that children learning English do come to produce correct questions and negative sentences and learn to correctly interpret sentences without any instruction from the adults in their environment is an important point that we will return to below.

Exercise 5

Reconsidering child sentences Go back to the child sentences in Exercise 4 and review your choices in light of the previous discussion.

Language Development at Age 4 and Beyond
By the age of 4, children have acquired all the basic morphological and syntactic patterns and processes—that is, the morphosyntax—of their language. However, there are important linguistic developments that take place beyond the age of 4. Vocabulary continues to increase rapidly into the school years, and even adults can expand their vocabularies throughout their lifetime. In the pre-school and early school years, children acquire many new skills in their ability to use language. While very young children can have conversations, they obviously lack the conversational skills of adults. Gradually, children learn to maintain or change conversational topics and take the perspective of the person they are conversing with (Brinton and Fujiki, 1984). They learn about the

First Language Acquisition


narrative structure of stories, so they can understand and also produce stories (Crais and Lorch, 1994; Johnson, 1995). They come to understand that different people must be addressed in different ways and that there are social conventions governing the use of language. Non-literal language, such as sarcasm and metaphor, comes to be both understood and produced (Capelli, Nakagawa, and Madden, 1990; Karadsheh, 1991). All of these developments, which we call pragmatics, allow children to successfully use their language appropriately in their language community. Arguably the most important acquisition beyond the age of 4 is that of metalinguistic skills. This refers to the ability to think of language as an object, beyond its simply being a device for communication. People with metalinguistic skills can separate the form of language from its content and address language form independently. There are a number of different forms of metalinguistic awareness that develop. One is phonological awareness, which is the consciousness that speech is actually composed of individual speech sounds. Four-year-olds can be taught to tap out the syllables of words (for instance one tap for but, two for butter, and three for butterfly), but not until the age of 6 can children perform the same task with individual speech sounds, for instance, three taps for cat, four for nest, and so on. (Liberman et al., 1974, were the first to demonstrate children’s phonological awareness in this way.) There are many other tests of phonological awareness, such as asking the child what play sounds like if you remove the “l.” Many studies have shown (e.g. Tunmer et al., 1988) that good phonological awareness skills predict successful early reading. Tests of phonological awareness are included in assessments of reading readiness and are taught in remedial reading programs. Since the basis of reading is the ability to know the speech sound represented by each letter (known as grapheme/phoneme correspondence), it is reasonable that an ability to perceive speech sounds is a necessary prerequisite. There are other metalinguistic skills that are predictors of successful early reading, such as syntactic awareness, which is the ability to discriminate well-formed sentences from ill-formed sentences and to correct the latter. Children exhibit syntactic awareness, for instance, if they can say that The monkey is climbing the tree up is incorrect and should be The monkey is climbing up the tree (Cairns et al., 2006).

Exercise 6

Detecting ambiguity A. Consider the following joke attributed to Groucho Marx: Last night I shot an elephant in my pajamas and how he got in my pajamas I’ll never know. B. In small groups, discuss the source of the humor of this joke. Share other jokes you know that hinge on word-play.

The ability to detect ambiguity, or multiple meanings, as you just did in Exercise 6, is not only the source of the humor of many jokes, but also related to reading skill (Cairns et al., 2004). Children exhibit this metalinguistic skill if they can report, for example, that The children saw the bat in the park has two meanings, based on the dual meaning of bat, and that the source of the humor in the joke in Exercise 6 is the fact that the phrase in my pajamas can apply either to the speaker or to the elephant. It appears that metalinguistic skills allow children to manipulate and process language in a way that is required for successful reading. The ability to understand written language rests on the same linguistic knowledge that underlies the ability to understand spoken language.


Graziano-King and Smith Cairns

By the time the developing child has become a young adult, acquisition is basically complete and the individual possesses full linguistic competence. An important ability of linguistically competent people is that they can discriminate grammatical from ungrammatical sentences, as you did in Exercise 1. Sentences are considered to be grammatical if they have been created by the morphosyntactic rules of the language. Every human language—including national languages and regional and social dialects—has morphosyntactic rules that generate the grammatical sentences of the language or dialect. Note that the grammaticality of a sentence does not depend on whether it is meaningful. Noam Chomsky created a sentence that has become famous in linguistic circles because it shows that a sentence that is nonsensical can still be grammatical: Colorless green ideas sleep furiously. Compare this to the ungrammatical: *Sleep green furiously ideas colorless. In Exercise 1, you used your metalinguistic skill and exhibited your ability to discriminate between grammatical and ungrammatical sentences, as children do when they exhibit syntactic awareness. Any theory of FLA must account for the fact that speakers are able to recognize ungrammatical sentences in their language. This is problematic for a theory of FLA because children are never told which sentences are ungrammatical. Yet they can’t assume that if they haven’t heard a sentence, it must be ungrammatical! Throughout our lives we produce and hear brand-new sentences that we have never heard before—sentences that are perfectly grammatical. How can children learn what is ungrammatical only from hearing grammatical sentences? We will revisit this important question when we consider how a first language is acquired.

How a First Language is Acquired
Now that you have some information about the language production of babies and children, we can ask how language is learned. To approach this question, we first consider two myths about language learning.

Myth 1: Children Must Be Taught Language
Nothing could be further from the truth. First, speakers are not overtly aware of the morphosyntactic rules that make up their linguistic competence, so how could caregivers possibly teach what they are not conscious of? Second, caregivers rarely try to teach language. They might teach some words and something about the appropriate social norms of language use, for example to say please and thank you, but little else. Instead, we just have conversations with our children and they are able to create a grammar—an internal set of morphosyntactic rules that generate the sentences of the language. Even word meanings, apart from simple reference, are not explicitly taught. Earlier we mentioned that one of us had a child who (at 2½) thought being a “friend” meant sitting on the floor with him, putting together a jigsaw puzzle. By the time he was 4, he knew exactly what a friend was, without having been taught the nuances of friendship. If you think that language acquisition depends upon children being taught language, consider cultures such as Kaluli and Samoan, where speech to children is not modified in any way to accommodate children’s interest or desire to be conversational partners, and yet language is acquired (Schieffelin and Ochs, 1986). Tied to the notion of teaching is the notion of correction. Caregivers might correct for content (No, that’s not a horse, it’s a cow), but they rarely correct grammar (Brown and Hanlon, 1970). Even if they do, they do not do it consistently, every time the child produces some ungrammatical form and, in any case, it does not help. We know a young mother who was horrified when her child said things like eated and goed and tried to

First Language Acquisition


correct him each time. It did absolutely no good. There is a classic story in the acquisition literature in which a mother tries to get her child to say Nobody likes me instead of Nobody don’t like me. After several tries, the child exclaims, “Oh, nobody don’t likes me!” (McNeil, 1966). Because a first language is not taught, and because teaching is often associated with learning, linguists talk about a first language as being acquired instead of learned. It is this point that Maria Montessori makes in the quotation that opened this chapter.

Myth 2: Children Simply Imitate Adults
The fact that children produce forms they have never heard, such as eated and goed, is evidence that they could not simply be imitating. Also, they are unable to imitate forms that are beyond their current level of linguistic ability. We saw examples of this in our own children’s speech. In one case, the child was read the story “The Gingerbread Man” every night and repeated each sentence as his mother read it. However, because he had not yet reached the stage of internal negation, when attempting to repeat the line Run, run as fast as you can, you can’t catch me—I’m the Gingerbread Man!, he consistently recited, “Run, run as fast as you can, not you catch me—I’m the Gingerbread Man!” There is a great deal of experimental evidence that children’s imitations are not verbatim repetitions, but reflect their own developing internal grammar (Lust et al., 1996). Although very few aspects of language are “taught,” and even though language acquisition is not dependent on correction and involves much more than imitation, we do know that to acquire language, children need to hear the language of their community. This is why deaf children don’t acquire spoken language, although they can acquire sign language—which is a fully human language (see Petitto, 1994)—if they are exposed to it. Understanding how language is acquired, then, requires considering what children do in the context of their linguistic environment.

The Linguistic Environment
The task of children who are learning language is to take the speech that they hear in the environment and create the lexicon and morphosyntactic rules (the grammar) of the language being acquired. We will begin by describing the speech the child is exposed to, which provides the data or input to the child for the grammar-creating process. Then we will move to consideration of how the child carries out the construction of a grammar from that input. The speech the child hears is, in many respects, very “noisy.” We mentioned earlier the fact that words are run together, so there is no obvious beginning and end to the words in the sentences the child hears. It is also true that words are pronounced differently by different people and in different situations. Consider a very informal sentence like Whadchadolasnight? and a reply, like Wentt’thuhmovies, which is not even a full sentence. Sentence fragments, such as appear in conversations, are not complete sentences, although they are grammatical phrases or clauses. Children have the remarkable ability to isolate lexical items, identify bound morphemes, and perceive the syntactic patterns of the speech that they hear around them. At the beginning of this chapter, we pointed out that acquisitionists differ with respect to the relative contributions of the environment and the developing child to the task of FLA. People who are primarily interested in what the child extracts from the environment point, quite correctly, to the child’s ability to recognize patterns (which extends far beyond linguistic patterns) and to the influence of the frequency with which particular language forms occur in the speech of the environment. Other acquisitionists


Graziano-King and Smith Cairns

are more interested in those aspects of language and language organization that come from the child; this has been our emphasis in this chapter. However, as we noted earlier, whatever emphasis a particular acquisitionist will favor, there is general agreement that FLA is achieved through an interaction between the developing child and the speech of the child’s language community.

What Children Bring to the Task of Language Acquisition
The “natural” nature of language acquisition gives us an important clue as to how children acquire language. If you think of it, language development is much like other developmental skills, such as walking. No one needs to teach babies to walk; they just do it when they are ready. Children do, however, have to be taught to do things like tie their shoes, snap their fingers, and ride a bike. This is because walking is “natural” to humans, but bike riding is not. Similarly, developing language is “natural” to humans in a very profound way. If children fail to acquire language, it is considered a pathology, as if they failed to begin walking or feeding themselves. Furthermore, just as they acquire the ability to walk on a similar timetable everywhere, all children, no matter what language they are acquiring, go through basically the same stages of language acquisition. They all acquire the patterns of their language and often over-use them. We can also get some insight into language development by examining adult language. There is by now no doubt that language knowledge is stored in the brain. For most right-handed people, it is stored in the left cerebral cortex. Damage to the language centers of the brain, by injury or a stroke, results in impaired language. So we think that as children are acquiring language and it is being stored in the brain, maturation of language is tied to maturation of the brain and its neural circuitry. This explains the common features of language acquisition among all people. Human beings have human brains. Language is stored in the brain. Therefore, everyone with a human brain will acquire language in a similar way. The work of the acquisitionist Dan Slobin (1985) and others indicates that children acquire language similarly no matter what their native language is. Their early sentences have common characteristics and they make similar kinds of systematic errors. These similarities are the result of the fact that all children have similar acquisition principles. All developing brains accept linguistic input and organize that input into the lexicon and grammar of their native language. If for some reason children haven’t acquired a first language by the age of about 12, they can never fully acquire the syntax and morphology of any language. The period from birth to (approximately) 12 is, therefore, known as the critical period for language acquisition. There is a famous case of a girl, given the pseudonym Genie, who was locked in her room and did not hear speech for 13 years (Curtiss, 1977; Rymer, 1993). When she was discovered by social workers, attempts were made to teach her language. She learned many words and was able to communicate to some extent using just her vocabulary and some strategies for forming sentences, but her syntax and morphology never fully developed. Since deaf children learning sign language go through the same stages of language acquisition, deaf children deprived of sign language during the critical period will not, in adulthood, become fluent “speakers” of sign language. Like Genie, they will learn words and a rudimentary syntax and morphology, but not the complete morphosyntax of their language. Well, you might ask, if all humans store language in the brain in a similar way, why doesn’t everyone speak the same language? This is actually a very good question. The diversity of human language is probably related to the geographical location of peoples, as well as to their history. Language is dynamic; it is always changing, so as different

First Language Acquisition


languages have changed in different ways, they have grown to be very different. Depending on how you count, there are between 3,000 and 5,000 different human languages. However, the really interesting fact is that, as much as languages differ, they are also very much the same. All human languages seem to be “cut from the same mold” or share what has been called Universal Grammar (UG). It seems to be that the human brain is set up to organize language in a particular way—all languages have phonology, morphology, and syntax—no matter what the “details” of an individual language are. Furthermore, there are many ways a human language can not be; that is, there are structures that are impossible in any human language. For example, there are a limited number of ways that human languages can make questions. English does it by moving words around. Recall our example of making The kitten is chasing a string into a question by moving is to the front, for Is the kitten chasing a string? Many languages move things around; others put suffixes or prefixes onto key words to indicate a question. But no language makes a question by reversing the words in a sentence, or even by moving the second or third word to the front (or the end). No language has a word that has a different meaning on different days of the week. All the many ways languages cannot be are called constraints on human language. What did John eat ice cream and? (example 2(a) from Exercise 1) violates a constraint on all human languages against breaking up coordinates (such as ice cream and cake). It may seem counter-intuitive, but the range of possible human languages is actually quite small compared to the range of impossible languages. So, we can answer the question What do children bring to the task of acquiring a language? in the following way: The developing brain is structured in such a way that it will organize the speech of the child’s environment into a grammar that is structured as all human languages are structured and violates none of the constraints. In the light of this understanding of the development of grammar, let us return to the topic of grammaticality and ungrammaticality. A sentence is said to be grammatical in a particular language if it can be generated by the grammar of that language. People who know the language can recognize as grammatical any sentence that can be generated by the grammar of their language and as ungrammatical any sentence that violates that grammar. Thus, it is not mysterious that people can recognize grammatical sentences they have never heard and identify as ungrammatical sentences whose ungrammaticality has never been pointed out to them.

In this chapter we considered some of what a child needs to know in order to know a language, and we looked at the course of FLA. Table 11.2 summarizes milestones covered in the early linguistic development of children.

Table 11.2: Milestones in Language Development
At birth: Infants have been attuned to language in the womb, especially their mother’s voice and general prosodic characteristics of their target language. First 12 months: Infants acquire the ability perceptually to distinguish the phonemes of their language from those of other languages.


Graziano-King and Smith Cairns In the second half of the first year, segmental babbling develops, along with the melodic characteristics of the target language. In the second half of the first year, infants develop the ability to segment individual words from running speech. 1–2 years: The first word appears generally between 10 and 15 months, although there is individual variation. Children produce one-word “sentences” until they have acquired a vocabulary of around fifty words. After the vocabulary has expanded to fifty words, children begin combining them into two-word sentences. At this point most children undergo a vocabulary “spurt,” during which more words are learned very rapidly. 2–4 years: Bound morphemes are gradually acquired. Sentences become longer, but early ones lack many function words and bound morphemes. Around the age of 3, complex sentences begin to be produced. Beyond age 4: Sentences become more complex and no longer lack function words or bound morphemes. Conversational skills mature and become more adult-like. Metalinguistic skills develop.

We also considered the linguistic environment, as well as what children bring to the task of language acquisition. A critical feature of FLA is its similarity among children acquiring the same language as well as children acquiring other languages. This fact allows us to conclude that all human children create their lexicons and grammars in the same way, through an interaction of the linguistic input of the environment and the child’s developing brain—this is the “painstaking teacher” that, as Maria Montessori (1967) notes, exists inside every child—to which no expert, no language instructor can be compared.

Bates, E., Marchman, V., Thal, D., Fenson, L., Dale, P., Reznick, J. S., Reilly, J., and Hartung, J. (1994) Developmental and stylistic variation in the composition of early vocabulary. Journal of Child Language, 21(1), pp. 85–123. Benedict, H. (1979) Early lexical development: Comprehension and production. Journal of Child Language, 6(2), pp. 183–200. Berko, J. (1958) The child’s learning of English morphology. Word, 14, pp. 150–177. Brinton, B. and Fujiki, M. (1984) Development of topic manipulation skills in discourse. Journal of Speech, Language, and Hearing Research, 27, pp. 350–358. Brown, R. and Hanlon, C. (1970) Derivational complexity and order of acquisition in child speech. In Hayes, J. R. ed. Cognition and the development of language. New York, Wiley-Blackwell, pp. 11–21, 37–53.

First Language Acquisition


Cairns, H. S., McDaniel, D., Hsu, J. R., and Rapp, M. (1994) A longitudinal study of principles of control and pronominal reference in child English. Language, 70(2), pp. 260–288. Cairns, H. S., Schlisselberg, G., Waltzman, D., and McDaniel, D. (2006) Development of a metalinguistic skill: Judging the grammaticality of sentences. Communication Disorders Quarterly, 27(4), pp. 213–220. Cairns, H. S., Waltzman, D., and Schlisselberg, G. (2004) Detecting the ambiguity of sentences: Relationship to early reading skill. Communication Disorders Quarterly, 25(2), pp. 68–78. Capelli, C. A., Nakagawa, N., and Madden, C. M. (1990) How children understand sarcasm: The role of context and intonation. Child Development, 61(6), pp. 1824–1841. Carey, S. (1978) The child as word learner. In Halle, M., Bresnan, J., and Miller, G. A. eds. Linguistic theory and psychological reality. Cambridge, MIT Press, pp. 264–293. Crais, E. R. and Lorch, N. (1994). Oral narratives in school-age children. Topics in Language Disorders, 14(3), pp. 13–28. Curtiss, S. (1977) Genie: A psycholinguistic study of a modern-day “wild child.” New York, Academic Press. Darwin, C. (1877). A bibliographical sketch of an infant. Mind, 2(7), pp. 285–294. Dromi, E. (1987) Early lexical development. Cambridge, Cambridge University Press. Hirsh-Pasek, K., Golinkoff, R. M., and Naigles, L. (1996) Young children’s use of syntactic frames to derive meaning. In Hirsh-Pasek, K. and Golinkoff, R. M. eds. The origins of grammar: Evidence from early language comprehension. Cambridge, MIT Press, pp. 123–158. Huttenlocher, J. and Smiley, P. (1987) Early word meanings: The case of object names. Cognitive Psychology, 19(1), pp. 63–89. Johnson, C. J. (1995) Expanding norms for narration. Language, Speech, and Hearing Services in Schools, 26, pp. 326–341. Karadsheh, R. (1991) This room is a junkyard!: Children’s comprehension of metaphorical language. Proceedings of the Biennial Meeting of the Society for Research in Child Development, April 1–3, 1991, Seattle, USA. Klima, E. S. and Bellugi, U. (1966) Syntactic regularities in children’s speech. In Lyons, J. and Wales, R. J. eds. Psychological papers. Edinburgh, Edinburgh University Press, pp. 183–208. Liberman, I. Y., Shankweiler, D., Fischer, F. W., and Carter, B. (1974) Explicit syllable and phoneme segmentation in the young child. Journal of Experimental Child Psychology, 18(2), pp. 201–212. Lieberman, P. (2006) Toward an evolutionary biology of language. Cambridge, Harvard University Press. Lust, B., Flynn, S., and Foley, C. (1996) What children know about what they say: Elicited imitation as a research method for assessing children’s syntax. In McDaniel, D., McKee, C., and Cairns, H. S. eds. Methods for assessing children’s syntax. Cambridge, MIT Press, pp. 55–76. McNeil, D. (1966) Developmental psychology. In Smith, F. and Miller, G. A. eds. The genesis of language. Cambridge, MIT Press, pp. 15–84. MacWhinney, B. (1991) The CHILDES Project: Tools for analyzing talk. Hillsdale, Erlbaum. Mandel, D. R., Jusczyk, P. W., and Pisoni, D. B. (1995) Infants’ recognition of the sound patterns of their own names. Psychological Science, 6, pp. 314–317. Marchman, V. A. and Bates, E. (1994) Continuity in lexical and morphological development: A test of the critical mass hypothesis. Journal of Child Language, 21(2), pp. 339–366. Marcus, G. F. (1995) Children’s overregularization of English plurals: A quantitative analysis. Journal of Child Language, 22(2), pp. 447–459. Marcus, G. F., Pinker, S., Ullman, M., Hollander, M., Rosen, T. J., and Xu, F. (1992)


Graziano-King and Smith Cairns Over-regularization in language acquisition. Monographs of the Society for Research in Child Development, 57(4), pp. 1–182. Mehler, J., Jusczyk, P., Lambertz, G., Halsted, N., Bertoncini, J., and Amiel-Tison, C. (1988) A precursor of language acquisition in young infants. Cognition, 29(2), pp. 143–178. Montessori, M. (1967) The absorbent mind. New York, Holt, Rinehart, and Winston. Nittrouer, S., Studdert-Kennedy, M., and McGowan, R. S. (1989) The emergence of phonetic segments: Evidence from the spectral structure of fricative–vowel syllables spoken by children and adults. Journal of Speech and Hearing Research, 32(1), pp. 120–132. Petitto, L. A. (1994) Are signed languages “real” languages? Evidence from American Sign Language and Langue des Signes Québecoise. Signpost (International Quarterly of the Sign Linguistics Association), 7(3), pp. 1–10. Rymer, R. (1993). Genie: A scientific tragedy. New York, Harper Perennial. Schieffelin, B. B. and Ochs, E. (1986) Language socialization. Annual Review of Anthropology, 15, pp. 163–191. Slobin, D. I. (1985) Crosslinguistic evidence for the language-making capacity. In Slobin, D. I. ed. The cross-linguistic study of language acquisition: Volume 2, theoretical issues. Hillsdale, Lawrence Erlbaum, pp. 1157–1249. Stern, D. N., Jaffe, J., Beebe, B., and Bennett, S. L. (1975) Vocalizing in unison and in alternation: Two modes of communication within the mother–infant dyad. In Aaronson, D. and Rieber, R. W. eds. Developmental psycholinguistics and communication disorders. New York, Annals of the New York Academy of Sciences, pp. 89–100. Thiessen, E. D. and Saffran, J. R. (2003) When cues collide: Use of stress and statistical cues to word boundaries by 7- to 9-month-old infants. Developmental Psychology, 39(4), pp. 706–716. Tunmer, W. E., Herriman, M. L., and Nesdale, A. R. (1988) Metalinguistic abilities and beginning reading. Reading Research Quarterly, 23(2), pp. 134–158. Weinberg, A. (1990) Markedness versus maturation: The case of subject–auxiliary inversion. Language Acquisition, 1(2), pp. 165–194. Werker, J. F. and Lalonde, C. E. (1988) Cross-language speech perception: Initial capabilities and developmental change. Developmental Psychology, 24(5), pp. 672–683.

Other Resources
Cairns, H. S. (1996) The acquisition of language (Pro-ed studies in communicative disorders). Austin, Pro-Ed. Fletcher, P. and MacWhinney, B. (1996) The handbook of child language. Oxford, Blackwell. Hoff, E. (2005) Language development. 3rd edn. Belmont, Wadsworth Thompson Learning. Pinker, S. (1994) The language instinct. New York, William Morrow. Pinker, S. (2000) Words and rules: the ingredients of language. New York, HarperCollins.

Fenson, L., Dale, P. S., Reznick, J. S., Bates, E., Thal, D. J., and Pethick, S. J. (1994) Variability in early communicative development. Monographs of the Society for Research in Child Development, 59(5), pp. 1–173. McDaniel, D., Cairns, H. S., and Hsu, J. R. (1990) Binding principles in the grammars of young children. Language Acquisition: A Journal of Developmental Linguistics, 1(1), pp. 121–138.

First Language Acquisition


The human language series: 2 (1995) Acquiring human language—Playing the language game. Ways of Knowing [video: VHS]. Wild child: The story of feral children (2002). Dir. Jonah Weston, Optomen TV [video: VHS].

CHILDES (2003) The Child Language Data Exchange System (CHILDES) [Internet]. Available from: [Accessed June 15, 2009]. Language Acquisition Research Center (2008) Language Acquisition Research Center: Hunter College, Department of Psychology [Internet]. Available from: http://www.hunter.cuny. edu/littlelinguist/ [Accessed June 15, 2009]. Linguist List (2007) Linguist List [Internet]. Available from: [Accessed June 15, 2009].

This Chapter Explores:
History of ASL Fingerspelling Iconic and Arbitrary Signs Phonology

ASL A Visual Language

Miako Villanueva, Deanna Twain, and Laura Leigh Wood

Morphology Noun to Verb Derivations Compounding Syntax

The truth about ASL is that until you really know it, you cannot appreciate it. Unfortunately, its stigma as an inferior language coerces us not to think of it as a real language until we learn enough of it to see that it is equal to, yet different from, spoken language. Oliva, 2004, p. 132

The authors of this chapter are second language learners of American Sign Language (ASL). We are excited to have the opportunity to explore some of the basics of ASL linguistics here and are indebted to the members of the deaf 1 community who have shared with us their rich culture and language. We hope that this chapter serves as an informative “appetizer” for those who do not yet know ASL and whets your appetite to look more deeply and seek more information, ideally leading to your own personal interaction with the deaf community, so that you too may truly know and appreciate it. MV: I was first introduced to American Sign Language and deaf culture when my interests in psychology, languages, and how people understand and communicate with one another converged in an ASL course I originally signed up for “just for fun” in college. I had the privilege of having a deaf teacher who incorporated deaf history, ASL jokes, and her own personal experiences as a deaf person into the course, imparting to us so much more than “the signs for things,” as I had originally expected to learn. My formative ASL years were spent as a dorm supervisor at the Colorado School for the Deaf and the Blind, where I engaged in a true immersion experience like no other. As an instructor of linguistics at Gallaudet University, I continue to learn from the deaf community, my colleagues, and especially my students every day. DT: My first exposure to American Sign Language came more suddenly than expected. As I

ASL: A Visual Language


emerged from the subway to register for a class in ASL at the New York Society of the Deaf, I was struck by the sight of two groups of young men, one on the south side of 14th Street and the other on the north side. They were all in the middle of a very heated argument in sign language. Hands were flying, faces were red, and sounds were noisily spitting back and forth across the street. It was a mesmerizing and exciting sight; I had never seen anything quite as clear and passionate, and I was hooked. My original reasons for wanting to “learn signs” were immediately eclipsed by a desire to study and truly learn this new and beautiful language, and I entered into my introductory ASL class with more serious intentions. I was fortunate to learn ASL from all-deaf instructors. I completed every class offered, including the interpreter training program, ending up with a Certificate of Interpreting. LLW: After my first semester of ASL at Marymount Manhattan College, I was feeling overly confident in my language skills and I decided I would work at a deaf camp for kids that summer. Somehow I had completely forgotten that I had (1) never met a deaf person, and (2) only had one semester of ASL. I showed up the first day of camp and quickly discovered I was the weakest signing staff member. Throughout that summer of miscommunications and unforgettable learning experiences, I realized that what I had thought would be a simple language to pick up was so much more. Since then my understanding of ASL, and my appreciation of it, have continued to grow.

Exercise 1 Before you read further, take this true/false (T/F) quiz. Then, after you’ve read the chapter see if your answers have changed. 1 ASL is a universal language. 2 ASL signers from different states have different “accents” and dialects. 3 ASL uses the same grammar as English. 4 Deaf people prefer to be called hearing impaired. 5 ASL signs are pantomimed gestures. 6 ASL is easier to learn than spoken languages. 7 ASL can be used to discuss abstract concepts. TF TF TF TF TF TF TF

This chapter outlines the main concepts in the linguistic study of American Sign Language (ASL), a language used by deaf people in the United States and a large part of Canada. While the study of languages has been around for centuries, the vast majority of research has focused on spoken languages; approaching the signs used by deaf people as full-fledged, natural languages in their own right and therefore equally worthy of linguistic study is a relatively new concept. The first documented linguistic studies of signed language in the United States were carried out in the late 1950s and early 1960s by a team of researchers at Gallaudet University, led by William Stokoe. Stokoe and his colleagues conducted groundbreaking research, applying linguistic principles to what was at the time referred to simply as “signing” (Stokoe et al., 1965). Their work provided evidence that the signs used by deaf people were not simply pantomimed actions or spoken language produced with the hands; in fact, there are rules for both how to create the individual signs (phonology and morphology) and how to put the signs together to form sentences (syntax or grammar). Showing that signed communication shares all the characteristics that define other natural languages proved that what Stokoe et al. (1965)


Villanueva, Twain, and Wood

termed American Sign Language is indeed a true language and laid the groundwork for the field of ASL linguistics (Valli et al., 2005). In one chapter it is, of course, impossible to cover all of the aspects of ASL linguistics or to do justice to the remarkable number of discoveries that have been made over the past fifty years in this still-emerging field. Therefore, the focus here is on the main areas of linguistic inquiry and topics that are most of interest to a wide audience. Before we delve into the linguistic concepts, however, it is important that we look back to the past, to situate ASL in its appropriate historical context and to understand its relationship to other signed languages and to spoken English.

History of ASL
Many people are surprised to learn that American Sign Language is not a universal language. In fact, there are many different signed languages used by communities of deaf people all over the world; over 120 signed languages have been documented by linguists, and it is likely others are in use that have not yet been identified (Ethnologue, 2009). Signed languages develop naturally out of humans’ need and desire to communicate with one another, just like with spoken languages, so anywhere there are groups of deaf people in close contact with one another for an extended period of time, a signed language is likely to develop. One of the main places that deaf people have historically come together is in residential schools, and thus most signed languages can trace their roots back to a particular location and a specific school. The roots of ASL reach back to the early 1800s and to Paris, France. In the early 1800s, a man named Thomas Hopkins Gallaudet traveled from the United States to Europe to study the educational methods used with deaf students there. Gallaudet had met a young deaf girl, Alice Cogswell, and become fascinated with trying to communicate with her and understand how she perceived the world. Alice’s father was convinced that Alice and other deaf children in the United States would benefit from a formal education, and he encouraged Gallaudet to further study methods of teaching the deaf in Europe. Gallaudet first went to England, but the Braidwood family who controlled the schools there were extremely protective of their methodology, which focused on teaching students using the oral approach, through speech and lipreading, rather than by using signs. After many frustrating months, it happened that Abbé Roch-Ambroise Sicard, director of the school for the deaf in Paris and notable authority on the education of the deaf, came to London to give demonstrations showing his success with the manual approach—a philosophy of deaf education that uses signed language as the mode of communication—using signs to communicate between teachers and students. Gallaudet attended a demonstration, put on by Sicard and two of his most notable students, Jean Massieu and Laurent Clerc, and was invited to return to Paris and learn the methods used there (Lane et al., 1996). After months of study in Paris, Gallaudet was ready to return to the States. He convinced Laurent Clerc, who had been one of his deaf teachers at the school in Paris, to accompany him back to the United States to help him establish a school for the deaf. The two men worked together at learning each other’s languages; Gallaudet continued to improve his French Sign Language (Langue des Signes Française, LSF) and Clerc began learning English. Together they founded the first school for the deaf in the United States, in West Hartford, Connecticut, in 1817. The school, initially named the Connecticut Asylum for the Education and Instruction of Deaf and Dumb Persons, began with a class of seven students and expanded quickly; thirty-one students hailing from ten states

ASL: A Visual Language


were enrolled by the end of the first year. (The school is still open, now called the American School for the Deaf, and currently claims over 4,000 alumni.) Many of these original students brought their own signs with them to the school in Hartford. These were home signs that had been developed by the students and their families in order to communicate; in some places small deaf communities existed and had developed signs (see Groce, 1985, for a description of Martha’s Vineyard as one example). The language that developed at the school was a combination of the students’ home signs and the LSF taught by Clerc. Over the years, graduates of the school and teachers who went there for training established schools for the deaf all over the United States, taking the signs they had learned at the West Hartford school with them (Lucas et al., 2003) (see Figure 12.1). In 1864, Thomas Hopkins Gallaudet’s son, Edward Miner Gallaudet, founded the first college for deaf students, which later became known as Gallaudet University. It was there almost 100 years later that William Stokoe and his colleagues first labeled the signs used by the deaf community as American Sign Language. A linguistic study comparing a sample of 87 signs in modern LSF and ASL found that 58 percent of them were cognates, meaning that the form of signs for the same concepts in the two languages are similar (Woodward, 1978). Linguists cite cognates as evidence that languages are historically related. For example, English and German are historically related and share many cognates: night and Nacht, false and falsch, idea and Idee, to name just a few. Similarly, ASL and LSF are historically related through their common ancestor, the signs used by Laurent Clerc in Paris and later in the United States.


Signed instruction spread from Clerc’s school throughout the country

Source: Lucas and Hogue, 2004, used with permission of C. Lucas and R. Hogue.


Villanueva, Twain, and Wood

An American in Paris
LLW: On a recent trip to Paris, I saw two deaf men signing on the street. Remembering the history of ASL, and having no knowledge of either spoken French or LSF (Langue des Signes Française, French Sign Language), I approached them to see if we could communicate in any way. Although our languages are indeed different, we were able to have a conversation and found that we were generally able to understand the concepts we were each trying to express.

The history of ASL demonstrates how signed languages develop naturally where groups of deaf people live out their daily lives together and interact, and how signed languages are passed down through these communities. Because this natural development occurs separately in different areas, there is not one universal signed language, but many distinct signed languages recognized throughout the world. The example that often is most striking to Americans is the fact that ASL is not at all similar to British Sign Language (BSL); although English is the written and spoken language of both Britain and the U.S., the signed languages used in the two countries are not related. This point also makes it clear that ASL is not simply a signed version of English; if that were the case, one would expect the British Sign Language and American Sign Language to be very similar, which in fact they are not. Even the signs used to represent the written English alphabet through fingerspelling are completely different: BSL uses a two-handed fingerspelling system, while fingerspelling in ASL is one-handed (see Figure 12.2).

Linguistics of ASL
Now that we have established the historical context of ASL, it’s time to turn our attention to the structure of the language itself. Stokoe’s two ground-breaking publications, Sign Language Structure in 1960 and A Dictionary of American Sign Language in 1965, used linguistic principles to scientifically prove that American Sign Language meets the full criteria of natural languages and should be classified as a fully developed language. Since then many scholars have done significant work to continue to explore the linguistic aspects of ASL, including its phonology, morphology, and syntax, as well as the sociolinguistic factors that impact and influence ASL. The sections covered below provide a glimpse into some of the linguistic aspects of ASL. There are whole books on ASL phonology (Coulter, 1993), ASL syntax (Neidle et al., 2000; Liddell, 2003), and indeed an entire series on the sociolinguistics of ASL (Lucas, 1995). Instead of covering any one area in detail, we hope this survey of the core aspects of ASL linguistics will inspire you to look further into the areas that interest you most; a list of references, with additional suggested readings and materials, is provided at the end of the chapter. Because sign language is a visual and spatial language instead of an auditory one, it presents “a challenge to theories of linguistic universals” (Wilbur, 1979, p. 6). Ongoing linguistic discoveries about ASL and other signed languages continue to foster critical inquiry and expand our understanding of human language.

The factor of signed languages that makes them most different from spoken languages is of course modality. Signed languages are perceived in a visual modality while spoken

ASL: A Visual Language


FIGURE 12.2 alphabets

The British Sign Language (BSL) and American Sign Language (ASL) manual


Villanueva, Twain, and Wood

languages are perceived in an auditory modality. This may seem straightforward on the surface, but it has a whole array of implications for the ways that languages are structured.

Looking Deeper
DT: As mentioned above, my first vivid exposure to ASL was watching two groups of young deaf men argue across a busy New York City street. As I watched that street scene years ago, I was awestruck, but I was not immediately aware of the complexity and completeness of ASL as a language. Rather, like many newcomers, I’m sure I thought it was more like a form of pantomime, similar to the game of charades. But once I looked below the surface and began to really understand, it became clear that ASL is a complex and sophisticated language complete in itself; so much more than what it seemed on the surface to me that day.

Many people who are not familiar with signed languages believe that they are simply systems of gesture or pantomime. This mistaken notion comes from the assumption that all signs are iconic. A sign or word is iconic if its form is related to its meaning. For example, in English there are many words that are iconic, such as boom, ring, woof woof, ding-dong, whoosh, and other such words that describe sounds. All of these words have in common that the way we say the word in English is related to how the thing we’re describing sounds. If you’ve studied other languages, though, you probably learned fairly early on that different languages use different words to describe the same sounds. The sound of a dog barking in Italian is described as bau bau, while in Russian it is gav gav, and in Japanese wan wan. All of these forms are iconic—the form of the word represents the sound it describes—and yet each language chooses a different way to express it. When it comes to signed languages, iconicity is more apparent, because the language is expressed in a visual, rather than auditory modality. For spoken language, words can be iconic if they sound like what they represent; “the aural/oral modality is suited for iconic representations of sound images (including sounds of animate origin, such as human or animal vocalizations, and sounds of inanimate origin, such as explosions, rustles, or bells)” (Taub, 2001, p. 64). For signed languages, which are perceived in the

Exercise 2

1 Consider the following words: HOUSE, CAT, BOOK, EXPERT, WHY, FINE, WRONG, LIGHT. Try to predict if the ASL signs for these words are iconic or not. Now continue reading. Were your predictions correct? Are the signs iconic in the ways you expected? What aspects of the concepts they express allow for or discourage iconic representations in ASL? 2 Look up the following signs in several different sign languages (do an Internet search or try starting at English_TOC). Compare how they are represented iconically: (a) (b) (c) (d) TREE HOUSE CAT WOMAN

ASL: A Visual Language


visual modality, signs that look like what they represent are iconic; anything that is visible can potentially be described with iconic signs. Think about it; in a visual modality you can show the shape of a house, a cat’s whiskers, or a person opening a book. In fact, in ASL the signs for HOUSE, CAT, and BOOK (shown in Figure 12.3) are all iconic.

FIGURE 12.3a HOUSE—iconically represents a roof and walls

FIGURE 12.3b CAT—iconically represents a cat’s whiskers

FIGURE 12.3c BOOK—iconically represents the action of opening and closing a book


Villanueva, Twain, and Wood


Arbitrary signs—no iconic representation

Because ASL is produced visually, it’s easy to represent objects iconically, but this does not mean that anywhere near all the signs in ASL are iconic. Many of the signs in ASL, like EXPERT, WHY, FINE, WRONG, and LIGHT (shown in Figure 12.4) are arbitrary, meaning there is no overt relationship between the form of the word and its meaning. So signed languages are not simply pantomime or gesture; they contain both iconic and arbitrary forms, just like all other languages, and the fact that there are relatively more iconic forms in signed languages than in spoken languages is simply by virtue of the visual modality being well suited for representing visually perceived objects and actions. “The smaller amount of iconicity in spoken languages, which has been attributed to the inferiority of iconic representations, could just as well have been attributed to the inferiority of the spoken modality in establishing iconic representations” (Taub, 2001, pp. 66–67). As noted above, signed languages develop naturally wherever deaf people are in ongoing contact with one another. Of course, the initial attempts at communication between people who do not yet share a language are necessarily highly iconic and gesture based. “One way to explain the paradox of a language that has its roots in iconicity yet is abstractly structured is to observe its changes over time. ASL signs exhibit diachronic development in the direction of increasing abstract formation constraints” (Klima and Belugi, 1979, p. 67). Over time, as certain gestures and iconic representations are used over and over again, they become conventionalized within that language community. Gradually, the widely accepted and frequent gestures transition from spontaneous pantomimes into standardized forms with specific meaning which function as lexical items: they become the words of the new language (Armstrong, 1999; Janzen and Shaffer, 2002).

ASL: A Visual Language


Linguists who study the diachronic changes in ASL signs have shown that the tendency is for signs to shift from being highly iconic signs to being much more arbitrary. Once a sign has been established, it is then regularized over time and usage to fit within the linguistic system. Frishberg (1975) points out several patterns in these changes, such as centering the content information in the hands and thus reducing head or body movements; limiting the size of the signing space, making it smaller and more standardized; and regularizing the handshapes and movements in two-handed signs so that both hands produce similar action, thus decreasing the amount of visual distraction. These same types of changes, from iconic to arbitrary, can be seen in newer signed languages as they develop, most notably presented in the research on the development of Nicaraguan Sign Language over the past few decades (Kegl et al., 1999; Morford and Kegl, 2000).

In the field of linguistics, phonology is often defined as the study of the sound system of language. In fact, the word is related to the Greek root phone, meaning ‘voice.’ In spoken language, phonology is the study of the speech sounds—the smallest building blocks of words. Each phoneme, or individual speech sound, does not have meaning on its own, but when combined with other phonemes can create a meaningful unit (a morpheme, but we’ll get to that later). Signed language linguists challenge the view of languages as only spoken, and have thus taken a wider view of the term phonology and expanded its scope so that it applies to all human languages. Taking this broad approach, any word can be broken up into phonemes, small component parts that individually do not have meaning. In English, phonemes are the individual sounds that combine to form the words; in ASL, the phonemes are the individual parameters (Stokoe et al., 1965) that combine to form a sign: namely handshape, location, movement, orientation, and certain specific facial movements known as non-manual signals. The study of phonology deals with which phonemes a given language uses, how phonemes combine to form words, and how phonemes behave in a particular language. Every sign in ASL can be analyzed in terms of these parameters: • • • • Handshape: how your fingers and thumb are arranged during the sign. Location: where you make the sign. Movement: where and how your hands move through space. Orientation: which way your palm is facing during the sign.

For example, the sign MOTHER in ASL is produced using a ‘5’ handshape with the palm facing left and moving toward the chin until the thumb touches the chin, and then moving slightly away and returning so that the thumb touches the chin again, as shown in Figure 12.5. Note that one-handed signs are produced using the dominant hand. If you’re right handed, you would sign MOTHER with your right hand and your palm facing left. If you’re left handed, you would sign it with your left hand and your palm facing right. The hands are not the only articulators used in producing ASL signs; signers also use their faces to add meaning. Examples of uses of eyebrow and head positions to mark sentence types and grammatical structures are provided in the syntax section later on. Some signs also include non-manual signals as a parameter of the signs themselves; if these signs are produced without the appropriate facial movements, they


Villanueva, Twain, and Wood



are incorrect. For example, the sign BITE is produced by closing a ‘C’ handshape on the dominant hand around a ‘B’ handshape on the non-dominant hand (see Figure 12.6). At the beginning of the sign, the signer’s mouth is open; as the dominant hand closes onto the non-dominant hand, the signer also closes her mouth, keeping the lips apart to show the biting action with her teeth. The non-manual signal must be produced so that it co-occurs with the action of the hand. A few other examples of signs that require a non-manual signal are RELIEVED (blow out air making a ‘phew’ sound), NOT-YET (slightly open mouth with tongue protruding a tiny bit), and SUCCESS (purse lips together then separate them and expel the breath making a ‘pah’ sound), shown in Figure 12.7. Just as in spoken languages, changing just one of the parameters can potentially make an entirely new word; these words that differ by only one parameter are called minimal pairs. Some examples of minimal pairs in English are bat and hat, pin and pan, and cap and can, where the meaning difference is created by a one-phoneme difference in the same word position. In ASL, the difference between the signs in a minimal pair can be the change in only the location, for example as in the signs ONION and APPLE, or perhaps a change in only the handshape, as in APPLE and CANDY. Look at the pictures of these signs in Figure 12.8 and you will see how similar certain signs can be to one another, with only the difference of a single parameter.



ASL: A Visual Language



Signs that require a non-manual signal


Villanueva, Twain, and Wood


Examples of minimal pairs

ASL: A Visual Language


So Close and Yet . . .
LLW: During that first overly confident summer of working at a camp for deaf students, my deaf co-counselors ensured that I learned every “not taught in class” sign and of course all the “dirty words.” Amongst those words was “orgasm.” I could not wait to take all of the new signs back to my class. One night the campers decided we would sleep under the stars. As usual, I fell asleep quickly, exhausted from trying to function in this new deaf world. All of a sudden I felt myself being shaken and heard the campers shrieking. Sleepily I opened my eyes to see the girls frantically signing: “Orgasm!” “Orgasm!” “Orgasm!” They pointed to their fellow camper huddled in her sleeping bag, awkwardly holding her arm. I was rapidly trying to sign “privacy” and “alone” and so on when I realized ten sets of eyes were blankly staring at me. One of the girls, finally realizing that I didn’t understand this situation, grabbed my finger and pointed to the camper’s arm and fingerspelled B-U-G and mimed chomping her teeth up and down. The camper was not having an orgasm as I had thought, but rather had been bitten by a large bug. As my deaf co-counselors explained the next morning (after they finished laughing), the signs ORGASM and BUG indeed have the same hand shape but are located spatially in different ways and require different facial expressions. Space and facial expression are an integral part of ASL, and that is one minimal pair I will never forget!

Exercise 3

Using an ASL dictionary,2 compare the parameters of the following pairs of signs. Which parameters do the two signs in each pair have in common? Which parameters are different? If only one parameter is different, that’s a minimal pair! 1 2 3 4 5 6 7 8 RED—CUTE SIT—CHAIR SUMMER—DRY BICYCLE—SHOES RESPONSIBILITY—BOSS ENJOY—HAPPY MOTHER—FATHER CHILDREN—THING Source: Adapted from Valli et al., 2005: 21.

Morphology is the study of the smallest meaningful units in language and how they combine to make more complex words. People who are interested in morphology want to understand the ways that a language uses smaller units to build larger units, i.e. they study word formation. In this section we will explore two ways in which ASL uses patterns of signs to create new words: deriving nouns from verbs and compounding.


Villanueva, Twain, and Wood

Nouns from Verbs
One way in which ASL creates new words is by deriving some nouns from verbs (Supalla and Newport, 1978; Baker-Shenk and Cokely, 1991). ASL does this by using the same handshape, location, and orientation of the sign for the verb, but repeating the movement and making it shorter and faster. As always, illustrations and examples help make the relationship between these verbs and nouns clear. Let’s start with the verb SIT. The sign SIT is made in front of the signer with both hands in a ‘U’ handshape (the first and middle fingers extended and together) with the palms facing down. The dominant hand starts above the non-dominant hand and moves downward in one smooth motion so that the bottoms of the dominant fingers are in contact with the tops of the non-dominant fingers (see Figure 12.9a). To produce the related noun sign CHAIR (Figure 12.9b), the signer uses the same location, handshape, and orientation as SIT, but produces a shortened movement repeated in quick succession. This same change—repeated shortened quicker movement—can apply to several different verbs to produce semantically related nouns (Supalla and Newport, 1978). Other examples of related noun–verb pairs, seen in Figure 12.10a–f, include: FLY and AIRPLANE, PRINT and NEWSPAPER, and SELL and STORE. For all of these signs, the verb is produced with one strong movement and the related noun is produced with a shorter, yet repeated movement. The sign FLY uses a handshape with the thumb, first finger, and pinky extended. The dominant hand begins in a position near the signer’s


An example of a noun–verb pair

ASL: A Visual Language


shoulder and moves forward and slightly upward. The handshape iconically represents an airplane’s nose and wings, while the movement iconically depicts the plane moving through the air. To produce the sign AIRPLANE, the signer uses the same handshape, location, and orientation but adds a repeated movement, as shown in the photographs in Figure 12.10. The same relationship can be seen in the examples of PRINT/NEWSPAPER and SELL/STORE shown in Figure 12.10 c/d and e/f, respectively.

FIGURE 12.10

Some examples of minimal pairs (continued overleaf )


Villanueva, Twain, and Wood

FIGURE 12.10


Just as a compound is formed in spoken language when two smaller words are used together to create a new word, for example: sweet + heart = sweetheart, ASL and other signed languages also have compounds. For example, similar to English, the signs HOME (Figure 12.11a) and WORK (Figure 12.11b) are both signs that can stand on their own in their own context. However, if you compound the signs and produce them together, the new word formed is ‘homework’ (sign shown in Figure 12.11c). You may notice when looking at the pictures for these signs, that the entire sign HOME and the entire sign WORK are not produced in the compound sign

ASL: A Visual Language


FIGURE 12.11

The signs HOME and WORK and their compound form

HOMEð WORK. There are, in fact, three rules that apply to the initial signs to change them into the final form of the compound (Liddell and Johnson, 1986): 1 2 3 The first contact rule—if an original sign in a compound touches the body, only that contact is kept; any other movements from the original sign do not appear in the compound. The single sequence rule—any repetition of movement in the original sign is deleted when the sign is used in a compound. The weak hand anticipation rule—the non-dominant (or ‘weak’) hand is already in position during the first part of a compound, even though the original first sign may be a one-handed sign.


Villanueva, Twain, and Wood

Let’s clarify these rules a little by looking at our example again: HOMEð WORK. The sign HOME on its own consists of an ‘O’ handshape touching the face close to the chin and then moving upwards and touching on the upper cheek. So the sign HOME contacts the body twice, once at the chin and once on the upper cheek. However, the first contact rule states that for any sign that touches the body only the first contact should be kept in the compound. Therefore, when the sign HOMEð WORK is produced, the hand does not touch the upper cheek; the ‘O’ handshape touching the chin is the only part of the sign HOME that appears in the compound. The sign WORK is made in the space in front of the signer using both hands in fist handshapes with the palms facing down. The dominant hand starts above the nondominant hand and moves downward to hit the non-dominant hand twice. (This is very similar to the movement in the sign CHAIR, which was shown in Figure 12.9b.) Because the movement in WORK occurs twice, when it is put into a compound, the single sequence rule applies, and the movement in the compound occurs only once. The sign HOME is a one-handed sign, but because the sign WORK requires both hands, when the compound HOMEð WORK is produced, the weak hand anticipation rule applies, and the non-dominant hand moves into position at the very beginning of the sign HOMEð WORK. So the sign HOMEð WORK is produced by touching the chin (first contact rule) while moving the non-dominant fist out in front of the signer in panel 2 (weak hand anticipation rule) and then moving the dominant hand down to form a fist handshape that hits the non-dominant hand only once (single sequence rule). Not all ASL compounds are the same as those found in English. In fact, most of them are very different. Table 12.1 shows a list of some ASL compounds and their meanings. Remember that for all of these compounds, the three compounding rules apply, resulting in the appropriate changes in form between the two original signs and the newly formed compound sign. If you search ASL dictionaries for the signs listed in Table 12.1, you’ll be able to see these principles at work.
TABLE 12.1 Some ASL compounds ASL compound GOODð NIGHT GOODð ENOUGH TRUEð WORK SLEEPð SUNRISE SAYð NAME THINKð MARRY THINKð TOUCH GIRLð FRIEND BOYð FRIEND English translation good night hardly adequate truly, seriously oversleep mention to believe be obsessed girlfriend boyfriend

Source: Adapted from Klima and Bellugi, 1979; Valli et al., 2005.

Exercise 4

Compounds Match the ASL compounds in the column on the left with their meanings in the column on the right. Think about the meanings of the two original signs and see if you can guess what the compound form means.

ASL: A Visual Language ASL compound 1 MOTHERð FATHER 2 THINKð SAME 3 GIRLð SAME 4 BOYð SAME 5 BEDð SOFT 6 GIRLð BABY 7 BOYð BABY 8 THINKð EXPAND English translation (a) brother (b) daughter (c) son (d) pillow or mattress (e) sister (f) parents (g) to envision (h) agree


Syntax involves the study of how multiple words are put together to form phrases, clauses, and sentences. The major aspect of syntax that distinguishes signed languages from spoken languages is the way that different types of sentences are marked visually rather than through word order or vocal intonation. In spoken languages, the vocal intonation is often used to distinguish which kind of sentence we are expressing. For example, consider the many different pitch patterns that could be used with the sentence You like linguistics. Imagine a situation in which someone would say each of the following examples. Think about how their meanings are different, based simply on the difference in intonation.
(a) (b) (c) (d) (e) ‘So, you like linguistics. What other classes do you like?’ ‘You like linguistics? I’m taking it now and I really like it.’ ‘You like linguistics? I would have never guessed that about you.’ ‘You like linguistics? I can’t stand it.’ ‘You like linguistics? Of all the things you could like, you really like linguistics?!’

The sentence could be said with a fairly stable intonation, as in (a), in which case we would understand it as a statement of fact. In a casual situation a speaker might add a rising intonation at the end of the sentence, as in (b), marking it as a question rather than a statement (a strict grammarian would tell you that the English question should start with the word “do” but often in casual speech the “do” is not used). In addition to using intonation to distinguish between casual questions and statements, English speakers also use intonation patterns to emphasize certain words in the sentence, yielding the variety of different meanings shown above. English speakers also use word order to mark different types of sentences. For example, declarative statements are generally produced in subject–verb–object order, imperative commands are stated without an overt subject, yes/no questions start with a form of do and then the subject and an uninflected verb, etc. In ASL, signers use their faces to mark these kinds of syntactic and intonational differences. Since the hands are the most salient articulators, linguistically relevant movements that are not produced with the hands are called non-manual signals (Baker-Shenk and Cokely, 1991). In addition to the non-manual signals discussed in the phonology section, non-manual signals are also used to mark syntactic structures. Non-manual signals on the face include position of the eyebrows, where the eye gaze is directed, the mouth configuration, and whether the cheeks are puffed or not. In addition, the signer can nod his or her head at different points in a sentence and with different levels of intensity to mark some types of sentence and discourse structures. A shake of the signer’s head indicates negation; it can either negate the signs produced simultaneously


Villanueva, Twain, and Wood

with the headshake or negate the entire utterance. Signers also use their shoulders and the position of their body, for example leaning forward or backward, to mark syntactic structures. Let’s look at some examples.

Declarative Statements
For declarative sentences that provide information, the face and body are in a neutral position. In many instances, signers nod their heads to mark the end of the sentence and/or to add emphasis to the truth of the statement. This is the unmarked, or most basic, form of sentence in ASL; the other forms described below all include some kind of non-manual marking to distinguish them.

In many ASL sentences, the signer begins by mentioning some entity that will be important in what follows. This is called topicalization. The topic is separated from the main clause and marked with different facial non-manuals, including raising the eyebrows and pulling the head back, as shown in Figure 12.12. After the topic has been stated, the signer relaxes her face and body back to a neutral position to make a statement related to the topic. Often there is also a short pause between the topic and the statement portion of the sentence.

Yes/No Questions
To ask a yes/no question in ASL, the signer raises her eyebrows, widens her eyes, and tilts her head and body slightly forward. In addition the signer’s shoulders may be raised and the last sign of a question may be held longer than other signs. For example, in Figure 12.13a the signer is producing the sign READY with her hands; at the same time she non-manually marks this as a yes/no question by raising her eyebrows, widening her eyes, and gazing directly at the addressee to elicit a response. In Figure 12.13b the signer uses the same non-manual marking with the sign GO-AHEAD to ask whether she should proceed or wait. Notice, however, that the handshapes are different in Figures 12.13a and 12.13b. This question could be understood to mean Shall I proceed?, May I go ahead?, Is it okay if I begin?, etc. (The label q after the brackets indicates a question.)

FIGURE 12.12

‘That street? [It’s] closed’

ASL: A Visual Language


FIGURE 12.13

‘Are you ready?’ and ‘May I go ahead?’

Wh-questions, sometimes called content questions, in ASL include a wh-word: WHAT, WHEN, WHERE, WHO, WHY, or HOW. In addition to using these signs to mark questions, non-manuals co-occur with the question. An example is shown in Figure 12.14. The signer squints her eyes and scrunches her eyebrows together. Her head is tilted to the side. (The label whq after the brackets means wh- question)

FIGURE 12.14

‘How are newspapers made?’

Rhetorical Questions
Rhetorical questions are questions that will be answered by the person who asked the question. English speakers use rhetorical questions for certain discourse purposes, especially in lecture formats; in ASL they are used more often in everyday conversation. The rhetorical question itself typically includes a wh-question word and is marked by the eyebrows being raised with the head either tilted or shaking slightly. In Figure 12.15, the signer starts with a topic and then marks the three signs FOOD MAKE WHO as a rhetorical question; the rhetorical question is labeled with ‘rhq’ after the brackets encasing the clause. She immediately signs the answer to her question by using a pronoun (pointing to the side) and fingerspelling the name SALLY.


Villanueva, Twain, and Wood

FIGURE 12.15

‘Tomorrow night’s dinner will be made [by] who? Sally!’

Commands (also called imperatives) are signed by making direct eye contact with the addressee. The intent is to tell the addressee to do something, so the facial expression is typically serious and may include a frown. In Figure 12.16, the signer topicalizes the phrase THAT REPORT and then commands that the addressee must complete it by next Friday. She maintains direct eye contact with the addressee throughout the command and includes a stern facial expression at the end. In the glosses below the pictures, the imperative clause is labeled with an asterisk at the beginning and the end.

FIGURE 12.16

‘That report must [be] completed by next Friday’

Conditional marking is used when signers want to express ‘if-then’ situations. In ASL the first part of the sentence sets up the ‘condition’ and the second part states the result. The non-manual marking on a conditional clause includes raised eyebrows and the head tilted back and often to one side. Between the conditional and the result statement, there can be a short pause and an eye gaze shift. The non-manual marking occurs during the signs of the conditional clause only. In the example in Figure 12.17, the conditional clause includes the verb PLAY and then the fingerspelled noun ‘chess.’ (Facial expressions in the rest of Figure 12.17 are emotive and emphatic, not grammatical.) Because the conditional includes a verb, the meaning of the conditional is akin to saying “if and when X happens, . . .” The conditional clause is labeled in the glosses with ‘cond’ following the bracket encasing the clause.

ASL: A Visual Language


FIGURE 12.17

‘If/when I play chess, I never lose’

Recall from the earlier discussion of iconicity that signers’ hands can be used to represent objects. The examples listed there, HOUSE, CAT, and BOOK, are all “shape-for-shape iconic representations” (Taub, 2001, p. 69), also called “substitutive depiction” (Mandel, 1977: 65). Note that the three example signs iconically represent slightly different aspects of the objects they describe: the silhouette of a house, just the whiskers of a cat, and the opening and closing of a book. In addition to representing objects in static ways, signers can also depict the size and shape of objects, the relative locations of things, and how things move. These types of signs are particularly difficult to classify and analyze using traditional linguistic terminology developed for spoken languages, as is evidenced by the many labels that have been used to describe them: “classifiers” (Frishberg, 1975; Supalla, 1986), “polymorphemic verbs” (Engberg-Pedersen, 1993), “polycomponential verbs” (Schembri, 2003), and “depicting verbs” (Liddell, 2003). These signs are clearly iconic but certain aspects are also quite conventionalized within particular signed languages. A few examples are given in Figure 12.18. In Figure 12.18a the signer is showing the size and shape of a metal pipe. She uses the ‘O’ handshape and puffs her cheeks to emphasize the diameter of the pipe; if it were a smaller pipe, the ASL convention would be to use an ‘F’ handshape, and if it were flimsy, she would use a different non-manual with the sign. In Figure 12.18b the signer depicts the locations of two dorms relative to each other; the handshape and palm orientation the signer uses are the conventionalized way to depict buildings in ASL. The sign in Figure 12.18c is a depiction of a person walking a long distance. The first finger extended upwards represents an upright person; the movement through the signing space represents the fact that the person walked from one location to another; both the up and down wiggling of the finger and the signer’s facial expression add the meaning that the distance walked was relatively far. In Figure 12.18d the handshape and movement depict objects moving rapidly forward as a group; it is understood from the context of the story that in this particular case the signer is using this sign to represent a group of boys running across a field. Signers can also show people’s actions, retell what someone said previously, or represent a character in a story.
[In] the most mimetic, least stylized instances . . . signers are free to do any action at all, with the understanding that their movements represent the actions of some referent person, i.e. someone else, or themselves at a different time. This iconic device is roughly analogous to quoted speech in spoken languages—but for signed languages, what is


Villanueva, Twain, and Wood

FIGURE 12.18

Different types of depiction in ASL

ASL: A Visual Language reported is not sound but body movement, both linguistic and nonlinguistic, i.e. both signing and other actions. (Taub, 2001: 75)


Examples of this are shown in Figure 12.19. In Figure 12.19a the signer is depicting someone pulling up a blanket and pretending that he has been asleep. In Figure 12.19b the signer is depicting the dorm supervisor angrily asking who snuck out. The signer shows that he is representing two different characters through the change from a) to b) in his facial expression, eye gaze, and body position. The depiction of people’s words and actions in signed languages is labeled constructed dialogue and constructed action (Metzger, 1995), similar to current analyses of what was formerly labeled “reported speech” in spoken languages

FIGURE 12.19

Depictions of two different characters in a story

This chapter has provided a glimpse of American Sign Language as it is understood through linguistic research. As languages produced in the visual modality, signed languages are able to capture aspects of the world around us and communicate them in ways not accessible through spoken languages in the auditory modality. Simultaneously, they employ many of the same structural features and systematicity that is found in all languages throughout the world. As such, the unique features and qualities of signed languages continue to stretch our definition of language while reinforcing our understanding of the linguistic systems that people use to communicate. The more we broaden our perspectives to include signed languages, the more we allow ourselves to know them and to become truly able to appreciate them.

1 “Deaf ” with a capital D is often used to refer to individuals who use American Sign Language and consider themselves to be culturally deaf. The word “deaf” written with a lowercase d is used as an all-encompassing term referring to people who have a hearing loss. This chapter will use the lowercase d throughout, except in the names of schools. 2 There are many options for ASL dictionaries. Check out your library or bookstore for ASL dictionaries to borrow or purchase. An Internet search will yield several online ASL dictionaries such as, http://, and aslpro.cgi.


Villanueva, Twain, and Wood

Armstrong, D. F. (1999) Original signs: Gesture, sign, and the sources of language. Washington, Gallaudet University Press. ASL Pro. (2009) ASL dictionary [Internet]. Available from: aslpro/aslpro.cgi [Accessed July 4, 2009]. Baker-Shenk, C. and Cokely, D. (1991) American sign language: A teacher’s resource text on grammar and culture. 4th edn. Silver Spring, T.J. Publishers. British Sign Language Resources. (2009) Learn British sign language [Internet]. Available from: [Accessed July 4, 2009]. Comm Tech Lab. (2009) American sign language browser [Internet]. Available from: http:// [Accessed July 4, 2009]. Coulter, G. R. (1993) Current issues in ASL phonology (Phonetics and phonology 3). San Diego, Academic Press. Engberg-Pedersen, E. (1993) Space in Danish sign language: The semantics and morphosyntax of the use of space in a visual language. Hamburg, Signum Press. Ethnologue. (2009) Ethnologue: Languages of the world [Internet]. Available from: [Accessed June 28, 2009]. Finger Spelling. (2009) Finger spelling [Internet]. Available from: http://www.finger [Accessed July 4, 2009]. Frishberg, N. (1975) Arbitrariness and iconicity: Historical change in American Sign Language. Language, 51(3), pp. 696–719. Groce, N. E. (1985) Everyone here spoke sign language: Hereditary deafness on Martha’s Vineyard. Cambridge, Harvard University Press. Handspeak. (2009) ASL dictionary online [Internet]. Available from: http://www.handspeak. com/sign/index.php [Accessed July 4, 2009]. Janzen, T. and Shaffer, B. (2002) Gesture as the substrate in the process of ASL grammaticization. In Meier, R., Quinto-Pozos, D., and Cormier, K. eds. Modality and structure in signed and spoken languages. Cambridge, Cambridge University Press, pp. 199–223. Kegl, J., Senghas, A., and Coppola, M. (1999) Creation through contact: Sign language emergence and sign language change in Nicaragua. In DeGraff, M. ed. Comparative grammatical change: The intersection of language acquisistion, Creole genesis, and diachronic syntax. Cambridge, MIT Press, pp. 179–237. Klima, E. S. and Bellugi, U. (1979) The signs of language. Cambridge, Harvard University Press. Lane, H., Hoffmeister, R., and Bahan, B. (1996) A journey into the deaf world. San Diego, Dawn Sign Press. Liddell, S. K. (2003) Grammar, gesture, and meaning in American Sign Language. Cambridge, Cambridge University Press. Liddell, S. K. and Johnson, R. E. (1986) American Sign Language compound formation processes, lexicalization, and phonological remnants. Natural Language and Linguistic Theory, 4(4), pp. 445–513. Lucas, C. (1995) Sociolinguistics in deaf communities. Washington, Gallaudet University Press. Available from: [Accessed June 29, 2009]. Lucas, C. and Hogue, R. (2004) The sociohistorical context for lexical variation in ASL. Poster presented at Theoretical Issues in Sign Language Research 8, September 30–October 2, Barcelona, Spain. Lucas, C., Bayley, R., and Valli, C. (2003) What’s your sign for pizza? An introduction to variation in American Sign Language. Washington, Gallaudet University Press.

ASL: A Visual Language


Mandel, M. (1977) Iconic devices in American Sign Language. In Friedman, L. A. ed. On the other hand. London, Academic Press, pp. 57–107. Metzger, M. (1995) Constructed dialogue and constructed action in American Sign Language. In Lucas, C. ed. Sociolinguistics in deaf communities. Washington, Gallaudet University Press, pp. 255–271. Morford, J. P. and Kegl, J. A. (2000) Gestural precursors to linguistic constructs: How input shapes the form of language. In McNeill, D. ed. Language and gesture. Cambridge, Cambridge University Press, pp. 358–387. Neidle, C., Kegl, J., MacLaughlin, D., Bahan, B., and Lee, R. G. (2000) The syntax of American Sign Language: Functional categories and hierarchical structure. Cambridge, MIT Press. Oliva, G. A. (2004) Alone in the mainstream: A deaf woman remembers public school. Washington, Gallaudet University Press. Schembri, A. (2003) Rethinking ‘classifiers’ in signed languages. In Emmorey, K. ed. Perspectives on classifier constructions in sign languages. Mahwah, Lawrence Erlbaum Associates, pp. 3–34. Stokoe, W. C. (1960) Sign language structure: The first linguistic analysis of American Sign Language. Buffalo, University of Buffalo Press. Stokoe, W. C., Casterline, D. C., and Croneberg, C. G. (1965) A dictionary of American Sign Language on linguistic principles. Washington, Gallaudet University Press. Supalla, T. (1986) The classifier system in American Sign Language. In Craig, C. G. ed. Noun classification and categorization. Amsterdam, John Benjamins, pp. 181–214. Supalla, T. and Newport, M. (1978) How many seats in a chair? The derivation of nouns and verbs in American Sign Language. In Siple, P. ed. Understanding language through sign language research. New York, Academic Press, pp. 91–131. Taub, S. F. (2001) Language from the body: Iconicity and metaphor in American Sign Language. Port Chester, Cambridge University Press. Valli, C., Lucas, C., and Mulrooney, K. J. (2005) Linguistics of American Sign Language: An introduction. 4th edn. Washington, Gallaudet University Press. Wikisign. (2009) Langue: signes du monde/English TOC [Internet]. Available from: http:// [Accessed July 4, 2009]. Wilbur, R. B. (1979) American Sign Language and sign systems. Baltimore, University Park Press. Woodward, J. (1978) Historical bases of American Sign Language. In Siple, P. ed. Understanding language through sign language research. New York, Academic Press, pp. 333–348.

Other Resources
Ballin, A. (1930) The deaf mute howls. Los Angeles, Grafton Publishing Company. Benderly, B. L. (1980) Dancing without music: Deafness in America. Garden City, Anchor Press/Doubleday. Cohen, L. H. (1994) Train go sorry: Inside a deaf world. Boston, Houghton Mifflin. Cyrus, B., Katz, E., Cheyney, C., and Parsons, F. M. (2005) Deaf women’s lives: Three self portraits (Deaf Lives Series, Volume 3). Washington, Gallaudet University Press. Emmorey, K. (2003) Perspectives on classifier constructions in sign languages. Mahwah, Lawrence Erlbaum Associates. Fox, M. (2007) Talking hands: What sign language reveals about the mind. New York, Simon and Schuster.


Villanueva, Twain, and Wood Jacobs, P. G. (2007) Neither–nor: A young Australian’s experience with deafness. Washington, Gallaudet University Press. Lane, H. L. (1984) When the mind hears: A history of the deaf. New York, Random House. Lane, H. L. (1999) The mask of benevolence: Disabling the deaf community. San Diego, Dawn Sign Press. Lane, H. L. and Philip, F. (1984) The deaf experience: Classics in language and education. Cambridge, Harvard University Press. Molloy, P. (2004). Life and deaf. New York, Universe. Neisser, A. (1990). The other side of silence: Sign language and the deaf community in America. Washington, Gallaudet University Press. Padden, C. A. and Humphries, T. L. (1988) Deaf in America: Voices from a culture. Cambridge, Harvard University Press. Padden, C. A. and Humphries, T. L. (2005) Inside deaf culture. Cambridge, Harvard University Press. Sacks, O. (1989) Seeing voices: A journey into the world of the deaf. New York, HarperCollins. Spradley, J. P. (1985) Deaf like me. Washington, Gallaudet University Press. Vasishta, M. (2006) Deaf in Delhi: A memoir. Washington, Gallaudet University Press. Zazove, P. (1993) When the phone rings, my bed shakes: Memoirs of a deaf doctor. Washington, Gallaudet University Press.

Audism unveiled. (2008). Dir. Ben Bahan, H-Dirksen Bauman, and Facundo Montenegro San Diego, Dawn Sign Press [video: DVD]. Through deaf eyes. (2007) Dir. Diane Garey and Lawrence R. Hott. Washington, WETA and Florentine Films/Hott Productions in association with Gallaudet University [video: DVD].

Wyatt A. Mangum
This Chapter Explores:
Bee Communication Waggle Dance Round Dance

Animal Communication The “Language” of Honey Bees

Shaking Signal Tremble Dance Design Features of Language

One bee says to another, “I have found the most marvelous flowers.” The second bee exclaims, “How strange—I understand you! Researchers say we don’t have a language!” Surprised, the first bee blurts out, “I heard we do have a language!” Perplexed, the second bee laments, “How would they know if we did have a language?” “I don’t know,” says the first bee, “Maybe they would compare it to their language.” Still perplexed, the second bee wonders, “So they must know what a language is.” Momentarily both bees are silent, then the first one says, “Let them worry about all that. I want to tell you about these wonderful flowers.” Eavesdropping on two fictitious bees, from a book not yet written by W. A. Mangum

As long as I can remember, bees have always captured my imagination. Yes, I did get stung occasionally, but that did not deter me. At age 10, I acquired my first hive and during high school I kept 125 hives. I delighted in watching bees with three glass hives in my bedroom, which sometimes leaked bees into the house. Thankfully, I had a very understanding mother. Since the study of bee behavior is inherently interdisciplinary, I needed to study certain fields in biology, like genetics and ecology. To analyze the data from bee experiments, I needed to study mathematics and statistics. The way bees communicate with dances is a continual source of amazement. By taking just a few measurements as I watch bees dance in a glass hive (now I have 30 of these hives in a special building), I can figure out where they are foraging, say a quarter-mile west of the hive, perhaps in a field of flowers. With dozens of bees dancing in my hives, I can in effect “listen in” on what would be analogous to their “conversations” about where to find the best nectar-bearing flowers. Those wonderful experiences led me to ponder if their communication is a language and how that question could be answered, or if it is even possible to


Mangum answer it. So now my love for bees has led me into a new field of study, linguistics. Although I am light-years away from being a linguist, I am a honey bee scientist always ready for a new intellectual adventure.

Extensive research shows that honey bees have a sophisticated communication system where they dance to indicate the location of food and other resources. But is this communication a language? It is, after all, commonly called the dance language. We could, as suggested by our fictitious bee, compare bee communication with human language. That premise supposes we know what a language is; otherwise, how could the comparison be made? Before we can attempt this comparison, we need to learn about the fascinating, surprising, and sometimes strange micro-world of bees dancing in the dark. Then we will compare their communication with various design features thought to characterize a language, although there is not complete agreement among linguists on what this list should include. We may even find ourselves pondering, “What is a language?” Let us begin our journey to the hive by conducting a simple thought experiment.

The Experiment
Place a small dish of sugar water on a table outside. Sugar water tastes sweet to a bee, like nectar from a flower. It may take several days for a bee to find this rich food source, a real treasure trove of free bounty. Most bees search for flowers up to a couple of miles from the hive. That makes the dish just a tiny dot in all that land. Finally—and remarkably—out of nowhere a bee appears. She (the worker bees are female) hovers near the dish for a few moments, appearing to inspect the situation. Then she lands. Finding the sugar water, she eagerly drinks her fill. Her abdomen even swells as she loads up. She holds only about a drop, a tiny bit for us, but a big amount for a little bee. When she can drink no more, she takes off, flying slowly, straining to lift her heavy load. She circles the table a couple of times, and then she disappears as a tiny speck vanishing in the open sky. After waiting all that time, just one bee came. Maybe in a few more days another bee will by chance find the dish. That does not happen. It does not take that long . . . After a mere half hour, a bee appears. Perhaps it is just the same bee returning for more food. That would not be too surprising since many animals can remember where they find food and return to the same place. Then something astounding happens that has forced people to reevaluate the capabilities of these little insects. In about a minute another bee appears; right after her a couple more arrive; then four or five more descend on the dish. They crowd around the dish frantically drinking quickly and flying off. Still more bees arrive in short order. After just an hour, dozens of bees, too many to count, mob the dish, gobbling up the sweet drink drop by drop, and haul it back to their hive. After the miniature melee subsides, and the last bee flies off, all the sugar water is gone. The bees have licked the dish squeaky clean—not even a trace of stickiness remains. What happened? After days of waiting just to see one bee at the dish, how did so many of them find the dish so quickly? We now know that the first bee communicated to the others the location of the dish. That would mean that bees with their tiny brains, less than the size of a pinhead, have some method of communicating fairly complicated directional and distance navigational information. Do bees have some kind of language for doing that?

Animal Communication: Honey Bees


A bee having found a flower patch could lead some of her nest mates back there. This leading behavior would not involve much communication. In the early 1930s, a French naturalist, Julien Françon, showed that true honey bees (bees in the genus Apis), like the ones in North America and Europe, do not lead recruit bees to the food source. In his book with the captivating title The Mind of the Bees (1939), Françon described how he captured a paint-marked bee as she left the hive to go back to a sugar water dish. Since she was just inside the hive, she had the opportunity to communicate the location of the dish to other bees. Now captured, however, she could not lead the recruit bees to the dish. The recruit bees came anyway (Françon, 1939). So if honey bees do not lead each other to food sources, how do they communicate the location of the dish? This leads to a series of questions: • What is the bees’ communication system? • Does this system have any features in common with language? • Do bees have language? We will turn now to some classic research in linguistics on the definition of language. We need to look more closely at the crucial criteria that allow a communication system to be considered language by linguists.

Design Features of Language
Most linguists consider human language a unique type of communication system, thus calling language species specific. While communication is the broader concept, language encompasses more specific properties. Beginning in the 1950s, linguist Charles Hockett set out to describe what makes human language different from the communicative behavior of a number of other animals (Hockett, 1977). He developed a set of criteria that generally refer to abstract properties of language. The list of design features that Hockett developed has varied through the years, but we will focus on seven of Hockett’s design features of language: (1) mode of communication, (2) discreteness, (3) interchangeability, (4) cultural transmission, (5) arbitrariness, (6) displacement, and (7) productivity. (Each of these design features is defined later in this chapter.) Since Hockett’s pioneering work, other criteria have been developed for language, but Hockett’s system continues to inform the field of linguistics. With these design features in mind, let us return to the historical progression of understanding bee communication.
Ask four or five people to define language. Also ask them to name any other species (besides humans) they believe have a language facility. Ask your informants to try to distinguish between language and communication. How many of your interviewees said language is specific to our species (a belief that language is species specific)? Now look up the words “language” and “communication” in two or three dictionaries. Compare the answers from these sources.

Exercise 1

Historical Observations
From at least 1655, European observers noticed something rather odd about bees returning to the hive (or to swarms in search of new nest sites) (Crane, 1999). Among the bustle of bees crowded on the combs, foragers sometimes perform curious little dances. As the bees dance, one to about five bees, called recruit bees, follow closely behind the




A waggle dance

Note: The arrows denote the dancer’s path. Note the two return loops. The wavy arrow indicates where the lateral abdominal wagging occurs during the straight run part of the dance. The dancer appears blurry because she is performing that wagging motion. Source: Chittka, L. (2004) Dances as Windows into Insect Perception. PLoS Biol 2(7): e216 doi:10.1371/journal.pbio.0020216

dancer. Some dancing bees run around in a kind of squashed figure-eight pattern about the size of a nickel (see Figure 13.1). They even vibrate their abdomens in a lateral wagging motion where the crossing of the squashed figure eight forms a straight line— the wavy line in Figure 13.1, actually called the straight run or waggle run. The return semicircular loops of the figure-eight pattern (when the bee does not vibrate her abdomen) are where the bee runs back to the starting point to begin vibrating her abdomen again. The bee also alternates the return loops. If early observers looked closely, they would have seen the following dance pattern, shown in Figure 13.1, which is called a waggle dance: • • • • • vibrate the abdomen in the straight run; loop right, coming back to the starting point; vibrate the abdomen again in the same manner; loop left, back to the starting point; and repeat the pattern

During the “straight-run” part of the dance the bee actually makes a long stride (Tautz et al., 1996), but I will refer to it as a “run” since this usage is common in the bee literature.
Exercise 2 Try to find all the components of the waggle dance I listed above in Figure 13.1. Try explaining it to someone who has not yet encountered this information. Which aspects are not as obvious as others, both to you and those to whom you explain the parts of the dance?

Animal Communication: Honey Bees


To make matters even more perplexing, other bees appear to perform another kind of dance. They run around in little circles, changing direction from circling clockwise to counterclockwise or vice versa. Had people noticed, the bees would have been even more baffling because sometimes forager bees dance and sometimes they do not dance. So in addition to why bees dance there is another, subtler question: How do they know when to dance? While looking in their hives, American beekeepers of the early 1900s noticed these strange dances too, sometimes referring to the dancers as “waltzing bees.” Beekeepers thought the bees were just full of joy and happy to tell nest mates about newly found flowers (Root, 1908). While this anthropomorphic (human-centered) view is not correct, nevertheless a vague impression existed of the bees having some form of communication. However, like listening to an unfamiliar foreign language, no one could understand it. Then in the 1920s the mysteries of these curious dances started to be explained in detail.

Deciphering the Waggle and Round Dances
The Waggle Dance
Austrian biologist Karl von Frisch came to realize through a series of experiments that the bees were indeed using a communication system far more sophisticated than anyone could have foreseen. The bees were communicating the location of nectar and pollen sources with directional and distance information (von Frisch, 1967). (Bees also communicate the location of new nest sites, as well as resin and water sources. We will concentrate on just the communication of nectar sources in this chapter.) Von Frisch found that if a food source is a long distance from the hive (more than 100 meters), the bee runs the squashed figure-eight pattern of the waggle dance, named such because the bee waggles (vibrates) her abdomen during the straight run part of the dance. The waggle dance conveys the direction and distance to the food source.

Direction Information The forager bee perceives the angle between her line of flight and the sun’s azimuth. The sun’s azimuth is the location on the horizon directly beneath the current position of the sun. (There is also evidence that a bee can perceive the flight direction from the pattern of polarized light in the sky, but this is beyond the scope of this chapter. For references see Dyer, 2002.) One way to think of this angle to the food source is as follows: Imagine you are standing by the beehive with your left hand pointing to the horizon right under the sun (the sun’s azimuth) and your right hand pointing to the food source. The angle between your arms is the angle to the food source relative to the sun’s azimuth, which is the direction of your left arm (see Figure 13.2). How does a bee communicate this horizontal angle inside a dark hive while dancing on a vertical comb, conditions very different from outside the hive? The direction to the food source relative to the sun’s azimuth is represented by the angle of the straight (waggle) run part of the dance relative to the vertical direction (up). For example, if the food source is 90 degrees to the left of the sun’s azimuth, then the direction of the waggle run is 90 degrees to the left of up (see Figure 13.3, III). The YouTube video (–7ijI-g4jHg) also shows a waggle dance in this same direction, and then to another location 60 degrees to the right of the sun’s azimuth (YouTube, 2009). Figure 13.3 shows directions to other food sources. If the food is in the same direction




The angle between the food source and the sun’s azimuth

Note: From the hive (the empty white box), the person’s left arm points to the sun’s azimuth, and her right arm points to the food source. The angle from her left arm (the sun reference) to her right arm is the direction to the food.

as the azimuth (0 degrees from it), then the waggle run is straight up the comb (see Figure 13.3, I). But, if the food is in the opposite direction from the azimuth (180 degrees from it), then the waggle run is straight down (see Figure 13.3, V). Interestingly, it is not yet completely understood just how the recruit bees perceive this angle as they follow the dancer (Michelsen, 1999). Hopefully, a future scientist will give us a better understanding of that perception. Nevertheless, knowing the direction to the food is an important piece of information in finding it. But how far away is the food? Or more precisely, how does a bee indicate the distance to the food source?

Distance Information Several aspects of the waggle dance could indicate distance, but as with the dance angle, it is not completely understood how the recruit bees perceive distance information (Michelsen, 1999). Nevertheless, several lines of research are investigating this question. Researchers who feed bees sugar water at known distances from the hive find that the tempo of the dance is a fairly obvious indication of distance to the food source. In fact, using just a stopwatch, von Frisch counted the number of dance circuits during 15 seconds. (A circuit was defined as a waggle run plus the following return loop a bee makes back to her previous position just before starting the next waggle run.) Von Frisch found that more circuits (in 15 seconds) indicated close distances to the sugar water locations. Conversely, dances with fewer circuits indicated longer distances. For example,

Animal Communication: Honey Bees


FIGURE 13.3 Examples of various directions to food sources and the corresponding waggle dances
Note: A waggle run straight up the comb indicates food is in the same direction as the sun (I). A waggle run at an angle ∝ to the left of the vertical indicates the food is at an angle ∝ to the left of the sun, (II). A waggle run 90 degrees (β) to the left of the vertical indicates the food is 90 degrees to the left of the sun (III). Notice as the angle between the sun and food increases, the angle between the waggle run and the vertical increases by the same amount (IV). A waggle run straight down the comb indicates the food is opposite the sun (V). A waggle run now at an angle δ to the right of the vertical indicates the food is at an angle δ to the right of the sun (VI).

on one occasion with the sugar water at 100 meters from the hive, von Frisch counted ten circuits in 15 seconds. With the sugar water at 500 meters, there were only six circuits (von Frisch, 1967). Then to see how fast the number of circuits decreased as the distance increased, von Frisch plotted his data with the (average) number of circuits on the y-axis and the distance on the x-axis. Figure 13.4 shows the plot. Notice that starting at 100 meters, as the distance increases, the number of circuits decreases rapidly and then levels off for much larger distances. Interestingly the decrease is not in a straight line, but rather in a curved fashion.
Exercise 3 You now know something about bee dances and their correspondence to information about food sources. Do you think these dances qualify as language? What else do you still need to know to determine your answer?

More detailed observations of the waggle dance reveal that other parts of it could also be used to perceive the distance. For slower dances, the forager bee spends a longer time



FIGURE 13.4 The average number of dance circuits per 15 seconds as a function of the distance from the hive to the sugar water feeding station in the waggle dance
Source: Redrawn from von Frisch, 1967.

wagging her abdomen during the straight run part of the dance. While she wags her abdomen, she also emits a high-pitched buzz so that slower dances have longer periods of buzzing. Either the duration of the abdominal waggling or this special buzzing could be used as a distance measurement. Currently we just do not have a complete understanding of what the recruit bees perceive. In addition to the direction and distance information, which indicates the approximate location of the nectar, the recruit bee can smell the flower’s odor adhering to the dancer’s body. This odor helps the recruit bee find the exact location of the particular flowers.

The Round Dance
Look again at Figure 13.4; notice that von Frisch had no data points for distances less than 100 meters. Why this omission? If the dish is less than 10 meters from the hive (this distance depends on the subspecies of the bee), then the bee performs another dance that observers must have been seeing before von Frisch explained its meaning. The round dance, as it is called, is the dance in which the bees run around in little circles, changing direction from circling clockwise to counterclockwise or vice versa (see Figure 13.5, 1). The round dance was originally thought to stimulate recruit bees to search for food near the hive, without providing directional information. The recruit bees would be guided by the flower’s odor detected from the dancer. Recent research has shown, however, that the round dance does contain directional information, but the recruit bees seem to make poor use of it since they search in all directions close to the hive (Dyer, 2002). (From 10–100 meters, bees perform transitional dances as shown in Figure 13.5, 2–5. For more details see von Frisch, 1967.) Now that we know how the dancers are behaving, let us take a closer look at the recruit bees that receive the direction and distance information from a waggle dance.

The Behavior of the Recruit Bees As a forager bee dances, the recruit bees follow closely behind her, usually for a few dance circuits. From the YouTube video mentioned above (watch it again—maybe a few

Animal Communication: Honey Bees


FIGURE 13.5 Dance behaviors observed near the hive (I–V). A round dance (I). Transitional dances for intermediate distances termed sickle or crescent dances (II–V). A waggle dance (VI)

more times), you can see the bees following the dancer. These are the recruit bees. Notice how they can come briefly to the dancer and then go away. The recruit bees can emit a brief, high-pitch piping sound signal that tends to stop the dancer. This piping was once hypothesized to be a begging signal where the dancer would stop and give the recruit a sample of nectar (Pastor and Seeley, 2005). The nectar would include the same flower odor and help the bee find the flowers, particularly when they are far away and the external odor clinging to the forager’s body has diminished (von Frisch, 1967). While recent research has shown that bees following the dance do pipe and stop the dancer, Pastor and Seeley (2005) found that the piping is not a begging signal; its function, however, is not currently known. After following a dance, a recruit bee somehow “translates” the coded information, along with a memory of the flower odor, into directional and distance information to the nectar. When a recruit bee flies from the hive, she, like the first bee that came to our dish, disappears into the sky. Her flight path to the flowers or sugar water dish was for a long time invisible to us. Scientists wondered whether she flies straight to the food source using dance information or instead uses predominantly odors to find the food. If odors guide her to the flowers, she would perform other flight behaviors to keep the odor source of information available to her. For example, she might first fly downwind, then turn around and fly upwind so that the odors are coming to her. She could follow the odors upwind that emanate from the flowers (or other odors that were near the plants) that she smelled on the dancer’s body. Bees are very sensitive to odors, and their ability to find food using just odors should not be underestimated. (For a detailed description, see Wenner and Wells, 1990.) New technology has been decisive in providing a better understanding of what recruit bees are doing. Recruit bees, having just followed a dancer, had their flight paths recorded with a new kind of radar called harmonic radar. Unlike conventional radar that tracks large objects (like planes) high up in the sky, harmonic radar can track small objects (like bees) close to the ground. Using harmonic radar, we can see how the bees fly



to the sugar water dish. While the recruit bees’ flight paths showed spatial variation, most of them indicated a fairly direct flight to the artificial feeder (with unscented sugar water) where the dancer had foraged (Riley et al., 2005). These flight paths indicate that bees are using the information from dances.

The Big Picture of the Bees’ Dance Communication
To understand the intriguing consequences of the bees’ dance communication, let us take an ecological approach. A colony of bees in a temperate climate faces a critical and complicated survival problem. Before cold temperatures stop the bees from foraging, they must store roughly 50 pounds of honey in the hive. During the winter, the bees feed on this honey. If they do not produce enough honey, they starve—hence the critical part of their problem. The nectar they need to make this honey is, of course, in the flowers, but getting enough nectar is not easy. The flowers are scattered in patches over a vast expanse of land comprising their foraging range. The nectar from some plants is sweeter (contains more energy) than the nectar of other plants. For some plants, a bee must visit dozens of flowers before obtaining a full nectar load, while other plants have so much nectar in their flowers, a bee can fill up from just one flower. If all the foragers go to the best nectar-bearing plants, bees of the same colony will become overcrowded on these flowers. They could ignore other nectar sources that are almost as good. To collect the nectar efficiently, several thousand foragers need to be dispersed over these variable nectar sources. To make matters more complicated, these plants are in effect “changing” their locations because they are constantly coming into bloom and dying off in different places. Even the foragers themselves are constantly changing by dying and being replaced by new ones, who are unaware of the current food sources. How can the foragers be allocated among the plants so they collect nectar most efficiently? That is the complicated part of the problem. Yet, as we will see, the bees have an ingeniously simple way of dispersing the foragers, a way that relies on their dance communication.

To Dance or Not to Dance
As mentioned previously, upon returning to the hive, a foraging bee sometimes performs a dance. (Note that the dance is not some robotic reflex that must be performed after each foraging flight. The situation is more complicated than that.) Her decision to dance is important because the choice is about whether to communicate to other bees or to keep silent. If she does not dance, the bee unloads her nectar and returns for another load or quits foraging at that location. Internally she appears to have a way of measuring the (sugar) profitability, or quality, of the nectar source, which probably accounts for such factors as its closeness to the hive, sweetness of the nectar, and collecting efficiency. Bees dance after visiting rich food sources (Seeley, 1995). Thus, the dance communication is about the best food locations found by the foraging bees. While these conditions depend mainly on the food source, the conditions in the colony are also important in a bee’s decision to dance. To understand this colony component, we need a more sophisticated understanding of how a colony of bees collects nectar and produces honey. Although we have seen bees on flowers collecting nectar, this collection is just part of the honey-making process. A colony of bees consists typically of about 50,000 bees with several thousand foragers. When foragers return with nectar loads in their crops (a crop is a temporary storage vessel), they do not themselves deposit the nectar in the cells as one might think. Rather,

Animal Communication: Honey Bees


in the crowded hive, a forager encounters other bees whose job is to process the nectar into honey. The forager transfers her nectar load to one or more of these processor bees. Thus, thousands of bees are collecting nectar and thousands of bees are processing nectar into honey. Most importantly, the rate at which bees collect nectar needs to be matched with the rate at which bees process nectar (Seeley, 1995). From recent research, it appears that if the forager bee finds processor bees quickly, that unloading experience has a positive influence on her decision to dance (Seeley, 1995). Another study suggests the stimulus to dance is the number of processor bees unloading a forager (de Marco, 2006). For example, when numerous processor bees are readily available, perhaps even waiting on more foragers to arrive, this imbalance is an indication that more foragers should be recruited (by dancing) because the colony’s processing capacity is greater than its collection capacity. Besides using the waggle dance to increase the number of foragers, a forager can apparently stimulate non-forager bees to become foragers with a behavior called the shaking signal. When a forager displays the shaking signal, she vibrates her body up and down for a couple of seconds, usually with her front legs grasping the recipient bee. A forager may perform the shaking signal on up to 20 bees per minute throughout the hive. Shaking signals are commonly observed during consistent abundant forage or when an abrupt increase in nectar availability occurs after a dearth. Thus the waggle dance and the shaking signal are two ways a colony can boost the number of foragers (Seeley, 1995). What happens if the opposite problem occurs? What if the colony’s nectar collecting exceeds its processing capacity? Now foragers spend lengthy amounts of time searching for relatively scarce processor bees to unload them. After experiencing a long search time, a forager performs another kind of dance, called a tremble dance. When a bee performs a tremble dance, she walks among other bees trembling her body in a somewhat erratic manner. This dance stimulates bees to become processors so that the colony’s nectar processing capacity will increase to match its collection capacity. In addition to stimulating an increase in the processing capacity, the tremble dance has an inhibitory effect on waggle dancing, thereby reducing a colony’s nectar collection (for more details, see Seeley, 1995). Now that we understand how the shaking signal and the tremble dance function to adjust the nectar collection and processing capacity, let us return to the waggle dance to see how the colony allocates its foragers among the various food sources.

Quality of Food
When bees dance, the dances are not just randomly scattered throughout the hive, but rather most of them occur on the combs near the hive entrance, a place known as the dance floor. Among the dancers on the dance floor are novice recruit bees or experienced forager bees that may not be actively working a particular flower patch. Thus the dance floor is where numerous bees congregate to, in a sense, communicate about food sources. (Also a bee never dances alone. She only dances in the presence of other bees, suggesting this behavior is inherently a social activity.) With most dances occurring near each other, it is tempting to think recruits would go from dance to dance comparing them in an attempt to find the ones indicating the very best food sources. Bees returning from more profitable sources perform more “lively” dances, information that recruits could use for such a choice. Despite this available information, bees do not compare dances (Seeley and Towne, 1992). This lack of choice is important because it concerns another aspect of the waggle dance we have not yet considered—the duration of the dance, that is, how long a bee dances after a foraging trip.



Bees convey the quality of the food source through the duration of the dance. Longer dance durations indicate higher quality food sources (Seeley, 1995). Even the liveliness of the dance can be considered part of the duration and an indicator of the richness of the food source. (See Seeley et al., 2000, for details.) Since recruit bees do not compare dances, appearing instead to attend to them at random, dances that last longer have more opportunity to direct additional recruits to the better nectar sources. Nectar sources of lesser quality elicit dances that do not last as long and acquire fewer recruits. In this way, the foragers distribute themselves among nectar sources of different profitabilities. With their dance communication, the bees are tracking the changing nectar sources, deploying foragers to new blooms, and efficiently distributing foragers among the best nectar sources. Think about what is missing here. No bees function like “bosses” or “generals” (not even the queen bee) to tell the other bees where to forage. Each forager has only a limited knowledge of the “foraging terrain.” Yet together these “mere insects” solve the difficult problem of efficiently dispersing a colony’s foragers among variable nectar sources.
Exercise 4 Before we discuss Hockett’s (1977) design features of language, look at Table 13.1. We now know that components of the waggle dance correspond to the direction and distance to the food source, as well as the quality of the food. We also know that different dances are used if a food source is relatively close to the hive. Finally, bees are able to adjust their dances depending on the number of forager and processing bees needed at the moment. How much of this behavior would you categorize as language, and why? Try to think of evidence for both sides of the argument.

Is the Bees’ Dance Communication a Language?
How can we decide if the bees’ communication is a language or a simple communication like a dog barking? After all, numerous animals communicate with sounds and gestures, but are these different languages? The problem is that there is no completely accepted definition of a language. To return to the chapter’s opening dialogue, we are struggling with the second fictitious bee’s concern that you should know what a language is before determining whether bee communication is a language. Your work on Exercise 1 might help convince you that it is difficult to obtain a commonly agreed upon definition of language. Even without a “universal” definition of language we can, nevertheless, compare the bees’ communication to a list of key design features thought to characterize language. (Note that even with this list there is not complete agreement on which
TABLE 13.1 Bee Dance Components: Forms and Functions Waggle Dance Longer Distance to Food Source Direction Information = Angle of straight run to sun’s azimuth Distance Information = Tempo of dance Quality of Food = Duration of dance Round Dance Shorter Distance to Food Source Shaking Signals and Tremble Dance Excite and Inhibit Forager and Processing Bees

Animal Communication: Honey Bees


features should be included.) The ones listed here were originally proposed by Charles Hockett in the 1950s and 1960s, although they have since been modified.

A Mode of Communication
A mode of communication is the means for transmitting messages. For humans the mode is predominantly by sound. For bees several characteristics of the waggle dance could serve as communication modes, though further research is needed here. A likely mode is the sound of the buzzing produced by the dancer during the waggle run. Another possible mode is tactile because the bees following the dancer can touch her with their antennae. One line of research investigated the airflow patterns generated by the dancer (Michelsen, 1999; see Dyer, 2002, for commentary). Could the pulses of air felt by the recruit bees (but without a direct contact with the dancer) be a kind of tactile mode? During the waggle run, the dancer produces vibrations transmitted through the wax comb, which could be felt through the feet and legs of the recruit bees (Dyer, 2002). Could this comb vibration be a mode? If so, the bees would be communicating through the comb. There was even quite an unusual line of research that involved measuring the temperature of the thorax (a bee’s middle segment). During a waggle run, the thorax of the dancer becomes hotter. Quite surprisingly, the sweeter the sugar water (a measure of nectar quality) collected by the dancer, the hotter her thoracic temperature (Stabentheiner, 1991). Could a bee’s radiant (thoracic) body heat be a communication mode for the quality of the food source? Yet, here is something to ponder. As explained above, after finding more profitable (better) nectar locations, bees waggle dance for longer times. As we have seen, dances with longer durations convey information about the location to more recruits. This simple method directs recruits to better food sources and consequently away from poorer quality food sources (Seeley, 1995). Is the dance duration (as in repeating the dance over and over) a communication mode? Specifically, can repetition be a kind of communication mode? What about a visual mode? Just considering the honey bees in North America, the answer seems to be—no visual mode, because they nest in cavities. After all, the bees are dancing in the dark. However, this conclusion is not true for all dancing honey bees. For Apis florea, an Asian open-nesting honey bee, the dancer is in full-lighted view of the recruits. For their dances, vision is a communication mode. Even when Apis florea performs the waggle run, she arches her abdomen almost vertically, a posture that makes her more easily seen by the recruit bees. In contrast, when bees (Apis mellifera) in North America dance, they do not raise their abdomens in this manner (Gould and Gould, 1988). So the modes depend to some extent on the particular species of honey bee. As opposed to having one or two modes (for example tactile and sound) perceived separately, bees might need to perceive the modes for the navigational information together in some sort of configuration. Some modes could even transmit the same information. This redundancy could help reduce navigational errors or function as a kind of backup system. These communication modes also depend on what the recruit bees can perceive (and even if they perceive it at all). A bee’s perceptional world in many respects is quite different from ours. As we try to understand a dancing bee’s communication modes, our human world-view might hinder us.

Discreteness is a language property where complex messages are constructed from smaller discrete units. For example, a sentence is built from numerous smaller units (e.g. phrases, words, morphemes, and phonemes).



For the bees’ waggle dance, it is tempting to think of the message as comprised of two “smaller” pieces of information, namely the direction and distance. We might then conclude that the bees’ communication system displays the property of discreteness, at least in a limited sense. It could also be that whatever components of the dance convey direction and distance might only make sense to the recruit bee if both are conveyed simultaneously as a unit. For example, if someone gives you directions and you hear the distance but not the direction, you still understand the distance because it is a discrete piece of the instruction. It is possible that, for a bee, the distance does not make any sense by itself, unless she also knows the direction. In that case direction and distance are not discrete units, weakening an argument that the bee communication system has the property of discreteness. However, if bees convey direction and distance as a unit and the flower odor is also included, as it is under natural conditions, is discreteness now a property? Bees can find unscented sugar water, so unlike the assumption above in which distance and direction might be conveyed as a unit, repeated evidence strongly suggests that bees can use dance information without requiring information from odor. Odor, then, might be a discrete element in bee communication.

Interchangeability is the ability of an individual to both send and receive messages. For example, with human languages, individuals can send messages, usually by speaking, and receive messages, usually by listening. In some forms of animal communication, a message is restricted to flow one way, where one individual must always be the sender and the other must always be the receiver. For example, when a queen bee flies from the hive to mate, she emits a chemical message through the air to attract the male bees. A queen can only send the message, and the males can only receive it. This form of communication does not possess interchangeability. What about the bees’ dance communication? Is it interchangeable? Here we need to be careful. An individual bee can send a message, if she is the dancer, or receive the message, if she is the recruit. In a particular dance episode, a bee can only be a dancer or a recruit—not both. The navigational information flows only from the dancer to the recruit. In that sense the communication is not interchangeable. In a bit of related communication, a recruit bee can send a dance-stopping message to the dancer (Pastor and Seeley, 2005). Again, for a particular dance episode, this communication is not interchangeable. Notice that in a particular human conversation both the sender and receiver are not generally restricted to sending certain messages, but as far as we know that flexibility is not true with bee communication.

Cultural Transmission
With cultural transmission, some aspects of the communication need to be learned through communicative interactions with another user of that system. For example, children learn discourse patterns from members of their speech community. This learning interaction is apparently not needed with the bees’ dance communication. The ability to perform and follow dances appears to be hard-wired into the bees’ nervous system. Bees reared without contact with dancing bees can correctly perform and follow dances on their first opportunity (Gould and Gould, 1988). While this innate ability is quite remarkable, it is also inflexible in that learning new communication patterns is impossible.

Animal Communication: Honey Bees


One way to observe this inflexible lack of learning is to rear bees of one subspecies (race) in a hive of a different subspecies. Each subspecies has it own dialect, which would be its own way of converting dance information into a distance measurement. What happens when a recruit bee that genetically originated in Italy (Apis mellifera ligustica) follows a dancer that genetically originated near Yugoslavia (Apis mellifera carnica)? (That would not happen naturally because geographic barriers separate the subspecies, but it can be created experimentally.) Each bee follows her own dialect, and they misinterpret each other’s dances. There is no learning the other bee’s dialect. Here are interesting questions to think about. Even if a bee could learn another dialect, how could she? What kind of innate capacities would be involved in learning it?

Arbitrariness refers to the lack of a logical or predictable connection between the form and meaning that make up a linguistic sign. Think of it as, linguistic sign = form + meaning. For example, to drive cars, we learn that green indicates the way is clear to proceed. The color “green” is the form and “proceed” is the meaning. There is no logical connection between “green” (form) and “proceed” (meaning) since that choice of color is essentially arbitrary. To see the arbitrariness of this connection, consider another transportation system. When railroads started to develop in Britain, well before the invention of the automobile, the meaning of green was “caution” and the meaning of “proceed” was a white light (Solomon, 2003). Conversely, if the meaning is derivable or predictable from the form, then the form and meaning are not arbitrary or nonarbitrary. An extreme degree of nonarbitrariness is the iconic sign, which is composed of a related representation of the meaning. For example, a cartoon drawing of a deer indicates the meaning of an actual deer. How would this concept of arbitrariness apply to the bees’ dance? After deciphering the waggle dance, some investigators interpreted it as a ritualized or symbolic form of the actual foraging flight. For example, longer durations of the waggle run part of the dance “looked” like a longer symbolic flight to a more distant food source. For our investigation, we could say the dance is an iconic form of the foraging flight. But let us think carefully here. This conclusion came after deciphering the waggle dance. None of that appears to have been obvious to people observing the dances before von Frisch, and he made some mistakes when interpreting the dance (which he corrected). Furthermore, as we learned, some beekeepers had only a vague notion of the dances announcing a food source. Not surprisingly, then, the forms of the various components of the waggle dance (number of circuits, angle of the waggle run) would be expected to have arbitrary connections to their meanings (direction and distance to the food) since it was not obvious what these things really meant. Another measure of the distance to the food is the duration of just the waggle run part of the dance (another human choice for the form, but perhaps accessible to the bee). Again from experimentation, we know that longer waggle durations are associated with longer distances. But it is fair to say the connection between the duration of the waggle run and the distance was not obvious. Rather, finding the connection was more like breaking a code where duration of the waggle run could be the bee’s “word” for distance. From experimentation we know that the distance from the hive to the food (meaning) can be estimated from the number of circuits per 15 seconds (a human choice for the form). No iconic or obvious connection seems to exist between the circuits and distance, and in that regard the form–meaning connection seems arbitrary. Indeed, the dance dialects underscore this arbitrary connection because different



FIGURE 13.6 Comparing the waggle dance circuits per 15 seconds as a function of the distance for two subspecies of honey bees, a European bee (Apis mellifera carnica, triangles) and an African bee (A. m. adonsonii, diamonds)
Note: For the same distances, the subspecies differ in the number of dance circuits. Source: Redrawn from von Frisch, 1967.

subspecies have different dance-to-distance conversions. However, it is not obvious without learning the procedure (making a graph like Figure 13.6) how a particular distance can be predicted from the number of circuits/15 secs. In that sense the form– meaning connection is nonarbitrary. What could be the form for the food direction? One could argue that at first there is no obvious form, so whatever it is, the connection to the food direction must be arbitrary. With some contemplation, one could suspect the form is related to the direction of the waggle run itself because it is the straightest part of the dance and could be used as a directional pointer. In that regard, perhaps there is a degree of nonarbitrariness here. Nevertheless, how the bees use the waggle run to indicate the food direction should be part of the form, and that part is not obvious. It had to be determined experimentally that the angle of the waggle run relative to vertical direction corresponded to the angle to the food relative to the sun’s azimuth. In that sense there is a degree of arbitrariness. Nevertheless once the procedure is known, we can predict the food’s direction from the waggle run angle. Thus in this sense the form–meaning connection is nonarbitrary. Finally, on a vertical comb, bees appear to measure angles from the vertical direction. Their choice of the upward vertical direction (form) for the reference direction (meaning) seems to be arbitrary since presumably bees could have chosen the downward vertical direction or some other direction.

Displacement is a property of language in which communication can occur about things not present in space or time. We can talk about events that happened in the past or that could happen in the future. The bees’ dance language exhibits displacement, though in a somewhat limited sense. When a bee dances, the message is about the food source she visited last. While that

Animal Communication: Honey Bees


foraging did occur in the past, it occurred very recently. That might be the end of the discussion on displacement except for some new research on how the recruit bees access dance information (Biesmeijer and Seeley, 2005). The key thing to keep in mind is that some recruit bees have been previous foragers. Using a glass-walled hive, these researchers followed numerous foragers over their foraging lifetimes. Some observations were expected, such as the finding that 60 percent of novice foragers were recruited with waggle dance information to find their first food source instead of going out and searching on their own. Biesmeijer and Seeley (2005) also found that following waggle dances for the discovery of new food sources accounted for only 12–25 percent of all interaction with the dances. On the other hand, following waggle dances for reactivating foragers or confirming the productivity of a source accounted for 75–88 percent of interactions with dances, by far the majority. For the concept of displacement, the dancer is communicating something that happened in the very recent past, that is, the location of her most recent foraging trip, which happened maybe within the past 15 minutes. That is not necessarily true about the bees following the dancer. Conceivably she is receiving information on a food location that she visited maybe a day ago. Thus, assessing displacement in bee communication, the message could be something in a recruit’s more distant past. It could also be argued that bee dance communication does not exhibit displacement. The dance could just be an elaborate posture for indicating a message like, “Fly 300 meters at 45 degrees to the right of the sun.” Since the recruit bee is physiologically ready to forage, when the bee gets to that location, she forages. This behavior could be interpreted as analogous to a bee exhibiting a certain posture that indicates to other bees “Groom me.” The conclusion would be that the dance communication, like the “Groom me” posture, conveys nothing about the past and does not exhibit displacement.

Productivity is a feature that allows the free and easy production and understanding of any number of messages that have never been expressed previously. These messages can express novel ideas. Human language has this open-ended quality. What about the bees’ dance communication? In principle, bees can communicate the coordinates of numerous new food locations within their foraging range. But that is just a consequence of the geometry of a plane where any point can be located by its distance from the origin (taken as the hive) and an angle (direction to the food). “New” messages are always in the context of a (two-dimensional) resource location. In that sense there is a limited level of productivity. As far as we know, honey bees cannot formulate truly novel messages. For example, honey bees (specifically Apis mellifera and probably the other Asian species in the genus Apis) cannot form a new message that contains a third dimension, namely the height of the food source. Some other social bees, however, can indicate the height (Nieh and Roubik, 1998), so that is not an impossible feat.

Let us return to the two bees confused about whether they have a language. Although I made them converse like humans (How exactly did I do that? Think of the design features), we now know that bee communication is vastly different from the way that people communicate. Yet when comparing bee communication to the design features of a language, there are similarities. Bees do have numerous possible modes of communication (sound, touch, air pulses, etc.). One can make a reasonable argument



that the waggle dance exhibits discreteness (distance, direction, perhaps odor too). For interchangeability, a bee can send or receive a message, but she cannot do both in a particular dance episode. Cultural transmission appears not to apply to bee communication. On the other hand, arbitrariness, displacement, and perhaps productivity all apply (to a limited extent). Thus in some aspects bees seem to have a communication system with features of human language. In other ways, these features are lacking. Debating these points helps us better understand human language. We also push Hockett’s (1977) design features of language into ways we might not normally think about them. Isn’t it amazing what we can learn about human language from understanding bee communication?

Exercise 5

What do you think? Break up into seven small groups. Each group should take a Hockett (1977) design feature covered in the chapter and list the arguments for and against calling bee communication language.

Biesmeijer, J. C. and Seeley, T. D. (2005) The use of waggle dance information by honey bees throughout their foraging careers. Behavioral Ecology and Sociobiology, 59(1), pp. 133–142. Chittka, L. (2004) Dances as windows into insect perception. PLoS Biology 2, pp. 898–900. Crane, E. (1999) The world history of beekeeping and honey hunting. London, Taylor and Francis. de Marco, R. J. (2006) How bees tune their dancing according to their colony’s nectar influx: Re-examining the role of the food-receivers’ ‘eagerness.’ Journal of Experimental Biology, 209, pp. 421–432. Dyer, F. C. (2002) The biology of the dance language. Annual Review of Entomology, 47, pp. 917–949. Françon, J. (1939) The mind of the bees. London, Methuen. Gould, J. L. and Gould, C.G. (1988) The honey bee. New York, Scientific American Library. Hockett, C. F. ed. (1977) The view from language: Selected essays, 1948–1974. Athens, University of Georgia Press. Michelsen, A. (1999) The dance language of honey bees: Recent findings and problems. In Hauser, M. D. and Konishi, M. eds. The design of animal communication. Cambridge, MIT Press, pp. 111–131. Nieh, J. C. and Roubik, D.W. (1998) Potential mechanisms for the communication of height and distance by a stingless bee, melipona panamica. Behavioral Ecology and Sociobiology, 43, pp. 387–399. Pastor, K. A. and Seeley, T. D. (2005) The brief piping signal of the honey bee: Begging call or stop signal? Ethology, 111(8), pp. 775–784. Riley, J. R., Greggers, U., Smith, A. D., Reynolds, D. R. and Menzel, R. (2005) The flight paths of honeybees recruited by the waggle dance. Nature, 435, pp. 205–207. Root, E. R. (1908). [Editorial] Gleanings in bee culture, Vol. 36, No. 1–24. New York, A. I. Root. Seeley, T. D. (1995) The wisdom of the hive: The social psychology of honey bee colonies. Cambridge, Harvard University Press. Seeley, T. D. and Towne, W. F. (1992) Tactics of dance choice in honey bees: Do foragers compare dances? Behavioral Ecology and Sociobiology, 30(1), pp. 59–69.

Animal Communication: Honey Bees


Seeley, T. D., Mikheyev, A. S., and Pagano, G. J. (2000) Dancing bees tune both duration and rate of waggle-run production in relation to nectar-source profitability. Journal of Comparative Physiology: A neuroethology, Sensory, Neural, and Behavioral Physiology, 186(9), pp. 813–819. Solomon, B. (2003) Railroad signaling. Minneapolis, MBI Publishing. Stabentheiner, A. (1991) Thermographic monitoring of the thermal behavior of dancing bees. In Goodman, L. J. and Fisher, R. C. ed. The behaviour and physiology of bees. Wallingford, UK, CAB International, pp. 89–101. Tautz J., Rohrseitz, K., and Sandeman D. C. (1996) One-strided waggle dance in bees. Nature, 382, p. 32. von Frisch, K. (1967) The dance language and orientation of bees. Cambridge, Harvard University Press. Wenner, A. M. and Wells, P. H. (1990) Anatomy of a controversy: The question of a “language” among bees. New York, Columbia University Press. YouTube. (2009) Bee dance (waggle dance) [Internet]. Available from: com/watch?v=–7ijI-g4jHg [Accessed June 17, 2009].

Other Resources
Friends of Washoe. (2009) Meet Washoe [Internet]. Available from: http:// [Accessed June 17, 2009]. A website about Washoe, a chimp who was raised by scientists and taught American Sign Language. Premack, D. (2009) Publications [Internet]. Available from: ~premack/Publications.html [Accessed June 17, 2009]. A website with references related to chimp/language studies. YouTube. (2009) Alex the talking parrot [Internet]. Available from: watch?v=R6KvPN_Wt8I [Accessed June 17, 2009]. This clip and others on YouTube show Alex, a parrot said to have learned language.

Language and Communication Science
The text finishes with an exploration of linguistics as it is employed by speech–language pathologists, neurolinguists, and psychologists. Part V offers us four chapters by therapists and researchers in the science world. Ann D. Jablon, in Chapter 14, shows us how a speech– language pathologist diagnoses and treats speech and language disorders. Readers work through case studies with the author to understand the process of diagnosis and treatment. Cecile L. Stein, in Chapter 15, looks more narrowly at language, with her investigation of how children tell stories (narratives) in different parts of the world. Chapter 16, by Yael Neumann, Linda Carozza, and Anastasia Georgiou, is a fascinating look into linguistics applied to neurological research. The brain as it ages, be it along a normal progression or into dementia and/or Alzheimer’s disease, can be better understood by employing psycholinguistic research. The book ends with a chapter by Marion Blank and Mary Beth Cull on a subject we hear much about these days but need to understand better: autism. Chapter 17 shows us the history and clinical view of what is now called Autism Spectrum Disorders in children.

Ann D. Jablon
This Chapter Explores:

Communication Disorders A Personal Perspective

The Professions of Speech–Language Pathology and Audiology Communication, Speech, Hearing, and Language The Classification of Speech, Hearing, and Language Disorders

The Nature of Language Disorders in Children Case Studies of Children with Speech and Language Disorders Principles of Treatment Language Disorders in Adults Language Differences

The function or purpose of language . . . is to allow for an efficient, effective means to develop and exchange thoughts, feelings, ideas, and information. Kohnert, 2007

How on earth did I become a speech–language pathologist when all I wanted to do was write novels? It was the love of language and the desire to be of help to people that led me to speech–language pathology. So, after a brief stint as an English major and an equally brief stint as a psychology major, I stumbled into the field of speech–language pathology, which asks questions about what prevents children from learning language, what happens when disease or accidents cause people to lose their ability to communicate using language, and how can we create the changes that will help them to communicate to the best of their abilities. I have had an exciting career as a speech–language pathologist working primarily with school-aged children, their teachers, and their families. I have spent equally many years teaching and supervising students who are preparing to enter the profession. I would like to express my thanks to Radmila Dym, a speech–language pathology major at Marymount Manhattan College, who served as my research assistant during the preparation of this chapter. She did fine work and I enjoyed our collaboration.



Introduction: The Professions of Speech–Language Pathology and Audiology
The purpose of this chapter is to introduce readers to communication disorders in general. I will focus primarily on speech and language disorders of children to illustrate the concepts presented and to demonstrate how speech–language pathologists address the needs of individuals with communication disorders. Speech–language pathologists (SLPs) are professionals who channel their knowledge of speech and language to help children develop their ability to learn to use language, and to help adults recapture their ability to use language when the typical process has been subverted. This is compelling work because language is the hallmark of our humanity. The inability to acquire language or the loss of language isolates individuals from our human community. Speech–language pathologists devote their professional lives to facilitating the changes that may alter the course of isolation when speech, language, or hearing use fails. Audiologists are professionals who are concerned with hearing and balance disorders.
Audiologists also select, fit, and dispense amplification systems such as hearing aids; they prevent hearing loss through providing and fitting hearing protective devices, consultation on the effects of noise on hearing, and consumer education; and they can serve as expert witnesses in litigation related to their areas of expertise. Some audiologists conduct research on hearing, tinnitus, and the balance system. (ASHA, 2009a)

The American Speech–Language–Hearing Association (ASHA) is “the professional, scientific, and credentialing association” for speech and hearing professionals (ASHA, 2009a). You can learn a great deal about the professions by exploring ASHA’s website ( There is a specific link on the ASHA site for students to learn about the professions and the career prospects in these fields. There is also a link for the public, which includes important information on communication and communication disorders. I will refer to specific documents throughout this chapter and will ask you to read the sources to support some of the activities and exercises below. Not only do speech and hearing professionals help people who have communication disorders, but they also contribute to the knowledge that science has developed regarding human communication. A basic premise in linguistics is that the study of language provides insights into how the mind works, i.e. what makes us human. Some of the knowledge that we have about speech and language comes from observations of communication disorders. Thus, not only does information from the scientists studying language help clinicians provide better care for individuals with language difficulties, but also, clinical knowledge of how language breaks down furthers the understanding of language and the organization of language within the brain. One branch of study that investigates the nature of language, and other cognitive processes, within the structures and functions of the brain is called neurolinguistics. Neurolinguistics is an interdisciplinary science, which means that linguists, speech–language pathologists, audiologists, psychologists, neuroscientists, among other scientists, all contribute to our understanding of human language.

We may all suffer moments of communication failure: we can’t think of the word we want to say; we can’t understand a passage we’ve just read; we can’t get our words out

Communication Disorders


smoothly; we lose our voices; or we mix up the sounds we mean to say. Sometimes we experience temporary losses of hearing or a ringing in our ears (tinnitus) due to exposure to loud noises (after concerts, for example). Individuals with speech, hearing, and language disorders struggle to express their thoughts and/or understand and hear the thoughts and ideas of others expressed through speech and language on a daily basis.
Can you recall, perhaps in a class or in a social situation, when you failed to express ideas and communicate effectively? How did you imagine others perceived you, or judged you, at that moment? Write down your feelings. The feelings that accompany a communication failure have been called the hidden component of a language disorder (Van Riper and Erickson, 1996). Now reflect on what communication failure means to those who experience it beyond mere moments. Discuss the impact you think this may have on their lives. Why do you think the professions of speech– language pathology and audiology are important? Remember the quotation that began this chapter.

Exercise 1

Communication, Speech, Hearing, and Language
An important distinction that speech, language, and hearing scientists and clinicians adhere to is the distinction among communication, speech, hearing, and language.

Communication encompasses many systems for transmitting a message. For example, people use gestures, alone or to accompany the words and sentences of their messages. They use facial expressions, eye contact, distance from the listener, silence, and many other patterns to convey meaning. These are nonlinguistic or non-verbal aspects of human communication. Sometimes the verbal message and the non-verbal message contradict each other.
Say the same sentence (for example, “I really love your haircut”) using different non-verbal indicators. Did you roll your eyes or change your vocal inflection? Observe others when they are speaking with you. Collect five examples of conflicting verbal and non-verbal messages. Discuss these in class.

Exercise 2

Speech, on the other hand, is as specific as communication is broad. Speech is a sound signal that transmits the message from a speaker to listener. It is the primary mode of conveying linguistic messages. Other ways that we encode our linguistic messages include writing and signed languages. Speech is often called an overlaid function because it uses structures of the body that serve a primary biological purpose. For example, the tongue, lips, teeth, and palate of our mouths (the oral cavity) serve the biological purpose of ingesting food and preparing the food for digestion for survival. Yet, these same structures have been adapted for the overlaid function of making the speech sounds.



Exercise 3

Say different sounds and notice which structures of the oral cavity are moving and changing shape to make the different sound patterns of your language. If you speak more than one language, try it in each language and notice how the movements change and make a different sound. Additionally, you may try this with a partner. Record each other’s facial movements. Or, record yourself and listen to differences in the sounds or prosody (the rhythm and intonation of speech).

Hearing is crucial to the understanding of the speech signals that are transmitted to convey messages. For individuals who are deaf or hard of hearing, the development or perception of speech is compromised. Many individuals who are deaf use signed languages to communicate. Signing language is an alternative modality to speaking language. In the history of human communication, as well as currently, there are many communities that use signed language as a primary mode of expression (see, for example, Everyone Here Spoke Sign Language: Hereditary Deafness on Martha’s Vineyard; Groce, 1988). In order to ensure that the opportunity to develop speech, or to develop a system of signed language, or to develop both, is available to individuals who are born deaf or profoundly hard of hearing, at this writing, forty states have passed laws requiring all newborns to have a hearing screening. Audiologists, as we discussed, are the professionals who are primarily responsible for identifying and assessing the hearing ability of individuals.

Exercise 4

Can you recall having your hearing tested? Can you describe what happened? Did you know that noise is a common factor in impairing hearing? Go to the National Institute on Deafness and Other Communication Disorders (NIDCD) website (http:// to learn more about the effects of noise on hearing and how to protect hearing health. How will you protect your hearing health now that you have more information? Write down two actions that you will take to ensure that you will protect your hearing, or the hearing of family and friends.

Language, as I noted earlier, is considered a uniquely human capacity. Language has been defined as a symbolic system, consisting of rules and a set of arbitrary symbols that convey meaning through spoken, written, or signed modalities. The rules of language capture the way we combine the arbitrary symbols to make meaning. For example, “duck” is an arbitrary symbol that means a particular type of water fowl. It is arbitrary because our language community has agreed that this symbol stands for that water fowl in the real world, but the word “duck” does not embody the actual properties of the fowl. We could have agreed on any other word to denote this water fowl. If we combine the words “See the duck,” we have one meaning; if we combine the words “The duck sees,” we have a different meaning. This very simplistic example captures the essence of how language works. The nature and complexity of language are discussed in more depth in a number of other chapters.

Communication Disorders


The importance of making the distinction among communication, speech, hearing, and language is to illuminate the discussion that follows. Speech, language, and hearing are all interconnected and all contribute to how effectively we communicate. Often a problem in one area is associated with or coexists with a problem in another area. That is why we use the label “communication disorders” to encompass the entire range of problems and causes of failures in communication. But we need more specificity if we are to understand how best to treat the problems of individuals.

Classifying Communication Disorders
We’ve just learned that speech, hearing, and language can be distinguished from one another. Similarly, communication disorders are classified in ways that reflect these distinctions. Speech, hearing, and language disorders stem from many causes. There are developmental disorders that cause speech and language difficulties, for example, Autism Spectrum Disorders, Down syndrome, or Specific Language Impairment. There are acquired causes of speech and language difficulties such as injury or illness that causes brain damage that results in a loss of language known as aphasia. We will examine some of these causes below. We will consider the implications of speech and language difficulties for individuals and their families. Speech, language, and hearing professionals use various systems to classify communication disorders. ASHA (1993) is an excellent paper that defines communication disorders and communication variations. In the organization’s paper, the reader will find definitions for the categories that are used by speech and hearing professionals. The information for retrieving this paper is cited below. The first definition, that of a communication disorder, demonstrates the interrelatedness of all of the components we’ve discussed. The ASHA paper states:
A communication disorder is an impairment in the ability to receive, send, process, and comprehend concepts or verbal, nonverbal and graphic symbol systems. A communication disorder may be evident in the processes of hearing, language, and/or speech. A communication disorder may range in severity from mild to profound. It may be developmental or acquired. Individuals may demonstrate one or any combination of communication disorders. A communication disorder may result in a primary disability or it may be secondary to other disabilities. (ASHA, 1993)

(See–00208.html for full definitions of each category.) One common system that is used to categorize communication disorders is based on the symptoms that cause communication failure. Thus we can classify a communication disorder as a speech, language, or hearing disorder.

Speech Disorders
Speech disorders are further classified as disorders of sound production, fluency, and voice. We will consider each type of disorder.

Disorders of Sound Production The failure to be able to produce the sounds of one’s language typically can result in a reduction in the ability to be understood (intelligibility). Substituting one sound for



another (saying “w” instead of “r”), omitting a sound when it should appear (saying “ca” for “cat”), distorting the way a sound is produced (using too much air flowing over the sides of the tongue when saying “sh”), or adding sounds (saying “balue” instead of “blue), all characterize the way that sound production can be atypical. If a person’s speech includes many atypical patterns, then communication can be impeded.

Disorders of Fluency You may be familiar with the term “stuttering.” This is one type of fluency disorder, which is also categorized as a speech disorder. According to ASHA,
A fluency disorder is an interruption in the flow of speaking characterized by atypical rate, rhythm, and repetitions in sounds, syllables, words, and phrases. This may be accompanied by excessive tension, struggle behavior, and secondary mannerisms. (ASHA, 1993)

A wonderful video, which shares the feelings and insights of individuals who stutter, is called Transcending Stuttering: The Inside Story (2004).
Watch the video as a class. Based on this viewing, write down your definition of stuttering. What do you think the main characteristics of stuttering are? What is the message of this film? For further information, access the Stuttering Foundation of America website ( Here you will find factual information on the causes and treatment of stuttering. You will learn about many successful people who live with a fluency disorder, and you will learn about the impact of this disorder on the lives of famous and ordinary people.

Exercise 5

Disorders of Voice One way to define voice is “the right to be heard.” But what if your voice did not allow you to be heard? What if you were a student, a teacher, a salesman, a professional actor or singer and your voice did not allow you to be heard? I am sure you have had the experience of losing your voice, laryngitis, when you suffered with a cold or your allergies overcame you. This is a temporary inconvenience. ASHA defines a voice disorder as
characterized by the abnormal production and/or absences of vocal quality, pitch, loudness, resonance, and/or duration, which is inappropriate for an individual’s age and/or sex. (ASHA, 1993)

The most extreme loss of voice is caused by the removal of the larynx due to disease, usually cancer. The larynx comprises the cartilaginous and muscular structures located in the neck that house the vocal folds. However, the loss of voice is devastating to all, since we depend upon our voices “to be heard” at every stage and in every aspect of our lives. A useful site for understanding our voices and the disorders that can affect them is the Voice Foundation (2009). Here is what the Voice Foundation has to say about voice disorders:
According to the National Institute on Deafness and Other Communication Disorders of the National Institutes of Health, approximately 7.5 million Americans have trouble

Communication Disorders using their voices. Despite widespread voice problems, too little is known about the vocal disorders that cut short careers, impair the speech process, devastate and even threaten people’s lives. (Voice Foundation, 2009)


Hearing Disorders
Our ability to hear connects us to the world in every way. A hearing disorder can be a barrier to this connection and can affect speech and language. Essentially, a hearing disorder is the result of impairment in the auditory system (the structures and functions of the ear or the nerves, which connect the ear to the auditory centers in the brain). There are two broad classifications of hearing disorders. A person may be classified as deaf or hard of hearing. According to ASHA,
Deaf is defined as a hearing disorder that limits an individual’s aural/oral communication performance to the extent that the primary sensory input for communication may be other than the auditory channel. (ASHA, 1993)

An individual who is classified as hard of hearing may have a permanent loss of hearing or a fluctuating hearing loss. For example, children may have ear infections, which cause a temporary loss of hearing. Another cause of a fluctuating hearing loss may be excessive wax (cerumen) buildup in the ear canal. Some permanent losses can be caused by exposure to loud noises. There are certain occupations that place individuals at risk for noise-induced hearing loss. For example, musicians are at risk for this type of hearing loss. Can you think of other occupations that place individuals at risk? If a person is hard of hearing, he or she still uses the “auditory channel as the primary sensory input for communication” (ASHA, 1993). Sometimes, an individual who has a hearing disorder will use hearing aids to increase his or her ability to hear sound. Whether a person can successfully use hearing aids depends on the cause of the hearing loss. You can read more in-depth information on hearing loss on the ASHA website (ASHA, 2009c) and on the Mayo Clinic website (, which is an excellent source of information on many health issues. Try taking the self-test on the ASHA website (ASHA, 2009c). Answering these questions will increase your awareness of the warning signs regarding hearing loss. The rest of this chapter will not discuss the details of speech and hearing disorders, but the reader is cautioned that a category such as “language disorders” does not exclude the presence of a coexisting speech or hearing impairment. Having cautioned you that I am focusing on only one aspect that underlies a communication disorder, I will turn our investigation to speech–language pathology and the topic of language disorders.

Speech–Language Pathology Revisited
Some common concerns of speech–language pathologists, one group of professionals who study, diagnose, and treat individuals with language disorders, are: • to determine the cause of a language disorder; • to determine the nature of a language disorder; • to determine whether a language difficulty stems from an impairment or from a



difference (such as using a dialect unfamiliar to the community or learning the community’s language as a second language); • to assess the impact of a language disorder in the different settings in which the individual communicates; • to apply scientifically tested methods to habilitate or rehabilitate language use. Many speech–language pathologists are employed in schools all over the country because children who fail to develop language in a typical way are at risk for academic failure. Other settings in which speech–language pathologists are employed include hospitals, rehabilitation centers, and long-term care facilities. Some speech–language pathologists work with families in their homes to provide early intervention when infants are at risk for developing language disorders or when adults are unable to travel to receive treatment.
The American Speech–Language–Hearing Association (ASHA) has a website with information for students, the public, and professionals. Go to the website ( Select the “public” tab. Dividing the class into two teams, have one research the job and settings of speech–language pathologists; the other team will research audiologists. Discuss the similarities and differences between the two branches of the profession.

Exercise 6

In the field of speech–language pathology, the most important goal of providing services is to ensure that individuals communicate as effectively as they are able. In order to do so, the speech–language pathologist must identify where the communication system has broken down when a person is struggling to send messages and to understand the messages conveyed to him/her.

Speech–Language Pathology and Linguistics
Because language is a uniquely human capacity through which we communicate, the failure to develop language or the sudden loss of language has a profound effect on individuals and their families. The field of linguistics proposes theories of language that describe the knowledge that speakers/listeners have of language. The field of speech– language pathology uses this information to understand the needs of children who fail to develop the language system in a typical way (developmental language difficulties) and of those who have developed typically but through disease or injury lose the ability to use language to communicate effectively (acquired language difficulties). Both fields of study inform each other. One example of the important relationship that exists between the two fields is exemplified by the seminal work Language Development and Language Disorders (Bloom and Lahey, 1978). Bloom and Lahey proposed a taxonomy that categorized language into three components, which encompass the traditional five components of phonology, morphology, syntax, semantics, and pragmatics. The components they propose are form (related to the patterns of speech sounds, or phonology; word-formation, or morphology; and sentence structure, or syntax), content (the meaning of words and sentences, i.e. semantics), and use (the socially appropriate norms of language use in a culture, i.e. pragmatics). Speech–language pathologists use this taxonomy (or the traditional five components) to guide their assessments and treatment of children with language disorders.

Communication Disorders


Let’s consider how this might work. The information in the case studies presented below is based on actual children whom I treated. I have changed their names to ensure their confidentiality.

Case Study I
Perry was 3 years old and enrolled in a preschool program. His teachers were concerned about his progress in the program and advised his parents to have his language evaluated. Upon evaluation, Perry demonstrated mastery of the appropriate forms for his age. He articulated his sounds clearly, did not omit any sounds, and was able to use blended sounds where appropriate (phonology). He used a variety of endings to indicate plurals, past or present actions, in addition to using function words (e.g. articles, conjunctions, prepositions) indicating morphological abilities at age level. He was able to structure sentences into appropriately ordered utterances (syntax). However, despite Perry’s elaborate and sophisticated vocabulary for a 3-year-old, his teachers and peers had difficulty understanding the meaning of his utterances. On the Monday after Thanksgiving, Perry was playing with the dollhouse, arranging furniture and narrating the family’s activities. A striking example of a typical utterance was: “The children saw them floating by the windows.” Here Perry was referencing the floats he had just seen in the Thanksgiving Day parade. Note that his syntax and morphology are intact. However, without context, a listener would not know what he was referring to and, as he elaborated on his theme, what he was talking about. This is an example of a deficit in the content realm of language. In addition, Perry had difficulty marshalling his knowledge of the form of language to accommodate the different contexts for language use. He had difficulty with establishing a topic, launching into a semantically bizarre presentation during show and tell. No topic and statements like the example above left his teachers and classmates puzzled about his message. This was typical of Perry’s communication. Now let’s consider a different pattern.

Case Study II
Billy, age 6, had a large vocabulary accompanied by extensive world knowledge (content). He had an excellent ability to use language across many settings: in school, with his peers, with familiar adults, etc. However, at age 6, Billy demonstrated many sound errors (phonology); he did not use articles, pronouns, or correctly mark the tense of the verbs in his sentences (syntax and morphology). In contrast to Perry, who had language difficulties characterized by deficits in the content and use domains of language, Billy’s deficits were in the form domain of language. As a beginning reader, Billy fell way behind others in his grade. These cases illustrate the benefit of using a taxonomic system that pinpoints the area with which a child is having difficulty. While both children may be said to have a language disorder, they have very different strengths and weaknesses. Understanding the nature of the language disorder is important because it guides teachers, speech–language pathologists, psychologists, and families toward the appropriate treatment, predicts possible future problems, and guards against misdiagnosis. In Perry’s case, a psychological evaluation done prior to his language evaluation wrongly concluded that Perry had a severe personality disorder. However, by correctly identifying Perry’s difficulties as a language disorder, and not a personality disorder, his family and the professionals involved with them were able to provide the best placement and treatment to ensure his progress. Perry was placed in a school for children with language



impairments and received intensive language intervention, which focused on developing his pragmatic and semantic abilities. While Perry remained in schools for children with special needs because of his language disability, it was recognized that he did not require psychiatric care. In Billy’s case, the identification that his language problems were in the form domain also predicted that Billy would struggle with reading. Children with persistent reading failures will struggle academically as school becomes increasingly dependent on the student’s ability to learn to read and then read to learn. Billy’s language intervention focused on developing his syntactic, morphological, and phonological skills, as well as the language skills which are critical for the development of reading.

Case Study III
Here is information about Elizabeth, a 6-year-old child with a language disorder. Elizabeth uses complete and complex sentences. She knows many words, but sometimes uses them inappropriately. For example, she calls sisters “twins.” She has clearly articulated speech and makes no grammatical errors when formulating her sentences. Observations of Elizabeth’s behavior include the following specific, but typical behavior. • Upon entering the room, she noticed the SLP sitting on the stool she usually occupied. Instead of greeting the members of the group, she shouted, “Hey, that’s my chair.” • When another child was telling a story, she interrupted, by speaking more loudly than the other child, to begin her own story.
Using the form–content–use taxonomy, indicate the nature of Elizabeth’s language disorder.

Exercise 7

Language Disorders
We looked at three children who were diagnosed with language disorders. Now we will consider a formal definition. One definition, and there are many, captures the pervasive and persistent nature of language disorders. Bashir states:
Language disorders is a term that represents a heterogeneous group of either developmental or acquired disabilities principally characterized by deficits in comprehension, production, and/or use of language. Language disorders are chronic and may persist across the lifetime of the individual. The symptoms, manifestations, effects, and severity of the problems change over time. The changes occur as a consequence of the context, content, and learning tasks. (Bashir, 1989)

We will take the definition apart and examine examples to illustrate Bashir’s important observation of the nature of language disorders and to understand some of the language disorders that can occur. The first claim is that language disorders are a heterogeneous group. Well, we examined above how language may deviate from our typical expectations in form, content, and/or use (FCU). But that analysis did not address the causes of the deviations. If a child has been diagnosed with Autism Spectrum Disorders (ASD), a developmental

Communication Disorders


disability that manifests in atypical development of social skills, communication, and may include ritualistic behavior, he or she may manifest atypical development in any one or all of the elements of FCU. But what if a child is diagnosed with Down syndrome? We still anticipate that atypical language will develop and can affect FCU, albeit in predictably different ways from children with ASD. Yet, we classify both children as manifesting a language disorder. Both Perry and Elizabeth were diagnosed with ASD. However, Billy presents with yet another causative category of language disorder. Children with language deficits like Billy’s are often diagnosed as having a Specific Language Impairment (SLI). When a child experiences difficulty in acquiring language and continues to have difficulty in understanding spoken language, in expressing spoken language, in reading and written language expression, but experiences no other deficits (i.e. motor, sensory, emotion, or cognitive), he or she is diagnosed with Specific Language Impairment. In contrast to a language disorder stemming from a known cause, such as Down syndrome (DS), which is associated with a chromosomal abnormality, SLI has no known cause. SLPs diagnose a child with SLI when they observe no other deficits beyond the deficits in language. Referring back to Billy, evaluation results indicated that Billy’s development was typical and intact in the sensory, motor, cognitive, and emotional domains. Again, contrasting SLI characteristics with DS characteristics, it is generally found that children with DS experience delayed cognitive and motor development, while children with SLI experience delayed language development but no other observable delays. It is important to note here that the identification of a language disorder due to any cause does not prevent individuals from leading successful lives. It is, in fact, due to understanding the different manifestations of this heterogeneous group known as language disorders that intervention can begin early and be structured to fit the child’s needs. Secondly, Bashir (1989) notes that a language disorder may be developmental or acquired. The examples above are categorized as developmental language disorders, which means that the language disorder manifests during the developmental period. In contrast, acquired language disorders are those that arise from a disease or injury and indicate that the disorder was not present at birth. An example of an acquired language disorder is one in which problems with language production or comprehension result from a traumatic brain injury (TBI). TBIs occur in children and adults. The National Institute for Neurological Disorders and Stroke (NINDS) reports that “half of all TBIs are due to transportation accidents involving automobiles, motorcycles, bicycles, and pedestrians” (NINDS, 2009). Other causes include falls, violence, and sports injuries. Aphasia is also an acquired language disorder. Aphasia is defined as a deficit in the comprehension or production of language that arises from brain damage. Stroke is the most common cause of aphasia, but aphasia can also be caused by tumors or infections that damage the language centers of the brain. In addition, aphasia may arise as the result of a TBI. So, you may ask, why aren’t the language deficits caused by TBI all considered aphasia? As indicated, some are, but some of the communication deficits that result from a TBI have to do “with the more subtle aspects of communication, such as body language and emotional, non-verbal signals” (NINDS, 2009). The deficits that stem from TBI go beyond the language domain. They include the loss of cognitive functions such as memory, reasoning, and thinking. Sensory processing can also be affected. Cognitive and sensory deficits can contribute to difficulties that individuals who have suffered a TBI have with using language. On the other hand, the primary deficit in aphasia is the loss of the ability to produce or understand language. Not only are speech production and



comprehension affected in aphasia, but reading and writing are commonly affected as well. Other cognitive processes such as memory, thinking, etc. are not affected. Bashir (1989), in his definition, next addresses the types of deficits that can be present: “comprehension, production, and/or use.” Using the categories of language disorders that we have discussed thus far, we can examine the types of language deficits that occur within these disorders. If we examine aphasia as an example, we see that it has long been observed that, depending on where in the brain the damage occurs, a predominant type of language loss will result. In the middle to late 1800s two neurosurgeons, Paul Broca and Karl Wernicke, working independently, helped establish an understanding of brain function. Building on their early observations, we have learned that, for most people, the left hemisphere of the brain is specialized to perform language functions. An area of the left hemisphere in the frontal lobe of the brain is called Broca’s area. When a person’s brain is damaged through stroke, trauma, or other reasons as described above, and the damage is in the region of Broca’s area, the person will most likely have difficulty expressing language, or will, in Bashir’s (1989) words, have a problem with language production. On the other hand, if the brain damage occurs further back in the brain, in the temporal lobe in the area called Wernicke’s area, then, most likely, a predominant problem with language will be in understanding language. Figure 14.1 shows these two cortical areas in a drawing of the human brain. Since the time of Broca’s and Wernicke’s observations about these brain sites, our knowledge about brain functions and our knowledge of the results that damage to the brain can have on language have transcended this mere dichotomy of function. Current research in neuroscience, which has been aided by advances in technology that allow scientists and clinicians to view images of the brain during the performance of various tasks (e.g. listening to words, recalling words), indicates that Broca’s and Wernicke’s areas appear to be important for language production and comprehension; however, these two areas cannot entirely account for the complexity of language. Neuroscientists think that language is widely located within the brain and that these areas cross-communicate with each other. Consider the following:
Obviously there are visual and manual components to language, for reading and writing. Where does sign language fit in? How do you explain a patient whose only deficit is an


Traditional cortical language areas, inferred from the results of strokes

Communication Disorders inability to name tools? He can describe the use of a hammer but not its name. How does a collection of syllables—a person’s name—trigger the face, personality, birthdate, or voice of that person in your memory? (Molavi, 1997)


Examining the predominant difference between Broca’s aphasia and Wernicke’s aphasia helps us to understand that language disorders can be differentiated not only by cause but also by the types of deficits a person may have. Reviewing the case studies, we see that Perry and Elizabeth had difficulty in the appropriate use of language, while Billy had difficulty in the production of language. The distinction here is that Billy was able to formulate his language in a way that was appropriate to the context and the listener, but his syntax and morphology deficits caused his language productions to be ungrammatical. For example, when describing a picture of a father helping his son to put on a Halloween costume, Billy said, “A man dress a boy.” Contrast this to Elizabeth’s description of a picture of her father reading to her and her siblings. She said, “Father is reading book about ghosts to his children.” Both children were 6 years old at the time. Billy and Elizabeth were fully able to comprehend what was said to them. Elizabeth, in contrast to Billy, did not use language in a way that was appropriate to the context. Recall that when Elizabeth entered the room, she did not greet the others. Instead, she shouted, “Hey, that’s my chair.” If we look at yet another example of a 6-year-old, whose language difficulties were in comprehending what was said, a different pattern emerges. Jason’s most frequent response to a question was: “What? What did you say?” Jason’s hearing was fully intact. His difficulty was in understanding the meaning of language. In addition to asking for a repetition of what was said to him, Jason attempted responses during conversations that often indicated that he had not understood. For example, in response to the question, “Where’s Al?” Jason responded that Al was fine. This exchange occurred when Jason was 15 years old. This exchange takes us to the last point in our definition in our attempt to understand language disorders. To illustrate the chronic, persistent, yet changing manifestations of a language disorder, we can track Jason’s development and growth.

Case Study IV
When I first met Jason, he was a charming, handsome 3-year-old whose language development was delayed. At 3, when the expectation is that children are using mostly fully formed sentences to express themselves, Jason was using two or three words to convey his thoughts. Jason was diagnosed with SLI. By the time Jason was 5, he entered kindergarten and his expressive language abilities had improved. He used full sentences but still struggled to find the words to convey his meanings. Sometimes, his word-finding difficulties were so great that Jason gave up trying to express himself and said, “Oh, never mind.” In school, his behavior reflected his frustration with his impaired expression and comprehension. He often misunderstood directions, and, as indicated above, often responded incorrectly to questions. Jason was a highly intelligent child, and his frustration with his limited language capabilities caused him to act out. Despite the fact that Jason received language therapy and was making progress, he did not function well in his fast paced, large public school classroom. Through long hours of advocacy for her child, Jason’s mother was able to move him to an inclusive class setting in his school. An inclusive setting is one in which children with varying types of impairments are in regular education classes with typically developing peers. The structure of the support system varies depending on the school. In



Jason’s case, he was assigned an aide who worked with him during class time. While sitting alongside of Jason, the aide was able to take the time to explain instructions to him so that he could complete his assignments. In addition, he received language therapy while the other children had recess or free time. While these were supportive structures, which Jason needed, his poor communication skills and his lack of time for unstructured interactions with his classmates created social problems for Jason. Thus, he made no friends. While Jason’s oral language abilities continued to grow, he began to lag behind in reading and writing acquisition. This illustrates the notion that language disorders change in response to the learning tasks. While Jason’s expressive language skills were not markedly different from his peers’, his language impairment was expressed in his difficulty learning to read. In addition to seeing the speech–language pathologist at school, he began to see a reading specialist. His mounting frustrations caused him to withdraw even further socially. He refused to do his work in school; and he would not complete his homework. In third grade, Jason’s mother transferred him to a private school with small classes. By this time, Jason was well behind in reading and writing. His classroom performance was poor despite the small classes. Jason no longer had his classroom aide and the accelerated curriculum assumed reading skills that Jason did not have. Again, he was assigned to sessions with a learning specialist. Because he could not follow the lessons, he was frequently disruptive in the classroom. In addition, he refused to cooperate with the learning specialist in his private sessions. Jason was asked to leave the school. At this time, Jason was diagnosed with a language-learning disorder (LLD), difficulties with age-appropriate reading, spelling, and/or writing skills. He returned to his public school, where he was placed in a small class for children with special needs.
Exercise 8 Reread Jason’s case study. Using the information presented, write out examples of the chronic and changing nature of Jason’s language disorder. Knowing that Jason was diagnosed initially with SLI, which means that he had no cognitive, motor, emotional, or sensory deficits, how did Jason’s language disorder begin to pervade his overall functioning?

Treating Communication Disorders
Does your college or university have a speech, language, and hearing clinic? Observing treatment given by speech–language pathologists, who are licensed by their states and certified by ASHA, is the best way to learn about treatment. If your institution doesn’t have a clinic, you may be able to arrange a visit to observe the speech–language pathologist in the local schools or in your hometown school.

Treatment Principles
One of the most important principles in treatment is the establishment of rapport between the clinician and the client. Rapport can be defined in many ways. In the therapeutic relationship, you should be able to observe mutual respect between the client and the clinician. You should be able to observe the clinician’s regard for the client. Does the clinician appear prepared? Is he or she sensitive to the client’s changing needs within the session, i.e. does he or she seem flexible? Does the clinician take time to explain the purpose of the session and the results once the session is completed? Is the clinician patient?

Communication Disorders


There are principles and phases in treatment of speech and language disorders. I will use the treatment of a speech disorder to illustrate how therapy works. The first phase in treatment, which relies on an evaluation that pinpoints the areas where deviance from expected behavior occurs, is the establishment of the behavior. For example, if a child has trouble producing speech sounds (an issue of unintelligibility, i.e. cannot be understood because of many sound production errors), the information from the evaluation will indicate which sounds are a problem. Also, the information from the assessment will indicate which sounds are used and how they are used. The SLP will use this information to target the sounds to establish. There are many considerations that govern the difficult decision of “where to begin.” Among the factors that will direct the clinician are questions such as “What are the easiest sounds for the individual to establish?” This allows the most rapid success and encourages the client to continue on in the process. Another question that might guide the decision of what sound/sounds to establish first is what sounds will increase my client’s ability to be understood. Often multiple sounds are introduced at the same time in treatment, particularly if there are many sound production errors that are impeding communication. Age-norms have also played a role in governing a clinician’s decision where to begin. So if a 5 year old is not producing sounds expected for his/her age, the clinician may choose these sounds to target for establishment. Once the client learns how to produce the sounds, the goal shifts to stabilizing the use of the targets. The final phase of treatment is the generalization phase. The client has learned the sound patterns, can produce them consistently in the learned contexts, and now needs to use the new sound patterns all of the time in all settings (home, school, playground, etc.). So let’s apply this to an example. I will use a simple example (but like all the others presented in the chapter, this child was a client of mine) so that you can easily work through the analysis. It is important to note that most children and adults who are treated by speech–language pathologists have far more complex speech and language problems than this example indicates. Another important aspect to note is the impact even this mild speech problem presents for the child. There is no communication disorder that is trivial to the person experiencing it.

Case Study V
Jamie was a 9-year-old fourth grader who had transferred to a new school. He transferred there because it was discovered that he was highly intelligent and was not sufficiently challenged by the curriculum in his former school. His teacher contacted me because Jamie had difficulties with producing his speech sounds (phonemes). Here are some examples of how Jamie pronounced his words. The words he attempted to say are presented first and his production of the words is given in parentheses.
Robot (wobot), rich (wich), merry (mewe), brown, (bwown), curb (cawb), car (caw)

Which sound was Jamie having difficulty with? If you said the “r” sound, you have correctly analyzed his difficulty. Now, let’s plan his remediation program. First, the correct production of the sound must be established. Jamie is having difficulty with the “r” sound in every position in the word (beginning, middle, and final). Did you notice that? The SLP would analyze how he was using his articulators (lips, tongue, teeth, palate, and lower jaw) to produce this sound. I did this and I noticed that Jamie’s tongue was lying too low in his mouth and that he was rounding his lips when he said the “r”. Thus Jamie is substituting the



articulation of one sound “w” for another sound “r” because he is not using his lips and tongue correctly. Try contrasting the way you use your lips and tongue to make the two sounds and notice the different positions that the tongue and lips are in when you are doing this. I had Jamie do this to establish the different use of his tongue and lips to produce these two sounds. I also had him listen to the contrast between the two sounds so that Jamie could discriminate between them. Learning to hear and feel the differences helped Jamie to establish the correct way to form the “r” sound and to distinguish it from the “w” sound. The establishment phase includes establishing the correct production of the sound all by itself (isolation) and then in words in each position (beginning, middle, and end). After Jamie could do this, his therapy concentrated on using a variety of words with the “r” sound in all positions in phrases and sentences, e.g. “The robot is big.” Each aspect of treatment increased the complexity of the task to challenge Jamie to retain his learning and to help maintain and stabilize his production of the “r” sound. All throughout his treatment, I consulted with his parents and his teachers. I gave Jamie and his parents exercises to do at home. I provided his teachers with ways to monitor his use of his new sound in the classroom and to help him use the new sound. This collaboration sets the stage for generalizing his use of the sound to all of the settings that Jamie engaged in. An extremely important issue in Jamie’s case was that what was a minor speech production problem had escalated into an emotional and social problem for Jamie. When his teacher first contacted me regarding Jamie’s sound production problem, her main concern was the impact it was having on his well-being and success as a person. Jamie was acting out in response to his classmates’ teasing him about his speech. He was spending class periods hiding in the closet; he refused to speak in class. He was not making friends, which is essential for children who are transferring to a new environment. Despite his intellectual capabilities, Jamie was failing in school because of his speech problem. I consulted with the school psychologist to keep her apprised of Jamie’s progress in speech therapy, and we both worked with Jamie to help him identify his communication goals and to accomplish them. Jamie’s case is particularly important to consider because the importance of speech, language, and hearing is, as we saw in Kohnert’s (2007) quotation at the beginning of the chapter, “to allow for an efficient, effective means to develop and exchange thoughts, feelings, ideas, and information.” Even when a speech or language problem seems small, as in Jamie’s case, the impact can be huge. Jamie’s case illustrates a number of general principles that apply to the treatment of any communication disorder. Treatment is based on a thorough analysis of the client’s communication strengths and weakness. Therapy (treatment) builds on the client’s strengths in order to develop new skills. Once a new skill is established, the client and the clinician work to increase the use of the new skill until the behavior is stabilized. The goal is always to improve the client’s ability to communicate effectively his or her thoughts and feelings and understand those of others. Throughout this process, collaboration with the client and with those in his or her environment to provide support is essential. Treatment, then, is important so that individuals with speech, language, and hearing difficulties can find a way to communicate with their fellow human beings because we all want to fit in with our families, our social groups, our schoolmates, and our work colleagues. And the ability to communicate through speech and language is essential to achieving this primary human goal.

Communication Disorders


Communication Disorders and Adults
While I have addressed the speech and language disorders of children in this chapter to help you develop an understanding of this field of study, communication disorders can strike individuals of any age, from the beginning to the end of life. We considered this briefly when we looked at the classification of communication disorders. Some of the materials to which I referred (e.g. the video Transcending Stuttering, 2004) focus on adults as well as children. There have been films made about adults and memoirs written by adults about the impact communication disorders have had on their lives. I have listed some of these profoundly affecting works in the Other Resources section of this chapter. Other chapters in this book address adult issues as well. Perhaps the big difference between developmental and acquired disorders is that an acquired disorder alters the life of the individual in a way that was not predicted. Let’s consider one last case.

Case Study VI
I met Molly when she enrolled as a student in the college where I teach. At the time, I was the director of a support program for students with learning disabilities, and Molly’s rehabilitation team contacted me because they thought she was ready to resume her college career. Here’s Molly’s story. At the age of 19, Molly was enrolled in a very large, highly competitive university and was both a scholar (on the way to a Ph.D. in philosophy) and an athlete. Molly was in a car accident and suffered a traumatic brain injury that left her unable to run or to complete her education. After several years of rehabilitation, Molly made enough progress to suggest that she might be ready to complete her education, slowly and in a different environment. I teach at a small college, and because we had a strong support program, this seemed like a good choice for Molly’s return to her studies. The end of the story is that Molly did complete her education, though in an entirely different field of study. Molly eventually was awarded a master’s degree in linguistics. She is married and lives a happy life in a small town where she works. The middle of the story is that she was detoured from her original path, where she had been a high achiever. She had to refashion her goals and rebuild her life, one very small step at a time; one class a semester at first. Her first foray into graduate school was not a successful choice. But Molly’s determination to find her place finally allowed her to triumph. These are the adult stories. Careers and relationships are established; and then are altered drastically by events that cannot be predicted or controlled. Yet, through treatment new paths are forged and lives go on differently but fulfillingly. This is the promise that the professionals in speech–language pathology make to all people with communication disorders.

Language Differences
Not all variations in language and speech are disorders. SLPs are charged with the task of determining who is eligible for services, that is, who needs the assistance of a speech and language professional. SLPs do not regard accents and dialects as deviations to be treated. However, there are individuals who are interested in “modifying” their accents or in changing their dialects for a variety of reasons. There are SLPs who specialize in assisting individuals who wish to do this because their careers or their life situations would be enhanced by making these changes. The most obvious group who want to



change the way they speak are actors, whose careers demand that they speak in a particular way or use a particular accent or dialect. Others are workers who are not native speakers of a language and want to alter their speech in order to advance their careers. While speech–language professionals can and do provide these services, it is most important to note that it is because an individual chooses to seek these services and not because speech and language clinicians regard any accent or dialect as inferior or as a communication disorder. ASHA issued a statement to guide professionals in making ethical decisions about serving clients. The position statement, Social Dialects (ASHA, 1983) can be accessed at

The last thoughts I would like to leave you with are these. Many people are compelled to engage in the helping professions, of which speech–language pathology and audiology are a part, because they want to be of service. What one fails to see at the outset of this pursuit is the benefit that accrues to oneself. Here’s what I mean. I have learned more about communicating effectively, about interpersonal relationships, about courage, about determination, about almost anything you can think of, from my clients. I appreciate the beauty of expressing ideas and feelings through language and speech in all of their variations, whether they are natural variations or variations imposed by a “communication disorder.” I have experienced the joy of others’ successes and the deflation of defeat. I have long-standing relationships with some families I met throughout my thirty plus years of practice and memories of others that I no longer have contact with. Every interaction I have had with my “clients” has shaped my view of life. I didn’t know it then, but I do know it now; that’s why I became a speech–language pathologist. Yes, we are agents of change, but, for the most part, the change is in ourselves. One final statement: communication requires both a sender and a receiver. It is a reciprocal relationship. If communication fails, the responsibility falls on each person engaged in the act.

American Speech–Language–Hearing Association (1983) Social dialects [Position Statement, Internet]. Available from: [Accessed June 18, 2009]. American Speech–Language–Hearing Association (1993) Definitions of communication disorders and variations [Relevant Paper, Internet]. Available from: docs/html/RP1993–00208.html [Accessed June 18, 2009]. American Speech–Language–Hearing Association (2009a) American Speech–Language– Hearing Association (ASHA) [Internet]. Available from: [Accessed June 18, 2009]. American Speech–Language–Hearing Association (2009b) Dean Garstecki: Statement [Internet]. Available from: [Accessed June 18, 2009]. American Speech–Language–Hearing Association (2009c) Hearing loss [Internet]. Available from: [Accessed June 18, 2009].

Communication Disorders


Bashir, A. S. (1989) Language intervention and the curriculum. Seminars in Speech and Language, 10(3), pp. 181–191. Bloom, L. and Lahey, M. (1978) Language development and language disorders. New York, Wiley. Groce, N. E. (1988) Everyone here spoke sign language: Hereditary deafness on Martha’s Vineyard. Cambridge, Harvard University Press. International Stuttering Awareness Day Online Conference (2005) Transcending stuttering: The inside story [Internet]. Available from: pws8/schneider8.html [Accessed June 18, 2009]. Kohnert, K. (2007) Language disorders in bilingual children and adults. San Diego, Plural Publishing. Mayo Clinic. (2007) Hearing loss: Definition [Internet]. Available from: http:// [Accessed June 18, 2009]. Molavi, D. W. (1997) Neuroscience tutorial [Internet]. Available from: [Accessed June 18, 2009]. National Institute of Neurological Disorders and Stroke (2009) Traumatic brain injury: Hope through research [Internet]. Available from: detail_tbi.htm#127653218 [Accessed June 18, 2009]. National Institute on Deafness and Other Communication Disorders (2008) Have wise ears for life! [Internet]. Available from: wiseears.asp [Accessed June 18, 2009]. Our Time Theatre Company (2009) Our Time: An artistic home for people who stutter [Internet]. Available from: [Accessed June 18, 2009]. The Stuttering Foundation (2009) The stuttering foundation [Internet]. Available from: [Accessed June 18, 2009]. Transcending stuttering: The inside story (2004) Dir. Phillip Schneider. New York, National Stuttering Association [video: DVD]. Available from: play?docid=3654482395702325018 [Accessed June 18, 2009]. Van Riper, C. and Erickson, R. L. (1996) Speech correction: An introduction to speech pathology and audiology. 9th edn. Boston, Allyn and Bacon. Voice Foundation. (2009) Voice Foundation [Internet]. Available from: http://www.voice [Accessed June 18, 2009].

Other Resources
This chapter focuses on the stories of individuals. I would like to refer you for further reading to the memoirs and films that tell individuals’ stories. There are many, but here are a few.

Bauby, J. (1997) The diving bell and the butterfly: A memoir of life in death. New York, Vintage International. DeBaggio, T. (2002) Losing my mind: An intimate look at life with Alzheimer’s. New York, Free Press. Douglas, K. (2002) My stroke of luck. New York, HarperCollins. Grandin, T. and Scariano, M. M. (1996) Emergence: Labeled autistic. 2nd edn. New York, Warner Books. Robison, J. E. (2007) Look me in the eye: My life with Asperger’s. New York, Crown Publishers.



Children of a lesser god. (1986) Dir. Randa Haines. Hollywood, Paramount Pictures [video: VHS]. The diving bell and the butterfly. (2007) Dir. Jean-Jacques Beineix. New York, Cargo Films [video: DVD]. My left foot: The story of Christy Brown. (1989) Dir. Jim Sheridan. Dublin, Ferndale Films [video: DVD]. Sound and fury. (2000) Dir. Josh Aronson. New York, Aronson Film Associates [video: VHS].

Cecile L. Stein
This Chapter Explores:

Analyzing Narratives An Example of Cross-Cultural Research Methods

Personal Narratives Macrostructure Narratives and Child Development Perspective Taking and Frame of Mind Fictional Narratives Story Grammar

Analyzing Content Perspective Taking Story Resolution A Cross-Cultural Application

We live in discourse as fish live in water.
Lodge, 2008

Story telling is a primary act of mind.
Hardy, 1977

As a speech-language pathologist and teacher in the area of language assessment and language disorders, I became increasingly interested in the importance of the narrative in our lives. I saw typically developing language users express their thoughts as they entered different stages of their language development, progressing from single words to sentences, and finally to narratives where they talked about a funny incident, a party, a problem, and created stories. Here was the narrative as a window into the organization and strength of the child’s language system and mind. However, I primarily saw children who were having difficulty achieving this important narrative skill, and needing treatment to obtain it. Then I had the exciting experience of researching narratives in a culture quite different from the mainstream American culture in which I had lived all my life. I compared the narratives of rural American children with those of Bhutanese children. This was an opportunity to analyze the commonalities and differences in children’s storytelling development. What was similar and what was different? Did our Western constructs of storytelling hold up in a culture that only recently had opened itself up to Western civilization? Collecting and analyzing narratives has become for me a fascinating means by which to view the child’s language development, the values and concepts of his or her culture, and, consequently, the child’s mind.



What are narratives? A narrative is simply an oral or written discourse that describes events, sequenced in time and conveying meaning. This chapter focuses on two genres, or types, of narratives: personal narratives, which relate real past experiences, and fictional narratives, which consist of invented stories that are not about real events. Researchers in Western cultures have extensively explored narratives in children with and without language impairment, and have described certain common components (parts) at the core of personal and fictional narratives. Their work has led to a variety of tests used in areas of education, communication sciences and disorders, and neuropsychology. Professionals in these and other fields continue to rely on analyses of personal and fictional narratives as a valuable method for assessing language development and treating language disorders. First, I will review some of the key components currently used to analyze personal and fictional narratives. At the same time, I will discuss methodological issues, exploring and explaining contradictory findings in narrative research. These discussions will appear in boxed text throughout the chapter. Second, I will show how an analysis of the narratives of children from a non-Western culture reveals commonalities with and differences from their Western counterparts.

Personal Narrative Components
One method of analyzing narratives looks at their macrostructure, or general organization properties. Labov and Waletzky (1967), referred to as L&W from here on, identified key macrostructure components of personal narratives, the narratives of “every-day life.” These components are listed in Table 15.1. The structure begins with the “Abstract” (what the point of the narrative will be). The speaker “Orients” the listener to the who, where, and when of the story, and continues to the “Complicating Actions” that lead up to the story’s “high point” (climax). “Evaluation” is the emotional information that a speaker conveys, giving us a glimpse into his or her attitudes about the narrative.
TABLE 15.1: Components of Personal Narratives Abstract: a brief summary informing the listener what the main point of the narrative will be. Orientation: information about who participated in the event, where and when the event occurred. Complicating Action: chronological presentation of the events that lead up to the climax or high point of the narrative. Evaluation: emotional information showing the speaker’s attitude towards the narrative. Resolution: this element occurs after the evaluation’s high point and resolves the high point action in some way. This prepares the narration for its ending. Coda: this is often thought of as a formalized ending (“that’s it” or “the end”) and returns the listener to the present.
Source: Adapted from Labov and Waletzky, 1967.

Analyzing Narratives


The “Resolution” then resolves the high point action, and the narrative ends in some type of formalized manner (“Coda”), such as “The End.” Examples of Labov and Waletzky’s personal narrative components follow:
1 Abstract Example: Well my picnic didn’t turn out so great. 2 Orientation Example: Me and my mom and dad went on a picnic last Sunday. 3 Complicating Action Example: We carried our stuff over to the picnic table and we took out all the sandwiches and sodas and then it started to rain real hard. 4 Evaluation Example: And then it started to rain real hard. (We will discuss the Evaluation component in greater detail in the section on childhood development, p. 300.) 5 Resolution Example: It didn’t look like it was going to stop raining so we decided to go back home. 6 Coda Example: I hope we have better luck next time.

Exercise 1

Personal narrative components Read the following personal narrative (written by a 13-year-old boy, from my files). Identify each component of the narrative by consulting Table 15.1. I was about eleven years old at the time when I saw a real hero. I was walking with my friends to get an ice cream in town when I saw a beautiful dog, maybe a Labrador Retriever, jump out of an open van. The traffic stopped short, but the van kept on going. The owner of one of the shops happened to see this and ran out in the street to get the dog before it got hit by the oncoming traffic. He was lucky that he himself didn’t get hit. Kids on the street hurried over and hugged the dog. He was so beautiful. We were all amazed at the risk the store owner took when he ran into the street—in the middle of all that traffic. He was my hero that day. So, we all waited for the driver of the van to come back, but that never happened. Finally, the owner called the police to come and get the dog. I sure wished I could have taken it home, but my mom probably would have said no.

At this point, let’s take our first look at important methodological issues.

Issues of Methodology
Research methodology may itself affect the narrative. For example, L&W (1967) asked their speakers to respond to the following question: “Were you ever in a situation where you thought you were in serious danger of getting killed?” (L&W, 1967, p. 14). This type of “dramatic” question triggered a telling of a temporal sequence of events, leading to an evaluation containing a high point. Schegloff (1997) asks us to look at the differences in the way narratives are told in reference to the question or topic posed, the presence of the listener, whether the


Stein listener is the experimenter or a co-narrator, the narrator’s experience with the task, and other aspects of the experiment’s design. With this caution in mind, consider if the question had been: “Were you ever in a situation where you met an old friend while on vacation?” Do you think the speaker would necessarily provide the same type of dramatic high point or even a high point at all? Another question to ask is: How does the speaker’s experience with the topic affect the narrative? Bamberg and Damrad-Frye (1991) speculate that first-person narratives (that is, first-hand experience with the topic) may have accounted for the appearance of the Evaluation around the high point in 4-year-old children’s narratives in Peterson and McCabe’s study (1983). (We will discuss Evaluation in the next section.) Yet another question relates to the strength of the child’s language system. While language impaired children may have difficulty telling narratives (McCabe et al., 2008), the difficulties may vary. Some children may omit various macrostructure and microstructure components. Others may demonstrate weak story organization.

Narratives and Child Development
Researchers applying L&W’s macrostructure components to the analysis of children’s narratives have found that children begin their narrative structural development between 2 and 3 years of age (Hudson and Shapiro, 1991). They then take years to master all the components of the personal narrative. The various stages of development are particularly marked by a child’s use of the Evaluation (emotional meaning) component. Evaluations become increasingly common as the child grows. While children as young as 2 years begin to use evaluations, Peterson and McCabe (1983) found that half of all comments from children between ages 4 and 9 were some form of evaluation. In addition, Evaluations can occur anywhere in a narrative that the speaker expresses emotional information. For example:
In an Orientation: I was so excited. In the Complicating Action: And then it started to rain real hard.

By the age of 5, children can typically produce a narrative in conversation that contains the emotional information or “attitude of the narrator towards the narrative” as a whole. This is called a classic “high point” narrative (Hudson and Shapiro, 1991: 99). The high point is a crucial juncture in the narrative that reveals the essential emotional meaning of the event. Often the speaker may pause at the appropriate moment “to suspend the action, create tension and suspense, convey the emotional significance of the story” (Peterson and McCabe, 1991: 42). McCabe (1996) states that a high point evaluation is distinguished from other evaluations in that it is “concentrated.”
Example: We were so soaking wet and our shoes got all squishy and I thought we’d never get dry. Example: I was still at school . . . and he went in and tore down all my pictures that I painted and he tore them up. And he broke one of my best, my very best doll . . . (McCabe, 1996)

Analyzing Narratives


Perspective Taking and Frame of Mind
Perspective taking, the ability to see another person’s point of view, attitude, and emotions, is a skill that continues to grow throughout childhood and into adulthood. In narratives, speakers can express their own perspectives on a situation or emotional state, or their impression of someone else’s emotional state or frame of mind. Bamberg and Damrad-Frye (1991) found that 5-year-old children could express frame of mind to “local” events, that is, to particular incidents within the narrative. An example would be: “and the tiger felt pity on him,” in a story by a 7-year-old, where a boy encounters a tiger in the woods (from my files). By age 9, references to frame of mind, particularly to emotions, become a “prevalent evaluative device” (Bamberg and Damrad-Frye, 1991). As the child grows (and in adulthood), he/she demonstrates greater perspective and knowledge of how the mind works. For example, in the final line of the “dog story” in Exercise 1, the speaker says: “I sure wished I could have taken it home, but my mom probably would have said no.” The narrator demonstrates perspective taking, inferring how his mother would react. Researchers disagree on when certain features become established. Berman and Slobin (1994) found that temporal sequencing (ordering events by time) and causal connections were not present in the narratives of 3-year-olds; by age 9, these features appeared established. In contrast, Ochs and Capps (2001) found temporal sequencing and causal connections in the narratives of 3-year-olds.

Issues of Methodology
When research articles find differences in outcome, it is important to look at the methodology: what were the subjects in the study asked to do? In the Berman and Slobin study, children were asked to tell a story based on the wordless picture book Frog, Where Are You? Ochs and Capps (2001), on the other hand, asked for personal narratives. It is possible that the 3-year-olds were unable to relate to the storyline because they had no experience with the story event (searching for a runaway frog). Consequently, they told their stories by simply stating the actions as represented on each page, without connecting the previous pages with upcoming ones. These were isolated event descriptions (Berman, 1995). In reference to Ochs and Capps’ study, telling a personal narrative is telling an event that already occurred, where the child speaker knows the sequence of events and possibly the causal connections. However, further research suggests an alternative view of the wordless picture book task. Bamberg (cited in McCabe, Bliss, Barra and Bennett, 2008) suggests that telling the story events depicted in the wordless picture book is “rooted” in social interaction and conversation, rather than being simply a cognitive, mental activity task.

Fictional Narrative Components
Story Grammar
During your own narrative beginnings, you probably recall creating fictional stories and at some point realized that you needed a beginning, middle, and an end. Gradually, you learned other components to enhance your storytelling. By age 6, children usually



internalize a macrostructure containing the components valued by their culture (Peterson and McCabe, 1983). These macrostructure components are called story grammar, and are used to analyze fictional stories created by children without the aid of a model story. A picture or sequence of pictures may or may not be provided as a prompt. Stein and Glenn (1975) established the model of story grammar components shown in Table 15.2. This model is based on the premise that fictional stories contain a problem or an element that initiates the story’s action and establishes the character’s goal—in other words, that fictional stories are goal oriented. The structure begins with a “Setting” (the who, where, and when of the story). The storyteller follows with the “Initiating Event” (the problem or event that starts the story’s action). The story’s character follows with an “Attempt” to solve the problem and the “Consequence” of the Attempt (successful or not). If unsuccessful, the character may initiate another attempt, followed by another consequence. Finally, the story leads to a “Resolution,” where no further attempts are made and the story begins to wind down. The story may have an “Ending” with a short phrase, or simply “The End.” For an episode to be complete, it must contain at least an Initiating Event, an Attempt, and a Consequence. Two components in story grammar are optional: the “Internal Response” and the “Internal Plan.” In the Internal Response, the storyteller states feelings, attitudes, thoughts, and decisions in reaction to the initiating event and to the resolution. These statements reveal the character’s frame of mind, motivations to act, and insight into the problem and solution. In the Internal Plan, the narrator shows thoughts or decisions that might be made to solve the problem, but does not include the Attempt (or action). Children pass through various stages in their acquisition of story grammar. These phases initially omit various components of story grammar; the initial phases may be only setting description or sequences of actions. Once the three elements of the Initiating Event, Attempt, and Consequence are present, the child has produced a complete “episode” (Hughes et al., 1997).1 By the fifth grade, children reportedly are able to produce a story containing complete story grammar, with a logical and coherent plot.
TABLE 15.2: Story Grammar Components Setting (S): includes the who, when, and where of the story, and introduces the main character. Initiating Event or Problem (IE): sets the story in motion and usually establishes the goal for the story (resolving the problem). *Internal Response (IR): an emotional statement that reflects how a character feels in response to the initiating event. *Internal Plan (IP): a statement of an idea that might fix the problem. Attempt (A): some action or series of actions taken by the main character in an effort to solve the problem. Consequence (C): the event or series of events following the Attempt and causally linked to it, whether successful or not. Resolution (R): the final state or situation triggered by the initiating event; it does not cause or lead to other actions or states. End (E): a sentence or phrase that clearly states the story is over.
Source: Adapted from Stein and Glenn (1975). Note: * The Internal Response and Internal Plan are optional.

Analyzing Narratives


The two optional components (Internal Response and Internal Plan) more commonly appear in the narratives of 7-year-olds than in the narratives of younger children. In the example that follows, you will see the story grammar components identified at the end of each numbered line. The story begins with a Setting and immediately presents the Initiating Event. You will then see a series of Attempts followed immediately by a Consequence, until the story reaches its Resolution and End. (Note that oral narratives are commonly segmented into lines of Communication Units. A Communication Unit comprises an independent clause and its modifiers (Hughes et al., 1997). 1 2 3 4 5 6 7 8 9 10 11 Once me and my dad went into the woods (S) and we didn’t know which way was back (IE) And then we took one way (A) and that way was covered by a tree (C) And we went over the tree (A) and we saw a clearing—a beautiful site (C) and you could see everything (C) you could see the school (C) And we went back up (A) and then I thought it’s the other way (IP) and then we went up (R) and went home (E) (Stein, 2004)
Story grammar components Try analyzing the following narrative according to story grammar components. Keep in mind that an utterance might contain more than one story grammar component. For example, in line 11 of the above narrative (“and then we went up (R) and went home (E)), both the Resolution and the End appear in the same sentence line. Remember, too, for a narrative to have complete story grammar, there must be a goal. You’ll find the goal represented by the Initiating Event or the Attempt to solve the problem. The story below is transcribed as spoken by a 7-year-old child, without changing word use, grammatical errors, or instances of lack of clarity: 1 2 3 4 5 6 7 8 9 10 11 12 13 One day there was a boy and he was walking along and he saw a woods He walked inside and he couldn’t find a way out He tried asking the animals but then he just remembered that they can’t talk And then he kept on trying He climbed up a few trees to see if his head could reach out of the wood But they were too small And he kept on walking and saw some light and followed it And he was keep [sic] on walking and he was out of the woods And went home and lived happily ever after The end (Stein, 2004)

Exercise 2

(Italics + [sic] indicates the narrator’s wording, and not a typographical error.)



Issues of Methodology
Pictures, especially those in a wordless picture book, present an intact story with the elements of story grammar already established. The child needs to be attentive to the pictures and integrate the details. Presenting the same pictures to many subjects gives us a way to compare an individual child’s narrative progress with his or her age group. In contrast, a self-generated story requires the narrator to display his or her knowledge of story grammar without picture prompts. By using a story grammar scoring system, self-generated stories can be coded for the presence of the macrostructure. Do you think stories would be more complete (contain more story grammar components) and contain more information when children are asked to create an original story without picture prompts than with these prompts?2

Exercise 3

Collecting a fictional narrative Collect a fictional narrative from a child. Ask something like: Tell me a story about a child who got lost in the woods or . . . about a child who lost his dog. Loss and villainy are two common themes found in children’s stories (Westby, 1994). Analyze your story according to story grammar components. Did you find the optional Internal Response and Internal Plan? These components might appear in a 7-yearold’s story. When they do, they give you a valuable look into the child’s mind. Keep a copy of this narrative for use in other exercises in this chapter. (Note for all exercises: if you don’t have access to children, then ask an adolescent or an adult.)

Analyzing Content
Beyond macrostructure, one can gain valuable information regarding a storyteller’s attitude and reaction to his or her narrative by examining various aspects of the narrative’s content. Content analysis may focus on a wide range of subjects depending on the source material and the researcher’s interests. Below we return to perspective taking and resolutions, which are two aspects of content analysis.

Perspective Taking (PT) and Frame of Mind (FOM) (Revisited)
It is this writer’s view that the story grammar model is by nature a problem solving model. The Initiating Event sets up the problem, which is or is not resolved by the end of the story. However, looking beyond macrostructure to story content can provide an interesting view of a child’s attitude and reaction to a story’s events. The Internal Response is the place where FOM and PT can be revealed through adjectives, such as “happy,” “sad,” “excited,” or verbs, such as “jumped up and down,” “shouted,” convey emotion.
Evaluating frame of mind and perspective taking Review the fictional narrative you collected in Exercise 3 for examples of perspective taking.

Exercise 4

Analyzing Narratives


Story Resolution
Another useful method for analyzing narrative content focuses on narrative Resolutions. These resolutions have been found to vary according to culture (Westby, 1994), and are considered an important part of children’s narratives. Providing a resolution is “the most remarkable accomplishment in narratives between the ages of 5–6” (McCabe, 1996, p. 66). Kongas-Maranda and Maranda (1970) identified five stages of story resolution or ending. In their classic model (Table 15.3), the main character is considered the “minor power” who faces a “threat” or a more powerful force or being. That “force” is the “major power.” The levels of story resolution reflect the main character’s attempt to overcome the threat and the degree of success.
TABLE 15.3 Story Ending Levels Level 0: no threat to the character or change in the character’s condition occurs; there is no interaction with a greater power. Level 1: a greater power overcomes the main character (minor power), who does not attempt to respond to this greater power. Level 2: the minor power attempts a response, but fails in his efforts. Level 3: the minor power attempts a response and succeeds in overcoming the threat, returning to his original state. Level 4: the minor power succeeds in overcoming the threat, destroying the threat, and significantly improving his original state to a superior condition.
Source: Adapted from Kongas-Maranda and Maranda (1970)

Exercise 5

Analyzing the resolution Analyze the story ending level of the fictional narrative you collected in Exercise 3.

A Cross-Cultural Application to Narrative
While we have been discussing narratives in our Western-oriented culture, it is interesting and important to see what narratives are like in cultures other than our own. By so doing, we can appreciate cultural similarities and differences, and transcend viewing the world through our own, culture-based lens. This section will look at the Bhutanese children’s narratives I collected (Stein, 2004) to see what we can learn about another culture’s style of storytelling and the method of collecting a narrative. In viewing the outcomes of the research that follows, I will focus on three issues: story macrostructure, perspective-taking, and resolution. As we explore these three areas, we will also ask how methodology might affect the analysis of the resulting narratives. The brief overview of Bhutanese culture that follows will help you understand the results of my study.

Introduction to Bhutanese Culture
Bhutan emerged from feudalism and isolation in the 1960s when its monarchy decided to open up the country to the West to bring in technology and improve living conditions, health, and education. Today, the Bhutanese people are not rich, but there is no



harsh poverty or homelessness. Whether relatively urban or rural, households are large, consisting of many relatives and visitors. The Bhutanese are a highly religious, Buddhist people. Religion is an intrinsic part of their everyday lives (Schicklgruber, 1997). At the same time, the Bhutanese are also a practical people who voice their opinions about improvements to be made in their society. Men and women share similar responsibilities, although the family property is inherited through the women. Families are large and close knit, with relatives living within the same household. Children spend time with their parents, often accompanying them as the adults work on the farms. They may stay up late while adults are conversing and telling folktales before children go to sleep (Penjore, 2005). Children begin school in kindergarten where the language of school instruction is in Dzongkha, the dominant language of Bhutan. They also begin learning English in kindergarten. A good part of the curriculum is increasingly taught in English as children reach the higher grades. At the time of this study, children in Bhutan were not evaluated for learning or language disabilities. Anyone having academic difficulty was given extra help. In general, Bhutanese children have less storybook exposure than Western children, except in families where parents have been well educated. However, oral folktales are deeply rooted in Bhutanese society (Dorji, 2002; Penjore, 2005). These tales transmit their culture’s values, moral codes, manners, and religious concepts. They often serve as entertainment, particularly in rural regions where there is no television or radio (Penjore, 2005). Bhutanese stories often depict people interacting with spirits, demons, animals, and the gods. While characters also interact with nature, such as praying to trees in the forest to protect them, they also warn children of dangers in nature, such as wild animals. In other words, stories “play an indispensable role in the life of Bhutanese society in general” (Dorji, 2002).

Research on Bhutanese Children’s Narratives: Goals and Participants
Given this background, the goal of my research was to explore the type of oral, fictional narratives told by 7-year-olds, the age at which story grammar would be established in a Western culture. Having had little exposure to Western-type stories, would the Bhutanese children structure their self-generated narratives in ways similar to or different from this Western story grammar construct? The 7-year-old (second grade) children chosen for the study by the school’s headmaster, according to what was judged to be “good” academic performance, lived in Thimphu, the capital of Bhutan. Although, as stated above, the children spoke Dzongkha, Bhutan’s dominant language (closely related to Tibetan), some had migrated from other regions of the country and so were bilingual. Bilingualism did not interfere with the children’s performance, as confirmed by the two translators of the children’s narratives. I compared these children to rural American children who were age-matched, second graders from a small farming and small industrial community in Pennsylvania. These were white, monolingual children from working class families, chosen by the school’s principal, and identified as “average” students, without apparent learning, emotional, or neurological disabilities. Hunting was an important aspect of life in this region, and children often accompanied their parents on hunting expeditions. This population I believe to be a closer approximation to the Thimphu children than an urban American population would have been.

Analyzing Narratives


Story Grammar Findings
Some studies of non-Western stories have claimed that such stories cannot be analyzed according to a Western story grammar model because they de-emphasize goal orientation. For example, Matsuyama (1983) found an interesting pattern in an analysis of 150 Japanese folktales. Twenty-five percent of the stories contained two episodes: one with a goal-structure and one without any goal. Typically, there was no goal-structure for the first main character, but there was for the secondary character. In light of these findings, the results of my Bhutanese study were fascinating and surprising. The same percentage of Thimphu and Pennsylvania children told stories containing complete story grammar (70 percent in each group). The Thimphu children’s stories were as goal oriented as the Pennsylvania stories (remember that it is necessary to have a story goal in order to apply story grammar analysis). Of all the components, the (optional) Internal Plan was the least utilized by both groups, but the (optional) Internal Response (where the emotional state or frame of mind is revealed) was equally present in both populations. Why were the findings so similar in Story Grammar macrostructure? Could there have been something in the methodology that resulted in or triggered story grammar in the Bhutanese narratives?

Issues of Methodology
After collecting warm-up personal narratives from each of the Thimphu and Pennsylvania children individually, I asked each to create a fictional story about “a child who got lost in the woods.” The task was not based on picture stimuli or a wordless picture book. By asking the children to create their own stories, they could tap into their own personal and cultural experiences, previously heard stories, and imagination. The resulting stories might be a combination of all these aspects of story generation. In my view, the children were, therefore, not restricted by the content of a stimulus picture or set of pictures, although they were restricted to the given topic.

Two possibilities suggest themselves. One is that story grammar is an intrinsically problem solving method of story narration. The next logical step after the problem (Initiating Event) is to make an attempt to solve the problem. The methodology in this study may have triggered story grammar. That is, each child was asked to tell a story of a child who got lost in the woods: therefore, the problem is already established by the topic: lost in the woods. Consequently, the narration proceeds logically through attempts to “rescue” the child, successfully or not. Hence, once again, it is important to look at what subjects are asked to do in research experiments. However, there is a second possible explanation for the goal-oriented stories as told by the Bhutanese children. The Bhutanese stories contained many secondary characters: parents, police, friends, demons living in the woods, and animals. These adults and friends often served as the rescuers for the child main character, who was lost in the woods. Consequently, although the child (lacking in authority) may not have made Attempts to solve the problem, others attempted to come to the child’s rescue; hence, the goal-structure was carried out by other (secondary) characters. Therefore, methodology comes into play again. If the researcher only rated the child’s attempts at rescue or lack thereof, the narrative could appear to have no goal-structure: the child did nothing to solve the problem. If, on the other hand, the total story is evaluated where even minor



characters are rated for Attempts to solve the problem, then the story’s narrator demonstrated goal orientation.

Perspective Taking
The second issue I considered in my study, that of perspective taking, revealed similarities and differences between the Pennsylvania and Thimphu children. Both groups demonstrated awareness of their story characters’ frame of mind. They exhibited this form of perspective taking through the Internal Response, where the narrators revealed their characters’ reactions to being lost in the woods. The Bhutanese stories contained highly charged emotional terms, with words like: “fear,” “felt,” “pity,” and “shouted.” Perspective taking in the form of recognizing parental anxiety for the lost child was expressed. For the Pennsylvania stories, words such as “afraid,” “uh-ee” (expressing joy), “more,” and “more afraid” occurred. The difference, however, was in quantity: the Bhutanese stories contained statistically more emotionally charged words than the Pennsylvania stories. Due to the presence of secondary characters, the Bhutanese children were able to demonstrate their awareness/perspective of the frame of mind of other characters. They expressed the anguish and horror of a parent unable to find a missing child. Since the Pennsylvania stories did not contain any active, secondary characters who could assist in the Attempts to help the child, we lack information as to this aspect of perspective taking.

Major differences were also found in story ending level. A majority of the Bhutanese stories ended in the death of the main character, and/or secondary characters (70 percent). That is, although these main or secondary characters attempted to solve the problem, they were unsuccessful, in that they were killed (Level 2, according to Kongas-Maranda and Maranda, Table 15.3). In contrast, the Pennsylvania stories had a high level of successful outcomes in which the character returned safely (70 percent at Level 3). While these results are interesting, it is important to look beyond the classification to interpret the reasons for differences. Why was death so dominant in the Bhutanese stories? A complete explanation of this finding is beyond the scope of this chapter, as it requires further discussion of Bhutanese culture and Buddhist beliefs. Recall, though, that Bhutanese folktales attempt to transmit their culture, religion, and warn children of life’s dangers (Dorji, 2002; Penjore, 2005). Consequently, it is not surprising that threats to life would appear in the children’s narratives as a response to the topic of “a child lost in the woods,” a topic representing the possibility of danger.
Story ending level, frame of mind, and secondary characters Analyze the following fictional story as told by a 7-year-old Bhutanese child for the presence of frame of mind, secondary characters, and story ending level. Story grammar components appear to the right of each line; the gaps in lines 5 and 18 indicate that the speaker omitted a pronoun, a linguistic feature permitted in Dzongkha when the person being referred to is clear; in this case the pronoun is “her.” 1 One day a girl wanted to go home. (S + IE) 2 She went through a forest. (A)

Exercise 6

Analyzing Narratives 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 She went all over. (A) Her father and mother looked everywhere. (S + A) They could not find . (C) They reported to police (A) and searched everywhere. (A) The girl did not know what to do. (IR) She was sitting and crying. (IR + C) An evil spirit came and asked her: “What are you doing?” (S) Who are you? (S) She said: “I don’t know. (IE) I can’t find my parents.” (IE) So the demon led her to her [the demon’s] house. (A) The police reached the demon’s house. (S + A) They asked: “Did you see a girl called Tashi?” (A) The demon answered: “I do not know.” (C) I just cut her into pieces and made into a curry.” (C) Her parents came by. (S + A) They searched the house. (A) They found her in a small bottle. (R) (Stein, 2004) (Alternating bolded print represents a new episode; the episode may be complete or incomplete.)


Exercise 7

Further analysis of frame of mind Try collecting a narrative from a child, asking for more than one character. Do you see more references to frame of mind than you did in the narrative you collected for Exercise 3? Was there greater conflict now that you introduced more characters? The presence of more characters in a story may lead to more frame of mind evaluations due to the possibility of greater interaction among the characters.

To review the results of the study analyzing narratives of the Thimphu and Pennsylvania children: both groups of children demonstrated story grammar structure. However, perspective taking and resolutions revealed significant differences. Both groups demonstrated perspective taking. However, only the Bhutanese stories contained numerous secondary characters, resulting in narratives containing frame of mind for these secondary characters.

So, what is the relevance of story grammar and personal narrative structures to the present and the future? Analyzing narratives beyond the components of story grammar or personal narratives provides an in-depth means of viewing a culture as well as individual development in areas such as perspective taking and frame of mind. Readers may be interested in research on perspective taking in children’s development as well as in language disordered populations, such as people on the autistic spectrum (Losh and Capps, 2003). This chapter has demonstrated that careful analysis of the methodology used to



collect narratives is important. What the researchers ask children to do, where the narrative collection occurred (within the family, at school, with friends), the child’s socioeconomic level, and what kinds of narrative a child experiences daily can all influence research findings. In the study presented, we saw that the children of an Asian culture told stories that could be analyzed according to the Western construct of story grammar, perhaps indicating that the methodology used triggered some general, shared cognitive skills in storytelling: in this case, problem solving skills. Without data, one would not and should not assume that the narrative construct of one culture applies to another. Hardy (1977) called narrative “a primary act of mind,” that is, a cognitive task. However, narratives are more than that. Narratives are also rooted in conversation and experience, as Bamberg (2008) proposes. Narrative structure links the social world with the cognitive. As the other opening quotation by academic novelist David Lodge (2008) says, we live in a medium of narratives. What we learn from typically developing children can be applied to the treatment of language disordered populations, who continue to struggle to tell their stories. Analysis of narratives provides a means of evaluating and teaching narrative components to those struggling to communicate their ideas logically, and with perspective. Typically developing language users absorb their cultures’ narrative values and styles. The atypical child needs guidance to do so.

1 The reader is referred to Hughes et al. (1997) for a sequence of story grammar development identified as Story Structure. This sequence is based upon the completeness of the story grammar components, and is especially helpful for anyone in the field of language disorders, where identifying stages of narrative development is important in determining language growth. 2 Losh and Capps (2003) found that typically developing children, when compared to children on the autism spectrum, used more Evaluations in their personal narratives, rather than in their fictional narratives. These fictional narratives were based on the wordless picture book Frog, Where Are You? In other words, the personal narratives meant more to the children than the events in the picture book; this makes sense. In turn, it may be the case that a child’s own, self-generated fictional story or personal narrative may yield more Internal Responses or Evaluations because it may have more personal meaning for the speaker than a picture or wordless picture book. The issue may not be whether fictional or personal narratives create more Evaluations, but whether creating one’s own story may produce more story grammar components than responding to a stimulus picture.

Bamberg, M. (2008) Sequencing events in time or sequencing events in storytelling? From cognition to discourse—with frogs paving the way. In Guo, J. Ervin-Tripp, S., and Budwig, N. eds. Festschrift for Dan Slobin. Mahwah, Lawrence Erlbaum. Bamberg, M. and Damrad-Frye, R. (1991) On the ability to provide evaluative comments: Further explorations of children’s narrative competencies. Journal of Child Language, 18(3), pp. 689–710. Berman, R. (1995) Narrative competence and storytelling performance: How children tell stories in different contexts. Journal of Narrative and Life History, 5, pp. 285–314. Berman, R. and Slobin, D. (1994) Relating events in narrative: A crosslinguistic developmental study. Mahwah, Lawrence Erlbaum. Dorji, T. (2002) Folktale narration: A retreating tradition. Journal of Bhutan Studies, 6, pp. 5–23.

Analyzing Narratives


Hardy, B. (1977) Narrative as a primary act of mind. In Spencer, M. M., Warlow, A., and Barton, G. eds. The cool web: The pattern of children’s reading. London, Bodley Head, pp. 12–23. Hudson, J. and Shapiro, L. (1991) From knowing to telling: The development of children’s scripts, stories, and personal narratives. In McCabe, A. and Peterson, C. eds. Developing narrative structure. Mahwah, Lawrence Erlbaum, pp. 89–136. Hughes, D., McGillivray, L., and Schmidek, M. (1997) Guide to narrative language: Procedures for assessment. Eau Claire, Thinking Publications. Kongas-Maranda, E. K. and Maranda, P. (1970) Structural models in folklore and transformational essays: Approaches to semiotics. The Hague, Mouton. Labov, W. and Waletzky, J. (1967) Narrative analysis: Oral versions of personal experience. In Helm, J. ed. Essays on the verbal and visual arts: Proceedings of the 1966 annual spring meeting of the American Ethnological Society. Seattle, University of Washington Press, pp. 286–338. Lodge, D. (2008) Deaf sentence: A novel. New York, Viking Adult. Losh, M. and Capps, L. (2003) Narrative ability in high-functioning children with autism or Asperger’s Syndrome. Journal of Autism and Developmental Disorders, 33(3), pp. 239–251. McCabe, A. (1996) Chameleon readers: Teaching children to appreciate all kinds of good stories. New York, McGraw-Hill. McCabe, A., Bliss, L., Barra, G., and Bennett, M. (2008) Comparison of personal versus fictional narratives of children with language impairment. Journal of Speech-Language Pathology, 17, pp. 194–206. Matsuyama, U. K. (1983) Can story grammar speak Japanese? The Reading Teacher, 36(7), pp. 666–669. Ochs, E. and Capps, L. (2001) Living narrative: Creating lives in everyday storytelling. Cambridge, Harvard University Press. Penjore, D. (2005) Folktales and education: Role of Bhutanese folktales in value transmission. Journal of Bhutan Studies, 12(12), pp. 47–73. Peterson, C. and McCabe, A. (1983) Developmental psycholinguistics: Three ways of looking at a child’s narrative. New York, Plenum. Schegloff, E. A. (1997) Narrative analysis thirty years later. Journal of Narrative and Life History, 7(1–4), pp. 97–106. Schicklgruber, C. (1997) Gods and sacred mountains. In Schicklgruber, C. and Pommaret, F. eds. Bhutan: Mountain fortress of the gods. Boston, Shambala Publications, pp. 159–174. Stein, C. (2004) Narratives of Bhutanese and rural American 7-year-old children: Issues of story grammar and culture. Narrative Inquiry, 14(2), pp. 369–394. Stein, N. L. and Glenn, C. G. (1975) An analysis of story comprehension in elementary school children: A test of a schema. In Freedle, R. O. ed. New directions in discourse processing. Norwood, Ablex, pp. 53–120. Westby, C. (1994) The effects of culture on genre, structure, and style of oral and written texts. In Wallach, G. P. and Butler, K. G. eds. Language learning disabilities in school-age children and adolescents: Some principles and applications. Needham Heights, Allyn and Bacon, pp. 180–218.

Other Resources
Early Groundbreaking Work on Narratives
Heath, S. B. (1983) Ways with words: Language, life, and work in communities and classrooms. Cambridge, Cambridge University Press.


Stein Propp, V. Y. (1968) Morphology of the folktale. 2nd edn. Austin, University of Texas Press.

Suggested Journals for Further Reading of Narrative Development
Brain and Language Journal of Child Language Journal of Narrative Inquiry Journal of Applied Psycholinguistics Journal of Speech and Hearing Research Seminars in Speech and Language Topics in Language Disorders

Suggested Readings on Narratives in Varied Populations and Tests of Narrative Development
Berman, R. (1997) Narrative theory and narrative development: The Labovian impact. Journal of Narrative and Life History, 7(1–4), pp. 235–244. Catts, H. W., Hogan, T. P., and Fey, M. E. (2003) Subgrouping poor readers on the basis of individual differences in reading-related abilities. Journal of Learning Disabilities, 36(2), pp. 151–164. Choden, K. (1998) Women in the city. In Schicklgruber, C. and Pommaret, F. eds. Bhutan: Mountain fortress of the gods. Boston, Shambala Publications, pp. 253–262. Coelho, C. (2007) Management of discourse deficits following traumatic brain injury: Progress, caveats, and needs. Seminars in Speech and Language, 28(2), pp. 122–135. Gillam, R. B. and Pearson, N. (2004) Test of narrative language. Austin, PRO-ED. Griffin, T., Hemphill, L., Camp, L., and Wolf, D. (2004) Oral discourse in the preschool years and later literacy skills. First Language, 24(2), pp. 123–147. Murfett, R., Powell, M., and Snow, P. (2008) The effect of intellectual disability on the adherence of child witnesses to a “story grammar” framework. Journal of Intellectual and Developmental Disability, 33(1), pp. 2–11. Peterson, D., Gillam, S., and Gillam, R. (2008) Emerging procedures in narrative assessment: The index of narrative complexity. Topics in Language Disorders, 28(2), pp. 115–130. Renfrew, C. E. (1991) The bus story: A test of continuous speech. 2nd edn. Oxford, Old Headington. Schneider, P., Hayward, D., and Dubé, R. V. (2006) Storytelling from pictures using the Edmonton Narrative Norms Instrument. Journal of Speech-Language Pathology and Audiology, 30, pp. 224–238. Westerveld, M. and Gillon, G. (2008) Oral narrative intervention for children with mixed reading disability. Child Language Teaching and Therapy, 24(1), pp. 31–54.

Suggested Reading on Bhutan
Infojeunesse. (2009) Infoyouth [Internet]. Available from: [Accessed June 20, 2009]. Kingdom of Bhutan. (2009) Visitor information [Information]. Available from: http:// [Accessed June 20, 2009]. Kuensel Online. (2009) Kuensel online [Internet]. Available from: http://

Analyzing Narratives

313 [Accessed June 20, 2009]. This is a weekly Bhutanese newspaper, written in English. Ministry of Education Bhutan. (2008) Ministry of education [Internet]. Available from: http:// [Accessed June 20, 2009]. Royal Government of Bhutan, Ministry of Education. (2000) Annual statistical report. Thimphu, Bhutan.

This Chapter Explores:

Neurolinguistics and Psycholinguistics: Contributions to Understanding Healthy Aging and Dementia

Yael Neumann, Linda Carozza, and Anastasia Georgiou

Neural Changes in Healthy Aging and Dementia of the Alzheimer’s Type (DAT) Linguistic Changes in Healthy Aging and Dementia of the Alzheimer’s Type Differential Diagnosis

Research Methods in Studying Language in Healthy Aging and Dementia of the Alzheimer’s Type Behavioral: Priming and Narrative Discourse Analysis Neuroimaging: Event-Related Potentials

It takes a long time to become young.
Pablo Picasso

YN: As a researcher, I am intrigued with the complex study of brain-language relationships, and as a speech-language pathologist I am inspired by the creative therapeutic efforts used to help restore speech and language functions to maximum capability. But how does brain damage affect particular speech and language functions? And, how does speech-language therapy effect improvements in neural processing? These are some general questions that drive the clinical research I do. It is my hope that in answering these questions clinicians will be able to make better evidence-based decisions in their clinical work so that the ultimate goal of helping people with neurological damage/disease live better quality lives, from the standpoint of human communication, can be achieved. LC: I was interested in what people say and how they say it for as far back as I can recall. I loved hearing stories and eventually reading and writing my own. The study of language is both imaginative and scientific when it is encapsulated in the field of linguistics and conversational analysis. I was inspired to work in this area by the many speakers who have so generously shared their narratives with me, both in and out of the therapeutic environment.

Understanding Healthy Aging and Dementia


AG: As a Greek Cypriot speaking both Modern Greek and the Greek Cypriot dialect, I was amazed by the variety of dialectal differences that exists in each region of my small country, the island of Cyprus. It was not until my college years that I started researching the ancient roots of my primary language. As a speech-language pathologist in a Skilled Nursing Facility, I work with typical aging adults, adults with aphasia after a traumatic brain injury, and adults with the diagnosis of Alzheimer’s disease, including other forms of dementia.

People throughout the world are living longer than ever before. In fact, a statement by the National Institute on Aging (NIA) in July 2008 reported that individuals age 80+ years are the fastest growing segment in the United States, predicting that by 2030 this “oldest-old” group could grow to 10 million (National Institute on Aging, 2009). This expected increase in the aging population naturally leads to interest regarding how to maintain quality of life and eliminate or, at least, delay the onset of disease or disability. The focus of this chapter is on the linguistic changes that occur in healthy aging and in dementia of the Alzheimer’s type (DAT), also referred to as Alzheimer’s disease (AD) related dementia. In doing so, we explore an area of linguistics called neurolinguistics, an interdisciplinary science that investigates the structures and functions of the brain underlying the use of language and that is concerned with the mechanisms of the human brain in the comprehension and production of spoken, signed, or written language. The chapter begins with a neurolinguistic review of the changes that occur in the human brain and the declines evidenced linguistically in both healthy aging and dementing illness, which is the area of research done by the chapter authors. Additionally, we discuss the differential diagnosis, the process of systematically looking at a cluster of symptoms to determine what type of condition a patient has (whether language changes associated with dementia, forms of neurological disease affecting language, or healthy aging). The second half of the chapter focuses on psycholinguistic evidence in studying linguistic aspects of healthy aging and dementia-related language changes. Two categories of experimental research methodologies are explored. The first focuses on behavioral methods, e.g. priming and narrative discourse analysis, and the second on a neuroimaging technique, called Event-Related Potentials. These methods are highlighted in the context of experimental linguistic studies.

Exercise 1

Test your aging IQ: Take this NIA quiz (twenty-eight questions) to determine how much you know about normal aging: index.php.

Healthy Aging
The brain is an extraordinary organ. Its adult weight is only about 3 pounds, yet it seems to effortlessly carry out all our bodily functions, both volitional (under conscious control), such as moving, speaking, thinking, and seeing, and non-volitional (under automatic, unconscious control), such as breathing, digestion, and blood circulation. The normal brain contains 100 billion neurons (nerve cells) and 100 trillion synapses (gaps between neurons where nerve impulses are transmitted and received) that carry out these complex processes. However, as the physical body ages, so does the brain. The rise


Neumann, Carozza, and Georgiou

of neuroimaging techniques such as Positron Emission Tomography (PET scanning), Computerized Tomography (CT scanning), Functional Magnetic Resonance Imaging (fMRI), and Event-Related Potentials (ERPs, a type of electrophysiology), to name a few, has allowed scientists to uncover the structural and functional changes that occur in the brains of living people. The result of these neural changes is seen as modest linguistic variations in healthy aging, and as more serious breakdowns in conditions associated with dementia (with specific brain changes associated with different underlying etiologies). Our exploration of this topic begins first with healthy aging, followed by an account of language changes associated with dementia, particularly when related to Alzheimer’s disease origin, since it is crucial to understand what to expect in the normal process in order to appreciate the differences that characterize the disordered state.

Neural Changes
The brains of older people are distinct from those of younger people in various ways. First, it is common that in the aged brain some neurons shrink, particularly large ones in areas of the brain involved in memory, learning, planning, and other higher-level mental activities. Second, synapses are lost, largely due to reduced levels of mental stimulation. Third, abnormal structures called neurofibrillary tangles (insoluble twisted fibers that increase inside the neuron) and beta-amyloid plaques (dense, insoluble deposits of protein and cellular material that build up between nerve cells) develop, although in much smaller amounts than in Alzheimer’s disease. Lastly, with age, damage by free radicals (a kind of atom, molecule, or ion that readily changes other nearby molecules) increases.

Linguistic Changes
These neural changes have an impact on some aspects of linguistic abilities, particularly in the oldest-old population (individuals aged 80+ years). The affected linguistic domains include: lexical retrieval (the ability to name specific nouns and verbs), auditory comprehension (the ability to understand spoken language), and spoken discourse (the ability to produce sentences, converse, and/or relate narratives). Interestingly, this picture of linguistic decline is subject to great inter-individual variability as there are many instances of older-old adults who do not manifest these deficits at all. In contrast to these language changes, there are linguistic areas such as phonology (the study of the sound system of a language), syntax (rules for how to combine sequences of words to form acceptable sentences), and vocabulary, that are generally preserved, and sometimes even continue to improve, as with vocabulary skills, into very old age.

Lexical Retrieval
Of the three main linguistic areas experiencing decline with age, the most pronounced change occurs in lexical retrieval. Findings reveal that although naming abilities decline significantly in individuals in their eighth decade, subtle changes have also been documented in individuals in as early as the late fourth decade. Additionally, the increased difficulty in word retrieval is hypothesized to be due to breakdowns in the ability to retrieve the phonological shape of the words (James and Burke, 2000; White and Abrams, 2002). However, some studies suggest there is also breakdown in retrieval of the semantic (meaning) information, albeit to a distinctly lesser extent than that evidenced in dementia.

Understanding Healthy Aging and Dementia


Exercise 2

Ask a younger adult (20–40 years) and adult who is 50+ years to name as many animals as they can think of within one minute. Write down the responses. How did the two adults compare in the ease and speed with which they retrieved the animal names, and in the number of words retrieved?

Auditory Comprehension
Another linguistic domain compromised by the aging brain is that of auditory comprehension. This difficulty in understanding spoken information, e.g. following oral directions, is different than hearing loss in that it is not due to an auditory system deficit as the underlying cause. It is also not as obvious a deficit as that of a specific word retrieval breakdown, since typical daily communication is often redundant, providing cues to compensate for any failure in comprehension. Research studies suggest that healthy older adults, at around the sixth decade of life, experience a decline in understanding auditorily presented sentences. Although this finding might partially be due to deficits in hearing that typically occur in healthy aging and/or to cognitive conditions that commonly occur in older adults, an auditory comprehension decline is still clearly indicated separate from these two possible influences.

Spoken Discourse
Spoken discourse is the third linguistic area that demonstrates a clear difference in younger vs. older adults. Research findings suggest that older adults use grammatically simpler sentences that are less informative and more fragmented than younger adults (Marini et al., 2005). Additionally, older adults (75+ years) tend to use more units of irrelevant content and have missing or ambiguous referents, which makes their narratives less coherent (Marini et al., 2005). Conversational-interaction style also evidences changes with age, particularly in the 77–88 year group but also in the 60–74 year group (James et al., 1998). For example, older adults tend to be verbose and do not maintain topic, nor use appropriate turn-taking skills or clear referencing. At the end of the chapter, a description of a project by Carozza and Georgiou demonstrates how discourse analysis was used as a behavioral method to study linguistic changes in healthy aging and Alzheimer’s related language changes. In sum, the normal human brain undergoes changes with aging. These neural alterations result in linguistic declines, particularly affecting word retrieval, auditory comprehension, and spoken discourse. However, despite these normal changes, evidence suggests that linguistic competence in healthy aging should generally remain adequate throughout one’s lifespan.

Dementia of the Alzheimer’s Type (DAT)
Let’s look at two questions frequently asked by lay people, along with their answers. 1 2 What’s the difference between dementia and Alzheimer’s? What else can cause dementia?

The terms are not interchangeable. Dementia is a general term for cognitive decline, e.g. loss of memory and other intellectual abilities, caused by physical changes that occur


Neumann, Carozza, and Georgiou

in the brain and that have a serious impact on activities of daily living (Sungaila and Crockett, 1993). There are various types or causes of dementia, some that are treatable, such as dementia caused by side effects from medication, depression, or vitamin B12 deficiency; others that are less responsive to treatment, such as dementia of the Alzheimer’s type, multi-infarct dementia (MID), frontotemporal dementia, and Pick’s disease. The focus here is on Alzheimer’s disease as it is the most common cause of dementia. We are primarily concerned with dementia that is characteristic of individuals with underlying neurological disease associated with neurofibrillary plaque accumulations and plaque formation in the brain. This phenomenon was first described by Alois Alzheimer in the early twentieth century. At present, the large majority of dementias is associated with an Alzheimer’s disease diagnosis. Remember, then, that persons with dementia may or may not have Alzheimer’s disease as the underlying reason, but all persons with dementia related to Alzheimer’s disease by definition demonstrate dementia symptoms. Dementia refers to a process of cognitive decline due to neurodegenerative processes; it does not specify what the exact cause is.
In order to get a first-hand view of commonly held notions about Alzheimer’s disease and dementia, ask a younger and older adult what they think of when you mention these terms. Compare the perceptions given by the younger and older person. How are they similar and/or different?

Exercise 3

As you may have learned from the above exercise, there are many misconceptions about the terminology used to describe language disorders in the aging population. For instance, you may have found that the general public commonly interchanges the terms “dementia” and “Alzheimer’s disease.” (Interestingly, some published articles also use the generalized term “dementia” to describe the cognitive and behavioral characteristics of the progressive neurological condition regardless of the underlying disease process (Sungaila and Crockett, 1993). Additionally, you may have heard some differences in perception among the younger vs. older generation, e.g. the use of the term senility among older adults. In the past, it was assumed that older people suffered from senility. This condition purportedly was responsible for all declines seen in language and cognition with age. See Table 16.1 for a differentiation among terms related to conditions in the aged population. We know today, however, that neuropathological conditions underlie changes in
TABLE 16.1 Differentiation of Terms Related to Aging Term Senility Dementia Definition an “old-fashioned” term for decline in non-specific cognitive functions due to aging a general term for cognitive decline, e.g. loss of memory and other intellectual abilities, which seriously impacts activities of daily living a brain disorder that is progressive and degenerative; due to neuronal death and tissue loss, particularly in areas of the brain which are responsible for memory and thinking; it is the major cause of dementia

Alzheimer’s disease

Understanding Healthy Aging and Dementia


behavior and language. Neuropathological means abnormal changes in how the brain is physically structured and how parts of the brain communicate with one another. A gerontologist and/or neurologist are key practitioners in the correct identification and treatment of the neurological diseases that may affect communication, cognition, and/or motor status depending on sites of lesion (damage) in the nervous system. Essentially our language abilities and/or disabilities are indicators of our brain functioning. Therefore, language “quirks” that someone is exhibiting may actually be signs of an illness, warranting further exploration. For instance, when family members start noticing something amiss with an older person’s language abilities, e.g. the older person is forgetful of names, becomes confused in describing personal events, has a hard time holding a meaningful conversation, etc., they may become alarmed and suspect the possible onset of dementia in general. Identification of these changes in language and cognition is medically significant, indicating possible changes in the physical structure of a person’s brain.

Neural Changes
Dementia related to Alzheimer’s disease is a brain disorder that is progressive and degenerative. It results in neuronal death and tissue loss throughout the brain, particularly in cortical areas responsible for memory and thinking, e.g. hippocampus (a brain area involved in the formation of new memories). It is only post-mortem (after death), by viewing the brain tissue under a microscope, that scientists can definitively diagnose dementia’s underlying pathology. These investigations reveal an abundance of two types of abnormal structures, beta-amyloid plaques and neurofibrillary tangles, which build up in a more extreme and destructive way than in normal aging. The consequences of these changes are neuronal destruction and death, in particular causing selected neurons to stop functioning, lose connections with other neurons (due to blocked synaptic transmission), and finally die.
Exercise 4 Take this interactive tour created by the Alzheimer’s Association to learn more about how dementia affects brain functions:

To date, science has remedies to “slow down,” but not reverse or cure, the deficits associated with dementing illnesses. Therefore, many dementia patients will initially only have mild cognitive impairments that will evolve into full-blown dementia. Although it is currently unknown what causes the onset of this disease, we do know much more about Alzheimer’s disease, e.g. some gene markers, how to better diagnose and manage behavioral aspects, and what possible drug treatments can “slow down” its progression, thanks to the fast paced development of Alzheimer’s disease research over the past twenty-five years. Continued research will hopefully bring us closer to a future where we will be able to detect dementia in its various forms at earlier stages and eventually prevent or cure this devastating disease.

Cognitive–Linguistic Changes
As Alzheimer’s related dementia begins its course, the neural damage occurs first in areas of the brain involved in memory, language, and reasoning functioning. Only later in the disease are physical abilities affected. Thus, it is not surprising that the first clinical sign, and what is often considered the hallmark of dementia, is memory


Neumann, Carozza, and Georgiou

loss. Memory skills continue to decline with the progression of the disease, making it difficult for the patient with a dementing condition to remember the past or consider the future. In the linguistic realm, it is primarily the semantic aspects of language, e.g. comprehension and word choice, which are affected and get progressively worse, thereby also affecting meaningfulness of discourse. This semantic degradation will affect various linguistic domains, e.g. reading, writing. Phonology and syntax, however, are relatively spared at earlier and middle stages of language deterioration associated with Alzheimer’s disease. Additionally, some basic pragmatic skills are maintained, such as making eye contact and using formulaic language; other higher-level pragmatic abilities, however, such as making inferences, begin to deteriorate from the middle stage of dementia onwards. Lastly, various higher-level thinking skills, e.g. organization of thoughts, logical thinking, ability to inhibit impulsive actions, and attention span, are disturbed. These skills progressively deteriorate with disease development.

Differential Diagnosis of Dementia of the Alzheimer’s Type from Healthy Aging
Differential diagnosis of dementia of the Alzheimer’s type from healthy aging is particularly difficult at the early stage of the disease as the memory and linguistic problems that manifest are initially mild. The symptoms typically appear as normal forgetfulness or word retrieval problems that occur in healthy older adults, thus leading to delays in getting appropriate diagnostic testing early on. However, one key component to help families realize when a “senior moment” is a more serious problem is to consider whether their loved one has increased difficulty in making sense of the world and/or carrying out everyday life activities, such as driving, shopping, and paying bills. For example, it is normal for healthy older adults to occasionally forget where their car keys are and what an acquaintance’s name is; however, it is not typical to forget how much change you get back from a cashier, how to use the telephone, or find one’s way home. Other signs of mild Alzheimer’s related dementia, and other dementias in general, can include: taking longer to finish daily tasks, using poor judgment, and demonstrating mood and personality changes, such as increased anxiety. Overall, the picture of dementia associated with Alzheimer’s disease is very distinct from that of healthy aging even at early stages in the progression of the disease when differential diagnosis is most difficult. The significant domains of progressive decline in Alzheimer’s related dementia include memory, language (particularly semantic aspects), and reasoning/higher-level thinking skills.

Exercise 5

Before going on to the next section, do the following activity to increase your personal connection to the topic of dementia and Alzheimer’s disease. You will see how dementia is portrayed in the media as well as learn about people in the public eye who have had dementia. • Watch The Notebook, The Quick Brown Fox, Away from Her or The Memory of a Killer, all movies dealing with dementia related to Alzheimer’s disease. How would you characterize the language of dementia associated with Alzheimer’s disease based on these films? What might be inaccurate about the way Alzheimer’s disease is portrayed in such movies?

Understanding Healthy Aging and Dementia


Psycholinguistic Methodology
The experimental study of linguistics is a great way to learn more about how we humans communicate both under normal conditions and when breakdowns occur. Findings from such studies contribute to important theories and models that shed light on brain processing. This experimental side of linguistics is referred to as psycholinguistics. Language researchers search for experimental paradigms that can describe and explain the process of language changes in pathological populations. These may take the form of (1) descriptive research that is qualitative in nature, research focusing on subjective and in-depth analysis of a given topic, often to gain insight into people’s motivations, behaviors, value systems; and (2) empirical research which is quantitative in nature, research focusing on objective and numerical analysis of data. Because of the distinct nature of each human brain and the fact that the same disorder will have somewhat different manifestations in different individuals, much of the research has focused on single case studies. There are also rigorous experiments in which a wellselected population is tested, however, with the result being that much stronger “predictive” statements can be made regarding results and observations made in the study. In this chapter, the language samples collected in the work of Carozza and Georgiou fall under the domain of qualitative research of a pilot nature, while the neuroimaging and reaction time data collected in the work of Neumann (2007), Carozza (2006), and Neumann, Obler, Gomes, and Shafer (2009) are quantitative research. Methodology is the term that scientists use to describe the actual experimental steps used as a procedure in analyzing participant responses to different tasks presented to them or during scientific observation. Two particular methodologies, behavioral and neuroimaging, are reviewed. First, behavioral methodologies of priming and narrative discourse analysis will be reviewed, followed by a neuroimaging methodology employing Event-Related Potentials (ERPs), a type of electrophysiology. Additionally, examples of linguistic experiments that study language abilities and/or disabilities are presented. The reader will appreciate from these studies how linguistics can inform our understanding of the brain and play an important role in differential diagnosis of neurological diseases.

Behavioral Methodology Priming Priming is a common psychological research methodology that is probably being used in various laboratories across departments on your campus right now. In fact, you may have been recruited and/or taken part in a priming study as a healthy young adult participant. So what exactly is priming? Priming can be described in general as the facilitation of performance by exposure to related knowledge. It is a subspecialty in the field of psycholinguistics that studies how exposure to a specific item facilitates recall of that item and other similarly stored items in the mental lexicon. Priming behavior may take many forms; however, the method focused on in this section is language-related priming. Within this category of methodology is a task in which word recognition response times are assessed under varying conditions of semantic relatedness. The examiner may manipulate the degree of relatedness through selection of stimuli presentations. This is analogous to a game of association. For example, if your friend says the word “doctor,” you will likely recall the word “nurse” or some other closely related noun. This is true for all highly associated semantic concepts, e.g. “salt”–“pepper,” “table”–“chair,” “dog”–“cat.” This phenomenon accounts for the rapid ease with which one can


Neumann, Carozza, and Georgiou

communicate ideas since, with maturation, the brain develops multiple organizational strategies allowing us to access vocabulary and meaning quickly and without conscious effort. This organization of mental referents, with semantic and possibly orthographic (spelling) representations that allow for processing of vocabulary and semantic concepts, is termed the mental lexicon. Different types of organizations of the mental lexicon exist, e.g. organizing words based on having similar rhyming patterns, orthography, and/ or high associative relationships. This system of grouping is part of the normal operations of an information processing system that allows the brain to cross-reference in an efficient manner. The early work of Collins and Loftus (1975) provided a foundation for the use of priming in future experimental studies. One type of experimental design that uses priming is the lexical decision task. In this paradigm, what’s being measured is the speed of response in making a choice, of whether the second word in a word pair is a real word or non-word. The first word of the pair serves as the prime and the second as the target. It is expected that the response time (RT) in recognition of a target is greatest when the first word has “primed” the recall of the second due to high semantic associativeness between the two words. For example, if the prime item is “woman” and the following target item is “man,” it is expected that participants will be able to quickly make a lexical decision of the target word “man” due to the semantic relationship of this word pair. This is alternately called facilitative, associative, non-attention-dependent, or related priming. On the other hand, if a more distantly related word is used as the prime, the RT for recognition of the target will take longer or be inhibited, which is also known as attention dependent. This is hypothesized to be due to the brain’s need to suppress related words that are activated from the prime, while at the same time trying to retrieve the target from another point in the semantic field. Researchers can make finer distinctions within these processes by manipulating different experimental variables. One example is that of modulating the inter-stimulus interval (ISI), the amount of time between the prime and target. In making ISI adjustments, the researcher can account for factors, such as attention and memory, that contribute to processing in a priming task. A short ISI (less than 500 milliseconds [ms], is associated with facilitatory priming, while a long ISI (greater than 500 ms) is associated with inhibitory priming. Priming research in healthy older adults has demonstrated that patterns of facilitative and inhibitory priming are essentially intact throughout the lifespan. Normal aging may cause a slowing in RTs but the patterns remain consistent; namely, related items prime more quickly than unrelated items in both longer and shorter ISIs (cf. Burke, Mackay, Worthley, and Wade, 1991, for a good example of such research).

Exercise 6

Play a game of association with a young adult, e.g. classmate, friend. Present the following words, one at a time, and ask what word they think of when they hear the target word. Write down the associated words they recall. Then play the same game with an older adult (65+ years). Compare and contrast the word lists obtained. 1 2 3 4 5 6 water fish apple spaghetti desert dentist

Understanding Healthy Aging and Dementia 7 8 9 10 party exercise freedom love


Priming has also been used to study linguistic processes in disordered populations. But why would linguistic researchers be interested in “low-verbal” individuals, who generally have a very limited speech output and severe, irreversible brain damage? One reason is that studying language differences and changes reveals many interesting and important findings regarding normal processes. Second, detailing linguistic aspects of the disorder will lead to better diagnosis and remediation. One way of addressing the challenge of early detection and diagnosis of Alzheimer’s related dementia has been through priming. Research has found that aberrant patterns of priming—absent, less than normal (hypopriming), and greater than normal priming (hyperpriming)—were implicated in patients with Alzheimer’s disease (Carozza, 2006). Specialists see these failures in the efficiency of information processing operations as underlying the more common complaint of “forgetfulness” and “word-finding problems” that are usually described by the families of the patients or the patients themselves in some instances. Why would people with dementia related to Alzheimer’s disease show different RTs in priming studies compared to normal populations, and from one dementia priming study to another? One explanation is that there is great variability across patients in the brain damage associated with Alzheimer’s related dementia. Despite these individual variations, however, as a group there is a general pattern of expected cognitive decline that allows researchers to make strong predictions regarding language behavior. What is less clear is the underlying reason for these differences in priming behavior within the Alzheimer’s related dementia population. One understanding is that a change in priming may somehow be an early indicator of dementia of the Alzheimer’s type. Another is that the neurologic disease in this disorder itself may cause poor communication within important brain centers, thereby directly affecting how well a patient can recognize a word and any of its meanings and associations. It is not known if this is caused by (1) reduced amounts and poor transmission of brain chemicals called neurotransmitters (a physiological cause); (2) neuronal tissue damage and loss, which is related to actual brain atrophy (a physical cause); and/or (3) an abnormal buildup of fatty plaques and tangles along cell bodies in the brain (a combination of structural and functional causes). Unfortunately, as dementia characteristically progresses, all of these situations are present, and patients are rendered essentially mute and are ultimately unable to perform priming or any directive test of language function. An area of priming research that is increasing in interest is that aimed at determining specific patterns in the speed and accuracy of language processing to separate out normal aging, mild cognitive impairment unrelated to dementia, and/or specific diseases that may have dementia as a correlate. This is an extremely viable avenue of investigation; however, no such current data are available that can reliably distinguish priming changes necessarily associated with one or another of these conditions. What is known, however, is that there appear to be a sensitivity and vulnerability of rapid access to stored language that are being tapped when priming changes are detected in rigorous scientific experiments. Therefore, it is of utmost importance to be aware that the purpose of this section is not to set up priming as an end-all approach to assessing dementia, Alzheimer’s disease, or any other age-related changes in language function. It


Neumann, Carozza, and Georgiou

is rather to show that investigations aimed at such differentiations are underway and appear fruitful, albeit not yet conclusive. As we learned in the first part of the chapter, not all people with dementia necessarily have Alzheimer’s disease as the underlying etiology. Veteran scientists interested in this area of research therefore describe the type of dementia they are studying. This is at times very challenging as there are many similar and overlapping characteristics of the different types of dementia. A highly valuable next step in research would be to develop definitive protocols that delineate the various dementia types and lead to better care. It is important to keep in mind that priming, though demonstrating clear differences in RT speeds in patients with Alzheimer’s disease vs. healthy older adults, cannot definitively distinguish dementia, Alzheimer’s disease, and other related causative conditions. However, priming can provide clues to hypothesis building and problem solving with these elusive conditions. The function of even normal memory and language access and retrieval is not agreed upon by available research findings. Furthermore, in a real medical sense, it is still the case that definitive diagnosis of certain neurological diseases, Alzheimer’s disease among them, is not possible unless a patient’s brain is autopsied after death. This is sometimes done by medical researchers with family consent in order to add to the body of knowledge.

Narrative Discourse Analysis Another type of behavioral method used in linguistic studies is that of collecting a language sample via storytelling and analyzing aspects of the person’s ability to demonstrate normal discourse patterns. These investigations are helpful in identifying what linguistic features of output are considered normal, as in healthy aging or not, as in disordered states. Carozza and Georgiou collected language samples from individual participants in order to begin to determine if conversational analysis in a free speech situation with a theme-based conversational approach was a potential method to explore. The rationale for the research drew upon the fact that many individuals diagnosed with dementia are community-dwelling, possibly living at home with caregivers, and may be attending day centers, but not necessarily in diagnostic and treatment centers. (The reasons we find most people with dementia living at home are many, including the fact that dementia is not considered a “restorable” medical diagnosis.) However, the individuals require and deserve the attention of the scientific community so that their quality of life may be enhanced through a skilled conversational approach, which is a potential outgrowth of conversational analysis work. This procedural methodology has been extensively used by many researchers, such as Hannah Ulatowska, who, along with her colleagues, has studied discourse in normal old age, aphasia, and dementia (e.g. Ulatowska et al., 1998), as well as Bayles, Kaszniak and Tomoeda (1987), who have written descriptive analyses of communication changes associated with normal aging and dementia. Small, Geldart, and Gutman (2000), and Orange, Lubinski, and Higginbotham (1996) are also among the many research groups that have described various aspects of communication between demented individuals and their caregivers in the scientific literature, all of which contributes to evidence-based practice and clinical education. Lastly, you might look at the work of Welland, Lubinski, and Higginbotham (2002) for research in discourse comprehension. TimeSlips creative storytelling project is a group storytelling activity designed for people diagnosed with dementia due to probable Alzheimer’s disease (remember, currently this disease can only be definitively diagnosed post-morbidly). Its purpose is to improve the quality of life for both participants and their caregivers by encouraging participants’ use of narrative skills beyond their memory loss. Speech-language

Understanding Healthy Aging and Dementia


pathologists (SLPs) use this technique therapeutically with the objective of maximizing and maintaining participants’ functional communication and long-term memory skills. For example, in using TimeSlips as an intervention, SLPs would include the use of prompts and cues to encourage communication with appropriate eye contact, turntaking, and topic maintenance in story-building, as well as provide caregiver training and family support (Pore and Reed, 1999). The therapy technique uses generic pictures, preferably colorful ones, to trigger participants’ long-term memory, imagination, and creativity in storytelling (for sample pictures, see the TimeSlips website: http://, or Yahoo Images). However, this technique can also be used to collect discourse samples in the analysis of language changes that occur over time and/or in different populations. On this book’s accompanying website, details can be found of a pilot project, conducted by Linda Carozza and Anastasia Georgiou, using the TimeSlips method. Transcripts of discourse samples from both healthy older adults and individuals with dementia demonstrate what discourse differences exist in healthy aging as compared to dementia. In a different project, Ramanathan (1995) collected personal narratives from a single person with dementia, over the span of one and a half years, to study the changes that occur in narrative discourse with progression of the disease. The study used the method of schematic analysis, a systematic review of the components of a language sample, based on the person’s repetitions in the narrative, to analyze the participant’s discourse. This type of analysis focuses on such aspects of narrative structure as topic maintenance, logical storyline cohesion, thematic ties, specific vocabulary, and syntactic skills. Findings revealed a progressive decline in linguistic abilities characterized by memory deficits, word-finding difficulties, abstract reasoning deficits, minimal to no turn-taking skills, inability to maintain topic of conversation, and short length of utterance. Additionally, Ramanathan (1995) noted that the participant’s discourse concentrated mostly on topics related to major events in her life, and was characterized by simple utterances, repetitions, incoherence, and general lack of semantic coherence. Furthermore, deterioration of cognition was apparent over time.

Exercise 7

Use a neutral picture from the TimeSlips website,, to derive narratives from a healthy aging adult, e.g. family member or neighbor. Record the stories collected, including the elicitations and prompts provided. Compare and contrast the data collected from the healthy older person to that of narratives from the Alzheimer’s disease group (A1) given on the website. Note: This activity could be done individually or as a group project.

Neuroimaging Methodology Event-Related Potentials Improvements in neuroimaging techniques over the past few decades have allowed for enhanced and direct investigation of brain functioning in living patients. As a result, studies of language and cognitive processes in both healthy individuals and those with brain damage have been conducted, advancing our understanding of what occurs in the brain during normal vs. abnormal conditions. The focus of this section is on a neuroimaging technique called Event-Related Potentials (ERPs) that is presented within the context of a linguistic project by one of the authors of this chapter, Yael Neumann (Neumann, 2007; Neumann et al., 2009), studying word retrieval in healthy


Neumann, Carozza, and Georgiou

aging. (ERPs are the particular components within the waveform elicited in connection with particular events, e.g. making a decision, seeing a picture, etc.; electrophysiology is a general and very broad term for this type of technique of obtaining electrical activity through the scalp in response to different types of processing. It includes any method that records electrical function from the body, e.g. recording single-cell electrical activity is also electrophysiology—so it is not limited to scalp-recorded activity.) ERP methodology has its strengths and weaknesses, as do other neuroimaging techniques. Its main advantage is its excellent temporal (timing) resolution (in millisecond [ms] duration), while its main disadvantage is its poor spatial (location) resolution. Thus, this technique is most useful when answering scientific questions of “when” in real-time brain processing, rather than precisely “where” in the brain that processing is localized. Other advantages of ERPs include the fact that it is relatively low in cost, noninvasive, and provides an online measure of brain processing even without a behavioral response. So, for example, ERPs will detect when in time the brain retrieved phonological information to name a picture, even prior to a person’s overt response of naming. The method requires placement of a sensor net with electrodes on the scalp that picks up the small electrical activity (brainwaves) that the brain emits during processing. As these electrical signals are somewhat dampened en route from the brain through the skull and skin, ERP signals are very small and need to be amplified and averaged over many trials so that the resulting waveform can be more clearly seen. We look for components. Components are the parts of the waveform that occur at a given time (ms) over particular scalp sites, and that are linked to particular cognitive processes; further, these components may be positive (P) or negative (N) peaks. The N200 component was the component the Neumann (2007) study focused on. This component typically appears in the waveform as a negative peak that occurs at around 100–300 ms over fronto-central scalp sites and indexes response inhibition (e.g. withholding a button-press response) (see Figure 16.1).


N200 waveform

Source: Taken from Neumann, 2007 (figure 2). Note: Raw N200 No-go and Go waves (seen from approximately 301–450 ms) as demonstrated in the grand average (GAV) data from younger participants on the segment task. The N200 latency is later here than the typical 100–300 ms time window since type of language processing and word length are taken into account. x-axis = time (ms); y-axis = amplitude (µV-microvolt)

Understanding Healthy Aging and Dementia


Various cognitive and linguistic tasks can be mapped onto the inhibition response (N200) as a way of detecting when a particular neural processing occurs in real time. For instance, Schmitt, Munte, and Kutas (2000) used the N200 component to study the time course of retrieving semantic and phonological information in healthy young adults in order to further the understanding of lexical access for speech production. They did this by having participants either press (“Go” trials) or withhold pressing (“No-go” trials) a button in making a semantic decision (does a presented picture depict an object or animal) and a phonological decision (does the entity’s name start with a vowel or consonant). Since a semantic and phonological decision was mapped onto the Go/No-go decision, the resulting N200 was assumed to reflect when a participant encoded the given information. Thus this design is a useful measure to study language processing in real time. Neumann (2007) similarly applied this methodology with a focus on studying retrieval of phonological sublevels—segmental (phonemes), and syllabic information —in order to further the field’s understanding of why adults experience increasingly more word-retrieval problems with age. Behavioral research has indicated that this age-related difficulty in naming lies at the phonological (sound) level (Burke et al., 1991). However, the precise phonological breakdown, e.g. retrieval of individual sounds, syllabic structure, etc. in encoding a word for speech production, has not been localized. Thus, the Neumann (2007) study aimed at providing refined temporal information regarding two phonological substages, segment encoding and syllable encoding, which are implicated in lexical access problems. These two types of encoding will be defined as we review the instructions given to participants of the experiment. The study included 16 healthy younger adults (ages 23–40 years) and 16 healthy older adults (ages 68–80 years). Participants performed an implicit picture-naming task (they were asked to retrieve the picture name without labeling it out loud) while making a phonological decision about the word. They were asked to either Go (push a button) or No-go (don’t push a button) in making a decision. For example, let’s say the picture of a penguin was presented. In the segment task, if participants were instructed to push the button if the picture name ended with a /n/ sound and not push the button if the picture name ended with a /r/ sound, then participants should ‘Go’ on this trial. In the syllable task, if participants were asked to push the button if it is a one-syllable word and withhold a button-press response if it is a two-syllable word, then they were expected to ‘No-go’ on this trial. Results of the study indicated that older adults had greater difficulty than younger adults in phonological retrieval for naming on both the segmental and syllabic retrieval conditions. This finding helps explain why, with age, there is an increase in word-finding difficulties, as greater difficulty in accessing specific phonological information (as opposed to general processing delays with age, measured by a sensory-related ERP component) stalls the continued process of encoding for articulation. The implication of this finding is that healthy older adults might benefit from phonological practice, mainly of segmental and syllabic information, such as word games and crossword puzzles, to improve retrieval. Additionally, the design employed in this study can be used to study naming problems in brain-damaged populations so that therapeutic remediation can be specified to the appropriate level of difficulty. Lastly, investigations into the effects of specific treatment on brain processing will allow for more focused remediation.


Neumann, Carozza, and Georgiou

In conclusion, this chapter reviewed what changes occur in neural structure and function in healthy aging and dementia of the Alzheimer’s type, and how these changes affect cognitive–linguistic skills. Understanding these differences will facilitate earlier detection of a possible neuropathology. Additionally, behavioral methods, namely, priming and narrative discourse, in addition to a neuroimaging method, namely electrophysiology, were presented as examples of scientific tools used in the study of brain-language relationships in healthy aging and dementia related to Alzheimer’s disease.

Our consideration of Alzheimer’s related language changes, in contrast to other neurolanguage disorders such as aphasia (a disruption in language processes, e.g. speaking, comprehension, reading, writing, due to an acute and sudden localized brain injury that is usually, but not always, stroke related), is that the progressive language decline in Alzheimer’s related dementia is essentially a “living experiment.” The dissolution of human language that occurs in individuals who have Alzheimer’s disease, starting with subtle higher level changes, progressing to the loss of rote speech, and finally to mutism, although occurring in the reverse of its natural development, is a basic model by which students can begin to conceptualize this process. However, readers interested in learning more about the vast amount of research in the study of other neurolanguage disorders, such as aphasia, are referred to the Academy of Neurologic Communication Disorders and Sciences website (

References (2009) Alzheimer’s disease [Internet]. Available from: http://alzheimers.about. com [Accessed June 28, 2009]. Academy of Neurologic Communication Disorders and Sciences. (2009) Academy of Neurologic Communication Disorders and Sciences (ANCDS) [Internet]. Available from: [Accessed June 28, 2009]. Alzheimer’s Association. (2009) Brain tour [Internet]. Available from: brain/01.asp [Accessed June 28, 2009]. Bayles, K. A., Kaszniak, A. W., and Tomoeda, C. K. (1987) Communication and cognition in normal aging and dementia. Boston, Little, Brown. Borod, J. C., Goodglass, H., and Kaplan, E. (1980) Normative data on the Boston Diagnostic Aphasia Examination, Parietal Lobe Battery, and the Boston Naming Test. Journal of Clinical Neuropsychology, 2(3), pp. 209–215. Burke, D. M., MacKay, D. G., Worthley, J. S., and Wade, E. (1991) On the tip of the tongue: What causes word-finding failures in young and older adults? Journal of Memory and Language, 30, pp. 542–579. Carozza, L. C. (2006) Lexical-semantic priming abilities in early Alzheimer’s disease: A comparison within attention-dependent operations. Acta Neuropsychologica, 4(4), pp. 313–326. Collins, A. M. and Loftus, E. F. (1975) A spreading activation theory of semantic processing. Psychology Review, 82(6), pp. 407–428. James, L. E. and Burke, D. M. (2000) Phonological priming effects on word-retrieval and tip-of-the-tongue experiences in young and older adults. Journal of Experimental Psychology: Learning, Memory, and Cognition, 26(6), pp. 1378–1391.

Understanding Healthy Aging and Dementia


James, L. E., Burke, D. M., Austin, A., and Hulme, E. (1998) Production and perception of “verbosity” in younger and older adults. Psychology and Aging, 13(3), pp. 355–367. Marini, A., Boewe, A., Caltagirone, C., and Carlomagno, S. (2005) Age-related differences in the production of textual descriptions. Journal of Psycholinguistic Research, 34(5), pp. 439–463. National Institute on Aging. (2009) Tips from the National Institute on Aging [Internet]. Available from: .htm [Accessed June 28, 2009]. National Institute on Aging. (2009) What’s your aging IQ? [Internet]. Available from: [Accessed June 28, 2009]. Neumann, Y. (2007) An electrophysiological investigation of the effects of age on the time course of segmental and syllabic encoding during implicit picture naming in healthy younger and older adults. Ph.D. thesis, CUNY Graduate Center. Neumann, Y., Obler, L., Gomes, H., and Shafer, V. (2009) Phonological vs. sensory contributions to age effects in naming: an electrophysiological study. Aphasiology, 23(7–8), pp. 1028–1039. Orange, J. B., Lubinski, R. B., and Higginbotham, D. J. (1996) Conversational repair by individuals with dementia of the Alzheimer’s type. Journal of Speech and Hearing Research, 39(4), pp. 881–895. Pore, S. G., and Reed, K. L. (1999) Alzheimer’s disease. In Pore, S. G. and Reed, K. L. eds. Quick reference to speech-language pathology. Gaithersburg, Aspen, pp. 116–122. Ramanathan, V. (1995) Schematic understanding: Evidence from Alzheimer’s discourse. Communication Theory, 5(3), pp. 224–247. Ramanathan-Abbot, V. (1994) Interactional differences in Alzheimer’s discourse: An examination of AD speech across two audiences. Language in Society, 23(1), pp. 31–58. Schmitt, B. M., Munte, T. F., and Kutas, M. (2000) Electrophysiological estimates of the time course of semantic and phonological encoding during implicit picture naming. Psychophysiology, 37, pp. 473–484. Small, J. A., Geldart, K., and Gutman, G. (2000) Communication between individuals with dementia and their caregivers during activities of daily living. American Journal of Alzheimer’s Disease and Other Dementias, 15(5), pp. 291–302. Stern, C., Prather. P., Swinney, D., and Zurif, E. (1991) The time course of automatic lexical access and aging. Brain and Language, 40(3), pp. 359–372. Sungaila, P. and Crockett, D. (1993) Dementia and the frontal lobes. In Parks, R., Zec, R., and Wilson, R. S. eds. Neuropsychology of Alzheimer’s disease and other dementias. Oxford, Oxford University Press, pp. 235–264. Time Slips (2009) TimeSlips: Creative storytelling project [Internet]. Available from: [Accessed June 28, 2009]. Ulatowska, H., Chapman, S., Highley, A., and Prince. J. (1998) Discourse in healthy oldelderly adults: A longitudinal study. Aphasiology, 12(7–8), pp. 619–633. Welland, R. J., Lubinski, R., and Higginbotham, D. J. (2002) Discourse comprehension test performance of elders with dementia of the Alzheimer type. Journal of Speech, Language, and Hearing Research, 45, pp. 1175–1187. White, K. K. and Abrams, L. (2002) Does priming specific syllables during tip-of-the-tongue states facilitate word retrieval in older adults? Psychology and Aging, 17(2), pp. 226–235.

Other Resources
Bayles, K., Kaszniak, A., and Tomoeda, C. (1987) Communication and cognition in normal aging and dementia. Boston, Little, Brown.


Neumann, Carozza, and Georgiou Neely, J. H. (1991) Semantic priming effects in visual word recognition: A selective review of current findings and theories. In Besner, D. and Humphreys, G. W. eds. Basic processes in Reading: Visual word recognition. Mahwah, Lawrence Erlbaum, pp. 264–336.

Abrams, L., White, K. K., and Eitel, S. L. (2003) Isolating phonological components that increase tip-of-the-tongue resolution. Memory and Cognition, 31(8), pp. 1153–1162. Albert, M. S., Heller, H. S., and Milberg, W. (1988) Changes in naming ability with age. Psychology and Aging, 3(2), pp. 173–178. Ardila, A. and Rosselli, M. (1989) Neuropsychological characteristics of normal aging. Developmental Neuropsychology, 5(4), pp. 307–320. Barresi, B. A., Nicholas, M., Connor, L. T., Obler, L. K., and Albert, M. L. (2000) Semantic degradation and lexical access in age-related naming failures. Aging, Neuropsychology and Cognition, 7(3), pp. 169–178. Cross, E. S. and Burke, D. M. (2004) Do alternative names block young and older adults’ retrieval of proper names? Brain and Language, 89(1), pp. 174–181. Goulet, P., Ska, B., and Kahn, H. J. (1994) Is there a decline in picture naming with advancing age? Journal of Speech and Hearing Research, 37(3), pp. 629–644. Heine, M. K., Ober, B. A., and Shenaut, G. K. (1999) Naturally occurring and experimentally induced tip-of-the-tongue experiences in three adult age groups. Psychology and Aging, 14(3), pp. 445–457. MacKay, A. J., Connor, L. T., Albert, M. L., and Obler, L. K. (2002) Noun and verb retrieval in healthy aging. Journal of the International Neuropsychological Society, 8, pp. 764–770. Martin, A. and Fedio, P. (1983) Word production and comprehension in Alzheimer’s disease: The breakdown of semantic knowledge. Brain and Language, 19(1), pp. 124–141. Meyer, D. E. and Schvanaveldt, R. W. (1971) Facilitation in recognizing pairs of words: Evidence of a dependence between retrieval operations. Journal of Experimental Psychology, 90(3), pp. 227–234. Nebes, R. D., Boller, F., and Holland, A. (1986) Use of semantic context by patients with Alzheimer’s disease. Psychology and Aging, 1(3), pp. 261–269. Nebes, R. D., Brady, C. B., and Huff, F. J. (1989) Automatic and attentional mechanisms of semantic priming in Alzheimer’s disease. Journal of Clinical and Experimental Neuropsychology, 11(2), pp. 219–30. Orange, J., Lubinski, R., and Higginbotham, D. J. (1996) Conversational repair by individuals with dementia of the Alzheimer’s type. Journal of Speech and Hearing Research, 39, pp. 881–895. Ramsay, C. B., Nicholas, M., Au, R., Obler, L. K., and Albert, M. L. (1999) Verb naming in normal aging. Applied Neuropsychology, 6(2), pp. 57–67. Rastle, K. G. and Burke, D. M. (1996) Priming the tip of the tongue: Effects of prior processing on word retrieval in young and older adults. Journal of Memory and Language, 35(4), pp. 585–605.

Simpson, J. (2002) Elderspeak—Is it helpful or just baby talk? [Internet]. Available from: [Accessed June 30, 2009].

This Chapter Explores:
Autism Spectrum Disorders Social Interaction Language

Language of Children with Autism The Two Worlds Underlying Verbal Communication

Marion Blank and Mary Beth Cull

Verbal Communication Discourse Hyperlexia

Although I could speak I often didn’t use language in the same way as others and often got no meaning out of what was said to me . . . I had a whole system of relating that I considered ‘my language.’ It was other people who did not understand the symbolism I used, and there was no way I could or was going to tell them what I meant. Williams, 1992, p. 29

MB: My story begins four decades ago, at the start of my career. A colleague suggested that I might find it interesting to meet a certain child. And so I met a 4-year-old boy with autism who could not respond to even the simplest of questions but was nevertheless adeptly reading the front page of the New York Times. Since that time, as I pursued many areas of language, I have maintained an interest in understanding and helping children with autism. I was delighted when I was offered the opportunity to write this chapter, since it gives me the opportunity to bring together a range of concepts central to understanding the lives of these extraordinary children. MBC: When I began in the field, autism was already a “hot topic.” A wealth of academic research and a variety of treatment models were steadily appearing. Keen to be a part of what was happening, I began working as a therapist in a home-based intervention program. Following that, I moved to New York. Committed to building on what I had learned, I was fortunate to join the newly created interdisciplinary program at Columbia University. That led me to my working with Dr. Blank, and to this chapter.


Blank and Cull

Autism . . . A term that floods today’s world—through movies, books, ads, and every other channel possible. The non-stop coverage makes it hard to believe that the concept of autism did not even exist until the middle of the twentieth century. The neglect is particularly striking given that the syndrome presents behavior that is hard to miss. For example, a person—with no foreknowledge of autism—seeing real individuals such as Amanda Baggs (on YouTube) or the fictional character Raymond Babbitt in the movie Rain Man would not have a shred of doubt that these persons are markedly different from most other people. So, how could autism have gone “unseen” for so long? The answer probably rests with the paucity of concepts that were available for interpreting developmental phenomena. For a start, children’s behavior was not well understood. The very concept of childhood—as a unique period in the life cycle—is relatively recent (see Aries, 1962). Hence, there were few, if any, tools for studying and interpreting children’s behavior.

Exercise 1

View a film or video depicting someone with autism. Possibilities are Rain Man, Molly, or In My Language (Amanda Baggs on YouTube: watch?v=JnylM1hI2jc). Focusing on language, facial expression, gestures, tone of voice, emotions, and social miscues, list ten specific behaviors you see as different.

With the developmental process of childhood essentially invisible, differences in development were even more invisible. Child psychiatry did not even exist as a discipline in the United States until 1930, when Leo Kanner was appointed to the faculty at Johns Hopkins University. And it was Kanner, an émigré from Austria, who first used the term autism in a paper where he detailed the developmental course of eleven children he had treated (Kanner, 1943).

A Core Constellation: Social Interaction, Language, and Verbal Communication
Autism, a syndrome that involves the biological, psychological, and social spheres, affects every major realm of behavior, including motor functioning, sensory hypersensitivities, sleep patterns, and digestion. The breadth of its effects is one of the reasons that the term Pervasive Developmental Disorders (PDD) is now commonly used for classifying children with symptoms of autism. In his groundbreaking 1943 paper, Kanner identified a constellation of behaviors. Not all children with autism will display severe difficulties in all areas. However, the clusters outlined by Kanner continue to be accepted as hallmarks of the syndrome. 1 2 Social skills: social interaction was askew with the children seeking an “extreme autistic aloneness that, whenever possible, disregards, ignores, shuts out anything that comes to the child from the outside.” Language: several children did not develop language, though most (eight of the eleven) did. In those eight, however, the language was focused on “nouns identifying objects” (“naming”), adjectives indicating colors, and numbers. In those who displayed more extended language, i.e. sentences, that language often represented “parrot-like repetitions of heard word combinations.”

Language of Children with Autism



Communication: the children showed patterns where their language failed “to convey meaning to others.” (Kanner, 1943, pp. 242–243)
Read Kanner’s 1943 paper. This classic paper is available at ( From Kanner’s writeup, identify five types of behaviors or symptoms that represent domains other than the three noted above that he states are associated with autism.

Exercise 2

In the following year, another Austrian, Hans Asperger (1944), identified a separate group of children who showed significantly reduced social interaction and communication skills. His cohort functioned at a higher level of language and, in some cases, had notably outstanding skills in other areas (often math and music). Asperger also brought to the fore difficulties in communication that did not directly involve language, such as problems in “eye gaze, gestures, posture, voice quality, prosody (melodic quality of speech)” (Frith, 1991, p. 10).

The Spectrum Concept
Once recognized, the disorder attracted attention from behavioral scientists eager to understand the forces behind this unique constellation (Rutter et al., 1971). At the time, however, the number of reported cases was low (about one in 2,500 children). A disorder that rare meant that interest in autism was confined largely to academia. That has now changed. The numbers keep skyrocketing, with recent figures suggesting the rate may be as high as one in 150, according to the Centers for Disease Control and Prevention’s Autism and Developmental Disabilities Monitoring (ADDM) Network (2009). There is considerable debate as to whether this increase actually reflects a combination of greater awareness and better diagnosis—children who previously would have been classified as mentally retarded may now be classified as having autism. There are others who believe that the rising numbers reflect a true “epidemic” caused by factors such as environmental toxins (Blaxill, 2004). Along with the increase in numbers, there has been an increasing recognition of large individual differences among children with autism. Some individuals show the full range of symptoms, while others show only a few. To capture this variety, the American Psychiatric Association (2000) created the concept of a spectrum: specifically, Autism Spectrum Disorders (ASD). Currently, Autism Spectrum Disorders (ASD) is used synonymously with Pervasive Developmental Disorders (PDD), the concept that we discussed earlier. The term Autism Spectrum Conditions (ASC) is also used. In keeping with the current practice of naming the individual before naming the condition that the individual has (person-first language), children on the spectrum are not referred to via labels such as “autistic children.” Rather, terms you will hear are ones such as “children with autism” or “individuals with ASD.” Based on the severity of the problems, a child is placed at different points along the spectrum. Classic autism is at the extreme lower end, while a stronger command of language places a child at the higher end. Asperger Syndrome (which is sometimes used synonymously with high functioning autism) is also part of ASD. Asperger Syndrome is a condition marked by difficulties in social interaction and repetitive patterns of behavior or interests; however, individuals with this syndrome, in contrast to others on the autism spectrum, show a relative absence of language delays and difficulties.


Blank and Cull

Despite the varying abilities, children on the spectrum continue to display the constellation that Kanner outlined in the realms of social behavior, language, and communication. Our goal here is to provide an overview of these areas and guidelines as to how they relate to one another.

Understanding the Constellation—Step 1: Social Cognition
The current understanding of autism owes a great deal to advances that have taken place in the social sciences over the past several decades, particularly in social cognition and infant perception. For generations, infants’ minds were thought to be masses of “blooming, buzzing confusion” (James, 1890, p. 462). The assumption was that babies, at birth, lacked any significant cognitive skills. In the 1970s, this view was challenged by research showing that infants possess an array of abilities, termed social cognition, that include particular skills in social interaction (e.g. making eye contact, attending to faces, listening to voices, categorizing speech sounds, etc.) (Cohen and Salapatek, 1975). This research encompassing areas such as gesture language, joint attention, mirror neurons, and theory of mind shows social cognition flowering throughout the first two years. The end result has been a long-delayed recognition of the richness of the infant’s preverbal social cognition. The decreased level of social initiation of children with autism contrasts with the pull towards social contact that marks typical infants (Stern, 2002). The exact nature of the differences is now being defined through the use of paradigms for studying social cognition in neurotypical infants (a term coined within the autism community to describe individuals whose neurological development follows what most people conceive of as normal).

Gesture Language and Joint Attention
When describing their concerns, parents of young children with ASD are likely to offer comments such as:
“She doesn’t ever seem to point the way other kids do.” “He doesn’t make eye contact.”

These comments signify problems in areas known as gesture language (Colgan et al., 2006) and joint attention (Bruinsma et al., 2004). • Gesture language refers to the rich system of intricate movements that, without speech, conveys much of our thoughts and feelings. While some of the actions are conscious (e.g. shaking one’s head to indicate “no”), many are produced without awareness or intent (e.g. a facial expression of surprise; a body posture of dismay, etc.). • Joint attention, by contrast, contains definite communicative intent. The goal, often conveyed via gesture language such as eye gaze or pointing, is to reach out to another person to share an experience. An infant under a year, for example, who is looking at something and wants a parent to do likewise, may take the adult’s face and move it so that it is positioned to do what the infant desires. Both gesture language and joint attention involve the rapid transmission of complicated information. In this respect, they are similar to the split-second decision-making

Language of Children with Autism


described in the book Blink (Gladwell, 2005). We see how people, relying on their social cognition, can instantly size up a situation and make valid judgments—without being aware of how they are doing this, or even that they are doing this. A similar process happens repeatedly in normal parent–child interaction. The initial research in this sphere focused on mother–child interaction. From the outset, it demonstrated that mothers, without explicit training, are equipped to identify and interpret the endless array of small, subtle, non-verbal behaviors that infants produce. Similarly, from research on attachment behavior, we have long known that mother–child interaction is bidirectional (Ainsworth et al., 1978). The adults are not the sole actors in the exchange. Babies are active participants as well. They are equipped not only to interpret the nonverbal behaviors produced by the caregivers, but also to produce comparable behaviors of their own. In ASD, however, the babies are not “performing” in the way that adults intuitively expect. This causes a profound alteration in the parent–child relationship and all interpersonal relationships that follow. Ironically, because the interpersonal system is designed to function without thinking, the parent may often be unaware that something is wrong. He or she may experience a sense of unease, but be unable to identify its source or significance. This situation is one reason why autism—even when present from birth—often remains undiagnosed in the first two years of life.

Mirror Neurons
Gesture language, joint attention, and other social skills appear as part of the baby’s intense desire for interpersonal contact. However, motivation alone is not enough. The skills also come into play because the infant possesses an ability to sense “what people are about.” Consider, for example, the following. An 8–10-month-old baby, in