Culture and Language Use 9789027207791, 9027207798

The ten volumes of "Handbook of Pragmatics Highlights" focus on the most salient topics in the field of pragma

742 120 2MB

English Pages 280 [296] Year 2009

Report DMCA / Copyright


Polecaj historie

Culture and Language Use
 9789027207791, 9027207798

Table of contents :
Culture and Language Use
Editorial page
Title page
LCC data
Table of contents
Preface to the series
1. Introduction: The pragmatics and metapragmatics of routine formulae
2. The anatomy of aisatsu
2.1 Marking social relationships
2.2 Marking contextual boundaries
3. The social acquisition of aisatsu
3.1 Aisatsu in school contexts
3.2 Aisatsu in workplace contexts
4. The pragmatics of aisatsu within the metapragmatics of aisatsu
5. Exploring intercultural routine formulae
Anthropological linguistics
1. Preliminaries
2. Early history
3. Early types of research
4. Continuity
5. The Sapir-Whorf hypothesis
6. Ethnographic semantics
7. The ethnography of communication
8. Sociolinguistics
9. Recent research and current directions
Franz Boas
1. Introduction
2. Biography
3. The Americanist tradition
4. Time perspective in aboriginal languages
5. The Handbook of American Indian Languages
6. Phonetics vs. phonemics
7. Assessment
Cognitive anthropology
1. Historical background
2. Reconstruing cognitive anthropology
Componential analysis
1. Introduction
2. The structuralist tradition
3. Linguistic anthropology
4. Generative and typological studies
5. Natural semantic metalanguage (NSM)
6. Other trends and problems
Cultural scripts
1. Introduction
2. Semantic primes: The language of cultural scripts
3. High-level scripts: Examples and observations
4. Communicative style: Some further examples
5. Culture-specific concepts in cultural scripts
6. Rhetorical speech practices
7. Scripts for ways of thinking and feeling, and for beliefs
8. Scripts for social models
9. Scripts for non-verbal communicative practices
10. The accessibility and practicality of cultural scripts
11. Closing note
1. Introduction: Culture as an interdisciplinary project
2. Defining and redefining culture: An overview
2.1. The historical transformation of the culture concept
2.2 What is not culture
2.3 What is culture
2.3.1 The whole is the sum of parts: The mentalist approach
2.3.2 A map of and for behavior: The behaviorist approach
2.3.3 The metaphor of culture as text: The semiotic approach
2.4 Culture is a verb
3. Culture as ideology: From a consensual to a differentiated view of culture
3.1 The construction of the cultural other
3.2 The politics of cultural difference in stratified multicultural societies
4. Doing cultural analysis: A critique of the method
4.1 The dominance of intensive fieldwork
4.2 The subjectivity of participant observation
4.3 Doing and writing ethnography as interpretation and invention
5. Culture as communication: The ‘linguistic’ turn
5.1. One language — one culture?
5.2 Language as a means to cultural resources
6. Cross-cultural and intercultural analysis in pragmatic research
6.1 Cross-cultural pragmatic research: The culture principle
6.2 Intercultural pragmatics: Towards a critical reading
7. Conclusion: The pragmatics of recovering culture
1. Introduction
2. Ethnography by example
3. Ethnography by contrast
4. Discourse
Ethnography of speaking
1. Development and main characteristics
2. Areas of inquiry
3. Issues and debates
Firthian linguistics
Folk pragmatics
1. Folk linguistics
2. Folk pragmatics
3. Directions for future research
1. Definitions
2. Dimensions of variation and comparison
2.1 Linguistic forms
2.2 Sources of honorific expressions
2.3 Honorification roles and participant roles
2.4 Deployment and ideology
3. Pragmatics and semantics
4. Trends in theory and research
Wilhelm von Humboldt
1. Life
2. Humboldt’s philosophy of language
3. Language and thought
4. Language and world
5. Language and languages (superior and inferior ones)
6. Language, culture, and creativity
7. Language, dialogue, and pronouns
8. Humboldt and the Idéologues
Intercultural communication
1. Background: Language and culture
2. Intercultural communication: The emergence of a field of inquiry
3. The concept of culture
4. Loci of culture-in-communication
5. Methodological (sub)discourses of the field
6. Conclusion
1. Introduction
2. The construction of interviews and questionnaires
3. Research on interviewing
4. Pragmatic underpinnings of interviews
5. Interviewing and institutional authority
6. Conclusion
Bronislaw Kasper Malinowski
1. Biographical sketch
2. The study of culture
3. Fieldwork
4. Theory of language
5. An appraisal
Phatic communion
Edward Sapir
1. Introduction
2. Biography
3. Sapir on language, culture and personality
3.1 Language
3.1.1 Americanist text tradition
3.1.2 Linguistic form
3.1.3 Form-feeling
3.1.4 Linguistic relativity
3.2 Culture and personality
4. Sapir’s importance to pragmatics
Benjamin Lee Whorf
1. Introduction
2. Whorf’s life and work
3. Whorf’s perspective on linguistics
4. The linguistic relativity principle
5. Whorf’s influence
The series Handbook of Pragmatics Highlights

Citation preview

Culture and Language Use

Handbook of Pragmatics Highlights (HoPH) The ten volumes of Handbook of Pragmatics Highlights focus on the most salient topics in the field of pragmatics, thus dividing its wide interdisciplinary spectrum in a transparent and manageable way. Each volume starts with an up-to-date overview of its field of interest and brings together some 12–20 entries on its most pertinent aspects. Since 1995 the Handbook of Pragmatics (HoP) and the HoP Online (in conjunction with the Bibliography of Pragmatics Online) have provided continuously updated state-of-the-art information for students and researchers interested in the science of language in use. Their value as a basic reference tool is now enhanced with the publication of a topically organized series of paperbacks presenting HoP Highlights. Whether your interests are predominantly philosophical, cognitive, grammatical, social, cultural, variational, interactional, or discursive, the HoP Highlights volumes make sure you always have the most relevant encyclopedic articles at your fingertips.

Editors Jef Verschueren

Jan-Ola Östman

University of Antwerp

University of Helsinki

Volume 2 Culture and Language Use Edited by Gunter Senft, Jan-Ola Östman and Jef Verschueren

Culture and Language Use

Edited by

Gunter Senft Max-Planck Institute for Psycholinguistics

Jan-Ola Östman University of Helsinki

Jef Verschueren University of Antwerp

John Benjamins Publishing Company Amsterdam / Philadelphia



The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences — Permanence of Paper for Printed Library Materials, ANSI Z39.48-1984.

Cover design: Françoise Berserik Library of Congress Cataloging-in-Publication Data Culture and language use / edited by Gunter Senft, Jan-Ola Ostman, Jef Verschueren.        p. cm. (Handbook of Pragmatics Highlights, issn 1877-654X ; v. 2) Includes bibliographical references and index. 1.  Language and culture. 2.  Pragmatics.  I. Senft, Gunter, 1952- II. Östman, Jan-Ola. III. Verschueren, Jef. P35.C825    2009 306.44--dc22 2009012596 isbn 978 90 272 0779 1 (pb; alk. paper) isbn 978 90 272 8930 8 (EB) © 2009 – John Benjamins B.V. No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publisher. John Benjamins Publishing Company • P.O. Box 36224 • 1020 me Amsterdam • The Netherlands John Benjamins North America • P.O. Box 27519 • Philadelphia PA 19118-0519 • USA

Table of contents

Preface to the series




Introduction Gunter Senft


Aisatsu 18 Risako Ide Introduction: The pragmatics and metapragmatics of routine formulae  18 1. 2. The anatomy of aisatsu  19 2.1 Marking social relationships  20 2.2 Marking contextual boundaries  21 3. The social acquisition of aisatsu  23 3.1 Aisatsu in school contexts  23 3.2 Aisatsu in workplace contexts  24 4. The pragmatics of aisatsu within the metapragmatics of aisatsu  25 Exploring intercultural routine formulae  26 5. Anthropological linguistics Ben G. Blount Preliminaries  29 1. 2. Early history  30 3. Early types of research  30 4. Continuity  31 The Sapir-Whorf hypothesis  32 5. 6. Ethnographic semantics  33 7. The ethnography of communication  35 8. Sociolinguistics  36 9. Recent research and current directions  37



Handbook of Pragmatics Highlights

Franz Boas Regna Darnell 1. Introduction  41 2. Biography  41 3. The Americanist tradition  42 4. Time perspective in aboriginal languages  43 5. The Handbook of American Indian Languages  45 6. Phonetics vs. phonemics  46 7. Assessment  47


Cognitive anthropology Stephen C. Levinson 1. Historical background  50 2. Reconstruing cognitive anthropology  53


Componential analysis Cliff Goddard 1. Introduction  58 2. The structuralist tradition  58 3. Linguistic anthropology  60 4. Generative and typological studies  60 5. Natural semantic metalanguage (NSM)  62 6. Other trends and problems  65


Cultural scripts Cliff Goddard 1. Introduction  68 2. Semantic primes: The language of cultural scripts  69 3. High-level scripts: Examples and observations  69 4. Communicative style: Some further examples  71 5. Culture-specific concepts in cultural scripts  72 6. Rhetorical speech practices  74 7. Scripts for ways of thinking and feeling, and for beliefs  75 8. Scripts for social models  76 9. Scripts for non-verbal communicative practices  77 10. The accessibility and practicality of cultural scripts  77 11. Closing note  78


Table of contents

Culture Srikant Sarangi 1. Introduction: Culture as an interdisciplinary project  81 2. Defining and redefining culture: An overview  81 2.1 The historical transformation of the culture concept  81 2.2 What is not culture  83 2.3 What is culture  84 2.3.1 The whole is the sum of parts: The mentalist approach  85 2.3.2 A map of and for behavior: The behaviorist approach  85 2.3.3 The metaphor of culture as text: The semiotic approach  86 2.4 Culture is a verb  87 3. Culture as ideology: From a consensual to a differentiated view of culture  87 3.1 The construction of the cultural other  88 3.2 The politics of cultural difference in stratified multicultural societies  89 4. Doing cultural analysis: A critique of the method  91 4.1 The dominance of intensive fieldwork  92 4.2 The subjectivity of participant observation  92 4.3 Doing and writing ethnography as interpretation and invention  93 5. Culture as communication: The ‘linguistic’ turn  95 5.1 One language — one culture? 95 5.2 Language as a means to cultural resources  96 6. Cross-cultural and intercultural analysis in pragmatic research  97 6.1 Cross-cultural pragmatic research: The culture principle  97 6.2 Intercultural pragmatics: Towards a critical reading  99 7. Conclusion: The pragmatics of recovering culture  99


Elicitation Gunter Senft


Ethnography Michael Agar 1. Introduction  110 2. Ethnography by example  112 3. Ethnography by contrast  117 4. Discourse  118



VIII Handbook of Pragmatics Highlights

Ethnography of speaking Kristine L. Fitch & Gerry Philipsen 1. Development and main characteristics  121 2. Areas of inquiry  123 3. Issues and debates  125


Fieldwork Gunter Senft


Firthian linguistics Jan-Ola Östman & Anne-Marie Simon-Vandenbergen


Folk pragmatics Nancy Niedzielski & Dennis R. Preston 1. Folk linguistics  146 2. Folk pragmatics  149 3. Directions for future research  151


Honorifics Judith T. Irvine 1. Definitions  156 2. Dimensions of variation and comparison  157 2.1 Linguistic forms  157 2.2 Sources of honorific expressions  161 2.3 Honorification roles and participant roles  163 2.4 Deployment and ideology  165 3. Pragmatics and semantics  167 4. Trends in theory and research  168


Wilhelm von Humboldt Brigitte Nerlich & David D. Clarke 1. Life  173 2. Humboldt’s philosophy of language  174 3. Language and thought  175 4. Language and world  177 5. Language and languages (superior and inferior ones) 6. Language, culture, and creativity  179 7. Language, dialogue, and pronouns  180 8. Humboldt and the Idéologues  181



Table of contents

Intercultural communication Volker Hinnenkamp 1. Background: Language and culture  185 2. Intercultural communication: The emergence of a field of inquiry  186 3. The concept of culture  188 4. Loci of culture-in-communication  189 5. Methodological (sub)discourses of the field  192 6. Conclusion  197


Interview Charles Briggs 1. Introduction  202 2. The construction of interviews and questionnaires  202 3. Research on interviewing  203 4. Pragmatic underpinnings of interviews  204 5. Interviewing and institutional authority  205 6. Conclusion  207


Bronislaw Kasper Malinowski Gunter Senft 1. Biographical sketch  210 2. The study of culture  214 3. Fieldwork  215 4. Theory of language  217 5. An appraisal  220


Phatic communion Gunter Senft


Edward Sapir Jeroen Vermeulen 1. Introduction  234 2. Biography  235 3. Sapir on language, culture and personality  237 3.1 Language  238 3.1.1 Americanist text tradition  238 3.1.2 Linguistic form  239




Handbook of Pragmatics Highlights


3.1.3 Form-feeling  240 3.1.4 Linguistic relativity  241 3.2 Culture and personality  242 Sapir’s importance to pragmatics  244

Taxonomy Robert E. MacLaury†


Benjamin Lee Whorf Penny Lee 1. Introduction  256 2. Whorf ’s life and work  258 3. Whorf ’s perspective on linguistics  260 4. The linguistic relativity principle  263 5. Whorf ’s influence  265




Preface to the series

In 1995, the first installments of the Handbook of Pragmatics (HoP) were published. The HoP was to be one of the major tools of the International Pragmatics Association (IPrA) to achieve its goals (i) of disseminating knowledge about pragmatic aspects of language, (ii) of stimulating various fields of application by making this knowledge accessible to an interdisciplinary community of scholars approaching the same general subject area from different points of view and with different methodologies, and (iii) of finding, in the process, a significant degree of theoretical coherence. The HoP approaches pragmatics as the cognitive, social, and cultural science of language and communication. Its ambition is to provide a practical and theoretical tool for achieving coherence in the discipline, for achieving cross-disciplinary intelligibility in a necessarily diversified field of scholarship. It was therefore designed to provide easy access for scholars with widely divergent backgrounds but with converging interests in the use and functioning of language, in the topics, traditions, and methods which, together, make up the broadly conceived field of pragmatics. As it was also meant to provide a stateof-the-art report, a flexible publishing format was needed. This is why the print version took the form of a background manual followed by annual loose-leaf installments, enabling the creation of a continuously updatable and expandable reference work. The flexibility of this format vastly increased with the introduction of an online version, the Handbook of Pragmatics Online (see While the HoP and the HoP-online continue to provide state-of-the-art information for students and researchers interested in the science of language use, this new series of Handbook of Pragmatics Highlights focuses on the most salient topics in the field of pragmatics, thus dividing its wide interdisciplinary spectrum in a transparent and manageable way. The series contains a total of ten volumes around the following themes: –– –– –– –– –– –– ––

Key notions for pragmatics Philosophical perspectives Grammar, meaning and pragmatics Cognition and pragmatics Society and language use Culture and language use The pragmatics of variation and change


Handbook of Pragmatics Highlights

–– The pragmatics of interaction –– Discursive pragmatics –– Pragmatics in practice This topically organized series of paperbacks, each starting with an up-to-date overview of its field of interest, each brings together some 12–20 of the most pertinent HoP entries in its respective field. They are intended to make sure that students and researchers alike, whether their interests are predominantly philosophical, cognitive, grammatical, social, cultural, variational, interactional, or discursive, can always have the most relevant encyclo­pedic articles at their fingertips. Affordability, topical organization and selectivity also turn these books into practical teaching tools which can be used as reading materials for a wide range of pragmatics-related linguistics courses. With this endeavor, we hope to make a further contribution to the goals underlying the HoP project when it was first conceived in the early 1990’s.  

Jan-Ola Östman (University of Helsinki) & Jef Verschueren (University of Antwerp)


A project of the HoP type cannot be successfully started, let alone completed, without the help of dozens, even hundreds of scholars. First of all, there are the authors themselves, who sometimes had to work under extreme conditions of time pressure. Further, most members of the IPrA Consultation Board have occasionally, and some repeatedly, been called upon to review contribu­tions. Innumerable additional scholars were thanked in the initial versions of handbook entries. All this makes the Handbook of Pragmatics a truly joint endeavor by the pragmatics community world-wide. We are greatly indebted to you all. We do want to specifically mention the important contributions over the years of three scholars: the co-editors of the Manual and the first eight annual installments, Jan Blommaert and Chris Bulcaen were central to the realization of the project, and so was our editorial collaborator over the last four years, Eline Versluys. Our sincerest thanks to all of them. The Handbook of Pragmatics project is being carried out in the framework of the research program of the IPrA Research Center / Antwerp Center for Pragmatics at the University of Antwerp. We are indebted to the university for providing an environment that facilitates and nurtures our work.  

Jan-Ola Östman (University of Helsinki) & Jef Verschueren (University of Antwerp)

Introduction Gunter Senft Max Planck Institute for Psycholinguistics, Nijmegen

Anthropology is the discipline which is centrally concerned with the concept of culture (see Sarangi, this volume), and linguistics is the discipline which is centrally concerned with language, languages and how their speakers use them. Bronislaw Malinowski (1920:78) pointed out that “linguistics without ethnography would fare as badly as ethnography without the light thrown in it by language” and Charles F. Hockett (1973: 675) varied this theme emphasizing that “[l]inguistics without anthropology is sterile; anthropology without linguistics is blind”. In what follows I first characterize Wilhelm von Humboldt’s contribution to the study of culture and language use – after a brief reference to Johann Gottfried Herder. Then I discuss some of the ideas of scholars like Malinowski, Boas, Sapir, Whorf and others who have been shaping the field and briefly outline some of its most important traditions, methods and topics. I finish with presenting some examples of interdisciplinary research which dealt, or still deals, with the interrelationship between language use, culture and cognition (see also Senft: 2006a). Before I start, however, I would like to point out that for me the overall topic of this volume – culture and language use – defines the research domain of the subdiscipline “anthropological linguistics” in its broad sense.1 In Johann Gottfried Herder’s prize-winning essay for the Berlin Academy of Sciences on “The Origin of Language” [“Über den Ursprung der Sprache”] we read the following rather enthusiastic passage on comparative linguistics: The analogies of all savage languages confirm my statement: every language is wasteful and poor in its own way, all in a specific manner. As the Arab has so many words for stone, camel, sword, snake (things with which he lives), so is the Ceylon language rich in compliments – according to the inclination of its people, its titles and its verbal pomp [...]. In Siam there are eight manners for saying “I” and “we”, depending on whether the master is talking with the servant or the servant with the master. The language of the savage Caribs is almost divided in two languages for women and men, and both have

1.  I use and understand the term “anthropological linguistics” as synonymous with the terms “ethnolinguistics” and “linguistic anthropology”. It goes without saying, however, that these terms can be used to signal different starting points for approaching the interdiscipline and for indexing the status of both disciplines within the interdisciplinary enterprise. See Foley (1997) and Duranti (1997). For a more narrow understanding of the tradition of “anthropological linguistics” from a North American point of view see Blount (this volume).


Gunter Senft

different names for the most commonest things – bed, moon, sun, bow – what a surplus of synonyms! But these Caribs have only four terms for colours, with which they have to refer to all other colours – what poverty! The Hurons have always a double verb for an animate and inanimate object; thus “see” in “to see a stone” and “see” in “to see a person” are always two different expressions; just follow this principle for the whole nature – what wealth! In the Peruvian main language the sexes refer to each other in such a strangely separate manner that the sister of the brother and the sister of the sister, the father’s child and the mother’s child are called differently; nevertheless, this language has no real plural! Each of these systems of synonyms is so closely related with custom, character and origin of the nation; but the inventive human mind characterizes itself everywhere. (Herder 1770: 149f. [my translation, G.S.]).2

Herder clearly understands language as the expression and manifestation of a speech community’s culture – and he thus addresses both linguists and anthropologists. Herder refers to examples, topics and themes that have always been, and still are, of great interest for anthropological linguistics. Moreover, in his last sentence he even emphasizes the importance cognitive sciences have for anthropological linguistics – a fact which we have realized within the humanities just recently again after all these years and which we now celebrate as the “cognitive turn” especially within linguistics and ethnology. Shortly after this general, though programmatic passage Herder directly addresses us linguists and drives it home to us that it cannot be sufficient to describe a language according to its formal rules in a grammar only. If we do so, we […] snatch its formalities [but we] have lost its spirit, we learn their language and do not feel the living world of their thoughts [...] There the blunt laws of the grammarians are said to be the divine which we venerate, and we forget the true

2.  The quoted passage reads in the original as follows: “Die Analogien aller wilden Sprachen bestätigen meinen Satz: jede ist auf ihre Weise verschwenderisch und dürftig, nur alle auf eigne Art. Wenn der Araber für Stein, Kamel, Schwert, Schlange (Dinge, unter denen er lebt!) so viel Wörter hat, so ist die ceylanische Sprache, den Neigungen ihres Volks gemäß, reich an Schmeicheleien, Titeln und Wortgepränge [...]. In Siam gibt es achterlei Manieren, ich und wir zu sagen, nachdem der Herr mit dem Knechte oder der Knecht mit dem Herren redet. Die Sprache der wilden Kariben ist beinahe in zwo Sprachen der Weiber und Männer verteilt, und die gemeinsten Sachen, Bette, Mond, Sonne, Bogen, benennen beide anders – welch ein Überfluß von Synonymen! Und doch haben eben diese Kariben nur vier Wörter für die Farben, auf die sie alle anderen beziehen müssen – welche Armut! Die Huronen haben jedesmal ein doppeltes Verbum für eine beseelte und unbeseelte Sache, so daß Sehen bei “einen Stein sehen” und Sehen bei “einen Menschen sehen” immer zween verschiedene Ausdrücke sind; man verfolge das durch die ganze Natur – welch ein Reichtum! In der peruanischen Hauptsprache nennen sich die Geschlechter so sonderbar abgetrennt, daß die Schwester des Bruders und die Schwester der Schwester, das Kind des Vaters und der Mutter ganz verschieden heißt; und doch hat eben diese Sprache keinen wahren Pluralismus! Jede dieser Synonymien hängt so sehr mit Sitte, Charakter und Ursprung des Volks zusammen; Überall aber charakterisiert sich der erfindende menschliche Geist”. (Herder 1770: 149f.)


divine nature of language, which formed itself in its heart with the human mind. (Herder 1770: 173 [my translation, G.S.]).3

And with this position Herder again seems to be of immediate interest for us, as recent debates within linguistics on endangered languages and their adequate documentation reveal. Without reference to Herder, but certainly under the influence of the ideas within the contemporary philosophy of language (Heeschen 1972: 29ff) Wilhelm von Humboldt developed his conception of language.4 As Brigitte Nerlich and David D. Clarke (this volume) point out, Humboldt’s thinking rooted in the philosophical tradition which begins with Leibniz; he was also strongly influenced by Kant and Fichte. For Humboldt (1830–1835: 426) language is ‘the creative organ of thought’ (“das bildende Organ des Gedanken”). The difference between languages represents ‘not only one of sounds and signs, but a difference of world views itself ’ (“nicht nur eine von Schällen und Zeichen, sondern eine Verschiedenheit der Weltansichten selbst” (Humboldt 1820: 20)). And ‘in every language [rests] a specific world view’ (“in jeder Sprache [liegt] eine eigenthümliche Weltansicht” (Humboldt 1830–1835: 224, 434)). Heeschen clearly worked out where Humboldt sees the foundation for the difference of languages: During the formation of language the given objective as well as the subjective point of view becomes valid – the point of view the speaker actually takes with respect to the world; a language puts down a world view in its vocabulary. But not just the material the world spreads out in front of the senses is processed subjectively, the forms of understanding, too, cannot but appear in subjective refraction in the language; thus the grammatical view causes even bigger differences, because it creates the distinction of whole word groups, paradigms, and syntactic categories. At a certain moment the point is reached where the by now complete language gains power which is independent from the individual and which – on the basis of world view and grammatical form – predetermines the direction, in which the individual can move. In the same measure, in

3.  This reads in the original: “[Wir] haschen [zwar] ihre Formalitäten [aber wir] haben ihren Geist verloren, wir lernen ihre Sprache und fühlen nicht die lebendige Welt ihrer Gedanken [...] Da sollen die stumpfen Gesetze der Grammatiker das Göttliche sein, was wir verehren, und vergessen die wahre göttliche Sprachnatur, die sich in ihrem Herzen mit dem menschlichen Geiste bildete” (Herder 1770: 173). 4.  Humboldt liked Herder’s poems, but it should be pointed out that he did not think high of Herder as a philosopher. It is obvious that Herder did not know much about foreign languages; moreover he had no idea of how to analyse and interpret linguistic data (see Wirrer 1996). Thus, as Heeschen (1972: 35) points out, Humboldt cannot be seen at all as being in the succession of Herder. Humboldt was highly influenced by Kant (see e.g., Humboldt 1820: 3f.); however, he refers to him explicitly only once in his linguistic writings (Humboldt 1830–1835: 593; see Heeschen 1972: 36).



Gunter Senft

which the nations have created the language, the created reacts on them in a paralysing or in an inspiring way. (Heeschen 1972: 255f. [my translation, G.S.])5

Humboldt (1830–1835: 434) describes this situation as follows: With the same act with which [Man] spins language out of himself he spins himself into this language, and every language draws a circle around the nation it belongs to. Getting out of this circle is only possible by stepping over into the circle of another language at the same time. (Humboldt 1830–1835: 434 [my translation, G.S.])6

Thus, as Nerlich and Clarke (this volume) point out, Humboldt came up with a rather complex language relativity thesis. However, his relativity thesis does not imply that humans are captives of their specific languages. On the contrary, ‘the system coagulated to world view and grammar provides the individual with material for a new wealth of ideas and forms’ (“das zum Weltbild und zur Grammatik geronnene System wird dem Individuum Material zu neuer Ideen- und Formenfülle” (Heeschen 1972: 255) [my translation, G. S.]). Language improves in the course of its development as an ‘organ of thinking’ in the same way as it deepens thinking. Works of literature and science come into being and allow for an additional wealth of innovative ideas. Spontaneity of the creative powers become apparent in the use of language [...] in the dialectics between primary, sensual world-outlook, world view and poetic individual world-outlook the freedom of language use – language as ‘energeia’ – gains its space. (Heeschen 1972: 256 [my translation, G.S.])7

5.  This reads in the original: “Während der Bildung der Sprache macht sich sowohl der objektiv vorgegebene wie auch der subjektive Standpunkt geltend, den der Sprecher zur Welt tatsächlich einnimmt; eine Weltansicht legt die Sprache in ihrem Wortschatz nieder. Aber nicht nur das Material, das die Welt vor den Sinnen ausbreitet, wird subjektiv verarbeitet, auch die Formen des Verstandes können nicht anders als in subjektiver Brechung in der Sprache erscheinen; die grammatische Ansicht bedingt so noch viel größere Unterschiede, weil sie gleich die Verschiedenheit ganzer Wortgruppen, Paradigmen und syntaktischer Kategorien schafft. Einmal ist der Punkt erreicht, wo die nun vollständige Sprache eine vom Individuum unabhängige Macht gewinnt und durch Weltbild und grammatischen Bau Bahn und Richtung vorherbestimmt, in der sich der Einzelne bewegen kann. Denn in dem Maße, in dem die Nationen die Sprache geschaffen haben, wirkt das Geschaffene auf sie lähmend oder begeisternd zurück” (Heeschen 1972: 255f.). 6.  This reads in the original: “Durch denselben Act, vermöge dessen [der Mensch] die Sprache aus sich heraus spinnt, spinnt er sich in dieselbe ein, und jede zieht um das Volk, welchem sie angehört, einen Kreis, aus dem es nur insofern hinauszugehen möglich ist, als man zugleich in den Kreis einer anderen hinübertritt” (Humboldt 1830–1835: 434). 7.  This reads in the original: “[Die] Spontaneität der schöpferischen Kräfte erweist sich im Gebrauch der Sprache [...] in der Dialektik von primärer, sinnlicher Weltansicht, Weltbild und poetischer individueller Weltansicht verschafft sich die Freiheit des Sprachgebrauchs, die Sprache als Energeia, Raum” (Heeschen 1972: 256).


Insights like this one justify why Nerlich and Clarke (this volume) characterize Humboldt’s theory of language as a ‘pragmatic’ theory in which the notion of the ‘act of speaking’ is central. They point out that Humboldt always emphasized the pragmatic grounding of language: language is not something which exists independently of its users. Trabant (1986) has shown that this insight has far-reaching consequences – for both linguistics and anthropology: For only by researching languages as they are used by individuals, in texts – the place where the physiological power of language meets the speaking individuals’ dynamic force – linguistics can answer that question which is the general basis for Humboldt’s anthropological search, namely the question of what is imagination and genius, the question of how man creates something new. (Trabant 1986: 203 [my translation, G.S.])8

Researchers who exclude in their linguistic analyses the aspect of actual language use within a specific speech community or who consider it marginal, linguists who are only interested in the structure of a language from a system linguistics point of view, scientists who understand rules and algorithms of grammar as the sole object of their research have to bear the scorn and the scolding expressed in the xenion “The linguist” – which was meant for Adelung (see Trabant 1986: 197): You may anatomize language, however, its carcass only; spirit and life fleetingly escape the crude scalpel. [my translation, G.S.]9

8.  This reads in the original: “Denn nur indem sie die Sprachen in ihrem Gebrauch durch die Individuen untersucht, in den Texten, diesem Ort der Begegnung der physiologischen Macht der Sprache und der dynamischen Gewalt der sprechenden Individuen, kann die Sprachwissenschaft jene der anthropologischen Suche Humboldts generell zugrundeliegende Frage beantworten, die Frage nämlich nach der Einbildungskraft und dem Genie, die Frage, wie der Mensch Neues schafft” (Trabant 1986: 203). 9.  This reads in the original: “Anatomieren magst du die Sprache, doch nur ihr Kadaver; Geist und Leben entschlüpft flüchtig dem groben Skalpell.” A “xenion” is a satirical poem that consists of two lines only. Johann Wolfgang Goethe and Friedrich Schiller wrote and published this xenion in their “Musenalmanach für das Jahr 1797”. Johann Christoph Adelung (1732–1806), the target of this poem, was a highly influential German-language scholar.



Gunter Senft

On the basis of Humboldt’s understanding of language it has to be emphasized, however, that it is the sense of language itself that remonstrates against being treated as a dead skeleton: The sense for language, the sense for that sense which Humboldt defined as the individual organ of thinking and as the sounding instrument of unison, linguistic phantasy stimulated by genius and study seems to require the mediation meant with the word “human” if linguistics should make sense. (Trabant 1986: 207 [my translation, G.S.])10

Language has to be seen first of all as a cultural achievement and as a cultural tool. Language is a mirror of the culture of its speech community. And – as Nerlich and Clarke (this volume) in their critical, condensed aperçu on Humboldt’s pragmatic philosophy of language point out, Humboldt emphasized the strong interrelationship between language, culture and cognition. Thus it is evident that we can trace the request for, and the foundation of “anthropological linguistics” back to the 18th and 19th century. In the 20th century we find Bronislaw Malinowski, one of the founders of modern cultural anthropology, as the great apologist of anthropological linguistics (see Young 2004, Senft, this volume, Senft 2006b). In 1920 he stated explicitly that “[...] there is an urgent need for an Ethnolinguistic theory, a theory for the guidance of linguistic research to be done among natives and in connection with ethnographic study [...]” (Malinowski 1920: 69). And in his famous programmatic introduction to his monograph “Argonauts of the Western Pacific” he emphasized the following: [...] the goal of ethnographic field-work must be approached through three avenues:

1. The organisation of the tribe, and the anatomy of its culture must be recorded in firm clear outline. The method of concrete statistical documentation is the means through which such an outline has to be given. 2. Within this frame, the imponderabilia of actual life, and the type of behaviour have to be filled in. They have to be collected through minute, detailed observations, in the form of some sort of ethnographic diary, made possible by close contact with native life. 3. A collection of ethnographic statements, characteristic narratives, typical utterances, items of folk-lore and magical formulae has to be given as a corpus inscriptionum, as documents of native mentality.

10.  This reads in the original: “Der Sinn für die Sprache, der Sinn für jenen Sinn, der als individuelles Organ des Denkens und als klingendes Instrument der Über-Ein-Stimmung von Humboldt bestimmt wurde, die durch Genie und Studium angeregte linguistische Phantasie, scheint nämlich die mit dem Wort des Menschlichen gemeinte Vermittlung zu fordern, wenn Sprachwissenschaft Sinn haben soll” (Trabant 1986: 207).


These three lines of approach lead to the final goal, of which an Ethnographer should never lose sight. This goal is, briefly, to grasp the native’s point of view, his relation to life, to realise his vision of his world. (Malinowski 1922: 24f.)

Here Malinowski drives it home to his discipline that it just cannot pursue its research interests without linguistics.11 As Senft (this volume) points out, Malinowski understood language ‘in its primitive function’ as a mode of behaviour, as a mode of action, rather than as a countersign of thought – and he illustrated this understanding by the concept he called “phatic communion”, a type of speech which serves to establish personal bonds but does not serve any purpose of communicating ideas (for a critical discussion of this concept see Senft, this volume).12 In Malinowski’s pragmatic theory of meaning the insight that the meaning of a word lies in its use is central. For him meaning is function within context. Understanding context requires that the researcher interested in the relationship between culture and language use has to do fieldwork (see Senft, this volume). Like Boas (see below) Malinowski insisted that adequate data could only be collected in the field. Only participant observation combined with other tools of data gathering, like for example interviews (see Briggs, this volume) or specific elicitation procedures (see Senft, this volume), enables the researcher ‘to grasp the native’s point of view’. Thus, all fieldworkers and ethnographers (see Agar, this volume) set out to learn meanings and contexts which lie outside the concepts and habits of prior experience.13 Malinowski’s aim to understand the interaction between culture and meaning (see especially Malinowski 1923), his theory of context of situation which bound language to the situational moments and cultural contexts of use (see Agar, this volume) also laid the foundation for the ‘British school’ of linguistics, also know as ‘Firthian linguistics’ (see Östman & Simon-Vandenbergen, this volume). This approach tried to tie down context and make it operationally approachable. Firth strongly advocated for a linguistics which studies language as a form of meaningful human behavior in society. With his approach he was taking initial steps into a new field of linguistics, namely pragmatics.

11.  For a detailed description of Malinowski’s linguistic interests and his ideas for a theory of language see also Senft 2005. 12.  By the way, Malinowski’s concept of “phatic communion” became central for Robin Dunbar’s research on the origin of language (see, e.g., Dunbar 1996). 13.  The contributions of Cliff Goddard to this volume present techniques for the analysis of word meanings (see his article on “Componential Analysis”) and for articulating culture-specific norms, values and practises (see his article on “Cultural Scripts”).



Gunter Senft

In 1975 Michael Silverstein repeats Malinowski’s request for integrating linguistics into anthropology in his discussion of the relationship between the two disciplines. Lamenting about the actual state of the art of both fields he states the following: On the one hand [...] the pursuit of anthropological studies without the use and investigation of the native language of the people being studied is unthinkable in theory, although all too frequently the case in practice. On the other hand, the pursuit of grammatical studies without the understanding of the function of the speech forms being studied is actually impossible in theory, although again linguists have simply assumed that this is the correct and necessary approach. (Silverstein 1975: 157f.)

But why then was anthropological linguistics or linguistic anthropology confronted at all with problems concerning its legitimation? In his attempt to reconstruct the history of anthropological linguistics, Agar points out that Franz Boas, the founder of American anthropology, also understood language as an unalterable prerequisite for his research: “[...] language was then [...] a part of anthropological field work, and the point of fieldwork was to get to culture. Culture was the destination; language was the path; grammar and dictionary marked the trail” (Agar 1994: 49; see also Agar, this volume). Or, as Regna Darnell (this volume) has it, Boas saw language as a symbolic form through which culture becomes accessible to study. Like Humboldt – who quite strongly influenced him – Boas was convinced that languages have an inner form, that they deserve to be described in their own terms. With his recognition that Indo-European categories actually distorted languages on which they were imposed began cultural relativism, as Darnell (this volume) notes. She also states that the so-called Sapir-Whorf hypothesis of the relationship of habitual thought to linguistic categories (see below) has its roots in the Boasian insistence on the unique perceptual patterning of each language. However, Boas is also one of the founders of descriptive-structural linguistics that found its first important expression in Leonhard Bloomfield’s (1935) monograph “Language”. With this book linguistics in the USA raised from a subdiscipline of ethnography, from a tool for anthropological field research to an independent discipline in its own right. However, the new discipline was no longer interested in the actual use of language and in language in its ethnographic context. Linguistics now was defined as the study of the sound system and the grammar of a language; there is no room for semantics in Bloomfield’s “Language” – this field is delegated to psychology and ‘science’. The further development of American structural linguistics led to the complete neglect of actual users of language (or better: English) and of contexts, in which speakers use their language (resp. English).14 In 1965 Noam Chomsky proclaimed

14.  Bickerton (1971: 457) polemically refers to Chomskyan linguistics as “linguistics without a speaker”. See Senft (1982: 1–5).


in his famous book “Aspects of the Theory of Syntax” the “ideal speaker/listener in a completely homogenous speech community” (Chomsky 1965: 3) whose language competence linguists describe and analyse via introspection. Linguists attempting to describe language in the lowlands of empirical research were disregarded within this paradigm. The various reformulations of Chomsky’s theory can still claim important status (and many positions at universities) in linguistics, probably also because ‘the analysis of a few hundred examples from languages one hardly understands is just a dirty business compared with the development of clean theories’ (see Klein 1979: 95; also Senft: 1991: 43–45). But let us come back once more to Franz Boas. He was not only one of the founders of American structuralism. He was also the influential teacher of Edward Sapir, who – together with Benjamin Lee Whorf – developed an interest for researching the indigenous languages of the North-American Indians. Edward Sapir kind of impersonated Malinowski’s ideal fieldworker: he was president of both the American Anthropological Association and the Linguistic Society of America! Like Humboldt Sapir was convinced that language is essentially dynamic and he spoke of a language’s genius, like Malinowski he insisted in studying language in the context of its use, and like Boas he was convinced that every language has its own unique way of conceptualizing social reality (see Vermeulen, this volume). Whorf also investigated language as a cultural phenomenon. With Malinowski he shares the conviction that language is a form of behavior, however for Whorf this behavior is of significantly mental nature. Thus, he theorized about the role of language in cognition and the place of ‘linguistic thinking’ in our understanding of language as a whole (see Lee, this volume). Whorf ’s cooperation with Sapir, their descriptions and analyses of indigenous languages of the NorthAmerican Indians, and Whorf ’s perspectives on linguistics resulted in the formulation of the linguistic relativity principle: We are [...] introduced to a new principle of relativity, which holds that all observers are not led by the same physical evidence to the same picture of the universe, unless their linguistic backgrounds are similar, or can in some ways be calibrated. (Whorf 1958: 5)

This idea with respect to the interrelationship between language and thought became famous as the “Sapir-Whorf hypothesis” – as the linguistic relativity principle was referred to in the 1950s (see Lee, this volume, also Lee: 1996) – and it is striking to see many parallels in the understanding of the relationship of language, experience and thought between Herder and Humboldt on the one hand and Sapir and Whorf on the other.15

15.  Note again that this does neither mean nor imply that there is any kind of a clear line of succession with respect to the scholars mentioned here. Humboldt’s ideas are much more refined and differentiated than Whorf ’s.



Gunter Senft

With research instigated by the Sapir-Whorf hypothesis, with the rise of American sociolinguistics and its efforts to understand, describe and analyse variation in language (with William Labov as probably its most important representative), and with the research within the “ethnography of speaking”-paradigm founded by John Gumperz and Dell Hymes (see Fitch and Philipsen, this volume; also Niedzielski and Preston, this volume) who rooted this tradition in theories and insights presented by Malinowski, Boas and Sapir anthropological linguistics gradually won recognition again – not only within American (see Blount, this volume) but also within European linguistics. With the rise of sociolinguistics dialectology – a linguistic subdiscipline traditionally rather open for anthropological linguistic ideas – regained importance by concentrating much more on researching spoken language in everyday contexts and use16 than on developing language atlases, finding isoglosses, and collecting most typical and archaic expressions for various dialects. Moreover, this sociolinguistic turn and the reception of the “ethnography of speaking” paradigm also revived the study of “intercultural communication” again – a field which gets more and more important in our times of globalization (see Hinnenkamp, this volume). Finally, the reception of Austin’s and Searle’s ideas with respect to speech act theory resulted in the strengthening of “pragmatics” as the subdiscipline of linguistics that researches rules and regulations which determine the choice of specific, situation-adequate varieties or registers in the social interaction of speakers.17 Michael Silverstein points out that researching the function of speech behavior is one of the central aims of anthropological linguistics. In sharp contrast to the Chomskyan “mainstream”-linguistics of that time he states […] that the study of grammar cannot in principle be carried on in any serious way until we tackle the ethnographic description of the canons of use of the messages corresponding to sentences. Reformulating this result, we may say that grammar is open-ended, not closed, and a part of the statement of the total meaning of a sentence is a statement of the rules of use that are involved in proper indexicality of elements of the message. This means, again, that if we call the ‘function’ of a sentence the way in which the corresponding message depends on the context of situation, then the determination of the function of the sentence, independent of its propositional value, is a necessary step in any linguistic analysis. Thus a theory of rules of use, in terms of social variables of the speech situation and dependent message form, is an integral part of a grammatical description of the abstract sentences underlying them. Rules of use depend on ethnographic description, that is, on analysis of cultural behavior of people in a society. Thus, at one level we can analyze sentences as the embodiment of propositions, or of linguistic meanings more generally; at another level, which is always implied

16.  See for example Ruoff (1972); also Senft (1982: 3, 163–173). 17.  The rise of pragmatics was also documented by the foundation of the “International Pragmatics Association” in 1986.


in any grammatical description, we must analyze messages as linguistic behavior which is part of culture. [...] a valid description of a language by grammar demands description of the rules of use in speech situations that are structured by, and index, the variables of cultures. (Silverstein 1975: 167)

The close relationship between anthropological linguistics and pragmatics is obvious. Especially recent developments within pragmatics and anthropological linguistics allow Bill Foley (1997: 29) to state that “[...] the boundary between pragmatics and anthropological linguistics or sociolinguistics is impossible to draw at present [...]”. Thus, in linguistics there are good and well grounded hopes that anthropological linguistics – “[...] this somewhat neglected topic [...]” (Trudgill 1997: xiii) – finally gains its due importance. The development of cognitive anthropology was certainly responsible for the fact that anthropological linguistics could not be ousted in anthropology. Agar (1994: 81) characterizes the approach of cognitive anthropologists as follows: “The cognitive anthropologists took Whorf ’s basic idea and developed a way to discover culture, a way to use the surface of language to get and document the culture that it expressed. They changed Whorf from a theory to a method”.18 Stephen Levinson (this volume) defines cognitive anthropology as “the comparative study of cognition in its full cultural and linguistic context [...]”, as “the indepth understanding of conceptual domains, primarily through language, together with comparison across unrelated languages and cultures”. For Levinson the aim to research the implications of linguistic diversity to thinking also goes back to Humboldt, Wundt, Boas, Sapir and Whorf; moreover, cognitive anthropology owes much to the “ethnography of speaking” paradigm and the “ethnoscience” approach of the 1950s. In 1991 the German Max-Planck-Society founded the “Cognitive Anthropology Research Group” with Stephen Levinson at its director. Since 1991 a group of linguists, anthropologists, and psychologists has been researching the interrelationship between language, culture and cognition in a number of semantic domains, like for example “Space”. For the group, which was integrated into the Max-Planck-Institute for Psycholinguistics as the Department of “Language and Cognition” in 1996, the study of cultural and linguistic variation serves as the starting point for researching central epistemological questions. The group’s research interests center on questions like the following: Are there differences between fundamental

18.  This is, of course, somewhat of an exaggeration. One should not forget – as Robert E. Maclaury reminds us in his contribution on “Taxonomy” that ever since Boas we can observe an anthropological linguistic interest in categorization. Many anthropologists and linguists were and still are concerned with the crosscultural comparison of ways that lexically labelled categories are internally organized and related to each other within a domain. Especially cognitive anthropological research has been dealing for a long time with ethnobiological taxonomies, with componential analyses of kinship systems, with inventories of (basic) color terms and body part terms.



Gunter Senft

semantic parameters in specific domains of the lexicon of different languages? Are these differences dependent on cultural factors or can they be referred back to cultural phenomena? Can one infer on the basis of these lexical-semantic differences that there are also differences between various languages with respect to the cognitive conceptualizations of their speakers or, more generally, that there are differences in the realm of cognitive processes that are decisive for the speakers of these languages? Or to put it more bluntly: Does a specific language influence the ways its speakers’ think? (see Senft 1994: 414). Among the colleagues who closely cooperate with this group are Jürg Wassmann and Pierre Dasen, who have been working together as an interdisciplinary team for a long time. They have certainly contributed crucially to the cognitive turn not only within German anthropology (see e.g., Wassmann & Dasen 1993; 1998). In their cooperation, the psychologist and the anthropologist completely agree that linguistic research is inevitable for solving certain problems – although Wassmann (1993) showed that “actions speak louder than words” – at least sometimes. In addition, we should not forget that much of the merit to understand language, culture and cognition again as interdependent domains of one interdiscipline and to research its mutual dependencies is due to psycholinguists and representatives of the “cross-cultural psychology” subdiscipline who refer in their publications explicitly to pioneers like Herder, Humboldt, Wundt, Boas, Malinowski and Lévi-Strauss (see Slobin 1967; Berry & Dasen 1974: 6; Lonner & Triandis 1980: 1; Berry 1980: 7; Klineberg 1980). The linguists and psycholinguists around Dan Slobin, Susan Ervin-Tripp and John Gumperz in Berkeley were especially interested in the “cross-linguistic study of the acquisition of communicative competence”. The representatives of the “crosscultural psychology” – especially followers of the psychology of Jean Piaget’s and Bärbel Inhelder’s Geneve school like Pierre Dasen, scientists like Gustav Jahoda, and some of Jerome Bruner’s associates at the Center of Cognitive Studies at Harvard University such as Patricia Greenfield, Michael Cole, Sylvia Scribner and their co-workers took the interdependence between language, culture and cognition for granted. They were convinced that psychological hypotheses – especially hypotheses in developmental psychology – proposed in researching populations within one culture and one language community could only claim to be general and universal if they were tested in intercultural research. Psycholinguists, linguists and anthropologists cooperated in this interdiscipline that was understood as a subdiscipline of psychology.19 With their interdisciplinary orientation they contributed fundamentally to questions of developmental psychology and to research on “learning, memory, verbal-logical problem

19.  For further representatives of this interdiscipline see the contributions to the six volumes of the “Handbook of Cross-Cultural Psychology” edited by Triandis and others. See also Senft (2003).


solving, and logical inference” especially in the 1980s; however, their findings are still of high relevance for psychologists, (psycho-) linguists and (cognitive) anthropologists. The “cross-cultural studies on the acquisition of communicative competence” initiated by Slobin and others developed into the “crosslinguistic study of language acquisition”. This is a field extremely important for psycholinguistic language acquisition research, but also for general and comparative linguistics, for language typology and for the cognitive sciences generally.20 The crosslinguistic study of language acquisition clearly shows how much a discipline – here linguistics and psycholinguistics – can gain approaching certain questions and problems in an interdisciplinary way. Another subdiscipline that developed out of the “cross-cultural studies on the acquisition of communicative competence” is “developmental pragmatics” which was founded by Elinor Ochs and Bambi Schieffelin (1979). Ochs and Schieffelin took up the plea for interdisciplinary research. Influenced by Clifford Geertz’s (1973: 6) “thick descriptions” they developed independent methods for data collection and for the anthropological linguistic transcription of data on the verbal socialization of children in different cultures (see B. Schieffelin 1979; Ochs 1988). On the basis of the careful, subtle and sophisticated processing of the data they can minutely and in great detail analyse the complex processes of verbal socialization. Both scientists do not only cooperate with each other after their field research, they also cooperated in their field sites with anthropologists: On Samoa, Elinor Ochs worked together with Alessandro Duranti, and Bambi Schieffelin researched the Kaluli of Papua New Guinea together with Steven Feld and Edward Schieffelin. The publications of both research teams document all advantages of interdisciplinary cooperation between linguists and anthropologists (see e.g., Feld 1982; Feld & B. Schieffelin 1982; E. Schieffelin 1976; B. Schieffelin & Feld 1998; Duranti 1981; Duranti & Ochs 1996). This is also true for the research of Penelope Brown and Stephen Levinson. Their theory with respect to the universality of politeness phenomena (Brown & Levinson 1978) caused many case studies and comparative research on this topic – especially within linguistic pragmatics, but also in anthropology and ethnology. The contributions to this handbook on “Honorifics” by Judith Irvine and on “Aisatsu” by Risako Ide present excellent overviews of specific domains within this research domain. One of the highly ambitious interdisciplinary projects, in which anthropologists and linguists cooperated for 8 years (1993–2001) not only with each other but also with demographers, historians, geologists, botanists and archeologists was the Dutch project “The Irian Jaya Studies”, a priority programme financed by the “Dutch Organisation for Scientific Research (Nederlandse Organisatie voor Wetenschappelijk Onderzoek)”

20.  See the 5 volumes edited between 1985 and 1997 by Dan Slobin and published under the title “The Crosslinguistic Study of Language Acquisition” by Erlbaum, Hillsdale.



Gunter Senft

and supervised by Wim Stokhof and Jelle Miedema in Leiden. In this comprehensive interdisciplinary project many researchers from various universities in the Netherlands and in Indonesia collaborate with each other. Moreover, they had also close links to scientists in Australia, in the USA and in Germany. The research results presented so far are absolutely spectacular (see, Miedema et al. 1999; Miedema & Reesink 2004). This project is exemplary for the tremendous benefits research programmes can derive from interdisciplinary cooperation. In 1997 the linguist William Foley published his monograph “Anthropological Linguistics – An introduction” and in the same year the anthropologist Alessandro Duranti published his monograph “Linguistic Anthropology”. Although the titles already signal the two different starting points from which the authors approach the interdiscipline, their understanding of anthropological linguistics and/or linguistic anthropology is strikingly similar. On the one hand, Foley (1997: 3) defines anthropological linguistics as [...] that sub-field of linguistics which is concerned with the place of language in its wider social and cultural context, its role in forging and sustaining cultural practices and social structures [...]. Anthropological linguistics views language through the prism of the core anthropological concept, culture, and, as such, seeks to uncover the meaning behind the use, misuse or non-use of language, its different forms, registers and styles. It is an interpretative discipline, peeling away at language to find cultural understandings.

He understands this discipline as “the study of how humans make meanings together in social interaction through conventional transgenerational cultural and linguistic practices” (Foley 1997: 81). Duranti, on the other hand, writes: Whether or not they see themselves as doing linguistic anthropology, the researchers from whose work I extensively draw are all concerned with the study of language as a cultural resource and with speaking as a cultural practice [and] rely on ethnography as an essential element of their analyses [...] What unites them is the emphasis on communicative practices as constitutive of the culture of everyday life and a view of language as a powerful tool rather than a mirror of social realities established elsewhere [...] (Duranti 1997: xv) [...] linguistic anthropology [is] presented as the study of language as a cultural resource and speaking as a cultural practice [...]. [...] it examines language through the lenses of anthropological concerns [...] (Duranti 1997: 2, 4)

If anthropological linguists and linguistic anthropologists agree again in their general definition of their interdisciplinary domain, then this old and periodically rediscovered discipline will finally gain its due importance for both linguists and anthropologists again. If we look at the work of the anthropological linguists and linguistic and cognitive anthropologists and see what insights were gained in these interdisciplinary research


projects, how many innovations have been emanating from them, and how this interdisciplinary research reacts upon the specific disciplines involved, and if we look at the contributions to this handbook then we cannot but conclude that anthropological linguistics which researches interdisciplinarily the interrelationship between language, language use, culture and cognition is the field “where the action is”. And if someone wants to be there, or was there once, or is there simply has to agree with Hockett (1973: 675): “Linguistics without anthropology is sterile; anthropology without linguistics is blind”.

References Agar, M. (1994). Language shock – understanding the culture of conversation. William Morrow and Co. Bartsch, R. & T. Vennemann (Eds) (1975). Linguistics and neighboring disciplines. North-Holland Publishing Co. Berry, J.W. (1980). Introduction to methodology. In H.C. Triandis & J.W. Berry (Eds): 1–28. Berry, J.W. & P.R. Dasen (Eds) (1974) Culture and cognition: Readings in cross-cultural psychology. Methuan. Bickerton, D. (1971). Inherent variability and variable rules. Foundations of Language 7: 457–492. Bloomfield, L. (1935). Language. Allen & Unwin (2nd revised edition). Brown, P. & S.C. Levinson (1978). Universals in language usage: politeness phenomena. In E.N. Goody (Ed.): 56–289. Cap, P. (Ed.) (2005). Pragmatics today. Peter Lang. Chomsky, N. (1965). Aspects of the theory of syntax. MIT Press. Deutsch, W., T. Herrmann & G. Rickheit (Eds) (2003). Psycholinguistik – Ein internationales Handbuch Psycholinguistics – an international handbook. Walter de Gruyter. Dunbar, R. (1996). Grooming, gossip and the evolution of language. Faber and Faber. Duranti, A. (1981). The Samoan fono: A sociolinguistic study. Pacific Linguistics. ——— (1997). Linguistic anthropology. Cambridge University Press. Duranti, A. & E. Ochs (1996). Use and acquisition of genetive constructions in Samoan. In Dan. I. Slobin, Julie Gerhardt, Amy Kyratzis, Jiansheng Gua (Eds): 175–189. Feld, S. (1982). Sound and sentiment. Birds, weeping, poetics and song in Kaluli expression. University of Pennsylvania Press. Feld, S. & B. Schieffelin (1982). Hard words: A functional basis for Kaluli Discourse. In D. Tannen (Ed.): 350–370. Flader, D. (Ed.) (1991). Verbale Interaktion. Studien zur Empirie und Methodologie der Pragmatik. J.B. Metzler. Foley, W.A. (1997). Anthropological linguistics. An Introduction. Blackwell. Geertz, C. (1973). The interpretation of cultures. Basic Books. Goody, E.N. (Ed.) (1978). Questions and politeness:strategies in social interaction. Cambridge University Press. Heeschen, V. (1972). Die Sprachphilosophie Wilhelm von Humboldts. Unveröffentliche Dissertation: Ruhr-Universität Bochum. Herder, J.G. (1770). Über den Ursprung der Sprache. In: Herders Werke in fünf Bänden. Band 2, 91–200. Aufbau Verlag (= 1978). Hockett, C.F. (1973). Man’s Place in Nature. McGraw Hill.



Gunter Senft Humboldt, W. von (1820). Über das vergleichende Sprachstudium in Beziehung auf die verschiedenen Epochen der Sprachentwicklung. In: Wilhelm von Humboldt. Werke in fünf Bänden. Herausgegeben von A. Flitner & K. Giel. 1963. Band III. Schriften zur Sprachphilosophie. 1–25. Wissenschaftliche Buchgesellschaft. ——— (1830–1835). Über die Verschiedenheit des menschlichen Sprachbaues und ihren Einfluß auf die geistige Entwicklung des Menschengeschlechts. Berlin: Königliche Akademie der Wissenschaften. In: Wilhelm von Humboldt. Werke in fünf Bänden. Herausgegeben von A. Flitner & K. Giel. 1963. Band III. Schriften zur Sprachphilosophie. 144–367. Wissenschaftliche Buchgesellschaft. Klein, W. (1979). Untersuchungen zum Spracherwerb ausländischer Arbeiter. Tätigkeitsbereicht für die Gesamtdauer des Projekts. 1.04.1974-30.06.1979. DFG-Az.: Kl 337/1, 3, 5. Nijmegen: Mimeo. Klineberg, O. (1980). Historical perspectives: Cross-cultural psychology before 1960. In H.C. Triandis, W.W. Lambert (Eds): 31–67. Lee, P. (1996). The Whorf theory complex: a critical reconstruction. John Benjamins. Lonner, W.J. & H.C. Triandis (1980). Introduction to basic processes. In H.C. Triandis & W. Lonner, (Eds): 1–20. Maccoby, E., T.M. Newcomb & E.L. Hartley (Eds) (1958). Readings in social psychology. Holt, Rinehart & Winston. Malinowski, B. (1920). Classificatory particles in the language of Kiriwina. Bulletin of the School of Oriental Studies, London institution, Vol. I, part IV: 33–78. ——— (1922). Argonauts of the Western Pacific. Routledge & Kegan Paul. ——— (1923). The problem of meaning in primitve languages. In C.K. Ogden & I.A. Richards: Supplement I. 296–336. Miedema, J., C. Odé & R.A.C. Dam (Eds) (1999). Perspectives on the Bird’s Head of Irian Jaya, Indonesia. Proceedings of the conference, Leiden 13–17 October 1997. Rodopi. Miedema, J. & G. Reesink (2004). One head, many faces. New perspectives on the Bird’s Head Peninsula of New Guinea. KITLV Press. Ochs, E. (1988). Culture and language development. Language acquisition and language socialization in a Samoan village. Cambridge University Press. Ochs, E. & B.B. Schieffelin (Eds) (1979). Developmental pragmatics. Academic Press. Ogden, C.K. & I.A. Richards. The meaning of meaning. A study of the influence of language upon thought and of the science of symbolism. Kegan Paul, Trench, Trubner. (Fourth edition revised 1936). Ruoff, A. 1972. Grundlagen und Methoden der Untersuchung gesprochener Sprache. Niemeyer. Schieffelin, B.B. (1979). The give and take of everyday life - Language socialization of Kaluli children. Cambridge University Press. Schieffelin, B.B. & S. FELD (1998). Bosavi – English – Tok Pisin dictionary (Papua New Guinea). Pacific Linguistics. Schieffelin, E.L. (1976). The Sorrow of the lonely and the burning of the dancers. St. Martins Press. Senft, G. (1982). Sprachliche Varietät und Variation im Sprachverhalten Kaiserslauterer Metallarbeiter. Peter Lang. ——— (1991). Mahnreden auf den Trobriand Inseln. Eine Fallstudie. In D. Flader (Ed.): 27–49. ——— (1994). Ein Vorschlag, wie man standardisiert Daten zum Thema “Sprache, Kognition und Konzepte des Raumes” in verschiedenen Kulturen erheben kann. Linguistische Berichte 154: 413–429. ——— (2003). Ethnographic methods. In W. Deutsch, T. Hermann & G. Rickheit (Eds): 106–114. ——— (2005). Bronislaw Malinowski and Linguistic Pragmatics. In P. Cap, (Ed.): 139–155. ——— (2006a). Völkerkunde und Linguistik – Ein Plädoyer für interdisziplinäre Zusammenarbeit. Zeitschrift für Germanistische Linguistik 34: 87–104.


———(2006b). Review article: A biography in the strict sense of the term – Michael Young, ‘Malinowski: Odyssee of an anthropologist 1884–1920, Volume 1’. Journal of Pragmatics 38: 610–637. Silverstein, M. 1975. Linguistics and anthropology. In R. Bartsch & T. Vennemann (Eds): 157–170. Slobin, D.I. (Ed.) (1967). A field manual for cross-cultural study of the acquisition of communicative competence (Second draft, July 1967). Berkeley: University of California, ASUC Bookstore. Slobin, D.I., J. Gerhardt, A. Kyratzis & J. Gua (Eds) (1996). Essays in honor of Susan Ervin-Tripp. Erlbaum. Trabant, J. (1986). Apeliotes oder Der Sinn der Sprache. Wilhelm von Humboldts Sprach-Bild. Wilhelm Fink. Tannen, D. (Ed.) (1982). Analyzing discourse: Text and talk. Georgetown University round table on language and linguistics 1981. Georgetown University Press. Triandis; H.C. & J.W. Berry (Eds) (1980). Handbook of cross-cultural psychology. Volume 2: Methodology. Allyn and Bacon. Triandis, H.C. & W.W. Lambert (Eds) (1980). Handbook of cross-cultural psychology. Volume 1: Perspectives. Allyn and Bacon. Triandis, H.C. & W. Lonner (Eds) (1980). Handbook of cross-cultural psychology. Volume 3: Basic Processes. Allyn and Bacon. Trudgill, P. (1997). Series editor’s preface. In William Foley, Anthropological linguistics. An Introduction. p. xiii. Blackwell. Wassmann, J. (1993). When actions speak louder than words: The classification of food among the Yupno of Papua New Guinea. Quarterly Newsletter of the Laboratory of Comparative Human Cognition 15: 30–40. Wassmann, J. & P. Dasen (Eds) (1993). Alltagswissen. Der kognitive Ansatz im interdisziplinären Dialog. Universitätsverlag Freiburg (Schweiz). ——— (1998). Balinese spatial orientation: Some empirical evidence of moderate linguistic relativity. Journal of the Royal Anthropological Institute (Incorporating Man) 4. 689–711. Whorf, B.L. (1958). Science and linguistics. In E. Maccoby, T.M. Newcomb & E.L. Hartley (Eds): 1–9. Wirrer, J. (1996). Stereotypen über europäische Völker in Herders ‘Ideen zur Philosophie der Geschichte der Menschheit’. Germanoslavica III (VIII): 107–131. Young, M. 2004. Malinowski: Odyssee of an anthropologist 1884 – 1920. Yale University Press.


Aisatsu Risako Ide University of Tsukuba

1.  Introduction: The pragmatics and metapragmatics of routine formulae As aptly put by Coulmas (1981: 1), conversation is a structured activity, and a large part of it consists of enacting routines. We greet and say goodbye to one another, we introduce ourselves, we thank and apologize, we make requests, we exchange good wishes, we give advice, we seek information, etc. – all of these are conducted within a range of conventionalized, pre-patterned expressions. To take an example, knowing how to greet is an important part of the communicative competence necessary for someone to be a member of any speech community. From an ethnological perspective, greetings are means to negotiate the rights to mutual access, and they are regarded as universal conducts taking place regardless of differences in culture or society. However, close analysis of these conducts within their own cultural system reveals that greetings are often rule-governed sequences, and that rituals of encounter are “cultural acts” themselves, reflecting the values in a localized sense. Whether it be Keith Basso’s (1970) report on prolonged interaction among the Apaches or Judith Irvine’s (1974) account of the elaborate ritual encounters among the Wolof, the study of greeting rituals from an ethnographic point of view can reveal fundamental norms and expectations of how language is used or not used in a given community of practice. In Japanese society, the rituals of encounters and other pragmatic activities are performed vis-à-vis the larger metapragmatic framework of aisatsu; this entry describes the significance of this concept in understanding Japanese pragmatics. The term aisatsu may be loosely translated into English as ‘greetings and farewell’. However, in addition to the notions of ‘greeting’ and ‘farewell’, aisatsu contains a wider range of pragmatic acts such as ‘thanking’, ‘apologizing’, ‘introducing onself ’, ‘making congratulatory remarks’, ‘giving speeches’, and so on. Generally speaking, aisatsu refers to a wide variety of fixed verbal and nonverbal formulae as well as ritual conducts that mark encounters in various contexts from the everyday to the cetremonial. Used under certain social contexts, all of these pragmatic acts can be uniformly explained as functioning on a metapragmatic level of aisatsu. While aisatsu refers to the ritual of encounters that permeates the everyday conduct of Japanese society, it also forms a nexus of social etiquette linguistically and non-linguistically speaking, functioning as a lubricant in keeping social relations smooth and in a culturally appropriate manner.


This overview of aisatsu first describes the different pragmatic functions of aisatsu in section 2, and then provides an ethnographic account of how aisatsu conducts are acquired, metapragmatically speaking, in section 3. Siding with the idea that pragmatic notions are only interpretable within the larger picture of metapragmatics (Coupland & Jaworski 2004; Verschueren 2004) I will provide a culture-specific case study of the workings of pragmatics in the frame of metapragmatics. 2.  The anatomy of aisatsu The term aisatsu was originally a Buddhist term that meant ‘dialogue’, ‘exchange of words’ or ‘question and answer’ and was used in reference to the practice of judging the extent of one’s enlightenment. Aisatsu does not correspond to any specific English word or notion, as none covers the ranges of the verbal and non-verbal behaviors referred to by the notion of aisatsu. There is a similar notion in Korean, viz. insa. (The Chinese term wènhòu refers to greeting and responding words and behaviors. However, it does not cover the range of public speech as discussed hereon.) Anthropologist Takie S. Lebra (2004: 52) explains that aisatsu consists of “various patterns of conduct, often focused on speech acts accompanied by expected bodily and facial expressions, to signal one’s good wishes and courtesy to other”. Typical types of aisatsu used on a daily basis are words of greeting such as ohayoo ‘good morning’, oyasumi ‘good night’, as well as departing expressions such as shitsurei shimasu ‘[I will] excuse myself ’, or matane ‘see you’. Besides these greeting routines, thanking and apologizing expressions such as arigatoo or doomo are also regarded as a part of aisatsu. On the non-verbal side, aisatsu may include the act of bowing or head-tilt in recognition of others. It may be an act of shaking hands, or bows accompanying words of thanking or apologizing, the exchange of name cards in a business context, and so forth. Whether verbal or non-verbal, aisatsu is typically exchanged among people in talk in action. However, aisatsu can also be extended to the spirits of the ancestors, pets and other animals, personified objects such as natural sceneries like mountains and rivers. Summarizing the social functions of aisatsu, Suzuki (1981: 46) notes that speakers make use of aisatsu (1) to confirm that some kind of interaction is about to take place; (2) to maintain a friendly relationship with the interlocutor; and (3) to ritualistically maintain the interpersonal relationship in a smooth, non-problematic matter. In other words, aisatsu functions to establish “phatic communion” among the interlocutors, in the sense of Malinowski (1923). While functioning to establish phatic communion, I would like to re-categorize the pragmatic functions of aisatsu based on the following two features: (i) marking social relationship between interlocutors; and (ii) marking contextual boundaries.



Risako Ide

2.1  Marking social relationships In her ethnographic account of Wolof greetings in Senegal, Irvine (1974) demonstrated that the exchange of greetings revealed the social relationship among the greeting parties, such as hierarchical relationships and kin member or non-kin member. Likewise, the practice of aisatsu may mark social relationship among the interacting parties, which can be divided into the following three subcategories: hierarchical relationship; in-group/out-group boundaries; and social interdependence. Social hierarchy such as age and social status difference is one factor that influences the choice of aisatsu expressions. For instance, one may say ohayoo ‘good morning’ to a person of equal or younger age, but one would choose the respect form, ohayoo gozaimasu, to a person older or with relative social distance. Gokuroosama is an aisatsu expression rewarding someone’s work, also functioning as a farewell formula. However, this expression may be extended only from the higher status to the lower, whereas the reverse use will sound quite impolite. The boundaries between uchi ‘in-group’ and soto ‘out-group’ can also be marked through the exchange of aisatsu (Abe 1999: 99). The greeting ohayoo ‘good morning’ may be exchanged among in-group members such as within families and between close friends, while konnichiwa ‘hello/good afternoon’ is typically avoided among in-group members as this expression does not promote the casual feeling sensed in the former expression. It is partially for this reason that college students on campus or employees in the workplace tend to use ‘good morning’ when greeting their in-group members even when they meet up during the afternoon periods. The third trait of aisatsu as a marker of social relationship is evidenced in the abundant routine formulae that index social interdependence among interlocutors. Yoroshiku onegai shimasu (lit. ‘please look after me’) is a typical aisatsu expression used when meeting someone for the first time or joining a new group. Okage samade ‘thanks to you’, osewani narimasu ‘I’ll be under your care’ are also routine aisatsu formulae that symbolically implies the inherent feeling that one is indebted to others in society. A symbolic display of the sense of indebtedness towards others is considered as polite and formal and used as linguistic etiquette within Japanese discourse. Aisatsu expressions denoting social interdependence are also used as parts of politeness strategies under circumstances such as requesting, asking for favors, turning down requests, giving warnings, and so on. For example, aisatsu expressions marking social interdependence and/or indebtedness are typically used to frame a request in a business context, giving an impression that one is grateful for the other’s service. Lebra (2004: 53) mentions how e-mail programs use aisatsu as in the following polite warning: “We are sorry, but (mooshiwake arimasen ga) this text cannot be changed.” Functioning in a similar manner, aisatsu is often repeated time and again in its usage. For example, aisatsu of encounters and farewells are repeated over and over


during a stretch of time especially when the encounter is formal as in the repetitious act of bowing. Similarly, thanking and apologizing can be repeatedly referred to over a stretch of time. Thanking and apology expressions such as konomaewa doomo ‘thank you the other day’ and senjituwa shiturei shimasita ‘I am sorry for the past trouble’ indicate that aisatsu is not just a one-time act, but a repeated action to confirm the social interdependence with each other. This habit of repeated aisatsu may be strange from a cross-cultural point of view. For instance in Korea, repeated thanking is normally perceived negatively, as a sign of greed or wanting more of the favor. However, as discussed in Naotsuka (1980) as well as by Yim & Ide (2004: 9), repeating aisatsu is a social indexical gesture to confirm and re-enforce the continuing social connection and interdependence among newly acquainted parties. On the non-verbal side, the annual exchange of New Year’s cards and the bi-annual gift exchanges called ochuugen and oseibo is a form of aisatsu, denoting social interdependence and indebtedness. In the case of the New Year’s card, an average individual sends 50 to 100 post-cards to one’s friends and acquaintances. Unlike regular mail, New Year cards are supposed to arrive on New Year’s day. Thus, the post office holds the cards until January 1st, making sure that they are not delivered before moving into the new calendar year. The message on the card typically consists of routine expressions, acknowledging one’s indebtedness to the other during the past year and asking for continuing friendship and guidance in the coming year. 2.2  Marking contextual boundaries The forms and contents of general routine formulae are regulated by the context of occurrence, i.e., of when to use it and under what circumstances. Many of the aisatsu routine formulae are highly context-bound in nature. For instance, a greeting expression will be selected based on the time of the day, whether morning, noon, early evening or later. The expressions tadaima ‘I’m home’ and okaeri ‘welcome back’ are only to be used by the party who has just returned and the other party who is greeting the one who returns. As an entire day may start with the aisatsu and end with the aisatsu, a season or a year can be bounded with aisatsu. The life cycle may be interspersed with aisatsu in a symbolic sense irrespective of whether it is announcing one’s school entrance, marriage, or retirement. The more formal the context is the more formulaic aisatsu becomes. These types of aisatsu are called goaisatsu and can be distinguished from everyday aisatsu on the basis of their formal and public nature, and their ritual obligation as part of social etiquette. (The prefix go- in goaisatsu is a respect marker.) Compared to everyday aisatsu, goaisatsu semiotically marks the rituality and formality of the event that takes place. The goaisatsu is typically longer in format, poetic and formulaic, and may be accompanied by a set of non-verbal behaviors that marks



Risako Ide

the particular context of interaction. Under different contexts, goaisatsu may be glossed as ‘self-introduction’, ‘public or formal speech’, or ‘congratulations and condolences.’ For instance, when an individual moves into a new community of networks from an older one, the entering is typically marked with a conduct of self-introductory goaisatsu. When joining a new work place, one may be asked to give a goaisatsu speech to one’s new colleagues. Moving into a new neighborhood, the newcomer is due for a goaisatsu in which one makes visits to the doors of the neighbors. These visits are accompanied with small gifts of hand-towels or sweets as a token of goaisatsu. Typically, the gifts are wrapped in noshi paper, with a printing of the newcomer’s family name and they function as a ritualistic introduction to the new community. Other occasions in which goaisatsu takes place are formal and/or public gatherings such as weddings and funerals, entrance and graduation ceremonies, congresses and conventions. The opening and closing speeches of such events are referred to as goaisatsu and are typically performed by the oldest and/or the highest ranked of the participants. The format of these speeches or remarks is highly conventionalized, and originality is sought within a set of routine expressions and phrases. For instance, phrases such as honjitsuwa ohigaramo yoku ‘today being a day of luck’ are so highly context-bound ritualized formulae that they are restricted to be used only in public ceremonial events. In these social contexts, goaisatsu becomes a crucial constituent, without which the events themselves cannot materialize. The bi-annual practice of sending gifts to one’s acquaintances is often coined as ‘seasonal goaisatsu’. In the act of letter writing, the opening and closing aisatsu are to be selected to reflect the relationship between the two parties as well as the season or month of the time of writing the letter. This seasonal context-sensitive letter-writing tradition is also evidenced in the varieties of stationeries of different seasonal motifs utilized in letter-writing in Japan. Duranti (2001: 212) points out adjacency pair formats as being one of the criteria for identifying greetings across speech communities. He states that the greeting ‘hello’ normally requires a greeting back from the interlocutor and a lack of the second pair may indicate a failed speech event. However, under certain social contexts, an aisatsu does not need to be met with a second pair part. A welcoming aisatsu phrase like irasshai(mase), uttered by a member of staff at a retail store is an example of a situation in which no response is expected by the customers. Similarly, customers are not expected to respond to the clerk’s greetings or thanking remarks at convenience stores; this would constitute breaking the norms for the adjacency pair format. (There are episodes of how non-Japanese, such as Koreans and Americans, respond to these one-sided greetings given by the sales providers, which are sometimes bewildering or amusing to the sales personnel.) These cases of one-sided greetings indicate that the usage of aisatsu is regulated by the interlocutor’s social roles designated by the given context, whether student-teacher, or service provider and customers.


3.  The social acquisition of aisatsu As we have seen in the previous section, aisatsu has the property of being indexically linked to certain social contexts, while both appropriating the context of the event and functioning as speech acts. I will now turn to describe the metapragmatic discourse of aisatsu in Japanese society as a primordial site where the use of aisatsu gets trained. As Coulmas (1981: 4) states, conversational routines are tacit agreements that members of a community presume to be shared with every reasonable co-member. For instance, greetings and farewells are the duty of a fully socialized person, and failing to use them may be a mark that one is not fully human, lacking the ability to speak (Goffman 1967: 54). In Japanese society, the ability to perform proper aisatsu is not only associated with linguistic competence, but also with the individual’s qualification as an adequate and mature member of the society. This is evidenced by commonly used phrases like “one who can’t even conduct aisatsu” (aisatsu mo dekinai hito) or “excuse me for not doing aisatsu” (aisatsu mo shimasende), indicating that doing aisatsu is the unmarked, normalized and socially expected conduct. Likewise, failing to perform aisatsu properly when it is socially expected is to mark oneself not only as immature but also as uneducated and as lacking social regard. Thus, as Lebra (2004: 52) notes, aisatsu becomes a point of emphasis and attention in socializing infants, preschoolers, schoolchildren up through junior high school, job applicants, and new employees. In this chapter, we shall see how the practice of aisatsu is taught and acquired in the Japanese society as one of the primordial sites of moral etiquette. 3.1  Aisatsu in school contexts As revealed in Ochs and Schieffelin (1986), early mother-child verbal interactions play an important role in socialization through language. Clancy (1986: 216), in her study of mother-infant interaction in a family setting, observes that the goal of socialization in Japan is to promote unanimity in feeling that supports the norms of verbal agreement and empathy. In other words, this is to extend the care and courtesy to others, including the conducts of thanking and apologizing, i.e., doing aisatsu. The patterned, routine usage of aisatsu gets taught collectively starting at preschool, where the day starts and ends with the singing of the aisatsu songs. Lunch periods are interspersed with meal songs that include the aisatsu phrases itadakimasu ‘I shall receive the meal’ and gochisoosama ‘thank you for the meal’. When entering elementary school, class periods are interspersed with verbal aisatsu accompanied by bowing to the teachers and/or classmates in unison with the commands of the peer leaders. Metapragmatically speaking, aisatsu is regarded as a key training site for individuals to extend their courtesy to others in a pre-established manner. For instance, “Let’s Conduct Proper Aisatsu” may be a typical monthly slogan fixed by junior high school



Risako Ide

or the district Board of Education as part of moral education. During the month of the slogan, school teachers would stand at their school gates every morning to exchange crisp words of greetings with the students coming to school. Also, students may be trained to perform proper aisatsu under certain contexts, such as entering the faculty office. As McVeigh (2002: 127) described in his ethnographic study in a Japanese junior high school, students bow and say in a clear voice, shitsurei simasu ‘excuse me’ upon entering and shitsurei shimasita ‘I’m leaving now’ upon leaving the teachers’ office, unlike the casual entering of classrooms. Adequately executed aisatsu is an embodiment of sincerity and should be composed of proper posture and attitude, speed of speaking, use of pauses, as well as facial expressions and gaze direction. Even when teachers are not around for supervision, aisatsu may be practiced and enforced among the students themselves particularly in the context of club activities. In the student organization of sports or cultural activities, the hierarchical relationship among the senior and junior members requires the juniors to use proper aisatsu to the seniors, including the usage of honorifics in talking to them. Likewise, the practice of aisatsu and extending courtesy to others are nurtured at home and in school contexts, especially in a collective manner through repeated bodily praxis. And this pedagogical practice of aisatsu further continues into the workplace. 3.2  Aisatsu in workplace contexts When visiting Japanese department stores or large retail stores during their opening hours, the customers are greeted by the employees who line up at the entrance or at their corners, chanting their welcoming aisatsu with a bow. The command of aisatsu is regarded as the minimum requirement in doing good business, whether at a supermarket, bank, or in an international firm. Newly employed workers in companies and businesses typically go through a training period to learn basic business mannerisms, including how to conduct proper aisatsu. The training period may vary from a few days to up to a couple of months according to each company, but is a mandatory course necessary for the novice to go through before becoming a full-fledged member. Some companies may send their employees to business “manner” schools or invite expertise from such schools for that purpose. One business mannerism school with branch offices in Tokyo and Osaka provides seminars for recruits, spending a full two-hour of aisatsu training, including how to exchange name cards, how to bow, smile, stand, and so on. In Japanese society, the ability to perform proper aisatsu is regarded as an essential quality of a responsible, skillful and able shakaijin ‘social person’ and also as a desirable trait for an individual. Or, as noted in Lebra (2004: 54), a person’s character and trustworthiness can be judged by his/her ability for aisatsu. A 2007 online survey sponsored by a newspaper company reported that one of the top complaints that company


employees had towards their recruits was how the new employees did not and/or could not use aisatsu properly. Under these circumstances, it is not only worth the investment but an obligation for the company to mold new employees to conform to strict rules for proper appearance, posture and aisatsu. The significance of performing proper aisatsu is also evident in the so-called “How-to” books teaching the usage of aisatsu that are abundant in books stores. Filling the shelves of every book store are titles such as “Dictionary of Well-Received Aisatsu Expressions”, “Aisatsu and Speech Style of a Successful Businessman”, “How to Write Aisatsu Letters”, “Aisatsu and Speech at Weddings”, and so on.

4.  The pragmatics of aisatsu within the metapragmatics of aisatsu As we have seen, initiating and responding to proper aisatsu is collectively, repetitiously, and physically ingrained throughout socialization, in the contexts of the home, school, workplace, and beyond in Japanese society. And it is precisely through these bodily practices of aisatsu that the extension of public courtesy as well as social etiquette is learned in a ritualistic and symbolic manner. Initiating and responding to aisatsu is not only reciprocal, but is decided by the social relationship between the parties involved and it is sensitive to the social contexts where the aisatsu takes place. Silverstein (1993) discusses the notion of the metapragmatic as the nonreductive way of putting language (or grammar) in relation to social action, in the sense of a sociocultural ‘praxis’. In his words, “without a metapragmatic function simultaneously in play with whatever pragmatic function(s) there may be in discursive interaction, there is no possibility of interactional coherence, since there is no framework of structure” (1993: 36). Thus, the notion of metapragmatics highlights the fact that using language is a sociocultural practice with its goal being to enact social relations in groups and societies. In light of the notion of metapragmatics, the conduct of aisatsu itself is a sociocultural practice reflecting Japanese discursive ideologies. It is precisely because of this metapragmatic notion of aisatsu, that the expression “to do aisatsu” (aisatsu-suru) may vary according to the context of its usage. From a pragmatic point of view, the statement “allow me a word of aisatsu” (hitokoto aisatsu sasete kudasai) may refer to getting permission to conduct ‘self-introduction’ or exhibit one’s feeling of thanks or apology, reporting one’s change of status, depending on the context. This point becomes particularly prominent when observing the manners in which a typical expression of apology is used in Japanese public discourse. As analyzed in Ide (1998), the expression sumimasen, typically translated as ‘I’m sorry’ in English can function both to indicate the feelings of ‘gratitude’ and ‘apology’ depending on the context of speech. The fact that a single phrase can function both as an apology and as a thanking expression seems counter-intuitive from the perspective



Risako Ide

of speech act theory, where thanking and apologizing satisfy separate felicity conditions. However, an ethnographic look into the actual usage of this expression reveals that sumimasen (lit. meaning, ‘it’s not ended’) in discourse carried multiple pragmatic functions such as an attention-getting and leave-taking device, a request marker, a confirmation marker, and also a device to simply acknowledge the social indebtedness to one another. This last pragmatic function can be observed in the following situation. Suppose A is on a crowded bus. B standing next to A accidentally bumps into A and quickly says sumimasen to indicate the sense of apology. To this, B says sumimasen to A almost in a simultaneous manner. In Goffman’s (1967) framework of “interactual rituals”, A’s sumimasen can be interpreted as a remedial move, whereas B’s sumimasen, clearly not an apology, nor an expression of thanking, nor a request marker, etc. can be interpreted as a ritualistic supportive move, acknowledging A’s move to remedy the situation in a verbal manner. In other words, it is a ritualistic display of one’s awareness of indebtedness to the other. While proven unsatisfactory from a speech act point of view, sumimasen – having the pragmatic functions of both the remedial and the supportive – can be explained only through the larger metapragmatic framework of aisatsu. That is to say, the speaker does not always spell out what the self or other is doing in these situations, but frames the encounter as a performance of aisatsu routine. Consequently, at its basis, aisatsu functions to mark the social boundaries of the individual, especially in terms of building his/her social face and persona. Thus, it is important for the speakers of Japanese to have the pragmatic competence to manipulate aisatsu in public interaction. This way of marking the social relationship and the public context with proper aisatsu is a sociolinguistic practice engraved in the discourse style of Japanese people, deeply influencing the sense of ‘self ’ as well as ‘social interactional space’. 5.  Exploring intercultural routine formulae This overview has been an attempt to bind together the pragmatics and metapragmatics in explaining the Japanese interaction formulae of aisatsu, which is a non-translatable notion in any other language. As stated by Wierzbicka (1991: 151), “folk names” of speech acts and speech genres are culture-specific and provide an important source of insight into the communicative routines that are most characteristic of a given society. As we have seen in the previous section, remedial and supportive interchanges both function as displays of sincerity or truthfulness on behalf of the individual self and are used routinely and in a dialogic manner to affirm the relationship between social personas. However, these routine formulae as well as the speech acts themselves differ


in their form and function from culture to culture as indicated through such ethnographic studies as kiturim ‘griping’ in Israeli culture (Katriel 1991). While recent trends in the study of pragmatics have developed to involve much cross-cultural perspectives especially in the area of speech act studies, more ethnographically oriented studies are necessary to get a comprehensive picture of the metapragmatic level at which these interactional conducts take place. With such studies, we could make better sense of how different cultural models perceive such notions as personhood, social relationship, and the public interaction. Metapragmatics, or the language awareness by the speakers of a language, is something that is not measurable but only describable and to be accounted for. As noted in Coupland and Jaworski (2004: 19), “whenever language is used in a social context, its meaning is only interpretable through the social values and expectations that fill out that particular context of usage”. While there are more performance-oriented studies of ritualistic and formulaic greeting and farewell expressions from an ethnographic point of view, as in the case of Duranti (2001) on the Samoan case, more studies combining pragmatics with ethnography will help in describing the ground rules for understanding communicative action which permeates a given community of practices.

References Abe, K. (1999). Nichibei no aisatsu kotobano rinkaku [The profile of English and Japanese aisatsu expressions]. Kokubungaku [Japanese Literature] 44(6): 98–103. Basso, K. (1970). “To give up on words”: silence in Western Apache culture. In P.P. Giglioli (Ed.), Language and Social Context: 67–86. Penguin Books. Clancy, P. (1986). The acquisition of communicative style in Japanese. In E. Ochs & B. Schieffelin (Eds), Language Socialization Across Cultures: 213–250. Cambridge University Press. Coulmas, F. (Ed.) (1981). Conversational Routine: Explorations in Standardized Communication Situations and Prepatterned Speech. The Hague: Mouton Publishers. Coupland, N. & A. Jaworski (2004). Sociolinguistic perspectives on metalanguage: Reflexivity, evaluation, and ideology. In A. Jaworski, N. Coupland & D. Galasinski (Eds), Metalanguage: Social and Ideological Perspectives: 15–51. Mouton de Gruyter. Duranti, A. (2001). Universal and cultural-specific properties of greetings. In A. Duranti (Ed.), Linguistic Anthropology: A Reader: 208–238. Blackwell. Goffman, E. (1967). Interactional Ritual: Essays on Face to Face Behavior. Doubleday. Ide, R. (1998). ‘Sorry for your kindness’: Japanese interactional ritual in public discourse. Journal of Pragmatics 29: 509–529. Irvine, J. (1974). Strategies of manipulation in the Wolof greeting. In R. Bauman & J. Sherzer (Eds), Explorations in the Ethnography of Speaking: 167–191. Cambridge University Press. Katriael, T. (1991). Communal Webs: Communication and Culture in Contemporary Israel. SUNY Press. Lebra, S.T. (2004). The Japanese Self in Cultural Logic. University of Hawai’i Press.



Risako Ide Malinowski, B. (1923). The problem of meaning in primitive languages. Supplement to C.K. Ogden & I.A. Richards, The Meaning of Meaning: 146–152. Routledge and Kegan Paul. Mcveigh, B. (2002). Aisatsu: Ritualized politeness as sociopolitical and economic management in Japan. In R. Donahue (Ed.), Exploring Japaneseness: On Japanese Enactments of Culture and Consciousness: 121–136. Ablex Publishing. Naotsuka, R. (1980). Oubeijinga Chinmoku Surutoki. [When Westerners Become Silent]. Taishuukan Publishers. Ochs, E. & B. Schieffelin (Eds) (1986). Language Socialization across Cultures. Cambridge University Press. Silverstein, M. (1993). Encountering language and language encounter in North American ethnohistory. Journal of Linguistic Anthropology 6(2): 126–44. Suzuki, T. (1981). Aisatsu towa nanika. [What is aisatsu]. Aisatsu to Kotoba [Aisatsu and Words] 14: 34–46. Verschueren, J. (2004). Notes on the role of metapragmatic awareness in language use. In A. Jaworski, N. Coupland & D. Galasinsiki (Eds), Metalanguage: Social and Ideological Perspectives: 54–73. Mouton de Gruyter. Wierzbicka, A. (1991). Cross-Cultural Pragmatics: The Semantics of Human Interaction. Mouton de Gruyter. Yim, Y-C. & R. Ide (2004). Hashi To Chokkarak: Kotoba to Bunka no Nikkan Hikaku. [Japanese Chopsticks, Korean Chopsticks: The Comparison of Japanese/Korean Language and Culture]. Taishuukan Publishers.

Anthropological linguistics Ben G. Blount University of Texas, San Antonio

1.  Preliminaries Anthropological linguistics is the name applied to the subfield of anthropology that has language as its primary subject matter. In that regard, anthropological linguistics is simply the anthropological study of language, but efforts to distinguish that area of inquiry among other, related, ones encounter some difficulty. Terminology and distinctions among disciplines and sub-fields can be somewhat hard to follow. An effort at clarity follows. Anthropological linguistics, as a characterizing label, is often synonymous with linguistic anthropology, but a basic distinction is sometimes drawn between them. One facet of the distinction is historical. Anthropological linguistics is sometimes used to refer to the study of languages as linguistic phenomena, but for ends that are ultimately anthropological. That approach to language was characteristic of anthropology in the early decades of the present century. An anthropologist might have been interested, for example, in a reconstruction of pronominal forms, not just for linguistic history but because reconstructed pronominal forms might point to regional distributions and thus shed light on migrations of populations within a specified geographical area. The linguistic study of languages to answer anthropological questions was at one time the principal way that anthropologists approached language. Anthropological linguistics was the name applied increasingly to the entire field of study. Today it can be used to denote similarly oriented research, but not the entire field. The second facet of the terminological distinction is thus focus. An anthropological study of language, linguistically, is anthropological linguistics, distinguished from linguistics by the ultimate anthropological aim of the work and from linguistic anthropology in terms of its avowedly linguistic method. The distinction drawn between anthropological linguistics and linguistic anthropology can be seen in the two professional journals devoted to the anthropological study of language. The historically older of the two journals, Anthropological Linguistics, throughout its history has principally published articles that are based on linguistic data, that record and document linguistic information, especially on Native American languages. The more recently created Journal of Linguistic Anthropology tends to be broader in its scope, covering to a greater extent the entire of interests found in the anthropological study of language.


Ben G. Blount

The anthropology of language is a label that would reflect and represent the entire field of inquiry well, but unfortunately it is not widely used. In the meantime, anthropological linguistics and linguistic anthropology overlap and can thus be used interchangeably, as will be done here, but with the recognition that specialists can and often do draw a distinction between them. 2.  Early history Anthropology and linguistics have separate histories, growing out of different traditions. Anthropology traces its origins largely to social philosophy, whereas linguistics grew primarily out of philological interests in European languages. Although the two fields of inquiry have intersected at various points in their history, as in the nineteenth century anthropology of Paul Broca, one of the major long-term intersections began in the early twentieth century in the study of Native American societies in the New World. The sheer diversity of Native American societies and the sharp contrasts between them and the Old World origins of early American scholars led to interest in and eventually systematic studies of these societies. American anthropology arose, in fact, as a discipline dedicated principally to an understanding of the origins, distribution, and characteristics of the people and societies who were the original inhabitants of the continent. Unlike anthropology in European countries, which tended to be primarily or exclusively ethnological, several traditionally distinct disciplines joined into one in the United States to form an integrated approach to studies of the American Indian. Those disciplines included ethnology, archaeology, physical anthropology, and linguistics. Disciplinary boundaries were sometimes obscured. In the first two to three decades of the twentieth century, linguists, as students of Native American languages, could be found in either departments of linguistics or departments of anthropology (hence anthropological linguists). 3.  Early types of research The scholarship produced by the early anthropological linguists was of three major types. One type was based on descriptions of the linguistic structure and linguistic characteristics of individual languages, produced by scholars such as Franz Boas, Edward Sapir, Leonard Bloomfield, Morris Swadesh, and Mary Haas, among others. Dell Hymes, who has pioneered studies on the history of linguistic anthropology, refers to that genre of research as the Americanist tradition (1976, 1983). Hymes identifies the objective of that tradition, in an ideal conception, as “to know all that can be

Anthropological linguistics

known about, and by means of, Native American languages” (1983: 119). The research was conducted as fieldwork, typically a recording of texts from native speakers and analyses of the texts into linguistic form and genre. Much of the work was salvage in nature, to record as much as possible before the language disappeared. A related, and second type of scholarship in the first decades of research on American Indian languages was classification. The working out of the genetic relationships among the hundreds of languages was a daunting task (still by no means complete), hampered by the absence of good field descriptions of many of the languages. The first classification was produced by John Wesley Powell in 1891, and it contained 58 major taxa. Edward Sapir produced two classifications, one a compilation of what was known in the period approximately 1910–1920, producing 23 major taxa, and a more carefully considered one with 6 major taxa in 1929. The genetic classification of Native American languages contributed substantially to anthropological concerns and interests. One was the complementary nature of the work to archaeological reconstructions of population movements and relatedness of the groups. Another contribution was that the linguistic classifications served as an organizing framework for the large and initially chaotic mass of information that was accumulating about Native American populations. In the absence of other, clear organizing principles, thematic, areal, functional, or otherwise, the language family groups were the obvious, first-order units for systematizing and ordering ethnological data. The third type of scholarship produced in anthropological linguistics in the early part of this century is less topically and more methodologically focused than the previous two types. Linguistic analyses at the time were strictly structural in nature. Language was broken down, analytically, into constituent parts, which formed an inventory of units from which language was built. The procedures for the discovery of the basic ‘units’ were well established and rigorous, making them attractive and useful for analysis of social and cultural information as well. Not only was a part-whole type of approach possible, but pattern identification was facilitated and pursued. One important application of the procedures in early work was Boas’s (1911) and Sapir’s (1927) recognition that linguistic patterning was largely unconscious to the speakers of a language, which meant that the patterning would be less susceptible to secondary rationalization and thus valuable for accurate insights into cultural process. 4.  Continuity Although the centrality of linguistic description and of genetic classification to anthropological linguistics as a subfield has diminished with time, both types of endeavors still occur, e.g., James Crawford’s work on Cocopa (1983) and Joseph Greenberg’s reclassification of Amerindian languages (1987). In general, however, the more linguistic tasks



Ben G. Blount

that were once done by both linguists and anthropological linguists are now much more the province of linguists. That shift coincides with the movement away from primary interests among American anthropologists in Native American research. 5.  The Sapir-Whorf hypothesis Recognition of the patterning of language, and of the implications of pattern for culture structure and process, has grown in importance throughout the history of the subfield. An idea that dates back well into the nineteenth century, and which underlies the positions taken by Boas and Sapir, is that the patterns, and thus structure, of the language one speaks bears some influential, perhaps causal, relationship to the pattern of one’s thoughts. A clear expression of that idea can be found in Sapir’s famous book, Language, published in 1921. As is well known, one of Sapir’s students, Benjamin Lee Whorf, made the idea the cornerstone of his work, culminating in his well-known paper The relation of habitual thought and behavior to language (1956 [1941]). In both cases, the fundamental ideas are that individuals adapt to their social environments through the languages they speak and that the social environments are thus not isomorphic with each other. The similarity of their ideas led Harry Hoijer (1956) to christen them as the ‘Sapir-Whorf hypothesis’, a name and concept that continues to be engaging in anthropological linguistics. The Sapir-Whorf hypothesis is a subject with a deep and intrinsic interest: how much does the language one speaks influence the way one thinks? Much of the literature has been critical, oriented to disproving and rejecting the claims that the hypothesis purportedly makes. There has been a diversity of opinion as to what Whorf, in particular, actually meant, and that in turn has led to considerable misunderstanding. Secondary interpretations of Whorf have led to false ideas such as the existence of two hypotheses, a strong one that claims that the structure of the language an individual speaks determines the thought patterns and thus world view of the individual, and a weak one that reduces the claim from ‘determines’ to ‘influences’. A careful reading of Whorf does not reveal any such distinction, nor does it reveal any claims to determinacy per se. In a very clear exposition of Whorf ’s claims, John Lucy (1985) shows that Whorf argued that the ontological categories upon which important language distinctions are based, such as tenses and classifiers, are habitually used by speakers and that the habitual use itself predisposes them to see their physical and cultural world through the categories. For example, an individual who conceptualizes linear distance in terms of miles will tend to ‘think’ in terms of those units, whereas someone whose experience is with kilometers will ‘think’ in terms of that category. While that may appear to be a trivial distinction — simply different metrics with different labels to designate them — the difference in a Whorfian sense is more significant. In that perspective,

Anthropological linguistics

people may think of distance itself as different because of the particular metrics with which they are familiar. One kilometer may be a ‘walkable’ distance, whereas one mile may not be. Moreover, a kilometer is based on the unit of ‘meter’, and thus a kilometer may well be thought of as a thousand meters, and thus easily segmentable, whereas a mile is not based on a similar basic unit and is unlikely to be thought of as 5,280 feet or 1,760 yards. Although interest in Whorfianism has been recently renewed in anthropological linguistics (Gumperz & Levinson 1991), attention to the topic was less pronounced in the 1960s and 1970s. Two major new developments occurred, however, in anthropological linguistics during the early 1960s, leading to new interests and types of research. One of them was antithetical to what was seen as the extreme relativism of the Sapir-Whorf hypothesis, focusing on what societies have in common beneath, or through, the apparent diversity that they exhibit. That area of inquiry has had a sequence of labels, ranging from ‘the new ethnography’, to ‘ethnoscience’, to ‘ethnographic semantics’, among others. The second development was not in as direct an opposition to Sapir-Whorf but sought patterns in language at a new level and in novel ways. This area of inquiry was originally ‘the ethnography of speaking’, later extended to cover communication in a broader, semiotic sense as ‘the ethnography of communication’.

6.  Ethnographic semantics To take ethnographic semantics first, the fundamental idea is that the phenomena that are particularly important to a society are likely to be named and that the phenomena will be organized into a structured domain. The lexical units, the names of the items in a domain, are built upon and are thus reflexive of the domain. The investigation of the domain can thus be made in terms of the semantics of the lexical units. Anthropologists interested in types of kin relatives (a favorite target of research) would select kinship as a domain to be investigated ethnographically, elicit the universe of lexical items that named types of relatives, and then seek the features, or components, from which the meaning of the lexical items were constructed. In English kin terms, ‘father’ would be constructed from the features ‘male’, ‘ascending generation one’, and ‘lineal relative’, whereas ‘mother’ would be the same except for ‘female’, and ‘son’ would be the same except for ‘descending generation one’. Ethnographic semantics was best applied in cases in which domains were relatively clear in their scope and were elaborated lexically. In some of those areas, kinship for example, the results unfortunately proved to be of dubious value. The original objective of the type of research was to accurately depict how members of societies actually organize their cultural knowledge. The idea was not only to do better



Ben G. Blount

ethnography, in the sense of it being more explicit, reliable, and replicable, but to produce a description that would reflect the psychological reality of the speakers. In reality, what anthropologists typically produced were analyses of how one could organize a domain lexically. Whether the psychological reality of the speakers was represented was another issue altogether. The two success areas of ethnographic semantics are ones in which the problem of psychological reality was the best overcome. One of those is in color term research, and the other is the nomenclature and classification of botanical and zoological phenomena. The classic work on color terms was by Brent Berlin & Paul Kay (1969). They began their research with a comparison across 20 languages of the lexical terms for basic colors (primary, salient, and monolexemic terms such as red, blue, yellow, pink, etc.). The sample has now been extended to more than 100 languages in an ongoing project, the World Color Survey (Kay, Berlin & Merrifield 1992). Among the major findings of the study are that the appearance of basic color terms across societies follows an implicational scale. If a society has only two color terms, they will be light (white) and dark (black); if a society has three terms, the third one will be red; if there are four terms, the fourth one will be green or blue, and so forth through seven terms with another four following those but with less predictability. The data base is now rich enough that variation has been documented in the paths that are taken in the ‘unfolding’ of basic color terms in the blue, green, and yellow regions. Moreover, physiological bases for the perception of some of the terms have been established, but interestingly, the physiological bases and the ‘coverage’ of the color spectrum are not isomorphic. Societies, in other words, override biological constraints with cultural ones, and moreover, they appear to do so in predictable, systematic ways. Part of what may underlie the cultural systematicity is the ordering of the number of basic color terms along a dimension, more or less, of technological complexity. The greater the technological complexity of a society, the more likely that it will have a greater number of basic color terms. Although that result is somewhat controversial, the evidence does appear to point in that direction. The results from research on folk terminology and classification of plants and animals has been recently summarized by Brent Berlin (1992). Relatively small-scale societies that continue to live in close association with their traditional environments appear to name and classify the biota in highly similar, predictable ways. The total number of names for native plants and animals is approximately 500, and by far, most of those are at the folk generic level. That level includes names such as eagle, raccoon, lizard, oak, pine, and pansy. Generics, however, occupy one position in a hierarchy. At the next highest, coordinate level, is life form, e.g., bird, mammal, reptile, tree, flower, etc. Up the taxonomic hierarchy another level is the unique beginner, animal or plant. Coming back to folk generics and going down the hierarchy, first are varietals, such as white pine, pin oak, etc., and down one further are specifics, such as kinds of white

Anthropological linguistics

pine, kinds of pin oak, etc. The specifics of a taxonomy vary somewhat from society to society, but the basic hierarchic structure is the same. The remarkable similarity of ethnobiological nomenclatural and classificational systems is apparently due to the similar ways that members of the societies perceive their local biota. Perception of morphological discontinuities and of salience seem to be the bases of the similar taxonomies. People perceive the morphological traits and the discontinuities across different types of plants and animals in much the same way, regardless of the language they speak or their particular geographical location. Common perception and perceptual strategies produce the similar hierarchic structures. Complementary to the finding of perceptual primacy is the discovery that the more culturally salient a folk generic becomes, the more likely that varietal and even specific forms will be recognized and named. Cultivars, for example, are far more likely to have varieties and specific types than are plants which are not domesticated. Salience, in fact, may be the driving element that leads to perceptual recognition and distinction, although that has not been fully worked out and demonstrated. 7.  The ethnography of communication More or less contemporaneous with the emergence of the new ethnography, a second line of ethnographic inquiry began, the ethnography of speaking. Like the new ethnography, the objective was to do ethnography better, but whereas in the new ethnography the idea was to do more focused, fine-grained investigation, in the ethnography of speaking, the new element was the topic itself. Throughout its history, anthropological linguistics had focused on language structure and then used the discoveries about structure to relate language to other phenomena. The new ethnography also followed that pattern. In the ethnography of speaking, speech behavior became the subject matter, the topic to be investigated. In a classic article, Dell Hymes (1962) defined and developed a framework for the ethnographic study of the way that speakers of a community use their language resources. Function was given priority over form; form followed function rather than vice versa. Descriptions and analyses sought patterns of the way that speech was used, looking at patterns among types of speakers, types of situations, genres of speech, and variant functions of speech. The shift from relativity of language structure to relativity of speech use, literally opened up a broad, exciting new area of research. An early publication that helped to further define the area was a special publication of the American Anthropological Association, co-edited by John Gumperz & Dell Hymes (1964). The objective of the ethnography of speaking was a complete description of the speech resources available in a speech community and the patterns and functions that ensued from their systematic utilization. The scope of that objective was sufficiently broad as to require



Ben G. Blount

extensive periods of research and research effort. As a consequence, relatively few studies have purported to be complete ethnographies of speaking. Rather, studies have focused on some particular aspect of speaking within a community, such as political oratory (Bloch 1975; Brenneis & Myers Eds. 1984), law (O’Barr 1982), speech play (Sherzer 1983), and socialization (Blount 1972; Ochs 1988; Schieffelin 1990). 8.  Sociolinguistics The ethnography of speaking was eventually expanded to cover types of signs and symbols other than spoken language, and the name was also expanded to be the ethnography of communication. During the same period, the early to mid-1960s, another area of research had been developing from a variety of sources, including anthropology, linguistics, philosophy, and sociology. The subject matter of that area was the social nature of speech, focusing on the socially organized use and consequences of speech. The approach was micro-level, looking at the social characteristics of speaker and speech recipient and the form of speech to see how they constrained the message conveyed and the response to it. This area of inquiry was referred to as sociolinguistics. A basic tenet of sociolinguistics is that linguistic form and function are viewed not simply as correlates to social phenomena but as social phenomena themselves. In time, a number of sub-areas on inquiry arose, most of them areas of inquiry in their own right today. Those include conversation analysis, social interaction, discourse analysis, pragmatics, speech act theory, and ethnomethodology. What is of particular interest here is that developments in sociolinguistics were also developments in the ethnography of communication. In one sense, the subject matter of sociolinguistics was the same as part of the ethnography of communication, the interrelationship between speakers as social actors and language form in the service of social action. A publication that spelled out the mutually, interdependent roles of the ethnography of communication and sociolinguistics was Directions in sociolinguistics: The ethnography of communication (Gumperz & Hymes 1972). Other publications of the period that were especially noteworthy were Susan Ervin-Tripp’s overview of sociolinguistics (1969), David Sudnow’s collection of essays on face-to-face interaction and conversation analysis (1972), Michael Silverstein’s work on social functions of language (1976), and Erving Goffman’s studies of interaction ritual (1967), public behavior (1971), and experiential frames (1974). The period of the 1960s and 1970s was one of rapid development through cross-disciplinary research, by the aforementioned scholars, among others. The work was oriented toward an understanding of how language represents culture and how the representations are used to construct meaning among actors engaged in social interaction.

Anthropological linguistics

9.  Recent research and current directions The innovative research agendas in the 1960s and 1970s laid the foundation for a flowering of research topics in the 1980s and 1990s. The data bases for studies of color terms have grown considerably, and new issues have emerged. Particularly intriguing is the role of the term ‘yellow’ in the evolutionary sequence; it appears to have ‘wild card’ status, for reasons not yet understood (Kay, Berlin & Merrifield 1992). Studies of ethnobiological nomenclature and classification have continued to expand, with new issues also arising there. One of the central ones concerns the ultimate motivation for the taxonomies and their form. The preponderance of evidence seems to be for what is called the intellectualist position, that people name and classify the living world simply because they are confronted perceptually by them. The opposing perspective, the utilitarian position, is that the basis for naming and classifying is functional, that some utility is, or was, derived from the plant or animal named (Hunn 1985). Part of the difficulty in resolving the issue is that multiple taxonomies involving the same plants or animals exist (e.g., a plant may be a representative of a type, food, medicine, etc.), and sharp and clear separation of them is not easy, especially for the respondents in small-scale societies for whom the taxonomies exist. Ethnographic semantics has also been extended to new domains, such as place names (Hunn 1994), personal names (Blount 1993), onomatopoeia (Berlin 1992; Hays 1994). Perhaps ironical is the fact that ethnographic semantics has produced recent results documenting the commonality of human cognitive processes, given that in its early history it was sometimes criticized as too particularistic and engulfed in unimportant detail. Developments within other areas of language, culture, and society have become so numerous and diversified as to preclude easy and succinct summary. Themes can be identified, but even here studies do not fall easily into clear-cut categories. One fairly evident theme is that ethnography continues to be a mainstay for productive characterizations of language and speech behavior in selected communities. William Hanks’s study of Yucatec Mayan (1990), for example, provides rich information both about the nature of speech in their communities and about the use of deixis (pointing devices such as he, that, there) in the construction of their sociocultural world. Similarly, Alessandro Duranti (1994) has shown that a large variety of linguistic and rhetorical devices are used in political oratory among Samoans to negotiate and maintain sociopolitical status. Jonathan Hill (1993) has shown in his studies on the Curripaco Indians of Venezuela how members of the society creatively construct and reconstruct their social identity through ritual speech and music. As even these few examples show, ethnographic studies of speech communities in recent years have a thematic orientation. That orientation is toward the demonstration of the use of language form and behavior to the construction of social identity,



Ben G. Blount

necessarily in public fora and within contexts of interaction with co-members of society. The establishment of an identity must to some degree be through ritualized, traditionally sanctioned, and publically approved forms. How individuals establish, construct, re-affirm, and re-construct their social identities is one pole of the ethnographic objective. The other pole is the identification and description of the institutions that have emerged and evolved that provide the framework and the means for individuals to establish social place and personae. An excellent illustration of this theme is to be found in Simon Harrison’s ethnography of the Avatip people of New Guinea (1990), in which individuals vie to establish personal names of value and distinction. A central, fundamental part of that process is a public display in which individuals compete with each other in regard to their knowledge of names of individuals and their accomplishments and achievements. Closely related to the two themes noted, and cross-cutting them, is a third theme, pragmatics. Just as the linguistic units of language must be related to each other in structured ways in order to form larger units, units of language use must be related to each other in structured ways so as to allow participants to be and become aware of the structures. Efforts for analysts to discern the structures require attention to the unfolding of interaction, to look at the patterning of speech behavior in terms of the interrelationships of its units or components, much as in the ethnography of communication described above. In this more recent, pragmatic approach to language behavior, an additional consideration is brought to the fore, one that substantially increases the complexity of the task. That consideration is the question of how the units or forms are tied not just to each other but how they are also tied in form and function to the broader context in which the behavior is constructed. The grounding of language use in behavior and of the behavior in specific contexts is, however, only the first part of the analytic endeavor. Contexts are interactive, i.e., features that are present or brought into the present are not merely backdrops to serve as anchors for speech but can be selectively and strategically interwoven into the stream of behavior. Context can serve a variety of roles in the interpretation of the meaning of speech behavior, as the aspect of the real world to which utterances are tied (referenced), as setting parameters for the particular kind or type of externality to which the speech refers or in which it is played out, and as units incorporated into the creation of the messages themselves. Context can be thought of as having multiple layers. Major innovative research on this topic can be attributed to John Gumperz (1982), and a recent collection of articles, edited by Alessandro Duranti & Charles Goodwin (1993), addresses the various aspects and issues of context and serves as a good overview of this area of anthropological linguistics. The multilayer view of language use, social interaction, and the construction of meaning tied to time and space requires mechanisms to move between and among layers, to mediate between units and across functions. The search for mechanisms and

Anthropological linguistics

for an understanding of how they work constitutes a central part of contemporary anthropological linguistics. A leading figure in that enterprise is Michael Silverstein, whose classic work on shifters (1976) helped to initiate inquiry into the question of mechanisms (shifters being devices that cast or recast language from one dimension or domain to another, e.g., verbal tense). Much current focus is on indexicals, on language form that is tied to meaning not through direct naming (reference) but through association of specific and general, as in the case of the characteristics of an individual’s speech indexing his or her social class. Indexicality is not an occasional feature of language form; it is inherent in form-function relationships. Indexicality itself must be based on underlying cognitive capacity and process, although the specific nature of the bases are not well worked out (except perhaps for color terminology). Along with a much more complete account of the mechanisms of mediation, the specification of how language behavior is related to underlying principles of thought and meaning is a contemporary challenge for anthropological linguistics. The two tracks of anthropological linguistics, nomenclatural/classificational and ethnographic/sociolinguistic may be closer to each other in orientation and goals than might appear to be the case. In the categories most studied, color terms and folk biology, names and categories, despite their enormous surface linguistic diversity, appear to have universal bases. The pragmatics of language use, which shows even greater diversity, must also have common, underlying bases, toward which mechanisms such as indexicals must point. This is where anthropological linguistics is likely to make its advances and contributions in the future. We may eventually be in a position to fully develop and understand linguistic relativity in all its complexity.

References Berlin, B. (1992). Ethnobiological classification. Princeton University Press. Berlin, B. & P. Kay (1969). Basic color terms. University of California Press. Bloch, M. (1975). Political language and oratory in traditional society. Academic Press. Blount, B.G. (1972). Aspects of Luo socialization. Language in Society 1(2): 235–248. ——— (1993). Luo names. In S. Mufwene & L. Moshi (Eds), Topics in African linguistics: 131–140. John Benjamins. Boas, F. (1911). Introduction. In Handbook of American Indian languages, Part 1. Smithsonian Institute. Brenneis, D. & F.R. Myers (Eds) (1984). Dangerous words. New York University Press. Crawford, J.M. (1983). Cocopa texts. University of California Press. Duranti, A. (1994). From grammar to politics. University of California Press. Duranti, A. & C. Goodwin (Eds) (1993). Rethinking context. Cambridge University Press. Ervin-Tripp, S. (1969). Sociolinguistics. In L. Berkowitz (Ed.), Advances in experimental social psychology: 93–107. Academic Press. Goffman, E. (1967). Interaction ritual. Doubleday. ——— (1971). Relations in public. Harper & Row.



Ben G. Blount ——— (1974). Frame analysis. Harper & Row. Greenberg, J.H. (1987). Language in the Americas. Stanford University Press. Gumperz, J.J. (1982). Discourse strategies. Cambridge University Press. Gumperz, J.J. & D.H. Hymes (Eds) (1964). The ethnography of communication. Special issue of American Anthropologist, 66(2), Part II. ——— (1972). Directions in sociolinguistics. Holt, Rinehart & Winston. Gumperz, J.J. & S.C. Levinson (1991). Rethinking linguistic relativity reconsidered. Current Anthropology 32(5): 613–623. Hanks, W. (1990). Referential practice. University of Chicago Press. Harrison, S. (1990). Stealing names. Cambridge University Press. Hays, T. (1994). Sound symbolism, onomatopoeia, and New Guinea frog names. Journal of Linguistic Anthropology 4(2): 153–174. Hill, J. (1993). Keepers of the sacred chants. University of Arizona Press. Hoijer, H. (Ed.) (1956). Language in culture. University of Chicago Press. Hunn, E. (1982). The utilitarian factor in folk biological classification. American Anthropologist 84: 830–847. ——— (1994). Place names, population density, and the magic number 500. Current Anthropology 35(1): 81–84. Hymes, D.H. (1962). The ethnography of speaking. In T. Gladwin & W. Sturtevant (Eds), Anthropology and human behavior: 13–53. Anthropological Society of Washington. ——— (1976). The Americanist tradition. In W.L. Chafe (Ed.), American Indian languages and American Indian linguistics: 11–33. Peter de Ridder Press. ——— (1983). Essays in the history of linguistic anthropology. John Benjamins. Kay, P., B. Berlin & W. Merrifield (1992). Biocultural implications of systems of color naming. Journal of Linguistic Anthropology 1(1): 12–25. Lucy, J. (1985). Whorf ’s view of the linguistic mediation of thought. In E. Mertz & R.J. Parmentier (Eds), Semiotic mediation: 73–97. Academic Press. O’Barr, W. (1982). Linguistic evidence. Academic Press. Ochs, E. (1988). Culture and language development. Cambridge University Press. Powell, J.W. (1891). Indian linguistic families of America north of Mexico. Bureau of American Ethnology, Report 7 (1885–1886). Sapir, E. (1921). Language. Harcourt, Brace. ——— (1927). The unconscious patterning of behavior in society. In E.S. Dummer (Ed.), The unconscious: 114–142. Knopf. ——— (1929). Central and North American languages. Encyclopedia Britannica (14th Edn) 5: 138–141. Schieffelin, B. (1990). The give and take of everyday life. Cambridge University Press. Sherzer, J. (1983). Kuna ways of speaking. University of Texas Press. Silverstein, M. (1976). Shifters, linguistic categories, and cultural description. In K. Basso & H. Selby (Eds), Meaning in anthropology: 11–55. University of New Mexico Press. Sudnow, D. (Ed.) (1972). Studies in social interaction. Free Press. Whorf, B.L. (1956) [1941]. The relation of habitual thought and behavior to language. In Language, thought, and reality: 134–159. MIT Press.

Franz Boas Regna Darnell University of Western Ontario

1.  Introduction Franz Boas (1858–1942) was unquestionably the preeminent figure in twentieth century North American anthropology. He established the four-field structure of the discipline around cultural, physical, linguistic and archaeological studies of the American Indian and trained most of the fledging anthropologists who professionalized anthropology and moved it from government, learned society, and museum into the academy (Darnell 1998). Boas himself contributed to all four sub disciplines. His work on rapid changes in immigrant head form undercut eugenics arguments and diminished the importance of anthropometric measures of race. Race became an arbitrary category, best approached by anthropology as racism. Although he strongly supported Afro-American anti-racism, particularly through his friendship with W.E.B. Dubois (Baker 1998), Boas also had a personal stake in mitigating antiSemitism. His archaeological work was more perfunctory, and students interested in archaeology and physical anthropology were usually sent to Harvard. In the study of culture, Boas’s theoretical contributions revolved around his critique of evolution, his rejection of rationalist theories of human nature, his historical particularism, his insistence on rigorous ethnographic method, and his emphasis on “the native point of view.” This essay, however, focuses primarily on Boas’s wide-ranging contributions to Amerindian linguistics and their inseparability from his views on culture. 2.  Biography Franz Uri Boas was born in Minden, Westphalia, 9 July 1858, to a well-to-do Jewish textile merchant, Meier Boas, and his wife Sophie Meyer Boas, who conveyed to her children the ideals of the abortive 1848 revolution. Boas moved from the University of Heidelburg to Bonn to study physics, receiving his doctorate from Kiel in 1882 for research on the optics of sea water. At Kiel he studied geography with Theobald Fischer. After completing his mandatory military service, Boas moved to Berlin’s Royal Ethnological Museum to prepare with Adolph Bastian for an expedition to the Eskimo. His year among the Central Eskimo (1883–84) persuaded him that geography was a limit rather a determinant of culture. His intellectual interests permanently moved from


Regna Darnell

psychophysics to geography to ethnology (Stocking 1968). In 1885, he shifted his ethnographic focus to the Northwest Coast, where it remained for the rest of his life. In 1887, fearing the effects of German anti-Semitism on his professional aspirations, he relocated permanently in North America and married Marie Krackowizer, the daughter of an Austrian-born New York physician. Boas was unemployed or underemployed for much of the next decade, serving as an editor of Science, teaching psychology at Clark University, and arranging museum exhibits for the 1894 World’s Columbian Exposition in Chicago. His early NWC fieldwork was sponsored by the Bureau of American Ethnology and the British Association for the Advancement of Science, supplemented by museum collecting and assistance from his wife’s uncle Abraham Jacobi. In 1897, Jacobi also negotiated and partially funded Boas’s joint appointment at Columbia University and the American Museum of Natural History. This appointment provided the base from which Boas reworked the institutional structure of American anthropology. Although in retrospect this Boasian dominance of the discipline seems virtually inevitable, at the time his power base and influence seemed to develop only gradually. The Museum sponsored the Jesup Expedition to study Siberian and NWC cultures and provided fieldwork funding for many of Boas’s early students. In 1905, however, Boas resigned from the Museum and devoted himself exclusively to academic training. His students founded most of the major North American anthropology departments, beginning with the appointment of Alfred Kroeber at Berkeley in 1901. In 1898, Boas collaborated with W.J. McGee of the American government’s Bureau of American Ethnology to expand the American Anthropologist from a Washington-based local publication into a national journal. He fought (with only partial success) for full professionalization in the membership and organization of the American Anthropological Association, founded in 1902. During World War I, his pacifism caused him major problems, culminating in censure by the AAA for exposing espionage under the guise of anthropological research in Mexico (this censure was revoked only in 2004). During World War II, Boas’s anti-racism and anti-Semitism led him to take an active political stance. Boas died in Paris 21 December 1942 in the arms of Claude Levi-Strauss. 3.  The Americanist tradition The Americanist tradition (Darnell 1998, 2001) developed by Boas and his first generation of students remains healthy today (Valentine & Darnell Eds. 1999) and continues to distinguish North American study of particular languages and their relationships from that in other anthropological traditions. Americanist linguistics incorporates cultural context in its understanding of linguistic form (Hymes & Fought 1975; Murray 1994). Texts from native speakers of Aboriginal languages

Franz Boas

provide the data-base for both cultural and linguistic analysis. Because Boasian linguists often worked closely with a small number of remaining fluent speakers, they also attended to what Boas called “the native point of view” (his paramount linguistic student Edward Sapir glossed this as the “psychological reality” of linguistic forms). Language was a symbolic form through which culture became accessible to study. Boas himself documented an “extraordinary number and variety of North American languages,” including Eskimo, Coast Tsimshian, Kwakiutl (Kwak’wala), Nootka, Bella Bella, Heiltsuk, Chemakum, Upper and Lower Chehalis, Kootenai, Chinookan, Chinook Jargon, Keresan, and Sioux (Mithun 1996: 44). His grammatical sketches of Tsimshian, Kwakiutl, Chinook and Lakhota (with Ella Deloria) “were based on language in use, with examples drawn from the connected speech of texts” (Mithun 1996: 44). Boas’s texts from Chinook, Tsimshian, Bella Bella and Kwakiutl were produced in collaboration with local native speakers. George Hunt’s work with Boas on Kwakiutl exemplifies the method; Boas preferred to teach native speakers to write their language and produce texts for him. Recent scholarship has restored credit in these enterprises to the collaborators, and the texts are widely used today in Native American communities to revitalize and preserve traditional languages (Berman 1996). In his introduction to the inaugural issue of his journal, the International Journal of American Linguistics in 1917, Boas emphasized how little was known about Native American languages. In addition to texts, grammars and dictionaries of each language, Boas called for “records of daily occurrences, everyday conversations, descriptions of industries, customs and the like” (1917 in Boas 1940: 201). In the language of Herder, Steinthal, and Humboldt, Boas noted the “inner form” of languages that persisted despite borrowings and change. The categories of Latin and Greek were inadequate to capture the thought-worlds of Native American languages, which deserved to be described in their own terms (on the basis of texts). 4.  Time perspective in aboriginal languages Boas’s early NWC fieldwork used linguistic classification to organize the diversity of North American cultures and languages, proposing several new relationships (Wakashan and Haida/Tlingit) that were included in Powell’s 1891 classificatory synthesis. In the late nineteenth century, Boas accepted the need for survey fieldwork to map the basic diversity of languages and group them together into linguistic families. Later, however, he became increasingly sceptical about the possibility of distinguishing “archaic residue” from “diffusional cumulation” (Swadesh 1951). In principle, language was valuable for historical reconstruction because it was not subject to “secondary rationalization,” i.e., linguistic forms were rarely perceived consciously by speakers. But Boas’s work on the borrowing of NWC folklore elements persuaded him that the



Regna Darnell

origins of linguistic forms were inaccessible at great time depth and that many languages might have complex origin histories. This position was not in the first instance linguistic. Rather, it resonated with his critique of evolution and its unilinear developmental sequences (e.g., Boas 1896). Boas preferred to focus on particular cultural histories, believing that folklore elements exemplified the ease and arbitrariness with which cultural forms, including linguistic ones, could move and adapt themselves to new cultural contexts. If language were part of culture, then linguistic elements should behave no differently from cultural ones. Boas distanced himself from most of his students’ work on linguistic classification, which culminated in Edward Sapir’s reduction of the number of North American linguistic families from 55 in Powell’s 1891 Bureau of American Ethnology classification to only 6 in 1921 (Powell 1891; Sapir 1921, 1929). In the more elaborated 1929 version of the classification, however, Sapir outlined an intermediate classification of 23 units (Darnell 1990, 1998, 2001) acceptable to his more conservative colleagues and acknowledged that his deeper connections were somewhat more speculative and might require further detailed demonstration. Sapir’s classification reflected fieldwork over the intervening three decades, but it also applied the Indo-European comparative method to unwritten languages in a way that was impossible for Boas and Powell as self-taught linguists depending primarily on visual inspection of word lists to establish linguistic relationships. Sapir’s classification was used by ethnologists as a framework for culture history for several generations, but the current climate in American Indian linguistics is a more cautious one. Campbell and Mithun (1979) summarize the present standard of evidence in recognizing 62 language families, even more than Powell in 1891. Their criteria involve detailed statement of sound correspondences without consideration of geographical or cultural contexts (and thus evidence of the borrowing that bedeviled Boas’s efforts at linguistic reconstruction). Nor is it Sapir’s position, which depended on well-honed linguistic intuition and historical imagination to pose questions for further investigation of potential genetic relationships at great time depth. Today, American Indian linguistics exists in its own right, as part of general linguistics, not as the handmaiden to ethnology. This autonomy is perhaps one reason that linguistics has become increasingly peripheral to anthropology in North America, despite lip service to a four-subdisciplinary structure. Sound changes provided firm evidence of genetic relationship among languages. Sapir argued as early as 1916, in Time Perspective in Aboriginal American Culture: A Study in Method, that ethnology could reconstruct history only through language, because elements of language retained traces of their past history in ways that other parts of culture did not. Chronological sequences based on archaeology were unavailable for North America until the Pecos classification of 1927. Thus, linguistics necessarily served as handmaiden to ethnology and potentially revealed vast vistas of migration,

Franz Boas

settlement, and cultural contact. For Sapir, Boas’s failure to respond to this project constituted a lack of historical imagination. Boas described his work as divided into two major themes: history and psychology (Darnell 2001). History had to be done first, because it allowed anthropology to map the diversity of cultures in North America and their interrelationships (primarily within broad culture areas). The history that interested Boas was more spatial or geographical than temporal. Because history was based on the observed movement of cultural elements, individual agency was inaccessible for the pasts of cultures without written records. The anthropologist outlined historical generalizations that members-of-culture lacked the perspective to formulate for themselves. This comparative element part of Boas’s method was adopted by Claude Levi-Strauss in his structuralist studies of American myth systems, although, for Levi-Strauss, the regularities resided in the nature of the human mind rather than in history. Boas has been criticized strongly for this rather sterile approach to history. E.g., his recalcitrant student Paul Radin (1927) defined history in terms of what could be known through oral tradition. For Boas himself, however, issues of text, cultural context, and oral tradition fell under the rubric of “psychology” rather than “history.” Once the basic ethnographic description and historical classification of cultures was completed, after around 1910 by his own statement, Boas could turn to psychological questions of “the native point of view,” accessible through texts. Language, then, fell between linguistics and culture as a broader entity of which it was part. Both grammar and textual organization would reveal “the mind of primitive man” (1911b). This psychological emphasis led Boas’s later students, especially Sapir, Ruth Benedict, and Margaret Mead to investigate the relationship of culture and personality. Boas’s own intimate link between linguistic categories, ethnographic texts, and cultural standpoint was not, however, incorporated in most of this work. 5.  The Handbook of American Indian Languages As Honorary Philologist of the Bureau of American Ethnology, Boas used the Handbook of American Indian Languages (1911a) to operationalize his ideas about the embedding of psychological point of view in linguistic categories. The grammatical sketches were prepared primarily by Boas’s first generation of students and formed a key part of his platform to professionalize American anthropology under his personal direction and vision. Boas kept careful editorial control over personnel and attempted to impose a standard format which successors have called “the Boas plan for the study of American Indian languages” (Stocking 1974). Sapir’s dissertation grammar of Takelma, for example, was excluded from the first volume of the Handbook because it was too long and Sapir refused to modify it to fit the format. Boas ignored most of the Bureau linguists,



Regna Darnell

considering their work unprofessional. The “analytic” sketches were commissioned largely on the basis of available material by a competent linguist, and assessed in Boas’s framework. The resulting selection was assumed without further discussion to be representative of the psychological types of American Indian languages. Such psychological patterns could result either from diffusion or common origin. Although Boas had left behind the question of genetic relationship of American Indian languages, he never produced the promised psychological typology, deferring this project until adequate sketches of the significant languages were available. Some of his hesitation over psychological types may be attributed to the evolutionary overtones of labelling all American Indian languages as polysynthetic. The types he envisioned could not be ranked. In one sense, Boas was contributing to the continuing mapping project of the Bureau of American Ethnology, moving from the cataloguing of words to the analysis of grammatical constructs. In another sense, however, Boas argued that each language stood as an independent packaging of what Sapir’s student Benjamin Whorf would later see as an amalgam of “language, thought and reality.” The so-called Sapir-Whorf hypothesis of the relationship of habitual thought to linguistic categories has its roots in this Boasian insistence on the unique perceptual patterning of each language and the equal value and expressive capacity of every natural language. Cultural relativism, the most salient contribution of anthropology to twentieth century American public culture, began with Boas’s recognition that Indo-European categories distorted languages on which they were imposed. Sapir and Whorf took the argument much further than Boas did with the Handbook of American Indian Languages; Ruth Benedict would apply it to the unique patterning of cultures studied by anthropology. The argument was not so much for incommensurability as for internal coherence and value for members of a culture. The anthropologist or linguist, having an ethnographically induced multi-lingual awareness, could transcend the categories or his/her own or others’ cultures or languages (Darnell 2001).

6.  Phonetics vs. phonemics In “On Alternating Sounds” (1889), Boas realized that what appeared to be errors of second language speakers arose from the transfer of sounds habitually used in their native languages. His argument was directed against the “primitiveness” of any language or language speaker, but it also set the stage for perception as an important element in the recording of sounds. In 1912, Boas established a committee of the AAA, under his firm direction, to design a standard orthography for all American Indian languages. The report, four years later, was a compromise, with a simplified system useful for large bodies of text collected by linguistically naive investigators and a more detailed system permitting detailed phonetic analysis (Mithun 1996: 45).

Franz Boas

Although the seeds of a distinction between phonetic and phonemic analysis are implicit in the idea of alternating sounds, Boas did not yet conceive the sounds of a language as forming a pattern. This further step would come from Sapir’s “Sound Patterns in Language” (1925), elaborated as “the psychological reality of phonemes” (1933). The former appeared in the first number of Language, journal of the newly founded Linguistic Society of America. It was a watershed which thereafter divided Sapir the professional general linguist from Boas, the pioneer descriptive linguist. Boas’s later students were taught the value of retaining all possible information on endangered languages, while Sapir’s students were encouraged to forego phonetic detail in favour of more abstract phonemic analyses depending on the perceptions and intuitions of native speakers (although these were usually formulated only by the linguist). Sapir’s phonemic method won the day, producing an “American structuralism” that Hymes and Fought (1975: 997) consider far closer to the program of Leonard Bloomfield than of Franz Boas. The shared distinctive features of Sapirian and Bloomfieldian linguistics were to develop methods of structural description and apply them to both written and unwritten languages, to establish linguistics as an autonomous discipline, to record disappearing languages, and to clarify the genetic relationships of American Indian languages. Sapir, embroiled in the emergence of linguistics from its roots in anthropology and literature, did not share Bloomfield’s commitment to interdisciplinarity and public affairs.

7.  Assessment More than sixty years after his death, Boas’s intellectual heirs are still debating the legacy of his Americanist paradigm. Some have argued that he was atheoretical, others that he was inconsistent, none that he was not influential. In “The Study of Geography” (1887) Boas distinguished between the methods of natural scientists and historians, among whom he included geographers, anthropologists, and cosmologists. Inductive and deductive methods both had their place in science, depending on the topic to be studied. Later commentators have often emphasized one side of his thought and ignored its counterbalancing binary. Methodological caution was the tenor of his critiques; he attacked premature generalization rather than generalization per se. At the beginning of Boas’s career, linguistic and cultural methods and theories were not well integrated. His teaching program at Columbia privileged grammar, dictionary and texts as the products of field research and ideal dissertation topics, even for those students who were not linguists. Many of them complied, and they produced a corpus that underwrote the emergence of American Indian linguistics as a significant interface between anthropology and linguistics. Although North American anthropology departments are now moving away from linguistic anthropology as a separate subdiscipline, depending instead on linguists in independent linguistic departments,



Regna Darnell

those linguists who identify themselves as anthropologists tend to be Americanists. The Boasian link between language and culture remains powerful, although the interface area of communication, pragmatics and the study of meaning is no longer necessarily identified as linguistics.

References Baker, L.D. (1998). Rethinking Race at the Turn of the Century: W.E.B. Du Bois and Franz Boas. In From Savage to Negro: Anthropology and the Construction of Race, 1896–1954: 99–126. Columbia University Press. Berman, J. (1996). The Culture as it Appears to the Indian Himself: Boas, George Hunt, and the Methods of Ethnography. In G.W. Stocking (Ed.), Volksgeist as Method and Ethic. History of Anthropology 8: 215–256. University of Wisconsin Press. Boas, F. (1887). The Study of Geography. In F. Boas (1940): 639–647. ——— (1889). On Alternating Sounds. American Anthropologist 2: 47–53. ——— (1896). The Limitations of the Comparative Method of Anthropology. In F. Boas (1940): 270–280. ——— (1911a). Handbook of American Indian Languages. Bureau of American Ethnology Bulletin 40. ——— (1911b). The Mind of Primitive Man. Macmillan. ——— (1917). Introduction International Journal of American Linguistics. In F. Boas (1940):199–210. ——— (1940). Race, Language and Culture. Free Press. Campbell, L. & M. Mithun (Eds) (1979). The Languages of Native America: Historical and Comparative Assessment. University of Texas Press. Cole, D. (1999). Franz Boas: The Early Years, 1858–1906. Douglas and McIntyre. Darnell, R. (1990). Edward Sapir: Linguist, Anthropologist, Humanist. University of California Press. ——— (1998). And Along Came Boas: Continuity and Revolution in Americanist Anthropology. Benjamins. ——— (2001). Invisible Genealogies: A History of Americanist Anthropology. University of Nebraska Press. Hymes, D. & J. Fought (1975). American Structuralism. In T.A. Sebeok (Ed.), Current Trends in Linguistics: Historiography of Linguistics 13: 903–1173. Mouton. Mithun, M. (1996). The Description of the Native Languages of North America: Boas and After. In I. Goddard (Ed.), Handbook of North American Indians 17: Language: 43–63. Smithsonian Institution Press. Murray, S.O. (1994). Theory Groups and the Study of Language in North America. Benjamins. Powell, J.W. (1891). Indian Linguistic Families of America North of Mexico. Bureau of American Ethnology Annual Report for 1885–86: 7–139. Radin, P. (1927) [1957]. The Mind of Primitive Man. Dover. Sapir, E. (1916). Time Perspective in Aboriginal American Culture: A Study in Method. Canadian Department of Mines, Geological Survey, Memoir 90, Anthropological Series 13. ——— (1921). A Bird’s Eye View of North American Languages North of Mexico. Science 54: 408. ——— (1925). Sound Patterns of Language. Language 1: 37–51. ——— (1929). Central and North American Languages. Encyclopedia Britannica 5: 138–141. ——— (1933) [1949]. The Psychological Reality of the Phoneme. In D. Mandelbaum (Ed.), Selected Writings of Edward Sapir: 46–60. University of California Press.

Franz Boas

Stocking, G.W. Jr. (1968). Race, Culture and Evolution: Essays in the Historiography of Anthropology. Free Press. ——— (1974). The Boas Plan for the Study of American Indian Languages. In D. Hymes (Ed.), Traditions and Paradigms in the History of Linguistics: 454–484. Indiana University Press. Swadesh, M. (1951). Diffusional Cumulation and Archaic Residue as Historical Explanations. Southwestern Journal of Anthropology 7: 1–21. Valentine, L. & R. Darnell (Eds) (1999). Theorizing the Americanist Tradition. University of Toronto Press.


Cognitive anthropology Stephen C. Levinson Max Planck Institute for Psycholinguistics, Nijmegen

1.  Historical background All the various branches of social and cultural anthropology are as much concerned with variation in ideas as variations in cultural practices. It is a truism of the social sciences that one cannot study human behavior without studying the ideas and motivations that guide and drive it. Nevertheless, in all the social sciences barring linguistics (if such it is), the idea that systems of concepts should be explored through specialized technical methods, and that such systems of different cultural origin might be systematically compared in the search for underlying principles of organization, has never met with much welcome. It was to redress the balance that various movements to establish anthropological studies of cognition have been launched. Cognitive anthropology might be defined as the comparative study of cognition in its full cultural and linguistic context. It is the attention to the in-depth understanding of conceptual domains, primarily through language, together with comparison across unrelated languages and cultures that differentiates it from e.g., cross-cultural psychology (which tends simply to apply Western tests to non-Western populations). The term psychological anthropology on the other hand has come to denote largely the study of the affective side of human mental dispositions, through the success of the psychoanalytical tradition in American anthropology especially in the 1950s.1 Cognitive anthropology is not one of the main recognized branches of anthropology (which are, in the US at least, archaeology, physical anthropology, social and cultural anthropology and linguistic anthropology). Nevertheless the issues addressed have a long history, and many of the founding fathers of psychology (e.g., Wundt) and linguistics (e.g., Boas, Sapir) had an interest in the area, as of course did many of the foundational figures of anthropology (Lévy-Bruhl, Durkheim, Mauss, Rivers, and more recently Lévi-Strauss to name but a few). The designation ‘cognitive anthropology’ is though of more recent origin, although it predates the fashionable ‘cognitive sciences’. It became current as part of a movement,

1.  These generalizations are of course simplifications. See Shweder (1990) on distinctions between psychological anthropology, cultural psychology, cross-cultural psychology and ethnopsychology.

Cognitive anthropology

known equally as ethnoscience or the new ethnography, which emerged in American anthropology towards the end of the 1950s. At that time, a number of scholars proposed that the central object of study, namely (in the American tradition of cultural anthropology) ‘culture’, should be thought about not in terms of behavior or behavioral products, but in terms of a generative set of mental dispositions (see e.g., Goodenough 1964). Thus the anthropologist’s job should be construed as reconstructing the mental categories and dispositions that lead e.g., a native New Yorker to act, react, choose, categorize and speak like a New Yorker. Moreover, the reconstruction should proceed along tightly controlled lines: e.g., linguistic categorization should be investigated using formal methods of structural semantics, with parallel cognitive investigations. It should be noted that these ideas predated the rise of the cognitive sciences; they were thus an independent assertion of mentalism against the prevailing behaviorism of the day, and shared with the new sciences of the mind the parallel hybris that new formal methods would instantly recast the study of mind. This anthropological contribution was recognized in the early formation of the cognitive science movement, where it was proposed that it would play a central role (as in the original Sloan Foundation proposals with which serious funding of the movement began). And some of the intellectual contributions to cognitive science now seem so central to the larger movement that it is hard to believe that they were developed by anthropologists in the 1950s (an example is the use of semantic features, and predicateargument analysis, for the analysis of semantics and conceptual structure). But already by the early 1970s the cognitive anthropology movement was in decline: the hybris of the goals compared ill with the handful of detailed studies of a few, sometimes trivial, domains of human categorization, and the movement was energetically attacked by those (like Marvin Harris) who felt that cultural anthropology might be diverted from the practical study of e.g., economic and political systems. Later, the ‘objectivist’, ‘scientistic’ approach was also subject to the vitriol of those who believe socio-cultural anthropology is a hermeneutic, interpretative craft that can lay no claim to objective knowledge. Thus at the very moment when the cognitive sciences were coming into their own, cognitive anthropology all but succumbed to fratricide (or in the case of one early practitioner, Stephen Tyler, intellectual suicide). But before the movement lost center stage, it produced some landmark achievements. For example, Berlin & Kay (1969) demonstrated that there are significant universals in color terminology — an area which had appeared superficially to suggest rampant Whorfianism. Also, Berlin and collaborators established a systematic ethnobiology that continues successfully to the present day (Berlin 1992). And perhaps most importantly, fundamental insight into the structure of kin terminologies was achieved by Lounsbury, Conklin, Goodenough and others. A representative sample of the landmark papers are collected in the reader edited by Tyler (1969) (see also Casson 1981).



Stephen C. Levinson

Although this earlier North American movement survives in attenuated form (see e.g., Keller 1985; Holland & Quinn 1987), it is probably fair to say that from being a major provider of fundamental concepts to the cognitive sciences, it has become primarily a consumer. Linguistic anthropology, which always encompassed broader sociolinguistic issues, meanwhile turned away from the problems of lexical semantics, and has become more concerned with issues of verbal interaction and verbal art. Psychological anthropology on the other hand is broadening its interests beyond the affective, and may come to embrace the comparative study of cognition (see Stigler et al. 1990). Meanwhile, a number of parallel developments suggest that the time has come for a revitalization and reconstrual of the subject. One of those developments is the Francophile emphasis on cognitive issues in anthropology, in linear descent from Durkheim, Mauss, Hertz, to Lévi-Strauss, Goody, Sperber and younger scholars like Atran (1990) or Boyer (1993). Here structuralist perspectives in anthropology have bounced off inventive historians to emerge as a movement for the study of the historical and cultural setting for ideas. If one asks for example for an explanation for the rise of Western technology, current thought directs attention to the cognitive effects of widespread literacy, to the emergence of a form of discourse where ideas are legitimized not by reference to authority but to canons of reasoning and argumentation, and so on. Another parallel development has been the neo-Vygotskian investigations of practical knowledge by anthropologists associated with Cole & Scribner (1974; see also Wertsch 1985). Here the leading idea, that external practices might be in complex interplay with internal processes, has led to a series of empirical investigations of everyday practice, combined with a well-articulated if programmatic challenge to the cognitive sciences (see e.g., Lave 1988; Suchman 1987), where mental processes tend to be conceived of as devoid of cultural content or influence. The cognitive effects of literacy, schooling and formal mathematics can be investigated in interesting ways wherever populations (as in developing countries) are divided by differential exposure (see e.g., Rogoff & Lave 1984). The idea that ‘technologies of the intellect’ (Goody 1977, 1989) might be a focus for the study of cross-cultural differences is yet to be fully exploited. Apart from the obvious case of literacy, devices like the abacus (Stigler 1984), complex mnemonic systems like the quipu, machines that guide and ‘remember’ (like a loom that may build in measurements for cloths of different use), calendars that orchestrate planting and ritual seasons, navigational charts and devices and so on, have yet to be fully explored from this new point of view. A further related movement, still in its initial phases, has origins more in the intellectual line of Boas and Sapir, not to mention Whorf and Humboldt: it attempts to resuscitate the old questions about the relation of language to thought, and thus the implications of linguistic diversity to thinking (Lucy 1992a). Part of the impetus comes from cross-linguistic studies of child language: the semantic parameters used

Cognitive anthropology

to organize some domain can vary fundamentally (see e.g., Bowerman & Choi 1991). Thus when describing that domain the speaker’s attention must be distributed to features of the event or state to be described that are relevant to the language in question, a process Slobin calls ‘thinking for speaking’. How does a child come to learn these semantic parameters and the attentional habits that go with them? What are their consequences for ‘thinking for thinking’? Lucy (1992b) shows that speakers of languages like Yucatec, where nominals are more mass-like and plurals either not obligatory or not possible, are less likely to notice quantity in non-verbal memory tasks. Levinson and co-workers in the Cognitive Anthropology Research Group at the Max Planck Institute for Psycholinguistics have also been able to demonstrate ‘Whorfian’ effects in the spatial domain: speakers of languages which use different spatial parameters solve non-verbal spatial tasks in distinct ways (see e.g., Danziger & Baayen 1994; Brown & Levinson 1993). Yet another development, still new, has been a willingness to speculate on the evolution and origin of human intelligence. The origins of language was a favourite nineteenth century parlor topic, then taboo for two or three generations of scientists; now the taboo is broken with works by psycholinguists (Lieberman 1984) and sociolinguists (Bickerton 1990). More generally, there has been wide speculation that it was the complexities of human social organization that selected for ever higher forms of strategic reasoning (Byrne & Whiten 1988; E. Goody 1995). Recognizing complex intentions seems to require a special form of reflexive reasoning that has a logic of its own. Here there is a direct overlap with pragmatics, conceived as the study of the use of language in the inferential envelope provided by communicative settings. These partially convergent movements suggest that it is time to reconstrue the subject, as an area where the holistic vision of anthropology can contribute to the many specialized but fractionated sciences of the Mind. 2.  Reconstruing cognitive anthropology A new approach to the comparative study of human cognition can hardly ignore the enormous energy and effort and inventiveness which has characterized developments in the cognitive sciences over the last twenty years. The crucial step has been the treatment of the mind as an information processing device, with the running metaphor of successive generations of computing devices. The central questions in the cognitive sciences as a whole concern especially (a) epistemology — the question of how we know what we know, and (b) implementation — how we process information. The presumption through much of this work is that cognitive processes and even the overall architecture of cognitive content (e.g., the nature of internal representations, and their interactions) is essentially part of our biological endowment. In the one area



Stephen C. Levinson

where this might have been actively questioned, namely in the study of linguistic universals and variation, the dominant trend has been to assume that languages are all cut to the same pattern, or perhaps to some small set of patterns. Thus the cognitive sciences have placed, inter alia, the following questions on the agenda: 1. The nature of the underlying mental representations of the external world that are used to ‘compute’ e.g., the visual recognition of objects, the meaning of utterances, future actions, etc. 2. The degree to which both the representations and the computational operations over them are given by our biological endowment as opposed to being learned in the cultural and natural environment; here notions of ‘unlearnability’ have been central — can one show that certain conceptual structures by virtue of their formal character and the lack of negative feedback during learning cannot in principle be learned? 3. The nature of the learning involved in these processes. The comparative study of cognition has much to contribute to these questions. Without cross-cultural data there can only be presumptions about the intrinsic nature of representations and their universality (unless, as potentially in the case of low-level sensory perception, these are actually instantiated in fixed neural architecture). There is much that cognitive anthropology can do to set the record straight on the degree of semantic variation across languages, on the differing susceptibility of distinct populations to visual illusions or discriminations, to variations in verbal memory or mathematical reasoning or decision making. These are, despite the current lack of attention, indispensable kinds of information if the current goals of cognitive science are to be met. But there are more important tasks. An initial step is to redirect theoretical effort away from cognition viewed as the intrinsic property of single organisms, towards a view of cognition as a system for mediating the relation between organism and environment. A second step is to recognize that in the human case the environment is largely culturally constructed, and interactions with it are socially controlled. Thus we are dealing with the systematic interactions between two systems that are normally studied independently, namely cognitive systems and social systems. A third step is to attend to the complex feedback relations between these two systems: what individuals know, and how they think about it, is clearly dependent on cultural knowledge and practice (e.g., it can be shown that an abacus-user processes mathematical operations in a different way than those used to purely mental arithmetic, while solutions by writing depend crucially on the notational system). A fourth step is to explore the constraints which the intrinsic properties of each system, cognitive and social, place on the possible interactions between them. For example, some imaginable conceptual patterns are in principle ‘unlearnable’ — i.e., the underlying pattern cannot be

Cognitive anthropology

induced from finite samples. A far greater range are in practice unlearnable given the limitations of human memory and inference. Thus the properties of human cognition constitute a bottleneck which filters or restricts possible cultural ideas and patterns. In exactly comparable manner, constraints on social systems limit conceptual possibilities. Transmission of primary cultural information must take place within a time span (mostly in childhood and adolescence) before social reproduction; information and skills that are too complex to absorb in that limited period cannot be part of the common cultural background, as opposed to expert knowledge. Technologies for the transmission of knowledge also limit or enhance these possibilities. Complex societies with elaborate divisions of labour can seek to overcome some of these limitations by a division of cognitive skills. Indeed some social institutions can be thought of as themselves ‘distributed’ information processing machines, as in a bureaucracy, or when the crew of a ship jointly determines its course, each man processing specialized information and sending it on to the next (Hutchins 1990). A further step still is to ask how these complex interactions between cognitive and social systems have played a role in the co-evolution of the human brain in parallel with the range of social systems (Durham 1991). This way of thinking directs attention to the replication or transmission of ideas across individuals. How is this achieved? Obviously language is a crucial medium or channel, since it provides representations which are at once ‘private’ (aspects of individual cognition) and ‘public’ (aspects of shared knowledge). Nor is the transmissive power of language restricted to what is ‘said’ (more in a moment). But it is by no means the only medium of transmission. Anthropologists have long argued that the cultural shaping of events, behavior and the environment encodes covert public representations. But cultural information is not transferred only by ‘codes’, as Sperber (1987) has forcefully argued. It is also made available by our native inferential dispositions applied to cultural products and behaviors of all kinds. As Dawkins, Dennett and others have pointed out, the wheel carries with it wherever it goes the brilliant idea of the wheel. Artefacts are thus inspected for their underlying design intentions. In the same way we inspect human behaviors of all sorts for their intentional background. One category of human behavior is of peculiar importance, namely other-directed actions where the actions are so designed to make their underlying intentions clear. There may be little observable difference between just say weaving versus demonstrating how to weave, but the distinction is crucial. This category of actions constitute of course Gricean communicative acts (cases of meaning-nn), and involve very complex reflexive reasoning about what the agent intended the observer to think that s/he intended. The bulk of social learning must be transferred in this manner, since relatively little is explicitly explained. Language use also of course trades on this underlying communicative ability (Sperber & Wilson 1986), and for this reason what is not said but rather indexed or hinted at in the conduct of verbal interaction plays a crucial role in



Stephen C. Levinson

socialization (Ochs & Schieffelin 1990). Linguistic pragmatics and cognitive anthropology thus share a fundamental overlap of interests. The perceived degree of convergence between cognitive anthropology and the different branches of the study of language depends on one’s views on the role of language in cognition. Much modern speculation has assumed that internal conceptual representations are in part language-like, and thus that language is a mere encoding of language-independent conceptual representations of the environment. But recent ‘Whorfian’ findings (mentioned above) undermine this view. And some recent speculations about consciousness suggest that perhaps language plays an essential role, through its potential externalization of thoughts, in integrating otherwise unconnected mental capacities, those ‘inner conversations’ constituting the phenomenon of consciousness itself (Dennett 1991).

References Atran, S. (1990). Cognitive foundations of natural history. Cambridge University Press. Berlin, B. (1992). Ethnobiological classification. Princeton University Press. Berlin, B. & P. Kay (1969). Basic color terms. University of California Press. Bickerton, D. (1990). Language and species. University of Chicago Press. Bowerman, M. & S. Choi (1991). Learning to express motion events in English and Korean. Cognition 41: 83–121. Boyer, P. (Ed.) (1993). Cognitive aspects of religious symbolism. Cambridge University Press. Brown, P. & S.C. Levinson (1993). Explorations in Mayan cognition. Cognitive Anthropology Research Group, Working Paper 24. Byrne, R. & A. Whiten (Eds) (1988). Machiavellian intelligence. Clarendon Press. Casson, R. (Ed.) (1981). Language, culture and cognition. Macmillan. Cole, M. & S. Scribner (1974). Culture and thought. Wiley. Danziger, E. & H. Baayen (Eds) (1994). Annual Report of the Max Planck Institute for Psycholinguistics 1993. Max Planck Institute for Psycholinguistics. Dennett, D. (1991). Consciousness explained. Little, Brown. Durham, W. (1991). Coevolution. Stanford University Press. Goodenough, W. (1964). Cultural anthropology and linguistics. In D. Hymes (Ed.): 36–39. Goody, E. (Ed.) (1995). Social intelligence and interaction. Cambridge University Press. Goody, J. (1977). The domestication of the savage mind. Cambridge University Press. ——— (1989). The logic of writing and the organization of society. Cambridge University Press. Holland, D. & N. Quinn (Eds) (1987). Cultural models in language and thought. Cambridge University Press. Hutchins, E. (1990). The technology of team navigation. In J. Galegher, R.E. Kraut & C. Egido (Eds), Intellectual teamwork. Erlbaum. Hymes, D. (Ed.) (1964). Language in culture and society. Harper & Row. Keller, J. (Ed.) (1985). Directions in cognitive anthropology. University of Illinois Press. Lave, J. (1988). Cognition in practice. Cambridge University Press. Lieberman, P. (1984). The biology and evolution of language. Harvard University Press.

Cognitive anthropology

Lucy, J. (1992a). Language diversity and thought. Cambridge University Press. ——— (1992b). Grammatical categories and cognition. Cambridge University Press. Ochs, E. & B. Schieffelin (Eds) (1990). Language socialization across cultures. Cambridge University Press. Rogoff, B. & J. Lave (Eds) (1984). Everyday cognition. Harvard University Press. Shweder, R. (1990). Cultural psychology. In J. Stigler, R. Shweder & G. Herdt (Eds): 1–43. Sperber, D. (1987). On anthropological knowledge. Cambridge University Press. Sperber, D. & D. Wilson (1986). Relevance. Blackwell. Stigler, J. (1984). Mental abacus. Cognitive Psychology 16: 145–176. Stigler, J., R. Shweder & G. Herdt (Eds) (1990). Cultural psychology. Cambridge University Press. Suchman, L. (1987). Plans and situated actions. Cambridge University Press. Tyler, S. (Ed.) (1969). Cognitive anthropology. Holt. Wertsch, J. (1985). Culture, communication and cognition. Cambridge University Press.


Componential analysis Cliff Goddard University of New England (Armidale, Australia)

1.  Introduction Componential analysis (CA) in the broadest sense, also known as ‘lexical decomposition’, is any attempt to formalize and standardize procedures for the analysis of word meanings. CA often aspires to represent the cognitive or psychological reality of the speakers, and to shed light on correlations between language and culture. The idea that word meanings may be broken down into combinations of simpler components is an ancient one, supported by a range of facts. These include the efficacy of paraphrase, the intuitively felt relationships (such as antonymy, hyponymy, partonymy) between word meanings, the fact that sentences may be tautologous, contradictory or odd due to the interplay of the meanings of their constituent words. The assumption of decomposability underlies the definitional side of traditional lexicography. For expository purposes, methods in CA may be described under four headings.

2.  The structuralist tradition The application of structuralist principles to the lexicon was inaugurated by Trier (1934) and Porzig (1934) in their work on lexical fields (cf. Geckeler 1971) and consolidated by Hjelmslev (1943), who insisted that both the ‘expression-plane’ (phonology) and the ‘content-plane’ (meaning) of language be subjected to the same structuralist mode of analysis. Modern practitioners include Algirdas Greimas (1966), Bernard Pottier (1963, 1974) and Eugenio Coseriu (Coseriu & Geckeler 1974) in Continental Europe, John Lyons (1977) and Geoffrey Leech (1981) in Britain, and Eugene Nida (1964, 1975) and Adrienne Lehrer (1974) in the US. Basically, CA works by comparing and contrasting words within a semantic field, that is, a set of words in single conceptual domain, such as kin, parts of the body, colors or verbs of motion. A notation of semantic components (also called markers, features or semes) is devised to summarize the similarities and contrasts in the most economical way, a procedure analogous to distinctive feature analysis in phonology. Since Hjelmslev, a favourite example has been stage-of-life terms, such as those in (1).

Componential analysis

(1) man [human, +adult, +male] woman [human, +adult, −male] boy [human, −adult, +male] girl [human, −adult, −male]

Relations between semantic fields can be accommodated by means of additional features. For instance, the similarities between the words in (1) and stallion, mare, colt, filly can be shown by substituting horse (or equine) for human. A given feature may be absent or neutralised in some words, e.g., child and lamb may be analysed as [human, −adult] and [sheep, −adult], with gender absent or unspecified. CA of this type does not seek to be exhaustive. Only that part of the meaning is captured which is necessary to account for the meaning relations at hand. Because of this, the terms of the analyses change as the field is expanded or contracted. For instance, Pottier (1963) used the six features shown in (2) for the French equivalents of chair, armchair, stool, sofa and pouf. When in a later work he excluded pouf from consideration, the feature s6 became unnecessary. (2) s1 ‘with a back support’, s2 ‘with legs’, s3 ‘for one person’, s4 ‘for sitting on’, s5 ‘with arms’, s6 ‘made of rigid material’

CA analysts generally prefer binary features, but some allow multi-valued features when economy seems to demand it. For instance, Nida (1962) proposed the analysis in (3) for a set of verbs of motion, commenting that grouping the components in terms of the limbs involved, the order of movement and the relationship to the surface produces a more economical and revealing analysis than would be possible with exclusively binary components.


Limbs involved: Order of movement: Relation to surface:










4 limbs



111 or 222


1–3, 2–4

one foot not always on surface

one foot always on surface

one foot not always on surface

one foot not always on surface

2 limbs always on surface

Most CA analysts recognize a distinction between components which are operative across a whole semantic field (diagnostic components, markers or classemes), and idiosyncratic components specific to individual words (supplementary components, distinguishers or semes). Nida, for instance, acknowledges 〈speed〉 as an important supplementary component in the meaning of run, though it is not a useful feature in identifying differences across the whole field. The emphasis of this variety of CA is very much on succinctness of description.



Cliff Goddard

3.  Linguistic anthropology A largely independent structuralist tradition originated with the work of American ethnographers Floyd Lounsbury (1956) and Ward Goodenough (1956) on classificatory kinship categories, such as those found in the indigenous languages of North American and Australia. Viewed through the prism of the English kin system, such categories appear to denote an indefinitely long list of relatives, but structural analysis shows that unitary definitions may be ascribed using components such as gender, seniority, generation level, descent type (lineal vs. collateral) and relationship mode (consanguineal vs. affinal). An oft-cited example is Burling’s (1970) analysis of Njamal, where terms mama and kabali, for example, may refer to any of the string of individuals indicated in (4a), among others. (4) a.

Mama: Fa, FaBr, MoSiHu, FaFaBrSo, MoMoBrSo, etc. Kabali: FaMo, FaMoSi, FaFaBrWi, MoFaSi, MoMoBrWi, etc.

Taking account of the Njamal system of patrilineal moieties, however, Burling is able to state unitary meanings in terms of a small number of components, as in (4b). Here the binary symbols Me and Mo indicate membership in ego’s vs. opposite moiety, Sm and Sf for male and female sex of the referent, and Om and Of for the sex of the older member of the pair. The multivalued symbol G designates generational level. G+1, for instance, stand for the first ascending generation; G2 (where the superscript lacks a plus or minus sign) includes both “grand” generations equally. (4) b. Mama: Kabali:

G+1MeSm G2MoOf

For anthropologists, a major attraction of CA is that it promises to bring to light the conceptual dimensions of the indigenous system. Aside from kinship studies, CA has been used in linguistic anthropology to examine such diverse areas as colors, meal terms, illnesses and taxonomies of the natural world. 4.  Generative and typological studies The adoption of CA into generative linguistics was pioneered by Jerrold Katz in the late 1960s with the goal of integrating lexical semantics and formal syntax (cf. Katz & Fodor 1963; Katz 1972). Divergent trends were introduced by George Lakoff (1972) and James McCawley, leaders of the now defunct ‘generative semantics’ movement. The most prominent CA practitioner in mainstream generative linguistics today is Ray Jackendoff (1990, 1991, 2002). From the beginning, generative CA has shown a natural interest in the semantic analysis of verbs and predicate-argument structure. Even simple analyses like those in

Componential analysis

(5) can capture a range of semantic and syntactic properties. For instance, (5a) shows some verbs in the natural class of ‘causative-inchoatives’; (5b) makes explicit some of the semantic and argument-structure properties of the converses give and take; (5c) explains the polysemy of persuade. (5) a.

x kill y: x liquify y: x open y:

x cause [come about [y not-alive]] x cause [come about [y liquid]] x cause [come about [y open]]

b. x gives y to z: x takes y from z: c.

x [cause [z have y]] x [cause [z not-have y]]

x persuade y to do z: x persuade y that z:

x cause [come about [y intend [do z]] x cause [come about [y believe z]]

Current generative-style CA is characterised by a much richer formal structure of hierarchically grouped category types, shown in annotated ‘tree diagrams’ or labelled bracketings, and designed to effect an interface with the syntactic part of the grammar. Examples (6) and (7) show analyses of transitive chase and climb by Katz (1987) and Jackendoff (1990), respectively.

(6) chase:

((Activity)[NPS]) x

() (Physical)


(Movement) (Fast)

((Catching)[NP, VPS]) x

((Direction)[NP, VPS])



() ((Towards location of)[NP, VPS]) x


(7) climb: [Event GO([Thing ]i, [Path TO ([Place TOP-OF [Thing ]j])])]

Jackendoff ’s work has tended to evolve in the direction of increasing abstractness. For example, example (7) above includes a conceptual function [PathTO (PLACE)], but in subsequent work (Jackendoff 1991), this was further decomposed into the more abstract representation of features and functions shown in (8). (8) TO X=

+b, –i DIM Id DIR + Space BDBY ([Thing/Space X])



Cliff Goddard

Some typological studies into the lexicon have employed CA methods, such as Leonard Talmy’s (1985) work on the different patterns languages use to ‘package’ lexical meaning. English motion verbs, for instance, typically incorporate a manner component (e.g., run, swim, fly, roll), whereas Spanish motion verbs incorporate ‘path’ (e.g., entrar ‘move in’, salir‘move out’, pasar move through’, volver ‘move back’).

5.  Natural semantic metalanguage (NSM) The key idea behind NSM semantics, i.e., to use systematic reductive paraphrase as a method of semantic decomposition, goes back at least to Leibniz. It was revived in the twentieth century by Andrzej Boguslawski (1972, 2000) and Jurij Apresjan. The main theoretical figure today is Anna Wierzbicka (1972, 1985, 1992, 1996, 1997, 1999, 2003; cf. Goddard 1998; Goddard & Wierzbicka 2002). The NSM approach differs from other CA methods in several respects. First, it disavows formalisms (technical features, logical symbols, etc.) entirely, on the grounds that any formalism can only be explained and understood through natural language. Second, it posits the existence of a highly restricted inventory of semantic-conceptual primitives (semantic primes), hypothesized to have lexical (or morphemic) exponents in all languages. Third, it accords little importance to the notion of semantic fields, striving instead for exhaustive semantic analysis of each word in its own right. Fourth, it generates more detailed and elaborate decompositions. A recent version of the inventory of semantic primes (based on Goddard & Wierzbicka 2002) is shown in (9). The number of proposed semantic primes is in the mid-sixties. Semantic primes and their combinatorial patterns comprise a minilanguage which, according to NSM researchers, has the same expressive power as a full natural language, and as such constitutes the ideal tool for semantic explication. (9)

Substantives: i, you, someone, something/thing, people, body Determiners: this, the same, other/else Quantifiers: one, two, all, many/much, some Descriptors: big, small Evaluators: good, bad Mental predicates: want, think, know, feel, see, hear Speech: say, words, true Events and actions: do, happen, move Existence and possession: there is/exist, have Life and death: live, die Time: when/time, now, after, before, a long time, a short time, for some time, moment

Componential analysis

Space: where/place, here, above, below, side, near, far, inside, touching, be (somewhere) Logical concepts: not, maybe, if, can, because Intensifier, augmentor: very, more Taxonomy, partonomy: kind of, part of Similarity: like/way/as

The universality of the NSM prime inventory remains controversial, though crosslinguistic studies (Goddard & Wierzbicka 1994, 2002) indicate that it can be isolated in languages as diverse as Spanish, Malay, Lao, Mbula, Chinese, Polish, Russian, French and Korean. In any case, the semantic components listed in (9) are much clearer and more translatable than the technical markers, features and symbols used in other CA approaches. There is a large body of empirical-descriptive work in the NSM framework, on various languages, spanning lexical areas such as speech acts, emotions, artefacts, colors, mass nouns, kin, body parts, natural kinds and functional categories, as well as some areas of grammatical and illocutionary semantics. Some examples follow, chosen from domains of particular relevance to pragmatics. Emotion concepts are important because it is through them that people of a given culture interpret and negotiate much social interaction (Wierzbicka 1999; Harkins & Wierzbicka 2001). Explication (10) for English sad shows several features typical of the NSM work on emotions, in particular the subjective orientation and the reliance on what can be called a ‘prototypical cognitive scenario’. The scenario for sad involves an awareness that ‘something bad happened’ (not necessarily to me) and an acceptance of the fact that one can’t do anything about it. This is compatible with the word’s wide range of use; for example, that I may feel sad when I hear that my friend’s pet dog died, or when I think about some unpleasant bickering at work. Such explications can capture subtle nuances of meaning, making it possible to elucidate the similarities and differences between sad, unhappy, miserable, depressed, and so on. (10) X was sad X felt something bad like people feel when they think like this: “I know that something bad happened I don’t want things like this to happen I can’t think like this: I will do something because of it now I know that I can’t do anything”

Social categories, such as friend, mate, colleague, comrade, etc. encapsulate “prepackaged” social construals encoded in the lexicon. Naturally they vary greatly across cultures. For example, Wierzbicka (1997) argues that several hundred years ago, the



Cliff Goddard

English word friend designated an exceptional, intimate, and life-long bond but that the meaning has weakened as social relationships have become more flexible and broadly based. The contemporary concept can be explicated as in (11). (11) (my) friend I think about this person like this: “I know this person well when I do things with this person, I feel something good when I am with this person, I feel something good” I know that this person thinks the same about me

Explications (12a) and (12b) below show how two English speech-act verbs (threaten and warn) can be explicated, so as to make their similarities and differences explicit. Speech-acts too differ greatly across language and cultures. Since they represent a kind of cultural catalogue of interaction types, accurate explications can provide valuable input to culturally oriented approaches to pragmatics (Wierzbicka 2003; Goddard in press). Notice the ‘first-person’ format of the explications in (12), which is a characteristic feature of the NSM approach to performative verbs. (12) a. person X threatened person Y X said to Y something like this about something (Z): “I want you to know that if you do Z I will do something bad to you” X said this because X wanted Y not to do Z b. person X warned person Y about something (Z) X said to Y something like this about something (Z): “I want you to know that something bad can happen to you I think that it is good for you to know it if you know it you can do something because of this” X said this because X wanted Y to know it

As just illustrated, some types of lexical meanings (such as those for emotions, social categories, and speech-acts) can be explicated directly in terms of semantic primes. Other, more complex, types of lexical meanings (in particular, for words from the concrete lexicon) can only be explicated in stages, using both semantic primes and intermediate-level “semantic molecules” together. The term semantic molecule refers to a packet of semantic components which both exists as a lexical meaning in a given language and functions as a “chunk” in more complex meanings in that language. For example, explications for words such as cat, dog, horse, mouse, etc., begin with the component ‘a kind of animal’ (followed by further specifications of the characteristic habitat, size, appearance, behavior, and relation with people (Wierzbicka 1996)). The term ‘kind (of)’ is a semantic prime, but the same is not true of ‘animal’. Rather, it is a semantic molecule which requires its own explication in turn. This can be done as shown in (13).

Componential analysis

(13) animal a kind of living thing living things of this kind are like people in some ways, they are not like people in other ways some parts of their bodies are like parts of people’s bodies, other parts of their bodies are not like people’s bodies there are many kinds of living things of this kind

Likewise, explications for artefact words, such as cup, knife, umbrella, etc., require the concept of ‘make’ as a semantic molecule, because such words designate kinds of thing which ‘people make’. Explications for physical activity verbs, such as hit, eat, run, etc., require body-part concepts such as ‘hands’, ‘mouth’ and ‘legs’ as semantic molecules. The concrete lexicon, in other words, has a multi-level hierarchical structure in which complex meanings are built up from combinations of semantic primes and semantic molecules (themselves composed ultimately of primes). The NSM metalanguage of semantic primes can be used not only for semantic decomposition, but as a notation for writing cultural scripts, i.e., hypotheses about culturally shared assumptions, norms and expectations which help regulate interaction in different cultural settings (Goddard & Wierzbicka Eds, 2004; Goddard in press). By using simple cross-translatable expressions, this approach to cultural pragmatics (or ‘ethnopragmatics’) avoids the terminological ethnocentrism implicit in the use of English-specific technical terms and descriptors.

6.  Other trends and problems Other trends in CA are to be found in Artificial Intelligence research, e.g., Roger Schank (1973) and Yorick Wilks (1975), in the work of Russian linguists such as Igor Mel’čuk and Alekandr Zholkovskij (Mel’čuk & Zholkovskij 1970) within the Meaning-Text Model, and in cognitive psychology, as in the work of Phillip Johnson-Laird (Miller & Johnson-Laird 1976). All systems of CA face challenges in relation to clarity, completeness and universality. It seems indisputable that semantic analysis, if it is to be predictive and verifiable, must strive for maximum intelligibility, but CA is often either obscurely technical or unmanageably dense. Most CA systems are also unabashedly partial, being tailored to one area of lexicon at a time, and it is questionable whether they could be integrated to model overall semantic knowledge. A lack of attention to the issue of the universality or translatability of components renders most CA systems vulnerable to charges of ethnocentrism. CA has also incurred criticism, especially from the new ‘cognitive semantics’ movement, for its alleged inability to model vagueness and prototype effects, and its alleged arbitrariness and psychological implausibility.



Cliff Goddard

References Apresjan, J. (1974). Leksiceskaja Semantika — Sinonimeceskie Sredstva Jazyka. Nauka. [Transl. A. Arbor (1992), Lexical Semantics: User’s Guide to Contemporary Russian Vocabulary. Karoma Publishers Inc.] ——— (2000). Systematic Lexicography [Transl. K. Windle]. Oxford. Boguslawski, A. (1970). On semantic primitives and meaningfulness. In R.J.A. Greimas, M.R. Mayenowa & S. Zolkiewski (Eds), Sign, Language and Culture: 143–152. Mouton. Burling, R. (1970). Man’s Many Voices: Language in its Cultural Context. Holt, Rinehart & Winston. Coseriu, E. & H. Geckeler (1974). Linguistics and semantics. In T.A. Sebeok (Ed.), Current Trends in Linguistics: 103–171. Mouton. Geckeler, H. (1971). Strukturelle Semantik und Wortfeldtheorie. Fink. Goddard, C. (1998). Semantic Analysis: a practical introduction. Oxford University Press. ——— (Ed.) (In press). Ethnopragmatics: Understanding discourse in cultural context. Mouton de Gruyter. Goddard, C. & A. Wierzbicka (Eds) (1994). Lexical and Semantic Universals — Theory and Empirical Findings. Benjamins. ——— (Eds) (2002). Meaning and Universal Grammar — Theory and Empirical Findings, vols. 1 & 2. Benjamins. ——— (Eds) (2004). Cultural Scripts. Intercultural Pragmatics 1(2) (Special issue). Goodenough, W.H. (1956). Componential analysis and the study of meaning. Language 32(1): 195–216. Greimas, A.J. (1966). Sémantique structurale. Recherche de méthod. Larousse. [Transl. D. McDowell, R. Schleifer & A. Velie (1983), Structural Semantics: An Attempt at a Method. University of Nebraska Press] Harkins, J. & A. Wierzbicka (Eds) (2001). Emotions in Cross-Linguistic Perspective. Mouton de Gruyter. Hjelmslev, L. (1943). Omkring sprogteoriens grundlæggelse. Lunos. [Transl. F.J. Whitfield (1953), Prolegomena to a theory of language. Indiana University] Jackendoff, R. (1990). Semantic Structures. MIT Press. ——— (1991). Parts and boundaries. In B. Levin & S. Pinker (Eds), Lexical and Conceptual Semantics: 9–47. Blackwell. ——— (2002). Foundations of Language: Brain, meaning, grammar, evolution. Oxford University Press. Katz, J.J. (1972). Semantic Theory. Harper & Row. ——— (1987). Common sense in semantics. In E. LePore (Ed.), New Directions in Semantics. 157–234. Academic Press. Katz, J.J. & J.A. Fodor (1963). The structure of a semantic theory. Language 39: 170–210. Lakoff, G. (1972). Linguistics and natural logic. In D. Davidson & G. Harman (Eds), Semantics of Natural Language: 545–665. Reidel. Leech, G. (1981). Semantics: The study of meaning. Penguin. Lehrer, A. (1974). Semantic Fields and Lexical Structure. North-Holland Publishing Company. Lounsbury, F.G. (1956). A semantic analysis of the Pawnee kinship usage. Language 32(1): 158–194. Lyons, J. (1977). Semantics. Cambridge University Press. Mel’čuk, I.A. & A.K. Zholkovskij (1970). Toward a functioning Meaning-Text Model of language. Linguistics 57: 10–47. Miller, G.A. & P.N. Johnson-Laird (1976). Language and Perception. Cambridge University Press. Nida, E.A. (1964). Linguistic and semantic structure. In A.S. Dil (Ed.) (1975), Language Structure and Translation. Essays by Eugene A. Nida. Stanford University Press.

Componential analysis

——— (1975). Componential Analysis of Meaning. Mouton. Porzig, W. (1934). Wesenhafte Bedeutungsbeziehungen. Beiträge zur deutschen Sprache und Literatur 58: 70–97. Pottier, B. (1974). Linguistique Générale. Klincksieck. ——— (1963). Recherches sur l’analyse sémantique en linguistique et en traduction méchanique. Université de Nancy. Schank, R.C. (1973). Conceptual information processing. North-Holland. Talmy, L. (1985). Lexicalization patterns: semantic structure in lexical forms. In T. Shopen (Ed.), Grammatical Categories and the Lexicon: 57–149. Cambridge University Press. Trier, J. (1934). Das sprachliche Feld. Eine Auseinandersetzung. Neue Jahrbücher für Wissenschaft und Jugendbildung 10: 428–49. Wierzbicka, A. (1972). Semantic Primitives. Athenäum. ——— (1985). Lexicography and Conceptual Analysis. Karoma. ——— (1992). Semantics: Culture and cognition. Oxford University Press. ——— (1996). Semantics: Primes and universals. Oxford University Press. ——— (1997). Understanding Cultures through their Key Words. Oxford University Press. ——— (1999). Emotions across Languages and Cultures. Cambridge University Press. ——— (2003). Cross-cultural Pragmatics: the semantics of social interaction. Mouton de Gruyter. [Expanded 2nd edition, originally published 1991]. Wilks, Y. (1975). A preferential, pattern-seeking semantics for natural language inference. Artificial Intelligence 6: 53–74.


Cultural scripts Cliff Goddard University of New England (Armidale, Australia)

1.  Introduction The term ‘cultural script’ refers to a technique for articulating culture-specific norms, values, and practices in terms which are clear, precise, and accessible to cultural insiders and outsiders alike. This result is possible because cultural scripts are formulated in the Natural Semantic Metalanguage (NSM) of semantic primes, a highly constrained ‘mini-language’ of simple words and grammatical patterns which evidence suggests have equivalents in all languages. The technique emerged in the mid-1990s, growing out of ‘cross-over research’ between semantics and cross-cultural pragmatics. Recent publications include the collective volumes Cultural Scripts (Goddard & Wierzbicka Eds 2004) and Ethnopragmatics (Goddard Ed. 2006). Languages/cultures to which the cultural scripts method has been applied include various varieties of English (mainstream or “Anglo” English, Australian English, American English, Singapore English), Ewe, French, German, Japanese, Korean, Malay, Mandarin Chinese, Polish, Russian, Spanish, and Yankunytjatjara. A representative selection of studies is supplied in the references below. The main goal of the cultural scripts approach is to understand speech practices from the perspective of the speakers themselves. To achieve this requires one to work concurrently in cross-cultural semantics, because to understand speech practices in terms which make sense to the people concerned, one must be able to understand the meanings of the relevant culturally important words – words for local values, social categories, speech-acts, and so on. Important words and phrases of this kind often qualify for the status of cultural key words (Wierzbicka 1997). One of the attractions of the natural semantic metalanguage is that it can be used equally for writing cultural scripts and for doing cross-cultural semantics, thus enabling one to draw out the connections between them. While not disregarding evidence from ethnography, sociological studies, literature, and so on, the cultural scripts approach accords particular importance to linguistic evidence. Aside from cultural key words, other kinds of linguistic evidence which can be particularly revealing include: common sayings and proverbs, frequent collocations, conversational routines and varieties of formulaic or semi-formulaic speech, discourse particles and interjections, and terms of address and reference – all highly

Cultural scripts

interactional aspects of language. A wide variety of data-gathering methods can be used, including the classical linguistic fieldwork techniques of elicitation, naturalistic observation and text analysis, consultation with informants, native speaker intuition, corpus studies, role-plays, questionnaires, discourse-completion tasks, and examination of literary materials and other cultural products.

2.  Semantic primes: The language of cultural scripts The cultural scripts technique relies crucially on a metalanguage of about 65 semantic primes, i.e., simple indefinable meanings, which appear to be expressible by words or wordlike expressions in all languages; for example (to use English exponents) – someone, something/thing, people, say, do, think, want, good, bad, if, because. Research indicates that these and other semantic primes can be expressed equally well and equally precisely in other languages; and furthermore, that they have an inherent combinatorial grammar which also manifests itself equally in all languages, albeit with language-specific formal variations (Wierzbicka 1996a; Goddard & Wierzbicka Eds 2002; Peeters Ed. 2006). Semantic primes can therefore be safely used as a common code for writing cultural scripts, free from the danger of ‘terminological ethnocentrism’ and with maximum clarity and resolution of detail. In the words of the distinguished anthropologist Roy D’Andrade (2001:246), the method “offers a potential means to ground all complex concepts in ordinary language and translate concepts from one language to another without loss or distortion in meaning”.

3.  High-level scripts: Examples and observations Cultural scripts exist at different levels of generality, and may relate to different aspects of thinking, speaking, and behaviour. To illustrate, script [A] is arguably a high-level script (sometimes termed a master script) of Anglo culture, expressing a cultural preference for something like personal autonomy; cf. Wierzbicka 2003[1991]. Script [B] is arguably a master script of Russian culture, expressing a cultural endorsement of, roughly speaking, an ‘expressive’ stance in speech and action (Wierzbicka 2002a). Script [C] is arguably a master script of Colombian Spanish culture, endorsing displays of something like ‘personal warmth’ (Travis 2006, 2004). In all three cases, it can be argued that the high-level concern captured in these scripts is played out in detail by way of a whole family of related speech-practices in their respective speech cultures. High-level scripts such as these are often closely associated with cultural key words, such as English freedom (and free), Russian iskrennost’ roughly, ‘sincerity’, and Spanish calor humano ‘human warmth’, respectively.



Cliff Goddard

[A] A high-level Anglo cultural script connected with ‘personal autonomy’ people think like this: when a person does something, it is good if this person can think like this: “I am doing this because I want to do it” [B]

A high-level Russian cultural script connected with ‘expressiveness’ people think like this: it is good if a person wants other people to know what this person thinks it is good if a person wants other people to know what this person feels

[C] A high-level Colombian Spanish cultural script connected with interpersonal ‘warmth’ people think like this: when I feel something good towards someone, it is good if this person knows that I feel like this because of this, it is good if I do some things when I am with this person, it is good if I say some things when I am with this person

Given the heterogeneity of any society, it is obvious that not every member of Anglo, Russian or Colombian Spanish culture would accept or endorse scripts [A], [B] and [C], but the claim is that even those who do not personally accept or identify with the content of a script are nonetheless familiar with it, i.e., that it forms part of the interpretative backdrop to discourse and social behaviour in a particular cultural context. Particular cultural scripts are of course not necessarily unique to any given culture. On the contrary, similar or identical scripts can recur in many different cultures, reflecting similarities and affiliations at the ‘trait’ level between different cultures. It should also perhaps be stated that cultural scripts change and develop over time and vary across geographical and social space. Returning to scripts [A]–[C] above, it can be seen that they are hinged around evaluative components: ‘it is good if —’. Evaluative components can also take the form ‘it is not good if —’, ‘it is bad if —’, ‘it is not bad if —’, etc.; or other variants such as ‘it can be good if —’ and ‘it can be bad if —’. Many cultural scripts are of this general format. Another kind of framing component, useful for other scripts and in other contexts, concerns people’s perceptions of what they can and can’t do: ‘I can say (think, do, etc.) —’ and ‘I can’t say (think, do, etc.) —’. Lower-level, more specific, scripts are often introduced by ‘when’-components and ‘if ’-components, representing relevant aspects of social context. Script [D], for example, is a lower-level script linked with Anglo ‘personal autonomy’. It expresses the Anglo distaste for abrupt directives, reflected in many ways in the phraseology and discourse patterns of mainstream Anglo English (Wierzbicka 2006a: Ch. 2, 2006b).

Cultural scripts

[D] An Anglo cultural script for avoiding ‘strong’ directives people think like this: when I want someone to do something, it is not good if I say something like this to this person: “I want you to do it    I think you will do it because of this”

Mere possession of a common language does not mean that people necessarily share all their cultural scripts. Particularly with a global language such as English, there can be marked regional and social variations, as one would expect of societies with different histories and lived experiences. The cultural scripts of ‘non-Anglo’ Englishspeaking societies can differ extensively from Anglo English varieties. Wong (2004a, b) shows that Singapore English, for example, has no cultural scripts endorsing Anglostyle personal autonomy in the style of scripts [A] and [D], and, correspondingly, that Singapore English speakers employ a very different interactional style as far as directives are concerned. 4.  Communicative style: Some further examples Even different varieties of Anglo English, such as Australian English, British English, and American English, may have significantly different cultural scripts capturing different aspects of interactional style. For example, Wierzbicka (1999: Ch. 6) has argued that Anglo-American English, even more than other varieties of Anglo English, values and encourages people to routinely display ‘good feelings’ that they may not necessarily feel, and, conversely, to suppress ‘bad feelings’ whose display may be seen as serving no useful purpose or as being unpleasant for other people. This is reflected, for example, in what has been called the American ‘Smile Code’ (Klos Sokol 1997: 117): “In American culture, you don’t advertise your daily headaches; it’s bad form; so you turn up the corners of the mouth – or at least try – according to the Smile Code.” Another reflection of this attitude is the high frequency of the word great in American discourse (cf. Wolfson 1983: 93), both as a modifier (especially of the verb to look, e.g., You look great! or Your X (hair, garden, apartment, etc.) looks great!) and as a response particle, e.g., That’s great! or Great!. The importance of positive feelings is also reflected in the key word status of the words happiness and fun in American society. To capture these attitudes, in part, Wierzbicka has proposed the script given in [E]. [E] An Anglo-American cultural script for projecting ‘positive feelings’ in verbal interactions people think like this: when I say something to other people, it is good if these people think that I feel something good, it is not good if these people think that I feel something bad



Cliff Goddard

Scripts like the one presented in [E] touch upon the area sometimes termed ‘communicative style’. Indeed, a significant part of the corpus of work on cultural scripts concerns communicative styles and strategies in different languages and cultures. As another example, script [F] below has been proposed as one of a suite of Malay cultural scripts enjoining caution in speech and action generally, and in particular, caution about other people’s feelings (Goddard 1997, 2000). Many traditional Malay sayings and expressions echo this theme, such as jaga hati orang ‘look after people’s feelings’, memilihara perasaan ‘looking after feelings’, and bertimbah perasaan ‘weighing feelings’. Commentators on Malay culture invariably mention the value placed on showing consideration and concern for others, on interpersonal sensitivity, and on harmonious personal relations. [F] A Malay cultural script for verbal caution about others’ feelings people think like this: it is not good if when I say something to someone, this person feels something bad because of it because of this, when I want to say something to someone it is good if I think about it for some time before I say it

To a certain extent, this script overlaps with an Anglo script which encourages people not to ‘hurt other people’s feelings’ unnecessarily. However, the Malay script goes beyond this to spell out a specific strategy — namely, a period of premeditation before saying anything, and this creates a quite different communicative mode to Anglo ways of speaking. Other studies of communicative style using the cultural scripts approach include: Ameka (1994, 2006) on Ewe; Goddard (1992, 2006, 2009) on Yankunytjatjara and Australian English; Hasada (2006) on Japanese; Peeters (2000) on French and English; Wierzbicka (1996c, 1998) on Chinese and English, and German.

5.  Culture-specific concepts in cultural scripts To formulate certain kinds of cultural scripts optimally requires not only semantic primes, but also certain ‘semantic molecules’. Semantic molecules are complex word-meanings which function as chunks or units in cultural scripts and/or semantic explications. Many cultural scripts must include as semantic molecules the concepts of ‘men’, ‘women’ and ‘children’; for example, the rules governing the usage of T/V pronoun forms and other forms of address in European languages (Wierzbicka 2004). Script [G] gives a similar example from a different part of the world. According to Ameka and Breedveld (2004), this is an ‘areal cultural script’ shared by many languages of West Africa. It specifies that one cannot say a person’s name, when speaking to a person, if this person is not thought of as a child. In this script, the terms ‘child’ and ‘name’ are both semantic molecules (and are marked as such by the notation [M]).


Cultural scripts

Cultural script for name avoidance in adult address in West African languages people think like this: if I think about someone like this: “this person is not a child [M]”, when I want to say something to this person, I can’t say this person’s name [M]

The semantic molecules ‘men’, ‘women’ and ‘children’ may well be universal or near-universal in the world’s languages, since they seem to represent a widely shared system of basic social categorization (Goddard & Wierzbicka to appear). Cultural scripts may draw also on language-specific semantic molecules, representing language-and-culture-specific social categorizations. For example, Yoon (2004) has shown that certain Korean scripts make reference to the Korean social category of noin (roughly:) ‘respected old people’. The word noin is a cultural key word for Korean. Script [H] captures the culturally expected attitude of younger Koreans when they are with noin (Yoon 2004). This includes seeing noin as ‘above’ them, being aware of verbal and non-verbal constraints, a perceived inability to defy the expressed wishes of old people (and even a positive attitude towards complying with their will), and the perceived need for caution in order to avoid causing them any negative feelings. [H] A Korean cultural script for interacting with noin people think like this: when I am with some people, if these people are noin [M] I have to think like this: “these people are not people like me, these people are people above me because I am with these people now I cannot do some things, I cannot say some things, I cannot say some words if these people say to me: “I want you to do something”, I can’t say to them: “I don’t want to do it” if these people want me to do something, it will be good if I do it it will be very bad if these people feel something bad because of me”

Ye (2004a) has shown that many Chinese interactional norms hinge on the distinction between the Chinese social categories of shúrén (roughly:) ‘an acquaintance, someone known personally’ and shēngrén (roughly:) ‘a stranger, a non-acquaintance’. Script [I] shows one such script. It sets out the perceived obligation, in Chinese society, to enact a certain kind of greeting routine (dă zhāohu), when meeting after some time with people who are shúrén. [I] A Chinese cultural script for dă zhāohu routine with shúrén people think like this: when I see a shúrén [M], if I have not seen this person for some time I have to say something like this to this person: “I see you now because of this I know that you are doing something now I want to know more about it”



Cliff Goddard

if I say this, this person can think because of this that I feel something good towards this person if I don’t say this, this person can think that I feel something bad towards this person I don’t have to say something like this to a person if this person is not a shúrén [M]

It should be clear at this point that cultural scripts can ‘reach down’ to very specific details of communicative practice, such as name avoidance, greeting routines, responses to particular conversational moves, and the like. Naturally, these more specific scripts are often longer and more involved than higher-level scripts.

6.  Rhetorical speech practices Cultural scripts can be used to develop an improved description and interpretation of rhetorical speech practices such as (to use conventional labels): active metaphor, irony, sarcasm, hyperbole, and euphemism (Goddard 2004a; Wierzbicka 2002b, 2004). The problem with these conventional labels is that they gloss over major differences between languages and between contexts. Just as there is no unitary phenomenon of ‘directness’ in terms of which speech styles in different languages can be compared, neither is there any unitary phenomenon of ‘metaphor’ or ‘irony’. The concept sometimes termed ‘active metaphor’, for example, is an artefact of a particular cultural tradition which can be traced back to classical Greek rhetoric. It encapsulates a complex meaning which lacks precise equivalents in many, if not most, of the world’s languages. Furthermore, there are many cultures, such as the Yankunytjatjara of Central Australia, in which active metaphorizing is marginal at best (Goddard 2004a). Cultural scripts allow us to articulate hypotheses about shared culture-specific understandings of particular ‘ways with words’ without recourse to technical English-specific labels. Goddard (2004a) has argued, for example, that active metaphorizing in English can be understood in terms of script [J] below (presented here in a slightly revised and improved version). This script sums up a chunk of cultural common knowledge about Anglo speech practices; namely, that speakers sometimes knowingly use words which can express a meaning different from the intended meaning, with a view to making the listener think about what is being said – in more abstract terms, in the interests of securing cognitive engagement. The speech practice described in script [J] is not universal, but is linked with culture-specific goals of expressiveness, originality, and individuality. [J]

An Anglo cultural script about active metaphorizing and related speech practices people think like this: sometimes when a person wants to say something about something, this person says it with some words, not with other words, because this person thinks like this:

Cultural scripts

“I know that these words can say something else I want to say it with these words because if I say it like this, people will have to think about it I want this” it can be good if a person can say things in this way

7.  Scripts for ways of thinking and feeling, and for beliefs Cultural scripts are not necessarily confined to matters directly related to ways of speaking, or more generally, to communication. They may also be employed to articulate cultural preferences for particular ways of thinking and feeling; in other words, to describe aspects of cognitive style and emotional style. Wierzbicka (1999: Ch 6) has argued that Anglo-American culture favours a cognitive stance which may be dubbed ‘positive thinking’ (using a common ethno-description of the culture concerned). It can be partially portrayed as in script [K]. Traditional Chinese culture, by contrast, encouraged an attitude which can be partially captured in the ‘Middle Way’ script given in [L], associated with Buddhism and Confucianism. [K]

An Anglo-American cultural script for ‘positive thinking’ people think like this: it is good if a person can often think that something good will happen it is good if a person can often feel something good because of this

[L] A Chinese cultural script for the philosophy of the ‘Middle Way’ people think like this: when something very bad happens to me, it is good if I think like this: “something good can happen to me afterwards because of this” if I think like this, I will not feel something very bad this is good when something very good happens to me, it is good if I think like this: “something bad can happen to me afterwards because of this” if I think like this, I will not feel something very good this is good

Cultural scripts can also be employed to spell out widespread cultural beliefs — beliefs which may be profoundly explanatory of aspects of communicative practice. Goddard (1997) has argued that traditional Malay culture includes a high-level ‘belief script’ concerned with the concept of balasan, a noun derived from the verb balas ‘to return (to someone), return in kind’. It is presented in [M]. Numerous traditional Malay sayings convey the message that a person’s deeds, whether good or bad, will be repaid in kind: Setiap perbuatan, baik atau jahat, akan ada balasan ‘every deed, whether good or wrong, will have its balasan’. This script leaves it open as to whether the balasan ‘return’ is likely to come from other



Cliff Goddard

people, or in the form of an apparently inexplicable event (which can be interpreted as the will of God), or whether the balasan might not be apparent until the afterlife. As can easily be imagined, such a belief has far-reaching implications for attitudes and behaviours. [M]

A Malay cultural script on the balasan ‘return in kind’ for one’s deeds people think like this: good things will happen to a person if this person does good things bad things will happen to a person if this person does bad things

According to Ameka (1987; cf. Goddard & Wierzbicka 1997), Ewe culture includes a cultural belief that good things cannot happen to people without the intervention of supernatural beings, such as ancestral spirits or local divinities, and, ultimately God (Máwú). Such a belief can be modelled as follows. [N] An Ewe cultural script on the efficacious role of supernatural beings people think like this: good things cannot happen to a person, if beings of another kind don’t do some things

It is on account of this belief, Ameka argues, that verbal routines used in response to the news that various ‘good things’ have happened to a person include expressions such as: ŋúwò núwó wf df´! ‘Beings around you have worked!’, Máwú wf df´! ‘God has worked!’, and Tf gbéwó wf df´! ‘Ancestors have worked!’.

8.  Scripts for social models The two previous scripts were concerned with how things happen in the world. Another class of belief scripts which can be particularly pertinent to people’s ways of speaking and interacting can be termed ‘social models’; i.e., widely shared representations about what people are like, about what kinds of people there are, about what kinds of relations exist between people, and so on. Yoon (2004) has proposed a widely shared script of this kind for Korean culture. It is given, in a slightly adapted form, in [O] below. It presents a picture of society as consisting, broadly speaking, of two groups. One of these groups consists of ‘people above me’; and these people are seen as necessarily ‘people not like me’. The other group, i.e., those who are ‘people not above me’, is further seen as apportioned between some who are ‘people like me’ and others who are ‘people below me’. The script portrays what one may term a ‘vertical’ model of society. [O]

A Korean cultural script for a ‘vertical’ model of society people think like this: some people are people above me, they are not people like me other people are people not above me some of these other people are people like me some of these other people are people below me

Cultural scripts

The broad two-way division corresponds to the major cleavage in the highly elaborated Korean system of speech styles and honorification. One uses contaymal (polite, respectful language) with ‘people above me’ and panmal (plain, non-respectful language, lit: half language) with others. As described above, respected elder people (Korean noin) necessarily fall into the ‘above’ category, but so do many others such as teachers, doctors, monks and priests, and many people relatively older than oneself, even if they are not old enough to be noin. Scripts similar or identical to [O] appear to be widely distributed in East Asia. 9.  Scripts for non-verbal communicative practices Cultural scripts can also deal with non-verbal communicative practices, as indicated by the reference above to the possibility of different cultural functions of smiling. The semantics and ethnopragmatics of facial expressions have been discussed by Wierzbicka (1999: Ch. 4), Hasada (1996), and Ye (2004b, 2006). A great deal remains to be explored in this area, including applications of cultural scripts to other non-verbal practices, such as gestures and gesturing, body postures, touching and proxemics, voice and vocalization styles, and so on.

10.  The accessibility and practicality of cultural scripts Because cultural scripts written in semantic primes can be readily transposed across languages, including into the language of the people concerned, native speaker consultants can become involved in a very direct way with working and re-working cultural scripts. Native speakers from different cultures are often surprisingly interested in engaging in this kind of collaborative work, especially those who have had direct personal experience of intercultural ‘cross-talk’ and confusion. Of course, consultants need guidance and support in such work, if only because it is no easy matter to learn to express one’s ideas solely within a controlled vocabulary and grammar of cross-translatable words. Even so, the intuitive accessibility of cultural scripts means that native speakers can at least read (or hear) them, that they can understand them, and that they can respond to them without the intervention and mediation of the analyst. Cultural scripts are therefore potentially empowering for native speaker consultants. Their accessibility and transparency gives cultural scripts a big advantage over technical modes of description when it comes to real-world situations of trying to bridge some kind of cultural gap, with immigrants, language-learners, in international negotiations, or whatever. There is no need to begin with a ‘tutorial’ about collectivism vs. individualism, positive politeness vs. negative politeness, high context cultures



Cliff Goddard

vs. low context cultures, or other arcane academic concepts. Because cultural scripts interface more or less directly with simple ordinary language – in any language – they can be practically useful for the purposes of cross-cultural education and intercultural communication (cf. Goddard 2004b; Goddard & Wierzbicka 2007).

11.  Closing note Since the mid-1990s, one can see various improvements being worked through in the form and format of cultural scripts, and this process is still ongoing. Researchers are still finding out, via trial and error experimentation, about the range of different formats and structures which may be appropriate to material of different kinds from different settings. Even so, it seems reasonable to conclude that the cultural scripts approach offers a promising method for describing cultural norms and practices in a way which combines an insider perspective with intelligibility to outsiders, which is free from Anglocentrism, and which lends itself to direct practical applications in intercultural communication and education.

References Ameka, F.K. (1987). A comparative analysis of linguistic routines in two languages: English and Ewe. Journal of Pragmatics 11: 299–326. ——— (1994). Areal conversational routines and cross-cultural communication in a multilingual society. In H. Pürschel, E. Bartsch, P. Franklin, U. Schmitz & S. Vandermeeren (Eds), Intercultural Communication: 441–469. Peter Lang. ——— (2006). ‘When I die, don’t cry’ –The ethnopragmatics of “gratitude” in West African languages. In C. Goddard (Ed.), Ethnopragmatics. Understanding Discourse in Cultural Context: 231–267. Mouton. Ameka, F.K. & A. Breedveld (2004). Areal cultural scripts for social interaction in West African communities. Intercultural Pragmatics 1(2): 167–187. D’Andrade, R. (2001). A cognitivist’s view of the units debate in cultural anthropology. Cross-Cultural Research 35(2): 242–257. Goddard, C. (1992). Traditional Yankunytjatjara ways of speaking—A semantic perspective. Australian Journal of Linguistics 12: 93–122. ——— (1997). Cultural values and ‘cultural scripts’ of Malay (Bahasa Melayu). Journal of Pragmatics 27: 183–201. ——— (2000). Communicative style and cultural values—Cultural scripts of Malay (Bahasa Melayu). Anthropological Linguistics 42(1): 81–106. ——— (2004a). The ethnopragmatics and semantics of “active” metaphors. Journal of Pragmatics 36: 1211–1230. ——— (2004b). “Cultural scripts”: A new medium for ethnopragmatic instruction. In M. Achard & S. Niemeier (Eds), Cognitive Linguistics, Second Language Acquisition, and Foreign Language Teaching: 145–165. Mouton.

Cultural scripts

——— (2006). ‘Lift your game Martina!’ – Deadpan jocular irony and the ethnopragmatics of Australian English. In C. Goddard (Ed.), Ethnopragmatics. Understanding Discourse in Cultural Context: 65–99. Mouton. ——— (Ed.) (2006). Ethnopragmatics. Understanding Discourse in Cultural Context. Mouton. ———(2009). Not taking yourself too seriously in Australian English: Semantic explications, cultural scripts, corpus evidence. Intercultural Pragmatics 6(1): 29–53. Goddard, C. & A. Wierzbicka (1997). Discourse and culture. In T.A. van Dijk (Ed.), Discourse as Social Interaction: 231–257. Sage. ——— (Eds) (2002). Meaning and Universal Grammar – Theory and Empirical Findings. Vols I & II. John Benjamins. ——— (Eds) (2004). Cultural scripts. Special Issue of Intercultural Pragmatics 1(2). ——— (To appear). Men, women and children: The semantics of basic social categories. ——— (2007). Semantic primes and cultural scripts in language learning and intercultural communication. In F. Sharifian & G. Palmer (Eds), Applied Cultural Linguistics: Implications for second language learning and intercultural communication: 105–124. John Benjamins. Hasada, R. (1996). Some aspects of Japanese cultural ethos embedded in nonverbal communicative behaviour. In F. Poyatos (Ed.), Nonverbal Communication in Translation: 83–103. John Benjamins. ——— (2006). Cultural scripts: some glimpses into the Japanese emotion world. In C. Goddard (Ed.), Ethnopragmatics. Understanding Discourse in Cultural Context: 171–199. Mouton. Peeters, B. (2000). S’engager vs. to show restraint: Linguistic and cultural relativity in discourse management. In S. Niemeier & R. Dirven (Eds), Evidence for Linguistic Relativity: 193–222. John Benjamins. Klos Sokol, L. (1997). Shortcuts to Poland. IPS Wydawniclwo. Travis, C.E. (2004). The ethnopragmatics of the diminutive in Colombian Spanish. Intercultural Pragmatics 1(2): 249–274. ——— (2006). The communicative realization of confianza and calor humano in Colombian Spanish. In C. Goddard (Ed.), Ethnopragmatics. Understanding Discourse in Cultural Context: 199–231. Mouton. Wierzbicka, A. (1996a). Semantics: Primes and Universals. Oxford. ——— (1996b). Japanese cultural scripts: Cultural psychology and “cultural grammar”. Ethos 24: 527–555. ——— (1996c). Contrastive sociolinguistics and the theory of cultural scripts: Chinese vs. English. In M. Hellinger & U. Ammon (Eds), Contrastive Sociolinguistics: 313–344. Mouton. ——— (1997). Understanding Cultures through their Key Words. Oxford. ——— (1998). German ‘cultural scripts’: Public signs as a key to social attitudes and cultural values. Discourse & Society 9: 241–282. ——— (1999). Emotions across Languages and Cultures. Cambridge. ——— (2002a). Russian cultural scripts: The theory of cultural scripts and its applications. Ethos 30(4): 401–432. ——— (2002b). Australian cultural scripts – bloody revisited. Journal of Pragmatics 34: 1167–1209. ——— (2003) [1991]. Cross-Cultural Pragmatics. [Expanded 2nd edition]. Mouton. ——— (2004) Jewish cultural scripts and the interpretation of the Bible. Journal of Pragmatics 36: 575–599. ——— (2006a). The English Language: Meaning and Culture. Oxford. ——— (2006b). Anglo culture scripts against “putting pressure” on other people and their linguistic manifestations. In C. Goddard (Ed.), Ethnopragmatics. Understanding Discourse in Cultural Context: 31–65. Mouton.



Cliff Goddard Wolfson, N. (1983). An empirically based analysis of complimenting in American English. In N. Wolfson & E. Judd (Eds), Sociolinguistics and Language Acquisition: 82–95. Newbury House. Wong, J. (2004a). Cultural scripts, ways of speaking, and perceptions of personal autonomy. Intercultural Pragmatics 1(2): 231–248. ——— (2004b). The particles of Singapore English: a semantic and cultural interpretation. Journal of Pragmatics 36: 739–793. Ye, Z. (2004a). Chinese categorization of interpersonal relationships and the cultural logic of Chinese social interaction: an indigenous perspective. Intercultural Pragmatics (2): 211–230. ——— (2004b). The Chinese folk model of facial expressions: a linguistic perspective. Culture & Psychology 10(2): 195–222. ——— (2006). Why the ‘inscrutable’ Chinese face? Emotionality and facial expression in Chinese. In C. Goddard (Ed.), Ethnopragmatics. Understanding Discourse in Cultural Context: 127–171. Mouton. Yoon, K.-J. (2004). Not just words: Korean social models and the use of honorifics. Intercultural Pragmatics 1(2): 189–210.

Culture Srikant Sarangi Cardiff University

1.  Introduction: Culture as an interdisciplinary project Several disciplines within the social sciences and humanities (e.g., anthropology, sociology, history, linguistics, literary theory, philosophy, psychology) have accommodated the study of culture in their academic inquiries. A cursory glance at the volumes of literature written on the subject suggests that although these disciplines share the notion of culture in a rather loosely descriptive way, there is little agreement between them with regard to the exact nature of its theoretical and analytical underpinnings (see Keesing 1974 and especially Kluckhohn 1962 for an illuminating interdisciplinary debate on the concept of culture). Even within the discipline of anthropology which is centrally concerned with the culture concept, various branches of inquiry continue to be both united and divided in the way culture is conceptualized for investigation purposes. Following Barthes (1984) who considers interdisciplinarity in terms of creating a new object that belongs to no one, it is possible to view culture as a truly interdisciplinary project — the term means what we want it to mean in specific contexts of use. In this article my main task is three-fold: (i) to offer a theoretical reassessment of the notion of culture as it is deployed in various traditions, in particular within social/ cultural anthropology and cultural studies (Sections 2 and 3); (ii) to problematize the methods used in doing cultural analysis (Section 4); and finally, (iii) to relate the above theoretical and analytical discussions to the debate on the language-culture interface (Section 5) and the pragmatic analyses of cross-cultural and intercultural discourse (Section 6).

2.  Defining and redefining culture: An overview 2.1  The historical transformation of the culture concept Braudel (1980) identifies two historical developments with regard to the origin and fortune of the term culture. First, he draws attention to the indiscriminate use of the words civilization and culture until the eighteenth century, after which period the two words were in fierce competition. Towards the middle of the nineteenth century culture took on the specific meaning of ‘intellectual culture’ and dominated Western thought.


Srikant Sarangi

The second aspect has to do with the shift from the singular to the plural, from talking about culture to talking about cultures: “the triumphant plural of the nineteenth century is undeniably a sign of new ideas, new ways of thinking — in short, new times” (Braudel 1980: 181). In the medieval age, Thompson (1993: 2) writes, “the term ‘custom’ used to carry much of what is now carried by the word ‘culture’” to foreground the legitimatizing practices of traditional values and norms (see also Benedict’s 1935 formulation of culture as science of custom). More specifically, in the critical period, i.e., between the last decades of the eighteenth century and the first half of the nineteenth century, the term ‘culture’ underwent a substantial transformation of meaning, in comparison to other key terms such as ‘industry’, ‘democracy’, ‘class’ and ‘art’: Before this period, it had meant, primarily, the ‘tending of natural growth’, and then, by analogy, a process of human training. But this latter use, which had usually been culture of something, was changed in the eighteenth and early nineteenth century, to culture as such, a thing in itself. (Williams 1958: xvi)

Thus ‘culture’ has been transformed from a noun of process (cultivation of the human mind) to a noun of configuration or generalization — as an abstraction and an absolute (way of life). Let us spell out the duality involved in the notion of culture a bit further. Culture is often seen as something to do with arts and literature in the way Matthew Arnold conceptualized ‘high culture’ in the nineteenth century as acquaintance with ‘the best’ that has been known and said in the world. In contrast to this literary-moral tradition, a recent reading of culture is perhaps more tied up with ordinary ways of life, ways of living. In the first part of the twentieth century, many anthropologists seem to have adopted this transformed meaning of culture as a generalizable abstraction in their study of “societies historically as little related as possible to our own and to one another” (Benedict 1935: 12). For Gellner (1964: 18ff), this amounted to first using ‘primitive’ tribes as “a kind of time machine, as a peep into our own historic past” which later led to a study of these groups “for their own sakes and explained in terms of themselves” (see also Mead & Metraux 1953). Outside anthropology and history, one can trace the origin of the culture concept, as an academic subject, to the late 1950s, i.e., the works of, among others, Richard Hoggart (1957), Edward P. Thompson (1968) and Raymond Williams (1958, 1976, 1981) which suggest the extension of literary criticism to cultural phenomena. Moving away from the cultural elitism of Arnold, Eliot and Leavis, the British cultural studies tradition is rooted in the movement of the Workers’ Educational Association. As Williams (1958: 245) puts it: “great literature is indeed enriching, liberating, and refining, but man [sic] is always and everywhere more than a reader, has indeed to be a great deal else before he can even become an adequate reader”.


Culture is thus seen as encompassing a whole way of life, as ordinary, with an interest in mass entertainment (e.g., folk art, folk songs, festivals). This shift coincides with Bakhtin’s (1968) concept of ‘carnival’ and ‘the culture of the marketplace’, where artistic forms such as the above belonged to the borderline between art and life. As Bakhtin (1968: 7) suggests: the carnival does not distinguish between actors and spectators […] [it] is not a spectacle seen by the people, they live in it, and everyone participates because its very idea embraces all the people.

Moreover, in the context of class-based societies, attention was also shifted to culture produced by the popular classes as opposed to culture imposed on the popular classes in order to emphasise the reciprocal influence between the cultures of the dominant and the dominated groups (see Ginzburg 1980; Thompson 1968). This calls for a further distinction to be made with regard to the treatment of the culture concept within social/ cultural anthropology and sociology of culture (i.e., the emergent discipline of cultural studies which draws almost on all the above disciplines). While the former is primarily engaged in describing or defining the diversity of social practices in ‘distant’ cultures and in problematizing the methodology of doing cultural analysis, cultural studies is preoccupied with theorizing on contemporary cultural forms and processes (popular culture) by combining textual analysis with an inquiry into socio-political changes. (For an overview of how the cultural studies tradition has brought about a conceptual displacement of the culture concept through processes of institutionalization, politicization and commodification, see Bennett et al. 1981; Featherstone 1990, 1992.) 2.2  What is not culture Culture is often juxtaposed with nature to differentiate the man-made origin of everything cultural from the non-cultural, ‘natural’ (see Tylor 1891 for an evolutionary perspective on the condition of culture). By equating culture with practical reasoning, Sahlins (1976), for instance, would argue that nature is in fact the product of human action and cognition. From this viewpoint, culture is a detachable part of a human being, a possession […]: it shares with the personality the unique quality of being simultaneously the defining ‘essence’ and the descriptive ‘existential feature’ of the human creature. (Bauman 1973: 7)

Because culture is constitutive of reality, only humans are able to challenge reality in a self-reflexive fashion. Another related theme is the opposition between humans and animals with regard to who possesses culture, thus making it possible to talk about ‘lack of culture’, ‘uncultured’ or the ‘transmitting of culture’. From the perspective of evolutionary



Srikant Sarangi

anthropologists and human palaeontologists, as Geertz (1964) points out, the transition to humanity is both a cultural and biological process. Among others, Carrithers (1992) would take this position further to argue that sociality — as the intellectual capacity for complex social behavior (including speech, creativity, intentionality etc.) — distinguishes humans from animals. Unlike the instinctive or biologically determined behavior patterns associated with animals, human beings are seen as having culture, i.e., the ability “to communicate, to learn and to teach, to generalise from the endless chain of discrete feelings and attitudes” (Kroeber 1958, cited in Geertz 1964: 38). Whether animals actually organize their lives in cultural terms is of course a very open question, which lies outside the scope of our discussion. 2.3  What is culture Kroeber & Kluckhohn (1952) offer a taxonomy of available definitions of culture, organized along descriptive, historical, normative, psychological, structural and genetic lines. A typical definition is always inclusive, as can be seen from the following: Culture consists of patterns, explicit and implicit, of and for behavior acquired and transmitted by symbols, constituting the distinctive achievement of human groups, including their embodiments in artifacts; the essential core of culture consists of traditional (i.e., historically derived and selected) ideas and especially their attached values; culture systems may, on the one hand, be considered as products of action, on the other as conditioning influences upon further action. (Kroeber & Kluckhohn 1952: 357)

The above definition draws attention to (i) the totality implied in the culture concept (every aspect of social life absorbed in culture); (ii) the organizing principle underlying the social structure (culture seen as synonymous with social system); and (iii) the determined and the determining aspects of culture as far as individual behavior is concerned. What it does not capture is the relationship of cultural elements to one another and their relationship to what is considered ‘noncultural’ in a given society. To quote Braudel (1980: 188): “ ‘Cultures’ and their bundle of relationships, which are all so obvious that there is no point in analyzing them: they exist, period”. The interrelationship between the individual and society is a crucial aspect of culture: individuals are seen as living in their culture and the culture as lived by individuals (see Goodman 1967 for a full account). On closer scrutiny, however, it seems that the culture project, concerned as it is with entities and events, focuses on ‘wholes’ as opposed to individuals, thus overlooking the processual and relational aspects of the individual-social dynamics. Keeping in view what properties of culture are stressed in different definitions, it may be helpful to single out three dominant approaches — the mentalist, the behaviorist and the semiotic — and briefly summarize them.


2.3.1  The whole is the sum of parts: The mentalist approach As early as 1891 Tylor defines culture as “that complex whole which includes knowledge, belief, art, morals, law, custom, and any other capabilities and habits acquired by man [sic] as a member of society” (p. 7). Here culture is seen as a mental map capable of making sense of the world around us. In other words, culture is a precondition for group membership in a given society. This position, in a sense, underlines the deterministic aspects of cultural phenomena, although it recognizes the fact that individual actions are consciously executed along the lines of cultural scripts or schemata. Gramsci (1981: 193), among others, alerts us to the inadequacy of a view of culture as “encyclopaedic knowledge […] filed in the brain as in the columns of a dictionary” which participants draw on in actual communicative settings. By equating culture with thoughts, feelings, values and beliefs of individuals, and by assuming that it exists in the heads of individuals, this notion of culture is not only very static and abstract, but it also conflates various aspects of human capacities (such as knowledge, belief, attitude) which need to be kept separate for analytic purposes. We notice a dichotomy here in the attempt to explain the observable activities of human actors in terms of unobservable cultural dispositions. Wuthnow et al. (1984) point out that culture, so defined, has been caught up with the individual psyche rather than being a theoretical effort to investigate the notion of culture per se. In other words, by assuming that culture consists of thoughts and feelings, the focus is not on an understanding of culture, but on explaining it away: Culture is that residual realm left over after all forms of observable human behavior have been removed. It consists of the inner, invisible thought life of human beings either as individuals or in some difficult-to-imagine collective sense, as in notions of ‘collective purpose’, ‘shared values’, and ‘intersubjective realities’. What people actually do, how they behave, the institutions they construct […] however, are not a part of culture. (Wuthnow et al. 1984: 4)

A view of ‘culture’ as the most fluid, unconstrained, and least observable category of ‘non-behavior’ explains, at a theoretical level, either too much or too little. At the methodological level, many empirical researchers find such a conceptualization of culture devoid of any analytic significance, and thus, unhelpful. 2.3.2  A map of and for behavior: The behaviorist approach Boas may be seen as a departure from the cognitive, mentalistic approach which deals with cultural artefacts, towards a behavioristic approach which examines the relationship between the individual and his/her culture. Minimally speaking, within this approach culture is defined as learned behavior in the absence of explicit teaching. This means that members of a group have the choice of selection, although, given the regulatory nature of social structure, one is very much constrained as to what counts



Srikant Sarangi

as acceptable social practice. Inevitably, culture is equated with a preference for those patterns of communicative behavior which are valued within a social group, however defined. In a limited sense, such a view recognizes human agency and the possibility for social change in suggesting that behavior patterns, like values and beliefs, are not necessarily fixed, but socially and contextually shaped. To quote Peterson (1979: 159): While it [culture] was once seen as a map of behavior it is now seen as a map for behavior. In this view, people use culture the way scientists use paradigms […] to organise and normalise their activity. Like scientific paradigms, elements of culture are used, modified, or discarded depending on their usefulness in organizing reality.

More importantly, then, for analytic purposes of understanding, behavior — taken as social action — is a useful means for investigating the invisible dimensions of culture implied in the mentalist approach. 2.3.3  The metaphor of culture as text: The semiotic approach As the pioneer of interpretive anthropology, Geertz (1964, 1973) locates himself between the behavioral and mentalist schools by suggesting that culture is essentially a semiotic system, i.e., a system of symbolic meanings (see also Schneider 1976). Moving away from a positivistic to a hermeneutic stance, Geertz (1964: 39) introduces the metaphor of cultures as texts to underscore the “imposition […] of symbolic meaning upon reality”. Elsewhere, he characterizes cultural organization as an octopus “whose tentacles are in a large part separately integrated” (1973: 407). Here Geertz is of course speaking from the viewpoint of a cultural analyst, but as far as cultural agents are concerned, this view resonates with Halliday’s (1984: 8–9) idea of culture as a network of information systems, i.e., in semiotic terms: As speakers and listeners, we project the linguistic system on to the social system […] interpreting verbal meanings as the expression of meanings that are inherent in the culture. Any construct of cultural meanings — that is, any social context — is realized in the form of acts of meaning in the various semiotic modes of which language is one.

A semiotic view of culture allows us to move away, on the one hand, from the reification of culture as a super-organic reality (implied in Tylor’s formulation of ‘most complex whole’) and on the other hand, from a reductionist view which treats culture as learned behavior. Geertz (1973: 14) emphasizes that culture is not about explaining mental phenomena or social behavior, but about understanding social practices in context: Culture is not a power, something to which social events, behaviors, institutions, or processes can be causally attributed: it is a context, something within which they can be intelligibly — that is, thickly — described.


It is the ‘thick description’ which allows one to interpret cultural categories (such as ‘winks’, to use Geertz’s favorite example) by taking into account the context of production and perception. This suggests that meaning and coherence are not inherent in cultural phenomena, but a question of subjectivity. To give Keats’s observation a twist, they are something which lies in the beholder’s eye. 2.4  Culture is a verb Many scholars now acknowledge that any definition of culture is necessarily reductionist, as they turn their attention to problematizing the notion of culture as something which is constantly being made and remade. In other words, we need to think of culture as an active rather than passive process. In a paper titled ‘Culture is a verb’, Street (1993: 25) argues that culture is “an active process of meaning making and contest over definition, including its own definition”. Rather than ask the question ‘what is culture’ which amounts to stereotyping culture and fix its meanings, following Thornton (1988), there is a shift in focus from ‘what culture is’ to what ‘culture does’ — by extension, how, why and where we ‘do’ culture: An understanding of culture, then, is not simply a knowledge of differences, but rather an understanding of how and why differences in language, thought, use of materials and behaviors come about. (Thornton 1988: 25)

By the same token, culture is also used to consolidate sameness and distinctive forms of identity within sub-groups, although the boundaries of these sub-cultures are being redefined constantly. The study of these group processes can be facilitated, according to Braudel (1980), against the following set of criteria: (i) identification of cultural area or locus with its frontiers; (ii) attention to borrowings of cultural elements across social groups and (iii) importance of refusals to borrow. 3.  Culture as ideology: From a consensual to a differentiated view of culture “What really binds men [sic] together”, argues Benedict (1935: 11), “is their culture — the ideas and the standards they have in common”. Whether looked at in terms of mental phenomena, behavioral patterns or the semiotic mode, consensus seems to have been the most salient defining feature of culture in the early anthropological literature. As Kluckhohn (1962) points out, culture has been intricately related to the shared, and historically created ‘definitions of the situation’ rather than to the distinctive ‘ways of life’ which are their manifestations. It is this shared interpretive framework, with elements of congruence and predictability, which defines culture, no matter who the members are. Accompanying this feature of sharedness we notice, on the one hand, the notion of reproduction of certain dominant norms and values and, on the other



Srikant Sarangi

hand, the probable sanctions against deviance and rebellion. Put strongly, to be cultural means being distinct from the idiosyncratic and the aberrant. One difficulty with this consensual view of culture is that there is little explanation, in a genealogical sense, of why there is this homogeneity within a group and what exactly is shared by whom and why (the same problem also applies to the sociolinguistic notion of ‘speech communities’). Asad (1980: 614), for instance, challenges such a theory of culture which “gives logical priority to the system of authentic meaning supposedly shared by an ideologically-defined community, and independent of the political activity and economic conditions of its members”. In the words of Herzfeld (1992: 75), “such ideologies suppress differences created through time, conflate social identity with cultural sameness, and disguise the contingent history whereby they came about”. This points to a second difficulty: such a holistic view of culture, by invoking consensus and consistency, may serve to distract attention from the discontinuities and contradictions within the ‘whole’ (see Barth 1989; Thompson 1993). As Sider (1986) maintains, the ‘anthropological concept of culture’ with its emphasis on shared values is not very effective for understanding class-based societies and it needs to be replaced by a stress on cultural conflict. This leads us to the discursive aspects of the construction of the cultural other from the Western point of view (Section 3.1) and the significance of cultural differences in stratified multicultural societies (Section 3.2). 3.1  The construction of the cultural other The mention of the term ‘culture’ almost always invokes the notion of ‘the other’ being described (or inscribed as anthropologists would prefer to call it) from the viewpoint of the observer. This often means a dominant culture using colonial modes of representation of the cultural otherness — a form of cultural imperialism. The unreciprocal interpretation of other, in this case non-Westen, cultures has been challenged as unequal narratives (see Asad 1973, 1980; Said 1978, 1993). As Said (1993: xiii) points out, culture becomes a source of identity construction whereby the non-Western other is presented from the Western point of view, using a discourse of exclusion: Culture comes to be associated, often aggressively, with the nation or the state; this differentiates ‘us’ from ‘them’, almost always with some degree of xenophobia.

This is seen as an act of oppression, based on the principles of dichotomization and opposition, which lead to us/them discourses (realized as ‘they do not’ and ‘contrary to us’). This attempt to divide human reality inevitably results in essentializing the ‘other’, with the Westerner having the authority to determine what can be said and written about the other. Asad (1980: 607) draws attention to the difficulty arising from a theoretical preoccupation with essential human meanings and criticizes the “a priori totality which defines and reproduces the essential integrity of a given order”.


It appears then that the cultural other arises from and through the process of studying cultures; he/she/it does not precede that process. The latter tendency is manifested, according to Said (1978), in the overly systematic description of Orientalism through a ‘sheer knitted-together strength’ of the text. The text produces the object, attributes static images and thus gives way to one-sided generalizations in a stereotypical manner. Following the Foucauldian notion of discursive formations, Said concludes that Orientalist discourse ‘orientalizes the Orient’. 3.2  The politics of cultural difference in stratified multicultural societies It is possible to argue that in many multicultural societies, the ideology underlying the construction of minority group cultures based on the principle of differences runs parallel to that of cultural imperialism discussed above. The current cultural debate, especially within cultural studies, emphasizes that cultural differences have meanings, functions and histories. To quote Thornton (1988: 25): Contemporary cultural studies look at these meanings, functions and histories in order to understand the differences; they do not use the apparent ‘fact’ of differences to explain history.

By extension, attention to cultural differences alone would leave issues of social inequality unchallenged and prevent one from seeing that cultural practices are expressions of power relations (see Gilroy 1991 on how an over-integrated conception of culture overlooks the changing and dynamic political cultures). An illustrative example here would be the way in which discriminatory practices in multicultural societies are explained in terms of either cultural differences or racist attitudes, although both explanations are co-existing constructs. Let us consider the relevance of the cultural studies view of culture from the viewpoint of stratified multicultural societies. In the tradition of cultural studies, the definition of culture as ‘universal cultural practices’ gives way to culture as the ‘lived practices’ of social groups in definite societies (Hall et al. 1980). Thus the everyday perception of culture and its institutionalization occupies the center stage, with due attention paid to processes of both production and consumption. The cultural studies framework draws heavily on sociological theory — in particular, Gramsci’s (1971) notion of hegemony, Bourdieu’s (1991) notion of cultural and symbolic capital and Foucault’s (1980) view of exclusion and power being constituted in discourse. As far as Gramsci is concerned, culture (as hegemony and as ideology) is one of the means by which a dominant group upholds certain values not by force, but by consensus, and thus maintains control over the others. Cultural differences come to be viewed in hierarchical terms, but hidden in hegemonic sense (Gramsci) and reproduced via educational processes (e.g., Bourdieu’s notion of habitus and symbolic capital). For instance, the dominant group presents its style or mode as natural and



Srikant Sarangi

neutral and expects the minority group to accept this as common sense and to defend the given boundaries. Here we notice both sides of the hegemonic process — accommodation and resistance — as the dominant discourse forms are appropriated by the minority groups for strategic reasons (cf. Derrida’s 1967 notion of ‘supplement’, which at once adds to something and supplants it). This hegemony can also be illustrated by appealing to what Horne (1986: 183–4) calls ‘public culture’: In a modern society there is no longer the dichotomy of ruling-class culture and folk culture […] The culture that dominates the public scene is not a ruling-class culture of triumphant display, but a fabricated ‘public culture’ that purports to be the culture not just of the rulers but of all the people.

To illustrate his point, Horne refers to the way high art is being presented these days as cultural commodities for — not necessarily by — the masses. As Hewison (1995: 303) rightly points out, such a commodification of culture has resulted in squaring “the circle of culture-versus-society and culture-as-expression-of-society”. One of the main advantages of the Gramscian approach to culture is, however, that it brings to the fore what goes unchallenged in the construction and use of the notion of culture. Culture ultimately becomes a site of contention as the school of cultural criticism questions the ideological basis of cultural values. Notice that in the discourse of cultural imperialism (cf. above), culture is something that the dominant West had and the non-Western world lacked. By contrast, in the discourse of stratified multicultural societies, culture is something the minority groups have either in terms of trivialities (ethnic dress, music etc.) or in terms of a system of absences (institutions). It is important to extend this duality to the debate about cultural identities. Hall (1990: 222) problematizes the very authority and authenticity to which the term ‘cultural identity’ lays claim, as follows: Identity is not as transparent or unproblematic as we think. Perhaps instead of thinking of identity as an already accomplished fact, which the new cultural practices then represent, we should think, instead, of identity as a ‘production’, which is never complete, always in process, and always constituted within, not outside, representation.

We can relate Hall’s observation to the cultural processes involved in the construction of non-Western and minority identities. With regard to the former, culture is being directly linked with nationality and ethnicity in a rather essentialist way as part of the decolonization process. This results in people taking pride in their cultural uniqueness and demanding new monolithic identities (see, for instance, Schama’s 1987 account of the ways in which the Dutch forged an identity for themselves in the seventeenth century). If we turn to the multicultural scenario, monolithic concepts of identity are being contested in strategic ways (see, among others, Goldberg 1994; Rattansi &


Westwood 1994; Wrench & Solomos 1993). The ambivalence is reflected in the way minority groups hold on to the cultural or ethnic labels constructed for them by the majority group, while also taking on new identities. This makes the point that cultural identities are not given, but formed, maintained and changed in society in order to fit emergent needs (see Roberts & Sarangi 1993, 1995, on how rhetorical strategies and discourse practices help to construct new cultural identities in stratified multicultural societies). Inherent in this view is the notion that cultural groups cannot be identified across ethnic, religious or other societal lines. By extension, culture is in a constant flux and its boundaries are not as rigid as many cultural analysts would like us to believe. 4.  Doing cultural analysis: A critique of the method One way of coming to terms with the theoretical nuances of the culture concept is to examine what the practitioners, i.e., cultural analysts, actually do. It is interesting that the problematic nature of the culture concept, at the theoretical level, is very much embedded in the methodological issues surrounding the analysis of cultures and cultural forms within social/cultural anthropology. The fundamental problem has to do with two assumptions cultural analysts make: (i) that well-defined cultural boundaries exist and correspond approximately with national, ethnic or regional boundaries; and (ii) that the analytic constructs and categories available in one language and culture (predominantly in English) are sophisticated enough to allow one to describe distinct cultures without distortion. The problem that the concept of culture is ‘unrealizable’ in empirical terms dates back to Sapir. Writing in the early 1930s, he points out that the seemingly effortless transition from talking about culture to identifying distinct cultures is more than difficult, it is impossible: The true locus (of behavioral patterns) which, when abstracted into a totality, constitute culture is not in a theoretical community of human beings known as society, for the term ‘society’ is itself a cultural construct which is employed by individuals who stand in significant relations to each other in order to help them in the interpretation of certain aspects of their behavior. The true locus of culture is in the interactions of specific individuals and, on the subjective side, in the world of meanings which each one of these individuals may unconsciously abstract for himself for his participation in these interactions. (Sapir 1949 [1932]: 515–16)

Sapir here formulates that groups or societies are nothing but abstractions of existing interpretations of behavior held by particular individuals in the course of particular interactions. This is how he shows his radical disavowal of integral cultures attached to discrete societies.



Srikant Sarangi

Despite such early warnings, cultural anthropologists, traditionally, have been seeking to arrive at generalizations about everyday social institutions within cultural groups. The assumption here is that culture (as social structure) is a whole which can be studied through a classification of its parts for the purposes of comparison with cultural patterns elsewhere. In most cases, the observer uses an a priori system of analytic categories to make sense of the observed and it is this classificatory system which imposes a uniformity on the cultural data. In sum, comparison, classification and generalization are the dominant tools of cultural analysis in the anthropological tradition (which are also prevalent in anthropological linguistics and subsequently in crosscultural and intercultural communication studies). In the remainder of this section I briefly describe the different elements of cultural analysis, followed by a commentary on the present debate about ethnography. 4.1  The dominance of intensive fieldwork Fieldwork is the first step to a cultural analysis of people in their social context. As Fabian (1983) elegantly puts it, it is mainly an intersubjective dialogue and communication between actor (i.e., informant) and ethnographer. Interaction with informants is crucial to the study of cultures as it is via informants that one gains access to the cultural constructions of reality. The dialogue usually takes place in the present time, outside the historic and political context of fieldwork. Since language (including paralinguistic and other non-verbal systems) is crucial to understanding cultural patterns, Malinowski (1922) and Mead (1939) in fact emphasized the importance of learning the native language for conducting fieldwork. ‘Going native’ and ‘empathy’ are seen as the tools to draw conclusions about the informants’ belief systems and social institutions. There are several problems with the fieldwork method, which, by definition, lacks systematicity in the observation procedure (but see Malinowski 1922 & Radcliffe-Brown 1948 on the influence of functionalism in relation to asking methodological questions). Further criticisms of the fieldwork method include: (i) the bias of the fieldworker, given his/her culture-specific analytical tools; (ii) the unreliability (including verification and accountability) of the fieldwork account which does not pay sufficient attention to communication between actors (see Sanjek 1990 for a detailed discussion of systematic fieldwork observation schedules). 4.2  The subjectivity of participant observation Seen as an improvement on fieldwork, within the method of participant observation, the anthropologist is considered an outsider entering a different culture. Unlike the fieldworker, the participant observer is not part of the observed. According to Clifford (1988: 34):


‘Participant observation’ serves as shorthand for a continuous tacking between the ‘inside’ and ‘outside’ of events: on the one hand grasping the sense of specific occurrences and gestures empathetically, on the other stepping back to situate these meanings in wider contexts.

Like the fieldwork method, participant observation lacks systematicity and observational rigor. In fact the metaphor of ‘observation’ in ‘participant observation’ suggests that prominence is given to the visual, as textual transcription/translation is relegated to a secondary status. Also missing from the account is the native point of view since participant observers do not use ‘privileged informants’ to access cultural knowledge. 4.3  Doing and writing ethnography as interpretation and invention Ethnography is defined as the way in which a sociocultural phenomenon is perceived and described against the background of the experiences of the observer. In this sense, it combines fieldwork and participant observation. As a research process, ethnography involves observing, recording and writing accounts of other cultures, paying particular attention to descriptive detail (i.e., thick description). Geertz (1973: 20) views all ethnographic description as interpretative social discourse: Cultural analysis is (or should be) guessing at meanings, assessing the guesses, and drawing explanatory conclusions from the better guesses, not discovering the Continent of Meaning and mapping out its bodiless landscape.

To put it differently, unlike scientific inquiries searching for ‘the truth’, cultural analysis is an interpretive enterprise concerned with multiple meanings. Ethnography, however, demands, on the one hand, a firsthand experience (what Geertz calls ‘finding our feet’, ‘being there’) and on the other hand, it requires the writer to distance himself/herself from the account to safeguard scientific objectivity. As Clifford (1988: 28) puts it, the ethnographers “did not speak as cultural insiders but retained the natural scientist’s documentary, observational stance”, which, for Fabian (1983: xii) amounted to “keeping object and subject apart”. This conflicting tension between extreme subjectivism and extreme formalism raises doubts about the authenticity of the ethnographic account. With the main focus on the ethnographer’s point of view (which is bound to be biased and selective), the informant, and thus the native point of view, is often silenced in the text (see Fabian 1983 on the silence and secrecy surrounding the ethnographic act). The authority of the ethnographic text has recently been challenged, thus bringing to question the conditions of its own inquiry and, by extension, the entire culture enterprise (for a critique of ethnography as a research process, see Clifford 1988; Clifford & Marcus 1986; Geertz 1988; Marcus & Fischer 1986). Influenced by developments in literary theory (in particular, Bakhtin 1981 & Ricœur 1983), the ethnographic narrative has been the object of inquiry: in what ways the rhetoric and the metaphors are



Srikant Sarangi

divorced from the representation of social reality. Spencer (1989), however, alerts us to the fact that literary critical theory is unhelpful in addressing the anthropological questions. The central issue revolves around the question of genre: not just in terms of selecting what to study, but also what kind of writing is acceptable and what authority the text has. In a sense, ethnographic accounts are more about the observer rather than the observed, as adjustments are made between the writer, the written-about and the reader. Geertz’s (1973: 9) observation that ethnographic texts are “our own constructions of other people’s constructions” poses the basic question about keeping description separate from interpretation. A related problem has to do with the imposition of formal order on social reality — consisting of unruly experiences — in the same way as a sculptor chooses a block of marble “from among all possible ones because the sculptor saw within it the latent image of his [sic] own project” (Sahlins 1976: 210). Leach’s (1954) attempt to construct a homogeneous Kachin identity based on an analytic category of ‘ritual language’ is a classic illustration of anthropologists’ preoccupation with an essentialist discourse. Among others, Asad (1980) vigorously criticizes this reductionist approach which presents the social order in an ‘authentic discourse’ constructed out of essential human meanings, without paying attention either to inter-subjective meaning-making (intentionality, force of utterance etc.) or to the ‘authoritative discourse’ which underlies the political and economic conditions. A similar point is made by Fabian (1983) when he claims that the predominant use of the present tense in ethnographic accounts implies a view of human behavior as conventional, predictable and rule-governed. A further issue concerns how ethnographers act as mediators who translate their experience into textual form through a process of textualization, thus playing down the interactional, dialogic aspects in the writing process. For Clifford (1986: 98), “ethnographic writing is allegorical at the level both of its content (what it says about cultures and their histories) and of its form (what is implied by its mode of textualization)”. Elsewhere he maintains that “textualization generates sense through a circular movement that isolates and then contextualizes a fact or event in its engobling reality” (1988: 38). There is thus a sense of displacement in the textual contextualization of experience, as the research process is separated from the text and other situational, dialogic aspects are also excluded from the final account. Said’s (1978) critique of the Orientalist ‘visions’ and ‘textualizations’ about ‘Islam’ and ‘the Arabs’ forcefully shows how the social object presented in the text is simply constituted in those texts. In addition to textualization, there is also the issue of temporalization. In pointing out the ‘politics of time’, Fabian (1983) proposes a theory of coevalness (sharing the same time) and the uses of time in anthropological discourse. He draws our particular attention to the denial of coevalness, by which he means “a persistent and systematic tendency to place the referent(s) of anthropology in a Time other than the present of the producer of anthropological discourse” (1983: 31). He goes on to conclude that


such temporal distancing and denial of coevalness are conditions of anthropological discourse which can be characterized as a form of allochronic discourse — “it is a science of other men in another Time” (1983: 143). In recent years, ethnographers have shifted from the metaphor of text to a metaphor of dialogue by introducing a dialogic interplay of voices and negotiation of meanings in their texts (see Sperber 1985 for an account of descriptive and non-descriptive categories, with an emphasis on the use of free indirect speech to guarantee the point of view of the actors). Of particular relevance here is Moerman’s (1988) exploration of the connections between ethnography and conversation analysis, based on Garfinkel’s ethnomethodological project (see also Carbaugh 1990). 5.  Culture as communication: The ‘linguistic’ turn 5.1  One language — one culture? Although most definitions of culture do not explicitly mention language in their list of culture contents or as an organizing principle, the interrelationship between language and culture (or, more precisely, between language form and thought patterns) has been passionately debated since the beginning of the twentieth century. At one end of the debate is the notion of ‘cultural determinism’, an anthropological construct, which holds that each culture presents a unique, coherent system that, in its own special way, shapes and moulds individuality and world view, including language behavior. A majority of anthropologists, however, would adopt a ‘cultural relativism’ position and accord culture a mediating role to account for differences across diverse communities. As one of the exponents of a ‘linguistic relativity principle’ (e.g., Boas 1911; Sapir 1949; Whorf 1956), Whorf posited that language is not a neutral medium for the expression of thought and therefore, one’s native language will influence one’s habitual ways of thinking. Whorf (1956: 67) strongly maintains, Thinking will be found to be fundamentally different for individuals whose languages are of fundamentally different types. Just as cultural facts are only culturally determined, not biologically determined, so linguistic facts, which are likewise cultural, and include the linguistic element of thought, are only linguistically determined.

In a similar vein, Sapir (1931: 578) viewed language as “a self-contained, creative, symbolic organization, which not only refers to experience […] but actually defines experience for us”. This notion of language channelling thoughts later led to (one-sided) interpretations of a so-called ‘Sapir-Whorf hypothesis’ in terms of linguistic determinism, of which two versions were circulated: (i) the first, which is a ‘weak’ version of



Srikant Sarangi

the linguistic relativity hypothesis, maintains that language facilitates one’s perception and experience of reality; (ii) the second, which is the ‘strong’ version, holds that the language of a community determines how native speakers codify, categorize and even create the cognitive structures. Whichever way it is presented, the linguistic determinism view claims that language actually determines our perception of reality and in this way perpetuates cultural differences. In Sapir’s own words (1929, in Mandelbaum 1949: 162): The ‘real world’ is to a large extent unconsciously built up on the language habits of the group. No two languages are ever sufficiently similar to be considered as representing the same social reality. The worlds in which different societies live are distinct worlds, not merely the same worlds with different labels attached.

We notice here an emphasis on differentiation between cultures, while suggesting homogeneity within cultural groups. Sherzer (1987), among others, observes that the so-called Sapir-Whorf hypothesis, as usually formulated, searches for isomorphisms between grammar and culture and views language as providing the means for thought and perception, and world view. But the limitations of such theoretical positions are apparent, given that Sapir’s work was very much preoccupied with lexical differences across exotic languages (linguistic change as a psychological process) and that he had no inclination towards performance data — how language is used in communicative settings (see Street 1993 on the problematic nature of the language-culture relationship). In passing, it is perhaps relevant to point out that the one-language-one-culture premise has also been dominant in sociolinguistics, with regard to language planning, maintenance of minority languages and nationalism (Fishman 1972). 5.2  Language as a means to cultural resources As Lévi-Strauss (1963) argues, language is a condition of culture because it is through language that one’s culture is learnt and sustained. In more specific terms, Bauman & Sherzer (1974: 8) point out: a careful focus on speaking as an instrument for the conduct of social life brings to the fore the emergent nature of social structures, not rigidly determined by the institutional structure of the society, but rather largely created by the strategic and goal-directed manipulation of resources for speaking.

This position suggests that language is not just a carrier of cultural values and norms, but that it plays a constitutive role in sustaining and changing cultural practices. Bloch (1991: 184), however, argues that the view of culture as being inseparably linked to language is based on two wrong assumptions: “on the grounds either that culture is thought and transmitted as text through language, or that culture is


ultimately ‘language-like’, consisting of linked linear propositions”. From another angle, the overreliance on language amounts to ignoring the non-linguistic elements from the cultural models (see, for instance, the pioneering work of Hall 1959 and others on non-verbal aspects of communication). In a broader framework, it is therefore instructive to approach the language-culture interrelationship in a discursive mode, because it is discourse that “creates, recreates, focuses, modifies, and transmits both culture and language and their interaction” (Sherzer 1987: 295). From the analytic perspective, there is little doubt about the significance of language — seen as discourse — as a methodological tool for the study of other cultures. If the one-language-one-culture principle is extended to its logical conclusion, we run into serious analytic problems as interpretation and description of other cultural forms using one’s own analytic categories can be highly suspect. This is where Hymes’s (1974) model of SPEAKING — setting, participants, ends, acts, sequences, keys, instrumentalities, norms, genres — becomes very useful for a comparative study of culture-specific speech events.

6.  Cross-cultural and intercultural analysis in pragmatic research In this section I briefly summarise two dominant research trends within pragmatics which are concerned with the culture-specific character of language use and communication: the cross-cultural and the intercultural dimensions. Following Knapp et al. (1987: 7), however, first a terminological clarification is in order: Traditionally, ‘cross-’, as in ‘cross-linguistic’ or ‘cross-check’ implies a comparison of phenomena. Therefore, the adequate use of ‘cross-cultural’ depends on the analyst’s perspective: if different cultures are to be compared with respect to the occurrence of, for example, a certain form of language use, the approach is ‘cross-cultural’, but if the focus is on ongoing interaction among members of different (sub)cultures, ‘intercultural’ communication is at issue.

To summarize, while ‘cross-cultural’ attends to abstract entities across cultural borders, the ‘intercultural’ deals with the analysis of an actual encounter between two participants who represent different linguistic and cultural backgrounds. 6.1  Cross-cultural pragmatic research: The culture principle The tradition of cross-cultural pragmatics research has its origin in the anthropological paradigm of cross-cultural juxtaposition, as a way of understanding the meaningmaking capacities across different cultures, including one’s own (see, for instance, Mead’s 1923 comparison of child rearing practices in America and Samoa). The interest



Srikant Sarangi

in looking at linguistic differences at the level of phonology, morphology and syntax across cultures dates back to the Lado-Fries school (see Lado 1957). However, the boundaries of contrastive linguistics have been extended since then to include not only structural comparisons but also comparisons at the socio-cultural level to show the relationship between language-in-use and the socio-cultural conventions prevalent in a society. The main area of interest here is the determining influence of language within cross-cultural research, and the goal has been to obtain comparable data from a large number of independent cultures, often with different language communities, on various communicative acts and then proceed to seek predictable and significant correlates. With this in mind, Wierzbicka (1991), while endorsing the culture-specific aspects of meaning construction in the Sapir-Whorf mold, makes a strong methodological case for the use of ‘universal semantic primitives’ to study cross-cultural differences in communicative styles (see Wierzbicka 1991 for a comprehensive account of areas studied cross-culturally). This cross-cultural comparison is done in two ways. First, one examines how a specific linguistic activity is carried out across linguistic communities. Brown & Levinson’s (1987) comparative account of how politeness strategies are realized differently in different languages is a classic example of this kind of cross-cultural inquiry. Other cross-cultural studies (e.g., Keenan 1976; Schieffelin & Ochs 1986; Scollon & Scollon 1980, 1983; Tannen 1984) mainly look at rhetoric patterns for structuring interpersonal conduct such as norms of quantity (i.e., self-presentation according to social expectations), norms in responding to interlocutors, differences in conversational control mechanisms like turn-taking, and formulaic and interactional routines. Secondly, one studies the different realizations of, for example a particular speech act, in a common second language by learners with different first-language backgrounds (e.g., Blum-Kulka & Olshtain 1984; Blum-Kulka, House & Kasper 1989; Kasper & Blum-Kulka 1993; Garcia & Otheguy 1989). In a sense, it is through the study of cross-cultural differences that one comes to understand the universals in language use. Furthermore, as Garcia & Otheguy (1989) point out, cross-cultural studies have implications for successful interethnic communication. A major problem with this line of inquiry, however, is that cross-cultural juxtaposition of the kind suggested here encourages a static, one-sided account of cultural categories, which over-emphasizes consistency within a cultural group. Also, by attributing differences in pragmatic realizations to cultural differences, there is no attempt to understand culture, except for using it as a tool to describe differences in pragmatic principles and/or their linguistic realizations. As Hofstede (1980: 11) points out, “crosscultural studies proliferate in all social sciences, but they lack a theory of the key variable: culture itself ” (see his analysis of cultural differences across four dimensions — power distance, uncertainty avoidance, individualism and masculinity — across forty countries).


6.2  Intercultural pragmatics: Towards a critical reading Most of the studies carried out under the labels ‘intercultural’ or ‘interethnic’ communication, which are mainly concerned with face-to-face encounters between individuals from different cultural backgrounds, can be brought together under the ‘intercultural pragmatics’ label (see Sarangi 1994 for details). Approaching intercultural analysis from a cultural-anthropological perspective, the first school (e.g., Brislin 1981; Gudykunst 1983, 1991) assumes that cultural problems — for example, perceptions, attitudes, stereotypes, prejudices, beliefs, values, and thought-patterning itself — are more important than linguistic problems, although there is an acknowledgement of the significant role of language in manifesting these cultural differences. As with the cross-cultural studies, culture is here used unproblematically as a static concept, analytically separate from language. The principle of cultural difference, by endorsing fixed value systems and resources, does not differentiate further the degree of difference between cultural groupings and the consequences of hierarchical differences. In short, culture becomes the necessary and sufficient explanation for intercultural mismatches. The second school, rooted in the tradition of interactional sociolinguistics (Gumperz 1982a, 1982b), offers an analysis of some of the fundamental principles underlying communicative style differences in inter-ethnic communication. Gumperz, for instance, claims that non-native speakers will transfer some of the mechanisms of their native language patterns and discourse expectations (referred to as ‘native contextualization strategies’) which may not correlate with the conventions established in the target language. In the interactional context this may result in the distortion of the communicative intent, as different ways of structuring information may receive different valuation in two different cultures. One significant development in recent years is that many researchers are taking on a cultural perspective in the study of interactional settings as they address various pragmatic issues such as contextualization cues, politeness and face wants, relevance/adequacy of information exchange, force/intent of speech acts (see Thomas 1983, 1984; for a more updated account of this mode of inquiry, see the Special issues of Multilingua 1994 and Pragmatics 1994; Scollon & Scollon 1994).

7.  Conclusion: The pragmatics of recovering culture Hall (1959: 53) suggests that “culture hides much more than it reveals, and strangely enough what it hides, it hides most effectively from its own participants”. This statement quite aptly summarizes the elusive nature of the culture concept from the viewpoint of cultural agents in given societies as well as cultural analysts bounded by different


100 Srikant Sarangi

disciplines. The main thrust of this paper has been to lay bare, from an interdisciplinary perspective, some of the hidden processes of the culture puzzle rather than privilege a particular reading. One can only endorse here Burke’s (1992: 123) warning about the limits of the two main strands of culturalism — cultural construction vs. culture as a system of shared meanings: It would be difficult to deny the reductionism implicit in some traditional approaches to culture, Durkheimian and Marxist, but the reaction in the opposite direction may well have gone too far. The current emphasis on cultural recreativity and on culture as active force in history needs to be accompanied by some sense of the constraints within which that creativity operates. Rather than simply replacing the social history of culture by the cultural history of society we need to work with the two ideas together and simultaneously, however difficult this may be.

Returning to the present scenario in cross-cultural or intercultural studies, it seems that the contested, multivoiced characterization of culture as it is currently understood in the fields of socio-cultural anthropology and cultural studies needs to be adequately operationalized in pragmatic research. There is certainly a danger if pragmatic accounts of cross-cultural and intercultural discourse embodies an essentialist view of culture and uses it as a taken-for-granted variable in understanding and describing communicative differences. Hinnenkamp (1987: 176) quite rightly observes: Culture as adapted in most linguistic subdisciplines has unfortunately become a passe partout-notion: whenever there is a need for a global explanation of differences between members of different speech communities the culture-card is played — the more ‘distant’ in geographic and linguistic origin, the more ‘cultural difference’!

By implication, the goal for discourse-oriented research in intercultural settings is not to use culture as an explanator of communicative behavior in an unproblematic way, but to make attempts to understand how and when ‘culture’ — in the sense of discourse practices and rhetorical preferences — plays an active role in shaping and influencing our meaning-making endeavours (see, among others, Blommaert 1991; Meeuwis & Sarangi 1994; Roberts & Sarangi 1993 for a critical approach which challenges cultural ideologies embedded in pragmatic and sociolinguistic inquiries). Following from the discussion above, it may be fruitful if, for instance, crosscultural pragmatists were to draw on and compare published ethnographic research on other cultural contexts, rather than treat cultural differences in a static fashion to explain variable communicative styles. By the same token, intercultural pragmatics, as a field of inquiry in its own right, could benefit from the emergent discipline of cultural studies by accommodating a differentiated notion of culture which challenges the idea of cultural values and beliefs as separable from power and political realities. The cultural studies paradigm acknowledges the rhetoric of difference but pays attention to power — something that crosscultural or intercultural research has often overlooked,


although power remains a key notion in pragmatics. To conclude, pragmatics (with the wide range of analytic tools emerging from the study of context and communicative behavior) offers the possibility of not just explaining culture away, but discovering what culture does, historically and in contemporary terms.

References Asad, T. (Ed.) (1973). Anthropology and the colonial encounter. Ithaca Press. ——— (1980). Anthropology and the analysis of ideology. Man 14: 607–27. Bakhtin, M.M. (1968). Rabelais and his world. MIT Press. ——— (1981). Discourse in the novel. In M. Holoquist (Ed.), The dialogical imagination: 259–442. University of Texas Press. Barth, F. (1989). The analysis of culture in complex societies. Ethnos 54(3/4): 120–142. Barthes, R. (1984). Jeunes Chercheurs. In Le Bruissement de la langue, 97–103. Seuil. Bauman, R. & J. Sherzer (Eds) (1974). Explorations in the ethnography of speaking. Cambridge University Press. Bauman, Z. (1973). Culture as praxis. Routledge. Benedict, R. (1935). Patterns of culture. Routledge. Blommaert, J. (1991). How much culture is there in intercultural communication? In J. Blommaert & J. Verschueren (Eds), The pragmatics of intercultural and international communication: 13–31. John Benjamins. Bloch, M. (1991). Language, anthropology and cognitive science. Man 26: 183–198. Blum-Kulka, S. & E. Olshtain (1984). Requests and apologies. Applied Linguistics 5(3): 196–213. Blum-Kulka, S., J. House & G. Kasper (Eds) (1989). Cross-cultural pragmatics. Ablex. Boas, F. (Ed.) (1911). Handbook of American Indian Languages, Part 1. Smithsonian Institute. Bourdieu, P. (1991). Language and symbolic power. Polity Press. Braudel, F. (1980). On history. Weidenfeld & Nicolson. Brislin, R. (1981). Cross-cultural encounters. Pergamon. Brown, P. & S. Levinson (1987). Politeness. Cambridge University Press. Burke, P. (1992). History and social theory. Polity Press. Carbaugh, D. (Ed.) (1990). Cultural communication and intercultural contact. Erlbaum. Carrithers, M. (1992). Why humans have cultures. Oxford University Press. Clifford, J. (1986). On ethnographic allegory. In J. Clifford & G.E. Marcus (Eds): 98–121. ——— (1988). The predicament of culture. Harvard University Press. Clifford, J. & G.E. Marcus (Eds) (1986). Writing culture. University of California Press. Derrida, J. (1967). L’ecriture et la difference. Seuil. Fabian, J. (1983). Time and the other. Columbia University Press. Featherstone, M. (Ed.) (1990). Global culture. Sage. ——— (1992). Cultural theory and cultural change. London: Sage. Fishman, J.A. (1972). Language in sociocultural change. Stanford University Press. Foucault, M. (1980). Power/Knowledge. Pantheon Books. Garcia, O. & R. Otheguy (Eds) (1989). English across cultures — cultures across English. Mouton de Gruyter. Geertz, C. (1964). The transition to humanity. In S. Tax (Ed.), Horizons of anthropology: 37–48. Aldine.


102 Srikant Sarangi ——— (1973). The interpretation of cultures. Basic Books. ——— (1988). Works and lives. Stanford University Press. Gellner, E. (1964). Thought and change. University of Chicago Press. Gilroy, P. (1991). It ain’t where you’re from, it’s where you’re at. Third Text (Winter): 3–16. Ginzburg, C. (1980). The cheese and the worms. Routledge & Kegan Paul. Goldberg, D.T. (Ed.) (1994). Multiculturalism. Blackwell. Goodman, M.E. (1967). The individual and culture. Dorsey Press. Gramsci, A. (1971). Selections from the prison notebooks. Lawrence & Wishart. ——— (1981). Culture. In T. Bennett, G. Martin, C. Mercer & J. Woollacott (Eds) (1981). Culture, ideology and social processes: 193–197. Batsford/Open University Press. Gudykunst, G.B. (Ed.) (1983). Intercultural communication theory. Sage. ——— (1991). Bridging differences. Sage. Gumperz, J.J. (1982). Discourse strategies. Cambridge University Press. ——— (Ed.) (1982). Language and social identity. Cambridge: Cambridge University Press. Hall, E.T. (1959). The silent language. Doubleday. Hall, S. (1990). Cultural identity and diaspora. In J. Rutherford (Ed.), Identity: 222–237. Lawrence & Wishart. Hall, S., D. Hobson, A. Lowe & P. Willis (Eds) (1980). Culture, media and language. Hutchinson. Halliday, M.A.K. (1984). Language as code and language as behavior. In R.P. Fawcett, M.A.K. Halliday, S.M. Lamb & A. Makkai (Eds), The semiotics of culture and language, vol. 1: 3–35. Frances Pinter. Herzfeld, M. (1992).The social production of indifference. University of Chicago Press. Hewison, R. (1995). Culture and consensus. Methuen. Hinnenkamp, V. (1987). Foreigner talk, code switching and the concept of trouble. In K. Knapp, W. Enninger & A. Knapp-Potthoff (Eds), Analyzing intercultural communication: 137–180. Mouton de Gruyter. Hofstede, G. (1980). Culture’s consequences. Sage. Hoggart, R. (1957). The uses of literacy. Penguin. Horne, D. (1986). The public culture. Pluto Press. Hymes, D.H. (1974). Foundations in sociolinguistics. University of Pennsylvania Press. ——— (Ed.) (1964). Language in culture and society. Harper & Row. Kasper, G. & S. Blum-Kulka (Eds) (1993). Interlanguage pragmatics. Oxford University Press. Keenan, E.O. (1976). The universality of conversational implicature. Language in Society 5: 67–80. Keesing, R.M. (1974). Theories of culture. Annual Review of Anthropology 3: 73–97. Kluckhohn, C. (1962). Culture and behavior. Free Press. Kroeber, A.L. & C. Kluckhohn (1952). Culture. Peabody Museum. Lado, R. (1957). Linguistics across cultures. University of Michigan Press. Leach, E. (1954). Political systems of highland Burma. Bell. Lévi-Strauss, C. (1963). Structural anthropology. Basic Books. Malinowski, B. (1922). Argonauts of the Western Pacific. Dutton. Marcus, G.E. & M.M.J. Fischer (1986). Anthropology as cultural critique. University of Chicago Press. Mandelbaum, D.G. (1949). Selected writings of Edward Sapir in language, culture, personality. ­University of California Press. Mead, M. (1923). Coming of age in Samoa. Morrow. ——— (1939). Native languages as field-work tools. American Anthropologist 42(20): 189–205. Mead, M. & R. Metraux (Eds) (1953). The study of culture at a distance. University of Chicago Press. Meeuwis, M. (Ed.) (1994). Critical perspectives on intercultural communication. Special issue, Pragmatics 4(3).


Meeuwis, M. & S. Sarangi (1994). Perspectives on intercultural communication. In M. Meeuwis (Ed.): 309–313. Moerman, M. (1988). Talking culture. University of Pennsylvania Press. Pauwels, A. (Ed.) (1994). Cross-cultural communication in the professions. Special issue, Multilingua 13(1/2). Peterson, R. (1979). Revitalizing the culture concept. Annual Review of Sociology 5: 137–166. Radcliffe-Brown, A.R. (1948 [1922]). The Andaman Islanders. Free Press. Rattansi, A. & S. Westwood (Eds) (1994). Racism, modernity and identity on the Western front. Polity Press. Ricœur, P. (1983). Time and narrative. University of Chicago Press. Roberts, C. & S. Sarangi (1993). ‘Culture’ revisited in intercultural communication. In T. Boswood, R. Hoffman & P. Tung (Eds), Perspectives on English for professional communication: 97–114. City Polytechnic, Hong Kong. ——— (1995). But are they one of us? Multilingua 14(4): 363–390. Sahlins, M. (1976). Culture and practical reason. University of Chicago Press. Said, E. (1978). Orientalism. Pantheon Books. ——— (1993). Culture and imperialism. Chatto & Windus. Sanjek, R. (Ed) (1990). Fieldnotes. Cornell University Press. Sapir, E. (1929). The status of linguistics as a science. Language 5: 207–214. [Mandelbaum 1949: 160–166.] ——— (1931). Conceptual categories in primitive languages. Science 74: 578. [Hymes 1964: 12.] ——— (1949). Culture, language and personality. University of California Press. Sarangi, S. (1994). Intercultural or not? In Meeuwis (Ed.): 409–428. Schama, S. (1987). The embarrassment of riches. Collins. Schieffelin, B.B. & E. Ochs (Eds) (1986). Language socialization across cultures. Cambridge ­University Press. Schneider, D.M. (1976). Notes toward a theory of culture. In K. Basso & H. Selby (Eds), Meaning in anthropology: 197–200. University of New Mexico Press. Scollon, R. & S. Scollon (1980). Linguistic convergence. Academic Press. ——— (1983). Face in interethnic communication. In J.C. Richards & R.W. Schmidt (Eds), Language and communication: 158–188. Longman. ——— (1994). Intercultural communication. Blackwell. Sherzer, J. (1987). A discourse-centered approach to language and culture. American Anthropologist 89: 295–309. Sider, G. (1986). Culture and class in anthropology and history. Cambridge University Press. Spencer, J. (1989). Anthropology as a kind of writing. Man 24: 145–164. Sperber, D. (1985). On anthropological knowledge. Cambridge University Press. Street, B.V. (1993). Culture is a verb. In D. Graddol, L. Thompson & M. Byram (Eds), Language and culture: 23–43. BAAL/Multilingual Matters. Tannen, D. (1984). The pragmatics of cross-cultural communication. Applied Linguistics 5(3): 189–195. Thomas, J.A. (1983). Cross-cultural pragmatic failure. Applied Linguistics 4(2): 91–112. ——— (1984). Cross-cultural discourse as ‘unequal encounter’. Applied Linguistics 5: 226–235. Thompson, E.P. (1968). The making of the English working class. Penguin. ——— (1993). Customs in common. Penguin. Thornton, R. (1988). Culture. In E. Boonzaeir & J. Sharp (Eds), Keywords. David Philip. Tylor, E.B. (1891). Primitive culture. Murray.


104 Srikant Sarangi Whorf, B.L. (1956). Language, thought and reality. Wiley. Wierzbicka, A. (1991). Cross-cultural pragmatics. Mouton de Gruyter. Williams, R. (1958). Culture and society. Chatto & Windus. ——— (1976). Keywords. Fontana. ——— (1981). Culture. Fontana. Wrench, J. & J. Solomos (Eds) (1993). Racism and migration in Western Europe. Berg. Wuthnow, R., J. Hunter, A. Bergesen & E. Kurzweil (1984). Cultural analysis. Routledge.

Elicitation Gunter Senft Max Planck Institute for Psycholinguistics, Nijmegen

The technical term ‘elicitation’ is derived from the classic Latin verb elicere (to coax, entice, call forth, summon, extract, induce, provoke). In linguistics (as in sociology, social psychology, and other social sciences) elicitation is the general term for describing various methods of directed data collection and thus for corpus construction. With the rise of the interest in dialects, and thus in spoken languages, linguists had to develop means for gathering their data. One of the first, and classic, means they came up with were questionnaires with sentences written in the standard language that had to be translated (in general by teachers and priests) into the local language. The pioneer for this kind of research was probably Georg Wenker with his 40 ‘Wenker-sentences’ which he started to send off in the German Rhineland in 1876 (Knoop et al. 1982: 47ff). Comrie & Smith’s famous Lingua descriptive studies: questionnaire (1977) and other comparable publications can well be regarded as continuing in one way or another this tradition of data gathering. However, although questionnaires can be extremely helpful, at least for starting data gathering procedures, they are of little use if the linguist is interested in how the language is really spoken in co-present interaction. Already 40 years before Wenker, Johann A. Schmeller emphasized the relevance of what we now call participant observation and field research — the interaction between linguists and their informants — for the collection of speech data (Schmeller 1855). These two types of data collection — asking questions (or just a question) following a questionnaire on the one hand and participant observation together with intensive field research on the other hand — mark the two extremes in linguistic data elicitation. However, this does not imply that these extremes are mutually exclusive. Fieldworkers have to use as broad a variety of elicitation procedures as possible in their linguistic field research. It goes without saying that scientific data are always collected according to specific research interests and purposes. Linguists must decide on –– –– –– ––

what kind of speech data they want to elicit in what group(s) of informants in which situations and settings within which speech communities and cultures.

106 Gunter Senft

The respective interests then are the guidelines for choosing the adequate elicitation method(s) and for defining a sample of informants that should be (as) representative (as possible). Another basic decision that has to be made is whether the linguist’s informants may or may not all the time be aware of the fact that their speech and their speech behavior is being observed. If linguists want to learn something about the (inflectional) morphology and the syntax of a language they have to start with the (sometimes tedious) elicitation of the respective morphological and syntactic patterns (see e.g., Foley 1991). This kind of elicitation is quite similar to other kinds of data elicitation that rely mainly on questionnaires. If linguists are interested in, e.g., the lexicon of color terms in various languages, they just can confront their informants with the 329 color chips provided by the Munsell Color Company — as Berlin & Kay (1969) did — and ask them to name the colors of the chips presented as stimuli in front of a tape recorder. If linguists are interested in formal styles of articulation they may ask their informants to read a wordlist (of minimal pairs) or a text out to the researcher in front of a microphone which then even helps to mark this situation as being formal. If linguists with special interests in pragmatics want to investigate the realization of speech act patterns such as requests and apologies crossculturally, and if they also want to investigate similarities and differences between native and non-native speakers’ realization patterns in these speech acts, they can devise controlled elicitation procedures like discourse completion tests — as Blum-Kulka & Olshtain in their ‘cross-cultural study of speech act realization patterns’ project did. The discourse completion tests used in this project consist of incomplete discourse sequences representing socially different situations. Before relatively brief discourse sequences in the form of incomplete dialogues are presented to the consultants, the situative context of the dialogue is outlined so that the setting, the social distance between the interlocutors and their status relative to one another is specified. The consultants are then asked to complete the dialogue, thereby providing the speech act aimed at in the given context (Blum-Kulka & Olshtain 1984: 198) — in this case a request and an apology. The consultants’ answers to these discourse completion tests allow for inferences with respect to preferences speakers have for realizing requests for action among persons of the same and different social status on the one hand and for inferences with respect to the appropriateness of apologies in the given situation on the other hand. Moreover, the cross-cultural design of this study also allows for answering the question whether there are differences in the types of strategies speakers choose to realize the respective speech acts under the same social constraints across languages and what these differences actually look like. If researchers want to find out how spatial relations are encoded in various languages, they can confront their informants with a kit of stimuli such as the one


developed for this purpose by the Cognitive Anthropology Research Group at the Max-Planck-Institute in Nijmegen. This kit contains, i.a., two sets of identical photographs together with the objects actually photographed, drawings, and toys. With these stimuli researchers can ask their informants to play matching games in front of video camera and microphone. In these games, one informant (the director) describes what is shown on a photo in such a way that the other informant (the matcher) can either find the same photo within a series of similar photographs or reconstruct the described spatial configurations with toys. The game situation asks for verbal interaction that centers on the spatial conceptualisations and their expressions in the lexicon of the various languages that are investigated (see e.g., Levinson 1992). If linguists are interested in narratives, they can ask their informants — be it children or adults — to look, e.g., at a book of 24 pictures with no written text that presents a story (e.g., the so-called ‘frog’-story) and then tell this story to another person while being video-filmed and tape-recorded (Berman & Slobin 1994). Linguists may also ask their informants to watch a movie (like e.g., ‘the pear film’) and then, after even telling the informants that the researchers are interested in studying how people talk about things they have experienced, ask the informants to tell about the movie to people who have not seen it in front of a video camera and/or a tape recorder (Chafe 1980). Both elicitation methods permit verbal interaction between the informants. However, if linguists do not like the idea that their informants are always aware of the fact that they are being observed and that their speech is being recorded, they have to find some ways of overcoming what Labov so aptly called the observer’s paradox: “The aim of the linguistic research in the community must be to find out how people talk when they are not being systematically observed; yet we can only obtain these data by systematic observation” (Labov 1972a: 209). Linguists may get the permission to just leave a tape recorder somewhere in a room in their informant’s house for a whole day and to record whatever is being said there. Of course, the tapes have to be renewed every hour or so and the risk is quite high that nothing is said in this room for a long time, but — as Ruoff (1973: 116) reports — the chances to document “how people talk when they (think or forget that they) are not systematically observed” are not too bad. However, the data gathered in this way are more documented by chance than elicited in the strict sense of the term. Labov developed and described a number of techniques to overcome the observer’s paradox. One of these techniques is the use of rapid and anonymous observations (see also Labov 1972b: 117) which Labov applied in his study on The social stratification of [r] in New York City department stores (Labov 1972a: 43–69). In three stores with different social prestige the interviewer approached an informant asking for directions to a department on the fourth floor. The informant normally responded to this question with the (elliptic) utterance ‘fourth floor’. The interviewer then pretended to have not


108 Gunter Senft

understood the informant and thus elicited a second utterance, this time spoken in careful style under emphatic stress. After this encounter the interviewer noted down some information about the informant and the use of (r) in preconsonantal and final position in casual and emphatic styles of speech. However, even if linguists decide to elicit speech data in interviews, they can prepare and structure these interviews in such a way that they not only result in the elicitation of comparable speech data but that they also provide situations that more or less guarantee the documentation of data that are as ‘natural’ as possible. These interviews are usually called structured intensive interviews, and they are best prepared on the basis of the linguist’s participant observation (Senft 1982: 17–70). In periods of participant observation linguists — like anthropologists — should attempt to immerse themselves into the daily lives of their informants in a kind of field research situation. On the basis of their experiences in this situation the researchers cannot only get acquainted with their future informants, they also have the chance to get a better understanding of what they are asking their informants about. This understanding and the fact that there is already a certain kind of relationship established between interviewer and informant may transform the structured interview into a talk between acquaintances where it does not really matter whether there is a tape recorder running or not. If linguists want to elicit ‘the natural speech data’ they should keep Labov’s general advice in mind: A field worker who stays outside his subject, and deals with it as a mere excuse for eliciting language, will get very little for his pains. Almost any question can be answered with no more information than was contained in it. When the speaker does give more, it is a gift, drawn from some general fund of good will that is held in trust by himself and the field worker. A deep knowledge implies a deep interest, and in payment for the interest the speaker may give more than anyone has a right to expect. Thus the field worker who can tap the full linguistic competence of his subjects must acquire a detailed understanding of what he is asking about, as well as a broad knowledge of the general forms of human behaviour. (Labov 1972b: 114ff; see also Ruoff 1973: 83)

In linguistics, elicited data certainly help to answer a number of specific questions; however, as Duranti (1981: 9 and 162ff) points out, elicitation sessions are speech events that as such influence the kind of language used. Therefore, it should go without saying that linguists aiming at describing the language and speech behavior of a certain speech community as completely as possible just cannot do without additional data that document their informants’ daily verbal communication in face-to-face interactions. For additional information about elicitation I would like to refer the interested reader to Ammon et al. (Eds) (1988) (vol. 2, Chapter 8 on ‘Elicitation methods’), Craig (1979), Dixon (1984), Malinowski (1922), Mayntz et al. (1976), Samarin (1967), Shopen (Ed.) (1979), and Whyte (1943).

Elicitation 109

References Ammon, U., N. Dittmar & K.J. Mattheier (Eds) (1988). Sociolinguistics. De Gruyter. Berlin, B. & P. Kay (1969). Basic color terms. University of California Press. Berman, R.A. & D.I. Slobin (1994). Different ways of relating events in narrative. Erlbaum. Blum-Kulka, S. & E. Olshtain (1984). Requests and apologies. Applied Linguistics 5: 196–213. Chafe, W. (Ed.) (1980). The pear stories. Ablex. Comrie, B. & N. Smith (1977). Lingua descriptive studies: Questionnaire. Lingua 42: 1–72. Craig, C.G. (1979). Jacaltec — field work in Guatemala. In T. Shopen (Ed.): 3–57. Dixon, R.M.W. (1984). Searching for aboriginal languages. University of Queensland Press. Duranti, A. (1981). The Samoan fono. Australian National University. Foley, W.A. (1991). Field methods. In K. Malmkjaer (Ed.), The linguistics encyclopedia: 121–127. Routledge. Knoop, U., W. Puschke & H.E. Wiegand (1982). Die Marburger Schule. In W. Besch et al. (Eds), Dialektologie: 38–92. de Gruyter. Labov, W. (1972a). Sociolinguistic patterns. University of Philadelphia Press. ——— (1972b). Some principles of linguistic methodology. Language in Society 1: 97–120. Levinson, S.C. (1992). Primer for the field investigation of spatial description and conception. Pragmatics 2: 5–47. Malinowski, B. (1922). Argonauts of the Western Pacific. Routledge & Kegan Paul. Mayntz, R., K. Holm & P. Huebner (1976). Introduction to empirical sociology. Penguin. Ruoff, A. (1973). Grundlagen und Methoden der Untersuchung gesprochener Sprache. Niemeyer. Samarin, W. (1967). Field linguistics. Holt, Rinehart & Winston. Schmeller, J.A. (1855). Sogenanntes Cimbrisches Wörterbuch. Joseph Bergmann. Senft, G. (1982). Sprachliche Varietät und Variation im Sprachverhalten Kaiserslauterer Metallarbeiter. Lang. Shopen, T. (Ed.) (1979). Languages and their speakers. Winthrop. Whyte, W.F. (1943). Street corner society. University of Chicago Press.

Ethnography Michael Agar Ethnoworks LLC, New Mexico

1.  Introduction Ethnography is a term that refers to an epistemology, a kind of representation, and a research method. In this article the methodological aspects will be featured. Ethnography as method differs in substantial ways from the usual notion of ‘scientific research’. As a way to begin, a few brief descriptions might help. First, ethnography has its longest history of practice in anthropology, where it has always been the research means towards the end of learning about culture. Second, it is often described using metaphors like ‘student’ or ‘apprentice’ or ‘child’, by which is meant that an ethnographer sets out to learn a way of living, one with which s/he has little prior familiarity. Finally, many would summarize ethnography as the study of ‘context’ and ‘meaning’. In other words, what appears on the surface means something other than what one initially thinks, and it occurs because of circumstances of which one is unaware. Ethnography sets out to learn meanings and contexts which lie outside the concepts and habits of prior experience, to construct and test representations of this new knowledge, and to offer those representations as a characterization of culture. Several different origins for ethnography are celebrated in the histories of the field. Harold Conklin (1968), a founder of ethnographic semantics, mentions Fray Bernardino de Sahagun, the sixteenth century chronicler of the Spanish conquest who described, then later advocated, the rich indigenous cultures that the Conquistadores overran. One can go further back than Sagahun: Tacitus’ description of the Germanic tribes during the Roman era also serves as a frequently mentioned ancestor. Everett C. Hughes (1960), a founder of ethnography in American sociology, prefers his ancestors to be more recent. He writes of the ‘social survey’ movement around the turn of the century in the US and Europe, a movement meant to document and then change the dismal slum conditions in the urban industrial West. Language has always been linked to ethnographic research and the cultural representations that it produces. Another ancestor, the French philosopher Joseph-Marie Degerando, is held up by the Peltos in their review of ethnography as the first author of a field methods book, considerations on the various methods to follow in the observation of savage peoples. Degerando writes that “the first means to the proper knowledge of the savages,


is to become after a fashion like one of them; and it is by learning their language that we shall become their fellow citizens” (see Pelto & Pelto 1973: 241). The tie between ethnography and language has been around for quite some time. Two of the founding fathers of modern ethnography, a German immigrant to the US and a Polish one to the UK, also included ideas about language as a necessary aspect of ethnography in their foundational work. Franz Boas, the founder of American anthropology, looked out at his adopted country and called for urgent ethnographic work with the ‘disappearing’ Native American cultures. His style of ethnography emphasized interviews, the request to older Native Americans to recount their memories, and those experts often spoke unwritten languages. For Boas, language was the critical means to the cultural end, and he founded the International Journal of American Linguistics and wrote the Handbook of American Indian Languages to establish this important fact. His orientation — language as a means to conduct ethnography — set the frame for much of the history of linguistic anthropology that followed (for examples of writings by Boas and other historical figures, see Hymes 1964). Bronislaw Malinowski, considered the founding father of participant observation, spent extensive periods of time on the Trobriand Islands and wrote what are still considered ethnographic classics. Because his view of ethnography included language as a seamless part of everyday life, he created his theory of ‘context of situation’, a theory that bound language to the situational moments and cultural contexts of its use (cf. Malinowski 1935). His work flourished in the UK, first with J.R. Firth and then with Michael Halliday, in a theory that leads directly to the kind of research that many pragmatists conduct today. This article reflects an American anthropological bias because of the author’s background, and because space limitations prevent a consideration of the way the research tradition took shape in other national contexts. Other avenues of exploration that mutually link the understanding of a people — ‘ethnography’ comes from the Greek for ‘folk description’ — and the understanding of a language are of course available. The writings of Vico, von Humboldt, and Bakhtin come to mind as only a few of many prominent examples. And other disciplines have contributed — the sociological tradition of ethnography in the US, for example, gave rise to ethnomethodology, which in turn spawned conversation analysis, another important participant in the intellectual territory of modern pragmatics. The point for now is that ethnography has always included a focus on language, and that much of current pragmatics has ethnographic origins — a point that could be further elaborated with more recent examples, such as Hymes’ ‘ethnography of speaking’ or Gumperz’ work on ‘discourse strategies’. In this brief summary, though, I only hope to define ethnography, first by example, and then by contrast. The moral of the story, the one I will return to in the conclusion, is that ethnography gives pragmatics the possibility to shift from unstable, top-heavy interpretations to stable, grounded



Michael Agar

ones. What pragmatics gives to ethnography are vivid displays of how culture comes to life, but that side of the story is neglected for reasons of space limitations, and because those who practice pragmatics are already aware of that.

2.  Ethnography by example First, let me present a definition by example, a story that reveals how language and ethnography combine, in practice, to yield representations that tack back and forth between linguistic detail and history and society (Agar 1994). One bit of detail that caught my attention when I worked in Austria in 1989 was the term Schmäh. It had registered when I was in Vienna three years earlier. I’d heard the term, but in the hectic flow of a new language I’d tagged it with a rough gloss and left it for a future date. But in rapid succession after my arrival in 1989, I heard it several times, saw it used in a book review and a newspaper article, looked it up in a couple of dictionaries, and read about it in a guide book. It looked central, slippery, and interesting, and I had no idea what it really meant. I was assigned a lecture course at the Linguistics Institute at the University of Vienna, and I decided to show the students some of the ways an ethnographer would go about looking at language and then ask them to collect some data. I considered the concept of Schmäh as a candidate. I decided to try it out. At a lunch with some Austrian friends, I told them I was thinking about looking at Schmäh in my class. You can’t imagine the laughter, which of course was the first sign that I’d made the right choice. Everyone at the lunch said it was a good idea. I asked if they knew what Schmäh meant. Of course they did. But then, as we discussed it, all kinds of disagreements followed. Schmäh was Viennese. No it wasn’t, it was Austrian, or universal. It was something men did. No it wasn’t. It was more characteristic of the lower classes. No it wasn’t. It was telling jokes, picking up a woman in a bar, manipulating a situation, what politicians did, a way of life. I sat back and listened and realized that Schmäh was not the sort of lexical item that lent itself to a straightforward dictionary entry. I asked the students in my course to do three assignments. I’ll describe them briefly: 1. A systematic interview in the tradition of cognitive anthropology around the concept of Schmäh. Such interviews take some schema, like a taxonomy or case grammar, place the concept in the center of it, and then pose systematic questions that represent relationships and place the answers to those questions in the appropriate slots. 2. A collection of anecdotes of Schmäh use encountered in everyday life. The notes that result are like the field notes traditionally collected in participant-observation.


3. An informal interview about Schmäh. Such interviews allow the native speaker to discuss the concept in whatever way they choose. Methods of discourse analysis can be applied to such data to make explicit the underlying folk-theory that contains the concept. The students were beginners, and I make no claims to a finished study. But from our discussions over the semester and the oral reports that the students delivered at the end, I’d like to summarize a few of the results. Schmäh is first of all a basic cultural premise, a Whorfian way of looking at life, a general attitude, and it is in this sense, I think, that one talks about the Viennese Schmäh. A notion that repeatedly appeared in interviews and conversations was that Schmäh was a way of looking at things, often described by contrast with the expression that it was nicht ernst, not serious, but a not-seriousness of a particular kind. Schmäh as worldview rests on irony, on the fact that things are not as they appear, on the difference between dream and reality, to use another Austrian cliché. And reality is cruel, full of harmful events and ill-intentioned others. Several interviewees mentioned that Schmäh was a way to deal with grisly reality, a way to convert this reality into humor, a release of hardship — real or imagined — through laughter. In this sense, Schmäh has something to do with schwarzes Humor, black humor, and several people mentioned this. But the difference is that, while a particular Schmäh attitude at a particular moment might also count as an example of black humor, black humor isn’t used to describe a general orientation to life. Schmäh is. At this general level, Schmäh isn’t directly connected to particular bits of discourse. Rather, it labels a wide-ranging premise with implications for numerous situations that do connect in a more intimate way. But the term doesn’t just label the basic premise; it labels two specific situations as well. In the first kind of situation, Schmäh is a humorous comment or exchange that arises from the details of the moment. Many interviewees emphasized that not everyone can perform Schmäh; it is a skill requiring intelligence and wit. And several mentioned that it is not ‘telling jokes’ — a description that people sometimes give at first — because jokes are pre-scripted and Schmäh is not. An example would be a student telling an American visiting professor teaching in German: “You know, your English has gotten much better since you came to Vienna”. There were arguments in class and disagreements among interviewees as to whether the Schmäh is bösartig or gut gemeint, roughly, done with good or bad intentions. Among the histories of the term that interviewees offered, one mentioned the Austrian wife of the French king, Marie Antoinette. When told that the poor had no bread, she uttered her famous line, usually translated into English as ‘let them eat cake’. The interviewee pointed out that Maria Theresa’s daughter had just uttered a Schmäh, but the French, as history has shown, didn’t see it that way.



Michael Agar

A second use of Schmäh is to label a lie, again with disagreement over whether it is ill- or well-intentioned, a lie that is linked to some personal, instrumental end. Again the folk history is interesting, since interviewees claimed that this version of Schmäh had roots in the monarchy, where one often had to manipulate people with authority over them to get something done. This Schmäh is something different from the first type. Such a Schmäh may or may not be funny, and the listener may or may not know that a Schmäh is in progress. Another example, one that is funny, obvious to the listener, and an illustration of how a Schmäh may be nonverbal, goes like this. A man pulls up and parks in a loading zone. A policeman tells him he can’t park there. The man walks into the coffee house, brings a chair out, and puts it in the car, drives around the block, parks again, and carries the chair into the coffee house. In this case the policeman laughed and let him park. The lie was obvious and humorous, but the goal was achieved all the same. To sum it up, Schmäh is a view of the world that rests on the basic ironic premise that things aren’t what they seem, what they are is much worse, and all you can do is laugh it off. Such an attitude is hardly unique to Vienna. What is unique to Vienna is that the premise, with all its complicated strands, is puttied into a single piece of language, and that rich piece of language is, in turn, used as a badge of identity. The Schmäh worldview finds expression in at least two different situations that are also labelled Schmäh: one, a humorous exchange that grows out of the moment that is based on a negative portrayal of the other; two, a deception designed to attain some instrumental end. Both specific examples fit the general philosophy — things are not what they seem, what they are is bad, but the fact that the difference exists is not to be taken seriously. When Schmäh is lifted out of the language, the associations that come with it drag along the raw material for a complicated but coherent set of meanings with links to history and culture. The ease with which the term is used in Viennese discourse to characterize situations and persons and verbal and written expressions testifies to its centrality and power, as are the disagreements when people discuss what it means. Schmäh is a laughing surface laid over an ugly world, a way of seeing and at least two different ways of talking within it. Now, this story — or rather the experience that it represents — is only one fragment of the overall ethnographic picture, but it contains several key characteristics of the way ethnography works. First, I was there. When an ethnographer takes an interest in some corner of the world, he or she goes out and encounters it first hand. In the jargon this is called participant observation, an awkward term that simply codes the assumption that the raw material of ethnographic research lies out there in the daily activities of the people you are interested in, and the only way to access those activities is to establish relationships with people, participate with them in what they do, and observe what is going on.


A second key characteristic is that something happened I didn’t understand. I would never have predicted that the term Schmäh would turn into a lever that pried deeply into Austrian society. Participant observation makes it possible for surprises to happen, for the unexpected to occur. Such problems in understanding are called rich points. When a rich point occurs, an ethnographer learns that his or her assumptions about how the world works, usually implicit and out of awareness, are inadequate to understand something that happened. A gap, a distance, between two worlds has just surfaced in the details of human activity. Rich points, the words and actions that signal those gaps, are the unit of data for ethnographers, for it is this distance between two worlds of experience that is exactly the problem that ethnographic research is designed to locate and resolve. A third characteristic: once a rich point occurs, another working assumption sets up what an ethnographer does next. That assumption is one of coherence. The rich point, you assume, isn’t their problem; it’s your problem. The rich point doesn’t mean that they’re irrational or disorganized; it means that you’re not yet competent to understand it. There is, you assume, a point of view, a way of thinking and acting, a context for the action, in terms of which the rich point makes sense. Your job, as an ethnographer, is to find out what it is, model it in some way, and check the model out in the subsequent words and actions of the group. The Schmäh story contains within it, then, three important pieces of ethnography. You do participant observation to establish the situations where rich points can occur, and, when they do occur, you assume coherence on the part of those who produced it and set out to reconstruct that coherence. Participant observation makes the research possible; rich points are the data you focus on; and coherence is the guiding assumption to start you off on the research that those rich points inspire. Once you’ve talked and observed and figured out the rich point, you’ve only begun the job. Once you’ve marshalled a host of ethnographic methods — and here the entire literature on methodology becomes relevant — and figured out the frame that shows how a simple lexical item ties into so many situations and meanings, you take that frame forward into conversations and try it out. In the case of Schmäh subsequent use of the frame leads down some interesting trails, as the story showed. As one uses the frame, one not only works to validate it, but also to modify it through use. And, in this case, the modifications shows how Schmäh is a key to understanding other rich points as well. The original frame carries one well beyond it. As one validates and modifies the frame through use, one enriches its content and learns more about its scope of application. The frame fills in the original distance between ethnographer and group, lets one see and understand and act in ways that now make sense, from the language speaker’s point of view. The ethnographer has assembled a fragment of culture, for that’s what culture is, knowledge one constructs to show how acts in the context of one world can be understood as coherent from the point of view of another world.



Michael Agar

This notion of ‘culture’ calls for some comment, since the traditional use of the concept has to do with something ‘those people’ have, something learned as a consequence of group membership and shared with other members of that group. The old concept doesn’t work anymore for several reasons, because of growing appreciation for the mix of human universals and cultural differences, because of variation in complex society, because of a clearer sense of what happens when one human sets out to understand some others. The issues are beyond the scope of this summary, but the notion of ‘culture’ offered above is one better grounded in ethnographic practice (see Rosaldo 1989; Hannerz 1992; Roseberry 1991 for other discussions). The story of Schmäh contains within it some of the key elements of ethnographic research. A full study, of course, is more complicated than this. The business of finding rich points and building frames, in actual ethnographic work, goes on at different levels in different ways, all the time, and one continually works to weave the frames into ever grander systems of understanding, to account for ever increasing numbers of rich points in different domains of life. And I haven’t described the appropriate methods to systematically organize and represent what you’ve done. Such expertise requires years of training. But the story does sketch the contours of the research territory, and it shows how an initial focus on a simple term, followed by an investigation of numerous other meanings and contexts, inspired the pragmatic wherewithal to understand and participate in numerous different speech events. I’ve used my own work here to exemplify how ethnography and language work together. There are, of course, many other examples. Let me mention just one, an example that represents more of a traditional anthropological research setting. In his recent book, Alessandro Duranti (1994) describes how he went off to Samoa in search of ergativity. He found it, but not with the frequency or distribution that he expected. The book then leaves the issue of grammar and sets off into an ethnographic investigation of a speech event, the people and issues that bring it to life, and the community in which it occurs. At the end of the book, we return to ergativity, only now the distribution of its use and its tie to evaluations of responsbility for actions in conflict situations comes clear. A linguistic puzzle is solved. The trick is that most of the pieces of the puzzle came from the ethnography rather than the linguistics. That is the thematic moral of the story in this article. Ethnography sets out to study contexts and meanings, and so do pragmatists. The relationship between the two fields is fundamental, as indicated by the earlier sketch of some of the historical figures. The problem, as mentioned at the beginning of this article, is that pragmatists might build interpretations on top of a few focused transcripted moments of language. These representations necessarily appeal to more general notions of social identities and the community in which they take their shape, other contexts and meanings, and the results of the analysis apply by inference to numerous other instances of speech that are not


examined. In the earlier jargon of this section, rich points are identified and frames are built. But the links to other rich points are neglected, and the frames lack depth and breadth. Ethnography is the method that is meant to solve just this problem. 3.  Ethnography by contrast In the rhetoric of social research, ethnography has long been considered somewhat marginal. Such evaluations are a product of comparison with the so-called ‘received view’, a particular notion of what social research is that is deeply embedded and highly valued in the American tradition. When ethnography is measured against this tradition, it generates a long list of deficits. The problem is that the evaluation is a bit like measuring computer capacity in cubic feet. You can do it, but it somehow misses the point. (For discussions of ethnography that develop this comparison, see Agar 1985. Briggs 1986 and Mischler 1986 discuss ethnographic interviews in the same comparative frame of reference, and they both do so off of a language-oriented base). The world of social research is more sophisticated now. Ethnography is understood to be an alternative, something distinct from the received view, not an inferior form of it. When a research question pertains to the distribution of some measurable variables, or to the test of a prior hypothesis, then the received view is the obvious way to go. When the research question is more global — who are these people and what are they saying — then ethnography provides the answers. The differences between the two approaches appear at several levels. For example, in the received view ‘data collection’ is one stage in a linear process, something that happens after the variables in the hypothesis have been operationalized and before the analysis gets done. In ethnography, in contrast, data collection and analysis are continual and dialectic. Data from participant observation or interview are analyzed. Then, based on that analysis, more data is collected, which leads to more analysis, and so on throughout the research process. A second example that follows from the first: in the received view, a ‘hypothesis’ is a clearly stated relationship among two or more variables. In ethnography — if we extend the definition of ‘hypothesis’ to mean ‘an idea to check out’ — one tests multiple linked hypotheses continually. At any particular moment during the research, several questions based on previous work are on the front burner, and relevant data — intentionally sought or serendipitiously encountered — are watched for to test them out. As another example, consider the issue of sampling. In received view research, the sample is typically specified before the fact. In ethnographic research, two considerations guide sampling. First, because of the emphasis on ongoing high rapport relationships, random sampling makes no sense at all. One has to work with people who are willing to spend the time. Second, significant dimensions of population variation are



Michael Agar

learned only after the research has started. For these reasons, ethnographic samples are emergent. They are constructed as the research develops, with choices of what kinds of people to work with made as the variation becomes clear. And ethnographic samples are known after rather than before the fact. One keeps a record of the sample as it develops, so that comparison of the ethnographic sample with already available population descriptions can be made later to gauge the ethnographic sample’s representativeness. Yet another difference. The received view emphasizes the quantitative measurement of variables; ethnography is after the description of a system. In ethnography, the units and relationships that make up the system are learned rather than being known a priori, and they involve the presence or absence of qualities — hence ‘qualitative’ — that may not lend themselves to measurement in terms of quantity. As a final example, ‘theory’ in the received view refers to a coherent literature about some domain of human life, a literature that consists of concepts and the propositions that link them. Theory, in fact, is the source of the hypotheses that the received view sets out to test. Ethnography is often associated with ‘description’ rather than ‘theory’, but that is an error. Theory in ethnography grows out of the data in an emergent way, since it organizes the concepts and relationships uncovered during ethnographic research. In the now famous phrase introduced by Glaser & Strauss (1967), ethnography produces ‘grounded theory’, whereas the received view derives a hypothesis from an a priori theory. ‘The culture of group X’ isn’t description; rather, it is a theory of the group grown from the ethnographic data gathered during a study of X. There is a fundamental switch that gets thrown when one moves from the received view to ethnography, a switch that each of the preceeding examples suggest, and that switch goes by the name of control. The power of the received view rests on control, all the way through the research process, the control of the researcher. With ethnography, control is handed back over to the persons and situations you are interested in, and the job involves locating concepts to account for the ethnographic data that the world produces when you let it.

4.  Discourse Ethnography, then, offers a distinct kind of research logic, one that differs substantially from the usual notion of ‘science’. The study of language and ethnography have always gone hand in hand. In fact, a view of ethnography based on linguistic rich points and frames is not so different from some pragmatic approaches to the analysis of discourse. That link is clearly established by Sherzer (1987) when he points out that the relationships between language and culture only materialize in discourse. At some point, all pragmatics work must leave the world of available linguistic detail and call up interpretive assumptions about the worlds of speakers and hearers, worlds


in the sense of biography and immediate circumstance, and also in the broader sense of social identity, culture, and political economy. The problem here is that the analysis often rests on a single transcript fragment, at most a couple of them, but the material that supports an ascent into a broader understanding of community is not available. The result is what was called at the beginning of this article an ‘unstable, topheavy’ interpretation. Visualize the usual pragmatic analysis as a triangle, with its apex balanced on a transcript segment; it rises from the data and broadens out to include interpretations of situation and identity and society that show the significance of the linguistic detail. Now consider the same transcript segment in the context of an ethnographic study. It would be one of a number of segments, some of which would be transcripts, some of which would be notes from participant observation, and some of which might be media clips, or works of literature, or even recent statistics of GDP and an excerpt from the national tax laws. Now the base of the triangle rests on several different segments, and the interpretation ascends and narrows at the apex, supports and validates it, so that the high level interpretation that results — call it culture, call it ideology, call it fundamental premise — is firmly supported across a range of experiences. In this sense, then, ethnography plays a role in pragmatics by grounding interpretations through an emergent analysis across many segments of data in a quest for a high level coherence that connects with multiple moments of language. Pragmatics combined with ethnography is still about language in use; but it adds to its repetoire an ability to systematically ground and articulate that elusive link between communication, identity, and society that gives pragmatics a more solid base for its interpetations.

References Agar, M. (1985). Speaking of ethnography. Sage. ——— (1994). Language shock. Morrow. Briggs, C. (1986). Learning how to ask. Cambridge University Press. Conklin, H.C. (1968). Ethnography. In D.L. Sills (Ed.), International encyclopedia of the social sciences, vol. 5: 172–178. MacMillan. Duranti, A. (1994). From grammar to politics. University of California Press. Glaser, B. & A. Strauss (1967). The discovery of grounded theory. Aldine. Hannerz, U. (1992). Cultural complexity. Columbia University Press. Hughes, E.C. (1960). Introduction. In B.H. Junker (Ed.), Fieldwork: v–xv. University of Chicago Press. Hymes, D.H. (Ed.) (1964). Language in culture and society. Harper & Row. Malinowski, B. (1935). An ethnographic theory of language and some practical corollaries. Supplement to Coral gardens and their magic. Dover Press. Mischler, E.G. (1986). Research interviewing. Harvard University Press.


120 Michael Agar Pelto, P.J. & G.H. Pelto (1973). Ethnography. In J.J. Honigmann (Ed.), Handbook of social and cultural anthropology: 241–288. Rand McNally. Rosaldo, R. (1989). Culture and truth. Beacon Press. Roseberry, W. (1991). Anthropologies and histories. Rutgers University Press. Sherzer, J. (1987). A discourse-centered approach to language and culture. American Anthropologist 89(2): 295–309.

Ethnography of speaking Kristine L. Fitch & Gerry Philipsen University of Iowa/University of Washington

1.  Development and main characteristics The ethnography of speaking is concerned with describing ways of speaking, as they construct and reflect social life within particular speech communities. It is concerned further with developing cross-culturally valid concepts and theories for interpreting and explaining the interaction of language and social life. The object of study within this tradition is situated discourse: how speaking is organized and conceptualized within a given community. ‘Ethnography’ refers to fieldwork, supplemented by techniques developed in other areas of study (notably conversation analysis, history, and pragmatics), which produces a written description of the way of life of a group of people. Its focus is on the observed patterns of speaking, and the symbols and meanings, premises, and rules applied to speaking within a given community. The intellectual roots of ethnography of speaking (or ethnography of communication; the terms are used interchangeably by many scholars) can be traced to anthropologists and linguists such as Bronislaw Malinowski, Franz Boas, and Edward Sapir. Attention to connections between language use and social context in that early work became a starting point for development of the enterprise within a network of scholars from different disciplines at Berkeley and Stanford in the early 1960s. From that group, the work of Dell Hymes (Hymes 1962, 1972, 1974) has been its programmatic impetus. Before that time, established approaches to linguistic description took the phonology and grammar of a language as the principal frames of reference, an approach which privileged attention to linguistic signs within a closed linguistic system. Deviation from the formal system was dismissed as free variation, as error in performance, or as idiosyncrasy. An objective and accomplishment in the ethnography of speaking has been to “push back the boundaries of ‘free variation’ ” (Murray 1983: 260) to demonstrate that such residual variance is ordered, and thus meaningful. As in linguistics, patterns of speaking had been neglected in anthropological descriptions of cultures, in favor of concerns such as political, economic, and kinship systems in which speech (and other modes of communicative activity) were taken for granted, rather than studied in their own right. The ethnography of speaking was designed to remedy the neglect of speaking, both in grammar and in ethnographic studies of culture. Since the appearance of a special issue of American Anthropologist,


Kristine L. Fitch & Gerry Philipsen

compiled by John Gumperz and Dell Hymes in 1964, ethnography of speaking as both a methodology, and as a model of how human beings construct shared meaning from communal and individual experience, has been extended across diverse questions and disciplines in the social sciences and beyond. Since that time, ethnographies have proliferated of legal, educational, medical, and organizational settings, of interpersonal relationships centered around particular roles and tasks in such settings, of communication within and between groups of people distinguished by their age, disability, ethnicity, or social class. Social issues arising from (or reflected in) the diversity of language use across speech communities, such as language policy, discrimination, and stigma and its attendant social consequences have also been examined ethnographically. A foundational premise of the ethnography of speaking is that diversity in the systems of language use is to be explored in all of its complexity. Societies differ as to what communicative resources are available to their members, in terms of languages, dialects, registers, routines, genre, artistic formulas, and so forth. They also differ in how these resources are patterned in use, in the work done (and doable) by and through speech and other communicative means, and in the evaluation of speaking as an instrument of social action. The initial formulation of the ethnography of speaking (Hymes 1962) included a framework for describing the particularities of ways of speaking in diverse speech communities. It was designed to provide an emic/etic framework: an acontextual format for discovering, describing, and comparing cases. That original framework, as a result of application and testing in fieldwork, was revised extensively (Hymes 1972). Important extensions included further development of the social units of description proposed in the original scheme; a typology for characterizing societies as to the quantitative and qualitative importance of speaking; formalized procedures for rule-discovery and rule-statement; and expansion of the number of factors in speech events. Sixteen components comprised the resulting descriptive framework, grouped into 8 main entries, to be remembered by way of the mnemonic device, SPEAKING: S (situation: setting and scene); P (participants: speaker/sender, addressor, hearer/receiver/audience, addressee); E (ends: outcomes, goals); A (act sequence: message form and message content); K (key); I (instrumentalities: channel, forms of speech); N (norms: of interaction and of interpretation); G (genres). The framework is intended to provide not so much a checklist of things to describe, as an initial set of questions and descriptive possibilities in the study of ways of speaking in particular communities. It is also intended to provide a format for comparison across communities, a set of categories for the discovery of similarities and differences. The SPEAKING framework captures, as well, a defining characteristic of the enterprise in that it was never intended to be, or become, a general theory of the possible relationships among the components. Although comparison across case studies is one of the central theoretical moves of the ethnography of speaking, abstraction

Ethnography of speaking

from the complexities of particular cases into universal, independent principles is not. The attention to specific, emic accounts is one aspect of ethnography that most clearly distinguishes it from its linguistic forebears. 2.  Areas of inquiry Some primary areas of inquiry currently pursued within the ethnography of speaking may be described as they reflect the social units of description put forth in Hymes’ formulation. One aspect of pragmatics is that of speech acts, which focuses on the actions performed through speech. From the perspective of the ethnography of speaking, speech acts and their meanings are to be discovered in particular speech communities. This approach contrasts with philosophical discussions of meaning, which rely on notions of logical relationships between language and the world that are, presumably, independent of particular languages and cultures. From the Hymesian view, the speech act level implicates both linguistic form and social norms, and thus is necessarily culturally variable and specific. Many ethnographic studies have examined particular speech acts cross-culturally, both to discover the cultural particulars of speech act performance, and to explore what commonalities might exist. Directives have been investigated in this way, as well as apologies and compliments. One such study, of directives, is Blum-Kulka’s (1990) investigation of family dinner table conversation among Israelis, US Americans, and American immigrant families in Israel. She describes culturally varied perceptions of children’s identities within the family as they are reflected in different styles of directive use. Although parents in all three groups were highly direct when speaking to children, Israeli parents used nicknames and endearments to soften their directives whereas Americans used first names and conventional politeness forms, such as ‘please’. A further contrast between the two groups involved socialization: Israeli parents explicitly taught rules of correct language use, reflecting a concern for maintenance of Hebrew, whereas Americans were more concerned with conversational management, reflecting a concern for formation of relationships through talk. The differences in directive use across cultures illustrates both the commonality of form — directives issued to children were more direct than those that parents used with other adults in every case — and the culturally distinctive styles of directive performance that reflected cultural beliefs about how language is to be used and what aspects of language use are explicitly evaluated and corrected during socialization. Verschueren (1989) has provided a useful cross-language scheme for the study of ‘linguistic action verbs’, such as ‘to speak’ or ‘to say’. Based on a survey of 81 languages, he has discovered that each provides a vocabulary item for the word or words that can be glossed as ‘to speak’ or ‘to say’. Seventy-seven of the 81 languages include a word


124 Kristine L. Fitch & Gerry Philipsen

for the English ‘to talk’. Such a scheme has important heuristic potential for inquiring cross-culturally into the distinctive means of speaking and their meanings because it provides a universal framework for identifying and, subsequently, comparing and contrasting, the particularities of local communicative resources. Another prominent level of analysis within the ethnography of speaking is that of speech events, locally defined contexts for speaking, each of which has an internal structure which differentiates it from other events in a community. Of particular interest are those contexts or activities which are associated with indigenous terms for talk. Carbaugh (1989) discusses 50 such terms drawn from 11 different cultures, and proposes several features of communication, as it functions and is understood within a given culture, that are indexed by such terms for talk. Those features include messages about the prevailing mode of communication (direct or indirect), messages about the nature of society and relationships, common premises about the nature of personhood, types of persons associated with types of talk, and so forth. Carbaugh’s study suggests that the presence of lexical terms for given activities indicates that the activity is both recognized as distinctive, and accorded some importance, within the community. As a recognizable configuration of elements in the SPEAKING framework, the speech event represents a level of analysis that preserves information about the social system as a whole from a firm basis in the particulars of interaction. An example of such a study is the case of salsipuede ‘leave if you can’ (Fitch 1990/91), a leave-taking ritual among urban Colombians. In salsipuede events, attempts to depart from social events are repeatedly rebuffed by the hosts or other guests at the gathering. Refusals of leave-takers’ desires, and denial of their reasons for departure are expected moves within the speech event. They constitute symbolic enactments of the interpersonal bonds formed during the social gathering. The presence of such a speech event, in which individuals’ attempts to leave are ritualistically overridden by appeals to stay connected to others, enacts a broader cultural theme of the significance of interpersonal bonds over and above the desires of individuals. A contrasting example of a speech event is ‘communication’, a ritual among some upper middle class Americans (Katriel & Philipsen 1981), in which unique ‘selves’ and ‘relationships’ are formed through close, flexible, supportive speech. To engage in this speech event, interlocutors disclose personal information and validate the other’s images of self. Speech about one’s unique self and experiences not only informs the listener about those experiences, but gives the speaker an opportunity to express (and thereby develop or enhance) his or her personal uniqueness. These examples suggest not only that something more than information transmission is being accomplished with speech, but also a difference, across groups, in what is accomplished and what values are celebrated through language use. Finally, ethnographic studies of speech communities have described ways of speaking and language codes as definitive of social group boundaries. Such investigations

Ethnography of speaking

are often presented as a series of accounts of the speech practices of a given community. Katriel (1986, 1991), for example, details features of Israeli society that illustrate distinctive threads of Jewish tradition and recent political history, as they each shape everyday talk. Philipsen (1992) describes the patterns and functions of speech in a working class neighborhood of Chicago. Speech in that community is predominantly a means to symbolize one’s place in a local social hierarchy defined in terms of ethnicity, gender, age, and place of residence. In a similar way, a comparative study of preschools in Japan, China and the US (Tobin, Wu & Davidson 1989) explored culturally distinctive orientations toward language use and the desired relationships between individuals and society, as they were formed in the context of early educational experiences. They note that insofar as preschools are embedded in communities, nations, and cultures, they both reflect and affect social change, particularly changes in the structure of the family. They found that Japanese teachers, for example, emphasized language use as a medium for expressing group solidarity and shared social purpose. By contrast, US teachers encouraged children to use words to express autonomy, to form friendships with other individuals, and for problem-solving and cognitive development. In China, enunciation, diction, memorization and self-confidence in speaking and performing were stressed. The researchers note that preschool as a site for socialization of young children has burgeoned in each of these societies during the past 30 years. They concluded that, in all three cases, preschools were essentially conservative institutions, responding to changes in the family rather than (as is sometimes charged) bringing them about. In examining preschools as institutions charged with instilling cultural values, including beliefs about language use, relationships and society, this study illustrates the diversity of such beliefs across distinctive speech communities. 3.  Issues and debates A recently prominent thread of ethnography loosely defined as ‘cultural studies’ raises important issues about the nature of culture, the appropriate position of ethnographers vis-à-vis the communities they study, and perhaps most characteristically, of the relations of power revealed and maintained through cultural practices. ‘Cultural studies’ examine the ways in which ideologies are upheld or suppressed through the symbols, themes, and practices of popular culture (see e.g., Carey 1989; Clifford & Marcus 1986; Hall 1980). It draws upon humanistic traditions as freely as anthropological ones, and searches out culture in its mediated, folk, and high forms more readily than its instantiations in everyday talk. Cultural studies approaches culture as a creation of the power structure, such that to reveal culture is necessarily to expose power at work maintaining the hegemony in a society of a culture. By contrast, the Malinowski-Boas-Sapir-Hymes


126 Kristine L. Fitch & Gerry Philipsen

ethnographic tradition approaches culture as an organic product of a people’s history and experiences. Although power imbalances may certainly exist and characterize a given system, there is no assumption before the fact that such is the case, or that this is to be the exclusive focus of ethnographic description. From each view there is danger in the other of distortion of ‘culture’ and an attendant set of practices designed to avoid such inaccuracies. Within the cultural studies paradigm, a primary concern is that the act of observing cultural practices and committing those observations to writing privileges the researcher’s point of view. Such researchers, scholars within the cultural studies paradigm note, are usually outsiders studying down, i.e., investigating a less-powerful group in ways that members of their own group would not tolerate. Beyond exploiting the observed group, such ethnographic practice risks reifying the status quo and legitimating power imbalances. Some ways in which those risks are minimized, from this perspective, are (a) studying marginalized and oppressed groups whose voices have typically been muffled in the formation of social structure; (b) allowing the members of the group to speak for themselves as much as possible, rather than presenting sterilized glosses altered to fit into predetermined (or analyst-generated) analytic categories; (c) encouraging, even demanding, straightforward discussion by the ethnographer of his/her own position in the ethnographic scene, including emotional reactions, relationships with members of the culture being studied, and personal involvement in the scene. From the Hymesian ethnographic paradigm, little of this is truly novel, nor relevant in every case. The danger in presuming that power imbalances permeate every social scene is that such a presumption violates the practice of discovering culture from the discursive practices of natives. Two examples may give a flavor of cultural studies ethnography. Kondo (1990) describes a Japanese working-class community, in which small businesses enact ‘family’ and the orientation toward family is flavored by a business ethic in which the heaviest emphasis is on continuity. Kondo describes the effort to mold selfhood and obligation toward the family and work into a seamless whole, and the discontinuities, tensions, contradictions and failures of the reality. A central theme in this account is that culture is ever thus: never fixed essences which can be discovered by an ethnographer, but rather a site of everyday contests over meaning and assertions of identity within those struggles. In describing the Japanese ‘self ’ as fluidly constructed within cultural forces of oppression and creativity, Kondo shows how it is that “[i]dentity is not a fixed ‘thing’, it is negotiated, open, shifting, ambiguous, the result of culturally available meanings and the open-ended, power-laden enactments of those meanings in everyday situations” (1990: 24). Her account captures both the domination of the shacho, or boss, and the obligation to care for employees’ needs outside the workplace as well as inside it. Kondo describes parallels to this structure within the family, embodied in a mother’s self-sacrifice offered in exchange for obedience and loyalty from her children.

Ethnography of speaking

Throughout this account is a consistent theme of how unrealistic it would be to speak of the Japanese self or the working class Japanese family as though they were unitary phenomena. Instead, there are shifting currents of motive, relationship and construction of meaning played out within a particular, definable Japanese system of assumptions and communicative resources. Paul Willis’ Learning to labour (1977) focuses on the resistance of working class ‘lads’ to the oppression of a British school, which attempts to mold them into compliant workers who will unproblematically fill menial, even degrading occupations. Willis details creative ways in which working class ‘troublemakers’ within the school retain a sense of importance and identity in the face of an oppressive system in which they are viewed as ‘deviant’. He makes the point that the sector of the educational system he studied is designed to reproduce a class structure marked by inequality on the backs of obedient, unquestioning laborers. This intention is thwarted at least partially by the counterdefinitions of the prospective laborers, by way of identification with a broader working class worldview. From the differentiated view of the working class, practice is valued over theory and ‘having a laff ’ is valued over blind conformity to the system, despite the objective costs of nonconformity. In this way, subordination gives way to creativity and a kind of counterdomination and cultural continuity, defined as opposition to conventional evaluations of manual labor in contemporary capitalist society. Both Kondo and Willis studied working class communities in industrial societies. Each represented cultural meaning as disputed territory rather than as shared understanding that served as grounding for social action. In doing so, they constitute conflict-oriented ethnographies as opposed to the consensus-oriented ethnographies more closely associated with the Hymesian tradition. Of course, work in this tradition has also provided a principled basis for embracing diversity within communities of discourse, with its emphasis on the speech community as an organization of diversity. Furthermore, such work has given principled attention not only to multiple voices within a community, but also to the critical potential of ethnography (e.g., Basso 1986; Huspek 1986; Katriel 1986, 1991). A second issue of considerable debate concerns integration of theoretical paradigms focused on universal aspects of language use with ethnographic studies. Because the thrust of the ethnography of speaking has been to map the cultural and linguistic relativity of language use, universal features or aspects of language use have been a secondary (or more commonly, nonexistent) concern. At times, however, ethnographers have focused productively on universalist theories, i.e., ones that begin with an assumption or claim that particular components inherent in communicative practice are presumably (or observably) universal, apart from theoretically specified areas where variation across cultures is expected to occur. One thread of this debate is exemplified by Rosaldo’s (1982) critique of speech act theory, grounded in the particulars of an ethnographic case. Rosaldo criticized


128 Kristine L. Fitch & Gerry Philipsen

speech act theory as developed by Searle (1975), claiming it reflected a specific cultural perspective rather than universal relationships between linguistic form, context, and meaning. Much of her argument concerns Searle’s proposition that directives (attempts on the part of one person to compel the actions of another) are ordinarily performed indirectly, because uttering them directly would be awkwardly forceful and rude. Based on fieldwork among the Ilongot tribe of the Philippines, Rosaldo concludes that there is nothing inherently forceful or rude about straightforward commands. Among the Ilongot, such commands are far more commonplace than requests phrased more indirectly, because of a conceptualization of personhood and of the orderliness of human life itself which is crucially dependent on explicit guidance of the actions of others. Beyond their function of instructing persons in what actions should be performed, commands are the basis for formation of important relationships. Two other universalist approaches bear mention. The first of these is politeness theory, which proceeds from an observation that speech acts may be performed directly or indirectly. An assumption of this theory is that interaction involves choices that will be received as more or less polite, and that this is a universal aspect of interaction, although which matters raise politeness issues, and how those issues are addressed through language use, is culturally defined. The theory (and related work, e.g., on the cross-cultural realization of speech acts ‘CCSARP’, cf. Blum-Kulka, House & Kasper 1989) formulates a set of uniform factors that serve as a starting point for investigation of spoken life within a culture, and for grounding contrasts across cultures. It centers around a system of categories for description of speech acts intended to capture patterns of language use that supersede (or provide some way to encode) language differences, as well as contextual variation. There is a presumption that cultural beliefs and values can be deduced from patterns of directive use in a speech community, based on the pragmatic structure proposed in the theory. A final universalist trajectory in ethnography involves the use of tape-recorded and transcribed conversation as a data base for ethnographic analysis, a move which engages ethnographers to some degree with conversation analysis. As a method of communication study, conversation analysis (CA) focuses on identifying principles and devices for structuring and coordinating talk that presumably apply across languages, cultures, and situational specifics. To the extent that there is observed variation in the specific aspects of those principles and devices, those variants may be selected for further analysis and cultural explanations may be sought to account for them. Michael Moerman’s Talking culture (1988) and a subsequent colloquy in the journal Research on Language and Social Interaction (1990/91) have established the blending of ethnography and CA as a promising, if still somewhat tentative, venture. The utility of conversation analysis to ethnography, Moerman noted, is that “traditional explainers of social action as ‘class’, ‘ethnicity’, ‘values’, etc., are not things, but processes — processes manipulated or, more radically, composed during the course

Ethnography of speaking

of interaction … [Transcription] can hold the smoke of interaction still for study” (1988: 2–3). At the same time, Moerman argued, CA that remains devoid of cultural context loses its power to capture speakers’ situated meaning. A barrier that continues to constrain attempts to incorporate CA into ethnography is fundamental disagreement about what counts as ‘context’, and what kinds of claims can be made from information that comes from ‘context’. Although there is variation among practitioners, CA generally restricts consideration of context to the specifics made relevant in a particular text. The commitment is to a talk-intrinsic notion of context (see Mandelbaum 1990/91; Schegloff 1987), in contrast to ethnography’s talk-extrinsic view of context, which includes reactions to spoken events as well as the events themselves, metacommunication, and so forth. A recent collection of essays (Duranti & Goodwin Eds. 1992) explores the notion of context further, bringing together research and discussion from conversation analysts and ethnographers into a useful exploration of some of the issues involved: levels of organization of talk on which context may be differently relevant; the perspective of the participant in determining what counts as context; figure/ground relationships, and so forth. A more ethnography-oriented collection by Auer & DiLuzio (Eds) (1992) is also relevant. So far, the endeavor to blend CA with ethnography remains more an intriguing challenge than a realized accomplishment.

References Auer, P. & A. di Luzio (Eds) (1992). The contextualization of language. John Benjamins. Basso, K.M. (1986). Portraits of ‘the Whiteman’. Cambridge University Press. Blum-Kulka, S. (1990). You don’t touch lettuce with your fingers. Journal of Pragmatics 14: 259–288. Blum-Kulka, S., J. House & G. Kasper (Eds) (1989). Cross-cultural pragmatics. Ablex. Carbaugh, D. (1989). Fifty terms for talk. In S. Ting-Toomey & F. Korzenny (Eds), Language, communication and culture: 93–120. Sage. Carey, J. (1989). Communication as culture. Unwin Hyman. Clifford, J. & G. Marcus (1986). Writing culture. University of California Press. Duranti, A. & C. Goodwin (Eds) (1992). Rethinking context. Cambridge University Press. Fitch, K. (1990/91). A ritual for attempting leave-taking in Colombia. Research on Language and Social Interaction 24: 209–224. Hall, S. (1980). Cultural studies. Media, Culture, and Society 2: 57–72. Huspek, M. (1986). Linguistic variation, context, and meaning. Language in Society 15: 149–163. Hymes, D. (1962). The ethnography of speaking. In Anthropology and human behavior: 15–53. Anthropological Society of Washington. ——— (1972). Models of the interaction of language and social life. In J. Gumperz & D. Hymes (Eds), Directions in sociolinguistics: 35–71. Holt, Rinehart and Winston. ——— (1974). Foundations in sociolinguistics. University of Pennsylvania Press. Katriel, T. (1986). Talking straight. Cambridge University Press. ——— (1991). Communal webs. State University of New York Press.


130 Kristine L. Fitch & Gerry Philipsen Katriel, T. & G. Philipsen (1981). ‘What we need is communication’: ‘Communication’ as a cultural category in some American speech. Communication Monographs 48: 301–317. Kondo, D. (1990). Crafting selves. University of Chicago Press. Mandelbaum, J. (1990/91). Beyond mundane reason. Research on Language and Social Interaction 24: 333–350. Moerman, M. (1988). Talking culture. University of Pennsylvania Press. Murray, S. (1983). Group formation in social science. Linguistic Research, Inc. Philipsen, G. (1992). Speaking culturally. State University of New York Press. Rosaldo, M.Z. (1982). The things we do with words. Language in Society 11: 203–237. Schegloff, E. (1987). Between macro and micro. In J. Alexander, B. Giesen, R. Munch, & N. Smelser (Eds), The micro-macro link: 207–234. University of California Press. Searle, J.R. (1975). Indirect speech acts. In P. Cole & J.L. Morgan (Eds), Syntax and semantics, vol. 3: 59–82. Academic Press. Tobin, J., D. Wu & D. Davidson (1989). Preschool in three cultures. Yale University Press. Verschueren, J. (1989). Language on language. IPrA Papers in Pragmatics 3: 5–144. Willis, P. (1977). Learning to labour. Columbia University Press.

Fieldwork Gunter Senft Max Planck Institute for Psycholinguistics, Nijmegen

The Compact Oxford English Dictionary (1991 Ed.) defines ‘fieldwork’ as “a comprehensive name to describe the practical side of research in archaeology, linguistics, the social sciences etc., carried out in the areas concerned, as distinguished from theoretical or laboratory investigation”. In the disciplines mentioned, but also in anthropology, in anthropological linguistics or ethnolinguistics, in dialectology, in sociolinguistics, in branches of linguistic pragmatics, in sociology, social psychology and in behavioral ethology, ‘fieldwork’ is the cover term for a complex of methodological approaches to data gathering. The term subsumes a variety of ways of collecting data ‘in the field’, such as elicitation, interviews, and participant observation where the observer interacts with the observed. Thus, fieldwork typically involves the presence of the researcher among what s/he has defined as her/his subject: society at large (whatever this may be), a group of people, a speech community, a restricted category of well-specified individuals serving as ‘consultants’, etc. Ideally, the researcher is physically present in her/his field. There are certain cases, however, where either some technical apparatus, as e.g., recording equipment that is used for data gathering purposes or a trained local consultant who is collecting data for the researcher, may serve as substitutes for the scientist (thus making her/his presence indirect, so to speak). That the researcher has to be extremely careful in selecting this local consultant is a point which will be discussed below. As already indicated by the list of disciplines where fieldwork is done, it must be mentioned that fieldwork can also include forms of unobtrusive observation in which the observer, e.g., the archeologist, collects data without interaction with any kind of (animate) participants at all. Finally, it should be emphasized that this methodological approach is hardly ever used as the sole source of information for the scholar. It is mostly combined with other forms of data collection such as experimentation, archive research, corpus analysis, etc. Disciplines such as archaeology, ethology, psychology, and sociology can certainly (and proudly) refer to their pioneers with respect to fieldwork done, and field-research methods used. However, by far the most, if not all, modern forms of fieldwork in these various disciplines owe much more to both cultural and social anthropology and anthropological linguistics with respect to the development of this complex of methods than to any other discipline within the social sciences (Goldschmidt 1976). Thus,


Gunter Senft

besides the early demands for fieldwork or field research and participant observation as methods of data gathering in dialectology by researchers like Johann A. Schmeller (who referred to his ‘field-trips’ to remote valleys in the Alps as Kundfahrten, i.e., ‘reconnaissance trips’) in the first half of the 19th century, it was the pioneering fieldwork of anthropologists that promoted this method and approach especially within the social sciences. Probably the most important advocate of fieldwork was the anthropologist Franz Boas who undertook his first one-year-long scientific expedition to Baffin Island in 1883–84. All his students, including B.A.L. Kroeber, R.H. Lowie, P. Radin, and especially E. Sapir, did fieldwork themselves and established this method in American anthropology. The famous Cambridge Anthropological Expedition to Torres Straits, organized by W.H.R. Rivers and A.C. Haddon, the ‘Südsee-Expedition der Hamburgischen Wissenschaftlichen Stiftung 1908–1909 im Bismarck- Archipel’ (Expedition to the South Seas of the Hamburg Scientific Foundation in the Bismarck Archipelago in 1908–1909), the research on the Andaman Islands by A.R. Radcliffe-Brown, and especially the exemplary studies of the Trobriand Islanders’ culture by Bronislaw Malinowski (see e.g., Malinowski 1922, 1936), but also scientists like M. Gusinde, P. Schebesta, O. Raum, and R. Thurnwald document a similar development with respect to the importance of fieldwork in British and German anthropology. The concept of fieldwork-based ethnography, developed in anthropology during the first half of our century, had strong influences on ethnomethodology (Garfinkel 1967) and the ethnography of speaking/communication framework in (socio-)linguistics (Gumperz & Hymes 1972; Hymes 1964, 1978). To emphasize it once more, fieldwork in general is a complex of methods used for data gathering in the social sciences that subsumes and incorporates a broad variety of means for data collection — both in Western and in Non-Western societies (the latter of which are often referred to — from a somewhat ethnocentric point of view — as ‘exotic’ societies). It goes without saying that methods of fieldwork vary according to the specific disciplines; however, the target of fieldwork is more or less well-delimited in each case: to collect a set of ‘natural’ data, which then serve as the material for various kinds of analyses. Moreover, although the situations for, and of, fieldwork in the various disciplines are certainly different and do vary, the basic problems the researcher encounters and has to overcome seem to be quite similar to those problems anthropologists and anthropological linguists have to face and solve when they are doing fieldwork. Comparing this author’s experiences doing sociologically and sociolinguistically oriented fieldwork with Italian, Spanish and German workers in Heidelberg (Heidelberger Forschungsprojekt 1977), social-psychologically, socio- and psycholinguistically oriented fieldwork with German metalworkers in Kaiserslautern (Senft 1982), and human-ethologically and anthropological-linguistically oriented fieldwork with Trobriand Islanders in Papua New Guinea (Senft 1986, 1992), differences become really marginal. Therefore, in what follows, fieldwork will be discussed mainly from


the anthropological and (anthropological-) linguistic point of view (though this may result in some form of bias). Anthropological and linguistic fieldwork requires first of all the researchers’ participant observation in the (speech) community and in the culture they want to investigate. Linguists and anthropologists can best study and describe a language, a language variety, a culture, or a subculture if they live within a city, a suburb, a village etc. where they can learn as much of the social customs, the language and language use of the people as possible. Especially for linguists, a good understanding of language use in its social context(s) turns out to be extremely important because the social context(s) of language use directly affect(s) aspects of language structure. The definition of their roles as participant observers is the result of ongoing interactive processes between the researchers and the communities in which they live and work. If the researchers work in communities with relatively great cultural and linguistic distance to the communities from which they come, the scientists have to pass a (more often than not difficult) initial phase of a first orientation in their new environment. The researchers’ main aim must be to get accepted — and somehow ‘socialized’ by their new communities. Especially in Non-Western cultures it may help if researchers can show photographs to the members of the community with which they want to live that document their personal (family) ties and aspects of their way of living within their own (Western) culture and community. During this phase fieldworkers start to learn the language, try to live within, and understand, the new, the ‘other’, pattern of life, and make first contacts with the people speaking the language and living (in) the culture they want to investigate. As a rule of thumb, the greater the cultural and linguistic distance between the researchers and the communities under investigation, the longer the fieldwork should last. Where a new language and culture must be learned, fieldwork should be done for at least a year; ideally, the researchers should stay in their respective fields for a longer period of time and also return to their fields after periods of data analysis and phases of planning further research on successive field trips. After this first phase of orientation — which is necessarily open and rather ‘unstructured’ and thus asks for the researchers’ abilities to respond and react ‘pragmatically’ and spontaneously to their new life situation — there usually comes a more exploratory, and also a more structured, phase where the researchers are now in a position to start their documentation of the respective cultures and languages. It is in this phase that linguists may start with pattern elication and where anthropologists start with their routines of data gathering such as taking a census and recording genealogies. In the beginning of this phase, all researchers are almost completely dependent on so-called ‘main consultants’ whom they should select very carefully. This selection of the main consultant(s) can be a rather delicate matter. There are societies where it is considered improper if a consultant — and especially if the main consultant — is


134 Gunter Senft

the opposite sex to the fieldworker. Moreover, it can be extremely difficult for both the researcher and the consultant to change their cooperative relationship if it turns out that there are better candidates for the role of main consultant than the one the researcher had chosen first. Before fieldworkers in such a situation actually change their main consultant(s) they must know whether this switch will be regarded as an acceptable act in the community or whether it will be interpreted as social rejection or even a rebuff. Good main or primary consultants should be interested in teaching the fieldworker something about their language, language use and culture. They should be accepted within their own community. They should be deeply rooted in their own language and culture, but at the same time they should be so open-minded that they can cope with the misunderstandings caused by clashes between their own experience and the fielworkers’ different social backgrounds. To overcome these misunderstandings requires mental alertness, a communicative and also a somewhat extrovert personality, and a general feeling of mutual sympathy in both the fieldworker and the consultant. If there is a contact language in the area where the fieldwork is conducted, it is quite helpful if the main consultant also has a good command of this language (if the fieldworker does so, too, of course). However, researchers have to be very careful in not selecting someone as their main consultant who speaks this contact language, but otherwise has a rather ‘marginal’ status within her or his own community. The longer researchers live in their fields the less dependent they become on these first main consultants. The linguistic and/or anthropological information which they can then gather with the help of more and more consultants offers a somewhat more representative sample and data corpus for the scientists’ specific research interests. However, fieldworkers should always keep in mind that their sex may open up or completely close down certain roads of special cultural information. It has been my own experience during my fieldwork on the Trobriand Islands that it is best to be in the field as a male and female team or even as a family. In this way, one can gather more information about the different ‘worlds’ of male and female, married and unmarried than one could hope for alone in the field. The longer the researchers are in the field, the broader will, and can, be the variety of data gathering techniques employed to reach their research aims. Interview and elicitation techniques can become more and more sophisticated, it will be possible to get more natural data by documenting (on film and/or audio- and video-tape) everyday or ritualized social interaction patterns. To a certain extent, field research can now also become even ‘experimental’. It is at this stage that researchers can rather highly structure their fieldwork in accordance with their respective research interests. Data gathered in the framework of a primarily linguistic research project, however, should (ideally) always encompass as many different types of text as possible. In linguistics, elicited data certainly help to answer a number of specific questions. Elicitation methods are characterized by standardized formats of interaction between the


researcher and the consultants: usually, the researcher asks a number of her/his consultants to answer standardized questions, to perform standardized tasks, or to solve standardized problems that are targeted at precise data types or themes. However, as Duranti (1981: 9 and 162ff) points out, elicitation sessions are speech events that as such influence the kind of language used. Therefore, the data collected should also include narratives about personal experiences, conversations of people in face-to-face interaction, legends, myths, songs, poems, nursery rhymes, and other forms of oral literature and verbal interaction. Thus, ethnographic methods, characterized by ‘natural’ and spontaneous interactions documented among the consultants and between the researcher and her/his consultants should always and ideally go hand in hand with these elicitation methods. Needless to say, these two methods constitute the two main poles around which fieldwork in pragmatics revolves. It is usually best to transcribe the gathered data in the field, first with the help of consultants, later under the control of consultants: even if it is possible to transcribe some texts alone back home in one’s study, it is too often just too easy to miss something while transcribing like this. All the data gathered during the researchers’ fieldwork form the linguists’ or anthropologists’ corpus on which they rely for their description and analyses of the respective languages and cultures. It goes without saying that the fieldworkers’ data collection which finally results in this corpus is always guided, if not governed, by theoretical considerations which clearly state what purposes the corpus should serve. However, fieldwork certainly is the area in the social sciences where researchers can experience the dialectics between theory and practice to its extremes. On the one hand, the practice of the field constantly forces researchers to reshape, criticize and reformulate their starting hypotheses and theories — on the other hand, it is these hypotheses and theories that help the fieldworker to not get ‘drowned’ in a ‘sea’ of sheer data. It is exactly here that fieldwork as one of the main sources of data-collection in the social sciences has to face criticism — especially with respect to its epistemological and methodological status (see e.g., Ellen 1984). In what follows I exaggerate ‘arguments’ to illustrate the extremes of the conflicting positions. For many fieldworkers criticism of their research comes especially from those theoreticians (fieldworkers often refer to them as ‘armchair’ anthropologists, linguists, or sociologists) who abhor the idea of wading through the morass of empiricism which may demolish their neat theories developed without contradiction and safeguarded by hardly vulnerable immunizing strategies. These theoreticians often denounce fieldworkers as being adventurers who flee their own culture (or laboratory or library) of which they have become estranged and then return from their romanticized ‘field’ with highly subjective data that claim scientific status though they can neither be falsified nor verified by others in a repeat study. Because some of the results the fieldworker reports are — necessarily — based on subjective impressions, this fact is used by those critics to question the use of fieldwork and all its results in general. They argue that fieldwork is an extremely personal


136 Gunter Senft

experiment which cannot be reproduced and thus cannot claim having any kind of scientific status. Moreover, this subjectivity is said to leave the fieldworkers’ theory formation in a ‘mysterious twilight’ (see e.g., Kohl 1979). However, most of these critics simply ignore the fact that fieldwork is not just one method, but — as emphasized above — a complex of methods researchers use for data collection purposes. Although fieldwork — in any discipline and in any field — is certainly a deeply personal experience, many of the specific methods can now be reproduced, falsified or verified. Moreover, by describing the conditions and situations in which fieldworkers gathered the data on which they base their analyses, they have no problems to come up with the methodological standards of their respective discipline. Many of the usually polemic discussions between fieldworkers on the more practical, empirical side and theoreticians on the more theory-dominated side are useless and void. Good theory needs practical proof, and good empirical research, good fieldwork is impossible without good theory. By now, I assume, it is quite evident that fieldwork requires the researchers’ familiarity with a broad variety of data collection techniques (see, e.g., Foley 1991; Mayntz et al. 1976; Ruoff 1973; Samarin 1967). Moreover, fieldworkers have to be familiar with systems for transcribing data and especially with the various machines (and at least some of their technical intricacies) that can be used for data documentation (like various kinds of tape recorders, film- or video cameras, microphones, etc., see Goodwin 1993). However, the planning phases of much fieldwork also ask for the researchers’ abilities to interact with bureaucratic institutions and their representatives to obtain research permits. Every field trip should be planned as carefully as possible — and researchers who plan fieldwork in foreign countries must first inquire whether or not these countries allow research without a special permit. It goes without saying that wherever a research permit is required, it must be requested — although the processing of this request may be the first test for the researcher’s patience and perseverance. Most national linguistic, sociological, psychological, archeological, and anthropological associations can provide first useful information with respect to these questions. Contact with the respective researchers’ national embassies or consulates in countries where they plan to do fieldwork is also extremely helpful. If fieldwork is planned in countries that can become problematic for the researchers’ health — because of extreme climate zones or deseases like malaria — good medical information, all the necessary vaccinations, an excellent personal medicine kit for the field — especially for more remote areas — and some practical medical knowledge are a must (for basic information see e.g., Werner 1990). Finally, fieldwork does not only ask for the researchers’ physical fitness, it also requires a psychologically balanced and stable personality. The researchers’ psychological stability is first tested — especially when alone in a rather remote field — when they have to deal with possibly very different local concepts of behavior, hygiene, time,


privacy, etc. This initial phase of fieldwork is often described as ‘culture shock’ — and every field researcher will experience some variety of it. Overcoming this culture shock requires a basic willingness to understand, and to learn something about, the ‘other’; it requires patience, but also the researchers’ marking of their unequivocal position with respect to the question of how far they are willing to immerse themselves into the participant part of their observation. On the one hand, there is a danger that fieldworkers remain too distanced and detached in their field and establish a subject–object relation between themselves as the observers and the ‘others’, the observed in their field. On the other hand, if the researchers identify themselves too much with the community under investigation — maybe as a strategy to overcome their culture shock — they are in danger of giving up the distance which is necessary for any scientific description that claims to be objective. In its extreme, this identification with the community under investigation results in a situation that is called ‘going native’. A researcher ‘gone native’ is a kind of living oxymoron. As soon as researchers have gone native it is impossible for them to do any kind of sound research. But even back home from the field, the researchers’ psychology has to pass many trials, especially with respect to their ethical commitments to the communities with which they work. Here researchers are more often than not forced to make decisions that directly affect the personal rights of their consultants. It is difficult to suppress some extremely interesting bits of information that may provoke a vivid discussion in one’s academic peer group because their publication may violate one’s consultants’ trust in the confidentiality of the information. If in doubt, consider the other side of the experience. Just imagine a group of Papuas with penile sheaths, bows and arrows, and netbags, of course, coming to your village, town or suburb to learn something about your language and customs and document as much interpersonal interaction as possible with video- and tape recorders. How would you welcome these researchers from a different ‘world’? Would you care to talk to them and teach them your language? What would you tell these Papuan researchers? What would you allow them to tape-record and film? What would you tell them confidentially after a longer period of time? And would you like to read about this confidential information if you happen to find it printed in a scientific journal? To conclude, fieldwork covers a broad variety of scientific methods for data gathering. However, fieldwork is not only a methodological approach to the resolution of a problem of scientific interest, but also — and maybe first of all — an approach to our fellow human beings. For additional information about field research I would like to refer the interested reader to Agar (1980), Barley (1983), Boas (1911), Bohannan (1964), Bouquiaux & Thomas (Eds) (1976), Dixon (1984), Healey (Ed.) (1975), Howell (1990), Jahoda et al. (1933), Lehmann (1976), Malinowski (1967), Nida (1949), Powdermaker (1966), Rubinstein (Ed.) (1991), Sapir (1921), Wilkins (1992), Whyte (1943), and Wylie (1957).


138 Gunter Senft

References Agar, M. (1980). The professional stranger. Academic Press. Barley, N. (1983). The innocent anthropologist. Penguin. Boas, F. (1911). Introduction to ‘Handbook of American Indian languages’. Bulletin 40(1): 1–83. Bureau of American Ethnology. Bohannan, L. (1964). Return to laughter. Doubleday. Bouquiaux, L. & J.M.C. Thomas (Eds) (1976). Enquête et description des langues à tradition orale. SELAF. Dixon, R.M.W. (1984). Searching for aboriginal languages. University of Queensland Press. Duranti, A. (1981). The Samoan fono. Australian National University. Ellen, R.F. (Ed.) (1984). Ethnographic research. Academic Press. Foley, W.A. (1991). Field methods. In K. Malmkjaer (Ed.), The linguistics encyclopedia: 121–127. Routledge. Garfinkel, H. (1967). Studies in ethnomethodology. Prentice-Hall. Goldschmidt, W. (1976). Feldforschung. In W. Berndsdorf (Ed.), Wörterbuch der Soziologie: 226–229. Fischer. Goodwin, C. (1993). Recording human interaction in natural settings. Pragmatics 3(2): 181–209. Gumperz, J. & D. Hymes (Eds) (1972). Directions in sociolinguistics. Holt, Rinehart & Winston. Healey, A. (Ed.) (1975). Language learner’s field guide. Summer Institute of Linguistics. Heidelberger Forschungsprojekt (1977). Die Erlernung des Deutschen durch spanische und italienische Arbeiter. University Osnabrück. Howell, N. (1990). Surviving fieldwork. American Anthropological Association. Hymes, D. (1978). The ethnography of speaking. In J. Fishman (Ed.), Readings in the sociology of language: 99–138. Mouton. ——— (Ed.) (1964). Language in culture and society. Harper & Row. Jahoda, M., P.F. Lazarsfeld & H. Zeisel (1933). Die Arbeitslosen von Marienthal. S. Hirzel. Kohl, K-H. (1979). Exotik als Beruf. Heymann. Lehmann, A. (1976). Das Leben in einem Arbeiterdorf. Ferdinand Enke. Malinowski, B. (1922). Argonauts of the Western Pacific (1978 Ed.). Routledge & Kegan Paul. ——— (1935). Coral gardens and their magic, vol. 2 (1966 Ed.). Allen & Unwin. ——— (1967). A diary in the strict sense of the word. Athlone Press. Mayntz, R., K. Holm, P. Hübner (1976). Introduction to empirical sociology. Penguin. Nida, E.A. (1949). Morphology. The University of Michigan Press. Powdermaker, H. (1966). Stranger and friend. Norton. Rubinstein, R.A. (Ed.) (1991). The correspondence of Robert Redfield and Sol Tax. Westview Press. Ruoff, A. (1973). Grundlagen und Methoden der Untersuchung gesprochener Sprache. Niemeyer. Samarin, W. (1967). Field linguistics. Holt, Rinehart & Winston. Sapir, E. (1921). Language. Harcourt Brace. Senft, G. (1982). Sprachliche Varietät und Variation im Sprachverhalten Kaiserslauterer Metallarbeiter. Lang. ——— (1986). Kilivila, the language of the Trobriand Islanders. Mouton de Gruyter. ——— (1992). As time goes by…: Changes observed in the Trobriand Islanders’ culture and language, Milne Bay Province Papua New Guinea. In T. Dutton (Ed.), Culture change, language change: 67–89. Australian National University. Werner, D. (1990). Where there is no doctor. Macmillan.


Wilkins, D. (1992). Linguistic research under aboriginal control. Australian Journal of Linguistics 12: 171–200. Whyte, W.F. (1943) Street corner society. University of Chicago Press. Wylie, L. (1957). Village in the Vaucluse. Harvard University Press.


Firthian linguistics Jan-Ola Östman & Anne-Marie Simon-Vandenbergen University of Helsinki/University of Ghent

‘Firthian linguistics’ gets its name from John Rupert Firth (1890–1960), the main proponent of an approach to language which was developed at the School of Oriental and African Studies at London University. (Cf. Firth 1930, 1934, 1937, 1957a, 1957b, 1968). The views of this London group of linguists differed on a number of important points both from those held by the American post-Bloomfieldians and from those of the European Saussureans, and they became so influential in Britain in the 1940s and 1950s that the approach is also often referred to as the ‘British school’ of linguistics. In this overview we will only deal with issues of Firthian linguistics that are closely related to pragmatics. The foundation for Firthian linguistics was laid by Bronislaw Malinowski, whose seminal article The problem of meaning in primitive languages (1923) attempted to get to grips with the interaction between culture and meaning. As an anthropologist and ethnographer his concern was with discourse as it functions in a particular situation. His research on the language and culture of the Trobriand Islanders led him to the conclusion that one cannot understand the meaning of messages unless one takes into account the situation in which they are uttered. He therefore emphasized the importance of the context of situation, embedded within the total culture, and in his own descriptions he particularly focussed on features that were direct reflections or indications of that context of situation. At the same time Malinowski found that the function of language is frequently not primarily the transmission of ‘ideas’. Sometimes, the main purpose of talk is simply to establish or maintain a camaraderie relationship and to show affect. Malinowski is perhaps most famous for having developed the notion of ‘phatic communion’, a type of linguistic interaction which primarily aims at keeping the communication channel open.1 Phatic communion was also the instance par excellence of what Malinowski had in mind when he described language as a ‘mode of action’. Firth was influenced by Malinowski’s ideas on language as part of the social process, and developed a theory of meaning in which context plays a central role. For him, however, meaning cannot be handled at one level only, as it radiates throughout language;

1.  Many later discussions use the term ‘phatic communication’, which for Malinowski would have been a contradiction in terms.

Firthian linguistics

consequently, the analyst has to deal with different modes of meaning, at different levels. Meaning, in his view, is function within context, and the technique for describing it is a serial contextualization of forms. The principal components of this total meaning are the phonetic, the phonological, the morphological, the syntactic, the lexical, the collocational (i.e., the characteristic verbal company a word keeps), and the ‘semantic’ functions. By the latter he meant the contextualization of an utterance in the ‘context-of-situation’, which is itself embedded in the context of culture. Thus, for instance, the meaning of the English form bɔ:d has to be described at different levels. First, the sounds have meaning as selections from the total sound system and the combination is used as distinct from other possible forms such as bɔ:t, pɔ:t, etc. The form can then further be described on the basis of its ‘paradigm scatter’, for instance as a verb or a noun. Once it has been contextualized as the participle of a verb, it has morphological meaning. Syntactic meaning can further be assessed for instance by intonation (e.g., Bored?). Collocational meaning may be added by indicating that bored often goes together with stiff. Finally, if we contextualize the word in a particular social situation, in which speaker A asks speaker B Bored stiff?, the utterance receives what Firth calls a semantic function. Situations themselves take their meaning from the context of culture. Firth repeatedly emphasized, though, that meaning in this sense can either be dealt with in a ‘descending’ order (beginning within the context of situation) or in the opposite order (as illustrated above). Note also that for Firth it would not make sense to say that Are you has been ellipted from the beginning of Bored stiff?, since that would presume a view of meaning not affected by context. Although Firth borrowed the concept of context of situation from Malinowski as one level in his hierarchical system of meaning interpretation, there is one very important difference between the ways Malinowski and Firth saw context. Whereas Malinowski was interested in the actual existing features of context, Firth saw the context-of-situation as an abstract frame of reference, which the linguist invents. The linguist decides which features are going to be important for the analysis of language in (a particular) context. All in all, Firth was very much an adherent of what was at the time called the hocus pocus view of linguistic analysis: linguistic categories are constructs, imposed on language in the hope of getting a better understanding of what is going on. Such constructs do not exist independent of their creator. This implies that every notion or term in linguistics — including his own — are only analytical tools which can and should be changed whenever this is felt to be necessary. The interrelated factors that Firth felt could be abstracted out as important categories for a contextual frame of reference were –– –– –– –– ––

the relevant features of participants, personalities [roles], and persons the verbal actions of participants, personalities, and persons their non-verbal actions relevant objects the effect of the verbal action.


142 Jan-Ola Östman & Anne-Marie Simon-Vandenbergen

Firth also made practical use of this view of context in his work on teaching foreign languages for ‘restricted contexts’, e.g., for teaching Japanese phrases and war-terminology during the second world war. At the time, Bartlett (1932) had indeed suggested the importance of background knowledge, of what we today would call frames, schemata and scripts, for appropriate understanding. As a psychologist, Bartlett was of course particularly interested in memory, storage, and retrieval. This direction of research has more recently had a boom in cognitive approaches to language, and it has in fact become one of the cornerstones for much present-day work on cognition. It can well be argued, though, that Firth’s specification of a set of general categories for his context-of-situation was an early, comparable attempt to build up a schema and a metaschema at that — on a more socio-anthropological basis.2 Thus, his interest was the same as Bartlett’s, and the same as modern-day pragmatics: how to tie down context and make it operationally approachable. But whereas Bartlett and his cognitive descendants of today focus on the actual interrelation between linguistic and cognitive structures, Firth — by insisting on the construct nature of linguistic tools — was more interested in developing a system on the basis of linguistic manifestations. This is a neglected type of approach even in present-day studies in pragmatics. A good respresentative of the Firthian way of approaching context is Mitchell (1957), where the author analyzes The language of buying and selling in Cyrenaica, showing not only the importance of the features that Firth mentioned, but also in more detail, the different verbal correlations with the different stages of a commercial transaction: e.g., salutation, and bargaining. Firth’s approach to language was ‘monistic’. He openly rejected the then widely held view of language as expressive or communicative of inner mental states, which requires introspection as a tool. By looking upon meaning as the process of what happens when a form is used in context, Firth avoids the problem of having to deal with the duality of word and thought. By studying language as meaningful behavior, and in particular the social aspects of meaning, the linguist can stick to the observable, without having to subscribe to a behavioral view of language. Firth’s view also implies that he would have no theoretical reason to add a term like pragmatics to his toolbox. However, his insistence on the importance of embedding every utterance in a ‘context-of-situation’ would surely today render him a position in the center of pragmatics. So much so, in fact, that one could argue that the reason why Firth has been considered vague and abstract in his writings, both by his contemporaries and by subsequent theoreticians, is really that he was doing pragmatics. In his 1966 evaluation of Firth’s contribution to semantics, Lyons

2.  In a slightly different context, Firth (1957a: 190) talked about how “the constructs or schemata of linguistics enable us to handle […] language events”.

Firthian linguistics

criticizes Firth for using the term ‘meaning’ in what in Lyons’s opinion is a very idiosyncratic sense: Firth’s use of ‘meaning’ comes close to a notion of appropriateness or acceptability in context; and, furthermore, it misses a lot, since Firth does not directly address issues having to do with reference, sense, or significance. Firth, it seems to us, as he was further developing Malinowski’s (1923: 307) ‘ethnographer’s perspective’ that an “utterance has no meaning except in the context of situation”, was de facto taking the initial steps in a new field of linguistics, i.e., pragmatics. Bolinger (1968: 525–526) said about Firth’s view of language as it was presented in The tongues of men (1937) that it was “closer to the heart of sociolinguistic interests today than to any of the theoretical preoccupations between then and now”. At a time when mainstream linguistics focussed on language as a unified system to be studied in isolation from social factors, Firth advocated a linguistics embedded in society. He was avant-garde in his insistence on the need to develop methods for describing language variation, in his view on the language user as a social person who accumulates a bundle of social roles, in his insight that verbal interactions are rulegoverned and conditioned by a particular culture, and in his plea for investigation into the mechanisms of culture contact and culture transmission. In sum, for Firth the goal of linguistics was to study language as a form of meaningful human behavior in society. Firth’s approach to meaning can also be seen in his adherence to the dictum of ‘meaning implies choice’ — on all levels. That is, although his context-of-situation determined the ultimate meaning, there is no meaning of lower-level elements to be embedded in the context unless some other element could have been chosen. The notion of meaning-implies-choice was later systematized and formalized by M.A.K. Halliday (e.g., 1967–8) in his theory of systemics. Although this may seem a clear semantic issue, Östman (1986) has shown how the notion can fruitfully be applied to the implicit level of language, too, whence the choices that have to be made are among categories stipulated by a small set of pragmatic parameters. Thus, not only meaning, but also function implies choice. This is also in consonance with Firth’s view that in addition to the referential function of a word, one also has to take into account its effects, and the purposes and attitudes (towards the topic and the addressee) embedded in it. There are two terms in the literature that are used to designate linguists connected with Firth: ‘Firthians’ and ‘Neo-Firthians’. These labels are sometimes confusing. A Firthian would typically be one of Firth’s colleagues and students; and, it seems to us, anyone who subscribes to the kind of view Firth proclaimed with respect to linguistic constructs as being hocus-pocus constructs, and with respect to the importance of meaning as many-layered with context-of-situation as the ultimate determining factor. The term Neo-Firthian has been used in two senses, to refer to what we called Firthians above, and to refer to the various directions in which Firth’s work has been


144 Jan-Ola Östman & Anne-Marie Simon-Vandenbergen

developed — primarily in Great Britain and in Australia — since the 1960s. From a pragmatic perspective, the two most prominent of these approaches are the Birmingham School of discourse analysis, and, in particular, the approach to language called Systemic Functional linguistics as developed by Halliday. Halliday was a student of Firth’s and has on various occasions expressed his indebtedness to Firth.3 Systemic Functional linguistics, too, approaches language primarily as a multi-functional social phenomenon, and Halliday has further developed Firth’s notion of ‘system’: in Systemic Functional grammar the concept of system is the formal construct of a system network, in which each system is a set of interrelated options. However, since Halliday’s thinking about language has also to a very important extent been influenced by Hjelmslev, the Prague School, Malinowski and Whorf, the label ‘Neo-Firthian’ is somewhat misleading. Malinowski’s view that language in its structure mirrors the uses to which it is put,4 combined with Whorf ’s interest in grammatical categories as reflections of conceptual orientations in society have led to Halliday’s ‘functionalism’, in the sense that grammar is seen as naturally related to meaning, i.e., there is a link between the text, the linguistic system and the situation. However, Halliday differs from both Malinowski and Firth with respect to the concept of the context of situation; to Halliday (cf. e.g., 1978), the context of situation is not the material environment (no matter how abstract) but a semiotic structure. By treating both text and context as semiotic phenomena, or ‘modes of meaning’, Halliday manages to reveal interesting links between them. The Firthian tradition in linguistics has been dealt with extensively in Mitchell (1975) and Monaghan (1979).

References Bartlett, F.C. (1932). Remembering. Cambridge University Press. Bolinger, D. (1968). Aspects of language. Harcourt, Brace, Jovanovich. Firth, J.R. (1930). Speech. Benn. ——— (1934). Linguistics and the functional point of view. English Studies 16: 2–8. ——— (1937). The tongues of men. Watts.

3.  For instance, in his early 1961 article he specifically said that the model he is presenting (then known as the Scale and Category model) is directly based on Firth’s ideas, and an attempt to systematize these ideas. 4.  “The grammatical categories […] are the reflection of the […] outlook imposed by man’s struggle for existence […]” (Malinowski 1923: 328). However, Firth’s outspoken hocus-pocus view of the status of linguistic tools, constantly stressing the importance of ‘renewal of connection’ between models and linguistic behavior, makes him more cautious in his evaluation of the direct behavioral relevance of linguistic categories.

Firthian linguistics

——— (1957a). Papers in linguistics, 1934–1951. Oxford University Press. ——— (1957b). Studies in linguistic analysis. Blackwell. ——— (1964). The tongues of men & Speech. Oxford University Press. ——— (1968). Selected papers 1952–1959. Longman. Halliday, M.A.K. (1961). Categories of the theory of grammar. Word 17(3): 241–292. ——— (1967–8). Notes on transitivity and theme in English. Journal of Linguistics 3(1): 37–81, 3(2): 199–244, 4(2): 179–215. ——— (1978). Language as social semiotic. Edward Arnold. Lyons, J. (1966). Firth’s theory of ‘meaning’. In C.E. Bazell, J.C. Catford, M.A.K. Halliday & R.H. Robins (Eds), In memory of J.R. Firth: 288–302. Longman. Malinowski, B. (1923). The problem of meaning in primitive languages. Supplement I to C.K. Ogden & I.A. Richardsm, The Meaning of meaning (8th Edition, 1946/1989): 296–336. Harcourt, Brace, Jovanovich. Mitchell, T.F. (1957). The language of buying and selling in Cyrenaica. Hesperis 44(1–2): 31–71. ——— (1975). Principles of Firthian linguistics. Longman. Monaghan, J. (1979). The Neo-Firthian tradition and its contribution to general linguistics. Niemeyer. Östman, J.-O. (1986). Pragmatics as implicitness. UMI no. 86–24885.


Folk pragmatics Nancy Niedzielski & Dennis R. Preston Rice University/Okhoma State University

1.  Folk linguistics Folk Linguistics (FL) aims to discover and analyze beliefs about and attitudes towards language at every level of linguistic production, perception, and cognitive embedding by collecting and examining overt comment about it by non-linguists, goals first presented in Niedzielski and Preston (2000). The current re-emergence of interest in FL dates from a 1960 conference; Henry Hoenigswald (1966: 20) said that we should be interested not only in (a) what goes on (language), but also in (b) how people react to what goes on (they are persuaded, they are put off, etc.) and in (c) what people say goes on (talk concerning language). It will not do to dismiss these secondary and tertiary modes of conduct merely as sources of error.

Researchers in FL have employed a number of techniques in their attempts to elicit these “secondary and tertiary modes of conduct”. In general, respondents are presented with and/or encouraged to discuss or respond to areas of language concern that expose not merely their traditional, prepackaged notions, but also the processes that govern their thinking. From this perspective, folk belief about language is a dynamic process that allows non-specialists (i.e., persons with no formal training in linguistics) to express their understandings of their linguistic environment. The relation of FL to other approaches to the study of language is illustrated in Figure 1. The top of this triangle (a) characterizes Hoenigswald’s “what goes on”, and the a' above it is the area that represents the bulk of what most subfields of linguistics are concerned with: linguistic competence. The bottom line of the triangle (b1 − bn) represents a continuum of consciousness with regard to comments about and reactions to not only language use but also language topics in general. The rightmost side (bn) focuses on automatic processes, those largely outside conscious awareness, and has been the domain of the social psychology of language or language attitude studies, although not all studies in that tradition have been of such non-conscious responses. The leftmost side (b1) is the area of concern in FL and is made up of conscious, deliberative acts. Although the boundary between the conscious and automatic domains is not precise, and recent work in more cognitively oriented social psychology questions positing different background beliefs for the two (e.g., Bassili & Brown 2005), the distinction

Folk pragmatics a′ - cognitive states and processes which govern a What people say a

Conscious reactions to and comments on language



Unconscious reactions to language

b′ - cognitive states and processes which govern b Figure 1.  The position of folk linguistics and language attitudes in the general study of language

is still worth maintaining, particularly since work on language attitudes often finds a mismatch between implicit and explicit responses (e.g., Kristiansen 2006). Although FL typically studies more deliberative responses, it hopes to contribute to work in the social psychology of language by providing a more explanatory basis for attitudinal rankings based on underlying folk theories. Finally, just as there is an a', there is a b', the underlying beliefs and belief systems which lie behind folk expressions about language, and FL hopes to determine these, in addition to the overt responses that make up the observable data. In FL, however, the dichotomy between the conscious and unconscious does not fully characterize the levels and types of awareness of the linguistic units that may stimulate responses. The availability of different structural units of language has been outlined (e.g., Silverstein 1981), and, more recently, Preston (1996) has characterized both levels and modes of such awareness: topics can be unavailable (those the folk will not, perhaps cannot, comment on); available (discussed only if they are carefully described); suggestible (although seldom initiated by the folk, nevertheless commented on without elaborate description), and common (frequent public topics of folk linguistic discussion). It is also the case that the folk may or may not be accurate in their discussion of linguistic phenomena (although, as in all ethnographic studies, this has no bearing on the value of the data). In addition, the detail of folk awareness of an object may be either global or specific. For example, an accent may be described without detail (…they all talk funny…), while some linguistic phenomena are characterized with great specificity (… It’s like you’re trying to clear your throat while you’re speaking). Finally,


148 Nancy Niedzielski & Dennis R. Preston

the degree of control that a respondent possesses in the performance of a linguistic unit under discussion is variable and might not be correlated with any of the other types of awareness. A respondent who says someone ‘talks funny’ might not be able to mention any specific feature that contributes to that characterization, but might be able to give a convincing imitation. In sum, FL awareness is not only a matter of degree of consciousness, but one of mode or type as well. Even an inaccurate characterization of a linguistic fact or a poor mimicry of a variety demonstrates some kind of awareness, and for every act of language perception, the mode and degree of awareness is an open and interesting question. Furthermore, awareness of linguistic phenomena is governed by salience, regardless of the level of availability of a structural type or the kind of response elicited. Sibata (1971: 372) suggests that attention is drawn to language primarily when it differs from one’s own, although we suspect that notice is also triggered by a speaker using a form different from that which a hearer expects him or her to use. Whatever the triggering event, it must somehow move the respondent away from the normal, communicative practice of language, redirecting the focus to one on language itself. FL is perhaps most closely aligned with the ethnography of speaking, where work in a variety of contexts has led to an enriched understanding of linguistic behavior. FL also benefits the ethnography of speaking by providing richer detail about the folk ethnography surrounding language itself, data perhaps too often derived from the observation of performance than from the elicitation of opinion. It is clear, however, that FL subject matter has been valued in the ethnographic tradition. If the community’s own theory of linguistic repertoire and speech is considered (as it must be in any serious ethnographic account), matters become all the more complex and interesting. (Hymes 1972: 39)

FL research has also contributed to a and a' – the general linguistics peak of the triangle. Plichta (2004), for example, noted that folk comment on the nasal character of Northern, urban U.S. pronunciation was ignored by linguists on the basis of its unscientific character (e.g., Labov’s comment in Hoenigswald 1966: 23–24). After a careful acoustic investigation of two vowels involved in the Northern Cities Chain Shift, however, Plichta discovered that a nasal formant was, in fact, a feature that accompanied the repositioning of these vowels, a strong confirmation of the utility of folk comment even in matters of general and descriptive linguistics. Although it is not directly represented in the triangle of Figure 1, research on language variation and change also benefits from research into folk beliefs about linguistic phenomena. Since FL belief is an integral part of the ethnography of a speech community, any research that depends on an understanding of a community will need FL information as much as it needs any other demographic and/or linguistic characterization. It is difficult to imagine not wanting to know what members of a speech

Folk pragmatics

community believe about the linguistic phenomena that are under investigation in the study of variation and change. Finally, FL contributes to applied linguistics. An obvious application is to language learning issues, given that the folk express numerous concerns in their discussions of second language learning. It seems logical to suggest that what people believe about how they learn language, how difficult the target language is, and what particular talents they believe they have or lack in learning a second language, as well as what the social outcomes of learning are, will aid those designing curricula, training teachers, and writing textbooks in order to adapt to or at least reflect an understanding of the learner’s conception of the language learning process (e.g., Pasquale & Preston 2006). 2.  Folk pragmatics If we appreciate the value of FL in general, it ought to be straightforward to suggest that it is worthwhile doing at every linguistic level and from every research point of view. That is, just as there is folk phonology, syntax, and semantics, there is folk pragmatics. Just as there is folk psycholinguistics, there is folk sociolinguistics. In Figure 1, therefore, pragmatic performance (which for us includes the performance of the hearer as well as the speaker) is as much a part of a as any other level of linguistic performance, and pragmatic competence is just as surely embedded in a'. If one takes a broad view of pragmatics, one that includes conversation and text organization as well as various communicative practices such as politeness and deference strategies, as we do, then there is little doubt that the domains of these areas in a' are various and complex and made even more so by the recognition that pragmatic competence includes receptive as well as productive abilities. What might be taken, therefore, as language regard, the term we prefer for the range of responses to language (the entire b1 to bn continuum that is the base of Figure 1), is all too easily confused with linguistic responses, things we believe are a part of a. For example, when a speaker insults a hearer, the hearer may be angry, upset, annoyed, aghast, disheartened, disappointed – all emotional characteristics that might seem very similar to the affective sorts of responses one associates with the right handside of the base of Figure 1. We would argue, however, that such an act and its uptake do not belong to the area of FL or even language attitudes. The hearer has used his or her pragmatic competence to understand whatever the speaker has said as an insult (whether direct or indirect, whether correctly interpreted or not). The emotional response to being insulted is part of the hearer’s usual emotional response system, one not tied to a focus on language. We are well aware of the fact, however, that language regard factors (whether from the attitudinal or FL side of the b continuum) may interact with pragmatic strategies


150 Nancy Niedzielski & Dennis R. Preston

in such a way as to redirect them, but we do not believe that this redirection is unique to pragmatic reinterpretation (e.g., Niedzielski 1999). In the above insult scenario, for example, let’s suppose that the insult was made in a voice that the hearer recognized as Upper Class Southern British. Putting aside how the hearer’s language attitudes might have been formed, let’s also assume that he or she has an enormously negative opinion of such speech and that the insult is, therefore, taken as extremely harsh and the resentment of it is considerably heightened. This, it seems to us, is a clear example of a language regard mechanism interacting with the pragmatic interpreter, at least if one agrees that degree of insult is a pragmatic concern; we certainly do, and, in fact, we could also describe cases in which the language regard factor is so strong that an insult is ameliorated enough to make it no longer taken as one but understood as joking, teasing, or even an act of solidarity. We also do not concern ourselves at length here with the question of the status of language features themselves in language regard situations. We are well aware of the fact that language performances that trigger regard responses, whether conscious or unconscious, may be clues to the identity of the speaker and that that identity, rather than the language itself, constitutes the actual attitude object. Edwards (1982: 20) summarizes this correlation for many social psychologists when he notes that “people’s reactions to language varieties reveal much of their perception of the speakers of these varieties”. There is evidence, however, that long-standing linguistic performances may awaken regard characteristics without attachment to the groups or individuals who characteristically use them. In Milroy and McClenaghan (1977), for example, respondents who could and could not accurately identify the home region of the speaker had the same affective responses to a variety of English dialects from Great Britain and Ireland. It has been widely assumed that an accent acts as a cue identifying a speaker’s group membership. Perhaps this identification takes place below the level of conscious awareness.… Presumably by hearing similar accents very frequently [one] has learnt to associate them with their reference groups. In other words, accents with which people are familiar may directly [italics in original] evoke stereotyped responses without the listener first consciously assigning the speaker to a particular reference group. (Milroy & McClenaghan 1977: 8–9)

We leave behind, therefore, the question of whether language regard responses are responses to a language performance or towards a speaker identity awakened by a linguistic fact. Our concern is with the left-hand side of the b continuum of Figure 1. Real FL regard responses are ones in which language as language is the topic. When pragmatic concerns in language are the topic, then folk pragmatics has come into play. We also distinguish, therefore, between folk pragmatics as a subcategory of FL and what Mey (1993: Chapter 13) calls metapragmatics. He cites You did a great job, and I’m

Folk pragmatics

not being polite as containing, in the latter half of the sentence, a metapragmatic statement (270). We distinguish, however, between what a sentence is about and what it refers to. The sentence Mey cites refers to linguistic politeness, but it is not about it. The speaker does not discuss, reflect on, elaborate, or even describe how linguistic politeness works or is constituted. We doubt if even the following extension illustrates a slip into folk pragmatics: You did a great job, and I’m not being polite, ‘cause I’m not the kind of person who goes around saying things I don’t mean. Consider this segment of conversation (with back-channels, pauses, and false starts removed for clarity) between folk respondent and a Taiwanese fieldworker who feels embarrassed about asking whether a sibling is older or younger, although his cultural background makes it important for him to know this information. The segment is clearly about politeness.   H = Taiwanese linguistic fieldworker   G = Middle-age male European American Upper Midwestern respondent

H: But – so here sometime I’m confused when people say “Oh … Angela” … She’s my – sister. Then I’m thinking is sister is younger or is the older G: Don’t be afraid to ask. Ask. And nobody’s offended. And they’ll say “Oh … oldest or eldest” or they’ll say “Youngest,” yeah. Don’t feel afraid. It’s a good way to start a conversation. Really. “Do you have (a) brother or sister?” You know, you ask, you know “older or younger, elder or younger.” H: I mean … this is not impolite, right? G: No, no, it’s not improper.

This is not only a conversation about the politeness of conversational topics but even contains advice about its usefulness as a conversation starter. It is thus an example of folk pragmatics, rather than metapragmatics in Mey’s sense. On the other hand, many pragmaticists (e.g., Silverstein 1993) would clearly include what we call folk pragmatics within the larger category of metapragmatics. 3.  Directions for future research Though pragmatics as such did not constitute a section of focus in Folk Linguistics (Niedzielski & Preston 2000), when we review the data for that work (and data collected for it but not included in it, like the above segment), we are struck by the number of concerns discussed by our respondents that could have been so classified. There is a need to delve more carefully into this area. In conclusion, we have taken the mention of folk belief at face value from discoursal and quantitative surveys and tried to catalog them by topic and investigate them for what they reveal about the kinds of beliefs the folk hold about language. We believe that an interesting foray could be made into such data with an aim towards collecting folk pragmatic comment as well.



Nancy Niedzielski & Dennis R. Preston

One of our aims in doing FL, however, has been to suggest that a general folk theory of language is often the accompaniment to any expression of folk belief. Although we do not claim to have the final word on such a theory, we believe a great deal of our quantitative and discoursal data from the U.S. suggests that an underlying folk theory of language stands in stark contrast to the one held by most professional linguists. Figure 2 shows this relationship. Linguists believe that language is a concrete instantiation in the mind/brain space of individuals; it most certainly has social effects, but its home is within the individual. Linguists abstract from this concreteness to such notions as dialect and language, but anyone who has ever struggled with the distinction between the two will agree that the question of which closely related varieties are dialects and which are languages is resolved by political, geographical, and even religious norms, not by structural ones. Not so the folk, at least not the U.S. folk we have investigated. For them, the language has a cognitively exterior reality, although they know that individuals possess something language-like, but that possession may either (responsibly) reflect the

THE LANGUAGE A Folk Theory of Language Good Language

Ordinary Language


A “Linguistic” Theory of Language

Dialect #1


Dialect #3

Dialect #2

Idiolect #1

Idiolect #2

Figure 2.  Folk and linguistic theories of language



Folk pragmatics

abstraction that is the real language or it may (irresponsibly) diverge farther and farther from it until it reaches the ultimate linguistic irresponsibility of dialect or simply error. We believe that the search for such underlying theories helps us understand a great deal of folk involvement in public questions about language (teaching, learning, evaluating others’ behavior, even judging their intellect) and that, even if FL did not shed light on questions that more often engage language professionals, the important revelations it may make about language and public life would make its pursuit worthwhile. We doubt if this pursuit will be different in the search for folk pragmatic data, but folk theories of pragmatics may hinge on a different construct than that of an ideal or correct external reality. In a recent discoursal investigation of Japanese attitudes towards the English and Japanese languages, Imai (2000) shows how a careful analysis of discoursal structure helps in understanding speaker beliefs about and attitudes towards language that are clearly pragmatic in nature. She triggered the conversation she analyzed by asking, simply, “What do you think about the differences and similarities between Japanese and English?” She characterized the portion of the conversation (between two respondents) she analyzed as an argument (following Schiffrin 1985; Preston 1993, 1994), a discourse genre which consists, essentially, of positions, disputes, and supports. One important part of Imai’s investigation lies in her discovery that a young female respondent, Y, bases her comments on language on a theory of what might be called ‘social use’. Briefly, she does not regard aspects of a language to be viable parts of it unless they are used in ordinary conversation. Evidence for this interesting folk theory comes mostly from this respondent’s support moves rather than from any positions she takes in the argument. Her interlocutor, for example, asserts that English has more words than Japanese, but Y disputes that position and supports her dispute with the claim that “Americans don’t use difficult words”. When her interlocutor asks if it isn’t the case that books can be linguistically difficult, Y notes that she is concerned only with conversation. Later she also notes that she does not consider phone calls from salespersons to be authentic language either, since scripted calls are also filled with difficult and incomprehensible language, the sort she has not encountered in face-to-face interaction. What is most interesting, however, is Imai’s eventual interpretation of Y’s folk theory when it turns evaluative. Y’s continuing support for the notion that authentic language is based on conversational usage leads her to be critical of what she sees as an American insensitivity to demands for flexibility in language use.

Y: This is not about the words and probably it is because of the national traits, but, well, I don’t know how to say this, but sometimes if I said something and they didn’t understand, they say they don’t understand, right? And if they say “say it again”, a Japanese would change the words or make it simpler= S: ((laughter))


154 Nancy Niedzielski & Dennis R. Preston

Y: =We try to make it simpler and explain, don’t we? Americans repeat exactly the same thing. All: ((laughter)) Y: They are not very flexible, you know?

Imai suggests that Y’s theory of good language includes sensitivity to the needs of the interlocutor, and she clearly finds Americans lacking in this respect. This is an extremely interesting notion to us since, as shown in Folk Linguistics, the prescriptive notion attached to language among U.S. respondents nearly always hinges on what might be called schoolroom correctness. At least Imai’s respondent Y suggests that Japanese respondents may base evaluative notions of language more in the area of speaker and hearer rights and responsibilities. If that is so, it may even prove to be the case that the underlying representation of language itself for Japanese speakers is not the idealized, cognitively external code held to be the essence of language by U.S. respondents, and its relevance to pragmatic concerns is obvious. Perhaps it is not surprising that the two examples of folk pragmatics we have cited here come from cross-cultural settings. We would like to paraphrase Sibata’s notion of why language calls attention to itself (i.e., when the speaker’s system differs from our own) for the pragmatic domain by suggesting that pragmatic concerns are foregrounded or become a potential target for FL in cases of pragmatic failure. Such cases are clearly not restricted to cross-cultural interactions, and we believe such a rule of thumb may be a good way to begin this investigation, one that we hope will bear fruit in the near future.

References Bassili, J.N. & R.D. Brown (2005). Implicit and explicit attitudes: Research, challenges, theory. In D. Albarracín, B.T. Johnson & M.P. Zanna (Eds), The handbook of attitudes: 543–574. Lawrence Erlbaum. Edwards, J.R. (1982). Language attitudes and their implications among English speakers. In E.B. Ryan & H. Giles (Eds), Attitudes towards language variation: 20–33. Arnold. Hoenigswald, H. (1966). A proposal for the study of folk-linguistics. In W. Bright (Ed.), Sociolinguistics: 16–26. Mouton. Hymes, D. (1972). Models of the interaction of language and social life. In J.J. Gumperz & D. Hymes (Eds), Directions in sociolinguistics: The ethnography of communication: 35–71. Holt, Rinehart and Winston. Imai, T. (2000). Folk linguistics and conversational argument. Paper presented at New Ways of Analyzing Variation (NWAV). East Lansing, MI, October. Kristiansen, T. (2006). Social meanings and subjective processes: A presentation of theories and methods from the Næstved studies. Paper presented at Approaches to the Study of Folk Linguistics, Sociolinguistic Awareness and Language Attitudes. Centre for Research on Bilingualism, Stockholm University. Mey, J. (1993). Pragmatics. Blackwell.

Folk pragmatics

Milroy, L. & P. Mcclenaghan (1977). Stereotyped reactions to four educated accents in Ulster. Belfast Working Papers in Language and Linguistics 2: 1–11. Niedzielski, N. (1999). The effects of social information on the perception of sociolinguistic variables. Journal of Language and Social Psychology 18(1): 62–85. Niedzielski, N. & D.R. Preston (2000). Folk linguistics. Mouton de Gruyter. Pasquale, M. & D.R. Preston (2006). The folk linguistics of language teaching and learning. Paper presented at the 2006 Second Language Research Forum (SLRF). University of Washington, Seattle. Plichta, B. (2004). Interdisciplinary perspectives on the Northern Cities Chain Shift. Unpublished Ph.D. dissertation. Michigan State University. Preston, D.R. (1993). The uses of folk linguistics. International Journal of Applied Linguistics 3(2): 181–259. ——— (1994). Content-oriented discourse analysis and folk linguistics. Language Sciences 16(2): 285–330. ——— (1996). Whaddayaknow?: The modes of folk linguistic awareness. Language Awareness 5(1): 40–74. Schiffrin, D. (1985). Everyday argument: the organization of diversity in talk. In T.A. van Dijk (Ed.), Handbook of Discourse Analysis 3: 35–46. Academic Press. Sibata, T. (1971). Kotoba no kihan ishiki. Gengo Seikatsu 236: 14–21. [English quotations and page references are taken from this article translated as ‘Consciousness of language norms’ in T. Kunihiro, F. Inoue & D. Long (Eds) (1999), Takesi Sibata: Sociolinguistics in Japanese contexts: 371–377. Mouton de Gruyter.] Silverstein, M. (1981). The limits of awareness. Sociolinguistic Working Paper No. 84. Southwest Educational Development Laboratory. Silverstein, M. (1993). Metapagmatic discourse and metapragmatic function. In J. Lucy (Ed.), Reflexive language: 3–58. Cambridge.


Honorifics Judith T. Irvine University of Michigan, Ann Arbor

As forms of social deixis engaging linguistic structure with social function, linguistic honorifics offer a rich field of investigation for students of pragmatics. Recent developments in research have deepened the theoretical basis of our understanding of honorifics. They have also shown that honorific phenomena are more widespread in the world’s languages, more various in form, and more creatively deployed, than was once supposed. 1.  Definitions Linguistic honorifics are forms of speech that signal social deference, through conventionalized understandings of some aspect(s) of the form-meaning relationship. One must specify linguistic honorifics since deference may be conveyed in ways other than in speaking and writing, although those other modalities are beyond the scope of the present discussion. Further, one must specify that honorifics are forms of speech: that is, the deference is understood (by the members of some community of discourse) as an inherent property of particular linguistic expressions, not just the consequence of an expression’s deployment in social life. Finally, the signalling of deference is conventional in that it is not merely the literal product of an expression’s compositional semantics, but requires some additional set of shared understandings. All languages provide the means for speakers to express respect for other persons — even if only in expressions like I respect you — but not all languages have the specially conventionalized forms we call honorifics. Potentially, conventions for the expression of deference might involve any level of linguistic organization. For the most part, however, the term ‘honorifics’ is applied only to special lexicon or morphology. Those forms, when examined individually, may be termed honorific items, or honorific alternants, since they alternate with nonhonorific forms which, apart from the expression of deference, supposedly ‘mean the same thing’ (see example (2) below, and discussion in Section 3). The honorific items’ cooccurrence with other such items, their engagement with other linguistic phenomena such as a characteristic phonology or syntax, and their possible involvement in special discourse structures longer than the sentence, give rise to what may be termed honorific registers (see Agha 1993). Although a register for conveying respectfulness might be constructed on some other basis, a language lacking specifically honorific lexicon or morphology is seldom said to ‘have honorifics’.


It is important to distinguish honorifics (or, honorification), a convention of linguistic form, from politeness, a communicative effect brought about by a person’s comportment in a particular social interaction.1 To put this another way, honorification belongs to sentence-meaning; politeness belongs to utterance-meaning. Only by making this distinction can one account for (e.g.) sarcastic uses of honorifics, uses whose purpose may be quite the opposite of polite (Levinson 1983: 93). Although some signal of deference is necessarily ‘in play’ when an honorific form is deployed, that signal can be framed in ways that displace it from the speaker’s personal comportment, and make it not actually count as deference in the interaction at hand. Other concepts related to honorifics include language levels, a term applied to cases such as Javanese (see example (2) below), where registers are relatively discrete — at least, their speakers so conceive of them — and form a graded series along a scale of respectfulness. Honorific pronouns, and their associated person-marking verbal expressions, are often called formal, although that term presents difficulties (see Irvine 1979). Some languages include a system of pejoratives — forms that (by convention) are derogatory — while other languages have no conventional pejorative system other than the derogation implied by the absence of honorifics. It has been suggested (Levinson 1988) that the existence of pejorative systems might depend on the presence of honorifics, although this has not been shown, since our information on the distribution and forms of pejoratives is too fragmentary. And while some honorific systems make extensive use of tropes of self-pejoration, many do not. 2.  Dimensions of variation and comparison Comparing sociolinguistic systems worldwide suggests (at least) several dimensions along which honorifics vary. These dimensions of variation will be illustrated with examples. 2.1  Linguistic forms What kinds of linguistic phenomena are recruited as the formal means for honorification? Items such as honorific titles and terms of address are found in languages having no other conventionally honorific forms. Apart from titles, some sociolinguistic systems recruit only personal pronouns (and any verb forms, etc., marked for person); other systems have repertoires, large or small, of lexical alternants; still other systems have honorific particles or affixes; and some systems present combinations of these types. Probably the most familiar examples of languages featuring honorific pronouns are European languages such as French and Russian, where second-person plural 1.  Agha (1993) makes a similar distinction; but what I have called ‘honorification’ he terms deference. In discussions of social organization, however, ‘deference’ so often refers to aspects of comportment and their social consequences that it may be misleading here, where comportment must be distinguished from semantics.


158 Judith T. Irvine

pronouns function as honorific singulars, and languages such as German and Spanish, whose honorific second-person pronouns derive from third-person forms. Many languages around the world have similar phenomena (for a survey, see Head 1978). Here is an example from Amharic (Ethiopia), in which second-person singular male pronouns show four degrees of respect, and the most honorific second-person forms are derived from third-person stems:

(1) Amharic personal pronouns (Hoben 1976): Person



Insulting Familiar Intermediate Honorific 1st 2nd, Male anci 2nd, Female 3rd, Male 3rd, Female

ïne antε anci ïrsu ïrswa

ïňňa (royal) ïrswo ïrswo ïrsaccεw ïrsaccεw

antu antu

ïňňa ïnnantε ïnnantε ïnnεrsu ïnnεrsu

Other languages show more extensive systems of lexical alternants, which may or may not include pronouns. Probably the best-known (to linguists) of such cases is Javanese, where sets of graded lexical alternants define ‘levels’ of respect to an addressee. Locally known as krama (respectful speech), madya (middle), and ngoko (basic, unrefined speech), these levels of speech style admit further gradations. As the sentence in (2) shows, the combinations of graded alternants define as many as six levels, even though most of the lexical sets, considered individually, actually have fewer alternants:2

(2) Javanese ‘language levels’ (Errington 1988): KRAMA:

1. menapa 2. menapa

nandalem mundhut sekul semanten panjenengan mendhet sekul semanten


3. napa 4. napa

sampeyan sampeyan

mendhet njupuk


5. apa 6. apa

sliramu kowe

mundhut sega njupuk sega

semono semono


Question-marker you


that much

sekul semonten sega semonten


Translation Did you take that much rice?

Many other languages have respect vocabularies that distinguish honorific from ‘ordinary’ lexical items, but lack Javanese’s elaborate system of degrees of honorification. 2.  Note that Errington (1988) does not call the language levels ‘honorifics’, but instead reserves that term for lexical items expressing respect for a referent rather than an addressee — unlike some other scholars of Javanese,who use‘honorifics’for both.Although a distinction between reference and address forms is important, cross-linguistic comparisons of ‘honorifics’ are better served by including both.


Examples would include Samoan (Duranti 1992) and Zulu/Xhosa (Herbert 1990b; Irvine 1992). Some languages construct honorific expressions morphologically, relying on a system of honorific affixes. Nahuatl, for example, uses special verbal affixes (the prefix on-, the reflexive mo-, and the ‘reverential’ tzīnōa) as well as an honorific suffix -tzīn for nominals, to create several levels of honorification (Hill & Hill 1978). The following expression, for example (based on the verb stem pīalīa, ‘have’), shows four levels:

(3) Nahuatl verbal affixes of respect (Hill & Hill 1978: 128):

ticpia ticompia ticonmopīalīa ticonmopīalihtzīnōa

‘you have it’ (familiar) ‘you have it’ (somewhat honorific) ‘you have it’ (honorific) ‘you have it’ (reverential)

Many African languages having noun classification systems, especially the central and eastern Bantu languages, recruit the classificatory affixes to distinguish degrees of honorification. ChiBemba, for example, a Bantu language of Zambia, has some 20 morphologically distinct noun classes, some of which have fairly consistent semantic or derivational functions, and some not. Many stems can be assigned to more than one class, thus shifting the obligatory class-marking prefix and effecting some semantic or grammatical difference corresponding to the shift, such as a change in the referent’s number or size. For nouns with human reference, ChiBemba shows both honorific and pejorative forms by shifting the noun-class assignments. In example (4), as illustrated for the stem -kaši ‘wife’, the plural-human class (class 2) serves as honorific singular; the ordinary singular (class 1) becomes disrespectful, and the classes otherwise marking small size (class 12), large size (class 7), and singularity (class 5) become pejorative:

(4) ChiBemba honorific and pejorative prefixes (singular referent; based on the stem -kaši, ‘wife’):3 Noun


Class Value

abakaši umukaši akakaši ičikaši ilikaši

‘(respectable) wife’ ‘wife’ ‘(insignificant) wife’ ‘(gross) wife’ ‘(egregious [?]) wife’

2 1 12 7 5

honorific disrespectful insult insult ‘a little derogatory’

While the pejoration conveyed by the size classes may be just a consequence of their semantics, the (pragmatic) honorification conveyed by class 2 overrides its (semantic)

3.  ChiBemba examples come from interviews with ChiBemba speakers at Brandeis University.


160 Judith T. Irvine

plurality. In (5), plural prefixes (ba- and aba-, class 2) are used with a referent explicitly specified as singular, to express respect for an older sibling: (5)

ChiBemba honorifics (pragmatics, not semantics, determines prefix selection): bamo abakalamba baandi baleelya isabi 2-one 2-older sibling 2-my 2(subject)-tense-eat 9a-fish ‘My one older sibling is eating fish’ (or, ‘One of my older siblings is eating fish’)

Finally, a particularly complex system of honorifics can be found in Japanese. In addition to honorific suffixes for terms of address, and pronoun alternants, Japanese has honorific affixes attaching to other parts of speech; moreover, the honorific morphology of verbs intersects with other aspects of the verb’s morphology and syntax. For example, a ‘politeness’ affix -mas- intersects with tense and negation:

(6) Japanese honorific verb morphology (Inoue 1979): Neutral



taberu tabeta tabenai tabenakatta

tabemasu tabemashita tabemasen tabemasen deshita

eat ate does not eat did not eat

nonpast, affirmative past, affirmative nonpast, negative past, negative

Syntactic complexities arise when the ‘respectful’ prefix o- (go- for Sino-Japanese words) combines with verbal compounds, some of which conventionally signal selflowering or other-raising. For instance, compounds formed with o- … -ni nar (literally ‘become’) are understood as other-raising, while compounds formed with o- … su (literally ‘do’) are understood as self-lowering:

(7) Japanese honorific verbal compounds, based on the stem kak-, ‘write’ (Inoue 1979; hyphens divide morphemes):

kai-ta kaki-mashi-ta o-kaki-ni nari-mashi-ta o-kaki shi-mashi-ta

‘wrote’ ‘wrote’ (‘polite’) ‘wrote’ (other-raising) ‘wrote’ (self-lowering)

The construction of these honorific compounds further depends on the syntactic location — subject or object — of an honored NP (Harada 1976; Shibatani 1990). Although the Japanese system is evidently one of the world’s most elaborate, it is not clear where the boundary lies between specially ‘honorific’ forms, on the one hand, and (on the other) expressions of broader interactional strategies such as indirectness, which might contribute to polite comportment but are not formally conventionalized in the same way as honorifics are (Conventionalization, in any sociolinguistic system, is probably always a matter of degree; and whether alternants ‘mean the same thing’ involves language ideology as well as denotation).


2.2  Sources of honorific expressions What is the basis on which honorific expressions are constructed? There seem to be two main possibilities: tropes (or circumlocutions; that is, expressions constructed from elements occurring in nonhonorific registers, and accorded conventionally honorific meanings) and linguistic borrowing. As we have already seen in ChiBemba, in a great many languages an honorific register poaches on nonhonorific expressions of plurality and/or size, to create honorific tropes. Thus (to take a well-known example) the French pronoun vous, though literally plural, is also used as respectful, or socially distant, singular. Some other languages take pluralization much further. In Moré (Burkina Faso), for instance, the more pronouns one pluralizes, the more honorific is one’s speech; so, while pluralizing only second-person pronouns is ‘polite’, pluralizing all other pronouns too is ‘honorific’ (Lehr et al. 1966). What tropes a particular language employs in honorific registers depends, presumably, on the particular system of cultural ideas about deference and sociability. Hill & Hill (1978) suggest that Nahuatl is probably unusual in that its honorific suffix -tzīn may also mark diminutive or affectionate meaning, depending on the context.4 Perhaps in such cases it is the affection, rather than the smallness, that motivates the trope. But one should not suppose that honorific tropes reveal their own motivation, without demanding any further investigation. Agha (1993) warns against taking honorific tropes at face value and constructing a picture of speakers’ world-view on that basis. The caution comes from his study of Lhasa Tibetan, where many honorific items are formed as lexical compounds whose ‘pragmatic operator’ — the lexeme signalling honorification — denotes a body part, even though the compound does not. Thus the honorific alternant for ‘price’, for example, is formed by attaching to the nonhonorific alternant for ‘price’ an honorific stem for ‘tongue’. Does this mean that Tibetans perceive the universe in anatomical terms? No, Agha argues; instead, the honorific construction links the referent to the honoree’s person, since (as Tibetan consultants pointed out) one honors persons, not inanimate objects or abstractions (Agha 1993: 144–145). That honorifics concern the honoree’s person is a common theme in honorific tropes, and it accounts for why body parts, corporeal activities, and personal possessions are so frequently the focus of honorific vocabulary. In another frequentlyfound theme, honorific tropes are euphemistic, disengaging the respected person from

4.  Although tropes involving diminutives do seem to be less frequent as honorifics than their opposite (tropes involving plurality or augmentatives), the Nahuatl case is not unique. Some honorific items in Zulu, for example, also use diminutives, such as umFazazana, literally ‘little woman’, as a respectful alternant for iNgungumbane, ‘porcupine’. The porcupine is shown respect, lest she damage the speaker’s garden (Bryant 1949: 220).


162 Judith T. Irvine

unpleasantness and from the concrete, mundane, messy details of everyday life. Both themes seem to be at work in example (8), from Shilluk ‘royal language’, which relies on tropes — some transparent, some obscure — from ordinary language to construct honorific ways of referring to the person, personal activities, and personal possessions of the king and certain other high-status beings:5

(8) Shilluk ‘royal’ language (predicated of the king; Pumphrey 1937):

kwai amak mith yi Jwok

ORDINARY USAGE ‘go out to graze’ ‘come to anchor’ ‘be nursed by God’

HONORIFIC USAGE ‘sleep’ ‘sit on the ground’ ‘be [locative]’

Another kind of source for honorific items is linguistic borrowing. In Javanese, for example, some of the honorific lexical alternants, as well as some affixes attaching to a stem and rendering it honorific, derive, ultimately, from Sanskrit (and the era of Hindu-Buddhist suzerainty in Indonesia). In Persian, many honorific expressions derive from Arabic, the language of religious texts (Beeman 1986); while in Hindi many derive from Persian, the language of Mughal conquerors of North India. In these instances, then, expressions borrowed from a language of high prestige have brought that prestige along with them into the honorific register. A different kind of borrowing seems to have taken place in the Nguni languages (Zulu, Xhosa, SeSwati) of Southern Africa. What was borrowed was not lexicon but phonology; and the source languages were probably not considered prestigious. In the respect vocabularies of these languages, click consonants are especially frequent, but clicks were not originally part of the Nguni sound repertoire. Instead, they were borrowed from the Khoisan languages, indigenous to southern Africa at the time the Nguni languages arrived there. The respect vocabularies, called hlonipha, were apparently the clicks’ point of entry (Herbert 1990b; Irvine 1992). In recent years largely restricted to women in conservative rural communities, the hlonipha vocabulary is a means of avoiding uttering the names of particular kinsmen or (in the past) the names of kings and chiefs — persons to whom special respect must be shown. Since the names were themselves meaningful expressions, the practice of hlonipha meant avoiding uttering any words that were included in those names (or close homophones of such words). To substitute a click consonant for one or more of the consonants occurring in the to-be-avoided name-word would have ensured that

5.  The ‘unpleasantness’ theme will not account for all examples, however, at least not without more information about this little-described sociolinguistic system. Why should the Shilluk honorific register recruit the ordinary-language word for ‘pebble’ to refer honorifically to the king’s head, and refer honorifically to the king’s pipe with the ordinary-language word for ‘mud’?


one could not possibly be uttering the name itself, since the clicks were (in the early years of the process) quintessentially ‘foreign’ sounds. And to the extent that Ngunispeakers regarded Khoisan speakers as low-status, the use of these Khoisan sounds would have suggested self-lowering on the Nguni-speaker’s part. Some examples of hlonipha words are illustrated in (9). (Note that clicks were not the only means of masking the relevant consonants; another process seems to have involved shifting non-coronal consonants to [+Coronal]. Moreover, some hlonipha words were constructed by circumlocutions — i.e., on a semantic, rather than a phonological, basis.)

(9) Zulu hlonipha vocabulary (Doke & Vilakazi 1958):

‘graze; weave’ ‘be dejected’ ‘affair’ ‘my father’ ‘hippopotamus’ ‘our’ ‘chief ’ ‘annoy’ ‘swing’



aluka jaba indaɓa uɓaɓa imvuɓu -ithu inkosi nenga lenga

acuka gxaba injušo utšatša incuɓu -itšu inqoɓo, inqotšana (dim.) cenga cenga

c, q, x = clicks (gx = voiced click)

ɓ = implosive bilabial stop

Sparse though the relevant historical information may be, the Nguni material has nevertheless afforded some opportunity to explore the historical dynamics of an honorific system. (Herbert 1990a has some relevant information on acquisition also, though of click sounds rather than of the hlonipha system, now rarely used). In the literature on honorifics such discussions are rare. Still, Errington (1988: 115) points to a process in Javanese that may be much more widespread: an inflationary pressure, in which honorific alternants tend to lose honorific value as a result of recurring patterns of strategic use. Irvine (1992) notes that Zulu (among other cases) parallels Javanese in this regard, since some Zulu hlonipha words may have lost their specifically respectful value and entered the everyday vocabulary. 2.3  Honorification roles and participant roles How is honorification distributed over the structure of participant roles (speaker, addressee, referent, etc.) in an act of speaking? Honorification implies a role structure, relating the giver of deference to its recipient. With Agha (1993) we may term the giver of deference the origo of honorification, and the being toward whom deference is directed the focus of honorification. This role structure must be mapped onto the


164 Judith T. Irvine

social situation in which the act of speaking occurs. Usually the origo is the speaker, who shows deference to someone else; but other mappings are possible, such as when a speaker temporarily takes on someone else’s point of view (performing the deference someone else owes; see Agha 1993: 135; Errington 1988: 160; Inoue 1979: 287).6 As for the focus of honorification, sociolinguistic systems differ as to how it is normally mapped onto a speech situation. In ChiBemba and Samoan, for example, the focus is the referent of the honorific expression; in some other systems it is the addressee. Javanese provides ways of honoring both, through different subsystems: while the ‘language levels’ (speech styles) are addressee-focused, there is also a set of referent-focused alternants, deployed independently of the speech styles (Errington 1988). Of course, honorific second-person pronouns, in any of these systems, honor addressee and referent simultaneously. A few honorific systems focus on still other participant roles. In the Australian ‘mother-in-law’ registers, the focus is neither addressee nor referent, but a bystander: these registers are to be used by a male speaker when he finds himself in the presence of particular in-laws, who may not be addressed directly anyway. Similarly, among the Mossi, special forms of speech are used in the presence of a chief, a member of the chief ’s family, or some other government official, whether the high-ranking person is addressed or explicitly referred to or not, and even when the person actually spoken to would otherwise receive ordinary, familiar speech (Lehr et al. 1966: 5). Related, perhaps, to bystander honorifics are cases where the honorific focus is further away (remote focus). The Nguni respect vocabularies mentioned earlier are examples, since one must not utter the name-sounds of the honoree regardless of one’s company at the moment of speaking. Another kind of remote-focus honorific register is illustrated in western Africa in the Gbe languages (Capo 1979) and Bamum (Delafosse 1922), where chiefs and kings sometimes dictated sets of lexical substitutions that were to be employed by their followers as a display of respect and loyalty. Although the speech register thus established was mainly characteristic of court life, a leader’s adherents also used his lexicon elsewhere, as a sign of loyalty, an emblem of party affiliation, and a means of secrecy. Finally, usages such as the royal ‘we’ suggest that it is also possible for the speaker to be the honorific focus. More often, however, respect accrues to the speaker of honorifics not explicitly (as honorific focus) but indirectly, as a byproduct of usage. A Javanese speaker who uses the krama inggil ‘high krama’ deferential speech style not only signals deference to an addressee, but also displays his/her own refinement through this delicacy of comportment, sensitivity to others, and knowledge of special linguistic

6.  As Agha (1993) points out, some apparent ‘violations’ of norms or consistency in honorific usage can be explained on this basis.


forms. Krama speech is supposed to be governed by an ethic of self-control, emotional restraint, orderliness, and concern for the feelings and wishes of others; supposedly, it is a style of self-effacement. Yet, the fact that it is also implicitly a display of refinement and knowledgeability informs speakers’ strategies of register use. These strategies depend, in part, on the supposition that knowledge of Javanese ‘language levels’ is not evenly distributed across the speaker population. As Geertz (1960) observed, only the traditional elite (the priyayi) then controlled the higher levels. Peasants made do with lower sets of styles. The use of high krama necessarily indexes membership in the elite, if only they know the forms well enough to use them. Presumably, many other sociolinguistic systems have similar properties: if the honorific items are many and their sources obscure, then their use is a mark of erudition. In such cases the implicit pragmatic effect of deploying honorific expressions will be selfraising, no matter how (explicitly) other-raising, or self-lowering, the honorific forms are supposed to be. (See Irvine 1992 for discussion of explicit vs. implicit pragmatic effects of honorifics, in different sociolinguistic systems.) Thus the relationship of honorifics to participant roles in an act of speaking is complex. Sociolinguistic systems vary as to how origo and focus — deferrer and honoree — map onto participant roles. They also vary as to whether there is an additional, implicit mapping — honoring the incumbent of some participant role other than the (explicit) honorific focus — that is necessarily entailed in the use of their honorifics. 2.4  Deployment and ideology What is the distribution of honorific usages over speakers and social situations — who uses honorifics, and when? What are the kinds of communities and social relationships in which honorific systems occur, and what are the ideologies of language that inform honorific usage and its activation? As the wide range of languages having honorifics indicates, there can be no simple relationship between linguistic honorifics and societal type. Nevertheless, it has sometimes been supposed that honorifics would most likely be found in complex, hierarchical societies, perhaps triggered by the social dynamics of royal courts (see, e.g., Wenger 1982: 161ff.). Within Europe, for example, Brown & Gilman’s (1960) influential paper on pronouns linked honorific pronouns with institutionalized power relationships, starting (they suggested) in the late years of the Roman empire: in the fourth century, the emperor came to be addressed in the plural, because there were then actually two emperors; this pattern persisted and spread, crystallizing in the courts and class stratification of late medieval and early modern Europe; and the pattern filtered down into everyday and domestic settings, as a kind of echo or metaphor of court power relations. Subsequent changes in social structure and ideology have shifted the class distribution of linguistic usages.


166 Judith T. Irvine

This picture of honorifics’ evolution is vulnerable to several objections. First, cases such as the Australian bystander honorifics indicate that linguistic honorifics are not limited to complex or stratified societies. They can occur in egalitarian contexts as well, as an expression of respect and social distance. Moreover, several subsequent studies of European pronouns have criticized Brown & Gilman’s specific historical assertions, pointing out (among other things) that their paper’s notion of ‘power’ was oversimplified, and that sociological trends in early modern Europe did not follow the course that had been claimed (see Agha 1994 for discussion). Of course, a Roman Empire source cannot account for honorifics outside Europe anyway. Finally, an argument deriving honorifics from court life cannot account for crosslinguistic variation in the ways honorifics are distributed over a larger population and over other settings. In the Nguni languages, for example, court usage of the hlonipha respect vocabulary was probably subsequent to domestic usage, and its distribution has always been linked to considerations of gender and of genre. Its use today is restricted to domestic talk among conservative rural Xhosa women; meanwhile, a separate system of eulogistic talk — izibongo praise-poetry — is associated with men and with courts and public settings (see Irvine 1992). No such strong connection with gender has been shown for Javanese. There, as we have already noted, the use of honorifics and language levels among the traditional elite differs from the usage of peasants (and of a ‘modern’ elite; Errington 1992). Yet, participants rank these usages along a single scale, envisioning a society-wide scheme of linguistic ‘refinement’, despite differences in usage resulting from differences in speakers’ social circumstances. Meanwhile, in Europe itself, the distribution of ‘formal’ pronouns across classes, genders, and settings is complicated and variable. Ideologies concerning a concept of ‘respect’, and its relation to those classes, genders, etc., have also been variable, and even mutually opposed. This discussion shows that honorifics, as sociolinguistic patterns, are not just a reflection of macrosociological structures. Instead, like other deictic forms, honorifics are indices pointing to aspects of an interactional arena. Speakers’ comportment in that arena is informed by their understanding of the social forces that impinge upon them, the social positions to which they may aspire, and their interpretation of the relevance of any of these to the situation at hand (see Section 4 for further discussion.) Worldwide, then, honorific systems inhabit many different kinds of societies and participate in different kinds of social dynamics. Although all honorifics involve some notion of ‘respect’, many questions still remain concerning what ‘respect’ (and respectful comportment) entails in particular sociocultural systems. Other questions must concern the nature and extent of the community of discourse in which a particular honorific system is found, and the homogeneity (or differentiation) of that community’s language ideologies. As a final problem concerning the deployment of honorifics and the ideology/ideologies associated with them, let us consider whether ethnographic cases may contrast at the level of norms about norms, as it were: the respect for respectfulness itself.


For example, a fine-grained interactional analysis of the Samoan respect vocabulary and its contexts of use (Duranti 1992) reveals that these items can be used to coerce ‘respectful’ behavior in one’s interlocutors. To deploy respect vocabulary is to presuppose — perhaps creatively, and therefore, in a sense, coercively (see Silverstein 1976) — that the situation demands ‘respect for tradition’ from everyone. What the honorifics index, then, is not so much the speaker’s deferential attitude toward a particular addressee or referent, but the speaker’s invoking of respect for tradition, the tradition of which honorifics are a part. Notice, however, that although the Samoan honorifics can be used in novel ways, they would not be coercive if speakers could not expect a high level of compliance with the norms they invoke. In some other sociolinguistic system, such norms might not be shared; or they might be opposed by some counternorm, to the effect that some situations, or some social identities, call for norm-breaking behavior or a display of disrespect. Might some such contrast in ideologies of respectful comportment coincide with a contrast between, say, preindustrial and ‘modern’ societies? No such supposition is warranted. Normative displays of pejoration occur in many kinds of societies, while analyses of honorific usage in contemporary Japan suggest a high level of ideological homogeneity and adherence to norms of respectful comportment. (However, the literature on Japanese honorifics is weighted toward the investigation of beliefs about usage; this is not the same thing as investigating the data of usage itself, as Agha 1994: 295 points out.) 3.  Pragmatics and semantics Linguistic honorifics offer interesting perspectives on relationships between pragmatics and semantics. On the one hand, as we have already said, honorific alternants supposedly ‘mean the same’ as their corresponding nonhonorific forms. That is, a certain referent may be denoted either by an honorific or a nonhonorific expression, the difference lying in the expressions’ pragmatic value and not in their denotational adequacy. Indeed, this semantic relationship is probably a necessary condition for identifying forms as ‘linguistic honorifics’ at all. On the other hand, honorific expressions do not necessarily have the same denotational scope as their nonhonorific alternants. That is, they do not necessarily organize a semantic domain at the same level of detail. Moreover, other complexities arise if honorifics are constructed via tropes, drawing on expressions from a nonhonorific register. Examples such as (5), from ChiBemba, show how the pragmatic value of conventional honorific tropes can override the expression’s ‘literal’ semantics. Yet, the ‘literal’ ingredients of a trope need not disappear without trace, even when the trope is conventionalized. So it is not really accurate to say that the semantics of expressions in the honorific register is simply ‘the same’ as in the nonhonorific register, varying ‘only’ in context of


168 Judith T. Irvine

use. Instead, there is a complex relationship between the two, with the pragmatic value imposing a second-order semantics. Some of the best-explored cases of semantic relationships between registers come from the Australian ‘mother-in-law’ respect registers (see, e.g., Dixon 1971; Haviland 1979). In a kind of variation on the theme of disengagement from the concrete, the honorific register is semantically impoverished or ‘bleached’, in that its lexicon makes fewer denotational distinctions than nonhonorific registers do. Instead, the honorific register makes do with a single, thus relatively abstract, lexeme to denote referents for which the nonhonorific register has several lexemes. For Dyirbal, a language of North Queensland, Dixon (1971: 437) estimates that the mother-in-law register (called Dyalŋuy) has only about a quarter as many words as the everyday register (called Guwal). Some examples of semantic bleaching in Dyalŋuy are shown in (10). Where the actual sound-shapes used in the Dyalŋuy respect register came from is not known, although some may have been borrowings recruited from the everyday speech of some other tribe (Dixon 1971: 439). (10) Semantic ‘bleaching’ in Dyirbal (Dixon 1971):

GUWAL dyambun bugulum mandidya gidya gaban nudin gunban baygun dyindan banyin

Gloss ‘long wood grub’ ‘small round bark grub’ ‘milky pine grub’ ‘candlenut tree grub’ ‘acacia tree grub’ ‘cut deeply, sever’ ‘cut less deeply, cut out’ ‘vigorously shake, bash’ ‘gently wave or bash’ ‘bash a log against a tree’

DYALŋUY dyamuy ” ” ” ” dyalŋgan ” bubaman ” ”

Agha (1988, 1993) notes a similar process of ‘denotational fading’ in the semantics of honorifics in Tibetan. It is not likely, however, that all honorific systems, worldwide, share this particular type of semantic relationship with their nonhonorific alternants. The Shilluk examples, like some others employing circumlocutions, seem instead to show a kind of elaboration. Still, whether bleaching or elaborating, honorific and nonhonorific registers are not exact denotational equivalents, even if their speakers so conceive of them. For Dixon (1971), it is the nonequivalence of these registers that makes them especially interesting, since the one implies an indigenous semantic analysis of the other. 4.  Trends in theory and research Although some of the phenomena discussed in this paper have long been mentioned in descriptive grammars, it is probably fair to say that honorifics and language levels

Honorifics 169

only began to capture wider attention in the early 1960’s (but see Gonda 1948). Brown & Gilman’s (1960) paper on European pronouns, Geertz’s (1960) description of Javanese ‘linguistic etiquette’, and Martin’s (1964) discussion of speech levels in Japan and Korea — juxtaposed, in Hymes’ (1964) influential anthology, with Newman’s (1955) paper on Zuñi sacred and slang registers — are among the prominent works of this period. The basic structure of honorific items and registers, their complex relationships along axes of reference and address, and their embedding in sociocultural systems, began to emerge. Brown & Gilman’s study has been especially widely read, and their discussion of power and solidarity, as dimensions of social relationships having a bearing upon language use, has been echoed in many sociolinguistic studies whether concerned with honorifics or not. Also important was their observation that the ‘power’ dimension predicted asymmetrical exchanges (a speaker gives T pronouns but receives V pronouns, e.g.), while the ‘solidarity’ dimension predicts symmetrical exchanges. Moreover, insofar as sociolinguistic patterns are part of social life, usages are affected by the historical forces impinging upon social structures and ideologies (especially, in the case of honorific pronouns, ideologies about status and community). Much of the research inspired by these early works falls into two types. One type of study concentrates on the structure and/or ideology of an honorific system, as these might emerge in linguistic elicitation sessions. These studies pay relatively little attention to the complexities of the system’s deployment, or to variability among users’ ideologies and social agendas. The other type of study focuses on the distribution of usages over speakers and situations. In earlier decades especially, distributional studies often took a correlationist perspective, viewing patterns of language use as a direct reflection of macrosociological structures. Tending to equate the distribution of honorifics (and other linguistic forms) with their meaning, reflectionist studies do not allow much consideration of variation in speakers’ viewpoints and competencies, of what strategic agendas are possible, or of how meanings and social relationships are creatively negotiated. Critiques of reflectionist approaches are abundant (see Irvine 1985 and Agha 1994 for relevant literature and discussion). A work that significantly shifted the theoretical ground for research on honorifics and related phenomena is Brown & Levinson’s widely-cited essay on language and politeness (1978, revised and reissued in book form 1987). Brown & Levinson emphasized an interactional approach, focusing on the strategies an individual might pursue, following a calculus of rational choice based upon considerations Brown & Levinson deemed to be universal: Gricean maxims of conversation, and a concern with the maintenance of (or threat to) ‘face’ — one’s public self-image, and one’s personal freedom of action. In this model, politeness consists in avoidance or mitigation of threats to face. Honorifics are said to represent ‘frozen’ outputs of politeness strategies, outputs that are “direct grammatical encodings of relative social status between participants” (1987: 23, 276).

170 Judith T. Irvine

While it is an important advantage to center the analysis of deictic forms (including honorifics) in interaction, and Brown and Levinson’s essay offers a plethora of interesting hypotheses, their approach has been criticized on several grounds, including the weaknesses of rational-actor models in accounting for social forms, and the authors’ exclusion of cultural conventions and conceptual differences from theoretical consideration, with concomitant difficulty in accounting for these (see Agha 1994 for literature and discussion.) For honorifics in particular, Agha (1994: 288) points out that if honorifics grammatically encode relative social status, it is difficult to explain speakers’ creativity in using them to alter relationships and achieve special effects (as, e.g., documented in Duranti’s 1992 Samoan research). Moreover, while honorifics obviously have something to do with deference, their deployment is not predictable simply as politeness, for reasons already mentioned. Anthropologically-oriented researchers, most notably Agha and Errington, tend to favor an approach that accords with Silverstein’s (1976, 1979) semiotic perspective. Bringing a Peircean (and Jakobsonian) semiotics into linguistic, social, and cultural consideration, Silverstein has emphasized the complex indexicality of deictic forms and their potential creativity in social interaction. That is, linguistic forms such as honorifics are multifunctional (and multiply indexical); the conditions of their production are complex, as are their potential interpretations. Their pragmatic effects are subject to social construction, and cannot be taken as predictable a priori. While this approach grounds honorifics (as deictic forms) in social interaction, it interposes a mediating framework of culturally-variable metapragmatics and ideologies of language.7 Research on honorifics — as a topic in its own right, potentially independent of research on politeness — has been on the upswing in recent years, as evidenced in a 1988 NSF-sponsored conference on honorifics, in the appearance of book-length treatments (e.g., Errington 1988), and in Agha’s (1994) literature review. Studies of particular honorific systems are now beginning to include observation of honorific usage in natural conversation (e.g., Duranti 1992; Hill & Hill 1986; Agha 1993). Other current concerns include the relationship of honorific registers to other kinds of registers and other languages (Hill & Hill 1978, 1986);8 historical dynamics of honorific systems (Herbert 1990b); and comparative studies (Irvine 1992). Meanwhile, other

7.  See, especially, Silverstein’s (1979) discussion of Javanese language levels in his paper on language ideology, and papers by Silverstein, Errington, Hill, and Irvine in Kroskrity, Schieffelin, and Woolard (Eds) (1992). 8.  Obviously, studies of honorifics usually compare honorific with nonhonorific registers. But they usually do so along a single scale of honorification, rather than looking at a more complex register repertoire as a system — let alone considering how register usage relates to code-switching in multilingual settings.


interesting topics such as the acquisition of honorifics, or their relations with genre, have scarcely been broached.

References Agha, A. (1988). Honorific register and systems of deference in Lhasa Tibetan. Rev. paper, annual meeting of the American Anthropological Association. ——— (1993). Grammatical and indexical convention in honorific discourse. Journal of Linguistic Anthropology 3: 131–63. ——— (1994). Honorification. Annual Review of Anthropology 23: 277–302. Beeman, W.O. (1986). Language, status, and power in Iran. Indiana University Press. Brown, P. & S.C. Levinson (1978). Universals in language usage. In E. Goody (Ed.), Questions and politeness: 56–311. Cambridge University Press. ——— (1987). Politeness. Cambridge University Press. Brown, R. & A. Gilman (1960). The pronouns of power and solidarity. In T. Sebeok (Ed.), Style in language: 253–276. MIT Press. Bryant, A.T. (1949). The Zulu people. Shuter & Shooter. Capo, H.C. (1979). Notes on language differentiation. Anthropological Linguistics 21: 419–42. Delafosse, M. (1922). Naissance et évolution d’un système d’écriture de création contemporaire. Revue d’Ethnographie et des Traditions Populaires 8: 11–19. Dixon, R.W.M. (1971). A method of semantic description. In D. Steinberg & L. Jakobovits (Eds), Semantics: 436–471. Cambridge University Press. Doke, C. & B.W. Vilakazi (1958). Zulu-English dictionary (2nd Ed). Witwatersrand University Press. Duranti, A. (1992). Language in context and language as context. In A. Duranti & C. Goodwin (Eds), Rethinking context: 77–100. Cambridge University Press. Errington, J.J. (1988). Structure and style in Javanese. University of Pennsylvania Press. ——— (1992). On the ideology of Indonesian language development. In P. Kroskrity, B. Schieffelin & K. Woolard (Eds): 417–426. Geertz, C. (1960). The religion of Java. University of Chicago Press. Gonda, J. (1948). The Javanese vocabulary of courtesy. Lingua 1: 333–76. Harada, S.I. (1976). Honorifics. In M. Shibatani (Ed.), Syntax and Semantics, vol. 5: 499–561. Academic Press. Haviland, J.B. (1979). How to talk to your brother-in-law in Guugu Yimidhirr. In T. Shopen (Ed.), Languages and their speakers: 161–239. Winthrop. Head, B. (1978). Respect degrees in pronominal reference. In J. Greenberg (Ed.), Universals of human language 3: 151–212. Stanford University Press. Herbert, R.K. (1990a). The relative markedness of click sounds. Anthropological Linguistics 32: 120–138. ——— (1990b). The sociohistory of clicks in Southern Bantu. Anthropological Linguistics 32: 295–315. Hill, J. & K. Hill (1978). Honorific usage in modern Nahuatl. Language 54: 123–155. ——— (1986). Speaking Mexicano. University of Arizona Press. Hoben, S.J. (1976). The meaning of the second-person pronouns in Amharic. In M.L. Bender, J.D. Bowen, R.L. Cooper & C.A. Ferguson (Eds), Language in Ethiopia: 281–288. Oxford University Press. Hymes, D.H. (Ed.) (1964). Language in culture and society. Harper & Row.



Judith T. Irvine Inoue, K. (1979). Japanese: A story of language and people. In T. Shopen (Ed.), Languages and their speakers: 241–300. Winthrop. Irvine, J.T. (1979). Formality and informality in communicative events. American Anthropologist 81: 773–790. ——— (1985). Status and style in language. Annual Review of Anthropology 14: 557–581. ——— (1992). Ideologies of honorific language. In P. Kroskrity, B. Schieffelin & K. Woolard (Eds): 251–262. Kroskrity, P., B. Schieffelin & K. Woolard (1992). Language ideologies. Special issue, Pragmatics 2/3. Lehr, M., J. Redden & A. Balima (1966). Moré: Basic course. Foreign Service Inst. Levinson, S.C. (1983). Pragmatics. Cambridge University Press. ——— (1988). Some putative universals of honorific systems. Paper, Honorifics Conference, Portland OR. Martin, S. (1964). Speech levels in Japan and Korea. In D. Hymes (Ed.): 407–415. Newman, S. (1955). Vocabulary levels: Zuñi sacred and slang usage. Southwestern Journal of Anthropology 11: 345–54. Repr. (1964) in Hymes (Ed.): 397–406. Pumphrey, M.E.C. (1937). Shilluk ‘royal’ language conventions. Sudan Notes and Records 20: 319–321. Shibatani, M. (1990). The languages of Japan. Cambridge University Press. Silverstein, M. (1976). Shifters, linguistic categories, and cultural description. In K. Basso & H. Selby (Eds), Meaning in anthropology: 11–56. University of New Mexico Press. ——— (1979). Language structure and linguistic ideology. In R. Clyne, W. Hanks, & C. Hofbauer (Eds), The elements: 193–247. Chicago Linguistic Society. Wenger, J.R. (1982). Some universals of honorific language with special reference to Japanese. Ph.D. Diss., University of Arizona.

Wilhelm von Humboldt Brigitte Nerlich & David D. Clarke University of Nottingham

1.  Life Wilhelm von Humboldt (1767–1835) was the brother of the explorer and anthropologist Alexander von Humboldt, who later on in life supplied Humboldt with specimen of exotic languages. Humboldt studied at the universities of Frankfurt-on-Oder and Göttingen. After 1789 he lived for periods in Erfurt and Weimar where he knew Goethe and Schiller. In 1794 Humboldt moved to Jena with his family. One reason for this move was that Humboldt and his family could join up again with Schiller and his family. In fact, Jena was at that time Germany’s literary and philosophical capital. In 1794 Fichte, taking over from Reinhold took up a professorship there, to be succeeded a few years later, in 1798, by Schelling who in turn was followed, in 1801, by Hegel. On the literary side one could meet early Romantics, such as the Schlegel brothers and Novalis. But it was also possible to leave this heady atmosphere behind and plunge into concrete natural science, such as dissecting corpses and studying anatomy, an opportunity Humboldt did not miss. These friendships form the backdrop for Humboldt’s first attempts at studying language which were published just before the turn of the century. After this phase of creative writing, he moved to Berlin where he had some employment at the Supreme Court of Justice. Between 1798 and 1799 he lived in Paris, after that in Spain, and from 1801 to 1808 he held the post of Prussian Resident Minister in Rome. These and other visits abroad permanently influenced his thinking in the fields of politics, history and philosophy. His most effective practical work was undertaken as the Director of Education and Instruction in the Prussian Ministry of the Interior. It should however be emphasised that, despite assertions to the contrary (cf. Dove 1881), Humboldt was never Minister of Education, as, in any case, this office did not exist in Prussia at that time. During his time in office (1808–1810) Humboldt founded the Friedrich Wilhelm University in Berlin, now the Humboldt University. In this connection he formulated the German university ideal of the unity of research and teaching. Humboldt also reorganised the Neuhumanistische Gymnasium (senior secondary schools with a stress on the classics) in Prussia. Between 1810 and 1819 he served the Prussian government in several different offices. He was, for example, Prussian Envoy to Austria, represented Prussia at the Vienna Congress in 1815 after Napoleon’s defeat at Waterloo, was Prussian Envoy to London, and finally Minister of

174 Brigitte Nerlich & David D. Clarke

Social and Communal Affairs. In 1819 he resigned from government service and from then until his death he devoted himself to writing in the seclusion of the Humboldt mansion at Berlin Tegel. During that time he wrote his most important works on the philosophy of language. (On Humboldt’s life, read Haym 1856; Dove 1881; Leitzmann 1919; Knoll & Siebert 1967; Sweet 1978–80; Manchester 1985; Seuren 1998). 2.  Humboldt’s philosophy of language Humboldt’s thinking on language is rooted in the German philosophical tradition which begins with Leibniz. Like Humboldt, Leibniz was one of those German thinkers who combined interests in science with interests in philosophy and linguistics. Both Leibniz and Humboldt not only speculated on language but also did concrete empirical research. In Leibniz’s famous reply to Locke’s Essay concerning Human Understanding (1689), entitled Nouveaux Essais sur l’entendement humain (written in 1703–05, but only published in 1765), Leibniz had entered into an imaginary dialogue with his English colleague. It appears that Leibniz and Locke agreed upon the following points: 1. Language is an instrument of society, and emerges because human beings are social animals. 2. Words have two functions: recording our own thoughts and communicating them to others. 3. Language must work with ‘ease and quickness’ (cf. Aarsleff 1982: 63). However, they disagreed on the following points: 1. For Locke words were arbitrary labels of ideas, whereas for Leibniz they were somehow natural; they showed the essence of things, their origin. 2. Locke rejected (just as Condillac after him) the doctrine of innate ideas inherited from Descartes, whereas Leibniz accepted it to some extent. One can see in Humboldt’s writing a continuation of Leibniz’s criticism of Locke with Kantian and Fichtian means. Although he never discussed Locke’s theory of language to any greater extent, it is clear that Humboldt must have known it thoroughly, and this not only through Leibniz. At the time of Humboldt’s first appearance on the philosophical scene, thinkers like Hamann and Herder (to name just two) were attacking Kant’s Critique of Pure Reason with weapons borrowed from British empiricist philosophers like Francis Bacon and Locke, and Humboldt certainly knew ‘his’ Herder (cf. Herder 1772, 1799). He also knew the works of the French philosophers who based their theories of language on an interpretation of Locke. As we have seen, Humboldt even went to France in 1798 in order to gain an understanding of them. The philosophy

Wilhelm von Humboldt

of language that Humboldt developed between 1795 and 1836 laid the foundations for a ‘pragmatic’ theory of language, in which the notion of the ‘act of speech’ is central. The articulatory act of speaking is the hub of Humboldt’s philosophy of language, just as the semiotic act had been for Locke and the act of reason for Kant: “[l]anguage proper is based on the act of its real production” (Humboldt 1963 [1830–35]: 418–419; cf. Nerlich 1995). 3.  Language and thought Humboldt first expressed his new insights into the nature of language in his sixteen theses on Thinking and Speaking, written in 1795/96 under the direct influence of Fichte’s treatise on the origin of language (cf. Humboldt GS, VII: 581–583; cf. Fichte, Von der Sprachfähigkeit und dem Ursprung der Sprache, 1795). Although still clinging, to some extent, to a ‘pre-Humboldtian’ sign theory, where signs represent thoughts (cf. Schmitter 1987: 80–81), Humboldt’s later, more revolutionary, philosophy of language already shines through. For Humboldt, language (or as he would later say ergon, language as form) is the product of two ordering processes combined in every act of speaking (energeia, or language as formation): these two ordering processes are reflection (a term inherited from Herder) and articulation.1 It should be stressed that this essay ‘on thinking and speaking’ was written very early on in Humboldt’s life, well before his so called linguistic period (1820 onwards) properly speaking had started (cf. Schmitter 1992: 299–301). However, as early as 1795 Humboldt had already mapped out the field of study he was to explore philosophically and empirically later on: the relationship between language, thought, and culture. The essay takes the form of sixteen very brief theses, of which we shall reproduce the first six in Manchaster’s translation: 1. The essence of thinking consists in reflecting, i.e., in distinguishing the thinking from that which is thought about. 2. In order to reflect, the spirit must in its progressive activity stand still for a blink, to gather what was just represented into a unity, and in this manner to posit it as an object against itself.

1.  In his Essai sur les langues du nouveau continent of 1811–12 Humboldt stressed: “C’est l’idée de son articulé qui renferme tout ce qu’il y a de grand et de mystérieux dans les langues. Ce n’est point ici l’endroit de développer cette idée, mais tout raisonnement métaphysique dans le langage doit partir de là. C’est en suivant cette route qu’on reconnoit véritablement que la parole devient tellement l’intermédiaire entre l’homme et l’univers, que c’est elle qui le crée devant ses yeux, et le rend capable en même temps lui-même de concevoir et de sentir son ouvrage.” (GS, III: 322)


176 Brigitte Nerlich & David D. Clarke

3. These units, of which it can form several in this manner, it in turn compares with each other, and separates and combines them according to its need. 4. The essence of thinking consists, therefore, in making sub-divisions in its own course; thereby to form totalities out of certain portions of its own activity; and these formations singly among themselves, but all together as objects, to posit against the thinking subject. 5. No thinking, not even the purest, can take place but with the aid of the universal forms of our sensuality; only in them can we comprehend it and, as it were, hold fast. 6. Now, the sensuous signification of units, into which certain portions of thinking are united, in order, as parts to be posited against other parts of a bigger totality, as objects to the subject, is called in the broadest understanding of the word, language. (GS VII, 581–583; transl. Manchester 1985: 35) To understand these theses, one has to grasp the significance of the following important relationships: (1) between thinking and reflection; and (2) between subject and object. Humboldt had inherited the concept of the first relationship from Kant and Herder, the concept of the second one from German idealism in general, particularly Kant and Fichte. Following Kant, August Wilhelm Schlegel and Schelling introduced the concept of transcendental subjectivity into the philosophy of language, and they conceived it as an active force mediating between the merely empirical subject and object. As Formigari writes in her article ‘De l’idéalisme dans les théories du langage’ (Formigari 1988), this introduction of a new concept had important consequences: the linguistic mediation, which 18th-century empiricist philosophers used to consider essential to the mind’s activity, is now conceived as the work of a mind that is in itself active or ‘spontaneous’. Humboldt’s philosophy, in turn, attempted to reconcile the spontaneity of the subject with the conditioning role of language. Through reflection, that is, literally, bending over the stream of unorganised thought, the mind breaks the flux of thinking up into different units. These thought units are concepts. But the mind can only divide and cut out these concepts by using signs. The concepts are thereby fixed, and fixed in a sensual, embodied form, that is, in the form of articulated sounds. This means that sounds or words are not mere signs of thought in the sense of mere representations of thoughts. Just as the word could not exist without the concept, the concept could not arise, let alone be retained and fixed without the word. In a way, the sign is constituent of thought, in as much as thought in the form of reflection is constituent of the sign. To use an image introduced by Saussure, the reflecting mind cuts and divides at one and the same time the mass of thought and the mass of sounds, it ‘articulates’ both. After this cutting up of both masses, the sound and the concept cannot be considered in isolation anymore — they are like two sides of a sheet of paper. The process of chopping and dividing is called by

Wilhelm von Humboldt

Humboldt, as well as later on by Saussure, ‘articulation’. In their theories this concept does not only refer to the vocal articulation of sound, but to the dividing up of the stream of thought in the formation of concepts and the creation of linguistic units (what Saussure calls ‘articuli’).

4.  Language and world With this first act of articulation (and reflection), which is at the same time the act of language creation, human beings emerge from the numbness and darkness of pure instinct and desire. They awaken to self-consciousness, because the articulated word is an exteriorised part of themselves, but a part that they can regard, now that it is exteriorised, as an object. This object is not a ‘thing’ in the ordinary sense of the word, but the result of a free (inter-subjective) act. It is a unit of a concept and a sound. In the ongoing process of articulation, triggered by this first act of reflection, and by the combination and comparison of the units thus created, the mind gradually creates a language, a certain form of language. Language as a whole is thus the ultimate form of object that the subjects posit outside themselves. In fact, language, through its intricate link with thought and its influence on thought, is the ultimate form of objectivity. Once created, it tinges the perception of every real ‘object’ in the world. All other objects of thought and culture are formed and shaped by it. But not only that: the subject itself, that is, the ultimate form of subjectivity, is shaped by this form of language as well. This point is made clearer in Humboldt’s ‘Introduction to the Kawi Language’, published posthumously in 1836: In speech the energy of the mind breaks a path through the lips, but its product returns through our own ears. The idea is [thus] translated into true objectivity without being withdrawn from subjectivity. Only language can do this […] [Moreover,] just as the particular sound mediates between the object and man, so the whole language mediates between him and the nature that works upon him from within and without. He surrounds himself with a world of sounds in order to assimilate the world of objects. (GS VII, 55f) (transl. in: Kretzmann 1968: 392)

This objectivity, the world, created by the intricate collaboration between language and thought, has therefore always a subjective tinge. It is not that there is here a world and there a language that represents it — they are each constitutive of each other and this through the activity of the subject. This is the reason why, according to Humboldt, every language provides a different world-view or conception of the world. Each world is constituted relative to its language and the users of that language, a relativity thesis that is quite a bit more complex and complicated than the


178 Brigitte Nerlich & David D. Clarke

one popularised not so much by Sapir and Whorf themselves (cf. e.g., Lee, 1996), but by their more indiscriminate followers.

5.  Language and languages (superior and inferior ones) One of Humboldt’s greatest achievements was to overcome the prevailing type of 17th- and 18th-century universal grammars, which had achieved universality at the expense of linguistic differences. He proposed instead a genuinely comparative grammar, where Latin was not the model against which to hold all other languages. Each natural language, he believed, was characterised by its own ‘inner form’ (the way it articulates concepts and sounds), expressive of the ‘psyche’ of a nation within which it had developed and which it bound together. Unfortunately, Humboldt replaced Latin by Sanskrit and other morphologically complex languages as the model languages against which others could be compared, a conception which led him into the deep waters of Eurocentrism and racism. His typology of languages, put forward in the ‘Introduction to the Kawi Language’, is therefore based on value judgements. To understand this, we have to go back briefly to the topic of the origin of language. At this mysterious moment of the conjoint articulation of sound and thought, the future of every language is determined. Some nations are lucky. They put all their creative energy into this moment of concept and language formation, others (like the Chinese) are less lucky and miss the opportunity of doing so. The continuous (social) language making, following that moment, depends on this once-off spontaneous act of language creation; it cannot really alter it, at least the morphology and syntax — in the shaping of semantics we have more freedom. How should a language look like that would be a perfect medium for the continuous expression of new thoughts? The model of such a language was Sanskrit. It was well established by Humboldt’s times that Sanskrit (and therefore the Indo-European languages descended from it) had a particular morphology, where lexicon and grammar are intimately linked: root words express concepts, inflections indicate the relations between the elements in a sentence. There were however other languages which lacked such a structure, such as Chinese, which is monosyllabic and does not have inflections or means to express grammatical relations. There are languages in between these two extremes of inflectional languages on the one hand and isolating languages on the other: agglutinative and incorporating languages. Humboldt not only observed these structural differences, he used them to rankorder languages from superior Sanskrit to inferior Chinese. There are however other passages which contradict this racist point of view. Humboldt was a complex thinker and a thinker of his time. Let us come back to some less controversial but equally important parts of his philosophy of language.

Wilhelm von Humboldt

6.  Language, culture, and creativity As we have seen, every culture generates a form of language which constrains the linguistic creativity of the speakers. These constraints could be called the linguistic ‘form’, and the factors that build up and change this form could be called the linguistic ‘force’ — what Humboldt called ergon and energeia. Whereas language as form (or system) restricts the freedom of individual linguistic acts, there is a force that reorganises and reproduces this system all the time, and this force is our capacity of speech, which finds expression in our acts of speech. The task of the philosopher of language is to find out how this happens. As early as 1795 Humboldt wrote to his friend Schiller: [Language] has to possess at any moment of its existence the characteristics that make it a whole. Immediate exhalation of an organic being in its physical and spiritual form, it partakes of the nature of all organic phenomena, which is that Each thing in it can only exist through the Other, and Everything can only be through the Force that permeates the Whole. (Humboldt GS, IV: 3, our italics)2

The picture of language as a system of signs used for the mere representation of concepts is here undermined by the introduction of a new factor: the force of the act of speech. For the later Humboldt, signs are not arbitrarily linked to pre-existing meanings; rather “signs, as well as their corresponding thoughts, are formed and fashioned by us at one and the same time and in the same articulatory act of speaking” (Mueller-Vollmer 1992: 135; cf. Humboldt GS, III: 19). The same could be said about thought itself, and about the relation between subject and object, man and world, and subject and subject. They all emerge in the act of speaking and, as we always speak to somebody (and be it ourselves) about something, one could call this the act of communication or conversation. The traces of this emergence of language, world and selves in the act of conversation are, for example, the pronouns and the dual. It is particularly relevant to language that duality has a much more important role in it then anywhere else. All speech is based on interlocution (Wechselrede), in which the speaker always posits the addressee as the one person opposite him, even when there are more people around. […] To divide humanity into two classes, the natives and the enemies, is the basis of all primitive social bonding. (Humboldt 1963 [1827]: 137–138)

For Humboldt, language is a system of signs or an ‘organism’ which organises speech and is itself organised and reorganised through the repeated acts of speaking and

2.  All translations are mine (Brigitte Nerlich), unless otherwise stated.


180 Brigitte Nerlich & David D. Clarke

understanding, through the continued intellectual work of the interlocutors. It is at one and the same time ergon and energeia, and it can never cease to be anything but the ever-repeated mental labour of making the articulated sound capable of expressing thought. In a direct and strict sense this is the definition of speech on any occasion; in its true and essential meaning, however, we can also regard, as it were, only the totality of this speaking as the language. (Humboldt 1988 [1836]: 49; GS, VII:46).

7.  Language, dialogue, and pronouns Thus ‘the totality of speaking’ or energeia is the dialectical mirror image of the ‘totality of language as an organised whole’ or ergon. Language and the speaking subject constitute each other mutually in the very act of interlocution, in the exchange of words between the I and the thou. In externalising their thoughts in language, in objectifying themselves for another, the subjects constitute themselves as subjects and understand each other and the words spoken. However, real objectivity lies not in the thou, but in the s/he. For it [the word] to become something more then a mere object of deceit or a dreamimage, it must embody itself in somebody who hears and replies. This prototype of all languages is expressed by the distinction between the pronoun for the second and the third person. (Humboldt 1963 [1827]: 139)

The thou mediates between the spheres of subjectivity and objectivity, between I and s/he. The monologue of transcendental philosophy is thereby replaced by dialogue, a dialogue based on deictical elements and the context of common action. I and He are really two different things, and between them they exhaust everything, because they are in other words I and non-I. The Thou is however a He posited opposite the I. Whereas I and He are based on internal and external perception, the Thou is based on the spontaneity of choice. It, too, is a non-I, but not as the He situated in the sphere of all creatures, but in another, in that of common action. The He includes therefore, apart from the non-I, also a non-Thou, and it is not only opposed to one of them but to both. (Humboldt 1969 [1827]: 139)

The study of pronouns in terms of deixis and not anaphora sets Humboldt in opposition to the old grammatical tradition of general grammar (cf. Conte 1976: 628). In his Über die Verschiedenheiten des menschlichen Sprachbaues, Humboldt himself distinguished his own analysis of the pronouns from those found in the general grammars of his time. Those were for the most part based on logic and the pronouns have a purely representational function. They overlooked the fact that pronouns are at the heart of language insofar that they are selbstbezeichnend or ‘sui-referential’ (Humboldt 1963 [1827–29]: 203).

Wilhelm von Humboldt

Humboldt’s conception of the pronoun is thus founded on his general assumption that language and thought, subject and object, the language and the world are rooted in the act of speaking. As Mueller-Vollmer points out: “For Humboldt the I-Thou and the deictic It or This mutually and simultaneously constitute each other in the act of speaking” (1992: 138) — and this act of speaking is in fact the deed (Fichte’s Tathandlung) of the transcendental subject. Language is thus the expression of the purposeful activity of the human spirit. It was the task of Madvig, Wegener, Marty and others to get the transcendental subject down to earth and study the real subject in communicative interaction (cf. Nerlich & Clarke 1996).

8.  Humboldt and the Idéologues After this rather condensed aperçu of Humboldt’s pragmatic philosophy of language, a brief detour will engage with a debate that went on in the 1990s between several interpreters of Humboldt. According to some, especially Aarsleff (cf. Aarsleff 1977, 1988), Humboldt was heavily influenced by French philosophers of language, notably Condillac and the Idéologues. According to others (cf. Gipper & Schmitter 1979; Sweet 1978–80; Oesterreicher 1981, 1986; Trabant 1986), Aarsleff ’s claim cannot be substantiated by any textual evidence (for a more balanced account cf. Formigari 1988; Seuren 1998). However, the French theories of language put forward by Condillac on the basis of a post-Lockian philosophy of language, and the German theories of language put forward by Humboldt, based on a post-Kantian philosophy of language, as well as the theories of the origin of language, proposed by Herder and Condillac respectively, were not as incompatible as it seems (cf. Formigari 1994). For both traditions, the post-Condillacian and the postKantian one, it was an axiom that there could be no thought without language. However, there was one point on which it seemed no mutual understanding could be achieved, and Humboldt found this out for himself during his stay in Paris. In 1798 he had met Degérando, one of the leading Idéologues. He also had a five-hour talk with Cabanis, Destutt de Tracy, LaRomiguière, Le Breton and Sieyès, in which he tried to explain German metaphysics to these Idéologues, brought up in a more empiricist tradition of thought. Afterwards, Humboldt wrote that this conversation had been fruitless and unsatisfactory. He noted in his diary: The reason why we couldn’t agree was really the following. All philosophy is founded on the pure intuition of the I, beyond all experience; either explicitly, so that one proceeds directly from this assumption, as Fichte does; or one does so tacitly in showing that to explain phenomena one has to assume something like this, as does Kant. The Frenchmen just do not know this, they have neither a feeling nor a concept for it and so it was that we always were in two different worlds. (Quoted in Gipper & Schmitter 1979: 110; GS XIV: 486)


182 Brigitte Nerlich & David D. Clarke

In another text on Condillac Humboldt underlined even more how much value he attributed to the notion of the spontaneity of the transcendental subject, the starting point of all post-Kantian philosophies of language: He [Condillac] defines and analyses the operations of our soul. It is here however that the biggest flaw becomes apparent, namely that everything is explained as phenomenon, that the real spontaneity which cannot be explained any more, is ignored everywhere; and everything which springs from it is therefore relegated to a lower level. Examples: He has no knowledge whatsoever of self-awareness. (Ibid.: 109; GS XIV: 446) He ignores the true difference, that is, where synthesis adds something to the concept. In one word: the key-stone of metaphysics is missing, the feeling, or rather the spontaneous action (Thathandlung [sic]) of the I. (Ibid.: 109; p. 469)

This insistence on the transcendental deed of the transcendental subject was not only the reason why the French philosophers could not understand Humboldt, it was also the reason why Humboldt was forgotten after the advent of historical-comparative linguistics, which brought with it a more empirically and positivistically oriented approach to language. As Fiesel points out: “As soon as the terrain of Fichte’s idealism was abandoned, the historical approach had to replace the philosophical one in linguistics.” (Fiesel 1927: 67–68) However, in forgetting Humboldt’s sometimes idealistic philosophy of language, linguists also forgot to study the “pragmatic” aspects of language. Language was seen as evolving and changing autonomously like a biological organism, independently of its users, whereas in Humboldt’s philosophy of language the ‘organism’ of language was deeply rooted in the acts of its (joint) production. This pragmatic grounding of language was only rediscovered by the end of the 19th century and came to the fore in the 20th century when linguists, influenced by speech act theory, came to question 20th-century versions of the ‘autonomy’ of language and/or syntax. In Germany itself, the influence of Anglo-saxon speech act theory merged with a long tradition of postKantian, post-Humboldtian and post-Bühlerian thinking to produce philosophically and socially grounded theories of communication, conversation, action and interaction in the works of Apel and Habermas.

References Aarsleff, H. (1977). Guillaume de Humboldt et la pensée linguistique des Idéologues. In A. Joly & J. Stéfanini (Eds), La grammaire générale des modistes aux Idéologues: 217–241. Presses University de Lille. ——— (1982). From Locke to Saussure. Essays on the study of language and intellectual history. Athlone.

Wilhelm von Humboldt

Böhler, M. (1973). Nachwort. In M. Böhler (Ed.), Wilhelm von Humboldt: Schriften zur Sprache: 233–54. Reclam. Borsche, T. (1981). Sprachansichten. Der Begriff der menschlischen Rede in der Sprachphilosophie Wilhelm von Humboldts. Klett-Cotta. ——— (1989). Die innere Form der Sprache. Betrachtungen zu einem Mythos der HumboldtHerme(neu)tik. In H. Scharf (Ed.): 47–65. ——— (1990). Wilhelm von Humboldt. Beck. Burkhardt, A. (1987). Der Dialogbegriff bei Wilhelm von Humboldt. In R. Hoberg (Ed.): 147–173. Conte, M.-E. (1976). Semantische und pragmatische Ansätze in der Sprachtheorie Wilhelm von Humboldts. In H. Parret (Ed.), History of Linguistic Thought and Contemporary Linguistics: 616–632. De Gruyter. Fichte, J.G. (1971 [1795]). Von der Sprachfähigkeit und dem Ursprung der Sprache. In Philosophisches Journal I: 255–73; 287–326. (Repr. in: Fichtes Werke, Ed. by I. Fichte. Vol. VIII: 301–41. de Gruyter, 1971.) Fiesel, E. (1973 [1927]). Die Sprachphilosophie der Deutsche Romantik. Georg Olm Verlag. Flitner, A. & K. Giel (Eds) (1960ff). Wilhelm von Humboldt: Werke in Fünf Bänden. Wissenschaftliche Buchgesellschaft. Formigari, L. (1988). De l’idéalisme dans les théories du langage. Histoire d’une transition. Histoire Epistémologie Langage 10/I: 59–80. ——— (1994). La sémiotique empiriste face au kantisme. Mardaga. Gipper, H. & P. Schmitter (1979). Sprachwissenschaft und Sprachphilosophie im Zeitalter der Romantik. Narr. Haym, R. (1856). Wilhelm von Humboldt, Lebensbild und Charakteristik. Gaertner-Verlag. Herder, J.G. (1772). Abhandlung Über den Ursprung der Sprache. Voss. ——— (1799). Metakritik zur Kritik der reinen Vernunft. Erster Theil: Verstand und Erfahrung. Frankfurt & Leipzig. ——— (1877–1913). Sämtliche Werke, Ed. by B. Suphan et al. Weidmann. Hoberg, R. (Ed.) (1987). Sprache und Bildung. Beiträge zum 150. Todestag Wilhelm von Humboldts. THD-Schriftenreihe. von Humboldt, W. (1903–1936). Gesammelte Schriften, Ed. by A. Leitzmann. Behr. ——— (1963). Werke in Fünf Bänden, Bd. III, Schriften zur Sprachphilosophie. Wissenschaftliche Buchgesellschaft. ——— (1963 [1827–29]). Über die Verschiedenheiten des menschlichen Sprachbaues. In Humboldt (1963): III, 144–367. ——— (1963 [1827]). Ueber den Dualis. In Humboldt (1963): III, 113–43. ——— (1963 [1830/5]). Über die Verschiedenheit des menschlichen Sprachbaues und ihren Einfluss auf die geistige Entwicklung des Menschengeschlechts. In Humboldt (1963): III, 368–75. ——— (1988 [1836]). On Language: The diversity of human language-structure and its influence on the mental development of mankind. Cambridge University Press. Knoll, J.H. & H. Siebert (1967). Wilhelm von Humboldt. Politician and educationist. Inter Nationes. Kretzmann, N. (1967). Semantics, History of. In P. Edwards (Ed.), The Encyclopedia of Philosophy Vol 7: 358–406. Macmillan. Lee, P. (1996). The Whorf Theory Complex: A Critical Reconstruction. John Benjamins. Leibniz, G.W. (1981 [1765]). [Nouveaux essais sur l’entendement humain] New Essays on Human Understanding. Cambridge University Press. Leitzmann, A. (1919). Wilhelm von Humboldt. Charakteristik und Lebensbild. Niemeyer. Locke, J. (1975 [1689]). Essay on Human Understanding. Oxford University Press.


184 Brigitte Nerlich & David D. Clarke Manchester, M.L. (1985). The Philosophical Foundations of Humboldt’s Linguistic Doctrins. John Benjamins. Mueller-Vollmer, K. (1992). Wilhelm von Humboldt’s Introduction to the Kawi Language. Semiotica 93:1/2: 125–144. Nerlich, B. (1995). The notion of ‘speech act’ in German linguistics, philosophy and psychology between 1830 and 1970. In E. Fava (Ed.), Speech Acts and Linguistic Research: 1–20. Edizione Nemo. Nerlich, B. & D. Clarke (1996). Language, action, and context. The early history of pragmatics in Europe and America, 1780–1930. John Benjamins. Oesterreicher, W. (1981). Wem gehört Humboldt? Zum Einfluss der französischen Aufklärung auf die Sprachphilosophie der deutschen Romantik. In H. Geckeler (Ed.), Logos Semantikos: 117–35. de Gruyter. Pott, A.F. (1876). Wilhelm von Humboldt und die Sprachwissenschaft. Calvary & Co. Scharf, H.-W. (Ed.) (1989). Wilhelm von Humboldts Sprachdenken. Essen. Schlerath, B. (Ed.) (1986). W. von Humboldt. Vortragszyklus zum 150. Todestag. de Gruyter. Schmitter, P. (1986). Humboldt, Wilhelm von (1797–1835). In T. Sebeok (Ed.), Encyclopedic Dictionary of Semiotics I: 317–323. Mouton de Gruyter. ——— (1987). Das sprachliche Zeichen. Nodus. ——— (1992). “Maschine” vs. “Organismus”. Einige Überlegungen zur Geistes- und Sprachwissenschaftsgeschichte im 18. und 19. Jahrhundert. In A. Ahlqvist et al. (Eds), Diversions of Galway: 291–307. John Benjamins. Seiffert, L. (1968). Neo-Humboldtian Semantics in Perspective. Journal of Linguistics 4: 93–108. Seuren, P.A.M. (1998). Western Linguistics. An historical introduction. Blackwell. Steinthal, H. (1848). Die Sprachwissenschaft Wilhelm von Humboldts und die Hegelsche Philosophie. Dümmler. Sweet, P.R. (1978–80). Wilhelm von Humboldt: A Biography, 2 vols. Ohio State University Press. Trabant, J. (1986). Apeliotes oder Der Sinn der Sprache. Wilhelm von Humboldts Sprach-Bild. Fink.

Intercultural communication Volker Hinnenkamp Fulda University of Applied Sciences

1.  Background: Language and culture ‘The world is small!’, Columbus is said to have exclaimed when he finally reached the New World. Columbus’ ‘discovery’ is also said to have finally combined the two hemispheres, thus having created the one world we live in today — the beginning of that global village, so often cited, where we all can or have to communicate with each other. Of course, between Columbus and satellite communication there was a long way to go. At least the event was an important milestone for the westerners of the world. Intercultural (or crosscultural) communication did not start with Columbus’ experience. It is rooted in the evolutionary differentiation of languages and cultures, however uni- or polycentral cultural evolution took place, and it started with contact between whoever regarded other individuals and groups as different: ‘They’re different from us, they don’t do things the way we do them. They do them in a strange way’. Culture contact is as old as trade, wars, migrations, conquests and the like, actually as old as mankind. It is only since the emergence of the art of record keeping that we are provided with a kind of literature on the subject. Marco Polo’s and Columbus’ notes and diaries are just two popular samples in a long history of recording culture contact. The study of intercultural communication as a special field, not only within pragmatics, is quite new. But the relationship between culture and language is a longstanding topic of inquiry. One such trend may be traced back to the Humboldtian tradition where the language-and-culture link found itself expressed in terms of grammar, worldview, and national character. Wilhelm von Humboldt thought that “each language draws a circle around the nation to which it belongs” [“jede Sprache zieht um die Nation, welcher sie angehört, einen Kreis”] (1830–35, Vol. VII: 60), a circle hard to escape. Different versions emerge in the anthropological work of Franz Boas and Edward Sapir, in Benjamin Lee Whorf ’s ‘principle of linguistic relativity’, and (though not from a strictly relativistic viewpoint) in Bronislaw Malinowski’s concept of ‘context-of-situation’. Versions vary from a once-popular deterministic re-interpretation of Whorf to Sherzer’s (1987) discourse-centered approach, or Silverstein’s (1979) reading of Whorf ’s theory of grammatical categories in terms of language ideologies. But the topic itself has had an enduring presence for more than a century. It was not until the 1960s, with Dell Hymes’ seminal works on the ethnography of speaking (see e.g., Hymes 1974), that linguistics was provided with a systematic

186 Volker Hinnenkamp

methodology that could account for the interdependence of language, speech, communication and culture. Intercultural communication is about the confrontation of one language-culture link with another. More specifically, it is human beings bearing the whole burden of culture-in-communicating as individuals, who meet, converse, talk, have conflicts, struggle, i.e., communicate in face-to-face interaction. Michael Agar has coined the term ‘languaculture’, intended as “a reminder […] of the necessary connection between its two parts, whether it’s theirs, or yours, or, as it always is when it becomes personal, something that belongs to you both” (Agar 1993: 60). A term such as languaculture seems useful as long as the juxtaposition of language and culture (likewise language and society) indicates that language runs the risk of being stripped of all these links. But at least in pragmatics, the cultural and social are part and parcel of the term ‘language’, to the point where ‘intercultural communication’ should become nearly tautologous. 2.  Intercultural communication: The emergence of a field of inquiry Intercultural communication has become a very popular field. In addition to linguistic pragmatics, contributing disciplines include communication science, sociology, psychology, and anthropology. Language teaching and language learning are two areas of application to which intercultural communication is of particular interest. The study of intercultural communication as a multidisciplinary venture has its roots mainly in the 1960s in North America. Besides the academic research, there was a very strong orientation towards practice right from the beginning. After the emergence of contrastive linguistics, due to the new demands of foreign language learning in the wake of World War II, language learning alone was not sufficient for the new role of the United States of America in the the world. Robert Lado (1957) made a first attempt at expanding contrastive linguistics to contrastive culture analysis. Around this time, Edward Hall, one of the pioneers in this field, is said to have been the first who gave the term intercultural communication an audience (Trager & Hall 1954; Hall 1959; Hall & Whyte 1960). Meeting the challenge in the global race with communism meant exerting influence internationally. In the 1960s thousands of American volunteers went to the ‘developing countries’ to give technical, social and medical support through the Peace Corps programs. They were not only trained in the local languages, but they were also given culture training to prepare them for the encounter with the natives. Many of the still popular games in intercultural awareness training stem from these times. Another motivation for the emergence of the field was to be found in the American immigrant society itself, in the Civil Rights Movement, in the New Ethnicity of African Americans and other mainly non-white ethnic groups. “Even when we overcome natural barriers of language”, stated Porter & Samovar (1974: 4), “we could still fail to understand and

Intercultural communication

be understood. These failures both in the international arena and the domestic scene gave rise to the marriage of culture and communication and to the recognition of intercultural communication as a field of study”. Unlike with other trends in academia, Europe did not latch on before the 1970s and 1980s. Although colonial and postcolonial immigration, as well as labor migration from Southern European countries to Western Europe presented a languacultural challenge as early as the 1950s, urban multiculturalism was conceptually non-existent. Germany, e.g., responded with German language programs. Only when institutions became aware of the multicultural reality of immigrants having become established sections of the population, and when refugees widened the flow of immigrants, ethnic minorities were beginning to be ‘noticed’. Studies on Gastarbeiterdeutsch (Immigrant German, literally ‘Guestworker German’; Heidelberger Forschungsprojekt 1975; Keim 1978) were followed by research on intercultural communication between Germans and Turks, Greeks, Italians etc. (Rehbein 1985; Knapp, Enninger & Knapp-Potthoff 1987; Hinnenkamp 1990). Meanwhile continuous international and intercultural mobility has come to be seen as one of the characteristics of the age of globalization. This permanent mobility can be structurally differentiated in terms of voluntary and involuntary, the latter implying little or no choice as to place and length of stay. Rights and obligations are quite unbalanced for the different types of sojourners. Refugees of war, famine, disease, political oppression and ethnic cleansing will find themselves at the mercy of the host country, and so will most labor migrants, although many of them will have more options to choose from. International tourism, exchange programs and business negotiations are clearly to be situated at the beneficial or even mutually profitable end of the scale. Such structural differences do have an impact on intercultural encounters. Often there are even specific institutions provided for the different kinds of migrants, where special kinds of encounters between local and foreign people take place (Jupp, Roberts & Cook-Gumperz 1982; Roberts, Davies & Jupp 1993; Koole & ten Thije 1994). Many intercultural encounters can be reformulated in terms of minority-majority relations. This is also where the term ‘interethnic’ comes in, alluding to ethnic minorities and focusing on their differences from mainstream society (cf. Scollon & Scollon 1981; Erickson & Shultz 1982; Gumperz 1982a, 1982b; Streeck 1985; Kim 1986; Meeuwis 1994b). Here we also find the rationale for dealing with intercultural communication in theory and practice. Intercultural encounters are typically believed to demand extra effort of conversants in making themselves understood, or even to be bound to failure. In the literature there is a heavy emphasis on communication breakdown, communicative disruption, failure, trouble and misunderstanding. A term covering all these is miscommunication. Miscommunication and cultural or languacultural differences are thought to go hand in hand. Studying and preventing miscommunication has therefore become the main raison d’être of the field. Trouble and misunderstanding lurk in all kinds of inter­cultural communications, in international business as much as in international political relations;


188 Volker Hinnenkamp

the former may lead to the loss of a good business deal, the latter to diplomatic crises or even to war (“the fate of the earth depends on cross-cultural communication” — Tannen 1986: 30). In interethnic minority-majority encounters, misunderstanding may lead to (additional) discrimination and disadvantages in matters of employment, housing, health, and civil rights. The precise interplay between factors such as cultural or languacultural difference, power, social constraints, prejudice and ideological a priori’s, and their (causal) relationship to miscommunication phenomena, are prime topics for pragmatically oriented intercultural communication analyses (see Sarangi 1994). 3.  The concept of culture But intercultural communication itself may give rise to misunderstandings as well. It forms the breeding ground for a multitude of myths and illusions (Verschueren 1984). For one thing, culture in intercultural communication is far from being clearly and explicitly defined within the scientific community of intercultural communication researchers (see e.g., Blommaert 1991). Moreover, whatever definition of culture we adopt, we still have the difficulty of showing how communication at any given moment is bound by culture or how culture continuously finds expression in communication. Even this phrasing of the problem is misleading, because it suggests two separate entities — communication and culture — whereas it has to be shown that the one is an integral part of the other, that culture is to be found within the use of language, just as every Sprachspiel (language game), to use Wittgenstein’s phrase, is embedded in the Lebenswelt (way of life) of the speaker and his/her group. Separating ‘communication’ and ‘culture’ would imply that certain forms of communication could be a-cultural or culture-free, totally untouched by the communicator’s sociocultural background. The juxtaposition of language and culture can therefore only be a provisional, yet necessary, analytic form of meta-discourse. Culture in intercultural communication, as Agar (1993) emphasized, is nothing to be objectified in individuals and groups. Culture is at first a very personal experience when it comes to communication outside our taken-for-granted ways of acting and framing the world. Culture as a scientific term “is supposed to explain differences, to take rich points and make them understandable”, that is, it “is supposed to be an answer to the problem of understanding differences” (Agar 1993: 124–125). There are some classical definitions of culture, which even fit nicely into the concept of talking, acting and interpretation. One often quoted definition is by Ward H. Goodenough, who holds that a society’s culture consists of whatever it is one has to know or believe in order to operate in a manner acceptable to its members, and to do so in any role that they accept for any one of themselves. Culture, being what people have to learn as distinct from biological heritage, must consist of the end product of learning: knowledge, in a most

Intercultural communication 189

general, if relative, sense of the term. […] culture is not a material phenomenon; it does not consist of things, people, behavior, or emotions. It is rather an organization of these things. It is the form of things that people have in mind, their models for perceiving, relating, and otherwise interperting them. As such, the things people say and do, their social arrangements and events, are products or by-products of their culture as they apply it to the task of perceiving and dealing with their circumstances. (1964: 36)

Towards the end of this quotation, culture becomes slightly reified, though it is less the product than the producer and product at the same time. But otherwise such a cognitively oriented definition comprises many of the relevant aspects, particularly for a pragmatic perspective on communicative competence and mutual knowledge. But that is certainly not all we need for an understanding of culture in intercultural communication. Culture has become an issue in itself, a discourse to look at critically. The complexity of our urban cultures necessitates differentiation: “the term ‘culture’, with its cosy invocation of consensus, may serve to distract attention from social and cultural contradictions, from the fractures and oppositions within the whole” says E.P. Thompson (1993: 6). Questions arise as to whether there can be a sense of multiple cultural belonging in a person, just as we find multiple identities. If so, how do these find their expression in communication? If there are overarching national or ethnic cultures, how are they communicated? How does the relevance of culture to communication manifest itself? Are there ‘rich points’, as Agar (1993) calls them, such as culture-bound signals, offering a glimpse of the tip of the cultural iceberg? Or is culture-in-communication just a marginal phenomenon so that errors can be as routinely corrected as grammatical mistakes? Moreover, can culture-in-communication not easily be overridden by common membership? Are we, as a group of culturally mixed researchers and writers communicating on pragmatics, not a good example to show that professional identity is stronger than any cultural difference? Finally we have to realize that, e.g., under the impact of worldwide communication, media languages and cultures continuously cross old semiotic borders creating hybrid and transcultural codes and identities, questioning established concepts and definitions of culture (Welsch 1999; Karim 2003; Georgiou 2006). Different definitions and approaches imply different discourses. None holds the truth. But they do have a common task: to show, to analyze and ultimately to construct a good theory of how culture translates into human communicative interchanges, how communication is part of the phenomenon or condition we call culture, and how both affect interchanges between members of culturally defined entities.

4.  Loci of culture-in-communication There are different loci of culture-in-communication. Whatever layer of the interactional structure we analyze, it may be of cultural relevance to the communicating

190 Volker Hinnenkamp

parties. Resources of interaction are based on features and properties. These can be ‘localized’. Most of the studies in intercultural communication show how one or more locus is made relevant for participants in an encounter. Thus, culture may be located in the style of a speaker, in his or her ways of speaking, of structuring arguments or of sequencing information units. It may be located in aspects of behavioral competence such as politeness, deference, or proper conduct. It may be located in ‘language’ competence, in native vs. non-native proficiency. It may be located in nonverbal signals, such as gaze direction or territoriality. It may be located in switching between language varieties. It may be located in stereotyped behavior, in opinions, attitudes and worldviews. It may be located in the available power resources. Generally, it may be located in the use of any ‘brought along’ property of a person or his/her group in terms of ‘visible’ categories such as skin color, gender, language, dialect, or less visible ones such as nationality, ethnicity, religion and the like. But whatever the loci, in order to work as a resource, what is needed is the interactional ‘counterpart’ of deriving meaning from them, of inferring frames, of guiding contingent action as participants go along. One locus commonly focused on in intercultural communication research is related to language selection. At one end of the scale, intercultural encounters may take place within the same language, as described in Gumperz (1982a, 1982b), Scollon & Scollon (1981) and Erickson & Shultz (1982), where native speakers of (a majority variety of) English encounter people speaking English as a second language or a minority dialect of English (speakers from India or the Carribean; London Jamaicans, African Americans, Hispanic Americans, to name just a few). Differences may be perfunctory at first sight, but when it comes to mutual perceptions of difference and attributions, minor stylistic variants may turn out to have a major effect on the process of communication. In particular, the function of nonverbal signs in a cross- and intercultural perspective have been investigated intensively. As early as 1941 and without any reference to interculturality, David Efron compared some gestural sign language use amongst New Yorkers with South Italian and East Jewish backgrounds and described their ‘hybridization’ under the impact of the urban ‘melting pot’ (Efron 1972). The main pioneer in this context is Edward Hall, who compared many elements of nonverbal ‘silent language’ crossculturally (Hall 1959). As another example, Erickson & Shultz (1982) have investigated, by way of micro-ethnography, how minor nonverbal differences in the synchronization and rhythm of head-nodding in conversations between black and white participants may influence the outcome of a counselling situation. At the other end of the scale, interlocutors may not have a language in common. Sometimes a provisional language has to be created, as in initial culture contact situations (Hewes 1974), resulting in pidginization processes or in ‘foreigner talk’ (Hinnenkamp 1982; Roche 1989; Jakovidou 1994). Further, we find languages used as lingua francas with no native speakers on either side (Jordan & Fuller 1974; Meeuwis

Intercultural communication

1994a; Knapp & Meierkord 2002), as well as native-nonnative interaction, which is one of the most typical intercultural contexts. An example is Kotthoff (1989), who contrasts the discourse of argumentation between Germans and Americans in a German university context in terms of deficits in competence versus pragmatic differences. Native-nonnative speakership is an important issue in interlanguage pragmatics (e.g., Varonis & Gass 1985), and intercultural aspects are gaining more and more prominence (for an overview cf. Wagner 1996; Barron 2001). It seems that culture is more easily referred to when discussing encounters with little ‘linguistic’ problems. On the other hand, the more language and communication have to function as a medium of pure content transmission, the less culture is emphasized. Perhaps this corresponds to a culturally defined need of focusing on the ‘content’, without the luxury of pragmatic and cultural ‘frills’. However, more research on native-nonnative communication is expected to show that this habitual perspective is untenable. Using cultural features for struggles over resources of prestige, power and dominance is part and parcel of intercultural interactions. How this works has been shown in the literature (Tsuda 1986; Hinnenkamp 1989; Meeuwis 1994b). Struggles over resources are most often institutionally conditioned. Personal or institutional power, involving the possibility to exert force, overrides any cultural feature as candidate resource for further action. This is why so many contact situations, e.g., in the history of European expansionism and later in North-South relations, can hardly be called ‘intercultural’ communication in the strict sense of the term (see e.g., Blommaert 1990; Eelen 1993). Even the understanding of the others, that holy concept of so many optimistic intercultural practitioners, has been all too often instrumentalized as a strategic weapon to dominate or eliminate other cultures, as Todorov (1982) has convincingly shown with reference to the Spanish conquest of Mexico. The exertion of power was also involved in pidginization and creolization processes. Pidgins and creoles have so far been seen as just the linguistic outcome of frequent contact situations. Of course contact is an important factor. Culture, however, also played an important role, although of course, adjustment to the new situation was permanently forced upon natives or slaves, so that they had no choice on how to exploit resources to their own advantage (Stoller 1979; Siegel 1987). Colonizers and slaveholders were affected much less by the change. Natives or slaves had to give up their languages and their cultures, and had to adapt. So intercultural communication is not necessarily to the better of human kind. It can also be to the disadvantage of those groups who were the disadvantaged all along. A particularly interesting topic in this respect is the way in which the institutionalized masterservant relations in a colonial context gave rise to the construction of particular varieties of local languages (mostly labeled as lingua franca, sabir, patois and so on), and how the colonial power relations were inscribed in the structure and the patterns


192 Volker Hinnenkamp

of usage of these language varieties (see, e.g., Fabian 1986 and 1991 on Swahili in the Belgian Congo). 5.  Methodological (sub)discourses of the field The intercultural discourse within linguistic pragmatics comprises various types of methodological subdiscourses. Within intercultural communication research several trends have been established in peaceful competition and sometimes peaceful ignorance of each other. Many of them share a number of shortcomings. For one thing, definitions of the topic are often skipped altogether, or only found implicitly. Second, culture remains an object of interpretation, and is not shown to be a resource. Third, there is a strong inclination towards ideology, anecdotes, stereotypic knowledge and assumptions over empirical evidence. The different approaches we come across are not mutually exclusive, but for analytical reasons they can be nicely contrasted with each other.1 Two approaches in particular have been of special relevance in pragmatics. The first one is the contrastive approach, represented in contrastive pragmatics in general, or more specifically in contrastive discourse analysis, contrastive rhetoric, contrastive sociolinguistics, or contrastive textology. Underlying this approach is the classical leitmotif of contrastive analysis that difference means a potential source of errors, learning difficulties, and interference. “Cross-cultural communicative interference, then, is the result of the negative transfer into the L2 context of L1 sociolinguistic and interactional rules due to a contrasting interdependence between speech behaviour and cultural world-view and value system”, as Loveday puts it (1982: 2). For Riley (1989: 234) “[p]ragmatic errors are the result of an interactant’s imposing the social rules of one culture on his communicative behavior in a situation where the social rules of another culture would be more appropriate”. So the rationale of contrastive pragmatics lies in language learning issues. Central to contrastive pragmatics is speech act analysis. The Cross-Cultural Speech Act Realization Project (CCSARP), to date the largest research project to systematically inquire into the cultural specificity of speech act behavior, was set up to investigate cross-cultural variation of the two speech act types of ‘requesting’ and ‘apologizing’ across such various speech communities (and cultures?) as the German, British, 1.  Only some major works will be introduced in the sections to follow. For an overview of the literature, the reader is referred to a comprehensive bibliography of the field (Hinnenkamp 1994) which provides more than 800 references, two thirds of its entries covering the years from 1983 to 1994. The bibliography is organized in six main sections: (1) Introductions; (2) Theory and methodology; (3) Interactional studies; (4) Contrastive studies; (5) Special areas; and (6) Applications. There are also two other selective bibliographies: Loos (1991) and Dirven & Pokuta (1994).

Intercultural communication

American, Australian, Danish, Israeli and French. Both speech acts are regarded as particularly pertinent in terms of politeness and ‘face work’ (as defined by Goffman 1955 and Brown & Levinson 1987) and in terms of culture-specific versus universal phenomena. Data were mainly gathered by a Discourse Completion Test, where subjects had to fill in dialogical replies in 16 different types of situations, selected according to the parameters of power, distance and rank. Some authentic data support the experimental findings. The goals of the project can be taken as paradigmatic for many other investigations into culture-specific speech act use, i.e., 1. to investigate the similarities and differences in the realization patterns of given speech acts across different languages, relative to the same social constraints (cross-cultural variation), 2. to investigate the effect of social variables on the realization patterns of given speech acts within specific speech communities (sociopragmatic variation), 3. to investigate the similarities and differences in the realization patterns of given speech acts between native and nonnative speakers of a given language, relative to the same social constraints (interlanguage variation). (Blum-Kulka, House & Kasper 1989: 12f.) Another major project, the PIXI project (Aston 1988; Gavioli & Mansfield 1990; Anderson 1994), involves contrastive discourse or contrastive conversation analysis. Its main objective was to “identify regularities of socially ‘unmarked’ collaborative talk in two cultures”, i.e., English and Italian bookshop encounters, “analysing eventual occurrences of conflict in terms of deviation from these regularities” (Gavioli & Mansfield 1990: xviii). The analyses of the project focused on: contrastive aspects of negotiation, of joint and progressive text production by participants; contrasting underlying scripts; examining contrasts in the lexicogrammatical and intonational forms as well as recurrent features such as laughter and interruptions; the sequential organization of requests and responses; patterns of openings and closings and accounting practices; participants’ strategies of dominance and control, stressing the ways in which identities and statuses were established and maintained through such strategies. Both the CCSARP-project and the PIXI-project stand for advanced research within contrastive pragmatics. Both are empirically oriented, though CCSARP uses ethnographic data only as control data in addition to their experimental design. The PIXI data are authentic: 379 naturally occurring conversations recorded in bookshops in England in Italy, processed mainly by means of ethnomethodological conversation analysis. Yet both projects, advanced as they are in methodology, sampling and analysis, are not intercultural in the strict sense of the term, but rather crosscultural (and would not claim otherwise). The latter denomination emphasizes the comparative aspect of two cultures, rather than the interactional perspective. We may assume that different phenomena will have to be accounted for when an Italian customer negotiates with an English bookseller, or if a German requests something from a speaker of


194 Volker Hinnenkamp

Hebrew. In such cases, the respective conventions will not simply be carried over to the encounter, but changes or adaptations will take place; and in case of conflict, repairs, accounting practices and the like will be used. Acting in an intercultural situation may be very different from what a researcher may imagine. Michael Roberts, as a historian, has shown how an entire history can be contextualized in the culminating ‘ethnic’ speech act “I am a Sinhalese” in a casual verbal exchange during a cricket match in Sri Lanka. Roberts’ analysis “underlines the manner in which the imponderabilia of everyday transactions can utilise and portray deeply rooted historical perceptions, and in so doing, may contribute to the reformulation, reproduction, and transmission of these imprints from the past” (Roberts 1985: 407). A contrastive approach is not able to reveal such intercultural problems, which only show up in real, authentic face-to-face intercultural interactions. It has to be stated, though, that it is sometimes hard to draw the line between contrastive and interactional approaches. Especially in what is now called “cross-cultural” or “intercultural pragmatics” – even with an international journal with that label – we find an enormous range of data-driven empirical studies and (meta)theoretical reflections which go beyond “contrastive pragmatics”. Sometimes the term “ethnopragmatics” (Goddard 2006) is used, sometimes the relevance of “communicative genres” (Günthner 2007) is at issue. It is logical, therefore, that the second major approach is interactional, focusing on authentic data, not only within one’s group or society, but derived from groups in contact, in confrontation, with the ‘cultural other’. Interactional or interpretive sociolinguistics is one tradition that provides us with analyses of face-to-face intercultural, interethnic and interracial encounters. The first scholar to mention here, is John Gumperz (1982a, 1982b). Gumperz dived straight into multicultural society, analyzing encounters between white British and American majority members and ethnic minority members in situations of counselling, job interviews, committee negotiations and the like, i.e., in institutional settings relevant to urban multicultural societies. What he found was that simple interpretations based on ‘low’ linguistic levels such as pausing, rhythm, tempo, volume, pitch and accent, intonation and prosody in general, could produce major misunderstandings. One of Gumperz’s well-known examples goes as follows: In the staff cafeteria at a major British airport, newly hired Indian and Pakistani women were perceived as surly and uncooperative […] Observation revealed that while relatively few words were exchanged, the intonation and manner in which these words were pronounced were interpreted negatively”. Instead of using ‘correct’ British intonation in so simple a question as “Gravy?”, intended to ask the customer if s/he wanted some gravy added to the meal, the Indian assistant would pronounce the word with falling intonation. However, this intonation was “not interpreted as an offer but rather a statement, which in the context seems redundant and consequently rude. (Gumperz 1982a: 173)

Intercultural communication

Gumperz calls these signals in situated communication ‘contextualization cues’, because they support interpretive frames of how to understand an utterance or an act. In Gumperz’s words: “a contextualization cue is any feature of linguistic form that contributes to the signalling of contextual presuppositions” (1982a: 133). Part of a speaker’s tacit sociocultural competence is the knowledge of how to employ and how to understand the cues; they will be conventionalized and taken for granted. When all participants understand and notice the relevant [contextualization] cues, interpretive processes are then taken for granted and tend to go unnoticed. However, when a listener does not react to a cue or is unaware of its function, interpretations may differ and misunderstanding occur. It is important to note that when this happens and when a difference in interpretation is brought to a participant’s attention, it tends to be seen in attitudinal terms. A speaker is said to be unfriendly, impertinent, rude, uncooperative, or to fail to understand. Interactants do not ordinarily notice that the listener may have failed to perceive a shift in rhythm or a change in pronunciation. Miscommunication of this type, in other words, is regarded as a social faux pas and leads to misjudgements of the speaker’s intent; it is not likely to be identified as a mere linguistic error. (Gumperz 1982a: 132)

Gumperz’s approach is relevant to pragmatics far beyond the field of inter­ cultural  communication, because contextualization cues are general signalling devices for creating and sustaining any kind of conversational and communicative involvement as such. Quite a large body of research has grown out of it (cf. Auer & di Luzio 1992). One of the advantages of the contextualization approach is the detectability of many of the signalling cues through microethnographic methods. The approach is able to link language and culture on a level of “shared typifications that enter into the signalling and use of activity types in interaction, as well as systems of contextualization conventions”, thus arriving at “interactively defined notions of culture [that] can be studied by empirical means” (Gumperz 1992: 51f). Susanne Günthner is another empirical intercultural communication researcher who works within the paradigm of interactional or interpretive sociolinguistics. Her work deals with conversations between Germans and Chinese learners of German in institutionalized settings, such as counselling situations, and in informal settings, such as social gatherings (Günthner 1993a). Günthner concentrates among other things on (contextualization) differences in discourse organization and in structuring arguments; she shows that conventions of recipient behavior of the Chinese differ sharply from German usage; she inquires into the function of proverbial uses in discourse. Her data show that “even Chinese with a good command of the German language rely on their own contextualization conventions, which are partly different from the German ones” (Günthner 1993b: 302).


196 Volker Hinnenkamp

Hinnenkamp’s works in interactional sociolinguistics (1987, 1989, 1991) widens the perspective of intercultural communication by his attempt to fully integrate society into the language-and-culture link. Hinnenkamp investigates ‘pretexts’ in relation to created contexts, and asks: “How and why is it possible that the said is sayable, that it can be said that way, that it is permissible to be said that way, and that it can be understood that way?” (Hinnenkamp 1991: 93). He resorts inter alia to the French sociologist Pierre Bourdieu and his credo “What speaks is not the utterance, the language, but the whole social person” (Bourdieu 1977). Similar to what happens in Michael Roberts’ approach, single speech acts can take the function of ‘rich points’ in terms of transforming knowledge of societal pretexts into discourse. According to Bourdieu, language, culture and ethnicity are negotiable goods on the market of symbolic capital. There is linguistic, cultural and social capital ‘convertible’ into discursive practice. Hinnenkamp applies this to a casual Turkish-German conversation, where he shows how a sudden switch into an emblematic, stereotypic kind of foreigner talk (Türkischmann Du? — ‘Turkish man, you [= tu]?’) transforms an ongoing exchange completely: culture, ethnicity, language and social position become an issue and are structurally reflected in the new allocation of discoursal space and in unequal rights and obligations, resulting in a disharmonious end of the encounter. The process shown in the example is one of active ethnic categorization, it is an example in which the minority member’s position in society can be turned into a resource for social distinction on a hierarchical ladder. The German interlocutor was a beggar, down at the bottom of society; the Turkish interlocutor was an immigrant worker, not quite that low in the hierarchy. For the beggar, the strongest symbolic profit to be gained was ethnic capital, since ethnic difference is the only irretrievable category able to draw a sharp distinction. Hinnenkamp (1987, 1989) provides more examples where it is shown that time and again an immigrant worker’s German competence is functionalized as the main resource in culturalization, ethnification and discrimination. There are of course other approaches in addition to the ones discussed so far. One of them is the ‘anecdotal’ or ‘case study approach’. It focuses on perceived or alleged cultural differences, mostly with reference to so-called ‘critical incidents’ which are given an illustrative or at least introductory role to reflect on intercultural communication. There are some theoretical insights which have been gained by this method, especially as to patterns of interpretation. Such critical incidents can certainly help to reflect on ‘rich points’ of an intercultural encounter, or, at least of another culture (Agar 1993). Actually, the majority of the publications in the field are of this kind; studies based on hard empirical data of authentic face-to-face encounters, crossculturally as much as interculturally are a minority. Another — less prominent but very critical — approach does not so much focus on the interpersonal communicative exchange level between ‘cultural others’, but views intercultural communication as being constructed by way of individual,

Intercultural communication

interpersonal and institutional discourses, i.e., by way of the textual presentation of communicative events as intercultural communication. That is, the more lay people (in everyday talk, jokes, interviews), or specialists (in institutional settings like the media, administration, politics), or scientific observers (in research, publications, academic discussions) talk about culture, ethnicity and intercultural communication, the more they help construing it as a discursive reality. This approach could be regarded as part of the sociology of knowledge or as a constructivist approach. Here we may situate Blommaert (1994), who criticizes the ideological impact of an allegedly neutral concept of culture within academic and social-pedagogical discourse (cf. also Blommaert & Verschueren 1993), and the work of Tom Koole & Jan ten Thije (1994), who investigate team sessions of educational advisers in view of the way in which Dutch institutions react toward the entry of immigrant employees. The latter approach is more or less in line with an emerging critical tradition of intercultural communication research, inspired by critical theory and critical linguistics (see e.g., Fairclough 1989). Despite the fact that most authors within this critical tradition acknowledge the pioneering nature and quality of Gumperz’s work, they emphasize societal and institutional conditions of disadvantage and discrimination against immigrants or ethnic minorities, criticizing a linguistic and culturalist bias in Gumperz’ approach (see e.g., Kandiah 1991, and most of the contributions to Meeuwis 1994b). Their aim is also to open the analysis of intercultural encounters to insights from present-day social and cultural theory, in particular those related to ideology and hegemony (Meeuwis & Sarangi 1994; Holliday, Hyde & Kullman 2004). 6.  Conclusion What is certainly needed in the pragmatics of intercultural communication is more investigations into face-to-face encounters between members of different cultural or ethnic minority groups. These investigations should cover the relevant groups of societies, such as immigrants and refugees in most Western European countries. Moreover, these investigations should take into account the real localities of interaction. There are many places in society — institutions such as hospitals, or market places, or schools — where different settings will probably lead to quite different forms of intercultural contact. The commercial sphere, one which is most strongly characterized by a structural integration of ethnic minority people (as opposed to education or health care) has not been tackled yet and is very likely to yield new and perhaps surprising results. Another very interesting type of interculturality has been emerging under the impact of multilingualism from immigration, such as the development of local hybrid languages and ethnolects and the stylizations of these languages. Rampton (1995) and Dirim & Auer (2004) have shown how ‘language crossing’ even affects groups


198 Volker Hinnenkamp

that do not ‘possess’ these languages. In a series of studies, adolescents’ use of hybrid language mixing have shown how immigration and hegemonic language demands were responded to by creating new social and identity spaces by minority speakers (e.g., Hinnenkamp 2003; Eksner 2006). These studies are quite at the beginning. They require in particular ethnographic methods. What we also need is an investigation of how children acquire cultural awareness and concepts of cultural otherness in discourse. How do concepts of strangeness, ethnicity, cultural outsiderdom develop in interaction among members of ethnically homogeneous and ethnically heterogeneous groups, how can processes of “otherization” or “integration” be reconstructed? An incipient study was already tackled by Hewitt (1986) when he looked at the mutual influence of ethnic styles in black and white friendship groups among children. Finally, and on a more theoretical level, the methodological development of the field should not be neglected. In particular, the conceptual apparatus used in the analysis of intercultural communication is far from unproblematic (as has already been mentioned above), and should be under constant scrutiny. Questions of methodology and epistemology are hardly ever touched upon, despite the fact that recognizing intercultural communication as an object of investigation would also imply that analyzing intercultural communication (or, in general, data from another languaculture) is an instance of intercultural communication in its own right (Fabian 1990; Blommaert 1991; Otten, Allwood et al. 2009). Intercultural communication is a new and booming field. I have mentioned a few — subjectively chosen — studies, mostly empirical ones. Some very interesting insights have been gained, finding their way into various channels of publication such as the new journal Intercultural Pragmatics. Many questions remain unanswered; many more have yet to be formulated.

References Agar, M. (1993). Language shock. William Morrow. Anderson, L. (1994). Accounting practices in service encounters in English and Italian. In H. Pürschel (Ed.), Intercultural communication: 99–120. Aston, G. (Ed.) (1988). Negotiating service. Cooperativa Libraria Universitaria Editrice Bologna. Auer, P. & A. Di Luzio (Eds) (1992). The contextualization of language. John Benjamins. Barron, A. (2001). Acquisition in Interlanguage Pragmatics: Learning how to Do Things with Words in a Study Abroad Context. John Benjamins. Blommaert, J. (1990). Lumumba, Hammarskjoeld, and the 1960 Congo Crisis. Afrika Focus 6(2): 97–118. ——— (1991). How much culture is there in intercultural communication? In J. Blommaert & J. Verschueren (Eds): 13–31. ——— (1994). Ideologien in interkultureller Kommunikation. Sprache und Literatur 25/74: 19–38.

Intercultural communication 199

Blommaert, J. & J. Verschueren (1993). The rhetoric of tolerance, or, what police officers are taught about migrants. Journal of Intercultural Studies 14/1: 49–63. ——— (Eds) (1991). The pragmatics of intercultural and international communication. John Benjamins. Blum-Kulka, S., J. House & G. Kasper (Eds) (1989). Cross-cultural pragmatics. Ablex. Bourdieu, P. (1977). The economics of linguistic exchanges. Social Science Information — Information sur les sciences sociales 16/6: 645–668. Brown, P. & S.C. Levinson (1987). Politeness. Cambridge University Press. Dirim, İ. & P. Auer (2004). Türkisch sprechen nicht nur die Türken: über die Unschärfebeziehung zwischen Sprache und Ethnie in Deutschland. Gruyter. Dirven, R. & G. Pokuta (1994). Intercultural communication: A selective bibliography. Linguistic Agency University of Duisburg (L.A.U.D.), Series B Paper No. 244. Eelen, G. (1993). Authority in international political discourse. Text 13: 29–63. Efron, D. (1972) [1941]. Gesture, race and culture. Mouton. Eksner, H.J. (2006). Ghetto ideologies, youth identities and stylized Turkish German: Turkish youths in Berlin-Kreuzberg. Lit-Verlag. Erickson, F. & J. Shultz (1982). The counselor as gatekeeper. Academic Press. Fabian, J. (1986). Language and colonial power. Cambridge University Press. ——— (1990). Power and performance. University of Wisconsin Press. ——— (1991). Accident and method in the study of intercultural communication. In J. Blommaert & J. Verschueren (Eds): 33–50. Fairclough, N. (1989). Language and power. Longman. Garfinkel, H. (1967). Studies in Ethnomethodology. Prentice Hall. Gavioli, L. & G. Mansfield (Eds) (1990). The PIXI Corpora. Cooperativa Libraria Universitaria Editrice Bologna (CLUEB). Georgiou, M. (2006). Diaspora, Identity and the Media: Diasporic Transnationalism and Mediated Spatialities. Hampton Press. Goffman, E. (1955). On face work. In Interaction ritual. Doubleday Anchor. Goddard, C. (Ed.) (2006). Ethnopragmatics: Understanding Discourse in Cultural Context. Mouton de Gruyter. Goodenough, W. (1964). Cultural anthropology and linguistics. In D.H. Hymes (Ed.), Language in culture and society: 36–39. Harper & Row. Günthner, S. (1993a). Diskursstrategien in der interkulturellen Kommunikation. Niemeyer. ——— (1993b). German-Chinese interactions. Pragmatics 3: 283–304. ——— (2007). Intercultural communication and the relevance of cultural specific repertoires of communicative genres. In: Kotthoff, H. & H. Spencer-Oatey (Eds), Intercultural Communication. (Handbook of Applied Linguistics. Vol. 7) 127–152. Mouton de Gruyter. Gumperz, J.J. (1982a). Discourse strategies. Cambridge University Press. ——— (1992). Contextualization revisited. In P. Auer & A. di Luzio (Eds): 39–53. ——— (Ed.) (1982b). Language and social identity. Cambridge University Press. Hall, E.T. (1959). The silent language. Doubleday. Hall, E.T., W.F. Whyte (1960). Intercultural Communication. Human Organization 19/1: 5–12. Heidelberger Forschungsprojekt ‘Pidgin-Deutsch’ (1975). Sprache und Kommunikation ausländischer Arbeiter. Scriptor. Hewes, G.W. (1974). Gesture Language in Culture Contact. Sign Language Studies 4: 1–34. Hewitt, R. (1986). White talk black talk. Cambridge University Press. Hinnenkamp, V. (1982). ‘Foreigner Talk’ und ‘Tarzanisch’. Buske. ——— (1987). Foreigner talk, code-switching and the concept of trouble. In K. Knapp, W. Enninger & A. Knapp-Potthoff (Ed.): 137–180.

200 Volker Hinnenkamp ——— (1989). Interaktionale Soziolinguistik und Interkulturelle Kommunikation. Niemeyer. ——— (1990). ‘Gastarbeiterlinguistik’ und die Ethnisierung der Gastarbeiter. In E. Dittrich & F.-O. Radtke (Eds), Ethnizität, Wissenschaft und Minderheiten: 277–298. Westdeutscher Verlag. ——— (1991). Talking a person into interethnic distinction. In J. Blommaert & J. Verschueren (Eds): 91–110. ——— (1994). Interkulturelle Kommunikation. Groos. Hinnenkamp, V. (2003). Mixed language varieties of migrant adolescents and the discourse of hybridity. Journal of Multicultural and Multilingual Development 24/1 & 2: 12–41. Holliday, A., Hyde, M. & J. Kullman (2004). Intercultural Communication: An Advanced Resource Book. Routledge. von Humboldt, W. (1830–35). Über die Verschiedenheiten des menschlichen Sprachbaues und ihren Einfluss auf die geistige Entwicklung des Menschengeschlechts. In W. v. Humboldt (1903–1920) Gesamtausgabe der Preußischen Akademie der Wissenschaften, Vol. 7: 1–344. Hymes, D.H. (1974). Foundations in Sociolinguistics. University of Pennsylvania Press. Jakovidou, A. (1994). Funktion und Variation im Foreigner-Talk. Narr. Jordan, B. & N. Fuller (1975). On the non-fatal nature of trouble. Semiotica 13/3: 11–31. Jupp, T.C., C. Roberts & J. Cook-Gumperz (1982). Language and disadvantage. In J.J. Gumperz (Ed.): 232–256. Kandiah, T. (1991). Extenuatory sociolinguistics. Multilingua 10: 345–379. Karim, K.H. (Ed.) (2003). The Media of Diaspora: Mapping the Globe. Routledge. Keim, I. (1978). Gastarbeiterdeutsch. Narr. Kim, Y.Y. (Ed.) (1986). International and intercultural communication annual, Vol. 10. Sage. Knapp, K., W. Enninger & A. Knapp-Potthoff (Eds) (1987). Analyzing intercultural communication. Mouton de Gruyter. Knapp, K. & Ch. Meierkord (Eds) (2002). Lingua Franca Communication. Lang. Koole, T. & J.D. Ten Thije (1994). The construction of intercultural discourse. Rodopi. Kotthoff, H. (1989). Pro und Kontra in fremden Sprachen. Peter Lang. Lado, R. (1957). Linguistics across cultures. University of Michigan Press. Loos, E.F. (1991). Auswahlbibliographie zur interkulturellen Kommunikation. Linguistic Agency University of Duisburg (L.A.U.D.), Series B Paper No. 220. Loveday, L.J. (1982). Communicative interference. IRAL 20: 1–16. Meeuwis, M. (1994a). Nonnative-nonnative intercultural communication. Multilingua 13: 59–82. ——— (Ed.) (1994b). Critical perspectives on intercultural communication. Special issue, Pragmatics 4(3). Meeuwis, M. & S. Sarangi (1994). Perspectives on intercultural communication. In M. Meeuwis (Ed.): 309–313. Oleksy, W. (Ed.) (1989). Contrastive pragmatics. John Benjamins. Otten, M., J. Allwood, M. Assumpta Aneas, D. Busch, D. Hoffman & M. Schweisfurth (2009). Qualita­tive research on intercultural communication. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research FQS 10/1, n° 30. php/fqs/issue/view/30 Porter, R.E. & L. Samovar (1972). Communicating interculturally. In L. Samovar & R.E. Porter (Eds), Approaches to intercultural communication: 4–24. Wadsworth. Rampton, B. (1995). Crossing. Longman. Rehbein, J. (Ed.) (1985). Interkulturelle Kommunikation. Narr. Riley, P. (1989). Well don’t blame me! On the interpretation of pragmatic errors. In W. Oleksy (Ed.): 231–249.

Intercultural communication

Roberts, C., E. Davies & T. Jupp (1993). Language and discrimination. Longman. Roberts, M. (1985). Ethnicity in riposte at a cricket match. Comparative Studies in Society and History 27: 401–429. Roche, J. (1989). Xenolekte. de Gruyter. Sarangi, S. (1994). Intercultural or not? In M. Meeuwis (Ed.): 409–427. Sherzer, J. (1987). A discourse-centered approach to language and culture. American Anthorpologist 89: 295–309. Siegel, J. (1987). Language contact in a plantation environment. Cambridge University Press. Silverstein, M. (1979). Language structure and linguistic ideology. In R. Clyne, W. Hanks & C. Hofbauer (Eds), The elements: 193–247. Chicago Linguistic Society. Stoller, P. (1979). Social interaction and the development of stabilized pidgins. In I.F. Hancock (Ed.), Readings in creole studies: 69–79. Story-Scientia. Streeck, J. (1985). Kulturelle Kodes und ethnische Grenzen. In J. Rehbein (Ed.): 103–120. Tannen, D. (1986). That’s not what I meant. William Morrow & Co. Thompson, E.P. (1993). Customs in common. Penguin. Todorov, T. (1982). La conquête de l’Amerique. Seuil. Trager, G.L. & E.T. Hall (1954). Culture and communication. Explorations: Studies in Culture and Communication 3: 137–149. Tsuda, Y. (1986). Language inequality and distortion in intercultural communication. John Benjamins. Varonis, E. & S.M. Gass (1985). Miscommunication in native/nonnative conversation. Language in Society 14: 327–343. Verschueren, J. (1984). Linguistics and crosscultural communication. Language in Society 13: 489–509. Wagner, J. (1996). Foreign language acquisition through interaction – A critical review of research on conversational adjustments. Journal of Pragmatics 26/2: 215–235. Welsch, W. (1999). Transculturality – the puzzling form of cultures today. In: Featherstone, M & S. Lash (Eds), Spaces of Culture: City, Nation, World: 194–213. Sage. Whorf, B.L. (1956). Language, thought and reality. MIT Press.


Interview Charles Briggs University of California at Barkeley

1.  Introduction Interviewing constitutes one of the most fascinating and most poorly investigated realms of pragmatic inquiry. One reason for its importance to the field is its ubiquity as a means of obtaining information. Practitioners in a broad range of disciplines rely on interviews. The widespread use of interviewing by journalists, providers of social services, physicians, and employers as well as the emphasis that politicians and corporations place on surveys point to the central role that interviews play in creating the institutional structures of modern societies. It is thus remarkable that relatively little research has explored the pragmatic underpinnings and effects of interviewing. This hiatus reflects the common assumption that interviews are relatively simple, straightforward, and well understood. The limited number of works that investigate the pragmatics of interviewing in depth rather point to its discursive complexity and to crucial gaps in our understanding. Practitioners may also be reluctant to subject interviews to too profound a critique in view of their efficacy as means of imbuing social scientific discourses with authority. 2.  The construction of interviews and questionnaires The major mode of structuring interviews discursively is the recursive use of questionanswer pairs, often with follow-up Q-A sequences (generally termed ‘probes’). The participation framework of interviews is likewise organized around a central, asymmetrical opposition: the interviewer asks the questions, the respondent answers them, and the interviewer then signals when s/he considers the response adequate (especially by asking a new question). Formal or structured interviews are pragmatically distinct from informal, unstructured ones. The former involve the use of a predetermined set of questions, and their presentation by an interviewer is standardized as much as possible: questions are to be read as printed and presented in the same order. The standardization of responses may be maximized through the use of closed questions in which the interviewee must choose between preselected alternatives. In survey interviews, professional social scientists write questions that reflect their research interests and chosen methodology. The researchers then hire a staff of interviewers, who are generally not

Interview 203

social scientists, instruct them in how they are to present questions, and assign lists of interviewees that are produced by sampling techniques. Formal interviews are thus structured by an absent party, one who also controls rights to interpret the discourse, a fascinating sort of ventriloquism (see Haraway 1992). The use of written questionnaires that are completed by the respondent standardize the process of obtaining information by dispensing with a face-to-face interview. Informal interviews are generally conducted by researchers themselves. While lists of questions are often prepared in advance, exact wordings and the order of presentation emerge in the course of the interview; as a result, the discourse produced in such settings is often structured more by the social interaction — as guided by the researcher’s interests — than by discursive constraints imposed by predetermined questions. Since the range of possible responses is less constrained, respondents are often invited to use a wider range of discursive forms (such as narratives). Unstructured interviews are generally associated with qualitative research and structured interviews using closed questions with quantitative approaches; the work of Labov (1972a, 1972b) and other sociolinguists suggests, however, that open questions posed during relatively unstructured phases of an interview may also be analyzed quantitatively.

3.  Research on interviewing While sociologists have long taken the lead in focusing attention on the problematics of interviews (see König 1966; Hyman et al. 1954; Manyntz et al. 1969), much of this research has been limited to attempts to identify sources of ‘bias’ or ‘distortion’. Such approaches assume that it is possible for the interviewer to constitute, at least in ideal terms, a means of obtaining the ‘individual true value’ — the ‘real’ or ‘unbiased’ view — of the respondent (see Brenner 1981). Practitioners have also concentrated their efforts on attempts to increase the ‘reliability’ and ‘validity’ of interview data. Reliability refers to the degree of invariance that is achieved when the same procedures are repeated, while validity points to the accuracy of a given technique in measuring the phenomena in question. As Hyman et al. (1954) argued in a classic study, standard interview techniques are oriented much more toward reliability than validity; focusing so intently on reducing inter-interviewer variation creates a strong force for methodological conservatism. Two sociologists who specialize in pragmatic approaches to language, Aaron Cicourel and Allen Grimshaw, offered trenchant critiques in the 1960s and 1970s. Grimshaw (1969, 1969–1970) criticized language ideologies that underlie sociological research in that they picture speech as a transparent vehicle for extracting ideas or attitudes from the mind of a subject and transmitting them onto the printed page. He emphasized the need to envision the role of language in research not only as an obstacle in the investigation of non-linguistic phenomena but as a crucial source of information on social processes.

204 Charles Briggs

Cicourel (1964, 1982a, 1986) conducted in-depth research on the pragmatics of interview discourse. In landmark but, unfortunately, less often cited work on the use of interviews in documenting fertility practices in Argentina, Cicourel (1974) tape-recorded interviews and conducted follow-up interviews with interviewers and interviewees. He thus demonstrated the indexical links between the form and functions of questions and responses and the social and communicative dynamics of the social interactions that constitute interviews. He has also documented the importance of intertextual relations with discourse that emerges in other contexts in shaping the structure and content of medical interviews (Cicourel 1982b, 1992). Once the pragmatic complexity of interview data is recognized, the assumptions that underlie such notions as bias, distortion, reliability, and validity becomes apparent. Mishler (1986) argues that researchers commonly see interview data as behavior that can be analyzed using stimulus-response models associated with scientific experimentation. The pragmatic reductionism of received interview practices also spring from Western ideologies of language (see Kroskrity et al. 1992; Joseph & Taylor 1990) that treat verbal interaction as a transfer of referential content from one party to another, as if participants had no interests or communicative foci that interfere with their playing the roles of interviewer and respondent (see Back & Cross 1982; Clark & Schober 1992; Dijkstra & van der Zouwen 1977, 1982: 3–4; Foddy 1993: 13–14). The information obtained in this manner is seen as a set of stable ‘social facts’ that have an objective existence independent of the linguistic and contextual settings in which they are ‘expressed’ (see Karp & Kendall 1982). Interviewing provides a valuable source of data on the ideologies of language that underlie social scientific research, particularly in that conceptions which have been banished from the realm of explicit theory are often preserved implicitly in ‘purely methodological’ spheres. 4.  Pragmatic underpinnings of interviews The interview format privileges conscious, explicit models of thought, speech, and action; it is thus constrained by limits imposed within a particular communicative system on what can — and may — be formulated in referentially explicit terms. Silverstein (1981) has argued that three characteristics render a given type of discourse maximally susceptible to conscious formulation: unavoidable referentiality, where the referential function is clearly foregrounded; surface segmentability, where the meaning of forms accrues to units that are susceptible to received modes of syntactic and semantic analysis; and relative presupposition, in which indexical relations are grounded in preexisting conditions rather than fashioned creatively in speaking. These features are maximized through standard interview techniques, which privilege the transfer of referential information in highly codable terms; while indexical relations in general are

Interview 205

suppressed (except connections between questions and responses), creative indexes are particularly incompatible with the goal of controlling stimulus-response relations in interviews. This close match between the interview and the features that render speech susceptible to conscious analysis — at least for bourgeois speakers of Indo-European languages — has a number of crucial implications. First, in analyzing interview discourse, researchers generally overlook the role of features that are not referential, surface segmentable, and/or relatively presupposing. Since such features significantly shape the meaning of responses and the ways they relate to questions, the pragmatic reductionism that is built into standard practices for coding interview data renders interpretations of the meaning of responses highly selective and often misleading. Second, the attempt to isolate referential content from the interactional setting of a particular interview motivates efforts to suppress discursive routines for establishing shared meaning. Such prohibitions are institutionalized in instructions to survey interview teams; here conversational routines for repair and clarification, local negotiation of topical shift, and the recipient design of utterances are deemed sources of ‘error’ and ‘bias’ that must be eliminated (see Briggs 1986; Mishler 1986; Suchman & Jordan 1992). Third, invocation of the interview frame selects negatively for communicative events in which the referential function is not foregrounded and which are less surface segmentable and more creative. Since some social forms are much less susceptible to statement in direct, overt, easily codable terms, projects in which interviews constitute the primary or sole mode of data acquisition will present critical problems of sampling. Briggs (1986) accordingly urges researchers to explore the norms that shape the pragmatics of discourse for a given population to determine their compatibility with the metacommunicative norms that underlie interviews. Paredes (1977) provides a poignant example of the way that gaps in metacommunicative norms result in highly flawed research when their presence goes undetected. Finally, several studies suggest that the pragmatics of standard interview techniques are hierarchical in that they grant the interviewer (and, in surveys, the researcher) a great deal of control over the social interaction (see Briggs 1986; Mishler 1984, 1986; Suchman & Jordan 1990). By reducing interview data to highly codable referential forms and by overlooking most indexical relations, researchers render their powerful role in shaping interview data invisible. 5.  Interviewing and institutional authority Recent attempts to incorporate issues of power and political economy into research on discourse suggest means of using research on the pragmatics of interviews in addressing

206 Charles Briggs

broader social and political questions. Foucault (1980) played a key role in shifting the study of discourse from the production of meaning in face-to-face interactions to concerns with relations of power in broader institutional contexts. He argued that compiling social information plays a crucial role in the discursive constitution of institutional power (Foucault 1977). Bauman & Briggs (1990; Briggs & Bauman 1992) argue that a vital part of the process of rendering discourse socially powerful is to gain control over its recontextualization — rights to determine when, where, how, and by whom it will be used in other settings. Interview discourse is maximally configured in terms of both form and content for recontextualization into the sorts of texts that the researcher anticipates creating — interviewees are granted very few rights over this process. In survey research, ‘instruments’ (lists of questions) and techniques for ‘implementing’ them maximize the social control of interviewers by the researchers who direct the study as well of interviewees by interviewers, creating hierarchies of discursive authority that also include individuals responsible for coding data. Such criticisms might seem inapplicable to the open-ended and informal interviews conducted by most ethnographers, since the ‘informant’ plays a greater role in determining the content and generic framing of discourse. Work by Clifford (1988; Clifford & Marcus 1986) and others points, however, to the responsiveness of ethnographic discourse to textual and institutional constraints and its use in constructing ethnographic authority. Researchers have also pointed to the issues raised by the discursive control exercised by interviewers for feminist research (Briggs 1986: 121–123; DeVault 1990; Oakley 1981; Reissman 1987). The ‘reluctance’ of potential interviewees to participate in a study, their ‘failure’ to answer questions in ways deemed acceptable by researchers, and the ‘failure’ of interviewers to ask questions as they have been worded by surveys are generally referred to in such terms as ‘incompetence’ or ‘error’. Such designations naturalize hierarchical relations in research situations and overlook the possibility that such ‘problems’ may in fact emerge from basic contradictions in the pragmatic underpinnings of interviewing, interviewers’ efforts to renegotiate power imbalances (particularly when they are required to ask questions that seem irrelevant or offensive), and interviewees’ attempts to resist the appropriation of control over the recontextualization of their own words. Drawing on sociolinguistic research (esp. Hymes 1974), Bourdieu (1991) argues that forms of communicative competence constitute symbolic capital. The acquisition of communicative competence is constrained by such gate-keeping institutions as schools and professional societies. Interviews are used by members of dominant sectors in furthering their institutionalized needs, such as the compilation of census information and the use of surveying in enhancing consumption or devising political rhetoric. Dominated communities are common targets for interview projects. Interviews currently form a crucial facet of the process of generating notions of the

Interview 207

public sphere (Habermas 1989), dominant discourses that shape cultural and political relations in national and transnational contexts (see Calhoun 1992; Appadurai 1990). Implicit models of interviewing construct notions of public culture in individualist ways in that they view questions and answers as discrete units of information and interviewees as bundles of separable demographic characteristics, behaviors, and attitudes; groups and societies can then be pictured as statistical aggregates (see Back & Cross 1982; Schuman & Presser 1981: 308). Interviews suggest the image of an anonymous citizen representing a segment of ‘the public’ by giving her or his opinions to a stranger who represents an institution, transforming interior states into verbal messages in an unconstrained and rational manner. This trope provides a powerful model of the relationship between ‘public’ and ‘private’, individual and group in bourgeois society. Dominant conceptions of interviewing augment and legitimate the role that researchers play in shaping the categories through which social life is not only perceived but created; by casting surveyors and interviews as expert conveyors of preexisting social facts, such images also help insulate interview techniques and the ideologies that sustain them from radical critique and fundamental change. 6.  Conclusion A number of researchers who rely on formal interview techniques have begun to take studies of the pragmatics of interviewing into account (see Aunger 1994; Foddy 1993; Groves et al. 1992). Nevertheless, these ‘new critics’ of interview techniques (such as Back & Cross 1982; Briggs 1986; Mishler 1986; Suchman & Jordan 1992) are often characterized as arguing that the indexical grounding of interview discourse in particular interactions renders interview data unsuitable for scientific analysis. Such responses assimilate pragmatically-based analyses of interviewing to the very theories of language that they have called into question. Rather than naively attempting to persuade researchers to abandon interviewing, such critiques challenge the native referentialist ideologies of language and individualistic ideologies of culture that underlie the production and interpretation of interview discourse. Pragmatic perspectives suggest that indexical dimensions of the contextualization and recontextualization of interview discourse are no less scientific or predictable than other types of social conduct. Rather than marginalizing awareness of pragmatic dimensions of interviewing by blaming them on the ‘incompetence’ of interviewers and respondents, defenders of interview techniques would be best advised to adopt models that help them discern the central role of discursive and interactional processes. In turn, research on interviewing offers students of pragmatics an extremely fertile ground for exploring new ways of connecting ‘micro’ and ‘macro’ analyses and of investigating the social power of pragmatics.

208 Charles Briggs

References Appadurai, A. (1990). Disjuncture and difference in the global cultural economy. Public Culture 2: 1–24. Aunger, R. (1994). Sources of variation in ethnographic interview data. Ethnology 33: 65–99. Back, K.W. & T.S. Cross (1982). Response effects of role restricted respondent characteristics. In W. Dijkstra & J. van der Zouwen (Eds): 189–207. Bauman, R. & C.L. Briggs (1990). Poetics and performance as critical perspectives on language and social life. Annual Review of Anthropology 19: 59–88. Bourdieu, P. (1991). Language and symbolic power. Harvard University Press. Brenner, M. (1981). Patterns of social structure in the research interview. In M. Brenner (Ed.), Social method and social life: 115–58. Academic Press. Briggs, C.L. (1986). Learning how to ask. Cambridge University Press. Briggs, C.L. & R. Bauman (1992). Genre, intertextuality, and social power. Journal of Linguistic Anthropology 2: 131–172. Calhoun, C. (Ed.) (1992). Habermas and the public sphere. MIT Press. Cicourel, A.V. (1964). Method and measurement in sociology. Free Press. ——— (1974). Theory and method in a study of Argentine fertility. Wiley Interscience. ——— (1982a). Interviews, surveys, and the problem of ecological validity. American Sociologist 17: 11–20. ——— (1982b). Language and belief in a medical setting. In H. Byrnes (Ed.), Contemporary perceptions of language: 48–78. Georgetown University Press. ——— (1986). Elicitation as a problem of discourse. In Sociolinguistics. Walter de Gruyter. ——— (1992). The interpenetration of communicative contexts. In A. Duranti & C. Goodwin (Eds), Rethinking context: 291–310. Cambridge University Press. Clark, H.H. & M.F. Schober (1992). Asking questions and influencing answers. In J.M. Tanur (Ed.): 15–48. Clifford, J. (1988). The predicament of culture. Harvard University Press. Clifford, J. & G.E. Marcus (Eds) (1986). Writing culture. University of California Press. Devault, M.L. (1990). Talking and listening from women’s standpoint. Social Problems 37: 96–116. Dijkstra, W. & J. Van Der Zouwen (1977). Testing auxiliary hypotheses behind the interview. Annals of Systems Research 6: 49–63. ——— (Eds) (1982). Response behavior in the survey-interview. Academic Press. Foddy, W. (1993). Constructing questions for interviews and questionnaires. Cambridge University Press. Foucault, M. (1977). Discipline and punish. Vintage. ——— (1980). Power/knowledge. Pantheon. Grimshaw, A.D. (1969). Language as data and as obstacle in sociological research. Items 23: 17–21. ——— (1969–1970). Some problematic aspects of communication in cross-racial research in the United States. Sociological Focus 3: 67–85. Groves, R. et al. (1992). Direct questioning about comprehension in a survey setting. In J.M. Tanur (Ed.): 49–61. Habermas, J. (1989). The structural transformation of the public sphere. MIT Press. Haraway, D. (1992). Promises of monsters. In L. Grossberg et al. (Eds), Cultural studies: 295–337. Routledge. Hyman, H. et al. (1954). Interviewing in social research. University of Chicago Press. Hymes, D. (1974). Foundations in sociolinguistics. University of Pennsylvania Press.

Interview 209

Joseph, J.E. & T.J. Taylor (Eds) (1990). Ideologies of language. Routledge. Karp, I. & M.B. Kendall (1982). Reflexivity in field work. In P.G. Secord (Ed.), Explaining social behavior: 249–73. Sage. König, R. (1966). Das Interview. Kiepenheuer & Witsch. Kroskrity, P., B. Schieffelin & K. Woolard (Eds) (1992). Language ideologies. Special issue of Pragmatics 2(3). Labov, W. (1972a). Sociolinguistic patterns. University of Pennsylvania Press. ——— (1972b). Some principles of linguistic methodology. Language in Society 1: 97–120. Manyntz, R. et al. (1969). Introduction to empirical sociology. Penguin. Mishler, E.G. (1984). The discourse of medicine. Ablex. ——— (1986). Research interviewing. Harvard University Press. Oakley, A. (1981). Interviewing women. In H. Roberts (Ed.), Doing feminist research: 30–61. Routledge & Kegan Paul. Paredes, A. (1977). On ethnographic work among minority groups. New Scholar 7: 1–32. Reissman, C.K. (1987). When gender is not enough. Gender and Society 1: 172–207. Schuman, H. & S. Presser (1981). Questions and answers in attitude surveys. Academic Press. Silverstein, M. (1981). The limits of awareness. Sociolinguistic Working Paper 84. Suchman, L. & B. Jordan (1990). Interactional troubles in face-to-face survey interviews. Journal of the American Statistical Association 85: 232–253. Tanur, J.M. (Ed.) (1992). Questions about questions. Russell Sage Foundation.

Bronislaw Kasper Malinowski Gunter Senft Max Planck Institute for Psycholinguistics, Nijmegen

1.  Biographical sketch “…he had an artist’s power to create with great integrative capacity a world of his own … and he had the true scientist’s intuitive discrimination between relevant and adventitious fundamental and secondary issues”, this kind epitaph, which Malinowski formulated in his obituary for Sir James George Frazer a year before he himself died, could equally apply to Malinowski, as Raymond Firth (1981: 137) so rightly emphasized in one of his articles on his teacher and colleague. Bronislaw Malinowski, one of the most important anthropologists of the 20th century, is generally recognized as one of the founders of social anthropology, transforming 19th century speculative anthropology into a field-oriented science that is based on empirical research. Malinowski is principally associated with his field research of the Mailu and especially of the Trobriand Islanders in what is now Papua New Guinea, and his masterpieces on Trobriand ethnography continue “to enthrall each generation of anthropologists through its intensity, rich detail, and penetrating revelations” (Weiner 1987: xiv). Bronislaw Kasper Malinowski (nicknamed Bronek & Bronio) was born in Cracow (then Austrian Galicia, now Poland) on 7 April 1884 as the only child of Jozefa (née Lacka) and Lucjan Malinowski. His father (1839–1898) was professor of Slavonic philology at the Jagiellonian University of Cracow and was well known not only as a philologist but also as an ethnographer specialized in Polish dialects and Silesian folklore and ethnology. He died of a heart attack at the age of 58 when his son (who was to die in the same way at the same age) was only 14. However, his mother Jozefa (‘Josephine’, 1848–1918), who came from a wealthy land-owning family and was a highly cultured woman and a good linguist, was much more important for Bronislaw’s development and education (see Wayne 1985, 1995), and the mother-son bond was extremely strong. Both parents belonged to a social class that Malinowski’s youngest daughter Helena Wayne (1985: 529) characterized as being something “between landed gentry and nobility, but certainly not aristocracy”. Bronislaw was a delicate child and had constant problems with his health: he nearly died of peritonitis and he had severe trouble with his eyes — he was even threatened with blindness. Therefore, after a year as an internal student he became an external student of the Jan Sobieski Gimnajum, one of Cracow’s best secondary schools. He worked at home and with the loving and caring help of his mother he managed to

Bronislaw Kasper Malinowski

brilliantly complete his schooling. Because of his health Malinowski and his mother — at the behest of his doctors — made various journeys south. They were already well travelled in central European countries, but in the time between 1899 and 1906 they went to Northern Italy, the Mediterranean, especially to the Dalmatian coast, Malta, Sicily, to North African countries, and to Madeira and the Canary Islands (Wayne 1985: 531). Nevertheless, his poor health continued to plague him throughout his life. In 1902 Malinowski went to Cracow University and studied first physics, mathematics and chemistry, and then philosophy and also psychology. In 1908 he was awarded his doctorate with the highest honours in the Austrian Empire (summa cum laude — sub auspiciis imperatoris) and was presented with a large gold and diamond ring from Kaiser Franz Josef (see Wayne 1985: 531). His Ph.D. thesis (On the principle of the economy of thought) examined the ‘second positivism’ of Richard Avenarius & Ernst Mach (see Paluch 1981: 284; Young 1987: 125); the doctorate was mainly a philosophical study with physics and mathematics as subsidiaries. After his doctoral work, Malinowski developed his interest in anthropology and ethnology. His health problem forced him to postpone the studies he had planned to undertake after his graduation and he started to read Frazer’s The Golden Bow (at this time only the first 3 of the 12 volumes had been published). Malinowski then went for three terms (‘semester’) to the University of Leipzig and studied ‘Völkerpsychologie’ with Wilhelm Wundt and historical economy and economics with Karl Bücher. He also worked in the chemistry laboratories of Wilhelm Ostwald. In 1910 Malinowski went to London with the South African pianist Annie Brunton whom he met in Leipzig, and he became a postgraduate student at the London School of Economics and Political Science (LSE). Here anthropology had been recently established as a discipline. Malinowski studied anthropology with Charles Gabriel Seligman (1873–1940) and sociology with Edward A. Westermarck (1862–1939). The influence of these great scholars on Malinowski’s thinking is excellently outlined in Weiner (1987). In 1913 Malinowski wrote his first book in English, The Family among the Australian Aborigines, as one part of his doctoral requirements. This monograph — like his second book, Primitive Religion and Forms of Social Structure, that was published in Polish (and in Poland) in 1915 — was written on the basis of published accounts, but he had already conceived plans to do anthropological field research. In 1914 Seligman managed to raise funds from the LSE and from the industrialist Robert Mond for Malinowski to do field research in the Western Pacific. Malinowski “went out to Australia with the British Association for the Advancement of Science… as a guest, and at the expense, of the Commonwealth Government of Australia” (Malinowski 1922: xix). On his arrival in Australia, war had been declared in Europe. As a Pole, Malinowski was a subject of the Austrian Emperor and thus an enemy alien. He nevertheless managed to get permission from the Australian authorities to proceed with his research. In September 1914 Malinowski sailed from Brisbane to British New



Gunter Senft

Guinea, after his friend Stanislaw I. Witkiewicz, who had planned to become his field work photographer, had left him and returned to Europe to fight for the Russians. He stayed for six months and — following Seligman’s advice — did about three months of field research with the Mailu on the south coast. In March of 1915 Malinowski returned to Australia after a brief trip to Samarai and Woodlark Island. He now wrote the monograph The Natives of Mailu: Preliminary Results of the Robert Mond Research Work in British New Guinea, published as Volume 39 of the Transactions and Proceedings of the Royal Society of South Australia in Adelaide in 1920 (Malinowski 1988). Together with his research on the Australian Aboriginal family, published in 1913, it earned him a doctor of science degree from the University of London in 1916. In May 1915 Malinowski left Australia for his second expedition to British New Guinea. Together with Seligman he had been discussing a possible fieldsite for this second field trip, and among the candidates were the Binandere people in the Mambare district, the Borowai, and especially the Rossel Islanders. Malinowski’s letters to Seligman reveal that he had the intention to make for the Mambare, but that he wanted to visit the Trobriands and stay there for a month on his way to this projected field site. It was R.L. Bellamy, the Assistant Resident Magistrate and Medical Officer of the district, who attracted Malinowski to the Trobriands. On his arrival on the Trobriands in July, Malinowski stayed with Bellamy, who taught him some basics of Kilivila (also: Kiriwina, Boyowa), the Austronesian language of the islanders (Senft 1986). Bellamy left the Trobriands after a month for the War, but Malinowski decided to stay there and he chose Omarakana, the village of the paramount chief To’uluwa, as the place to set up his tent. Already in September 1915 he mastered the language so well that he did not need the help of an interpreter any more (besides his mother tongue Polish he could also speak Russian, German, French, English, Italian, and Spanish). Malinowski stayed on Kiriwina, the largest of the Trobriand Islands, for nine months, then returned to Melbourne at the end of March 1916 (see Young 1984: 20) and started to write his article Baloma — the Spirits of the Dead in the Trobriand Islands (Malinowski 1974: 149–254). He managed to get a permit to continue his research on the Trobriands for another year, but he was not allowed to go anywhere else and thus had to give up his plans to briefly visit Rossel Island after his second field trip to the Trobriands. Seligman in England and Spencer in Australia managed to raise further funds for his research (from 1914 to 1920 Malinowski had a budget of 250 pounds a year), and Malinowski left Australia again in October 1917 for the Trobriands where he stayed until October 1918 (see Young 1984). It was during the second period of field research on the Trobriands, in January 1918, that Malinowski’s mother died. Back in Melbourne he continued working on his Trobriand material together with Elsie Rosaline Masson. Malinowski had met Elsie in 1916, and they married in March 1919. They had three daughters, Josefa Maria, Wanda, and Helen. A year after

Bronislaw Kasper Malinowski

their marriage the couple left Australia for good. They first spent some months in England, then they moved to the Canary Islands and lived for a year in a country villa in Tenerife. Here Malinowski finished his first monograph on the Trobriand Islands, The Argonauts of the Western Pacific, which was accepted by Routledge within three days and published in 1922. This book made Malinowski’s reputation. In 1921 Malinowski started to give lectures at the LSE. In 1924 he was appointed to a readership in anthropology, and in 1927 he accepted the offer to become the first chair in anthropology at the LSE. Between 1921 and 1927 the Malinowski family travelled to the south of France, to Poland and to the south of Tyrol — where in 1923 they bought a beautiful Alpine house in Oberbozen (Soprabolzano) that was to become the home of the family for 6 years. During this time Malinowski continuously commuted between London and Oberbozen. He was a brilliant teacher, attracted students from many disciplines and trained a generation of distinguished British anthropologists. Among his students were Raymond Firth, E.-E. Evans Pritchard, Isaac Schapera, Hortense Powdermaker, Edmund R. Leach, Meyer Fortes, Lucy Mair, and Ian Hogbin (see also Firth 1957b). The Director of the LSE, Sir William Beveridge, urged Malinowski to live in London — and in 1929 the family moved to a house in Primrose Hill. Moreover, following the advice of Beveridge, Malinowski also became a British subject. He remained in London for almost twenty years, but he continued to travel widely. In 1926 Malinowski was invited to the United States and spent 6 months there as a guest of the Rockefeller Foundation. He visited Ruth Benedict at Columbia and gave a summer course on anthropology at the University of California, Berkeley. He also visited the Hopi Indians and travelled to Mexico for the first time. The years between 1923 and 1938 were his most productive years as a writer and a teacher. Besides many essays and shorter theoretical works Malinowski published the other two of his three major monographs on the Trobriands, The Sexual Life of Savages in Northwestern Melanesia in 1929 and the two volumes of Coral Gardens and their Magic in 1935. Among his shorter essays on aspects of Trobriand ethnography Magic, Science and Religion (1925), Crime and Custom in Primitive Society (1926), and Myth in Primitive Psychology (1926) deserve special mention. With respect to his interdisciplinary research in anthropology and psychology his publications titled The Psychology of Sex (1923), Psychoanalysis and Anthropology (1924), and The Father in Primitive Psychology (1927) have to be mentioned as well. In her article on Malinowski, Rhoda Métraux (1968) lists 70 publications (see also Ellen et al. 1988: 210–227; Firth 1957b: 265–271). In 1930 Malinowski developed an interest in Africa. He travelled for four months through South and East Africa, visiting his students (Audrey Richards, the Wilsons, and Hilda Beemer) who were working on the Bemba, the Swazi and on other tribes. Among his African students was Jomo Kenyatta, who prepared his


214 Gunter Senft

book Facing Mount Kenya (published with an introduction by Malinowski) as a diploma thesis under his supervision. In 1933 Malinowski made his second visit to the USA, presenting the Messenger lectures at Cornell University. Two years later, in September 1935, his wife Elsie died. In 1936 Malinowski was awarded the honorary doctor of science degree from Harvard University, and in 1938 he spent his sabbatical leave in the USA. In September 1939 the director of the LSE, Sir A.M. Carr-Saunders, advised Malinowski to stay in the States because of the unclear wartime future of his university. Malinowski followed this advice and brought his daughters to America. In the same year he became a Bishop Museum Visiting Professor of Anthropology at Yale (where the brilliant teacher had difficulties with students who — according to Métraux (1968: 546) — “were far less mature than his students in London”). In 1939 he also married Valetta Swann (née Hayman-Joyce), an English painter 20 years his junior. Together with his second wife and the Mexican anthropologist Julio de la Fuente he spent the summers of 1940 and 1941 in Mexico doing fieldwork studying the Zapotec Indians’ peasant market in Oaxaca. Early in 1942 Malinowski was appointed professor of anthropology at Yale, but on the 16th of May 1942 he suffered a heart attack and died in New Haven, Connecticut.

2.  The study of culture As Métraux (1968: 541) pointed out so pertinently, “Malinowski’s primary scientific interest was in the study of culture as a universal phenomenon and in the development of a methodological framework that would permit the systematic study of specific cultures in all their peculiarities and open the way to systematic cross-cultural comparison”. Central and recurrent themes in his research were the following topics: the family, kinship, culture change, anthropology and psychology, the integrity of culture; the complex interrelationship of the society, the culture and the individual, and the systematic nature of culture. In his 1931 article on “Culture” in the Encyclopedia of the Social Sciences Malinowski states that for him “culture comprises inherited artefacts, goods, technical processes, ideas, habits, and values”. Moreover, besides social organization the concept of culture also includes “the set of forces impinging on the individual born into each society” (Richards 1957: 21). For Malinowski this idea of the “… ‘social heritage’ is the key concept of cultural anthropology” (Malinowski 1931: 621). He was convinced that human beings have biological needs that culture satisfies, that culture is an instrumental reality derived from human needs: “…rites, beliefs, and customs, however extraordinary they appear to an observer, actually fill ‘needs’, biological, psychological, and social” (Richards 1957: 18). Therefore anthropology is the science that has the task to “study the ‘use’ or ‘function’ of the customs, institutions, and beliefs which formed part of each culture” (Richards 1957: 16). Each culture represents for him a closed system, and therefore all cultures are comparable.

Bronislaw Kasper Malinowski

On the basis of these ideas he developed his functional theory of needs “in which each basic human biological need triggers a cultural response” (Weiner 1987: xxx). Functionalism “aims at the explanation of anthropological facts at all levels of development by their function, by the part which they play within the integral system of culture, by the manner in which they are related to each other within the system, and by the manner in which this system is related to the physical surroundings. It aims at an understanding of the nature of culture, rather than at conjectural reconstructions of its evolution or of past historical events…”. Moreover, functional theory “… insists… upon the principle that in every type of civilisation, every custom, material object, idea and belief fulfils some vital function, has some task to accomplish, represents an indispensable part within a working whole” (Malinowski 1926: 132, 133). Understanding a culture therefore presupposes the understanding of such functions (for a synthesis of his ideas on functional theory see Malinowski 1944; see also Young 1987: 132ff). However, as Métraux (1968: 541) points out, although “the idea of ‘function’ is a key concept throughout his work … his use of the term was open-ended, exploratory, and subject to continual modification”. His theoretical thinking was very much influenced by Westermarck and Seligman, of course, but also by Wilhelm Wundt, Karl Bücher, James Frazer, Alfred Haddon, William Rivers and R.R. Marrett, by the French sociological school, especially by Emile Durkheim and Marcel Mauss (though he did not like their abstract notions of society), by Richard Thurnwald, by Richard Gregory, Havelock Ellis, A.H. Gardiner, Julian Huxley, C.K. Ogden, Cyril Burt, S.S. Myers, J.C. Flugel, W. Powys Mathers, G.H.L.F. Pitt-Rivers, and J.H. Oldham (see Firth 1957a). He reacted strongly against the speculations of evolutionists and diffusionists like Lewis Henry Morgan, Herbert Spencer, Edward Burnett Taylor, Fritz Graebner, Wilhelm Schmidt and other representatives of the Kulturhistorische Schule and their Kulturkreislehre, against Lévy-Bruhl’s theory of primitive mentality, and against Freud’s theory of psychoanalysis. On the basis of his theoretical ideas about culture he studied a broad range of cultural aspects and institutions — mainly, if not almost exclusively, with respect to the culture of the Trobriand Islanders — and he challenged in, and with, his work many theories on core-concepts of anthropology such as kinship, marriage, exchange, and ritual. For Malinowski, functionalism was a research tool, “the prerequisite for field-work and for the comparative analysis of phenomena in various cultures” (Malinowski 1944: 175), and therefore his theory had to include a general theory of how to do fieldwork.

3.  Fieldwork Malinowski was not the first anthropologist who did intensive field research — Lewis H. Morgan studied the Iroquois, Franz Boas the Kwakiutl, Carl Strehlow as well as Spencer and Gillen Australian Aborigines in the field — but through “his example and teaching, fieldwork became the ‘constitutive experience’ of anthropology, the


216 Gunter Senft

‘central rite of the tribe’ ” (Young 1987: 124). As Leach (1957: 120) points out so aptly, Malinowski, the ‘fanatical theoretical empiricist’, developed a field technique that was unique because of the “severely curtailed use of the professional informant” and because of “the theoretical assumption that the total field of data under the observation of the field worker must somehow fit together and make sense”. He was convinced that observation without theory is impossible. However, he also insisted on the principle that theory can only be falsified or verified on the basis of the observed and that the observed will always lead to a refinement or to a reformulation of basic assumptions in the field researcher’s theory. Already in his 1916 article Baloma — the Spirits of the Dead in the Trobriand Islands Malinowski (1974: 237f.) emphasized that the traditional gathering of ‘pure facts’ in the field is ‘pure ‘collectioneering’ of data’ if it is not accompanied by “the interpretation which sees in the endless diversity of facts general laws; which severs the essential from the irrelevant; which classifies and orders phenomena, and puts them into mutual relationship”. Thus, “field work consists only and exclusively in the interpretation of the chaotic social reality, in subordinating it to general rules”. In his introduction to the Argonauts of the Western Pacific he clearly formulates the basic lines of his approach to field research and the final goal of an ethnographer: …the goal of ethnographic fieldwork must be approached through three avenues: 1. The organisation of the tribe, and the anatomy of its culture must be recorded in firm clear outline. The method of concrete statistical documentation is the means through which such an outline has to be given. 2. Within this frame, the imponderabilia of actual life, and the type of behaviour have to be filled in. They have to be collected through minute, detailed observations, in the form of some sort of ethnographic diary, made possible by close contact with native life. 3. A collection of ethnographic statements, characteristic narratives, typical utterances, items of folklore and magical formulae has to be given as a corpus inscriptionum, as documents of native mentality. These three lines of approach lead to the final goal, of which an Ethnographer should never lose sight. This goal is, briefly, to grasp the native’s point of view, his relation to life, to realise his vision of his world. We have to study man, and we must study what concerns him most intimately, that is, the hold which life has on him…. In each culture we find different institutions… To study the institutions, customs, and codes or to study the behaviour and mentality without the subjective desire of feeling by what these people live, of realising the substance of their happiness — is, in my opinion, to miss the greatest reward which we can hope to obtain from the study of man. … Perhaps as we read the account of these remote customs there may emerge a feeling of solidarity with the endeavours and ambitions of these natives. Perhaps man’s mentality will be revealed to us, and brought near, along some lines which we never have

Bronislaw Kasper Malinowski

followed before. Perhaps through realising human nature in a shape very distant and foreign to us, we shall have some light shed on our own. (Malinowski 1922: 24f.) Thus, Malinowski introduced the concept and the method of “participant observation” into anthropology, being convinced that “alien cultures had to be explored ‘from the inside’ to make most sense” (Young 1987: 131). That this ambitious concept necessarily puts field researchers in a position where they have to face the strains of field research is very explicitly and incredibly frankly documented in Malinowski’s posthumously published Diary (Malinowski 1967), which Raymond Firth in his new introduction to the 1989 edition (re-)evaluates as follows: “It is not merely a record of the thinking and feeling of a brilliant, turbulent personality …. it is also a highly significant contribution to the understanding of the position and role of a fieldworker as a conscious participator in a dynamic social situation” (Firth 1989: xxxi). Moreover, besides the anthropologist’s role as a “participant observer”, the ethnographer’s linguistic competence in, and competent use of, the native language is for Malinowski another and an equally important basic requirement to fulfil the anthropologist’s task “to give a full description of language as an aspect and ingredient of culture” and “to translate the native point of view to the European” (Malinowski 1935, vol. II: xxf.). 4.  Theory of language Malinowski became very much interested in linguistics when he found that he could not realize his project of writing a grammar of Kilivila because he had no linguistic training and because he was — rightly — convinced that the grammatical categories offered by the linguistic theories of his time did not fit for the description of a language like Kilivila (Malinowski 1920: 74, see also Senft 1994a). As early as 1920 he explicitly stated the following: …there is an urgent need for an Ethno-linguistic theory, a theory for the guidance of linguistic research to be done among natives and in connexion with ethnographic study… A Theory which, moreover, aims not at hypothetical constructions — ‘origins’, ‘historical developments’, ‘cultural transferences,’ and similar speculations — but a theory concerned with the intrinsic relation of facts. A theory which in linguistics would show us what is essential in language and what therefore must remain the same throughout the whole range of linguistic varieties; how linguistic forms are influenced by physiological, mental, social, and other cultural elements; what is the real nature of Meaning and Form, and how they correspond; a theory which, in fine, would give us a set of well-founded plastic definitions of grammatical concepts. (Malinowski 1920: 69)

Besides coining the term ‘ethnolinguistics’, Malinowski emphasizes in his first ‘linguistic’ paper that “grammar can be studied only in conjunction with meaning, and meaning only in the context of situation” (Nerlich & Clarke 1996: 320).


218 Gunter Senft

Malinowski’s linguistic interests “centered on language as a mode of behavior and on problems of culturally determined meaning” (Métraux1968: 524). The second volume of Coral gardens and their magic (Malinowski 1935) presents his “ethnographic theory of language”. Before this he published two linguistic papers, one on Classificatory particles in the language of Kiriwina (Malinowski 1920; see Senft 1996: 200f.) and one on The problem of meaning in primitive languages (Malinowski 1923; see Senft 1995b). Malinowski developed his ethnographic theory of language mainly in connection with his attempts to translate the Trobriand Islanders’ magical formulae. He realized that the Trobriand Islanders believed in the power of the words in the magical formulae: they used these magical formulae to reach certain aims with the firm conviction that they can thus influence and control nature and the course of, and events in, their lives. Thus, in the domain of magic language is doing something, it has certain effects, it has power and force. Malinowski (1922: 432) summarized this observation as follows: “Magic is … an instrument serving special purposes, intended for the excercise of man’s specific power over things, and its meaning, giving this word a wider sense, can be understood only in correlation to this aim”. As Nerlich & Clarke (1996: 321) rightly infer, Malinowski explicitly equates here meaning with pragmatic function — and this is typical for his way of looking at language functionally and contextually with semantics as the starting point for linguistic analyses. He characterized his — pragmatic — theory of meaning as a theory that insists on the “linking up of ethnographic descriptions with linguistic analysis which provides language with its cultural context and culture with its linguistic interpretation. Within this latter … [Malinowski has] … continually striven to link up grammar with the context of situation and with the context of culture” (Malinowski 1935: 73). Malinowski was influenced by the work of the German linguist Philipp Wegener (Wegener 1885, see also Nerlich & Clarke 1996: 318) and familiar with the works of other linguists like Humboldt, Lazarus, Meinhof, Müller, Jespersen, Paul, Steinthal, Tregear, Wundt, Oertl, Temple, and Tucker (see Malinowski 1920: 71f., 74f.). On the basis of this linguistic background speech is for Malinowski first and foremost part of the context of situation in which it is produced, language — in its primitive function — has an essentially pragmatic character (Malinowski used the term ‘pragmatic’ himself, see, e.g., Malinowski 1935: 45), and “meaning resides in the pragmatic function of an utterance” (Baumann 1992: 147). For Malinowski (as well as for Wittgenstein) the meaning of a word lies in its use. Thus, to study meaning one cannot examine isolated words but sentences or utterances in their situative context: “the real understanding of words is always ultimately derived from active experience of those aspects of reality to which the words belong” (Malinowski 1935: 58). Malinowski emphasizes that language — at least in its primitive function — has to be regarded as a mode of action (Malinowski 1923: 296; see also Firth 1957: 94); and that to understand the use of a complex speech situation requires the understanding of the situation in which it occurred and the action it accomplished.

Bronislaw Kasper Malinowski

This position can certainly be described as a “radical functionalism and contextualism” (Nerlich & Clarke 1996: 323). Malinowski (1923: 296, 309ff) illustrates how the meaning of utterances can be determined in what he calls “the essential primitive uses of speech: speech in action, ritual handling of words, the narrative, ‘phatic communion’ (speech in social intercourse)”; he emphasizes his main position as follows: “language in its primitive function and original form has an essentially pragmatic character; … it is a mode of behaviour, an indispensable element of concerted human action … to regard it as a means for the embodiment or expression of thought is to take a one-sided view of one of its most derivate and specialized functions” (Malinowski 1923: 316; see also Firth 1957: 94; Langendoen 1968: 21ff). Moreover, Malinowski is convinced that language “serves for definite purposes, that it functions as an instrument used for and adapted to a definite aim”. Malinowski exemplifies the essentially pragmatic character of language by referring to two situations from his Trobriand experience — a fishing expedition (Malinowski 1923: 310–312) and the verbal guiding of a boat into a reef channel in complete darkness (Malinowski 1935: 58f.) — in which he noted that “words have to be uttered with impeccable correctness and understood in absolutely adequate manner in … situations where speech is an indispensable adjunct to action” (Malinowski: 1935: 58). Malinowski sums up his analysis of the linguistic actions he observed during the fishing expedition as follows: All the language used during such a pursuit is full of technical terms, short references to surroundings, rapid indications of change — all based on customary types of behaviour, well-known to the participants from personal experience. Each utterance is essentially bound up with the context of situation and with the aim of the pursuit, whether it be the short indications about the movements of the quarry, or references to statements about the surroundings, or the expression of feeling and passion inexorably bound up with behaviour, or words of command, or correlation of action. The structure of all this linguistic material is inextricably mixed up with, and dependent upon, the course of the activity in which the utterances are embedded. The vocabulary, the meaning of the particular words used in their characteristic technicality is no less subordinate to action. For technical language, in matters of practical pursuit, acquires its meaning only through personal participation in this type of pursuit. It has to be learned, not through reflection but through action. … The study of any form of speech in connection with vital work would reveal the same grammatical and lexical peculiarities: the dependence of the meaning of each word upon practical experience, and of the structure of each utterance upon the momentary situation in which it is spoken. Thus the consideration of linguistic uses associated with any practical pursuit, leads us to the conclusion that language in its primitive forms ought to be regarded and studied against the background of human activities and as a mode of behaviour in practical matters. (Malinowski 1923: 311f.)

It is obvious that Malinowski here emphasizes and stresses “action at the expense of structure and system” (Nerlich & Clarke 1996: 333). He even argues further that this


220 Gunter Senft

“adaptation, this correlation between language and the uses to which it is put, has left its traces in linguistic structure” (Malinowski 1923: 327). Therefore, for Malinowski “the categories of universal grammar are reflections of universal human attitudes toward life and are brought out by the universally found conditions under which children grow up in the world” (Langendoen 1968: 27). Thus, these “categories of universal grammar must underlie categorizations implicit in nonlinguistic human behavior” (Langendoen 1968: 36). In the second volume of Coral gardens and their magic Malinowski developed the central idea of his theory, namely “that the meaning of utterances is provided by the context of concurrent human activity” (Langendoen 1968: 30). He once more points out that “the real linguistic fact is the full utterance within its context of situation” (Malinowski 1935: 11). And he emphasizes “that the context of situation may enable one to ‘disambiguate’ sentences that are semantically ambiguous” (Langendoen 1968: 32; see Malinowski 1935: 32). Langendoen presents a rather fair and competent criticism of Malinowski’s linguistic views. However, he seems to underestimate the importance of what J.R. Firth (1957: 118) referred to as Malinowski’s “outstanding contribution to linguistics”, namely “his approach in terms of his general theory of speech functions in contexts of situation, to the problem of meaning in exotic languages and even in our own”. Malinowski certainly had a major impact on English linguistics in the first half of our century. And within linguistics, anthropology and anthropological linguistics some of Malinowski’s ideas about language continue to be thought-provoking, and — with explicit reference to Malinowski — social scientists have started ‘rethinking context’ (Duranti & Goodwin 1992). 5.  An appraisal Malinowski’s work and his theory of language and culture has been amply criticized and discussed (Firth 1957b, Weiner 1987, Kohl 1987, Agar 1994, Nerlich, Clarke 1996: 317–335). In what follows I will give an assessment and appraisal of the master of Trobriand ethnography on the basis of my own field research on the Trobriand Islands. I cannot but completely agree with Michael Young (1987: 138) that Malinowski “was an incomparable fieldworker and master ethnographer”. The only reliable linguistic data I found in the literature preparing for my first 17 months of field research on the Trobriands in 1982/83 came from Malinowski’s linguistic publications and from his anthropological linguistic remarks in his ethnographic masterpieces on the Trobriand Islanders. Bits and pieces of Kilivila linguistics that I found in Capell, Lithgow, and Greenberg turned out to be either utterly wrong or extremely speculative (see Senft 1991: 27, 46). Moreover, I had the quite romantic feeling when I first set foot on the Trobriands in 1982 that it was like stepping right into the picture so vividly

Bronislaw Kasper Malinowski

presented in Malinowski’s Trobriand ethnography (Senft 1992: 68). I could easily verify major aspects of his exceptionally thorough ethnographic description of Trobriand culture in my own experience as a participant observer. Although the Trobriand culture — as well as the Kilivila language, of course — have been affected by numerous changes since Malinowski worked and lived there (see Senft 1992, 1997a, b), I have always found it worth my while to check my observations and insights with Malinowski’s publications. There are two excellent ethnographic restudies on the Trobriands. First of all Harry A. Powell’s An Analysis of Present Day Social Structure in the Trobriand Islands, his 1957 Ph.D-thesis, “is envisaged as supplementary to Malinowski’s published data” and presents “a theoretical interpretation of Trobriand kinship and marriage relations different from that developed by Malinowski” (Powell 1957: Abstract of thesis). Second, the results from the 1978 “Kula and Massim Exchange Conference” in Cambridge published by Leach & Leach (1983) present an excellent reassessment of Malinowski’s classic study of 1922. One of the few ethnographic niches Malinowski left, the ‘female world’ on the Trobriands, was congenially filled by Annette Weiner (1976, 1988). It is extremely difficult to find other such niches (see e.g., Senft & Senft 1986; Senft 1994b: 65f) with respect to Trobriand ethnography (but not, of course, with respect to linguistic research on Kilivila). Moreover, I found it equally difficult to falsify major aspects of his ethnography. However, there is one — hotly debated — topic in Malinowski’s description of Trobriand sexuality where I am convinced that Malinowski either made a gross mistake or played ‘career politics’ (or even took his peer group for a ride?) — I am referring to the controversy over Trobriand “virgin birth”. In 1983 my wife stayed with me for 11 months on the Trobriands. We then had no children, and after my wife had learnt to speak Kilivila, the women started to discuss contraception with her. The women of Tauwema told my wife that they had two means of contraception that are both based on a mixture of herbs that grow in the bush; to this mixture a little bit of water is added. Some women but also some men know how to prepare this fluid based on the herbal composition. Once the women had talked about this topic, I could easily verify this information with two of my male informants, namely with Weyei, the weather-magician of Tauwema (see Senft 1985a, 1997b), and Vapalaguyau, who both were very proud of the expertise which they had inherited from their ancestors; however, they did not want to show me how and with which herbs they produce this mixture (I respected their reservations and did not urge them for further information about something that is as personal and secret as magic). Anyhow, there are two modes of application for this contraceptive. Either, before the coitus the woman drips the fluid on a small sponge and then inserts it into her vagina placing it in front of her os uteri — the herbal composition is spermatocidal and thus prevents contraception. Or, the woman drinks the herbal composition in a more hydrous solution. The problem with this second mode of application


222 Gunter Senft

is that the ratio between the herbal mixture and the water is rather delicate: the contraceptive effect of the drink may either last for a few days only or for years — and if the herbal mixture is too highly concentrated it may even cause sterility. Such a long term effect of the contraceptive almost endangered the marriage of a loving couple — however, when we returned to the Trobriands in 1989 (with our then 2 and 4 year old children), the woman who six years ago poured out her troubles to me and especially to my wife proudly presented us her two children. (I would like to note here that the “yam or Dioscorea” — the most important part of the Trobriand diet — “was long known by certain Mexican Indians to have a contraceptive effect. In 1993 Dr. Russell Marker … determined the molecular structure of diosgenin, a stereoid substance with progesteronic effect derived from the yam root. Based on this information, Organon, a leading producer of contraceptive pills, uses the diosgenin from Mexican yam roots as the raw material for some of its products” (de Revai: 1992) — but this is just an aside). The fact that the Trobriand Islanders know about natural contraceptives and that this knowledge is traditional is — to my mind — a clear and convincing counter-argument against Malinowski’s claim which he first made in his very first publication on the Trobriands in 1916, Baloma; the Spirits of the Dead in the Trobriand Islands (Malinowski 1974: 220–237), which he used like a beat of a drum to introduce his “sex book” (as he himself and his first wife called it, see Weiner 1987: xxxii), and which he elaborates in Chapter 7 of The Sexual life of Savages in North-Western Melanesia, namely that the Trobriand Islanders are ignorant of the role of the ‘pater’ as ‘genitor’. Given Malinowski’s excellent command of Kilivila and the incredible amount of magical formulae he managed to collect and to translate — texts that represent secret and inherited knowledge which was up till 1983 well guarded by the experts and almost exclusively bestowed on relatives (i.e., within the matrilineal line, see Senft 1997b) — it is very hard for me to understand why the master of Trobriand ethnography did not hit upon the fact of Trobriand natural contraceptives, but took the Trobriand “myth” of conception and “virgin birth” not as a kind of ideology with the function (!) to diminish discrimination of extramarital births and to allow obviously cuckolded men to keep their face (see Weiner 1976: 122; Sprenger: 1997: 61ff). Note also that the incident Weiner reports about a man who returns after his year-long absence to his village and to his wife and who reacts extremely jealous when his wife presents him with a new born child, is just another confirmation of the fact that this man obviously knew about physical paternity. I can only explain this — for me rather open and blunt — mistake of Malinowski’s in two ways. Either he became the victim of the Trobriand Islanders’ love of making fun of people — with their metalinguistic vocabulary they also differentiate the so-called ‘Biga Sopa’, the lying or joking or indirect language (see Senft 1985b, 1986: 125) — and

Bronislaw Kasper Malinowski

they really took him for a ride, or — but this alternative is a somewhat nasty imputation — he used this “exotic” claim to promote his academic career even further. In her introduction to the latest edition of The Sexual Life of Savages Weiner (1987: xxvii) discusses Malinowski’s relation to psychoanalysis in detail and refers to Stocking (1986) with a footnote, in which she states: “According to Stocking (pp. 32–33) Malinowski’s response to psychoanalysis may have been more than intellectual. Rivers, the most eminent figure in British anthropology, died suddenly in 1922, leaving a vacuum in leadership. Stocking suggests that because Rivers had a long interest in psychoanalysis, Malinowski in taking up psychoanalytic debates, strengthened his bid against the diffusionists William Perry and Elliot Smith, to become Rivers’s successor” (Weiner 1987: xlii, fn. 19) — and the claim in ‘the sex book’ could then be interpreted as the final culmination of such a strategic move within academic politics (however, I want to emphasize that this assumption is nothing but a nasty imputation. An even nastier imputation would be to accuse Malinowski of having taken his peer group for a ride with this ‘exotic fact’ — I still find it extremely interesting and sometimes quite hilarious to note that — with the exception of Anna Weiner — all people engaged in this debate have never set foot on the Trobriand Islands). However, this obvious mistake does not and cannot diminish Malinowski’s incredible achievements within his Trobriand ethnography. With respect to the style in which he presents his insights and analyses of Trobriand culture I have to agree with Robert Redfield (1974: 9) who pointed out that no “writer of our times has done more than Bronislaw Malinowski to bring together in single comprehension the warm reality of human living and the cool abstractions of science. His pages have become an almost indispensable link between the knowing of exotic and remote people as we know our own neighbors and brothers, and conceptual and theoretical knowledge about mankind”. According to Mrs. Seligman, Malinowski once said proudly “Rivers is the Rider Haggard of anthropology; I shall be the Conrad” (Firth 1957: 44), and I must confess that I read the books of the Polish ethnographer with the same suspense as the books of his fellow Polish novelist. With respect to an overall assessment of Malinowski I would like to point out once more that he is one of the most important anthropologists of our century, that he is one of the founders of social anthropology who introduced the concept of “participant observation” into anthropology and who insisted on the ethnographer’s linguistic competence in, and competent use of, the language spoken in the culture studied as another absolutely necessary prerequisite for doing sound anthropological research. It is extremely difficult to do justice to such a “protean character” (Firth 1989: xxi) like Bronislaw Kasper Malinowski in a rather short handbook article — the expected publication of Michael Young’s biography on this great anthropologist and ethnographer will be more than welcome.


224 Gunter Senft

References Agar, M. (1994). Language Shock — Understanding the Culture of Conversation. William Morrow and Co. Baumann, R. (1992). Text and discourse in anthropological linguistics. In W. Bright (Ed.), International encyclopedia of linguistics: 145–147. Oxford University Press. Duranti, A. & C. Goodwin (Eds) (1992). Rethinking Context. Cambridge University Press. Ellen, R., E. Gellner, G. Kubica & J. Mucha (Eds) (1988). Malinowski between two worlds. The Polish roots of an anthropological tradition. Cambridge University Press. Firth, J.R. (1957). Ethnographic Analysis and Language with Reference to Malinowski’s Views. In R. Firth (Ed.): 93–118. ——— (1957a). Introduction: Malinowski as Scientist and Man. In. R. Firth (Ed.): 1–14. ——— (1981). Bronislaw Malinowski. In S. Silverman (Ed.), Totems and teachers. Perspectives on the history of anthropology: 101–140. Columbia University Press. ——— (1989) [1967]. Second Introduction 1988. In B. Malinowski, A Diary in the Strict Sense of the Term. xxi–xxxi. The Athlone Press. Firth, R. (Ed.) (1957b). Man and Culture — An Evaluation of the Work of Bronislaw Malinowski. Routledge & Kegan Paul. Kohl, K.-H. (1987). “Der Verdammte der Inseln” Bronislaw Kaspar (sic) Malinowski (1884–1942)”. In K.-H. Kohl, Abwehr und Verlangen: zur Geschichte der Ethnologie. 39–62. Edition Qumran im Campus Verlag. Langendoen, D.T. (1968). The London School of Linguistics: A Study of the Linguistic Theories of B. Malinowski & J.R. Firth. The MIT Press. Leach, E.R. (1957). The Epistemological Background to Malinowski’s Empiricism. In. R. Firth (Ed.): 119–137. Routledge & Kegan Paul. Leach, J.W. & E. Leach (Eds) (1983). The Kula — New Perspectives on Massim Exchange. Cambridge University Press. Malinowski, B. (1920). Classificatory particles in the language of Kiriwina. Bulletin of the school of Oriental studies, London institution, Vol. I, Part IV: 33–78. ——— (1922). Argonauts of the Western Pacific. Routledge & Kegan Paul. ——— (1923). The problem of meaning in primitve languages. In C.K. Ogden & I.A. Richards, The meaning of meaning. Supplement I. 296–336. Kegan Paul. ——— (1926). Anthropology. In Encyclopedia Britannica. 13th Edn. Sup. Vol. 1: 131–140. Benton. ——— (1929). The Sexual Life of Savages in Northwestern Melanesia. Routledge & Kegan Paul. ——— (1931). Culture. In Encyclopedia of the Social Sciences, Vol. 4: 621–646. Macmillan. ——— (1935). Coral gardens and their magic. Allen & Unwin. ——— (1944). A Scientific Theory of Culture and other Essays. University of North Carolina Press. ——— (1967). A Diary in the Strict Sense of the Term. The Athlone Press. ——— (1974). Magic, Science and Religion and Other Essays. Souvenir Press. ——— (1988). Malinowski Among the Magi: The Natives of Mailu. Routledge. Métraux, R. (1968). Bronislaw Malinowski. In International Encyclopedia of Social Sciences Vol. 9: 541–549. Crowell Collier & Macmillan. Nerlich, B. & D.D. Clarke (1996). Language, Action and Context — The Early History of Pragmatics in Europe and America, 1780–1930. John Benjamins. Paluch, A.K. (1981). The Polish Background to Malinowski’s Work. Man 16: 276–285. Powell, H.A. (1957). An Analysis of Present Day Social Structure in the Trobriand Islands. Ph.D. thesis, University of London.

Bronislaw Kasper Malinowski

De Revai, M.L., M.D. (1992). Letter to the Editor: Forum Trobriand Islands. National Geographic 182:5, Nov. 1992. Richards, A.I. (1957). The Concept of Culture in Malinowski’s Work. In R. Firth (Ed.): 15–31. Routledge & Kegan Paul. Redfield, R. (1974). Introduction. In B. Malinowski: 9–13. Souvenir Press. Senft, B. & G. Senft (1986). Ninikula — Fadenspiele auf den Trobriand Inseln, Papua Neuguinea. Baessler Archiv, Beiträge zur Völkerkunde NF 34, 93–235. Senft, G. (1985a). Weyeis Wettermagie — eine ethnolinguistische Untersuchung von fünf wettermagischen Formeln eines Wettermagiers auf den Trobriand Inseln. Zeitschrift für Ethnologie 10 (1): 67–90, (2) erratum. ——— (1985b). How to tell — and understand — a ‘dirty’ joke in Kilivila. Journal of Pragmatics 9: 815–834. ——— (1986). Kilivila — The Language of the Trobriand Islanders. Mouton de Gruyter. ——— (1991). Mahnreden auf den Trobriand Inseln — Eine Fallstudie. In D. Flader (Ed.), Verbale Interaktion — Studien zur Empirie und Methodologie der Pragmatik : 27–49. J. B. Metzler. ——— (1992). ‘As Time goes by…’: Changes observed in Trobriand Islanders’ culture and language, Milne Bay Province, Papua New Guinea. In T. Dutton (Ed.), Culture change, language change — case studies from Melanesia. 67–89. Pacific Linguistics, Australian National University. ——— (1994a). These “Procrustean” Feelings… — Some Of My Problems In Describing Kilivila. Semaian, 11: 86–105. ——— (1994b). Spatial Reference in Kilivila: The Tinkertoy Matching Games — A Case Study. Language and Linguistics in Melanesia 25: 55–93. ——— (1996). Classificatory Particles in Kilivila. Oxford University Press. ——— (1997a). Magic, missionaries and religion. Some observations from the Trobriand Islands. In T. Otto, A. Borsboom (Eds), Cultural dynamics of religious change in Oceania. 45–59. KITLV Press. ——— (1997b). Magical Conversation on the Trobriand Islands. Anthropos 92: 369–391. Sprenger, G. (1997). Erotik und Kultur in Melanesien. Eine kritische Analyse von Malinowskis “The Sexual Life of Savages”. LIT-Verlag. Stocking Jr., G.W. (1986). Anthropology and the science of the irrational: Malinowski’s encounter with Freudian psychoanalysis. History of Anthropology 4: 13–49. Wayne, H. (Malinowska) (1985). Bronislaw Malinowski: the influence of various women on his life and works. American Ethnologist 12: 529–540. ——— (Ed.) (1995). The Story of a Marriage. The Letters of Bronislaw Malinowski and Elsie Masson. Vol. I: 1916–20. Vol. II: 1920–35. Routledge. Wegener, P. (1885). Untersuchungen über die Grundfragen des Sprachlebens. Max Niemeyer. Weiner, A.B. (1976). Women of Value, Men of Renown. New Perspectives in Trobriand Exchange. University of Texas Press. ——— (1987). Introduction. In B. Malinowski The Sexual Life of Savages in North-Western Melanesia. xii–xlix. Beacon Press. ——— (1988). The Trobriand Islanders of Papua New Guinea. Holt, Rinehart & Winston. Young, M.W. (1984). The Intensive Study of a Restricted Area, Or, Why Did Malinowski Go to the Trobriand Islands. Oceania 55: 1–26. ——— (1987). Malinowski and the function of culture. In D.J. Austin-Broos (Ed.), Creating culture. Profiles in the study of culture: 124–140. Allen & Unwin.


Phatic communion Gunter Senft Max Planck Institute for Psycholinguistics, Nijmegen

In the first supplement to C.K. Ogden’s & I.A. Richard’s The meaning of meaning, Bronislaw Malinowski discusses ‘The problem of meaning in primitive languages’. Three years after his first explicitly linguistic paper on ‘Classificatory particles in the language of Kiriwina’ (Malinowski 1920) and 12 years before the publication of the second volume of his book Coral gardens and their magic (Malinowski 1935), where he presents his ‘Ethnographic theory of language’, this is the second primarily linguistic paper of the master of Trobriand ethnography, who emphasized from the very beginning of his research the importance of linguistics for anthropology in general and for ethnography in particular (see e.g., Malinowski: 1915: 501; 1922: 1–25). Malinowski developed his ethnographic theory of language mainly in connection with his attempts to translate the Trobriand Islanders’ magical formulae. He characterized his — pragmatic — theory of meaning as a theory that insists on the “linking up of ethnographic descriptions with linguistic analysis which provides language with its cultural context and culture with its linguistic reinterpretation. Within this latter […]” Malinowski has “[…] continually striven to link up grammar with the context of situation and with the context of culture” (Malinowski 1935: 73). For Malinowski speech is part of the context of situation in which it is produced, language — in its primitive function — has an essentially pragmatic character, and “meaning resides in the pragmatic function of an utterance” (Bauman 1992: 147). In his contribution to Ogden’s & Richard’s book Malinowski also refers to the conception of ‘context of situation’ which is so important for his theory of language. In the central section of this article he emphasizes that language — at least in its primitive function — has to be regarded as a mode of action rather than as a countersign of thought; and that to understand the use of a complex speech situation requires the understanding of the situation in which it occurred and the action it accomplished. Malinowski then introduces the concept of ‘phatic communion’ into linguistics. Discussing language used in what he calls “free, aimless social intercourse”, mentioning “inquiries about health, comments on weather” (Malinowski 1936: 313), and greeting formulae, Malinowski points out the following: […] to a natural man another man’s silence is not a reassuring factor, but on the contrary, something alarming and dangerous […]. The breaking of silence, the communion of words is the first act to establish links of fellowship, which is consummated only by the

Phatic communion

breaking of bread and the communion of food. The modern English expression, ‘Nice day to-day’ or the Melanesian phrase ‘Whence comest thou?’ are needed to get over the strange unpleasant tension which men feel when facing each other in silence. After the first formula, there comes a flow of language, purposeless expressions of preference or aversion, accounts of irrelevant happenings, comments on what is perfectly obvious […]. There can be no doubt that we have a new type of linguistic use — phatic communion I am tempted to call it, actuated by the demon of terminological invention — a type of speech in which ties of union are created by a mere exchange of words […]. Are words in Phatic Communion used primarily to convey meaning, the meaning which is symbolically theirs? Certainly not! They fulfil a social function and that is their principal aim, they are neither the result of intellectual reflection, nor do they necessarily arouse reflection in the listener […]. Each utterance is an act serving the direct aim of binding hearer to speaker by a tie of some social sentiment or other. Once more, language appears to us not as an instrument of reflection but as a mode of action […]. […] ‘phatic communion’ serves to establish bonds of personal union between people brought together by the mere need of companionship and does not serve any purpose of communicating ideas. (Malinowski 1936: 314–316)

Thus, Malinowski’s concept of ‘phatic (from Greek phatos, ‘spoken’) communion’ highlights — what ethologists would call — the ‘bonding function’ of language (Senft 1987: 111–112). Konrad Ehlich (1993: 317) quite plausibly interprets Malinowski’s use of the word ‘communion’ with its religious connotation as a means for emphasizing the intensity of this type of speech. Malinowski’s concept was borrowed and slightly modified by Roman Jakobson (1960) in his expansion of Karl Bühler’s (1934) ‘organon model of language’ to refer “to that function of language which is channeloriented in that it contributes to the establishment and maintenance of communicative contact” (Lyons 1977: 53–54). I assume it is most probably because of Jakobson’s rather influential paper that nowadays most linguists and anthropologists refer with the technical term ‘phatic communication’ to Malinowski’s concept. However, the term ‘phatic communication’ that many writers have used to refer to Malinowski’s concept is not really an alternative to the term ‘phatic communion’. As Adam Kendon (personal communication) points out, the term ‘phatic communication’ is probably used because people tend to forget the more general meaning of the term ‘communion’; it is precisely that achievement of ‘rapport’ through the use of speech — a kind of communion, indeed — that Malinowski emphasized, and this is different from what is often thought to be the meaning of ‘communication’. Before I discuss the concept in more detail, I want to note that ‘phatic communion’ should not be confused with Austin’s ‘phatic act’ which is defined as “the act of uttering certain vocables or words, i.e., noises of certain types belonging to and as belonging to a certain vocabulary,


228 Gunter Senft

in a certain construction, i.e., conforming to and as conforming to a certain grammar, with a certain intonation, &c.” (Austin 1975: 92). To briefly summarize again, based on Malinowski’s definition and influenced by Jakobson’s concept of the ‘phatic function’ of verbal communication, the terms ‘phatic communion’ (and ‘phatic communication’) are generally used to refer to utterances that are said to have exclusively social, bonding functions like establishing and maintaining a friendly and harmonious atmosphere in interpersonal relations, especially during the opening and closing stages of social — verbal — encounters. These utterances are understood as a means for keeping the communication channel open. It is generally claimed that phatic communion is characterized by not conveying meaning, by not importing information; thus, phatic utterances are described as procedures without propositional contents. Greeting formulae, comments on the weather, passing enquiries about someone’s health, and other small talk topics have been characterized as prototypical examples for phatic communion ever since Malinowski’s coining of the term. However, a search on the literature reveals that research dealing explicitly with ‘phatic communion’ is a relatively neglected area in linguistics. Why is this so? First of all it has to be mentioned that with the exception of ‘Firthian linguistics’ (Mitchell 1957, 1975; J.R. Firth 1957) and M.A.K. Halliday’s work, Malinowski’s functionalist pragmatic ideas about language had little influence in Europe. With respect to the USA, Noam Chomsky’s student Terence Langendoen presented in 1968 a criticism of Malinowski’s linguistic theory, arguing amongst other things that in ‘The problem of meaning in primitive languages’ Malinowski “failed to prove that the meaning of utterances is in any way related to contexts of situation” (Langendoen 1968: 25). Nevertheless, in the USA, Malinowski’s ideas about speech as action certainly had much influence on the ethnography of speaking paradigm as well as on discourse and conversation analysis. But it is exactly with one of the leading figures of the ‘ethnography of speaking’ approach that we find a rather severe objection against one aspect of Malinowski’s concept of ‘phatic communion’. With the definition of the concept, Malinowski also claimed that his outline of a semantic theory is “throwing some light on human language in general” (Malinowski 1936: 310). However, this only slightly hedged claim that concepts of his theory of language are universal, is explicitly refuted by Dell Hymes (1967, 1972, 1974). Hymes refers to the ‘ethnographic record’ that suggests that “phatic communication is far from universally an important or even accepted motive” (Hymes 1972: 40). Hymes mentions Sapir (1949: 16, 11) and quotes Gardner (1966: 398) who points out that the Paliyans of south India “communicate very little at all times and become almost silent by the age of 40. Verbal, communicative persons are regarded as abnormal and often as offensive” (Hymes 1972: 40). In addition, on the basis of his own personal fieldwork experience, Hymes points out that with the “Wishram Chinook of the Columbia River […] one does not talk when one has nothing that needs to be

Phatic communion 229

said” (Hymes 1974: 127). However, Hymes does not clarify whether the “almost silent” Paliyans never exchange greetings, e.g., under certain circumstances; and he does not elaborate either on the concept of what it is with the Wishram Chinook that “needs to be said”. David Crystal (1987: 10–11) points out that cultures vary greatly in the topics which they permit as phatic communion. The weather is not as universal a conversation-filler as the English might like to think! For example Rundi women (in Burundi, Central Africa), upon taking leave, are quite often heard to say, routinely and politely, ‘I must go home, or my husband will beat me’.

However, this critricism is somewhat incoherent. One of the things one may do in leave-taking, even if one is English, is to refer to some sort of external reason that explains that one is no longer free to stay. Adam Kendon (personal communication) pointed out that for example henpecked husbands in pubs have been known to talk along the same lines: “The old lady will have it in for me if I don’t scapa”. Hence the Rundi women are doing just what English people do — although what constitutes the unavoidable circumstance that compels a person to leave is a matter of cultural variation. Thus, we can summarize that the universality of the concept of ‘phatic communion’ as well as the universality of conversational topics that are claimed to be characteristic for phatic communion have been questioned by writers like Hymes and Crystal. We have to conclude that this issue is indeed a matter for further research. However, this discussion with respect to the universality of the concept only partly helps to explain why we find only a few studies that explicitly deal with ‘phatic communion’ as defined by Malinowski. Is there something wrong with the concept, one is tempted to ask? To find an answer to this it may be helpful to go back to Malinowski’s definition and see what topics he mentioned as typical examples for phatic communion. Besides “inquiries about health, comments on weather”, and the “modern English expression, ‘Nice day to-day’ ” he refers to “the Melanesian phrase ‘Whence comest thou?’ ” (Malinowski 1936: 313, 314). How do the Trobriand Islanders — with whom Malinowski worked and lived together and who inspired his linguistic theorizing — express this ‘Melanesian phrase’ in their language? And does this expression really function as an act of phatic communion (only)? In Kilivila, the Austronesian language of the Trobriand Islanders, we find the following greeting formulae that are similar to our European greetings, which can probably be traced back to the influence of European and Australian missionaries: bwena kaukwau (good morning), bwena lalai (good day), bwena kwaiyai (good afternoon), and bwena bogi (good night) (Senft: 1986). However, these formulae are only used in rather formal situations. The general greeting formula on the Trobriand Islands consists of the question ambeya or its shortened form ambe which literally translates as ‘where’ and which can be glossed here either as ‘where are you going to?’ or, according to the situational context, as ‘where are you coming from?’, or, if you like, as ‘whence

230 Gunter Senft

comest thou?’. In a social encounter situation the participants ask each other this question and answer it as accurately and as truly as possible, usually in the form of complex serial verb constructions such as Bala bakakaya baka’ïta basisu bapaisewa (‘I will go, I will take a bath, I will come back [to the village], I will stay [there], I will work’). This greeting formula certainly signals friendliness and opens the communication channel. But what is really happening? With this form of greeting Trobriand Islanders signal and assure persons addressed — and greeted — in this way that they can completely rely on their status as members of a community where everybody cares for the other. This implies a person’s security within the community’s social net, and this also guarantees a person’s secure way to her or his destination, a secure stay, and a secure way back home to where he or she came from. Keeping these functions in mind (which, by the way, are not — etic — interpretations but represent my phrasing of explanations my Trobriand consultants gave me at a very early stage in my field research), the meaning of these greeting formulae becomes evident: if anything may happen to Trobriand Islanders on their way — be it by the influence of evil spirits or because of ‘black magic’, or because of a more ‘profane’ accident like breaking one’s leg on the small, narrow, and stony paths or being hit by a falling coconut — they can be sure that their whereabouts are (roughly) known and that people will search for them and help them. Thus, the greeting formula cannot only be regarded as a ‘ritual’ of friendly encounter; as a binding ritual it also signals security within the social net of the community in which it is used. I understand this kind of greeting as a form of ‘ritual communication’ on the Trobriand Islands (Senft 1987: 107–108, 1991: 245–246; see also Senft in Eibl-Eibesfeldt & Senft 1987: 92–94, and Huxley 1966). Without going into more detail here, I want to emphasize that even on the Trobriand Islands “the Melanesian phrase ‘Whence comest thou?’ ” obviously conveys more than the social function of creating a bond between speaker and addressee, because the phrase initiates routine exchanges that are rich in information — and one wonders why this function escaped the great ethnographer’s attention. The observation that there is generally more behind an utterance which is said to serve only a phatic function, also holds for all of the rather few studies that explicitly deal with the concept of ‘phatic communion’. Thus, in his anthology Conversational routine we find a contribution by Florian Coulmas in which he emphasizes for Japanese that although “most apologies observed in everyday interaction” seem to be “desubstantialized routines with no semantic content, merely functioning as means of ‘phatic communion’ ” the situation is more complex: “the functionally similar employment of apology and gratitude expressions must be seen as a significant reflection of social values and attitudes prevailing in Japanese culture” (Coulmas 1981: 87). In the same volume, John Laver points out that the “linguistic behavior of conversational routines, including greetings and partings, as well as pleas, thanks, excuses, apologies and small talk, is part of the linguistic repertoire of politeness”

Phatic communion

(Laver 1981: 290) as analyzed by Brown & Levinson (1978). Discussing utterances of phatic communion, he finds that besides the two social functions already mentioned by Malinowski (viz. to “defuse the potential hostility of silence” and to allow participants in a social verbal encounter “to cooperate in getting the interaction comfortably under way”: Laver 1981: 301), these linguistic routines also have a third and probably more important function in the initial phase of conversation: “phatic communion […] allows the participants to feel their way towards the working consensus of their interaction […], partly revealing their perception and their relative social status” (Laver 1981: 301). In an earlier, and most important paper for the discussion of Malinowski’s concept, Laver (1975) elaborates on all “communicative functions of phatic communion” in detail. In this paper, he first points out that “the fundamental function of the […] communicative behavior that accompanies and includes phatic communion is the detailed management of interpersonal relationships during the psychologically crucial margins of interactions” (p. 217). He then describes and analyzes the functions of so-called ‘phatic communion’ utterances in the opening and closing phases of interaction, especially with respect to the transition phases from “noninteraction to full interaction” and from “interaction back to noninteraction” (p. 232), as well as the role of phatic communion with respect to interactional consensus and as a kind of ‘rite of passage’. After his minute analyses of all the functions of phatic communion, Laver (p. 236) concludes the following: […] phatic communion […] serves to establish and consolidate the interpersonal relationship between two participants […]. phatic communion is a complex part of a ritual, highly skilled mosaic of communicative behaviour whose function is to facilitate the management of interpersonal relationships. The information exchanged between the participants in this communicative process is not primarily referential information, but rather is indexical information about aspects of the participants’ social identity relevant to structuring the interactional consensus of the present and future encounters. The function of phatic communion thus goes beyond the creation, in Malinowski’s phrase, of ‘ties of union’: it certainly does serve to establish such broad ties in that the tokens of phatic communion are tokens exchanged in the ritual transactions of psychosocial acceptance, but it also provides the participants with a subtle tool for use in staking indexical claims which shape and constrain their detailed relationship in the crucial marginal phases of encounters when their psychological comfort is most at risk.

Thus, Laver modifies and broadens Malinowski’s concept, emphasizing and proving again that “language is used to convey more than the propositional content of what is said” (Levinson 1983: 42). It was mentioned above that Malinowski’s ideas about language had much influence on discourse analysis and conversation analysis. Thus, although Malinowski may have failed to fully define a linguistic concept which he observed and for which he


232 Gunter Senft

coined the term ‘phatic communion’, he may still be regarded as one of the founding fathers for an important approach in linguistics that has since refined, rephrased and expanded his own ideas and theories about language. Moreover, at least within the study of language use in Oceania, Malinowski’s concept of “language as a mode of action” and his “emphasis on understanding utterances within their context of situation” (Watson-Gegeo 1986: 149) are still highly valued and appreciated. In general, it seems that within linguistics, anthropology, and anthropological linguistics some of Malinowski’s basic ideas about language and culture continue to be thought-provoking. With explicit reference to Malinowski as “an ethnographic precursor” (Goodwin & Duranti 1992: 14), social scientists have for instance started ‘rethinking context’ (Duranti & Goodwin 1992). Thus, although some of his ideas about language — such as the concept of ‘phatic communion’ — have to be modified and redefined, there still seems to be much in Malinowski’s research from which we can learn and profit.

References Austin, J.L. (1975). How to do things with words (2nd Edn). Oxford University Press. Bauman, R. (1992). Text and discourse in anthropological linguistics. In W. Bright (Ed.), International encyclopedia of linguistics: 145–147. Oxford University Press. Brown, P. & S.C. Levinson (1978). Universals in language usage. In E.N. Goody (Ed.), Questions and politeness: 56–289. Cambridge University Press. Bühler, K. (1934). Sprachtheorie. Fischer. Crystal, D. (1987). The Cambridge encyclopedia of language. Cambridge University Press. Coulmas, F. (1981). ‘Poison to your soul’. In F. Coulmas (Ed.): 69–91. ——— (Ed.) (1981). Conversational routine. Mouton. Duranti, A. & C. Goodwin (Eds) (1992). Rethinking context. Cambridge University Press. Eibl-Eibesfeldt, I. & G. Senft (1987). Studienbrief ‘Rituelle Kommunikation’. Fernuniversität — Gesamthochschule Hagen, Fachbereich Erziehungs- und Sozialwissenschaften. Ehlich, K. (1993). Kommunikation. In H. Glück (Ed.), Metzler Lexikon Sprache: 315–317. Metzler. Firth, J.R. (1957). Ethnographic analysis and language with reference to Malinowski’s views. In R. Firth (Ed.), Man and culture: 93–118. Routledge & Kegan Paul. Gardner, P. (1966). Symmetric respect and memorate knowledge. Southwestern Journal of Anthropology 22: 389–415. Goodwin, C. & A. Duranti (1992). Rethinking context. In A. Duranti & C. Goodwin (Eds): 1–42. Gumperz, J.J. & D.H. Hymes (Eds) (1972). Directions in sociolinguistics. Holt, Rinehart & Winston. Huxley, J. (1966). A discussion of ritualization of behaviour in animals and man. Philosophical Transactions of the Royal Society of London, Series B, Biological Sciences No. 772, Vol. 251: 247–526. Hymes, D.H. (1967). Models of the interaction of language and social setting. Journal of Social Issues 23/2: 8–28. ——— (1972). Models of the interaction of language and social life. In J.J. Gumperz & D. Hymes (Eds): 35–71. ——— (1974). Foundations in sociolinguistics. University of Pennsylvania Press.

Phatic communion

Jakobson, R. (1960). Linguistics and poetics. In T.A. Sebeok (Ed.), Style in language: 350–377. MIT Press. Langendoen, D.T. (1968). The London school of linguistics. MIT Press. Laver, J. (1975). Communicative functions of phatic communion. In A. Kendon, R.M. Harris & M.R. Key (Eds). Organization of behavior in face-to-face interaction: 215–238. Mouton. ——— (1981). Linguistic routines and politeness in greeting and parting. In F. Coulmas (Ed.): 289–304. Levinson, S.C. (1983) Pragmatics. Cambridge University Press. Lyons, J. (1977). Semantics, vol. 1. Cambridge University Press. Malinowski, B. (1915). The natives of Mailu. Transactions and Proceedings of the Royal Society of South Australia 39: 494–706. ——— (1920). Classificatory particles in the language of Kiriwina. Bulletin of the School of Oriental Studies 1/4: 33–78. ——— (1922). Argonauts of the Western Pacific. Routledge & Kegan Paul. ——— (1935). Coral gardens and their magic, 2 vols. Allen & Unwin. ——— (1936). The problem of meaning in primitive languages. In C.K. Ogden & I.A. Richards, The meaning of meaning, [1923], Supplement I: 296–336. Kegan Paul. Mitchell, T.F. (1957). The language of buying and selling in Cyrenaica. Hesperis 44: 31–71. ——— (1975). Principles of Firthian linguistics. Longman. Sapir, E. (1949). Language. In Selected writings of Edward Sapir (D. Mandelbaum, Ed.): 7–32. University of California Press. Senft, G. (1986). Kilivila. Mouton de Gruyter. ——— (1987). Rituelle Kommunikation auf den Trobriand Inseln. Zeitschrift für Literaturwissenschaft und Linguistik 65: 105–130. ——— (1991). Prolegomena to the pragmatics of ‘situational-intentional’ varieties in Kilivila language. In J. Verschueren (Ed.), Levels of linguistic adaptation: 235–248. John Benjamins. Watson-Gegeo, K.A. (1986). The study of language use in Oceania. Annual Review of Anthropology 15: 149–162.


Edward Sapir Jeroen Vermeulen Utrecht University

1.  Introduction The work of the American linguist and anthropologist Edward Sapir (1884–1939) is characterized by its impressive many-sidedness. Besides linguistic and anthropological topics Sapir wrote extensively about such diverse subjects as psychiatry, literature and music. In a certain way his oeuvre is fragmentary. He has written only one book, Language, in 1921, but many articles and essays (see Mandelbaum 1949). Much has been written about him and his work, mostly in article or essay form. These articles deal with various aspects of his biography, of his work and of the traditions which he followed and changed (see for instance contributions in Cowan et al. 1986). There are, nonetheless, few attempts to present a more comprehensive account of his views. In a range of publications (and throughout his own work), Dell Hymes, a major figure in anthropological linguistics himself, has offered a detailed, insightful and historical perspective on Sapir’s work through in-depth analyses of aspects of his work. Penetrating analyses of his work have also been written by Zellig Harris (1984 [1951]), Michael Silverstein (1986) and Richard Handler (1983). In addition, a biography of Sapir by Regna Darnell (1990) has been published, in which concrete events of his life are described against the background of their historical, intellectual and organizational contexts. The fragmentary, essayistic and only sometimes comprehensive way in which Sapir’s work has been written about could perhaps be related to Hymes’ observation that “[I]ndeed, there is no point or publication in his career that can be safely taken as representing ‘Sapir’s view’. Sapir’s view of the relation between language and culture, between linguistics and anthropology, was a continuously changing one” (1983: 160). Handler, on the other hand, writes: “Thus there are two moments, inextricably linked, in Sapir’s intellectual personality: his desire for self-expression and his delight in form” (Handler 1986b: 444). I will use Handler’s phrase as Leitmotiv, concentrating on Sapir’s emphasis on form in relation to his emphasis on individual expression as central themes in his work. This article starts with a presentation of Sapir’s biography (Section 2) and continues with discussions of the respective fields he worked in (Section 3). Section 3 is organized in a way that echoes the canonical reception of his work as presented in the Mandelbaum reader (which first appeared in 1949) of

Edward Sapir

selected writings of Edward Sapir in language, culture and personality. In the last section the link to contemporary work in pragmatics is made. 2.  Biography Edward Sapir was born 26 January 1884 in Lauenberg, Pomerania (Prussia), now Lebork in Poland (Darnell 1990: 1). His parents were Lithuanian Jews who never took German nationality. Sapir’s native language was Yiddish. When Edward was 5 years old, the Sapirs emigrated to the United States. He attended school in New York. He received his BA from Columbia University in 1904, and earned his master’s degree there in Germanic philology. At Columbia he met Franz Boas and enrolled in his anthropology courses. Boas introduced him into the field of American Indian anthropology and linguistics. Immediately following the completion of his MA, he carried out fieldwork on Wishram Chinook in Washington (1905) and Takelma in Oregon (1906). In 1907–8 Sapir was a research associate in anthropology with Alfred Kroeber at the University of California (Berkeley), where he worked on Yana. Until 1910 he worked on Ute/Southern Paiute as a fellow at the University of Pennsylvania. In the meantime he had submitted his decription of Takelma grammar as a dissertation to Franz Boas, which was published only in 1922 in Volume 2 of the Handbook of American Indian Languages. In 1910, due to Boas’ intervention, he got his first important position as the head of the newly established Anthropological Division of the Geological Survey of Canada, housed in the Victoria Memorial Museum in Ottawa. His task was to build up anthropology in Canada along the lines of the Boasian program. Brock, the Director of the Survey, wanted “a thorough and scientific investigation of the native races of Canada, their distribution, languages, cultures etc., and to collect and preserve records of the same” (Darnell 1990: 42). Sapir came to Ottawa full of enthusiasm and energy. He married Florence Delson, with whom he had three children. He would stay in Ottawa for 15 years; in fact the Ottawa years were his most productive ones in terms of empirical work on Amerindian languages, of publications and of developing ideas on the relations between language, culture and personality (cf. Silverstein 1991: 183; Handler 1983). He conducted fieldwork on a number of languages including Nootka and Sarcee, the latter being his first work on a language of the Athabaskan family, to which Sapir would be committed during his career (Darnell 1990: 238–261; Krauss 1986). His commitment not only lies in the details of grammatical description, but he also, through Athabaskan, “worked out his position on aesthetics and personality variations in cultural tradition” (Darnell 1990: 238). This illustrates his (Boas-inspired) interest in the relevance of linguistics to culture, which was stated explicitly in his “Time Perspective in Aboriginal American Culture: A Study in Method”, published in 1916.


236 Jeroen Vermeulen

During the years 1913 to 1920 he worked on a classification of American Indian languages, based on the assumption that linguistic method could make suggestions about historical relations. As a result of that work he presented his six unit classification of American Indian languages (published in 1921), which was received with scepticism. In that same year his only book, Language, appeared. In this rich book he wrote about his general views on language and related topics, like culture and literature, in a way accessible to a non-linguistic public. After a few years in Ottawa his initial enthusiasm about the possibilities of his job diminished. He felt isolated in Ottawa and longed, somewhat nostalgically, for an artistically and intellectually more inspiring environment, like the one he experienced during his student years at Columbia University. His feelings of isolation and frustration were enhanced by the fact that his anthropological research on the native people of Canada was hampered by lack of financial support due to the First World War. On top of these professional worries, his wife Florence’s mental and physical illnesses were of great concern; she died in 1924. In this period he began to write poetry and to compose music; he published one volume of poetry, Dreams and Gibes in 1917. He also wrote quite a number of literary reviews in important American journals like The Dial and Poetry. Many of the ‘Boasians’ wrote poetry in these days. Among them was Ruth Benedict, with whom Sapir exchanged and discussed poems (cf. Darnell 1986a & 1990: 151; and esp. Handler 1986a). During these years he became increasingly interested in issues of culture and personality (see for instance his view of culture in the famous “Culture, genuine and spurious” from 1924). In 1925 Sapir was offered a position at the University of Chicago. He was asked to lead a new anthropology program in the Department of Sociology and Anthropology, together with Fay-Cooper Cole. The years in Chicago (until 1931) were his happiest ones. He had numerous students attending his graduate courses in linguistics and the very popular course called “The psychology of culture”.1 Together with the political scientist Harold Lasswell and the psychiatrist Harry Stack Sullivan, Sapir tried to establish an interdisciplinary program on the issue of personality and culture (‘life histories’). The attempt was not very succesful in terms of institutionalization (cf. Darnell 1986a); but Sapir was soon recognized as one of the most influential figures in American anthropology and in the increasingly professionalizing field of linguistics (he became the president of the American Anthropological Association as well as of the Linguistic Society of America). In his personal life a happy and important change took place in September 1926 when Sapir married his second wife Jean McClenaghan.

1.  This course was intended to be the basis for a book with the same title; a book which he never finished himself, but which was reconstructed and edited by Judith Irvine, 1994.

Edward Sapir

In 1931, in spite of the inspiring intellectual climate in Chicago, he accepted a financially and professionally attractive Sterling professorship of Anthropology and Linguistics at Yale, having become disillusioned with the amount of administrative effort demanded of him at the University of Chicago. The years at Yale appeared to be difficult years (cf. Darnell 1998), not in the least due to the anti-semitic climate there. Nevertheless, he was surrounded again by numerous students (some of them followed Sapir from Chicago to Yale), such as Stanley Newman, Morris Swadesh, Walter Dyk, Harry Hoijer, Mary Haas and Benjamin Lee Whorf. Here he wrote much on his mature ideas on the interplay between culture and personality. Besides these culture-and-personality themes he returned to work on Semitic and Indo-European languages. In 1937–38 Sapir suffered from a series of heart attacks; he died of heart disease on 4 February 1939. 3.  Sapir on language, culture and personality Understanding Sapir’s work and views on language means to historicize them in the context of the genesis of American anthropology under Boas’ leadership at the turn of the century. Franz Boas (1858–1942) emigrated in 1886 from Germany to the United States. By then he was influenced by the linguistic ideas of the filosopher-linguistdiplomat Wilhelm von Humboldt, via his writings and via his student the linguist Heymann Steinthal (cf. Koerner 1990; Drechsler 1988).2 In the Humboldtian linguistic tradition, emphasis was placed on the linguistic classification of a diversity of, especially non-Indo-European, languages and on the relationship between language structure and culture. The idea was that language played a fundamental role in the conceptualization (‘articulation’) and construction of reality (cf. Erickson et al. 1997). As Boas conceived this, via Steinthal: “(…) the form of thought is molded by the whole social environment of which language is part” (1974 [1904]: 28). Through the study of language unconscious categories of thought could be uncovered: “No other manifestation of the mental life of man can be classified so minutely and definitely as language” (ibid.). The statement expresses Boas’ motive for applying linguistic method to anthropological research and explains why he committed himself all his life to a linguistic classification of American Indian languages of North America. The empirical stance in Boasian thinking on language and culture must be seen as a reaction against “the evolutionary prejudice which instilled itself into the social sciences towards the middle of the last century (…)” (Sapir, 1921: 123). Boas was very cautious in making abstract generalizations too quickly. This did not mean that he 2.  This influence is largely a question of reconstruction and comparison: Boas rarely referred to him or other intellectual predecessors (see Koerner, 1990: 114; but see Boas, 1974 [1904]).


238 Jeroen Vermeulen

had a pure relativistic view on language and culture. He held that all languages reflect the history and culture of their communities — a universalistic theme in his work, as Hymes & Fought (1975: 81) write. In Section 3.1 a discussion of some issues of Sapir’s linguistic work will be presented against the background of Boas’ views. Sapir’s views on culture and personality will then be discussed (in Section 3.2). 3.1  Language 3.1.1  Americanist text tradition Given the importance of language for the conceptualization of reality and thus given the inseparability and language of culture, Boas developed and used linguistic method to study and understand cultural phenomena. This led to a “text-oriented field method” (Darnell 1990: 267; see also Leeds-Hurwitz & Nyce 1986), to which the collection of authentic texts was central. Authentic here meant that texts should be recorded and studied in their historical context of appearance. As Silverstein (1986: 77ff) notes, the view Boas had on the nature of language as a historical and cultural object had implications for the nature of its study. In contrast to objects of physical science knowledge of language could not be objectively determined, but could be determined only in terms of the native speaker’s own subjective point of view. The correct way for an analyst of language to approach his/her object was, as Boas himself put it in the Introduction to the Handbook of American Indian Languages, “as though an intelligent Indian was going to develop the forms of his own thoughts by an analysis of his own form of speech” (1911: 81). This meant, as Sapir once put it in a letter to Cole (25 April 1938, cited in Darnell 1990: 251), “the stuff in the raw, as felt and dictated by the natives”. Each language must be described in its own terms, in its own time and place, as part of a historical continuation of its community’s tradition. A language is a historical object in the sense of an “unbroken transmission of subjective intuitions” (Silverstein 1986: 78). Sapir followed Boas’ approach in his “Time Perspective” of 1916, where he used linguistic method in reconstructing ethnological phenomena. Historical understanding was the proper goal for ethnology (see Handler 1983: 212). Where knowledge of the historical connections is missing no theoretical categories should be imposed on the raw data. Both linguistic and ethnological data should not be considered in isolation, but in their historical context. The contextual approach to ethnological artifacts also became a way of representing material in museums, for instance in Sapir’s Victoria Memorial Museum in Ottawa. Of the Boasians, Edward Sapir was the most brilliant linguist. He applied linguistic techniques with more sophistication than any other, including Boas himself. Because of that, he was more daring in applying these methods in reconstructing genetic relations between languages and cultures. See for example his Six Unit

Edward Sapir

Classification of Amerindian Languages, about which Boas was explicitly negative (see Darnell 1986a: 562ff). Sapir took great pains at publishing the texts, which were the basis for ethnological and linguistic research. The importance of detailed text analysis (always in combination with its contextualization) in understanding human behavior (linguistic or otherwise) lies in trying to “know what is the precise manner and articulation of the doing” (Sapir 1927: 547). 3.1.2  Linguistic form Sapir was very much concerned with linguistic form, more than with functional aspects of language. Throughout his book Language (1921) he argues for the dominance of form over function, although “[T]o say in so many words that the noblest task of linguistics is to understand languages as form rather than as function or as historical process is not to say that it can be understood as form alone” (Sapir 1929: 152). It is important to note here, that by ‘function’ Sapir means conceptual meaning. In Sapir’s sense form and function are relatively independent. As an example, in English ‘five men’ indicates plural in two ways, by way of internal modification of the vowel ‘man’ → ‘men’ and by way of the number ‘five’. The formal, grammatical, process of internal modification is redundant with respect to the functional meaning. As Sapir says: “Now form lives longer than its own conceptual content” (Sapir 1921: 98). Although language is above all an instrument for “ideation” (ibid.: 38), on a deeper level of analysis “irrational form, form for form’s sake” (ibid.: 98) seems to dominate (see also Handler 1986b: 445). This discussion could be summed up in his statement: “Were a language ever completely ‘grammatical’, it would be a perfect engine of conceptual expression. Unfortunately, or luckily, no language is tyrannically consistent. All grammars leak” (Sapir 1921: 38). In Language Sapir classifies linguistic elements according to the types of concepts or meanings they express (see Contini-Morava 1986 for a discussion of Sapir’s theory of grammar). A basic distinction is that between radical and grammatical elements. The radical elements express relatively concrete concepts such as objects, actions and qualities (Sapir 1921: 93). Grammatical elements (mostly affixes) are more abstract concepts that “put[s] upon the fundamental [i.e., ‘radical’] concept a formal limitation” (ibid.: 26). Grammatical elements, contrary to radical ones, can not occur in isolation, but are always attached to a radical element. In a sentence radical and grammatical elements can be combined by making use of, what Sapir calls, grammatical processes. In Chapter 4 of Language Sapir lists six types of processes: word order, composition, affixation, internal modification, reduplication, and accentual (stress or tone) change. Grammatical processes are to be distinguished from phonological processes. It is only the former processes, that have functional significance, i.e., that produce meaningful differences (ibid.: 62). There is, for example, no meaningful difference in the s in bag-s and book-s (pronounced /z/ in the former, /s/ in the latter) as there is in the noun house versus the verb to house. Languages can be classified to the extent they make use


240 Jeroen Vermeulen

of specific grammatical processes. For instance, English makes use of affixation as a grammatical process but not of reduplication; reduplication is used in Hottentot and Chinook; change of tone is used as grammatical process in Shilluk (ibid.: 77–80). Sapir adopted the notion of grammatical process from Boas (1911: 23), who used it as a basis for the comparison of languages, in reaction to evolutionist classifications which all too simply used terms like ‘incorporating’ or ‘polysynthetic’. An important difference with Boas is the fact that Sapir stressed that grammatical processes are a particular means of expressing grammatical meaning (cf. Contini-Morava 1986: 347ff). The grammatical patterning has psychological significance for the speakers of a language in the sense that speakers have a special sensitivity to linguistic form. Linguistic forms are not only objective or logical, but also subjectively felt. We can illustrate this distinction by placing the analytical concepts of radical and grammatical elements against Sapir’s definition of the word. His definition is as follows: “… the word is one of the smallest, completely satisfying bits of isolated ‘meaning’ into which the sentence resolves itself ”. Note here the qualification ‘satisfying’, which points to the psychological meaningfulness for the language user (Sapir 1921: 34). And then: “… the radical and grammatical elements of language, abstracted as they are from the realities of speech, respond to the conceptual world of science, abstracted as it is from the realities of experience, and (…) the word, the existent unit of living speech, responds to the unit of actually apprehended experience, of history, of art” (ibid.: 32/33). Here Sapir refers to the grammatical system of a language in relation to the way its speakers, unconsciously, experience its forms and to the “feeling for form” as a creative, artistic, capacity they derive from it. Sapir often uses the notion of ‘form-feeling’ (next to ‘intuition’ and ‘aesthetic’) with respect to linguistic form and pattern.3 Sapir sums up his position in Language: “Every language, then, is characterized as much by its ideal system of sounds and by the underlying phonetic pattern (…) as by a definite grammatical structure. Both the phonetic and conceptual structures show the instinctive feeling of language for form” (p. 56). 3.1.3  Form-feeling The concept of ‘form-feeling’ is best elaborated in Sapir’s phonological work. He distinguishes between a ‘purely objective system of sounds’ (ibid.: 55), its physiological or phonetic character, and an inner (‘ideal’) meaningful pattern or system, which he calls ‘phonemic’ in his seminal article “Sound Patterns in Language” (1925). Here he develops the idea that sounds are speech sounds only if they are elements in a sound pattern. A sound pattern is “the inner configuration of the sound system of a language,

3.  Hymes (1983: 152) notes that the source is undoubtedly the Italian philosopher Croce — see also Handler (1986b: 436) for a further discussion of the notion and its source.

Edward Sapir

the intuitive ‘placing’ of the sounds with reference to one another, (Sapir 1925: 35/36). These sounds (phonemes) are psychologically real for the speakers of a language: speakers have a definite feeling for the formal patterning of sounds in their language. He discusses the psychological reality of the inner pattern of phonemes in an article with that title in 1933. One of his examples is the spelling of the word pá ßa’ ‘at the water’4 by Tony Tillohash, native speaker of Southern Paiute (Sapir 1933b: 48–50). When Sapir asked him to divide the word in syllables, Tony spelled it pa (pause) pa’, very much to Sapir’s astonishment. Apparently Tony ‘heard’ a more abstract form: pa — is a root that changes a subsequent stop (here p) in a spirant (ß). It is the sound that Tony ‘heard’, that Sapir qualifies as the subjectively and psychologically true sound: an ‘organic’ sound or phoneme. Here we see that a speaker’s unconscious feeling for form can be brought to his consciousness (see also Sapir 1921: 55). On this point Sapir differs from Boas, who said (1911: 67) “that the linguistic classifications never rise into consciousness” and if they do, they “give rise to secondary reasoning and to re-interpretation”. Notice that for Sapir, on the contrary, this reasoning is leading to the recovery of the true sounds in the case of phonology. Hymes and Fought mark the development of the phoneme and ‘sound patterns’ as a step from “Boas’ analytic approach to structural analysis proper”, where the “status of elements [is defined] explicitly and consistently in terms of internal relations” (1975: 82; see also Hymes 1983: 147/8). 3.1.4  Linguistic relativity With regard to the specific formal patterning and the form-feeling of a language, Sapir speaks of the language’s ‘genius’, or ‘secret’, its own “type or plan or structural ‘genius’ ” (Sapir 1921: 120). Nevertheless, he holds the view that every language is capable of expressing in its forms all the concepts and ideas it needs; this is what he calls the ‘formal completeness’ of any language. More poetically he expresses this universal feature of language as: “When it comes to linguistic form, Plato walks with the Macedonian swineherd, Confucius with the head-hunting savage of Assam” (Sapir 1921: 234). In this formal sense language is complete in itself; this means for Sapir that language is an autonomous, irreducible, system. As said earlier, form, for Sapir, is relatively independent from its function. The formal patterns of language have, so to say, a life of their own, due to the inertia of linguistic form. At the same time, however, linguistic form is not as cold as steel, but is the living unit of human experience. Linguistic form is psychologically real and meaningful for the speakers of a language. Unconsciously, or naïvely, language functions thus as our ‘guide to social reality’ (Sapir 1929: 162). Every

4.  To be analyzed as: voiceless labial stop; stressed long a; voiced bilabial spirant; unstressed short a; final aspiration.


242 Jeroen Vermeulen

language has its own way of conceptualizing social reality by using and accepting the “fixed habits of speech as guides to an objective understanding of the nature of experience” (Sapir 1924b: 155). He repeats this statement 5 years later: “The fact of the matter is that the ‘real world’ is to a large extent unconsciously built up on the language habits of the group. No two languages are ever sufficiently similar to be considered as representing the same social reality” (1929: 162). This is what Sapir calls the “relativity of the form of thought” (1924b: 155). His student Benjamin Lee Whorf explicitly followed and explored this path of linguistic relativity. This never lead (neither with Sapir, nor with Whorf) to simple or rigid ideas about the relation between language, thought and culture (see for a discussion of Sapir’s view on language in relation to culture, Hymes 1983: 150ff). Language is essentially dynamic, or to use a Humboldtian notion, Energeia, and not Ergon, a completed and closed work. In his article “Language” (1933a) Sapir defines language at the cutting edge between its formal and denotational capacities and the intuitively based expressive use a speaker makes of it. The dynamics of language and of the language-culture relationship is located on this cutting edge. 3.2  Culture and personality An important insight of Sapir with respect to form was that he realized “more than any American anthropologist, that the persistence of recognizable form … is greater in language than in culture” (Hymes 1983: 150). Sapir’s formal orientation led him to such important notions as pattern and the phoneme, as we have seen in the last section, which brought him to a structuralist approach to language. But as also Handler (1986b: 210) notes, Sapir transcends structuralism in emphasizing the individual’s intuition for pattern, his form-feeling, and thus his concern for individual experience and creativity. In using language the speaker is led (not coerced) by the ‘configurational pressure’ (Harris 1984 [1951]) of the grammatical patterns, but at the same time this allows him to take part in social action. Sapir also developed a theory of culture along these same lines: an approach that can be called structural for its analysis of patterning, but that has great concern for and insight in individual behavior. Already in 1917 he criticized the concept of a ‘superorganic’ (a social reality like ‘society’ that transcends the individual) by arguing that individual and social behavior can be distinguished only analytically, for all behavior is the behavior of individuals (Sapir 1917a). Sapir does not need the concept of a ‘superorganic’ to explain social behavior as something else than individual behavior. The collective forms of language and of culture are elements of patterned wholes, that are experienced and unconsciously known by the individual members of a community. He elaborated this idea in an article called “The Unconscious Patterning of Behavior in Society” (1927). There he states about the knowledge an individual has of the patterns: “In a sense it is well known to him. But this knowledge is not capable of conscious

Edward Sapir

manipulation in terms of word symbols. It is, rather, a very delicately nuanced feeling of subtle relations, both experienced and possible” (ibid.: 548, orig. italics). The notion of ‘unconscious patterning’, that is central to both his linguistic and cultural views, allows Sapir to elaborate on the relationship between individual creativity and given cultural forms (see Handler 1983: 210). According to Handler he developed this insight in connection with the practice of his poetry and other forms of art: “Sapir sought in art a way to unite form and feeling, cultural givens and subjective experience (…)” (ibid.: 211). Handler (ibid.: 218–221) discusses two essays in literary theory, both published in The Dial in 1917, the same year that Sapir published his only volume of poems. The essays are titled “The Twilight of Rhyme” (1917b) and “Realism in Prose Fiction” (1917c). In the first essay he explores the relation between individual creative expression and traditional, cultural or linguistic, forms. He argues that the artist must find a balance between the two in order to express himself in a sincere way: “Just as soon as an external and purely formal aesthetic device ceases to be felt [sic!] as inherently essential to sincerity of expression, it ceases to remain merely a condition of the battling for self-expression and becomes a tyrannous burden (…)” (Sapir 1917, cited in Handler 1983: 219). The argument is not only central to Sapir’s theory of art, but also to that of culture. In “Culture, Genuine and Spurious” (1924), a genuine culture is described as a culture in which there is a balance between the needs of individuals to express and unfold themselves and the forms that their culture generates. A culture is genuine, when these two form an integrated whole. Sapir says: “The genuine culture is not of necessity high or low; it is merely inherently harmonious, balanced, selfsatisfactory. It is the expression of a richly varied and yet somehow unified and consistent attitude toward life, an attitude which sees the significance of any one element of civilization in its relations to all others. It is ideally speaking, a culture in which nothing is spiritually meaningless (…)” (Sapir 1924a: 315). Note here that Sapir speaks of culture in terminology that presents a culture as a work of art (see also Handler 1983: 225). The second essay is about describing reality in literary prose. Sapir here discusses and rejects two literary techniques of realism, which he calls ‘rigorous’ and ‘objective’ (1917: 504, cited in Handler 1983: 220). Objective realism is the way in which the narrator takes an outside position from which he is able to give an overview of the characters and their action. It aims at giving an objective account of reality, but fails to give the reader a sense of identification with the characters. In rigorous realism the author tells the story from the point of view of one character only, trying to present an inner truth. Sapir offers a third technique for prose fiction: a story should be told from several perspectives, each character giving his own portrayal of the world. That would present a true ‘objective’ understanding of social reality: “for may not objectivity be defined as the composite picture gained by laying a number of subjectivities on top of one another…?” (ibid.).


244 Jeroen Vermeulen

Here we meet a second central point in his view on culture: “understanding human interaction from multiple ‘inner’ points of view” (Handler 1983: 221). The locus of culture is not something ‘superorganic’ like society, but must be sought for in the interaction of individuals or in what Sapir calls “interpersonal relations”, following Harry Stack Sullivan (Sapir 1939: 579). And ‘individual’ here should not be understood as “a biologically defined organism” but as “that total world of form, meaning and implication of symbolic behaviour, which a given individual partly knows and directs, partly intuits and yields to, partly is ignorant of and swayed by” (Sapir 1932: 518). In “The Emergence of the Concept of Personality in a Study of Cultures” (1934) Sapir argues that cultural patterns must be replaced in the real-life situations from which they first have been abstracted in order to really understand their meaning for the individuals. And then: “The more fully one tries to understand a culture, the more it seems to take on the characteristics of a personality organization” (Sapir 1934: 594). As he describes his concept of personality, he states, somewhat ironically, that there is nothing about it an anthropologist should be afraid of, and at the same time he brings back the notion of culture to more humble proportions: personality is “a distinctive configuration of experience which always tends to form a psychologically significant unit and which, as it accretes more and more symbols to itself, creates finally that cultural microcosm of which official ‘culture’ is little more than a metaphorically and mechanically expanded copy” (ibid.: 595).

4.  Sapir’s importance to pragmatics The interdisciplinary versatility of Edward Sapir makes him a pragmaticist in optima forma, when we use the definition of pragmatics (in this handbook) as the cognitive, social and cultural study of language and communication. We can make the characterization more precise by connecting Sapir to one of the formative traditions of pragmatics, the ethnography of speaking. As shown in Section 3.1, Sapir adhered to the text tradition of Boasian anthropology and he can be said to having contributed greatly to this approach. In all of his work he insisted on studying language in the context of its use, in the cultural and historical matrix of its appearance and development. In doing so he, directly through his students and indirectly, after his death, through his writings, influenced many anthropologists and linguists. He is one of the originators of the ethnography of communication and of other forms of “socially constitued linguistics” (Hymes 1974: 196) such as interactional sociolinguistics (Gumperz 1982). His formal orientation to contextualized texts is also an explicit inspirational ‘model’ for exciting work that is being done in contemporary anthropological linguistics by, e.g., Silverstein & Urban (1996), Hanks (1989) and Briggs & Bauman (1990). This work may be summarized as the study of processes of entextualization in the ‘histories of

Edward Sapir

discourse’ and the study of metapragmatics (see Lucy 1991). It is also highly influenced by the work of Mikhael Bakhtin (1981) on polyphony and intertextuality. A remarkable resemblance of Bakhtin’s views and Sapir’s work on language and culture can be seen in Sapir’s literary article “Realism in Prose Fiction” (1917c — see Section 3.2). It would be interesting for theory formation in pragmatics to further investigate the resemblance (and differences) between the theoretical frameworks of Sapir and Bakhtin (for Bakhtin and Whorf, see Schultz 1990). Sapir’s work in culture and personality did not have a good fate (see Darnell 1986), maybe in part due to his early death in 1939, but maybe also because in those days the intellectual climate did not favor such, as we now say, ‘constructivist’ thinking. Nevertheless, his work on culture and personality is still relevant for many issues that lie at the heart of present-day pragmatics.

References Bakhtin, M. (1981). The dialogic imagination. University of Texas Press. Bauman, R. & C. Briggs (1990). Poetics and performance as critical perspectives on language and social life. Annual Review of Anthropology 19: 59–88. Boas, F. (1904). The history of anthropology. Science 20: 513–524. (reprinted in G. Stocking, Jr. (Ed.): 23–36). ——— (1911). Introduction. In F. Boas (Ed.), pt. 1: 1–83. Boas, F. (Ed.) (1911). Handbook of American Indian languages, Part 1,2 and 3. Government Printing Office. Contini-Morava, E. (1986). Form in language: Sapir’s theory of grammar. In W. Cowan et al. (Eds): 341–369. Cowan, W., M.K. Foster & K. Koerner (Eds) (1986). New perspectives in language, culture and personality: proceedings of the Edward Sapir Centenary Conference. John Benjamins. Darnell, R. (1986a). Personality and culture: the fate of the Sapirian alternative. In G.W. Stocking Jr. (Ed.): 156–183. ——— (1986b). The emergence of Edward Sapir’s mature thought. In W. Cowan et al. (Eds): 553–588. ——— (1990). Edward Sapir: linguist, anthropologist, humanist. University of California Press. ——— (1998). Camelot at Yale: the construction and dismantling of the Sapirian synthesis, 1931–39. American Anthropologist 100(2): 361–372. Drechsler, E.J. (1988). Wilhelm von Humboldt and Edward Sapir: analogies and homologies in their linguistic thoughts. In W. Shipley (Ed.), In honor of Mary Haas: 225–264. Mouton de Gruyter. Erickson, J., M. Gymnich & A. Nünning (1997). Wilhelm von Humboldt, Edward Sapir, and the constructivist framework. Historiographia Linguistica XXIV(3): 285–306. Gumperz, J.J. (1982). Discourse strategies. Cambridge University Press. Handler, R. (1983). The dainty and the hungry man: literature and anthropology in the work of Edward Sapir. In G.W. Stocking Jr. (Ed.): 208–231. ——— (1986a). Vigorous male and aspiring female: poetry, personality, and culture in Edward Sapir & Ruth Benedict. In G.W. Stocking Jr. (Ed.): 127–153.


246 Jeroen Vermeulen ——— (1986b). The aesthetics of Sapir’s Language. In W. Cowan et al. (Eds): 433–454. Hanks, W.F. (1989). Text and textuality. Annual Review of Anthropology 18: 95–127. Harris, Z.S. (1984 [1951]). Review of Selected writings of Edward Sapir in language, culture and personality, ed. By D.G. Mandelbaum. In K. Koerner (Ed.): 69–114. Hymes, D.H. (1974). Foundations in sociolinguistics: an ethnographic approach. University of Pennsylvania Press. ——— (1983). Linguistic method in ethnography: its development in the United States. In Essays in the history of linguistic anthropology: 135–244. John Benjamins. Hymes, D.H. & J. Fought (1975). American structuralism. Mouton. Irvine, J. (Ed.) (1994). Edward Sapir: the psychology of culture. A course of lectures. Mouton de Gruyter. Koerner, K. (Ed.) (1984). Edward Sapir: appraisals of his life and work. John Benjamins. ——— (1990). Wilhelm von Humboldt and North American ethnolinguistics: Boas (1894) to Hymes (1961). Historiographia Linguistica XVII 1(2): 111–128. Krauss, M.E. (1986). Edward Sapir and Athabaskan linguistics, with preliminary annotated bibliography of Sapir’s work on Athabaskan and Na-Dene. In: W. Cowan et al. (Eds): 147–190. Leeds-Hurwitz, W. & J.M. Nyce (1986). Linguistic text collection and the development of life history in the work of Edward Sapir. In W. Cowan et al. (Eds): 495–532. Lucy, J. (1991). Reflexive language: reported speech and metapragmatics. Cambridge University Press. Mandelbaum, D.G. (Ed.) (1949). Selected writings of Edward Sapir in language, culture and personality. University of California Press. Sapir, E. (1916). Time perspective in aboriginal American culture: a study in method. In D.G. Mandelbaum (Ed.) (1949): 389–462. ——— (1917a). Do we need a ‘superorganic’? American Anthropologist 19: 441–447. ——— (1917b). The twilight of rhyme. Dial 63: 98–100. ——— (1917c). Realism in prose fiction. Dial 63: 503–506. ——— (1921). Language: an introduction to the study of speech. Hartcourt Brace. ——— (1922). The Takelma language of southwestern Oregon. In F. Boas (Ed.), pt. 2, 1–296. ——— (1924a). Culture, genuine and spurious. In D.G. Mandelbaum (Ed.): 308–331. ——— (1924b). The grammarian and his language. In D.G. Mandelbaum (Ed.): 150–159. ——— (1925). Sound patterns in language. In D.G. Mandelbaum (Ed.): 33–45. ——— (1927). The unconscious paterning of behavior in society. In D.G. Mandelbaum (Ed.): 544–559. ——— (1929). The status of linguistics as a science. In D.G. Mandelbaum (Ed.): 160–166. ——— (1932). Cultural anthropology and psychiatry. In D.G. Mandelbaum (Ed.): 509–521. ——— (1933a). Language. In D.G. Mandelbaum (Ed.): 7–32. ——— (1933b). The psychological reality of phonemes. In D.G. Mandelbaum (Ed.): 46–60. ——— (1934). The emergence of the concept of personality in a study of cultures. In D.G. Mandelbaum (Ed.): 590–597. ——— (1939). Psychiatric and cultural pittfalls in the business of getting a living. In D.G. Mandelbaum (Ed.): 578–589. Schultz, E.A. (1990). Dialogue at the margin: Whorf, Bakhtin and linguistic relativity. The University of Wisconsin Press. Silverstein, M. (1986). The diachrony of Sapir’s linguistic description; or, Sapir’s ‘cosmographical’ linguistics. In W. Cowan et al. (Eds): 67–110. ——— (1991). Problems of Sapir historiography. Historiographia Linguistica XVIII(1): 181–204. Silverstein, M. & G. Urban (Eds) (1996). Natural histories of discourse. The University of Chicago Press.

Edward Sapir

Stocking, G.W., Jr. (Ed.) (1974). A Franz Boas reader: the shaping of American anthropology, 1883–1911. Basic Books. ——— (Ed.) (1983). Observers observed: essays on ethnographic fieldwork. University of Wisconsin Press. ——— (Ed.) (1986). Malinowsky, Rivers, Benedict and others: essays on culture and personality. University of Wisconsin Press.


Taxonomy Robert E. MacLaury

Taxonomy, in the present context, is largely an anthropological linguistic interest in categorization. It affords a fluid interface with psychologists, linguists, anthropologists, and other specialists concerned with the crosscultural comparison of ways that lexically labelled categories are internally organized and related to each other within a domain. Certain linguistic anthropologists have specialized in one or another of the ubiquitous domains, such as plant classification or color terms (Berlin 1992; MacLaury 1995), where they have established replicable elicitation procedures, demonstrated widely recurrent patterns among categories, proposed models of processes and systems that may explain the patterns, and attempted to test the models for cognitive validity. The domain-specific concentrations have brought methodology to the forefront of taxonomy, setting it apart from non-numerical approaches to categorization (Lakoff 1987; Turner 1967; Lévi-Strauss 1966). Other rubrics of the method-oriented program are ethnoscience (Sturtevant 1964), ethnographic semantics (Colby 1966), and folk classification (Conklin 1972). Taxonomy fits under the umbrella of cognitive anthropology, which highlights additional processes, such as decision making (Young & Garro 1994), that need not be linguistic. Anthropological linguistic concern with taxonomy traces back to Boas’s (1881) inaugural dissertation on Eskimo classes of sea-water color and to Bartlett’s (1928, 1940) field studies of color naming and plant taxonomy in Malaysia. Conklin (1954, 1955), under Bartlett’s counsel, pioneered the systematic study and formal modeling of semantic domains with his fieldwork on color and plant taxonomies among the Hanunùo of Mindoro. The approach became ethnoscience when Conklin’s (1962) concise statements found parallels. Lounsbury (1964) proposed extension rules for Crow and Omaha kinship systems; Goodenough (1964) proffered componential analyses of Lapp and English kinship systems; Frake (1961) delineated Subanun disease taxonomy. The initial joint effort was united under the assumption that categories are model-theoretic sets of features; the features are given to algebraic expression and constitute dimensions across a domain; a contrasting category is formed at each dimensional intersection, like a phoneme. Kay (1966) summarized the formal properties of three domains: paradigm, taxonomy, and tree. Pike’s (1967) notions of emic versus etic, which had circulated since the 1950s, became axiomatic to the stated goal of representing native view. Toward this end, some practitioners experimented with exacting elicitation procedures that featured questions about set-inclusion and contrast, but they produced sterile results (Black 1969). Investigators of taxonomy backed

Taxonomy 249

themselves into a corner by proclaiming that their formal approach would extend to the whole of culture (see Burling 1964 for a critique). An ideology of strong linguistic relativity pervaded ethnoscience to the same extent that all anthropologists believe that culture is relative. However, the ide a that linguistically sanctioned categories shape thought is tautological because categories are thought. Yet the circularity was masked by asking ‘Does language determine thought?’ Substantial programs aimed at probing the language-thought relation became a main impetus of taxonomy, producing classic descriptions of categorical semantics, such as Casagrande & Hale’s (1967) experiments with Papago folk definitions. Predictably (albeit in retrospect), the mirage of language-thought influence was never pinned down (see Lucy 1992 for another view). Berlin & Kay (1969) chose color as the apt domain in which to scrutinize the doctrine of linguistic relativity. There had been at least fifteen years of interdisciplinary concord that color constituted the paradigmatic example of semantic and cognitive arbitrariness. For instance, Ray (1953) used a spectrogram to measure color-term ranges at a narrow band of middle brightness among speakers of eight American Indian languages, all of whom differed substantially from each other and from English. He declared that color perception itself is in flux and that languages improvise categories of any conceivable configuration to impose order upon it. Gleason (1961) used equivalent examples of relative color categorization in the opening chapter of his popular introductory linguistics textbook (following Bloomfield 1933 & Hjelmslev 1953). Psychologists Brown & Lenneberg (1954) prefigured cognitive anthropology with experiments that correlated codeability and memorability of Munsell color chips; they asserted that the former linguistic characteristic caused the latter mental capacity. Berlin & Kay (B&K) adopted stimuli and method from Lenneberg & Roberts (1956), who had investigated the codeability of Zuni color categories by asking subjects to encircle the ranges of pre-elicited terms on acetate laid over a 320-chip Munsell array; they had also asked for a focus of each term. B&K expand ed the array with achromatic (white-grey-black) chips but otherwise adopted this method to interview 20 foreign students at Berkeley and 40 Tzeltals in Mexico. Their sample further included verbal reports of color terms from 78 languages. Observations were confined to ‘basic color terms’ (BCTs), the simplest abstract names most easily invoked and generally applicable whose ranges are not encompassed by others. They found that BCTs are universally focused on the same colors but may number from two to eleven in a language. Their foci are predictable from their number, which implies they evolve in a constrained order in any language. As normal people see color in only one way, all will name what they see accordingly as an augmentation of societal complexity creates the functional need to increase vocabulary. B&K’s statement countered the entrenched relativity so forcefully that it ignited debate that continues to the present (Hickerson 1971; Conklin 1973; Sahlins 1976;

250 Robert E. MacLaury

Tornay 1978; Wierzbicka 1993). It resonated with Chomsky’s universalist-psychological model of syntax. Four developments followed. First, field tests in Dani (Heider 1972), Bellonese (Kuschel & Monberg 1974), Aguaruna (Berlin & Berlin 1975), Binumarien (Hage & Hawkes 1975), Futunese (Dougherty 1975), and Eskimo (Heinrich 1974) confirmed B&K’s seven evolutionary stages but showed that BCTs name all colors at any time. Berlin & Berlin (1975) and Kay (1975) added predictable boundaries to their stages, drawing the lines around purest hues as though people always name hue but never favor brightness. The formulation ignored findings that antedated dogmatic relativism, such as Geddes (1946), who found that Fijians mainly categorize brightness. Second, cognitive psychologist Rosch (1973) formulated her prototype notion after testing B&K’s prediction of focality among Dani speakers. She suggested that Brown & Lenneberg’s (1954) correlation of codeability and memorability derived from an underlying perceptual purity of the specific hues that attract B&K’s universal focus clusters: panhuman perception determines both language and cognition while neither language nor cognition is causal. Rosch’s prototype model of the category inherited this naive realism from B&K and retained it until she wrote its obituary (Rosch 1978), conceding that it does not accommodate crosscultural differences or dynamics within systems (the enduring popularity of the model notwithstanding). Third, Zollinger (1972) attributed universal color-term focality to a neural process model. De Valois & Abramov (1966) had monitored its analogue in the post-retinal pathway of Macaque monkeys. Kay & McDaniel (1978) thus proposed that people categorize color by applying Zadeh’s (1965) fuzzy-set operations to neural response: categorization is classical logic with a decimal point appended. This formalism shared with Rosch’s prototype structure the inability to explain why color categories change even though colors do not. Burgess et al. (1983), in attempting to model semantically graded Tarahumara color-term qualifiers, discovered that values of membership skew toward a focus as a category divides; the plastic manipulations contravened their prediction that the values would conform to perceptually based fuzzy-set curves. Fourth, MacLaury (1991, 1992, 1995) enhanced observation of color categorization with a three-part elicitation in which a subject names loose chips in random order, selects foci, and maps the range of each color term on the Munsell array with rice grains; as the latter often proceeds by spontaneous pauses and steps (in response to ‘Put rice on more of X-term’), it reveals something of the internal organization of a category. The improved data warrant a two-part account of color-category dynamics. First, a shift in selective emphasis between the perceived similarity and difference among colors drives change; the shift is part of an individual’s private response to external events, such as social conditions or impingement of novelty. Thus, in B&K’s evolutionary order, distinct hues, such as red and yellow, are separately categorized before similar hues, such as green and blue. Second, perceptions of pure colors and


the emphases on similarity and difference are used as coordinates to construct a category by analogy to physical coordinates in space-time as though the category were a point of view. This ‘vantage theory’ of categorization accounts for various processes of change (e.g., near synonymy → coextensivity → inclusion → complementation) as well as regularities brought out by the methodological advance. For example, Hage & Hawkes (1975) found that Binumariens name one broad color category with two terms whose ranges are coextensive but that, nevertheless, are focused widely apart. Speakers of Mesoamerican languages interviewed by the new method show, in addition, that terms of such coextensive pairs have a ‘dominant’ member that is abundantly named and centrally focused in opposition to its ‘recessive’ counterpart whose range is a bit narrower and more skewed; the pattern is statistically significant. The vantage model attributes this dominant-recessive pattern to distinct arrangements of the same set of coordinates; the dominant view assigns priority to similarity while the recessive view favors the differences. Finally, the model addresses the universal conversion of brightness categories to hue categories; the more that observers attend to differences, the more they notice hue. Many linguists and anthropologists think of taxonomy as mainly the folk classification of plants and animals; it is a scheme that relates categories by hierarchical inclusion and by contrast on any level. For example, Berlin, Breedlove & Raven (1973), after reviewing a worldwide sample of native plant taxonomies, propose six possible ‘ranks’ (often diagrammed as vertically arranged). Folk generic taxa are mid-most and primary. They are the categories, such as oak, sparrow, or cow, that people find easy to envision wholly and concretely. Some languages encode only taxa of this rank (Lee 1979), and children learn to name its taxa before others (Dougherty 1978; Stross 1973). Below the folk generic rank are specifics and varietals, such as Angus cow and black Angus cow. These taxa are based on contrasting characteristics, and their names are compounded. Above the folk generic rank are life forms, such as tree, bird, or animal, and a unique beginner, such as thing, entity, or phenomenon. Most languages use life forms, which may be semantically complex and broadly applicable, as is ‘animal’ or as are cases of polysemy that breach taxonomic ranks (Hage, Hawkes & Miller 1976). Few languages name a unique beginner, fewer yet with only one term. Intermediate taxa reside between generics and life forms, e.g., bovine. A hierarchical taxonomy represents one pole of a typological continuum whose opposite consists of a single-rank taxonomy in which every category is extendable from a typical meaning to others that resemble it. Neither pole has a perfect representative among unschooled people, and the systems that gravitate markedly toward the single-rank, extendible type are only confirmed in Northern California and Oregon in the USA. Bright & Bright (1965) find that Yurok, Karuk, Smith River Athapascan, and neighboring languages name a focal taxon as well as others with a single term, attributing all to one ‘sphere of influence’. Hunn & French (1984) describe how Sahaptins name a


252 Robert E. MacLaury

residual generic taxon with a simple lexeme and a coordinate generic taxon with the lexeme plus a qualification, such as ‘like-’ or reduplication. Sahaptins name few life forms or specifics. Comparison of folk taxonomies reveals predictable patterns suggesting that all people, to some extent, see and classify external reality in the same way. The best evidence regards polytypy, the inclusion by generic taxa of specific taxa (regardless of whether the latter are separately named). Berlin (1992) finds that agriculturalists name more polytypic specifics than do foragers; domesticated generics are lexically subdivided more than others, which, in turn, decrease in polytypic naming in correlation with whether a taxon is protected as a resource, merely significant as such, or insignificant. It seems that people will subclassify a generic category to the extent that they pay attention to it, although the special treatment in itself may proliferate physical variety. Apart from whether polytypy is named, Berlin finds that languages everywhere show significantly similar ratios of monotypic to polytypic generics, about 80:20. Berlin, Boster & O’Niel (1981), in a study of Jívaroan folk ornithology, find among genetically related bird species that those which have evolved pronounced physical differences tend to be classified as monotypic significantly more than those which are morphologically less diversified. The latter are especially subject to an exchange of names, as are the color categories that are named from coextensive slants. But Berlin, Boster & O’Niel’s study of bird taxonomy also typifies weaknesses that pertain throughout much of the research on folk biological semantics (not to fault only these authors). Following Hays (1976), they regard the mixing of names to be evidence of covert intermediate taxa while they pass over this opportunity to develop a model of the cognition that operates within a category. They use stimuli solely for identification: asking subjects to name one specimen after another (in this case, their skins), like naming loose color chips. They forego classification: asking each subject to sort out all specimens called X-name, recording the result, and repeating the request until the subject insists that no specimens of that name are left, like mapping color chips. In other studies, Rosch et al. (1976) and Dougherty (1978) show that taxa change rank with variation of expertise or culturally inspired attention, although this flexible property of folk classification remains unaddressed by the predominant model. Hunn (1982) offers that some folk biological classes are based on the utility of taxa from a human perspective, but Berlin (1992) counters that all classes of a hierarchical taxonomy are based on intrinsic features of form and behavior; utilitarian classes only crosscut a taxonomy. The work on folk biology is tethered to a conviction that universals derive from naming unedited perceptions of the external world. Although Berlin & Kay’s concept of BCTs also espoused this objectivism, the work with color since has recognized cognition, such as selective emphasis and analogy, while, nevertheless, most color ethnographers accept that universal percepts, such as unique hues, must be part of the model


(see Van Brakel & Saunders 1996 for an interdisciplinary forum on this latter point). But folk biological classification moves in a different direction. On the one hand, it remains a proving ground for a positivist philosophy of semantics, thus determining the data that will excite interest and the hypotheses that will be tested against them. On the other hand, it assists a worldwide drive to record disappearing local plant knowledge for practical and commercial purposes (e.g., Shaman Pharmaceuticals), which may employ some linguistic anthropologists. Taxonomy subsumes more than only color and folk biology. Researchers are beginning to innovate methods to investigate the taxonomies of other domains, for example, de León (1992) uses dolls and other props to elicit Tzotzil concepts of space and body parts. In widespread languages, such as those of Mayan or Austro-Asiatic stocks, obligatory words or affixes classify referents according to shape, array, or even social status whenever they are counted (Becker 1975; Berlin 1968), a domain called ‘numeral classifiers’. Burling (1970) shows in Thai that distinct numeral classifiers arouse different images of the same referent, for example ‘one river’ as a hair-line (on a map), a segment (at a bathing spot), a link (between villages), or a conduit (to the sea). The enduring aim of taxonomy has been to bring the description and modeling of crosscultural categorization within the reach of replicable procedure over and above techniques based on participant observation, verbal elicitation, and informed analysis, however astute they may be.

References Bartlett, H.H. (1928). Color nomenclature in Batak and Malay. Papers of the Michigan Academy of Sciences, Arts, and Letters 10: 1–52. ——— (1940). The concept of genus. Bulletin of the Torrey Botanical Club 67(5): 349–362. Becker, A.L. (1975). A linguistic image of nature. Linguistics 165: 109–121. Berlin, B. (1968). Tzeltal numeral classifiers. Mouton. ——— (1992). Ethnobiological classification. Princeton University Press. Berlin, B. & E.A. Berlin (1975). Aguaruna color categories. American Ethnologist 2: 61–87. Berlin, B., J.S. Boster & J.P. O’Niel (1981). The perceptual bases of ethnobiological classification. Ethnobiology 1: 95–108. Berlin, B., D. Breedlove & P.H. Raven (1973). General principles of classification and nomenclature in folk biology. American Anthropologist 75: 214–242. Berlin, B. & P. Kay (1969). Basic color terms. University of California Press. Black, M. (1969). Eliciting folk taxonomies in Ojibwa. In S.A. Tyler (Ed.): 165–189. Bloomfield, L. (1933). Language. Holt, Rinehart & Winston. Boas, F. (1881). Beiträge zur Erkenntniss der Farbe des Wassers. Doctoral Diss., University of Kiel. Bright, J.O. & W. Bright (1965) Semantic structures in Northwestern California and the Sapir-Whorf hypothesis. American Anthropologist 67: 249–258. Brown, R. & E.H. Lenneberg (1954). A study in language and cognition. Journal of Abnormal and Social Psychology 49: 454–462.


254 Robert E. MacLaury Burgess, D., W. Kempton & R.E. MacLaury (1983). Tarahumara color modifiers. American Ethnologist 10: 133–149. Burling, R. (1964). Cognition and componential analysis. American Anthropologist 66: 20–28. ——— (1970). Man’s many voices. Holt, Rinehart & Winston. Casagrande, J.B. & K.L. Hale (1967). Semantic relations in Papago folk-definitions. In D. Hymes & W.E. Bittle (Eds), Studies in Southwestern ethnolinguistics: 165–196. Mouton. Colby, B.N. (1966). Ethnographic semantics. Current Anthropology 7: 3–32. Conklin, H.C. (1954). The relation of Hanunòo culture to the plant world. Ph.D. Diss. Yale University. ——— (1955). Hanunòo color categories. Southwestern Journal of Anthropology 11: 339–344. ——— (1962). Lexicographical treatment of folk taxonomies. In F.W. Householder & S. Saporta (Eds), Problems in lexicography: 119–141. Indiana University Research Center in Anthropology, Folklore, and Linguistics. [Reprinted in S.E. Tyler (Ed.): 41–59.] ——— (1972). Folk classification. Yale University Dept. of Anthropology. ——— (1973). Color categorization (review of Berlin and Kay 1969). American Anthropologist 75: 931–942. De León, L. (1992). Body parts and location in Tzotzil. Zeitschrift für Phonetik, Sprachwissenschaft und Kommunikationsforschung 45: 570–589. De Valois, R.L. & I. Abramov (1966). Color vision. Annual Review of Psychology 17: 337–362. Dougherty, J.W.D. (1975). A universalist analysis of variation and change in color semantics. Ph.D. Diss. University of California. ——— (1978). Salience and relativity in classification. American Ethnologist 5: 66–80. Frake, C.O. (1961). The diagnosis of disease among the Subanun. American Anthropologist 63: 11–32. Geddes, W.R. (1946). The color sense of Fijian natives. British Journal of Psychology, General 37: 30–36. Gleason, H.A. (1961). An introduction to descriptive linguistics. Holt, Rinehart & Winston. Goodenough, W. (1964). Componential analysis of Könkämä Lapp kinship terminology. In W. Goodenough (Ed.): 221–238. ——— (Ed.) (1964). Explorations in cultural anthropology. Holt, Rhinehart & Winston. Hage, P. & K. Hawkes (1975). Binumarien color categories. Ethnology 24: 287–300. Hage, P., K. Hawkes & W. Miller (1976). ‘Eagle’=‘bird’. American Ethnologist 3: 481–488. Hays, T.E. (1976). An empirical method for the identification of covert categories in ethnobotany. American Ethnologist 3: 489–507. Heider, E. Rosch (1972). Probabilities, sampling, and ethnographic method. Man 7: 448–466. Heinrich, A.C. (1974). Colour classification of some Central Canadian Eskimos. Arctic Anthropology 11: 68–72. Hickerson, N. (1971). Review of Berlin and Kay (1969). International Journal of American Linguistics 37: 257–270. Hjelmslev, L. (1953). Prologomena to a theory of language. Waverly Press. Hunn, E. (1982). The utilitarian factor in folk biological classification. American Anthropologist 84: 830–847. Hunn, E. & D. French (1984). Alternatives to taxonomic hierarchy. Ethnobiology 3: 73–92. Kay, P. (1966). Comments on Colby. Appended to Colby (1966): 20–23. [Reprinted in S.E. Tyler (Ed.): 78–90.] ——— (1975). Synchronic variability and diachronic change in basic color terms. Language in Society 4: 257–270. Kay, P. & C. Mcdaniel (1978). The linguistic significance of basic color terms. Language 54: 610–646. Kuschel, R. & T. Monberg (1974). ‘We don’t talk much about colour here’. Man 9: 213–242.


Lakoff, G. (1987). Women, fire, and dangerous things. University of Chicago Press. Lee, R.B. (1979) The !Kung San. Cambridge University Press. Lenneberg, E.H. & J.H. Roberts (1956). The language of experience. International Journal of American Linguistics 22: 2 (Memoir 13). Lévi-Strauss, C. (1966). The savage mind. University of Chicago Press. Lounsbury, F.G. (1964). A formal analysis of the Crow and Omaha-type kinship terminologies. In W. Goodenough (Ed.): 351–93. Lucy, J.A. (1992). Language diversity and thought. Cambridge University Press. MacLaury, R.E. (1991). Social and cognitive motivations of change. Language 67: 34–62. ——— (1992). From brightness to hue. Current Anthropology 33: 137–186. ——— (1995). Color and cognition in Mesoamerican languages. University of Texas Press. Pike, K.L. (1967). Language in relation to a unified theory of the structure of human behavior. Mouton. Ray, V.F. (1953). Human color perception and behavioral response. New York Academy of Sciences (series 2) 16: 98–104. Rosch, E.H. (1973). On the internal structure of perceptual and semantic categories. In T.E. Moore (Ed.), Cognitive development and the acquisition of language: 111–144. Academic Press. ——— (1978). Principles of categorization. In E. Rosch & B.B. Lloyd (Eds), Cognition and categorization: 28–48. Erlbaum. Rosch, E., C.B. Mervis, W.D. Gray, D.M. Johnson & P. Boyes-Braem (1976). Basic objects in natural categories. Cognitive Psychology 8: 382–439. Sahlins, M. (1976). Colors and cultures. Semiotica 16: 1–22. Stross, B. (1973). Acquisition of botanical terminology by Tzeltal children. In M.S. Demonson (Ed.), Meaning and Mayan languages: 107–141. Mouton. Sturtevant, W.C. (1964). Studies in ethnoscience. In A.K. Romney & R.G. D’Andrade (Eds), Transcultural studies in cognition: 99–113. Special issue of American Anthropologist 66(3). Tornay, S. (Ed.) (1978). Voir et nommer les couleurs. Laboratoire d’Ethnologie et de Sociologie Comparative. Turner, V.W. (1967). The forest of symbols. Cornell University Press. Tyler, S.A. (Ed.) (1969). Cognitive anthropology. Holt, Rinehart & Winston. Van Brakel, J. & B.A.C. Saunders (1996). Colour. Behavioral and Brain Sciences 18. Wierzbicka, A. (1993). Semantic primitives and semantic fields. In A. Lehrer & E.F. Kittay (Eds), Frames, fields, and contrasts: 209–227. Erlbaum. Young, J.C. & L.C. Garro (1994). Medical choice in a Mexican village. Waveland Press. Zadeh, L.A. (1965). Fuzzy sets. Information and Control 8: 338–353. Zollinger, H. (1972). Human color vision as an interdisciplinary research problem. Palette 40: 1–7.


Benjamin Lee Whorf Penny Lee PO Box 39, Maylands WA 6931, Australia

‘There is no need to apologize for speech, the most human of all actions.’ (Whorf 1941a: 220)

1.  Introduction That the so called ‘pragmatic’ perspective in linguistics, with its current accommodation of cognitive as well as social and cultural dimensions of language use, came into its own significantly in reaction to the narrow sterility of exclusively syntax-focused approaches to analyzing language is not generally contested. Pragmaticians operate in the understanding that the full subtlety and intricacy of language as a human phenomenon can only be appreciated when there is also willingness to investigate the mysteries of its operation as the central means by which people communicate, societies are brought into being and sustained, and cultures are developed and transmitted. The question for pragmatically oriented linguists has been how to make a start in the face of an appreciation of language so comprehensive and daunting in its evident complexity. The response, for the time being, has been to concentrate mainly on the many ways in which meaningful language is context-bound and functional in those situations in which it is produced, contexts in which humans typically also deploy a range of inferential processes fundamental to making language work. In these respects, a very great deal has been accomplished in some fifty to sixty years, most of it only dimly anticipated or not anticipated at all by linguists of the mid-twentieth century such as Benjamin Lee Whorf (1897–1941). Yet Whorf offered perspectives on language very compatible with the broader understanding of linguistics that now characterizes European pragmatics.This paper provides a brief introduction to Whorf, his life, his influence, and the current relevance of his theories about language in cognition and society. Whorf may have introduced the term ‘ethnolinguistics’ in 1938, defining it for his purposes as a major subdivision within linguistics as a whole. Although Malinowski had used the adjective in the 1920s, the first published use of the noun did not appear until 1947. Whorf made ‘configurative’ or ‘straight’ linguistics (the formal identification, analysis, and description of patterning observable in language data) the essential

Benjamin Lee Whorf

starting point for investigating language as a cultural phenomenon. He never doubted the relevance of morphosyntax to language used in its social context and apparently took for granted that study of syntactic patterning on its own is not complete without semantics. He began to develop a gestaltic theory of reference to anchor exploration of relationships between language and experience. Focusing on universally available ‘isolates of experience’ and their culturally specific operationalization as ‘isolates of meaning’ (Whorf 1940a: 208, 1940b: 164), this theory provides a crucial key to the logic of the ‘linguistic relativity principle’ (1940a: 214) which has to do with the way experience is variably construed as a function of different linguistic resources. Although Whorf does not seem to have known the writings of Charles Morris (1901–1979) or Karl Bühler (1879–1963) whose key works, written at the same time as he himself was active gave impetus to modern pragmatics, his deep respect for Edward Sapir (1884–1939) ensured that his own linguistics as it matured during the 1930s would have the focus on language as a key element of human behavior that characterizes the anthropological linguistics tradition and developments from it. That part of Whorf ’s oeuvre which is most directly pertinent to pragmatics is largely found in the well-known collection edited by John B. Carroll (1916–2003) in 1956. An interesting insight into Whorf ’s overall contribution may be gained by starting with a sophisticated schematization of linguistics he developed around 1938. It exists in part as an outline guide for fieldworkers entitled “Language: Plan and conception of arrangement” (Whorf 1938) and in part as a comprehensive draft report. The outline was planned as a supplement to an extensive guide for ethnologists developed by George P. Murdock (1897–1985) in the Yale Department of Anthropology. Whorf circulated it privately, Murdock having declined to use it in full. The report (Whorf & Trager 1938) was also written for the department but apparently never completed. Although it purports to detail research Whorf undertook with his colleague and friend, George L. Trager (1909–1992) during the 1937/38 academic year (when they jointly taught one of Sapir’s courses while he was on leave) it is also an extensive supplement to the outline, as Whorf indicates in some introductory remarks. Held from 1979 as a largely handwritten document in the Yale archives, it was relatively difficult to access prior to publication as an appendix to Lee (1996). By Whorf ’s own account, the report was written entirely by him after consultations with Trager, correspondence with colleagues Morris Swadesh (1909–1967), Mary R. Haas (1910–1996), and Charles (Carl) F. Voegelin (1906–1986), and discussions with students Charles F. Hockett (1916–2000) and Norman A. McQuown (1914–2005). Whorf ’s schematic plan for linguistic investigation is presented as a table divided into two basic sections: ‘Grammar’ (which includes ‘Phonology’and ‘Semasiology’ with a note (Whorf 1938: 126) that “actually these merge”) and ‘Stylistics’. Phonology is further divided into ‘Phonemics’ and ‘Morphophonemics’ (or ‘Phonologic process’) while the semasiology section is expanded over seven pages under broad


258 Penny Lee

headings: ‘The sentence’, ‘The word’, and ‘The lexeme’. Syntax is included as ‘Intrasentential linkage’, with ‘Intersentential linkage (external syntax)’ and ‘Predication’ following directly below. The Yale report has a predominantly ethnolinguistic focus, dealing only briefly with ‘the corner-stone of linguistics’ (phonemics and morphophonemics) and, in more detail, with grammatical classes (both overt and covert) before going on to deal extensively with ‘Configurative Linguistics and Cultural World Outlook — “Ethnolinguistics”’. Matters as diverse as the ‘The configuration of experience as seen in language’, ‘Coordination between language and culture’, ‘Behavior patterns as correlated with language’, and the study of ‘Supra-linguistic and quasi-linguistic mentality’ are covered, often with illuminating illustrative examples not repeated in the published writings. Central to all Whorf ’s work is his theorizing about the role of language in cognition and the place of ‘linguistic thinking’ in our understanding of language as a whole. The core ideas in this regard are found in an article he wrote in 1937 entitled “A linguistic consideration of thinking in primitive communities”. (The term ‘primitive’ was, of course widely used in the earlier part of the twentieth century, although it is interesting to note that Whorf made it clear, e.g., with the subheading “So-called Primitive Mentality” in the Yale report, that he used it with reservations. He clarified (Whorf & Trager 1938: 266) that: “What has mostly been meant by the term ‘primitive mentality’ is ‘any cultural mentality other than the SAE [Standard Average European] cultural mentality’ ”.) Whorf ’s theoretical use of the term ‘agreement’ (see Whorf 1940a: 211–214, 1941a: 238, Lee 1996: 224–250, and below) is also important for understanding the way he brought cognitive, social, and cultural aspects of language in use into alignment with each other. 2.  Whorf’s life and work What is not always realized is that even in his own time Whorf was a significant if somewhat enigmatic figure. He came into linguistics proper as a largely self-taught investigator of Mayan inscriptions, translator of historical Nahuatl texts, and student of Hebrew, which, as a young man, he believed to hold linguistic keys (in the form of ‘fundamental root ideas’ at the core of conceptual activity) to understanding the human mind. After meeting Sapir in 1928, he began to take up and develop his ideas, attending his classes at Yale from 1931. His respect for the contribution of Franz Boas (1858–1942) and Leonard Bloomfield (1887–1949) is also important for understanding the range and tenor of his linguistics in its mature form.When he died Whorf was most widely known as an authority on Mayan and Aztec. He was also a well respected member of major American linguistic and anthropological organizations and committees and the acknowledged authority on the Hopi language.

Benjamin Lee Whorf 259

Although a short period of Mexican fieldwork in 1930 was funded by the American Social Science Research Council, and although he also received several academic fellowships (Carroll 1956, Lee 1996), Whorf ’s economic security throughout his working life was assured by employment with the Hartford Fire Insurance Company. Here his role significantly involved fire prevention counseling, as well as the more usual work of insurance adjustment and sales. He specialized in chemical factories — having completed a degree in Chemical Engineering at MIT (Massachusetts Institute of Technology) in 1918 — and was well respected in the company, rising to Assistant Secretary and being given considerable leeway to undertake his linguistic work. Whorf ’s conviction that language factors were sometimes implicated in carelessness leading to fires and other industrial accidents suggests that a profoundly pragmatic approach to studying language was in place early. This balanced his equally strong interest in the role of language in cognition, an interest informed by his own original insights as well as by ideas shared with Sapir, whom he regarded as having “done more than any other person to inaugurate the linguistic approach to thinking and make it of scientific consequence, and moreover to demonstrate the importance of linguistics to anthropology and psychology” (Whorf 1937: 78). Whorf also read European and Asian philosophy, being influenced by Alfred North Whitehead (1861–1947) and Henri Louis Bergson (1859–1941) on the one hand and writers in the Theosophical tradition (which draws on Hindu and Buddhist thought) on the other. He was familiar with (and critical of) modes of logic which, in his opinion, blindly accepted grammatical categories of English and other European languages as correlating with a supposed organization of reality intrinsic to the order of nature. In spite of his business career and the practicalities it involved, Whorf was nonetheless, apparently, a supremely impractical person in terms of his personal life. He was absent-minded, never learned to drive, was said not to know one end of a screwdriver from the other, and left all practical aspects of the management of their married life, including the car and the welfare of their three children, to his wife, Celia Inez Peckham (1901–1997) who, in spite of his frequent absences from home, made his scholarly work possible through her general competence and ensured that his papers were preserved and published where feasible after his death. She was supported in the latter endeavor by her children, Raymond Ben (1921–1973), Robert Peckham (1924–1986), who provided Carroll with documents for the 1956 collection, and Celia Lee (b.1930). Born on 24 April 1897, at Winthrop, Massachusetts, into a family with predominantly intellectual and artistic interests, Whorf died in Wethersfield, Connecticut, on 26 July 1941 of a cancer he had been fighting since a major operation in late 1938. Most of his best known papers were written between 1937 and early 1941 when further writing became impossible. Explicit formulations of the linguistic relativity principle appeared only in 1940, after the famous memorial to Sapir completed in

260 Penny Lee

1939 and entitled “The relationship of habitual thought and behavior to language” (Whorf 1941c). Controversies about what he might or might not have meant in relation to linguistic relativity have raged for over fifty years, tending to marginalize him as a major contributor to linguistic theory and often involving dismissal of his work as popularist or scholastically flawed. At the heart of the debate, very frequently, is failure to differentiate between his broad interest in linguistic thinking in all its manifestations and his more specific treatment of linguistic relativity as one manifestation of that activity. Whorf ’s intellectual legacy is found in correspondence and notes (Whorf 1979) as well as in his published writings. It makes up a coherent ‘theory complex’ (Lee 1996), wide-ranging and subtle in its implications for the study of language. He was unusual for his time in that he addressed the question of mental phenomena squarely, while at the same time remaining firmly anchored in a conception of language as behavior, albeit behavior of significantly mental nature. Had Whorf ’s ideas been better understood in the 1950s and 60s they might perhaps have provided a more successful challenge than was offered to the also mentalist but overly narrow shift of focus brought about by Chomskyan structuralism with its relegation of culture and society to the periphery of scholarly attention and its abrogation of responsibility to explore the dynamics of language in use. 3.  Whorf’s perspective on linguistics The central question for Whorf was not so much that of modern pragmatics: What is it to use language? as: What is it to be a languaging being? Whorf did not conceive of language as a tool in the conventional sense but as a neurological or physico-chemical matter of linkages and connections in the brain (see Whorf 1937: 66–69; Lee 1996: 53–64). He noticed that these patterns, acquired in the course of linguistic enculturation, can have concomitant (and observable) repercussions in apparently nonlinguistic behavior. He argued that whenever language-engendered connections in the brain are activated, at whatever depth of consciousness or in whatever form, the outcome may be validly regarded as an instance of “thought insofar as it is linguistic”, the essence of such activity being the highly specific overall state of ‘rapport’ or ‘linkage’ characteristic of the particular configuration of linguistic knowledge of the person concerned (Whorf 1937: 67–68). His conception of language as a cognitive phenomenon was thus very different from the algorithmic paradigm taught in most linguistics departments in the latter part of the twentieth century. Nevertheless, like other proponents of the linguistic science of his era, Whorf was an empiricist through and through. He took for granted that the fundamental task for language analysts is to search for objectively discernible patterning in linguistic data

Benjamin Lee Whorf

of all kinds, from the most fundamentally physical (e.g. in phonetics) to the most elusively symbolic (e.g. in religious thinking). He went beyond his mentors, however, writing of ‘pattern’ and ‘patternment’ in ways which quite remarkably anticipate the preoccupation with pattern recursivity and ubiquity that characterizes chaos theory today (Lee 1997: 235). He similarly transcended the cruder associationism of behaviorist reasoning, anticipating its more sophisticated forms in modern connectionism and parallel distributed processing theory (Lee 1996: 50–54). In particular, he went well beyond Bloomfield, with his refusal to deal with nonobservable mental entities and processes, but not Sapir who, nearly twenty years earlier, had pointed out that language “consists of a peculiarly symbolic relation — physiologically an arbitrary one — between all possible elements of consciousness on the one hand and certain selected elements localized in the auditory, motor, and other cerebral and nervous tracts on the other” (Sapir 1921: 10). Whorf (e.g., 1941b: 231) also developed Sapir’s (1921: 156n, 1925) ‘points in the pattern’ insight into the cognitive organization of culturally acquired systems of behavior. He developed a somewhat holographic model for conceptualizing the structure and activity of the internalized linguistic system, one which also has much in common with Hockett’s (1987) ‘resonance’ theory of morphology (see Lee 1996: Ch. 2). The idea that linguistic enculturation is a matter of cognitive entrenchment (in a physicalist sense) of language patterns of all kinds is thus fundamental to understanding Whorf ’s linguistics and its relevance today, including its relevance to the linguistic relativity issue. To clarify how constraints brought about by entrenchment of language patterns work, Whorf offered a pioneering illustration of the strictly constrained patterning of phonemic combinatoriality by demonstrating the structure of the English monosyllable (Whorf 1941a: 223–230, 1942: 254–256). He argued that, just as speakers of any language are unconsciously bound by habitual patterning in the phonological domain, so are they also both licensed and trammeled in every other domain of linguistic patterning. Neurolinguistically entrenched constraints associated with habitual linguistic behavior may perhaps be overridden unconsciously when other ways of languaging are learned. Metalinguistic awareness, including conscious attention to what is involved in thinking nonhabitually, can facilitate conscious overriding of these patterns. Although the social implications of varietal diversity are now studied, there is another sense in which cognitively entrenched language has profound social and cultural effects. Whorf argued that learning a particular language involves learning to isolate some ‘bits of experience’ (Whorf & Trager 1938: 261) while ignoring others equally available to human senses, but less salient because not brought into focus by linguistic processes that pick them out of the ongoing flow of experiential data as ‘figure’ or as aspects of ‘ground’. He developed these ideas derived from gestalt psychology further, contrasting events conceived as happening in the ‘external field’


262 Penny Lee

of experience with phenomena inside the body, the ‘internal’ or ‘egoic’ field (Whorf 1940a: 208, 1940b: 164). Whorf ’s reasoning here is most accessible at the level of concrete and generally ‘figural’ phenomena. One language or specialist sublanguage may pick out certain objects or actions for explicit naming where another does not. For speakers of any language other languages may be experienced as having gaps with regard to what seems to be the natural order of things. We should note that the process of picking out isolates of experience, whether manifest as figure or ground, is not mediated only by linguistic activity in cognition. It is fundamentally a feature of the natural selectivity of perceptual processing in general, a process in which personal and cultural influences may intervene to create specific habits of attention that vary from individual to individual or from one cultural group to another. Primary levels of selection and perception should not be objectively different for speakers of different languages, according to Whorf ’s reasoning. But epistemologies or, for that matter, any general understandings about experience, may be different because these are subjectively, socially, and culturally developed and differently configured on the basis of (at least partly linguistic) habits of noticing and reasoning. Such differences remain out of awareness in most cases unless conscious metalinguistic attention to them is developed. Whorf emphasized that it is not necessary to be able to use another language for awareness to develop; it is enough to be alerted to systematic differences between languages for our more global conceptual “picture(s) of the universe” (1940a: 214) or “views of the world” (1941a: 221) to shift slightly and our understanding of experience as a whole to expand. The processes by which languages help to configure experience include subtle grammatical activity as well as naming. Whorf gave an example of a different conceptual, and thus experiential, focus fostered by inflectional processes in Hopi. He pointed out that: “We ‘unbutton’ a coat, Hopi ‘causes inner-plural separation’ of it (unit term ‘separate’ with inner-plural and causative inflections) without any allusion to such bits of experience as buttons” (Whorf & Trager 1938: 261). Indeed, he considered that Hopi, through a number of coordinated grammatical processes, generally promotes interpretations of experience that give greater salience to factors of dynamism in events than English. While it might be argued along lines developed by Langacker (e.g., 1991) that buttons are necessarily evoked when the coat-removing schema is activated, it is an empirical matter unlikely to be determined in the context of modern Hopi bilingualism whether monolingual speakers of Hopi and English would allocate different salience at the primary experiential level to the buttons or the equally figural repeated movements involved in opening the front of the garment. The gestaltic shift that can be experienced here is similar to that experienced in the classic illustration where focus shifts between a vase shape and that of two faces in profile. Although verbal, but nonplural, ‘undo’ in English may also promote isolations from experience more similar to the Hopi than explicit reference to buttons (including ‘unbutton’), the

Benjamin Lee Whorf

configuration of salience in what is foregrounded and backgrounded in each alternative is different, illustrating the subtlety of the focal shifts involved. 4.  The linguistic relativity principle It is important to come to terms with the isolate construct because it is central to Whorf ’s formulation of the linguistic relativity principle and critical to understanding why he has been so frequently misunderstood in relation to it. When one looks closely at his definitions (Whorf 1940a: 214, 1941a: 221) it is clear that he took the relativity principle of physics as his starting point. According to that analogy, patterns of language entrenched in cognition as part of the machinery of interaction with environmental stimuli provide (culturally acquired) frames of reference through which observers of an event come to have different understandings of it. These differences are conceived as analogically parallel to those brought about (according to the relativity principle of physics) by the positioning of observers in different coordinate systems, moving at different speeds relative to each other (Alford 1981; Lee 1996: 86–89; Foley 1997: 192–193). Whorf (1940b: 164) believed that universals of perception offer us a “canon of reference for all observers, irrespective of their languages or scientific jargons, by which to break down and describe all visually observable situations, and many other situations also” and against which linguistically generated variables may be considered. Stating that: “Our problem is to determine how different languages segregate different essentials out of the same situation”, Whorf (1940b: 162) argued that the isolate notion and the figure/ground conceptual frame enable investigators to “analyze or ‘segment’ the experience first in a way independent of any language or linguistic stock, a way which will be the same for all observers”. He took it as axiomatic that the data of experience as perceived in either the external or the egoic domain are the same across the species, a point which has often been missed but which is essential to his reasoning (see Lee 1996: Ch. 3, 2000). As Harré & Krausz (1996) point out in a general treatment of relativism, something must be held constant for talk of relativity to make sense. To the degree that people unconsciously segregate different constellations of isolates out of a situation, it is difficult not to conclude that they must also differ in the way they subjectively experience and think about it. Objective evidence of such subjective differences at the individual level can be sought in nonlinguistic behavior, e.g., sorting or remembering arrays of objects differently in line with cross-linguistic differences in grammatical categories that are obligatorily activated in relation to a particular referential domain. Demonstrations of linguistic effects in early cognitive development by Slobin (1996) and Bowerman (1996) are particularly convincing, given that linguistic influences on patterns of attention to, and classification of, experiential data should


264 Penny Lee

not be available prior to language acquisition and might be expected to show changes over time as the internalized system develops. With regard to the much larger ‘worldview’ question, Whorf was very much more cautious than has often been claimed. He considered that it is only where a number of patterns of linguistically generated differences, both lexical and grammatical, exist and only where these have been in place as “fashions of speaking” within a cultural tradition for an extended period of time (Whorf 1941c: 159), that we can begin to answer the broad question of whether there are “traceable affinities between (a) cultural and behavioral norms and (b) large-scale linguistic patterns” (p. 138), i.e., the ‘worldview’ question. This point is frequently overlooked. Whorf ’s use of the isolate construct to provide a realist basis (in terms of the cognitive processing constraint involved in figure/ground differentiation) for the referential value of morphemes seems to have anticipated more recent insights into the expression of similar constraints at discourse level. Given that the figure/ground phenomenon is relevant in terms of focus and periphery at every level at which attention can be focused, abstraction of ‘situations’, ‘events’ and ‘referents’ from ongoing experience as ‘ideas’ (Chafe 1994), or ‘conceptual events’ (Grace 1987), can also be thought of as involving isolates of experience, albeit of a more comprehensive kind than Whorf ’s. What Chafe, with his discovery that intonation units present ‘one new idea at a time’, and Pawley & Syder (1983), with their delineation of the ‘one clause at a time’ processing constraint, bring out more explicitly than Whorf is the degree to which linguistic structure reflects temporal as well as focusing and differentiating restrictions. The value of these insights in relation to the linguistic relativity question is that they further clarify the nature of the dialectical tension between universals of human cognitive processing and cultural variation in the way significance is abstracted from the ongoing flow of experiential data. Whorf maintained that it is imperative that linguists collect as much information as possible about the world’s languages so that the diversity of ways humans segment experience linguistically and build culturally-specific systems of logic based on their grammars can be mapped. He provided detailed categories of analysis and ethnolinguistic exemplars in his outline and report as guides to help free linguists from the influence of categories of grammar and reference entrenched during their own linguistic socialization. The critical point, he considered, is not to be alarmed about the existence of linguistic constraints, but to celebrate, through “a linguistic consideration of thinking” (Whorf 1937: 81) “what the human mind can do” (Whorf 1940a: 215). Whorf also believed that the fact of linguistic relativity helps us come to terms with the understanding that: “Whenever agreement or assent is arrived at in human affairs, and whether or not mathematics or other specialized symbolisms are made part of the procedure, THIS AGREEMENT IS REACHED BY LINGUISTIC PROCESSES, OR ELSE IT IS NOT REACHED” (Whorf 1940a: 212 orig. caps.). We will not, he

Benjamin Lee Whorf 265

said, enjoy the same “picture of the universe” unless our “linguistic backgrounds are similar, or can in some way be calibrated” (p. 214). He (p. 211) was impressed by: “The amazingly complex system of linguistic patterns and classifications” that people “must have in common before they can adjust to each other at all”. Three kinds of calibration can be considered (Lee 1996: 225–228). Full calibration of understanding involves attending to the detail of what we say, first checking to see that both interlocutors are talking about ‘the same thing’ in ordinary terms, e.g., whether ‘next Friday’ means the one ‘this’ week or ‘next’ week, a matter of usage. Secondly, calibration involves checking resonances entrenched in interlocutors’ internalized systems, e.g., ‘jargon’ in my idiolect is a relatively technical term while for someone else it might have strongly pejorative connotations. Finally, in more technical cross-linguistic terms, calibration involves checking isolates of meaning, considering what isolates of experience they might reflect, and determining degree of overlap with ‘bits of experience’ picked out in other languages and in danger of being lost in translation. 5.  Whorf’s influence There was heated debate about Whorf ’s theories even during his lifetime, but not in relation to linguistic relativity. As noted earlier, the articles most significant in popularizing those ideas were published only in the last months of his life (or posthumously). Even his most widely known piece, the memorial for Sapir (Whorf 1941c), was not published until the year he died, though written two years earlier. No, the controversies he generated in the 1930s and 40s were over his strenuously promoted, well published (see bibliography in Whorf 1956), and provocative opinion that Mayan hieroglyphs include phonetic elements, a view since vindicated, although not in terms of his reasoning. Even if it had been more widely known, much of what he wrote about linguistic relativity fitted the Herder/Humboldt tradition of inquiry into relationships between language, experience, and thought. (See Koerner 1995 for a comprehensive overview and bibliography of relevant literature and Lucy 1992 for discussion of Boas’ and Sapir’s treatments of the ‘linguistic analysis of experience’ issue). Even Whorf ’s pioneering ideas about the nature of the internalized system may be understood in part as developing the notion of the ‘inner form’ of a language, which also goes back to earlier European thought. It was not until after 1949, when Trager, through the Foreign Service Institute, published four, then in 1952, a total of five of Whorf ’s articles as a booklet for use in cross-cultural training, that controversy about linguistic relativity began to escalate. As well as using these articles in training, the FSI made them available to the general public for several years. Further popularization occurred when the General Semanticists

266 Penny Lee

began to promote Whorf ’s work through their journal and elsewhere (see Joseph 1996 for an interesting account of Whorf ’s links to this group). A flurry of research activity involving psychologists as well as anthropological linguists in the early 1950s built on Whorf ’s ideas and continued into ensuing decades, albeit under increasing pressure from generative linguistics. The most detailed coverage and evaluation of the linguistic relativity research tradition is found in Lucy (1992) while excellent reviews by Hill (1988) and Hill & Mannheim (1992) are also recommended. Conferences dealing explicitly with Whorf ’s ideas were also held in the early 1950s (Lee 1996: 17–18). Hoijer’s (1954) report of the 1953 Language in Culture conference, attended by philosophers, psychologists, and anthropologists as well as linguists, makes especially interesting reading as it includes transcripts of discussions as well as articles by the main participants. The interdisciplinary nature of Whorf ’s thinking attracted a diverse following, but perhaps the most difficult hurdle for many of his admirers was to escape the confines of their own discipline’s worldview when tackling his ideas. His background in the physical sciences was also an impediment to many although, as we have seen, its influence needs to be taken into account if he is to be properly understood. The term ‘Sapir-Whorf Hypothesis’ came into regular use around this time and is probably still the most frequently used, in spite of the fact that neither Whorf nor his predecessors regarded what they observed about the way languages differentially isolate universally available aspects of experience for utilization within their respective systems of reference as anything other than an empirical matter, i.e., not a hypothesis (Alford 1981; Grace 1987). A scientific principle is a general law based axiomatically on evidence held to be incontestable. Few linguists or translators would deny that languages do provide differently configured resources for talking about experience and that although translation is possible, it often involves circumlocution, non-idiomatic formulations and loss of conceptual subtleties in favor of others. Whorf developed his reasoning by exploring and elaborating on such facts. A distinctive feature of Whorf ’s influence has been that, more often than not, it seems to have been mediated by secondary commentary rather than direct access to his own writings. This mediation includes several attempts to systematize debate about linguistic relativity. For instance, while Lenneberg & Roberts (1956) argued that Whorf was misguided in several respects, they also introduced the notion of ‘the language of experience’ as a basis for empirical research and discussion. In doing so they came interestingly close to the isolate idea, while managing to convince readers that Whorf ’s reasoning did not warrant closer attention. Lenneberg’s insights about ‘codification’ were particularly significant in subsequent research on color, probably the most robust of the sub-fields of ‘Whorfian’ research over several decades. (Whorf himself did not make an issue of the cross-linguistic treatment of color.) In 1959 Trager offered a more comprehensive systematization of ‘the Whorf hypothesis’, again ignoring Whorf’s logic, this time in favor of his own schematization

Benjamin Lee Whorf

of linguistics, including ‘metalinguistics’ as the field of inquiry about relationships between language and the rest of the world. Trager argued that a ‘point by point’ linking of features of language with related features of culture, arrived atthrough rigorously structured ethnography, was needed as the basis for further inquiry. In the meantime, Brown’s (1958) chapter on ‘linguistic relativity and determinism’ was becoming well known, including its discussion of a ‘strong’ and ‘weak’ version of the ‘hypothesis’ and, by implication, the phenomenon itself. An extremely influential debunking of Whorf by philosopher Black (1959) was also widely read in the 1960s and 70s, its more extreme misreadings of Whorf waiting to be identified by Ellis (1993), although they had also been highlighted in the 1991 thesis version of Lee (1996) which includes detailed evaluations and refutations of Black. Fishman’s (1960) systematization, prepared for behavioral scientists, was particularly sophisticated and influential in sorting out several relevant variables and providing yet another framework for setting up and evaluating empirical research. Discussing impacts of ‘verbal habits’ on ‘other kinds of behavior’, he stressed, a little confusingly, that ‘the Whorfian hypothesis’ did not concern ‘verbal determinism in general’ as much as ‘relative linguistic determinism’, i.e., effects of contrasting characteristics of specific languages. In arguing this way, he missed the significance of Whorf ’s point about the entrenchment of all linguistic patterning in cognition and its pervasive effects. His ‘four levels of the Whorfian hypothesis’ elaborate the strong/weak thesis in some detail, a discussion also continued authoritatively by Penn (1972) in a monograph originally prepared in 1966. Penn (p. 10) identified the ‘extreme hypothesis’ as the notion that “thought is dependent on language” and the ‘mild’ that “the categories of a language influence the cognition of its speakers”. Her discussion is historically and philosophically grounded in more detail than many others of the time. It focuses on the relevance of innatism to relativity but suffers from problems pointed out by Grace (1987: 118) who noted that many Whorf commentators have been faced “with the dilemma of either finding a way of setting aside the questions raised by Whorf or having their own assumptions … called into question”. Failure to notice Whorf ’s focus on linguistic thinking is again fundamental. Hymes’ (1966) argument that a ‘relativity of use’ needs to be taken into account is particularly relevant to pragmatics. Whilst much of what he says about the matter is thought-provoking, it is not explicitly related to Whorf ’s original, physics-based definitions. As in much of the rest of the literature, one also suspects at times that the word ‘relative’ is simply being used to mean ‘different’. Certainly, different people, even within the same society, do use language differently, both cognitively and socially. The same language may be used for different purposes in different societies and different languages may be used for different purposes within one society. All these phenomena need to be investigated for a fuller understanding of language. What also needs to be clarified is precisely what cognitive processes or effects are being posited in discussions


268 Penny Lee

about differences in use, how they are presumed to operate, and how they may intensify, overlay, or mitigate the basic selection and abstraction of experiential isolates as a function of linguistic resources. Silverstein (e.g., 1979) is another whose influence has been considerable. Like Hymes, he is justifiably credited with significantly helping to keep interest in Whorf ’s ideas alive through decades of denigration. Referring to Whorf as “one of the most misunderstood writers of the century” (1979: 193), Silverstein’s claims that he was extending Whorf ’s thinking nevertheless seem at times to have more to do with his own undoubted intellectual creativity and linguistic perspicacity than Whorf ’s original ideas. Some misreadings in the 1979 article include reference (p. 194) to Whorf ’s use of Einstein’s relativity principle as an “unfortunate metaphor”. The alternative term ‘uncertainty principle’ is advocated without, however, an explanation of precisely how (Heisenberg’s?) principle is “in keeping” with Whorf ’s original formulation. Again, while defining linguistic ideologies as “any set of beliefs about language articulated by the users as a rationalization or justification of perceived language structure or use” (p. 193) Silverstein overlooks an essential element of Whorf ’s argument. Whorf ’s point was that such activity is not generally able to be articulated as a set of beliefs, although deliberate fostering of metalinguistic awareness may facilitate such articulation and partially overcome the unconscious effects of unexamined language patterning in cognition. Silverstein’s insights into constraints on metalinguistic awareness are, however, pertinent and useful and much of his work does indeed extend Whorf ’s thinking in productive ways. Other significant publications in the1970s include the (1976) Pinxten collection on universalism versus relativism which, among other interesting papers, brought Gipper’s contestation of Whorf’s allegations about an alternative space/time conceptualization of the universe embodied in the Hopi language to the English-speaking world. Rossi-Landi’s (1973) avowedly Marxist and sometimes overdrawn overview of the relativity debate was also known in some quarters. Although Malotki’s (1983) report on Hopi claimed to demolish Whorf’s reasoning permanently, Lee (1991) showed that in one area at least, Malotki’s data support Whorf’s reasoning and clarify some of his claims. In 1984 Hasan offered a perceptive and sophisticated application and extension of Whorf ’s ideas in an essay on the use of language as a resource to deal with what it, itself, creates. Drawing on Hjelmslev, Wittgenstein, Halliday, Rorty, and others, she deconstructed some common misreadings of Whorf and invited readers to take him seriously. Grace’s (1987) exploration of ‘the linguistic construction of reality’ (contrasted with the ‘mapping view’) makes explicit links to Whorf while Schultz (1990) brought Whorf and Bakhtin into engagement in a provocative way. Lakoff (1987) offered readers of his popular book a rather equivocal but basically admiring treatment. Pinker (1994), with his carelessly documented and sweeping opinions about Whorf ’s ideas, based apparently only on secondary and out-of-date

Benjamin Lee Whorf 269

sources, stands out rather anachronistically in the context of the new surge of interest in Whorf (see Lee 1996: 19–20, 139, 154 for refutations of his claims). In retrospect, there have been four main problems in relation to Whorf ’s ideas. First, his own explicit definitions of the linguistic relativity principle have rarely been taken as foundational. Even when quoted, they are often passed over in favor of alternative formulations developed by the commentators. Lucy (1992) and the 1996 Gumperz & Levinson volume provide recent major examples of this tendency. Foley (1997), by contrast, explains Whorf ’s logic carefully but then, somewhat less convincingly, attempts to weld Lucy’s reasoning about analogical structure (see below) to that discussion. Duranti (1997) sidesteps most of what is important in Whorf ’s contribution; a disappointing vagueness characterizes this section of an otherwise useful book. Whilst none of these authors had access to Lee (1996), Lee (1994) was available to some, although not used. Secondly, the Sapir memorial article, with its arresting anecdotes about insurance claims and its discussion of the ramifications of large-scale fashions of speaking in Hopi and European societies, is too often taken as the central document with respect to linguistic relativity, even though crucially significant additional articles that include Whorf’s definitions, are found both in the 1956 collection and the Yale archives. Lucy (1992) developed a detailed theoretical account of the workings of the ‘hypothesis’ based on Whorf’s discussion in that article of analogical processes but dismissed much of value in the Technology Review articles. Although much of Lucy’s reasoning is valid in its own terms, it is worth noting that analogical processes that manifest in language have more to do with the projection of established patterns of meaning-making onto experience than the original segregation and abstraction of isolates from experiential data. The mutually reinforcing relationship between abstractive and projective activity in linguistic thinking is dealt with in Lee (1996: 198) on the basis of points made by Sapir and Whorf. The third problem which often diverts attention away from what Whorf really said about linguistic thinking is the unnecessary dichotomizing of ‘language’ and ‘thought’ as separate phenomena in order to show definitive ‘influences’ one way or another (see Lee 1996: 66–72). A good deal of sloppy reasoning can result where precise definitions are not used as the basis for discussion, but even where clarification is included, readers are often left in the dark about the relationship of what is said to Whorf ’s original reasoning. The fourth problem has to do with arguments about relativism versus determinism and the notion that there are ‘strong’ and ‘weak’ versions of the ‘hypothesis’. (Note the incongruency of talking about strong and weak versions of a principle.) Although Whorf himself made seemingly contradictory statements that are used to license this debate, careful reference to the context of each contentious remark will often reveal modulations which make a difference. For Whorf, the fact that humans think did not mark what is most distinctive about the species. “The beasts may think, but they do not talk” he said and, further: “There is

270 Penny Lee

no need to apologize for speech, the most human of all actions” (Whorf 1940b: 220). What is distinctive about human cognition, as we have seen, is ‘thought insofar as it is linguistic’. He warned also that: “Words and speech are not the same thing” and, in a possibly rather Fregean manner, that “the patterns of sentence structure that guide words are more important than the words” (Whorf 1942: 253). In comments written just a few months before he died he seems to anticipate some of the foundational insights of modern pragmatists when he states that: Even the lower mind has caught something of the algebraic nature of language; so that words are in between the variable symbols of pure patternment … and true fixed quantities. That part of meaning which is in words, and which we may call “reference,” is only relatively fixed. Reference of words is at the mercy of the sentences and grammatical patterns in which they occur. And it is surprising to what a minimal amount this element of reference may be reduced. The sentence “I went all the way down there just in order to see Jack” contains only one fixed concrete reference: namely “Jack.” The rest is pattern attached to nothing specifically; even “see” obviously does not mean what one may suppose, namely, to receive a visual image. (Whorf 1942: 259)

References Alford, D.K. (1981). Is Whorf ’s relativity Einstein’s relativity? Proceedings of the Berkeley Linguistics Society 7: 13–26. Black, M. (1959). Linguistic relativity: The views of Benjamin Lee Whorf. Philosophical Review 68: 228–238. Bowerman, M. (1996). The origins of children’s spatial semantic categories: Cognitive versus linguistic determinants. In J. Gumperz & S. Levinson (Eds): 145–176. Brown, R. (1958). Linguistic relativity and determinism. In Words and Things, Ch. vii. The Free Press. Carroll, J.B. (1956). Introduction. In Whorf (1956): 1–34. Chafe, W. (1994). Discourse, consciousness and time: The flow and displacement of conscious experience in speaking and writing. The University of Chicago Press. Duranti, A. (1997). Linguistic Anthropology. Cambridge University Press. Ellis, J.M. (1993). Language, thought and logic. Northwestern University Press. Fishman, J.A. (1960). A systematization of theWhorfian hypothesis. In Behavioral Science, 5, 323–339. Foley, W.A. (1997). Anthropological linguistics: An introduction. Blackwell. Grace, G. (1987). The linguistic construction of reality. Croom Helm. Gumperz, J.J. & S. Levinson (Eds) (1996). Rethinking linguistic relativity. Cambridge University Press. Harré & Krausz (1996). Varieties of relativism. Blackwell. Hasan, R. (1984). What kind of resource is language? Australian review of applied linguistics 7(1): 57–85. Hill, J.H. (1988). Language, culture and world view. In F.J. Newmeyer (Ed.), Linguistics: The Cambridge survey. Vol. 4. Language: The socio-cultural context: 14–36. Cambridge University Press.

Benjamin Lee Whorf

Hill, J. & B. Mannheim (1992). Language and world view. In Annual Review of Anthropology 21: 381–406. Hockett, C.F. (1987). Refurbishing our foundations: Elementary linguistics from an advanced point of view. John Benjamins. Hoijer, H. (1954). Language in culture: Conference on the interrelations of language and other aspects of culture. University of Chicago Press. Hymes, D. (1966). Two types of linguistic relativity. In W. Bright (Ed.), Sociolinguistics: Proceedings of the UCLA sociolinguistic conference, 1964; 114–167. Mouton. Joseph, J. (1996). The immediate sources of the ‘Sapir-Whorf hypothesis’. Historiographia Linguistica 23: 365–404. Koerner, E.F.K. (1995). The Sapir-Whorf hypothesis: An historico-bibliographical essay. In Koerner (Ed.), Professing Linguistic Historiography: 230–240. John Benjamins. Lakoff, G. (1987). Women, fire, and dangerous things: What categories reveal about the mind. Stanford University Press. Lee, P. (1991). Whorf ’s Hopi tensors: Subtle articulators in the language/thought nexus? Cognitive Linguistics 2(2): 123–147. ——— (1994). New work on the linguistic relativity question. Historiographia Linguistica 21(1/2): 173–191. ——— (1996). The Whorf theory complex: A critical reconstruction. John Benjamins. ——— (1997). Language in thinking and learning: Pedagogy and the new Whorfian framework. Harvard Educational Review 67(3): 430–471. ——— (2000). When is ‘linguistic relativity’ Whorf ’s linguistic relativity? In M. Pütz & M. Verspoor (Eds), Explorations in linguistic relativity: 45–68. John Benjamins. Lenneberg, E.H. & J.M. Roberts (1956). The language of experience: A study in methodology. Waverley Press. Lucy, J.A. (1992). Language diversity and thought: A reformulation of the linguistic relativity hypothesis. Cambridge University Press. Malotki, E. (1983). Hopi time: A linguistic analysis of the temporal constructs in the Hopi language. Mouton. Pawley, A. & F.H. Syder (1983). Natural selection in syntax: Notes on adaptive variation and change in vernacular and literary grammar. Journal of pragmatics 7: 551–579. Penn, J.M. (1972). Linguistic relativity versus innate ideas: The origins of the Sapir-Whorf Hypothesis in German thought. Mouton. Pinker, S. (1994). The language instinct: The new science of language and mind. Allen Lane. Pinxten, R. (1976). (Ed.), Universalism versus relativism in language and thought. Mouton. Rossi-Landi, F. (1973). Ideologies of linguistic relativity. Mouton. Sapir, E. (1921). Language: An introduction to the study of speech. Harcourt Brace & World. ——— (1925). Sound patterns in language. Language 1, 37–51. Schultz, E.A. (1990). Dialogue at the margins: Whorf, Bakhtin and linguistic relativity. University of Wisconsin. Silverstein, M. (1979). Language structure and linguistic ideology. In P.W. Clyne & W. Hanks (Eds), The elements: A parasession on linguistic units and levels: 193–247. Slobin, D.I. (1996). From “thought and language” to“thinking for speaking”. In Gumperz & Levinson (Eds): 70–98. Trager, G.L. (1959). The systematization of the Whorf hypothesis. In Anthropological Linguistics 1(1): 31–35.


272 Penny Lee Whorf, B.L. (1937). A linguistic consideration of thinking in primitive communities. In Whorf (1956): 65–86. ——— (1938). Language: Plan and conception of arrangement. In Whorf (1956): 125–133. ——— (1940a). Science and linguistics. In Whorf (1956): 207–219. ——— (1940b). Gestalt technique of stem composition in Shawnee. In Whorf (1956): 160–172. ——— (1941a). Linguistics as an exact science. In Whorf (1956): 220–232. ——— (1941b). Languages and logic. In Whorf (1956): 233–245. ——— (1941c) [written 1939]. The relation of habitual thought and behavior to language. In Whorf (1956): 134–159. ——— (1942). Language, mind, and reality. In Whorf (1956): 246–270. ——— (1956). Language, thought and reality: Selected writings of Benjamin Lee Whorf. Edited by J.B. Carroll. M.I.T. Press. ——— (1979). Benjamin Lee Whorf papers. Manuscripts and archives. Yale University Library. Whorf, B.L. & G.L. Trager (1938). Report on linguistic research in the department of Anthropology of Yale University for the term Sept. 1937 to June 1938. In B.L. Whorf (1979).


A Aarsleff, H.,  174, 181 Abe, K.,  20 aboriginal languages,  43–45 Abramov, I.,  250 abstract representation,  61 accessibility and practicality of cultural scripts,  77–78 active metaphor,  74 adjacency pair formats,  22 Agar, M.,  7, 8, 11, 110, 112, 117, 137, 186, 188, 189, 196, 220 Agha, A.,  156, 161, 163, 164, 166, 167, 168, 169, 170 aisatsu,  18 aisatsu, types of,  19 aisatsu expressions,  20 aisatsu in school,  23–4 aisatsu in workplace,  24–25 aisatsu routine formulae,  21 Alford, D.K.,  263, 266 Allwood, J.,  198 Ameka, F.K.,  72, 76 American Indian linguistics,  31, 44, 46, 47, 236, 237, 249 American structuralism,  9, 47 Americanist text tradition (Sapir),  238–239 Americanist tradition,  42–43 Amerindian languages, classification of,  31 Amharic personal pronouns,  158 Ammon, U.,  108 Anderson, L.,  193 anecdotal approach,  196 Anglo-American cultural script,  75 Anglo-American English,  71 anthropological bias,  111 anthropological fieldwork,  133 anthropological linguistics,  1, 2, 6, 8, 10–11, 14, 15, 29–39, 92, 131, 220, 232, 234, 244, 248, 257

anthropological linguistics, history of,  30 anthropological linguistics, types of research,  30–31 anthropological linguistics-linguistic anthropology distinction,  29 anthropology,  1 anti-racism,  41, 42 anti-Semitism,  41–42 Appadurai, A.,  207 applied linguistics,  149 Apresjan, J.,  62 areal cultural script,  72 articulation and reflection,  175, 177 Asad, T.,  88, 94 Aston, G.,  193 Atran, S.,  52 Auer, P.,  129, 195, 197 Aunger, R.,  207 Austin, J.L.,  10, 227, 228 B Baayen, H.,  53 Back, K.W.,  204, 207, 212 Baker, L.D.,  41 Bakhtin, M.M.,  83, 93, 111, 245, 268 balasan,  75–76 Barley, N.,  137 Barron, A.,  191 Barth, F.,  88 Barthes, R.,  81 Bartlett, F.C.,  142 ‘basic color terms’ (BCTs),  249 Bassili, J.N.,  146 Basso, K.M.,  127 Bauman, R.,  96 Bauman, Z.,  83 Baumann, R.,  218 Becker, A.L.,  253 Beeman, W.O.,  162 behaviorist approach,  85–86

‘belief script,’  75 Benedict, R.,  45, 46, 82, 87, 213, 236 Berlin, B.,  34, 37, 51, 106, 248–253 Berlin, E.A.,  250 Berman, J.,  43 Berman, R.A.,  107 Berry, J.W.,  12 Bickerton, D.,  8, 53 B&K’s seven evolutionary stages,  250 Black, M.,  248, 267 Bloch, M.,  36, 96 Blommaert, J.,  100, 188, 191, 197, 198 Bloomfield, L.,  30, 47, 249, 258, 261 Blount, B.G.,  1, 10, 29, 36, 37 Blum-Kulka, S.,  98, 106, 128, 193 Boas, biography of,  41–42 Boas, F.,  1, 7–12, 30, 32, 41–48, 50, 52, 85, 95, 111, 121, 125, 132, 137, 185, 215, 235, 237–241, 258, 265 Boas’s fieldwork,  43 Boas’s work on history and psychology,  45 Boguslawski, A.,  62 Bohannan, L.,  137 Bolinger, D.,  143 Boster, J.S.,  252 Bouquiaux, L.,  137 Bourdieu, P.,  196, 206 Bowerman, M.,  53, 263 Boyer, P.,  52 Braudel, F.,  81, 82, 84, 87 Breedlove, D.,  251 Breedveld, A.,  72 Brenneis, D.,  36 Brenner, M.,  203 Briggs, C.L.,  7, 117, 202, 205, 206, 207, 244 Bright, J.O.,  251 Bright, W.,  251

274 Index Brislin, R.,  99 Brown, P.,  13, 53, 98, 170, 193, 231 Brown, R.D.,  146, 165, 166, 169, 249, 250 Bryant, A.T.,  161 Bühler, K.,  257 Burgess, D.,  250 Burling, R.,  60, 249, 253 Byrne, R.,  53 C Calhoun, C.,  207 Campbell, L.,  44 Capo, H.C.,  164 Carbaugh, D.,  95, 124 Carey, J.,  125 Carrithers, M.,  84 Carroll, J.B.,  257, 259 Casagrande, J.B.,  249 Casson, R.,  51 Chafe, W.,  107, 264 ChiBemba honorific and pejorative prefixes,  159–160 Chinese cultural script,  73, 75 Choi, S.,  53 Chomsky, N.,  8, 9 Chomskyan linguistics,  8, 10 Chomsky’s theory,  9 Chomsky’s universalist-psychological model of syntax,  250 Cicourel, A.V.,  203, 204 Clancy, P.,  23 Clark, H.H.,  204 Clarke, D.D.,  3–6, 173, 181, 217–220 Clifford, J.,  92, 93, 94, 125, 206 cognitive anthropology,  11, 50–56, 112, 248, 249 cognitive anthropology, historical background,  50–53 cognitive anthropology reconstrual,  53–56 Cognitive Anthropology Research Group,  11, 53, 107 cognitive sciences,  2, 13, 50, 51, 52, 53–54 cognitive turn,  2 Colby, B.N.,  248 Cole, M.,  12, 52 color term research,  34, 37 communication and culture,  188

communicative competence,  206 communicative genres,  194 communicative style,  71–72 componential analysis (CA),  58 Comrie, B.,  105 Conklin, H.C.,  51, 110, 248, 249 Conte, M.-E.,  180 context and meaning,  110, 116 context of situation,  7, 111, 141–143, 185 contextual boundaries,  21–22 contextual frame of reference,  141 contextualization approach,  195 contextualization cues,  195 Contini-Morava, E.,  239, 240 contrastive and interactional approaches,  192–194 contrastive pragmatics,  194 conversation analysis,  111, 128 Cook-Gumperz, J.,  187 Coseriu, E.,  58 Coulmas, F.,  18, 23, 230 Coupland, N.,  19, 27 Cowan, W.,  234 Craig, C.G.,  108 Cross, T.S.,  204, 207 cross-cultural comparison,  98 cross-cultural pragmatics,  68, 97, 194 cross-cultural psychology,  12, 50 cross-cultural semantics,  68 Cross-Cultural Speech Act Realization Project (CCSARP),  128, 192 cross-cultural studies,  98, 99 cross-language scheme,  123 Crystal, D.,  229 cultural acts,  18 cultural analysis,  91–92 cultural determinism,  95 cultural difference, politics of,  89–91 cultural identity,  90, 91 cultural imperialism,  88, 90 cultural key words,  68, 69, 73 cultural otherness,  88–89 cultural relativism,  8, 46, 95 cultural script for name avoidance,  72 cultural scripts,  68 cultural studies,  83, 89

cultural studies ethnography,  125–126 culture,  51, 84, 85 culture, historical transformation of,  81–83 culture and language use,  1, 7, 8 culture as communication,  95–96 culture as ideology,  87–88 culture as interdisciplinarity,  81 culture-in-communication, loci of,  189–192 culture in intercultural communication,  188–189 ‘culture is a verb,’  87 ‘culture shock,’  137 culture-specific concepts,  72–74 culture-specific speech act,  193 D D’Andrade, R.,  69 Danziger, E.,  53 Darnell, R.,  8, 41, 42, 44, 45, 46, 234–239, 245 Dasen, P.R.,  12 data collection,  105, 117, 131–132, 135–136 data-gathering methods,  69, 134 Davidson, D.,  125 Davies, E.,  187 De León, L.,  253 De Revai, M.L.,  222 De Valois, R.L.,  250 deference,  157 deictic forms,  170 Delafosse, M.,  164 Dennett, D.,  55, 56 denotational fading,  168 deployment and ideology of honorifics,  165–167 Derrida, J.,  90 descriptive-structural linguistics,  8 developmental pragmatics,  13 developmental psychology hypotheses,  12 di Luzio, A.,  195 dialect and language,  152 Dijkstra, W.,  204 directives,  123 Dirim, Ý.,  197 Dirven, R.,  192

Index 275

discourse, susceptible to conscious formulation, characteristics,  204 discourse approach,  118–119 discourse completion tests,  106, 193 discourse strategies,  111 discursive constitution of institutional power,  206 Dixon, R.W.M.,  108, 137, 168 Doke, C.,  163 domain,  248 Dougherty, J.W.D.,  250, 251, 252 Drechsler, E.J.,  237 Dunbar, R.,  7 Duranti, A.,  1, 13–14, 22, 27, 37, 38, 108, 116, 129, 135, 159, 167, 170, 220, 232, 269 Durham, W.,  55 E Edwards, J.R.,  150 Eelen, G.,  191 Efron, D.,  190 Ehlich, K.,  227 Eibl-Eibesfeldt, I.,  230 Eksner, H.J.,  198 elicit speech data,  108 elicitation,  105 elicitation methods,  106, 134 elicitation of morphological and syntactic patterns,  106 Ellen, R.F.,  135, 213 Ellis, J.M.,  215, 267 emotion concepts,  63 Enninger, W.,  187 Erickson, F.,  187, 190 Erickson, J.,  237 Errington, J.J.,  158, 163, 164, 166, 170 Ervin-Tripp, S.,  12 ethnic capital,  196 ethnic categorization,  196 ethnicity,  90, 122, 125, 128, 190, 196, 197, 198 ethno-linguistic theory,  6 ethno-pragmatics,  194 ethnobiological nomenclature and classification,  35, 37 ethnographic fieldwork, goal of, avenues,  216 ethnographic interviews,  112–114

ethnographic practice risks minimization,  126 ‘ethnographic record,’  228 ethnographic role in pragmatics,  111–12 ethnographic samples,  117–118 ethnographic semantics,  33–35, 37 ethnographic theory of language (Malinowski),  226 ethnography,  33, 93–94 ethnography and language,  111, 116 ethnography by contrast,  117–118 ethnography by example,  112–117 ethnography of communication,  33 ethnography of speaking,  10, 11, 33, 35, 36, 111, 121–129, 132, 148, 185, 228, 244 ethnolinguistics,  217, 256 ethnomethodology,  111 ethnopragmatics,  77 ethnoscience,  11, 33, 51, 248, 249 evaluative components,  70 Ewe cultural script,  76 explications,  63–64 F Fabian, J.,  92–94, 192, 198 Fairclough, N.,  197 fanatical theoretical empiricist,  216 Featherstone, M.,  83 Feld, S.,  13 Fichte, J.G.,  3, 173, 175, 176, 181 fieldwork,  131, 132 fieldwork method,  92 Fiesel, E.,  182 Firth, J.R.,  7, 111, 140–144, 210, 213, 215, 217–220, 223, 228 Firthian linguistics,  7, 140 Firthians and Neo-Firthians,  143 Firth’s approach to language,  142 Firth’s approach to meaning,  143 Fischer, M.M.J.,  93 Fishman, J.A.,  96 Fitch, K.,  10, 121, 124 FL awareness,  148

Foddy, W.,  204, 207 Fodor, J.A.,  60 Foley, W.A.,  1, 14, 106, 136, 263, 269 folk and linguistic theories,  152 Folk Linguistics,  146, 151, 154 folk linguistics and language attitudes,  147 folk names,  26 folk terminology and classification,  34–35 folk theories of pragmatics,  153 folk theory of language,  152, 153 foreigner talk,  190 form-feeling (Sapir),  240–241 Formal interviews,  203 formal/structured interviews vs. informal, unstructured ones,  202 formal styles of articulation,  106 Formigari, L.,  176, 181 Foucault, M.,  206 Fought, J.,  42, 47, 238, 241 Frake, C.O.,  248 “free, aimless social intercourse,”  226 French, D.,  251 Fuller, N.,  190 functionalism,  215 G Garcia, O.,  98 Gardner, P.,  228 Garfinkel, H.,  132 Garro, L.C.,  248 Gass, S.M.,  191 Gavioli, L.,  193 Geckeler, H.,  58 Geddes, W.R.,  250 Geertz, C.,  84, 86, 93, 165 Gellner, E.,  82 generative CA,  60–61 generative semantics,  60 genuine culture,  243 Georgiou, M.,  189 Gilroy, P.,  89 Ginzburg, C.,  83 Gipper, H.,  181 Glaser, B.,  118 Gleason, H.A.,  249 goaisatsu,  21, 22

276 Index Goddard, C.,  7, 58, 62–65, 68–70, 72–76, 78 Goffman, E.,  23, 193 ‘going native,’  137 Goldberg, D.T.,  90 Goldschmidt, W.,  131 Gonda, J.,  169 Goodenough, W.H.,  51, 60, 188, 248 Goodman, M.E.,  84 Goodwin, C.,  38, 129, 136, 220, 232 Goody, E.N.,  53 Goody, J.,  52 Grace, G.,  264, 266, 267 grammatical elements,  239 Gramsci, A.,  85, 89 Greenberg, J.H.,  220 greeting formulae, meaning of,  230 greeting rituals,  18 Greimas, A.J.,  58 Grimshaw, A.D.,  203 grounded theory,  118 Groves, R.,  207 Gudykunst, G.B.,  99 Gumperz, J.J.,  10, 12, 33, 35, 36, 38, 99, 111, 122, 132, 187, 190, 194, 195, 197, 244, 269 Günthner, S.,  194, 195 H Habermas, J.,  182, 207 Hage, P.,  250, 251 Hall, E.T.,  97, 99, 186, 190 Hall, S.,  89, 90, 125 Halliday, M.A.K.,  111, 143, 144, 268 Halliday’s functionalism,  144 Handbook of American Indian Languages,  45–46 Handler, R.,  234, 235, 236, 238, 239, 240, 242–244 Hanks, W.F.,  244 Hannerz, U.,  116 Harada, S.I.,  160 Haraway, D.,  203 Harkins, J.,  63 Harré,  263 Harris, Z.S.,  234, 242 Hasada, R.,  72, 77 Hasan, R.,  268 Haviland, J.B.,  168 Hawkes, K.,  250, 251

Haym, R.,  174 Hays, T.E.,  37, 252 Head, B.,  158 Healey, A.,  137 Heeschen, V.,  3, 4 Heider, E.R.,  250 Heinrich, A.C.,  250 Herbert, R.K.,  159, 162, 163, 170 Herder, J.G.,  1–3, 9, 12, 43, 174–176, 181, 265 Herzfeld, M.,  88 Hewes, G.W.,  190 Hewison, R.,  90 Hewitt, R.,  198 Hickerson, N.,  249 hierarchical relationship,  20 high-level scripts,  69–71 Hill, J.H.,  37, 159, 161, 170, 266 Hill, K.,  159, 161, 170 Hinnenkamp, V.,  10, 100, 185, 187, 190, 191, 192, 196, 198 Hjelmslev, L.,  58, 144, 249, 268 hlonipha,  162, 163 Hoben, S.J.,  158 Hockett, C.F.,  1, 15, 257 Hoenigswald, H.,  146, 148 Hofstede, G.,  98 Hoggart, R.,  82 Hoijer, H.,  32, 237 Holland, D.,  52 Holliday, A.,  197 honorific expressions,  159 honorific focus,  164 honorific items,  156 honorific pronouns,  157 honorific registers,  156, 161 honorific titles,  157 honorific tropes,  161 honorification and politeness,  157 honorification roles,  163–165 Horne, D.,  90 House, J.,  98, 128, 193 Howell, N.,  137 Hughes, E.C.,  110 Humboldt, W. von,  3–6, 8, 9, 11, 12, 43, 52, 111, 173–182, 185, 218, 237, 265 Humboldt and idéologues,  181–182 Humboldtian linguistic tradition,  237 Humboldt’s life,  173–174

Humboldt’s philosophy of language,  6, 174–175 Humboldt’s theory of language,  5 Hunn, E.,  37, 251, 252 Huspek, M.,  127 Hutchins, E.,  55 Huxley, J.,  215, 230 Hyde, M.,  197 Hyman, H.,  203 Hymes, D.H.,  10, 30, 35, 36, 42, 47, 111, 121, 122, 123, 132, 148, 169, 185, 206, 228, 229, 234, 238, 240, 241, 242, 244, 267, 268 I I-Thou,  180, 181 Ide, R.,  13, 18, 21, 25 Imai, T.,  153, 154 in-group/out-group boundaries,  20 indexicality,  39 indexicals,  39 Informal interviews,  203 Inoue, K.,  160, 164 institutional authority, interviewing and,  205–207 interactional sociolinguistics,  99, 194, 196, 244 intercultural communication,  10, 78, 185–198 intercultural pragmatics,  99, 100, 194 intercultural routine formulae,  26–27 Intermediate taxa,  251 International Pragmatics Association,  10 interview and features rendering speech susceptible to conscious analysis,  205 Interviewing,  202 interviews, pragmatic underpinnings of,  204–205 Irvine, J.T.,  13, 20, 156, 159, 162, 163, 165, 166, 169, 170, 236 J Jackendoff, R.,  60, 61 Jahoda, M.,  137

Index 277

Jakobson, R.,  227 Jakovidou, A.,  190 Japanese honorific verb morphology,  160 Japanese honorific verbal compounds,  160 Japanese ‘self,’  126 Javanese language levels,  158 Jaworski, A.,  19, 27 Johnson-Laird, P.N.,  65 Jordan, B.,  190, 205, 207 Joseph, J.E.,  204, 266 Jupp, T.C.,  187 K Kandiah, T.,  197 Karim, K.H.,  189 Karp, I.,  204 Kasper, G.,  98, 128, 193 Katriel, T.,  27, 124, 125, 127 Katz, J.J.,  60, 61 Kay, P.,  34, 37, 51, 106, 248, 249, 250, 252 Keenan, E.O.,  98 Keesing, R.M.,  81 Keim, I.,  187 Keller, J.,  52 Kendall, M.B.,  204 kinship,  33 kinship categories,  60 Klein, W.,  9 Klineberg, O.,  12 Klos Sokol, L.,  71 Kluckhohn, C.,  81, 84, 87 Knapp, K.,  97, 187, 191 Knapp-Potthoff, A.,  187 Knoll, J.H.,  174 Knoop, U.,  105 Koerner, E.F.K.,  265 Koerner, K.,  237 Kohl, K.-H.,  136, 220 Kondo, D.,  126, 127 König, R.,  203 Koole, T.,  187, 197 Korean cultural script,  73, 76 Kotthoff, H.,  191 krama speech,  165 Krauss, M.E.,  235 Krausz,  263 Kretzmann, N.,  177 Kristiansen, T.,  147 Kroeber, A.L.,  42, 84, 132, 235 Kroskrity, P.,  170, 204

Kullman, J.,  197 Kuschel, R.,  250 L Labov, W.,  10, 107, 108, 148, 203 Lado, R.,  98, 186 Lakoff, G.,  60, 248, 268 Langendoen, D.T.,  219, 220, 228 language, culture, and creativity,  179–180 language, culture and cognition,  6, 11, 12 language, dialogue, and pronouns,  180–181 language, Malinowski view of,  219 language and culture,  95–96, 185–186 language and languages,  178 language and thought,  9, 175–177, 181 language and world,  177–178 “language as a mode of action,”  232 language competence,  9, 190 “language conveying more than what is said,”  231 language-culture interrelationship,  96–97 language in cognition, role of,  9, 56, 258, 259 language learning,  186 language levels,  157 language universals,  178 language use,  38 language use, universal aspects of,  127 language variation and change,  148 Lave, J.,  52 Laver, J.,  230, 231 Leach, E.R.,  94, 213, 216, 221 Leach, J.W.,  221 Lebra, S.T.,  19, 20, 23, 24 Lee, P.,  9, 178, 256–270 Lee, R.B.,  251 Leech, G.,  58 Leeds-Hurwitz, W.,  238 Lehmann, A.,  137 Lehr, M.,  161, 164 Lehrer, A.,  58 Leibniz, G.W. Leitzmann, A.,  174

Lenneberg, E.H. Lévi-Strauss, C.,  42, 45 Levinson, S.C.,  11, 13, 33, 50, 53, 107, 157, 169, 193, 231 lexical alternants,  158 lexical meanings,  64 lexicon of color terms,  106 Lieberman, P.,  53 linguistic action verbs,  123 linguistic anthropology,  1, 8, 14, 29, 30, 47, 50, 52, 60, 111 linguistic borrowing,  161, 162 linguistic competence,  146 linguistic determinism,  95–6, 267 linguistic fieldwork,  133 linguistic forms,  43–44, 157–160, 170, 179, 239–241, 240 linguistic honorifics,  156 linguistic ideologies,  268 linguistic politeness,  151 linguistic pragmatics,  13, 56, 131, 186, 192 linguistic relativity principle,  9, 95, 241–242, 257, 259, 263–265, 269 linguistic responses,  149 linguistic thinking,  9, 258, 260, 267, 269 linguistics, defined,  1, 8 literary techniques of realism,  243 Locke, J.,  174, 175 Lonner, W.J.,  12 Loos, E.F.,  192 Lounsbury, F.G.,  51, 60, 248 Loveday, L.J.,  192 Lucy, J.A.,  32, 52, 53, 213, 245, 249, 265–266, 269 Lyons, J.,  58, 142, 227 M MacLaury, R.E.,  248, 250 Malay cultural script,  76 Malinowski, anthropological studies,  211 Malinowski, B.,  1, 6, 7, 9, 10, 12, 19, 92, 108, 111, 121, 125, 132, 137, 140, 141, 144, 210–223, 226–229, 231–232, 256 Malinowski, biographical sketch,  210–214

278 Index Malinowski, functional theory,  215 Malinowski, study of culture,  214–215 Malinowski’s linguistic interests,  218 Malinowski’s pragmatic theory of meaning,  7 Manchester, M.L.,  174, 176 Mandelbaum, D.G.,  96, 234 Mandelbaum, J.,  129 Mannheim, B.,  266 Mansfield, G.,  193 Marcus, G.E.,  93, 125, 206 Mayntz, R.,  108, 136 Mcclenaghan, P.,  150, 236 Mcdaniel, C.,  250 Mcveigh, B.,  24 Mead, M.,  45, 82, 92 Meeuwis, M.,  100, 187, 190, 191, 197 Meierkord, Ch.,  191 Melcuk, I.A.,  65 mentalist approach,  85 Merrifield, W.,  34, 37 metaphors,  110 metapragmatic of aisatsu,  25 metapragmatics,  18–19, 25–27, 150–151, 170, 245 methodological subdiscourses,  192–197 Métraux, R.,  82, 213, 214, 215 Mey, J.,  150, 151 Miedema, J.,  14 Miller, G.A.,  65, 251 Miller, W.,  65, 251 Milroy, L.,  150 miscommunication,  187–188 Mishler, E.G.,  117, 204, 205, 207 Mitchell, T.F.,  142, 144, 228 Mithun, M.,  43, 44, 46 model-theoretic sets of features,  248 Moerman, M.,  128, 129 Monaghan, J.,  144 Monberg, T.,  250 morphophonemics,  257 motion verbs,  59 Mueller-Vollmer, K.,  179, 181 Murray, S.O.,  42, 121 Myers, F.R.,  36

N Nahuatl verbal affixes,  159 Naotsuka, R.,  21 Native American languages,  29, 30, 31, 43 native-nonnative interaction,  191 native-nonnative speakership,  191 natural semantic metalanguage (NSM),  62–65, 68 Nerlich, B.,  3–6, 173, 175, 179, 181, 217–220 Newman, S.,  169 Nguni respect vocabularies,  162, 164 Nida, E.A.,  58, 59, 137 Niedzielski, N.,  10, 146, 150, 151 Njamal system,  60 North American languages,  43 noun classification systems,  159 NSM approach vs. CA methods,  62 ‘numeral classifiers,’  253 Nyce, J.M.,  238 O Oakley, A.,  206 O’Barr, W.,  36 observer’s paradox,  107 Ochs, E.,  13, 23, 36, 56, 98 Oesterreicher, W.,  181 Ogden, C.K.,  215, 226 Olshtain, E.,  98, 106 one-language-one-culture,  96–97 O’Niel, J.P.,  252 organon model of language,  226–227 orientation phase,  133 origins of language,  53, 175, 178, 181 Östman, J.-O.,  7, 140, 143 Otheguy, R.,  98 Otten, M.,  198 P Paluch, A.K.,  211 Paredes, A.,  205 participant observation,  92–93, 217 participant observation and field research,  105

participant roles,  164–165 Pasquale, M.,  149 Pawley, A.,  264 Peeters, B.,  69, 72 pejorative system,  157 Pelto, G.H.,  111 Pelto, P.J.,  111 Penn, J.M.,  267 Peterson, R.,  86 ‘phatic act’ (Austin’s),  227–228 phatic communion,  7, 140, 226–227, 228, 230 phatic communion, functions of,  231 Philipsen, G.,  10, 121, 124, 125 phonemes,  241 Phonemics,  257 phonetics vs. phonemics,  46–47 pidgins and creoles,  191 Pike, K.L.,  248 Pinker, S.,  268 Pinxten, R.,  268 PIXI project,  193 Plichta, B.,  148 Pokuta, G.,  192 politeness theory,  128 Porter, R.E.,  186 Porzig, W.,  58 Pottier, B.,  58, 59 Powdermaker, H.,  137, 213 Powell, H.A.,  221 Powell, J.W.,  31, 43, 44 pragmatic functions of aisatsu,  19 pragmatic grounding of language,  5, 182 pragmatic theory,  5, 175 pragmatics,  7, 10 pragmatics and semantics,  167–168 pragmatics of aisatsu,  25–26 Presser, S.,  207 Preston, D.R.,  10, 146, 147, 149, 151, 153 probes,  202 prototypical cognitive scenario,  63 psycholinguistics,  13 psychological anthropology,  50, 52 psychological reality,  34, 43, 47, 58, 241

Index 279

psychology of culture,  236 Pumphrey, M.E.C.,  162 Q Q-A sequences. See probes questionnaires,  105 Quinn, N.,  52 R Radcliffe-Brown, A.R.,  92, 132 radical elements,  239 radical functionalism and contextualism,  219 Radin, P.,  45 Rampton, B.,  197 Rattansi, A.,  90 Raven, P.H.,  251 Ray, V.F.,  60, 249 Redfield, R.,  223 Reesink, G.,  14 Rehbein, J.,  187 Reissman, C.K.,  206 relative presupposition,  204 reliability,  203 research on honorifics,  168–171 restricted contexts,  142 rhetorical speech practices,  74–75 Richards, A.I.,  213, 214 Ricoeur, P.,  93 Riley, P.,  192 Roberts, C.,  91, 100, 187 Roberts, J.H.,  249 Roberts, J.M.,  266 Roche, J.,  190 Rogoff, B.,  52 Rosaldo, M.Z.,  116, 127, 128 Rosaldo, R.,  116, 127–128 Rosch, E.H.,  250, 252 Roseberry, W.,  116 Rubinstein, R.A.,  137 Ruoff, A.,  10, 107, 108, 136 S Sahaptins, naming generic taxon,  251–252 Sahlins, M.,  83, 94, 249 Said, E.,  88, 89, 103 salsipuede events,  124 Samarin, W.,  108, 136 Samoan honorifics,  167 Samovar, L.,  186 Sanjek, R.,  92

Sapir, biography,  235–237 Sapir, E.,  1, 9–11, 30–32, 43–47, 50, 52, 91, 95, 121, 132, 137, 178, 185, 228, 234–245, 257–259, 261, 269 Sapir, on language, culture and personality,  237–244 Sapir-Whorf hypothesis,  8, 9, 10, 32–33, 46, 95, 96, 266 Sapir’s importance to pragmatics,  244–245 Sapir’s linguistic classification,  44 Sapir’s view on language,  242 Sarangi, S.,  1, 81, 91, 99, 100, 188, 197 Saunders, B.A.C.,  214, 253 Schank, R.C.,  65 Schegloff, E.,  129 Schieffelin, B.B.,  13, 23, 36, 56, 98, 170 Schieffelin, E.L.,  13 Schiffrin, D.,  153 Schmäh,  112–116 Schmeller, J.A.,  105, 132 Schmitter, P.,  175, 181 Schneider, D.M.,  86 Schober, M.F.,  204 Schultz, E.A.,  245, 268 Schuman, H.,  207 Scollon, R.,  98, 99, 187, 190 Scollon, S.,  98, 99, 187, 190 Scribner, S.,  12, 52 scripts for non-verbal communicative practices,  77 scripts for social models,  76–77 scripts for ways of thinking, feeling, and beliefs,  75–76 Searle, J.R.,  128 seasonal goaisatsu,  22 semantic bleaching,  168 semantic functions,  141 semantic molecules,  64–65, 72 semantic primes,  62–65, 69 semantic relationship,  167–168 semiotic approach,  86–87 Senft, B.,  221 Senft, G.,  1–2, 6, 7–10, 12, 105, 108, 131, 132, 210, 212, 217, 218, 220–222, 226, 227, 229, 230 Seuren, P.A.M.,  174, 181

Sherzer, J.,  36, 96, 97, 118 Shibatani, M.,  160 Shilluk royal language,  162 Shopen, T.,  108 Shultz, J.,  187, 190 Shweder, R.,  50 Sibata, T.,  148 Sider, G.,  88 Siebert, H.,  174 Siegel, J.,  191 Silverstein, M.,  8, 10, 11, 25, 39, 147, 151, 167, 170, 204, 234, 235, 238, 244, 268 Slobin, D.I.,  12, 13, 53, 107, 263 ‘Smile Code,’  71 social acquisition of aisatsu,  23–25 social categorization,  63–64, 73 social/cultural anthropology,  6, 50, 51, 81, 83, 91, 100 social functions of aisatsu,  19 social heritage and cultural anthropology,  214 social interdependence,  20 social models,  76 social relationships,  20–21 social use,  153 socialization,  23 sociolinguistic systems,  157, 164, 165, 167 sociolinguistics,  36 sociolinguistics dialectology,  10 Solomos, J.,  91 spatial relations,  106–107 SPEAKING framework,  122 speech act analysis,  192 speech act theory,  10, 26, 36, 127–128, 182 speech act theory, critique of,  127–128 speech-act verbs,  64 speech acts,  123 speech as part of context of situation,  226 speech behavior,  10, 35, 37, 38, 106, 108 speech communities,  124–125 speech events,  124 speech patterns,  121, 125 speech practices,  68, 74 speech realization patterns,  106 Spencer, J.,  94, 212, 215 Sperber, D.,  52, 55, 95

280 Index Sprenger, G.,  222 stage-of-life terms,  58, 59 Steinthal, H.,  43, 218, 237 Stigler, J.,  52 Stocking, G.W.,  42, 45, 223 Stoller, P.,  191 Strauss, A.,  118 Streeck, J.,  187 Street, B.V.,  87, 96 Stross, B.,  251 structuralist tradition of CA,  58–59 Sturtevant, W.C.,  248 Suchman, L.,  52, 205, 207 surface segmentability,  204 Suzuki, T.,  19 Swadesh, M.,  30, 43, 237, 257 Sweet, P.R.,  174, 181 Syder, F.H.,  264 symbolic capital,  206 Systemic Functional grammar,  144 Systemic Functional linguistics,  144 T Talmy, L.,  62 Tannen, D.,  98, 188 Taxonomy,  248 taxonomy, hierarchical,  251 taxonomy, linguists and anthropologists view,  251 Taylor, T.J.,  204, 215 theory of coevalness,  94 theory of meaning,  140 thinking and speaking theses,  175–176 Thomas, J.A.,  99 Thomas, J.M.C.,  137 Thompson, E.P.,  82, 83, 88, 189 Thornton, R.,  87, 89 Tobin, J.,  125 Todorov, T.,  191 Tornay, S.,  250

Trabant, J.,  5, 6, 181 Trager, G.L.,  186, 257, 258, 261, 262, 265–267 transcendental subjectivity,  176 transitive analyses,  61 Travis, C.E.,  69 trends in honorifics,  168–171 Triandis, H.C.,  12 Trier, J.,  58 Trudgill, P.,  11 Tsuda, Y.,  191 Turner, V.W.,  248 Tyler, S.A.,  51 Tylor, E.B.,  83, 85, 86 U unavoidable referentiality,  204 Urban, G.,  244 V Valentine, L.,  42 validity,  203 Van Brakel, J.,  253 van der Zouwen, J.,  204 ‘vantage theory,’  251 Varonis, E.,  191 verbal interaction,  204 verbal/non-verbal aisatsu,  19, 21 verbs and predicate-argument structure,  61 Verschueren, J.,  19, 123, 188, 197 Vilakazi, B.W.,  163 von Humboldt, W.,  1, 3, 6, 111, 173, 185, 237 W Wagner, J.,  191 Wassmann, J.,  12 Watson-Gegeo, K.A.,  232 Wayne, H.,  210, 211 Wegener, P.,  181, 218 Weiner, A.B.,  210, 211, 215, 220–223 Welsch, W.,  189

Wenger, J.R.,  165 Werner, D.,  136 Westwood, S.,  91 Whiten, A.,  53 Whorf, B.L.,  1, 9, 11, 32, 46, 52, 95, 144, 178, 185, 237, 242, 245, 256–270 Whorf, life and work,  258–259 Whorf, perspective on linguistics,  260–263 Whyte, W.F.,  109, 137, 186 Wierzbicka, A.,  26, 62, 63–65, 68–78, 98, 250 Wilkins, D.,  137 Wilks, Y.,  65 Williams, R.,  82 Willis, P.,  127 Wilson, D.,  55 Wirrer, J.,  3 Wolfson, N.,  71 Wong, J.,  71 Woolard, K.,  170 working class communities,  127 Wrench, J.,  91 Wu, D.,  125 Wuthnow, R.,  85 Wylie, L.,  137 X ‘X-name,’  252 Y Ye, Z.,  73, 77 Yim, Y-C.,  21 Yoon, K.-J.,  73, 76 Young, J.C.,  248 Young, M.W.,  6, 211, 212, 215, 216, 217, 220, 248 Z Zholkovskij, A.K.,  65 Zollinger, H.,  250 Zulu hlonipha vocabulary,  163

In the series Handbook of Pragmatics Highlights the following titles have been published thus far or are scheduled for publication: 10 Sbisà, Marina, Jan-Ola Östman and Jef Verschueren (eds.): Philosophical Perspectives. ca. 250 pp. Forthcoming 9 Östman, Jan-Ola and Jef Verschueren (eds.): Pragmatics in Practice. ca. 250 pp. Forthcoming 8 Ledin, Per, Jan-Ola Östman and Jef Verschueren (eds.): Discursive Pragmatics. ca. 250 pp. Forthcoming 7 Jaspers, Jürgen, Jan-Ola Östman and Jef Verschueren (eds.): Society and Language Use. ca. 250 pp. Forthcoming 6 Fried, Mirjam, Jan-Ola Östman and Jef Verschueren (eds.): Variation and Change. Pragmatic perspectives. ca. 250 pp. Forthcoming 5 Brisard, Frank, Jan-Ola Östman and Jef Verschueren (eds.): Grammar, Meaning and Pragmatics. xv, 305 pp. + index. Expected August 2009 4 D’Hondt, Sigurd, Jan-Ola Östman and Jef Verschueren (eds.): The Pragmatics of Interaction. xiii, 254 pp. + index. Expected August 2009 3 Sandra, Dominiek, Jan-Ola Östman and Jef Verschueren (eds.): Cognition and Pragmatics. ca. 250 pp. Forthcoming 2 Senft, Gunter, Jan-Ola Östman and Jef Verschueren (eds.): Culture and Language Use. 2009. xiii, 280 pp. 1 Verschueren, Jef and Jan-Ola Östman (eds.): Key Notions for Pragmatics. xiii, 253 pp. Expected June 2009