Situation-Bound Utterances in L1 and L2 [Reprint 2012 ed.] 9783110894035, 9783110173581

This book focuses on a particular type of formulaic expressions called Situation-Bound Utterances (SBUs). Since the mean

185 81 5MB

English Pages 236 [240] Year 2002

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Acknowledgments
Introduction
Chapter 1 : Lexical and conceptual level
Chapter 2: A dynamic model of meaning
Chapter 3: Context and salience
Chapter 4: The origin of situation-bound utterances
Chapter 5 : Distinguishing features of SBUs
Chapter 6: Classification and interpretation of SBUs
Chapter 7 : Creativity and formulaicity
Chapter 8 : Conceptual socialization
Chapter 9: Easy to learn, hard to understand
Conclusion
Data Sources
Notes
References
Subject Index
Recommend Papers

Situation-Bound Utterances in L1 and L2 [Reprint 2012 ed.]
 9783110894035, 9783110173581

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Situation-Bound Utterances in LI and L2

W G DE

Studies on Language Acquisition 19

Editor Peter Jordens

Mouton de Gruyter Berlin · New York

Situation-Bound Utterances in LI and L2

by

Istvan Kecskes

Mouton de Gruyter Berlin · New York 2003

M o u t o n de G r u y t e r (formerly M o u t o n , T h e H a g u e ) is a Division of Walter de G r u y t e r G m b H δί Co. KG, Berlin.

©

Printed o n acid-free p a p e r which falls within the guidelines of the ANSI t o e n s u r e p e r m a n e n c e and durability.

Library

of Congress

Cataloging-in-Publication-Data

Kecskâes Istvâan. S i t u a t i o n - b o u n d u t t e r a n c e s in L I a n d L2 / by Istvâan Kecskâes. p. c m . — (Studies o n language acquisition ; 19) Includes bibliographical references a n d index. ISBN 3 11 017358 1 (alk. paper) 1. C o n v e r s a t i o n analysis. 2. C o n t e x t . (Linguistics) 3. Second language acquisition. 4. Bilingualism. I. Title. II. Series. P95 45. Κ4 2 0 0 2 401'.41-dc21 2002013889

ISBN 3 11 017358 1 Bibliographic

information

published

by Die Deutsche

Bibliothek

Die D e u t s c h e Bibliothek lists this publication in t h e Deutsche N a t i o n a l b i b l i o g r a f i e ; detailed bibliographic d a t a is available in the Internet at < h t t p : / / d n b . d d b . d e > .

© C o p y r i g h t 2002 by Walter de G r u y t e r G m b H δί Co. KG, D - 1 0 7 8 5 Berlin. All rights reserved, including those of t r a n s l a t i o n i n t o foreign languages. N o p a r t of this b o o k m a y be r e p r o d u c e d in any f o r m or by any means, electronic or mechanical, including p h o t o c o p y , recording, or any i n f o r m a t i o n storage a n d retrieval system, w i t h o u t permission in w r i t i n g f r o m t h e publisher. C o v e r design: Sigurd W e n d l a n d , Berlin. P r i n t e d in G e r m a n y .

To my wife, Tünde

Acknowledgments

In writing this book I was fortunate enough to receive encouragements and support from many different people. My wife, Dr. Tiinde Papp was there for me whenever I wanted to discuss a problem or an idea which needed to be elaborated on. Her patience, knowledge, insightful and inspiring remarks helped me a lot in developing the manuscript. It is good to have another linguist in the family. While I was writing the book I was teaching a doctoral seminar on pragmatics at the State University of New York at Albany. My international group of students in the class, Kristen Wilcox, Noritaka Mochida, Sam Harb, Jieun You, Veronique Martin and Saihua Xia were "guinea pigs", resources, advisors and critiques to the project. Their heated debates, interesting suggestions and perceptive remarks on several issues contributed to a great extent to the shaping of my understanding of situation-bound utterances cross-culturally. Some of the ideas I am discussing in this book were shaped in discussions with my colleagues at the Universität Autonoma de Barcelona, Spain while I was a visiting professor there in the spring of2001. I am especially grateful to Professor Maria Josep Sole-Sabater, Dr. Hortense Curell and Dr. Melissa Moyer whose suggestions and insightful remarks helped me clarify several issues concerning the complexity of meaning. I need to say special thanks to Professor Joshua Bear, Middle Eastern Technical University, Ankara, Turkey, who provided me with several interesting SBUs used in Turkish. My visit to Turkey and the wonderful hospitality of the Turkish Higher Education Council headed by Dr. Kemal Guruz and Dr. Barbaros Guncer enabled me to gain insight into Turkish life and culture. I have received continued support for the project from Dr. Anke Beck, Editor-in-Chief of Mouton de Gruyter and Dr. Ursula Kleinhenz, my editor, who strongly believed that I would be able to finish the manuscript in a timely manner even when I accepted a visiting professorship at the Universität Autonoma de Barcelona, Spain for eight months. I am very grateful to Professor Peter Jordens, Editor of the Series on Language Acquisition for his genuine interest in the project and very helpful suggestions. This book could not have been published without Andrew MacFarland, a graduate student of mine who did an excellent job editing the

viii

Acknowledgments

manuscript. Andrew's thoughtful and expert editing, good-humored and helpful suggestions made a considerable contribution to the successful completion of this project. Some portions of this book first appeared, in a different format, as articles in journals or chapters in edited volumes. I appreciate the publishers' permission to include revised versions or part of the following publications: Kecskes, I. 2001."The Graded Salience Hypothesis in Second Language Acquisition". In: Martin Putz, Susanne Niemeier & Rene Dirven (Eds.) Applied Cognitive Linguistics. Vol. I. Berlin/New York: Mouton de Gruyter: 249-271; Kecskes, I. 2000. "ACognitive-Pragmatic Approach to Situation-Bound Utterances". Journal of Pragmatics. 2000. Vol. 32. No. 6: 605-625; Kecskes, I. 2000. "Conceptual Fluency and the Use of Situation-Bound Utterances in L2". Links & Letters. Vol. 7: 143-158. I also thank R.S. Zaharna for permission to reprint the table which previously appeared in the following publication on the web: Zaharna, R. S. 1995. Understanding cultural preferences of Arab communication patterns. Public Relations Review. Internet Site.

Istvan Kecskes

Contents Acknowledgments

vii

Introduction

1

Chapter 1 : Lexical and conceptual level

15

Chapter 2: A dynamic model of meaning

31

Chapter 3: Context and salience

55

Chapter 4 : The origin of situation-bound utterances

79

Chapter 5:

99

Distinguishing features of SBUs

Chapter 6: Classification and interpretation of SBUs

115

Chapter 7: Creativity and formulaicity

135

Chapter 8 : Conceptual socialization

155

Chapter 9: Easy to learn, hard to understand

177

Conclusion

197

Data Sources

201

Notes

203

References

205

Subject Index

225

Introduction

This book grew out of a keen interest in an interdisciplinary approach to language development, and a strong belief that monolingual approaches and theories are not the best means to explain second language use and bi-and multilingual development . The emergence of another language directs attention to the bidirectional influence between the two languages, highlights the decisive role of the interplay of language and culture in shaping meaning, and shifts the explanatory movement from the linguistic level to the conceptual level. Why did I choose situation-bound utterances as the object of inquiry? The answer to this question is rather simple.

1. The first encounter I took an interest in formulaic expressions when I happened tobe one of the participants in a short conversation during my first visit to the U.S. in 1989. The conversation between a student (S) and me (K) at a university campus went as follows: ( 1 ) K: - Excuse me, could you tell me how to get to the library? S: - Go straight and take a left at the park. K: - Thank you. S: - You bet. I was surprised to get a response like " you bet" to my "thank you" because I had never heard that expression before. I concluded that Americans use "you bet" as a possible reaction to "thank you", and several further encounters with the expression justified my belief. However, the real surprise came when I wanted to use the phrase in the following situation: Soon after my arrival in the U.S., I was invited to give a presentation at a U.S. university. When the talk was over the chair of the session thanked me for the presentation, and I happily responded with a "you bet". Well, I think that utterance brought me quite a bit of success because everybody laughed at my pragmatic failure: the inappropriate use of a conversational routine. However, my

2

Introduction

experiencing of American "pragmatic acts" (Mey 2001) was not over yet. One day I went to a supermarket. When I was checking out the cashier welcomed me with a big smile and a happy "How are you doing?" I was surprised that she was interested in my well-being, but I wanted to be polite and started to tell her that I was OK, but since I just arrived in the U.S. I had to learn a lot of new things. Her smile changed into a strange look. That moment I was the one who got surprised. Apparently, she was not interested in my story at all. In that situation "How are you doing?" was only a kind of greeting, a sign that she noticed me. I am afraid I did not process the utterance that way. These events (and several similar ones) directed my attention to the formulaic expressions that I call "situation-bound utterances" (Kecskes 1997; 1998). I realized that these expressions are much more than just another group of formulas. The relationship of linguistic units and non-linguistics factors is always mutually complimentary in any kind of verbal interaction but the linkage between the two sides seems to be especially strong in prefabricated expressions. The investigation of situation-bound utterances (SBUs) can lead researchers to the heart and soul of the target language. These expressions are "windows" to the culture, thought patterns, social values, communication structures and institutions of the society in which the source language is used.

2. Formulaic language Hymes ( 1962) pointed out that an immense portion of verbal behavior consists of linguistic routines. Bolinger suggested that "speakers do at least as much remembering as they do putting together" (Bolinger 1976:2). Fillmore also found that "an enormously large amount of natural language is formulaic, automatic and rehearsed, rather than propositional, creative or freely generated " (Fillmore 1976:24). Altenberg (1998) went even further: he claimed that almost 80% of our language production can be considered formulaic. This might not necessarily be the case but one thing is for sure: Nowadays speakers in conventional speech situations tend to do more remembering than putting together. Our everyday conversation is usually restricted to short routinized interchanges where we do not always mean what we say. So a typical conversation which starts with the expression "How are you doing?" may go as follows rather than as it went in the supermarket when I was one of the participants:

Introduction

3

(2) Bob: - Hi Jim, how are you doing? Jim: - Fine, thank you. Just like the cashier in the conversation with me, Bob also does not really care how Jim is doing, and Jim is far from being "fine" but he has to say so because he is expected to. In this type of conventional interactions the gap between "what is said" and "what is communicated" seems to be getting wider.1 This phenomenon characterizes not only English but most modern languages. Here is a similar example from Hungarian:

(3) Pista: - Szevasz Zoli, hogy vagy? Zoli: - Jól, és te? Pista: - Nines okom panaszra.

'Hi, Zoli, how are you?' 'Fine, and you?' Ί have no reason to complain.'

We are usually in a hurry and have little time for expressing our thoughts and/or discussing our ideas with others so we rely mainly on conversational routines which help us pretend that we are engaged in "real" conversation. Some conversational routines are very often repeated, others are used less frequently. Coulmas (1981a) argued that frequency of occurrence has a crucial impact on the meaningfulness of linguistic expressions. The more frequent they are, the more meaningless they may become in terms of referential semantics. The compositional meaning of utterances often becomes of secondary importance and the functional aspect begins to dominate. Formulaic expressions have received a fair amount of attention within both theoretical and applied linguistics as well as in pragmatics and second language acquisition (c.f. Fillmore C. 1976; 1979; Bolinger 1976; Fillmore L. 1976; Coulmas 1979, 1981a, 1981b; Cowie 1988, 1998; Yorio 1980; Vihman 1982;Fonagy 1982; Bahnsetal. 1986; Bohn 1986; Kiefer 1985,1995; Raupach 1984; Moon 1992; Verstraten 1992; Nattinger and DeCarrico 1992; Sinclair 1991;Weinert 1995; Aijmer 1996; Miller and Weinert 1998; Giora 2001,2002; Wray 1998,2002).

4

Introduction

These studies discuss the same phenomenon from different perspectives, but most of them have something in common. They emphasize the important role of fixed expressions in speech production and stress the importance of their functional aspect, i.e., the fact that certain language sequences have conventionalized meanings which are used in predictable situations. This functional aspect, however, is different in nature in each fixed expression, which justifies the hypothesis of a functional continuum (Kecskes 1997) which contains grammatical units (for instance: "be going to") on the left, semantic units (c.f. "as a matter of fact"; "suffice it to say") in the middle and pragmatic expressions (such as situation-bound utterances: "welcome aboard"; "help yourself') on the right. The more we move to the right on the functional continuum the wider the gap seems to become between 'what is said' and 'what is communicated'. Language development often results in a change of function, i.e., a right to left or left to right movement of a linguistic unit on the continuum. Lexical items such as 'going to'can become grammaticalized, or lexical phrases may lose their compositionality and develop an 'institutionalized' function, suchas Til talk to you later'. Therefore pragmatic approaches to the study of formulaic language can provide a very important basis for their identification and categorization (cf. Coulmas 1981, Bahns et al 1986). The need for a developmental angle is stressed by Bahns et al (1986), Coulmas (1985), and Kasper (1994). The pragmatic dimension seems to be essential in explaining the development and use of formulae, and yet, as pointed out by Bahns et al. (1986), formulaic language has not been investigated extensively in relation to the development of pragmatic competence as a central learning task.

3. Situation-bound utterances (SBUs) This book focuses on a particular type of formulaic expressions called situation-bound utterances (Kecskes 1997; 1998, 2000a). SBUs are highly conventionalized, prefabricated pragmatic units whose occurrences are tied to standardized communicative situations (Coulmas 1981a; Kiefer 1985,1995; Kecskes 1997). I f , according to their obligatoriness and predictability in social situations, formulaic expressions are placed on a continuum where obligatoriness increases to the right,

Introduction

5

Situation-Bound Utterances will take the rightmost place because their use is highly predetermined by the situation (Kecskes 2000a). The pragmatic fonctions are not always encoded in these linguistic units, therefore SBUs often receive their "charge" from the situation in which they are used. It is generally this situational charge that distinguishes SBUs from their freely generated counterparts. Compare the following situations:

(4) Jill: - Bill, may I talk to you for a minute? Bill: - Sorry Jill, I am very busy now, but I'll talk to you later.

(5) Wife: - Jack, do you still love me? Husband: - Of course, I do. I'll talk to you later. (Conversation from the film Mary Higgins Clark's Lucky Day presented on PAX on April 30,2002) In (4) the expression "77/ talk to you later" appears to be freely generated as opposed to (5) where it is an SBU rather than a freely generated utterance. Several labels have been used to refer to these expressions in the relevent literature: "interaction rituals" (Goffman 1967) "routine formulae" (Coulmas 1979) "énoncés liés" (Fónagy 1982), "situational utterances" (Kiefer 1985; Kiefer 1995), "bound utterances" (Kiefer 1997), "institutionalized expressions " (Nattinger and DeCarrico 1992), and "situation-bound utterances" (Kecskes 1997, 1998, 2000a). The variety of terms can be explained not only by the difficulty of defining this particular type of pragmatic units but also by the fact that these expressions are discussed in different subfields of theoretical and applied linguistics, and sometimes authors seem to care relatively little about research on SBUs made outside their own respective field. The term "situation-bound utterances" is preferred by the author to any other term because this expression refers to the main characteristic feature of utterances being investigated: their boundedness to a particular situation. "Routine formulae" is too broad a category, "situational utterances" presupposes that there are utterances other than situational, and

6

Introduction

"institutionalized expressions" seems to be too specific a term. The French term "énoncés liés" used by Fónagy (1982) expresses best what these expressions are all about. Kiefer (1995) refers to Fónagy when explaining "situational utterances" but the term "situation-bound utterances" appears to be a closer equivalent to "énoncés liés" than "situational utterances". At this point it is important to explain the relation of SBUs to "conversational routines" (cf. Coulmas 1981; Aij mer 1996) on the one hand and to idioms on the other. Conversational routines have an inclusive relation to SBUs: all SBUs are conversational routines, but this is not so conversely because not all expressions labeled as conversational routines are SBUs. Aijmer argued that conversational routines are expressions which, as a result of recurrence, have become specialized or 'entrenched' for a discourse function that predominates over or replaces the literal referential meaning (Aijmer 1996:11). It is quite difficult to draw the dividing line between conversational routines and situation-bound utterances but there are some points where they differ. Conversational routines (CR) are function-bound rather than situation-bound. They can express one and the same particular function in any situation while SBUs quite often receive their charge from the situation itself. For instance, 'let me tell you something' or ' to tell you the truth' are CRs rather than SBUs. They can be uttered in any situation with the same function where they sound appropriate. However, expressions such as 'how do you do?' upon acquaintance or 'welcome aboard' as a greeting to a new employee make sense only in particular well-frameable situations. The tie of SBUs to a given situation charges their particular meaning which may become so dominant that the functional-situational meaning may take over as the most salient meaning of the expression. Conversational routines tend to have discourse functions which is a broader term than a situation-bound function. Discourse functions are not tied to particular situations, and can be expressed by conversational routines including not only SBUs but also expressions of turntaking, internal and external modifiers, discourse markers, connectors, and others. SBUs are usually idiomatized and less often used in their original literal meaning than in their situation-bound pragmatic function. Nattinger and DeCarrico (1989: 128) referred to them as "idioms with a pragmatic point". Consequently, SBUs can be considered as pragmatic idioms that take a unique place among idioms because they have several features that distinguish them from other idiomatic expressions which will be discussed in chapter five. SBUs are prefabricated units

Introduction

7

whose occurrences are very strongly tied to conventional and frequently repeated situations. We cannot help thinking of a particular situation even if we hear the following expressions without their routine context: 'Welcome aboard'; 'Please help yourself, etc. As mentioned earlier, the functional-situational meaning may become the most salient meaning of the expression. The meaning of SBUs without context will be paid special attention to when the graded salience hypo-thesis is discussed in chapter 3.

4. Theoretical framework The aim of this book is not to present a unified theory of SBUs, or analyze their occurrences in particular corpora. Rather, I would like to discuss some important issues that concern this unique group of pragmatic idioms. Because of the complexity of the inquiry I need to rely on findings in many disciplines such as psycholinguistics, theoretical linguistics, sociolinguistics, cognitive linguistics, communication studies, pragmatics, second language acquisition, bilingualism, etc. This interdisciplinary approach lets us consider the totality of the phenomenon and describe the development and use of SBUs in L l a n d L 2 . There are, however, some leading principles that serve to keep the cohesion and cohesiveness of the book.

4.1. The cognitive-pragmatic

approach

In present day's linguistic theory there are two main lines which define language differently. Cognitive linguistics claims that there is no separate language module. Language is only one way among several others that humans use to interact with the world. Langacker (1998:1) argued that "[Cognitive linguistics] contrasts with formalist approaches by viewing language as an integral facet of cognition (not a separate "module" or "mental faculty"). Insofar as possible, linguistic structure is analyzed in terms of more basic systems and abilities (e.g. perception, attention, categorization) from which it cannot be dissociated." Generative linguistics represents the exact opposite view, claiming that there is a unique mental faculty which has to be analyzed on its own terms. Recent research points to the fact that the truth is somewhere in the middle and neither position can be maintained in its extreme form (Newmeyer 1999).

8

Introduction

In this book I will discuss SBUs from a cognitive-pragmatic perspective which originates in cognitive linguistics but recognizes the uniqueness and certain independence of the language faculty. Since the meaning of SBUs is shaped by the interplay of linguistic and extralinguistic factors they can be best accounted for in a theoretical framework which represents a knowledge-for-use conception. Nuyts' (1993; 2001) cognitive-pragmatic perspective on language serves as a basis in shaping the paradigm in which SBUs are discussed in this book. The adjective 'cognitive' relates to the observation that language is a dimension of human mental activity. The other adjective 'pragmatic' relates to the observation that, in the context of the human behavioral repertoire, language has a specific role to play. Its primary role is to allow communication with other members of the species (c.f. Grice 1969; Nuyts 1993; Searle 1972). Kay (1983) claimed that people conceive of their knowledge of language as a part of their knowledge of the world. According to cognitive semantics, there is a conceptual system which represents the world as understood by us, speakers. Language has as one of its integral parts the world as understood by us through experience, and the world as understood is our only guide to the world as it is (Kövecses, 1993). The cognitive and the pragmatic dimensions of language, which are central to strong research traditions in the language sciences, are not just two separate issues; instead, they "are two faces of one phenomenon, which must be mutually interrelated and interdependent" (Nuyts 2001 : 3). Nuyts claimed that the cognitive- pragmatic perspective assumes that an adequate account of language in general, or of any linguistic phenomenon in particular, has to do full justice to both dimensions simultaneously, in an integrative way (Nuyts 2001 : xv). Although they are two faces of one phenomenon, the cognitive and the functional dimension have a different status in the investigation of language. The cognitive dimension is hidden in the black box of the human mind, and can only be accessed through observing what the system perceives and how it behaves. The functional dimension of language, however, is part of the observable behavior. Hence, Nuyts argued that "analyzing the functional dimension of language is logically antecedent to analyzing the cognitive infrastructure for language" (Nuyts 2001: 3-4). The two key properties of the cognitive-pragmatic approach as applied in this book are that 1) language research is bound to be concerned with the linguistic and conceptual dimensions at the same time, and that 2) lan-

Introduction

9

guage is a dynamic entity. Our targets are SBUs which are units of language that demonstrate a particular symbiosis of linguistic and conceptual features through the dynamism of their use.

4.2. The multilingual

perspective

This book focuses on SBUs in several languages taking a crosslinguistic approach. Examples throughout the book are from several different languages including but not restricted to English, Hungarian, Russian, French, German, Turkish, Chinese and Japanese. The true nature of SBUs can be investigated if attention is paid not only to native speakers of a language but also to non-native speakers. The analysis of the problems of non-native speakers acquiring and using SBUs in the target language makes it possible for us to have an insight into the interplay of linguistic and socio-cultural factors in the use of these pragmatic units because in second language acquisition the study of the language system is frequently cut off from its extralinguistic background. This is especially so in a foreign language environment. Classroom language instruction is not particularly efficient for language learners but it can be advantageous for researchers who may have the chance to investigate what happens when the language faculty is separated from its genuine source of experience. Adult L2 learners are expected to learn new labels and modify existing knowledge and concepts in most cases without continuous and direct exposure to the target language environment and culture in which the experience expressed by the linguistic signs of that language originated and are rooted. The investigation of second language development which significantly differs from LI development may shed light on phenomena and processes which remain unexplored in the LI theoretical paradigm. Therefore a multilingual approach is applied to most of the issues discussed in the following chapters. We argue that language-oriented research has been governed by a monolingual view for several decades and still is (Kecskes and Papp 2000a). Even the most influential theories of second language acquisition such as the Universal Grammar approach (e.g. Flynn 1996; White 1996) and interlanguage (e.g. Selinker 1972, 1992; Laksmanan and Selinker 2001) adopt the monolingual view which treats both languages as if they were the first language. Meara (1983, p. iv) is opposed to that view arguing that

10

Introduction

"there is no reason why a person who speaks both English and Spanish should behave in the same way as a monolingual speaker of either language." Grosjean (1985:467) suggested that the bilingual is a "specific speaker-hearer" and not two monolinguals in one body. Bilinguals and/or multilinguals have a unique competence that Cook (1991a) called "multicompetence" and Kecskes and Papp (2000a; forthcoming) referred to as the "dual language system" which contains a state of LI knowledge that is, in some respect, different from that of monolinguals. Cook (1992; 1993) claimed that people who know more than one language have a distinctive state of mind, "multi-competence", that is not equivalent to two monolingual states. He argued that multicompetence is not a final state of mind as the native monolingual's competence is but covers all stages of second language acquisition. A crucial question is when does the dual language system (DLS) begin to develop. It depends on several factors including proficiency, exposure to the language and culture, motivation, age and others. However, the level of proficiency in the target language is a decisive factor. A person who has taken first year university Spanish will hardly develop a separate system for Spanish. De Bot ( 1992) argued that the LI is usually flexible enough to add the emerging foreign language as an additional register to those already in existence. (Kecskes and Papp 2000a) claimed that intensive exposure to and regular use of the foreign or/and second language may lead to the emergence of a Common Underlying Conceptual Base (CUCB) which is responsible for the operation of the two or more languages. However, the development of the CUCB is dependent on proficiency in the L2 which has to reach a certain hypothetical threshold. Before that proficiency level is reached L2 learning is likely to be no more than a kind of educational enhancement that may only slightly affect cognitive development and not necessarily result in the emergence of a CUCB. If the exposure to the L2 is not intensive enough the L1 conceptual base may remain practically unaffected because students usually learn only new labels for existing concepts. The emerging L2 is incorporated into the classification system already available in the first language, and its operation relies on the previously developed conceptual system. In order for a dual language system to develop the conceptual structure needs to change from an LI-dominated conceptual base into a CUCB that is responsible for the operation of both languages. The CUCB was conceptualized (Kecskes and Papp 2000) as the basis and originator of all bilingual or multilingual linguistic actions, a'con-

Introduction

11

tainer' which includes everything but the two or more language systems (rules plus lexicon). It is in the CUCB where the socio-cultural heritage and previous knowledge of the learner are confronted with the new information entering into the CUCB through both language channels, and real-world knowledge mixes with academic knowledge developing into something that is frequently referred to as "socio-cultural background knowledge" (c.f. Adamson 1993; Kecskes 1994; Kecskes and Papp 2000a). The development of the CUCB goes together with what I call "conceptual socialization " that refers to the transformation of the conceptual system which undergoes characteristic changes to fit the functional needs of the new language and culture. During the process of conceptual socialization the LI-dominated conceptual base is being gradually restructured making space for and engaging with the new knowledge and information coming through the second language channel. This leads to the gradual development of a conscious awareness of how another culture is different from one's own culture, the ability to reflect upon this difference in language production, and the development of an identity that is the reflection of the dual culture. Thoughts originate in the CUCB, and then are mapped onto linguistic signs to get to the surface through either of the language channels. Due to the CUCB, information processing skills and educational attainment may be developed through two languages as well as through one language. Cognitive functioning and school achievement may be fed either through one language or through two languages if both are well developed. Monolinguals have an LI-governed conceptual base and one language channel while bilinguals have a CUCB which is a common knowledge base for both languages, and two language channels which are usually operational in both directions. The existence of the CUCB as well as the constant interaction of the two language channels make bi- and multilingual development and language use unique, and this is why neither of the participating languages can be compared to a monolingual system. Consequently, the acquisition and use of SBUs is entirely different in a monolingual speaker and in a bi- or multilingual speaker. This approach is in line with Grosjean's theory about "language mode" which is understood as the state of activation of the bilingual's languages and language processing mechanisms at a given point of time (Grosjean 1998; 2001). According to Grosjean bilingual speakers or listeners switch from monolingual mode to bilingual mode on the language mode continuum which makes code-switching a natural phenomenon in bilingual language production. Grosjean claimed that "given the high level of

12

Introduction

activation of both languages in the bilingual mode, not only can codeswitches and borrowings be produced but the base language can also be changed frequently, that is the slightly less activated language becomes the base language and vice versa."(Grosjean 2001: 6)." Grosjean's theory gives a compelling explanation of the operation of the two language channels. However, it does not explain what makes it possible for the bilingual speaker to move on the language mode continuum changing from monolingual mode to bilingual mode and vice versa. I argue that movement on the language mode continuum is facilitated by the CUCB which keeps both language channels available any time, and operates one or both language channels depending on the immediate needs of the bilingual speaker/hearer. The existence of the CUCB is supported by the language behavior of bilinguals in the monolingual language mode. Grosjean (2001) argued that "...there is considerable evidence in the literature that bilinguals make dynamic inferences (ephemeral deviations due to the influence of the other deactivated language) even in the most monolingual of situations (Grosjean 2001:7)". Monolinguals and multilinguals have the same type of cognitive mechanisms that are responsible for SBUs, but the content of these mechanisms is significantly different as will be discussed in the following chapters.

5. Objectives What can this book offer? How does it differ from similar titles? First of all, it does not focus on the whole range of conversational routines as Karin Aijmer did in her excellent book (Aijmer, Κ. 1998. Conversational routines in English. London: Longman). Nor does this book aim to review all research on formulaic language and acquisition issues as Alison Wray did in her recent intriguing book (2002) on "Formulaic language and the lexicon" (Cambridge: Cambridge University Press). Instead, the focus is on one particular type of prefabricated units with the purpose of discussing the characteristic features of SBUs in a cognitive-pragmatic framework. I also give a comprehensive view of these formulaic expressions from a cross-linguistic perspective, and relate new developments in cognitive research such as the multilingual approach (Cook 1992, Kecskes and Papp 2000; Kecskes 2002) and the graded salience hypothesis (Giora 1997; 2001) to the issue of use of SBUs in various languages. The focus is on meaning rather than

Introduction

13

structure, which gives me the opportunity to take a stand in the continuing debate on meaning and the relationship of context and lexical units. The comparison of a monolingual and multilingual approach and the application of the graded salience hypothesis to SBUs within a cognitive-pragmatic theoretical framework reveals that issues such as the role of context in shaping situational meaning, and the existence of common or similar cognitive mechanisms and know-ledge structures responsible for cognitive functions and speech behavior in different languages need revision. I put forward a Dynamic Model of Meaning (DMM) which aims at supporting this re-vision, and at-tempts to bridge the gap between the definition-based approach (c.f. Wierzbicka 1996) and the prototype theory (Rosch and Melvis 1975) by explaining how the lexical item unites the lexical level and the conceptual level. On the basis of recent research it is argued that context affects comprehension only after highly salient information has been accessed (c.f. Giora 1997, 2002). Search for the appropriate meaning stops if the information accessed initially is compatible with the context, and it continues, if it is not. This approach puts the issue of context-sensitiveness of SBUs into an entirely different perspective and leads us to the revision of the role of context in the development and use of SBUs. Why exactly did these utterances start to be used to express those pragmatic functions and not others? SBUs demonstrate better than any other linguistic unit that although there is a strong cognitive-linguistic interdependency it would not be wise to equate the conceptual level with the lexical level. The development of certain SBUs can be accounted for through cognitive mechanisms, and vice versa: learning an SBU for a culturally important category can linguistically reinforce the learning of the category itself. In sum, this book pays special attention to the following issues: -

the interplay of lexical units and context in shaping meaning, the role of salience in language processing, origin and extent of context-sensitiveness of SBUs, development of the situational functions of SBUs, role of SBUs in language acquisition, conceptual socialization as the result of the development of the dual language system, and - the use of SBUs in a second language.

14

Introduction

6. Cautionary note I would like to emphasize that this book is about situation-bound utterances that have always been and will always be present in human interaction. Each speech community develops a kind of speech ritual or speech etiquette which keeps changing depending on the ever-alternating needs of that particular speech community. Among these pragmatic idioms we can find expressions which corrupt our speech and thinking, and which are the result of a temporary linguistic decline that is characteristic of most language communities as a consequence of our industrial-technological, advertisement and consumption-centered society. Utterances such as 'that sucks', or "get lost" to express any kind of dislike, "buy one, get one free" in the paper, or 'your call is important to us' frequently heard on automated answering machines also fit into the scope of this investigation. However, these expressions come and go, and will never become an essential part of our language. (At least that is what is hoped.) The dynamic process of human language development is full of ups and downs. However, as a mirror and partly a wheel of society each language is capable of expressing the everchanging communicative needs of one (or sometimes even more than one) speech community.

Chapter 1 Lexical and conceptual level

SBUs are units of language which demonstrate a particular symbiosis of linguistic and conceptual features through the dynamism of their use. They amalgamate lexical, conceptual and contextual information, and demonstrate better than any other linguistic unit that there is a strong cognitive-linguistic. If we want to understand their function and use we will need to answer two questions: 1) How does the lexical information relate to the encyclopedic2 information in these pragmatic units?; Is the lexical level separate from the conceptual level?, 2) What model of meaning can explain the dynamic contextual behavior of SBUs? In this chapter we seek an answer to the first question.

1. Linguistic and conceptual knowledge 1.1. The cognitive approach In recent years contemporary linguistic theory has been paying more attention to meaning than ever before. This is especially true for cognitive linguistics whose primary concern is meaning. Mainstream cognitive linguists (c.f. Geeraerts 1997; Langacker 1987; Lakoff 1987) deny the existence of any boundary between real-world (encyclopedic, extralinguistic) knowledge and linguistic knowledge. Language is seen as a repository of world knowledge and a structured collection of meaningful categories which help us deal with new experiences and store information about old ones (Geeraerts 1997). Our linguistic knowledge is rooted in our experience with the world. This is what language in use expresses. So if language is a system for categorization of the world there does not seem to be any need to postulate a systemic and structural level of meaning that differs from the level where world knowledge is associated with linguistic forms.

16

Chapter 1

Langacker claimed that the distinction between semantics (linguistic knowledge) and pragmatics (extralinguistic knowledge) is "largely artifactual" and the only viable conception of linguistic semantics 1) avoids "false dichotomies" such as semantics versus pragmatics, and 2) takes an encyclopedic perspective (Langacker 1987: 154-166). Consequently, grammatical theory should incorporate information about cultural and sociolinguistic conventions that are necessary to interpret grammatical patterns. Langacker (1990) argued that it is pointless to analyze grammatical units without reference to their semantic value because grammatical structures do not constitute an autonomous formal system. Instead they are claimed to be inherently symbolic, "providing for the structuring and conventional symbolization of conceptual content (Langacker 1990: 1)." Lakoff insisted that one of the objectives of the cognitive approach is to "show how aspects of form can follow from aspects of meaning (Lakoff 1987: 491)." In the cognitive approach semantic representations which constitute the semantic pole of a linguistic sign are equated with "conventionalized conceptualizations" (Langacker 1988:94). This view integrates encyclopedic knowledge into the lexicon.

1.2. The generative view There have also been changes in the generative paradigm because a number of linguists have come around to the view that the lexicon is the central component of a person's internal grammar, with the syntax as subsidiary (Hudson, 1984). Chomsky claimed that parametric variation takes place in the lexicon (Chomsky, 1986,1989,1995). Chomsky's Minimalist Program is built on the claim that the process of forming structures has to start from the lexicon because lexical elements, will, on the whole, determine the content of any legitimate expression in a language. We start by selecting a set of lexical items from which the structural description is to be built (Chomsky 1995). The lexical learning hypothesis proposes that lexical items, along with their properties, trigger the reconstruction of a child's grammar; that is, grammars are lexically-driven (e.g. Clahsen, 1992; Klein, 1995; Weisenborn, Goodluck and Roeper, 1992). In the Minimalist Theory there is no farther level of representation between the Phonetic Form (PF) and Logical Form(LF). The LF, although explicitly related to the conceptual representation, has always been understood in generative

Lexical and conceptual level

17

grammar as a level of syntactic representation. This has not changed within the Minimalist program. Bierwisch, however, argued that there is a clear distinction between a word's purely linguistic meaning and the interpretation that a word may have relative to conceptual knowledge (c.f. Bierwisch 1981; Bierwisch and Schreuder 1992; Taylor 2000). This two-level model distinguishes between a linguistic-semantic level of meaning and an essentially non-linguistic, conceptual level of interpretation. So on the one hand, generativists consider grammars as lexically-driven, and the lexicon as the central component of a person's internal grammar which is linguistic-semantic in nature. Linguistic knowledge constitutes an encapsulated mental module which is independent of other skills and knowledge. On the other hand cognitive linguists argue that semantic structure is a conventionalized conceptual structure, and grammar is the conventional symbolization of semantic structure. Lexicon and grammar form a continuum of symbolic structures which contains no sharp dichotomies based on generality, regularity, or analyzability (Langacker 1990:102). Cognitive linguists believe that " knowledge of all kinds is integrated in the mind to such an extent that it does not make any sense to partition it into two distinct realms" (Goddard 1998:15).

1.3. Dynamism of meaning

There seems to be a problem with both approaches. The generative approach excludes the conceptual system as a decisive component in shaping grammar. The relationship between thought and language has been completely ignored in formulating theories of linguistic structure and behavior. Researchers focusing on monolingual LI development have been convinced that in the monolingual conceptual base everything is language-independent, containing mainly universal concepts that are lexicalized in different ways in different languages. The most extreme view on this issue is that of Chomsky who sees the lexicon of a language as a set of labels to be attached to concepts that are languageindependent and are determined biologically rather than culturally (Chomsky, 1987). He maintains that "acquisition of vocabulary is guided by a rich and invariant conceptual system which is prior to any experience (Chomsky, 1987:33)". Generative semantics identifies the word with what it stands for: the concept or conceptual structure, because it lists conceptual properties as essential properties of the word

18

Chapter 1

itself. The lexical learning hypothesis proposes that lexical items, along with their properties, trigger the reconstruction of a child's grammar. These labels, i.e., the lexical items are attached to concepts which are language- and culture-independent. As opposed to the generative approach the cognitivist view equates the meanings of linguistic forms ("semantic structures") with conceptualizations which are language- and culture- specific and considers the conceptual-semantic > grammatical sequence a continuum. In this paradigm, as Kövecses claimed, "it is unreasonable to talk about the autonomy of semantics (Kövecses 1993: 255)", because meaning is very often a function of the conceptual system, contextual information and cultural influences. Neither the generativist nor the cognitivist view appears to be acceptable in its present format because neither of them seems to catch the dynamism of meaning in which conceptual and lexical properties are brought together, and which amalgamates synchronic anddiachronic information. Taylor (2000:120) argued that if someone wants to understand an expression, s/he needs to rely not only on encyclopedic (conceptual) knowledge, but also on "knowledge of the language in which the expression is encoded." He asks the question: "what does a person who knows the meaning of a word' actually know?" Langacker's position is that knowing the meaning of, say, the verb "open" implies knowing the kinds of activities a person performs when s/he opens entities of various kinds (beds, books, wounds, doors, bank accounts, and letters). This approach entails a high degree of polysemy in the mental lexicon. According to Bierwisch's two-level model, this kind of encyclopedic knowledge is provided only in the act of interpreting the word in context, while the word meaning itself is underdetermined with respect to its possible interpretations in different kinds of context (Taylor 2000:121). This position seriously limits polysemy in the mental lexicon. Langacker does not draw a distinction between linguistics semantics and conceptual (encyclopedic) knowledge. For him meaning is equated with conceptualization. He characterizes semantic structure as conventionalized conceptual structure (Langacker 1991: 109). In contrast, Bierwisch makes a clear distinction between a linguistic level of meaning, and a non-linguistic level of conceptual representation. This approach relegates polysemy to the conceptual level. Words have several meanings because of the several possible conceptual interpretations. The goal is to collect all possible interpretations under a unitary semantic entry. Taylor pointed out that "on the

Lexical and conceptual level

19

two-level model, interpretation of an expression emerges through the interaction of the unitary semantic representation with conceptual information, relative to a context (Taylor 2000:130)." In Langacker's network model different senses of a word are organized in a network where not all the nodes are equally salient. Salience as it will be discussed in chapter 3, is connected with accessibility. The more accessible a sense is the more salient it is. The most salient node in the network may be considered as the prototype, or the prototypical sense of the lexical unit (c.f. Taylor 2000; Williams 1992). The Dynamic Model of Meaning (DMM) that I put forward in this book offers an alternative to the two mainstream approaches. DMM emphasizes that meaning can be understood only if we take into account its dynamic structure. This means that a lexical item both in context and without context (as a dictionary entry) assimilates present (synchronic) and past (diachronic) information, incorporates conceptual and lexical properties, and has both permanent and temporary aspects. The real issue is to what extent the lexical unit comprises pre-sent and past information, and to what extent it expresses permanent and temporary aspects when it is used for a particular purpose. This "extent" changes synchronically when the lexical item is used in one context or another, and it changes dischronically when the lexical item is an entry in the dictionary where it is put to symbolize something. Meaning has both relatively permanent and temporary elements, and does not let us make a clear dividing line between linguistic and encyclopedical knowledge. Additionally, it does not allow their full merging. Meaning is not just something that is general and abstract, and resides in dictionaries, or even in general symbolic representations inside people's mind. Meaning is situated in specific social and cultural practices, and is continually transformed in those practices. Gee claimed that "meaning is a matter of situated meanings, customized in, to and for context, used always against a rich store of cultural knowledge (cultural models) that are themselves "activated" in, for, and by contexts (Gee 1999: 63)."

1.4. Can the conceptual level be separated from the lexical level? Before we discuss the DDM we need to take a closer look at the lexical- conceptual debate. Slobin (1997) argued that language is a transmitter of real-world experiences, and that these experiences are

20

Chapter 1

filtered through language into verbalized events as it is demonstrated below. Real World Experience- > Language + Conceptual System—> Verbalized Events The main question is whether our experience with the real world filters through our conceptual system which contains language as the majority of cognitive linguists claim, or while fulfilling its "filtering function" language constantly interacts with the conceptual system as in the figure below which presupposes a certain kind of independence of both systems. Real World Experience—>Language System—>Verbalized Events

t

Conceptual System

1.4.1. Cognitivist views on separation As seen earlier the debate on the separation of the conceptual and lexical level does not necessarily go on between generativists and cognitivists, although it concerns the basic difference between the two perspectives. Rather, it involves linguists from both sides. Bierwisch is a generativist, and Wierzbicka is a cognitivist; still, they are on the same side although their reasons are different. Several researchers who are considered as cognitive linguists argued for keeping lexical semantics separate from conceptual semantics. Haiman claimed that while theoretically untenable, the distinction between dictionaries and encyclopedias has "the happy property of working very well in practice (Haiman 1980: 355)". Gibbs (1996) also said that in his estimate cognitive linguists would not argue against the idea that there is a mental lexicon that might be independently accessed during sentence processing. In his opinion part of the confusion about the role of cognitive structure in language use and processing results from the failure to distinguish between different levels at which cognition and language interact. Wierzbicka (1996) strongly believes that word meaning can and needs to be defined at the lexical level. She argued that the ac-

Lexical and conceptual level

21

quisition of another language demonstrates very well the importance of lexical semantics. According to Wierzbicka, "for anyone seriously trying to learn another language and understand another culture, the proposition that words cannot be defined can hardly be anything but bad news (Wierzbicka 1996:256). She said that the "belief that a dictionary definition represents nothing other than a selection from a (real or imaginary) encyclopedic entry, with the choice being determined by practical considerations and having no theoretical justification, leads to stagnation in lexical semantics (1996:336)." Wierzbicka makes the distinction between linguistic knowledge and encyclopedical knowledge because she thinks that linguistic knowledge is essentially shared between all speakers of a particular language, while encyclopedical (real-world) knowledge is not. She pointed out that there is linguistic evidence that the human mind itself draws a distinction between a mental dictionary and a mental encyclopedia (Wierzbicka 1996: 344). The nature of available linguistic evidence is summarized by Taylor (forthcoming) who argues that "acquisition is not a process of building up a concept from its constituent parts", but it consists in the "gradual elaboration of a knowledge network" (Quoted inPeeters 2000:16). Another important observation comes from Harnad (1990) who said: "Once one has the grounded set of elementary symbols provided by a taxonomy of names (and the iconic and categorical representations that give content to the names and allow them to pick out the objects they identify), the rest of the symbol strings of a natural language can be generated by symbol composition alone, and they will all inherit the intrinsic grounding of the elementary set (Harnad 1990: 343-344)." This implies that higher order symbols can be interpreted without direct acquaintance with reality. We should not think, however, that these higher order symbols are not grounded in experience at all. They, together with their underlying symbolic representations, are derived from the sensory representations and thus indirectly grounded in experience. Actually, the real issue here is not whether the content of lexical units is grounded in experience or not, because in most cases they are grounded in experience directly or indirectly. What is important for us is that linguistic symbols can entirely be cut offfrom their original source and live a life of their own which is independent from the conceptual system to a particular extent and from a particular perspective.

22

Chapter 1

Based on the instances of mono- and multilingual development described above it does not seem to be correct to claim that linguistic units (words) reflect specific, autonomous linguistic knowledge that is separate from various conceptual knowledge. However, it might very well be that some linguistic knowledge that is attached to the word form is partly autonomous from the rest of our conceptual system (Gibbs 1996). Croft and Woods argued that "it is not the case that any time we think we must conceptualize our experience the way that our language requires us to. But it is the case that any time we express our thoughts in language, we must conceptualize our experience in the way that our language requires us to. Cognition may be linguistically neutral, but language is not semantically neutral (Croft and Wood 2000: 55)". Slobin (1991) also made a similar point when he described "thinking for speaking" as the appropriate domain for the influence of language on thought. The issue here is not necessarily whether language influences thought or not but to what extent. The world is always given to us under some construal for our purposes. The question is: how much of this construal is provided by the language we speak? We will return to this issue at a later point.

1.4.2. Child language acquisition Two strong arguments for the separation of lexical semantics from conceptual semantics come from child language acquisition (c.f. E. Clark 1973; Nelson 1974; Rosch 1975) and bi- and multilingualism research (c.f. Kecskes 1998; Kecskes and Papp 2000a). Each explanation for children's acquisition of word meaning (semantic feature hypothesis of E. Clark, functional core hypothesis of Nelson, prototype theory of Rosch) emphasizes that children do not acquire complete knowledge of a word's meaning all at once. Rather, children initially may identify the meaning of the word with only one or two of its semantic features. They seem to learn words as labels only without the full concepts they represent. This fact can be interpreted in two ways. Chomsky claimed that the rate of lexical acquisition in children is so fast and precise that we have to think "that the child somehow has the concepts available before experience with language and is basically learning labels for concepts that are already part of his or her [innate] conceptual apparatus (Chomsky 1988:28)". Another possible explanation is that children develop lexical meaning and conceptual meaning

Lexical and conceptual level

23

parallel in constant interaction. So there is nothing a priori about the concept. Simply, the concept is not in the mind of children yet, or is not fully developed when the word is already used. The word seems to live its own life for a while as a relatively independent lexical unit without being fully associated with the concept it denotes. As James (1990) said, children may both overextend and underextend words. Overextension means that the child uses a word to refer to objects or events which are not part of the conceptual category for that word. For instance, they may use the word "dog" to refer to any four-legged animal. Underextension occurs when the child uses the word to represent a more restricted category than the full concept. For instance, the word 'flower' is used to denote only roses (James 1990:44). These examples demonstrate that there is hardly any one-to-one relationship between a conceptual category and a word that denotes it. Words are often used before their conceptual representation is in place. Initially the child may identify the meaning of the word with only one or two of its semantic features. Both underextension and overextension point to the fact that the linguistic unit and its conceptual representation are engaged in a unique interplay. Brooks (1978:68) argued that the first-language "learner is likely initially to acquire a concept representation based upon a particular instance. This may be the first instance encountered by the learner, and thus it may not necessarily be a 'best' instance. Subsequent encounters with other instances bring about changes that may alter the stored representation of the concept toward the direction of typicality. However, such changes will continue to depend upon the particular in-stances and situations encountered by the learner." Because the learn-er does not acquire the concept at the first encounter, several other encounters of the same concept in various contexts should follow until the concept is firmly established in the learner's mind. Gagne (1977:99) claimed, that "concrete concepts that are learned by young children are often given fuller meaning and greater precision when they are later brought to the formal level (as defined concepts) by learning in school." Thus concepts in the LI are not learned, rather they grow. Conceptual content constantly changes under the influence of input and the environment. The more frequently the learner is exposed to a certain word in the input, the more firmly the concept becomes established in the mind. But the development of meaning relations between words and concepts is more than a simple clustering of meaning aspects. As

24

Chapter 1

Vygotsky (1962:83) pointed out "...word meanings evolve. When a new word has been learned by the child, its development is barely starting: the word at first is a generalization of the most primitive type; as the child's intellect develops, it is replaced by generalizations of a higher and higher type ~ a process that leads in the end to the formation of true concepts."

1.4.3. Bi- and multilingual development A very similar process occurs in bi- and multilingual development in the first phase of which learners acquire new labels (words) from L2 which have no conceptual content but are attached to existing LI concepts. This is what is called the "lexical association" (Kroll and Stewart 1994; Kroll and Tokowitz 2001) period in which bilinguals are aware of the word specific properties of L2 lexical units rather than their conceptual features that distinguish them from their LI counterparts (Kecskes and Papp 2000a). Connecting the new L2 labels (words) to the existing LI concepts is especially common in foreign language learning and in the first phase of second language acquisition. In this developmental period there is a clear difference between lexical semantics represented by the new L2 words and conceptual semantics represented by the existing LI concepts. The L2 learner's real problem is that s/he is not familiar with the word-specific semantic properties of the new labels and the culturespecific conceptual properties3 of the L2 concepts with which the new L2 labels (words) are associated. This situation changes when higher proficiency in L2 results in concept mediation where the L2 word is supposed to have an immediate contact with the L2 concept (Kroll and Stewart 1994; Kroll and Tokowitz 2001). However, this is not exactly what happens. L2 learners concept mediation differs from LI learners concept mediation. In the LI direct contact is established gradually between the concept and the word that denotes it. In L2, however, the new L2 word will have an immediate contact with the modified version of the LI concept which is enhanced with the L2 culture-specific conceptual properties. When a native speaker of English is studying Spanish and acquires the word 'COMIDA' as an equivalent of 'LUNCH', s/he will not establish a new concept in the mind for 'COMIDA '. Rather the learner will modify the existing concept for 'LUNCH' by adding L2 (Spanish) socio-cultural information (culture-

Lexical and conceptual level

25

specific conceptual property) to the LI (English-based) concept. As a consequence, a hybrid concept with encyclopedic information from both cultures is established in the mind which is represented by the word 'LUNCH' through the English language channel, and the word 4 'COMIDA ' through the Spanish language channel. This process is called "reconceptualization" (Kecskes 1995 ; Kecskes and Papp 2000a), which shows very well that the linguistic (word) level and the conceptual level are two sides of one and the same phenomenon, but to equate them would be a mistake.

2. The tri-Ievel production system The separation of grammatical, lexico-semantic and conceptual levels is closely connected with the theory of modularity, according to which a person's linguistic knowledge constitutes an encapsulated mental module which is independent of other skills (c.f. Fodor 1983). Each level has its own importance and role and is inseparable from the other two with the lexico-semantic level playing an interface role between the conceptual and the grammatical level, while still having a particular kind of independence. The semantic level serves as a bridge which links the linguistic system to the conceptual system. So the semantic entry for a lexical item points into two directions. "On the one hand, the entry comprises specifications that allow the item to enter into syntagmatic relations with other items. The entry also contains specifications that make it possible for the word to identify conceptual units at the conceptual level (Taylor 2000:123)." There are also three levels presented (although from a different perspective) in a tri-level speech production model that was put forward by Levelt (1989). Levelt clearly separated the'containers'of declarative knowledge (knowledge base and the mental lexicon) and the'processors'that operate on that knowledge. According toLevelt's theory the conceptualizer, i.e., the message generator is responsible for three things: communicative intention; planning the content, and planning the form of the message. Content planning means "selecting the information whose expression may realize the communicative goals." (Levelt, 1989:5). Message generation is affected by the speaker's knowledge base consisting of encyclopedical knowledge, situational knowledge, and actual knowledge (what has been said before in the interaction). The output of the conceptualizer, which is a preverbal

26

Chapter 1

message, is fed into the next component called the formulator. The formulator is responsible for grammatical and phonological encoding which takes place through extensive interaction with the lexicon. In Levelt's theory the mental lexicon contains not only word meaning but also syntactic, morphological and phonological information about the lexical items. In the course of production the conceptual information in the preverbal message activates the appropriate lexical items which trigger the syntactic information attached to them. This is how grammatical encoding starts. As a result of that process, a surface structure emerges, which will further go through phonological encoding based on the information in the form components of words. The last step is when the articulator transforms the phonetic plan produced by the phonological encoder into overt speech. In Levelt's model there are two sources of declarative knowledge. One is the knowledge base with actual, situational and encyclopedical information, and the other is the mental lexicon. They stand as independent entities: the first interacts with the conceptualizer, and the second with the formulator. The conceptual level is connected with the grammatical level through the léxico- semantical system which is the lexicon. Although all these levels are interconnected, each has some kind of autonomy. It does not seem to be feasible to merge the grammatical level with the semantic level and ignore the conceptual level as done by generativists, or merge the conceptual level with the semantic level as the cognitivists do. According to the cognitive approach, the conceptual system that emerges from everyday human experience is the basis for naturallanguage semantics and word meanings which may not apply at all outside the relevant background assumptions since meaning and its frame are inseparable from one another (Sweetser, 1990). Meaning is rooted in human cognitive experiences: experience of cultural, social, mental, and physical worlds. Consequently, background knowledge which constitutes frames for meaning plays a very important role in shaping meaning of linguistic signs. Taylor argued that if the meanings of linguistic forms are equated with conceptualizations, and these conceptualizations are conventionalized in a language, then the conceptualizations are made available to speakers of a language by the language system that they have learned (Taylor 1993: 212).

Lexical and conceptual level

27

3. Relative independence of linguistic signs There is no doubt that human experience is the basis for meaning, and, to a certain extent, languages predispose their speakers to conceptualize experience in a way as described by Slobin (1991) and Croft and Woods (2000) above. However, we should not overestimate the role of either side. Language is the result of human experience which creates for itself a "monster" that claims a certain kind of independence from its source. This independence is partly demonstrated in that languages do not always reflect human experience adequately. This fact is acknowledged by cognitive linguists in, what is called, 'the perspective nature of linguistic meaning" that implies, as Geeraert says, that "the categorization function of language imposes a structure on the world rather than just mirroring objective reality. Specifically, language is a way of organizing knowledge that reflects the needs, interests, and experiences of individuals and cultures (Geeraerts 1997:8)." Harris (1988) speaks about factual misrepresentation and conceptual misrepresentation, and illustrates this through an example from French. French grammar assigns the masculine gender to the word 'professeur' which is a factual misrepresentation of the truth because there are women professors as well. It is also a conceptual misrepresentation because French speakers do not believe that only men can be teachers. So their idea is misrepresented in the language which still sticks to a historical fact. We can find a similar example in Russian where gender differences in professions are denoted by different words:

(6)

Masculine yuumejib βραπ napuxMOxep

Feminine ynumexbHUifa epauuxa napuxMaxepiua

[teacher] [doctor] [hairdresser]

However, when the profession of a person is given in a document such as passport or other, it is always the masculine form that is preferred even if the person is a woman. Linguistic signs seem to have a certain kind of independence, and as a result of this independence they have some relatively permanent elements in them. Therefore, they not only reflect human experience but

28

Chapter 1

also affect and filter human experience (Slobin 1997). What meaning they have within their system (linguistic competence) and outside their system (language use) is a crucial question. Meaning formation appears to be a two-way street. This is where thought and language meet. Human beings have a remarkable capacity to create words for concepts, which presumably indicates that such concepts do not depend for their existence on the préexistence of a word. The Russian concepts [perestrojka] and [glasnost], the French [détente], or the English [software] are some recent examples of word creations for newly emerged concepts. The appearance of these concepts can be explained by societal need: the concepts were there and they had to be labeled with a word in order for them to mean the same for each member of the given speech community, and what is more, in the case of these particular words, for the world's language community. As they were so strongly language-specific, speakers of other languages found it reasonable to borrow these words from Russian and French and adopt them to their own system rather than find new labels for them in their linguistic system. Words, however, are needed not only to symbolize concepts but also to stabilize them, keep them tidy and make them definable. This, however, can only be done if a word or formula has some elements in its meaning that are relatively constant, i.e., part of the language system and not totally dependent on actual use and context. Leibniz was right when he said: "Si nihil per se concipitur, nihil omnino concipietur", that is, 'if nothing can be understood by itself nothing at all can ever be understood (Leibniz 1903:430)." In sum, we can say that human communication would hardly be possible if words did not have any relatively permanent elements in their meaning, and meaning formation would only be the matter of on-line production. If, however, these relatively constant elements of meaning really exist, it is questionable what cognitive linguists claim that word meanings may not apply at all without the supporting cognitive structures and mechanisms. The main issue about meaning can be summarized in three questions: What is word meaning? How is it shaped? How does word meaning relate to the conceptual system and the grammatical system? These questions are important from the perspective of SBUs because they usually behave like words, and some linguists (c.f. Kiefer 1995; Kecskes 2000a; Wray 2002) maintain that they have to be listed in the

Lexical and conceptual level

29

lexicon. If we want to explain the meaning structure we will need to apply an approach that is the same or similar to the one we use for words and word meaning.

Chapter 2 A dynamic model of meaning

When language is used, its "magical" property is activated in two ways. When people speak or write they craft what they need to express to fit the situation or context in which they are communicating. But, at the same time, the way people speak or write, the words, expressions and utterances they use create that very situation or context in which the given communication occurs. Consequently, two things seem to happen simultaneously: people attempt to fit their language to a situation or context that their language, in turn, helped to create in the first place (Gee 1999). This dynamic behavior of human speech and reciprocal process between language and context basically eliminates the need to ask the recurring question: Which comes first? Is it the situation that the speakers are in (e.g. faculty meeting, car renting, dinner ordering, etc.)? Or is it the particular language that is used in the given situation (expressions and utterances representing ways of talking and interacting)? Is this a "faculty meeting" because participants are acting and speaking that way, or are they acting and speaking that way because this is a "faculty meeting"? Acting and speaking in a particular way constitutes social situations and these social situations require the use of a particular language. "Which comes first" does not seem to be a relevant question synchronically. Social and cultural routines result in recurring activities and institutions which give us the impression that they exist outside language. But these institutions and routinized activities have to be rebuilt continuously in the here and now. This is how the change, transformation and the power of language-in-action in the world can be accounted for (c.f. Gee 1999). Of course, building and rebuilding our world occurs not merely through language but through the interaction of language with other real life phenomena such as non-linguistic symbol systems, objects, tools, technologies, etc. The dynamic and multidimensional relation of lexical items and context results in meaning that is multiple, flexible and tied to culture. It is rooted in our recurring experience with the world. Mey said that "we bespeak the world, and it speaks back at us (Mey 1985:166)." Meaning is constructed in the dynamic interplay of context and lexical items with the context representing the actual, present, situational, ever-changing side and the

32

Chapter 2

lexical item(s) embodying previous experiences and relations. A theory of meaning should grasp the dynamism of this inter-play and explain its result. The Dynamic Model of Meaning (DMM) I am presenting here is an attempt to: 1) explain the role of a lexical item as an interface between the conceptual and lexical level, and 2) reconcile the two extreme approaches to meaning: one that may be represented by Wierzbicka's (1996) Natural Semantic Metalanguage (NSM) theory and the other that derives from Wittgenstein's (1953) claim that meaning is a function of use, and there is no uniform set of attributes that constitutes the meaning of a linguistic expression. There are only similarities between different uses of an expression. Consequently, it is not possible to grasp, or theoretically pin down the meaning of an expression. In contrast, the NSM approach states that each word has an invariant meaning, which can be represented by means of a semantic metalanguage, and the concept behind a word is primary with respect to its use in communication. Before the suggested solution is presented in the DMM we need to analyze the problem. So first we will have an overview of the relationship of concept and word, and how thought is generated as a result of their interplay.

1. The concept-word relationship 1.1. Two perspectives: semasiology and onomasiology When putting the relationship between concept and word into perspective, a distinction has to be made between two approaches. Kurt Baldinger described this distinction in the following way: " Semasiology [...] considers the isolated word and the way its meanings are manifested, while onomasiology looks at the designations of a particular concept, that is, at a multiplicity of expressions which form a whole (Baldinger 1980:278)." The semasiological perspective focuses on the word and explores how concepts (conceptual meanings) are associated

A dynamic model of meaning

33

with that word. So, semasiology target polysemy and the multiple applicability of a lexical item. In contrast, the onomasiological approach takes the concepts as its starting-point and investigates what words can express a particular concept. The subject of onomasiology is synonymy, name-giving, and the selection of an expression from among a number of possible alternatives (c.f. Geeraerts 1997). It is clear from this distinction that semasiology refers to processes of meaning while onomasiology involves naming.

1.2. What is in the word, and what is in the concept? As a result of the thought-word interaction in online production , thought usually undergoes several changes as it turns into speech. During this process conceptual categories (concepts) are mapped on linguistic categories (words). The problem is that there is no one-toone relationship between concepts and words. There have been long debates about how much autonomy a lexical unit can have. Is it just a form which reflects the conceptual system and is filled with meaning when it is actually used? Or does a lexical unit have some kind of autonomy which allows for some consistency in its semantic content and a two-way relationship of semantics with both the conceptual level and the lexical level involved? Or from the perspective of the conceptual system: should concepts be considered as stable mental representations, or they are flexible, temporary, context-dependent constructions (c.f. Lamberts and Shanks 1997:3)? The relative consistency view may be represented by Vygotsky's (1962) approach. He insisted that there is a unique symbiosis between thought and word through meaning which is the result of their interaction. He quoted O. Mandelstam: "I have forgotten the word I intended to say, and my thought, unembodied, returns to the realm of shadows." (Cited by Vygotsky, 1962:119). What Vygotsky wanted to say is that it is thought and word through which conceptualization and verbal formulation are united, and thought and word are amalgamated through word meaning. He emphasized that the relation between thought and word is a process, " a continual movement back and forth from thought to word and from word to thought. Thought is not merely expressed in words; it comes into existence through them (Vygotsky, 1962:125)". Since thought and word are not cut from one pattern, there are more differences than likenesses between them. Speakers

34

Chapter 2

cannot put thoughts on words directly like ready-made units because, as Vygotsky says, "the structure of speech does not simply mirror the structure of thought (Vygotsky, 1962:125)." Having reviewed the relevant literature Smith and Samuelson (1997) came to a different conclusion. They reported that the traditional search for constant concepts has not been successful. Several studies (c.f. Malt 1994; Smith and Sloman 1994) have suggested that acts of categorization are not simply repeated; they actually vary. Different tasks and contexts appear to create different categories. Barsalau (1993) demonstrated that individual acts of categorization do not necessarily require an already represented concept. Based on these claims Smith and Samuelson (1997) suggested a unified account of category stability and variability that is not built on the notion of fixed, represented concepts.

1.3. Conceptual system: what causes the

controversy?

The nature of the human conceptual system is currently a matter of conjecture and controversy. As we saw the cognitivist view states that our conceptual competence, in various respects, subsists in our linguistic competence, and is not separate from it (Ross, 1992:144). The conceptual system and linguistic system are related through the interdependence of concepts and words. Most of the controversy centers around the question of how concepts and words are related to one another. A natural language possesses an inventory of lexical forms, and these are mapped onto the concept network (Cruse, 1992:290). The question is: how does this mapping take place? In the former chapter it was reported that not all cognitive linguists accept that lexical items can be equated with concepts. It is true that concepts are represented by words, but this representation is rather contradictory. The contradiction is caused by two facts: 1) There is no one-to-one relationship between the concept and the word: the overall concept usually extends beyond the sections labeled with a word.

A dynamic model of meaning

35

2) The semantic domain of the word also contains specific properties which are not present in the concept. Sometimes there are several different lexical routes to the same concept. The Dynamic Model of Meaning offers a solution to these problems by explaining what is the common core in an word and the designated concept, what does not overlap in them, and how word-specific semantic properties and culture-specific conceptual properties are brought together in a lexical item when it is used in context. We will proceed from the outside to the inside, and first discuss what is not common in concepts and words, and then make an attempt to describe what they share.

1.3.1. Two approaches to categorization There are two approaches to human categorization: 1) the "classical approach" (linked with Aristotle) which is based on definition and the theory of invariant, and 2) the "prototype approach" (linked, in particular with Wittgenstein and Rosch) which postulates that "category boundaries are not necessarily definite (Mervis and Rosch 1981:109)", and "natural language concepts are inherently vague (Hersch and Caramazza 1976:273)". Wierzbicka (1966:148) argued that what we need is not a choice of one over the other but a synthesis of the two traditions. She made an attempt to do this synthesis in her Natural Semantic Metalanguage theory. This is also the goal of the Dynamic Model of Meaning, but the approach applied in the DMM differs from that of Wierzbicka's. However, before the difference is discussed, we need to have a closer look at the prototype approach first whereas Wierzbicka's NSM will be discussed later. Rosch and Mervis wrote about Wittgenstein's role in the foundation of prototype theory in the following way: " Wittgenstein [1953] argued that the referents of a word need not have common elements to be understood and used in the normal functioning of language. He suggested that, rather, a family resemblance might be what linked the various referents of a word. A family resemblance relationship takes the form AB, BC, CD, DE. That is, each item has at least one, and probably several, elements in common with one or more items, but no, or few, elements are common to all items (Rosch and Mervis 1975:574-575)."

36

Chapter 2

The human mind has a core meaning for a concept and relates variations to this central meaning. At the same time certain concepts are treated as archetypes, and others are seen in relation to these central concepts (Cook, 1997:91). Rosch(1977) postulated the "prototype theory" which claims that the mind puts objects into three levels of vocabulary. The basic level contains those objects that people notice immediately when they look at the world around them, such as [CHAIR], [TREE], [HOUSE], etc. These central prototypes are the ones already encountered. When we look at a street, for instance, we tend to see things as cars, houses, shops, etc. not just sheer shapes and forms. This is how our world is organized in our minds. Above the basic level we have a "superordinate" level, and below it there is a "subordinate" level. The superordinate level includes more general terms which are abstractions and are not directly available to our eyes. The subordinate level puts basic terms into more specialized categories. Cook (1997:91) argued that "the mind sees the world in three levels of abstraction going from the most general to the most specific, with the basic level coming in the middle as the most useful everyday term." For instance: (7) Superordinate: Basic: Subordinate:

BUILDING HOUSE TOWNHOUSE

ANIMAL DOG BULLDOG

1.3.2. Culture-specific conceptual properties The DMM makes an attempt to synthesize the "definition" approach and the "prototype approach" so it postulates that although concepts are vague, their definition is still possible if we concentrate on what is common in the concept and the word denoting the given concept and also try to describe what is clearly linguistic, and what is clearly conceptual in the concept-word symbiosis. We will start with the latter. As stated above, the overall concept usually extends beyond the sections labeled with a word because the conceptual content comprises encyclopedic background knowledge that is part of the native speaker's socio-cultural heritage. This is what makes many concepts culture-

A dynamic model of meaning

37

specific. These background elements constitute, what we call, "culture-specific conceptual properties" (Kecskes and Papp 2000a). These culture-specific conceptual properties (CSCP) are the attributes of the concept denoted by the lexical item. Encyclopedic knowledge is often expressed through these culture-specific conceptual properties. As mentioned earlier the English word 'LUNCIT is usually given as an equivalent of Spanish 'COMIDA '. This is definitely correct, if we ignore the socio-cultural differences. The concept denoted by the word 'COMIDA ' refers to the main meal of the day for Spaniards which they have between 1:00 and 4:00, and usually consists of three courses. The background knowledge to 'LUNCH' in American culture is entirely different: a quick, 30-minute meal between 11:30 and 1:00 consisting of different types of sandwiches. If a Spaniard sits down in an American restaurant to have his 'COMIDA', he will have an unpleasant surprise. However, it is important to note that not all concepts have these culture-specific conceptual properties, and sometimes, even if they do have them, these properties can be canceled by the context. The concept denoted by the word 'SALT' has no culture-specific properties. However, the word 'PUMPKIN'behave?, in a different way. It has a culture-specific conceptual property in American culture with a positive value: pumpkin pie is something that almost everybody loves, and parents often use this word to address their children: (8)

Father: - Listen Pumpkin, how about going for ice-cream? Margie: - Cool, let's go.

As seen in the Spanish example these culture-specific conceptual properties can be cross-culturally misleading. A conversation similar to the one above could not take place in the Hungarian culture where the equivalent of the English word 'PUMPKIN' is "TÖK" which has a culture-specific conceptual property with a negative value. Hungarian children would not be happy if they were addressed by the word "TÖK' which may mean "stupid". However, culture-specific conceptual properties can be canceled by the context. The CSCP (which has a positive value) that is attached to the concept denoted by the word 'PUMPKIN' in American-English can be canceled, for instance, by the following context: "Quite by accident Jack stepped on a pumpkin that broke into pieces and left a yellow spot on his pants." Not all aspects of a concept are lexicalized when it is use in a context. Some elements of the meaning potential of the concept may come

38

Chapter 2

to the fore; others may be canceled. Bibok and Németh (2001,2002) studied an interesting case in which not all the concept is lexicalized in an utterance. They focused on utterances with implicit arguments or predicates, and investigated how an argument of a verbal predicate (A), or a part of the verbal predicate (B) can be left implicit and how the references of lexically unrealized arguments and predicates can be recovered. For instance: (9) A. Paul gave (money) to Amnesty International. B. Mary began (reading) the book. Bibok and Németh (2001) and Németh (2002) argued that the identification of implicit arguments relies to a great extent on the encyclopedic knowledge stored in the concept denoted by a particular word in the utterance. This case demonstrates very well the dynamic structure of meaning, and that a word does not always denote the whole concept. As a similar example we can mention a rather interesting tendency in present day American-English. Several verbs seem to incorporate in their stem certain meaning aspects that used to be expressed, and many times, still are expressed by prepositions: (10) Thank you for shopping Albertsons's. (Thank you for shopping at Albertson's.) They always fly Northwest. (They always fly with Northwest.) Teach Turkey. (Teach in Turkey.)

1.3.3. Word-specific semantic properties The semantic domain of the word also contains specific properties which are not present in the concept. These word-specific semantic properties (WSP) make it possible for speakers to have alternative lexical access routes to a single concept (Cruse 1992). Several different

A dynamic model of meaning

39

words that are considered to be synonyms may refer to the same concept. WSPs are the result of recurring use of words in particular contexts. Originally WSPs are the results of the interplay of the lexical unit and context, and the best proof for category stability and variability. They usually derive from metaphors or other figures of speech, and in time they have become conventionalized. So first they might be "charged" to the word by the context, and after a time, as a result of frequent use and/or conventionalization, they may be attached to the word permanently and become the property of the word form. This means that whenever the word is used, this "color", this "stigmatism" ( and often "metaphorism") is there, which, basically, distinguishes the word from other words with similar meaning. Words with WSPs can create a more emotive, more "colored" context than words without WSP. Compare the following sentences: (11) Our dean died yesterday . Our dean passed away yesterday. John repaired the car.< > John fixed the car. Cruse (1992:291) argued that cognitive synonyms map onto identical concepts. The meaning properties that differentiate cognitive syn-onyms like 'GIVE UP'¡'CAPITULATE'/SURRENDER'; 'CHICKEN OUT'5 can be viewed as properties of the individual lexical units, as distinct from properties of the common concept. Wordspecific sem-antic properties will include such things as emotive coloring, stylistic value and various kinds of contextual affinities (see Cruse, 1990, for a more detailed discussion of word-specific semantic properties). According to the approach developed by Cruse (1990; 1992) lexical semantic relations (word-specific semantic properties) are opposed to conceptual semantic relations. For instance, if the word 'SURRENDER ' serves to activate the concept [SURRENDER] directly, without modulatory effect, whereas the word 'CHICKEN OUT ("to act in a cowardly manner") activates the same concept, but with a modulatory effect, then the word ' CHICKEN OUT stands in a meaning relation to the word 'SURRENDER'. Each of the other members of the synonymy group underlines a different side, a different aspect of one and the same concept [SURRENDER]: 'GIVE UP' implies the activity of the agent while 'CAPITULATE' involves "surrender" under prearranged conditions. According to Cruse's approach the whole meaning of a word comprises the associated concept (together with its pattern of connec-

40

Chapter 2

tions within the concept network) plus any word-specific properties. This is how thought and word are united through meaning. The problem with Cruse's definition is that he ignores culture-specific conceptual properties which can also be attached to the word, as a constituent of its "whole meaning". Words at the lexico-semantic level differ from one another in terms of word-specific semantic properties. Cruse talks about "plain word" when the word does not have any word-specific semantic properties, and "charged word" when the word has semantic distinguishing features.6 There are two problems with Cruse's approach. First, Cruse's interpretation presupposes that words have context-free meanings on the lexico-semantic level (This will be questioned in the next chapter). Second, in unbound discourse it is very often the situation itself that "charges" a word and activates the word-specific semantic property (Kecskes, 1997; 2000a). In this case the WSP is not carved (encoded) into the semantic domain of the word yet. This can be illustrated by the following example: (12) - Please bring that parcel here. - Sorry, it's too heavy. I can't move it. ("plain word") (13) -- We thatmy book asap. Go and get it. OK,need I'm on way. - Move it! (situationally "charged word") In the DMM a word-specific semantic property can be either a permanently-present or a situationally-activated element of the word meaning. In the words "chicken out", or "pass away" the WSP is permanently present: it is there even if the word is uttered out of context. In other cases the potential of the WSP is in the word form but it has to be activated by the context. As a result, the word gets "charged", that is to say, gets a certain kind of metaphorism, a special emotive, stylistic or contextual coloring in the situation where it is used (see [13] above). Since this difference is significant, I suggest that the two word-types be distinguished by different labels: When the word-spec-

A dynamic model of meaning

41

ific semantic property is encoded in a word, the lexical item should be referred to as a "loaded word" {[DIE] - 'PASS AWAY'; [SURRENDER] - 'CHICKEN OUT; [RUN] - 'DASH'; [INTERFERE] - 'BUTT IN'), and when the basic semantic function of a word is extended with the WSP the word should be labeled as "charged word". It has to be emphasized that word-specific semantic properties ('loads'), just like culture-specific conceptual properties, are NOT mandatory features that are attached to each lexical unit. There are lexical units which seem to be neutral from this perspective and have no word-specific semantic properties such as, 'DIVISION', 'LAW', 'EXAMPLE, 'DEPART', etc. There are also concepts which are relatively universal without culture-specific conceptual properties. Context can cancel culture-specific conceptual properties (as we saw in the case of 'PUMPKIN' above) but this hardly is the case with word-specific semantic properties when they are encoded in the word. Context can activate WSPs but hardly ever cancel them completely if they are encoded in the lexical meaning of the word (such as, "pass away", "chicken out", "dash ", etc.) and do not need contextual charge to come to the surface. This is due to the special hierarchy of processing which will be discussed later. When used in context, words generally keep their WSP which, however, can become vague as a result of contextual effects.

1.3.4. The complexity of word meaning In the DMM a lexical unit, may it be a word or an SBU, generally bring together the following meaning features: core-conceptual meaning, word-specific semantic properties, and culture-specific conceptual properties. Out of the three, only certain aspects of core-conceptual meaning are always present and mandatory although, to a certain extent, they can also be forced into the background by context as will be seen later. Culture-specific conceptual properties are conceptual semantics phenomena, while word-specific semantic properties are features of the word itself. They are a matter of lexico-semantics. Practically, the following permutations are possible:

42

Chapter 2

(14) A) WORD: core meaning (no WSP and/or CSCP) "Can you pass me the salt please? B) WORD: core meaning + CSCP Father: Listen pumpkin. let's go and buy some present for your mother's birthday. Betsy: OK, let's go. C) WORD: core meaning + WSP Bill: Do you know why Bob chickened out of the race? Steve: I have no idea.

D) WORD: core meaning + WSP + CSCP Excuse me for my butting in. I need to ask you a question. The president kicked off the new academic year with an excellent speech.

2. Coresense and consense in the DMM Now let us turn our attention to the common core of the concept-word symbiosis. The complexity of meaning justifies the introduction of two new terms in the DDM to denote the two faces of meaning that result from the diachronic and synchronic interplay of the lexical item and context: coresense and consense. The core meaning that is abstracted from possible contextual occurrences of the word is represented in the coresense of the word. Coresense has the basic conceptual information, (reflection of the concept) with or without culture-specific conceptual properties (CSCP) and with or without word-specific semantic properties (WSP) in use. Coresense as understood here is neither conceptual nor lexical but both. It is the interface between the lexical

A dynamic model of meaning

43

and conceptual level: coresense units the two levels. Coresense is an abstraction that is based on previous contextual experiences with the lexical unit. Coresense has a unique relationship with the word-specific semantic properties and culture-specific conceptual properties. WSPs link the coresense to the lexical level while CSCPs tie it to the conceptual level. Coresense should be distinguished from consense which is the situated, contextual meaning of the lexical unit. Coresense is the invariant while consense represents the possible variants. Consense realizes a particular aspect or aspects of the coresense when the word is actually uttered in a particular situation. For instance, consider the following two utterances about "coffee" (Gee 1999: 48): (15) "The coffee spilled, get a mop". (16)

"The coffee spilled, get a broom."

In sentence (15), the word "mop" triggers a meaning aspect of "coffee" which can be described as a "dark liquid", while in sentence (16) the word" broom" brings about another image of "coffee", which may be described as "dark dry beans". This dynamic change of aspects of meaning has led several researchers to assume that the human mind is a pattern recognizer and builder rather than a rule-following logiclike calculator. According to this view the mind operates primarily with flexibly transformable patterns extracted from former experience, and not with highly general or decontextualized rules (c.f. Barsalau 1992, 1993; Clark 1993, 1997; Gee 1992,1999; Minsky 1985). The mind works primarily by storing experiences and finding patterns in those experiences from which they were extracted and that shape how people engage with (and store in their minds) their subsequent experiences. Gee (1999) argued that words have situated meanings which have no supporting concepts behind them because a situated meaning is an image or pattern that speakers and hearers assemble "on the spot" as they communicate in a given context, based on their construal of that context and on their former experiences. I agree with Gee as far as the online production is concerned, but I am reluctant to accept that these recurring situated meanings do not result in any kind of knowledge domain which is responsible for coresense in the mind. Coresense is a diachronically changing feature of the word while consense is a synchronic feature. Coresense is a denotational, diachronic, relatively constant7 and objective feature, and it reflects

44

Chapter 2

changes in the given society, i.e., linguistic community while consense is actual, subjective, referential, connotational, and changes by situation and context. This actual situational meaning is the result of the interplay of conceptual mechanisms and structures that may create context, or are activated by context, and interact with context in shaping the actual meaning of a linguistic unit, may it be a word, a pragmatic unit or a sentence. The DMM postulates that the dynamic nature of meaning can be understood only if we conceptualize it as a phenomenon which has both a synchronic and a diachronic side that are inseparable and keep interacting all the time. Coresense is the reflection of the most typical, regular and salient instances of possible consenses of a lexical unit. It is an abstraction based on concrete experience, i.e., concrete occurrences of words. How does the understanding of coresense in the DMM differ from "core meaning" in linguistic semantics? In linguistic semantics, sense is identified with core meaning, and a difference is made between core meaning (denotative, conceptual, cognitive, etc.) and peripheral meaning (connotative, residual, etc.). Core meaning is characterized by a restricted number of properties and components that are taken to define a category in an adequate manner. Core meaning is the sense of the word which is taken to be the linguistically crucial aspect of meaning. Core conceptual components constituting the sense are all expected to be present in every case when the category is used appropriately. This leads to the claim that categories are defined by certain essential properties which must always be present (c.f. Kövecses 1993). So these essential properties constitute the sense of the word which is considered static to a certain extent (according to the traditional view) because it is always present whenever that word is used. I think this is where the problem lies. It is crucial how we conceptualize these relatively constant constituents of meaning. One thing is certain: they are not exactly static, and not all core conceptual components are present all the time. Coresenses keep changing diachronically and not sychronically as consenses do. Diachronic change is always much slower than synchronic change. Diachronic change always reflects the changes in the language use of a particular linguistic community. Words have histories. Gee (1999) argued that words have been in other people's mouths and other people's pens, they have been circulated through discourses and been part of specific historical events and episodes. Therefore, he claimed, words carry as potential situated meanings all

A dynamic model of meaning

45

the situated meanings they have picked up in history and in different settings and discourses. Let us take an example. The sense of the word 'GAY' in the fifties and sixties was entirely different from what we have today. Irwin Shaw had a short story published in the fifties, titled "Wistful, delicately gay" where the word "gay" is used throughout the story and has nothing to do with the salient meaning that the word has nowadays. Gradually the meaning of the word has changed from 'joyful' to denote a homosexual person. Presently that is the most typical, regular and salient meaning of the word. That is the meaning which comes to the mind of most English native speakers if they hear the word out of context. Why is that so? Because the "cultural models" (c.f. Gee 1999) or frames associated with the word have changed. The same is true to lexical units bigger than a word. For instance, the fixed expression 'PIECE OF CAKE' has "easy" as its most salient meaning, and not a reference to a part, or a slice of a cake. So it seems to make sense to modify what we mean by core meaning. I use "coresense"to emphasize a different approach to the phenomenon. Coresense is not necessarily the sum of the most essential properties of the given category, but a summary of the most familiar, regular, typical and (generally but not always) frequent usage of a word in a particular developmental period of the given language. Consequently, coresense is not a pure linguistic phenomenon because it depends on extralinguistic features such as familiarity, conventionality and frequency. It is an essential feature of the word that brings together conceptual, semantic and grammatical information. This means that in a particular period of the development of a given language the most salient, familiar, regular and typical meaning of that particular lexical unit is the one described as coresense. This sense, however, is (or may be) a temporary phenomenon because it may last only for a particular period of time. Changes in society may lead to change in the use of lexical units. Since 1989, when I came to the U.S. for the first time, the signs used at road construction sites have changed several times: from "Men at Work", to "Crew at Work", then to "My mother at Work". These chang-es demonstrate very well how language reflects upon changes in society. Therefore, coresense can be characterized by three adjectives: most salient, relatively constant, diachronic as opposed to consense which is actual, situated and synchronic. The word connects the conceptual and lexical system through coresense while consense makes a tie to grammar which is shaped as constituents (words) of an

46

Chapter 2

utterance require. Coresense represents a temporary regularity while consense refers to variation. Both of them have their role in the dynamic multi-dimensionality of language and language use. As Violi claims "...language possesses an inherent dynamic which constantly moves between stability and instability, regularity and change, because it is neither a completely stable system of representation nor merely a cluster of continuous variations without structure (Violi 2000: 112)." Coresense and consense represent the dynamism and plasticity of language. Consense enables us to cope with contextual changes, while core-sense, the regularity aspect of meaning guarantees the possibility of stable sets of meanings which make actual communication possible (c.f. Violi 2000). The figure on the next page demonstrates the Dynamic Model of Meaning. In the model coresense is the interface between the conceptual and lexical level. WSPs are links to the lexical level while CSCPs are ties to the conceptual level. Consenses are the variations of coresense in context. The contextual interpretation of coresense is expressed in a consense which is connected to other consenses to form an utterance. WSPs are either encoded in the word or charged by the context while CSCP are encoded in the word and can be either activated or canceled by the context. As the model shows the constant interaction of the lexical items and context creates meaning.

A dynamic model of meaning

WSPs Lexical Level

C O Ν Τ E Χ Τ

Conceptual Level CSCPs

Figure 1: The Dynamic Model of Meaning

47

48

Chapter 2

3.1. How does the DMM relate to other two-level models? Focus will be on two models which appear to be closest to the DMM: Wierzbicka's (1996) Natural Semantic Metalanguage (NSM) theory and Bierwisch's (1996; 1997) two-level model. The points of comparison are as follows: 1) how is the invariant- variants relation understood?, and 2) the role of context in shaping meaning. All three theories make an attempt to identify a common semantic core. The NSM postulates that each word has an invariant meaning, which can be represented by means of a semantic metalanguage through semantic primes. Bierwisch's model attempts to bring the different readings of a word under a single common denominator. The DMM identifies the coresense as an abstraction deriving from the concrete occurrences of a concept represented by a particular lexical item. The concept is accessible to direct observation only through concrete uses of the lexical item in question. Wierzbicka (1996) gives particular importance to definitions. She argued that a definition is a scientific hypothesis about the concept that is encoded in a given word. The components of the concept determine which aspects of a word's use are available and which are invariable. The goal of a definition is to grasp the invariable aspects of a word's use, i.e., its semantic invariant. Bierwisch's theory is based on an underdetermined single common denominator whose meaning is specified by the context which shifts the core meaning into various conceptual fields. Coresense in the DMM differs from both Wierzbicka's invariant and Bierwisch's unitary denominator because it is neither a lexical nor a conceptual entity but both: it units the lexical and conceptual level. At the same time it is not a "fuzzy set of meaning components" because it is the abstraction of previous occurrences. In contrast with the other two approaches coresense is the diachronic face of meaning which is only relatively static with a possibility of change in time. The possibility of change and development presupposes a canonical "sense" which is represented in the coresense in DMM. What distinguishes between the DMM and Wierzbicka's NSM theory is basically the difference in perspectives. The DMM tends to

A dynamic model of meaning

49

be a more "context-friendly" theory than the NSM. The DMM seeks answer to the following question: What is in the meaning of a lexical unit when it is in use? In the meantime the NSM attempts to answer what is in the meaning of a lexical item when it is not in context? I need to underline, however, that this does not mean at all that Wierzbicka denies the importance of context in shaping meaning. Rather, it is that her primary concern is proving the existence of semantic primes, decomposing words into semantic primes and constructing explications out of a set of semantic primes. Wierzbicka's theory is wordcentered rather than context-centered. In evaluating Wierzbicka's latest book (Harkins and Wierzbicka 2001) Sharoff (2002) pointed out that the book provides an example study of how various phenomena referring to emotions in various cultures can be treated by decomposing words into semantic primes. He noted, however, that the discussion in the book says relatively little about the actual behavior of lexical items and about cases in which they are really used. In contrast, the DMM is based on the dynamic understanding of context according to which the meaning of lexical items can create context and be created by it. In the DMM the relationship of coresense and consense is a reflection of contextual use.

3.2. Coresense and consense in second language acquisition Second language acquisition and bilingual development can contribute to the understanding of the relationship between coresense and consense. Words in an L2 are learned rather than evolve. When acquiring a word second language learners usually have to rely on either dictionary definitions or a particular occurrence of the word or expression in a given context. When they learn definitions they approach word meaning through coresense. If they focus on situated, contextual meaning they encounter consenses. Neither of these approaches is really effective enough without the other although context-based vocabulary learning is favored in the literature (c.f.. Honeyfield 1987; Willis 1996). The problem is that L2 learners usually approach word meaning through the label (form) encountering words on the linguistic level rather than on the conceptual level. This is especially true for classroom language learning. What L2 learners miss is not the coresense but word-specific semantic properties (that distinguish the word from other possible lexical routes to one and the same concept),

50

Chapter 2

and/or the culture-specific conceptual properties (that may be attached to the coresense in actual use)8 WSPs are language-specific and CSCPs are culture-specific. Word equivalency is usually established by comparing coresenses. Hungarian learners of English usually think that the equivalent of the Hungarian verb "rúgni" is the English verb "kick". This is true if we take only their coresense: "to strike out with the foot or feet". However, contextual use shows that the almost equivalent coresenses are added WSPs and CSCPs which are very different in the two languages. Compare: (17) Equivalent coresense : Péter berúgta a labdát a hálóba. Peter in-kick the ball-ACC the net-LOC. 'Peter kicked the ball into the net.' (18) Language-specific WSP in Hungarian: Pétert kirúgták. Peter out-kick-PAST. 'Peter was given the sack.' (19) Language-specific WSP in English: They kicked off the holidays with a party. 'Egy parti val kezdték a szünidöt.' A party-INST start-Past the holidays-ACC.

Another example: (20) Szeretem a futballt. Like-TRANS. the football-ACC. Ί like football.' At first sight, the two sentences are totally equivalent. This is not so, however, if we think of the CSCPs that are attached to the core-

A dynamic model of meaning

51

sense in both languages. The way football is played in the U.S. and in Hungary is entirely different. The examples demonstrate that almost no equivalency is possible if WSPs and/or CSCPs are involved in either language. So coresense is relatively easy to access if there is no culture-specific conceptual property attached to it. It is not difficult to find a close equivalent to the following English words in most European languages: 'house', 'coat', 'play', 'speak', 'big', and so on. The fact that it is easy to find a close equivalent shows that there are words with similar coresenses in other languages. However, consense, which is tied to a concrete, situated, contextual occurrence and the interplay of neighboring expressions, represents a problem. Not being familiar with the WSPs and CSCPs of a lexical item L2 learners usually recognize only the specific function of a word in a context rather than its actual contextual meaning. They rely on compositionality which leads them to the coresense of the word rather than to the actual consense. For instance, in the utterance "For his bad behavior John was grounded for two days", the expression "grounded" has not only a CSCP but also a WSP so it has to be positioned to lexical items such as "restrict", "indoctrinate", "teach a lesson", and others, in order for L2 learners to understand the meaning of the word properly. Because of problems discussed above vocabulary learning in L2 is more likely to be effective if it relies on the analysis and acquisition of both consenses and coresenses of lexical units.

3.3. Thought and word interaction in online production As a result of the thought-word interaction in online production, thought usually undergoes several changes as it turns into speech. Thinking requires adjusting word coresenses to one another so that consenses in which thought appears in speech form a coherent and cohesive whole. Consenses change in actual use depending on the neighboring words in the utterance, and as required by the context in which the utterance appears. The result of this process is consenses of words that make an appropriate utterance.9 Let us take an example from (Kecskes and Papp 2000a):

52

Chapter 2

(21) While driving home Mr. Brown hit a cat on the road. (22) We must hit the road soon if we don't want to be late for the wedding. It is the knowledge of the English language that supplies the coresense, and it is the knowledge of the world that is added to it to supply the consense of the word in the given context. The aspects of the coresense of 'HIT' in the dictionary are as follows: "1. to come in contact with forcefully; to strike, 2. To cause to make sudden and forceful contact; to knock; to bump. (Morris, W. 1976. The American Heritage Dictionary of the English Language. Boston: Houghton Mifflin. P. 625.) The consense in sentence (21) is "to kill something by car", and in sentence (22) "set out". Vygotsky (1962) said that a word in context means both more and less than the same word without context. It means more because the word may acquire new content from its environment. This is where certain aspects of coresense can be forced to the background by context. It means less because the context may limit and narrow its potential coresense. The relationship between consense and word is much looser than between coresense and word. Consense keeps changing synchronically because it is shaped by actual thought. Coresense changes diachronically. Sometimes it takes words long decades to change their coresense as we saw in the examples ('gay'; 'piece of cake') above. Kövecses (1993) argued that recently several linguists had found that some of the key aspects of our semantic competence, such as polysemy, metaphorical and metonymical meaning cannot be given an account in terms of sense as a fixed and minimal number of components. He was right. However, the approach presented above that considers word coresense and word consense as two equally important sides of the semantic-conceptual interface may erase this problem. The word form unites conceptual meaning and linguistic meaning into one. Context specifies what aspect(s) of the concept need to be present in an actual use, i.e., in the situational meaning of the word. This can be illustrated by the following example: (23) Bob: Bill:

- Bill, I'd really like to tell you about my trip. - OK, shoot.

A dynamic model of meaning

(24) Bob: Bill:

53

- Bill, can I use your gun? I think I can easily hit that target from here. - OK, shoot.

The meaning of the SBU changes by the use. In (23) the coresense of the word 'SHOOT' is supplemented by a WSP which is based on the cognitive metaphor 'argument is war" (Lakoff and Johnson 1980). The WSP is added to the coresense which creates the consense: the actual situational meaning of the expression. In (24) 'SHOOT" is used in its literal sense without a WSP. Encyclopedic knowledge which, in our interpretation, is the source of culture-specific conceptual properties of a lexical item (Kecskes and Papp 2000a) plays an important role in utterance meaning interpretation. The two-level model (lexical semantics and conceptual semantics) proposed by Bierwisch (1983, 1996) distinguishes between two types of knowledge, and accordingly, two levels of meaning representation. As mentioned earlier the theory postulates that word meanings composed of semantic components are often not fully specified on the linguistic level. This may sound correct if we think of coresense as potentiality rather than something that is fixed and clearly defined. This frequently (not always !) underdetermined coresense is strengthened and made concrete by means of encyclopedic knowledge evoked by neutral contexts, i.e., ones that do not require metaphorical interpretation. Bibok and Németh claimed that the specification of word meaning (consense) in context is "done by conceptual shift, which 'shifts' the core meaning into various conceptual fields, and by conceptual differentiation, which only 'differentiates' the core meaning in different ways within one and the same conceptual domain, thereby yielding literal meanings. In other words, interpretations mean mapping underspecified semantic meanings onto fully determined conceptual meanings on the basis of our encyclopedic knowledge (Bibok and Németh (2001b:313)". I do not agree fiilly with the claim that core meaning (coresense) is underdetermined. In my interpretation coresense is the sum of previous occurrences of the word, which in itself can have a context-creating force, because based on previous experiences it can bring about a standard context as will be explained in the next chapter. In sum, the DMM is a model that builds on the complexity, dynamism, modularity and the multi-dimensionality of meaning. The dia-

54

Chapter 2

chronic and synchronic aspects of meaning are equally important in the model. The former are represented in the coresense which is the interface between the lexical level and the conceptual level. Wordspecific semantic properties tie the coresense to the lexical level while culture-specific conceptual properties relate it to the conceptual level. The interplay of context and word results in consenses in which particular aspects of coresense as well as WSP and/or CSCP are expressed according to communicative needs. How is the standard context modified in actual use? Which aspects of coresense are highlighted in the given situation? What culture-specific conceptual properties and/or word-specific semantic properties are expressed in actual use, and how? These issues will be discussed in the next chapter where the focus is on the interplay of lexical items and context.

Chapter 3 Context and salience

The definition of SBUs says that they are tied to a particular context. It is important to explore the nature of this tie: how do SBUs relate to context? Grammars are said to be lexically-driven (e.g. Chomsky 1997; Clahsen, 1992; Klein, 1995; Weisenborn, GoodluckandRoeper, 1992). Are contexts also lexically-driven? This seems to be a relevant question in the lexical revolution. I argue that the answer to this question should be "yes", contexts are also lexically driven. This assumption can be explained by answering another question: What is the relation of a lexical unit and context in shaping meaning? Violi (2000) argued that 'context' is one of the most controversial terms in linguistic analysis. This controversy is created by two, basically unanswered questions: 1) how to define the border and limits of context?, and 2) how to define the relationship between lexical meanings and contextual components, which may be both textual (linguistic) and extratextual (extralinguistic)? Here we will discuss only the second question because that has direct relevance to the use of SBUs and that is what the DMM attempts to explain. In the previous chapters it was discussed how coresense relates to WSPs and CSCPs and consenses. Here the model's contextual components will be highlighted. The issue of the relationship of linguistic units and context is the central question of discourse analysis research. The traditional approach to discourse analysis can be described as the exploration of how language is used beyond sentence level, how meanings are constructed between sentences and how lexico-grammatical forms come to take on particular meanings in different contexts (c.f. Brown 1980; Pennycook 1994). The opposite position is articulated by Ball (1990: 3) who argued that " the issue in discourse analysis is why, at a given time, out of all possible things that could be said, only certain things were said." There is a clear shift here from description to explanation.

56

Chapter 3

As Pennycook (1994:116) said ".. The focus is not so much on how meanings are constructed between sentences, but rather on how meanings come to be articulated at particular moments, ...and there is an emphasis not so much on how language works once it has been uttered, but rather on how utterances come to be made, and how those choices are both produced and constrained." Luke, McHoul and Mey (1990: 40) expressed a similar idea: "Paradoxical as it may appear to the structural linguist, we can show that discourse is not something that language does. Discourse is not a mere function of language. Rather discourse is, to put it crudely, the condition by which language as a structure or a system exists." This approach looks at how meanings are a product of social, cultural and political relationships and then turns to see how these may be realized in language. Consequently, the sequence may appear as follows: Meaning (required to be expressed as a result of the interplay ofsociopoliticalfactors and individual needs) > Lexical units (needed to express the required meaning) > Construction of a context in which the required meaning can be expressed by the selected lexical units. In the first two chapters, I emphasized that when language is used we both create context and situations (make things meaningful in certain ways and not others, [c.f. Ball 1990; Gee 1999]) and fit, adjust and adapt our language to these ongoing contexts and situations. These contexts and situations are regularly recreated in a relatively similar way, and may be institutionalized and/or routinized, and as a consequence they can stay in existence for a shorter or longer time in history because of the interactional needs that created them. Gee (1999) argued that in child language acquisition this process can be demonstrated very well because children learn how to fit their language to the context and situations primarily created by others in their socio-cultural groups. At the same time children learn how certain forms of language can create and transform such context and situations.

1. Externalist and internalist perspective on context The revision of the role of lexicon in language production and new approaches to discourse analysis have resulted in a change in our under-

Context and salience

57

standing of the relationship between lexical meaning and context which has been characterized by an underlying distinction between contextfree meanings and context-bound meanings. As discussed in chapter two, the relation between coresense and consense presupposes that meanings are never completely context-free but are usually indexed to some standard context of reference. Coresenses are abstractions of the most salient, familiar, regular and typical consenses. When a native speaker is asked to identify the meaning of a lexical unit out of context the person will remember the meaning on the basis of his/her previous contextualized experiences with the lexical unit. So if a native speaker is asked what the expression "take care" means, and s/he will say that it reminds her/him of some responsibility, this means that for her/him this lexical unit creates a context where s/he has to think of some responsibility to take care of. The point is that lexical units are generally context-bound: when we hear them they make us recall a particular context. Some of them may trigger several con-texts while others such as SBUs usually (not always) initiate one particular context. Consequently, when analyzing context we must ask two questions: 1) How may context contribute to change in the lexical meaning of a word or expression? (Context as selector of lexical features) 2) What context can a word or expression create when uttered? {Lexical units as creators of context) These questions reflect on two sides of one and the same phenomenon. In our conceptual-pragmatic paradigm and the DMM the lexical unit—context relation is a two-way street that represents the plasticity of language and the unique dynamism of language production: lexical units used by speakers create contexts (internalist perspective) and contexts specify meanings of lexical units (externalist perspective): Lexical Units

> Context

> Lexical Units

58

Chapter 3

1.1. Externalist perspective In the traditional externalist perspective10, both extralinguistic and linguistic context is considered to be something that in one way or other modifies word meanings. Context

> Lexical Unit (word, expression)

Context is seen as a selector of lexical features because it activates some of these features while leaving others in the background (c.f. Bibok and Németh 2001). For instance: (25) Please hear me out. (26) I did not hear anything at night because I was fast asleep. (27) Father to his son: "Don't do this again." Son: "OK, I hear you." Each of these utterances focuses on a different aspect of the coresense of the word "HEAR ": (25) listen to attentively, (26) perceive, (27) understand. The phrasal context of the utterance thus foregrounds some specific components of the semantic representation while leavi n g others in the background. This is what makes metaphorical use possible in (27). Context pushes some aspects of the coresense of the word into the background, and selects the perception feature only. The process is similar to what happens to the word "SEE" when we say " I see what you mean." It is important to note that such selection is possible only because there is some sort of pre-existing standard representation that is called "coresense" in the DMM. Standard representation, regularity and semantic systematicity, however, have been un-der fire in cognitive psychology and linguistics as was pointed out in the previous chapters. James (1890) suggested that we should not use the expression "concept" altogether because ".. .For an identical sensation to recur it would have to occur the second time in an unmodified brain. But as this, strictly speaking, is a physiological impossibility, so is an unmodified feeling an impossibility (William James 1890: 232-3)." Smith and Samuelson argued that "... a successful theory of categories thus might require that we give up timeless abstractions such as concepts (Smith and Samuelson 1997: 190)."

Context and salience

59

Others suggested that we keep the idea of concept, just alter its meaning. Barsalau (1993) proposed that the word 'concept' be used to refer not to internal representations but to the thought made at the moment. This idea is strongly supported by cognitive linguists who equate semantic representation with conceptual representation and suggest that individual acts of categorization do not require an already represented concept. Children's word learning demonstrates that they create categories online out of their history of experience, the transient effects of preceding activity, and the details of the moment (Smith and Samuelson 1997:188). Categories created online will be a unique mix of past and present. This approach suggests that the processes that make stability and variability may be the same. A strong version of externalist contextualism takes this line to the extreme and claims that meanings are specified entirely by their contexts, and that there is no semantic systematicity underlying them at all (c.f. Barsalau 1993; Gee 1999). Violi argued that " in this case, context does not merely have a modeling function on semantic selection, but becomes the one and only principle that regulates semantic organization. Meanings are, or at least can be, redefined at the moment of realization of each new utterance, and there is no such thing as semantic systematicity (Violi 2000:115)." However, it would b e a disaster for communication if this were indeed the case. People could not understand each other at all because contexts change all the time since they are user-oriented. What makes communication both predictable and unpredictable is that the lexical unit - context relationship is a two-way street: meanings are at one and the same time neither completely defined nor completely undefined. Bibok and Németh (2001b) following Bierwisch (1996; 1997) claimed that the meaning of lexical units is always underdetermined before they are put into context, and context specifies their meaning. In my opinion both the lexical units (representing former knowledge) and the context (representing both former and actual knowledge) have to do their own share in order for the utterance to be meaningful and appropriate in a given situation. This requires us to distinguish three degrees of determination of the meaning of lexical units. The first is when the lexical unit stands without context (which is hypothetical) and has a meaning potential only. The next degree of determination is when lexical categories participate in the shaping of preverbal thought. The third is when the thought appears in the form of an utterance, and the meaning of lexical items in the utterance becomes specified. The

60

Chapter 3

context-creating role of lexical units that are to be discussed below is made possible by these three degrees of determination.

1.2. Internalist perspective The internalist perspective on context considers lexical units as creators of context: Lexical Unit

> Context

In everyday life what we call "context" generally presents itself to us in highly structured and regular ways. Violi (2000) argued that our experience is developed through a regularity of recurrent and similar situations that we tend to identify with given contexts. Standard context can be defined as a regular situation which we have repeated experience with, and based on which we have our expectations about what will or will not happen, and on which we rely to understand and predict how the world around us works. Violi claimed that it is exactly these standard contexts that linguistic meanings allude. The relative stability of coresense is based on the underlying regularity of their contexts of reference. However, it is important to underline that this relative stability is brought about by events which are only relatively similar. No two events can exactly be the same (c.f. James 1890). When we talk about repeated events what we really refer to is the similarity of the frame (c.f. Fillmore 1976; Fillmore and Atkins 1992). For instance, 'reserving seats in a restaurant', 'greeting neighbors', 'ordering a pizza', 'renewing car insurance', etc. are scenarios in our life that are repeated with a certain regularity but none of these events are exactly the same because both the linguistic and extralinguistic context as well as the lexicalization of thoughts vary all the time. According to the relativity theory of meaning, (c.f. Barsalau 1992, 1993; Smith and Samuelson 1997; Gee 1999) situated meanings are tied to patterns and patterns are explained by "cultural models" which are rooted in the practices of socio-cultural groups to which speakers belong to (D'Andrade 1995; D'Andrade and Strauss 1992; Strauss and Quinn 1997; Gee 1999). Gee pointed out that "...bits and pieces of cultural models are in people's heads (different bits and pieces for dif-

Context and salience

61

ferent people), while other bits and pieces reside in the practices and settings of cultural groups and, thus, need not take up residence inside heads at all (Gee 1999:43)." So words and SBUs have not only situated meanings but also are associated with cultural models which is considered to be a fully or partially unconscious explanatory theory or "storyline" that is connected to a word. These cultural models explain why the given word has the situated meanings it has for a specific social and cultural group of people. Smith and Samuleson argued that objects can not map stably into a set of taxonomy, because objects of mental life are not themselves stable entities (Smith and Samuleson 1997: 173). In a similar manner, lexical items can not fit stably into one and the same frame because the frame itself is not a stable entity. This relative instability, however, does not mean the complete lack of semantic typicality. The DMM holds that regularity characterizes standard context which leads to semantic typicality. There seems to be no such thing as "context-free" meaning because the coresense of a lexical unit is an abstraction of consenses which occur in relatively similar situations. Consequently, it is not necessarily the context that constrains the condition of applicability for lexical insertion. The lexical units being inferential devices can create their own conditions of applicability. Supporting this view Violi said: "...it is not the existence of a given context that makes the use of the word possible, but the use of the word that initiates a mental process in the listener which seeks to construct a context in which its present use would be most appropriate (Violi 2000:117)." Context is characterized both by regularity and variability. The behavior of context justifies the relative independence of semantic level because the communicative (both oral and written) process always relates back to the system of language which is the necessary basis for all contextual realizations. The possibility of change and development presupposes a kind of canonical form, rather than denying it. A lexical unit can function as a trigger for possible new interpretations because it already has an underlying schema, a store of possible interpretations which represent the typical context to which the lexical unit is indexed. Words operate as devices which initiate the construal of some possible context for their interpretation. This device is usually so powerful that the appropriate conditions for interpretation are created even if initially they are neither in the linguistic nor the extralinguistic context. This is why it is possible to interpret a word introduced into a discourse even when all the semantic features of the word are po-

62

Chapter 3

tentially canceled by the perceived context. For instance, the coresense of the word "CAKE" is defined in the following way: "A sweet baked mixture of flour, liquid, eggs and other ingredients in loaf or rounded layer form." (Webster, p. 187.) Having these features in mind one might be quite surprised when reading a sentence like this: "The stoneshaped cake they had in the middle of the night was salty." In this sentence the linguistic context cancels some of the most important features of the word "CAKE", still we are able to interpret it although we may find that particular type of cake peculiar. Cases when some (or even all) of the most important semantic features of a lexical unit are canceled by the context are quite common in communication. One can look at these cases from two different perspectives. On the one hand it can be denied that a lexical unit has any stable meaning, i.e., there is no regularity in the semantic structure, and everything is in the context; meaning is created on line (c.f. Barsalau 1993). From another perspective, however, we can say that lexical units are attached to some kind of standard context of reference that represents the usual (normal, regular, most familiar) semantic potential of the word. Only because of these standardized links between a word and context, can lexical units function as inferential devices. Violi argued that "words force interpretations upon us and compel us to develop hypotheses about what a possible compatible context may be. Words, then, can be said to impose their own context on the situation at hand, rather than saying that their meanings are merely affected by the context in which they are uttered (Violi 2000:119)." At this point, I do not necessarily agree fully with Violi. As I explained earlier the relationship of lexical unit and context demonstrates the dynamism of language use which occurs both diachronically and synchronically. The meaning of a lexical unit can create context and be created by it. This fact ensures the dynamism of communication from the interlocutors' perspective. Based on previous experiences lexical units have built-in contexts which are activated when the given lexical item is used. The activated built-in context has to be matched to the extralinguistic context of the situation. If there is no match between the two, it may lead to miscommunication. The following example illustrates this point well. In class one of my students started to cough. A non-native speaker student said to him: "Bless you". Everybody started to laugh. When the SBU was uttered it created a context which did not match the actual context. People say "Bless you" when someone sneezes and not when someone coughs. Speakers use lexical

Context and salience

63

units thinking that these units will create the right context for their thoughts when they are properly associated with other lexical units according to the rules of grammar. This is where we can say that lexical units create context. However, the created utterance is expected to match the extralinguistic context.

2. Salience Geeraerts (2000) argued that systematic attention to salience phenomena is one of the major innovations of the cognitive approach. Prototype theory showed "how the various semantic applications that exist within the boundaries of one particular category need not have the same structural weight within that category (Geeraerts 2000:79-80)". Salience is where structure and use meet. Salience is ignored in traditional lexical semantics which focuses on the description of various meanings of lexical items and structural relations among those meanings. This kind of analysis gives a structured list of possibilities that language users may choose from in actual communication. In the meantime the actual choices of the language users and the study of communication processes leading to those choices are completely ignored. The dynamism and multi-dimensionality of meaning require us to study salience, which is the result of natural language use and unites the conceptual and lexical levels. The prototype theory and other cognitive approaches, however, include aspects of use in the description of lexical categories. It is emphasized that some senses are more salient than others because they are more readily chosen when using that category (c.f. Geeraerts 2000). As we will see later frequency is an important factor in salience but not the only one. Salience is essential in the analysis of SBUs because salience is usually seen as the structural reflection of pragmatic phenomena. This combination of semantic and pragmatic perspective requires a slightly different understanding of the notion of 'structure' than is customary. Geeraerts argued that "whereas linguistic structures would traditionally be seen merely as ordered sets of possibilities, adding pragmatic-based salience implies introducing probabilities rather than just possibilities (Geeraerts 2000:80)."

64

Chapter 3

2.1. Prior knowledge In the previous chapters it was claimed that the linguistic system cannot be equated fully with the conceptual system because each has a certain kind of independence. Words categorize experience and things in a particular way which are "carved" in the mind, and remain there relatively unchanged until external (extralinguistic) or internal (linguistic) reasons require a change. This experience-framing and defining role of lexical units results in a processing strategy that usually relies on former knowledge. Consequently, new information is always processed through existing information: incoming experience filters through the "results" of previous experience. All theories acknowledge the decisive role of prior knowledge in language production and processing. The debate is usually about the content and organization of this former knowledge. However, no matter what approach we follow there is a problem with prior knowledge: language generally FOLLOWS the changes in society and rarely PRECEEDS them. So what was "carved", encoded in the mind in a particular way at a particular time is difficult to erase or change even if the socio-cultural context has already changed, and requires the appropriate cultural models to be changed accordingly. Let us take a very old riddle as an illustration to this point (Giora 2002). "A young man and his father had a severe car accident. The father died and the young man was rushed to the hospital. The surgeon at the emergency room refused to operate on him, saying: "I can't. He is my son". How is this possible? Giora (2002) presented this riddle in a course whose orientation was feminist. I found it in a graduate class on multilingual development where most of the students were female. The riddle worked in both cases. In spite of much contextual evidence to the contrary (e.g., mostly female students in class; related classroom topic; the explicit context which eliminates the possibility that the surgeon is the young man's father) what came to the students' mind first was the contextually incorrect feature of the word 'surgeon', i.e., "male". For most people, no matter what the context implies 'surgeon' refers to a male rather than a female although gone are the days when this was the case

Context and salience

65

in real life. So language, as we saw earlier and see it now, does not always reflect real life because certain elements (words, expressions) of language are very slow to change and they reflect social facts that are obsolete and were true some time ago. Nothing can change this developmental characteristic of language; neither the purist approach of the French Academy nor the PC ("politically correct") campaign in the U.S. although each may have some modest success.11 The riddle refers to some unique features of language processing. Why is our mind so rigid with respect to parsing? I think because our language "trained" it that way. Yes, language trains our mind to a certain extent: at least that part of the mind which is dominated by representations of knowledge through concepts based in language.12 Parsing depends on the kind of "training", i.e., how certain experience was conceptualized when filtered through language. The mind will look for what it is used to: patterns, cultural models, and the like. If the mind is accustomed to the fact that the word 'surgeon' usually denotes 'male physician', it will look for this meaning. If the most frequent and familiar meaning of the lexical unit 'hold up' is 'rob', the mind will opt for this meaning first. If, however, the mind does not find this meaning it will not be confused, but look for other cues such as context. When, however, the training (i.e. linguistic categorization affecting conceptual categorization) is changed, the mind is also ready to change. But it is very difficult to give up what one is accustomed to. This is true for the mind as well, especially if the symbol (lexical unit) does not change, but means something different than usual, or a new item occurs to mark something that previously did not exist. Does this mean that our mind is completely trapped by language? Not really. As we saw earlier the thought and word relation is a dynamic two-way street. Construals in the mind are culture- rather than language- specific and change as required by the alternating socio-cultural environment. Giora (2002) argued that the way the riddle above is processed demonstrates our inability to instantly adjust to contextual information. ( I would add to this that the riddle also demonstrates that a word has a context-creating power. If it creates an ambiguous context interlocutors will have some problem figuring it out.) Context kicks in only after the most frequent, stereotypical, familiar meaning of the word or expression has been accessed. Available information is accessed initially, regardless of contextual frame or speaker's intent. Why is this so? Why does our mind display some rigidity with regard

66

Chapter 3

to parsing? What role does language play for our mind to function like this? How do compositionality of language and context interact? Can we ignore compositionality and focus only on the frame? Is it really always the literal meaning we process first? These are the main questions that will be addressed in the following sections.

2.2. Interactive view of lexical processing In recent linguistic research there is a strong tendency to support a view according to which in utterance processing it is not the context that has primary importance but the lexical representation embedded in the lexical unit. This approach is in accordance with the two-level semantic (lexical semantics vs. conceptual semantics) assumption and the DMM that were presented in the previous chapters. As a result of the latest psycholinguistic research a hierarchy of utterance interpretation has emerged which starts with the lexical meaning of the word or pragmatic unit. If it does not lead to the relevant interpretation, then the immediate context comes into play. If this does not result in a pragmatically appropriate interpretation either, the context ought to be extended (c.f. Bibok and Németh 2001). So the hierarchy of interpretation looks like this: Lexical interpretation

> Immediate context

> Extended context

Lexical interpretation relies on the coresenses of the word or expression with or without WSPs and CSCPs while immediate and/or extended context shapes the consenses based on the conceptual and encyclopedic information. Several researchers have come to support this hierarchy of utterance interpretation although they took different paths. While Giora (1997; 2002) called this interpretation the "Graded Salience Hypothesis", Bibok and Nemeth took a different route via the principle of relevance which states that "human cognition trends to be geared to the maximization of relevance (Sperber and Wilson 1995: 260)". This means that in communication we want to achieve more cognitive effects with less processing effort. Consequently, processing should start with the most obvious and most salient meaning. Examining the contextual behavior of ambiguous words Gibbs (1996: 33) came to the conclusion that people momentarily activate all the meanings of an ambiguous word, with context then working to clarify

Context and salience

67

its meaning. So context becomes operative only at a post-access stage guiding the selection of the contextually relevant meaning of the ambiguous word. Gibbs and other cognitive linguists (c.f. Kövecses) may not be right when they say that ALL meanings of an ambiguous word are activated at a time. The question is that which of the possible aspects of meaning Jacks infirst. Concretely, for instance, out of the five possible readings of 'HOLD UP ', why exactly does the 'to rob " interpretation kick in first? So how do we process information? What happens while parsing? Why does nobody misunderstand the following conversation?: (28) Jim: - ' Joe, can you fix my car?' Joe: - ' Piece of cake. ' If 'piece of cake' processed literally, it does not fit into the context. We, however, do not look at the context first, rather we opt for the more accessible information. This situation also suggests that context affects comprehension after highly salient information has been accessed. If it is compatible with the context, no further effort is needed as the principle of relevance requires. If, however, there is no match the search for possible alternatives continues. It is not rare that words or certain expressions have more than one very frequent, stereotypical, familiar and equally conventionalized aspects of coresense as in the following situation (Maley 1980:10) where the expression may have two almost equally salient readings:

(29) Sam: - 'Coming for a drink?' Andy: - 'Sorry, I can't. My doctor won't let me.' Sam: - ' What's wrong with you? ' In this situation the SBU "what's wrong with you?" is an inquiry about the health condition of Andy. Thisconsense is triggered by the

68

Chapter 3

contextual cue 'my doctor' which precedes the SBU. If, however, we change this contextual cue for 'my mother-in-law', the meaning of "what's wrong with you?" will change with it in the following way: (30) Sam: - 'Coming for a drink?' Andy: - 'Sorry, I can't. My mother-in-law won't let me.' Sam: - ' What's wrong with you? ' Here, the pre-text activates the appropriate consense of 'what's wrong with you?'. Prior context can affect the accessibility of the meaning of a word or an expression. A prior occurrence of a word ("mother-in-law") which is semantically or pragmatically (as above) related to a following word or expression ("what's wrong with you?") may ease the processing of that word: the selection of the context-appropriate consense out of the more than one possible salient consenses. So what is salience all about? What is the role of salience in language comprehension and production? How is it possible that in many cases contextual information does not affect initial access? What makes certain meanings so powerful that even a highly biased context is ineffective in blocking them?

3. Literal meaning versus figurative meaning Recent literature on figurative language has come up with new facts about language processing. Cognitive linguistics has directed attention to the differences and similarities of processing literal and figurative language. Since SBUs are pragmatic units whose literal meaning is often canceled for the sake of a special communicative function it is relevant to discuss the use of these units within the framework of contemporary theories of understanding figurative and literal language (e.g. Gibbs, 1984, 1994; Keysar, 1989; Blasko and Connine, 1993; Giora, 1997, 2002). According to the traditional approach, metaphors and other figures of speech are the property of language and problem for language because they are expressed in language and not in some other medium (Katzetal. 1998). Lakoff(1993) and others, however, maintained that

Context and salience

69

metaphors are a property of our conceptual system rather than a property of language. There is little doubt about the fact that novel metaphors are frequently novel only as a linguistic expression but not as a deeper conceptual relation (Katz et al. 1998). After studying the point at which certain metaphors enter a language Lehrer (1978) found that first a basic conceptual relationship is instantiated [e.g. PERSONALITIES are like TEXTURES], then based on this conceptual mechanism novel metaphors are generated. For instance, "he is a smooth talker", or " he is rough around the edges". Metaphors go back to a conceptual relation between their concrete source domain and (usually) abstract target domain. In this respect, they are conceptual rather than linguistic phenomenons. As explained earlier, language and thought do not coincide, and figures of speech are the results of the unique interplay of language and thought. Metaphors are the conventional or ad hoc lexicalizations of conceptual structures which result in WSPs. According to Winner, the recent experimental literature has made it clear that if "people were limited to strictly literal language, communication would be severely curtailed, if not terminated (Winner 1982:253)." The traditional approach to the literal versus figurative dichotomy is that figurative meaning must have a literal basis: figures of speech develop from literal meanings. What do people consider literal? "Literal" is defined in the dictionary in the following way: "In accordance with, conforming to, or upholding the explicit or primary meaning of a word or words of a text (Webster, p. 762.)" Literal is explicit, referential, compositional, but is it primary? What psychological reality does "primary meaning" have in the mind of native speakers? It is questionable whether "primary" always means "literal". In fact, "primary meaning" can be either figurative or literal as discussed above. Let me illustrate this with a story. I was supposed to have a meeting with one of my PhD students who is a native-speaker of English. She was late. When she entered my office, she said 'hello', and gave the following explanation: (31) "- Sorry. I was held up at a gas station. Not literally though." This linguistic behavior of the student raises the question: why did she find it important to add "not literally though", after the use of the

70

Chapter 3

lexical unit 'hold up'? Why did she think that I could misunderstand her ? I confronted her with these questions. She said that she did not want me to misunderstand her and think that there was actually a 'holdup' at the gas station. The student actually thought that the LITERAL meaning of the verb 'hold up" was "rob" which is clearly the FIGURATIVE meaning. According to the dictionary The American Heritage Dictionary of the English Language. (Morris, William. 1976:628), the lexical entry 'HOLD UP" has the following readings: 1) To prevent from falling; to support, 2) To present for exhibit; to show, 3) To last; stand up; endure, 4) To stop or interrupt; delay, 5) To rob. So why was this confusion in the mind of this native speaker of American English? I think what happened was that she equated the most frequent and familiar meaning, i.e., the most salient meaning with the literal meaning. Is this something that is unique for one speaker only, or is it a generalizable fact? I think it is a generalizable fact. The traditional linguistic approach claims that in most people's minds it is the literal meaning from which all other meanings derive. The belief about the primacy of the literal meaning is so strongly conventionalized that not everybody notices when there is a shift in the semantic structure of a word, and what was the most familiar, most frequent and most conventionalized of all possible consenses gives way to another consense which takes over as the primary, most salient but NOT as the literal meaning. The important thing is that it is absolutely irrelevant whether the primary consense is the literal meaning or the figurative meaning of the lexical unit. Primary meaning appears to be the most salient meaning which can be either literal or figurative as claimed above. What is important is that there is a shift in lexical representation which does not necessarily lead to a shift in conceptual representation. This fact also confirms the need to make a difference between lexical semantics and conceptual semantics. The literal-fig-

Context and salience

71

urative dichotomy makes sense for language analysis but not for language processing where it cannot be claimed that literal is always processed before figurative.

4. Characteristics of salience Lexical units, i.e., words and formulas usually have more than one meaning. As seen earlier their meaning structure can be approached from a synchronic and a diachronic perspective. These two perspectives are united in what is called "salient meaning" ( c.f. Geeraerts 1997; Giora 1997, 2002). The most conventional, frequent, familiar, or prototypical interpretation is the most salient meaning of a specific word, expression or utterance. A review of recent research (c.f. Giora 1997,2002) on this issue reveals that figurative and literal language use are governed by a general principle of salience. Salient meaning (which can be either literal or figurative) is processed first because it is conventional, frequent, familiar and usually enhanced by prior context. The graded salience hypothesis of Giora (1997) claims that different linguistic expressions (salient - less salient) may tap different (direct/parallel/ sequential) processes. Direct processing applies when highly salient meanings are intended. When, for instance, the most salient meaning is intended ( e.g., the figurative meaning of conventional idioms, or the special communicative function of SBUs), it is accessed directly, without having to process the less salient (literal) meaning first (Gibbs, 1980). For instance: (32) Store manager: - Jim, let me introduce you to Sarah Brown, our new sales associate. Jim: - Welcome aboard, Sarah. However, when a less rather than a more salient meaning is intended (e.g. the metaphorical meaning of novel metaphors, the literal meaning of conventional idioms, a novel interpretation of a highly conventional literal expression, or the literal meaning of a conventional SBU) comprehension seems to involve a sequential process, upon which the more salient is processed initially, before the intended

72

Chapter 3

meaning is derived (Blasko and Connine, 1993; Gibbs, 1980; Gregory and Mergler, 1990). Parallel processing is induced when more than one meaning is salient. Conventional metaphors whose metaphoric and literal meanings are equally salient are processed initially both literally and metaphorically (Blasko and Connine, 1993). For instance: (33) Major Willis: - Good morning, Captain Brown. Captain Brown: - Good morning, Major Willis. Welcome aboard. Sequential processing applies when language is used innovatively, as in the case of literal uses of highly conventional idioms. The more salient meaning is processed initially, before the intended meaning is derived. The graded salience hypothesis demonstrates that the literal/figurative divide cannot account for ease of processing, and the salient/non-salient continuum is suggested instead (Giora, 1997). The hypothesis requires that the standard pragmatic model be revised: Instead of postulating the priority of literal meaning, the priority of salient (e.g., conventional, familiar, frequent) meaning should be postulated (Giora 1997; 2002). Salient interpretation has unconditional priority over less salient interpretation. The most salient meaning of a word or utterance is always activated, and is always activated initially. The more salient the meaning is, the more difficult it is to reject as the intended meaning. In the case of conventional metaphors or expressions, whose figurative and literal meanings are as salient, both are processed initially. Salience has some characteristic features that are important for its proper understanding:

4.1. Salience in the DMM Salience demonstrates best how the dynamic relation of the lexical item and context works in practice. Salience is a category that is based on the contextual use of words and expressions. It establishes a kind of a hierarchy among possible interpretations. The more standard the contextual frame associated with the lexical item is the more salient the

Context and salience

Ti

given sense is. A change in salience, sooner or later, will result in a change in coresense.

4.2. Not the same for each member of a language community Salience is not something that is exactly the same for each member of a language community, but it certainly is the same for the majority. Some native speakers of English may argue that when they ask someone "how are you doing?", it is not a semantically empty expression for them because they mean what they ask. Moreover, one and the same person can use this expression in its literal or figurative sense depending on the situation or individual communicative need. This may be true in the case of certain people and certain situations but it can hardly be denied that this expression usually functions like a greeting or a conversation-opener, or just as a sign that the speaker noticed the recipient.

4.3. Dynamic phenomenon Salience is a dynamic phenomenon that is under constant change in language use, and relies on stored, prior knowledge. The transitional nature of salience may result in different interpretations for different generations. As usual the younger generation may have already picked up the relatively new salient meaning of an expression while the older generation still sticks to what they have been accustomed to. This can be illustrated by a situation in which I was one of the participants. In 1997 I was in a mall in Phoenix, Arizona and wanted to drink some good coffee. I saw an elderly lady using an Italian coffee maker at a bar counter to make espresso coffee. That was exactly what I needed, so I went up to her and asked: ' Can I get a double espresso?'. Her answer was: 'You certainly may.'' She taught me a lesson. For me the use of 'can' was a frequent, familiar and reasonable way to ask for anything including coffee. Not necessarily for her generation. That old lady was a "purist' in one way. She stuck to what was meant by "correct language" for her generation which did not necessarily accept the obvious change in the use of 'can' during the years. Another example from Russian can also illustrate this point. In

74

Chapter 3

Russian only the older generation uses the following friendly SBU upon acquaintance. You will rarely hear the underlined expression from the younger generation. (34) Uo3HOKOMbmecb, nootccuiyucma. 3mo Harn Hoebiü compydnuK Anna

AneKcandpoena, o Komopoü Η βοΜ Xoeopun. IJpoiuy Jiwöumb u otcanoeamb ! (Akisina and Formanovskaja 1982:46). 'Please get acquainted. This is our new colleague, Alia Aleksandrovna whom I have told you about. Please be nice to her.'

4.4. Salience is graded Salience is graded: a word, SBU or utterance may have more and less salient meanings. For instance, the word 'fix' may have the following meanings in the order as given in (Webster, p. 497): 1) 2) 3) 4) 5) 6) 7) 8)

to place or fasten securely to put into a stable or unalterable form to direct steadily, concentrate to establish definitely, specify to ascribe, allot to rectify, adjust to set right, repair to make ready, prepare

The internal hierarchical structure of the mental lexicon is not fixed but dynamic. However, lexicographers hardly follow the dynamism of the lexicon. What we usually find in the dictionary is a hierarchy of meaning which may have been correct several years or even decades ago. When 28 native speakers of American English were asked to write one sentence with each word that they heard from a list presented to them twice by the teacher, the word 'fix' generated one of the most interesting results. Meaning number 7 ("to set right, repair") emerged as most salient (32%), followed by number 8 (28%), and number 1 (19%). This grading hardly allows us declare that number 7 is the

Context and salience

75

most salient meaning. What must be said is that the word 'fix' usually has two salient meanings (number 7 and number 8). This was not the case, however, with the word 'bank', where the meaning "financial institution" clearly emerged as the most salient meaning (78%). Language changes result in a "reshuffling of salience" (Giora 2002). Giora argued that for computer people the salient meanings of the words 'desktop', 'windows', 'folder', 'paste', 'inbox', 'mail', etc. are metaphorical rather than literal. The literal meaning of these words which was once salient, has lost its priority.

4.5. Salient meaning depends on conventions Nunber, Sag and Wasow (1994:492) view conventionality as "a relation among a linguistic regularity, a situation of use, and a population that has implicitly agreed to conform to that regularity in that situation out of preference for general uniformity, rather than because there is some obvious and compelling reason to conform to that regularity instead of some other". The first part of this definition is acceptable. However, I cannot fully agree with the arbitrariness of conventionality. Whether conventionality is motivated or arbitrary is a particularly important issue for SBU research. The cognitive-pragmatic approach predicts that SBUs are motivated and not arbitrary. There is a cognitive reason why this and not another expression denotes a situational event in this or that culture, or why one culture finds it important to use an expression in a given situation while the other ignores that. For instance, "Please help yourself' is often used by Americans at the table to urge their guests to start to eat or take some more food. The original cognitive mechanism responsible for the situational meaning of the expression could be described as follows: Take as much as you wish. I don't want to help you because you know how much you need —> so help yourself to as much as you need. This culture-specific pragmatic property of the SBU is no longer maintained consciously. So no inferential reasoning is necessary to find out that the speaker asks you to "help yourself' not because he does not want to help you but because he thinks that you yourself know exactly how much you want to eat. The linguistic form has acquired a pragmatically motivated sense which became conventionalized.

76

Chapter 3

Conventionalization is a culture-specific phenomenon. What functions are conventionalized and how these functions are lexicalized is culture-dependent. This fact makes the interpretation of conversational routines very difficult for non-native speakers who usually rely on compositional meanings without the appropriate knowledge of culturespecific conceptual properties of the expression. Coulmas (1979:8) gave the example of a Japanese scholar who described his difficulties in adjusting to the American culture in the following way: "The 'please help yourself that Americans use so often had a rather unpleasant ringing in my ears before I became used to English conversation. The meaning, of course, is simply 'please take what you want without hesitation', but literally translated it has somehow a flavor of'nobody else will help you', and I could not see how it came to be an expression of good will (Doi, 1973:13)."

Salience is shaped by language use and relies on stored, prior information. Professor Doi and other non-native speakers may always find this expression rude because they lack in their conceptual base the conceptual mechanism and cultural information (CSCP) that is responsible for this SBU in English. If a non-native speaker only picks up the function of an SBU without understanding the socio-cultural background that the expression is rooted in, processing and use of that SBU will always be troublesome for the non-native speaker. This issue will be discussed in chapters 7 and 8.

4.6. Frequency of use Frequency of use has a major impact on salience. The word "drug" has two frequent meanings: 1) "medicine", and 2) "a narcotic, especially one that is addictive" (Webster, p.401). When 28 native speakers of English were asked to give the meaning of the word "drug" without context, meaning # 2 came in as most salient with 64% as opposed to meaning # 1) with 33%. Frequent use may bring about changes in salience: it can neutralize the stigma that was attached to the word. In this case it can be argued that the most

Context and salience

77

salient meaning of the word 'cool' is not "moderately cold" any more, but "excellent, superior" (Webster, p. 292) which was originally slang.

4.7.

Familiarity

All the factors described above are correlated by familiarity according to Giora (2002). She argued that familiarity seems to be the most crucial factor because a meaning that is frequent, conventional and prototypical is likely to be more familiar to a speaker/hearer than an alternative meaning which is low on these variables. Frequency and familiarity correlate in a unique way. Frequency can be general or attached to a particular register or situation. We say that the utterance " Hello, how are you?" is very frequent. That is true because the situation (meeting and greeting others) requiring the use of this (or a similar) expression occurs very often. I think there is a difference between word frequency (WF) and utterance frequency (UF). WF refers to the general use of words in any kind of situation. UF, however, is more register- and/or situation-bound. It is especially true if we take SBUs. It makes little sense to speak about the general frequency of utterances when they are usually register- and/or situation-bound. Consequently, the frequency of an SBU depends on the frequency of the given register or situation to which it is attached. Familiarity has an important impact on L2 acquisition which has not been investigated yet. Based on my own experience with thousands of L2 learners I speculate that L2 learners show more affinity to unique and motivated expressions in the new language than to more frequent expressions. When the LI is acquired the arbitrariness of lexical units dominates. However, in the process of L2 acquisition the learner already has a vocabulary to which s/he can compare the new, emerging vocabulary. Uniqueness, similarity and motivation are likely to be more significant for the acquisition process than frequency of occurrence. Mostly anecdotal examples can support this speculation with the exception of Wiley and Rayner's study (1998) which focused on native-speakers and claimed that a meaning maybe infrequent but familiar to the individual. L2 learners easily retain expressions that they like, and are motivated to learn. They have a selective attitude to learning new words and expressions. LI learners usually do not and cannot control how their affective filter works while acquiring their mother tongue: the acquisition process is usually subconscious, in-

78

Chapter 3

stinctive and automatic (c.f. Vygotsky 1962). This is not the case with L2 learners. Familiarity in L2 seems to depend more on learnermotivated and directed factors than in LI. Learning new words and expressions in an L2 is a part of pragmatic rather than grammatical skills (Kecskes 1998b). This issue will be discussed in chapter 9.

4.8. Reflexivity How does salient meaning relate to the internal interpretation of context and the cognitive theory of cultural meaning (D'Andrade 1997; Strauss and Quinn 1997; Gee 1999)? There is one thing for sure: these ideas are not antagonistic, in fact they seem to be supportive of each other. Words can create their own contexts because they represent cultural models that have psychological reality for the socio-cultural group which considers them as reflections of reality. This derives from the unique reciprocity between language and reality. Language does two things simultaneously: it reflects reality (the way things are), and constructs reality to be a certain way. Gee (1999:82) argued that 'reciprocity' is a good term to describe this property of language but 'reflexivity' is a more commonly used term because language and context are like two mirrors which face each other and "....constantly and endlessly reflect their own images back and forth between each other." In the next chapters we will focus on how the relationship of language and reality is reflected in the development and use of SBUs.

Chapter 4 The origin of situation-bound utterances

1. Linguistic solution to non-linguistic problems As far as their etymology is concerned, SBUs are freely generated phrases which have become delexicalized to a particular extent during their frequent use in certain predictable social situations. SBUs are reflections of expectations of the socio-cultural system. Speakers use these tools to relate to others in an acceptable way. Fónagy (1956, 1961) and Coulmas (1981a) argued that excessive currency usually corrupts expressiveness and diminishes meaningfulness. SBUs may lose their semantic transparency (or at least a part of it) because they are used to express certain pragmatic functions attached to particular situations only. They can become formulae for specific social interactions and thus decrease the processing overload. There are several reasons why SBUs are widely used in conversation:

1.1. "Speech economy (Hymes 1972)" Wray (1998, 2002) argued that the fact that formulaic expressions seem to appear and disappear suggests that they are not linguistic in origin but are a linguistic solution to non-linguistic problems. She identified two such problems: the need to achieve social interactional goals and the need to produce (and understand) language without causing a processing overload. This is in line with the principle of relevance which states that "human cognition trends to be geared to the maximization of relevance (Sperber and Wilson 1995:260)", and Sinclair's idiom principle, which says that the use of prefabricated chunks "..may .. illustrate a natural tendency to economy of effort (Sinclair 1991: 110)." This means that in communication we want to achieve more cognitive effects with less processing effort. The development of SBUs strongly supports the relevance principle and Wray's assumption. Formulaic expressions ease the processing overload not only because they are "ready-made" and do not require the speaker or hearer

80

Chapter 4

any "putting together" but also because their salient meanings are easily accessible in online production and processing. As it was discussed in the previous chapter the graded salience hypothesis demonstrates that the literal/figurative divide cannot account for ease of processing, and the salient/non-salient continuum is suggested instead (Giora, 1997).

1.2. "Phatic communion (Malinowski 1923) " According to Malinowski ( 1923) pre-coded routines serve the function of "phatic communion". They are tools which establish personal bounds among the members of a speech community. They can signify group membership. This function of SBUs is especially important in societies such as Japanese, Arabic or Turkish where ties to the past, tradition, and ancestors are especially important. Tannen and Öztek (1981:39) referred to this function as "rapport establishment". They argued that basically any type of social interaction is an occasion for establishing rapport between speakers. People generally use two strategies to build rapport: put oneself down and build the other up. This is especially important in speech communities such as Chinese, Arabic, Turkish, Thai where relative status is a key factor. The Turkish "estagfurullah " ["I ask pardon of God"] which is a very commonly used rejoiner is a good example. It serves to negate another's expression of self-abasement. If someone should say, for instance, "Oh, I am a fool", the interlocutor's response, "estagfurullah" would mean something like "No, that's not so" (Tietze 1963: 145).

1.3. SBUs reduce the complexity of the situation SBUs also help to reduce the complexity of the situation because they are recognized ways of interaction within a particular speech community so speakers know that they cannot get into trouble by using them in the appropriate situations. SBUs represent standardized ways of organizing interpersonal communication and as such their use ensures an appropriate, and accepted way of language behavior in certain recurring situations. Tannen and Öztek argued that "cultures that have

The origin of SBUs

81

set formulas afford their members the tranquility of knowing that what they say will be interpreted by the addressee in the same way that it is intended, and that, after all, is the ultimate purpose of communication (1981: 46)." Development and use of SBUs in frequently occurring social situations is universal in the sense that these expressions constitute a kind of'etiquette of speech' in languages (c.f. Kasper 1996b; Akisina and Formanovskaja 1982). People greet each other, apologize, make compliments and requests no matter what language they speak. Every language has a stock of conventionalized expressions for fulfilling these functions.

1.4. SBUs act as "fillers" Sorhus (1972:214) pointed out that prepatterned expressions are frequently used to give the speaker time to seek for words to express his/her ideas. Several other researchers (c.f. Coulmas 1981; Bolinger 1976) also argued that routine phrases can be drawn from the memory without particular effort, and their use may give the speaker the chance for planning conversation. We can agree with these statements although it is important to note that SBUs occur much more rarely in this function than other types of prefabricated units.

2. Where are SBUs from? The emergence of SBUs in human languages raises several questions which are expected to be addressed in the next chapters. These questions are interconnected and will be treated accordingly. How are conventional, frequently occurring situations lexicalized in different languages?; What restrictions do different socio-cultural systems impose on this kind of linguistic activity?; Why exactly these expressions and not others are selected or developed to express the given situational functions?; and, why is it that in certain languages these institutionalized units become semantically almost or fully empty while in other languages they keep their compositional meaning or at least a part of it in spite of their frequent use?

82

Chapter 4

2.1. How are conventional recurring situations lexicalized in different languages? Coulmas (1981) argued that conversational routines are tacit agreements that members of a community are supposed to share with every reasonable co-member. The development of these units in each human language is a natural consequence of communication. There is a particular need for them in every culture. Wierzbicka (1992; 1994) claimed that universal human cultural concepts are lexicalized in various forms in different languages. Although these routines coexist in languages, and the functions they express may be universal, the way they are lexicalized differs from one language to another. The Japanese expression "o-genki desu-ka " is usually given as an equivalent to American English " how are you?". However, Americans studying Japanese will soon discover how different these two expressions are both linguistically and pragmatically.13 The forms of expressing politeness or request in one language are usually quite different from those used to express the same or similar function in another language. Differences can be explained by responding to the following three questions: A) What particular cultural phenomena does the given language community find important to lexicalize? B) If a cultural phenomenon is lexicalized in one language, how does this relate to similar phenomena in other languages which also lexicalize the given phenomenon? C) What are the differences in lexicalization? Coulmas (1979) assumed that lexical differentiation in a language has socio-cultural relevance. If there is a high degree of differentiation in a particular lexical field the given domain seems to bear special relevance in the life of the given group of people. For instance, Turkish people find it important to use the SBU "Ag 'zindan yel alsin " ["may the wind take it from your mouth"]14 when someone has mentioned a possible disaster. Some other cultures find it unnecessary to use any SBU in similar situations. In most cultures people use SBUs when a friend or family member is setting off on a trip. The Turkish people say "Yolun acik olsun " ["May your way be open."], while the

The origin of SBUs

83

Hungarians say "Jó utat !" ["Have a good trip"]. So the two speech communities highlight different aspects of a similar situation. The Russian "(ÏKenaio SOM) xopotuero nymu" [I wish you a good trip]15 is closer to the Hungarian and American English expression ("Have a nice trip") than to the Turkish SBU. The core of the cultural concept can be the same but generally there is some knowledge domain attached to this core concept which is unique and characteristic only to the given culture (Kecskes and Papp 2000a). There is, for instance, a significant difference in the use of Japanese 'Sumimasen' and Hungarian 'Bocsánat ', although each has the main function of expressing apology. The choice of a lexical item or items to express the function of apology depends on how apology is conceptualized in that particular linguistic community. The notion of apology is broader in Japanese than in Hungarian or English. In Japanese apologies indicate the speaker's willingness to conform to social expectations and conventional rules. People in Japan dispose apologies for actual, anticipated or even assumed violations of conventional rules (c.f. Coulmas 1979). The use of 'Sumimasen' is the best way to maintain a face-conscious behavior that is one of the central values governing interaction in Japanese. Consequently, to find the right equivalent of this Japanese SBU in another language is a very difficult task because of the lack of similar socio-cultural knowledge and expectations. The equivalency of SBUs in different languages is not the matter of compositional semantics. Explanation must be sought in the sociocultural and conceptual rather than in the linguistic domain. The linguistic equivalent of an SBU in another language may sound awkward. A good example is when Hungarian TV anchors started to use "Legyen szép a napjuk ('May your day be nice')", " Legyen szép az estéjük! ('May your evening be beautiful')" expressions as borrowings (caiques) from their American colleagues who frequently say "Have a nice day", or "Have a nice evening" to their audience. These expressions sound perfectly OK in American culture where people have the necessary conventional knowledge to process them properly. In Hungarian culture, however, the linguistic literal equivalents of these SBUs lack cognitive support in the mind of Hungarian speakers so the expressions sound rather strange and out of place. The lack of supportive conventional knowledge responsible for the development and use of these expressions in other cultures does not mean that people in those cultures are less considerate or polite to their fel-

84

Chapter 4

low human beings. It simply means that there is no conventional rule that governs that particular situation, or if there is, it is entirely different. Another example can be the use of specific SBUs at the beginning of a meal in several different cultures such as "Bon Appetit " in French, "Guten Appetit" in German, or "Jó étvágyat!" in Hungarian, and the lack of an expression with similar function in American culture. This, however, does not mean that Americans are inconsiderate to others when eating. Simply, in American culture, people think that if you sit down to eat you have reason to do so: you have appetite. Japanese culture is considered to be very polite. Still, they do not find any reason to use SBUs similar to "Bon Appetit" in French, "Guten Appetit" in German, or "Jó étvágyat! " in Hungarian when they start eating, but probably for reasons other than those in American culture. Talmy said that "given the socio-cultural context of a speech community, vocabulary gaps are seldom accidental. They most commonly reflect lack of pragmatic need or lack of dynamism within a certain subportion of the socio-cultural universe of the speech community (Talmy 1978:235)."

2.2. Non-nativelike use of SBUs Language learners often make the mistake that they think what is expressed in their language should also be expressed in the target language. "How do you say this in English?", is one of the most frequent questions of learners of English. I think it is appropriate to respond with another question: "Why do you think that this can or needs to be said in English?" Bear suggested that the question that should be asked "is not how to say χ in English, but rather, whether there exists an English utterance y appropriate to the situation in which χ is used in Turkish (Bear 1987:32)." So the first step of inquiry is to figure out whether the target language may, can, or needs to express something similar to that which the speaker has in mind based on his/her own language. I heard one of the most illustrative examples from a French colleague who married an American woman. In American English "I love you" is so frequently used that in fact nobody knows what it really means. Its function changes with the situation. It can be a greeting, a saying "good-bye" to someone, an expression of affection, a "thank you", or something else. However, in French "Je t'aime" the literal

The origin ofSBUs

85

equivalent of " I love you" is used only in its very specific compositional meaning. "Je t'aime " is not an SBU in French while "I love you" in American English can be considered as a unique SBU that has more than one situation where its use is perfectly appropriate. So my colleague's American wife wanting to express her affection towards her father-in-law said to him "Je t'aime ", several times, which, of course, made the Frenchman very confused. This example demonstrates how important it is whether the underlying conceptual knowledge is universal or culture-specific. It certainly is culture-specific in this example. As we have seen, SBUs more than any other linguistic element allow us to have an insight into the thinking of people who belong to a particular language community because through these commonly accepted conventionalized expressions the individual speaker may affirm his/her belonging to a socio-cultural group. That is why it is said that the easiest way to discover whether someone is a native speaker of a language or not is to watch how that person uses SBUs. You can have native-like pronunciation, a rich vocabulary and confident use of linguistic rules but your proper or improper application of SBUs and idioms reveals where you are from, and how much you are a member of a particular linguistic and cultural community. Pawley and Snyder (1983) argued that insufficient knowledge about which expressions are appropriate and idiomatic is one of the clearest ways in which even highly proficient non-native speakers can mark themselves out as non-native. They gave the following examples: (35) is spoken by a Dutch flight attendant, and (36) is a more likely native version in the same situation.

(35) The Captain has just illuminated the seat belt sign, as an indication that landing is imminent. (36) The Captain's put on the seatbelt sign, which means we're about to land. Pawley and Snyder (1983) pointed out that (35) is entirely grammatical, and could be generated by a native speaker. Yet it would tend not to be. This is just not the way a native speaker would express the same idea. (35) is a thought generated in Dutch and shaped into English words according to the grammatical but not the lexical rules of English.

86

Chapter 4

The following short list of signs and notices written in English from hotels all over the world demonstrates how awkward the result can be when a particular knowledge structure from one language is packed into the lexical units of another: (37) In a Paris hotel elevator: "Please leave your values at the front desk." (38) In a Zurich hotel: "Because of the impropriety of entertaining guests of the opposite sex in the bedroom, it is suggested that the lobby be used for this purpose." (39) In a Yugoslavian hotel: "The flattening of underwear with pleasure is the job of chambermaid." (40) In a Bucharest hotel lobby: "The lift is being fixed for the next day. During that time we regret that you will be unbearable." (41) Advertisement for donkey rides in Thailand: "Would you like to ride on your own ass?" (From The Octopus. Oct. 1995. Campaign, IL. P. 144.)

3. Language distance and cultural distance Language distance which plays an important role in the interpretation of SBUs has an objective and a subjective side. The objective side is called language or linguistic typology which focuses on classifying languages according to their structural characteristics. Kecskes and Papp (2000a) argued that language typology distance or closeness usually goes together with cultural closeness or distance but there is no one-to-one relationship between the two because culture is generally significantly influenced by geographical distance. Before continuing discussion on this issue, we need to explain briefly how languages differ from each other typologically. Configurational languages (such as, English or French) have a bound word order which is governed by grammatical rules (c.f. (White, 1989; Kiss 1987). Phrase structure configurations generally encode the grammatical functions, and logical relations can only be computed at

The origin ofSBUs

87

computed at a virtual level of representation. For instance: (42)

Peter likes dogs.

No other permutation of this sentence is allowed because "Likes Peter dogs", "Dogs Peter likes", Likes dogs Peter", etc. are all incorrect. In contrast to configurational languages, non-configurational languages (such as, Russian or Hungarian) have complex morphology and a word order which is governed by pragmatic rather than grammatical rules (c.f. Kiss 1987). For instance, the sentence "Peter likes dogs" can have several equivalents in Hungarian, as follows: (43) Péter szereti a kutyàkat. ' Peter likes the dogs-ACC. 'Peter likes dogs.' (44) A kutyàkat Péter szereti. The dogs-ACC Peter likes. 'It is Peter who likes dogs.' (45) Péter a kutyákat szereti. Peter the dogs-ACC likes. 'It is dogs that Peter likes.' The difference between these sentences is in emphasis: which element of the sentence is emphasized by the speaker. Word order is governed by pragmatic rather than grammatical rules. These permutations are made possible by suffixes and endings which make clear how the words in the sentence are connected grammatically. It is important to note, however, that configurationality is a continuum rather than a dichotomy. There is no language that is only configurational or nonconfigurational. Even English, which is one of the most configurational languages, has some grammatical suffixes to express grammatical relations. (For instance: plural marker: book-books·, suffix of past tense: opened; suffix of progressive: opening) Linguistics typology is related to psychotypology through culture. Typologically Swedish (Indo-European, Germanic, configurational) and Finnish (Finno-Ugric, non-configurational) are farther from each other than Finnish and Estonian (Finno-Ugric, non-configurational).

88

Chapter 4

But culturally Swedish is closer to Finnish than to Estonian. The subjective side of language distance is constituted by the reflection of structural and cultural closeness in the human mind. Kellerman (1983) referred to the perception of language distance as "psychotypology". Language typology builds on objective facts; structural differences and similarities between languages. Psychotypology tells us how multilingual speakers perceive structural and cultural differences and similarities between their languages. My understanding of psychotypology described in Kecskes and Papp (2000a) differs at least on two points from Kellerman's interpretation. Kellerman (1983) suggested that transfer (he means mainly negative transfer) is seriously affected by the learner's perception of language distance. Transfer of a form from the LI to the L2 partly depends on how likely the learner thinks the form is to be acceptable in the other language. I assume that psychotypology plays an important role in multilingual development and the language use of bi- and multilingual speakers but not just as Kellerman proposed. In Kellerman's framework transferability is a relative notion which depends on the perceived distance between the LI and the L2, and the structural organization of the LI (c.f. Gass and Selinker, 1993). But the perceived distance is a two-way rather than a one-way process: it is not just the structural organization of the LI that has a decisive role but also the structural organization of the L2 . Language learners are sensitive not only to structural differences between languages but also to cultural and other differences. Kellermann also acknowledged the importance of these other differences. He referred to Widdowson (1980) who pointed out that differences other than typological, e.g., psychological and social should also be accounted for. However, Kellermann emphasized that those differences were beyond the scope of his study. Kecskes and Papp (2000a) claimed that in the mind of the language learner those differences including structural, social and cultural aspects are inseparable, and they all together have a cumulative effect on the decision making process of the learner. It may well be that assumptions about cultural distance between languages direct the language learner not exactly as it is expected on the basis of language typology. For instance, English native speakers studying Finnish and Turkish will probably feel that Finnish is closer to English than Turkish although both Finnish and Turkish differ from English to a great extent. They both have a very complex morphology unlike English. In spite of the similar linguistic differences, English speakers feel less

The origin ofSBUs

89

distance to Finnish because they think Finnish culture is closer to them than Turkish culture. The question is: how language and cultural distance as a whole with its objective and subjective aspects influences the acquisition process and the outcome of this process. The acquisition process is usually longer and more difficult if the two languages are remote both typologically and culturally. Native speakers of one language frequently believe that they will find certain other languages especially difficult to learn. They are guided by one of the main principles of learning according to which learning difficulty is tied to the degree in which the object of learning resembles something already known. In the previous chapter, it was discussed how important role of prior knowledge is in language production and processing. According to Odlin ( 1989) the cumulative effects of crosslinguistic and cross-cultural similarities and differences in the acquisition process can be assessed by the length of time needed to achieve a high degree of mastery of a language. Little research has been done to prove the validity of this hypothesis. As an example, Odlin (1989) referred to the schedule of courses of the Foreign Service Institute (1985) of the U.S. State Department which summarized the lengths of language courses offered to the members of the U.S. diplomatic corps. Here are some examples from the list: Language Arabic Bulgarian Chinese Czech Finnish Hungarian Japanese Korean Russian Turkish Danish French German Italian Norwegian

Number 44 44 44 44 44 44 44 44 44 44 24 20 20 20 24

90

Chapter 4

Rumanian Spanish Swedish

24 20 24

It can be questioned whether 44 weeks is enough for English speakers to acquire Arabic or Russian. But this is not an issue here. What is important is that languages in the first tier clearly differ from English, both linguistically and culturally while languages in the second tier are supposed to be much closer from both aspects. Typological differences, however, seem to play a more important role in course length than culture does. This can be demonstrated very clearly with some examples from the list. There is not much cultural difference between Finnish and Swedish for English native speakers but linguistically Swedish is closer to English than Finnish. The same can be said about Rumanian and Hungarian with Rumanian being a Neo-Latin language and Hungarian belonging to the Finno-Ugric language group with a very complex morphology. Analyzing language aptitude Child (1998) referred to a more elaborated version of the language distance chart developed by the Foreign Service Institute. This focuses on three "textual elements": A) Writing system as representation of spoken language B) Grammatical system as framework for communication C) Semantic system as cultural outlook. These three phenomena are then ordered in such a way as to indicate relative distance of foreign languages from English: Near (1), Middle (2), and Remote (3). Using this system Child (1998:7-9) characterized German as A1/B2/C2, Turkish as A1/B3/C3, and Japanese as A3/B3/C3. Linguistic and cultural distance, however, does not mean that there is no, or very little, positive knowledge and skills transfer between languages. Mace-Matluck, Hoover and Calfee's study ( 1984) suggested that transfer of underlying cognitive skills occurred between Chinese and English despite strong surface structure dissimilarities. They examined how variables such as length of stay, level of LI academic proficiency and amount of reading instruction received influenced the L2 (English) oral and written proficiency of Chinese stu-

The origin of SBUs

91

dents in Seattle. Their findings demonstrated a consistent beneficial effect of bilingual instruction and of instruction in Chinese on the development of English literacy skills. It is important to note, however, that very little research has been done to describe the effect of language distance in the symbiosis of two or more languages, and whether differences and similarities have any impact on language production through either channel: what is better from the perspective of bilingual development? If the two languages share similar typological and cultural features or the opposite: when they have hardly any features in common. Most researchers claim that language learners are sensitive to encoding and decoding differences between their LI and L2 and/or Lx. A study conducted by Kecskes and Papp (2000a) demonstrated to some extent that positive transfer from the FL to the LI is expected to be especially intensive, marked and beneficial if the LI and the FL differ from each other in configurationality. As far as the use of SBUs is concerned, they are more sensitive to cultural distance than to typological distance because they are built on a cognitive structure or frame which are culture- rather than languagespecific. If these structures or frames are similar in the two cultures a better understanding of the given target language SBU can be expected in the L2 learner. This is in line with Bortfeld's findings (forthcoming) who investigated the understanding of idioms by non-native speakers. If the conceptual structures which are responsible for the idiomatic expressions are shared by the two languages the different lexicalization usually does not make processing problematic. For instance, the use of the equivalent of the above mentioned 'Guten Appetit' in close European cultures such as Austrian, French, Hungarian (German influence), etc. is essential. In Eastern European cultures it is common practice to ask the guests to stay when they want to leave. Consequently, the Hungarian SBU "Maradjatokmég.koraiméghazamenni" ("Please stay; it is still early to go home") has equivalents in Slovakian or Czech although they are Slavic languages and significantly differ from Hungarian typologically. It would not make any sense to look for the equivalent of this expression in American culture where the social understanding is that the guest knows when s/he wants to leave and the host has no right to advise her/him on this issue. However, the closeness of socio-cultural background can be misleading sometimes. Kellerman (1984) has found in the case of Dutch learners of English and German that this closeness might well be an impediment to learning idioms. Dutch learners were reluctant to transfer idioms from

92

Chapter 4

Dutch into English or German when transfer would have worked. Further research is needed to investigate if this is true in the case of SBUs as well.

4. Why these utterances and not others? 4.1. Motivated SBUs An important question is why was it precisely these utterances that started to be used to express those pragmatic functions and not others. Why is it that exactly these phrases have obtained those functions and not others? These questions can be answered only within a cognitivepragmatic framework. The traditional "arbitrariness of sign" view in which linguistic meaning is divorced from the human conceptual system will not lead us anywhere (c.f. Fónagy 1956)16. It is certainly not arbitrary why these expressions fulfill the given functions and not others. As claimed before, SBUs are motivated expressions and demonstrate better than any other linguistic unit that there is a strong cognitive-linguistic interdependency. The development of certain SBUs can be accounted for through cognitive mechanisms, and vice versa: learning an SBU for a culturally important category can linguistically reinforce the learning of the category itself.17 What does the selection of SBUs for situation-specific communicative functions depend on? The traditional view emphasizes that the selection of a given form (be it a word, a phrase, or a sentence) for a particular function may simply be arbitrary in a given culture (Pawley and Syder, 1983). Following Searle (1975), Nattinger and DeCarrico (1992: 17-18) went one step further when they argued that the choice of a form may be, at least to some degree, predictable. This predictability can be based on syntagmatic simplicity and paradigmatic flexibility. According to Nattinger and DeCarrico ( 1992:17) lexical phrases with a relatively simple syntactic pattern and a flexible amount of lexical variation are favored for common pragmatic acts. The cognitivepragmatic approach refutes both explanations and claims that relationships between linguistic form and function reflect human conceptual structure and general principles of cognitive organization (c.f. Sweetser 1990; Nuyts 1992; Kövecses and Szabó 1996; Kecskes 2000). SBUs are usually based on conceptual metaphors or conventional implicatures. Their meanings are motivated and not arbitrary.18 They make

The origin ofSBUs

93

the fullest use of cultural knowledge common to the members of a speech community. If, for instance, we want to explain exactly why "give me a break" (and not "give me a rest" or "give me a pause") is used to function as an objection or rejection in American English we will have to examine the relationship between the given function 'objection' (or 'rejection'), the semantic domain of this phrase and the cultural knowledge attached to the expression. It has been argued (c.f. Lakoff and Johnson 1980; Johnson 1987; G. Lakoff 1987) that our cognition and hence our language operate metaphorically. This view can be used to explain polysemy and specific figurative usage not only at the word and idiom level, (as done, for instance, by Sweetser 1990; Kövecses and Szabó 1996) but also at utterance level. When we use "give me a break" as an objection rather than some other utterance it is not just a fact about the language but also about the cultural community that sees "asking for a break" as metaphorically standing for objecting to something. As a result of this metaphorical understanding, the linguistic expression acquired a pragmatically motivated new sense. There is a great deal of systematic motivation for the meaning of most SBUs, which is similar to what Kövecses and Szabó ( 1996) found in idioms. As mentioned above, several SBUs derive from conceptual metaphors which bring two distant domains (or concepts) into correspondence with each other. One of the domains is usually more physical than the other (which is thus more abstract). The correspondence is established for the purpose of understanding the more abstract in terms of the more concrete. Here are some conceptual metaphors from Kövecses (2002:30): LOVE IS A JOURNEY: "We'll just have to go our separate ways". THEORIES ARE BUILDINGS: "We have to construct a new theory" IDEAS ARE FOOD: "I can't digest all these facts". However, there are some metaphoric expressions whose origins are not as clear because it is not easy to find their source. The SBU "piece of cake" is a good example. This pragmatic unit first appeared in the written record in 1936, in Ogden Nash's Primrose Path: "Her picture's in the papers now, And life's a piece of cake." One source (www.takeourword.com) says that "piece of cake" may derive from

94

Chapter 4

"cakewalk", originally referring to a type of promenade dance contest of the late 19th century in America, especially among African-Americans. The contestants displaying the best walk won a cake as a prize. Soon after, the word "cakewalk" came to be used figuratively as anything that was stylish yet easy to do. By the middle of the 20th century, the expression had been transformed to a "piece of cake", al-though "cakewalk" also survived with the same meaning, as a headline from MSNBC "Cakewalk To Kabul" (November 15, 2001) illustrates. Another Internet source (www.pseudodictionary.com) also refers to the African-American origin of the SBU, giving the similar opinion that it derives from a dance contest in which couples made complex strutting movements, usually with high steps, and the winners won a cake. The dance was called the "cakewalk" and the expression "That takes the cake." derives from it. A related pragmatic unit is "easy as pie". These explanations look convincing; however, the source domain is not clear. Yet another explanation can connect metaphorically the easiness of a task (target domain) with the lightness of a cake (source domain). Some other typesof SBUs evolve out of implicatures. Grice(1968; 1975) made a distinction between conversational implicatures and conventional implicatures. Conversational implicatures are created ad hoc in the course of conversation and have to be calculated by the interlocutor. Mey (2001:46) argued that conversational implicature concerns the way people understand an utterance in conversation in accordance with what they expect to hear. Here is a "classic" example: (46) Amy: Bill:

- What on earth has happened to the roast beef? - Your dog is looking very happy.

Bill's response is a typical conversational implicature in which his opinion of what has happened is not directly expressed but is expected to be figured out by the interlocutor when he makes the contact between the roast beef and the happiness of his dog. Conversational implicatures are generated by a particular utterance for a particular receiver in a particular situation, so what is conversationally implicated is somewhat indeterminate and cancelable. For instance, in the situation above the dog is not directly accused of having eaten the roast beef.

The origin ofSBUs

95

Conventional implicatures significantly differ from their conversational counterparts. As Mey says, "they implicate by themselves, or "conventionally", a certain state of the world, regardless of their use (Mey 2001:51)." They create context rather than are created by it. They may become routine formulas and are used almost automatically. The conventional implicature of an expression is quite specific, and usually means the same for members of a speech community so its content is hardly cancelable. The speaker cannot deny that s/he meant what was said without involving herself/himself in a contradiction. These expressions are routinely interpreted in a derived (i.e. nonliteral) sense. For instance, "I'm not buying this" is often used by Americans to suggest that the speaker is not willing to accept something as it was said. The original cognitive mechanism responsible for the situational meaning of the expression is no longer maintained consciously, and no inferential reasoning is necessary to find out that the speaker means. The linguistic form acquired a pragmatically motivated sense which has become conventionalized. Recent literature on semantic change (i.e. Warren 1992; Bybee et al. 1994; Geeraerts 1997) demonstrated a growing tendency to treat the conventionalization of conversational implicatures as a separate mechanism of semantic change which is of the same nature as more traditional mechanisms such as metaphor and metonymy. Geeraerts (1997:83) argued that while regular implications apply to any member of a category, conversational implicatures do not involve the category as a whole but only a contextually salient subset. Of course, not all of them will develop into conventionalized implicatures because the process of conventionalization is dependent on several factors including frequency, expressiveness, typicality and others.

4.2. Non-native speakers ' difficulties processing implicatures Coulmas ( 1981 a: 8) argued that non-native speakers often have trouble with many routines because they try to interpret those expressions "on the basis of word meanings alone and without knowledge of cultural habits, customs, values, attitudes, etc." We need to be careful though because what Coulmas claimed does not relate to all idiomatic expressions. Non-native speakers have less problem with those that have more universal knowledge as their underlying cognitive mechanism suchas "what's up?", "no problem", or others (Kecskes 1998). Bort-

96

Chapter 4

feld (forthcoming) also came to a similar conclusion when investigating what mental images native and non-natives speakers of English form about familiar American English idioms. Non-native speakers did as well as native speakers where the underlying cognitive mechanisms were universal rather than culture-specific. There is a special group of conventional implicatures, however, that are especially problematic for non-native speakers (NNS): relevance implicatures. The uniqueness of this group is in that it is usually not the expression itself that is conventionalized but the function and/or the structure. When, for instance, there is a "yes/no" question, or a statement and the obvious answer from the perspective of the interlocutor is "yes", a relevance implicature may be used in a question format as in (47): (47) Amy: Don't you think Jim drinks a bit too much? Billy: Is the Pope Catholic? One and the same function ("Isn't this obvious for you?") can be lexicalized in a lot of different ways. This kind of implicature is not necessarily culture-specific but the way it is lexicalized can be, and this may result in an absolute misinterpretation of the implicature in L2 processing. Investigating the non-native speakers' skill in interpreting implicatures in American English. Bouton (1994:98) argued that in certain cases of the relevance-based implicatures the learners did not understand some points related to the nature of the utterance, the context, or both. He suggested that non-native speakers must learn the particular "culture points" on which the relevance implicatures are based. The pragmatic unit "Is the Pope Catholic!" is a good example of how a conventional implicature becomes a fixed expression. In a study (see a detailed description of the study later) that focused on the graded salience hypothesis in L2, SBUs that derived from conventional implicatures because they lacked the appropriate knowledge structures which were responsible for the selection of the given SBUs in English. Some of the explanations from respondents of different LI background and culture demonstrated very clearly that students relied on their LI socio-cultural background when processing relevance and misinterpreted the relevance implicature (47) when its content was too

The origin ofSBUs

97

specific in the American-English culture. Some examples: Russian student (2 years in US, female, age 23): much"

"Catholics drink too

Spanish student (6 months in US, female, age 27): Catholic people drink too much"

"Billy thinks that

Spanish student (8 months in US, male, age 22): Catholic with alcoholic."

"Billy confused

When the relevance implicature was expressed by an utterance whose processing required less specific knowledge the results showed that the connection was easier for NNSs to make. For instance in (48): (48) Alan: - Do you think Mr. Herbert will give us a quiz today? Bob: - Does the sun rise in the east? The results of Kecskes' study (2001) are in accordance with Bouton's (1994) conclusion: learners did not recognize the structural and functional relationships between the two relevence implicatures having similar functions. Another interesting example for the relevance implicature is (49), and (50). What is conventionalized here is not the expression but the function: "if you do not want to say anything negative about something/someone, talk about things that are positive about that person or thing." The obvious irrelevance of response is expected to indicate negative opinion to the interlocutor as in (49) and (50):

(49)

Mr. R: Mr. M: Mr. R: Mr. M:

-

Have you finished with Mark's term paper yet? Yeah, I read it last night. What did you think of it? Well, I thought it was well typed.

98

(50)

Chapter 4

Ann: - How do you like your teacher? Bill: - Well, he dresses nicely.

The results (Correct: [49] = 35%; [50] = 41%) demonstrated that students relying on the propositional meaning of the expressions had difficulty interpreting the intended meaning in these situations (Kecskes 2001). In (50) many students noticed only the positive propositional meaning of the expression and concluded either that Bill had a positive opinion about the way the teacher dresses (25%), or that Bill thought that the teacher was a good one (24%). This shows how important the lexicalization of a situational function is for speakers of different languages. Even if the compositional meaning is easy to process the lack of an adequate supporting conceptual structure does not always allow NNSs to process this type of SBUs properly.

Chapter 5 Distinguishing features of SBUs

1. SBUs as pragmatic idioms Kecskes (2000) argued that SBUs are pragmatic idioms which should be distinguished from other types of idioms. SBUs have to be related to predictability as phraseological units (for instance: 'spill the beans'; 'it's raining cats and dogs'; 'thin end of the wedge'; 'tip of the iceberg'), frozen metaphors (for instance: 'he drives me up the wall'; 'the ball is in your court now'; 'she has one foot in the grave') and lexical idioms (for instance: 'bring up'; 'turn down'; 'put up with'). It is not my goal here to give a detailed account of the definition and classification of idioms. Instead, I am going to focus on the relationship of those idiom-groups to SBUs which show the most similarity to our targeted pragmatic units. The most common definition of idioms states that they are linguistic expressions whose overall meaning cannot be predicted from the meanings of the constituent parts. "Idioms are conventionalized expressions in which the intended meaning often is difficult or impossible to recover from the words making up the expression (Katz et al. 1998:3)." In the utterances "He kicked the bucket", or "They could hardly make both ends meet", the compositional meaning of elements gives hardly any hint about the functional meaning of the expressions. Idioms can be placed on a continuum according to the extent that their overall meaning is predictable without context.19 On this continuum predictability increases to the left:

The problem with SBUs is that although most of them show as little (or no) predictability just like phraseological units some of them stand closer to lexical idioms or frozen metaphors. The extent of predictability is based on the extent of systematic conceptual motivation that underlies all idioms including SBUs. According to the traditional view, idioms are a matter of language and constitute a special set of the larger category of words (e.g. Carter and McCarthy 1988), and are independent of each other. In this

100

Chapter 5

approach the meaning of idioms is divorced from the human conceptual system and encyclopedic knowledge that speakers of a language share. As opposed to this view, cognitive linguists claim that most idioms are products of the conceptual system rather than simply a matter of language. Idioms are embodied in our conceptual system as a result of our experience with the world. Consequently, cognitive linguists conclude that idioms are conceptual rather than linguistic in nature. If we want to understand how the meaning of idioms is constructed we need to analyze the cognitive mechanisms that are responsible for idioms: metaphors, metonymies and conventional knowledge. These mechanisms provide motivation for the idiomatic meaning which is not arbitrary as claimed in the traditional view. Lakoff (1987) distinguished prediction from motivation which is a weaker notion than the former. There is no claim that based on the non-idiomatic meaning of an idiom (e.g. "kick the ball o f f ' for the idiom "kick off') we can fully predict what the idiomatic meaning ("start") is. Conceptual motivation is either stronger or weaker in the mind depending on how much psychological reality the expression may have for speakers of the given community. Based on the results of a study Gibbs (1990, 1994) claimed that conceptual metaphors have psychological reality and that they motivate idiomatic expressions. Subjects in the experiment were asked to form mental images of idioms, and then were expected to answer questions about their images. Gibbs showed that there was a significant degree of consistency in people's images and responses to the questions. The study demonstrated that people have tacit knowledge of the metaphorical basis for idioms. It has to be noted, however, that the nature of tie and motivation between the originating knowledge-base (metaphor, metonymy, or conventional knowledge) and the given idiom keeps changing under the influence of several variables including time, societal perception, frequency of use and others. If the originating knowledge-base is easily recallable, the motivation is stronger, if it is not then sometimes speakers may see no motivation at all in present-day English in the case of some idioms. There is certainly no problem finding the knowledgebase for the following idioms:

Distinguishing features of SBUs

101

(51) The murder sparked off the riot. The president kicked off the meeting with a short speech. You are history. Give me a hand. The same, however, cannot be said about the following expressions: (52) You're pulling my leg. John's uncle kicked the bucket. Don't be a wet blanket. It's raining cats and dogs. Kövecses (2001) mentioned an interesting example. Some native speakers of English disagreed with his analysis of the idiom "wet blanket " which he related to "extinguish the last sparks" arguing that the mapping 'end of fire > end of state/event' applies to both. The native speakers said that they no longer associate the idiom "wet blanket" with the FIRE-metaphor. This basically shows that after a particular period of time in language development, the tie between the originating knowledge-base and the idiom may first become vague and then be cut off forever, and the expression will have no psychologically established relation with the source anymore. This is how we can distinguish phraseological units from frozen metaphors and lexical idioms. The strength of the tie between the source and the expression can be put on a continuum where the strongest tie is on the left, and as we approach the right edge the strength of the tie decreases. Phraseological units appear to lose their tie with their original source while lexical idioms do not. Group A above can be considered as lexical idioms and Group Β comprises phraseological units. So (52) can be placed to the right edge of the continuum while (51) is close to the middle. The tie of the idiom with its originating knowledge-base is important from the perspective of the argument for relative independence of lexical semantics from conceptual semantics. In many phraseological units we do not have conceptual motivation for the meaning of the

102

Chapter 5

idiom at all (for instance: 'kick the bucket', 'beat around the bush', 'cats and dogs', etc.). Kövecses and Szabo (1996) pointed out that these idiomatic expressions are the most celebrated examples of idioms in the standard view because they are idioms that can be said to be simply a matter of language and not the conceptual system. This example shows very well how important historical perspective is in the analysis of meaning. The relative independence of a part of the semantic system can be justified synchronically but not diachronically. What phraseological units, lexical idioms and pragmatic idioms (SBUs) have in common in the cognitive paradigm is that most of them are based on cognitive mechanism and claimed to be conceptually motivated by the underlying metaphors, metonymies and conventional knowledge (c.f. Gibbs 1990; 1994; Kövecses 2001). Idiomatic expressions are physiologically motivated in the sense that the connection between a source domain and target domain is based on direct sensorimotor experience with the world, and that the source domains of the metaphors are concepts that emerge from direct sensorimotor experiences (Kövecses 2001). 1.1. Metaphor-based idioms Conceptual metaphors bring into correspondence two domains of knowledge: the source domain and a target domain. The source domain of metaphors is usually a familiar physical domain while the target domain is a less familiar abstract domain (e.g. Lakoff and Johnson 1980; Kövecses and Szabo 1996; Kövecses 2001). The source domain is used to make sense of the target domain. Kövecses and Szabo (1996:332) used the following way to analyze idioms: Linguistic forms: spit fire Meanings of forms: 'spit', "fire' Special idiomatic meaning: 'Be very angry' Cognitive mechanisms: Metaphor: ANGER IS FIRE Conceptual domains: FIRE and ANGER

Distinguishing features of SB Us

103

The same type of analysis can be used with SBUs: Linguistic forms: Welcome aboard Meanings of forms: 'welcome', 'aboard' Special idiomatic meaning: 'welcome to the new job'; 'glad to have you here' Cognitive mechanisms: Metaphor: SHIP IS JOB Conceptual domains: SHIP and JOB The metaphor 'SHIP IS JOB' has a historical significance which does not have psychological reality in the mind of native speakers of English any more. Sailing for the British was a way of life for centuries. Most jobs on the British isles were connected with sailing, trading and shipping. This is not so anymore but it certainly explains the etymology of the metaphor. Kövecses (personal communication) gave the following explanation: "I think that it is the Event Structure metaphor that largely accounts for the meaning of this idiom. In particular, relationships are vehicles and purposeful activities are journeys. The ship is a vehicle that corresponds to the new relationships the person will be participating in (this is similar to 'This relationship isn't going anywhere' in the 'love is a journey' metaphor). The metaphor has the interesting entailment that if relationships are vehicles, then the fate of the travelers in the vehicle is a shared fate. The expression 'welcome aboard' means or connotes, at least for me, that we will share the same fate (whether good or bad) from now onward. The second metaphor, 'purposeful activities are journeys', has as one of its mappings 'the beginning of a sea journey is the starting of a new job,' i.e., beginning to do all the purposeful activities that the job requires. How do we know that it is this particular mapping that applies to the idiom as used here? It is because in the 'ship' domain we say 'welcome aboard' to someone who joins us on a ship before we embark on a journey with that ship (Kövecses 2002: e-mail communication)."

1.2. Metonymy and conventional knowledge While a metaphor involves an "is" or "is understood" relationship between two distinct conceptual domains, a metonymy involves only one domain and expresses a "stand for" conceptual relationship between two entities within this single domain (c.f. Lakoff and Johnson 1980; Kövecses and Szabo 1996). According to Kövecses and Szabo (1996) the metonymy "THE HAND STANDS FOR THE ACTIV-

104

Chapter 5

ITY" is the basis for several idiomatic expressions such as "hold one's hand" ('wait and see'), "sit on one's hand" ('deliberately do nothing'), "join hands with somebody" ('cooperate with another person'), and the like. The SBU "give me a hand" can also be explained by this metonymy. Conventional knowledge as a cognitive mechanism can also be the source of many idiomatic expressions. People belonging to a given culture share information concerning conceptual domains. This shared everyday knowledge has been labeled in different ways in the literature: such as, for instance, "frame" or "scene" (Fillmore 1976b; 1977), "schema" (Langacker 1987), "idealized cognitive model" (Lakoff 1987), "cultural model" (Gee 1999) or "folk theory" (Holland and Quinn 1987). According to Fillmore (1977) frames and scenes are coherent organizations of human experience. They play a specific role in the development and use of SBUs because the meaning of an SBU usually applies within a relevant frame. The consenses of SBUs are pragmatically conditioned within situational frames. Their use is closely tied to the performance of specific meanings, so they are more or less fixed and predictable. Each prefabricated expression is linked to a communicative function the language user is motivated to perform. Kiefer (1995) suggested that SBUs be discussed within the confines of frame semantics because any account of stereotypical expressions calls for frames. Frames are interpretive devices by which we understand a term's deployment in a given context. Each situation has a script which is a sequence of events and actions, taking place in frames (van Dijk 1980:234). Gee (1999) claimed that the situated meaning of lexical items is associated with cultural models which are "storylines," groups of connected images, or informal "theories" shared by people who belong to a particular sociocultural group (D'Andrade 1995; D'Andrade and Strauss 1992; Holland and Quinn 1987). These cultural models are tied to each other in complex ways to create larger and larger storylines, and these connected networks of cultural models help to organize the thinking and social practices of socio-cultural communities (Gee 1999). SBUs have their own cultural models as seen in examples in the former chapters. They are often 'indexical' in the sense that their occurrence evokes a specific subevent of a particular script (c.f. Fillmore 1976; Coulmas 1979; Kiefer 1995). They can create their own context and are recalled (required) by context. Predictability and boundedness to a specific subevent is, of course, a matter of degree rather than abso-

Distinguishing features

of SB Us

105

lute. To what extent are situated meanings (consenses) of SBUs routinized? The more predictable the link between a subevent and an utterance is, the more bounded the expression is (c.f. Kiefer 1995). The following example (adapted from Genzel 1991:54) demonstrates how situation-bound utterances are tied to certain subevents: (53) Frame: general conventional knowledge connected with doctor's office. Script·. (Don is the customer, Bob is the receptionist) Subevents-. (a) D enters and greets the receptionist B, (b) Β looks up and inquires about D's goal, (c) D tells about his appointment, (d) Β asks him to fill in a form, (e) D thanks B, (f) Β reacts to thanks.

D: B: D: B: D: B:

- Good afternoon. - Good afternoon. Can I help you1? - I'm Donald Daniels. I have a 2 o'clock appointment. - Yes, Mr. Daniels. Please fill out this form. - Thank you. - You are welcome.

As it is illustrated in the italicized expressions in the dialog, a given domain of knowledge can often account for a particular idiomatic meaning in a direct way, i.e., without a metaphor or a metonym.

106

Chapter 5

2. Differences between idioms and SBUs 2.1. SBUs are strongly tied to standard situations SBUs are usually idiomatized and are less often used in their original literal meaning than in their routinized, situation-bound pragmatic function. SBUs can be considered as pragmatic idioms that take a unique place among idioms because they have some features that distinguish them from other idiomatic expressions. SBUs create their own frame and are hardly used outside their usual context. This is not the case with other idioms because they can appear in many different situations and their immediate context is usually created ad hoc rather than bound. Cruse (1997) argued that when idioms are used it is rare that the whole sentence is idiomatic. All the items except those which form a part of the idiom can be changed without destroying the idiomatic meaning. For instance: (54)

Jane pulled Marta 's leg about her boyfriend.

(55)

Peter is pulling Jim's leg about his bike.

(56)

They pulled my leg about my new mobile home all the time.

SBUs usually form one single unit which functions as an utterance. No change is possible within the unit and generally the immediate environment is quite predetermined. For instance:

(57)

Bob: - Thank you for your time. Clerk: - No problem.

(58)

Assistant: - Can I help you? Customer: - Thank you. I am just looking.

Distinguishing features ofSBUs

107

2.2. SBUs can have more than one salient meaning SBUs are the products of the process of conventionalization. As we discussed above, SBU may originate from both other types of fixed expressions or freely generated expressions. If they derive from freely generated expressions they may often keep their original compositional meaning. This is one of their unique features that distinguishes them from other idioms. While the situation-bound meanings of SBUs are often quite different from the sums of their parts they still may keep their compositional meaning in some situations. For instance: "Give me a hand", "I'll talk to you later", "Don't go away". In certain cases they can be processed both in a literal and figurative sense. However, this is usually not the case with idioms where the connection between compositional meaning and functional meaning is almost completely lost. For instance: 'wet blanket', 'spill the beans', ' touch base with', 'go Dutch', 'get down to the wire', etc. The DMM explains the difference between idioms and SBUs in the following way: Idioms : invariant (coresense) has only one variant (consense). SBUs: consenses may reflect upon different aspects of coresense. The next two situations demonstrate very well what happens when one and the same expression ("How are you?") is understood by the interlocutor as an SBU (situation [59]) or as an expression with a propositional meaning (situation [60]).In each society there are standardized communication situations which require the members of that particular speech community to react in an automatic manner (c.f. Coulmas 1981a). Routine, recurring situations require no real negotiations between individuals as the beginning of the following conversation demonstrates: (59) Intern: - How you doin '? Patient: - Pretty good, actually. Intern: - Good. I'm glad. What brings you into the clinic today?

108

Chapters

Patient: - Uhm, I need to.. I'm taking a trip to Ecuador and need to get a series of inoculations. Intern: - Okay. (Hospital Site 5-14-93 Tape 1. Data collected by Anita Pomerantz) This conversation between an intern and a patient begins "routinely". There is no misunderstanding: both the intern and the patient comprehend the situation properly and act accordingly: they use SBUs at the beginning of the conversation. If, however, for some reason, one of the interlocutors does not give the expected response, because s/he processes the expression literally this may confuse his/her partner. This point can be illustrated by an extract from David Lodge's novel Paradise News (cited in Mey 2001:143): (60) "The waitress, whose name, Darlette, was displayed on a badge pinned to the front of her apron, put a jug of iced water on the table and said brightly, 'How are you this evening, sirT'Oh, bearing up\ said Bernard, wondering if the stress of the day's events had marked him so obviously that even total strangers were concerned for his well-being. But he inferred from Darlette's puzzled expression that her enquiry had been entirely phatic.'Fme, thank you\ he said, and her countenance cleared (Lodge 1992:129)." Focusing on the propositional meaning of the question, Bernard did not give the usual "fine, thank you" type of response to Darlette's SBU which resulted in a puzzled expression on Darlette's face. Noticing this confusion Bernard found it important to go back to the "routine" and give a conventional response "Fine, thank you" to Darlette's SBU. This was his solution to a face-threatening situation. He thought it was better not to explain why he did not really feel "fine".

2.3. SBUs are more context-sensitive than other idioms Because of the possibility of different consenses described above, an SBU by itself is generally not enough of a meaning-carrier. Mey ( 1993)

Distinguishing features of SB Us

109

argued that what has happened prior to the utterance, and what is going to happen afterwards, is of as much, if not more, important as what is talked about within the frame of the utterance itself. SBUs are more context-sensitive than phraseological units or frozen metaphors. Idiomatic expressions such as 'kick the bucket', 'pull one's leg', and others always make the user or interlocutor think of their idiomatic meaning which is the most salient meaning in their case. They usually have no meaning other than that. So special effort is needed to cancel the idiomatic meaning when the expression is freely generated. For instance: (61) Bill kicked the bucket and the whole carpet got wet. (62) Jim is pulling Peter's leg because he has a cramp. As we saw above SBUs may often have a freely generated counterpart. Compare the following: (63) Larry King: - We have to take a break now. Don't go away. (64) Sam: - Don't go away. Your parents will be here any minute. In (63), the utterance "Don't go away" is an SBU while in (64) it is a freely generated expression. This is a case where parallel processing is induced because more than one meaning is salient, (c.f. Blasko and Connine, 1993). Context also defines to what extent an SBU is routinized and/or institutionalized. This is not the case with idioms where synchronically the co-text or context does not have an impact on how routinized an idiom is. Compare how routinized the same expression is depending on the context in the following examples: (65) George: - Uh-oh! The baby's going to cry again! Kate: - Take it easy, George. He's only yawning (Howard 1987:7).

110

Chapter 5

(66) Greg: - Well, I must go now. Bob: - Take it easy, man. (67) John: - Oh, God, I have lost again. Father: - Take it easy. You'll have another chance to win. The context-sensitivity of SBUs can be demonstrated cross-culturally. Why is it that in certain languages these institutionalized units become semantically almost or fully empty while in other languages they keep their compositional meaning or at least a part of it in spite of their frequent use? The following example demonstrates the problem: (68)

American English:

A: B:

- "How is it going?" - Fine, thank you."

(69)

Russian:

A: B: is OK.)

Κακ dena? HuneTo. Bee β ηορηόκε.

(70)

Hungarian:

A: B:

- "Hogy vagy?" - " Köszönömjol".

(How are things?) (No problem. Everything

(How are you?) (Fine, thank you.)

As we can see there is quite a bit of difference in the wording of this particular function in the three languages. The real difference, however, comes in after the SBU "Fine, thank you" in English or its pragmatic equivalents in the other two languages are uttered. In American English it is almost sure that "B" will not continue by explaining why he is "fine", while in Russian and Hungarian "B" usually proceeds with

Distinguishing features of SBUs

111

telling something about recent events in his life. In American culture it is not expected to justify the use of the SBU "Fine, thank you" because it is taken for granted that even if you have problems you do not necessarily show that to your interlocutor. This can be illustrated by a joke from the TV program "Saturday Night Live": (71) Agent One: - "How's your wife?" Agent Two: -"If I had a wife, she'd be fine!" However, in the other two cultures the speaker has a kind of obligation to support his positive answer with some information because in Russian and Hungarian cultures people are more likely to complain than say something positive about their life. Complaining is something that is almost as strong an expectation as giving no justification is in American culture . This difference in social expectations has a serious impact on the semantical content of the expression. In American English "Fine, thank you" becomes semantically empty and develops a specific pragmatic function while its Russian and Hungarian counterparts generally keep their propositional meaning because of how the conversation usually continues. In the company of freely generated expressions that are functionally and contextually connected to them and, in a way, continue the line of thought expressed by them, SBUs tend to keep their compositional meaning. If, however, they are surrounded by other prefabricated phrases there is a good chance that they become semantically empty.

3. Content of SBUs SBUs are essentially a social phenomenon. Coulmas argued that "the more tradition-oriented a society is, the more its members seem to make use of situational formulae (Coulmas 1981:11)." This raises the question of how the content of SBUs relates to the situation(s) they are used in. Depending on whether SBUs directly say something about the action, and/or participants or relate them to other situations or agents, I will make a difference between situation-bound routines ("nice to meet you"; "how are you?"; "take care"; "you are all set";" welcome aboard", etc.) and situation-bound rituals ("God bless you"). Situation-bound rituals are close to Goffman's "interaction rituals", a term with which Goffman referred to SBUs (1967). Situation-bound

112

Chapters

rituals usually relate the situation to other situations, events or agents. They are especially frequent in tradition-oriented cultures such as Japanese, Arabic, Chinese and Turkish. The use of situation-bound rituals is almost obligatory and usually no freely-generated phrases are acceptable instead. They behave like Yiddish psycho-ostensives about which Matisoff claimed that "often it is not so much that the speaker is using an emotive formula that actually belies his true feelings, as that the formula has become a surrogate for the true feeling, an almost automatic linguistic feature that constant usage has rendered as predictable and redundant as the concord in number between subject and verb (Matisoff 1979:6)." Yiddish psycho-ostensives just like SBUs refer to the speaker's attitude to what s/he is talking about. Tannen and Öztek (1981), however, called our attention to the fact that while in Yiddish culture the priority is on verbal inventiveness, (consequently, these emotive expressions are productive), in Greek and especially in Turkish culture situation-bound rituals are fixed set and are usually complete utterances. In tradition-oriented cultures situation-bound rituals are not considered insincere because these cultures seem to have agreed to accept the surrogate as evidence for the true feeling (Tannen and Öztek 1981). For instance in Turkish: (71) If someone mentions a bad event or disaster, this should be followed by an expressions to erase the effect: "Agzindan yel alsin "

'May the wind take it from your mouth.'

Tannen and Öztek (1981) argued that in Turkish many formulaic expressions deal with human powerlessness, and usually two sources of power are appealed to: God and the magic power of words. For instance, when asking for God's aid Turks say: (72) "Allah kolaylik versin "

'May God give ease'

In contrast to situation-bound rituals, situation-bound routines generally do not sound sincere, and both the speaker and interlocutor are aware of this. They usually say something that is directly related to the situation, participants or actions. They are standardized solutions to coordination problems where no negotiations are necessary. Situationbound routines serve an important function in giving people confidence

Distinguishing features ofSBUs

113

and behavioral certainty in conversation. Future-oriented cultures (like the American culture) as opposed to tradition-oriented cultures generally prefer the use of situation-bound routines to situation-bound rituals. This is one of the reasons why it is so difficult to find the English equivalent to Turkish or Japanese situation-bound rituals. The situations that prompt the use of SBUs in tradition-oriented cultures, in many cases, are not even recognized by native speakers of English as events which demand any verbal reaction (c.f. Bear 1987). To illustrate this point Bear (1987) listed several situations where English native speakers usually have nothing to say while the Turks find it necessary to use an SBU: (73) "sihhatler olsun" [ "may (it) be healthy"] To address someone who has just taken a bath. (74) "helal" choked.

["(it is) lawful, legitimate"]

To someone who

(75) "giile giile giy" ["wear laughingly"] acquired a new article of clothing.

To someone who has

(76) "giile giile otur" ["stay laughingly"] moved into a new house.

To someone who has

In contrast, English SBUs are more "down-to-earth", and there is less endeavor in them to establish rapport between participants of conversation. For the future-oriented Americans expressions, such as "See you soon", "I'll talk to you later", I look forward to seeing you again", "Why don't we have lunch tomorrow?", and others are quite natural. Zaharna (1995) argued that in Arabic culture, envisioning future activities as certainties sounds very unnatural. When Arabs speak about an action in the future they often use the situation-bound ritual "In sha allah " ("God willing") because it is only God who knows for sure what will happen in the future.

Chapter 6 Classification and interpretation of SBUs

1. SBUs behave like words Kiefer (1995) argued that the important role prefabricated expressions and sentences play in everyday discourse has blurred the distinction between "words" and "sentences". It can be no longer maintained that "words" are listed in the mental lexicon and "sentences" are freely generated. The lexicon contains a large number of prefabricated sentences and phrases which are associated with certain typical situations (c.f. Moon 1992; Verstraten 1992; Wray 2002). Kiefer (1995) suggested that the situational meaning of an SBU usually cannot be derived from its semantics, therefore, SBUs must be listed in the lexicon in the same way as words are listed there. Several researchers (c.f. Bolinger 1976:9; Nattinger and DeCarrico 1992:12; Plunkett 1993:45) assumed that the lexicon contains both complete formulaic sequences and some or all of their component parts. This means that the lexicon may have a separate entry for 'take it easy', and one for each individual word: 'take', 'it', 'easy'. SBUs, however, are pragmatic rather than lexical units so there must be reference in the lexicon to a particular frame or script which they belong to, or are most frequently used in. More precisely, each SBU must be characterized by the subevent of the script for the given frame with which it is associated. In the DMM frames represent the culture-specific conceptual properties of SBUs. As we saw in chapter 2 , CSCP are encoded in the lexical unit but are cancelable by context in words. This is not the case with SBUs: CSCP can rarely be canceled. As it was demonstrated in the previous chapter, SBUs seem to be more context-sensitive than words or other idioms. Situational meanings and pragmatic functions of SBUs may not apply at all outside the relevant frame and script because these functions and their frames are usually inseparable from one another. Take, for instance, 'you bet' or'piece ofcake'. It is hard to tell what they mean without a frame. Speakers usually name the most salient meaning when asked about the meaning of these expressions without context. They can obtain different meanings in different situa-

116 Chapter 6 tional frames, and create different contexts: their meaning changes by the situation.

1.1. Formula-specific pragmatic properties To explain the word-like behavior of SBUs I will use the Dynamic Model of Meaning that was presented in chapter 2 where the relationship of coresense and consense as well as word-specific semantic properties (WSP) and culture-specific conceptual properties (CSCP) were discussed. The DMM has to be adjusted to accommodate the pragmatic acts expressed by SBUs. Mey argued that the explanatory movement in a theory of pragmatic acts is from the outside in: "the focus is on the environment in which both speaker and hearer find their affordances, such that the entire situation is brought to bear on what can be said in the situation, as well as on what is actually being said (Mey 2001:221)." In the DMM situation-bound utterances also have a coresense and their instantiations in actual situations are the consenses. The coresense here is the communicative, pragmatic function that the SBU is capable of executing in a particular situation. The coresense is similar to Mey's (2001) "pragmeme" that he uses in his pragmatic act theory (PAT) to refer to a generalized pragmatic act which is concretely realized by "practs" in situational use. It is important to note, however, that Mey's theory has a more ambitious goal than the DMM. The PAT aims to explain pragmatic acts in general not just the use of a type of pragmatic expression. Situation-bound utterances have communicative functions which represent socio-cultural concepts. Their culture-specific conceptual properties (CSCP) are represented by their frames. When an SBU is used a particular frame kicks in. There are several possible pragmatic access routes (consenses) to each socio-cultural concept. These pragmatic access routes are represented by functionally synonymous SBUs. The relationship of these SBUs to the socio-cultural concept that they represent is similar to the relationship that words have with concepts.

Classification and interpretation

117

For instance: Socio-cultural concept: [when someone says "thank you" it must be recognized] Some SBUs representing this concept20: -

Don't mention it. It's OK. You are welcome. You bet.

Socio-cultural concept: [when you meet someone you not only greet him/her but want to sound friendly] Some SBUs representing this concept: - How are you doing? - How is it going? - Howdy. Socio-cultural concept: [you need to finish a conversation with a pleasant closing] - Nice meeting you. - It was good to talk to you. Each SBU has a coresense which represents the same concept. The several different pragmatic access routes are distinguished from one another by, what I call, "formula-specific pragmatic properties" (Kecskes 2000a). Just like words, SBUs can either be neutral, have no modulatory effect like 'it's OK or can have formula-specific pragmatic properties (FSP) like 'you bet ', or "howdy ". That is the reason why consenses instantiating the same coresense are not always interchangeable. (Nor are words):

118

Chapter 6

(77) - Professor Brown, thank you for the interesting lecture. - You 're welcome. ( "You bet " would not fit.) (78) - Professor Brown, how are youl ate.)

("Howdy " would not be appropri

According to the ΌΜΜ, formula-specific pragmatic properties can be either permanently-present or situationally-activated elements of an SBU's meaning. They are either encoded in SBUs or charged by the situation. Several SBUs have these pragmatic properties without the situation. Note, for example: 'howdy', 'you bet', 'welcome aboard', etc. However, many SBUs are not like this. Their formula-specific pragmatic properties (FSP) are not encoded in them but are charged by the situation, just like in the case of words. This is analogous to what Horn (1985) and Sweetser ( 1990) called "pragmatic ambiguity": when a single semantics (coresense of SBU) is pragmatically applied in different ways according to pragmatic context. The formula-specific pragmatic property is not encoded in the SBU but is charged by the situation in the following case: 79)

Rob: Debra:

- I think Ray was really rude to you yesterday. - Tell me about it. (Source: "Everybody loves Raymond". Sitcom. CBS. 3/12/2002)

The expression "Tell me about it" received its charge from the context. This SBU is often used without the FSP as below: (80)

Sally: - Greg, I just found an apartment to rent. Greg: - Really? Is it in the downtown area? Tell me about it.

Coulmas (1981b:73-79) argued that the literal meaning of these expressions is only suspended in use. However, a closer examination

Classification and interpretation

119

shows that it is not suspension of meaning that we have here but functional change or a pragmatic extension (FSP) of the basic semantic function (coresense). If SBUs are frequently used in a particular meaning they will encode that meaning, and develop a new formula-specific pragmatic property. This pragmatic property is getting conventionalized when basically it starts to mean the same thing for most native speakers. That is to say, when native speakers are asked what comes into their mind first when they hear a given expression, and their response will be very similar, we can say that the SBU has already encoded the formula-specific pragmatic property. If, for example, native speakers are asked what they think the meaning of "welcome aboard" or "piece of cake" is, most of them will not think of a ship or a cake which are supposed to be the primary reference of those expressions. What comes into their mind first is the meaning modified by the formula-specific pragmatic property of the given formulaic expressions. Consequently, those expressions have already encoded the formula-specific pragmatic property and do not need the situation to charge it. In a survey conducted with 34 native speakers of English at the University of Montana students were given a list often words and phrases out of context. The list contained items such as "cool", 'piece of cake', 'welcome aboard', 'gay', etc. Students were asked to use the given words and phrases in a sentence that comes first into their mind when they see the given item. Out of 34 respondents 27 used 'piece of cake' in the meaning 'easy', and 30 of them used 'welcome aboard' in the meaning "cordial greeting to a new member/members of a group'.

1.2. How does the DMM explain the use of SBUs? The DMM explains the situational behavior of SBUs in the following way:

(81) Policeman: - [(l)What's going on here?] [(2) What seems to be the problem?] [(3) What's up?] David: - He hit my car. Bob: - He stopped short for no reason.

120

Chapter 6

Policeman: (To Bob, the driver of the other car): - How fast were you going? Bob: - 1 don't know. Maybe 40. (Genzel 1991: 8) The dialog above happens after a traffic accident. The policeman is talking to the two drivers. He needs to address the participants of the accident and may choose from several SBUs which can substantiate this pragmatic function. Each of the three SBUs has a formula-specific pragmatic property that distinguishes it from the others. "What's going on here? " and "What seems to be the problem? " are likely to open up the same frame (CSCP) while "What's up? " appears to open up a different frame which is not appropriate in this situation. The CSCP that is attached to "What's up?" makes the use of the SBU appropriate between close friends, but not necessarily between a policeman and a participant of a traffic accident. So while the three SBUs have the same coresense and can express the same situated meaning their FSPs and CSCPs are different. The difference of FSPs allows substitution as in (1) and (2). If, however, the CSCPs are different, the two SBUs are not interchangeable.

2. Classification by specific pragmatic extension SBUs are similar to words in many respects but not in all. SBUs are the creations of language use, so in their case the explanatory movement is from the outside in. On the discourse level one of the main issues is the difference between what is said and what is communicated. SBUs reflect upon this difference very clearly. Their compositional meaning usually differs from their situated meaning so we need to take the literalfigurative dichotomy into account when we want to explain their use. SBUs do not have polysemy the way words do. They usually have one particular communicative function which is represented by their figurative meaning. In some cases both their figurative and literal meaning can be used. Then, however, they usually represent two different coresenses. I argued in chapter 1 that if we want to adapt Cruse's approach (1992) to the discourse level it needs modification because the theory ignores cases when a form, which is a "plain word" according to Cruse's theory receives a pragmatic charge from the situation. In the framework of the DMM it was suggested that the term "loaded" be used

Classification and interpretation

121

when the FSP is encoded in the SBU, and "charged" be applied when that particular pragmatic property is not encoded in the expression but is charged by the situation. This categorization will be used for SBUs which do not represent a homogenous group. According to the nature of their specific pragmatic extension the following types of SBUs can be distinguished:

2.1. Plain SBUs These SBUs usually have a compositional structure and are semantically transparent. Their situational meaning (consense) may only differ slightly from their propositional meaning because their pragmatic extension is minimal if any. Their most salient meaning can be computed from their compositional structure. For instance: (82) Assistant: Customer:

- Can I help you, Sir?. - Thank you. I'm just looking.21

(83) Bill: - Is that you, Jim? Jim: - Oh, hi Bill. It's so good to see you.

2.2. Charged SBUs An SBU may exhibit pragmatic ambiguity, in the sense that its basic function (coresense) is extended pragmatically to cover other referents or meanings (c.f. Sweetser 1990:1). For instance, this is the case with phrases such as "See you soon", which retains its original sense but is also conventionally (situationally) interpreted as a closing, i.e., a way to say good-bye to one's partner. It is the situation that is responsible for the charge of the formula-specific pragmatic property to this expression. If used out of context these phrases behave like ambiguous sentences because there is nothing that can distinguish the SBU from its freely generated counterpart. For instance, both the figurative and

122

Chapter 6

literal meaning of "get out of here" and "come on" are equally salient. When native speakers of English were asked to interpret "get out of here" without context in an experiment conducted by Kecskes (2001), 47% of the subjects interpreted it as "kidding" and 53 % responded with "leave". These percentages suggest that the expression has at least two almost equally salient meanings. This is when, according to the graded salience hypothesis, parallel processing occurs and the role of context becomes decisive. In the following situations (84/A) and (85/A) demonstrate the literal use of the SBUs while (84/B) and (85/B) stand for the figurative use of the expressions. In both cases the immediate context points out the appropriate meaning of the given expression which is charged with a FSP. (84/A) - Get out of here. I don't want to see you any more. (84/B) - John, I think you really deserved that money. - Oh, get out of here. (85/A)

- Come on, Jim, we will miss the plane.

(85/B)

-- ICome think on, I can't withplenty you. of time. yougo have

2.3. Loaded SBUs SBUs that are very frequently used such as greetings, closings, apologies and other rituals become delexicalized relatively easily. They lose their compositionality and are not transparent semantically any more. Their pragmatic function is more important than their original literal meaning which is difficult to recall if needed. They are closer to phraseological units than to other types of idioms. Loaded SBUs have formula-specific pragmatic properties which are present even if no particular subevent of a script is supplied. These SBUs are "loaded" with their new function which remains there, and is not dependent on the situation because it is encoded in the expression as a whole. They are pragmatic idioms whose occurrence is very strongly tied to conventional, frequently repeated situations. We think of a particular situation even if we hear the following expressions without their routine

Classification and interpretation

123

context: "Howdy"; "Please help yourself', etc, because their most salient meaning comprises the formula-specific pragmatic property as well.

3. Interpretation of SBUs Salient meaning of SBUs plays an important role in the classification described above. Therefore it is essential to investigate how the graded salience hypothesis works for native and non-native speakers of English. In a attempt to explore the validity of the graded salience hypothesis in L2 learners' language production, Kecskes (2001) conducted a cross-sectional study with 30 native speakers of English (NSs) as a control group and 51 non-native speakers of English (NNSs) from 12 countries who had studied English as a foreign language in their home country for at least four years and had spent from six months to two years in the target language environment. In the second part of the experiment students were expected to demonstrate their competence in the use of situation-bound utterances (SBUs) in three different tests. An analysis of the results of these tests will be used here to demonstrate the differences between native speakers and non-native speakers in the interpretation of different types of SBUs. Test 1. In Test 1 students were asked to describe the meaning of situationbound utterances without context: S/l. Take it easy. S/2. Piece of cake. S/3. Get out of here. S/4. Why don't you sit down? S/5. I'll talk to you later. S/6. Welcome aboard. S/7. Stick around. S/8. Tim is going out with that blonde. S/9. Please hear me out. S/10. I am sorry, I didn't get your name. S/l 1. Hang on for a second.

124

S/12. S/13. S/14. S/15.

Chapter 6

You are welcome. Help yourself. Give me a break. Give me a hand.

Test 2. In Test 2 students were given dialogs with SBUs most of which have both a literal and a figurative meaning as salient. The participants were expected to recognize the most salient meaning in the given situation. Some expressions from the previous list (S/2, S/3, S/5, S/6) were repeated here to find out how student responses change if the expressions are used in a situational context.

S/16. - Bill, I don't think I can agree with you. - OK, shoot. S/17. - Frank, I think you really deserved that prize. - Get out of here. S/18. - Hurry up. We will miss the train. - Come on, we have plenty of time. S/19. - Jim, do you think you can repair this coffee machine? - Piece of cake. S/20. - John, please clean up this mess on the desk. - Oh, give me a break, will you? S/21. - David, let me introduce you to our new secretary, Sally. - Nice to meet you, Sally. Welcome aboard. S/22. - I must be going now. See you soon. - OK, I'll talk to you later. S/23. - I think John was really rude to you yesterday. - Tell me about it.

Classification and interpretation

125

S/24. - Thank you very much for your kindness. - You bet. Test 3. The last test focused on SBUs in which the figurative meaning is usually the most salient meaning. Some of these expressions (S/25, S/29) are rarely used in their literal meaning; others (S/26, S/27), however, quite frequently keep their original meaning. In the dialogs below these formulae were given in their literal meaning, and participants were expected to consider the literal meaning of the expressions as the intended meaning in the given situation. S/25.

- Bill, you must show me how to use this word processor. - Sorry, I am busy. You'll have to help yourself.

S/26. - Jim, can I spend the night in your apartment? It's, kind of, too late to go home now. - You know you are always welcome. S/27. - Mary, why don 'tyou sit down? There is a chair over there. - That's OK, I prefer to stand. S/28.

- Andy, can you tell me why you left me in the store without a word? - Not now. Don't you see I am busy. I'll talk to you later.

S/29.

- What do you want me to do? - Stick around here until I get back.

S/30.

- 1 feel like going to town. - I'm not going out with you now. I have a lot to do.

S/31. - Tom, get me out of this hole, will you? - OK, give me a hand. Salience is primarily a socio-cultural rather than a linguistic phenomenon which is based on particular knowledge structures that are

126

Chapter 6

language- and culture-specific. Not having those knowledge structures in their Common Underlying Conceptual Base (Kecskes 1998;Kecskes and Papp, 2000a) NNSs were expected to have difficulty identifying the salient meaning of an expression in the L2, and consider the compositional and literal meaning of the utterance as salient rather than its pragmatically extended consense. The analysis of the results has revealed the following: Test 1. In the first test SBUs were listed without context. According to the GSH for information to be salient, i.e., to be foremost on one's mind, it needs to be stored or coded in the mental lexicon (Giora 1997; 2002). Stored information enjoys cognitive priority over unstored information such as new information or information inferable from context. It follows that it is the most salient sense of a word or expression that is directly computable from the mental lexicon before any extra inferences based on contextual information are put into play . This approach as explained in chapter 3 claims that context has a limited role in the activation of salient meanings because salience is primarily a matter of convention, frequency, and familiarity. Research ( Rayner, Pacht, and Duffy 1994, Giora 1997; 2002) has suggested that even when prior context is heavily biased in favor of the less salient (e.g. less frequent, familiar and conventional) meaning of an ambiguous expression, salient meaning is activated first. So salient meanings can-not be bypassed in LI processing. Consequently, this test aimed to reveal how stored information about the same phenomenon differed between native speakers and non-native speakers. Familiar metaphors and frequent SBUs often have both a figurative and a literal meaning that is salient. They activate both their literal and figurative meanings simultaneously if both meanings are equally salient. In the first test most expressions could have more than one salient meaning. However, in most cases the figurative meaning was expected to be considered salient. The subjects' responses are summarized in Table 1.

Classification and interpretation

127

Table 1. SBUs without Context NSs

NNSs

SBUs FIGURATIVE

LITERAL OR F2

FIGURATIVE

LITERAL ORF2

relax 83%

bye 17%

relax 75%

bye 18%

7%

Piece of cake.

100%

0%

90%

6%

4%

Get out of here.

kidding 47%

leave 53%

kidding 8%

leave 82%

10%

Why don't you sit down.

sit down 83%

let's talk 17%

92%

0%

8%

I'll talk to you later.

bye 70%

else 30%

75%

18%

7%

Welcome aboard.

soc.accept. 90%

ship 10%

soc.accept. 61%

airp. 22%

17%

Stick around.

100%

0%

73%

25%

2%

Tim is going out with that blonde.

dating 80%

derog. 20%

69%

6%

25%

Please hear me out.

listen 80%

don't interr. 20%

80%

0%

20%

Take it easy.

I am sorry, I didn't get your name.

96%

4%

76%

10%

MISINTE RPR

14%

128

Chapter 6

Hang on for a second.

100%

0%

88%

4%

8%

You are welcome.

no problem 83%

guest 17%

no problem 75%

guest 6%

19%

eat 41%

do yourself 29%

Help yourself.

Give me a break.

Give me a hand.

87%

13%

disagree 47%

stop kidding 53%

rest 31%

leave me 27%

42%

98%

2%

96%

2%

2%

30%

Both the NS and NNS responses support GSH. The majority of respondents found the most salient meaning without context. The numbers demonstrate how important the frequency of encounters with the expression is for both NSs and NNSs. Even the NSs made their decisions based on frequency when two salient meanings were possible. For instance: "Take it easy", "I'll talk to you later". Where parallel processing was needed because two meanings were equally salient (for instance: "Get out of here", "Give me a break"), NSs were divided in their decision since there were no contextual cues to rely on. NNSs, however, preferred literal meaning to figurative meaning, which makes sense, and supports our hypothesis about NNS language processing which usually prioritizes literal meaning. The expression "Welcome aboard" is a good example to demonstrate how much NNSs rely on their experience, and how important familiarity with the expression is when there is no context. A significant number of NNSs (22%) identified "Welcome aboard" with the expression they hear from the captain on an airplane. In some cases there are significant differences between NSs and NNSs. The expression "Give me a break" was interpreted by NSs as "I disagree with you" (47%), or "Stop kidding" (53%). The NNSs, however, thought that the intended meaning was either "Let me rest"

Classification and interpretation

129

(31%), or "Leave me alone" (27%). This also demonstrates thatNNSs are usually more comfortable with a meaning that is close to the compositional meaning of the expression. The fact that NSs and NNSs do not share a common socio-cultural background also led to different interpretations of one and the same expression. Quite a number of the NSs (20%) said that the expression "Tim is going out with that blonde" was derogative. None of the NNSs thought that way. It is worth noting though that this expression is not a fixed one and only the last part ("that blonde") may be responsible for the NSs feeling that it is derogative. Test 2. In the second test, context was biased for the figurative meaning of SBUs no matter whether the figurative meaning or the literal meaning of the expression was salient or both such as in S/16 ("OK, shoot."), S/17 ("Get out of here."), S/18 ("Come on".). There was absolutely no difference in how NSs processed these expressions so their results are not given here. The common interpretation of NSs is essential because the context was given this time. When parallel processing was expected to be applied because both the literal and figurative meanings of the expressions were equally salient, context played a decisive role. However, this was not necessarily the case for NNSs. The table below shows that their interpretation was based on frequency ("piece of cake", which is very frequently used in everyday conversation was processed 100% correctly), and contextual cues did not always help them to find the correct salient meaning if the salient meaning was not the literal meaning. For instance: S/16 ("OK, shoot."), and S/17 ("Get out of here.")

130

Chapter 6

Table 2. SBUs with Figurative Meaning SBUs

FIGURATIVE

LITERAL OR F2

MISINTERPRETATION

OK, shoot.

55%

0%

45%

Get out of here.

shut up 39%

leave me 24%

37%

Come on.

76%

10%

14%

Piece of cake.

100%

0%

0%

Oh, give me a break, will you?

don't bother 61%

want rest 18%

21%

Welcome aboard.

70%

5%

25%

I'll talk to you later.

84%

14%

2%

Tell me about it.

agree 49%

talk about it 22%

29%

You bet.

92%

0%

8%

More specific and less frequent and familiar expressions resulted in low scores such as "OK, shoot", and "tell me about it". It is interesting to compare the responses between two situations in Test 1. and Test 2. The presence of context in Test 2 changed the interpretation of these two expressions significantly.

Classification

and interpretation

131

"Get out of here" Test 1. S/3. "Kidding": 8%

"Leave": 82%

Misinterpr.: 10%

Test 2. S/17. "Shut up": 39% "Leave": 24%

Misinterpr.: 37%

S/17. - Frank, I think you really deserved that prize. - Get out of here. In fact the context did not seem to help NNSs interpret the SBU. When no context was given they stuck to the literal meaning of the expression. In context, however only a relatively small (39%) percentage understood the figurative meaning and made the correct interpretation. "Give me a break" Testi. S/14. "Rest": 31%

"Leave me": 27%

Test 2. S/20. "Don't bother me": 61% Misinterpr.: 21% S/20. -

Misinterpr.: 42%

"Want rest": 18%

- John, please clean up this mess from the desk. Oh, give me a break, will you?

Although NNSs hear the expression "give me a break" very often in everyday conversation they do not seem to understand the cognitive mechanism that is responsible for the meaning of this SBU. No salient meaning emerged as dominant in Test 1. In Test 2 the context was supportive but still the percentage of incorrect responses is too high: 39%. This shows that the coresense of the SBU is not quite clear for a significant number of NNSs. Test 3. In the third test SBUs were used in their less salient meaning which was their original literal meaning. Generally it is very difficult to find situations where an SBU which has already become a pragmatic idiom is used in its original literal meaning. This is what Gibbs (1980) called "literal uses of highly conventional expressions." Responses of NSs

132

Chapter 6

demonstrated that the less salient meaning was processed sequentially with no problem if the context was clear. In a couple of cases, however, where the situation was open-ended, and the context was unclear, sequential processing did not always occur in the responses of participants because some NSs and many NNSs directly processed the most salient meaning. From this respect the S/31 "Give me a hand" situation deserves special attention. S/31 - Tom, get me out of this hole, will you? - OK, give me a hand. The most salient meaning of "Give me a hand" is "Help me", that is to say, help me, the speaker. 24% of the NS respondents and 45% NNSs interpreted "Give me a hand" in that way: "Help me so that I can help you". Most of the NSs (76%), however, felt that the situational meaning was "Extend a hand (and I will pull you out of the hole)", which is not supposed to be the second salient meaning of the expression. In the case of the expression "I'll talk to you later" both the literal and figurative meaning can be salient. A part of the respondents interpreted the SBU literally: "When I have time I will explain to you why I left you in the store" while others understood it differently and gave the following interpretation: "Don't bother me now". Table 3. SBUs with Literal Meaning NSs S Β Us

Literal

Help yourself.

100%

You are always welcome.

100%

NNSs

Figurative.

Literal 100%

-

-

96%

Figurative -

Misintr.

-

-

4%

Classification and interpretation

Why don't you sit down?

100%

I'll talk to you later.

explain 57%

Stick around. I am not going out with you now. OK, give me a hand

-

don't bother 43%

100%

explain 57%

100%

-

don't bother 35%

133

-

8%

100%

busy 100% extend a hand 76%

help 24%

84%

10%

extend a hand 41%

help 45%

6%

14

%

The study confirmed the validity of the hypothesis which claimed that salience is language- and culture-specific and derives from certain knowledge structures that are essential parts of the conceptual base of both native and non-native speakers. The more frequent, familiar, conventional and prototypical a certain piece of information in a particular linguistic community is the more salient it is for that community. Salience is a very important part of the socio-cultural heritage of NSs. It is something that reveals a unique feature of the mind: Giora (2002) argued that rather than accessing the context-compatible information first, our mind opts for the most accessible information. When it fails to integrate with the context, we access the next available interpretation. It is this most accessible information that is culturespecific and makes language processing difficult for NNSs because they do not have direct access to the most accessible (for the native speakers) information in the target language since it is not stored or coded in their LI-dominated mental lexicon. What is stored and encoded there is what is salient in the native tongue and culture of language learners, and that usually does not work in the target language. NNSs can hardly apply the principle of salience in the target language. The lack or low level of conceptual fluency in the L2 forces NNS to rely on linguistic signs rather than conceptualizations while processing L2. The study demonstrated this very clearly. NNSs usually had no difficulty identifying the compositional meaning of SBUs. Problems occurred when literal meaning was not the most salient meaning. Then

134

Chapter 6

NNSs mapped target language expressions on LI conceptualizations, which often resulted in misinterpretation of expressions. Using the principle of salience, LI speakers processed figurative meanings directly without falling back on literal meanings. Adult L2 speakers, however, could hardly ignore the literal meaning even if the context required otherwise. This confirms the claim that the most salient piece(s) of information is processed first. For NNSs it is the literal meaning of the target language expression because they are likely to have encountered that first (and most frequently) when they studied English as a foreign language in their home country. Only slowly and gradually can this interpretation strategy be modified even if the learner had a longer stay in the target language country. The GSH claims that context affects comprehension after highly salient information has been accessed (Giora 1997; 2002). There is empirical research that supports the hypothesis that available information is accessed initially in LI processing, regardless of contextual fit or speaker's intent (Keysar 1998; Keysar, Barr, Balin and Paek 1998). This is not so in L2 processing where contextual cues seem to have priority over salience for the reasons discussed above. It is usually the linguistic context that NNSs rely on, and this is a direct consequence of the compositional interpretation of words and expressions in the target language in instructed L2 acquisition. The GSH was developed to explain LI processing. However, its application in second language acquisition can also be very useful since it may reveal and explain several important features of L2 language use and processing, which can contribute to better understanding of what goes on in the multilingual mind.

Chapter 7 Creativity and formulaicity

1. Creativity on sentence and utterance level The frequency of formulaic expressions calls for the review of the claim that language has to be seen as a formulaic-creative continuum (Bolinger 1976; Wood 1981; Pawley and Syder 1983; Langacker 1987; Nattinger and DeCarrico 1992). Coulmas pointed out that "in the enactment of verbal routines the creativity of language is socially canalized according to successful solutions of recurring verbal tasks, fixed by functional appropriateness and tradition (Coulmas 1979:3)." Based on recent research (c.f. Bahns et al. 1986; Cowie 1988; Rescorla and Okuda 1987; Nuyts, 1992; Sinclair 1991) Kecskes (2000a) argued that the formulaic-creative dichotomy makes sense at sentence level but it loses its significance at discourse level because sentence production cannot be equated with language production. Discourse level as opposed to sentence level requires a more complex understanding of creativity that is to be regarded as an interplay of grammatical rules, functional adequacy, situational appropriateness, stylistic preference and norms of use (Coulmas 1981a). Language is obviously organized both on the discourse and sentence levels with no opposition in evidence. Creativity was crucial in Chomsky's conception of what a linguistic theory should be able to explain: " an essential property of language is that it provides the means for expressing indefinitely many thoughts and for reacting appropriately in an indefinite range of new situations (Chomsky 1965:6)." Chomsky's original understanding of creativity was quite narrow: ".. the grammar of a language must, for empirical adequacy, allow for infinite use of finite means, and we assigned this recursive property to the syntactic component (Chomsky 1972:155)." Nuyts (1992) argued that this quantitative concept of creativity is based on recursivity which concerns the possibility to expand structures in length by repeating a structural element in its own internal pattern.22 Real creativity in linguistic behavior means much more than that. Language use does not stop at sentence level. When language is used for communication or for other purposes it usually works with units which are qualitatively and quantitatively more complex than sentences:

136

Chapter 7

oral and written texts or discourse. Creativity in the production of these higher level units can be regarded as an interplay of grammatical rules, functional adequacy, situational appropriateness, stylistic preference, and norms of use (Coulmas 1981a:6). Chomsky (1974) acknowledged the existence of two types of creativity - one which is called "mathematical" or "formal" (the recursive property of grammars), and the other which he calls the "creative aspect of language use". Realizing, however, that "the creative use of language should be none of the business for a mechanism which does not say how language is used" (Nuyts 1992:110) Chomsky referred creativity to the mysteries of performance. He wrote: "When we ask how humans make use of these cognitive structures, how and why they make choices and behave as they do, although there is much that we can say as human beings with intuition and insight, there is little, I believe, that we can say as scientists. What I have called elsewhere 'the creative aspect of language use' remains [..] a mystery to us (Chomsky 1975: 138)." For him textual creativity is not part of linguistic competence in the narrow sense; it belongs to the level of what he calls "creative aspect of language use". Chomsky (1974:28) argued that "the recursive property of generative grammars provides the means for the creative aspect of language use, but it is a gross error to confuse the two, as some linguists do." From a generativist perspective creativity is rule-governed. Chomsky (1974:29) argued that "creativity . . . presupposes a framework of rule", and that "if all constraints are abandoned, there can be no creative acts" (cited by Nuyts 1992:111-112). When Chomsky speaks about this rule-governed creativity he has syntactic creativity in mind. But human beings can be extremely creative in a rule-changing, and rule-breaking manner: this is the essence of much of literary creativity and discourse creativity. Indirect speech, implicatures, and flouting of maxims of the Cooperative Principle (Grice 1968) are all based on some rule-breaking conversational activities. Pragmatic acts often require us to use the language in unconventional ways which cannot be done without creativity. Pragmatics tells us to be as creative with language as possible. Mey (2001) argued that it is not a problem if we let ourselves be 'semantically shocked' if there is a reason for our action, and as long as we know, as language users what we are doing. The following examples (source: Levinson 1983) illustrate very well how creative we can be in language use:

Creativity and formulaicity

137

(86) Joke: 'what is said ' instead of 'what is communicated ' I came home last night and there was a car in the dining room. I said to my wife: - "How did you get the car in the dining room?" She said: - "It was very easy. I made a left turn when I came out of the kitchen." (87) Implying no wish to give an opinion Alison: - 1 do think that Mrs. Jenkins is an old windbag, don't you? Tim: - Huh, lovely weather for March, isn't it? (88) Indirect response Amy: Brian:

- How is John doing at his job at the bank? - Oh, quite well. He likes his colleagues and he has not been sent to prison yet.

Nuyts acknowledged that there is a need for creativity to follow a certain order — otherwise it would not be comprehensible (Nuyts 1992:112). But at the same time he argued that creative language use is the optimal exploitation of the structural possibilities available to reach all kinds of goals in as efficient a way as possible (Nuyts 1992: 111). This approach to creativity bridges the gap between structural well-formedness and communicative appropriateness, both of which are needed in the use of formulaic expressions. The only problem with this perspective is that after all lexical units require particular structures and not the other way round. Consequently, creativity begins with lexical choices which require particular structures to create a particular context or to fit into a given situation. No appropriate use ofSBUs is possible without the ability to communicate in a socially acceptable manner

138

Chapter 7

and/or to produce a coherent and cohesive text. But this creativity is governed by pragmatic rather than syntactic rules. Fixed expressions are already formed, so they need not be generated online. Creativity means using an appropriate utterance in a given situation and connect it with other utterances (generated or fixed) into a text or conversation both of which are higher level units of language than sentences. This creativity requires the ability to create coherent and cohesive text (either oral or written) from sentences/utterances according to the requirements of a given situation and a speaker's need. Breaking the rules is an important part of this creative process. As demonstrated in (86), (87) and (88) above, rules of conversation can be broken in a creative way. This cannot be said about sentence level where rules must be abided otherwise the result of sentenceformation is incorrect. Bouveresse (1974) claimed that the truly creative faculty is located on the level of the production of ideas. Use of language is always a creative process in text and discourse production (both oral and written). The creative-formulaic dichotomy loses its significance in discourse because any kind of "putting together" is a creative activity. Constraints on creativity can only be formulated if the functional dimension is taken into account. Speakers producing a text always have to evaluate, select and "put together". Differences at sentence and discourse level are only in the nature of creativity. To generate a sentence from selected words requires mainly grammatical creativity. Development of a text or conversation needs a qualitatively different, more complex type of creativity which has grammatical, logical and sociocultural aspects (Kecskes 2000a). In language use these aspects are all present at any given time but not to the same extent. Even when we use the most routinized expression we have to make a choice. This choice, however, is motivated by the socio-cultural aspect of creativity. Grammatical and logical aspects have several objective constraints while the socio-cultural aspect is more flexible and has more subjective and individual elements. Rules of communication can be canceled almost at any time if required by the speaker while this is not truth for the rules of grammar which are expected to be complied with at all times. Formulaic expressions constitute a very important part of this creative process. The significance of novel-sentence creating ability as opposed to the use of formulae has been overstated in the linguistic literature although Pawley and Syder called our attention to this fact

Creativity and formulaicity

139

rather early: " native speakers do not exercise the creative potential of syntactic rules to anything like their full extent, and... indeed, if they did so they would not be accepted as exhibiting nativelike control of the language. The fact is that only a small proposition of the total set of grammatical sentences are nativelike in form - in the sense of being readily acceptable to native informants as ordinary, natural forms of expression, in contrast to expressions that are grammatical but are judged to be 'unidiomatic', 'odd', or 'foreignisms' (Pawley and Syder 1983:193)." The overemphasis on novel, and full sentences has significantly influenced language teaching practice where it has done more harm than good because by expecting their students to use "full sentences" instead of chunks, patterns, and incomplete utterances that are very common in normal language production, teachers encourage students to sound unnatural. Compare the following: (89) Bob: Amy: Bob: Amy:

- Do you know where Sally lives? - In Singleton. - I s it far? - No, twenty minutes by bus.

(90) Bob: Amy: Bob: Amy:

- Do you know where Sally lives? - Yes, I do. She lives in Singleton. - Is it far from here? - No, it takes twenty minutes to get there by bus.

(90) is usually preferred to (89) by language teachers who want their students to use "full sentences" which sometimes may make people sound non-nativelike even if everything is correct in the utterances. In language production we rely on two of our abilities. We recall the prefabricated units that were stored in our mind from previous encounters with those units, and at the same time we also use our ability to create and understand entirely novel strings. These two abilities are always activated and applied according to our communicative needs. This leads to a language production where even the most customary language use

140

Chapter 7

can be changed for novel, freely generated expressions and vice versa. Instead of using the customary SBU " You are all set" after finishing business with the customer, the clerk or assistant can say something like " I think there is nothing else to attend to"; "Everything has been taken care of ", or else. Sinclair (1991) explained this process in his book where he argued that linguistic material is handled in two different ways. What he called the "open choice principle" means the selection of individual words in the sense as Chomsky understands creativity. Based on his work with COBUILD Sinclair went one step further with the formulation of "idiom principle" which emphasized the decisive role of formulaic combinations in language use: "The principle of idiom is that a language user has available to him or her a large number of semi-preconstructed phrases that constitute single choices, even though they might appear to be analyzable into segments. To some extent, this may reflect the recurrence of similar situations in human affairs; it may illustrate a natural tendency to economy of effort; or it may be motivated in part by the exigencies of real-time conversation (Sinclair 1991: 110)."

Sinclair also suggested that "the first mode to be applied is the idiom principle, since most of the text will be interpretable by this principle. Whenever there is a good reason, the interpretive process switches to the open-choice principle, and quickly back again. Lexical choices which are unexpected in their environment will presumably occasion a switch (Sinclair 1991: 114)." There is no doubt about the fact that language production uses the potential of both creativity and formulaicity. In fact, one can hardly exists without the other. Wray ( 1998:64) suggested that human language is "an uneasy compromise between a rule-based and a holistic system." Neither of these two systems can account for the structure of our output alone. As Wray said: "Without the rule-based system, language would be limited in repertoire, cliched, and, whilst suitable for certain types of interaction, lacking imagination and novelty. In contrast, with only a rule-based system, language would sound pedantic, unidiomatic and pedestrian (Wray 1998:64-65)."

Creativity and formulaicity

141

2. Role of formulaic sequences in language acquisition 2.1. Child language acquisition 2.1.1. Central role of formulaicity in L1 development Facts about human languages, the contexts in which first language acquisition typically occurs, and properties of the human brain jointly support the assumption that prefabricated chunks of language play a central role in first language acquisition (c.f. Miller and Weinert 1998; Wray 2002). Different research in different disciplines come together to support the view that first language acquisition involves the largescale use of ready-made chunks. Recent research in spontaneous spoken language has revealed that the role of formulaic expressions in language acquisition and use is more important than it has been found within the Chomskyan paradigm. Cowie (1988) reported that an analysis of authentic data yielded literally thousands of stable multi-word units. Perkins ( 1998) reviewed a range of research suggesting that children over the age of six make heavy use of formulaic language and do not construct syntactic structures online. When children make heavy use of ready-made clauses, phrases and collocations, they do exactly what adults do in spontaneous speech. Miller (1998) argued that Chomskyans ignore the fact that small children acquire language in a limited range of situations and invalidly project properties of the data in large dictionaries and grammars of English onto small children. Children can indeed create novel clauses, but they typically use the same phrases and even clauses over and over, because they deal with the same limited range of situations. Based on the analysis of the syntax and discourse organization of spontaneous speech in various languages Miller and Weinert (1998) reported that the syntactic and lexical properties of spontaneous speech are directly, even startlingly, relevant to theories of first language acquisition. They emphasized that "a fundamental distinction must be drawn between the acquisition of spontaneous speech, which is what children are exposed to as their mother tongue, and the learning of written language, which takes place in a very different fashion ((Miller and Weinert 1998:4)." Both the grammatical codes and the pragmatic factors which affect acquisition and learning are very different. The investigation of the spontaneous spoken language of very young children by Lieven, Pine and Baldwin

142

Chapter 7

(1997) revealed prominence to patterns built on particular lexical items and to recurring collocations of words that they classified on a scale of fixed-ness. According to Miller and Weinert (1998) the grammatical code used in spontaneous speech of adults is characterized by not only simple clausal and phrasal syntax, but also by a large body of ready-made phrases and collocations. These chunks extend from completely fixed formulas to structures that allow limited alterations. Recent research (i.e. Miller and Weinert, 1998, Wray, 1998, 2002; Perkins, 1998, Kecskes, 1998b) demonstrated that both children and adults heavily rely on prefabricated units in spontaneous language production. This requires the revision of the role of these expressions in both first and second language acquisition because the centrality of fully productive, complex rules is questionable. These studies argue for the pervasiveness of simple, unintegrated syntax with units at clause level and below, limited productivity rooted in the behavior of specific lexical items, preferred rather than novel sequences, as well as the "usual" stock of idioms and social routines (Weinert 2000). This affects the view of how we conceptualize the learning task, the possible end point(s) of learning, and especially the nature and role of rules. Miller and Weinert (1998) suggested that rather than starting with a powerful, global system and then limiting its use, we may postulate a simple system which is expanded conservatively to meet local demands. So the main issue is how learning from data is possible and how language may develop from a system of limited productivity. The idea that children learn from exemplars leads to assuming the storage of chunks and prefabricated units, and the importance of formulas in language development. The question is explored explicitly in research by Lieven, Pine and Baldwin (1997), Pine and Lieven( 1993), Pine and Lieven (1997), Rowland and Pine (1998), Tomasello and Brooks (1999), Tomasello (2000). These studies report several important observations: 1 ) there is a relationship between frozen units, vocabulary development and early productivity, 2) children use unanalyzed chunks to varying degrees, 3) there is evidence of a relationship to productivity which is lexically based,

Creativity and formulaicity

143

4) there is evidence of distributional learning rather than syntactic or semantic learning in early development, 5) there is a lack of abstract structures which would justify generativity (Weinert 2000). In the early stages of language acquisition children have very little or no notion of the word or the morpheme as a separate unit. They have a holistic approach to linguistic units. Considering formulaicity as central to early stages of first language acquisition Cruttenden (1981) suggested that children's approach to all aspects of language comprises two stages: first "item-learning" and then "system-learning". He said the following about the item-learning stage: "Compound words and noun phrases involving several morphemes are commonly learned by children as unanalyzed wholes. "Cup of tea" be learned as one unit and, when the plural morpheme has been learned, it may be pluralized as "cup of teas ". The child has learned the group of adult morphemes together with their reference, but has not learned that paradigmatic substitutions are applicable to each part... He has not learned to segment the expressions, nor has he learned the possibilities of commutation for any part of the expressions (Cruttenden (1981: 84)." Cruttenden explained his model in the following way: " an utterance must be assumed to be item-learned if none of its parts is used in different combinations. Thus the child who said "here he comes " used neither "here ", "he ", nor "comes " in any other utterances. Not were there any examples of similar structures (Cruttenden 1981: 84)."

2.1.2. Role of economy Several researchers (c.f. Hymes 1972; Wray 1998; 2002) emphasized the decisive role of economy in language development and production. There is psycholinguistic evidence that fixed expressions and formulas have an important economizing role in speech production (c.f. Miller and Weinert 1998; Wray 2002). Locke (1997) argued that formulaic language gives infants a set of 'starter' utterances that they need in restricted contexts to participate in adultlike social interactions. According to Nelson formulaic constructions "allow children to say more and more completely what they mean than they would if they had to construct and utterance from scratch (Nelson 1981:181-182)."

144

Chapter 7

Although at a later stage of the acquisition process the developing competence of the child makes it possible for him/her to create novel utterances the importance of prefabricated units does not fade away because the processing overload increases rather than decreases (Wray 2002). Besides, ready-made expressions play an important role in helping the child to learn and produce new forms. They serve like raw material for the analytic process that results in the development of grammar. Liven et al. argued that these memorized chunks "may provide the child with templates, which, although initially unanalyzed, have the potential for further subsequent analysis and conversion into lexically-based patterns (Lieven et al. 1992:307)." The development of grammar is supported by the processes of observation, analysis and generalization. The child recognizes two important things: 1) some part of a chunk is used in another utterance, and 2) one and the same chunk is used in different situations (c.f. Cruttenden 1981). This observation will enable the child to detach the pattern from its connection in context, and analyze and generalize it into regular syntactic rules (Nattinger and DeCarrico 1992). However, there are researchers who refute the idea that formulaic sequences play an important role in language development. Bates et al. (1988) considered formulae as useful communicational fillers which disappear once the grammar is at place. For Plunkett (1993) formulaic chunks in child language are the temporary product of erroneous segmentation. He thinks that when the child reaches the developmental stage at which he is capable of isolating and combining words, formulaic units will decline. Wray (2002:122) made an attempt to reconcile the two views by proposing that prefabricated strings can be analyzed, but only are on a piecemeal basis, and often not at all. Recent research has pointed out that child language acquisition is not just simply gradual development from gestalt to analytic. The nature of the acquisition process varies by individual learner differences and the need for analysis. Tomasello and Brooks assumed that "young children come equipped to move in either direction - part to whole or whole to parts...All children probably use both processes to some extent in different aspects of language acquisition (Tomasello and Brooks 1999: 166)." There are considerable individual differences between children in the use of the two processes. Wray (2002) argued that 'analytic' children focus on word-sized chunks very quickly, and use these as building blocks for their grammar while 'gestalt' children prefer longer sequences, which they will then

Creativity and formulaicity

145

have to break down. However, not all formulaic strings will be analyzed and broken down into their constituents by the child. Wray (2002:131) introduced the idea of 'needs-only analysis', postulating that children will simply analyze whichever strings need analyzing, to the extent that they need to, and no further. So word strings are treated holistically unless there is a specific reason to break them down, and even then, they are only analyzed to the minimal extent necessary. She built upon Bruner's (1983) assumption that children have a socio-interactional agenda rather than a grammar-learning one. Consequently, language is developed through a process of 'pattern detection' in a communicative context and not by mapping input onto innate grammatical concepts. This means that the child would prefer analysis only as a means of making the language communicatively pliable. Need-only analysis will result in the storage of simple and complex units with overlapping content in the mental lexicon. This is in compliance with those theories which postulate that the lexicon contains both complete formulaic expressions and their components parts as separate entries as described in chapter 6.

2.2. Second language

acquisition

In order to learn more about the role and use of formulas (including SBUs) in L2 acquisition it is essential to review what second and foreign language23 research has to say about the issue. L2 acquisition differs from LI acquisition to a significant extent. This is especially true if the target language culture is not present to support the acquisition process as is the case when the L2 is learned in a classroom environment. Vygotsky (1962) argued that first language and foreign language develop in reverse directions. In the mother tongue, the primitive aspects of speech (ready-made formulaic units) are acquired before more complex ones (grammatical rules). Foreign language shows the opposite development: The higher forms develop before spontaneous speech because foreign language learning is usually a conscious, task-oriented process based on controlled operations. Vygotsky (1962:109) said that "the child's strong points in a foreign language are his weak points in his native tongue, and vice versa." However, we have to be cautious with what Vygotsky said because he always talks about foreign language learning rather than the acquisition of an L2 in the target language environment.

146

Chapter 7

Although changing speaker behavior and growing interest in formulaic speech have had some impact on second language research, the view of L2 acquisition as a product of rule formation still prevails (Larsen-Freeman and Long, 1991, Weinert, 1995). Until now, most of the discussions on routine formulae have focused on their role in the acquisition of syntax: researchers have asked how learners go about analyzing those strings they pick up and memorize as wholes (Bahns et al. 1986). But the acquisition of an L2 is much more than just the development of a new system of signs, and the development of a dual language system can be understood only through a thorough analysis that focuses not only on the structure of utterances but also on their functions (Kecskes and Papp 2000a). When discussing the role of formulaic sequences in L2 acquisition we need to pay attention to several types of differences: - difference between the L2 and LI acquisition process, - differences between naturalistic and classroom environment, - differences between child and adult L2 learning, - difference in individual learners, - difference in the role of formulas.

2.2.1. Difference between the L2 and LI acquisition process There are two approaches to the role of formulaic expressions in L2 development. One suggests that just like in LI formulaic speech is the basis for creative speech and the emergence of grammar in L2 development (as understood by generative grammarians). According to the second view formulaic expressions do not participate in the development of L2 grammar. Several researchers (Clark 1974; W. Fillmore 1976; Pawley and Syder 1983; Bygate 1988) believe that second languages are acquired through a process by which learners acquire unanalyzed "chunks" of language (such as Bygate's "satellite units" or Pawley and Syder's "sentence stems") as formulae. Over time, they learn to break these units down into their constituent elements and com-

Creativity and formulaicity

147

bine them with other constituents in a variety of rule-bound ways. Wong Fillmore (1976) claimed that formulaic sequences are a major point of access for identifying component forms. Bygate (1988) argued against the notion that one moves from syntax to discourse. In fact, he demonstrated that a great deal of interactional talk consists not of complex grammatical structures, but of what he calls "satellite units". Satellite units are moodless utterances which lack a finite verb or verb group. They can consist of any dependent syntactic element including noun, adjective, adverb and verb groups, prepositional phrases, pronouns and subordinate clauses. An example from Bygate's data: Prepositional phrase: S12: at the door SI 1 : yes in the same door I think S12: behind the man who is leaving S i l : behind him Adjective group: S8: aha they're very polite S7: polite really polite that's one of their characteristics On the opposite side, Bohn (1986) claimed that formulaic language plays only a minor role in L2 acquisition and formulae in his data reflect the learners' current stage of morphosyntactic complexity, and they do not interact with syntactic development. Weinert ( 1995) also noted that basically there is no direct evidence to prove that chunks are related directly to subsequent target-like performance. Krashen and Scarcella (1978) arrived at a similar conclusion but for a different reason. They believe that memorized utterances and creative speech are produced in ways that are neurologically different and that, therefore, there can be no interface between them. They distinguish between 'routines ' and 'patterns ', to refer respectively to whole utterances learned as memorized chunks (e.g. 'Nice to see you'; 'Have a good day') and to utterances that are only partially unanalyzed and have one or more open slots (e.g. 'Can I have a...?; 'Would you like to...?'). Several studies (c.f. Shapira 1978; Schmidt 1983; Yorio 1989) conducted in an L2 environment found that the subjects's use of formulaic sequences was minimal.

148

Chapter 7

Granger argued that formulaic sequences are not analyzed: "it would ...be a foolhardy gamble to believe that it is enough to expose L2 learners to prefabs and the grammar will take care of itself (Granger 1998: 157-158)." Consequently, in adult naturalistic learners there is no compelling evidence of a type of progression that was identified for first language acquisition, i.e., where formulaic sequences were used initially as an aid to communication, then a process of segmentation supported the development of grammatical competence. Wray argued that "where the first language learner starts with large and complex strings, and never breaks them down any more than necessary, the postchildhood second language learner is starting with small units and trying to build them up (Wray 2002:206)." There is also evidence that non-native speakers use formulaic sequences for different purposes than native speakers. DeCock et al. ( 1998) reported that advanced adult L2 learners in some cases used even more prefabricated units than native speakers. They applied the idiom principle, but the prefabricated units they used were not necessarily the same as those used by the native speakers, they were not used with the same frequency, had different syntactic uses and often fulfilled different pragmatic functions.

2.2.2. Differences between naturalistic and classroom environment As argued earlier formulaic language can be considered as a dynamic response to specific communicative needs. In first language acquisition social-communicative needs requiring a comprehensible message that triggers the use of unanalyzed chunks. However, in second language development both the linguistic means and the needs they have to express are different, especially in classroom setting. Wray claimed that "classroom learners are rarely aiming to communicate a genuine message with a beneficial outcome to their physical, intellectual or emotional state, so there is no drive to use formulaic sequences for manipulative purposes (Wray 2002: 205)." There is no doubt about the fact that a naturalistic environment where the target language culture is present encourages a more holistic approach to language than the classroom. Willett (1995:499) argued that the real issue is how formulaic expressions "enable learners to construct positive identities and relations and manage competing agendas." The use of formulae is considered to be the key to socialization in a

Creativity

and formulaicity

149

naturalistic environment where using a fixed sequence before it is broken down into its constituents is often a necessity rather than a choice. Hakuta argued that "if learners always have to wait until they acquire the constructional rules for forming an utterance before using it, then they may run into serious motivational difficulties in learning the language, for the functions that can be expressed...would be severely limited (Hakuta 1976:333)." Wong Fillmore's (1976) Nora used formulaic expressions to sound nativelike with the expectation of being treated like a native speaker. If, however, the communicative pressure and the desire for acculturation is not present the reliance on formulaic expressions may decrease as it was the case in a German family that Bohn (1986) wrote about. Although those German children lived in the U.S. temporarily with their parents their acquisition process was different from that of Nora due to the environment which was bilingual rather than monolingual. Bohn suggested that the fact that the children were free to use both English and German at home eliminated the need to rely on formulaic stereotypes as much as other L2 learners in monolingual environments (Bohn 1986:199). Classroom language learning encourages an analytic, from-smallunits-to-larger-units approach. As a result both children and adults seem to be able to apply the analytic techniques learned in the classroom to holistically acquired strings. In the first period of class-room learning the use of formulae is encouraged, however, later in the learning process the development and use of formulaic language becomes of secondary importance and the knowledge of prefabricated units lags beyond the desirable level. This may be due to the quality and quantity of input that learners are exposed to in the classroom (Kecskes 1995). Irujo (1986) argued that idioms are often omitted in the language materials addressed to second language learners, and even if students are exposed to idiomatic speech through the media they have no change to practice the use of those expressions. Based on her EFL corpus studies Granger (1998) argued that EFL learners rely on a small number of familiar and safe sequences and cling to these expressions which they feel confident in using. Yorio (1989) came to a similar conclusion after examining the use of idioms, formulas and collocations in the writing of ESL students who could not pass the University's Writing Assessment Test. His findings demonstrated that the students tried to use formulaic expressions but they seemed to have little formal control over these units. Yorio, however, reported an interesting finding that does not fit the picture that we have

150

Chapter 7

about EFL learners. When comparing ESL learners who had both informal and formal exposure to English with EFL learners who had no direct access to the target language culture, Yorio (1989:65) found that the EFL students demonstrated not only better grammatical skills than the immigrant group but also produced more idiomatic expressions, and their written English, despite errors, was more authentic that of the immigrant students. How could this be? Was this just a "one-time shot", or should we revise what we know about idiomatic language use of ESL and EFL students? I think we can agree with Wray (2002:186) who pointed out that there were several uncontrolled variables in Yorio's study, and the difference in the quality and quantity of the language to which the ESL and EFL students had been exposed could bring about those results.

2.2.3. Differences between child and adult L2 learning Differences between child and adult learners of L2 are mainly due to two interrelated variables: age which determines to what extent is the L1 already at place when the L2 starts to emerge, and training of the mind by education. It is commonplace that children have a more holistic approach to the L2 than adults whose analytic perspective is supported by education. Despite the significant differences in the types of formulae used, purpose of use, and strategies of use of formulaic sequences by children and adult, we can agree with Ellis's claim that, at least in the early stages of L2 development, formulaic language "figures frequently in the speech of all learners, irrespective of their age (Ellis 1994: 85)." The experience of the very young second language learner will be very similar to that of the first language learner because the conditions of interaction which support the use of prefabricated elements and make their analysis unnecessary in several cases are almost identical. For a child to get the meaning through is more important than syntactic wellformedness. There is no doubt that unanalyzed chunks can support this goal. The introduction of literacy in school, however, will result in more analysis and breaking down of word strings. The older the child is the more his approach to the new language becomes similar to that of an adult L2 learner. This approach is more likely to be word- and rulecentered rather than formulae-centered. An interesting point is made by Wray (2002:200) who proposed as a central principle of her model that formulaic sequences are broken

Creativity and formulaicity

151

down primarily for the sake of their words. This means that language learners analyze formulaic sequences not to extract grammatical information but as a source of lexical material only. As a result they store the words that occur in the formulaic sequences but not the detailed grammatical information about how the words can combine with other words. When reconstructing the same formulaic sequences at a later stage of development the learners often make mistakes because of the lack of appropriate knowledge of grammar governing the use of those words in their dual language system. The word-centered approach encouraged by reading and writing and as a result the relative balance of words to formulaic chunks in the lexicon of a second language learner will differ significantly from that of a native speaker.

2.2.4. Difference in individual learners Most researchers acknowledge the decisive role of formulae in both first and second language acquisition. However, the development of the dual language system demonstrates individual learner variety in the creation and use of prefabricated units (Kecskes 1998b). DeCock et al (1998) and Granger (2000) provided extensive native speaker and learner data which demonstrated that learners have their own preferred sequences, a point that was made by Beebe and Giles (1984) and is empirically supported by some studies such as Kasper and Blum-Kulka, 1993; Bardovi-Harlig, 1996 and Kecskes 1999. Nick Ellis (1996) suggested that successful language learning is dependent upon the language learner's having a well-developed ability to perceive and remember sequences. If the prefabricated chunks are stored in the long term memory they will support language learning and processing because they may make the identification of input easier, and since they are available for analysis they can also contribute to the development of the grammatical system. Individual learners, however, differ in the ability to remember and use word sequences, which has immediate consequences for their language acquisition. Myles et al. (1998, 1999) studied the language use of English learners of French in British secondary schools for two years (from ages 11/12 to 13/14) through regular classroom observations, interviews and paired activities. The results showed individual learner differences in segmentation. The researchers noticed a correlation between formulaicity and creativity. Those learners who were most

152

Chapter7

able to use, and who continued to use, formulaic expressions were also the most creative in the later stages (Myles et al. 1999:76). Those students, however, who had difficulties recalling the chunks learned at an earlier stage were not creative either. Wray (2002) recommended to exercise caution when interpreting these results because the researchers could not control for the effects of explicit teaching. One thing is certain: The attempt "to resolve the tension between structurally complex but communicatively rich formulas on the one hand, and structurally simple but communicatively inadequate 'creative' structures on the other hand (Myles et al. 1999:49)" divides second language learners into several different groups in which individual learner differences play a decisive role. Schmidt (1992), who is interested in what underlies fluency, suggested that even errors may be stored and retrieved as wholes. This is an interesting suggestion that concerns the issue of analysis. Once learners start to segment data in a particular language area, this may not necessarily continue to develop and may be subject to fossilization. In other words, learners may only segment up to the point of their communicative needs. To develop target language idiomaticity seems to be impossible mainly because the role and use of prefabricated units in the dual language system significantly differs from those in the first language system. The dual language system contains not only target-like units but also learner-created formulaic chunks and sequences which may or may not be used to generalize, and which may or may not result in target-like language (c.f.. Bolander 1989, Weinert 1994).

2.2.5. Difference in the role of formulae Differences in explaining the role of formulaic expressions in L2 acquisition can be due to two factors: 1) methods of data collection (Bohn 1986) and 2) the broader (W. Fillmore 1976; Nattinger and DeCarrico 1992) or narrower (Coulmas 1979,1981; Bohn 1986) understanding of formulae. A common feature of all the studies is that they focus mainly on the structural features of memorized chunks and their role in the acquisition

Creativity and formulaicity

153

of syntax. It is almost completely disregarded that the use of these formulaic expressions is strongly affected by the development of pragmatic abilities in L2 which must be distinguished from that in LI, since the acquisition of these abilities in LI is tied to the general cognitive development while L2pragmatic skills are usually modifications of LI skills (Kecskes 2000a; Kecskes and Papp 2000a). This difference between LI and L2 has a decisive effect on the development and use of routine formulae and their role in the acquisition of syntax. Bates (1976:292) argued that children acquire a number of performative or pragmatic idioms "without having the flexible, conscious control over form-function relations." L2 learners also pick up a number of ready-made units before they can synthesize expressions of corresponding complexity. There is, however, a crucial difference between LI learners and L2 learners in this respect. In LI a sense of situational appropriateness of a given formula usually precedes its understanding. LI learners supported by their socio-cultural background knowledge are usually aware of the pragmatic aspects of the use of those formulae, which is usually not the case with L2 learners. Adult second language learners usually miss the target language socio-cultural background knowledge, because they live outside the target language community. Interferential pragmatic mistakes occur in their speech whenever socio-cultural background knowledge obtained in the L1 is applied in the use or production of L2 utterances. L2 learners only think that they are aware of the pragmatic aspects of the use of the target SBUs. What they know is based on their LI socio-cultural background knowledge which is often transferred to L2 production. In order to develop target-like use of SBUs L2 learners need as much direct interaction with the L2 culture as possible. Grabois and Pavlenko (1996) found that when immersed in the target language environment, many L2 learners develop close-tonative conceptual and metaphorical competence. Without direct exposure to the target culture, however, SBUs in L2 are rather learned than evolve. Syntactic development and pragmatic development do not go hand in hand in the L2. Especially in a foreign language environment syntactic development usually precedes pragmatic development. Bahns et al. (1986:720) reported that "the increase ofL2 knowledge seems to lead away from target-oriented verbal behavior to a phase of creative but sometimes pragmatically inadequate verbalizations." L2 learners tend to handle most fixed expressions as just another syntactic unit: they split

154

Chapter 7

them into parts and often come up with their own "creations". Pawley and Syder (1983) pointed out that L2 learners often assume that an element in a expression may be varied according to a phrase structure or transformational rule of some generality, when in fact no variation (or maybe, some very restricted one) is allowed in native-like usage. This assumption of adult L2 learners, however, depends on the type of fixed expressions and the semantic transparency of the expression. With respect to the acquisition process a dividing line has to be drawn between patterns, prefabricated chunks ("satellite units" of Bygate, "sentence stems" of Pawley and Syder, "lexical phrases" of Nattinger and DeCarrico) on the one hand and "routine formulas" (Coulmas 1979; 1981), "situational utterances" (Kiefer 1995) and "situation-bound utterances" (Kecskes 1997) on the other. Patterns can support the development of syntactic rules (al-though this has never been proven) by the learner breaking them down into their constituents, and in doing so augmenting his interlanguage system (Ellis 1985). This can occur because these memorized chunks have a compositional structure and, most of them are semantically transparent: What's this?; Can I have...?; There is no...; I wanna...; I don't know; etc. Language learners can figure out how these chunks are put together from words. However, this is not the case with situation-bound utterances that are pragmatic rather than syntactic units. They can hardly support the development of grammatical competence in second language acquisition because their function is usually different from what they say. They are like idioms but their use is situationally tied, which is not the case with idioms. If broken down into constituents they could do more harm than good to L2 learners because their functional meaning can hardly be figured out from the elements they contain. Even in the case of "plain" SBUs there is usually at least a slight difference between "what is said" and "what is communicated". Consequently, SBUs can hardly be expected to take part in the development of grammatical competence in L2.

Chapter 8 Conceptual socialization

In the previous chapter it was argued that in the beginning phase of their development second language learners relatively easily pick up formulaic expressions and use them to decrease the communicative pressure. When doing so they handle these expressions as functional units rarely paying any attention to their compositional meaning. There is nothing wrong with that, actually in acting so second language learners come close to what children do when acquiring their first language. Why is it then that while in LI development the result is nativelike use of formulaic language, in the L2 that is almost out of question? What is so different in the two acquisition processes? If we want to simplify the response to these questions we can answer with one phrase "socio-cultural environment" which is responsible for conceptual socialization. I use this term to distinguish the process from "language socialization" (c.f. Ochs 1988; Willett 1995; Mitchell and Myles 1998) which has its roots in anthropological linguistics. Language socialization research represented by a limited number of studies (c.f. Ochs 1988; Ochs and Schieffelin 1984; Willett 1995; Piatt and Trudi 1997) emphasizes that language and culture are inseparable because they are acquired together: each supports the development of the other. Here is what Ochs said about this issue: "It is evident that acquisition of linguistic knowledge and acquisition of sociocultural knowledge are interdependent. A basic task of the language acquirer is to acquire tacit knowledge of principles relating linguistic forms not only to each other but also to referential and nonreferential meanings and functions Given that meanings and functions are to a large extent socioculturally organized, linguistic knowledge is embedded in sociocultural knowledge. On the other hand, understandings of the social organization of everyday life, cultural ideologies, moral values, beliefs, and structures of knowledge and interpretation are to a large extent acquired through the medium of language.... Children develop concepts of a socioculturally structured universe through their participation in language activities (Ochs 1988:14)."

Ochs' and Schieffelin' work has focused on LI development. There are only a restricted number of studies that extend the paradigm to second language acquisition (i.e. Willett 1995; Piatt and Trudi 1997).

156

Chapter 8

Willett's (1995) conducted a longitudinal study with young classroom learners of ESL in an elementary school with an international intake. Based on her results she argued that language socialization is a complex process in which participants construct and evaluate shared understandings through negotiation. This process leads to changes not only in their identity but also in social practices. Conceptual socialization broadens the scope of the paradigm of language socialization which has its main focus on language developmental issues. Conceptual socialization differs from language socialization in that it emphasizes the primacy of mental processes in the symbiosis of language and culture, and aims at explaining the bidirectional influence of the two languages in second language development. The process of conceptual socialization is strongly tied to the emergence of the common underlying conceptual base that is responsible for the operation of two language channels (Kecskes 2000a; Kecskes and Papp 2000a). The child acquiring his/her first language lives in the culture that is responsible for the development of the knowledge base, image system, and concepts which give meaning to all linguistic signs that are used in the given language. This is not the case if the target language is acquired as a second language. Age, and direct or indirect access to the target culture and environment are decisive factors. The older the language learner is, and the less the exposure s/he has to the target culture the more s/he relies on the existing conceptual system and knowledge base developed as a result of LI acquisition. New knowledge, and new information are processed through existing knowledge. The language learner can pick up an SBU or other fixed expressions as a functional unit based on observations of use. What s/he cannot pick up, however, is the "load", the culture-specific information which that expression may carry. This is what happened to me when I picked up the expression "you bet" as I described in the "Introduction". I was completely aware of its function. What I missed was the FSP and CSCP of the expression, the culture-specific cognitive mechanism that was responsible for the fact that Americans can use this pragmatic unit as one possible response to "thank you" in particular cases. In what particular cases, and why exactly in those particular cases is the use of "you bet" appropriate? That information was not in my knowledge base so I used the expression inappropriately. The acquisition of fixed pragmatic expressions in a second language requires the knowledge of the socio-cultural back-

Conceptual socialization

157

ground of the target language which is acquired through conceptual socialization.24 The process of conceptual socialization goes together with the emergence of a CUCB which is the "driving force" of the dual language system. Conceptual socialization refers to the transformation of the conceptual system which undergoes characteristic changes to fit the functional needs of the new language and culture. During the process of conceptual socialization the LI-dominated conceptual base is gradually restructured making space for and engaging with the new knowledge and information coming through the second language channel. This leads to the gradual development of a conscious awareness of how another culture is different from one's own culture, the ability to reflect upon this difference in language production, and the development of a dynamic social identity that is the reflection of the dual culture. Conceptual socialization has two sides which are inseparable: a skillside and a content-side. They are two sides of the same phenomenon. The skill-side means that conceptual socialization will be reflected in the actual language skills: structural well-formedness, language manipulation, sentence-structuring and lexical quality (Kecskes and Papp, forthcoming). Changes in both sides are qualitative rather than quantitative. The skill-side of the conceptual socialization process is measurable as it will be discussed in the next chapter. Changes in the contentside, however, can hardly be measured. They are qualitative changes in the content of what the language learner says and the way the language learner behaves in communication. The content of the language production is expected to give information about metalinguistic awareness, interactional style, knowledge base, and multicultural attitude.

1. Awareness of cultural differences Without a conscious awareness of how another culture is different from one's own culture, there is a tendency to see the differences in another culture through the prism of one's own culture. This may easily lead to the phenomenon of ethnocentrism. Norman Daniels (1975) argued that when differences are not perceived as differences, they are perceived as right or wrong. Conceptual socialization means that the second language learner can eliminate this problem due to his/her CUCB which

158

Chapter 8

operates the two language channels, and which contains the necessary information on the differences between the LI and L2 cultures that is demonstrated in the language behavior and production of the learner no matter which language s/he chooses for expressing his/her ideas in a particular situation. It is also conceptual socialization that makes codeswitching possible. Kecskes and Papp (2000) and Kecskes and Moyer (forthcoming) argued that code-switching is a natural consequence and product of the dual language system (DLS). Code-switching is characteristic of the bilingual language mode (Grosjean 1998; 2001) when both language channels are equally activated to express thoughts originated in the CUCB through either language channel according to the immediate needs of the speaker, depending on what context the speaker wants to create or in what context the speaker wants to fit his expressions or utterances. The next instance of code-switching data is taken from the Gibraltar corpus of written texts (Moyer ( 1992; 1998). In this text code-switching is used to construe the humoristic style of the text, which is a good example of the context-creating power of insertions and alternations. Lexical items and expressions are chosen to create a specific context in which the Spanish SBUs usually give a humorous tone to the conversation (translated by Melissa Moyer): (91) "Caramba con el Dia de la Hispanidad, what an invasion that was." [Good heavens with Columbus Day, what an invasion that was.]

It goes to show that the average Spaniard is not as obsessed about our Rock as some make out. Can you imagínate, dear Cloti, celebrating Hispanic Day in British Gibraltar! Just as well, y que sigan viniendo, dear Cynthia. Gosh, el Main Street parecía la Feria de La Linea con tanta gente all over the place. [Just as well, and let them keep on coming, dear Cynthia. Gosh, the Main Street looked like the fair in La Línea with so many people all over the place.]

Conceptual socialization

159

Como que los relations y el co-operation have never been better entre la gente de a pie. [Like relations and the co-operation have never been better with the ordinary citizen. ]

El que anda por las nubes is the mayor of La Linea who has broken relations with Gibraltar. [The person who is up in clouds is the mayor of La Linea who has broken relations with Gibraltar.]

Oh dear, when are ambassadors being withdrawn? If you ask my husband, he says that El Alcalde is out of touch, if not que se lo pregunte a los 30,000 who came here for the Spanish feast day. [Oh dear, when are ambassadors being withdrawn? If you ask my husband, he says that The Mayor is out of touch, if not he should ask the 30,000 who came here for the Spanish feast day. ]

You've hit the clavo on the head, my dear." [You've hit the nail on the head, my dear. ]

2. How is conceptual socialization reflected in the use of SBUs? Conceptual socialization is usually a dynamic process with ups and downs. It requires direct access to the target culture, personal motivation for acculturation and a regular use of the target language. How does this fact affect the L2 acquisition process and L2 use? Gee ( 1999:63) argued that the situated and local nature of meaning is largely invisible to us. It is easy for us to miss the specificity and localness of our own practices and think that we have general, abstract, even universal meanings. We come to think, when we have confronted no other languages, that 'sitting' is just sitting, 'drinking' is just drinking, 'over there' is just over there. In fact, the situated, social and cultural nature of meaning as described in the previous chapters often becomes visible to us only when we confront language-at-work in languages and cultures far distant from our own. This "confrontation" often occurs at

160

Chapter 8

the level of fixed expressions which are especially important in the first period of second language development. The language socialization studies underlined the importance of prefabricated chunks in the socialization process both in LI and L2 development. Ochs and Schieffelin ( 1984) pointed out that there is much direct teaching of the interactional routines ("elema") among the Kaluli in Western Samoa. Willett (1995) argued that in the first months ESL students relied heavily on prefabricated chunks which they picked up from their fluent English-speaking peers or from adults during routine events. Coulmas (1979:256-260) gave an excellent summary of difficulties L2 learners have when using routine formulae. He categorized pragmatic interferences according to the respective process or structural phenomenon giving rise to the mistake in question. A cognitive-pragmatic framework, however, requires a different approach to these pragmatic errors. This approach emphasizes that errors in the use of SBUs are the results of pragmatic rather than structural problems deriving from the incomplete process of conceptual socialization during which second language learners have to deal with the problems of culture-specific communicative functions, the formula-specific pragmatic properties of expressions and the differences in communication patterns.

2.1. Situational obligatoriness The situational obligatoriness of SBUs varies across cultures. This means that certain situations require the use of SBUs in a certain culture, which might not be the case in another culture. Americans easily can opt for a freely-generated phrase while Japanese or Turks may not do so. This maybe true the other way around. It is impossible to find a Russian, Japanese or Hungarian equivalent to the SBUs: "have a nice day", "have a good weekend", "I'll talk to you later" in English.. Non-native speakers have not only to memorize SBUs as linguistic units but also understand the socio-cultural background (cultural habits, customs, values, attitudes, etc.) these routine expressions are applicable in. This is where most of the transfer from LI occurs. Sometimes the communicative function is culture-specific which makes it quite difficult to give the functional equivalent of English "you bet", "welcome aboard", or Turkish "giilii giilü oturun " ("stay laughingly"), "gözünüz aydin " ("your eye bright"), "giile giile biiyütün " ("raise laughingly") and the like, in other languages because these

Conceptual socialization

161

phrases are the result of specific cognitive development in the given language and have specific pragmatic features that are called "formulaspecific pragmatic properties" in the DMM, and which are usually nonexisting or different in the other language(s). For instance, Turkish "gülii giilii oturun " ("stay laughingly") is used to someone who has just bought, rented, or moved into a new house. "Gözünüz aydin " ("your eye bright") is used to someone who has had the good fortune to be visited by a loved one who was far away. It is not that these situations are non-existing in other language. It is that other languages may not find it important to introduce a special SBU in those cases. Hungarians also say something when a friend or relative tells them that s/he has bought or moved into a new house. But they do not insist on the use of a particular SBU or SBUs in that situation. What they say is usually freely generated. It is hardly possible to find equivalents to these Turkish SBUs in American-English, Hungarian, Russian or other languages because they are the representatives of Turkish culture which finds it important to give voice to one's feelings in the situations described above. "Giilegüle büyiitün " ("raise laughingly") differs from the other two expressions in that it refers to an event which is considered to be joyful and very special in many cultures: birth of a child. This expression is used to the parents of a newborn infant. This situation is lexicalized differently in many cultures. Conceptual socialization means, among other things, that the L2 learner becomes aware of these differences and does not look for an equivalent SBU where there is none, and does not try to use the communicative customs of Lx in Ly.

2.2. Socio-cultural values expressed by SBUs SBUs are cognitive-pragmatic units whose development is an essential part of the overall socio-cultural growth of the LI learner. SBUs very strongly relate to socio-cultural values. According to Albert (1968) a value system represents "what is expected or hoped for, required or forbidden. It is not a report of actual conduct but in the system of criteria by which conduct is judged and sanctions applied (Albert 1968: 288)." Even the slightest differences in the socio-cultural value system may result in differences in cognitive mapping and, as a consequence, in lexicalization. For instance, guests are equally valued in American,

162

Chapter 8

French and Hungarian society. But the SBUs that are used to welcome them demonstrate interesting differences: (92) English: Make yourself at home. French: Faites comme chez vous. ('Do as [you do] at home.') Hungarian: Érezd magad otthon. ('Feel yourself at home.') Although these expressions can be considered as functional equivalents, the use of different verbs (French "faites" is the imperative of the French equivalent of "to do", and the Hungarian "érezd" is the imperative of the Hungarian equivalent of "feel") demonstrates that each language highlights something else as important in one and the same situation. Coulmas (1979:262) argued that from the perspective of values it is important to analyze the original and literal meaning of SBUs because they may be quite revealing with respect to social values, and past and present history of the given group of people. In chapter 5 a dividing line was drawn between situation-bound rituals and situation-bound routines. It is especially rituals that give us information about the value system. Coulmas ( 1979) hypothesized that lexical differentiation is correlated with socio-cultural relevance. The richness of a language in a particular type of SBU suggests that the particular domain of life that the given SBUs refer to plays an important role in the life of those people who use the SBUs. This hypothesis can be justified by several examples. For instance, as we saw, the Turkish language is full of SBUs that are connected with religious thinking, or in Japanese several situationbound rituals are attached to "omoiyari" which Lebra defines as follows: "Omoiyari refers to the ability and willingness to feel what others are feeling, to vicariously experience the pleasure or pain that they are undergoing, and to help them satisfy their wishes (Lebra 1974: 38)."

Conceptual socialization

163

2.3. Metaphorical thinking As discussed in Chapter 5, SBUs can be "plain", "loaded" or "charged". The problem of non-native speakers is that they usually have difficulty understanding the "load" or the "charge" which are the formulaspecific pragmatic properties of an SBU. NNSs generally treat SBUs as unanalyzed units knowing only little (or nothing) about the metaphorical development (source domain and target domain) or the conventionalizing process of the implicature that may have developed into an SBU. Danesi (1992:493) argued that "the programming of discourse in metaphorical ways is a basic feature of native-speaker competence." Metaphorical thinking is always present both in production and comprehension. It is metaphorical thinking that dominates our discourse activities (Lakoff and Johnson 1980). Second language learners have trouble with the use of SBUs because although they use the appropriate configurations of the target language, they think in terms of their native conceptual system. This clearly demonstrate the incompleteness of the conceptual socialization process. Charteris-Black argued that when dealing with unfamiliar senses of expressions second language learners "are likely to look initially for literal readings based on their knowledge of established senses rather than on pragmatic knowledge (Charteris-Black 2002:108)." That is why the SBU "get out of here" will always be a "rude" way to mean "stop kidding or stop pulling my leg" for Japanese, Chinese or Hungarian speakers of English. The salient meaning of this expression for NNSs is its literal meaning which sounds rude. As the study (Kecskes 2001) presented in chapter 6 showed one of the problems of second language learners is that this general principle of salience changes language by language. This requires the L2 learners to make the necessary adjustment to accommodate the new principle which is not easy for several reasons. NNSs usually cannot use parallel processing the way native speakers do. For NNSs the most salient meaning of an expression is what the phrase "says" (semantic interpretation) and not what it "communicates" (pragmatic interpretation). This is why "get out of here" sounds rude for NNSs. They should learn to interpret this meaning by inference just like the native speakers do. What L2 learners can rely on in this learning process is not only the clear understanding of the pragmatic function of the utterance but also the formula-specific pragmatic property of the expression. They should know that when an American says "I'll talk to you later" s/he does not

164

Chapter 8

necessarily mean that. This SBU functions as a closing in conversation. The meaning of the whole phrase should be experienced by the learner as a part of habitual action and situational context. Many language teachers argue that SBUs should be learned prior to and independent of their individual word meanings and internal structure. Does this mean that L2 learners should not try to understand the compositionality of SBUs? The answer is "no" because if students can figure out the meaning of an SBU by themselves or with the help of the teacher, they will have a link from the pragmatic function to the literal wording which may have reference to the etymology of the expression, which, as Coulmas (1979) claimed, can be important for understanding the target culture. It will also help NNSs learn the link between the source domain and target domain that will support the appropriate use of the SBU, which comprises not only the clear understanding of its situational function but also the understanding of its formula-specific pragmatic properties. Discussing the use and learnability of idioms Irujo (1993) and Kövecses and Szabó (1996) basically emphasize the same thing. They argue that the understanding of semantic transparency of idioms will facilitate their learning in the classroom. Kövecses and Szabó claimed that the transparency of idioms arises from knowledge of their underlying cognitive mechanisms (metaphor, metonymy, conventional knowledge). Johnson suggested that "broader inclusion of figurative language in L2 instruction would enhance the communicative proficiency of L2 students and provide them with insight into the conceptual systems of the L2 culture (Johnson 1996: 237)."

3. Communication Patterns During the process of conceptual socialization language learners are expected to familiarize themselves with the communication patterns of the target language. This is not easy because cultures have distinct preferences for structuring requests, apologies, persuasive and appealing messages and other pragmatic acts. Kecskes and Papp (2000a) argued that communication patterns are dependent upon the culture and the structural organization of language.

Conceptual socialization

165

3.1. Structural organization of language A language that is a subject-prominent, configurational, syntacticallydominant, grammatical word-order (GWO) language such as English will definitely have a presentation format different from that of a language which is topic-prominent, non-configurational, pragmaticallydominant and pragmatic word-order language (PWO) like, for instance, Russian. The question is how this organizational difference is reflected in the presentation style. Rhetorical differences are multidimensional and usually include factors such as the inductive-deductive dimension, high-context versus low-context, linear-nonlinear organization, and topic-centered versus topic-associating style. Having reviewed the findings of several researchers (Kaplan, 1966; Kintsch and Greene, 1978; Fish, 1980; Glen and Glen, 1981;Kachru, Y. 1982; 1986;Purves, 1986; Odlin, 1989; Sato, 1989) concerning presentation formats in different cultures we have come to the conclusion that the further a language can be placed to the right on the configurational or GWO versus non-configurational or PWO typological continuum, the greater the chance is that this language will have an inductive, non-linear and topic-associating presentation format (Kecskes and Papp 2000a). However, this hypothesis needs further investigation because at present it is more like a speculation than a fact. Based on its configurational, subject-prominent, GWO and syntactically-dominant structure, the English presentation style appears to be deductive, low-context, linear and topic-centered rather than inductive, non-linear, high-context, and topic-associating. "Deductive" and "linear" do not need to be explained here but topic-centered does. To my knowledge the dichotomy "topic-centered versus topic-associating" presentational style was first used by Michaels (1981,1986) detailing a contrast in the discourse patterns of Black English Vernacular-speaking students and their Caucasian teacher. Michaels described the "topic-centered" style of white students in the same class as "tightly organized, centering on a single topic or series of closely related topics, with thematic development accomplished through lexical cohesion, and a linear ordering of events, leading quickly to a punch line resolution" (Michaels, 1986:102). On the other hand, he argued that the AfricanAmerican students "topic-associating" presentation style represents " a series of segments or episodes which are implicitly linked in highlighting some person or theme" (Michaels, 1986:103).

166

Chapter 8

Michaels spoke about oral presentation styles in an entirely different context than the one being discussed here. We, however, think that the "topic-centered" versus "topic-associating" continuum has its validity when languages are compared according to their presentation styles. It is enough to compare an article written on a linguistic topic in English by a native speaker of English and another article written also on a linguistic topic in Russian by a native speaker of Russian. In the English article the point is made at the very beginning of the article, and then come the facts that support the argument with a summary at the end functioning like a conclusion. In the Russian article, however, first comes a list of facts which are about the topic but it is not clear yet exactly how those facts are connected. Close to the end of the article the loosely connected facts are united in an inseparable whole supporting the main point that is just becoming clear for the reader. This presentation style used by a Russian native speaker writing an essay, composition or article in English can easily upset a native speaker of English who is not familiar with that kind of organization of text. Similar difference can be found between English and other languages. Lu (2001) pointed out that the fundamental root difference between Chinese and English lies in the fact that Chinese is a non-configurational, topic-comment language while English is a configurational subject-predicate language. This results in a presentation style in Chinese where the utterances lay stress on the cohesive relations of all parts in an utterance whereas in English the emphasis within an utterance falls on "who does what". In English the logical subject (action doer) and the grammatical subject are usually the same while in Chinese this is usually not the case. Lu (2001:219) illustrated this with an example: (93) Chinese: Caise katongpian yao shu disini de zui hao. Possible English translations: A. Disney color cartoon movies are the best of all. B. Disney produces the best cartoon movies. C. As for cartoon movies, Disney products are at the top.

Conceptual socialization

167

Lu explained that the utterance has its emphasis on "caise katongpian" ("colour cartoon movies") while the second part of the sentence "yao shu disini dezui hao " ("Disney products count as the best") makes a comment on the topic which is the first part of the sentence. In the English translations, however, "Disney" remains the focus because that is the doer of the action. In contrast, the Chinese sentence discusses the topic "caise katongpian", and does not put much emphasis on "Disney".

3.2. Cultures and communication patterns The effect of the structural organization of language on communication patterns seems to be in accordance with theoretical frameworks developed to distinguish differences between cultures.25 There is no room here to review all literature on this issue so only those frames will be mentioned which seem to have some special bearing on the use of SBUs in different cultures.

High-Context and Low-Context Hall (1976) distinguished between high-context and low-context cultures. He considered meaning and context as "inextricably bound up with each other (Hall 1982:18). " The difference between high and low context cultures is dependent upon how much meaning is encoded in the context versus in the linguistic code. Low-context cultures, such as the American culture, are inclined to place more meaning in the linguistic code and less meaning in the context. For this reason, communication tends to be specific, explicit, and analytical (Hall 1982; Ting-Toomey 1985). In high-context cultures, such as Arabic or Japanese, meaning appears to be embedded in the context rather than in the code. According to Hall "most of the information is either in the physical context or internalized in the person, while very little is in the coded, explicit, transmitted part of the message (Hall 1982:18)." Consequently, the interlocutor needs to look for the contextual cues in order to understand the meaning of the message. A speaker belonging to a high-context culture will tend to talk around a point rather than explicitly expressing the point, and expects his interlocutor to figure out what he really means. High-context cultures usually require a topic-associating rather

168

Chapter 8

than topic-centered style. The speaker is not expected to be specific because the details are in the context, not in the linguistic message so much of the "burden" of how to figure out meaning falls on the listener. In low context cultures, however, the speaker is supposed to convey the meaning accurately and thoroughly. According to this approach lowcontext cultures like the American culture would require the use of less SBUs than high-context cultures. This is however, not necessarily so. In chapter 5 we distinguished two main groups of SBUs according to their content: situation-bound rituals and situation-bound routines. Low-context cultures such as American seem to prefer routines to rituals while in high-context cultures it is the other way around. Highcontext cultures such as Arabic, Korean, or Turkish are usually tradition-oriented in which establishing rapport between participants is very important. Situation-bound rituals serve this communicative purpose very well.

Indirect and Direct Levine (1985) classified cultures in terms of directness versus indirectness, and clarity (univocal) versus ambiguity in communication patterns. According to Levine direct style requires clarity, accurate representation of facts, explicitness and communication techniques that aim to avoid emotional overtones and suggestive allusions. American cultural prefers clear and direct communication as evidenced by many common expressions: "Say what your mean," "Make your point", "Don't beat around the bush," and "Get to the point." In contrast, indirect style of communication is more ambiguous and emotionally rich. This mode of speaking is more likely to conceal or bury the message and it lacks clarity and precision in presenting the facts. Levine argued that "By alluding to shared experiences and sentiments, verbal associations can express and evoke a wealth of affective responses (Levine 1985:32)."

Doing and Being Kluckhohn and Strodtbeck (1961) proposed a cultural divide on the basis of two different value orientations. One orientation focuses on activity, the other on being and becoming. The activity orientation puts

Conceptual socialization

169

emphasis on activities which results in accomplishments that are measurable. Stewart (1972:34) called this type of orientation "doing." He pointed out that features of "doing" cultures are characteristic for the American culture which underlines the importance of achievement, visible accomplishments, and measurement of achievement. Several SBUs express this kind of cultural orientation such as "How are you doing?", "What's happening?", "Is everything all right?" Zaharna (1995) claimed that an important characteristic feature of the activity orientation is that words are expected to match actions. He referred to Gold who stated that "symbolic statements should not be a substitute for action (Gold 1988:171)." "Being" cultures such as the Chinese, Japanese, or Arab cultures are, in some respect, the opposite of "Doing" cultures. Okabe (1983) compared the Japanese "being" culture to the American "doing" culture. He pointed out that achievement and development are not so important in a traditional vertical society such as Japan where it is the individual's birth, family background, age and rank that really matter. Okabe claimed that for a person living in a "being" culture, such as Japanese, "what he is" carries greater significance than what he does (Okabe 1983:24)." According to Zaharna (1991) in Arabic culture, the equivalent of "How are you doing?" is literally "What is your condition?". To which one would be expected to respond with reflecting upon one's emotional or physical state, or Thank God for his blessings.

Linear and Non-linear

Carey Dodd (1992) distinguished between linear and non-linear thought framework. He argued that the linear cultural pattern highlights beginnings and ends of events, is object-oriented rather than people or event-oriented, and is empirical in its use of evidence. The linear framework focuses on presentation of singular themes, i.e., one point is followed by the second point, which is followed by the third, etc. Points or facts are presented sequentially, in a linear progression. People using this presentation framework usually build an argument in a "step by step fashion" instead of "throwing things in all at once (Dodd 1992:163)." The message structure of non-linear cultures is developed in the opposite way. It generally has multiple themes, is expressed in oral terms and heightened by nonverbal communication. Dodd argued that the non-linear thought framework includes the "simultaneous

170

Chapter 8

bombardment and processing of a variety of stimuli" so people using this framework would think in images, not just words (Dodd 1992: 163)." People and events are more important than time orientation so segmentation of time is usually missing. Differences in communication patterns can be described on a continuum on the left edge of which we can place cultures which are lowcontext, direct, doing and linear such as English or German. On the right edge of the continuum we can find high-context, indirect, being and non-linear cultures like Turkish or Arabic. Cultures such as Russian or Japanese are closer to the right edge while French, Swedish or Dutch are closer to English. It is interesting to note that the communication patterns continuum does not differ too much from the language typology continuum which was discussed in chapter 3. Based on this fact we can speculate about the close tie between the organization of a language and the communication patterns that develop in that particular culture. A configurational, grammatical-word order and topic-centered language such as English will develop communication patterns that characterize low-context, direct, doing and linear cultures while a non-configurational, pragmaticword order and topic-associating language such as Arabic or Turkish will prefer communication patterns that characterize high-context, indirect and non-linear cultures. The use of SBUs, however, shows us that we must be careful with this type of generalization. American English is considered to be a direct and low-context language. Still, native speakers use a rather high number of SBUs which either derive from implicatures (see chapter 4) which do not necessarily characterize direct style, or they are compositionally almost empty and have only a pragmatic function that plays the role of indicator rather than information-provider in communication. We can argue, of course, that these SBUs are functionally rather direct if interlocutors are able to understand the "hidden directness" in the indirect expressions.

3.3. Role of language from a cultural perspective Zaharna (1995) argued that two cultures may have fundamentally different perspectives in how they view the role and function of language. This fact makes conceptual socialization difficult because it requires the learner to develop an alternative approach to how they use a language. Message design is shaped not only by cultural factors but

Conceptual socialization

171

also by socio-historical forces. Stock (1991) pointed out that written language was viewed primarily as a means for record keeping and documentation from a Western historical perspective. Language was used as an instrument for conveying information across space and time. This required emphasis on accuracy of content. According to Stock (1991) and Zaharna (1995) this historical root bears through today because most Americans view language as a medium for conveying or "transmitting messages." Even the names of the basic communication components in models of communication: "sender," "message" and "receiver" reflect this emphasis on transfer of information. In a low-context culture such as American English users expect language to be specific, accurate and factual. In message production preference is given to clarity, objectivity, and directness. In comparing Arabic culture to American English Zaharna underlined that "the major socio-historical forces that influenced the role of Arabic for the Arabs stem from the language's association as an art form, a religious phenomenon, and an identity tool. These forces appear to have shaped the role of the Arabic language in an entirely different fashion than English. Rather than viewing language as a means for transferring information with a stress on factual accuracy, language appears to be a social conduit in which emotional resonance is stressed. (Zaharna 1995 : 4)." Hitti (1958:90) claimed that "hardly any language seems capable of exercising over the minds of its users such irresistible influence as Arabic... The rhythm, the rhyme, the music produce on them the effect of what they call 'lawful magic' (sihr halal)." There are three major socio-historical forces associated with the Arabic language: poetry, religion, and nationalism. Each of them requires symbolism which is embedded in the very essence of the Arabic language. When using this symbolic system the communicator cannot be disengaged from either the message or the audience. This is totally in contrast with the very function of the written word (to record, preserve and transmit) which presupposes that the speaker may be physically separated from his audience. Cohen pointed out that Arabic ".. is a social instrument-a device for promoting social ends as much as a means for transmitting information (Cohen 1987:31)." Zaharna(1995) gave an excellent outline of the dominant communication preferences for the American and Arabic cultures in the following table:

172

Chapter 8

Table 4. Cultural Variations of Messages In American and Arabic Communication

Socio-Historical Influences

Hall (1976)

Levine (1985)

American

Arab

Legal Documentation

Poetry, Islam

Record Preservation

Nationalism

Need for Accuracy

Reliance on Symbols

Technical, Concrete

Emotional Resonance, Abstract

Language Used to Transmit Information

Language Used to Create Social Experience

Low-Context

High-Context

Meaning in Message

Meaning in Context

Explicit

Implicit

Include Details in Message

Details in Context

Speaker Responsible for

Listener Responsible for

Message Comprehension

Responding Message

Direct/Univocal

Indirect/Ambiguous Indirect, Circular

Direct, to the Point Ambiguous Clear Embellishments Valued Simplicity Valued Subjective(Deliberately Use Emotion) Kluckhohn and Strodtbeck (1961)

Objective (Avoid Emotion) Activity/"Doing" Emphasize Action Measurable Action Tie Between Word and Deed

Dodd (1982)

Linear

"Being and Becoming" Emphasis Relationship in Social Context Words for Social Effect Configuration/Non-Linear

Conceptual socialization

Cultural Preferences for Messages

Organized with Beginning and End

Organization Not Stressed

Object-Oriented

People and Event-Oriented

Literate Society

Oral Society

Written Word Valued

Oral Experience Valued

Singular Experience

Group Experience

Factual Accuracy Stressed

Imagery and Sounds Stressed

Logic and Coherence

Emotional Resonance

Speaker Detached from Audience

Speaker and Audience Linked

Analytical Reasoning

Intuitive Reasoning

American

Arabic

Simplicity

Repetition

Accuracy

Imagery

Understatement

Exaggeration

Actions

Symbols

173

The decisive effect of culture on communication patterns is emphasized in several studies. Wu pointed out that Chinese culture teaches people to "never reveal his or her thoughts and feelings (Wu 1996: 146)." This results in a communicative style that is characterized by implicitness, indirectness and an endeavor to save face and leave room for free manouevre. Lu argued that "the embedding of real meanings into seemingly vague and indirect utterances mirrors the Chinese cultural tradition of leaving sufficient room for future changes and for one's own convenience of handling cases (Lu 2001:215)." The influence of culture on communication patterns is so strong that even if the conceptual socialization process is very advanced and the individual has high proficiency and excellent skills in the L2 his or her

174

Chapter 8

interaction with Westerners is severely blocked by the limits imposed by cultural factors. According to Lu (2001) the influence of the traditional Chinese culture is so far-reaching and persistent that even second- or third-generation Americans of Chinese descendants are un-able to fully ignore it although their English proficiency is very close to native English speakers. Many of them do not speak Chinese any more and totally depend on English as the tool of thinking and communication. "Nevertheless, their speech acts are still in the shadow of culturally governed modes of thinking, talking and behaving (Lu 2001: 216)." In the next chapter we will discuss how the factors discussed here affect the use of SBUs and what kind of errors they induce.

4. Social identity The process of conceptual socialization is expected to lead to a degree of shared socio-cultural knowledge that is important for L2 learners to be able to communicate with native speakers cooperatively. Mitchell and Myles suggested that in certain settings the communication of NNSs with NSs may result in new forms of speech events, "in which old norms do not automatically apply, and are reconstructed in the course of interaction (Mitchell and Myles 1998: 168)." Encounters and communication opportunities with the NSs affect the development of social identity to a great extent. Tajfel defined social identity in the following way: "That part of an individual's self-concept which derives from his knowledge of his membership of a social group (or groups) together with the emotional significance attached to that membership (Tajfel 1974:69 cited in Hansen and Liu 1997: 567-568)." Conceptual socialization requires a dynamic, both individual- and group-oriented understanding of social identity as described in the works of Pierce (1995) and Bremer et al. (1996). Pierce emphasized that social identity integrates the language learner and the learning context. She described social identity as "multiple, a site of struggle, and subject to change (Pierce 1995:20)." In the process of conceptual socialization there is a constant interaction between language, context and identity. Pierce speaks about this dynamic process as follows: "I foreground the role of language as constitutive of and constituted by a language learner's social identity. It is through language that a person negotiates a sense of self within and across different sites at different points in time, and it is through language that a person gains access to -

Conceptual socialization

175

or is denied access to - powerful social networks that give learners the opportunity to speak (Pierce 1995:13)." The European Science Foundation (ESF) project on adult migrants learning a range of L2s informally in several European settings is a very good source for analyzing the L2 learners' linguistic development over time. Some researchers (Bremer et al. 1996) also used this project to investigate migrants' encounters with a variety of gatekeepers. They paid attention to social factors such as face and self-esteem that are closely tied to social identity. Their examples represent a large scale from resistance, reassertion of the speaker's LI identity to the assertion of a positive native-like identity. Movement on the scale is tied to several variables depending on the environment, target language group and the individual. In her study Willett (1995) described how three girls who were ESL students in an elementary school managed to develop identities as 'good students' through cooperation and mutual support. Referring to cases similar to this Mitchell and Myles concluded that "learner's attitudes and identity are dynamic, negotiable, socially contextualized, and subject to change, even within the framework of individual interactions (Mitchell and Myles 1998:170)."

Chapter 9 Easy to learn, hard to understand

1. UseofSBUs

1.1. Developmental stages According to Kasper and Blum-Kulka "one area where insufficient control of pragmalinguistic knowledge is particularly obvious is that of pragmatic routines (1993:9)." Wray (2002:186) argued that "formulaicity may be best seen as a strategy that can be employed at various times to enable a particular stage to be reached." Second language learners use formulaic expressions at every stage of their language development but to a different extent. As discussed earlier, the first stage of learning shows heavy reliance on formulaic expressions just as in LI acquisition. Learners pick up a certain limited set of fixed expressions paying attention to their function rather than compositionality. A sense of situational appropriateness of a given formula usually precedes the conscious control over form-function relations in the expression. In contrast, the intermediate level appears to be dominated by the learners' endeavor to create novel strings, and the presence of formulae is rather scarce. As Bahn et al. (1986) pointed out the growing L2 proficiency may lead away from target-oriented verbal behavior to a phase of creative but sometimes pragmatically inadequate verbalizations. The advanced level is characterized by a more balanced used of creativity and formulaicity where the learner can manipulate his/her language according to his/her wish to sound nativelike or not.

1.2.

Idiomaticity

SBUs are reflections of idiomatic language use which, as we saw earlier, is the clear sign of nativelike competence. Idiomatic language use can be defined as the ability to choose from among the possible synonymous lexical units (lexical routes) the ones that are best supported by the

178

Chapter 9

given culture-specific frame and the speech customs of the given language community. Both the open choice principle and idiom principle of Sinclair (1991) require idiomatic language use. Even when a single lexical unit ( a word, or a prefabricated phrase) is chosen, its selection is determined by its compatibility with other units in the utterance. The examples (39, 40, 41) in chapter 4 demonstrated what happens if the principles of idiomatic language use are denied. If we focus only on the use of SBUs, the most typical errors made by L2 learners can be summarized as follows:

1.2.1. A near-literal translation of an SBU from the L1 into the L2. House (1989) referred to a case where a German used "Excuse me, please" instead of the more appropriate "I am sorry" because the former is a translation of the German SBU "Entschuldigen Sie bitte ". Coulmas (1979) also gave a German example in which Americans translate their SBU "you are welcome" into German "Sie sind willkommen ", which sounds awkward.

1.2.2. No SBU is used in the L2 but the learner uses one based on LI The use of "Good appetite" at the table is a good example for this case. Another example is when I let American ladies go through the door before me. I open the door and say: "Please go ahead", or "After you". These two SBUs usually cause confusion: not because ladies cannot process these simple "plain" SBUs in which I mean what I say, but because it is not necessarily customary for a male to allow a female to precede him through the door in the U.S.

1.2.3. Situational load of the expression (FSP) is misunderstood The SBU is functionally appropriate but the actual situational load of the expression (FSP) is misunderstood. This is what happened when "You bet" was used inappropriately as a response to "Thank you." Another example to illustrate this case is when a student walks into a professor's office with a smiling face and greets him with a casual expres-

Easy to learn, hard to understand

179

sion such as "Hi Henry, what's up?" (See a similar example in the data of the study described below.)

2. A cross-sectional study Kecskes (2000b) conducted a cross-sectional study which aimed at investigating three closely related research questions concerning the process of conceptual socialization and the use of SBUs by Non-Native Speakers of English (NNSs):26 1 ) How much do NNSs rely on their L1 conceptual system when using SBUs? 2) To what extent is SBU selection affected by individual preferences? 3) Does the use of SBUs in the study confirm the existence of different developmental stages in the use of SBUs?

2.1. The subjects and the tests Data were collected from 88 non-native speaker (NNS) students and 33 native speakers of English (NS). NNSs represented ten countries (Japan, Spain, China, India, Nepal, Thailand, Laos, Malaysia, Russia, Mexico) and, at the time they participated in this study, they had been in the U.S. for at least six months. All NNS participants of the survey had a history of starting to learn English as a foreign language in their home country. It was anticipated that this fact would seriously affect their use of SBUs because EFL students always have a more conscious approach to language production than learners who acquire the language in a naturalistic environment with or without instruction. Foreign language learning is usually characterized by an analytic, bottom-up approach to language production, conscious rather than automatic use of fixed expressions, LI pragmatic dependency, a dominant role of individual learner strategies (Kecskes, 1995), and the interplay of principles and learning with the structural properties of the target language (Jordens 1996). Discussing input and instruction in second language acquisition Jordens argued that " if we want input and

180

Chapter 9

instruction to be effective, we have to take into account that principles of learning and processing interact with the structural properties of language to be learned.... This interaction is conditioned by three types of linguistic constraint: domains of generalization, clustering, and prerequisite linguistic knowledge (Jordens 1996:442-443)." The impact of these factors on the learning strategies generally remains with the language learners for a while even if they have the opportunity to immerse in the target culture. Consequently, tasks designed for the study tried to take into account the students background and pervious experience with the target language. Since the goal of the study was to investigate the effect of L1 pragmatic system and use of individual strategies in L2 language production students should not be detracted from the real task by grammatical and lexical difficulties. All subjects, native speakers and non-native speakers alike were given three different types of tests: a Dialogue Interpretation Test (DIT), two Discourse Completion Tests (DCT), and a Problem Solving Test (PST). First the students were tested to determine the extent to which their interpretation of situation-bound utterances in American English was the same as that of native speakers who took the same test. The test items consisted of a dialogue containing an SBU that subjects were to interpret, and sufficient context to permit them to do so. For instance: (94) - Hurry up, we will miss the train. - Come on, we have plenty of time. (95) -- David, let menice introduce to our new secretary, Oh, hi Sally, to meetyou you. Welcome aboard. Sally. The second survey was a Discourse Completion Test that consisted of two parts and in which students were expected to respond to a question or a statement of the interlocutor. Those were utterances that are very frequently used outside the classroom. In the first DCT students were expected to give a response to questions such as:

Easy to learn, hard to understand

(96)

181

Can I talk to you after class ?

(97) How was the party last night? (98) Hey, ..What's up? In the second type of DCT participants had to complete the missing part of a short dialog. Students were encouraged to enter anything they thought was appropriate in the given situation: (99) Hi, John. How is it going? -

Listen, can I talk to you after class?

-

OK, see you after class.

The Problem Solving Test described a situation in which an utterance was expected to be made. The situations were frequent ones but required more culture specific knowledge than the DCTs. Students were asked to write whatever they would say in that particular situation. For example: (100) "You want to talk to your professor. You knock at his door, open it, and say what? " (101) "You are calling your friend Bob on the phone. What do you say when someone picks up the receiver at the other end? "

182

Chapter 9

2.2. Data analysis Students' answers were entered into a table for each test. The table contained the following columns: Table 5. Student's answers NS/NNS

LI

Years of English

Years in U.S.

"Hi, John,how is it going?"

"Can I talk to you after class?"

NNS

Russian

1.5

1.3

"Allright"

"Yes, sure"

NNS

Japanese

7

1

"Not bad"

"Sure"

This way it was possible to see vertically all students' answers at a time in each test and horizontally each student's responses to each question, utterance or task. Every answer was entered into the table as was produced by the speaker, i.e., no errors were corrected. NNSs' responses were compared to one another and to native speakers' responses. Data were analyzed for two variables: cultural specificity of SBUs and individual learner strategies in picking the appropriate SBUs, as it is in these two variables that the conceptual effect of LI and dominance of individual choice can be demonstrated best. Because of the nature of research questions and data (students' responses in this survey were unquantifiable) a qualitative approach was applied, which is called "focused description" in the relevant literature (Larsen-Freeman and Long, 1991:17). Findings are based on observations and careful screening of data for common features, patterns, similarities and tendencies in use. I have no room here to present all data so focus will be only on issues relevant to the subject of this book.

2.3. Main findings of the survey Conceptual socialization requires familiarity with the socio-cultural background of the target language. Elements of language that are cui-

Easy to learn, hard to understand

183

ture specific are especially difficult for L2 learners to acquire and use properly because not only linguistic but also extralinguistic factors such as learner's preference and social identity affect language use. Yoshida described this problem in the following way: "... although I might have knowledge of what to say with whom in what circumstances, that does not necessarily mean that I am able to perform accordingly. Moreover, even if I could perform in an "American" way if I consciously strived to do so, that does not mean that I feel comfortable doing so (Yoshida 1990:20)."

Another example which connects Yoshida's statement with what was said on the issue of social identity in the previous chapter was given to me by one of my former graduate students. She taught EFL at the Berlin Water Utility Co, Germany, where she met a water engineer (A) who had developed an innovative well-juvenating technique and was to present his findings in Colorado, and a lady (B) who headed the international project negotiations team. The American translator working on the text joked about A's complex and lexically dense writing style. "A" had prior experience with his American counterparts and understood that translation means not just a different system of signs but also different communication patterns, so he allowed his text to be re-worked into a streamlined, linear version. "B", however, was not as easy-going. She expressed deep concern about betraying her integrity, living a lie, if she adopted English-style discourse. She complained to my former student in the following way: "They realize our cultures are different; they know that I am German: why must I change? Why can't they accept me as I am?". Clearly "B" showed resistence to conceptual socialization and was not ready to change her LI-based identity in any way. The findings of this survey seem to confirm Yoshida's claim and the strong impact of social identity on language use including SBUs.

2.3.1. Cultural specificity Students' reactions observed as elicited through the Problem Solving Test (PST) and the Dialogue Interpretation Test (DIT) demonstrated that cultural differences played a significant role in the selection of SBUs. This was especially noticeable in the following two cases:

184

Chapter 9

SBUs with FSP and CSCP These SBUs usually have a very specific meaning. The gap between "what is said" and "what is communicated" is quite wide. They are commonly used by native speakers but NNSs either had difficulty with their interpretation in the DIT or chose to ignore them in the PST even if they perfectly understood their situational meaning. What made the use of these SBUs difficult for NNSs is the fact that these expressions receive their specific pragmatic charge from the situation. Their figurative meaning sometimes is so remote from their literal meaning that NNSs can hardly recognize the metaphorical use. As examples:

(102) "OK, shoot." [Go ahead ] (103) "Get out of here" [Don't fool me] (104) "Come on" [Relax] When asked to interpret these utterances NNSs relied on the literal meaning and compositional structure of the expressions which led to misunderstandings. What makes the interpretation of these SBUs even more complicated is that they are widely used in their literal meaning as well, as in the following: (105) Jim: - Hi, Roger. I hope I'm not interrupting anything... Roger: - Get out of here. Don't you see that I am busy? Reluctance of NNSs to use these expressions can be explained by the fact that cognitive mechanisms responsible for the metaphorical or figurative meaning of these phrases prove to be too culture-specific for NNSs. For instance, it would be hard to find another culture in which

Easy to learn, hard to understand

185

'telling one's opinion' is compared to 'shooting' ("OK, shoot") as in American-English.. Another typical example for cultural specificity is when students were confronted with the following task in the Discourse Interpretation Test: "There is a TV show on. The anchor pauses the program for a commercial. S/he wants the audience to continue watching the program after the break. What does s/he say?" Students were expected to remember cliches they hear on TV in this situation. Natives speakers came up with the most frequently used expressions such as: (106) "Stay tuned. We'll be right back." (107) "We'll have to take a break. Don't go away." (108) "Stick around." NNSs having spent less than two years in the U.S., however, did very poorly on this task. They could hardly recall anything that at least vaguely resembles the most frequently heard expressions. Although foreign students often watch TV, these utterances sound strange for them, and they just ignore them thinking that they will never need them. Here are some responses: (109) " Keep your channel" ( Japanese, 13/1)27 (110) "When we come back we will an action ..." (Spanish, 7/0.6) The correct answer was not a problem for those students who had spent more than two years in the target language country.

186

Chapter 9

Culture-specific situation Sometimes NNSs have the same or similar situational frame in their L1 but the expected scripts are different in the two cultures (Kecskes, 1998b). For instance, the expression "How are you doing?" in American English generally functions like a greeting. No other answer than "Fine, thank you" (or its pragmatic equivalents) is expected. In many other cultures, however, this expression means what it says: "tell me how you are doing". So the situational frame is the same as in American English but the script is different. Not recognizing this type of difference NNS participants of the survey often used their LI script. In the Discourse Completion Test (DCT) all native speakers reacted to the question "How are you doing?" with a short, positive response. None of them used a negative answer. This was not always the case in the NNSs'responses. Although most students'responses showed that they had acquired the positive attitude there were responses that were not exactly nativelike because students wanted to say too much. Here are two examples: (111) - How are you doing? - Hi. Pretty good. I am happy. How are you? (Japanese, 9/2.6) - Hi. I am fine, thanks. What about you? (Russian, 4/0.6) Culture specific situations make difficult for NNSs to figure out what style or tone is appropriate to use. This was especially obvious when students tried to distinguish between relaxed, informal American-style of speech and false intimacy. For instance, when NNSs were given the following task: "You want to talk to your professor. You knock at his door, open it and say what?", students having spent less than a year in the U.S. applied more or less their own socio-cultural rules:

Easy to learn, hard to understand

187

(112) - Hello Professor Brown, I am in your class, and I have some questions to ask you. Do you have time that I can ask you now? (Japanese, 9/2.9) (113) - Excuse me sir, are you free now because I need to talk to you for a while (Malaysian, 10/0.6) More advanced students, however, seem to have made false generalizations and sounded too relaxed and disturbingly intimate: (114) - Hi, ex. Randy (Chinese, 10/1.5) (115) - Hi, how are you, George? (Laotian, 2/2)

2.3.2. Individual learner strategies Individual learner strategies seemed to dominate SBU selection and use. SBUs are learned rather than evolve in the L2 (Kecskes, 1998b:305) which increases the role of conscious, individual selections. Cognitive mechanisms responsible for SBUs can differ to a great extent in the LI and L2. Pragmatic transfer is motivated not only by LI influence but also by individual learner differences. Speakers play an active role in selecting which SBU or SBUs to acquire for a particular situation. Their like or dislike of certain expressions and situations was demonstrated in their responses. Students often picked out some expressions from their inventory and started to use them in a formulaic way. Individual learners seem to prefer to set up their own pragmatic norms in the use of SBUs which can change subject to contextual variation and development over time. Kasper argued that "...monolingual native speaker norms - themselves products of negotiation rather than pregiven, and subject to social and contextual variation - are often in-

188

Chapter 9

adequate goals for nonnative speakers' production of L2 linguistic action and should be replaced by standards that recognize L2 learners as multilingual speakers (Kasper 1996a: 10)." The results of the survey con-firmed the findings of several other surveys (c.f. Bardovi-Harlig 1996; Beebe 1988; Kecskes 1999) according to which depending on the level of their conceptual socialization students can take an individual approach to the use of SBUs. Here are some examples from the data: Oversimplification In many cases NNSs preferred simple SBUs to more complex ones. However, they often misunderstood the function of simplicity in those expressions. As a result, their production often appeared to be oversimplified. They used one-word responses to questions where native speakers usually felt the need to add something to the straight response, or use a 'motivated' one-word utterance. Some examples: (116) - Can I borrow your pen ? NNS:

- Yes.

NS:

- Sure./Certainly./Yeah, sure./OK, here you go.

(117) - I have to go now. I'll call you tomorrow. NNS: NS:

- OK. - Okay, see you soon./OK, talk to you later./Alright, take it easy. This oversimplification was especially common among Asian students who had been in the US for less than two years. So they were not beginners and had quite a bit of experience in the target language culture. In spite of this they usually made a false generalization about American speaking style. In comparison to formulas in Asian languages, many American expressions might seem to be too short and laconic to these students. In consequence, they had the false impression

Easy to learn, hard to understand

189

that if they wanted to sound nativelike, they had to be as short-spoken as possible, as in: (118) thank you very much fine, thank you sure, you can

— thanks — fine « sure

Verbosity False generalization is a part of an evolving process which characterizes the development of NNS discourse style. Responses of NNSs with less than a year experience in the US demonstrated that they were usually more "verbose" than native speakers. They often seem to have broken the Gricean quantity maxim which says "Make your contribution as informative as required. Don't make your contribution more informative than required." (Grice, 1968). Many students (especially those coming from Malaysia, Nepal, Thailand ) sounded too verbose in situations where Americans expect only short responses, as follows: (119) - How was the party last night? Did you have a good time? - It was really nice. Oh yes. It was cool.(Spanish, 6 months in the US) - Yes, it was a great party. I enjoy meeting different people. (Thai, 6 months in the US) - Good. But I didn't have a good time. (Malaysian, 2 months in the US) - Yes, I had a happy time, because I could meet old friends and had good dishes. (Japanese, 6 months in the US)

(120) - Jamie, hi. Listen, I want to apologize for what happened yesterday. - Never mind. I think I have forgotten your stupid trick. (Russian, 6 months in the US)

190

Chapter 9

- Um um - who are you? I am sorry. I do not recognize you. (Spanish, 6 months in the US) Ignoring SBUs Advanced NNSs with long experience in the target language culture are expected to have made the necessary adjustments to the conceptual system of the target language. Their ignoring of SBUs where native speaker generally used conventional SBUs only, however, showed some deliberate break of communicative rules which can hardly be explained by the lack of conceptual socialization. For instance: (121) - I have to go now. I'll call you tomorrow. - OK, at what time are you going to call me tomorrow? (Mexican, 6/2.6)

(122) - Hey, what's up? - I got some trouble. Could you help me? (Japanese, 7/2) (123) - Today is my birthday. - Wonderful. One more year and you can come to the bars with me. (Laotian, 6/1.5) The survey demonstrated that NNSs have difficulties adjusting to the language-specific principles of salience in the target language and their conceptual socialization does not seem to have been completed yet. This problem is especially acute where parallel processing is needed to figure out whether the more salient meaning of an SBU is figurative or literal in the given situation. The use of SBUs demonstrated the existence of the three developmental stages. The first developmental stage

Easy to learn, hard to understand

191

is the period of strong LI-culture transfer, the second is usually characterized by false generalizations, and the third is when things are expected to fall into place. This three-stage developmental process explains why NNS production was characterized by both 'oversimplification' and 'verbosity'. Short answers characterize the language use of more advanced students having been in the US for one or two years while verbosity is frequent in the responses of beginners. Students living in the US for more than two years are expected to develop a more balanced use of SBUs which, in many cases, resembles that of the native speakers. But in fact, this is not what happens to many advanced NNSs. It is usually the third period when individual selection starts to play a significant role. Based on the results of this survey it can hardly be claimed that the more time students spend in the target language country the better their use of SBUs becomes. The language use of students with one year or under is strongly influenced by pragmatic transfer from the LI. Individual learner differences can be explained by the intensiveness and success of their previous studies of English as a foreign language. Students with at least two years of experience in the target language country seemed to use SBUs as a sign of their integration in the new language community. Their individual choices can not always be considered as signs of their pragmatic competence because they are significantly affected by factors such as cultural differences and individual learning strategies. The survey demonstrated that length of stay was less important than distance of cultures and familiarity with target discourse patterns. Individual selection was especially noticeable in the responses of advanced Asian students. Their production was grammatically correct and even appropriate but not necessarily nativelike which can partly be explained as a clear endeavor to emphasize their individual use of target language means.

3. Positive transfer In the previous chapter it was claimed that during the process of conceptual socialization second language learners have to deal with the problems of culture-specific communicative functions, the formulaspecific pragmatic properties of expressions and the differences in communication patterns. Immersion in the socio-cultural environment

192

Chapter 9

of the target language is not enough. Serious efforts are needed from the learner. Conceptual socialization means shaping an integrity which is ready to absorb both cultures. This is often reflected in a language behavior that is described in the "the intercultural style hypothesis " (Kasper and Blum-Kulka, 1993). This refers to a unique development of the CUCB in multilingual people. According to this hypothesis (Blum-Kulka, 1991; Blum-Kulka and Sheffer, 1993; Kasper and BlumKulka, 1993) speakers fully competent in two languages may create an intercultural style of speaking that is both related to and distinct from the styles prevalent in the two substrata, a style on which they rely regardless of the language being used (Kasper and Blum-Kulka, 1993). Kasper and Blum-Kulka claimed that the hypothesis is supported by many studies of cross-cultural communication, especially those focusing on interactional sociolinguistics (e.g. Gumperz, 1982; Tannen, 1985) and research into the pragmatic behavior of immigrant populations across generations (e.g. Clyne, Ball, and Neil, 1991).Our experiment with Hungarian students (Kecskes and Papp, 2000a; Kecskes and Papp 2000b) and some findings of a Finnish immersion program (Bjorklund, 1995) demonstrated that even students learning a foreign language intensively and/or studying content area subjects in the foreign language may develop some kind of intercultural style. This is, however, only a possibility and not a necessity. The process of conceptual socialization is based on positive transfer which is the result of the bidirectional influence of the two language channels and cultures (Kecskes and Papp 2000a). What is transferred is not some kind of accent, lexical units or structures but knowledge and pragmatic skills, and the result is noticeable in the language use of multilingual speakers: in better language manipulating skills, metalinguistic awareness, good interactional style, broader knowledge base, appropriate use of SBUs and multicultural attitude. This assumption is supported not only by my own experiment (Kecskes and Papp 2000) but also by Verhoeven' study (1994) who attempted to find empirical evidence for the linguistic interdependence hypothesis (Skutnabb-Kangas and Toukomaa, 1976; Cummins, 1976, 1979, 1991) which states that in bilingual development, language and literacy skills can be transferred from one language to another. He worked with Turkish children living in the Netherlands since infancy. During the longitudinal experiment Verhoeven monitored the development of lexical, morphosyntactic, pragmatic, phonological and literacy abilities in both the LI (Turkish) and L2 (Dutch) of the children.

Easy to learn, hard to understand

193

Verhoeven's findings demonstrated that there was almost no transfer at the level of lexicon and syntax. However, he found positive evidence for the interdependence in bilingual development at the level of pragmatic skills and literacy skills. It is important to note that these were multilingual speakers who were exposed to the L2 since infancy so it was essential that they developed a CUCB and went through conceptual socialization. We know very little about this positive transfer which involves pragmatic behaviors or other knowledge rather than words and structures. Kasper and Blum-Kulka ( 1993) argued that this phenomenon has been paid little attention by researchers because positive transfer usually results in successful communication and therefore it is less exciting to study. The other reason for neglect may be a methodological problem. It is extremely difficult to demonstrate positive transfer as a result of second language development because of the nature of the changes. Kecskes and Papp (2000a) argued that positive transfer is bidirectional and must be looked for in the LI production rather than in the L2 production because the LI production is also expected to change during conceptual socialization as a result of the development of the CUCB. I pointed out previously that conceptual socialization has two sides: a skill-side and a content-side. In the skills of language use it is possible to demonstrate qualitative changes that are quantifiable. In the article "How to Demonstrate the Conceptual Effect of the L2 on the LI?" Kecskes and Papp (forthcoming) claim that we cannot expect the L2—>L1 transfer to be the same as the L I — > L 2 transfer. The L2 effect will not result in any errors in L1 use. The problem is that the L2 influence on the LI is sometimes hardly "visible" at all in the literal sense of the word. This effect cannot be expected to occur in the form of some sort of structure or vocabulary transfer (although there may be such examples), rather it will influence the way in which LI is used, e.g., it will result in a more sophisticated use of the LI structures and vocabulary which may occur in the form of a positive change in literacy skills, text developing and manipulating skills, sentence-construction, and a more selective use of lexical items. So, again, this is not the transfer of concrete L2 structures and vocabulary items but the more elaborated use of LI structures and vocabulary items. Francis (2000) came to similar conclusion to that of Verhoeven"s (1994). He found that literacy skills learned through the dominant language (Spanish) can be applied to literacy tasks in another language (Náhuatl) that children understand, but in which they have not had the

194

Chapter 9

opportunity to practice reading and writing (Francis: 2000). Francis argued that this was possible because literacy skills are accessed from a Common Underlying Proficiency. Although Francis's research scenario significantly differs from that of ours his results also support the existence of a CUCB which makes conceptual transfer possible in either direction. As underlined earlier we cannot look for a direct effect of the L2 on the LI because there hardly seems to be anything like this. What we look for is a positive qualitative change in the use of the LI which is quantifiable. Positive qualitative change can be manifested in the form of improving literacy skills, text developing and manipulating skills, sentence-construction, and a more selective use of the vocabulary. Verhoeven's (1994) and Francis's (2000) experiment confirmed that positive transfer usually occurs after conceptual socialization reaches a certain level. The successful reach of this level depends not only on the development of L2 language proficiency but also on the willingness of the learners to acquire the new socio-cultural frameworks and make them a functional part of their CUCB. Stalnaker (1991) argued that shared mutual beliefs and assumptions are essential in communication and comprehension. Several researchers (c.f. Schachter, 1983; Gonzalez, 1987; Schmidt, 1993) have found that world views, beliefs, pragmatic assumptions, and values are almost always transferred from the LI to the L2 environments. Willingness, motivation and ability of individual learners to assume L2 socio-cultural beliefs and norms seem to play a decisive role in multilingual development. Consequently, an advanced non-native speaker cannot be expected "simply to abandon his/her own cultural world" (Barro, Byram, Grimm, Morgan and Roberts, 1993:56). Adamson (1988) pointed out that non-native speakers are often reluctant to accept and share the values, beliefs and presuppositions of an L2 community even if they have been living there for a long period of time and can speak the language quite well. Language proficiency is, of course, a crucial factor, but not the only factor. It has been argued that learners of high grammatical proficiency will not necessarily show concomitant pragmatic skills. Bardovi-Harlig (1996) observed that the range of success among students with a high level of grammatical proficiency is quite wide. Kecskes (1998b) argued that this variety in pragmatic proficiency can partly be explained by individual learner differences. As it was claimed earlier students seem to have more control over their pragmatic development than their grammatical devel-

Easy to learn, hard to understand

195

opment. They frequently learn pragmatic units and develop pragmatic attitude by choice, which they usually cannot do when learning grammar. Acton and Walker de Felix (1986) claimed that until proficient nonnative speakers reach an advanced acculturation stage (which is similar to what is called above " a certain level of conceptual socialization"), their language production is continually based on their LI world view and its socio-cultural framework. Hinkel (1995) demonstrated how cultural values are reflected in the use of modal verbs in the language production of non-native speakers of English. All these findings seem to point at one important fact: the language learner plays a very important and active role in the process of conceptual socialization. Investigating the use of situation-bound utterances (SBUs) by adult NNSs of English, Kecskes (1998b) found that individual learner differences play a decisive role in students' selection of SBUs in various speech situations. Linguistic behavior of adult NNSs with a high command of L2 appears to be dominated by their individual choice. This coincides with Beebe's opinion that "second language learners may never attain nativelike proficiency to the best of their ability because they may find that the reward of being fluent in the target language is not worth the cost in lost identification and solidarity with their own native language group" (Beebe, 1988:63). It should not be forgotten however, that individual learner endeavors are constrained differently in the second language and foreign language environments (Kecskes and Papp 2000a). In the foreign language environment the target culture is not directly present which makes conceptual socialization significantly slower. Learners in the second language environment, however, have the option to either immerse in the socio-cultural environment or ignore it to a certain extent. This possibility of choice is especially apparent in the use of SBUs. As the study (Kecskes 1998b) in the first part of this chapter demonstrated, NNSs often ignore SBUs in situations where NSs use them very frequently, or often picked up some expressions from their inventory and started to use them in a formulaic way. The results of the study confirmed Bahn et al.'s assumption that the increase of L2 knowledge seems to lead away from target-oriented verbal behavior to a phase of creative but sometimes pragmatically inadequate verbalizations (Bahn et al. 1986). Based on the different outcomes of Takahashi's (1996) study and Maeshiba et al.'s study (1996) Kasper (1996a) concluded that the learning environment may often play a central role

196

Chapter 9

in pragmatic development. Takahashi's research on the transferability of conventionally indirect request strategies was conducted with EFL students in Japan whereas participants in Maeshiba et al. were in an ESL context. Takahashi found that the L2 proficiency had very little effect on transferability ratings. Maeshiba et al., however, demonstrated no positive correlation between negative pragmatic transfer and proficiency. The study concluded that in pragmatics just like in other areas of interlanguage knowledge, negative LI transfer decreases as L2 proficiency improves.

Conclusion

The exploration of SBUs has given us the chance to discuss a number of theoretical issues that are on the research agenda of several fields including theoretical linguistics, applied linguistics, communication studies, sociolinguistics, psycholinguistics, diachronic semantics, pragmatics , second language acquisition and bilingualism: - the relationship of linguistic knowledge and encyclopedic knowledge, - prototype and invariant, - meaning as a result of the interplay of lexical units and context, -

salience,

-

literal and figurative meaning,

- creativity and formulaicity, - conceptual socialization. The interdisciplinary perspective and the multilingual rather than monolingual approach applied in this book has made it possible to examine these issues in their complexity. The attempt to bring a range of disciplines under one roof represented by a unique group of pragmatic units required the building and use of non-traditional approaches which resulted in the development of a dynamic model of meaning, the application of a dual language model, and the exploration of conceptual socialization. These three entities are mutually supportive. The diachronic and synchronic aspects of meaning are united in the DMM which explains how lexical items interact with context in shaping meaning in the dual language system which is responsible for conceptual socialization. SBUs represent an interface between language and culture. It was argued that language constructs as well as reflects culture. Treichler and

198

Conclusion

Frank ( 1989:3) claimed that language not only serves as the transparent vehicle of context or as the simple reflection of reality, but also participates itself in how that content and reality are formed, apprehended, expressed and transformed. As an important elements of language this is exactly what SBUs do. They both reflect and construct reality. The use of these prefabricated expressions is one place where the linkage between culture and language behavior becomes visible. Coulmas (1979:260) argued that societies have different attitude to prepatterned speech. In one culture standardized expressions that are frequently used by and familiar to each member of the given society enjoy respect and currency whereas in other cultures they may be considered unimaginative and simplistic because members of that group prefer individual creativity in language production. Coulmas (1979:260) put forward the idea that "societies vary along a multiplicity of continua whose extremes are marked by qualities such as originality vs. conformity; flexibility vs. ritualization; imagination vs. imitation; etc." He hypothesized that in a conformistic society where the pressure for normative language use in everyday conversation is high, the use of prefabricated units is expected to be more widespread than in another society, where the individual is supposed to demonstrate a marked personality which can be reflected in creative language use as well. As we saw the use of SBUs only partially supports this hypothesis because situation-bound rituals are usually preferred by conformistic societies while situationbound routines are widely used even in cultures without any "normative pressure"28 such as the American society. SBUs are powerful communicative means because they are not only required by institutionalized contexts but also can create contexts since the choice for the speakers from among possible communicative actions is much more restricted than in informal contexts. SBUs represent social knowledge, consequently their development is an essential part of conceptual socialization for second language learners because the proper use of SBUs can be expected only if learners become familiar with the cognitive mechanisms that are responsible for that particular social knowledge. Coulmas (1979) argued that cross-linguistic differences in the repertoires of routine formulae correspond to differences in the institutional structure of the respective societies. This is why SBUs are so important elements of the process of conceptual socialization. Through SBUs learners acquire more than just linguistic units. They are informed about the way of thinking of people, the value system and the institutional structure of the given society.

Conclusion

199

It was hypothesized that in bilingual and/or second language development the restructuring of the conceptual system results in the emergence of a common underlying conceptual base which supports conceptual socialization, a process during which the bilingual individual becomes able to use the routinized repertoire of both languages (Kecskes 1998; Kecskes and Papp 2000a). Conceptual socialization refers to the transformation of the conceptual system which undergoes characteristic changes to fit the functional needs of the new language and culture. A very important part of this process is the interplay of the two value-systems that are represented in the two languages. SBUs often represent social values, especially those which were called situation-bound rituals. It is not by chance that the Turkish language contains a lot of SBUs in which there is reference to the power of God. The abundance of fixed expressions originating from American football such as "kick off', "cut in", "touch down", etc. suggests that this branch of sport plays a special role in the social life of American people. SBUs carry not only meaning but also information about the life, history and thinking of people who have developed them to help communication in their language community. Therefore SBUs can play a very important role in language acquisition and education. The analysis of their compositional meaning, origin and use can give language learners invaluable knowledge about social values, the past and present of the target language community. There are fundamental differences in the production and comprehension of SBUs between native speakers and adult L2 learners. It was argued that our mind is trained to process new information through existing information so the impact of salience on the acquisition of subsequent languages is greater than we have ever thought. Salience is culture-specific, consequently the language learner has to make the necessary adjustments in her/his conceptual system in order to process information properly through the new language channel. Reconceptualization and modification of the existing salience system become possible only when a CUCB starts to emerge during the process of conceptual socialization. Interference appears to be a matter of salience rather than a matter of syntactic development, especially at a higher level of fluency. The syntax of a second language is learnable and teachable, but we can hardly say the same about the conceptual structure of an L2 (Kecskes 2000a). At least we do not know yet how to teach and/ or learn it. One thing is certain: we would need to develop methods other than we have been using to teach syntax. One of the main goals

200

Conclusion

of the book is to provoke further discussion and highlight those areas where further research is needed. The role of salience in second language development, the reflection of the language and culture interplay in the process of conceptual socialization, the teachability of cultural patterns through certain pragmatic units, and the characteristics of pragmatic acts in second language speech production are all issues that need further exploration. The new knowledge that this type of research will provide can help to better understand both first and second language acquisition, further advance a multilingual perspective of language production, and develop methods that can be effective in the teaching of second and foreign languages.

Data sources

Spoken Professional American English. million words).

Houston: Athelstan. (Two

Genzel, B. Rhona. 1991. Getting the hang of idioms and expressions. New York: Maxwell Macmillan. Howard, Julie. 1993. Idioms in American life. Englewood Cliffs: Prentice Hall.

Notes

1.

2. 3. 4.

5. 6.

7. 8. 9. 10. 11.

12.

13.

14.

15.

This does not mean, of course, that the original semantics of this type of utterances is always lost. There are situations when speakers mean what they say, i.e., one can sincerely be interested in how his/her partner is doing. The terms "encyclopedic" and "conceptual" are used here as synonyms. The terms "word-specific semantic property" and "culture-specific semantic property" will be explained later. This is not the case, of course, if the new word denotes a concept that does not exist in the LI. In that case, there is nothing to reconceptualize, so what happens is very similar to the process of LI development. Several encounters with the new word are expected to result in the development of the new concept. However, encounters are often substituted by definitions (mainly in school). That is why it was claimed that concepts in L2 are learned rather than evolve. (Kecskes and Papp, 2002a) These are my examples and not Cruse's. In an earlier work of mine (Kecskes, 1994) I referred to words with specific semantic properties by the term "loaded words". This is exactly the same concept referred to by Cruse with the term "charged words". When I wrote that article, I did not as yet know of Cruse's paper. "Relatively constant" as used here means "constant for an indefinite period of time. That is why Bibok and Németh (2001 b) argued that word semantic meaning is underspecified. We often ask: "In what sense is this word used in the sentence?" In explaining the externalist and internalist perspective on context I will rely on Violi's excellent summary of the issue (Violi, 2000). A good example for successful interference might be some words created as a result of the "feminist movement" in the U.S. : "flight attendant"; "police officer"; "chairperson"; "firefighter", etc. There are, of course, kinds of knowledge that do not involve language-like symbolic representations, such as motor skills or the ability to recognize faces or musical themes, and episodic memory for particular events. Literally, the Japanese expressions roughly means "Are you well?", rather than "How are you?", and is used less frequently than its American functional counterpart. All the Turkish examples in this book, if not otherwise specified, were provided by Professor Dr. Josua Bear, Middle East Technical University, Ankara, Turkey. The first part of the expression ")Kejiaio BaM" [ I wish you] is almost never used in conversation.

204 16. 17.

18. 19. 20. 21. 22.

23.

24.

25. 26. 27. 28.

Notes Fónagy discussed the role of Saussure's "arbitraire du signe" in his study published in 1956. Sweetser, referring to the issue of language shaping cognition, says that "...learning a word for a culturally important category could linguistically reinforce the learning of the category itself' (Sweetser, 1990: 7). Motivation here is to be distinguished from prediction as in Lakoff (1987). Here it is important to note that "without context" means that we want to focus on their most salient meanings. "Some" here and in the other examples refers to the fact that there are several other SBUs that can express the same function. The minimal pragmatic extension could mean something like " I am not ready to buy anything (yet)." Nuyts (1992:110) notes that Chomsky (1974) denied ever having had the intention to relate creativity and recursivity. But in the quote from Chomsky (1972:155) in Nuyts (1992) the suggestion for this correlation seems rather strong. There is considerable difference between a "second language" and a "foreign language" although these terms are usually inter-changeable in the relevant literature. The most important difference is the environment and the linguistic background of the learners (c.f. Hammerly, 1991; Kecskes and Papp 2000a). In the SL environment the target language is the LI of the community and SL students speak a variety of languages as their LI. In the FL environment the target language is remote and language learners usually speak the same LI. When "L2" is used in this paper it refers to both SL and FL where the distinction between them is not of primary importance. The term "conceptual socialization" was suggested to me by Dr. Josua Bear, Middle East Technical University, Ankara, Turkey after he read my book (Kecskes, I. and Papp, T. 2000. Foreign Language and Mother Tongue. Mahwah, NJ: Lawrence Erlbaum). He summarized with this expression ("conceptual socialization") how he understood my theory of the development of the dual language system described in the book I have relied on the study of Zaharna ( 1995) here. In the original study (Kecskes 2000b) only the first two research questions were investigated. The first number refers to the number of years of English studies while the second denotes the time spent in the U.S. This concept was used by G.H. von Wright (1963:12)

References Acton, W. and Walker de Felix, J. 1986 Acculturation and mind. In J. Valdes (ed.) Culture bound. Cambridge: Cambridge University Press. 20-32. Adamson, H. D. 1988 Variation theory and second language acquisition. Washington, DC: Georgetown University Press. AKHiiiHHa, A. A. and H.H. OopMaHOBCicaa 1982 PyccKuü peteeoù srnmem. MocKBa: PyccKHH 83ωκ. Albert, Ethel 1968 Value systems. In: The international encyclopedia of the social sciences. Vol. 16. New York: Macmillan. Altenberg, Bernd 1998 On the phraseology of spoken English: the evidence of recurrent word-combinations. In Cowie, A.P. (Ed.) Phraseology. Oxford: Clarendon Press, 101-122. Arnaud, Pierre, J.L. and Henri Bejoint (Eds.) 1992 Vocabulary in Applied Linguistics. London: Macmillan. Bahn, Jens, Hartmut Burmeister and Thomas Vogel 1986 The Pragmatics of formulas in L2 learner speech: use and development. Journal of Pragmatics. Vol. 10. 693-723. Baldinger, Kurt 1980 Semantic Theory: Towards a Modern Semantics. Oxford: Blackwell. Ball, S. J. 1990 Introducing Monsieur Foucault. In S. J. Ball (ed.) Foucault and Education: Disciplines and Knowledge. London: Routledge. Bardovi-Harlig, Kathleen 1996 Pragmatics and Language Teaching: Bringing Pragmatics and Pedagogy Together. In Bouton, L. (ed.) Pragmatics and Language Learning. Vol. 7. 21-41. Barro, Α., Byram, M., Grimm, H., Morgan, C., and Roberts, C. 1993 Cultural studies for advanced language learners. In: D. Graddol, L. Thompson and M. Byram (eds.) Language and culture. Clevedon: Multilingual Matters. 55-70. Barsalou, L. W. 1992 Cognitive psychology: An overview for cognitive scientists. Hillsdale, NJ: Lawrence Erlbaum. 1993 Challenging assumption about concepts [commentary]. Cognitive Development. 8. 169-180.

206

References

Bates, Elizabeth 1976 Language and Context: The Acquisition of Pragmatics. New York: Academic Press. Bates, E., Bretherton, I. and Snyder, L. 1988 From first words to grammar. Cambridge: Cambridge University Press. Bear, Joshua 1987 Formulaic utterances and communicative competence. Journal of Human Sciences. (Insan Bilimleri Degisi.) Middle East Technical University. VI/2: 25-34. Beebe, Leslie 1988 Issues in Second Language Acquisition. New York: Newbury House. Beebe, Leslie and Giles, H. 1984 Speech accommodation theories: A discussion in terms of second language acquisition. International Journal of the Sociology of Language. 46. 5-32. Beebe, Leslie, M., T. Takahashi, and R. Uliss-Weltz. 1990 Pragmatic transfer in ESL refusals. In R. Scarcella, E. Andersen, and S.D. Krashen (Eds.) On the development of communicative competence in a second language. 55-73. Biber, Douglas 1988 Variation across speech and writing. Cambridge: Cambridge University Press. Bibok, Károly and Enikö Németh 2001 a On the interaction between lexical and contextual information in the composition of utterance meaning. In Enikö Németh (ed.) Cognition in language use: Selected papers from the 7th International Pragmatics Conference. Vol. 1. IPRA: Antwerp: 12-25. 2001 b How the lexicon and context interact in the meaning construction of utterances. In Enikö Németh and Károly Bibok (eds.) Pragmatics and the flexibility of word meaning. Amsterdam: Elsevier: 289-321. Bierwisch, Manfred 1981 Basic issues in the development of word meaning. In Werner Deutch (ed.) The child's construction of language. London: Academic Press: 341-387. 1996 How much space gets into language? In Bloom, Paul, Mary A. Peterson, Lynn Nadel, and Merrill F. Garrett (Eds.) Language and Space. Cambridge, MA: MIT Press. 31-77. 1997 Lexical information from a minimalist point of view. In Chris Wilder, Hans-Martin Gartner and Manfred Bierwisch (eds.) The role of economy principles in linguistic theory. Berlin: Academie Verlag: 227-267. Bierwisch, Manfred and Schreuder, Robert 1992 From concepts to lexical items. Cognition. 42. 23-60.

References

207

Blasko, G.D. and C. Connine 1993 Effects of familiarity and aptness on metaphor processing. Journal of Experimental Psychology: Learning, Memory and Cognition. 19 (2), 295-308. Blum-Kulka, Shoshana 1991 Interlanguage pragmatics: The case of requests. In: R. Phillipson, E. Kellerman, L. Selinker, M.Sherwood Smith and M. Swain (eds.) Foreign/second language pedagogy research. Clevedon: Multilingual Matters. 255-72. Blum-Kulka, Shoshana. and H. Sheffer 1993 The Metapragmatic Discourse of American-Israeli Families at Dinner. In: Kasper, G. and S. Blum-Kulka (eds.) Interlanguage Pragmatics: 196-224. New York: Oxford University Press. Blum-Kulka, Shoshana, Juliane House and Gabriele Kasper (eds.) 1989 Cross-cultural Pragmatics. Norwood, NJ: Ablex. Bohn, Ocke-Swan 1986 Formulas, frame structures, and stereotypes in early syntactic development: some new evidence from L2 acquisition. Linguistics. 24. 185-202. Bolander, M. 1989 Prefabs, patterns and rules in interaction? Formulaic speech in adults learners' L2 Swedish. In K.Hyltenstam and L.Obler (eds.), Bilingualism across the lifespan: aspects of acquisition, maturity and loss. Cambridge: Cambridge University Press 73-86. Bolinger, Dwight 1976 Meaning and memory. Forum Linguisticum. 1:1-14. Bortfeld, Heather Forthcoming. What native and non-native speakers' images for idioms tell us about figurative language. Bouton, Larry 1994 Can NNS Skill in Interpreting Implicatures in American English Be Improved Through Explicit Instruction? - A Pilot Study. In Bouton, L. and J. Kachru. Pragmatics and Language Learning. Vol. 5. Urbana-Champaign: University of Illinois. 88-110. Bouveresse, Jacques 1974 On Linguistic Methodology. In H. Parret (m.) Discussing Language. The Hague: Mouton. Bremer, K., Roberts, C., Vasseur, M-T., Simonot, M., and Broeder, P. 1996 Achieving understanding: discourse in intercultural encounters. Harlow: Longman. Brown, H. D. 1980 Principles of Language Learning and Teaching. Englewood Cliffs: Prentice Hall. Bruner, J. 1983 Child's talk: learning to use language. New York: Norton.

208

References

Bybee, J., R. Perkins, and W. Pagliuca 1994 The Evolution of Grammar: Tense, Aspect, and Modality in the Languages ofthe World. Chicago: The University of Chicago Press. Bygate, Martin 1988 Units of oral expression and language learning in small group interaction. Applied Linguistics. 9(1): 59-82. Casad, H. Eugene (Ed.) 1996 Cognitive Linguistics in the Redwoods. Berlin: Mouton de Gruyter. Charteris-Black, Jonathan 2002 Second Language Figurative Proficiency: A Comparative Study of Malay and English. Applied Linguistics. 23/1. 104-133. Child, J.R. 1998 Language Aptitude Testing. Applied Language Learning. Vol.9. Nos. 1-2. 1-10. Chomsky, Noam 1965 Aspects of the theory of syntax. Cambridge, MA: MIT Press. 1972 Language and mind. New York: Harcourt. 1974 Dialog with H. Parret. In Parret, H. (mod.) Discussing language. The Hague: Mouton. 27-54. 1980 Rules and representations. New York: Columbia University Press. 1986 Knowledge of Language: Its Nature, Origin and Use. New York: Praeger. 1995 The Minimalist Programme. Cambridge, MA: MIT Press. Clark, A. 1993. Associative engines: Connectionism, concepts, and representational change. Cambridge: CUP. 1997 Being there: Putting brain, body, and world together again. Cambridge, MA: MIT Press. Clyne, M., Ball, M. and Neil, D. 1991 Intercultural communication at work in Australia. Complaints and apologies in turns. Multilingua. 10. 251-73. Cohen, Raymond 1987 Problems of Intercultural Communication in Egyptian-American Diplomatic Relations. International Journal of Intercultural Relations. Vol. 11: 31. Cook, Vivian 1991 Second Language Learning and Language Teaching. London/New York: Edward Arnold. 1993 Linguistics and Second Language Acquisition. New York: St. Martin's Press. Coulmas, Florian 1979 On the Sociolinguistic Relevance of Routine Formulae. Journal of Pragmatics. 3. 239-266. 1981 a Conversational routine: explorations in standardized communicative situations andprepatterned speech. The Hague: Mouton.

References 1981b Cruse, Alan 1990 1992

2000 Cowie, A.P. 1988

209

Routine im Gesprach: zurpragmatischen Fundierung der Idiomatik. Wiesbaden: Athenaion. Prototype theory and lexical semantics. In S.L. Tsohatzidis (ed.) Meaning and proptotypes: Studies in linguistic categorization. London and New York: Routledge. 382-402. Antonymy Revisited: Some Thoughts on the Relationship Between Words and Concepts. In Lehrer, A. and E.F. Kittay (eds.) Frames, Fields and Contrasts. Hillsdale, NJ: LEA. Meaning in Language: An Introduction to Semantics and Pragmatics. Oxford: Oxford University Press.

Stable and creative aspects of vocabulary use. In R.Carter and M.McCarthy (eds) Vocabulary and language teaching. London:Longman. 1998 (ed.) Phraseology. Theory, Analysis and Application. Oxford: Clarendon Press. Cruttenden, A. 1981 Item-learning and system-learning. Journal of Psycholinguistic Research 10,79-88. Cummins, Jim 1976 The influence of bilingualism on cognitive growth: A synthesis of research findings and explanatory hypotheses. Working Papers on Bilingualism. 9., 1-43. 1979 Cognitive/academic language proficiency, linguistic interdependence, the optimum age question and some other matters. Working Papers on Bilingualism, No. 19. pp. 121-129. 1991 Conversational and Academic Language Proficiency. AILA Review. 8. pp. 75-90. D'Andrade, R. 1995 The development of cognitive anthropology. Cambridge: Cambridge University Press. D'Andrade, R. and Strauss, C. (eds.) 1992 Human motives and cultural models. Cambridge: Cambridge University Press. Danesi, Marcel 1988 Neurological bimodality and theories of language teaching. Studies in Second Language Acquisition 10, 13-31. 1992 Metaphorical competence in second language acquisition and second language teaching: The neglected dimension. In J.E. Alatis (ed.) Georgetown University Round Table on Languages and Linguistics. Washington, DC: Georgetown University Press. 489500. Daniels, Norman 1975 The Cultural Barrier: Problems in the Exchange of Ideas. Edinburgh: Edinburgh University Press.

210

References

Dechert, H., Möhle, D. and Raupach, M. 1984 Second language productions. Tübingen: Gunter Narr. De Cock, S., S.Granger, G.Leech , and T.McEnery 1998 An automated approach to the phrasicon of EFL learners. In S.Granger (ed.) Learner English on computer. London and New York: Addison Wesley. 67-79. Dodd, Carley 1992 Dynamics of Intercultural Communication. Dubuque: Wm. C. Brown. Dreyfus, Hubert, Stuart Dreyfus, and T. Athanasion 1986 Mind over Machine. New York: Free Press. Ellis, Nick 1996 Sequencing in SLA. Phonological memory, chunking, and points of order. Studies in Second Language Acquisition, 18, 91-126. 1996 Analyzing language sequence in the sequence of language acquisition. Some comments on Major and loup. Studies in Second Language Acquisition, 18, 361-368. Ellis, N. and S.G. Sinclair 1996 Working memory in the acquisition of vocabulary and syntax: putting language in good order. The Quarterly Journal of Experimental Psychology, 49A (1), 234-250. Ellis, Rod 1985 Understanding Second Language Acquisition. Oxford: Oxford University Press. 1994 The study of second language acquisition. Oxford: Oxford University Press. Fillmore, Charles J. 1976a The need for a frame semantics within linguistics. SMIL, Skriptor, Stockholm, 5-29. 1976b Frame semantics and the nature of language. In Steven R. Harnad, Horst D. Steklis and Jane Lancaster (eds.) Origins and evaluation of language and speech. New York: Annals of the New York Academy of Sciences. 2-32. 1977 Scenes-and-frames semantics. In A. Zampolli (ed.) Linguistic Structure Processing. Amsterdam: North Holland, 55-81. 1979 On fluency. In C.J. Fillmore, D. Kempler and W.S-Y. Wang (eds.) Individual differences in language ability and language behavior. New York/London: Academic Press. Fillmore, Charles J. and Atkins, Beryl T. 1992 Toward a frame-based lexicon. The semantics of RISK and its neighbors. In Adrienne Lehrer and Eva F. Kittay (eds.) Frames, fields, and contrasts. Hillsdale, NJ: Lawrence Erlbaum. 75-102. Fillmore, Lily W. 1976 The second time around: Cognitive and social strategies in second language learning. Unpublished doctoral dissertation. Stanford University.

References Fish, S. 1980 Flynn, Susan 1996

Fodor, Jerry 1983 Foucault, M. 1972 Fónagy, Ivan 1982

211

Is There a Text in This Class? The Authority of Interpretive Communities. Cambridge, MA: Harvard University Press. A parameter-setting approach to second language acquisition. In Ritchie, W. and Bhatia, T. (eds.) Handbook of second language acquisition. San Diego: Academic Press: 121-158. The modularity of mind. An essay on faculty psychology. bridge: MIT Press.

Cam-

The Archeology of Knowledge. London: Tavistock.

Situation et signification. Pragmatics and Beyond. Amsterdam: John Benjamins. Francis, Norbert 2000 The Shared Conceptual System and Language processing in Bilingual Children: Findings from Literacy Assessment in Spanish andNahuatl. Applied Linguistics. 21/2: 170-204. Gass, Susan and Larry Selinker (eds.) 1983 Language Transfer in Language Learning. 112-133. Rowley, MA: Newbury House. Gee James Paul 1999 An introduction to discourse analysis. London: Routledge. Geeraerts, Dirk 1997 Diachronic Prototype Semantics. Oxford: Clarendon Press. 2000 Salience phenomena in the lexicon: A typology. In Albertazzi, Liliana (ed.) Meaning and Cognition. Amsterdam: John Benjamins. Genzel, Rhona B. 1991 Getting the hang of idioms and expressions. New York: Macmillan. Gibbs, Ray W. Jr. 1984 Literal meaning and psychological theory. Cognitive Science. 8, 575-604. Giora, Rachel 1997 Understanding Figurative and Literal Language: The Graded Salience Hypothesis. Cognitive Linguistics. 8-3. 183-206. 2001 Models of Figurative Language. Mahwah: Lawrence Erlbaum. 2002 On our mind: Salience, Context and Figurative Language. Oxford: Oxford University Press. Glenn, E. S. and C. G. Glenn 1981 Man and Mankind: Conflict and Communication Between Cultures. Norwood, NJ: Ablex. Goffman, Erring 1967 Interaction ritual: essays on face to face behavior. New York: Anchor Books.

212

References 1972

Gold, Ellen 1988 Gonzalez, A. 1987

On Face-Work: an Analysis of Ritual Elements in Social Interaction. In John Laver and Sandy Hutches on (eds.) Communication in Face-to-Face Interaction. Harmondsworth: Penguin, 319-346. Ronald Regan and the Oral Tradition. Central States Speech Journal. 39: 171.

Poetic imperialism or indigenous creativity. In: L. Smith (ed.) Discourse across cultures. Englewood Cliffs, NJ: Prentice Hall. 141-156. Granger, Sylviane 1998 Prefabricated patterns in advanced EFL writing: collocations and formulae. In A.P. Cowie (ed.) Phraseology: theory, analysis, and applications. Oxford: Clarendon Press: 145-160. Grice, H.Paul 1968 Studies in the Way of Words. Cambridge, MA: Harvard University Press. 1975 Logic and Conversation. In P. Cole and J.L. Morgan (eds.) Syntax and Semantics. Vol. 3: Speech Acts. New York: Academic Press. 41-58. Grosjean, Francois 1997 The bilingual individual. Interpreting: International Journal of Research and Practice in Interpreting: 2: 163-191. 1998 Transfer and language mode. Bilingualism: Language and Cognition:, 1: 175-76. 2001 The bilingual's language modes. In Janet L. Nicol (ed.) One mind, two languages. London: Blackwell. 1-23. Gumperz, John 1982 Discourse strategies. Cambridge: Cambridge University Press. Gumperz, John and Hymes, Dell 1972 Directions in sociolinguistics: The ethnography of communication. New York: Holt. Gumperz, John and Steven C. Levinson 1996 Rethinking linguistic relativity. Cambridge: Cambridge University Press. Hakuta, Kenji 1976 A case study of a Japanese child learning English as a second language. Language Learning. 26(2): 321-351. Hanks, William 1996 Language and Communicative Practices. Boulder, CO: Westview Press. Harkins, Jean and Wierzbicka, Anna (eds.) 2001 Emotions in Cross-linguistic Perspective. Berlin: Mouton de Gruyter.

References

213

Hartford, Beverly S. and Kathleen Bardovi-Harlig 1996 At your earliest convenience: A Study of Written Student Requests to Faculty. In Bouton, Larry F. Pragmatics and Language Learning. Vol.7. 55-71. Hatch, Evelyn 1978 Second Language Acquisition. A Book of Readings. Rowley, MA: Newbury House. Hersch, Henry and Alfonso Caramazza 1976 A fuzzy set approach to modifiers and vagueness in natural language. Journal of Experimental Psychology: General. 105/3:254276. Hinkel, Ely 1995 The use of modal verbs as a reflection of cultural values. TESOL Quarterly, Vol. 29. No.2. 325-243. Hitti, Phillip 1958 History of the Arabs. New York: Oxford University Press. Holland, D. and N. Quinn (eds.) 1987 Cultural models in language and thought. Cambridge: Cambridge University Press. Honeyfield, John G. 1987 Word frequency and the importance of context in vocabulary learning. In Michael H. Long and Jack C. Richards (eds.) Methodology in TESOL. New York: Newbury House: 318-325. House, Julianne 1989 Politeness in English and German: The functions ofplease and bitte. In S. Blum-Kulka, J. House and G. Kasper (eds.) Cross-cultural Pragmatics: Requests and Apologies. Norwwod, NJ: Ablex. Hymes, Dell 1972 Models of interaction of language and social life. In John Gumperz and Dell Hymes (eds.) Directions in Sociolinguistics. New York: Holt, Rinehart and Winston. Irujo, Suzanne 1993 Steering clear: Avoidance in the production of idioms. IRAL XXXI/3.: 205-219. James, William 1890/1950 The principles ofpsychology. Vol. 1. New York: Dover Publications. (Original work was published in 1890. New York: Henry Holt and Co.) Johnson, Mark 1987 The Body in the Mind. Chicago: University of Chicago Press. Johnson, J. M. 1996 Metaphor interpretation by second language learners. The Canadian Modern Language Review. 53:219-241.

214

References

Jordens, Peter 1996 Input and instruction in second language acquisition. In Peter Jordens and Josian Lalleman (eds.) Investigating second language acquisition. Studies on language acquisition. Berlin: Mouton de Gruyter. Jordens, Peter and Josian Lalleman (eds.) 1996 Investigating second language acquisition. Studies on language acquisition. Berlin: Mouton de Gruyter. Kachru, Yamuna 1986 Applied Linguistics and Foreign Language Teaching: A NonWestern Perspective. Studies in the Linguistic Sciences. Vol. 16. No. 1. 35-51. Kaplan, Robert 1966 Cultural thought patterns in intercultural education. Language Learning. 16:1-20. 1990 Writing in a Multilingual/Multicultural Context: What's Contrastive about Contrastive Rhetoric? The Writing Instructor. Fall 1990:7-18. Kasper, Gabriele 1996a Website: www.lll.hawaii.edu/nflrc/Strategies.html 1996b Linguistic etiquette. In Florian Coulmas (ed.) Handbook of Sociolinguistics. Oxford: Blackwell: 374-385. Kasper, Gabriele and Shosana Blum-Kulka 1993 Interlanguage Pragmatics. Oxford: Oxford University Press. Kay, Paul 1983 Linguistic competence and folk theories of language: Two English hedges. Proceedings of the Ninth Annual meeting of the Berkeley Linguistics Society, 128-137. Kecskes, Istvan 1995 Concept Formation of Japanese ESL/EFL Students. Second Language Research in Japan. IUJ. Vol. 6. 130-149. 1997 A Cognitive-Pragmatic Approach to Situation-Bound Utterances. Paper presented to Chicago Linguistics Society. March 7, 1997. 1998a The State of L1 Knowledge in Foreign Language Learners. WORD, Journal of the International Linguistics Association. Vol. 49. No. 3. December. 1998. 321-340. 1998b Situation-Bound Utterances from an Interlanguage Perspective. In Jef Verschueren (ed.) Pragmatics in 1998. Vol. 2. Antwerp: International Pragmatics Association. 299-310. 2000a A Cognitive-Pragmatic Approach to Situation-Bound Utterances. Journal of Pragmatics. 2000. Vol. 32. No. 6: 605-625. 2000b Conceptual Fluency and the Use of Situation-Bound Utterances in L2. Links and Letters. Vol. 7: 143-158. 2001 The Graded Salience Hypothesis in Second Language Acquisition. In: Martin Putz, Susanne Niemeier and Rene Dirven (Eds.) Applied Cognitive Linguistics. Vol. I. Berlin/New York: Mouton de Gruyter: 249-271.

References

215

Kecskes, Istvan and Tunde Papp 2000a Foreign Language and Mother Tongue. Mahwah, NJ: Lawrence Erlbaum Associates. 2000b Metaphorical Competence in Trilingual Language Production. In Cenoz, J. and U. Jessner (eds.) 2000. Acquisition of English as a Third Language",Clevedon: Multilingual Matters. Pp. 99-120. Forthcoming. How to Demonstrate the Conceptual Effect of L2 on LI? Methods and Techniques. In: Vivian Cook (ed.) The Effect ofL2 on LI. Clevedon: Multilingual Matters. Kecskes, Istvan and Melissa Moyer 2002 An Ethno-Conceptual Approach to Code-Switching. Paper presented at the "sociolinguistics Symposium 14." in Gent, Belgium. Kellerman, Eric 1983 Now you see it, now you don't. In Gass, S. and L. Selinker (eds.) Language Transfer in Language. Rowley, MA: Newbury House.l 12-133. 1984 The empirical evidence for the influence of the LI in interlanguage. In Davies, Α., Croper, C. And A. Howatt (eds.) Interlanguage. Edinburgh: Edinburgh University Press. 98-122. Keysar, B. 1989 On the functional equivalence of literal and metaphorical interpretations in discourse. Journal of Memory and Language. 28,375-385. Kiefer, Ferenc 1985 How to account for situational meaning? Quaderni di semantica. 2/85. 288-295. 1995 Situational Utterances. Keynote presented to the 5th International Pragmatics Conference at Brighton, UK. (Manuscript). Kintsch, W. and E. Greene 1978 The role of culture-specific schemata in the comprehension and recall of stories. Discourse Processes. 1:1-13. Kluckhohn, Florence 1963 Some Reflections on the Nature of Cultural Integration and Change. In E. Tiryakian (ed.) Sociological Theory, Values andSociocultural Change: Essays in Honor of P.A. Sorokin. New York: Free Press: 17-28. Kluckhohn, Florence and R. Strodtbeck 1961 Variations in Value Orientation. Evanston, IL: Row Peterson. Kövecses, Zoltan 1993 Minimal and full definition of meaning. In R.A. Geiger and B. Rudzka-Ostyn (Eds.) Conceptualization and mental processing in language. Berlin: Mouton. 247-266. 2000 Metaphor and Emotion. Cambridge: Cambridge University Press. Kövecses, Zoltan and Peter Szabó 1996 Idioms: a View from Cognitive Semantics. Applied Linguistics. Vol. 17. No.3. 326-355.

216

References

Krashen, Stephen and Robin Scarcella 1978 On routines and patterns in language acquisition and performance. Language Learning. 28: 283 300. Kroll, Judith and Stewart, E. 1994 Category interference in translation and picture naming: Evidence for asymmetric connections between bilingual memory representations. Journal of Memory and Language, 33, 149-74. Kroll, Judith and Tokowitz, Natasha 2001 The development of conceptual representation for words in a second language. In: Janet L. Nicol (ed.) One mind, two languages. London: Blackwell: 49-72. Lakoff, George 1987 Women, Fire, and Dangerous Things. Chicago: University of Chicago Press. Lakoff, George and Mark Johnson 1980 Metaphors We Live By. Chicago: University of Chicago Press. Laksmanan, Usha and Larry Selinker 2001 Analyzing interlanguage: how do we know what learners know? Second Language Research. 17.4: 393-420. Lamberts, Koen and David Shanks 1997 Knowledge, concepts and categories. Cambridge, MA: MIT Press. Langacker, Ronald W. 1987 Foundations of Cognitive Grammar, Volume 1, Theoretical Prerequisites. Stanford: Stanford University Press. Larsen-Freeman, Diane and Michael Long 1991 An Introduction to Second Language Acquisition Research. London/New York: Longman. Lebra, T.S. 1974 Reciprocity and the asymmetric principle: An analytical reappraisal of the Japanese concept of "on". In T. S. Lebra and W.P. Lebra (eds.) Japanese culture and behavior: Selected readings. Honolulu: University of Hawaii Press: 192-207. Lehrer, Adrienne and Eva F. Kittay (eds.) 1992 Frames, fields and contrasts. Hillsdale, NJ: Lawrence Erlbaum. Levenston, E. 1979 Second Language Acquisition: Issues and Problems. Interlanguage Studies Bulletin. 4. 147-160. Levine, David 1985 The Flight from Ambiguity. Chicago: The University of Chicago Press. Levinson, Stephen C. 1983 Pragmatics. Cambridge: Cambridge University Press. Lieven, E., Pine, J.M., and Barnes, H.D. 1992 Individual differences in early vocabulary development: redefining the referential-expressive distinction. Journal of Child Language. 19:287-310.

References

217

Lieven, E., J. Pine, and G. Baldwin 1997 Lexically-based learning and early grammatical development. Journal of Child Language. 24. 187-219. Lodge, David 1992 Paradise News. Harmondsworth: Penguin. Luke, Α., A. W. McHoul and Jacob Mey 1990 On the limits of language planning: Class, state and power. InR. B. Baldauf Jr. and A Luke (eds.) Language Planning and Education in Australasia and the South Pacific. Clevedon: Multilingual Matters. Mace-Matluck, B.J., Hoover, W.A. and Calfee, R.C. 1984 Teaching reading in bilingual children. Final report. Austin, TX: Southwest Educational development Laboratory. Maeshiba, N., Yoshinaga, N., Kasper, G. and Ross, S. 1996 Transfer and proficiency in interlanguage pragmatics. In J. Neu and S. Gass (eds.) Speech acts across cultures. Berlin: Mouton de Gruyter: 155-187. Maley, Allen 1980 Teaching for Communicative Competence: Reality and Illusion. Studies in Second Language Acquisition. Vol. 3. No. 1: 11-16. Malinowski, B. 1923 The problem of meaning in primitive languages. In Ogden and Richards (eds.) The meaning of meaning. London: Routledge and Kegan Paul. Malt, B. C. 1994 Water is Not H20. Cognitive Psychology. 27: 41-70. Matisoff, James A. 1979 Psycho-ostensive Expressions in Yiddish. Philadelphia: ISHI Publications. Meara, Paul 1980 Vocabulary acquisition: neglected aspect of language learning. Language Teaching and Linguistics: Abstracts. 13., 221-242. Mervis, C.B. and Eleanor Rosch 1981 Categorization of natural objects. Annual Review of Psychology. 32:89-115. Mey, Jacob 1985 Whose Language? A Study in Linguistics Pragmatics. Amsterdam?Phialdelphia: John Benjamins. 2001 Pragmatics. London: Blackwell. Michaels, Sarah 1981 "Sharing time": Children's narrative style and differential access to literacy. Language in Society. 10. 423-442. Michaels, S. 1986 Narrative presentations: An oral preparation for literacy with first graders. In J. Cook-Gumperz (ed.) The social construction of literacy. Cambridge: CUP. 84-116.

218

References

Miller, Jim 1998

Acquiring Spontaneous Spoken Language: the Role of Simple Syntax and Ready-Made Phrases. Paper presented at the 6th International Pragmatics Conference in Reims, France. Miller, Jim and Regina Weinert 1998 Spontaneous Spoken Language. Syntax and Discourse. Oxford: Oxford University Press. Minsky, M. 1985 The society of mind. New York: Simon and Schuster. Mitchell, Rosamond and Florence Myles 1998 Second Language Learning Theories. London, New York: Arnold. Moon, Rosamund 1992 Textual Aspects of Fixed Expressions in Learners' Dictionary. In Arnaud, P.J.L. and H. Bejoint (eds.) Vocabulary and Applied Linguistics. London: Macmillan. Morris, William, (ed.) 1976 The American Heritage Dictionary of the English Language. Boston: Houghton Mifflin Company. Moyer, Melissa 1992 Analysis of code-switching in Gibraltar. Unpublished PhD Dissertation. Universität Autonoma de Barcelona. 1998 Bilingual conversation strategies in Gibraltar. In Peter Auer (ed.) Code-switching in conversation. Language, interaction and identity. London: Routledge: 212-234. Myles, F., Hooper, J. and Mitchell, R. 1998 Rote or rule? Exploring the role of formulaic language in classroom foreign language learning. Language Learning. 48 (3): 323-363. Myles, F., Mitchell, R. and Hooper, J. 1999 Interrogative chunks in French L2 : a basis for creative construction? Studies in Second Language Acquisition. 21 (1): 49-80. Nattinger, James R. and Jeanette S. DeCarrico 1992 Lexical Phrases and Language Teaching. Oxford: Oxford University Press. Nelson, K. 1981 Individual differences in language development: implications for development and language. Developmental Psychology. 17 (2): 170-187. Nunberg, G., Ivan Sag and Terry Wasow 1994 Idioms. Language. 70 (3), 491-538. Nuyts, Jan 1992 Aspects ofa Cognitive-Pragmatic Theory of Language. Amsterdam: John Benjamins. Ochs, E. 1988 Culture and language development: language acquisition and language socialization inaSamoan village. Cambridge: Cambridge University Press.

References

219

Ochs, E. and Β. Schieffelin (eds.) 1984 Language acquisition and socialization: three developmental stories and their implications. In Shweder, R. and LeVine, R. (eds.) Culture theory: essays on mind, self and emotion. Cambridge: Cambridge University Press: 276-320. Odlin, Terence 1989 Language Transfer. Cambridge: Cambridge University Press. Okabe, R. 1983 Cultural Assumptions of East and West: Japan and the U.S. In W. Gudykunst (ed.) Intercultural Communication Theory. Beverly Hills, CA: Sage. 1628.

Pavlenko, Aneta 1996 Bilingualism and Cognition: Concepts in the Mental Lexicon. In Pavlenko, Aneta and Rafael Salaberry (eds.) Cornell Working Papers in Linguistics. Volume 14: Papers in Second Language Acquisition and Bilingualism. 1-22. Pawley, Andrew and Syder, F.H. 1983 Two puzzles for linguistic theory: Nativelike selection and nativelike fluency. In Richards, J.C. and Schmidt, R. W. (eds.) Language and Communication. London: Longman. 191-226. Pennycook, Alastair 1994 Incommensurable Discourse? Applied Linguistics. Vol. 15. No.2: 115-138. Perkins, M. 1998 Productivity and formulaicity in language development. In C. Schelletter, C. Letts and M. Garman (eds.) Proceedings of the 1996 Child Language Seminar. Reading: University of Reading. Pierce, Bonny Norton 1995 Social identity, investment, and language learning. TESOL Quarterly. Vol. 29. No. 1:9-31. Pine, J. and E. V. Lieven 1993 Reanalysing rote-learned phrases: individual differences in the transition to multi-word speech. Journal of Child Language, 20, 551-571. 1997 Slot and frame patterns and the development of the determiner category. Applied Psycholinguistics, 18, 123-138. Platt, E. and Troudi, S. 1997 Mary and her teachers: a Grebo-speaking child' place in the mainstream classroom. Modern Language Journal. 181: 28-49. Plunkett, K. 1993 Lexical segmentation and vocabulary growth in early language acquisition. Journal of Child Language. 20:43-60.

220

References

Politzer, Robert 1978 Errors of English speakers of Germanas perceived and evaluated by German natives. Modem Language Journal. 62,253-261. Purves, Allen C. 1986 Rhetorical communities, the international student, and basic writing. Journal of Basic Writing. Vol. 5. No.l. 38-51. Radden, Gunther 1995 Motivation metaphorized: The case of coming and going. In B. Korponay and P. Pelyvas (eds.) Studies in Linguistics III. Debrecen: Kossuth University. Raupach, Manfred 1984 Formulae in second language speech production. In H.W. Dechert, D. Mohle, and M. Raupach (eds.) Second Language Productions. Tubingen: Narr. 114-137. Rescorla, Leslie and S. Okuda 1984 Lexical development in second language acquisition: initial stages in a Japanese child's learning English. Journal of Child Language, 11,689-696. 1987 Modular patterns in second language acquisition. Applied Psycholinguistics. 8. 281-308. Rosch, Eleanor and C. B. Mervis 1975 Family resemblance: studies in the internal structure of categories. Cognitive Psychology. 7:573-605. Ross, James 1992 Semantic Contagion. In Lehrer, A. and E.F. Kittay (eds.) Frames, fields, and contrasts. Hillsdale, NJ: Lawrence Erlbaum. Rowland, C. F. and J. Pine 1998 Subject-auxiliary inversion errors and wh-question acquisition: "what do children know?". Child Language Seminar, University of Sheffield. Sato, Charlene J. 1989 A nonstandard approach to standard English. TESOL Quarterly. Vol. 23. No.2. 259-283. Schachter, J. 1983 A new account of language transfer. In S. Gass and L. Selinker (eds.) Language transfer in language learning. Rowley, MA: Newbury House. 98-111. Schmidt, Richard W. 1983 Interaction, acculturation, and the acquisition of communicative competence: a case study of an adult. In N. Wolfson and E. Judd (eds.) Sociolinguistics and language acquisition. Rowley, MA: Newbury House: 137-174. 1993 Consciousness, learning and interlanguage pragmatics. In G. Kasper and S. Blum-Kulka (eds.) Interlanguage Pragmatics. Oxford: OUP. 21-42.

References Searle, John 1975

221

Indirect speech acts. In P. Cole and J. Morgan (eds.) Syntax and Semantics. Vol. 3. Speech Acts. New York: Academic Press.

Selinker, Larry 1972 Interlanguage. International Review of Applied Linguistics. 10:209231. 1992 Rediscovering interlanguage. London: Longman. Shankoff, Serge 2002 Review of Harkins and Wierzbicka (eds.) 2001 Emotions in Crosslinguistic Perspective. Linguist List. April 24, 2002. Shapira, R.G. 1978 The non-learning of English: case study of an adult. In E.M. Hatch (ed.) Second language acquisition: a book of readings. Rowley, MA: Newbury House: 246-255. Sinclair, John 1991 Corpus, concordance, collocation. Oxford: Oxford University Press. Skutnabb-Kangas, Touva and P. Toukomaa 1976 Teaching migrant children mother tongue and learning the language of the host country in the context of socio-cultural situation of the migrant family. Tampere, Finland: Tukimuksia Research Reports. Slobin, Dan 1991 Learning to think for speaking: Native language, cognition, and rhetorical style. Pragmatics. 1.7-25. 1996 From 'thought and language' to 'thinking for speaking'. In Gumperz, J.J. and S. Levinson (eds.) Rethinking linguistic relativity. 7097. Cambridge, UK: Cambridge University Press. Smith, E. E. and S. A. Sloman 1994 Similarity versus rule-based categorization. Memory and Cognition. 22: 337-86. Smith, Linda B. and Larissa Κ. Samuelson 1997 Perceiving and remembering: Category stability, variability and development. In Koen Lamberts and David Shanks (eds.) Knowledge, concepts, and categories. Cambridge, MA: MIT Press: 161196. Sorhus, Helen B. 1977 To hear ourselves - implications for teaching English as a second language. English Language Teaching Journal. 31 (3) April: 211221.

Stalnaker, R. 1991 Stock, B. 1991

Pragmatic presuppositions. In S. Davis (ed.) Pragmatics. Oxford: Oxford University Press. 471-484. The Implications of Literacy. Princeton: Princeton University Press.

222

References

Strauss, C. and Quinn, N. 1997 A cognitive theory of cultural meaning. Cambridge: Cambridge University Press. Stewart, Edward 1972 American Cultural Patterns. Chicago: Intercultural Press. Swan, Michael 1985a A critical look at the Communicative Approach (1). ELT Journal. Vol. 39/1. January. 2-11. 1985b A critical look at the Communicative Approach (2). ELT Journal. Vol. 39/2. April. 76-87. Sweetser, Eve 1990 From Etymology to Pragmatics. Cambridge: Cambridge University Press. Tannen, Deborah 1985 Cross-cultural Communication. In T.A. Van Dijk (ed.) Handbook of Discourse Analysis, vol.4. 203-215. London: Academic Press. Tannen, Deborah and Oztek, Piyale, C. 1981 Health to Our Mouths. Formulaic Expressions in Turkish and Greek. In Florian Coulmas (ed.) Conversational Routine. The Hague: Mouton. Taylor, John R. 1993 Some pedagogical implications of cognitive linguistics. In R.A. Geiger and B. Rudzka-Ostyn (eds.) Conceptualizations and Mental Processing in Language. Berlin/New York: Mouton de Gruyter. 200-223. 2000 The network model and the two-level model in comparison. In Bert Peeters (ed.) The lexicon-encyclopedia interface. Amsterdam: Elsevier: 115-143. Forthcoming Concepts, or: what is it that a word designates? Manuscript. University of Otego. Tietze, Andreas 1963 Turkish Literary Reader. Bloomington: Indiana University. Tomasello, M. 2000 Do young children have adult syntactic competence? Cognition 74. 209-253. Tomasello, M. and Brooks, P.J. 1999 Early syntactic development: a construction grammar approach. In M. Barrett (ed.) The development of language. Hove: Psychology Press: 161-190. Treichler, Paula A. and Francine Wattman Frank 1989 Introduction: scholarship, feminism, and language change. In Frank, F. W. and P. A. Treichler (eds.) Language, gender, and professional writing: Theoretical Approaches and Guidelines for Nonsexist Usage. New York: Modern Language Association of America. Commission on the Status of Women in the Profession: 134.

References Turner, Mark 1987

223

Death is the mother of beauty. Chicago: University of Chicago Press. Van Dijk, Teun A. 1980 Macrostructures. Hillsdale: Lawrence Erlbaum Associates. Verhoeven, L. 1994 Transfer in Bilingual Development: The Linguistic Interdependence Hypothesis Revisited. Language Learning. 44:3. 381-415. Verstraten, Linda 1992 Fixed Phrases in Monolingual Learners' Dictionaries. In Arnaud, P.J.L. and H. Bejoint (eds.) Vocabulary and Applied Linguistics. London: Macmillan. Vihman, Marilyn M. 1982 Formulas in First and Second Language Acquisition. In Obler, L.K. and L.Menn (eds.) Exceptional Language and Linguistics. New York/London: Academic Press. Violi, Patrizia 2000 Prototypicality, typicality, and context. In Albertazzi, Liliana (ed.) Meaning and Cognition. Amsterdam: John Benjamins. 103-123. Warren, B. 1992 Sense developments: a contrastive study of the development of slang senses and novel standard senses in English. Stockholm: Almquist and Wiksell International. Weinert, Regina 1995 The Role of Formulaic Language in Second Language Acquisition: A Review. Applied Linguistics. Vol. 16. No. 2. 180-205. 2000 Introduction. In Kecskes, I. and R. Weinert (eds.) Formulaic language in first and second language production. (Unpublished manuscript) White, Lydia 1996 Universal Grammar and second language acquisition: current trends and new directions. In Ritchie, W. and Bhatia, T. (eds.) Handbook of second language acquisition. San Diego: Academic Press: 85120. Widdowson, Henry 1980 Models and fictions. Applied Linguistics. 1/2. 165-170. Wierzbicka, Anna 1992 Semantics, Culture, and Cognition. New York/Oxford: Oxford University Press. 1994. "Cultural Scripts": A Semantic Approach to Cultural Analysis and Cross-Cultural Communication. In Pragmatics and Language Learning. Vol. 5. 1-25. 1996 Semantics. Primes and universals. Oxford: Oxford University Press.

224

References

Willett, J. 1995

Becoming first graders in an L2: an ethnographic study of L2 socialization. TESOL Quarterly. Vol. 29 (3): 473-503. Williams John N. 1992 Processing polysemous words in context. Evidence for interrelated meanings. Journal of Psycholinguistic Research. 21:193:218. Willis, Dave 1996 The Lexical Syllabus. London: Collins Cobuild. Winner, Ellen 1982 Invented worlds: The Psychology of the arts. Cambridge, MA: Harvard University Press. Wittgenstein, Ludwig 1953 Philosophical investigations. Oxford: Blackwell. 1980 Remarks on the philosophy of psychology. Oxford: Blackwell. Wood, M 1991 A Definition of Idiom. Manchester: Centre for Computational Linguistics. Wray, Alison 1998 Formulae in LI, L2 Adult Native Speakers and Aphasies: Are They All the Same? Paper presented at the 6th International Pragmatics Conference in Reims, France. 2002 Formulaic Language and the Lexicon. Cambridge: Cambridge University Press. Wright, Georg Henrik von 1963 Norm and action. A logical enquiry. London: Routledge and Kegan Paul. Yorio, Carlos 1980 Conventionalized forms and the development of communicative competence. TESOL Quarterly, XIV/4: 433-42. 1989 Idiomaticity as an indicator of second language proficiency. In K. Hyltenstam and L.K. Obler (eds.) Bilingualism across the lifespan. Cambridge: Cambridge University Press. Yoshida, K. 1990 Knowing vs. behaving vs. feeling: Studies on Japanese bilinguals. In L.A. Arena (ed.) Language proficiency. New York: Plenum Press. Zaharna, R. S. 1991 The ontological Function of Intercultural Communication. : A CrossCultural Analysis of American and Palestinians. Howard Journal of Communications. 3: 87-98. 1995 Understanding cultural preferences of Arab communication patterns. Public Relations Review. Internet Site.

Subject Index Articulator, 26

Bilingual, development, 49, 91, 192, 193 Bidirectional influence, 2, 156, 192

Categorization, 4, 7, 15, 27, 34-5, 59, 65, 121 Code-switching, 11, 158 Cognition, 7, 20, 22, 60, 79, 93,203 Cognitive, dimension, 8 mechanisms, 12, 13, 92, 96, 102-3, 164, 186-8 Cognitive-pragmatic, approach, 7, 8, 75,92 framework, 12, 92, 160 perspective, 8 Common Underlying Conceptual Base (CUCB), 10, 12, 126-7, 157-9, 192-5, 198-9 Compositionality, 4, 51, 66-7, 122, 164, 177 Concept, mediation, 24 Conceptual, base, 10, 11, 17, 76, 157, 198 components, 44 content, 16, 23-4, 36, level, 2, 13, 15, 17, 18, 19, 26, 43,48-9, 54 motivation, 99, 100, 102 representation, 16, 18, 23, 59, 70 shift, 53 socialization, 11, 13, 155-60, 164, 170, 173, 174, 179, 182-3, 188, 190-5, 197-9, 204 system, 8, 11, 17, 18,20, 22, 25-6, 28, 33-4, 47, 64, 69, 100,

102, 156-8, 163-4, 179, 190, 198 Conceptualizer, 25-6 Configurationality, 87, 91 Context, externalist perspective, 56-9, 204 internalist perspective, 56-7, 60, 204 Continuum, 4, 12, 17, 18, 71, 80, 87, 99, 101, 135, 165-6, 170 Conventionalized, conceptualizations, 161 Conversational routine, 2, 3, 6, 12, 82 Consense, 42-6, 49, 51-5, 57, 61, 668, 70, 104-5, 107-8, 116-7, 124, 126 Coresense, 42-6, 48-55, 57-8, 60-2, 66, 73, 107, 116-21, 131 Creativity, 135-40, 151, 153, 177, 197-8, 205 Cross-linguistic, approach, 9 Cultural specificity, 182-3, 185 Culture, doing and being, 168-70, 172 direct and indirect, 168, 170, 172 high-context and low-context, 165, 167-8, 170, 172 linear and non-linear, 165, 169, 170, 172 Culture-specific conceptual property (CSCP) 37,424, 50-1,54-5,66, 76, 115-6, 120, 156, 184

Diachronic, 18-9, 42-5,48, 52, 62, 71, 102, 197 Discourse, 6,40, 44-5, 55-6, 61,115, 120, 135-6, 138, 141, 147, 163,

226

Subject index

165, 180, 183, 185-6, 189, 191 Distance, cultural, 86, 88-91 language, 86, 88, 90-1 Dual Language System, 10, 13, 146, 151, 152, 158, 197, 205 Dynamic Model of Meaning (DMM), 13, 19, 32,35-6, 40-2, 44, 4749, 53,55, 57-8,61,66, 72, 107, 115-6, 118-20, 161, 197

Economy, 79, 140, 143 Encyclopedic, 15, 16, 18, 19, 21, 256, 36-8, 53, 66, 100, 197, 203 Ethnocentrism, 157 Extralinguistic, 9, 15, 16, 55, 58, 60-4, 182

Formulaic expressions, 2, 3, 4, 12, 79, 113, 119, 135, 137-8, 141, 1456, 148-9, 152-3, 155, 177 language, 2,4, 12, 141, 143, 14750, 155 sequences, 115, 141, 144, 147-8, 150-1 speech, 146 Formulaicity, 135, 140-1, 143, 151, 177, 197 Formula-specific pragmatic properties (FSP), 117-22, 156, 178, 184 Formulator, 25-6 Frame, 26, 45, 60-1, 65-6, 72, 91, 1046, 109, 115-6, 120, 149, 178, 186 Functional, aspect, 3,4 continuum, 4

Graded salience hypothesis, vii, 12, 13,66,71-2, 80, 96, 122-3

Idioms, 6, 7, 14, 71-2, 85, 91-3, 96, 99, 100-2, 106-8, 115, 122, 142, 149,153-4, 164 Idiomatic expressions, 6, 91, 100, 102, 104, 106, 150 Implicatures, conventional, 94-6 conversational, 94-5 Institutionalized, expression, 5 Intercultural style hypothesis, 192 Interdependency, 13, 92 Interface, 25, 32, 42, 46, 52, 55, 147, 197 Interlanguage, 9, 154, 196 Invariant, 17, 32, 35,43, 48, 107, 197

Knowledge, actual, 25, 59 background, 11, 36-7, 153 conceptual, 17, 22, 85 conventional, 100, 103-4 declarative, 25-6 encyclopedic, 16, 18, 37-8, 53, 100, 197 linguistic, 15-17,21-2, 25, 155, 180, 197 pragmalinguistic, 177 prior, 64, 73, 89 real-world, 11 situational, 25

Language, configurational, 86-7 non-configurational, 87 Language-specific, 28, 50, 91, 190 Lexical association, 24 item, 4, 13, 16,18, 19, 25-6, 3Ιό, 41-2, 46, 48-51,53,54, 59, 61-2, 72, 83, 104, 142, 158, 193, 197 Lexicalization, 60, 69, 82, 91, 98, 161

Subject index Linguistic, behavior, 69, 135, 195 level, 2, 18, 49,53 typology, 86 unit, 2 , 4 , 5 , 13, 15,22-3,44, 55, 92, 143, 160, 198 Linguistics, cognitive, 7, 8, 68 generative, 7

Meaning, conceptual, 22, 32, 41, 52-3 core, 42,44-5, 53 figurative, 68-9, 71, 120, 124-6, 128-32, 134, 197 linguistic, 17, 27, 52, 60, 92 literal, 6, 53, 66, 68, 70-2, 75, 106, 118, 120, 122, 125, 12831, 134, 163, 184 referencial, 6, 155 situated, 19, 43-5, 60-1, 1045, 120 Mental, encyclopedia, 21 dictionary, 21 lexicon, 18, 25-6, 74, 115, 126, 133, 145 Metaphors, 39, 68-9, 71, 92-3, 99, 100,102,109, 126,218 Metaphorism, 39,40 Metonymy, 95, 100, 103-4, 164 Minimalist, Program, 16, 17 Theory, 16 Mode, bilingual, 11, 12 language, 11, 12, 158, 197 monolingual, 11, 12, Monolingual, system, 11 view, 9 Multicompetence, 10 Multi-dimensionality, 46, 53, 63 Multilingual, perspective, 9, 199 Multilingualism, 22

227

Naming, 33 Nativelike proficiency, 195 Natural Semantic Metalanguage (NSM), 32, 35

Onomasiology, 32-3 Overextension, 23 Oversimplification, 188, 191

Patterns, 2, 16, 43, 60, 65, 139, 142, 144, 147, 154, 160, 164-168, 170, 173, 182-3, 191, 199 Phatic communion, 80 Phraseological units, 99, 101-2, 109, 122 Pragmatic act, 92, 116, 136, 164 competence, 4,191 dimension, 4, 8 expression, 4, 116, 156 extension, 119-21,205 function, 5, 6, 13, 79, 92, 106, 112, 115-6, 120, 122, 148, 163-4, 170 idioms, 14, 99, 102, 106, 112, 134, 153 unit, 4 , 5 , 9 , 15,44, 68, 93-4, 96, 156, 161, 195, 197, 199 Pragmatics, 3, 7, 16, 136, 196-7 Pragmeme, 116 Prefabricated expressions, 2, 104, 115, 198 Principle, idiom, 39, 140, 148, 178 of relevance, 67 open choice, 140, 178 Processing, parallel, 72, 109, 122, 128-9, 163, 190 Processing, sequential, 132 Prototype theory, 35-6, 63 Psychotypology, 87-8

228

Subject index

Reconceptualization, 25 Reflexivity, 78 Rituals, interaction, 5, 112 situation-bound, 112-4, 162, 168, 198 speech, 14 Routines, conversational, 3, 6, 12, 82 cultural, 31 linguistic, 2 situation-bound, 112, 169, 198 Salience, 7, 12, 13, 19, 55, 63, 66, 68, 71-6, 80, 96, 122-3, 125-6, 1334, 164, 190, 197-9 Second language acquisition, 3, 7, 9, 10, 24, 49, 134, 142, 145, 151, 155, 179, 197, 199 Semantics, conceptual, 20, 22,41, 53, 66, 70, 101 generative, 17 lexical, 21-2, 53,63,66, 70, 101 linguistic, 16, 44 natural language, 26 Semasiology, 32-3 Social, identity, 175-6, 183 values, 2, 163, 198 Socio-cultural, background, 11, 76, 91,97, 129, 152, 160, 182 system, 79, 81 Socialization, conceptual, 11, 13,

155-60, 165, 171, 174-5, 179, 182-3, 188, 190-5, 197-9, 205 language, 155-6, 160 Synchronic, 18, 19,42-5, 50-3, 62, 71, 102, 109, 197

Topic-associating, 166-8, 171 Topic-centered, 166-9, 171 Transfer, 88, 90-2, 153, 160, 172, 187, 191-4, 196 Turn-taking, 6

Underextension, 23

Verbal predicate, 38 Verbosity, 189, 191

Word, charged, 40-1, 203 loaded, 41, 203 Word-specific semantic properties (WSP), 24,38-9, 40-1,43,49, 54, 116, 203