You are on page 1of 196

Developing Narrative and Discourse Competence


This Page Intentionally Left Blank

Developing Narrative and Discourse Competence


Edited by
Pennsylvania State University, University Park

BogaziGi University, Istanbul, Turkey


University o British Columbia, Vancouver, Canada f


Mahwah, NewJersey


The final camera copy for this work was prepared by the editors and therefore the publisher takesno responsibility for consistency or correctnesstypographical style. of However, this arrangement helps make publicationof this kindof scholarship to possible. Copyright 0 zoo 1 by Lawrence Erlbaum Associates, Inc. All rights reserved. No part of this book may be reproduced form, by in any photostat, microform, retrieval system, or any other means, without the prior written permission of the publisher. Lawrence Erlbaum Associates, Inc., Publishers 10 Industrial Avenue Mahwah, New Jersey 0743 o Cover design by Kathryn Houghtaling Lacey Library of Congress Cataloging-in-Publication Data

0-8058-3292-0 0163-2809

Books published by Lawrence Erlbaum Associates are printed acid-free paper, on and their bindings are chosen for strength and durability.

Printed in the United States of America







Setting the NarrativeScene: How Children Begin to Tell a Story


2 .

Representation of Movement in European Protugese: A Study of Childrens Narratives

H A N N A J A K U B O W I C Z B A T O R ~ OA N D I S A B E L H U B F A R I A

3. Why Young American English-Speaking Children Confuse Anger and

Sadness: A Study Grammar in Practice of


4. A Crosscultural Investigationof Australian and Israeli Parents Narrative

Interactions With Their Children



The Acquisition of Polite Language by Japanese Children




Interactional Processes in the Origins of the Explaining Capacity



7. Childrens Attributionsof Pragmatic Intentions andEarly Literacy


Author Index

Subject Index



Bogaziqi University, Istanbul, Turkey Tel Aviv University, Israel Clark University, Worcester, Massachussetts Open University, Lisbon, Portugal


University o Lisbon, Lisbon, Portugal f


University o British Columbia, Vancouver, Canada f


Keio University, Tokyo,Japan Pennsylvania State University, University Park


University o British Columbia, Vancouver, Canada f


Hebrew University and BjeitBerl College, Israel


Universite' Nancy andUniversite' ParisV-CNRS


Macquarie University and University Melbourne, Australia of

vi i

This Page Intentionally Left Blank


The chapters in volume deal with discourse development, with this an emphasis on narrative, from ages 1 1/2 to 10, ranging over sevenlanguages. They and were developed from 7 of the 276 presentations at the Seventh International Congress of the International Association for the Study Child Language of (IASCL) in Istanbul, Turkey,in July 1996. That meeting was a broadly international assembly of 350 participants from 41 countries, representing more than 30 languages, who are contributing to the development a scientific of tradition in thefields of general linguistics, sociolinguistics, psycholinguistics, and intercultural communication, as they pertain to childrens language acquisition. During theCongress, participants shared exciting research projects and exchanged a broad spectrum of viewpoints with honesty, depth, and respect. Our Turkish hostswere outstanding in their warmth and organization. IASCL continues to invite international applications fromall professionals who share an interest in enhancing the understanding childrens acquisition of and use oflanguages.


This Page Intentionally Left Blank



BogaziFi University

Pennsylvania State University


University of British Columbia

The chapters in this volume reflect recent directions of thinking the area of in childrens discourse development, with an emphasis on narratives. Each contribution shows that empirical workin the last decade has focused finer on distinctions regarding the effects on development of discourse genres, different elicitation techniques and communicative contexts, literacy and schooling and, of course, age, language, and culture. Each chapter addressesissues concerning the interrelations between social, cognitive, and affective capacities and processes in discourse.Finally, each raises theoretically challenging questions regarding how and when new representations are constructed to support new complexities in narrative and discourse more generally. A comprehensive theoretical frame calls for a conceptualization of discourse as an interactional space that promotes the development higher level metalinguistic, metaof representational, and metapragmatic operations. The chaptersby Ruth Berman, Hanna Batorko and Isabel Hub Faria, Gillian Wlgglesworth and Anat Stavans,and Ken Reeder focus on aspects of childrens narrative ability, their productions in differentgenres, in different modes of expression and in relation to the input they in different cultures. They all get look at form:function relations, that is, the interface between specific devices within a language and narrative development. Chaptersby Edy Veneziano and Keiko Nakamura, on the other hand, deal with more specific discourse capacities such as justifications of oppositions anduse of politeness indicators as thesedevelopinthemoregeneralcontextofconversation.Michael Bambergs contribution cuts across the two groups, as it deals with childrens narrative accounts of emotion, from a dynamic discursive but perspective. From a methodological point view, issues of relating to task conditions, levels of assessment,cultural differences, and the like are brought under consideration,


A K S U - K O ~ ,N E L S O N , A N D J O H N S O N

suggesting that someof the factors typically identified as methodological should perhaps be approached from different perspective: They mightreveal undera lying processes themselves be determinants acquisition. or of

With high appropriateness, the book begins with a chapter childrens own on beginnings. Berman (chap.1) analyzes and discusses how children start their oral narratives. She argues that such story beginnings depend heavily on general cognitive development, and provide insight not only into childrens knowledge of narrative structure, but on their story planning also abilities and their representations of their listeners and task. The stories examined are told by children at ages 3, 5, and 9 and adults. Narrative structures, including the specific linguistic forms used to express scene-setting and motivational elements, are compared for narratives based on a wordless picture book(Frog, where are you?, Mayer, 1969) and fora story constructed about a personal experience (a fight) in two modes: speechand writing. These comparisonsare supplemented with results from an additional body Hebrew and English of data from varied storytelling tasks. Berman observes age-related developments in setting the spatiotemporal frame the story, the amount of background for and motivational information about the characters, the demarcation the of setting from the storyline events, metatextual comments relating the narrated events to the storytelling situation, and the formal means used to serve these functions. For example, the proportion of setting clauses increased withage in thepersonalconflictnarratives,butnotinthepicturebooknarratives. Furthermore, use of specific linguistic forms varied with narrative genre and mode of elicitation, suggesting the effects of schooling and literacy (Hicks, 1991; Kuntay, 1997; McCabe &Peterson, 199la). Berman addresses the intriguing theoretical question of whychildren require a long time acquiring cerafter tain forms, including past tense verbs and temporal terms suchafterwards, as before they usethem effectively and flexibly in their dynamically constructed on-line narratives. The findings overall show that speakersrely on the formal options made available to them by the typologicalstructure of their language, and that these are first effectively exercisedat the local clause level, then at the global discourselevel (Berman & Slobin, 1994; Hickmann, 1991,1995). Batoreo and Faria (chap. 2), whoalsoinvestigatedchildrensstory openings, start with an interesting review of how narratives can be begun in ways that utilize the most available, most reliable features of a particular language, in this case, European Portuguese. In accord with the central themes of this volume, they examine the interfaces between language-specific devices, cognitive development, and the initiation and elaboration narratives, using of data from preschoolers, school-age children, and adults. Their methods employ



open-ended, textless storybooks now familiar to narrative development researchers worldwide, the Cat and Horse stories (Hickmann, 1982). Their 5-year-olds proved have available all lexical and morphosyntacticdevices to the used by skillful adult narrators. However, they did not sensitively deploy these devices for the spatiotemporal framingof the stories, introducing the characters, or providing new informationabout themas the narratives proceeded.By age IO, the childrens adaptation of lexical and morphosyntactic devices to narrative purposes was much closer to the adult model, revealing mature form:function relations. The authors affirm the importance of concomitant advances in the cognitive capacity to represent the in termsof his or her listener informational needs and the capacity to organize events in a goal-directed temporal-causal sequence (Berman& Slobin, 1994; Hickmann, 1995; McCabe & Peterson, 19 91a; Slobin, 1996). Open for future specification are how particular discourse-learning contexts and particular cognitive advances after age 5 contribute to advances in the flexibility and sophistication of narratives. Wigglesworth and Stavans (chap. 4) deal with the less studied question of the natureof narrative input children receive from their parents two different in languages andculturalbackgrounds-English-speakingAustralianand Hebrew-speaking Israeli-nicely complementing Bermans chapter. Age- and language-related analysesof how the parents narrated Frog, where are you? (cf. Berman, this volume) to their children revealed differences in the use of a variety of interactive strategies and linguistic acts such as rhetorical questions, personal digressions, affective and evaluative comments and the like, in addition to actually advancing the action the story (McCabe Peterson,lg91b; of & Ninio & Snow, 1996; Scollon & Scollon, 1981). For 3-, 5-, and 7-year-old children and their parents, there were clear cultural differences in interpreting the storytelling task. Althoughall of the story sessions were collaborativeto some degree, the Australian English-speaking parents had a more conversational, dynamic style, suggesting that they took thetask to be an interpersonal exchange activity as well as story construction, whereas the Israeli parents displayed a more strictly literacy-oriented approach in their storytelling. This difference is paralleled in the performance of the children; Israeli 7-year-olds contributed to the actual story line with a literary-oriented performance, whereas their English-speaking agemates conversed with parents as well as the actively adding to the plot line. Cultural styles, thus, a bearing on the have narratives children eventually produce, and input important not only in is exposing children to language-specific form:function relations in narrative, but also to cultural expectations and specific modes ofexecuting the task. Wlgglesworth and Stavans also show that despite thesedifferences, universal patterns canbe identified in the behaviorof parents of 5-year-olds who, both in groups, distinctly emphasized story structure and story organization in telling their children stories. Noting the critical status 5 years of age for of developments in discourse-organizational capacities, as reflected in the shift



from a fragile plot-structure to a hierarchically organized one (Berman & Slobin, 1994), the authors suggest that parents in both cultures are equally sensitive to the cognitive-linguistic readiness of their children and respond accordingly. What remains to be furtherspecified is the particular interplay between universaland culturally specific determinants of behavior. Reeders contribution (chap. also has childrens narrative competenceas a 7) main focus, but in the written mode. Reeder investigated school childrens writing skills in two genres (story and expository description) a function retell as of their ability to attribute pragmatic intentions to others. Attribution ability was determined by interviewing the children about speakers intent after they viewed a puppet scenario. Reeder found that advances in understanding pragmatic intent supported advances in written narratives at around 8 7 to years in both genres. The effects werestronger for personalized narrativesthan expository descriptions, which have less interpersonal and contextual support. Skill in identifying the intentionsof speakers requiresunderstanding therelation between the linguistic features utterances andaspects of participants of epistemic states (knowledge, attitudes, assumptions), that metapragmatic is, competence. As Reeder shows,the effects of a complex set of developmental advances such as increased differentiationsof the feelings, motives, intents, agreements, and conflicts of varied participantsis reflected in increased narrative sophistication inboth spoken and written modes. Taking the broaderview across allthe chapters in this volume, one theme is strongly evident. To wit, new cognitive, social, and language representationsare only graduallywoven into flexible, differentiated narrative skills fitted to many varied contexts goals. and


The studies summarized earlier, which explore various aspectschildrens of developing narrative abilities, emphasize bidirectionalinfluences, particularly between linguistic and cognitive development. The focus research reported of in the following three chapters, on the other hand, requires, in addition, a consideration of the interrelations between social and emotional development, because advances in these domains reciprocally feed into and bring about also changes in languageand cognition. The inquiries into the expression of anger and sadness (chap.3),justification of opposition (chap.6 ) )and the learning of forms of politeness (chap. all placelanguage development within broader 5) the context of social-communicative development whereby child learns to take the up different positions or roles. In chapter 3 Bamberg takes a discursive approach to the development of emotion, arguing that language use plays a dynamic, foundational role in the social construction of emotions suchanger and sadness. He defines emotions as as meaningful positions taken by a personto signify self-other relationships, up



performed in the form of discourse actions. His analysis of the grammarsof anger and sadness (i.e., the linguistic construction types that mark the discursive positions of being angry and being sad) highlights the interface between linguistic structures, discourse genres (personal involvement narrative vs. explanatory discourse) and social positions in the discourse of children between ages 4 and 11. The empirical data revealing these processes are childrens explanations, for example, whatit means to be sad, and narratives centered on personal experiences anger and sadness. The context of creating of a first-person sadness narrative proved problematic for to 6 -year-olds, who 4did not clearly differentiate attributing blame andeliciting sympathy in their first-person sadness and anger narratives. However, in their explanations of anger and sadness, younger as well as older (7 to 11 years) children clearly differentiated sadness and anger. Bambergs account of the contrasts between the developmental patterns the first-person narratives and explanationsrests in on the childrens developing ability to use construction types todifferentially mark discursive stances. Successful narratives requirethe complex coordination of suchlinguistic components with other personal, social, emotional, and contextual factors. Emotion talk functions both to inform the about what child emotions are, and how they are dealt with in the social-emotional realm (Eisenberg, 198 6 ; Miller, 198 6 ; Schieffelin & Ochs, 198 6). Thus, the language that unfolds in aclear and appropriate narrativereflects the childs sense of how self should interplay with the current listener-participants, emotionalsocial context, and shared expectations communicative goals. of Veneziano (chap. 6), who utilizes longitudinal data from children between ages 1 1/2 and z 1/2 to explore early developmentalstages of childrens justification and persuasion, also presents a dynamic approach. Children first learn to be convinced by their mothersjustifications; soon after, they themselves become effective at producing justifications and explanations to persuade their mothers give in to within a conflictual situation. Mappingrichly to the themes of this volume, Venezianos data and discussion show that childrens justifications of oppositions reflect their growing pragmatic, conversational, social abilities (Peterson & and McCabe, 1992; Reimann, 1998; Snow, 1989; Sorsby &Martlew, 1991; Watson, 1989). In managing oppositions children learn to take up positions vis-a-vis the other, making rightful demands or requesting concessions, thereby creating for the other a positionbe filled. Insisting their initial intention in contrast with to on responding to the others opposition, children between self- and othermove centered perspectives acquire competencein recognizing the and others mental states very early Discourse actsand self positions thusdevelop ininteraction, on. viaprocessesverysimilar tothoseobserved byBamberginthecaseof emotion-talk. These observations are tied to an acquisition model in which communication advances rest on childrens concomitant developments in flexibility of cognitive-linguistic representations and in representations of their conversational partners. These twin developments feed into narrative



conversational sequences in which children have the intent to persuadethrough specific, explanatory, argumentative expressions-another person seen as holding adifferent, conflicting position. Dynamic negotiation part of the is discourse-making process. s necessary, whentheir viewpoints have not been A accommodated in prior conversational moves, children demonstratepersistence, redundancy, and/or reformulations. Partly because there is a delay between childrens increased responseto their mothers explanations and childrens the own entry into effective explanation-making,Veneziano argues for a gradual process of complex abstractions. The children learn interrelated patterns rich, from maternal verbal justifications: narrative devices, their pragmatic impact, and the emotional cognitive states of mother and the relative to those the child. of The significance of parent-child interaction advances in conversational for discourse and pragmatic knowledge is also exemplified chapter 5. Politeness in in Japanese childrens discourseis charted developmentally by Nakamura, who investigated the acquisition of the different registers by I- to 6 -year-olds. She uncovers thechilds increasing sensitivity to the social-contextual variables that underlie the pragmatics polite linguistic forms (Andersen, 8 6 ; of 19 Clancy, 198 6 ; Ervin-Tripp, 1 9 7 9 ) . In both spontaneous speech and role play in pretense contexts, Nakamura observes that Japanese children display a know how polite language use as early as 1 year by using nonverbal means and verbal formulas in contextually appropriate ways. Between ages 2 and 5 Japanese children progressto produce forms whose use depends on awareness ofinloutgroup membership, power, age, sex, social setting, and the nature of the information communicated. Japanese childrens control of politeness forms and their understanding of the pragmaticrules that govern shifts in register develops gradually withage and with exposure to discourse demonstrating such A use. significant aspect of process is again this the gradual unfolding form:function of relations, whereby different ways of showing deference are formally expressed with a range linguistic forms and conversational of strategies. Like emotions, politeness is a domain inwhich language functions toinform thechild about social-cognitive distinctions. Nakamura suggests that children gradually learn to comprehend psychological feelings such as consideration for others and discernment which underlie use of the different politeness forms in the process of socialization to theuse of registers. The complexity socialunderstanding and of cognition required to support the more advanced levels of politeness,relates closely to other discussions in this volume of how communicative development interfaces with related developmental changes inother domains.

This overview shows that thequestions raised in the present volume multiare faceted, leading to the exploration of the interfaces between the cognitive,



social, and linguistic skills that feed into pragmatic competence. Without exception, all the contributions invoke the dynamic processesof discourse as the developmental context for the concomitant changes that occur in these domains. As argued across different chapters such as Venezianos, Bambergs, and Nakamuras, early in development, theseprocesses result in a know how of the mind that entails a practical understandinginterpersonal relations. of The next level of representations that emergelater in development involve an understanding of the know that type. These know that representations support new complexities discourse, with narrative a case inpoint. Across of as thedifferentcontributionsand,inparticular,inBerman, Reeder, and Bambergs chapters, older childrens grasp of the pragmatics interpersonal of communication are examined and further complex abstractionsare proposed as a foundation for success in narrative performance. Despite many similarities across the authors, however, it should also be recognized that theparticular kindsof representations children construct often are seen variously. For example, Bambergand Veneziano stress emergent dynamic processesthat comprise any conversational and narrative performance. In their respective frameworks, they present self- and cognitive-processes as deeply intertwined with pragmatic, social, intentional, and linguistic representations. In contrast, other authors examine childrens increasingly complex discourse and narrative skills after ages 5 to 6 in terms of intercoordinations between metarepresentational and metalinguistic operations, and between theseand social-cognitive knowledge at the know that level. The metarepresentational capacity to reflect on the other in terms of his or her beliefs and intentions, and the metalinguistic skill of relating aspects of the code to different interpersonal functions constitute the components the of higher metapragmaticcapacities that can perhaps best be accounted for within a general theory mind framework(Aksu-Kog & Alici, 19 g g ; Astington, of 1990; Nelson, 1996; Perez-Leroux,2000).

Finally, the chapters in the present volume point someissues emerging from to methodological considerations. example, variations in performance to For due differences in the nature the task, such as method of elicitation (spontaneous of narratives vs. those supported with pictures) or context communication of (role play in pretense activity vs. experimental situations), are now treated in their own right as sources of information about processesor determinantsof development.Berman,forexample,findsthatdevelopmentsinsetting information are first elaborated in make-believe narratives, then in personal experience accounts, and finally in narratives basedon picture books. Thus, genre and elicitation method interact and make a difference to the amount of information produced.




The same observation, furthermore, brings to focus the importancethe of pretend worldas an assimilatory zoneor a special operational space where new capacities related to story structuring are first tried out (Garvey, 1986; Nicolopoulou, 1995; Piaget, 1951; Vygotsky, 1967). Nakamuras role-play procedure withtoys suggeststhat sophistication in politeness carried specific by grammatical and pragmatic devices, emerges earlier than many experiments reveal. In experimental contexts role-play tasks often place cognitive and perspective-taking demands on children that result in scphistication in less language use than that observed in naturalistic discourse. What we obtain in experimental settings, which typically put the child in a reflective-interpretive mode, appears to be a different capacity, the meta version the kind of of competence that researchers aim to measure.taken to represent linguistic If instead of metalinguistic competence, such findings attenuate the of ages emergence. These methodological issues, which underscore the fact that different tasks may be tapping different cognitive-linguistic capacities at different levels, are likely to betaken up in future researchand discussions. The context play, of opening up the possibility for deferred imitation of adult speech and behavior, is noted to be important the operationof processes of for acquisition. Research (Clancy, 1985; Day, 2 0 0 0 ) points again to the role of imitation, albeit in the form of representing anothers utterance, in the development of pragmatic knowledge. Role-play speech often repetition entails of an utteranceby preserving its formal properties to large extent. A s noted by a Nakamura, repeating dialogues heard from adults similar contexts, children in grasp the underlying linguistic rules, deduce the relevant interpersonal factors, and generalize toother contexts with increasing productivity appropriateand ness. These processes maylead toinsight into the distinctions that the linguistic forms embody and thus promote metapragmatic awareness. Lastly, as evident in most chapters the volume, crosscultural differences of emerge as a fascinating sourceof diversity that needs to be taken into account in addition to crosslinguistic variation. The Berman and Wigglesworth and Stavans chapters,however, bytaking up culturaldifferences in their own right, shift a once methodological concern to status of an interesting independent the variable, in line with a whole body of research carriedout in the ethnographic tradition (Clancy, 1986; Heath, 1983; Hough-Eyami& Crago, 1998; Ochs & Schieffelin, 1995; Schieffelin & Ochs, 1986).

Aksu-KO$,A., & Alici, D. M. (199 9 , April). Marking of uncertainty and understanding beliefstates:
Childs theory ofevidentiality.Paper presented at the 30th Child Language Research Forum, Stanford University. Andersen, E.S. (1986). The acquisition of register variation by Anglo-American children. In B. Schieffelin & E. Ochs (Eds.), Language socialization across cultures (pp. 153-16 1). Cambridge, England: Cambridge University Press.



Astington, J. W. (1990). Narrative and the childs theory of mind. In B. Britton &A. Pellegrini (Eds.), Narrative thought and narrative language(pp. 151-171). Hillsdale, NJ: LawrenceErlbaum Associates. Berman, R.A., & Slobin, D. I. (Eds.). (1994). Relating events in narrative: A crosslinguistic developmental study. Hillsdale, NJ: Lawrence Erlbaum Associates. Clancy, P. (1985). The acquisitionof Japanese. In D. I. Slobin (Ed.), The crosslinguistic study of language acquisition, Vol. 1. Thedata(pp. 373-534). Hillsdale,NJ: Lawrence Erlbaum Associates. Clancy, P. (1986). The acquisition of communicative style in Japanese. In B. Schieffelin &E. Ochs (Eds.), Language socialization across cultures (pp. 213-250). Cambridge, England: Cambridge University Press. Day, C. ( 2 0 0 0 ) . A developmental perspective on modal verb use by French speaking children.In K. E. Nelson, A. Aksu-KO$,& C. E. Johnson (Eds.), Childrens language, Vol. 11: Interactional contributions to language development. Hillsdale, NJ: Lawrence Erlbaum Associates. Eisenberg, A. (1986). Teasing: Verbal play in two Mexican0 homes. In B. Schieffelin & E. Ochs (Eds.), Language socialization across cultures (pp. 182-198). Cambridge, England: Cambridge University Press. Ervin-Tripp, S. (1979).Childrensverbalturn-taking.In E. Ochs & B. Schieffelin (Eds.), Developmentalpragmatics(pp. 391-414). NewYork Academic Press. Garvey, C. (1986). Peer relations and the growthof communication. InE. Mueller & C. Cooper (Eds.), Process and outcome in peer relationships (pp. 329-345). Orlando, F1:Academic Press. Heath, S. B. (1983). Ways with words:Language, life and work in communities and classroom. Cambridge, England: CambridgeUnversity Press. Hickmann, M. (1982).The development ofnarrative skills: Pragmatic and metapragmatic aspects of discourse cohesion. Unpublished doctoral dissertation,University of Chicago. Hickmann, M. (19 91). The development of discourse cohesion: Some functional and crosslinguistic issues. In G. Pieraut-Le Bonniec & M. Dolitsky (Eds.), Language bases. ..discourse bases: Some aspects of contemporary French-language psycholinguistics research (pp. 157-185). Amsterdam: Benjamins. Hickmann, M. (1995). Discourse organizationand thedevelopment of reference to person,space, and time.In P. Fletcher &B. MacWhinney (Eds.), handbook ofchild language(pp. 195-218). The Oxford, England:Blackwell. Hicks, D.(1991). Kinds of narrative: Genre skills among first graders from twocommunities. InA. McCabe & C. Peterson (Eds.), Developing narrative structure (pp. 55-87). Hillsdale, NJ: Lawrence Erlbaum Associates. Hough-Eyami, W., & Crago, M. (1998). Three interactional portraits from Mohawk, Inuit and White Canadian cultures. In Aksu-Kog, E. Erguvanli-Taylan, A. S. ozsoy, & A. Kuntay (Eds.), A. Perspectives on language acquisition: Selectedpapers from the VIIth International Congressfor the Study of ChildLanguage (pp. 124-139). Istanbul: BokaziSi University Press. Kuntay, A. (19 97). Extended discourse skills of Turkish preschool children across shifting contexts. Unpublished doctoral dissertation, University of California, Berkeley. McCabe, A., & Peterson, C.(Eds.). (1991a). Developing narrative structure. Hillsdale, NJ: Lawrence Erlbaum Associates. McCabe, A., & Peterson, C. (19 9 1b). Getting the story: longitudinal studyof parental styles in A eliciting narratives and developing narrative skill. In A. McCabe & C. Peterson (Eds.), Developing narrative structure(pp. 217-255). Hillsdale, NJ: Lawrence Erlbaum Associates. Miller, P. (1986). Teasing as language socialization and verbal play in a white working-class community. In B. Schieffelin & E. Ochs (Eds.), Language socialization across cultures (pp. 199-212). Cambridge, England Cambridge University Press. Mayer, M.(1969). Frog, where areyou!New York Dial Press. Nelson, K. (1996). Languagein cognitive development: The emergence of the mediated mind. Cambridge, England: CambridgeUniversity Press.



Nicolopoulou, A. (1995). Narrative development in social context. In D. I. Slobin, J. Gerhardt, A. Kyratzis, & J Guo (Eds.), Social interaction, social context, and language: . Essays in honor of Susan Lawrence Erlbaum Associates. Ervin-Tripp(pp. 21-36). Mahwah, NJ: Ninio, A., & Snow, C. (1996).Pragmatic development. Boulder, CO: Westview Press. Ochs, E.,& Schieffelin, B. (1995).Theimpactoflanguagesocializationongrammatical development. In P. Fletcher & B. MacWhinney (Eds.), The handbookofchild language (pp. 73-94). Oxford, England: Blackwell. Perez-Leroux, A. ( 2 0 0 0 ) . Subjunctive mood in Spanish child relatives: Atthe interface of linguistic and cognitive development. In E. K. Nelson,A. Aksu-Koc, &C. E. Johnson (Eds.), Childrens lunguage, Vol. 11: Interactional contributions tolanguage development. Hillsdale, NJ: Lawrence Erlbaum Associates. Peterson, C.,& McCabe, A. (1992). Parental styles of narrative elicitation: Effect on childrens narrative structure and content. First language,^, 299-322. Piaget, J. (1951). The language and thought of the child. London: Routledge. (Original publishedin 1923) Reimann, B. (1998). Maternal questionresponses in early child-mother dialogue. In A. Aksu-Kog, E. Erguvanli-Taylan, A. S. ozsoy, & A. Kuntay (Eds.), Perspectives on language acquisition: Selected papers from the Vllth International Congress for the Study of Child Language (pp. 108-123). Istanbul: Bogaziqi University Press. Schieffelin, B., & Ochs, E. (1986). Language socialization across cultures. Cambridge, England: Cambridge University Press. face in interethnic communication. Norwood, Scollon, R., & Scollon, S. (1981). Narrative, literacy and NJ: Ablex. Slobin, D. I. (19 96). Two waysto travel: Verbsof motion in English and Spanish. In M. Shibatani & S. A. Thompson (Eds.), Essays in semantics (pp.195--217). Oxford, England: Oxford University Press. Snow, C. E. (1989). Understanding social interaction and language acquisition: Sentences are not enough. In M. Bornstein& J. Bruner (Eds.), Interaction in human development(pp. 83-103). Hillsdale, NJ: Lawrence Erlbaum Associates. Sorsby, A.,& Martlew, M. (19 91). Representational demands in mothers to preschool children talk in two contexts: Picture book reading and a modeling task. Journal of Child Language~B, 373-396. Vygotsky, L. S. (1967). Play and its role in mental developmentof the child. In J. S. Bruner, A. Jolly, & K.Sylva (Eds.), Play: Its role in development and evolution (pp. 537-554). New York Basic Books. Watson, R. (1989). Literate discourse and cognitive organization: Somerelations between parents talk and 3 year oldsthought. Applied Psycholinguistics, IO, 221-236.

Setting the Narrative Scene:

How Children Begin to Tell a Story


Tel Auiu University

This chapter deals with facet of its one ambiguous title. It concerns how children begin the stories that they produce, rather than how first learn or when they they first begin to tell stories. The segmentof a narrativetext that constitutes start its or opening has been the subject considerable interest in literary theory, and of the exposition is generallyrecognized as a critical component of narrative fiction (e.g., Oz, 1996; Said, 1978; Sternberg, 1978). s a psychological counterpart A to this notion, thesetting constitutes an integral part of narrative structure in cognitively oriented story grammar analyses (Rumelhart, 1975; Shen, 198 8). Discourse linguists, too, have paid attention to elements that the narrative set scene, analyzedas orientation in Labovs (1972) study of personal-experience narratives, or as initial background information in Reinharts (1984, 1995) discussions of literary and othertexts. In this line, Polanyi (1985) referredto contextualizing state clauses in her analysis of conversationally embedded narratives. Labov identified the orientationas belonging to the narrative rather than theevaluative elements that constitute story, whereas Reinhart suggested a that scene-setting elements constitute partof the narrative background,as distinguished from its foreground. In line withReinharts proposal, I have suggested that scene setting, or background orientation, may include both interpretive evaluative elements and informative descriptive elements as precursors to the third typeof narrative element, the sequentialor eventive elements that make up the story plotline (Berman, 1997a). Researchersagree that the opening a story of typically relates to thestate-ofaffairs existing prior to the onset the plot. A s such, it provides a backdrop of to the ensuing chain events by specifying the who, where, when, why ofthe of and events to be reported. In the present context, story setting is defined as serving several functions, termed here presentative, informational, and motivating, respectively. The presentative function serves to introduce the characters that will be referred to subsequently as participants in events. Theinformational framingfunction provides aspatio-locative and/or temporal framework the for events. The motivating function explains what sets the chain of events in motion andwhy an accountof these events relevant to the is hearerheader or of


interest to the narrator. These three functions the shared aimof orienting have the audience toward what to come. is This analysis suggeststhat how to start a story constitutes an important feature of the development of both narrative knowledgeand storytelling performance among children (Berman, 1995;Reilly, 1992). The ability to provide adequate background setting informationis of considerable interest for research on narrative developmentfor a numberof reasons. First, understanding how the different functions of narrative setting develop should throw light on important cognitive abilities that relate to how childrendevelop a representation of the listener (Berman Slobin, 1994, p. 6 o 4). This shows that theytake & into account audience needs in the case in point, by providing adequate background information to the story they about to tell (Menig-Peterson & are McCabe, 1978). Giving a suitable setting to the story means that the narraalso tor can construct a text autonomously, means of a self-sufficient monologic by narrative rather than through interlocutor queries, prompts, and other scaffolding devices. Moreover, it requires preplanning of the text a whole, as which in the case of narrative discourse implies a hierarchical, global view of the chainof events that are about to related. These are complex cognitive debe mands, whichtake a long time to evolve. Related to the development of these abilities is command of narrativespecific knowledge. Being ableto provide adequate setting information and motivation will depend on moreoverall narrative competence, in the sense of global plot-organization or action-structure,as defined by cognitive theories of narrative discourse (Giora Shen, 1994; Rumelhart, 1975;VanDijk, 1976). & That is, children must have recourse to a narrative schema, with an initiating event or problem, one or more episodes directedat solving that problem, and an eventual resolution. This suggests that in order to begin a story adequately, children need to be to structure the of the text appropriately. able rest The present studywas undertaken on the assumption that, with the age, scene-setting elements provided narrative texts will change along three interto related dimensions: amount, content, and expression. Thus, young preschool children provide little or no such information (Peterson, 1990; UmikerSebeok, 1979). And while children from as young as age 4 provide some orienting background information to the stories they produce, younger children relate to fewer, and to different, types of setting functions than do older storytellers (Kernan, 1977; Peterson McCabe, 1983). One aim of the & present study, then, is to go beyond these relatively few studies that have analyzed childrens story beginningsby extending the analysis to a database consisting of picturebook based narratives as well as personal-experience accounts, comparing preschoolers with school-age children well as adult as storytellers. A second aim, one to thebest of my knowledge not addressed in prior studies of scene-setting, considers the linguistic used to express this forms component of narrative discourse.

Setting the

Narrative Scene

The present studyis thus embedded in earlier work that has focusedon the relation between linguistic forms and narrative functions across development (Berman, 1996,1997~; Berman & Slobin, 1994; Slobin, lgg3).And itconsiders three interrelated developmental predictions. First, whatwe termed scenesetting elements (Berman & Slobin, 1994, p. 86) will change with age both in quality and quantity. In quality, preschool childrenwill mention different types of setting elements than older speakers; example, they may make for explicit reference to participants but not to motivations (who vs. why) and they will mention place more than time(where vs. when). And in quantity, settings will constitute a larger proportion texts produced by older speakers. Second, of elicitation context or genre will affect the nature of setting elements provided, but this effect will be lessmarked with age. Third, with age the linguisticforms used for scene-settingwill become less formulaic, more explicit, and more varied, in meetingdifferent narrative functions. Findings are basedon monologic narrativetexts produced by children and adults in different elicitation settings. Narratives elicited with the pictured storybook, Frog, where are you? by Mercer Mayer (19 6 g ) , constitute thefrog stories and accounts a personal experience of elicited by asking subjects, Have you ever had a fight or quarreled with someone?Tell me about it, constitute the fight stories. Supplementary data come from an additionalof oral and set written fight stories elicited from older school children adults, from other and personal-experience accounts, where childrenwere asked to tell a story about something that had happened to them, and from based on pictures and texts picture-series. The bulk of these analyses are from texts produced by speakers of Israeli Hebrew, on the assumption that the language in which they conare structed has little effect on the quality and narrative functions setting eleof ments, when speakers share similar literate, western-type cultural backgrounds of the lundconsidered here. Our findings for Hebrew-specific linguistic forms used for the narrative functions marking story openings, transition from of the scene-setting to narrativeevents, and the distinction between background setting elements and narrative events can and should, however, be extended for comparison with other languages.


Earlier research has shownthat children favor some types of setting elements over others. Specifically,they tend to provide more framing information at an earlier age about theplace rather than the time at which events place, and took they give relatively little information about participants and even less about background motivations or reasons for the events (Peterson,1990; Peterson & McCabe, 1983). Such studies have typically relatedto personal-experience narratives, whereas the present analysis starts by considering the different kinds of setting elements provided the basis of a picturebook (the frog story). on


Percentage of Subjects Mentioning Different Setting Elements in the Hebrew and English Frog Stories, by Age (N = 24 per agegroup)
Setting Element
3 yrs

T A B L E 1.1

5 Y

9 yrs

92 79 33

who[by noun]= boy 75 where = jar when = at night why = feeling



Distribution of Setting Elements in Englishand Hebrew-Language Frog Stories

The picturebook Frog, where ure you? (Mayer, 19 6 9) has been the basis for numerous studies of narrative development indifferent languages and from different perspectives over and above the database relied on here (Berman & Slobin, 1994). These includeboth published works(e.g., Bamberg & DamradFrye, 1991; Bazanella & Calleri, 199 1; Berman, 1993; Kail & Hickmann, 19 92) and doctoral dissertations (Herman,1996; Kern, 1997; Wlgglesworth, 1992). Yet to thebest of my knowledge, the natureof the background scene-setting to this story has not been the subjectseparate studyapart from considerations of of referent introduction(Kail & Sanchez-Lopez, 1997) anda brief note o n formulaic openers (Berman& Slobin, 1994, p. 86). The 24-picture booklet in question tells a story about a boy and dog whogo out insearch of their pet his frog, which escapes from its jar during the night, when the and dog are boy sleeping. The setting scene is provided by the first picture in the booklet, which stands alone, to the rightof the title page. It shows the boy and his dog at the foot of their bed, looking at the frog in its glass jar. Table 1.1 gives the breakdown of setting elements mentioned in the frog storytexts produced by speakers of (American) Englishand (Israeli) Hebrew. The figures in Table show a marked 1.1 increase with age in elements counted as belonging tothe setting, together with a change the typeof such elements in mentioned by different age groups. Only half the youngestchildren, aged 3 to 4, introduce the main protagonist,boy, by an appropriatelyexplicit noun phrase, the not just as he, compared with nearly the older all subjects. Few children provide the relevantspatial setting for the frog, as being inside a jar from whichit subsequently escapes; even fewer subjects mention that the events took at night place (as shown by the moon shining in at the window). These findings to clearly seem confirm the prediction, particularly because very few first children provide evaluative or motivational elements to the backgroundfor how theboy came to set have afrog, and why he might want go out looking for it. Examples such to of motivational elements taken from theEnglish-language textsare given in (I).

Setting the

Narrative Scene


Motivational Elements in English Language Frog Story Texts

a. Well-there was a little boy, he liked his-pet frog and his his pet dogvery much [boy,aged 521 b Um-once there was like-a dog-who liked a frog, but the frog didnt . like him, so he decided.. . [girl,aged 9;7] c. This is a story about a boy-a dog-and his frog. Right now, in the beginning of the story-hes-the boy and his dog are just basically admiring his frog, looking at the frog in the jar. The frog looks kind of happy-obviously hes not real satisfied with his existence, because when .. [female adult] .

in The figures in Table1.1indicate aclear age-related development the ability to begin a story with scene-setting information. these figures need to be But evaluated qualitatively as well. First, in relation to character introduction, the task at hand, where both child and investigator have the picturebook open in front of them, may bias the cognitive issue of shared knowledge and required level of informativeness in referring tothe main protagonist. Indeed, this has been shownto be the case in aseries of studies on childrens ability to introduce story characters using same picturebook, but comparing design with a this our situation where thereis no mutual knowledge shared between the subjectnarrator and theinvestigator-audience (Hickmann,Kail, & Roland, 1995;Kail & Hickmann, 1992; Kail & Sanchez-Lopez,1997). Second, over halfthe 9 -yearold and adult subjects do in explicitly mention at night, nighttime as the fact temporal setting for their narratives. However, they do so not only atthe outset in talkingabout thefirst picture but subsequently, to introduce event which the initiates the plot, for example: At night, when the boy and his dog were sleeping, the frog jumped outof the jar and escaped.Thisin marked contrast is to the preschoolers, only one of whom started to describe the event saying by Then one day at night [sic]. Moreover, the examples in (1) are not typical of the frog-story sample: Their settings are more elaborate than most, including the adultstexts. Again, this couldbe task related, inasmuch as both narrator and interlocutorhave the same pictures in front them. of The figures in Table 1.1thus need to be somewhat hedged. Methodologically, picture-based elicitationscreate problems for character introduction and the need for providing detailed background information (see the Comparisons Across Elicitation Settings section). This analysis also suggests that, inprinciple, story-setting elements cannot bedefined by a prescribed of categories such list as who, where, when, and why. Rather, story settings need to be analyzed in relation to the particular story that unfold. Inthe case of the frog story, this will means taking into account the relationship between the and the frog, as boy motivating theevents to come. This was minimally achievedby subjects who started outby saying that theboy hasor keeps a frog, that he thinks the is frog cute, more elaborately by those who described how the frog came in the to be boys possession (he found it, gotit as a present). Again, almost none of the


3- to 4-year-oldsrefer to motivating circumstances, about one third at 5 to age 6 (37%), and twice as many 9-year-oldsand adults.
Nature of Setting Elements in Hebrew-Language Fight Stories

The dataset for this analysis also includes 12 Hebrew-speaking childrenat ages 3,5,7, and 9, compared with a group 1 2 adults. To elicit a fight story, of subjects were asked if they knew what a fight quarrelwas, and to tell about one or they had been involved in. Scene-setting elements were defined as all material that preceded the event that initated the quarrelfight, that is, any verbalrefor erence to when, where, why, or under what circumstances, as well as with whom, the altercation took place. These elements analyzed into the folwere lowing categories, ranging from to most explicit, from to moreelaboleast less rated, fromjuvenile to mature,as in (2).

Ranking of background-setting elements in personal-experience narratives:


No background 1.1 Background element provided by adult input 1 2 InitialEvent . 1.3 Initial Event plus formulaic opener, for example, paam once,yom exad one day 2 . Minimal informational or framing background 2.1 Name of antagonist plus specifying sex or relationship to narrator = protagonist 2 2 Mention of place time of initial event . or 3.Specificframinginformation 3.3 Specification of a particular time and/or place 3.4 Temporal distancing to specify circumstancessurrounding initial event 4. Motivational background, scene-setting orientation 4 1 Temporal distancingto set events in past time . off 4.2 Metacognitive orientation the actof storytellingand/or the nature to of quarreling

1 The texts used in the Hebrew-language data-base elicited from different groups subjects were of in the various studies referredto in these sections. However, it seems legitimate compare results to across these populations, since the subjects all shared the following background they are children of educated, middle-class speakers Hebrew as a first language (like adultsubjects who serve as of the controls in each study); the preschoolers attend Hebrew nurseryschool or daycare fromthe age of z and enter kindergarten age 5 to 6 ;and the at schoolchildren are in gradeschool from 11 or 12years 6 to of age, they enter junior in Grade , and complete high school the end Grade 12. high 7 at of z Instructionswere worded as follows: yodea ma z la-riv, rneriva? hayitapaarn bi-rneriva?saperli uta alze,saperalpahrnfle ravt(a). Doyou knowwhat it is to-quarrel, have a-quarrel?Have you ever been ina-quarrel?Tell me about tell about a time when it, you-quarreled.If the subject hesitated, a prompt was provided saper li sipur alrivfle hayita bo. Tell me a story about a quarrel/fight =took part you were in.

Setting the

Narrative Scene

This ranking, as suggested, corresponds to clear developmental pattern in a amount andtype of background setting information providedby respondents. Thus, as illustrated by the excerpts in (3),the youngest children in our sample, 3- to 4-year-old nurseryschoolers (mean 3;6), typically gave no background age at all, but plunged straight into a report the events. The only exceptionsto of this were when children used a formulaic opener such words meaning once, as o n e day, as semantically nonspecific markers of discourse initiation (see the Story Openers section) gave the name of the antagonist, without further or identifying comment. Older preschoolers, this sample 5-year-old kinderin gartners (mean age 5;4), in some cases did the same, but many them also of added a locative frame, specifying the place where the incident occurred, as illustrated in (4). (3) No background scene setting a. ravti i m Elad ve baxiti I-quarreled with Elad and I-cried [Adi, girl, 3;5] b. paam Orly hi natna li beita a1 ha-rosh i m sirgadol kaze Once Orly she gave me a kick the-head with (a) big kinda pot [Yafit, girl, on 3;10] (4) Minimal background scene setting a. ba-gan yalda axat daxfa oti me-ha-nadneda At kindergarten agirl pushed me off-the-swing [Efrat, girl, 5;4] b. etmol ba-gan ravti i m xavera sheli Roni Yesterdayat-kindergarten I-quarreled with my (gir1)friend Roni [Meital, girl, 5;4] In contrast, some preschoolers, nearly all the 7-year-old (Grade 2) and all the 9-year-old (Grade 4) school children, provided additional framing information, often in the form some surrounding circumstances or event, as in of (sa), or being highly by specific about the exact placeor time as in(5b) and (5c), respectively. Older childrenalso quite oftengave two or more different types of framing information(5d, 5e).

More specific framing by circumstances, time, and/or place a. y o m exad sixakti xevel ba-xacer One day I-played was playing] jumprope in-the-yard[Galit, girl, 5;1] [= b. paam hayinu ba-kantry ba-brexa, ve haya li mishkefet kazot. .. Once we were at-the-sports-center in-the-pool, and I had kinda goggles [Liron, boy,5;6] c. hayom le yaradnu ba-maalit mi-safia... Today when we took the-elevator down from Granny. ..[Amit, boy, ; 6 ] 7 d. paam haya li vikuax im ima ze haya ba-telefon, le anicilcalti laavoda sheli,
shela. ..


Once I had (an) argument with my Mom, it was on-the-phone, when I called her at work [Dafna, girl, ] 9;2 e. ani vexaversheliRonen halaxnula-xanut matanot liknot matana layomuledet she1 Dan. .. Me and my friend Ronen wentto-the gift storeto buy (a) gift for Dans birthday [Tal, boy, ; 5 ] 9 Interestingly, these more elaborated settings illustrated in (5) were also quite generally set off explicitly from the onset element, or plot initiation, by an overt segmentation marker, typically in the form expressions such as ve-azand of then or pitom suddenly(see the Transition Marker section). In addition, proficient narrators, mainly adults also some g-year-olds as but shown in (5d) and (5e), typicallyprovided some background motivation for the quarrel. They did this by talking about the relations between the antagonists beyond this specific incident, by distancing themselves from the events, by going back in time and setting the reported incident ingeneral frame of a more memories, and alsobyvolunteeringmetacognitive o r metatextual comments relatingto fight scriptsin general, or to the storytelling situation. These strategiesare illustrated in (6).
(6) Maturely elaborated scene settings

nizkarti. ani ravti im baxur ie haya iti baxeder bakibuc, kie hayiti gar bakibuc. a1 ze ie huhaya maklit lo flirim a1 ha-kaseta ieli, hu haya (h)ores li et ha. kasetot .. I remember quarrelling with fellow that was my roommate on the kibbutz, a when I was living the kibbutz, he used record songsoff my tape-cassettes, on to

ruining them [yuval, male, 221 b. be-bet sefer yesodi haya paam yeled i e hecik li. karu lo Zohar S. hu nahag laruc axaray ve lehatrid oti... In grade school there once was a kid that gave me trouble. He was called Z S. He used to run after me and bother me [yair, male, 281 c. tov, anixoshevetle-saper a1mashehume-ha-gan.haytalixaveranoratova, Enav, b a - p . ve kol hazman hayinu ravot a1 miney shtuyot ... Okay, I think Ill tell you about something from kindergarten. I had a very close friend, Enav, in kindergarten. And all the timeto quarrel about all we used kinds of stupid things [Havatselet, female, 22.1 d. racitie ani asaper lax sipur a1 riv, aval loh ravti im af exad. ani loh ish ?e rav
im axerim. ve afpahm loh ravti im af exad. Kaasti P O ve sham, kabsu alaypo ve sham, betor yeled xatafti makot... You wanted me to tell you (a) story about a quarrel, but I never quarreled with

anybody. Imnot a person that quarrels with others. And Ive never quarreled with anyone.Ive gotten mad hereand there, people have gotten madat me here and there, a kid, I got beaten up [Eran, male, as 241

Setting the

Narrative Scene

The picture that emerges from these excerpts, one reinforced by findis that ings from other personal-experience accounts in Hebrew as in English (e.g., Peterson & McCabe, 198 3) is o f clearly age-related patterns in the ability to take the listener into account in providing adequate background information and scene-setting orientation. Thesecan be summed up in terms of four developmental phases: juvenile, transitional, structured, and proficient: (a) Immature, juvenile narrators provide no setting elements atall, or else only formulaic starters, as in the example (3). in This is consistent with other studies have that noted that young preschool children tend to give little or no background setting, but instead start their stories with immediate action, whether they making up fictive stories (Pradl, 1979), embedding them in conversational interaction (Minami, 1996), or basing them on familiar scripts (Seidman, Nelson, & Gruendel, 1986). This is followed by (b) a transitional phase, when minimal information is provided to identify the relevant participants (in the case of the fightstory sample, the antagonist) the location o f the event. or Next, older, school age children typically provide (c)structured scene-setting frames by specific identification of the place and/or time of the events, combined with some temporal distancing, and with sequential events being clearly set off by overt markingof the transition from background to foreground, plot initiating events. Finally, (d) maturely proficient narrations are not onlyfully structured and temporally distanced by means of initiating elements that distinguish the events to be reported from the time of their reporting, and place they often contain personalized or other evaluative commentary concerning the relationship between the participants, the narrators attitude to the events reported and to others like them, and/or to the of storytelling and reporting act on these events.


The two sets of analyses presented here,on thefrog stories and fight stories, respectively,differ along a number of dimensions, although both deal with narrative texts produced by similar groups of subjects (see note 1). These differences suggest that analysis of scene-setting elements must take account of the particulark i n d o f story being told. In the present for examcase, ple, the frog story is based on the script of an adventure story, and it is in the genre of childrens storybooks. The fight stories, in contrast, are basedon the script of a conflict situation and belong to the genre personal-experience of accounts. In fact, the type of task and the context of text-elicitation turns out to have an impact on the amount as well as the nature of the setting elements provided across different ages. This was revealed by analysis of the overall

1 0


amount of scene-setting, analyzed as the proportion of clauses serving this function across the two types texts.3 A s shown in Table 1.2,the prediction of of development with age in amount of scene setting was not confirmed for the frog story sample. The first line of Table 1.2 shows a consistently low proportion of clauses dedicated to scene setting (the first picture out of a total 24 pictures in the book), between 4% to 7% of all clauses across age-groups. In marked contrast, as shown in the second lineo f Table 1.2, the fight story reveals a clear and consistent rise in mean proportion of setting clauses with age, up to10% among preschoolers, around one quarter at school and over age, one third among adults. These findings are robust,as they tally with findings for similar typesof elicitation across other populations. Analyses of frog story texts in other languages from the Berman and Slobin (1994) study, together withFrench data elicited by the same methodology Kern(1997), reveal similar trends. by The first picture, providing the background antecedent to the plot-initiating of the event frog escaping from its jar, yields the same figures for adult narrators English low in (mean of 5.5% of all clauses in the sample), Spanish (mean 7.6%), and French (mean 7.3%). In marked contrast to these figures, analysis of setting low the element in a range other fight stories of elicited from other Hebrew-speaking school children and adults reveal a closely parallel trend to the original fight story sample in Table 1.2. This additional database consisted other Hebrewof language fight stories elicited in much the same asthe originalset from eight way second graders (aged 7 S), 12 fourth graders (aged 9 to lo), and students to 12 (aged 17 to 18) and adults,each of whom produced two narratives about two separate experiences with a quarrel orfight, one in speech and onein writing (balanced for order modality). In these stories, similar to figures in the of fight the second lineof Table 1.2, the secondand fourthgraders produced anaverage of 2 0 % to 25% background settingclauses out of the total clauses in their narratives, whereas the adults devoted many as one third as (33%) to one (49%) of half their narratives to background setting clauses. Further evidence for these general trends for personal experience stories is that therewas no notable difference between the figures for the narratives producedwriting compared with in speech.4
3 The clause, defined in Berman and Slobin (1994) as any unit that contains a unified predithat expresses a single situation (activity, event, state) constitutes a unit of analysis highly cate ... relevant to the characterizatior. o f narrative texts in both form and content (p. 660). 4 Written narratives reveal another, unique feature marking background scene-setting which also has developmental consequences: More than half o f the adults but almost none f the children o marked off their story-setting from the initial episode or enabling event graphically, by a separate paragraph. In fact, one adult, a young computer scientist called Itay, set off his first paragraph by a heading in the margin with the word reka backgroundand started his second paragraph (both indented, with a two-line space between them) with the one-word heading in the margin ha-maase the incident. I am grateful to Nurit Assayag o f the Tel Aviv University project on developing literacy for bringing this subject to attention. my

Setting the

Narrative Scene


Mean Percentage o f Setting Clauses out ofTotal Clauses in Hebrew Frog Stories Compared With Hebrew Fight Stories, by Age ( N = 12 per age group)
Age Group
type Story
Frog story Fight story


3 yrs
6.6 6.4

5.2 9.8

4.5 20.6

Adults 9 yrs
5.5 27.7 7.2 36.4

These figures appear, moreover, highly consistent withother findings for amount of background setting material compared with overall text length as defined by number of clauses across a range of other materials, inEnglish and in Hebrew. The analyses in the rest of this section derive from detailed but less similarly motivated examinationsof background setting elementsin narrative texts producedby three distinct methodsof elicitation: narration of content of picture series, recapitulation of personal-experience accounts, and makebelieve, fictive stories. The database for picture series consists of Hebrewlanguage materials basedon three sets of four pictures each toldby preschool children aged 4 5 , and 6, compared with IO -year-olds and adults (Berman& Katzenberger, 19 9 8; Katzenberger, 19 9 4), on aHebrew replication of the cat and horse series used by Hickmann and her associates (Hickmann,19 91; Hickmann, Hendriks, Roland, & Liang, 1996), which elicited texts from 15 Hebrew speakersaged 5,7, and11 years compared with 10 adults (Kahanowitz, 19 9 5, and on oral ) stories basedon a series of pictures about avisit to the zoo by seven Hebrew-speaking preschoolers, and written versions from 14 adults (Berman, in press). The database personal-experience accountsconsists of for a range of Hebrew-language materials on various subjects elicited from children aged 3 to 12 and adults (described in Berman, 1995,1997a).These are supplemented by written andspoken versions of two different fight stories by 48 Hebrew-speaking grade-schooland high-school studentsand 16 adults, and combined with the English-languagetexts in the appendix to Peterson and McCabe (1983). And the materials surveyed for make-believe stories are based, again, on a range of Hebrew-language texts culled from various sources, combined with the largenumber of stories collectedby Pitcher and Prelinger (1963), as published in Sutton-Smith(1981). This surveyof a wide range of materials providesstrong support for my earlier claim regarding intertask differences in childrens narratives (Berman, 1995, pp. 295-302). There, the issue at handwas how the typeand context of elicitation affects childrens ability to give expression to principles of narrative discourse organization. In developmental terms, this ranges from immature expression of isolated events, via encoding of temporal sequence local and



relations of causality, on to a global, hierarchically organized action structure. Comparison across different narrativegenres and varied methods elicitation of (recounting of familiar script, of a personal experience, of the contents of a picture series, a pictured storybook, and film without words) yielded the a conclusion that narrative notdevelop along a uniformly linear curve. ..[since] divergent results emerge in different settings and across different tasks (Berman, 1995, p. 298). In the present context, wish to make I an analogous claim for the development childrens ability to start a story of by providing relevant, and adequate, background, scene-setting information. That is, here as in other domains of development, task effect needs to be taken into account. Children proved able demarcate setting elements better and earlier to i personal experience accounts than in narratives based a picturebook n on story. They did as young as age 3 when they were free to tell about anything so that had happened to them, but only from around 5 when asked to tell age specifically about afight theyhad experienced. In general, personal experience accounts appear to provide more authentic contexts for elaborating on scene setting than picture-based elicitations. examples in(7) are based on aseries The of six pictures depicting what happens to two children visiting a when the zoo, monkey snatches theice cream from the younger child (Berman, in press). They suggest that young childrenmay not bother to provide setting elements at all in picture-series narrations.

Opening clausesof three childrens picture-series based stories zoo


Once there were kids and they went with their to thezoo . . two mom . [Doron, boy,6;8] 2 . PO ani role yeled ve yalda ve xayot ve az hem... Here I see a boy a girland animals and then they .. and .[Tibi, boy, 5;8]

paam hayu shney yeladim ve hem halxu im ima shelahem le-gan xayot

hem be-gan xayot vepo hem mistaklima1 aryeim.

Here theyre at the and here they are looking at lions[Batya, girl,4;8] zoo ...

Although they all clearly recognized the zoo script situation in the pictures, only from school werechildren able to provide any kind scene setage of ting as a frame for theirtexts, and only childrenaged 6 and older produced stories organizedaround an acceptable action structure. This finding consisis tent with results for another set of picture seriesas analyzed in Katzenberger (1994). In contrast to these picture-series elicitations, nonpicture-basedfictional accounts, where children are asked to create imaginary stories, appear to provide a particularly rich context for expressing scene-setting abilities (as early can be inferred from what kindergarten children do in a pretend-reading task, as in Segal, 19 9 6). After all, in prose literature,background exposition plays a crucialrole in constructionof narrative texts. Against this background,I suggest that sometypes of narrative-elicitation

Setting the

Narrative Scene


tasks and certain communicative situations promote earlier,and richer, exwill pression of background settings will others. These be ranked as in (8), than can from most toleast likely to elicit appropriate background material from relatively early on.

Ranking of scene-setting evocation, by narrative genre, elicitation method, and communicative context on own imagination b. basedon known filmor book 2. Veridical Personal-experienceaccounta.outsideinvestigator b. familiar interlocutor Picture-storybook a. basedmutual no knowledge 3. Fictive b. mutual knowledge Picture-series based a. no mutual knowledge 4. Fictive b. mutual knowledge


Make-believe based fantasy a.

We can thus explain the disparity Table 1.2 between the amountof textual in material given over to background setting elements in the frog story datab (type 8.3b) compared with thefight storymateria1s (type as a function 8.2a) of the differences narrative genreand in the elicitation context. in I suggest that, in general, as indicated byits ranking in(8.1), make-believe fantasy will be the first type of narrative in which children will provide some scene-setting informa tion. In fact, it is in the context of familiar materials that young preschoolers have seen, heard,or hadread to them(as listedin 8.1b), that they will first acquire the conventional markings story openers such once upon atime or of as its Biblical style classical Hebrew counterpart hayo huya paam be was once = once there was(see Story Openers section). Because the interlocutor needs to be introduced to the fantasy world being created or recreated in the narrative text, adequate background information essential for orienting the audience to is what is about to told. be And this is more critical in the ofa story that uncase is familiar to both narrator and audience, as in (8.1a), than to one they shared have knowledge of (e.g., a favorite fairytale or well-known fable) as in (8.1b). Next in rank as setting-evocativeare personal experience accounts. Here, the factor of mutual knowledge is critical. A s shown by the excerpts from the 3-year-olds in (3) compared with those from the 5-year-olds shown (4), young children in often fail to provide the minimal referential information necessary for an unfamiliar investigator,as in (, in situations no mutualknowledge, to of identify the participants in the event. the other hand, fight stories elicited On in in Hebrewin a situation where school children recounted a personal experience to a friend or classmate, often one who had been present at the event, ( situation 8.2b), even teenagers felt no need to specify details of the other participants identity, beyond their names, they tended to provide minimal and locative framing if the event took place at school. these high-school students, But



like the adultstelling a fight story to a friend, gave overt expressionto other all setting elements, by providing suitably detailed and distanced temporal framing. They also often gave extensive motivational background in terms the general of relationship between the narrator and the other participant(s) in the events, personal predilections of narrator and/or other the participants, and so on. In contrast to such rich setting-evocation in self-constructed narratives, whether fictive or veridical, picture-based elicitation procedures of pictured storybooks (8.3) or picture-series (8.4), appear less accessible to explicit far verbalization of setting elements. Thisis shown, as noted, for therelatively small amount of background information provided in a wide range of frogstory texts in different languages, even among adults. And Katzenbergers (1994) large sample of Hebrew-language texts elicited on thebasis of several different picture series from 4-,5-, and 6-year-olds compared with lo-yearolds and adults, reveal that from age 5 children often give some standard , opener suchas the word foronce (Story Openers section) plus a minimal referent introduction withoutany additional background information prior to the initiating event. The single exception was one 1 0 -year-old, nearly halfof whose clausesabout a series of four pictures showing a woman in a hat store buying a new were given over to distancedand motivational background hat setting.5 True, situations of no mutual knowledge, thatis, (8.3a) and (8.4a), enrich the amount and form referent introduction, as noted for frog storyof based studies of this kind (Setting Elements section), and as such theyare better suited to meeting the presentative functionof establishing story background. But they, too, fail to stimulate much in the of the other two functions of way story settings: locative and particularly, temporal framing and evaluative motivation. Thisis noteworthy, because picture-based elicitations have yielded particularly rich analyses of childrens narrative abilities across different languages. These findings highlight a general point of both principle and methodology: Different types of elicitation procedures and communicative contextspromote, or at least allow expression some types abilities earlier or more than others to, of (Berman, in press). True, in developmental terms, onceboth narrative competence and storytelling performance are well established, older and more proficient narrators will prove less susceptible to effects of task and context with respect to narrative story-setting as in other domains. Nonetheless, the ranking tentatively proposed in is worth examining under suitably (8) controlledconditions,usingcomparablematerials (e.g., based onthe same theme, topic, or script) across different types of tasks and different developmental phases,to furthertest the prediction thatsetting evocation will be strongest atlevel (%la),weakest at (8.4b).
5 I am grateful to Dr. Katzenberger for making her summary and illustrationsof these data available to me.

Setting the Narrative Scene



This final set of analyses focuseson the linguistic forms used for three related narrative functions: to mark narrative openings, to demarcate the transition from scene-settingto narrative events or episode inception, and todistinguish between backgroundsetting elements and narrative events.
Story Openers

An early development in marking linguistic form:function relations is use of a temporal term identifying the as inthe narrative mode. Thus text Pradls (1979) study based on thelarge Pitcher and Prelinger (1963) corpus of fictive stories children were asked to tell, noted that only 20% of the 2-year-olds began their stories with a formal opening device, whereas this increased to nearly threequarters (73%) among the 5-year-olds. This is consistent with findings from the frog-story sample in English and Hebrew, in which nearly all the preschoolers plunged directly into picture description narration, or whereas 23 out of the 24 fourthgraders(9-tolo-year-olds)providedsomeintroduction,and two-thirds of them used whatwe termed formulaic opening expressions such as once, once upon a time, one day or their Hebrew equivalents,as described next (Berman& Slobin, 1994, pp. 74-75).6 A term suchas once (Hebrew paam) specifies that something happened, and it happened in the past. A quarter the 3-year-olds and nearly half the of 5-year-olds opened theirfight stories with this word,or with a similar expression, one time (Hebrew paam axat),one day @om exad), and an occasional yesterday(etmoZ). So did the7- and 9-year-old school children, but theiruse of these terms differed importantly from that of younger children. Among preschoolers, these temporal openers invariablytext initial, whereas among were older children they could be internal. For example, 7-year-old David (7;6) text started off with h a l a x t i b la-ken im Saar I went once to theclubhouse with Saar, and 9-year-old Etti (9;7) started her story with asaper lax riv e h a p li ani ; im axoti ha-gdola. paam m a t . ..Ill tell you about a quarrel I had with my older sister. One time. ... Moreover, occasionally among 5-year-olds invariably and with older children, these openers were accompanied by additional setting information about time and/or place, in line with the general developmental trends noted earlier. In contrast, early use of these narrative openerswas very restricted in function: The children started a storyas having some generic off
6 One of these fourth graders but only of the Hebrew- and English-speaking adults started two out with thehighly narratively oriented comment Its a story about. ..This might reflect cultural conventions, since in Kerns (1997) French sample, which used the same elicitation procedures as the Berman and Slobin (1994) study, overhalf the adults althoughonly one 11-year-old and none o f the younger children started their narratives with comment along the lines Cest a of

une histoire de . . .



prior-to-the-present temporal location, but failed to anchor it withinany specific time frame.Use of these openers is thus essentially formulaic,rather than well motivated in terms of the semantics of temporality or the discourse function of temporal framing. But they do show that children are familiar with conventional narrative devices marking story beginnings in their culture. In the picture-based frog-story corpus, noted, use of similar expressions, as such as English once upon a time, occurred at the beginningof many of the texts produced by 9-year-olds but in almost none of the younger childrens. The most typical such opener was in the form paadpaam axathaya yeled once/one time (there)was a-boy (used by no fewer than 6 outof 12 Hebrewspeaking 9-year-olds and 3 out of 12 older children, aged 11 to 12). These expressions serve a rather different, though no lessstereotypic, culturally conventional function than the terms paam/paam mat oncelone time in the personal-experience fight stories of the younger, preschool children; they mark the startof a childrens fairytale or fictional storybook account. Moreover, four of the 11- to 12-year-old sixth-grader frog stories, none among the but y-yearold fourth graders, used an archaically flavored literary type opener in Biblicalstyle Hebrew, in the form of hayo haya paam be was once = there once was. This is the classic opener for Hebrew childrens literature in fairytales and fables, so that this finding ties in with what was noted in the previous section about the important effect of genre on narrative settingor expositions. These conclusions are supported analysis of make-believe stories written by by third-grade 8- to 9-year-olds compared with sixth-grade to 12-year-olds 11asked to make up a story about a child who meets a strange creature on a journey (Argeman, 199 6). Of the younger children,two started outwith yeled exad a boy = child, two with yom exad one day, most (7 out of 12) with paadpaam mat once/one time, and one with the fable-marking hayo haya there was once. An almost identical breakdown marked the openers of the make-believe stories written by the older group11-year-olds,except that two of of them started with theclassically flavoredhayo haya opener. This suggests that mode of elicitation (written versus spoken) and narrative genre (makebelievefableversus picturedadventurestory) willevokeearlier, more widespread use of strictly conventionalized story openers. That is, these children were manifesting knowlege not only of the narrative as a type of text, but of literacy-related awareness of subgenres of narrative as well. Moreover, the different linguistic and textual or situational contexts in which a single term such as paam once is used as a story opener supports earlier findings with regard to form:function relations in language acquisition in general, and in the development of narrative abilities specifically (Berman, 1996,1998; Berman & Slobin, 1994; Slobin, 1993). Early uses of a given linguistic form will serve a restricted range of functions, whereas the same superficially identical forms may serve different functions across development. With age, the range of forms used to indicate that a story is about tobegin not

Setting the

Narrative Scene


only becomesmore varied and more personalizedor less stereotypic, the forms also tend to be more explicit markers of a particular narrative genreand/or as function. Thus, for example, the young man referred to in marking off as note 4 the background section not only a separate paragraph but also by explicit by mention of the word reka background in the margin of his written story introduced an oral narrative means of this same word when telling a friend by of his about a quarrel had been involved in at work. In the opposite direction, he that of the same form serving different functions with age, none of the adult narratives, whether frog-bookbased or personal-experience fight-story accounts, startedout with temporal markers paam (axat) like once, one time or yom exad one day. As noted earlier, adults start out their stories with more specific temporal andlor locative framing, e.g.,1 halaxti lemale delek Yesterday I went to take gas [Hanan, man, 251, lifnev xamesh shanim horay, baali ve ani yacanu le-tiyul...Five years ago my-parents, husband, andI went on a trip ...[Sara, woman, 401, be-bet sefervesodi haya paam yeledAt grade school there oncewas a boy [Yair, man, 281; or else they make some metacognitive comment on their recall or reconstruction of the event, for example, ani rotse le-saper al mikre riv I want to tell about a quarrel incident [Shlomo, man, 321, tov, ani xoshevet le-saper mashehu me ha-ganOkay, I think Ill tell about a1 something from kindergarten [Havatselet, woman,Temporal adverbs such 223. as yom exad, paamone day, once do occur in the adult narratives, but in a different place, not at the outset, and for a different purpose to markbackground setting off from plot initiation,as next discussed. Transition Markers Consider, next, how narrators mark off or otherwise indicate the boundary between scene setting and plot inception. In writing, this may begraphically marked by means paragraphing (see note 4), but in spoken texts, of some overt linguistic form is needed to perform this kind of segmentation. Young preschool children use overt, conventional linguistic means to mark story openings even prior to the development of a well-structured narrative schema, but in contrast, the transition from setting to the events which start the story per se is often blurred and not clearly marked in their texts.Table 1.3shows the expressions used to mark the transition from introductory setting to the plotline chain of events in the Hebrew frog-story texts, where the frogdepicted is escaping fromits jar while the boy dog asleep in the bed nearby. and are There is an almost complementary distribution between the younger and older speakers in marking the transition, shown the different clustering of by the figures in eachcolumn of Table 1.3,and by the figures in bold, which stand for the favored means at each group. Three-year-olds favor zero marking, age 5-year-oldsprefer and, and g-year-olds rely on an explicitly temporal expression. The fact that young children generally provide overt marker of no



T A B L E 1.3 Markers of Transition to Plotline Chain of Events in Hebrew Frog Story Texts, by Age (N= 1 2 per agegroup)

Zero marking ve and (ve) hine and here(on) az then, so pitom suddenly axarkax afterwards yam exad one day benatayim meanwhile balayla at night balaylakie atnightwhen bizman ;e while (that)

3 Yrs

5 Yrs

9 yrs




narrative-event inception indicates failureto distinguish between background setting and foreground plotline.If 3-year-olds do mark subsequent events versus prior states, they the vague, general connector ve and, in away as yet use laclang in conventional syntaxor semantic content, and meeting normative no narrative function (Berman,1996; Peterson & McCabe, 1991). Older speakers almost always mark the boundaryexplicitly, either bythe general episode markerone day, or by more specific terms for points in time (at night) or duration (while the boy slept). Adults use a wider range forms of than otherage groups, and they avoid sequential expressions suddenly,and like then, after that, which are favored by school-age children. Besides, where the youngest and oldest groups share surface forms, these serve quite distinct functions. Zero marking in the case of the 3-year-olds is indicative of their picture-by-picture description of isolated scenes and events. Adults who fail to use an overt marker of plot inception rely on other devices to mark the transition from background setting to foreground plot-a switch in verb tenseor a shift from stative to dynamic predicates7 The transition markers used by the 5- and 9-year-olds, again, illustrate more general trends in the development of narrative form-linguistic function relations. Five-year-old children reveal
7 The locative hine and here also functions differently in the younger and older texts: the For little children, it has a deictic spatial function, correspondingthe of the temporal to use deictic now, like French voici; for adults, Hebrew hine is anaphoric, it marksoff a given point in thechain of events under discussion. A similar switch from a deictic to an anaphoric, discourse-motivated function occurs with the time word arshav, much likeits English counterpart now(e.g., boyis The in d a n g e r m that the owl has been disturbed).

Setting the Narrative Scene


command of narrative sequentiality through markers linear clause-chaining: of multifunctional and, along with sequential terms like then, suddenly, afterwards. Nine-year-olds are more like adults in segmenting background setting from plot initiation through of specificallytemporal termslike meanwhile, use that night. In the fight-story sample, too, 3-year-olds only occasionally marked the off initial chain of events a sequential term suchas az then, so, axarkax afierby wards, or pitom suddenly. These expressions servethis transition-marking function in the bulk of the childrens fight stories from age 5 years up, for example, g hitxilu caakot sohhen (there) started shouting [Shay, boy, 5;0], pitom yeled exad shovav kafatsSuddenly a naughty kid jumped (out)[Galit, girl, 5;1], az axarey ze hu omer And so after that he said[Tomer, boy, 7;5]. The adults rarely used the termpitom, which serves as a typical marker of episode initiation in childrens storybooks. And if they did,it was the more literary, high-register equivalent Ie-feta all-of a sudden (e.g., le-feta xash Avi be-ra of-a-sudden, Avi was taken bad = ill [Sara, woman, 40]), in line with what was found for the Hebrew frog-story sample as well (Berman & Slobin, 1994, p. 301). In contrast tothe children, adults mainlyused the punctual term (az) yom exad (then) oneday as transition markers; for example, an adult fight story of 38 clauses long about howin junior high theyused to throw things down on people in the street below started with background introductory IO clauses, then switchedto theinitial event as follows: azyom exad lakaxti tapuax Se heveti me ha-bayitSo one day I-took an apple that I-brought from home ... [Udi, man, 231. In general, temporal markers transition from background to of plot onset used by older speakers are more specific, for example, boker exad one morning [Sarit, woman,211, corresponding to the transition-marking ba-layla at night of the frog story. They tend,also, to be more detailed, and often introduce an embedded temporal clause, for example, ba-yom bo hexel ha-kurs on-the-day on-which started the-course [Idan, man, 221, yom exad Se tiyalti it0 One daywhen I was out with-him= the dog [Shlomo, man, 311, yom exad, kje hu hecik li One day, when he bothered me [Yair,man, 281. A similar preference for a particular form to mark the shift from scene setting to the startof the action among older Hebrew speakers was even more marked in another narrative task. Three- and 4-year-old preschool children, 11-year-old sixth graders and adults were asked to make up a story based on a large picture showing an old carrying a man sack of fish walking toward a house where a woman and children stand waiting on the porch (Ben-Haviv,19 9 6). More than half the younger children started their texts with the expressions paam once or yom exad one day, showing that they knew they supposed were to tell a story. The school children and adults withonly one exception used similar expressions, for example,paam axat one time, yorn exad one day, boker exad one morning at a point three to four clauses into their narrations. Thesetemporaladverbsservedtoindicateaswitchfromscene-setting



background description to reporting the narrative chain of events. In marking this transition, too, the identical linguistic forms different narrative funcserve tions at different phases in the developmentstorytelling abilities. of
TenseIAspect Shifts

This heading concerns the ability encode rhetorical alternations between to background setting and foreground plot elements. I examined of tense/asuse pectshiftingto distinguish story introductions from the chain of plotline events, since grammatical aspectis recognized as a key means for distinguishing foreground and background elements in narrative Berman & Slobin, 19g 4, pp. (e.g., 6-9; Chvany, 1985; Hopper, 1982; Labov, 1972; Reinhart, 1984). Modern Hebrew, unlike theclassical Biblical language, does not mark aspectual distinctions grammaticallyby inflections on the verb, so that todays Hebrew speakers need only mark the inflectional distinction between finite marked for past verbs compared with present and future forms (Berman & Dromi, 1 84; Ravid, 19g 5, 9 pp. 42-45). Two relevant findings emerged from large-scale crosslinguistic our studyinthisrespect.First, we found almost no evidencefor linguistic compensation, defined as expressing by lexical means notions that are not morphologically grammaticized in the language. More specifically, we noted that with regard to verbal aspect,found onlyrare instances of we attempts in German and Hebrew to add distinctions punctuality or durativity that are of notmarkedgrammaticallyinthelanguage(Berman & Slobin, 1994, pp. 621-622). Second, in the Hebrew frog-story corpus, narrators tense use shiftingas a means of global discourse organization,of whose functionsis one to set offbackground settings .. . from the central body of the plot(Berman & Slobin, 19 94, p. 29 5).But narratively motivated deployment tense shiftingis of restricted in severalways. First, tense shifts are only from past to present or present to past, depending which tense the text anchored in in general. on was Second, onlysome outof the 16 adults in the Hebrew sample (Berman, 1988) shifted tenses to a noticeable extent; and only some of these didso for the purpose of distinguishing background settings from the main plotline. Third developmental perspective, none of the Hebrew-speaking children used tense shifting as a device for global narrative organization thefrog-story sample; in for example, school-age children it asa local device to express the temporal used relations between complementclauses and their matrix predicates (Shen & Berman, 19g 7). Global text-level tense shifting a peculiarly adultdevice, was not employed even by g - to lo-year-old fourth graders with full command of complex syntax global narrative action structure. and For purposes of the present study,examined a range other narrative texts I of produced by Hebrew-speaking children and adults see whether tense shifting to would serve to distinguishbackgroundsettingororientationfromthe foreground narrativeevents. My hypothesis was that in nonpicture-based narra-

Setting Narrative the Scene


tives, present-tense benoni intermediate participial forms would function to set off background situations from foregroundevents. And indeed, the findings for the Hebrew frog-story corpus were strongly confirmedin these other samples, too: Only older speakers, only someof them, used tense shifting to serve the and narrative function marlung off story beginnings from continuation. This of their confirms the prediction that, with the particular elicitation setting exerts less age, effect on narrative production than among younger subjects. Moreover, in one sample, in which texts were elicited from older, teenage children, they behaved more like the adults in this connection than did younger grade schoolers. The impact of increased exposure different types of narrative and other types and to text well-establishedliteracy evidently makes 12-year-olds more familiar with a range of cultural conventions rhetorical options of the narrative genre. and A second findingalso went beyond what emerged from the Hebrew frogstory sample. Narrators used a rangeof other formal options in addition to present/past-tense shifts to distinguishsetting from story, and they didthis similarly in quite different contexts. These included(a) an interview-type situation, in which Israeli adults were asked to tell about their experiences high school in and the army; (b) elicitation setting, in which childrenand adultswere asked an to pretend theywere telling a story based on large colored picture to their a friends or their pupils at nursery school; and personal-experience accounts (c) as represented by the fight-story sample. Inall these settings, narrators whoare characterizable as proficient used a range of devices to distinguish between predications in the background settings and those describing narrative plotline events.This distinction is achieved by shifting between the small repertoire of relevant tense/aspect marking forms in the language: present tensebenoni intermediate forms, whichare also participial in function; past tense forms inflected for person as wellas number andgender, which cover the whole range of English past-tense forms -progressive and perfect as wellas simple; and the complex form of haya + benoni wadwere participle, equivalent to English would do, used to do. The excerpts in (9) and (IO) illustrate tense/aspect switches used to distinguish setting from narrative-event predicates in our Hebrew narrative sample: from present-tense (participial) forms for background setting to the past tense event recounting(9.1); shifting from past to in present tense in a historical or narrative present (9.2); shifting between the complex form of habitual past followed by the unmarked simple past tense (10.1); and simple past followed complex habitual by past, as in (10.2).


Switches between Present [Participial]and Past Tense

1 .

Present Past Tense [Present = Participial] a. misphaxat Yisraeli & ornnarn misphaxa ktana. y & ba rak saba, savta, shney yeladim ve zug horim. azzo & mixphaxa rneusheret ve smexa. hasaba ve ha-savta & ba-kfar ve le-yadam shoxenet brexat dagim. & exad ha-saba hexlit. ..



The Yisraeli family h actuallyjustasmallfamily.Theyonly a grandpa, grandma, two kids parents. So this & one happy and conand tented family. The grandparents live in a village, with a fishpond that lies nearby. One day, the grandpa decided...[Pnina, girl,n, 7th grade, pic ture-based fiction-continues all in past tense] b. tov. ani m t i e nogim li ba-dvarim, j e mitaskim li im hadvarim baxeder. yom exadaxoti Yael hexlita . . . Okay,itwhen(people) touch my things, mess around with the things in my room. One daymy sister Yael decided ...[Hila, girl,13,7th grade, fight story] c. lifney beerex shvuayim noda li, l e ani nosaat le-xu/ be-taarixi e & & im mivxan be-anglitsheli, aznigashti la-rnora le-anglitsheli ., , About two weeks agoI learned thatI g~ am going abroadon a date that = conflictswith my English test, [= then] I-went-up to my English teacher ...[Merav, girl,16, nth grade]

Past Present [Present = Historical, Narrative Present]

etmol halaxti le-male delek ba-oto ha-tsahov k s e w lesham, bederex ve klal ani memaleki ha-ovdim be-taxanot ha-delekmitaclim laasot et avooto betox datam. az vacati me haoto lakaxti et ekdax ha-delek ve ve hamexonit, hitxalti le-male delek ve hu mistake/ alay kaxa, amarti lo j e ani m e lemale shemen,azar li le-male shemen. axarev ze ani& le-shalem lo im ha-viza, kmo i e ani meshalem bederexklal, ve bederex klalani tamid Sam lev, bekicur, hu-r li i e ha-mexir ha-kolelhu ... YesterdayI went to fill up the yellow car with gas, and when I there, I

present tense for rest of story, until last 8 clauses out of 80 = the coda, also introduced by to cut a long story shortq

usually do it because the guys working there too lazy to do it properly. So I ofmycar and took the hose and inserted it, I started to fillup, and he looksat me in a weird kind of way, and I told him I oil as well, (he)helpedmewith that. After that, I to pay him,with my credit card, likeI always W ,I usually watch what he hto cut a long story , short, he me the price is . .[Hanan, man, 25, fight story, continues in .

It is not by chance that there are three examples (9.1) of shifting from in present to past, from older school children, but only one example of shifting from pastto present, from an adult. general, across our database, there were In far more examples the first than ofthe latter shift between the tense of two forms. This might seem surprising, because past tense (basically perfective though also possible with durative predications in Hebrew) might seem better suited to the anterior nature background, scene-setting situations. However, of as noted, in Hebrew present-tense forms function as nontensed participialso als in complement and adverbial clauses expressing attendant circumstances. Thus, from late schoolage, but notbefore then, narratorsshowed the impact of

Setting the Narrative Scene


Biblical and other literary fiction for expressing narrative temporality and background-foreground distinctions by means of the participial-(generic or durative) present for setting versus use of the morecompletive, sequential pasttense forms fornarrative events. The reverse example in (9.2) is lesstypical and reflects a highly individual Damon Runyonish type ofstyle, which only afew adults and none the younger subjects adopted. of As noted, shifting between present to past and past present to mark off to story scene-setting from story plotline only onedevice used by proficient is Hebrew-speaking narrators. Another is by contrasting the simple or perfective past with the complex habitual, durative past form. This is overtly marked by combining the past-tense form the verb huyu be (or any of its alternants in of 1st and 2nd person, singular plural) as an auxiliary with the participial, prevs. sent tense form the main verb(which agrees with the subject in number and of gender). This quite common verb form never once occurred in the Hebrew frog-story sample, and it is extremely rare in the conversational usage of preschool children through age 5 (Berman & Dromi, 1984). But it does serve proficient narrators as a rhetorical option for formally marking off background settings from the foreground narrative chain events, or vice versa, as illusof trated in (10).

Shifting between Simple (Perfective) and Complex (Habitual) Past Tense:

1 .

Habitual Past [=haya waslwere+ Benoni Participle] Simple Past a. ze shana,yeladim lifney ve ba-kita sheli h a p osim shtuyot, mitkashrim habaytave ze. Gyeled e x a d d v ie ani hitkasharti elav ... It was last year,and the kids in my class were doing crazy-things, & people at home and so on. [=then] one kid thought that I (had) called him.. .[Tal, boy, 12;5,7th grade] b. li xavera axatie hi hayta mexatetet baaf ve ani haviti koseset cipornayim, hayinuvoshvot axat leyad ha-shniya, ve haviti rava ita, haviti omeret la, haviti tsoeket aleha, ve hi havta omeretli ....azpaam halaxti ita makot bemizderon bet ha-sefer
I had a friend in first grade, that was uicking to pick] her nose, and [=used

I was biting my nails, we were sitting next each other,and I was arming to with her, I was saying to her,I was shouting at her, and she saving to was me ....So once I simply got into a fight with her the school corridor. in [Shani, woman,23,continues all the rest in simple past tense]
2 .

Simple Past Habitual Past

loh & lahemmusagex le-tape1 be-tinok,azhemotale-imuts,aval ze sipur axer legamrey, ex ve lama hi e a kax. ha-sabasheli axarey kama shanim nisa le-baxura tseira bat 17, ve & lahem od shney yeladim beyaxad. ha-aba sheli &l be-mosdot, & lo aba ve ima aval hu kmo ba-sipurim im ima xoregetrashaitie hayta otam be-arba ba-boker ve meira



otam me ha-mita, kulam h a p kamim be-arba ba-boker lenakot et ha-bayit, hayta rneshugaat ha-aba s h e l i m r le-kibuts, hu& sham shanim, Dapash sham et ha-ima sheli ... They didnt have a clue how to take care of a baby, g a ~ e so they her up for
adoption, but that is a whole other story, how and why it hapened that they did so. My grandfather after a few years married a young girl17, of and they had another two kids together. My father up in institutions, h e had a mother and father but he like in the storybooks with a wicked stepmother who was waking used to wake] them up at [= four in the morning and (m) throwing them out of bed, of them were getting all up at four a.m. to clean the house, she was compulsive about cleanliness .. met my ..At some uoint my dad moved to a kibbutz, he years there, mother there [Chaya, woman, telling lifestory to a friend] 33,

Along with useof tense/aspectshiftingtodistinguishbackground circumstances from foregroundedevents, more mature or proficient Hebrew narrators alternate predicates two additionalways. First,they rely heavilyon in the verb h a y to indicate both copulabe and possessive havein background clauses, in contrast to the lexically specific verbs that they prefer in the sequential part of the narrative. Second, narratorsuse stative-durative verbsas background predicates, and activity or event verbs elsewhere; that is, they make use of inherent aspector Aktionsarten distinctions toset off background from foregrounded events. These findingsfor how Hebrew speakers alternate across predicate types in order to markoff different components of their stories illustrate several more general themes. In crosslinguistic terms, speakers will rely on the formal options made available to them the typological structure of their native by language, rather than seeking to use compensatory periphrastic means for marking distinctions not made their grammar. On the other hand, in proficient speakers, and they alone, resort to full range of textual devices for marking a relevant distinctions, across a range of forms whichis not immediately obvious from grammatical oreven lexical analysis at the level of the single sentence. Furthermore, proficient narrators deploy these devices in a way that is not accessible to younger, less proficient speakers in constructing narrativetexts. Besides, even among fully proficient narrators, use of these devices is optional rather thanobligatory. Narrative texts in Hebrew sound perfectly well formed if they are constructed entirely in past or in present tense, or without any surface marking of habitual past aspect contrasting with simple past tense. However, the ability to exploit such rhetorical options gives the narratives constructed by skilled narrators a textual flavor, a richnessand variety whichare the hallmark of good storytellers and storytelling.

Setting the Narrative Scene



This study has confirmed findings of prior research on narrative development to theeffect that youngpreschool children do not appear to recognize the need to provide their audience with relevant background information. Subsequently, at a more structured,middle-level phase of development, narrators provide at least the minimal background information needed to frame events in place and time, and they occasionally add motivation for the events that will ensue. However, metacognitive comments on the task itself and/or onits thematic content or on the script (say, of an adventure storyor a personal experiitself ence with a conflict situation) are given only by mature narrators,reflecting a quite different type and level of communicative competence.Moreover, as I have noted elsewhere (Berman, 1988,1995), the greatest individual variation is found at the two extremes, among theyoungest children and the adults. Some adults tell stories as straightforwardlyinformative and well structured as school childrens, while other adults devote as much as 5 0 % of their texts to background before proceeding to the onset the action. of The question of what childrens narrative abilities can tellus about their knowledge of language is not a simple one, because narrative construction is a domain in which linguistic structure interacts in complex ways with general cognitive faculties (Berman & Katzenberger, 1998; Shatz, 1984). These include the ability to give expression to aninternalized narrative schema in the formof an action structure organized around goal or problem, attempts tomeet this a goal, and a resolution. Also dependent oncognitive underpinnings is the ability to provide adequate and appropriatebackground information to set the scene for the story that about to unfold. Nonetheless, certain common themes is emerge to illuminate how children develop the ability to linguistic forms use for meeting such narrative functions. These themes are shared by the findings of the large-scale crosslinguistic frog-story study of Dan Slobin and ourcollaborators (Berman& Slobin, 1994); by the analysis of the expression of temporality and connectivityin five different contexts used for narrative elicitation among Hebrew-speaking subjects (Berman, 19 9 5; and by the morespecialized ) study of story-beginnings presented here. First, from the pointof view of form:function relations, the same surface forms (e.g., the Hebrew counterparts of once, one day) fulfil different narrative functions with age. Moreover, some forms initially serve in only restricted contexts, but with time come meet a wider range of narrative functions. to Thus, young childrenuse stereotypic lexical items to introduce their stories, whereas mature narratorsrely on less conventional rhetorical devices to set off background orienting elements from the mainstoryline. Among the youngest narrators, the distinction between background and foregrounded elements is often unmarked orinitially blurred, whereas subsequently it is marked by relatively nonexplicit additive or temporalexpressions like those meaning and



(then), after that. Only later in development is the transition from scene setting to plot onsetclearly marked by explicit lexical as wellas grammatical devices, including tense/aspectshifting in some cases. Second, and relatedly, most of the relevant linguistic forms are available from early on, for example use ofpast tense marking verbs or lexical markers of of temporal sequencelike one time, afterwards. Yet even where childrendo have command of the relevantlinguistic forms at the level of the simple clauseand, later on, for relating adjacent clauses, it takes them a long time to learn how to deploy these formsboth flexibly and appropriately in the context of extended discourse. In the present context, they need to know which linguistic forms to use in order to distinguish background scene-setting elements from the foreground chainof narrative events. And they mustdo so by using appropriate lexical markers of the transition andby flexible shifting between predicate semantics, tense, and aspect in background versus foreground elements. Furthermore, some forms do not appear to be used at until quite late. all Examples includeuse of the past perfect in English and Spanish (Kupersmidt, 199 6 ; SebastiPn & Slobin, 1994), use of syntactic passives in Hebrew (Berman, 1997b) and,as shown here, use of Hebrew habitual past aspect marking. These findings provide strong motivation further examination of the more general for issue of late acquisitions and the need to accountfor the delay in emergence of some forms compared with others (Berman, 1998; & Avidor, 1998). Ravid These findingspoint to the importance including adultsubjects as a basis of for comparison and for evaluating the range options used by proficient of speakers in different types of narratives. The present study shows that we should include teenage narrators, as was done tosuch fine effect by Labov too, (1972). Adolescents in general, and high school students in particular, can illuminate in importantways how developing narrativeabilities and linguistic form:narrative function relations are affected by school-based literacy and increased exposure toand awareness of different types of narrative genres and the rhetorical options suited to each one. It seems to through to adulthood take until this knowledge further incorporated into a personal is style and the narrative stance that each individualselects to deploy in any given context. Next, as in other domainsof development, task effect is relevant here, too. Children proved able mark off setting elements betterand earlier in personal to experience accountsthan in narratives basedon a picturebook story. And they did so as young as age 3 when they were free to tell about anything that had happened to them, but only from around 5 when asked to tell specifically age about a fight they had experienced. In general, personal experience accounts appear to provide more authentic contexts for elaborating on scene setting than do picture-based tasks. These preliminary findings indicate that, noted, the as methodological and developmental issue of task effect could be illuminated by in-depth, suitably controlled studies how andwhen each setting element of is expressed across different narrative genresand indifferent elicitation settings.

Setting Narrative the Scene


Additional avenuesfor further research that emerge from this study are indepth examination of crosslinguistic and crosscultural differences that might affect how scene-setting circumstancesas distinguished from plotline events are expressed across development, example, in languages with for rich tense/aspect distinctions or in cultures with highly conventionalized formats for this purpose. Finally, as a possible sourceof new insights in the domain of general as well as developmentally motivated narrative research, it would seemof interest to compare such analyses of scene-setting elements with extent and way the in which children and adults give expression to the coda in different types and contexts of narrative production.

Research on which this study based was funded by grants from the US.-Israel is Binational Science Foundation to R.A. Berman and D. I. Slobin of the University of California, Berkeley, from theLinguistics Program of the National Science Foundation to D. I. Slobin, and from the Israel Science Foundation to R. A. Berman and D. D. Ravid of Tel Aviv University. The authoris indebted to students who participated in her seminar on narrative development Aviv Tel at University in the years 199 5 1 9 7 for their assistance in providingdata, toDr. -9 Irit Katzenbergerfor her invaluable help in data collection and analysis, and to Iris Levin and Yeshayahu Shen of Tel Aviv University as well as to the editors of this volumefor their valuable comments earlier drafts. Responsibility for reon maining inadequaciesrests with the author alone.

Argeman, K.(1996). Sipurey bdaya shel talmidim be-alpe U bixtav [Schoolchildrens make-believe stories in speech and writing]. Tel AvivUniversity seminar paper. Bamberg, M,, & Damrad-Frye, R.(19 91).O n the ability to provide evaluative comments. Journal of Child Language, 18,689-710. Bazanella, C., & Calleri ,D. (19 9 1). Tense coherence and grounding in childrens narratives. Text, 11, 175-187. Ben-Haviv, 0. (1996). Hashvaa beyn rexiv ha-reka be-sipureyhem she1 yeladim bney 3,7, ve 12 shanim ve shel mevugarim [A comparison of the setting component in the narratives of children aged 3,7, and years and adults]. Tel Aviv University seminar paper. 12 Berman, R.A. (1988). On the ability to relate events in narrative.DiscourseProcesses, 11, 469-497. Berman, R.A. (1993). The development language use: Expressing perspectives on a scene. InE. of Dromi (Ed.), Language and cognition: A developmentalperspective(pp. 172-201). Norwood, NJ: Ablex. Berman, R.A. (1995). Narrativecompetence and storytelling performance: how children stories tell in different contexts. Journal ofNarrative andLife History, 5,285-314. Berman, R.A. (1996). Form and function in developing narrativeskills. In D. I. Slobin, J. Gerhardt, A. Kyratzis, & J. Gu (Eds.), Social interaction, social context, and language (pp. 243-268). Malwah, NJ: Lawrence Erlbaum Associates. Berman, R.A. (1997a). Narrativetheory andnarrative development:The Labovian impact. Journal ofNarrative and LifeHistory, 7,235-244.



Berman, R. A. (1997b). Preliterate knowledgeof language. In C. Portecovo (Ed.), Writing development:An interdisciplinary view(pp. 61-76). Amsterdam: John Benjamins. Berman, R. A. (1997~). Developing form/function relations in narrative text production. Lenguas Modernas, 24, 45-60. Berman, R. A. (1998, January). Emergence and mastery in language development. Talk presented to Linguistics Colloqium, Tel Aviv University. Berman, R. A., (19 9). Bilingual proficiencylproficient bilingualism: Insights from Hebrew-English 9 narratives. InG. Extra &L. Verhoeven (Eds.), Bilingualism and migration (pp. 187-208). Berlin: Mouton de-Gruyter. Berman, R. A. (in press). The role of context in developing narrative abilities: The frogstory findings in lightof other narrative genresand elicitation settings. In S. Stromqvist & L. Verhoeven (Eds.), Relating events in narrative: Variation acrosslanguages, cultures, and genres. Mahwah, NJ: Lawrence Erlbaum Associates. Berman, R. A., & Dromi, E. (1984). O n marking time without aspect in child language. Papers and Reports on Child Language Development, 23,21-32. Berman, R. A., & Katzenberger, I. (1998). Cognitive and linguistic factors indevelopment of picture-series narration. In A. G. Ramat& M. Chini (Eds.), Organization of learnerstexts, Studia italiani i linguistica theoratica e applicata(special issue), 27,21-47. Berman, R. A., & Slobin, D. I. (1994). Relating events in narrative: A crosslinguistic developmental study. Hillsdale, NJ: Lawrence Erlbaum Associates. Chvany, C. V. (1985). Backgrounded properties and plotline imperfectives: Towards a theory of grounding. In M.S. Flier & A. Timberlake (Eds.), The scope of Slavic aspect (pp. 247-273). Columbus, Ohio:Slavica. Giora, R.,& Shen,Y. (19 94). Degrees of narrativity and strategies of semantic reduction. Poetics, 2 , 2 447-458. Herman, J. (1996). Grenouille, whereare you? Cross-linguistic transfer in bilingualkindergartners learning toread. Unpublished doctoral dissertation, Harvard Graduate School of Education. Hickmann, M. (1991). The development of discoursecohesion: Some functional and crosslinguistic issues. In G. Piiraut-Le-Bonniec & M. Dolitsky (Eds.), Language bases ..discourse bases: . some aspects of contemporary French language psycholinguistic research (pp. 157-185). Amsterdam: John Benjamins. Hickmann, M , , Hendriks, H.,Roland, F., & Liang, J. (1996). The marking new information in of childrens narratives: A comparison of English, French, German, Mandarin Chinese. Journal of Child Language, 23, 59 1-6 2 0 . Hickmann, M,, Kail, M,, &Roland, (1995). Cohesive and anaphoricrelations inFrench childrens F. narratives as a function of mutual knowledge. First Language, 15,277-300. Hopper, P. S. (1982). Aspect and foregrounding in discourse. In T Giv6n (Ed.), Discourse and syn. tux: Syntax and semantics, Vol. 12 (pp. 213--241). New York Academic Press. Kahanowitz, S. (1995). Darxey ha-rifrur be-sipurey yeladim U mevugarim ]Expression of reference in the stories childrenand adults]. Tel AvivUniversity seminar paper. of Kail, M,, & Hickmann, M. (1992). French childrens ability to introducereferents in narratives as a function of mutual knowledge. First Language, 12,73-94. Kail, M,, & Sanchez-Lopez, I. (1997). Referent introductions inSpanish narrativesas a function of contextual constraints:A crosslinguistic perspective. First Language, 17,103-130. Katzenberger, I. (1994). Hayexolet lesaper sipur al piy sidrat tmunot:hebetimkognitiviyim, leshoniyim, ve hitpatxutiyim [Cognitive, linguistic, and developmental factors in the narration of picture series]. TelAviv University unpublished doctoral dissertation. Kern, S. (1997). Comment les enfrantsjonglent avec les contraintes communicationnelles, discursives et linguistiques dans la production dune narration [How children juggle communication, discourse, and linguistic constraints in the production a narrative]. Unpublished doctoral disof

Setting Narrative the Scene


sertation, Universite Lumikre-Lyon 2 . Kernan, K. T.(1977). Semantic and expressive elaboration in childrens narratives. In S. Ervin-Tripp &C. Mitchell-Kernan (Eds.), Child discourse(pp. 91-104). New York: AcademicPress. Kupersmitd, J. (1996). Tense-aspect marking in Spanish-Hebrew bilingual narratives. Tel Aviv University seminar paper Labov, W. (1972). The transformation of experience in narrative syntax. In W. Labov (Ed.), Language in the inner city(pp.355-3516), Philadelphia: University of Pennsylvania Press. Mayer, M. (1969). Frog, whereareyou!NewYork Dial Press. context oftheir Menig-Peterson, P,, & McCabe,A. (1978). Childrens orientation of a listener to the narratives. DevelopmentalPsychology, 14, 9 2 - 9 2 , Minami, M. (1996). Japanese preschool childrens narrative development. First Language, 16, 339-364. Oz, A. (1996). Matxilim sipur[The story begins]. Jerusalem: Keter Publishers. Peterson, C. (1990). The who, when, and where of early narratives. Journal ofchild Language, 1 , 7 433-456. Peterson, C.,& McCabe, A. (1983). Developmentalpsycholinguistics:Three ways oflookingat a childs narrative. New York:Plenum. Peterson, C.,& McCabe, A. (1991). Linking connective use to connective macrostructure. A. In McCabe & C. Peterson (Eds.), Developing narrative structure (pp. 29-54). Hillsdale, NJ: Lawrence Erlbaum Associates. Pitcher, E. G., & Prelinger, E. (1963). Children tell stories: A n analysis offantasy. New York: International Universities Press. Pradl, G.M. (1979). Learning howto begin and enda story. Language Arts, 56,21-25. Polanyi, L.(1985). Telling the American story. Norwood, NJ: Ablex. Ravid, D. (1995). Language change in child and adult Hebrew: Psycholinguisticperspectives.Oxford, England: OxfordUniversity Press. Ravid, D., & Avidor, S. (1998). Acquisition of derived nominals in Hebrew: Developmental and linguistic principles. Journal ofchildLanguage, 25, 229-266. Reilly, J. S. (1992). How totell a good story: The intersection of language and affect in childrens Life narratives. Journal ofNarrative and History, 2,355-377. Reinhart, T. (1984). Principles of gestalt perception in the temporalorganization of narrative texts. Linguistics, 22,779-809. Reinhart, T. (1995). Mi-tekst le-mashmaut: Emcaey haaraxa [From text to meaning: Strategies of evaluation]. In Y Shen (Ed.), Cognitive aspects of narrative structure (pp. 4-37). Tel Aviv . University: The Porter Institute [inHebrew]. Rumelhart, D. E. (1975). Notes on a schema for stories. In D. G. Bobrow & A. Collins (Eds.), Representation and understanding: Studies in cognitive science (pp. 211-236). New York: Academic Press. Said, E. A. (1978). Beginnings: Intention and method.Baltimore: Johns HopkinsUniversity Press. Sebastian, E., & Slobin, D. I. (1994). Development of linguistic forms: Spanish.In R. A. Berman & D. I. Slobin (Eds.), Relating events in narrative: A crosslinguistic developmental study (pp. 239-284). Hillsdale, NJ: Lawrence Erlbaum Associates. Segal, R. (1996). Kriat sefer keilu bekerev yaldeygan bi-tnaey divuv shonim[Kindergartners, pretend-reading intwo different conditions]. Tel AvivUniversity seminar paper. Seidman, S., Nelson, K., & Gruendel, J. (1986). Make believe scripts: The transformation ERS in of fantasy. In K. Nelson (Ed.), Event knowledge: Structure and function in development (pp. 161-188). Hillsdale, NJ: Lawrence Erlbaum Associates. Shatz, M. (1984). A song without music and other stories: How cognitive process constraintsaffect childrens oral and writtennarratives. In D. Schiffrin (Ed.), Meaning,form, and use in context: Linguistic applications (pp. 313-324). Washington, DC: GeorgetownUniversity Press. Shen,Y. (1988). The X-Bar grammar for stories: Story grammar revisited. Text, 9,415-467.



Shen, Y., & Berman, R.(1997). From isolated event to global action-structure. InJShimron (Ed.), Psycholinguistic studies in Israel: Language acquisition, reading and writing (pp.119-145). Jerusalem: The Magnes Press, Hebrew University [in Hebrew]. the Slobin, D. l. (1993). Passives and alternatives in childrens narratives in English, Spanish, German, and Turkish. In B. Fox & P. J Hopper (Eds.), Voice: Form and function (pp. 341-364). . Amsterdam: Benjamins. Sternberg, M.(1978). Expositional modes and temporal ordering infiction. Baltimore: John Hopkins University Press. Sutton-Smith, B. (1981). Thefolk stories ofchildren. Philadelphia: University Pennsylvania Press. of Umiker-Sebeok, D. J (1979). Preschool childrens intraconversational narratives. . Journal o f c h i l d Language, 6,91-109. Van Dijk,T. (1976). Philosophy action and theory narratives. Poetics, 5,287-338. of of Wlgglesworth,G. (19 92). Investigating childrens cognitive and linguistic development through narrative. Unpublished doctoral dissertation,La Trobe University, Australia.

Representation of Movement in European Portuguese:

A Study of Childrens Narratives

Open University, Lisbon, Portugal



University of Lisbon, Lisbon, Portugal

A significant part of recent language acquisition research has focusedon the organization of information in discourse, with special reference to person, space, and time (Berman & Slobin, 1994; Hendriks, 1993; Hickmann, 1995; Smoczynska, 19 92; see also Berman, chap. 1, this volume). The acquisition of spatial terminology that differs markedly between typologically different languages has beenof particular interest (Bowerman, 1985,198 9,1996; Choi & Bowerman, 19 9 1).At the same time, new wave of research on language and a space has uncovered enormous variation in the linguistic coding spatial of relationships (Goddard, 1998). The emphasis has been, on the one hand, on crosslinguistic variation in spatial semantics and,on the other, on the semantic primes of spaceproposedwithinAnna Wierzbickas natural semantic metalanguage(Goddard, 1998;Wlerzbicka, 1996,1998; see also our discussions inBatorko, 1998b; Batorko & Duarte, 1998). Acquisition research on the organizationof information in discourse has particularly focused on two importantissues: the marking of information status and the groundingof information in discourse. In both domains, three recurrent observationsthat mustbe taken into account in any model mother of tongue acquisition are reported. These are a relatively late developmental progress in discourse organization, interrelations among the utterance and discourse levels of analysis, and a combination general cognitive developof mental patterns with language-specific ones (Hickmann, p. 215). 19 95, Taking into consideration both cognitively and linguistically oriented studies it can be postulated that early acquisition is based not only on universal sensorimotor concepts but also on the particular language being acquired. Earlierresearchacross a number oflanguagesrevealed that childrens production of locative expressions is determined by cognitive complexityand


B A T O R E O 6: F A R l A

follows a similar sequence that can be summed up in three stages: (a) in, on, b under, beside; ( ) between, back, front, with featured objects; (c) back, andfront with nonfeatured objects (e.g., Johnston& Slobin, 1977; Slobin, 1973). Since the 1 9 8 0 s however, in contrast to the findings mentionedearlier, and in line with the language typology developedby Talmy (1983,1985) andfollowed in Bowermans research (Bowerman,1989; ChoihBowerman, 1991),it has been shown that from very early ages children are highly sensitive to thespecificities of their language. They acquire different kinds of linguistic devices in order to mark locative states versus dynamic actions, expressingboth motionin general, as in jump or fry around and change of location, as in jump into X or fly out of Y. These devices may be both global andlocal in character (Hickmann,199 5). Local devices include lexical and morphological markers of motion such as posture mrbs, prepositions, adjectives, adverbials, particles, deictics, case markings, andso forth, whereas global devices include word order and event conflation mechanisms correlated with other language-specific factors. Some examples of language-specific local devices of lexical and morphological character are the morphologically complex and transparent Portuguese motion verb atravessargo overlacross = a travls + ar (spatial preposition across + infinitive marker), the English particle over (Example I),the Polish preposition do into and genitive case marker - U indicating direction (Example 2 . ) Conflation of Motion andPath in Portuguese atravessar (go + over/across) or conflation of Motion and Manner English drivego by car(Example l), on in the other hand, instantiate global devices. Note that in the case of the Portuguese motion verbatruvessar go over/across we observe both a global device of event conflation (Motion and Path) and a complex combination of local devices such as a preposition, an adverbial and an infinitive marker in a morphologically transparent verb form.


Portuguese: English:

Atravessdmos a de ponte go across 1st pl.pastbridge the


carro car

We drove the overbridge dom + -U go into house/home Gen To enter the house
do domu

( ) 2



It is considered to be easier for children to interpret sentences when their language has a richand transparent morphology(e.g., Italian or Polish) than when it depends moreon word order to express grammatical relations(e.g., English; Hickmann, 9 5). Comprehension studies in different languages have 19 also shown that, regardless their age, native speakers the cuesthat are the of use most available and reliable intheir language, example, word order English for in and lexical or morphological cues in Polish, suggesting a model in which

Representing Movement in Portuguese


children learn how functions completeand fuse in relation to available forms (Hickmann, 1995; Smoczynska, 1992). In view ofthe foregoing claims and thefact that Portuguese is a morphologically rich Romance language,we hypothesize that in theprocess of acquiring their native tongue, Portuguese speakerswill behave according to the Polish rather than the English model. This means that Portuguese learners will follow lexical and morphological cues that the most available and reliable in their are language ratherthan some global cues such as word orderas inlanguages of the SVO type.

Our research (Batorko, in press) focusing on the expression of space in EuropeanPortuguese follows Talmys (1975, 1983,1985)workonthe structuring of space in language.Talmy defined a motion situationas one in which one object (Figure) is moving or located with respectto another object (Ground), whereas the respect in which the object located or moving to is another object functionsas the Path (Talmy, 1975). Hefurther postulated that event conflation in the clause is the object of linguistic variation. Some languages, such those in the Germanicfamily, combine Motion with Manner as in the verb (as in English drive in Example 1). Others such as Romance languages (Portuguese, Spanish, French, combine Motion with Path in etc.) (as Portuguese atravessar go across in Example I), expressing Manner other by means suchas relativeand infinitive (or gerundive) clauses (seeExamples 4 and 7 ) .Thus, it follows that, whereas English speakers elaborate the trajectories that protagonists follow in their displacement through space, Portuguese, Spanish, or French speakers provide simpler descriptions of displacements with less elaborate paths and morestative information situating the protagonists. This means that in English stative locations must be inferred from paths as in Example (3) from the storyFrog, where are you? (Mayer, 196 9; see Berman & Slobin, 1994), in which the frogs inferior positionis presented as a result of the boys movement on the path directed down.

The boy put the frog down into a jar

In Portuguese or Spanish paths mustbe inferred from path verbs (e.g., meter = put inside) and static locations (e.g., haver em baixdhaber abajo = there be (located) down). Example (4) presents the equivalents in Portuguese and Spanish of Example (3).


Portuguese: 0 meninometeu o sapono frasco que havia la em baixo the boy put the frog inlon the jar that there was there below The boy put the frog down into a jar


B A T O R ~ O& F A R I A


El nifio metid rana en la

elfrasco que

habia abajo

the put frog the that boy the in/on jar there The boyput thefrog down into jar a

was below

Our research (Batorko, 1gg5,1996,1998a, 19g8b, press) shows, however, in that it issometimes dangerousto overgeneralize the basic characteristics one of language group to all the others classified as its members. Although Romance languages are considered to be more stative than Germanic (Slobin,1989), the degree of stativity may vary from one language to another. example, in For Portuguese strongly marked stativity is not really acceptable in case of some verbs such as subir move up, whereas it is in their Spanish formal cognates (Example 5).

English: He (=the climbedtree boy) a arvore Portuguese: 0 subiu uma a menino the
arribasubido Esta Spanish:


movedup (PERF)


tree a
un arbol

located/moved is

up (IMPERF) totop the

tree a

Portuguese: 0 meninosubido arvore em da estd cima the boy located/moved is the boy (IMPERF) located is the

up (IMPERF) top the the of tree top to see of the tree


0 menino estdarvore da cima em 0 menino ver empoleirou-se para

boy himself the perched


holes the

The difference shown by the Romanceexamples in (5) is not only transparent between the dynamic and thestatic perspectives but also on the level of aspectual marking. Thus, in Spanish, the predicateestu subido- is located moved up is both stative and resultative, that is, imperfective in character (present tense), whereas the Portuguese correspondent predicate is either resultative and perfective (simple past tense), as in subiu moved up or empoleirou-se perched up, or stative, that is, imperfective (present tense), in as esta em cima is located + on thetop. It is important to bear in mind that when refer to language-specific we factors what is taken into consideration is not a supranational linguistic system-such as a Portuguese language-but a given linguistic variety used in a given community, for example, European Portuguese or Brazilian Portuguese. Being a transcontinental language officially spoken in different countries, Portuguese is subject to great linguistic variation. Furthermore, the distinction between standard varieties of European versus Brazilian Portuguese has been maintained due to specificity not only of lexical choice but also of morphosyntactic structure (Faria & Duarte, 1989). As for the study of spatial

Representing MovementPortuguese in


expressions, significant differences can be observed between the two basic varieties of Portuguese, as illustrated in Example (6).
(6) European

H cavalo urn d
(0 det.) galloping horse [gerundive]

a galopar

Portuguese:is horse there a (prep.)

cavalo Brazilian Portuguese: have

to gallop [infinitive clause]


As the example shows, the difference between the two main varieties

of Portuguese lies in the selection of the determiner,in the choice of the existential introductory verb-haver there be/have or ter have-and on the morphosyntactic level. The infinitive clause is preferred in the European variety (though thegerundive construction still possible both in the standard is variety and, especially, in some dialects), whereas in Brazil the gerundive is used (Batorio, 1998b, Batorio& Duarte, 199 8). The beginningof a narrative contains a lot new information given in the of first or the first few utterances. Such information packing presents a heavy processing load, which can be dealt with in different ways, as illustrated in Example (7).

Data from Batorko (in press)-adult narrators

cavalo Hd que a. urn there that in horsewalks a is the meadow cavalo urn Havia passei pelosprados a galopar

there was
Era urna


to gallop[infinitiveclause]
que estava a galopar no campo estava galopando [infinitive or gerundive]

vez urn cavalo

there was horse galloping a that was in once upon a time

Era urna vez urn cavalo que estava cerca preso nurna

the fields

there was horse a that was imprisoned in the corral once upon a time There was a horse galloped inthe fieldslin the meadowlin corral that the

b. cavalo Andava urn horse aused

cavalo urn Andava

correrpelos a prados


run across the meadows


used to golwalk across a horse the meadows There was a horse used to run across the meadows that c.
Urn andava cavalo a correr pelosprados

horse aused to run across the meadows There was a horse used to run across the meadows that


B A T O R ~ O6: F A R I A

Brazilian Portuguese

Data from Horse and Cat Stories in Guimarzes (1994): Aqui tern urn correndo cavalo (5 years) here have a horse [gerundive] running There is a horse running here Tern urn cachorro drvore passarinho e na (5 years) have a dog and o det in birdie the tree There area dog anda bird in the tree
e tinha o cachorro que puxou o gat0 pelo rabo. (7 years) and have the that dog pulled he cat by the tail There wasa dog that pulled thecat by histail Aqui tern urna drvore drvore nessa e tern urn ninho de(1o years) here have a tree in tree andthis have a nest of
passarinho pousar.

birdie to sit There is a tree and in the tree there nest for a bird to sit a is b. Data from Chavegatto, Souza, Silveira, Silva, and Cassano (in press): Ld em Buzios, praia nada Ferradura, tern uns (Adult) ThereinBuzios, at thebeach of Ferradurahavesome
bares ern cima.

bars top in/on In Buzios in the upper of the Ferradura beach there are some bars part Portuguese,thus, offers a number of ways forthecombinationof new animate and inanimate referents in differentroles in the same utterance, focusing on their existential introduction (once upon a time there an NP) was as well as some extra information,packed in an additionalclause (relative and/or infinitive/gerundive) of either locative (that was on the top of, in the fields, etc.) or active character (that liked walking/running/jumping, etc.). Note that all this information-existential and locative, of stative or dynamic character-can be recurrent, thatis, it is quite common toindicate more than one location(e.g., there was a bird in the nest in the tree). This particular language characteristic, common for instance to Portuguese and Spanish, contrasts with direct one-clause-packing typical of the Germanic languages (e.g., There was an NP running in the meadow; Hendriks, 1993,pp. 85-86).

Our research (Batoreo, 1996) set in the theoretical parameters referred to earlier, examines spatial reference in narratives produced by childand adult speakers of European Portuguese in a context where the interlocutors lack mutual situationalknowledge. Narrative productions were elicited with two

Representing MovementPortuguese in
T A B L E 2.1


Characteristics ofthe Samples

G (5 years) I G2 (7 years)
G 4 (Adults)



7;6 4;4 7;3



-5;11 7;11

1 G3 ( 0years)

10;1- 10;10


18 47


Note. From Batorko (1996).

picture stories: the Horse Story and the Cat Story (Hickmann, Each story 198 z).1 is based on a sequence of pictures2 in which different animals are depicted in interactive roles that determine the status of the characterin the protagonistcategory. In the Horse Story there horse, cow, and a bird, and in theStory are a a Cat a mother bird with her ones, a and adog. Although the general frame little cat, is very similarin both cases, the stories differ in the role various protagonists can assume in the storyin relation to the other animals. In first story the horse the is the main protagonist; it runs in thefields, falls down and gets help from his friends, a cow and a bird. In the secondwe really do not know who the main story, character is. The title is Cat Story, but there is no cat inthe first picture, in which we can only spot a mother bird and some little birds in a nest in tree. The cat a appears onlyin the second picture when the mother bird away.The cat tries flies to grab the baby birds in but cannot the nest reach them because a dog comes and pulls him down from the Finally, the birdsare safe when the cat runs away, tree. chased by the dog,and the mother bird comes to take care of little ones. back her The data consistof the Batorko corpus (in press), composed of 120 narratives produced by 6 o monolingual European Portuguesesubjects: 30 adults and 30 children (half males and half females) of 5,7, and 10 years of age, with 10 children in each group (Table2.1). The children were tested in a kindergarten and a primary school in the center of Lisbon. All data used in present the study were recorded, transcribed, and coded according to the CHILDES system (Faria & Batorko, 199 4; MacWhinney, 199 4).
1 The materialswereoriginallydesignedforHickmann(1982).Theyaredescribedin Hickmannsstudies(e.g.,Hickmann,1991,1g95)anddevelopedby,amongothers,Hendrlks(1993), Hendriks and Hickmann (1998) and Smoczynska (1992). Bokus (1996) andalso our discusSee sioninBatorko1995,1996,and1998a. * Picture-by-picturesynopsis ofthe Horse Story: A horse is running in the near afence;( 2 ) (I) field The horse looksacross the fence at a cow; (3) The horse jumps fence with the in the background the cow and a bird on fence;(4) The horse stumbles on fence and falls. The cow and bird watch; The the the (5) cow bandages up the horses leg while bird holds a firstaid the kit. Picture-by-picture synopsis the Cat of Story: (1) A bud is sitting in a nest which is on a branch a tree; 2) bird flies away of ( A and a comes up cat to the tree; The cat watching the empty (4) The cat climbs the as a dog (3) sits nest; tree watches; (5)The dog pulls the tail, as the bird cats flies back; (6)The dog chases the cat away, as the bird hovers at the nest.


B A T O R ~ O6: F A R I A

100 P

80 70 -F-.
Percentage of 50 occurrences


I .


40 30 20 10


-.. " "

.. " * .".

cow + Bull + BulllOx

. . .Bird ...



- 0 5Yrs




Age Groups
FIG. 21 ..

Horse Story: Introducing the protagonist category.


The study of the representation movement in language of involves, on the one hand, ananalysis of nominal reference to determine the linguistic realizations of Figure and Ground. On the other hand, it on verbal reference to define refocuses lations between Figure and Ground, yielding different semantic-syntactic classes of verbs, such as appear-in-the-stage verbs (appear, come, go, turn up,etc.), action verbs (walk, run, jump, moveup, etc.), and perceptionverbs (watch, see, spot, etc.). In the present study three variables, (i) story, (ii) language, and (iii) age, were taken into consideration.By (i) the story variable mean the storystructure we and reference to the protagonistswith explicit variation in their hierarchical differentiation. By (ii) the language variablewe mean specific semantic and syntactic characteristics European Portuguese, of typologically and contrastively exemplified earlier. The three variables were taken into considerationin trying to answer the following qutstions: When and how do Portuguese children provide a spatial anchoring for the story, on one hand, and a spatial frame in their set narratives, on the other? linguistic terms this In involves determining (i) the semantic roles assigned to Figures and Grounds; (ii) the kindsnominal and verof bal reference; (iii) thespecific lexical and morphological means used to establish new spatial information; the kinds of syntactic constructions (iv) selected.To answer these questions shall discuss the form:function relations investigated we we in our narratives (Batorko, in press) in relation to these three relevant variables.
The Story Variable

In this sectionwe discuss the story structure and reference to the protagonists, stressing the distinction between spatial anchoring and setting the spatial of frame. Spatial information given throughout the narrative provides a spatial




Percentage of
OcrC uT



Representing MovementPortuguese in


*-. * .
S .

. . .. -" .

," .


40 30 20


" " "


Mother Bird


10 0 5YE
FIG. 2 2 ..



i IOyrs

Age Groups

Cat Story: Introducingthe protagonist category.

anchoring for the story, whereas the spatial frameis established with information given at the beginningof the story for the setting, to provide background information in theabsence of mutual knowledge. This distinction has to do with therole different protagonistsplay in the story. Setting the spatial frame at the beginning of the narrative involves introducing the main protagonist, whereas spatial anchoring has to with all the categories, protagonistas well do as instrumental, not only the introduction butacross the whole narrative. in The datareveal that narrators are more likely to introduce the main protagonist (or more than one in the Cat Story), that to set a spatial frame, than is, to introduce the otherprotagonists at the beginning. Itis observed from Fig. 2.1 for the Horse Story and 2.2 for the Cat Story that story structure no Fig. has effect on spatial framing.All narrators, at all ages, in both stories mention the main protagonists 9 o YOto 10 o % of the time. No matter what their age is, narrators, without failure,refer to the horse as well as to the bovineanimallexicalized as a vaca 'cow', a boi 'bull', or a touro 'odbull' in the Horse Story and to the cat and the dog in the Story. Protagonists lower on thehierarchy are Cat mentioned with less frequency as compared to the main protagonists. This difference is more clearly observable in the Horse Story (Fig. 2.1), in which the roles protagonists play are clear cut, compared to the as Cat Story (Fig.2.2), which has two main protagonists and a third one situated close to themin very the protagonist hierarchy. In the Horse Story the bird with the firstaid kit is mentioned less frequently (70% to 80% by 5-year-olds and adults and 50% to 6 0 % by 7- and lo-year-olds;Fig. 2.1),whereas the mother bird in the Story Cat is referred to by8 0 % to 9 0 % of the subjects, suggesting that this character is situated very close to the major characters on the protagonist hierarchy (Fig. 2.2). There are some small interesting age differences: For the younger but children (5- and 7-year-olds) the animals areof equal importance ( 9 0 % all t o l o o %of reference), whereas lo-year-olds and adults distinguish between


B A T O R ~ O& F A R I A



Percentage of




Types of constructions


FIG.2.3. Percentage of

different types of structures used for the introduction the of horse category in the Horse Story different age groups. by

refermce to thefirst two characters, the cat and dog (10 o %) and the next one, the mother bird ( 8 0 % to 90%; Fig. 2 2 .Thus children as well as adults .) differentiate between main and secondary or tertiary characters, that is, between spatial framing and spatial anchoring. Comparing Figs 2.1 and 2 2 it ., can be observed that thepercentage of the introductory constructions lower is for the characters lower in status on the Protagonist hierarchy: The lower the role, the later in the process of acquisition the spatial anchoring takes place. A s discussed in the Theoretical Framework section, in order to introduce a protagonist, syntactically complex constructionsof existential and/or locative character (both static and dynamic) are required in European Portuguese. When there is a clear-cut protagonist division as in the Horse Story, the main character is introduced at the very beginning withexistential verbs huverthere be and ser be, as in pragmatic construction umu vez once upon a time the era there was; with stative constructions withestar to be located (estar + a Infinitive, estar + Gerundive); or with a descriptive motion verb such gallop, as run, jump, as shown in Example (7). Figure 2.3 represents the proportion of different types o f structures usedby different age groups to introduce the main protagonist in the HorseStory. When there is not a clear-cut protagonistdivision, as in the Cat Story, the main protagonists-both the cat and the dog-are introduced by adults with the appear-in-the-stage verbs of the unaccusative type, such as aparecer appear, surgirturn up, vircome, get in, chegarcome, aproximar-segetcloser, as illustrated in Example(g).

Representing MovementPortuguese in



Data from Batoreo (in press)-adults

Chegou urn gato

came a cat
Apareceu urn cdo

appeared a dog
Aproxirna-se urn gat0

comes up a cat
A s Examples (7) and (9) show, constructions used in the adult grammar for

spatial anchoring (i.e., introducing a protagonist in the middle narrative) of the focus on new information (the new protagonist) andsyntactically locate it to the right of the verb as a postverbalindefinite subject: a V+ Subject construction. Let us compare the examples (IO): in

V + Subject V
Era urna vez urn passarinho . .

once upon a time there was


a birdie...
urn cavalinho q u e . . .

there was

a horsiethat.. .
urn passarinho ern cirna de

was located

a birdieon the top of.. .

urn gat0


a cat

+ (Indefinite) Subject type,it clearly stands out as a marked one in anSVO

pro-drop language suchas European Portuguese. This is significant for the fact process of acquisition because a child learning the language has to acquire this construction as different from the canonical order master it properly if he and or she wants to be a successful storyteller in his or her mother tongue. This is observed to be a long complex process; definite SVO constructions constiand tute approximately 40% of all the introductory constructions produced 5by year-olds, whereas they make less than 10% of the adult introductions. 5 up At years we find four different types of constructions, some of which stop being functional in narrative introductions in later years. These are thefollowing: Definite Subject+ Verb, Indefinite Subject + Verb, Verb+ Definite Subject, and Verb + Indefinite Subject, illustrated in Example (11).

If in the adult grammar the predominant introductory construction V is of

Data from Batoreo (in press)-5-year-olds a.DefiniteSubject+Verb

0 caval0 estava a galopar sernpre, sernpre, sernpre

the horse was galloping always, always, always


B A T O R ~ O6: F A R I A

Urn passarinho estava no ninho

a horsie was located in the nest c. Verb

Chegou o ciio

+ DefiniteSubject

appeared the dog d. Verb

Encontrou urn gat0

+ Indefinite Subject (standard adult)

(he) found/met a cat

This rich pattern of introductory constructions observed atage 5 does not develop in a linear way into the target adult grammar. Having the morphoall logical and lexical markers the adults European Portuguese children deploy use, them in syntactic constructions that adult speakers very seldom produce or even never use. As a matterof fact, all the constructionsexemplified in (11) can be found in the adult data, but constructions (lla, and c) are extremely b, infrequent (approximately 5% of all the setting constructions in general), whereas children use them quite productively. The really standard construction (lid) is oftype V + Indefinite Subject. However, and in contrast to theexamples in (II), there are also child constructions that never appear in the adult narratives: At different stages of their development children test different syntactic possibilities and then drop them when they happen not to suit the input received. Notice that this is precisely what happens with some ofthe uses verb sertobe in European Portuguese(Fig. 2.4). As Fig. 2.4 shows, both adults and children telling the Horse Story frequently use a partially frozen pragmatic construction, erauma vez once upon a time there was, with a ser to be form in Past Imperfective, era was(approximately 50% of all the introductory constructions). The of this construction is, use however, the only similarity between the two groups because the children, in general, prefer other constructions with the verb tobe, whereas adults ser choose another to be verb, haver, used in a clearly existentialconstruction. Children of different ages prefer differentser constructions, some of which are only observed for a short period (e.g., at one age level in our study). Thus only in the data of 5-year-oldswe have a descriptive, strictly deictic construction, do aqui there is, which is totally absent from the more developed discourse whe lack of mutual knowledge (and, therefore, ofdeictic indication) expected lack is (Example 12).

Data from Batorko (in press) 5-year-olds Aqui L urnpassarinho cornfilhotes

here is a birdie with little one

Aqui L urn gat0

here isa cat

Representing Movement

in Portuguese


Era uma vez... (once... )


Types of introductory expressions

(There be)


F I G . 2.4. Percentage of existential relative constructions used for the introduction of the horse category in the Horse Story.

Although the construction itself is well formed with an indefinite subject in postposition, children make incorrect of it in the narrative context. ages use At 7 and 10 we observe some examples of ser to be in its Imperfective form, era was (and atage 7, even in Present--t is), that can possibly be seen as the first part of the once upon a time construction era uma vez (Example 13). In these constructions, unusual from the target grammar point of view, the syntactic construction itself is wellformed, again with the postverbal Indefinite Subject.

Data from Batorko (in press)

Era urn caval0 que ia U correr (7- and lo-year-olds) was (IMPERF) a horse that was running

e urn cuvulo que estd a correr (7-year-olds) is (PRESENT) a horse that is running

As for the verb haverto existential constructions, noneof the lo-year-olds be use it, though younger children so occasionally (approximatelylo%), and do adults use it in half of all the introductory constructions they produce (Fig. 2.4). A s for the Cat Story, the data show that, with the appearance of of age, one the main protagonists, the cat thedog, comes to be expressed not as a target or of a perceptual (or physical) encounter of another protagonist but as one who shows up independentlyof other protagonists, requiring an appear-in-thestage verb (Fig. 2.5). In other words, older children are morelikely to introduce the dog a in sentence like a dog came up than the cat saw a dog, which is characteristic of younger ones. This trend suggests that with age the children become less


B A T O R ~ O6c F A R I A

100 90

; :

1 " " .

(Cat) Perceptual verbs

verbs (Dog)

-"Appear" type verbs (Cat)





Age Groups
F I G . 2.5. Percentage of "Appear" type and perceptual verbs used in introducing thecat and dog categories in the CatStory by age.

dependent on making reference to their perceptual world-even from the protagonist's point of view when one animal spots another-in describing the existential phenomena of appearance and disappearance. may be that with It age children become less dependent on tracingevents from the intranarrative standpoint suchas the protagonist perspective, trying to establish a narrativeindependent perspective. However, this is not the case in the Horse Story, in which the second protagonist, thecow, is always introduced as a result of a perceptual or physical encounter with the main protagonist by all the age groups, and is introduced either by a perceptual verb (ver'see', olhar'look', avistar 'spot', etc.) or a physical encounter one (encontrar ' m e e t h d ' ) , as exemplified in( 4 . 1)

Data from Batoreo (1gg6)"adults


Do outro lado da cerca viu urna on theother side ofthe fence saw a 'on the other side the fence he saw a cow' of
dentro de urna sebe.

vaca. cow

urna vaca b. Encontrou met in a cow 'he met a cow in the corral' The LanguageVariable


SomelanguagesystemicspecificitieswerediscussedintheTheoretical framework section, from the typological point of view designating European Portuguese as aPathandMotion-conflatinglanguage,andfromthe contrastive point of view opposing it to the Portuguese variety from Brazil. In

Representing Movement

in Portuguese


the present sectionwe focus our attention on the discourse level, discussing (1) the semantic roles the referents play, and (2) the syntactic constructions they appear in.
The Language Variable: SemanticRoles Assigned to Figures and Grounds. Regardless of age,the narrators of our corpusshow a functional motivationin choosing different linguistic forms to introduce the Figures, that is, the mainor secondary protagonists. The difference is marked not only syntactically, as shown in the examples in thelast section, but also at the semantic level. The well-defined protagonist in the Horse Story is clearly introduced as agentive, whereas the second and the third charactersgain their agentivity only at the end of the story. In the Cat Story, in which three animals compete the status for of main protagonist in the narrative, the semantic roles are also less stable. These animals are introduced either dynamically, in sentence frames requiring Agents (AGT) and Patients (PAT; e.g., the cat is trying to grasp them, the dog is chasing the cat) or statively, in frames requiring Experiencers (EXP; e.g., the mother birdis in thenest with thechicks, the cat is sitting by the tree), as illustrated in Example (15).

Data from Batorko (in press)-adults urnpassaro nurn ninho corn osfilhotes que vai d procura da cornida. there is a bird in a nest with the children that o is goingto find the food EXP(bird)

0 gat0 olha para a drvore, v&o ninho, v& os passarinhos, p8e-se a observar

the cat looks at the tree, the nest, sees the birdies, becomes to observe 0 sees 0 0 cat) EXP(cat) EXP(cat) AGT(cat)
corneqa a trepar a drvore e tenta chegar a0 ninho onde estiio ospassarinhos.

starts to climb the tree 0and to get to the nest where are located the bird tries AGT(cat)
Entretanto chega o passaro rnde corn a cornida para osfilhotes e

in meanwhile comes the mother-bird with the food for the children and AGT(bird) there isa dog who0 see (in0 the cat0 him pulls the tail 0to takeoff the nest him EXP(dog) EXP(dog) PAT(cat) AGT(dog) PAT(cat) AGT(dog) PAT(cat)
ha urn ciio que ao ver o gat0 lhepuxa o rabopara o tirar do ninho.

0 gat0 foge e o ciio corre atras dele eo pdssaro-rniie

corn a cornida para osfilhos.

pousa junto aosfilhos

the cat flees and the dog runs after him and the mother-bird sits by the children PAT(cat) ) AGT(dog) AGT(cat) with the food the children for In the Cat Story presented Example (IS), the roles keepchanging throughout in the narration. Thus the cat comes up to the tree (Agent), then is sitting he


B A T O R ~ O6: F A R I A

(Experiencer) by it and watching (Agent) the birds,and after that he is climbing (Agent) the tree, being finally pulled down (Patient) and chased (Patient) by the dog. The dog, being the counterpart in this situation, turns up (Agent) in the second part of the story to protect (Agent) the nest from the cat and make (Agent) him (Patient) flee. In the HorseStory, in which the protagonist is well defined (Example 16), the roles alternate only at the end of the story, where Agents turn into Patients and vice versa: For instance, the running horse (Agent) falls down (Experiencer) and is helped (Patient) by his friends, while the cow spotted (Patient)in the meadow bandages (Agent) his leg. Note that up no significant systematic changes been observedby age. have

Data from Batoreo (in press)-adults

Hd um cavalinho que vai correrpelosprados

there is a horsie thatis goingto run across the meadows 0 EXP(horse) AGT(horse)
Chegauma a cerca
0 comes up

e do outro da lado cerca

v& urn touro

to a fence on the other side the fence 0 sees a bull/ox and of

umpasarinho. Elee salta cerca cai a


Em estd cima cerca da

on the top of the fence is located a birdie he jumps the fence and AGT(horse) EXP(bird) EXP(horse) E depois o touro ajudd-lo vai e opassarinho tambtm. andafterthebull/ox is going to helphimandthebirdiealso AGT(bird) PAT(horse) AGT(cow)


The same Ground can be assigned different roles in relation to different protagonists: While it i s an entity for one it can beplace for another. For a instance, thefence in the Horse Story play a different role for the horse can and for the bird (Example 16). the first itis an obstacle to be For jumped over and thus functionsas an entity, whereas for the second is a location.Similarly, it different conceptualizations can be observed in the of the tree in the Cat case Story (Example 15). Some narrators perspectivize it as an independententity, introducing the tree at the beginningof the story(e.g., There very was a tree, and there was a nest with little birdies it). The majority, though, conceptualin ize it as a place to locate the mother birdwith her babies. The strategies of the assignment of semantic to Figures and Grounds roles may be evaluated from the point of view the perspective that narrator is adopting in different parts of of the the story. Finally, asExamples (7), (g), (U), and show, only the verbs mo(14) of tion, but not the verbs of location, allow the omission of the Ground. Location verbs require the specification a place, whereas some typesof motion like the of one triggeredby the appearance-in-the-stageverbs (Example g ) do not. It is not easy to trace a clear-cut dependencebetween the semantic roles assigned to Figures and Grounds and spatial framing and anchoring on the

Representing Movement

in Portuguese


narrative level. Experiencer protagonists existential constructions and in Agents with verbs of movement seem tobe characteristic of spatial framing, whereas Agents, Patients, and Experiencers turn up in spatial anchoring throughout the story.

The Language Variable: Morphosyntactic Marking. Typical syntactic structures used in narratives have already been discussed in previous sections (e.g., see Example.10). Inthe present section we focus on morphosyntactic marking by discussing some aspectual distinctions transcategorically related to the expression of space in discourse. As was discussed earlier, overload of information can be avoided distributing it indifferent positions of initial utterances, by using variousstrategies to introduce a new referent. European Portuguese adult storytellers distribute equally their predilection between existential/locative constructions, on the one hand, and descriptive ones withmotion verbs, on the other (Fig. 2.3). As the examples in this chapter show, a new referent is introduced with the construction + Indefinite Subject rather than with the V canonical SVO word order that occurs lessthan 6 % of allthe adultnarratives. in A new referent can also be introduced a construction with a motion verb suc in as andar go (general verbof movement) or correr run, or passear walk (see Examples 7b and c and descriptive constructions in Fig. 2.3). Existential, locative, and action verbs are either in past imperfective (haviathere was, era was, estava waslocated, andava was going)or in present (imperfective in existential and locative verbs: h6 there is, estu is located). The two storiesin Examples (15) and (16) are coincidentally both told in present (imperfective only) and startwith an existential construction with huthere The imperfecis. tive past and present canalso be frequently interchanged the same story. The in same European Portuguese speaker can start his narration in one tense, continue in another, and then come back to the (Example 17). first one
(17) Data from Batorio (in press)-adults

uma Era

vez do seu ninho

umpassarinho que estava

em cirna
on the top


PAST Entretanto chega

there once a a was upontimebirdie of its nest

that was located

urn gato e o passarinho quando v& inthemeanwhilecomesacatandthebirdiewhenseesthecatflees

PRESENT OgatoFcaparado a olharpara o ninho

o gat0 foge

the cat stays still looking the nest at


[ ..] intervening text in Present Tense) . (all




Opassarinho que vinha corn a minhoca na boca chega ao ninho ... thebirdiewho came/was coming with the worm in themouth arrived is arriving at the nest

The use of the imperfective serves for the description of the background of the story against which the main episodes, expressed in perfective, are going to take place. In the Cat Story, the introduction of the mother bird in the nest is generally in the imperfective, setting the background, whereas the main protagonists, the cat and the dog, introduced inperfective forms (present are or past) of unaccusative verbsthat express change of state and do notrequire any specification of place (as, e.g., in came cat). The use of a generalmotion.verb a such as andargo in the imperfectivesettings (Examples 7band c) shows some peculiarities, as the past imperfective form here is always iterative, being translated as used to go. Thus, if we have the first sentence of Example (7b; andava urn cavalo a correr pelos prados) we understand that thehorse used to run. But in the case of the second sentenceof the same example (andavaurn cavalopelos prados) the andava form is polisemic: It communicates the iterativeness of the general movement (go) and can be translatedliterally as used to go or used to move (general movement), which in English probably corresponds rather toused to r n . Its complex nature results in relatively late u acquisition and is most typical of adult discourse. Typically the imperfective is used in spatial framing, giving informationon background, whereasin spatial anchoring both imperfective and perfective are used. As Example (14) shows, the specification of the place (with an adverbial or prepositional phrase) can be sentence initial or sentence final. Thisa is strategy to deal withthe excess ofpacked information on the sentence and level is not rare among languages. As has already been stated, the target adult grammar described in this section is not acquired in a linearly progressive way, in which linguistically and cognitively simpler constructions are always acquired before more complex ones. The acquisition the adult grammar discussed in the nextsection. of is
The Age Variable

A first global look at the importance ageseems to suggest that 6 to 7 of years is significant for changes at both thecognitive and the linguistic levels. At this age children start to reorganize their discourseand begin to use adultlike cognitive and linguistic structuring strategies. This is reflected in the use of existential/locative structures (Figs. 2.3 and 2.4). It takes children quite some time to use adultlike realizations such as the haverthere beconstructions, as children aged 7 to1 still 0 show a very clear preference strategies other than those for used byadults.

Representing Movement

in Portuguese


Our study shows that children have difficulties establishing and maintaining spatial frames according to the rules of the target grammar, represented here by the adult sample, until late, approximately 10. The interesting thing age is that children atage 5 actually have all the basic morphosyntactic andlexical information at their disposal. Even the marked constructions that violate the canonicalsyntactic order of an SVO language-and marked strongly for the introduction of information-are largely mastered at this That is, children age. can construct meaningful sentences built with the correct lexico-morpholog material (Examples 11-13); however, they have problems in using them in appropriate discourse contexts, our sample of narratives produced in the as absence of mutual knowledge shows. In other words, only some these syntacof tically correct and well-built utterances are used adequately at the discourse level. The strategies childrenuse to verify the contextual adequacy their of utterances arecontinuous and start to used when they can be really construct a successful narrative, which in our sample, is not before the age of 5;6. With the morphosyntactical devices at their disposal, children test different in uses different contexts and stick to those that the input shows as contextually adequate andsuccessful. All the others (as Examples 11,1z, and 13 show) are dropped out or retested in different contexts until they prove successful in one. This processof testing and retesting competing structures does not happen at a particular age but continues throughout acquisition after the age of5.

The results of the present research show the relevance ofthe three variables proposed previously and provide some answers to theinitially posed hypotheses and questions, leading to the following conclusions. In acquiring their native language European Portuguese children accomplishfollowing cognitive the and linguistic tasks: (a) They discoverthe discourse strategy of spatial anchoring, and especially setting of the spatial frame; @I)they learn to situate expressions of spatial anchoring inall the sentence positions used by adults, especially at the very beginning of the utterance; (c) they learn syntactically postpose subjects, to especially indefinite ones. These conclusions only partially confirm our initial hypothesis, according to which Portuguese speakers acquiring their language would behave line native in with the lexical and morphological cues model. In fact we have evidence supporting both models-the one based on lexical and morphological cues in morphologically rich languages (suchas Polish or Italian, see Hickmann, 1995; Smoczynska, 1992) and the onebased on word order (such as English)integrated in the process of language acquisition. These observations us make reformulate our initial hypothesis. Portuguese children select lexical items according to the type of event conflation their language represents, that is, fusing Motion and Path, and they choosetypes of cues-lexical, morphological, all




and syntactic-that are available, reliable, obligatory in this process. and This process is basically stable at age 5. What is really acquired after this is the age contextual appropriatenessof the linguistic structures, whereby context we mean the discourse situation, characterizable psycholinguistic and socioby linguistic parameters. At the linguistic level our discourse analysis shows that European Portuguese should not be classified as the same type as morphologically rich languages, predicted to be acquired by morpho-lexical cues. European Portuguese children showthat at an early period of development around theage of 5, they can use syntactically marked constructions such as postpositioned indefinite subjects for introducing new information in discourse, revealing the importance of syntactic cues as much as the initially hypothesised lexical and morphological ones. However, they do not use the marked constructions skillfully in appropriate contexts for certain discourse purposes in the same way as adults beforeage of 9 or IO.Thus, our results confirmour hypothesis at the discourse level, showing that structures acquired accordance with the in syntactic word-order model becomefully functional later than structures acquired in accordance with lexical-morphological model the that is claimed to be more valid for European Portuguese. Our results corroborate findings from recent crosslinguistic research concerning linguistic marking local and global levels for talking on about space. Recent research examines childrensuse of spatial devices in discourse across languages, showing that typological differences such as those suggested by Talmy affect what spatial information is focused upon and how the flow of information in discourse organised both on local and on the is the global level (Hickmann, 1995,p. 210). We propose that further research should be devoted to a thorough understanding of theinterrelationsbetweenaspectsof childrens cognitive development and developments in discourse strategies, well as acquisition of as specific linguistic phenomena, especially those typologically determined. By general cognitive aspects,we mean spatial and temporal conceptualization as well as scripts and narrative schemata. Developments of discourse strategies include the acquisition of global and local principles guiding discourse organization. Acquisitionof specific language phenomena concern, in particular, the interdependence among different linguistic expressions space in relation to of categories such tense, aspect, possession, as existence, and nonexistence.

This research is part of project PRAXIS =I/BD/5260/95, financed by Junta Nacional de InvestigaqHo Cientifica e Tecnologica.The authorswould like to acknowledge Professor Maya Hickmann and Cambridge University Pressfor authorization to publish the Horse and Cat Story pictures.

Representing MovementPortuguese in


Batorio, H. J. (1995). Spatial expression in childrens narratives: A study in European Portuguese. In I. H. Faria & M. J. Freitas (Eds.), Studies on the acquisition ofPortuguese (pp.191-206). APL: Edipes Colibri. Batorio, H. J. (1996). Spatial relations in European Portuguese childrens narratives. In S. Contento (Ed.), Psycholinguistics as a multidisciplinary connected science, Vol. I1 (pp.zz5--230). Proceedings o f the Fourth ISAPL International Congress, Societl Editrice Ponte Vecchio, I1 Bologna, Italy. Batorio, H. J. (1998a). Acquisition o f spatial expression in European Portuguese childrens narratives. Polish PsychologicalBulletin, z(I), 47-57. Batoreo, H. J. (1998b, May). Language typology and semantic primitive of space: Evidencefrom European Portuguese. Paper presented at Primeiro Encontro de Linguistica Cognitiva. Porto, Portugal. Batoreo, H. J. (in press). Contribuipio para a caracterizapio da interface expressdo linguisticaCognipo espacial no Portuguts Europeu. Abordagem Psicolinguistica da Expressdo do Espa~o em Narrativas Provocadas [Toward the characterization o f the interface between linguistic expression and spatial cognition in European Portuguese: A psycholinguistic approach to the expression o f space in elicited narrative discourse]. Fundapo Calouste Gulbenkian. Batorio, H. J, & Duarte, I. (1998, September). Aberturas de narrativas e primitivos semtlnticos de . posse [Narrative setting constructions and semantic prime o f possession]. Paper presented at Encontro Nacional da Associapio Portuguesa de Linguistica, Aveiro, Portugal. Berman, R., & Slobin, D. I. (1994). Different ways ofrelating events in narrative: A crosslinguisticdevelopmental study. Hillsdale, NJ: Lawrence Erlbaum Associates. Bokus, B. (1996). Narrative space structuring at the preschool age: Findings on monologic and dialogic discourse. In C. E. Johnson & J. H. V. Gilbert (Eds.), Childrens Language, Vol. g (pp. 197-20 8). Mahwah, NJ: Lawrence Erlbaum Associates. Bowerman, M. (1985). What shapes childrens grammars? In D. I. Slobin (Ed.), The crosslinguistic study oflanguage acquisition, Vol.2: Theorethical issues (pp. 1257-1319). Hillsdale, NJ: Lawrence Erlbaum Associates. Bowerman, M. (1989). Learning a semantic system: What role do cognitive predispositions play? In M. L. Rice & R . L.Schiefelbusch (Eds.), The teachability oflanguage (pp. 329-363). Baltimore: Paul H. Brooks. Bowerman, M. (1996). Learning how to structure space for language: A crosslinguistic perspective. In P. Bloom, M. A. Peterson, L. Nadel, & M. F. Garret (Eds.), Language and space (pp. 385-436). London: MIT Press. Chavegatto, V. C., Souza, R. M. N., Silveira, M,, Silva, E., & Cassano, L. B. (in press). Figuraqties na representapo do espaGo e do tempo em descripes de cenirios em portuguCs [Representation o f space and time in scene description in Portuguese]. Actus do Encontros do GTDescripo do Portuguts, Rio de Janeiro. Choi, S., & Bowerman, M. (1991). Learning to express motion events in English and Korean: The influence o f language-specific lexicahation patterns. Cognition, 41,83-121. Faria, I. H., & Batorio, H. J (Eds.). (19 9 4). CHILDES: Uma Adaptqdo para o Portuguis Europeu (1f . versdo) [Adaptation proposal of CHILDES to European Portuguese, 1st version]. Unpublished manuscript. Faria, I. H., & Duarte, I. (1989). 0 paradox0 davariapo: Aspectos do Portugues Europeu [Paradox o f 21-27. variation: Aspects of European Portuguese]. Revista Internacional de Lingua Portuguesa,~, Goddard, C. (1998). Universal semantic primes of space-a lost cause? In R. Driven (Ed.), Humboldt and WhorfRevisited.Amsterdam: Benjamins. Guimarties, A. M. (1994). Desenvolvimento da linguagem da crianqa na fase de letramento art [Language development in children in the phase o f literacy]. Cadrnos de Estudos Lingiitsticos,26,


B A T O R ~ O& F A R I A

Hendriks, H. (1993). Motion and location in children) narrative discourse: A developmental study of ChineseandDutch. Unpublished doctoral dissertation, Rijksuniversiteit of Leiden. The Netherlands. Hendriks, H., Hickmann, M. (1998). Reference spatiale et cohesiondiscours: Acquisition la & du de langue par Ienfant et par ladulte [Spatial reference discourse cohesion: Acqusition lanand of guage by children and adults]. In M. Pujol-Berche, L. Nussbaum, & M. Llobera (Eds.), Adquisicidn de lenguas extranjeras: Perspectivus actuales en Europa (pp. 150-161). Spain: C. I. D. Linea Metodologica de Edelsa. Hickmann, M. (1982). The development ofnarrative skills: Pragmatic and metapragmatic aspects of discourse cohesion.Unpublished doctoral dissertation. University Chicago. of Hickmann, M. (1991). The developmentof discourse cohesion: Some functional and crosslinguistic issues.In G. Pieraut-Le Bonniec & M. Dolitsky (Eds.), Language bases, discourse bases (pp. 157-185). Amsterdam: Benjamins. Hickmann, M. (1995). Discourse organization and the development reference to person, space of and time. In Fletcher & B. MacWhinney(Eds.), The handbook ofchild language(pp. 194-218). P. Cambridge, MA: Basil Blackwell. Johnston, J., & Slobin, D. I. (1977) The development of locative expressions in English, Italian, Serbo-Croatian andTurkish. Journal ofchild Language, 6,529-545. MacWhinney,B. (199 4). The CHILDES Project: Toolsforanalysing talk.Pittsburgh: Carnegie Mellon University. Mayer, M. (1969).Frog, whereareyou?NewYork Dial Press. Slobin, D. I. (1973). Cognitive prerequisitesfor the development of grammar. In C. Ferguson & D. I. Slobin (Eds.),Studies ofchild language developrnent(pp.175-zl1). New York Halt, Rinehart & Winston. Slobin, D. I. (1989). Factors of language typologyinthecrosslinguisticstudy of acquisition. Unpublished manuscript, University California atBerkeley. of Smoczynska, M. (1992). Developing narrative skills: Learning introduce referents in to Polish. Polish Psychological Bulletin, 23(2),103-120. Talmy, L. (1975). Semanticsand syntax of motion. In J. P. Kimball (Ed.), Syntax and semantics 4 (pp.181-238). New York AcademicPress. Pick Talmy, L. (1983). How language structures space. InL.H. & L.P. Acredolo (Eds.), Spatial orientation: theory, researchand application (pp. 225-282). New York Plenum Press. Talmy, L. (1985). Lexicalization patterns: Semantic structure inlexical forms. In T. Shopen, S. Anderson, T.Giv6n, E.Keenan, & S. Thompson (Eds.), Language typology and syntacticfield work, Vol. 3. (pp. 57-149). New York Cambridge University Press. Wlerzbicka,A.(1996). Semantics, primes and universals. Oxford, England: Oxford University Press. Wlerzbicka, A. (19 9 8). Anchoring linguistic typology in universal semantic primes. Unpublished manuscript, Australian National University at Canberra, Australia.


The Cat and Horse Picture Stories (Hickmann, 1982; published by permission of Cambridge University Press).

Representing MovementPortuguese in


1 3

This Page Intentionally Left Blank

Why Young American EnglishAnger and Sadness: Speaking Children Confuse

A Study o f Grammar in Practice

Clark University


The findings in this chapterwere not anticipated when I began this line of research afew years ago. Originally, I did not intend to investigate and compare childrens accounts ofemotion situations,in particular their accountsof anger and sadness situations. Rather,began with the aim of investigating how chilI dren orient themselves toward-so to speak-the same situation,but from different genre perspectives. Morespecifically, I was interested in childrens accounts of situations in which they usedthe personal pronounI (in order to refer to pastevents, or personal experiences) contrast to accounts events in in of which athird person (she or he) went through thesame experience. In addition, I compared those two genres to one in which a generalized personor the (one, generalized you) acted or was acted upon. In short, original investigation my aimed at a genre comparison (a) personal narrative, (b) third-person narraof tive, and (c) explanatory discourse. The idea to employ emotion situations su as being angryor being sad came up the attempt to find a situation that in was ecologically meaningful for both younger and older children,and was of the same kind,so I could compare the linguistic devices used accordingto the age ofthe childrenand according to the genre was targeted by the children. that This in mind we asked 8 o (American English-speaking) children ranging from preschool to third grade (ages 4-10) to tell u s about o n e time whenyou were angry/sad/scared/happy-prompting for the account a personal experiof ence; in addition,we asked them to imagine little boy or girl and togive us an a account for one time, when o r he was angry/sad/scared/happy, she and last, we simply asked them to explain, whatit means to be angry/sad/scared/happy.This resulted in 12 interview questions, which were randomly assigned, all revolving around different perspectives on thoseparticular emotional situations. In the course of interviewing the children,we stumbled across something that would reverse the weconceptualized childrens abilities to use linguistic way



constructions for discourse purposes. repeatedly heard some the children We of maintaining that oneor the other question had already been answered, which initiallydid not surprise because the battery questions could us, of easily confuse interviewees, particulary if they did not fully concentrate. However, when scrutinizing the data more thoroughly, we realized thatthesekindsof confusions occurred solely when we had asked to give anger or sadness accounts, but never with anyo f the others. When asked children in these situations we to give us an answer anyway, their accounts were most often word-for-word repetitions. In addition, we noticed that these confusionswere more typical for the youngerthan for the olderchildren. This preliminary evidence seemed to hold some water, although it is not in agreement with what one would expect based research reportsof childrens on emotion knowledge(cf. Stein & Trabasso, 1992;Stein, Trabasso, & Liwag, 1992). According to those reports, children young as age 3 are ableto perform at a as high level of proficiency in figuringout thedifferent components thatlead to emotional states such angry and sad and whattypically follows from them. as The only encouragement to probe deeper into this observation came from an thropological reportsabout a number ofAfrican languages, which, least at the at lexical level,do notseem to differentiate between what is divided accordingto the English lexicon anger and sadness (cf. Davitz, 19 6 9 ; Heelas, 198 6 ; Leff, into 1973; Matsumoto, 1994).Thus, this original accidental stumbling across some childrens confusions of angry and sad launched into a us closer look athow the accounts of sadness and angerwere linguistically constructed by younger (and o1der)American English-speaking children, and what these accounts actuwere ally used for when it came to a comparison between different genres. the To clarify how young children actually come toconfuse two so-called basic emotion concepts-at least in the genre of narrative accounts (although this genre is highly relevant forself accounts and identitypresentations)-I first discuss some general tenets of the relationship between language, thought and emotion, and their relationship for developmental studies. Then, I show in more detail howmy study of emotion talk led to the differentiation two of different grammars, that of anger and that of sadness, and to how young childrens confusion between these two grammars can accounted for. In my be concluding section, I take the relationship between narrating and emotion up talk with the somewhat radical argument thattalk is more foundational than traditionally credited, not only for the we make sense of emotions, but way also-at least to adegree-for how we actually feel.

Generally speaking, talk about emotions, that talk in which emotions are is, thematized, seems to imply that emotions are objects or entities that have an

Grammar in Practice


existence outside talk and apart from language of in general. In this, they are very similar to ourfolk notions of events that seemto have their existence outside of talk, but can be referred to-just like emotions-in and through talk. Events and emotions could have taken place inthe pastor they can beimagined; they can be of a personal nature, that is, the teller could be centrally (or peripherally) involved,or they canbe ofa completely detached, impersonal nature, where the teller is not simultaneously thematizing himself herself, as in accounts of or emotions ofothers, past or imagined, or as inexplanations, definitions, or other more detached situations such incard sorting as tasks (cf. Lutz, SS). 19 The question that immediately comes to mind, however, is how we know what emotions are and what they mean, and more specifically, how children learn the meanings? In order to answer this question, may be thrown back we onto language and emotiontalk as the sources and possibly even resources that tell us what we know about emotions and theyare dealt with in the social, how communicative realm. A way to avoid the issue ofdealing with language and emotiontalk as somewhat foundational to our understanding of emotion, would be by way of borrowing from a theory natural perception. In this theory, emotions are of not really learned. They are bodily experiences that are directly sensed and differentiated into alimited number of emotion categories. What is learned are the language-appropriate labels for these categories. And although much our of everyday talk about emotions and feelings seems torely on this theory, anyone who has struggled with a foreign language knows the emotioncategories that we learned with our first language are not the sameas in any other language: Natural perception cannot automatically read from bodily sensations the off categories that are considered meaningful for the speakers of particular languages. Thus, seem to be thrown back onto we language as one of the sensegiving foundations whenit comes to emotioncategories. A second route to avoid taking in any way as a foundational factor for talk the constitution of emotions as meaningful entities, although by far more sophisticated than the theory natural perception, relies on the intuition that of all humans have emotions, and that the particular languagethat we learn as our first just carves up the emotion spectrum differently from anyother language, leaving us with the impression that our(first) language doesit somewhat more naturally, whereas other languages are somewhat derived. This theory is actually quite similar to the one developed for color categories, has been proposed in its most sophisticated version by Wlerzbicka (1992,1994,1995) and, morerecently, also Goddard (1997). The basic tenets on which this theory rests are cognitive universals. In short, resting on the assumption that human cognition (the mind) differentiate between the different emotion categories can and translate emotion terms fromlanguage into another useof a limited one by set of (cognitive-semantic) universals, the foundational capacity for making sense is attributed to cognition, not to language.



Asimilartheoreticaladvanceontherelationshipbetweenemotion, language, and cognition, although not from a cognitive-semantically universalist point of view, has been madeby Stein and her associates (Stein & Levine, 19 90; Stein & Trabasso, 19 92; Trabasso & Stein, 1993). In her approach, emotions areviewed as tied into relationsbetween people, although they are approached as a representational system of the goal-plan-outcome knowledge that is held to regulate and coordinate the relations between people. Thus, knowledge of goals and plans is assumed to form the core prerequisite for making sense of others, and it figures foundationally in explaining and accounting for ones own actions, that is, inthe process in making senseones of own self. According to Stein and her associates, this type of knowledge is acquired relatively early,at aroundage 3. Atthis point,children are assumed to be apt to successfully differentiate between the components of actions goal and plans that lead to (English) anger, sadness, fear, or happiness (Stein, Liwag, & Wade, 19 9 7). In contrast to Wierzbicka and her colleagues, who view emotions as a semantic domainthat governs the patterns discourse, Stein uses narratives of of real life emotion situations and subjects them to online questions online for reasoning. However, similar to Wierzbicka, Stein and her colleagues use discourse data to analyze language in its ideational, representative function, that is, asa moreor less transparent window into the conceptual underpinnings of what their talk is about. The aboutness of talk (or what is behind the talk) is taken as basic, irrespective whether the speaker wants to be understood as blaming someone else or saving face. That is, the directive force of language (the interpersonal function)is not considered to beof immediate relevance to the meaningof the emotion account, nor to the meaning situatedness of of the the emotion, nor to an emotiongeneral. Thus, what the work Wierzbicka in of and Stein share is a theoretical proceeding from the abstract to the concrete: The meaning emotion is a foundational concern its application in situated of for expressions or displays, and those foundational for situated verbal accounts. are Although there has beenan abundance of theorizing during the1990s on the relationship between language, cognition, and emotion, most itof nevertheless has centered on the more narrow relationship between (emotion) concepts (emotion) and words. although And ethnomethodological approaches to emotion talk in other cultures/languages (Basso, 1992; Lutz, 1988; Ochs, 1988,1996; Schieffelin, 1990; Schieffelin & Ochs, 1986) have repeatedly underscored the situatedness and cultural contextualization of emotion talk for theway emotions make sense and can enter as meaningful entities the interactionsof participants, their contributions nevertheless have mostly gone unheard they have been misconstruedas dealing predominantly or with concepts and words. Developmental studies (with only exceptions) few have predominantly targeted emotion concepts and, in this sense, are very much in line with mainstream developmental attempts to contribute to a

Grammar in Practice


debate regarding by what age children have emotion concepts (or atleast the basic ones), have a theory of mind, the basic narrative components, other or psychologicalobjects. It is also interesting to note that not muchthis debate of has been incorporated into standard language acquisition volumes, ranging from general psycholinguistic introductions to the acquisition the lexicon of (see, e.g., Clark, 1993; Fletcher & MacWhinney, 1995; Gleason Ratner, 1993; & Hoff-Ginsberg, 1997), which may be due to the that the groundsof what fact actually develops and what it is that facilitates development, are rather murky: Is it concepts that develop and at one point anotherbecome mapped onto or the appropriatelinguistic forms, or is it linguistic forms that develop, channeling thoughts and cognitions toward socially appropriate ways of making sense? And further,is making sense primarily a more reflective, conceptual activity, or can it be also described in morepractical, participatory terms? In this chapter, the attempt is made to break out the cycle of describing of language use as principally based on cognitive terms or conceptual entities. Starting from the assumption that emotion displays deeply embedded in are our humanway of displaying ourselvesas situated selves in situations with others, we do notdeny that emotion displays havephysiological (bodily caused reactions) correlations, or that situationscan be conceptually structured and talked about. However, in order to determine how emotion displays gain their meaning as meaningful events, we cannot solely rely on physical reactions to stimuli or toconceptual structures in themind of private individuals as foundations. Rather, bodyand mind occupy (jointly) a social space in unfolding episodes, communicating the (relational) position a self vis-&-vis others. of This is where joy, anger, shame, surprise, and thelike materializeas meaningful positions taken up by a person purposely. Thus, these positions are displayed as actions that are purposely taken up to signal and signify a selfother relationship. In this way, wecan study the kindsjudgments, aesthetics, of morality, and prudence that are expressed in emotion displays. We can determine what interactionally has led up to an emotion display, and what the display has accomplished. In this sense then, emotions like conversations are (Hard & Gillett, 1994), andaccounts of emotion situations typically work up the aesthetics, judgments, and morality involved in such situations. In sum, the present chapter does not take narrative accounts of emotion situations as windows into some (underlying) conceptual (mind) physical or (body) foundations of human meaning making, but as windows to the positionings that are being performed in the formnarrative actions. In these of positionings narrators provide the audience with an order that they can so convince, blame, save face,that is, practically orient the audience to an order or within which judgments,aesthetics, and morality are purposely arranged. And
1 And although we also have emotions as private individuals, with oneelse around, self and no other remain the unit around which selfhood and otherness constructed. are



since these narrativesare constructed for an audience, the window metaphor should have been replaced by the better imageof a signpost or pointer: The actual performanceof the narrating act orients the audience to attend to the order constructed. And because the researcher participant in this construcis tion processas interviewer, the construed ordermay also be considered as part of the research situation itself. It should be noted that this account of the relationship between language, emotion, and other, and mind and body self is centrally dialogical. Inasmuch as language is alwaysan embodiedact and always centrally dialogical, it orients selfhood and otherness to one another in a foundational way (see Bamberg, 1999, for further discussion on the centrality of language). It should also be noted that the approach presented in this chapter bears heavily on the notion development: In contrast to mapping changes over of out time of childrens uses of words (semantic structures) or childrens linguistic applications of conceptual structures, and claiming that is what develops, this we see the issue of development much moreclosely tied to the issue of participating in (linguistic) practices. And because these practices are no longer conceived of as structures that have their existence apart from the person (or within the person as internalized or matured mental structures), but rather as embodied discourse activities, the trajectoryof language development (here, emotion talk) is no longer constrained to a single domain, such as lexical development, but closely interwoven with the development of the person a as whole. We return tothis issue in theconcluding section.

In what follows I will extrapolate the linguistic devices that are typically employed in the construction two typesof situations, being angry and being of sad. I do not detail the findings for each single age group, but for contrasting purposes compare the older children (the thirdgraders, mean age g;1) with the younger age groups (preschoolers, mean 5 2 collapsed with kindergartners, age ; , mean age 6;i). In addition, central to our discussion is the genre of lived experience (past-tense, first-person narratives), but we also briefly consider the explanatory genre for comparative purposes, neglecting here a more detailed discussion of the third-person genre (see Ozgaliqkan,1997, for a report of the genre findings forall four emotions). The term grammar for the characterization of form:function relationships has intentionally been chosen, the one on band to index an affinity toWlttgensteins use of the term grammar in his theory of language games (Wittgenstein, 1953), and on the other to take a critical position vis-&-visthe use of the term grammar as a system outsideand prior to the person and his or her use of forms for practical purposes (functions). Thelinguistic forms thatare taken to constitute a grammar a for

Grammar in Practice


particular population (here age group of children) are purely descriptive listings of formal devices, that is, they arenot meant to be in any way exclusive nor exhaustive; rather, they taken as (formal) indicesof functional orientaare tions to direct the hearer or audience to attend to a particular discursive position (see Bamberg, 1991,1997; Talbot & Bamberg, 1996). I first list the devices constituting thegrammars of anger and sadness for First-Person the two G e n r e and then briefly compare them with the grammars for the same emotions in the Explanatory Genre. The Grammar ofAngerin the First-Person Genre The linguistic devices employed by older younger children, that is, all and by age groups, to construct angry situations in which they were (made) angry typically consist of:
(i) a highly individuated agent (my sister-see Example I, and a highly ) individuated undergoer (me); (ii) a marking of the action as highly transitive; (iii) a positioningof the Ias the recipient and target the action in the of direct object slot; and (iv) a positioning of the other (the agent) in subject slot.

These four features apply consistently to all of the verbal accounts, and Examples (1) and (2) may serve to illustrate how angeris constructed in terms of these four linguistic construction types:
I was in the room
and my sister kicked me and it went right into the rib bone when

my sister slapped me across the face

just because she didnt letme in her room and I wanted to play a game but she didnt let me and slapped me across the face

In terms of the discursive purposes for which these lexicosyntacticdevices are employed, we can tentatively draw up two general orientations: On the one hand, the constructionof a highly individuated targetof others actions may orient the audience toward empathy orsympathy, particularlyif the action is not sufficiently motivated or justified. On the other hand, introducing o t h e r the as the topical focus in the position the syntactic subject opens her him to of or become subjected to blame, again, particularly whenthe action was unmotivated or unjust. In the anger accountsof children across all age ranges, this topical focus on the perpetrator (for the purpose of attributing blame) overshadows, so to speak, the discursive purpose of eliciting empathy for the



victim. Or, in other words, the construction anger in American English of consists (developmentally from very early on) of two discursive purposes: blaming and eliciting empathy, with the latter subordinated to the former.
The Grammar o f Sadness in the First-PersonGenre

Typical of the accounts of older American children (the third gradersin our study, who are very much in agreement with theway adults construct sadness accounts) aretwo different construction types:
(i) positioning the other in subject position, as in Example (3), o r (ii) positioning the I in subject slot, as in Example (4):
( 3 ) it was when I was about5 or 4 years old

my biggest sister got into a accident car

so she died
because of a car accident and I was really sad for a few weeks (4) I was in Charlton and I moved to Worcester and I couldnt see my neighbors and their dogs

Whereas construction type (i) holds up the possibility to make other (here the my sister) the potential topicalfocus, and as such orients the discourse activity in its purpose toward blaming, this option is ruled out by two additional linguistic devices:
(a) the avoidance and downplay o f marking the other as agentive, and

of (b) the absence (bychoice o f predicate-type/Aktionsart) a target of the

activity referred to(dyingis atelic).

These two devices are similarly employed construction type (ii) Example in (see 4)) denying the I to achieve the status of a topical focus, which-in case the I really becomes the topic, withsome potential foragentivity-would open the door to a possible interpretation that the narratori s signaling that it had been his faultand thathe was intending to blame himself. Thus, it can be maintained that the grammar typically employed for the construction of sadness differs from the grammar of anger in degree of complexity: Whereas anger consistsof one (formal) construction type, but comprises the two discursive orientations empathy and blame as such (and requires a delicate balance between these two orientations), sadness is less complex in termsof discursive purposes, because it geared toward only is one discourse purpose, namely empathy, more complex in termsthe existence but of of two constructive options. In addition, taking the prototypical English construction type of the transitive scene, theconstruction types employed to

Grammar in Practice


orient toward eliciting empathy also can be characterized as more complex, as they are deviations from the prototype, because they require an additional downplay of the topical focus. In other words, after the subject (whichtypically is the topical focus) has been established, this focus has to be defocused subsequent clauses. The clause because of a car accident (in Example 3) illustrates this function, removing my sister from potentially becoming the topical focus, implying that she didnt really do anything; thisis not really about heras an agent inthe depicted event. Thus, although the construction of anger-in English-is more complex whenit comes toits discursive purposes, the construction of sadness is more complex in terms its actual linguistic of construction types. Turningnexttothedescription of how sadness was linguistically constructed in the first-person genre the younger children(the preschoolers by and kindergartners in study), we find their accounts structurally equivalen our with the anger accounts the American English-speaking subjects ofall age of groups (including their own). Examples and (6) illustrate this point: (5)
( 5 ) when Nikki hit me in the eye I was really really sad I cried for a whole halfan hour ( 6 ) my Mommy hit me she hit me in the eye and I was sad and cried

These accountstypically consist of two components: Thefirst part topicalizes the perpetrator by constructing ahighly transitiveevent, which is likely to be taken to orient the audience toward an attribution of blame to the agent. However, in the second part, the topicshifts from the other to the I, orienting the audience toward empathy the discursive purpose the twocomponents as of as a whole. However, the construction of the happening that can be held responsible for the emergence of sadness very much like that of an anger is scenario for the youngerchildren.
The Grammar ofAnger in the ExplanatoryGenre

The constructionof anger in explanatory discourses achieved by fivedifferent is construction types that mostly run in concert:

(i) an unspecified agent in subject slot, most often plural they (ii) the unspecified target of the activity described in direct object slot, most typicallyyou (where it remains unclear whether refers to an unyou specified hypothetical personor to the interviewer); (iii) an active verb which nevertheless is much less specific therefore less (and transitive) when compared with verbs used in the first-person genre (e.g.



doing something or hurting here in the explanatorygenre, versus hitting and kicking in thefirst person genre); (iv) the clause modus is most often marked by ifor when; in conjunction with (v) the present tense, taking the situation into the realm of the timeless and possible world.

Examples (7) and (8) illustrate these types:

( 7 ) ifsomeone hurts you and you getreally really really mad then you are angry ( 8 ) you are angry at someone because they did somethingto you and you didnt like it what they did

In more general terms, the construction types in the overall construct result of an anger scenario that is much less of a bounded event, less vivid, and presented from a much more detached perspective anger was constructed than in the first-personperspective. Seeking empathy from the audience blaming or the other for any transgression clearly do not matter. If there is a particular discursive orientation, it lies in describing or making what usually occurs explicit, though clearly from a detached vantage point. The audience led out is into a world of usual occurrences, distanced from the realmof the special occurrence of subjective experience that made Examples (1) through (6) tellable narratives. As already mentioned, the five construction types (i)-(v) employed for the depiction or explanation anger situations in general used in concert of are by the older children, that is, all of them materialize together. Younger children have difficulties in using all fivesimultaneously,and they also often slip after having given an account employing the markers for the explanatory typical genre intothe first-person genre, telling how this once happened to them (past tense, plus devices typical for the grammar being angry). of However, across the board, they all are able to employ at least a few of the earlier mentioned five construction types. In spite of these shortcomings terms of particular formal in devices typical forthe explanatory genre, younger childrens general competence to give anger explanations nevertheless relatively wellestablished. is
The Grammar of Sadness in the Explanatory Genre

Similar to the accounts given in the first-person genre, a sad situation is constructed by our oldersubjects as well as by the younger children terms of in either s o m e t h i n g [bad] is h a p p e n i n g t o you (as in Example g>, or in terms of you want something, but you cant have it (Example lo). Agentive others, who

Grammar in Practice


could be held responsible, veryrarely figure in these accounts,and if they do, they are always defocused as potential targets of blame. Example(9) illustrates the something bad is happening to you scenario by transforming in line4 the potential agents into an impotent a mere happening that you helpless it of leaves (most likely, because thereis no target for revenge):
( 9 ) when like someone calls you four-eyed

ifyou have glasses and you get not mad at them but ithurts your feelings and youre sad
(IO) like your favorite blanket was up high

where you couldntget it

The only difference between the younger and the older children in their construction of sad scenarios in the explanatory genre was that the younger children at times seemedto consider a descriptionof the behavioral display of being sad a sufficient explanation of what it means tobe sad as in crying, or if somebody cried. Thus, in spite some difficulties in sorting out the of linguistic complexities of the explanatory genre at an early age, children of all age groups clearly demarcated what it means to be angry from what it means to be sad. They construed the anger scenario a bounded unit as (with aclear beginning and an end, where the actionin the middle led to the end), which generalized and was presented from a detached discourse (purpose) orientation; construing the and latter as a nonagentive happening, entailing no telicity, and consequently no particular other as a target to be blamed. In both sadness and anger explanations, the discursive orientationsattributing blame or eliciting empaof thy were backgrounded, whereas the discursive orientation to describe gener occurrences from a detached perspective became essential interactive goal. the

Why Do Young English-Speaking Children Confuse Anger and Sadness?

Summariziig the insights gained from the earlier data, are now better we equipped for discussing the origins for childrens early confusions between angry andsad scenariosand for delineating someof the factors involved in the developmental process between preschool and the time age children reach third grade, that is, between the ages of 5 and 9 years. First, the evidence assembled clearly points toward the early constructions of sadness accounts in the first person genre as theissue for what we called confusion. In the attempt to determine how these accounts differ from developmentally later sadness accounts, and also how they differ from accounts of anger experiences(across



all age groups), we realized that anger accounts typically consist of the construction of a highly agentive other who is introduced as the discourse topic. This construction is in direct service of the discursive act of attributing blame. Because in English this discursive strategy is achieved by use ofthe prototypical sentence format that endorses the transitivityscene (Budwig, 1995; Hopper & Thompson, 1980), the construction of first-person anger scenarios is grammatically relatively easy. Sadness scenarios, however, require a deviation from this more prototypical syntactic format: If another person has been introduced in subject slot, and therefore is likelyto be taken to be the topic of the account, the narrator to deemphasize this persons agency in order to has avoid the invocation of blame. He or she needs to reorient the listener to an empathetic stance towards the person who gives the account.It is exactly this problem of reorientation which younger English-speaking children face in their accounts of their ownsadness experiences. Considering that thegeneralized person perspective is linguistically more complex than thefirst person perspective, and that the younger of the children struggled considerably in coming to grips with the timeless as-if modality of this genre, it should come a surprise that anger and sadness accounts are as relatively clearly differentiated in the explanatory discourse genre along the dimension of transitivity-agency. Thus, theconfusion in the younger children between being angryand being sadcannot be traced the general unreadiness to of linguistically presenting what has conceptually already been mapped out. Rather, the early underdifferentiation between the two types of accounts lies clearly rooted in the pragmatics of emotion talk, more specifically, in the inability of not clearly differentiating between how to mark the respective discourse purposes of attributing blame versus eliciting empathy. And we would like to maintain that this pragmatic inability is responsible for our early younger subjects responding to the interview questions a confused manner. in With regard to what itis that developmentally pulls the child out of this state of underdifferentiation toward a higher level of differentiation (and as such also to a higher level of integration), we have no hardevidence to say for sure. However, the way we were able to map out the developmental route from a clear state of underdifferentiation toa higher level of making sense of sadness. That is, as a process of appropriating the toolsnecessary to talk meaningfully about the social relationships in which emotions are embedded, points up some highly important underpinnings. First, itshows that modeling emotional development in terms of an internalizationprocess of learning how feel may to not be sufficient. Further, we were ableto draw out the limitationsmodeling of emotional development in strictly cognitive terms. shown by this study, As grammar, if understood correctly-that is, not as abstract principlesof a universalist nature, butas social know-how relevant for the constructionof social meaning to participate in conversational practices-plays an integral role in coming to grips with what emotions and what they are for in social do used

Grammar in Practice


communicative practices.* A such, learning touse the linguistic construction s procedures for socially appropriate purposesis part and parcel of our cultural practices. To view this process in terms of appropriation rather than internalization gives space for the dialectics involved in the developmental process in general: Ontheonehand,thegrammatical means-so-to-speak-are preformed. Theyhave their social existence before they put touse in social are practices. It is for exactly this reason that children (or others) can come use to these tools inappropriately such in their early sadness constructions in the as first person genre. However, these so-called tools are not predetermining and imposing theiruse apart and independent from their users. The child early on is practicing in a relatively autonomous way with these tools, assembling new construction parts with others that are already successfully in place. Thus, viewing this processof appropriating linguistic constructions in the determination of emotion meanings as an integral part of 1earningthe language adds an extremely relevant component to emotional development, probably one that is much more central than were able to imagine we thus far. In addition, and here admittedly enter more speculative we territory, having documented that thedifferentiation between anger and sadness accountstook place developmentally prior in the generalized-person, explanatory genre, before it could be appropriated in the first-person, past-experience narrative genre, one mightexpect some learning effects spilling over from practices in doing talk for being descriptive to doing for more involved, interpersonal talk purposes suchas blaming: saving face: or seeking empathy. This, however, should not be misunderstood meaning to imply that those latter purposes as are learned in more detached speech genresfirst. Not at all. But in caseswhere the linguistic procedures relevant for the construction processes highly of involved speech genres constitute a particular problem (such in the case of as constructing complex sad scenarios earlier), practices in more detached as speech genres might enable speakers to sort out form:function relationships and reintegrate them at a higher level of integration in more involved speech genres. Taking up on the findings by Stein and her associates that were discussed in the opening of this chapter, namely that children younger than the5- and 6-year-olds in our study were perfectly able to differentiate between the different components thatdistinguish (English) anger, sadness, fear, and happiness, we are nowin a better position to reconsider seeming contradiction to this our own findings, and tie it closer to the concerns of methodology and language development. Although one of the important differences between the two
z This is not meant to imply that emotions cannot occur outsidecommunicative settings, of that is, so-to-speak privately. But the private experience of emotions is by no means their sole and primary aspect. As Wittgenstein (1953)was able to convincingly demonstrate,if that were the case, we could not only not talk about them, we wouldntknow about themeither. but



studies under consideration is the age of the children, another considerationis the way the datawere elicited. Apart from these two aspects, however, there is a third issue, which concerns the question which aspects of performance we of take to represent relevant developmental strides. Note that the online interview technique, used by Stein and her associates, is traditionally employed as a cognitive procedureto test comprehension. For this purpose it is legitimate to interrupt the naturalconversational flow with questions that probechildrens real understanding. The discourse mode that created in this type of is interview resembles the way caregivers and children interact in a topic-elaborative style that is quite common in our culture, where the caregivers build bridges to test and teach knowledge. This type knowledge, though not necessarily of of an abstract nature, nevertheless accessed in a much more detached, quasiis descriptive, explanatory mode. In contrast to type of discoursemode, we this found in our own investigation the early confusionanger and sadness of scenarios to be grounded in the involved discourse mode, where itwas the primary goal to grammaticize the discursive force of thetwodifferent emotions. Thus,we do not see the findings of Stein and her associates contradicting ours. Rather, they complement our own findings in the that a sense more detached discourse mode in both investigations was found to facilitate a clear differentiation of what is underdifferentiated only in the involved first person genre. Thus, it seems important to note that while the marking of attributing blameand eliciting empathy does not seem to play a major role in the explanatory discourse, thereby facilitating an early mastery of differentiating between anger and sadness, it is the complexity of the linguistic construction types necessary for the differentiation betweenattributing blame and eliciting empathy that ultimately is in the way of an (equally) early mastery in the more involved, first-person discourse genre. Further, that this linguistic inability also showsup in the type confusion between anger sadness that led us of and originally to look deeper into the different types of accounts and how theywere made up in terms of linguistic construction types. However, I assume that the relevance of these findings is weighted quite differently in the two different frameworks. A more cognitively oriented approach is most likely to consider as relevant the point in time when the knowledge base for individual emotions (or other cognitive systems) can first successfully be tapped, because from thenon, all confusions can be(and need to be) explained in terms of situational performance constraints. In order to tap this basic knowledge developmentally early as possible, an elicitation as technique must be chosen that imposes little as possible situational and as contextualconstraints. Accordingly, withinthecognitiveframework of emotions, to ask children to construct emotion scenarios in the first person genre, might not count as the most efficient way, exactly for the reason that their personal involvement might interfere with their actual knowledge. In addition, these distortions all occur after the basic knowledge ofparticular

Grammar in Practice


emotions has already been firmlyestablished. Therefore, accordingto such a position, our findings and elaborate discussionsdo not contribute to how children establish their emotion knowledge. At best, they may contribute to how children apply their knowledge under difficult and adversarial circumstances. In contrast to the cognitive frameworkof emotions and its approach to knowledge acquisitionas the major developmental achievement, the discursive orientation views knowledge of the emotions rather the result or the product as of participation in cultural practices. The discursive approach to emotions is primarily interested in the processes through which cultural knowledge obtains its motivational force or individuals, and for this reason, a confusion between emotions in particulardiscursive settings is of utmost interest, since it offers insight into the developmental process of how the cultural directives of emotions are sorted out. Consequently, comparisons between performances in different practices or discourse settings are of extreme interest for studies that focus on development as a form of cultural learning. And the findings and discussions presentan important starting point within this orientation. Emotions as Positions Taken in Practices by Use of Linguistic Construction Types One possibly puzzling concern lies hidden behind much of what has been discussed thus far: Why people, when do asked to give emotion accounts (of how they or others oncefelt), construe elaborate circumstances around happenings and events, that is, seek refuge the world actions? And why, in of when asked to construe events or happenings (in which they themselves or othersfigure as actors), do they deviate from the sequencing of actions and resort to references to feelings and emotions? This concern actually becomes more urgent when it concerns accounts in which the narrator seems implicated or is implicating to be someone else, that is, accounts that have been classified in this chapter as involved.Edwards Potter (19 9 2), in and response to seeming contradiction, this argue that in natural discourse, talk about events and happenings designed in is particular ways to allow inferencesabout mental life and cognition (p. 142) and- I would like to add-particularly inferences about emotions. Edwards and Potter also hold that theconverse is equally true. Assumptions about the world, what happened, andwhy it happened, are inferred from the the speaker way designed the emotions and motivations theactors. Thus, event construal and of
3 The potential argument that thereis no real confusion between the emotions, because obviously even younger children know what is (feelslike) and what anger (feels like), and that sad is the above at best accounts for a confusion of emotion talk,obviouslymisses the point.To know what it means be to angry, and to know what it means be sad are derived out of to not bodily sensations, at least not solely and not directly (unmediated). The meanings of anger and sadness are imbedded in talk, talk that is situated and simultaneously situating the interactantewith regard to how they want be understood and how they position themselves to as moral agents.



the construalof characters inner psychologiesare closely orchestrated with regard to one another. Accounting one, to a for large degree, indexes the other. However, these worlds of actors and events within which references to emotions are embedded do not stand their own. They versions produced on are in discourse situationsfor discursive purposes. The particular purposes can be manifold, such as to attribute blame to others (in order to save face or restore ones dignity), or to elicit empathy (in order to pursue joint retaliation or revenge). How the speaker positions him- or herself vis-i-vis the audience results in the particular construction design that given to the worldof actors is and events. Thus, the world the interactants (speaker-audience) of regulates the way the psychological reality ofthe actors are construedas well as what is happening in the eventsin which the actors have a role. As such, the approach I have proposed in the foregoing attempts to turn around the traditional,realist picture of the relationship between emotions, cognitions, and language. The realist picture starts from events happenings and as taking place in the world, to be re-presented in peoples thoughts and feelings, so that we can subsequently speak about these events as well as the thoughts, evaluative appraisals, andfeelings. I am suggesting an inversion of this relationship: In communication, which is the performative domain of social action, both events as well as stances toward them (evaluative or cognitive) are organized, not because they stored andavailable previous to are and outsideof any discursive purpose, be executed subsequently commuto in nication. Rather, eventsand theway they arethought about and valued by the speaker are constructs thatare borne outof the purpose talk. Consequently, of the way in which the purpose of manifests itself in the talk world of interactants is not a by-product, but rather the starting point a (discursive) analysis of for what is manifested in the talk in terms of the characters, their activities, and the evaluative position with regard to them. A s a concluding remark, let me briefly touch on an additional dimension regarding the notion of positioning that I alluded to earlier. In addition to the orchestration of characters with regardto one another at the level of what is being talkedabout (positioninglevel l and in addition to the orchestration of ) , the speaker-audience relationship (positioninglevel 2, the same linguistic ) construction types index how the speaker positionsor herself with regard himto the self(positioning level 3). Coordinating the content of talk with the discursive purposes for whichit is constructed forms the presupposition in what is commonly considered the constructionof self-identity. Whether the talk is actually about the self (the l or others(he or she), or about people ) , in general (one or the generalized you, as in the explanatory genre), italways reveals aspects of a moral order in the characters and audience are orchesway trated. The moral identity question What am in relation to the Good? I (Sarbin, 1995, p. 219) turns into a position with regard to ones own identity: Who am I? Thus, the constructionof characters in events at the level of

Grammar in Practice


content formation, the construction the speaker-audience relationship, and of the construction of ones self-identity are closely interwoven. And although traditionally psychologists start from the assumption the unity of the self, of and see narrative accounts and interactive relationships as orchestrated from and toward the purpose of maintaining this position, the approach that is schematically outlined in this chapter views this unity of a self (at least to a degree) as consisting of local achievements that are based on grammatical constructions for discursive purposes (Bamberg, 1997). In sum, what been offered inthis contribution to this volume is an attempt has to more clearly delineate between cognitive approaches to emotions (and emotion development) and their discursive counterparts. What has come to the in forefront this attempt is the role of language, in particular of grammatical constructs (construction types for discourse purposes) which form important building blocks in the formation the social constructs thatare achieved indevelopment of and interaction (e.g., emotions, intentions, memory, self, and identity). The close look at some these building of blocks, as illustrated in the study presented has here, led to an illumination what is involved inthe interplayof language practices, of emotions, concepts emotions, and their respective developments. of in

Thanks go to the students who participated this project, including: Andrea in Berger, Sunil Bhatia, Ayden Reynolds,and Michelle Sicard. In addition, I would like to thank Keith Nelson for his extremely helpful and clarifying comments on earlier drafts of this chapter.

Bamberg, M. (19 91). Narrative activity as perspective taking: role ofemotionals, negations,and The voice in the constructionof the storyrealm. Journal ofcognitive Psychotherapy, 5,275-29 0. Bamberg, M. (19 97). Positioning between structure and performance. Journal ofNarrative and Life History, 7.335-342. Bamberg, M. (1999). Language and communication-What develops? Determining the role of languge practices for a theory of development. In N. Budwig, 1. Uzgiris, & J.Wertsch (Eds.), Communication:Arena ofdeveloprnent(pp.55-77). Stamford, C T Ablex. Basso, E. B. (19 92). Contextualization in Kalapo narratives. In A. Duranti & C. Goodwin (Eds.), Rethinking context (pp. 253-26 9 . Cambridge, England: Cambridge ) University Press. Budwig, N.(1995). A developmental-functionalist approach tochild language. Mahwah, NJ: Lawrence Erlbaum Associates. Clark, E. (1993). The lexicon in acquisition. Cambridge, England: CambridgeUniversity Press. Davitz, J. (196 9). The language ofemotions. London: Academic Press. Edwards, D., &Potter, J (1992). Discursivepsychology. London, England:Sage. . Fletcher, P., & MacWhinney, B. (Eds.). (1995). The handbook ofchild language. Cambridge, MA: Basil Blackwell. Gleason, J. B., & Ratner, N. (I y 93).Psycholinguistics. Fort Worth,TX: Harcourt Brace Jovanovich. B. Goddard, C. (1997). Contrastive semantics and cultural psychology: Surprise in Malay and English. Culture and Psychology, 3,153-18 I.



Harrk, R., & Gillett, G. (1994). The discursive mind.Thousand Oaks, CA. Sage. Heelas, P. (1986). Emotion talk across cultures. InR. Harre (Ed.), The socialconstruction ofemotions. Oxford: Basil Blackwell. Hoff-Ginsberg, E. (1997). Language development. Pacific Grove, A Brooks Cole. C discourse. Language, 56,251-299. Hopper, P., &Thompson, S. (1980). Transitivity in grammar and Leff, J (1973). Culture and the . differentiation of emotionalstates. British Journal ofpsychiatry, 123, 299-306. Lutz, C. A. (1988). Unnatural emotions: Everyday sentiments on a Micronesian atoll and their challenges to western theory. Chicago, IL: University of Chicago Press. Matsamuto, D.(1994). People: Psychology from a cultural perspective. Pacific Grove, CA: Brooks Cole. Ochs, E. (1988). Culture and language development. Cambridge, England: Cambridge University Press. Ochs, E. (1996). Linguistic resources for socializing humanity. In J J. Gumperz & S. C. Levinson . (Eds.), Rethinkinglinguisticrelativity (pp. 406-437). Cambridge, England: Cambridge University Press. OzgaliSkan, $.(1997). Genre construction in emotional discourse.Unpublished masters thesis. Clark University, Worcester, MA. Sarbin, T. R. (1995). Emotionallife, rhetorics, and roles. Journal ofNarrative and Lije History, 5, 213-220. Schieffelin, B. B. (1990). The give and take ofeveryday life, Cambridge, England: Cambridge University Press. Schieffelin, B. B., Ochs, E. (1986). Language socialization. Annual Review ofAnthropology, 1, & 5 163-191. Stein, N.L., & Levine, L. J, (1990). Making sense out of emotional experience: The representation and use of goal-directed knowledge. In N. L.Stein, B.Leventhal, & T. Trabasso (Eds.), Psychologicaland biological approachesto emotion (pp. 45-73). Hillsdale, NJ: Lawrence Erlbaum Associates. Stein, N. L., Liwag, M. D., & Wade, E. (1997). A goal-based approach to memory for emotional events: Implications for therries of understanding and socialization. In R. D. Kavanaugh, B. Z. Glick, & S. Fein (Eds.), Emotions: The G. Stanley Hall Symposium. Mahwah, NJ: Lawrence Erlbaum Associates. Stein,N. L., & Trabasso, T. (1992). The organization of emotional experience: Creatinglinks among emotion, thinking and intentional action. In N. Stein & K. Oatley(Eds.), Cognition andEmotion (special issue), 6, 225-244. Stein, N.L.,Trabasso, T., & Liwag, M. (1992). The Rashomon phenomenon: The role of framing . and future-orientation in memories for emotional events. In M. M. Haith, J B. Benson, R. J. Roberts, Jr., & B.F. Pennington (Eds.), The development offuture oriented processes (pp. 409-436). Hillsdale, NI: Lawrence Erlbaum Associates. Talbot, J, & Bamberg, M.(1996). Affirmation and resistance of dominant discourses: The rhetori. cal construction of pregnancy. Journal ofNarrativeand Life History, 6, 225--251. Trabasso, T., & Stein, N.L. (1993). How do we represent both emotionalexperience and meaning? A review of Richard S. Lazarus"Emotions and adaptation'. Psychological Inquiry, 4,326-333. Wlerzbicka, A. (1992). Defining emotion concepts. Cognitive Science, 16,539-581. Wlerzbicka,A.(1994). Emotion, language, and cultural scripts. In S. Kitayama & H. R. Markus (Eds.), Emotion and culture(pp. 133-19 6). Washington, DC: American Psychological Association. Wlerzbicka, A. (1995). Emotion and facial expression: A semantic perspective. Culture and Psychology I 227-258. , Wlttgenstein, L. (1953). Philosophical investigations. NewYork Macmillan.

A Crosscultural Investigation of Australian and Israeli Parents arrative Interactions With eir Children


Macquarie University 4 University of Melbourne


Hebrew University 4 Beit Berl College

We tell children stories, relate the days events, and recapitulate missed television episodes. These are examples of narrative, the overt discourse manifestation of the recounting of eventsthat occur in daily life. We expect that people will be able to understand our narratives and, in turn, produce appropriate narratives when required. We assume these narratives will be relevant to the situation, coherent, and linguistically cohesive, building on common contextual knowledge shared by narrator and listener. How d o children learn tobe competent narrators? Most research on narrative discourse has centered around either adults or childrens productions of narrative structures of different types, from more structured narratives such storyas book reading to more open-ended narratives such as personal experience narratives. In recent years, numerous studies have focused on the development of childrens narrative ability, both in specific languages and crosslinguistically (e.g., Bamberg, 1987; Bamberg& Marchman, 1990;Berman & Slobin, 1994; Hickmann 1995; Karmiloff-Smith, 1979, 1981; McCabe & Peterson, 1991; Peterson & McCabe, 1983; Wigglesworth, 1997). Fewer studies have focused on the environment in which children are exposed to narrative discourse, reporting (a) cultural aspectsof narrative development (Brice Heath, 1983; Ochs, 1988; Polanyi, 1989; Scollon & Scollon, 1981; Schieffelin,1984); (b) the educational contribution the narrative discourse environment to of literacy development (Snow 1977); and mostrecently, (c)crosslinguistic and crosscultural features of narrative discourse (Blum-Kulka, 1997; Minami& McCabe, 1995).



Research concerned with narrativesas parental input has tended to take either a sociocultural or educational perspective. For example, Snowand an Dickenson (1990) found that among low-income families, there is variable access to literacy and narratives, and they speculatedon whether being good a narrator related to social class differences. Brice Heaths (19 83) famous study of Roadville and Trackton reported that African-American and Euro-American children in the southern United States are exposed to different forms and genres of narratives from birth.Brice Heath claimedthat thesocialization processes to which the children in each community were exposed predicted schoolsuccess more strongly than single factors such as formal language instruction or the amount of parent-child interaction. To date, very few studies have dealt with the types of narratives parents produce when the addressees their own children. Harkins (1992) showed are that mothers narratives told to their children varied according to both the mothers educational background and their perspective on the style of goal and the narrative. Peterson and McCabe (1996) undertook a longitudinal study in which they addressed the question of why children differ so much in the amount of orienting information they providetheir narratives. They found a in relationship between the amount of temporal andspatial information children provided in their narratives at age 3 with the amount of prompting these children received from their parents 1 year to 18 months earlier. In another longitudinal study, Haden, Reese, and Fivush (19 9 6 ) investigated the typesof comments and questions mothers directed to their children during joint story-reading activities. The mothers read both a familiar and an unfamiliar storybook to their children atages 3;3 and 4;8. When the children were5 ~ 0 , their literacy skills were assessed with a battery of tests. Haden et al. found that the mothers fell into one of three different types of storybook reading styles, which remained consistent time although not the typeof book. These over over authors also found a relationship between mothersstyle and childs print and story skills at 5;10. Minami and McCabe (1996)comparedinteractive personal-experiencenarrativesinJapaneseandAmericanmonolingual mother-child dyads, which revealed cultural and stylistic differences. Stavans (1996) analyzedparentsnarrativesdirected to their own monolingual Hebrew-speaking children. Herresults led her to propose a 3-stage model to describe the age-related strategies the parents deployedin their narrative interactions. These studieshave mainly regarded parents narrativesas one type of discourse input andchildrens narratives as a specific type of discourse production. Although they have yielded information about childrens narrative development, we know very little about how parent narrators and child listeners co-construct naturalistic, interactive narratives. The study reported in this chapter deals with narrative development from the perspective of the kind of narrative input that children from middle-class literate families receive from their parents in two different languages and

A Crosscultural Investigation


cultural backgrounds: English-speaking Australian and Hebrew-speaking Israeli. The study was motivated by Berman and Slobins (1994) extensive crosscultural research on the developmentof narrative ability in childrenas they face the cognitively complex task relating events as extended discourse. of It is meant to complement Berman and Slobins account with a detailed examination of the kind of input children receive from their parents. Our aim is to contribute to the growing body research that uses crosslinguistic and of crosscultural comparisons to discover similaritiesand differences in parentchild discourse.We hypothesize that the discourse narrative interactions of may importantly influence how children learn torelate narratives. Although narrative development is partially driven by universal aspects of cognitive development, it must also be driven by language-specific structures and/or cultural traits of the community thatuses the language. Thespecific objective of the study is to delineate featuresof Australian and Hebrew parental storytelling practices that may be conducive to the childrens acquisition of the cultural andlinguistic knowledge appropriate narrative construction. for In this chapter report on the natureof parent-child narrative interactions we produced in naturalistic conditions when parentstell the same story used to elicit the narratives in Berman and Slobin (1984; Mayer, 1969). In particular, we are interested in the means used parents to by address their children three at different ages, in two different cultures, and in two typologically different languages. Rather than examining the storyitself, we consider the nature of age- and language-related differences apparent in monolingual AustralianEnglish and Israeli-Hebrew parent-child discourse, shown byuse of: rhetoras ical questions; labeling as means to teach the child; personal digressions; affective and evaluative comments about the story; didactic or judgmental or moral comments; and a rangeinteractive strategies. of


The children and their parentswere monolingual Hebrew speakers in Israel and monolingualEnglish speakers inAustralia. The children were aged3, 5, and 7, and only single child from any family was included. Theage groups a one were specifically chosen to provide comparative data for Berman and Slobins (1994) crosscultural study. There were 10 parent-child dyads in each group.

Parents were askedto relate a story from the wordless picture book,Where Frog, Are You?to their child. This story depicts the adventures a boy, a dog, and a of frog. An outline of the storyprovided in the Appendix. This story chosen is was for the following reasons: provides a solid basis for comparison with previous It


W I C C L E S W O R T H 6: S T A V A N S

studies; the plot line sufficiently complex forall age groups; the narrative is structure provides ample opportunities to expand and elaborate episodes or to compress and encapsulate them; it provides a thematic basis thatis flexible enough to suit the different ages and culturalbackgrounds of the populations involved in the study.

Parents were provided witha copy of the book, an audio cassette, and a small cassette tape recorder (where needed) to enable them to record the storytelling at a convenient time in their own homes. They were askedto familiarize themselves with the book before telling the story the child and ensure thatno other to to children were present during the session. In addition, parents asked to were record the first attempt at the so that thesession would provide natural, untask, rehearsed data. It was stressed that the parent should be the one to the story, tell while interacting with the child ashe or she normally did whentelling a story.
Transcription and Coding

English-languagerecordings were transcribedinstandardorthography. Hebrew-language recordings were transcribed in a modified broad phonemic script. All utterances from both parent and child were included. Subsequent to this, each clause spoken by each parent and each child was classified into two different categories: story and conversational. We define each category as follows: These were clauses in which the parent provided information about the story. More precisely, they were statements referred directlyto activities that depicted in or inferable fromthe pictures of the book, or provided contextual or descriptive information, and descriptive statements. Included in this categoryas(but subcategories)were parents explicitly stated opinions about the events depicted (e.g., I think hes going to fall down), as well as comments about the motivations, feelings, or desires of the characters involved in the story (e.g., little boys very The
happy now). Parent-story:

In theory, it was possible for a parent to do no more simply relate the story than in the session, so that all clauses would be categorized as parent-story. In fact, this occurred in only one case, with the parentof a Hebrew 5-year-old) and even so, there were a number of interjections from the child.

Included in this category were any utterances by the child (including responses toadult questions) that contributed the ongoing story line, description, to or context depicted in or inferable from the pictures.

A Crosscultural Investigation


Parent-conversational: The parent conversational clauses were the focus of the analysis. Consequently, they were further categorized oneof four types: into
1 .

Story-related conversational clauses were defined as attempts by the parent to involve the child in the storytelling activity. Thus, this category included the following types of questions, which were designedelicit various types into of formation about the story represented in book the a. Wh-type questions requesting general or specific information, for example, Whats happening? Whats happening to the dog? Whats the boy doing there? b. Wh-type questions asking explicitly for the childs interpretation of the events depictedor an opinion about what was depicted in the book, for e ample, What do youthink is happening? Whatdo youthink is happening to the dog? Why he doing that? What do you is going to happen to the is think dog? c. Questions about the internal states of characters, for example,How d o y o u think hesfeeling? d.Labeling and y e s h o questions designed to focus the child on a some specific item depicted in the book, for example, is that? Is that agoWhat pher? What do you think that is? 2 . Focusing devices were clauses used by parents to draw the childs attentionto the shared activity,or to focus the childs attention on the task. Examples are look and listen clauses designed to draw the childs attention to something in the book, for example, Look at that!Look whats happening now! Look! 3 . Non-story-related conversational clauses consisted of interaction tokens that were related to something the child had previously said,that did not elicit but or state any new information related to the storytelling task. These were: a. Statements adding something to what the child had said; b. Clarification or repetition of a childs utterance; c. Correction of childs previous utterance; d. Confirmation of childs utterance as correct. 4. Asides were defined as any interaction component was unrelated to the acthat tual storytelling or a clear diversion from the storytelling task. This category had two basic subcategories: a. Comments, for example, This book has pictures. b. Digressions involved discussion of events tangential to the storytelling task, for example, you remember when you were stung by a disDo The bee? cussion was often quite and involved before the dyad returned the long to story activity. Child-conversational: Child utterances that did not specifically relate to the story utterances (including responses to adult questions) were assigned to this category.



Mean Numberof Clauses, Range,and sd for English-and Hebrew-Speaking Children at Ages3, 5, and 7 Years
Age in Years Mean



English Range

74.404 63.921 117.696

281.4 191.1 176.6

123-410 62-375 73-298

86.124 85.137 85.516



80-277 113-489



Overall Length of the DyadicSessions

The overall length of the storytelling sessions (measured as number of clauses produced by both parent and child) was computed. The mean number of clauses, range, and standard deviations for each age and language group are given in Table 4.1.A two-factor ANOVA revealed no significant effects for either language or age across the groups. However, there was a statistically significant difference forage (F = 4.401, df = 2 ~ 7p, = .ozzz) within the Hebrew dyads, with the 3-year-olds sessions being significantly longer that those of the 7-year-olds. Across languages, the Hebrew3-year-olds sessions were significantly longer than thoseof the English dyads (F = 6.93, df = 18, p = .0169). Substantial individual variation length of sessionamong the dyads noted in was across all groups. As a result, all the analyses that follow are represented proportionally in relation to overall length in order to neutralize this effect and to properly account qualitative characteristics of the data. for
Narrative-Interaction Clause Profiles for Each Language Croup

In order to compare the story-interaction styles of Israeli Hebrew-speaking dyads and Australian English-speaking dyads, is first necessary to sketch the it profile of clause distributionfor each groupseparately. The profile of the English-language narrative sessionsis shown in Fig. 4.1. The distributionof clausetypes suggests that: The three groups are heteroage geneous; 5-year-old children were exposed to a greater proportion o f story clauses from their parents than 3-year-old children = 3.552, p = 0.0427); (F sessions with the 3-year-oldswere characterized by more conversational input by the parents ( F = 5.424, p = .0105); and 7-year-old children made a greater contribution to the story than the 5-year-olds 4 . 4 2 7 , ~ 0.217). (F= = The profile of the Hebrew-language dyadic sessions showed significant no differences. This profile, shown inFig. 4.2, suggests that: theage groups are very

1 .oo 0.90

A Crosscultural Investigation




Chlld story

I converaatlonal Parent
Parent story

Child conversational



Five Seven

F I G . 4.1. Proportion of parent-story, parent-conversational, child-story, and child-conversational clauses used by English-language dyads.


" "

Parent conversational


. 5 a

H 2

0.70 --0.50 0.50


. .



0.20 -0.10 -.... .... ....

Heb three

Heb five

Heb seven

4.2. Proportion of parent-story, parent-conversational, child-story, and child-conversational clauses used by Hebrew-language dyads.

homogenous; both7-year-olds and their parents produced a slightly greater proportion of story clauses than those of the othertwo age groups;dyads with 3- and 5-year-olds produced more conversational clauses than those with 7-year-olds; and the proportion parent-story clauses was approximately of equivalent in all three age groups.


W I G C L E S W O R T H 6: S T A V A N S

0.70 1 0.60
0.40 E

b a

2 0.30




Proportion of parent-story clauses in English and Hebrew.

Comparisons of Australian-English and Israeli-HebrewNarrative Interactions by Clause Type

In order to provide a more detailed descriptioneach language group and of trace similarities and differences across the groups we examined each category in turn, comparing the language groups and the three groups. Statistical two age analyses are used to determine where differences were significant and where they were not. The resultingprofiles describe the storytelling practices the of two cultural groups.
Story Clauses

Although the parents told the story, some contribution by the children occurred in all groups. We first determined the number of story clauses contributed by the parents, and then those contributed the children. by A two-factor ANOVA showed that there was a significant difference in the percentage of story clauses usedby the Hebrew- and the English-speaking parents ( F = 11.304, df= 2,54, p = .0014). Compared to the English-speaking parents of 3- and ~-year-olds, the Hebrew-speaking parents tended to focu more on the story ( F = 10.063, df= ~ 1 8p , = .0053 with 3-year-olds; and F = 6.81, df= 1,18, p = .0177 with 7-year-olds), but this was not the case with the 5-year-olds, where the English-speaking parents exhibited a proportion similar to thatof the Hebrew-speaking parents (see Fig. 4.3). With respect to the childrens contributions to the story, there was a significant overall difference in the proportion of story clauses by children of

A Crosscultural Investigation


0.70 1

c 0

Oa60 0.50

English Hebrew

H3 year olds
5 year olds


2 0.30
0.1 0

7 year olds

Proportion of child-story clauses in English and Hebrew.


different ages (F = 4.292, df= 2,54, p = .0186) and in the different languages ( F = 6.858, d f = 2,54, p = .0114). Figure 4.4 depicts these differences by language and age. The English-speaking 7-year-old children contributed significantly more story clauses than the English 5-year-olds ( F = 4.427, df= 2,27, p = .0217). While both the English 3- and 7-year-olds contributed substantially more storyclauses than their Hebrew peers, only the difference between the 3-year-olds was significant (F= 6.294, df= 1 ~ 8 ,= .0219). This p is because there was considerable intragroup variance in the English 7-year-old group, withthe proportionof story clauses contributed by the children ranging from .056 to.461. Another comparison of the groups shows that, whereas the Hebrew-speaking childrens contribution of story clauses to the narratives increasedsteadilywithage,thecontributions of theEnglish-speaking 7- and 3-year-old children were greater than those of the English 5-year-olds. In fact, the English- and Hebrew-speaking 5-year-olds contributions were rather similar.
Parents Conversational Clauses

Although parents were requested to relate the story to their children, of all the narratives were to some degree cooperative, as can be seen from the contributions made by the children to the narrative story line. In cases, the most contributions by the children were elicited by the parents through arange of techniques. In thissection, we attempt todelineate the degree to which parental attempts at co-constructing the narratives differed across age groups or cultures. To this end we first examine the parents conversational clauses and then examine each of the four subcategories defined in the Method section.




0.60 0.50

3 year olds


5 year olds
7 year old6

2 0.30
0 . 10

F I G . 4.5.


Proportion of parent-conversational clauses in English and Hebrew.

This is followed by a brief outline of the number of the childrens conversational clauses that did not specifically constitute part of the narrative. Figure 4.5 compares the proportionof conversational clauses used by the parents. For theEnglish-languagegroup,itshowsthattheparents of 3-year-olds used significantly more conversational clauses than parents of 5-year-olds (F = 5.426, d f = 2 ~ 7p = .0105).In the Hebrew group, the differ, ences were less salient overall, with a pattern ofsteady-though not statistically significant-decrease in the use of conversationalclauses as the children got older. Within age groups, there were significant differences for the 3- and 7-year-olds, with the English dyads incorporating more conversation than Hebrew dyads. The conversational clauseswere subdivided into four different subcategories: story-related conversational, focusing methods, non-story-related conversational, and asides.
Story-RelatedConversationalClauses. Story-related conversational clauses were defined as a parents attempts tokeep the storytelling process dynamic by drawing thechild into it. There was a significant difference for language (F = 26.287, df= 2,54,p = .OOOI; see Table4 2 ; this was because the parents in the .) English-language dyads with children at ages 3 and 7 incorporated more story-related conversational clauses than theIsraeli parents withchildren at the sameages(F=17.796,df=2,27,p=.0005atage3andF=8.975,df=2,27,p= .0078 at age 7). One of the reasons for this is that theIsraeli parents primarily used story clauses to relate the events the narrative, which diminishes the of proportion of use of story-related conversational clauses. For example, the Hebrew-speaking parents used virtually no deictic type questions(e.g., What is


A Crosscultural Investigation


Number (and Proportion ofTotal Clauses for Each Age) of Parent-Conversational Clauses Subcategorized as Story-related, Focusing Devices, Non-story-related, and Asides, and Child-Conversational Clauses, for English and Hebrew Dyads
Age in



Story-related Focusing
315 ( l ) .7 131(.O7) 56 (.03) 23 (.01)

Non-story Child-Conv. Asides

7 (.OO)



2 1(.11) 1

5 3

260 ( ~ 3 ) 778 (.14)

203 (.11)

210 (.04)

7 (.OO) 134 (~37) 4 (.OO) 18 (.OO) 255 (l. .?) 571(.lo)

162 (.09)

244 (.12) 617 (.11)

75 (.04) 418 (.06) (.09)

66 (.04)

9 2 (.03)
1 (.01) 0

79 (.03)
20 (.01) 122 (.oz)

174 (.06) 125 (.07) 375 (.06) 76 (.04)


168 (.03)

23 (.01)

262 ( ~ 4 )

413 (.15)

203 (.U) 878 (.14)

that?) with their children, whereas English-speaking parents used them extensively with theirchildren, with significant differences across all agegroups for this subcategory. Focusing Strategies. Focusing strategies were clauses used by the parent to keep the childs attention focused on the task. English, these were tokens In such as Look, Look at that: and in Hebrew, tireh, tistakelhistakli. A 2 x 3 ANOVA of focusing strategies revealed a significant difference acrossages ( F = 22.285, df= 2,54,p< .oool),andlanguages(F= 10.478, df= 2,54,p< .oool). Looking at the age differences, within the English group there a dramatic was decrease in theuse of focusingclauses across theage groups, with asignificant difference betweenthe 3-year-olds and the othertwo groups ( F = 18.749, df= 2,27, p < .oool). The Israeli parents used this technique with the 3- and 5-year-olds (although not to the same extent as the parents of the English 3-year-olds)butverylittlewiththe7-year-olds(F= 6.466,df=2,27,p= .0051; see Table 4.2). Comparing across languages, there were significant differences between the 3-year-old groups, with the English-speaking children receiving significantly more focusingdevices than theHebrew speakers( F = 13.248, df= 1 ~ 8 ,= .0019), although the dyads with 5-year-olds fairly comparable. p were Non-Story-RelatedConversationalClauses. This category included comments by the parentin response to something the child had said previously but which did not provide any new information related to thestorytelling process.



There were generally veryfew instances of non-story-related conversational clauses, but despite the smallproportions, there was a significant effect for age ( F = 3.19, df= 2,54,p= 0.049) andlanguage(F= 15.692, 2.54,p= .0002), df= and there was an interaction effect ( F = 3.388, df= 2,54, p = .0411). These differences are manifest in the higher proportion used by parents of the Hebrew-speaking children at age levels. all
Asides. This category included comments unrelated to the story and digressions that deviated completely from the topic the story. Overall, there of was a significant difference in the use of asides between the language groups ( F = 9.605, df = 2,54, p = .0028),withtheEnglishchildrenhearing significantly more asides than the Hebrew children. However, this effect was only manifest at the group level for each language, with no significant differences either within the language groups, or within theage groups (see Table 4.2).
Childrens Conversational Clauses

Because the focus of the analysis for this chapter was on the parents interactions with their children, the childrens conversational clauses were not subcategorized. Children conversed with, responded to, their parents either or with a story-related statement (discussed earlier) with some other type or of clause. When we examined the contributionsby the children to the overall conversation (excluding those contributions that specifically related to the story), we found a similar pattern across groups, and with no significant all effect for language or age.
Parent-ChildSocialization and Linguistic Interface

When we compare the parents story-related conversational exchanges with th childrens contributions to the story Fig. 4.6), we find that at ages 3 and 5 (see there was a relatively even matching of parent-child exchanges; however, this was not the case for the 7-year-olds in either cultural group. At age 7, the children made substantially more contributions to the story than those solicited by their parents. Thus, appears that,in this naturalistic collaborative it task, there was great variability in individualstyle; the children could not resist contributing to the activity, particularly the 7-year-olds in both cultural groups. Perhaps this need to contribute andparticipate in the storytelling task at age 7 is naturalbecausechildrenatthis agein bothculturesare relatively independent readers and the storytelling task becomes a rather different activity (for both parent andchild) from what it is when the children are 3 and 5 years old. Seven-year-olds not only want to contributeto the story, but need to do so in order to maintain interest in the activity, which theymay otherwise


A Crosscultural Investigation


English parent story conv. English children story Hebrew parent story conv. Hebrew children story



0.14 -0.12

. 0.10 p

2 a


e 0.08 -0.06


0.04 -0.02




F I G . 4.6. Relationship between parent-conversational story clauses and child-story clauses in English and Hebrew,shown as proportion of total clauses.

perceive as childish and inappropriate. Itmay also be the case that the parents of the 7-year-olds allowed more of this active involvement because they perceived it to be the natural oftelling a story to a child way of that age.

A comparison of the profiles (Figs. 4.1 and 4.2) of both language groups suggests that the collaborative narrative session of parent-child dyads inthe Hebrew speakers' group was more homogeneousacross the children's ages than that of the English speakers' group, where there were salient differences across the ages. A s the main purposeof this studywas to investigate crossculturally and crosslinguistically the parental narrative styles in Hebrew-speaking and English-speaking parent-childdyads, we conducted an in-depth analysis of the parental speech. Singling the story clauses produced by the parent in the out dyad, we found homogeneity across the age groups for the Hebrew speakers, with the pattern matching that the parents of the English 5-year-olds. The of English 3- and 7-year-olds heard proportionallyfewer story clauses from their parents, but contributed proportionally more clauses to thestory, particularly the 7-year-olds. These results suggest thereis a crosscultural difference in the storytelling practices of Hebrew-and English-speaking parents.The storytelling practices



vary from cultureto culture in the the task is interpreted and conducted. way Storytelling appears to be perceived by Israeli Hebrew-speaking parents of children at ages 3 through 7 years to be a reading-based activity. Englishspeaking parents appear approach the storytelling to task differently according to the age of their child, or the demands of the child, by making storytelling a more interaction-basedactivity. The conversational components of the Hebrewdyads talk decreased with age, in both parents and children, with a concomitant gradual increase age with in the proportion of story clauses. In other words, the older the child the was, longer the story becameand the fewer conversational exchangeswere needed between parent and child. Given that at 7 years the Israeli children made considerably greater contributions to the actual storytelling than simply those requested of them by their parents, we may assume thatthese children were becoming increasingly involved in the storytelling activity, a trend that perhaps reflects the development of literacy skills. Although the English groups were much more variable across the age groups, the 5-year-olds (both English and Hebrew) interactions were more similar to each other than those of theother two age groups. Thismay suggest that, regardless of culture or language, parents of 5-year-olds universally deploy a similar proportion of their storytelling activity in interaction opposed to as telling the story. This pattern may be typical of parents style in using the narrative task as preparation for the formal educational system (i.e., school) or, alternatively, parents may make adjustments in keeping with storytelling interaction practicesto which the childrenmay have been exposed in kindergarten, preschool, or day care. These resultssuggest a greater focus the story for the Hebrew on dyads, and for the English 5-year-olds. In both languages there were fewer asides (i.e., comments and digressions) to the 5-year-old children, and the g-year-old English-language children heard fewer conversational clauses than their3- and 7-year-old compatriots. This leaves us with a situation in which the Hebrew children and the English 5-year-oldswere involved in similar narrative interactions, while the English 3- and 7-year-olds sessions followed a different pattern. Around 5 years appears to be a criticalage for bothcognitive and linguistic development. Socially, at thisage children are just coming toa major life up change, as they are either in the year immediately before school their first or in year of the formal education system in bothAustralia and Israel. Cognitively, this is also a critical age. In their extensive crosscultural examination of childrens narrative, Berman and Slobin (1994) argued that the 5-year-olds do not form a homogeneous group. There are a number of reasons for this. First, some 5-year-olds construct globally structured and thematically motivated narratives (p.6 S), whereas othersin this group relate only one or two of the major plot elements. Second, considering linguistic some childrenat this level,

A Crosscultural Investigation


age demonstrate the use of elaborate syntax and a rich lexicon (cf. KarmiloffSmith, 1981, 1985), whereas others produce juvenile-sounding text with impoverished linguistic devices(Berman Slobin, 1994, p. 42). Consequently, & Berman and Slobin argued that the narratives of the 5-year-olds in their cross-sectionalstudysuggestedthatthesechildren were ata mixed or transitional stage. This demonstrated by the variation they showed in the was types of narratives they told.The results of our study support these claims in terms of the 5-year-old childrens contributions to the story in both English line and Hebrew. The picture is slightly different when we look at what happened in the intradyadic interaction patterns beyond delivery the story line. Parents are of aware that their children need to become not only proficient storytellers,but also socially and cognitively adjusted to the environment which they are in growing up. It maybe that parents 5-year-olds are particularlyattuned to of this age as critical, so that the parents focusis on the story, howto tell it, and how to structure the story and informational content. Parents from both its languages inour study were focused on the structural and functional aspects of the narratives theywere telling to their5-year-old childrenand less concerned with the conversational procedures involved. It is important tostate that atage 5 in western cultures, most children are familiar withthe storytelling activity (both at home and outside) and, in fact, that activity is part of childrens entertainment repertoire(as opposed to other types discourse, suchas basic of explanations or definitions, which may be more typical for younger children or, alternatively, discussion of debatable problems,which may be more typical for older children).Five-year-olds are an authentic audience for narratives of this type; not only are theymore receptive to storytelling input, they are more also passive in interaction with the storyor theteller. We may hypothesize that the Australian parents are more concerned withthese issues when their children are 5, allowing greater flexibility and negotiation atages 3 and 7. The Israeli parents in our study, on the other hand, demonstrated a more focused approach to the task at hand acrossage groups. the As a result, in contrast to the 5-year-olds: the 3-and 7-year-olds parents from the two language backgrounds showed greater differences in their storytelling. At both 3 and 7, the English-speaking children made substantially greater contributions to the story than their Hebrew-speaking peers. Whereas the Hebrew-speaking parents and children across three age groups engaged all in similar storytelling interactions, the English speaking dyadsdid not. At age 3 the major difference was the conversational nature ofthe sessions determined by the parents of the Australian children, whereas the Israeli children participated in sessions that were less conversational and more story focused. Despite the difference in the proportions of narrative versus conversa tional input provided by Hebrew- and English-speaking parents, it may be that the greater elaboration and detail provided the Hebrew narratives in serves the




same interactional purpose the parental conversational approaches apparent as in the English narratives, namely, to retain youngchildrens attention eitherby creating informative narratives by monitoring whether the childfollowing or is the story line. Berman and Slobin (1994)noted that one of the features that distinguishes 3-year-olds narratives from those other age groups is the interactive nature of of their accounts. Thus, the American 3-year-olds used exhortations such as See! Look! throughout their accounts and frequently digressed from the content of the pictures. Berman and Slobin found that their 3-year-olds (American-born middle-class children) provided interactive and personalized accounts and required much more confirmatory prompting and encouragement from the interviewer than the other groups did. Renner age (1988) noted that these same3 -year-olds engaged the listener such a way that it was very in difficult to maintain silence throughout thenarrative. Combining this with the resultsfromthepresentstudy, we may postulatethatherAmerican English-speaking children mayhave learned their interaction style from their parents. A preliminary examinationof the Hebrew stories has suggested that the narratives addressed to different age groups illustrate different styles; for example, the narratives addressed to the 7-year-olds more literaturewere oriented. Such narratives often included literary opening and closing formula common in the literaryexperience of 7-year-old Hebrew-speaking children. The negotiative nature of the more interactional storytelling practice typical with 3- and 5-year-olds may be reduced or neutralized for 7-year-olds. That is to say, many things that parents of younger children may do in order to establish clarity, suchas negotiate the name of the main character with the child, are no longer done with the older group. Rather, the parentaware that is the naming of the main character with a common boys name, for instance, would not be contestedby the child. Hence, the older the child more the is, the parent takes for granted a common literary experienceand the more emphasis the parent places on developing and elaborating the plot line(Stavans, 1996). The case for English is slightly different. The English-speaking children heard significantlymore conversational clauses. The parents of the Australian 7-year-olds appeared to have to work at being able to maintain the narrative task. The 7-year-olds were keen, and enthusiastic to take over the storytelling activity. Although this also appeared to be the case with the 3-year-old English-language children, there was an important qualitative difference between these two age groups. The 3-year-olds contributions to the story resulted from questions and comments posed by their parents; for the 7-year-olds, this was not so clearly the case. Their contributionswere often unsolicited. These differences manifested by the two cultural groups require further investigation. They may be driven by language differences or by cultural

A Crosscultural investigation


differences that relate to how totell a story. Whereas this study has only attempted to outlinedifferences in the overallstructure of these storytelling sessions, the next focus of investigation is the storyclauses themselves. This analysis delineates the differences and similarities across and within the languages in the actual content of the stories the children are beingtheir by told parents. The present study suggests that different styles of storytelling are used in Israel and Australia, particularly when children are3 and 7 years old. Israeli parents of 3-year-olds in our studywere more focused on plot-line delivery, whereas the Australian parentswere more interested in involving the children in the activity. Inother words, for the 3- and 7-year-olds we see a storytelling difference between Israeli and Australian parentsas being more story-oriented for the former and more interaction-oriented for the latter. The difference in styles is negligible at age 5, which may bea reflection of an awareness by parents-in general-that children of this age must be well preparedforschoolandtheonset of formaleducation.Althoughthe differences in the group patterns are negligible atthis age, it is worth pointing out that while the Australian parents of 5-year-olds very focused were onthestorytellingtask,thechildrenseemedtobeconcentratingon the parental input to such an extent that their own contributions were minimal. As for the 7-year-olds, the parent-child relationa storytelling task more in is vulnerable by definition.At this age both Israeli and Australian children have been trained to read, and most of them are independent readers. Hence, the storytelling task may be somewhat odd for the child, who may no longer be accustomed to being told a storyby the parent. This may add an affective element; that is, either the childmay be insulted or bored by the parent telling the story and undermining thechilds ability to read, or alternatively, the child may be cuddled into the extra attention given. These affective variables act differently for each dyad in each language groupand across the groups.Still, the very suggestive pattern that emerges is that the 7-year-oldIsraeli children were neither too eager nor too involved in the storytelling, while the Australian 7-year-olds provided a lot input to the story and attempted take over the of to task. The findings here arenecessarily tentative. The group sizes at each age level were small; the parentswere professional middle-class parents may not be and fullyrepresentativeofthepopulationfromwhichthey were drawn. Nonetheless, the results suggest cultural differences in the manner in which do the two cultural groups approach the storytelling task, and there is some tentative support for the idea that these differences are reflected to theway children will approach a storytelling task. Clearly further investigation is needed. The next stage of the analysis of these datawill address the question of story content.This may provide some explanation of the cultural differences


W I C C L E S W O R T H 6: S T A V A N S

found in the approaches the Israeliand Australian parents,as well as theage of differences found within theAustralian data. In addition, data collection of parental storytelling practices from other cultures expand our knowledge will and awareness of the range of possibilities available.

Bamberg, M. (1987). The Acquisition ofnarratives: Learning to language. Berlin: Mouton de use Gruyter. The Bamberg, M,,& Marchman, V.(1990) What holds a narrative together? linguistic encodingof episode boundaries.Papers in Pragmatics, 4,58-121. Berman, R.,& Slobin, D. I. (1994). Different ways ofrelating events in narrative: crosslinguistic A developmental study.Hillsdale, NJ: Lawrence Erlbaum Associates. Blum-Kulka, S. (1997). Dinner talk: Cultural patterns ofsociability and socialization in family discourse. Mahwah, NJ: Lawrence Erlbaum Associates. Brice Heath, S. (1983). Ways withwords: Language, life and work in communities and classrooms. Cambridge, England Cambridge University Press. Haden, C., Reese, E., & Fivush, R. (1996). Mothers extratextual comments during storybook reading: Stylistic difference over time and across texts. Discourse Processes, 21,135-169 Harkins, D. A.,(1992). Parentalgoals and styles storytelling. In Demick, K. Bursik, & R.Dibiase of J. (Eds.), Parental Development(pp. 61-74). Hillsdale, NJ: Lawrence Erlbaum Associates. Hickmann, M.(19 95). Discourse organization and the development reference to person,space, of and time. In Fletcher &B. MacWhinney (Eds.), handbook ofchild language(pp.194-218). P. The Cambridge, M A Blackwell. Karmiloff-Smith, A. (1979). A functional approach to child language. Cambridge Studies in Press. Linguistics, W . Cambridge, England: Cambridge University Karmiloff-Smith, A. (1981). The grammatical marking of thematic subject in the development of language production. InW. Deutsch (Ed.), The childs construction of language (pp.121-147). New York AcademicPress. Karmiloff-Smith, A. (1985). Languageand cognitive processes from a developmental perspective. Language and Cognitive Processes, I, 61-85. Mayer, M. (196 Frog, where are youlNew YorkDial Press. 9). McCabe, A., & Peterson, C. (1991). Getting the story:longitudinal studyof parental styles in A eliciting oral narratives and developing narrative skill. In A. McCabe & C. Peterson (Eds.), Developing narrative structure 217-253). Hillsdale, NJ: (pp. Lawrence Erlbaum Associates. Minami, M,,& McCabe, A.(19 9 5). Riceballs and bear hunts: Japanese and North American family narrative patterns.Journal ofchildLanguage, 22,423-445. Ochs, E. (19 8 8 ) . Culture and language development: Language acquisition and language socialization Press. in a Samoan village. New York Cambridge University Developmentalpsycholinguistics: Threewaysoflookingata childs Peterson, C., & McCabe,A. (1983). narrative. New York Plenum. Peterson, C.,& McCabe, A. (1996). Parental scaffolding context in childrens narratives. C. E. of In Johnson & J. H. V.Gilbert (Eds.), Childrens language, Vol. 9 (pp. 183-196). Mahwah, NJ: Lawrence ErlbaumAssociates. Polanyi, L. (1989). Telling the American story. Cambridge, M A MIT Press. Renner, T. (1988). Development oftemporality in childrens narratives. Unpublished doctoral dissertation, Universityof California, Berkeley. Schieffelin, B. (1984). How Kaluli children learn what to say, what to do, and how to feel: An ethnographic studyof the development ofcommunicative competence. New York Academic Press.

A Crosscultural Investigation


Scollon, R., Scollon, S. (1981). Narrative, literacy andface in interethnic communication.Norwood, & NJ: Ablex. Journal ofchild Snow, C. (1977). The development of conversation between mothers and babies. Language, 4,1-22. Snow, C., & Dickenson, D. (1990). Social sources of narrative skills athome and at school. First Language, 10, 87-104. Stavans,A. (1996). Development of parental narrative input. Journal ofNarrative and Life History, Wigglesworth, G: (1997). Individual approaches to the acquisition of narrative. Journal ofchild Language, W , 279-309.
6, 253-280.

A Picture-by-Picture Descriptionof Frog, where are you? (Mayer, 196 g )

1 2

3 4 5 6 7 8 9
10 11 12


1 5


17 18 19
21 22 23 24

A boy and his dog are looking atfrog in a jar. a The boy andthe dog arein bed; the frog is escaping fromthe jar. The boy andthe dog wake upand lookfor the frog. The boy andthe dog searchthe room forthe frog. They look outof the window; the dog has the frog's onhis head. jar The dog falls outof the window, breakingthe jar. The boy has also climbed out the window and picks up the dog. of They go to the forest, the boy calling forthe dog. The boy searches in gopher hole. a A gopher comes outof the hole and bites him the nose. on The dog is sniffing at beehive; the boy is searching in tree. a a The boy is frightened by an owl; dog is chased by bees. the The boy is bya rock. The boy climbs up the rock and leans 0n"branches' (really deer antlers). The deer hiding behind rock picks upthe boy on its antlers. the The deer runsoff with the boy, the dog in hot pursuit. The boy and the dog are thrown down the cliff. They landat the bottom in a pond. The dog climbs on the boy's head. The boy hears something and tells the dogbe quiet. to They climb over an old log. They seetwo frogs sittingon a bank, oneof them the escaped frog. They seethat there are also number of baby frogs. a They takeone of the babies and leave waving the remaining frogs. to

This Page Intentionally Left Blank

The Acquisition of Polite Language by Japanese Children


Keio University

One fascinating feature of the Japanese language that is shared by many languages around theworld (e.g., Korean, Javanese,Vietnamese)is its elaborate system of polite language. Keigo, or verbal politeness in Japanese, is often described as involving two basic dimensions,namely, formality (whichreflects the psychologicaland/or social distance between participants)and honorifichumble language (which indicates respect and deference). Politeness in Japanese influences the formmany linguistic components, suchpronouns, of as verbs, adjectives, nouns, and conventional expressions. In addition, politeness is also marked by paralinguistic and nonverbal features such as intonation, physical distance, and posture, as well as conversational strategies such as indirectness and hesitancy. How doJapanese children learn how to master this complex system of linguistic politeness?

Japanese societyis often described as being a vertically structured society, and respect and deferenceto otherscan be expressed by the use of polite language (Nakane, 1973).The Japanese keigo polite language system divides keigo into four general categories: sonkeigo honorific-respectful language, kenjoogo humble language, teineigo formal language, and bikago beautification honorifics (for a full description the keigo system refer to Bunkachoo, 1974, of 1986; Kindaichi, Hayashi, & Shibata, 1988; Martin, 1g75). Although keigo functions as a marker of formality and respect, it can also be an indicator of psychological and/or social distance, as well as a mark of good breeding (Mizutani & Mizutani, 1983; Oishi, 1983). Postwar Japanese society,based on a more democratic system with a greater emphasis horizontal relations, said on is
1 Finally, along with these four basic categories, there large number greetings and polite are a of expressions that involve gratitude, apology, self-blame, and humility (Martin, 1964).




to have had a strong influence the current functions and ofkeigo. Many on uses linguists claim that keigo nowadays is less a marker of vertical relations (i.e., respect and deference) than an indicator of social and psychological distance (i.e., formality), and a marker etiquette and refinement(e.g., Bunkachoo, of 1986; Mizutani &Mizutani, 1983,1987; Okuyama, 1983). Keigo usage varies accordingto many speakedhearer characteristics, such as gender, age, social status, occupation, dialect, educational background, group membership, and degree of familiarity, well ascontextual factors such the as as topic of conversation and formality setting (Ide,1982,1990; Kikuchi, 1996; of Niyekawa, 199 1;Yoshioka, 199 5). Another key factor is strategic purpose, such as sarcasm. Levels of politeness cannot be determined the basis of one factor on alone, as all of the above-mentioned factors interact in an extremely complicated manner. As Maynard (19 9 0 ) wrote, choosing the appropriate style in different social encounters requires social sensitivity and experience (p. IS), which makes it particularly challenging for nonnative speakers to master.
Sonkeigo Honorific-Respectful Language

Honorific language,or sonkeigo (also referredto as subject honorifics),is used to show respect when one describing another personor his or her actions or is belongings. Sonkeigo is usually used when one is addressing or talking about a person of higher status (e.g., an out-group member such as the speakers superior), as it raises thereferents status by increasing the vertical distance between the speaker and the referent.* For example, the question shachoo wa irasshairnasu ka? could meanwill the (company) president In this example, go? one uses the honorific verb form irasshairnasu to describe the actionof the respected president, who, this case, could be either the addressee a thirdin or person referent. Some verbshave honorific lexical substitutes, whereas other honorific verb forms are made grammatically. Nominal referents take honorific suffixes such as -sun or -sarna (Tanaka-san), while many nouns can take honorific prefixes such as o-, go-, and on-to refer to actions and objects related to the respected person(e.g., go-iken opinion,o-sumai residence).A handful of adjectives and adverbs also may take the go- or o- prefixes (e.g., go-shinsetsu kind,o-kireipretty).

Kenjoogo Humble Language

Humble language (nonsubject honorifics) is used to depict the actions or belongings of the speaker or in-group memberin relation to the nonsubject
z In some cases, use of sonkeigo and kenjoogo is unrelated to the absolute amount of respect a person has fora referent, but occurs because the speakerconversing basedon conventional social is norms (Oishi,1974).

Polite Language in Japanese


referent in a deferent manner. Kenjoogo increases the vertical distance between the speaker and the referent by signaling the speakers lower status. For example, in watakushi wa mairimasu I will go, the humble first person pronoun watakushi is used with the humble verb mairimasu to go to describeones own actions with humility.Humble formsappear in verbs (both lexical substitutes and grammatically constructed forms), and a nouns have humble forms few (e.g., chichi for father) or can take humble prefixes, such as gw.3
Teineigo Formal Language

In general, teineigo (addressee honorifics) describes language is formal and that polite, as opposed to language that is informal and casual. Addressee honorifics are used to express the speakers politeness to, or distance from, the hearer,as opposed to honorific-humble language, which conventionally marks the speakers deference to an exalted referent (Kuno, 1987). Niyekawa (1991) described teineigo as involving the issue of style. Forexample, itis possible to speak in informalstyle and yet show respect to the referent.One can say sensei ga irassharu the professor will go in a conversation to a friend, using the honorific predicate irassharu to describe the actionof the esteemed professor, although the predicateitself is in informal style because it being used in a is casual context. of teineigo depends on the situational context determined Use as by factors such as the in-group:out-group relationships of the interlocutors (involving factors such as social position, power, age, and gender), thesocial setting involved, and the nature of the information being (e.g., Ide, 1982; given Martin, 1964; Mizutani & Mizutani, 1987). Teineigo predicates are usually marked with dew/-masupredicate endings, used with verbs, adjectives, copulas, which can be used a conversation and in with ones superiors orin a formal setting, such as an interview or a public speech. The informal -da style would be used in amore casual setting, perhaps in conversations with ones friends. There are also simple lexical choices and personal pronouns with varying degrees of formality (e.g., first person pronouns watakushi, watashi, atashi, boku, ore).
Bikago Beautification Honorifics
Bikago, or beautification honorifics, are used to refine beautify ones or language. They differ from other formsof keigo in that they are not used to express respect for the addressee or referent. Many of these forms are now accepted as the standard form (e.g., o-cha tea).Women, in particular,
3 It is said that kenjoogo is the most complicated formof keigo in terms of the limited kinds of verbs that have humble forms, the syntactic position of the exalted non-subject referent and the pragmatic, aswell as contextual constraints,involved (Mori, 19 93).



frequently use honorific prefixes for refinement (e.g., o-soba buckwheat noodles, o-sushi sushi). Bikago are also common in child-directed language (e.g., p h o n book, o-isu chair, o-uchi house) and are known to appear excessively in the speech of female kindergarten teachers, who use yoochien kotoba preschool language, asin o-juusu juice, and o-kaban bag.

In order to beyond mere linguistic competence use language effectively, go and children must undergo language socialization and develop communicative competence (Hymes, 1972). Politeness is an integral partof pragmatic competence. In order to comprehend and produce polite language, children must master thenecessary linguistic forms, and they must understand the pragmatic rules that govern shifts in register within each socialand situational context (Ervin-Tripp, 1977).As in the case of most studies on pragmatic and sociolinguistic development, much of the research on childrens acquisition of polite language focuses on English-speaking children. This research generally falls into two categories: childrens acquisition of social routines (e.g., greetings and polite expressions),and childrens acquisition of directivesand requests. Such research illustrates that social routines involving ritualized expressions, such as hi, thank you, and please, are the earliest forms of polite speech acquired by children (e.g., Greif & Gleason, 1980). Parents are often persistent in their efforts to socialize their children in the correct of language used in social use routines. Research on the comprehension of requests has shown even by ages 3 to that 4, children are able respond to mitigated requests and hints made peers to by and caregivers (e.g., Garvey,1975,1984; Holzman, 1972; Shatz, 1978). Research on the production requests has found that childrenable to produce a wide of are range of request forms by the time they to 5 years old (e.g., Bruner, Roy, & are 4 Ratner, 1982; Ervin-Tripp; 1977;Garvey, 1975,1984), and that the proportion of mitigated requests increases with between the ages 2 and 5, especially age when children are making requests to someone in power or in possession of desired items (e.g., Ervin-Tripp, 1982; Wllkinson, Wllkinson, Spinelli, Chiang, & 1984). Speaker-hearer characteristics such as age, familiarity, power,and role play a key role in influencing childrens choice of politeness forms (e.g., Andersen, 1978,1990; Ervin-Tripp, 1974; Ervin-Tripp &Gordon, 1986). As most research on childrens acquisition of polite language has focusedon the acquisitionof greetings and requests English-spealung children, by research on childrens acquisition of Japanese, a language with an elaborate politeness system, can help better our understandingof pragmatic development and communicative competence. However, surprisingly enough, thus far, very little research has been conducted on the acquisition of keigo by Japanese children.

Polite Language in Japanese


In general, Japanese researchers, mainly relying data from spontaneous utterances and parental questionnaires, believe that young children can only use informal verb forms and do not productive use of different politeness have registers until they receive formal instruction during elementary school (Mizutani & Mizutani, 1987). Morioka (1973) argued that while children may master reading or writing on their own, polite language has to be acquire within agiven social context, withan understanding of the relative positions of the speaker, hearer, referent, as well asan understandingof the setting and and conversational topic. In other words, the acquisition of proper gengo koodoo linguistic behavioris more difficult than the acquisition of gengo language itself. The assumption is that children seldomuse polite language because they do not encounter contexts in which they need to use keigo, and therefore, children need to be instructed in the of keigo across a wide variety of use contexts. Recently, the need for children to use keigo has been reduced drastically by the increasing tendency for children use the informal -da style to even to speak to elder members of family.4 the Other researchers, such as Bunkachoo (1986), Muraishi (1g73), Murata (19 8 3) and Nakamura(19 g 6,19 g reported that the earliest forms of polite 7) language used by children areteineigo (formal language), referent honorifics (e.g., using the suffix -San), polite expressions or greetings, and beautification honorifics (see Nakamura, 1997, for a full description of the literature on this topic). In general, acquisition honorific and humble forms has been reported of to occur extremelylate (Mackie, 198 3). One possible explanation for this that is the use of humble and honorificlanguage depends on complex interpersonal factors(e.g., uchi-soto in-group vs. out-group, jooge-kankei top-down relations), unlike theuse of teineigo, which relies more on situational context (Clancy, 1986; Nakamura, 1996). Japanese children must learn to function an extremely hierarchical society in by varying their language appropriately according the social context. It was to hypothesized that they would be able to use a large repertoire polite forms of (including honorific and humble forms) appropriately in role-play contexts, which allow children touse forms that may not spontaneously appeareveryin day interactions (Nakamura,199 6,1997). Such data would show that even young children are grappling with the complicated linguistic forms and sociointeractional relations underlying correct keigo usage and also provide us with invaluable insight regarding pragmatic developmentand childrens acquisition of polite language. This study addresses four basic questions: do Japanese children acquire How the four typesof keigo? Why are certain typesof keigo easier to acquire than
4 This is because nowadays intimacy weighsmore heavily than hierarchy in determining style choice in contemporary Japanese families. Not only do childrenno longer use keigo in the home, they are usually not expected older family members to do so either. by



others? What speakedhearer characteristics and contextual factors are children sensitive to when using keigo? How are children socialized to use keigo?

This study is part of an ongoing project to examine aspects of the acquisition and development of Japanese childrens language that reflect pragmatic and social knowledge. Thirty children (3 girls and 3 boys in each age group, ages 1 to 6) were observed in their homes during monthly visits over the courseof 1to 3 years (see Table 5.1).5 The children were recruited from middle-class families living in the Tokyo metropolitan area. Efforts were made to observe the same group of children across a wide range of activities (e.g., role-play, object construction, snacktime) with different interactants (e.g., parents, peers, siblings, unfamiliar adults). Research on English-speaking childrenhas shown that there are strong socioecological constraints childrens use of language on and has pointed to the need for researchers to collect data across different contexts (Cook-Gumperz & Corsaro, 19 7 7 ) .Each visit lasted approximately 6 o to g o minutes. Each session was audiotaped and videotaped, and the data was then transcribed and analyzed. During each visit, for approximately15-30 minutes, toy props were used to encourage the children to engage in pretend play (e.g., store clerk-customer, doctor-patient,trainconductor-passenger,teacher-student).Role-play contexts allowed to the children to act out different hierarchical relationships by assuming a variety social roles in which they were given the opportunity of to use different levels of politeness.6 In addition, questionnaires and interviews targeting issues pertaining to childrens use of language in different contexts and language socialization were administered to the mothers.

Even young children are clearly aware of the importance of polite language. Japanese children are able to use many different linguistic forms to mark politeness and often vary their language usage appropriately according to context. Nonverbal politeness such as bowing routines seem to emergefirst. Even 1-year-olds were able to bow at appropriate times, such when greeting as or thanlung others,or when saying goodbye.
5 Nakamura (1996) was based on data collected from 18 3-year-olds, 4-year-olds and 5-yearolds. The current study includes data from younger subjects 1-year-olds and 2-year-olds) as it (i.e., was discovered in Nakamura (1996,1997) that various aspects of keigo acquisition occur much earlier than previously reported. 6 For example, Andersen (1990) found that children often used polite language when playing the role of a mother in a family setting.


Polite LanguageJapanese in


Age, Gender, andNumber ofTarget Children, and Time Range ofvisits

~ ~~~


Age in Years

Number and Time Gender Range

3 girls l 3 boys 3 girls l3 boys
12 12

of Visits

3 girls l3 boys 3 girls I 3 boys

3 girls I 3 boys

months months 18 months to 3 years 12 months to 18 months 12 months

Aisatsu Greetings and Polite Expressions

As reported previously in the literature (Bunkachoo,198 6), aisatsu emerge extremely early, with children young as 12 months using aisatsu such as as baibai goodbye and doozo please,accompanied with the appropriate gesture (e.g., waving goodbye, bowing). Young children quickly expand their repertoire of greetings, using forms suchas ohayoo-gozaimasu good morning, konnichi wa hello, sayonara goodbye, and polite expressions such asarigatoo-gozaimasu thank you and gomennasai (Im) sorry. Seeing his teacher at the entrance tohis day-care center, Yuuki (z;z, male) said, with bow: a
sensei, ohayoo gozairnasu!

Good morning, teacher!

Similarly,Nao (z;o, male), bowed to thedaycare teacher as he left the classroom when his mother came to pick him up at the of the day: end
sensei, sayonara!

Goodbye,teacher! At an early age, the spontaneoususe of some polite expressions becomes second nature. In example, Ken (z;o, male) had been suffering from one a fever of 1030 for several hours, and his mother had been cooling his forehead with cold compresses. In a daze, whispered arigatoo thank you and gomen ne Ken (Im) sorry. Evenat this earlyage, very young children are to use the right able social expressions spontaneously and appropriately. Aisatsu appear frequently during pretendplay scenarios, in which children assume a wide variety of social roles.For example, while playing store, even 1-year-olds playing the role of store clerks were able to use the polite greeting irasshairnase! Welcome! when greeting customers. Two- and 3-year-olds were able to use more complicated routine expressions, such as maido arigatoo-gozaimam thank you (for your patronage) and mata oide-kudasai please come again.



As reported by Clancy(lg8 6 ) ,mothers and preschool teachersparticularly are keen on modeling suchexpressions and constantly provide direct instruction. In the preschool, children recite their greetings and politeexpressions in unison or in song at appropriate times during the school In fact, the acquisition of day. aisatsu is often explicitly listed by the teachers one of the main goals the as of preschool year. These greetingsand polite expressionsplay a key role in introducing and familiarizing young children to sonkeigo, kenjoogo, and teineigo basic forms, as aisafsu often consist of these different polite linguistic forms. In addition, by learning to use greetings and polite expressions children also learn to gradually comprehend the psychologicalfeelings which underlies aisatsu, such as omoiyari and sasshi consideration for others,and wakimae discernment.
Sonkeigo Honorific-Respectful Language

Children were able to address people with the honorific suffixes -San and -sama (e.g., Hashimoto-san Mr. Hashimoto). Somewere also able to use respectful nominal forms and prefixes (e.g., o-kyaku-sancustomer, o-uchi house,kochira here), but not always consistently. Some children invented creative honorific forms, such as Ryota (4;2, male), who addressed a monster as o-kayuu honorable monster, and Ken (3;0, male), who incorrectly referred old men to as 0-jii-San-not noticing thatthis term already had the honorific suffix -San incorporated in it. Polite request forms appeared frequently, but request forms with honorific predicates were more rare. Emi (4;6, female), pretending to be a doctor talking to ababys mother, said chotto o-machi-kudasai o-kaa-samaplease wait a little, Mother, using the honorific form for mother and apolite requestform of the verb matsu to wait. Ken (3;2, male), pretending to be a storekeeper, said mata oide ni natte kudasai please come again, using an honorific request form his as customer leaves the store. Many children were able to use such honorific request forms when in but somehad difficulty being consistent. role, As mentioned earlier, the acquisition aisatsu seems to play a key role in of familiarizing children with honorific forms. example, itterasshai,said when For someone leaves the house, consists of the honorific form for the verb go, to namely irassharu. Similarly, the expression irasshaimase! welcome!, used by children when playing the role of store clerksthey welcome customers uses as the same honorific predicateirassharu, which, in this case, means to come. Some of the childrenwere already using these expressions as early as 18 months. It seems that Japanese children first acquire expressions that happen to contain keigo forms and thengradually begin to grasp the linguistic rules that underlie keigo usage. At this early stage, childrendo not seem to separate the informational content of the utterance from the form of the expression. keigo With increased exposure to a variety of contexts, children seem gradually to learn to discriminate between different levels ofspeech.

Polite Language in Japanese


Children sometimes made errors in their use of honorific forms by using them todescribe their own actions. example, Akira (3;0,male) pretending For to be a customer making a take-out order on the phone, an honorific used noun form to refer to himself incorrectly:
boku wa


desu san

TOP customer HON HON I am a customer


At this age, Akira probably consideredo-kyaku-san to be the standard form, using it without being aware of the honorific natureof the prefix0- and the suffix +an. Similarly, Emi(4;9, female), pretending to be a doctor, misled by her was mothers use of the honorific form, using it to describe her own actions:
Mother: sensei irasshai- mas-u


POL doctor NONPASTQ be Doctor,are you (here)?


hai *irasshai-mas-u yes be POLNONPAST Yes,I am (here)

Emi used an honorific form of the verb to refer to herself, instead of using a humble or neutral form.After this turn the mother laughed, sayingimasu deshoo you mean imasu, implying that the neutral form the verb imasu of would bemore appropriate to describe Emis own actions. Children also made errors in the honorific forms used to describe their referents actions: Koosuke (4;7):
tsugi wa shuuten de Pozaimasu Next TOP last stop C0P:POL The next (stop) is the last stop oriT rem* kata wa go-chuui kudasai Get off POL people TOP HON attention please (Those) people getting off, please take care

Here Koosuke, pretending to a train conductor, be announced the stop with last an ultraformal copula,and then attempted touse the honorific form orirareru or o-ori-ni-naru to describe the actions the passengers, but did notsucceed of in making the formcorrectly (he did succeed in choosing an honorific form kata for people, and a polite request form). During role-play, children sometimes tried to use honorific forms, but often made morphosyntactic as well as pragmatic errors in their attempts to do so. Obviously the use of honorific forms is complex and difficult for young children acquire. to



Kenjoogo Humble Language

As in the case of honorific predicates, many examples of humble predicates used by the children were in formulaic, ritualized expressions, suchas ojama itashimmu I will humbly impose on you (when entering someones house), itadakimasu I will humbly receive this (stated beforemeals) or itte-mairimasu I will humbly go and come back (stated when one leaves the house). Furthermore, manyuses of humble predicates were variations of set routine expressions suchas Saeko (4;0, female) in Ichikawa to mooshimasu ked0 (This is) Ichikawa (humbly) speaking,used when answering the phone. s in thecase A of honorific predicates, the majority humble predicates that appear in early of child language seem to be those appear in formulaic, routine expressions. that It seems that these expressions serve the role of familiarizing the children to the humble forms, after which children gradually become capable them in of using different nonformulaic contexts. Children often attempted to use humble predicates when playing store scenarios, in which they portrayed the of storekeepers and customers. roles For example, when Ken (3;8, male) pretended to be a customer in a supermarket deciding between apples and tangerines, he said kochira o itudakimasu (Ill) take this one, using the humble verb itadakimasu to takeheceive and the polite form forkochira this one, Ken wasfamiliar with this humble verb through its use in the polite expression itadakimasu, which is used when one is about to receive a mealor starteating. Some children were able to use humble forms in slightly more creative contexts,especially during role-play.Forexample,Mayu (54, female), pretending tobe a telephone operator recording, used the humble verb form tsukawarete-orimasen instead of the nonhumble formtsukawarete-imasen to say that the telephonenumber dialed is currently out of service:
kono denwa bangoo wa tsukaw- are- te-orirnasen
This number phone TOP use PASS PRES.PROGNEG POL This telephone number not being used is

Although the children were able to use humble predicates, there were few examples of humble nouns (e.g., haha for mother). As Clancy (1985) proposed, children seem initially to use ritualized social routines to learn honorific and humblelanguage (especially predicate forms) within a specific context (e.g., ritualized social routines).Over time, with a wider repertoire of social experiences and a more sophisticated understanding the relevant of interpersonal factors, they are to generalize such forms to wider variety able a of social contexts, with increasing productivity appropriateness. and

7 Ken had been using this expression daily since was a 1-year-old. he

Jeineigo Addressee Honorifics

Polite Language in Japanese


By age 3, the children in this studywere able to use desul-masu forms in a context-appropriate manner, and were also able to use polite request forms (-te kudasai). Ken (3;6, male), inviting his visiting grandparents sit on the sofa, to said kochira ni o-suwari-kudasai please sit over here: using an appropriate polite request. Polite requests appeared frequentlyduring role-play. For example, Shin (3;7), pretending to a chef offering an omelet to a customer, said: be
hooku de tetabe- kudasai

fork eat with CONT please Please eat this with a fork In the beginning, young children frequentlyteineigo when they are repeating use adult utterances directly quotingadult or speech.As reported by Muraishi (1973), eventually theybegin to use formal language in order to respond to the of level the interlocutor (usually their mothers). When their mothers teineigo they use attempt to respond with the same of formality.With age, children become level better at initiating of these forms and them more consistently.Here we see use use Mika (3;2), pretending to be working at the counter a fast-food restaurant: of Mika:
irasshairnase...nani ga ii &U ka? welcomewhatSUBJ good C0P:POL Q Welcome. What would you like? nani ga




what SUBJgood wonder What is good? Mika: potato C0P:POL Q How about french fries?
hai poteto


U &



hambaagaa kudasai

yes potato and hamburger please Yes, please give me french fries and a hamburger. Mika:
hai chott0


yes little HON wait please Yes, please wait little a Mika was able to use formal predicate formsconsistently, without prompting, in additionto a polite greeting and anhonorific request form. Occasions in which children code-switch between different levelsof politeness for different purposes are another excellent source of data. Children showed that they were sensitive to factors such as the familiarity of the addressee. For example, some of the children used casual forms for their mothers and polite forms to address the experimenter. Asking if he can play with one of the experimenterstoys, Shin (3;6, male) says:



mother) (to

kore yat- ii?te

This do CONT good Can I do this? (to experimenter)

yat-temo ka ii ...kore? do CONC good C0P:POL Q this May (I) do this?

He used a more formal requestfor permission, and postposingof the object, showing some hesitancy reserve when addressing the experimenter. and Children also code-switch betweendifferent politeness levels, depending on other characteristics of the speaker and addressee. Here,Yuriko (5;4, female), pretending to call her house, told her younger sister to pass the phone to her mother:
mama kawat-te ni Mommy to change IMP

Change with Mommy

A few turns later, she called again, assuming therole of a well-bred lady who is
speaking with her friends daughter (played her youngersister): by
0- kaa- kawatsama teni kudasai HON mother HON to change CONTplease Please change with (your) mother

In this example she used the honorific form mother okaasama, with apolite for request form. Althoughthe two situationsare similar, Yuriko varied her levels of politeness according to the different roles that she and her sister were playing. Contextual factors also play an important role. Many children seemed to associate usage of desu/-masu forms with a formal style of presentation or identification modeof speech. For example, children usually usedteineigo when identifying themselves during self-introductions. Similarly, children often used teineigo in their phoneconversations, especially when they did not know the identity of the caller. Here we see Shin (3;7, male) in a conversation with his mothers friend, who called to speak withher: had
moshimoshi ...Shin-chan h. ..konnichi wa ...i-masu yo. Hello Shin DIM C0P:POL hello be POL NONPAST EMPH

hello . . is Shin (speaking) ...hello ...(she) i s (here). . this

Shin was ableto use desu/-masuforms consistentlyand appropriately. Children realize that certain contexts call for usage of formal language (e.g., interviews, public speeches), and are able to vary theirlevel of formality accordingly. As Muraishi (1973)reported, achilds use ofteineigo seems to be closelyrelated to his or her sense ofaratamari standing on ceremony during formalsituations.

Polite Language in Japanese


Furthermore, storytelling often encourages the use of teineigo. As Muraishi (1973) reported, even young children tend to end their narratives and stories with teineigo. The children in the study usually used teineigo to mark the voice of the narrator, while using da-style informal verb endings to mark dialogue.* Bikago Beautification Honorifics children use many bikago in both polite and casual language.For example, Nanami (3;4, female) said o-soto de asobu no ka naa? Are they going to play outside? when she a picture of two mice carrying knapsacks in her storysaw book. The children used words such go-hon book, o-uchihouse, o-mise also as store, o-ryoori cooking,o-shigoto work,and o-hashi chopsticks. However, in many cases, children had learned o- or go- form as the standard form of the the word and did notuse the word without the corresponding honorific prefix. In other words, theywere not using these forms honorific forms, but simply as as the standard form (i.e., the prefix was not isolated morphologically and productively). In general, girls tended to beautification honorifics more frequently use than boys. This pattern mirrors adult usage. Excessive use of these forms by women, especiallymothers and preschool teachers (e.g., o-ekaki drawing, o-kutsu shoes)hasrecentlybeenthetarget o f muchpubliccriticism (Bunkachoo, 1974). Some women, in their effort to sound refined and polite, overuse beautification honorifics, using unnecessary incorrect forms. and

As bikago are frequently used in child-directed speech,as one might expect,

As researchers such as Clancy (1985)) Iijima (1974), and Muraishi (1973)

proposed, it seems extremely likely that children initially associate certain polite linguistic expressions with a specific context (often ritualizedsocial routines suchasgreetings).Intheearlieststage,theygaininvaluablelinguistic experience by copying dialogues they hear around them (e.g., television programs, adult conversations), making up dialogues and imitating adults (e.g.,
8 This reflects the tendency forchildrens books to be writtenin teineigo, with informal language being the styleof choice for dialogue. In addition, although the children tended to use teineigo when retelling familiar stories (i.e., from books), they tended touse casula language when telling personal narratives. One possible explanation for this is that in the case of personal narratives, children are emotionallyinvolved in the storytelling task, and therefore do not teineigo, use which would serve as a distancing function, creating a more formal mood. This follows into Maynards (1992)theory that the of teineigo during narratives makes the speaker take a narrause tive external position, which the speaker describes the event as an outside in observer, while use of casual forms enables the speaker take a narrative internal position as the speaker were witto if nessing the scene right then and there.



train conductors,store clerks). Gradually they begin to grasp linguistic rules the that underlie these polite expressions, developing a basic understanding the of relevant interpersonal factors, and generalizing such forms a wider variety of to contexts with increasing productivity and appropriateness. Later, formal instruction provides children with concrete knowledge regarding specific linguistic forms.Late acquisition of honorific and humble forms probably is caused by thelexical, morphological, semantic, and social complexities ofthe politeness system. In particular, limited experience a simple network of and social relationships make itdifficult for children to master such forms.9 Only when children leave the protective environmentsof school and home for new roles in society are they ableto increase their proficiency in keigo, through trial and error,as well ascorrection, in new social contexts. Furthermore, unlike teineigo, honorific and humble language rarely appear in child-directed speechand are usually not heard in the home. The parentchild relationship has become hierarchical, and nuclear families are replacless ing extendedfamilies. Previously, with multigenerational families, both adults and children were required to use different levels offormality and politeness in their everyday home interactions (Bunkachoo,1988). Furthermore, changing standards and attitudes regarding polite language (i.e., a greater degree of equality in language usage) have made it even more difficult for children to master the system. For example, recently many school teachers chosen to have avoid keigo usage in their interactions with their students, hoping to base thei student-teacher relationships on intimacy, rather discipline. than In addition, many of the parents the children in this study admitted that of they themselves felt uncomfortable with their ownlevel of keigo proficiency and admitted that much their keigo knowledge had been acquired the of on job, in work-related situations. Interviews with the mothers of the target children revealed that mostof them felt that it was unnecessary for their childrento use sonkeigo honorific language and kenjoogo humble languageas long as they could use teineigo properly. With all these factors, it is surprising that children able to use keigo at all. are Even young children use greetings and polite expressions, beautification honorifics and teineigo spontaneously, in their daily interactions. In addition, they are able to some honorific/humble language in role-play contexts. use Role-play contexts give children the opportunity to engage in hierarchical relationships differing from their everyday roles. Such contexts illustrate that
g Two sociocultural concepts that are said to underlie the of keigo are (a) jooge-kankei use top-down relations, which refersto vertical human relations, such as the relationship between parents and children, managers and employees, or teachers and their students, and uchi-soto (b) in-grouplout-group,which refers to thesense of belonging in a group which exists in the context of the speaker-hearer dyad (e.g., Doi, 1973,Nakane, 1970). Such concepts canbe difficult to grasp, as an individual will often belong to many social groups (e.g., school, extracurricularactivity, family) simultaneously.

Polite Language in Japanese


Japanese children are sensitive to differences in politeness levels and that they are grappling with the difficult task of mastering honorific and humble language. Furthermore, childrenare able to make other more subtletypes of language adjustments to make their speech more polite. During polite speech, they use more hesitation, fewer sentence fragments, contracted forms, and colloquial sentence-final particles. As in the case of other linguistic forms, right after children acquire certain types of keigo forms, they become very rigid regarding the use of these forms, demanding that adults and other children use the right forms at the right also time. This is particularly true of aisatsu and teineigo. For example, Kunta(2;8, male) demanded thathis friend, who had bumped into him, apologize, saying gomen ne-tte itte say youre sorry.Similarly, when Ken (3;2, male) thanked his mother for cooking his favorite dumpling dish, saying arigatoo thank you, his mother responded with a simple yes. However, Ken was unsatisfied hai with this abbreviated response, saying arigatoo no atowa doo-itashimashite-tte iun da yo (you) should say youre welcomeafter (someone says) thank you. Furthermore, parents are often reminded their children tosay the right by premeal and postmeal greetings, namely itadakimasu (said before a meal)and o-gochisoo-sama (said after a meal).Several ofthe parentsalso commented that their children corrected them when they responded with uh-huh instead un of the more proper yes. It is clear that children atthis stage are eager touse the hai right polite forms and also demand that others use such linguistic forms appropriately. Another interesting observation is that children are very sensitive to trends in language use regarding politeness. For example, one common error among the children was use of the humble form, not a marker of humility for the as speaker, but rather as a general polite or refined form. For example, in a store scenario, Nana (3;0, female), asked her customer to wait for her change by saying:
mooshi-wake maarisen.

shooshoo o-machiitadake-masu


Excuse have POL NEG little HON wait receive HON POL Q (Im) very sorry. Would you please wait (humbly) a little? Upon close examination, onefinds that inthis case, the use of the humble form to describe the customers of waiting strays from its traditional usage as a act marker of the speakers humility. Thisreflects a current trend adult language in usage, in which the humble formis used as an overall polite or refined form. Children seem to be extremely sensitive to such trendsin adult languageusage. It is also clear that children aresensitive to factors suchas speaker/hearer characteristics, social context, and topicof conversation. For example, although most of the children usedteineigo in role-play contexts when answeringthe and phone, a few of them used teineigo and other formsof polite language (e.g., polite requests) during rough-and-tumbleplay. The childrenalso tended to use



teineigo in formal contexts, such as in child-researcher interviews and in presentations to theirpreschool classes. As previouslymentioned,Japanesechildren also makespeaker- and addressee-related adjustments,based on factors such as familiarity and age. The children observedin the study tended to casual language forms in their use interactions with peers and younger siblings, while using polite language with unfamiliar and older adults. Research on English-speaking children has also found sensitivity speaker-hearer characteristics such agein the selection to as of politeness forms.For example, Ervin-Tripp (1974) found 3-year-olds used that imperatives more often to children than to adults, and that when 3-year-olds used interrogative requests, they were usually adult-directed. Other studies have also confirmed that young children are sensitive to speaker-hearer characteristics such as age (or size), gender, degreeof familiarity, and role or status (e.g., mother vs. father; Andersen, 1990; Corsaro, 1979; Ervin-Tripp & Gordon, 1986; Ervin-Tripp, OConnor, & Rosenberg, 1984; James, 1978; Walters, 1981). One direction for futureresearch is to identify which factors children are sensitive to in selecting polite language forms. Mizutani Mizutani (1987) and suggested several factors determining the adult usage of keigo (i.e., familiarity, age, social relations, social status, gender, group membership, and situation). The data show that from an early children are sensitive to factors such age, as familiarity, age, and situation. One question is in what order are these factors ranked in childrens form choices?Is this order determinedby the relative importance of social factors within a speech community? Oris it determined by cognitive factors? example, most Japanese children seem to stress familFor iarity over age, by using informal and casual forms to their grandparents. However, familiarity andage seem to be closely intertwined when determining politeness levels, aschildren are more likely to use polite forms with unfamiliar elderlypersons, as opposedtounfamiliaryoungerchildrenorpeers. Furthermore, how is polite language usage related to childrens understanding of social relations and social structure? For example, children seem to find it easier to understand uchi-soto in-grouplout-group relations than jooge-kankei vertical/hierarchical relations. Do the contexts and factors to which children are sensitive change withage? Is polite language usage initially based characon teristics of the interlocutor(e.g., familiarity, age)or the context(e.g., store, school)? These issues need to be examined more closely. Wlth increasing age, children gain greater controlover polite speech and are able to switch between different politeness registers spontaneously in a contextappropriate manner. However, it is also important to note that there are enormous individual differences in childrens use of polite speech, which are closely tied to factors such the attitudes and expectations the childs parents of and teachers regarding polite language Although some mothers provided usage. constant instruction in the of polite language, other mothers only corrected use

Polite Language in Japanese


their childrens use of polite forms when they were addressing other adults (e.g., on the telephone, when greeting guests). is also a well-known fact that It keigo usage varies greatly according factors such as dialect, age, gender, and to class/occupation (e.g., Kokuritsu Kokugo Kenkyuujo, 1964). For example, older speakers are more likely to be more conservative regarding keigo usage, and urban dwellers are known to use more keigo than people living in rural areas. How such factors influence the acquisition do of keigo! Would children of families living in rural areas be less sensitive to keigo usage than children growing up in urban areas?Would children living with their grandparents be better at keigo usage than children growingup innuclear families?

The results of the study show even the youngest Japanese children are that able to use polite linguistic forms in certain and situations. They are to use roles able a wide variety of greetings and polite expressions, bikago beautification honorifics, as well as honorific suffixes such as -San and teineigo addressee honorifics..It seems that Japanese childrenfirst acquire formulaic, routine expressions (e.g., greetings and polite expressions) which happen to contain keigo forms and then graduallybegin to grasp the linguistic rules which underlie keigo usage. During the earliest stages, children do not seem to separate the informational content of the utterance from the keigo form of the expression. These expressions serve the role of familiarizing the children to honorific and humble forms, afterwhich children gradually become capable of using them in different nonformulaic contexts. Data from role-play contexts help to show that young children are often grappling with the complicated forms and sophisticated concepts underlying honorific and humble language usage. Children often make attempts touse polite linguistic forms when playing certain types of roles, such as those of storekeeper (e.g., addressing customers) and mother(e.g., addressing a doctor teacher). or Therefore, it is clear that the prevalent view that most children cannot keigo use forms stems from thefact that they do notneed to use them in their daily spontaneous interactions. When presented with a wide range of hierarchical roles and relationships in role-play contexts, children often show a precocious ability to use sophisticated keigo forms. In attempting to honorific and humble forms, use children make two types of error: errors in linguistic form, stemming from morphosyntactic, semantic, and lexical complexities;and errorsin attributing the honorific or humble form to the wrong referent, stemming from the interpersonal complexities underlying keigo usage. However, as many of these errors reflect the current controversies and changes occurring in the keigo system, it is surprisnot ing to see how difficult it is for Japanesechildren to master these forms. It is not the overall grammatical structure thatmakes polite language so elusive in Japanese, but rather, the combination of the sociointeractional



relations that underlie the wholekeigo system with itslexical, morphological, and semantic aspects. In fact, nonnative speakers often have the most difficulty with such pragmatic subtleties the language, rather than theuse of compliof cated vocabulary difficult grammatical constructions. In order to become and competent languageuser, a childmust master the various linguistic forms that mark different politeness registers and learn when and how to use them appropriately.Withage,aided by bettergrammatical skills andmore sophisticated cognitive abilities, well as a larger rangeo f social experiences, as children gradually expand their repertoire of polite language, becoming increasingly capable of communicating effectively in a wide varietysocial roles. of

I am grateful for the support this research has received from the U.S.-Japan Educational Commission (Fulbright Program), National the Science Foundation,theUniversity ofCalifornia,theAmerican Psychological Association, the Spencer Foundation, the Japan and Society for the Promotion of Science. preliminary version this chapter appeared in Nakamura9 6). A of (19 This chapterwas written whileI was at the Universityof California, Berkeley.

Andersen, E. S. (1978). W1 you dont snore, please? Directives in young childrens role-play spee 11 Papersand Reports on Child Language Development, 15,140-150. Andersen, E. S. (1990). Speaking with style: The sociolinguistic skills ofchildren. London: Routledge. Bruner, J., Roy, C.,& Ratner, N. (1982). beginnings of request. InK. E. Nelson (Ed.),Childrens The language,Vol. 3 (pp. 91-138). New York GardnerPress. Bunkachoo [Agency of Cultural Affairs]. (1974). Kotoba shiriizu I: Keigo [Language series1: Polite language]. Tokyo: Ministry of Finance Publishing. Bunkachoo [Agency of Cultural Affairs]. (19 86). Kotoba shiriizu2 Zoku-keigo [Language 4 : series 24: A sequel to Polite languagel .Tokyo: Ministry of Finance Publishing. Bunkachoo [Agencyof Cultural Affairs]. (1988). Kotoba shiriizu 28: Kotoba no henka [Language series 28: Language change]. Tokyo: Ministry of Finance Publishing. Clancy, P. M. (1985).The acquisition of Japanese. InD. I. Slobin (Ed.), The crosslinguistic study of language acquisition, Vol.I: Thedata(pp. 373-534). Hillsdale, NJ: Lawrence Erlbaum Associates. Clancy, P. M. (1986). The acquisitionof communicative style in Japanese. InB. Schieffelin & E. Ochs (Eds.), Language socialization across cultures (pp. 213-250). New York Cambridge University Press. Cook-Gumperz, J., & Corsaro, W. (1977). Social-ecological constraintson childrens cornrnunicative strategies.Sociology ofEducation, 5515-79. Corsaro, W. (1979). Young childrens conception status and role.Sociology ofEducation, 52, of 46-59. Doi, T. (1973). The anatomy ofdependence.Tokyo: Kodansha. Ervin-Tripp, S. M. (1974). The Comprehension and productionof requests by children. Papers and Reports in Child Language Development, 8,188-196. Ervin-Tripp, S. M. (1977).Wait for me, roller skate. In Ervin-Tripp &C. Mitchell-Kernan (Eds.), S. Child discourse (pp. 165-188). New York AcademicPress.

Polite Language in Japanese


Ervin-Tripp, S. M. (1982). Ask and it shall be given unto you: Childrensrequests, In H. Byrnes (Ed.), Georgetown UniversityRoundtable on Languages and Linguistics. Contemporary perceptions of language: Interdisciplinary dimensions(pp. 235-245). Washington,D C Georgetown University Press. Ervin-Tripp, S. M., & Gordon, D. P. (1986). The development of childrens requests, R. L. In Schiefelbusch (Ed.),Communicative competence:Assessment and intervention (pp. 61-96). San Diego, CA: College Hill Press. Ervin-Tripp, S. M,, OConnor,M. C., & Rosenberg, J. (1984). Language and power in family. In the C. Kramerae, M. Schulz, OBarr(Eds.), Language andpower(pp. 116-135). New York Sage. &W. Garvey, C.(1975). Requests and responses in childrens speech. Journalof ChildLanguage, z,41-63. Garvey, C. (1984). Children\ talk. Cambridge, MA: Harvard UniversityPress. Greif, E. B., & Gleason, J B. (1980). Hi, thanks, and goodbye: More routine information, . Language in Society, 9,159-166. Holzman, M. (1972). The use of interrogative forms in verbal interactions of three mothers and the their children.Journal ofPsycholinguisticResearch,I, 311-366. Hymes, D.(1972). On communicative competence, In J. Pride & J. Holmes (Eds.), Sociolinguistics: Selected readings(pp. 269-293). Baltimore: Penguin. Ide, S. (1982). Japanese sociolinguistics: Politeness womens language. Lingua, 57,357-385. and Ide, S. (1990). How and why women speak more politely in Japanese. Ide &N. H. do In S. McGloin (Eds.), Aspects ofJapanese womens language.Tokyo: Kuroshio Shuppan. Iijima, T. (Ed.). (1974).Keigo o doo shidoo sum ka [How to teach polite language]. Tokyo: Ministry of Finance Publishing. James, S. (1978). Effect of listener and situation on the age politeness of childrens directives. Journal of PsycholinguisticResearch, 7,307-317. Kikuchi,Y. (1996).Keigo sainyuumon [A re-introduction to keigo]. Tokyo:Maruzen. Kindaichi, H., Hayashi,O., & Shibata, T. (Eds.). (1988).Nihongo hyakka daijiten [An encyclopedia of Japanese]. Tokyo: Taishuukan. KokuritsuKokugoKenkyuujo[NationalJapanese Language Research Institute].(1964). Shoogakusei no gengo nooryoku no hattatsu [The development linguistic ability in elementary of school students].Tokyo: Meiji Tosho. Kuno, S. (1987). Honorific marking in Japanese and the word formation hypothesis of causatives and passives. Studies in Language, 11(1), 99-128. Mackie, V. C. (1983). Japanese children and politeness. Papers of the Japanese Studies Center, 6, Melbourne. Martin, S. E. (1964). Speech levels in Japan andKorea. In D. Hymes (Ed.), Language in culture and society (pp. 407-415). New York Harper and Row. Martin, S. E. (1975). A referencegrammarofJapanese.New Haven, CT:Yale University Press. Maynard, S. K. (1990). An introduction to Japanese grammar and communication strategies. Tokyo: Japan Times. Maynard, S. K. (1992). Discourse modality: Subjectivity, emotion and voice in theJapanese language. Amsterdam: John Benjamins. Mizutani, O.,& Mizutani, N. (1983). Nihongo notes. Tokyo: Japan Times. Mizutani, O., & Mizutani, N. (1987). How to bepolite in Japanese. Tokyo: Japan Times. Mori, J. (19 93). Some observations in humble expressions in Japanese: Distribution V (stem) of osuru and V(Causative) itadaku. In S. Choi (Ed.),JapaneseKorean linguistics, Vol. 3 (pp. 67-83. Stanford, CA: CSLI. Morioka, K. (1973). Keigo to kyooiku [Keigo and education]. In S. Hayashi & F. Minami (Eds.), Keigo kooza 7: koodoo no naka no keigo [Keigo lectures 7: Polite language in behavior] (pp. 197-253). Tokyo: MeijiShoin. Muraishi, S. (1973). Gengo hattatsu to keigo [Language development and polite language]. In S. Hayashi & F. Minami (Eds.),Keigo kooza 7:Koodoo no naka no keigo [Keigo lectures 7: Polite language in behavior](pp. 163-196). Tokyo: Meiji Shoin.



Murata, K. (1983). Kodomo no kotoba to kyooiku.Tokyo: Kanebo Shobo. Nakamura, K. (1996). The use o f polite language by Japanese preschool children. In D. Slobin, J. Gerhardt, A. Kyratzis, & J. Guo (Eds.), Social interaction, social context, and language: Essays in honor of Susan Ervin-Tripp (pp. 235-251). Mahwah, NJ: Lawrence Erlbaum Associates. Nakamura, K. (1997). The acquisition of communicative competence by Japanese children: The development of sociolinguistic awareness. Unpublished doctoral dissertation, University o f California, Berkeley. Nakane, C. (1970). Japanese society. Harmondsworth Penguin. Niyekawa, A. (199 I). Minimal essentialpoliteness: A guide to the Japanese honorific language. Tokyo: Kodansha. Oishi, H. (1974). Keigo no shikumi [The structure of polite language]. In Bunkachoo [Agency o f Cultural Affairs], Kotoba shiriizuI: Keigo [language series I: Polite language] (pp. 25-36). Tokyo: Ministry of Finance Publishing. Oishi, H. (1983). Gendai keigo kenkyuu [Contemporary keigo research]. Tokyo: Ootsuki Shoboo. Okuyama, M. (1983). Gendai keigo jiten [Contemporarykeigo dictionary] (Vol. 5). Tokyo: Tokyo-do Shuppan. Shatz, M. (1978). On the developmento f communicative understandings: An early strategy for interpreting and responding to messages. Cognitive Psychology, 10,z71-301. Walters, J (1981).Variation in the requesting behavior of bilingual children. InternationalJournal of . the Sociologyof Language, 27,77-9z. Wllkinson, L., Wllkinson, A., Spinelli, F., & Chiang, C. (1984).Metalinguistic knowledge o f pragmatic rules in school-age children. Child Development, 55, Z I ~ O - Z I ~ O . Yoshioka, Y. (1995). Keigo koodoo to kihan ishiki- Hichiku hoogen-iki ni okeru gengo koodoo hoosa kara [On the relationship between consciousness of norms and honorific expression in Japanese: from the language survey of Hichiku dialect]. Kokuritsu Kokugo Kenkyuujo Hookoku 110. Tokyo: Shuei Shuppan.

Interactional Processes in the Origins of the Explaining Capacity

Universite Nancy


and Universite Paris V - C N R S

One importantpragmatic slull necessary to handle interpersonal relations to is provide explanations and arguments forones own action, demands, or point of view. This chapter addresses the this capacity starts to be functional in way early mother-child interaction and the processes that might underlie its emergence. Previousstudies have shownthatveryyoungchildrencanprovide explanations and justifications their behavior, particularly other-directed of behavior such as requests and refusals, before they have acquired specific linguistic means, such as the connectivebecause, to express them (Bloom & Capatides, 1987; Dunn 1988; Dunn &Mum, 1987; Hood &Bloom, 1979). In a longitudinal study of four mother-child dyads, we have shown that this capacity emergesduring thesecond half of the second year, when children are in the late single-word period (Veneziano Sinclair, 19 9 5).Childrens early & explanatory and justifyingactivity, though still quite primitive in nature,is particularly interesting to study. It provides insights not only into childrens ability to establish links between their behaviorand thereasons for it, showing early interest in psychological and social causes (see also Dunn, 1993), but also into their capacity to express them verbally, using languagein a displaced and informative way. Moreover,itprovidesinsights into childrensgrowing pragmatic, conversational, and social abilities. Indeed, explanations and justifications, if well placed in the conversational exchange, may play an important role in the regulation interpersonal relations, because they may be of used adaptively to influence the interlocutors intentions andbeliefs so as to modify hisor herbehavior. In this sense, the study of the emergence early and development of the capacity to provide explanations and justifications adds crucial informationon childrens early adaptive capacitiesand on their understanding of someone elses mental states.



One particularly revealing case is that of oppositions,that is, casesin which contrasts arise between the desires, intentions,beliefs of children and their or interlocutors: One partner intends to do something, makes demands on the other, or asserts something, and the other makes his or her disagreement known. Such a pub1iccontrast (EisenbergGarvey, 19 81, p. 151) needs to managed, & be and that is where explanations or justifications might interveneas an adaptive interactional strategy to redirect the relationship tohave onespoint ofview or accepted with greater likelihood the other.In conversational analysis, explaby nations-as reasons, motivations, causes of the participants and actions-have been considered as a normatively required feature actions that are unexof pected or unlooked for (Heritage, 9 9 0 , p. 35).Studies of children their 1 in third and fourth year observed in thefamily environment show that childrens use of justifications in conflictual situations with their mothers increases with age (Dunn & Munn, 1987; Dunn, Slomkowsky, Donelan, & Herrera, 1995; Eisenberg, 19 g 2 ; Tesla &Dunn, 19 92) and suggest that justifications may indeed be determinant in convincing othersof the validity ofones position. Similar phenomena are reported for interactions among siblings (Dunn &Mum, 87; 19 Dunn et al., 1995; Tesla & D u m , 1992) and peers (e.g., Eisenberg& Garvey, 1981; Goodwin Harkness,1990; Pontecorvo, 1990; Shantz, 1987). Very few studies have looked at thesespecific phenomena in mother-child interaction with children younger than age 3 (Dunn & M u m , 1987; Haight, Garvey, & Masiello, 19 g 5; Kuczinski, Kochanska, Radke-Yarrow, Girnius& Brown, 1987; and, for one data point at2 ; 9 , the studies by Dunn and her collaborators cited earlier) even less for childrenin their secondyear ( D u m and & Munn, 1987; Kuczinslu et al., 1987). Moreover, none of these studies provide fine-grained longitudinal data on individual dyadsbut mostly pooled dataon relatively large samples. Because this chapter focuses the processes underlying the emergenceof on childrens explanatory capacity, it provides both fine-grained longitudinal data on individual dyads and data concerning the early periodsof development. It moreover concentrateson justifications of naturally occurring oppositions which, like justifications of requests, are among the first justifications appear to in childrens production (Veneziano & Sinclair, 19 g 5). In particular, theeffect of explanations and justificationsis investigated before and around the time children start to provide justifications, by comparing theway the dyad resolves contrasting interactional episodes in which one partner justifies his or her opposition to those in which she he or does not justify it: Does the childs provision of an explanation or justification of his or her opposition to the mother lead the latter give up her previous to action, demand, or statement more easily than otherwise? And, reciprocally, is the child sensitive to the mothers providing explanations or justifications of her oppositions?Moreover, when the contrast is not resolved immediately on the third turnby the partners renouncement of the originalproject, and the

Origins of the Explaining Capacity

11 5

initial contrast becomes thus an open conflict of intentions, howdo children go about insisting on their original intention or statement after the mothers opposition? And how do they persist opposing the mother case the latter in insists on her original intention or statement? Do children show adaptive behavior by modifying their formulation by adding an explanation justifior or cation to strengthen their case? The pattern of developmental resultsis then discussed in relation to what it can tell us about the interactionalprocesses that underlie the emergence the of explanatory capacity i the child. Itis argued that although these phenomena n are inherentlysocial and require social and communicative skills, their acquisition is best understood within a constructivist approach requiring the childs internal elaboration of the jointly constructed mother-child episodes.

Subjects and Method of Data Collection The data presented here come from the longitudinal studies two motherof child dyads-called Dyad A and Dyad C-observed at home during naturally occurring interaction. The language spoken at home French. During the was period under analysis the child inDyad A, a boy, was aged between 1;8.16 and 2;7.8; the child in Dyad C, a girl,was aged between ~ 3 . and 2;2.6. The two 2 families were of middle-class social background. Both children are second born to anopposite-sex siblingabout 3 years older.The boy was going to a kindergarten in the mornings andstayed with his mother in the afternoons; the girl did not attend any day care. the observational sessions were carried out, When the mother was available for the child and the oldersibling was absentfrom home. After some preliminaries, the sessionswere audio- and video-recorded during approximately one hour, and were held, whenever possible, every two weeks. In both studies, two observers2 were present during thesessions, one filming and the other sitting in a peripheral place in the room, taking notes and assuming a friendly nonintrusive attitude. For the boy, this situation was but maintained during the whole observational session; for the girl, during the second half hour, the observer took a more role. active The sessions included various types of play activities (block construcfree tion, ball playing, ritual games, manipulation of objects), book reading, spontaneous symbolic play and, sometimes, snacks around the kitchen table. Table 6.1 presents the number of sessions analyzed-corresponding to a total of 16 hours of recording forthe boy and 15 112 hours for the girl-and the age of the child at each recorded session.
1 For the girl, the last session, 4 months from the previousone, is an extrasession whose analysis is included here. z The authorwas one of them i both studies. n



Number of Sessions and Ageofthe Child at Each ObservationalSession

Session Number
1 2



Age of the Child at Session Dyad A C Dyad

181 ;.6 1;8.30 1;9.20 1;10.4 11.7 ;01 111 J. 11.5 ;11 132 ;. 131 ;. 1;4.0 1i4.14 1;4.26 13.9 1i5.23 164 ;. 1;6.22 176 ;. 171 ;.8 183 ;. 181 ;.5 &9.3 1;10.12 2;2.6

3 4

5 6
7 8 9
10 11

211 4.5 2p.14 2;2.28 231 ;.2 2;4.2 2;4.23
W.27 278 ;.

1 3 14 15 16

Data analysis

The data fromeach dyad was analyzed independently by persons. All the two naturally occurring oppositions from the childs and from themothers side were identified in the video recordings, and their interactional coursewas followed until the resolution of the contrast, namely, until oneof the partners accepted explicitly, or implicitly (letting go, changing the topic or center of attention, withdrawing from the situation), the others position (action, demand, statement, or opposition), a compromise was found and mutually or accepted. Both verbal nonverbal cueswere used to interpret the opposition and episodes and the sequences thus identified were carefully transcribed. They were then segmented into individual movements and entered on an analysis sheetrepresentingthestructureofthesequence while preservingthe information about its content. In most cases, this step required the repeated slow-movement viewingof the audio-visual material.
Definitions and Examples
Defining Criteria for the Occurrence an Opposition. Much of the general of framework and some of the specific categories found and distinguished here can

Origins of the Explaining Capacity


also be found in other studies oppositions andof conflicts (see, in particular, of Dunn & Munn, 1987; Eisenberg 8 Garvey, 1981). Although authors differ concerning the defining criteria conflict or a dispute, agreethat an opposiof a all tional move is necessary. Some include in their analyses of conflictsall occurrences containing a first opposition, considered already to be adversative, conflictu episodes (e.g., Dunn & M u m , 1987;Eisenberg, 1992; Eisenberg & Garvey, 1981); others limit the extent the termconflictto episodes containing of two oppositional moves (e.g.,Haight et al., 1995;Laugaa & Brossard, 1997; Shantz, 1987). In this studywe use the term opposition to refer to thesecond move of a sequence in which of the interactional partners one explicitly manifests dislike, disagreement, aversion,or any other behavior that is contrary to the immediately previous action, real potential proposal, demand, or statement the or of other interactional partner. cases of opposition have been considered. For All this study, however, negative answers to real yes/no questions as well as to sincere demands for confirmation or clarification concerning the interpretation of the childs previously manifested intention, cases in which the opposition to the other part of a mutually agreed game, and instances is of passive opposition in which one partner doesnt respond or attend to the others request for action or attentionwere been considered oppositions. They have been excluded because these responses areambiguous in the sense that they might not signify opposition: They might manifest lack of attention, or a lack of availability of any lund action other than the one the of child is engaged in, and so on. Given their ambiguity, we think that they should be dealt with separately from clear cases of oppositions and conflicts (see, however, Tardif, 1998, for a different stand this matter). on Following Eisenberg and Garvey (1981) the opposing partner is referred to as the opposer, and theopposed partner,as the opposee. The opposers and the opposees moves constitute the minimal pair required for an opposition to take place. This is not considered, however, to constitute a conflict a dispute or (if it is simple, as Dunn & Munn, 1987, call it), but simply the opposition of one partner. Indeed, although it often assumed that the is opposees initial move reflects an underlying clear-cut intentionon his or her side, itisnt until the third move of the oppositional episode (i.e., until the opposees second move) that the extent to which he or she cares about the initiating action, demand, or proposition becomes clearer. Thus, from the observers point of view, it is only at the time the thirdmove that an overt divergence of intenof tions or opinions may appear and a conflict be declared.

Ways of Expressing the Opposition. An opposition may be expressed verbally and/or nonverbally. Verbal oppositions may explicitly contain the negative particle non o , n accompanied or not by other verbalizations. When they do not contain th negative particle, they may be expressed in one of the followingways:



by a negative sentence: for example, tire pas dont pull, said while the child ne is trying to remove the hat from a dolls head; maman nenleve pas les souliers mommy doesnt remove shoes, said after the child has requested the mother the to remove her shoes;moi je vois pas une maman qui boit le biberon I dont see a mommy that drinks from the bottle, said after the child has said#Jbibon maman mommy# bottle; by a sentence which, by its content, goes againstor hinders the action or proposition of the opposee: for example, attends mettre les chaussettes dabordwait, on vu we are going put the to socks first, said while child tries to slip her feet into her the shoes; ilfaut le hisser you should leave it, said after the childs requestremove to the toy cradlescurtain; by a sentence introduced an adversative marker like but: for example, by mais mais cest l jaquette du cloun but it is the clowns jacket, said while child is a the trying to slip the clowns jacketon her rocking horses handle; by expressing directly the reason(s) for the opposition: for example,est trop elle petite cette baignoire its small this bathtub, while the child is trying to place too a relatively big dollinto a toy bathtub; cest tout cousu itsall sewn up, refusing the childs requestto remove the dolls shirt. The availability of videotaped material has allowed us to consider also nonverbal oppositions. To be included, nonverbal oppositions must present an active behavior of resistanceandlor an action that in the opposite goes direction of that requested or executed by the opposee. For example, the mother holds the child by the hips preventing her from going to her father, after she had got up saying papa.
Different Types of Opposition:Protests, Refusals, Denials, and Prohibitions.

Authors worhng specifically on oppositions often make distinctions relative to the content or topic of the opposition (see,e.g., Dunn 8r Munn, 1987; Eisenberg, 19 Haight et al., 19 5). 92; 9 Another way of distinguishing oppositions is consider the events that to trigger the oppositions. In this relation, Eisenberg (1992), for example, distinguishes four typesof speech acts: two kinds of request andtwo kinds of statement. In her analysis of negation in general, Bloom (1970) found a developmental progression in the semantic categories of nonexistence, rejection, and denial. Oppositions, as an interactionalcategory, can fall into the last two. For purposes of our developmental study, we thought itof interest the to distinguish oppositions according whether theevents triggering them to were expressed verbally or not, or constituted language events in themselves. Accordingly, we distinguished them into protests, refusals, denials, and prohibitions. Protests are oppositionsto events that are present in the situation
3 The symbol # means the presence of a pauseof between .5 and 1 second; the symbol,##l hn pause longert a 1 second andup to 2 seconds.

Origins of the Explaining Capacity

11 g

and that have been brought about by the others action; for example, the child says non and pushes back the hando f his mother, who had pressed one o f the buttons o f his taperecorder. refusals are oppositions toevents that are not in the situation but are expressed by the others verbalization formo f directives in the (orders, different kinds o f request), o f proposals for activities involving the addressee or of announcements of intentions involving the speaker; for example, the child says non, refusing the mothers suggestion togive the doll something to drink; the mother refuses the childs request to open a bottle, saying ah non on va pas louvrirah no we are not going to openit. Denials are oppositions to the others statements about events, which may or may not be present in the situation; for example, the child denies the mothers statement Fa cestfaitthis is done, talking about the dolls skirt, bysaying non # sepafesa no # its not done this; or the mother denies the statement ton lityour bed, childs talking about one o f the beds in the room, byreplying rnais cestpas rnon lit Fa but it is not my bed this. Prohibitions are a kind o f negative directive, or control act (Ervin-Tripp, Guo, & Lampert, 1990), a sort o f preemptive opposition to the others likely action. Along the dimension discussed here, they can be seen temporally displaced protests that, contrary to the latter, can as be performed only verbally. In our sample they are usually produced to prevent the repetition o f an undesirable action that has just been performed; for example, after the child has broken one f the toy cradles, the mother says tu o sais tu dois pas les casseryou know you must not break them. Justijications of Oppositions. Oppositions were distinguished according to whether theywere accompanied by a justification. A justification consistsa of statement that provides the reason (theexpzanans) for the opposition which, fromaninteractionalandpragmaticperspective, is supposedtobean explanandurn (an act that needs to be explained; Barbieri, Colavita, Scheuer, & 19 9 0 ; Berthoud-Papandropoulou, Favre, & Veneziano, 19 9 0 ; Veneziano & Sinclair, 1995). Logically, these kinds o f explanation are analogous to the regressive typeo f argument (Grize, 19 9 6 ; Schlesinger, Keren-Portnoy, & Parush, in preparation) in which the conclusion (here, the object o f the opposition, or the explanandurn) is put forward first and the argument(s) is (are) provided later in its support. Since we were dealing with the first manifestations o f explanations and justifications, operational criteriawere two retained (Veneziano, 1990; Veneziano & Sinclair, 1995): (a) the explanandurn may be marked in a nonverbal way, whereastheexplanansneedstobe verbalized4 (hence the possibility that oppositions may be nonverbal and that their verbal expression may contain only the justification); verbalization @) the o f the explanans is addressed to the interlocutor and cannot be confounded
4 This criterion may be judged too restrictive. Thus, Rojo-Torres (forthcoming) argues that nonverbal explanantia are producedby 3-to 4-year-old deaf children in peer interaction.



with the verbalization of an ongoing action or event. This latter criterion excludes ambiguous cases in which it is unclear whether childs production the is the verbalization of an ongoing actionor the justification of his opposition. The following are all examples of justified oppositions. each of the types For of opposition distinguished earlier provide one example produced the we by mother and one produced the child. by Examples of justified protests: In Example (1) the protest is produced by the mother, in Example (2) by the child: Mothersjustifiedprotest: The child is pulling the of a perfume bottle. The mother says #on p u t top non pas enlever # Fa cest impossible# sinon on le casseno # it cant be removed it is # impossible # otherwise it breaks. The child the bottle puts down and covers the top with a second cap, accepting the mothers protest. The mother justifies her protest appealing first the property of the by to topimpossible to remove-and then to negative consequence of childs action-the a the top breaks.

Childs justified protest: The mother places a doll (mother figure) in a toy chair. The child protests action by sayingnon # Id le papa no there thefather. The mother accepts by re# moving the doll she had placed letting the child put his doll there (the father and figure). The child justifies his opposition by stating a different plan, alternative to th one of the mother-not the mother, the father.

Examples of refusals accompanied by a justification:


Mothersjustified refusal: The child tends a toy to her mother saying more (request for more cup ko water). The mother refuses by saying tas assez bu Cno, you have non, drunk enough C. The mother justifies her refusal making an evaluation of what the child ha ready drunk and judging it sufficient, implying thatfurther water is required. no The child accepts implicitly as she abandons her initial request. Childsjustifiedrefusal: The mother suggests that the child use the water that is contained in a small bottle sayingId tas de [eau # Id dedansthere you have some water# in there. A A The child, holding toy bathtub, looks back at his mother says pas assez not a and enough and then goes the bathroomto get some water. The mother accepts to implicitly by letting carry out his action. him The child justifies his refusal with an evaluation, an estimate of water he the needs which is more than thatproposed.

Origins of the Explaining Capacity


Examples of denials accompanied by a justification:


Mothersjustifieddenial: a at to the Mother and child are looking picture book. The child, pointing image of a truck, says monsieur dedansman inside. The mother non # y a pas says de monsieur # F cest le moteurno # there is no man that is the motor. a # a different identity for the reThe mother justifies her denial by specifying ferred object. The child accepts her interpretation and starts talking about the motor.
(6) Childs justified denial:

The mother, taking the lead from the child, who was goingto make a said he joke, says jokingly aprh on va chercher la blague later we are goingto look for the joke. The child immediately says cestpas une blague# cest mes chosesno it is non not a joke # its my things. The child justifies the denial specifymga different identity for the referred by entity. Example ( 7 ) provides an illustrationof a mothers prohibition accompanied by a justification: Mothersjustifiedprohibition: The childlooks with interestat the place where cassettes are inserted into his Fisher-Price taperecorder and the mother says ld il fautpas mettre les doigts mais parce que ld tu vas tefaire mal but you must not your fingersthere because put there you are going hurt yourself, and then continues: ici ya va # Id but here to mais its ok# there, showing another part the taperecorder. The child accepts and of touches the place shown by the mother. The mother justifies her prohibition by bringing childsattention a negto the ative consequence of the prohibited action-to hurt himself; then she adds what the child cando instead.

Unfolding of the opposition sequence. Three main types of unfolding have been distinguished here. Thefirst one consists of third-turn resolution: The opposition is resolved immediately in the episodes third move where the opposee gives up explicitly (by complying verbally and/or in action) or implicitly (e.g., by abandoning ones position, not insisting, changing the subject) the initial position (action, demand, or statement). Examples to ( 7 ) are all (1) instances of third-turn resolution, where itis the opponents stand or will that prevails immediately, before anyclash is overtly declared. In thesecond typeof unfolding, the resolution the oppositionis delayed. of In the third turn, the opposee opens intermediary sequence, producing an an intermediary sequence move (ISM), that is, a neutral move signaling time of a halt, consisting often a clarification or confirmation request (comparable of to Garveys, 1977,side sequence). Example (8) shows an episode in which it is



the mother who opens an intermediary sequence, Example(9) an episode in which it is the child who does so:
(8) The mother opens anintermediary sequence: MI: on le met assis l sur le garage d

Ms initiating move

we seat him there the garage on (seating a clown on the top part a toy garage) of C1

childs protest
Ms ISM: clarification

(removing the clown from the garage)

Mz: non?

Cz: /ekun laba/

the clown there and thengoes towarda pile of toys, on the floor
M3: ah cestlrt bas #pour le clown! ah it isthere # for the clown

C clarifies, providing a justification of his protest (a different place forthe object) M explicitly accepts the Cs position

between mother andchild. The child In this example there no overt contrast is manifests his disagreement towards the mothers previous move but, in her third turn, the mother doesnt insist on her initial intent. However, the resolution of the opposition is delayed by the mothers request for clarification or confirmation of the childs opposition, followed by the childs clarification that provides at the same time simple justification for his previous protest(a a different place forthe clown). After the intermediary sequence, the mother explicitly accepts the childs position.

The child opens an intermediary sequence: Cl: 000 maman mommy warm childs initiating (to) move (looking at mother after having placed a baby dollinto a toy cradle)
MI: il n b pas tellement chaud Id, il na pas de couverture he is not so warm there,
Ms justified denial

he doesnot have any blanket Cz: hein ? Mz: il na pas tellement chaud Id
il a pas lespieds chaud au

Cs ISM: a clarif. req. M clarifies by repeating and adding justification

he is not warm there, so his feet are not warm warm) (to C3: (C doesnt insist and changes subject) C implicitly gives up her idea Here the resolution of the opposition is delayed by the childs request for clarification of mothers denial, followed by the mothers reinstatement for her the

Origins of the Explaining Capacity


previous turn (a negative property of the object), to which she adds another justification, and then the childs implicit acceptance of the mothers proposition. by The thirdtype of unfoldingis identified as a conflict: In the third turn, the opposee insists on his or her initial intentionor position, thus producing a conflictual move (CM). It is the production of such a conflictual movethat signals the presence two overtly different intentionalor belief states among of the interactional partnersand that opens an overt conflict. Example (IO) shows an episode in which it the mother who opens a conflict, is Example (11) an episode in which is the child who does it: it

The mother opens a conflict MI: on essuie les miettes? we wipeout the crumbs? (M approaches the childs bib her mouth) to Cl: non protests (shakes her negatively) head refuses Mz: on essuie les miettes? wipe we out the crumbs? (Ms first (M wipes childs mouth with the bib) the Cz: (grasps the pulls bib it and away) insists child M3: tu me la donnes ? you giveit to me? (pulls gentlythe bibto her side) c 3: nhnh (holds on the bib) M4: tu me l donnes a you giveit tome (pulls gentlythe bib to her side) c4: nhnh (holds onto the bib)
M5: tu la donnes d maman

Ms initiating move

C insists Mother

Ms action,
Ms proposal


mother insists

child insists mother insists

child insists mother insists

you giveit tomommy (pulls gentlythe bibto her side)



child insists

(holds onto the bib) M6: mh? tu me donnes? insists mother mh? you give tome? it (pulls gentlythe bibto her side) C6: nhnh insists child (holds on the bib) M7: back on the (sits chair and abandons) gives implicitly M in After the mothers first conflictual follow five moves by the childand four move additional moves by the mother, in which the two partners insist theirown on



positions: The mother wants recuperate the bib wipe thechilds mouth to to and the child wants to prevent the mother from accomplishing this. The conflict is finally solved by the mother, who abandons her project.

The child opens a conflict Cl: peux mettre # le tapis Id? Cs initiating move can put # the carpet there? (the child touchesthe carpet under the bed) MI: on vu pasle mettre maintenantA Ms refusal, with we arenot going to put it now A temporal delay child insists: Cz: oui yes Cs first CM M2: non mother insists child insists, providing CS: j k i envie moi a justification I feel like it M3: tu pourrais laver le bCbt M insists, proposing you could wash baby the a different activity C4: accepts mothers proposition and gives up C givesin implicitly the initial idea the carpet of

In this case, after the mothers refusal the childs permission request, the of child insists, starting an overt divergence of intentions. Themother repeats her refusal and the child insists with his original intention by providing a minimally distant justification. The conflict finally solved by mothers strategy to is the divert the childs attention with an alternative proposal and the childs acceptance of it. Intermediary sequences may sometimes arise after a conflict has been declared, and aconflict may arise after intermediary an sequence.
Coding. Opposition episodeswere transcribed and analyzed for structure, as follows. Each episode was coded for (a) the presenceabsence of a justificaor tion accompanying thefirst oppositional move, and (b) the pathway to the resolution of the opposition, namely, whether the opposition is resolved immediately in the third turn, after an intermediary sequence, or after conflictual moves or a mixture of conflictual and intermediary-sequence moves. It was also coded for type of opposition type of justification. When and an intermediary sequence, a conflict, or a mixture the two occurred, the of unfolding of the sequence coded forclues for assessingthe occurrence of a was resolution, and who gave in or whether a compromise reached. was

Developmental Periods
Alhough theage ranges during which the two children were studied are shifted by five months relative to each other, some developmental similarities can be detected during the first 10 months o the study. Indeed, we have been able to f

Origins of the Explaining Capacity


distinguish three corresponding periods: first period (from 1;8 to 1;lofor the a boy; from1;3 to 1;6 for the girl) in which the children produced justifivery few cations in general; a second period (from 1;lo for the boy; from to 1;8 to Z;I 1;6 for the girl) in which justifications produced more regularly appeared were but more as justifications of requests and of the childs own action; a third and period (from 2 ; to 2;4 for the boy; from 1;9 to 2 ; for the girl) in which a ~ ~ sizeable increase in the productionof justifications was observed, particularly in the justifications of oppositions.The data are thus presented according to these three developmental periods, to which has been added, for Dyad a A, fourth period covering the 2 months of the longitudinal study. last Mothers and Childrens Oppositions: Number and Justification First Oppositions and Types of

A total of 591 opposition episodes was identified and analyzed in the two studies, 254 for Dyad A (122 produced by the child 132 bythe mother) and and 337 for Dyad C (164 produced by the mother and 173 by the child). As can be seen in Table 6.2, during thefirst two periods both children justified their first oppositions very infrequently: Until 1;8.15 only four such justifications were observed for the girl, and until 2;1.15 only nine for the boy. A change was observed at the third period, starting at for the girl and at 1;9.3 2;2.14 for the boy, when respectively 56% and 48% of the first oppositions accompanied were by a justification. By contrast, from the beginning the observations, mothers justified their of first oppositions frequently and always to a greater extentthan thechildren: the girls mother in Dyad C, between 62% and87%; the boys mother inDyad A, between 59% and 81%. The difference is important and significant for the three periods forDyad C [ x 4 2 x 2) = 47.20, 63.20, p ( .OOI, respectively at ( Periods I and and 8.55, p <.01, at Period111, all for df= 11, and at the two 11, first periods for Dyad A (x2 (2 x 2) = 19.96,22.38, at Periods I and respectively, 11, p ( .oo1, df= 1; for Periods I11 and IV, n.s.). These results show that mothers ( justified their oppositions frequently before their children started show a well to similar behavior more than sporadically. Concerning the number oppositions, thetwo dyads presented a different of profile. In Dyad A, the mother opposed the child more often than the child opposed her during the two periods, whereas the reverse was first true during the last period, when the childseemed to be in a negative phase, frequently refusing the mothers initiatives. In Dyad C, on the other hand, the child opposed the mother more often than the mother did the child during the first period, w the reverse wastrue during the third period. Both dyads went through an intermediate period, during which the number oppositions evened out. of Concerning the typesof opposition (see Table 6.3), during thefirst two periods, more than three quarters the boys (Dyad A) oppositions were of

5 @

" N N


f l ?

0 C


0 . * . .' -


cc 0



Origins of the Explaining Capacity


protests; they decreases proportionally over time, while refusals increases correspondingly, andin period IV were more numerous thanprotests. Denials and prohibitions which, though numerous, were produced by themother less since the first period, appeared later in the childs production: denials in Period I11 and theonly two prohibitions in Period IV. For Dyad C, protests were the most numerous child oppositions during the first period, butalready during the second period refusals became the most frequent and continued to be during the third period. The data from this so child confirm the later emergence of denials, which appeared in Period I1 but were very infrequent throughout, and prohibitions, which of were not observed for this child during thestudy. These results suggest the existence of a developmental trend, according to which verbal representations acquire an increasing force and psychological reality of their own necessitating, as such, to let the other know what she that or he envisionsis not going to take place or that averbally created reality differs from a perceptual action-created one. It another facet ofthat same capacity or is that is considered to be at the base of childrens displaced, informativeuses of language (see also Veneziano & Sinclair, 199 5).
Justifications During the Unfolding of Opposition Episodes

The explanations and justifications considered earlier Table 6.2) are offered (see spontaneously with the first opposition, and are thus offered without any solicitation from the partner. mother and Do child producemore justifications of their oppositionsor of their insistenceon the initial intent, later on in the exchange, during conflictual or intermediary sequences, following a request fo justification or for confirmation or clarification? Indeed, justifications were produced later in the exchange but they were less numerous thanthose offered immediately. For Dyad A, these constituted 38% of all the justifications produced by the child in opposition episodes and of those produced by 24% the mother; for Dyad C, they constituted 10% of all such justifications produced by the child (all of them at the last period) and 27% of those produced by the mother. However, not all of these later occurring justifications were produced after a request for explanation arequest for confirmationor or clarification by the partner. Dyad A, only 4% of the total number justifiIn of cations of oppositions produced by the child occurred after the mothers request for explanation and after her request for confirmation clarifica11% or tion, while the corresponding figures for the mother are in both cases. In 1% Dyad C, none occurred after a request for explanation, and of the total 7% justifications were produced after a request for confirmation or clarification, by both the mother and the child. It should be noted that in oppositional episodes, veryfew explicit why explanation questions are found. Dyad A,a total of three questions were In why

h 0

= :

W 0 h * r n



observed for the child (in the last two periods), two of which triggered explanations introduced by parce que, and a total of 4 were found for the mother, three of which triggered explanations introduced by parce que (see the later sectionon Form of Explanation). Dyad C, only fourwhy questions were In found, all produced by the mother, nonefollowed by an explanation. Thus, at least in the developmental period under study, the great majority of explanations and justifications occurring in oppositional episodes offered were spontaneously, not only by the mothers (98% and 93% of the justificationsof oppositions forDyad A and Dyad C, respectively),but also by the children,who produced them mostly to justify their first oppositions most often without and any prompt from the adult (85% for theboy, and 93% for thegirl). This finding suggests that the processes underlying the acquisition this pragmatic capacity of are not based on conversational functioningof the simple adjacency-pair type.
Types of Explanation

Concerning the types of explanation offered (the explanantia), most o f those provided by the children take a minimal distance from the opposition they as simply state a different intention, willingness, or opinionconcerning activities, agents, objects, their state and their location, alternative to that acted out, requested or proposed by the mother. Eighty-six percent of the boys and 79% of the girls justifications and explanations fell into this broad category. Mothers offered instead a wider variety justifications, having dowith properties of of to the objects, physicalimpossibilities,consequencestobeavoided,rules, evaluations, or internal states. When mothers provided justifications during th unfolding of the episode, they, at times, provided interesting embedded and revolving chaining of explanations,as in Example (12):

Cl: nh nh (hands a baby doll to the mother, understood as a request undress the baby) to M. justifies her refusal M1 tu le laves avec la robe on narrive invoking the physical pas rt lenlever tu vois impossibility (reasonof you wash him with dress, we cant the satisfying the request take it off Cs ISM: reflection halt Cz: (looks at the dressed baby in her hand invokes the property of Mz: elle est cousue la robe C , the obj, explaining the on peut pas lenlever physical impossibility it is sewn the dresswe cant take off C, it given inM1 (reason b of reason a )and presented here as its consequence @ a)

Origins of the Explaining Capacity


C3: rnarnan ta

insists CM: Cs

mommytake (hands again the baby to the mother) M3: il faut le laver avec la robe C on peut , M insists on refusal(see pas lenlever, elle cousue est tu vois, below) on peutpas. you need to wash himwith the dress, we cant take it off, it is sewn you see, we cant (then the mother showsto C how to wash the dressed baby).

In M3 the motherjustifies again her refusal, chaining the two reasons already uttered earlier, and repeating the first one in a sort revolving movement of tightly linking reason a and reason to each other. The justificationof the b refusal (reason a) becomes itself the explanandurn that is explained by anew reason (reason b) which then serves as the condition for the subsequent consequence, whichis nothing other thana, the original justification of the refusal.
Forms of Explanation

To what extentare justifications and explanations of oppositions introducedby the causal connective parce because?None of the girls justifications of que oppositions were expressed by parceque and neither were any of the boys justifications of first oppositions. During the last two periods, the boy introduced 18.5% of the oppositions produced during the later unfolding of the opposition episodesby parce que (5 instances-7% of the total-ofwhich 3 were in response to themothers 4 pourquoi?why), a result suggesting that having to holdones stance in conflictual movements might solicit the child to draw on emerging resources. Given the developmental period under study here, not surprising to is it find that the connective parce que was absent or only marginally present in the childs production of justifications of oppositions. But how about mothers? Do they use it regularly when producing these kinds of justification? Our data show that mothers also introduced their justifications of oppositions with the connective parceque infrequently. Theboys mother (Dyad A) used it in13.4% of all the justificationsof oppositions she produced (10.4% of the justifications of first oppositions and 22.6% of those produced later in the episode) and the mother of the girl (DyadC) produced it only once, corresponding to about 0.6% of all the justifications. Thus, in the developmental period under study here, not only childproduced but also mother-produced justifications were infrequently marked in a linguistically specific way. Could the child nevertheless identify their presence in the mothers oppositional moves? This question is addressed in the next section.



Mothers and Childs Reaction to the Partners Opposition, According to Presence of Justification in First Oppositions

Cids h l Age

Per. I


Per. II


Per. H I

2;2.14 2;4.2

Per. IV



with with without with without with without without with without justification justification justification justification justification justification justification justification justification justification

M gives in immediately M opens an intermediarysequence

M insists and startsa conflict


35% 35% 30%

0% 0%




21% 50%


57% 24%



g 1%


0 Yo


without with without with without with without with without with justification justification justification justification justification justification justification justification justification justification

C gives in immediately copens an intermediarysequence C insists and starts a conflict





36% 9% 55%

83% 3%




65% 12% 23%


0 Yo

0Y o




82% 3Yo 15%


Per. I


Per. II

Per. III


without with without with without with without with without with justification justification justification justification justification justification justification justification justification justification
100% 0Yo

39% M gives in immediately M opens an intermediarysequence 28%

M insists and starts a conflict

23% 35%

67% 33%


0 Yo



32% 32% 36%

73% 27%
0 Yo

without with without with without with without with without with justification justification justification justification justification justification justification justification justification justification

C gives in immediately

0y o





C opens an intermediarysequence C insists and starts a conejct




54% 29%

6% 53%






Proportion o first oppositionswith justification and proportion of oppositionswithout justification, that are followed by third turn acceptance by the partners opening of an intermedif ary sequence,and by the partners insistence (opening of a conflict), by childs age and dyad.

Origins of the Explaining Capacity


Effects of Explanations and Justifications of Ones Oppositionson the Partner In order to verify whether providing an explanation or justificationaffects the opposees acceptance of the opposers position, first oppositions accompanied by a justificationwere distinguished from those that were not so accompanied for both mothers and children. These two classes offirst oppositions were then further classified according to the kindof third turn issue they were followed by: the partners acceptanceof the opposition; the partners opening of an intermediary sequence with anIMS; the partners insistence on his or her or original position that starts an overt conflict. Table 6.4 presents the results of this analysis separately for the mother and for the child, each of the two for dyads, for the developmental periods considered. Results of z x 3 chi-square analyses (z types of first oppositionx 3 types of resolution) performed on the data Dyad A, at each time period for each of partner, show that the mother behaved differently according to whether the childs first oppositionwas accompanied by a justification or not. This was so from the-time the child started to justify his first oppositions more than sporadically [ x 2 (z x 3), at Period I1 = 7 . 9 3 6 , p < .02,at Period 111 = 17:z3, p <.ool, at period IV= 1 7 . 7 9 , ~ . O O I , for df= 21. Specifically, the motherwas < more likely to give in immediately following a childs first opposition thatwas accompanied by a justification than after childs first opposition that was not a so accompanied; in contrast, she was more likely to insist (and thus to start a conflict) or open an intermediary sequence after achilds first opposition that wac not accompanied by a justification than after one that justified. The z x was z contingency tables comparing immediate to delayed resolution (after an intermediary sequence, a conflict, or both) show the same significance levels of as those mentioned earlier, at the three corresponding periods. The child was also more likely to give in to the mother immediately following a first opposition that was accompanied by a justification than after a first opposition that was not so accompanied, and to insist (and thus to start a conflict) after a mothers first opposition that was not justified than after a mothersfirst opposition that was accompanied by a justification. This tendency was clear in the first two periods [ x 2 (z x 2) for Period I = 8.89, for Period I1 = 7.0 9 , both withp . 0 1 , df=1], and in the expected direction not < but significant in the two periods [ p (z x 2) for Period111 = 3.4, for Period IV last = 3.45, .o5 <p <.IO for both, df = 13. We mentioned earlier that10.4 % of the mothers justifications of first oppositionswere introduced by parce que. Did the presence of this explicit marker make the justification more effective for the closure of the opposition episode? The comparison between justified oppositions introduced by because and those without it, performed the overall on data, did not show any significant difference in the childs following behavior, both types of justifications being equally effective in preventing the opening of a conflict by child [xz (2x 2) = 0.056, n.s.1. the



In Dyad C, as soon as the child started justifying her oppositions more than sporadically, a similar significant effect on the mothers subsequent behavior was found [xz (2x 3) for Period = 10.37, p <.01, df= 23. As for Dyad A,this I11 mother was more likely to give in immediately after the childs first opposition that was accompanied by a justification than after the childs first opposition that was not so accompanied [comparing immediate todelayed resolution, x2 (z x 2) = 5.6, p < .oz,df= 11. The childs behavior was also influenced by the presence of a justification inthe mothers previous opposition. Overall, this effect was highly significant [x2 (z x 3) = 16.52, p < .oo1, df= 2 , showing in 3 particular the tendencyof the child to open conflicts frequently after a less mothers justified oppositionthan after a nonjustifiedone [xz ( x 2)= 14.17, 2 p < .oo1, df= 11. Similar significant results were obtained at Periods I and 1 1 for PeriodI: xz (z x 3)= 6.78,p <.05,df= 2 x2 (z x 2 = 4.4z,p <.05,df= 1; ; ) forPeriodII:~z(zx3)=8.29,~<.ol,df=z;~2(zxz)=5.45,p<.oz,df=1but no significant difference was found at Period111, when the child showed a sizeable increase in her productionof justifications of oppositions. It should pointed out that be many oppositions arise play situations anddo in not concern issues of importance. In some cases, however, oppositions occur for events that have some impacton real life (spilling wateron the floor, throwing around objects, shutting lights and doors that should remain open, etc.).Analyses performed on the dataof DyadA show, asone mightexpect, that conflicts arose more in the latter situations than in situations where nothing particular was at stake [35.7% vs. 17.4% of the situations, difference that is significant: xz (2 x 2 a ) = 9 . 4 6 , <.01, for df= 11. However, it is interesting to note that in occasions ~ even that have realconsequences, the presence of a justification makes the partnerless likely to insist on his or her original proposal: In fact, only 12.5% of those oppositions that were accompanied by a justification(N= were followed by 48) an insistence of the partner, whereas 73% of those thatwere not accompanied by a justification (N 3 6) gave rise to aconflictual exchange, a difference is = that highlysignificant[cz(zxz)=z8.7,p~.oolfordf=1]. These resultssuggest that providing an explanation or justification of ones opposition turns out to an adaptive behavior functionswell in convincbe that ing the partner of the legitimacy opposition, even when whatis at stake of ones has some, though minor, impacton real life. Both children started to use this behavior regularlyin Period 1 1 whereas both mothers produced it abundantly 1, from the beginning of the study (PeriodMoreover, they show that at the I). time children rarely justified their oppositions, they seemed to take justifications into account when their mother produced them.
Childrens Insistence on Their Initial Intention and on Their Opposition

The analysis of episodes i which children opened a conflict and of those in n which they insistedon their opposition after the mother had started a conflict

Origins of the Explaining Capacity toadapttothe


mayprovideadditional interactional partner.

clues to childrens capacities

(a) When children do not immediately accept their mothers opposition and produce a first conflictual move, how they go about insisting on their initial do position (analysis of Cc, third move in sequences type CMCc and longer)? of the
Low-level adaptive responses were distinguished fromhigher level adaptive n responses. I low-level adaptive responses, children continue the action the

mother protested, realize what they intended to do,repeat the initial or move i n the sameway it was formulated. Higherlevel adaptive responsesmay be of two kinds. One is characterized by a change relative to the initial move: Children either reformulate their initialmove, add a verbalization or a vocalization to the initial action,or replace a vocalizationby a verbalization.The second kind comprises other-oriented responses that take into account and provide recognition of the mothers opposition by replying to it, refusing it witho n , or n insisting by sayingoui for the French si, or that enrich the initial move by a justification. For both children there was an early time period during which the children showed only mostly low-level adaptive behavior (Periodsand I1 for the boy; or I Period I for the girl) and later period in which they started to show highera level adaptive behavioras they reformulated the initial move, responded to the mothers opposition saying clearly that they disagreed withand/or countered it, the mothers opposition with ajustification of the initial move.For the boy, higher level adaptive behavior increased from in Periods I and 11, to 78% in 0% Periods I11 and IV,constituting a significant change (2 x 2)= 18.91, ( . O O I [ p p ( for df= 13; for the girl, it increased from in PeriodI, to 50% in Period11, to 23% 82% in Period111, the changebetween Period I and Period111 and thatbetween Period I and Periods I1 and I11 combined being significant[ p (2 x 2 = 6.04, ) p ( .oz, df = 1, for I vs. 1111. ( @) In sequences in which mothers start a conflict by insisting on their initial intention afterthe childs first opposition, and children keep on insisting on their opposition, howdo they do it (analysis of Ci,fourth move in sequences of type the MCMcCi and longer)? The trend observed forchildrens insistence on their initialintent was found also for childrens insistence on their opposition. In the two early periods, the boy didnt show clear adaptive behaviours, because most often (87.5% of the cases) he repeated the previously uttered performed the intended contrary non, action, or both. the later periods,on the contrary,he was more likely to In formulate his opposition differently,to add an explanation justification of it, or or torespond directlyto themothers previous conflictual move (68.5% of the cases), behaviors that were more finely adaptedto theprogression of the interaction. For the girl, higher adaptive behaviors constituted ofthe cases level 14% of insistence at Period I, increasing to 29% at Period 11; all of the few



occurrences at Period I11 were of the higher adaptive type. The difference between the early and late periods in terms of the presence of adaptive behaviors is significant for both children [xz (2 x 2) = 4.85 and 5.10, respectively, for theboy and thegirl, both p <.05 for df = 11.

The results presented show that mothers provided justifications of their oppositions abundantly from the beginning of the observations, whereas children provided them more than sporadically about months later. Both 6 mothers and childrenoffered typically unsolicited justifications, most of them accompanying their first opposition. The great majority of the justifications of opposition were not introducedby the causal connective parce que because, a finding that is particularly noteworthy for the mothers. Results also show that providing the explanation or justification of an opposition led the partner to give in more quickly and decreased the likelihood the occurrenceof an open of conflict, and this even when the issue had some impact onreal life. Mothers responded in this way to their childrens production of justifications from the time children started producing justifications more than sporadically. Children took their mothers justificationsof oppositions into accountwell before they started to produce justifications more than sporadically, thus learning tobe convinced before they learned to convince others. Thus, very early on in mother-child interaction, children seemed to learn that justifying ones opposition is an efficient behavior that renders the opposition more legitimate or acceptable for the opposed partner. What kind of acquisition model we need to account for the developmendo tal changes describedin this particular pragmaticuse of language?Veneziano and Sinclair (1995) argued that together with the appearance references to of the past, the cooccurring appearanceexplanations in general show of childrens use of the newly apprehended informative potentialities o f language. Two independent but functionally converging developments hypothesized to were underlie the appearance of these new uses of language. On the one hand, cognitive development allows children to use language to refer to entities and relations that are immediately perceptible the situation at the not in moment o f speaking. This development is in line with the general cognitive evolution highlighted by Piaget, according towhich-during the sensorimotorperiodchildren become less dependent on immediate perception andon the actions they perform. Similar to the reversal in the relationbetween the objects and the subjects action schemes that occurs the last part of the sensorimotor period, in there may also be a sort of reversal or, at least, a more reciprocal relation, between signifieds and signifiers: Signifiers acquire amental reality of their own that make possible movement not only fromsignifieds to signifiers but also the other way around, signifiers eliciting the corresponding signifieds.

Origins of the Explaining Capacity


Vygotsky (19 33/19 6 7 ) proposed a similar change for symbolic when he play, spoke about a reversal in the relation between objects or actions, on the one hand, andsymbolic meaning,on the other.The present resultson the development of protests,refusals, and denials, are in line with this hypothesis. On the other hand, there to be a sociocognitive developmental change seems in the way children apprehend their interlocutor, who starts to be as an viewed alter ego, that is, a person whose psychological states may be different from the childs own. Several pieces of evidence support the appearance of this change toward the end of the secondyear. For example, findings show that children take clues from others emotional states,make verbal reference to own and others internal states, and use language informatively to highlight aspects of the situation that are either not perceptually available, as in earlyreferences to the past and first justifications (Veneziano & Sinclair, 1995)~ or inherently subjective, as are the meaning transformations of early symbolic play (e.g., Bretherton, McNew, 8 Beeghly-Smith, 1981; Dunn, 1988,1991; Lamb, 1991; Musatti, Veneziano, & Mayer, 1998). Although this development may be ultimately linked to acquisitions knownthe literatureas childrenstheory of in mind, it has been considered reflect childrens know-how of the mind, a to sort of practical understanding that doesnt need conscious reflection on others feelings and mental states (Dunn, 1988;Veneziano & Sinclair, 1995). However, these psychological, internal states, of which children start have a to practical understanding, need to be taken into account in interpersonal relations, be it a question sharing them, countering them, or changing them of with adaptively persuasive behaviors. The pattern of results presented here brings additional evidence support in of the hypothesis of a developmental change in childrens attribution of internal states toa partner and in their capacity to take them into account by producing adapted behavior. Indeed, children not only start justify their to oppositions regularlyand effectively (from Period 111on, starting at for the 1;g girl and at 22 for the boy) but, at that time, they ; also present other kinds of behaviors that show adaptation to the interactional situation and waythe to the partner responds to them. When they insiston theirinitial intentin response to the mothers opposition, or on their opposition response to themothers in insistence, they do it by reformulating their previous move, by adding an explanation or justification, or by responding directly to the mothers opposition. These findings show that children in their third may be more year competent in adjusting to socially produced failure than other studies might suggest (e.g., Ervin-Tripp etal., 1990). In this connection one might wonder about the extent to which the interactional phenomena we have put into evidence require a representation of the partners mental states or whether an on-the-spot fine adjustment to the dynamically changing flow of interaction might suffice to account for the results reported. Although it is not yet legitimate to exclude such a low-level



interpretation, related results favor more the hypothesis that children into take account-although still in a primitive way-the other persons intentions or beliefs and that, at times, one of the goals of their behavior is to modify these intentions orbeliefs. One of the results that seems to plead more stronglyin favor of this interpretation is the fact that children, like mothers, offer most of their justifications spontaneously with their first opposition and, if they produce justifications during the later unfolding of the oppositional episode, they seldom do it after a mothers explicit request for explanation or after a clarification or confirmation request. Thus, already with their early justifications, children seem to forestall their mothers objections ratherthan simply reacting to them. The developmental changes in language and social representation are supposed to provide, first, the necessary cognitive base and, second, the necessary communicative thrust for the productionof explanations. They are not, however, sufficient to account for the details the emergence and the of further development of this pragmatic capacity. What are thespec$c processes that might underlie this functional acquisition, important for the developso ment of sociocognitive functioning? The results presented seem to exclude a simple and direct conversational model according to which, for example, explanations would arise fromchildrens appropriating for themselves part of what the mother provides in her previous utterance, or from replying to the mothers requests for justification. Indeed, we have seen that the mothers studied here ask very few why questions in relation to childrens oppositions and that these are not necessarily followed by an explanation by the child. Moreover, from the timethey appear in childrens production, justifications accompany mostly childrens first opposition and are thus produced within not a conversationally framed adjacency pair. Children seem to learn in a much more indirect manner, basing their own elaborations on the way their mother behaves toward them. The mother justifieshigh a proportion of heroppositions,andchildrenrespond differentially as a function of the presence of a justification accompanying the opposition. Such behavior suggests that childrencan treat these justifications at some level well before they start producing them regularly. The cues for this differential treatment are most likely semantic because, as we have seen, only a tiny number of mothers justifications are introduced the specific linguistic by marker parce que. If it is not thecase that children learn from mother-regulated conversational functioning leading to jointly constructed explanations, should assume the we working of a more extremesocially triggered model implying a process of internalization of the behavior produced the more competent partner? This by interpretation is hard to reconcile with the considerable (of at least about delay 6 months for both children) found between the time children manifest attention tomothers justificationsand thetime they start producing justifica-

Origins ofthe Explaining Capacity


tions themselves. Such a delay cannot be accounted by a lack of verbal means for of expression, as the early justifications children offer consist single-word, of successive single-word, orearly two-word utterances involving lexical items they had at theirdisposal much earlier (see Veneziano & Sinclair, rg g 5, for a more elaborate discussion of this point). Rather, the existence of the delay fits well with a constructivist interpretation, according to which the emergence of the explaining capacity requires childrens internal elaboration of the communicative and pragmatic significance of the justifications identified the in mothers oppositions. In addition to the delay between sensitivity to motherproduced justifications and the emergence of self-produced justifications, the overall pattern of results obtained for the two dyads provides further support for this interpretation. Indeed, pulling together the developmental results, two distinct periods can be clearly identified: (a) an early period in which children pay attention to and are influenced their mothers justificationsof opposiby tions but either do notproduce them or do it only sporadically;and (b) later a period in which they produce justifications regularly and show other-directed adaptive behavior in the later moves of the opposition episodes, become but somewhat less sensitive to their mothers provision of a justification. The existence of these two periods shows that children center themselvessuccessively on different aspects of the complex pragmatic situation, aprocess that often characterizes cognitive constructions: First they focuson the semantic content of their mothers justifications and are often persuaded by them and start to elaborate from them ideas to implemented in their own behavior; be then they focus on their own other-directed justifications, produced to affect the partner, who indeedis influenced by them. The constructivist hypothesis predicts the existence of a third period in which children both use effective adaptive behaviour and accept beingmore easily persuaded when the partner uses it. It is only then thatjustifications of oppositions are integrated into a coherent system in which they become two-way procedures (self-to-other directed and other-to-self directed) that effectively support theopposers can position (be it theself or theother), both topersuade andto be persuaded. In sum, we suggest that the acquisition of childrens early explanatory capacities requires general cognitiveand social developments: One relates to the growingpsychological reality of verbal signifiers that can then be usedin a displaced, informativeway and, the other to the attribution psychological of internal states, which can be different from the childs own, to theinterlocutor. The specific processes underlying the emergence and early development of childrens explaining capacity are supposedto be linked to adult-child interactional functioning, but notof the conversationally direct adjacency-pairs type. The mothers behavior and its treatment by the child, the fact that the child is coparticipant in the oppositional episodes and has vested interests to defend, and the childs internal organization of this material, are aspects of all the complex processes that might be involved in the emergence of childrens



explaining capacity. Other important variables, not taken into account here, such as the topicof the opposition, the emotional involvement, and the nature of the partner (DunnbMunn, 1987; Eisenberg, 1992; Haight et al., 19g5), as well as the nature of the explanations offered, might interact with the emergence of this capacity. Moreover, consideringthat mothers justifications are rarely introduced by the specific connectiveparce que, and that there are often several missing links be filled in order to understand them justifito as cations, further investigation is likely to reveal the implication of other cognitive abilities, particularly of specific types of inferential processes.

The research reported in this paper supported in part the Fonds national was by suisse de la recherche scientifique (grants no. 11-30927.91and no. 37304.93 to E. Veneziano and H. Sinclair). The authorwishes to thankKarine Coulon and Isabelle Gauthier for assistance in data analysis.

Barbieri, M S . , Colavita, F., & Scheuer, N. (19 yo). The beginning of theexplaining capacity. In G. Conti-Ramsden & C. Snow (Eds), Childrens language, Vol.7 (pp. 245-271). Hillsdale, NJ: Lawrence Erlbaum Associates. Berthoud-Papandropoulou, Favre, C., & Veneziano, E.(19 y 0). Construction et reconstruction I., des conduites dexplication [Construction and reconstruction of explanatory behavior]. Cahiers dAcquisitionet de Pathologie du Langage, 7/6,9-35. Bloom, L. (1970). Language development: Form and function in emerginggrammars. Cambridge, MA: MIT Press. Bloom, L., & Capatides, J B. (1987). Source of meaning in the acquisition of complex syntax: The . sample case of causality. Journal ofExperimental ChildPsychology, 43,112-128. Bretherton, I., McNew, S., & Beeghly-Smith, M. (19 81). Early person knowledge as expressed in gesLamb & tural and verbal communication: When do infants acquire a theoryo f mind? In M. E. L.R. Sherrod (Eds.), Infant social cognition (pp. 333-373). Hillsdale, NJ: Lawrence Erlbaum Associates. D u m , J. (19 88). The beginnings ofsocial understanding.Cambridge, M A : Harvard University Press. Dunn, J. (19 y1).Understanding others: Evidence from naturalistic studies of children. In A. Whiten (Ed.), Natural theories ofmind(pp. 51-6 1).Oxford, England Basil Blackwell. . Dunn, J (1993). Social interaction, relationships,and the development o f causal discourse and conflict management. European Journalofpsychology ofEducation, Special issue on Everyday life, social meanings, and cognitivefunctioning, 6(4), 391-401. Dunn, J & Munn, P. (1987). Development o f justification in disputes with mother and sibling. . , Developmental Psychology, 23,791-798. Dunn, J,Slomkowsky, C., Donelan, N., Herrera, C. 95). Conflict, understanding and relation. & (19 ships: Developments and differences in preschool years. Early Education andDevelopment, 6(4), 303-316. Eisenberg, A. (19 y 2). Conflicts between mothers and their young children. Merrill-Palmer Quarterly, 36(1), 21-43.

Origins ofthe Explaining Capacity


Eisenberg, A. R., & Garvey, C. (1981). Childrens use o f verbal strategies in resolving conflicts. DiscourseProcesses,4,149-170. Ervin-Tripp, S., Guo, J., & Lampert, M. (1990). Politeness and persuasion in childrens control acts. Journal ofPragmatics, 14,307-331. Garvey, C. (1977). The contingent query: A dependent act in conversation. In M. Lewis & L. Rosenblum (Eds.), Interaction, conversation and the development oflanguage: The origins of behavior(pp. 63-93). New York: Wiley. Grize, J.-B. (1996). Logique naturelle and communications [Natural logic and communication]. Paris, France: Presses Universitaires de France. Goodwin Harkness, M. (1990). He-said-she-said: Talk as social organization among black children. Bloomington & Indianapolis: Indiana University Press. Haight, W., Garvey, C., & Masiello, T. (1995). Playing with conflict: A longitudinal study o f varieties of spontaneous verbal conflict during mother-child interaction at home. Social Development, 4(1), 92-107. . Heritage, J (19 90). Interactional accountability: A conversational analytic perspective. In B. Conein, M. de Fornel &L. QuCrC (Eds.), Lesformes de la conversation [The forms of conversation] (Vol. 1, pp. 23-49), Paris: CNET. Hood, L., & Bloom, L. (1979). What, when and how about why: A longitudinal study of early expressions of causality. Monograph ofthe Societyfor Research in Child Development, 446) Serial no. 181. . Kochanska, G., Radke-Yarrow, M , , & Girnius-Brown, 0. (1987). A developmental Kuczinski, L , interpretation of young childrens noncompliance. DevelopmentalPsycholog~2&6),799-806. Lamb, S. (1991). Internal state words: Their relation to moral development and tQ maternal communications about moraldevelopment in the second year o f life. First Language, 11,391-406. Laugaa, A., &Brassard, M. (1997, May). Les conflits dans le5 dialogues mtre-enfant: Un contextepour le dkveloppement! [Conflicts in mother-child dialogues: A context for development?]. Paper presented at theAnalysis of Interactional Processes Workshop, Aix-en-Provence, France. Musatti, T., Veneziano, E.,& Mayer, S. (1998). Contributions o f language to early pretend play. Cahiers de Psychologie CognitiveKurrent Psychology of Cognition, 17(z), 155-181. Special issue edited by A. Dowker. Pontecorvo, C. (1990). Opposition, explication et invocation des regles dans la discussion en classe entre enfants de 5 ans [Opposition, explanation, and invocationof rules in class discussion among 5-year-oldsI. Cahiers dAcquisition et de Pathologie du Langage, 7/8,55-69. Rojo-Torres, M. (forthcoming). Communication entre jeunes enfants sourds et Cmergence de conduites dexplication [Communication amongyoung deaf infants and the emergence o f explanatory behaviors]. In S. Rayna & E. Veneziano (Eds.), Processus interactionnels, fonctionnements cognitifs et apprentissages. Schlesinger, I. M., Keren-Portnoy, T., & Parush, T. (in preparation). The structure ofarguments. Shantz, C. (1987). Conflicts between children. ChildDevelopment,58,283-305. Tardif, T. (1998, July). The sounds ofsilence in disputes between Beijing toddlers and their caregivers. Paper presented at the Sixth International Pragmatics Conference. Reims, France. . Tesla, C., & D u m , J (1992). Getting along or getting your own way:The development o f young childrens use of arguments in conflicts with mother and sibling. SocialDevelopment, ~(z), 107-121. Veneziano, E.(19 90). Les premieres manifestations des conduites dexplication [The first manifestations of explanatory behavior]. In I. Berthoud, C. Favre, & E. Veneziano (Eds.), Construction et reconstruction des conduites dexplication [Construction and reconstruction o f explanatory behavior. Cahiers dAcquisition et de Pathologie du Langage, 7/8,11-20. Veneziano, E., & Sinclair, H. (1995). Functional changes in early child language: The appearance o f references to thepast and of explanations. Iournal ofchild Language, 22,557-581. Vygotsky, L.S. (1933/1967). Play and its role in the mental development o f the child. Soviet Psychology,5, 6-18.

This Page Intentionally Left Blank

Childrens Attributions of Pragmatic Intentions and Early Literacy


University o British Columbia f

How good are young school-aged children at attributing intentionsgoals to or speakers in various discourse settings? Answers to this question may have important bearings upon the development writing ability in school-age of children. This chapter explores the links between underlying pragmatic attribution ability, childrens consciousness of that ability, and the early development of narrative and expository-descriptive writing.

A fundamental theoretical for studiesof linguistic pragmatic knowledge goal is to describe explicitly how conversational participants manage compute the to illocutionary intent of utterances and, in turn, respond some appropriate in way, verbally or otherwise. The participants task is complex, because they must attend not only to features of the linguistic utterance, but to relevant also aspects of participants knowledge, assumptions, and attitudes (Levinson, 1983). Inadditiontolinguistically conveyed informationaboutsuch participant features (henceforth, intentions), a range of extralinguistic cues is also employedintheinferential processinvolved. This is particularly problematic when littleexplicit information concerning pragmatic point is provided in the linguistic content of the utterance, as in the case of polite indirect speech acts (Ervin-Tripp8, Gordon, 1985), ironic or metaphoricusage (Winner, 1988), or idiomatic usage(Gibbs, 1987). Similarly, studies of childrens development of linguistic-pragmatic knowledge set out in part to trace the development of the abilities underlying the attribution of participants intentional states. Several lines of work in the development pragmatic abilities indicate of that such an enterprise is problematic. In a well-known single-subject case study,
1 43



Halliday (1975) notedthat when his subject used the form more + NP he was trying toconvey a requestfor that NP, whereas if he used the form (or three, two etc.) + NP, the child intended to convey an assertion of fact. At that stage Hallidays subject was mapping his own pragmatic intention communicative or goal onto linguistic forms a one-to-one manner, without in any requirement on his or his hearers part to employ background information in order to compute illocutionary point. However, Shatz (1978) observed her 2-year-old subjects responding with a complying action not only to linguistically possible request stimuli suchas can you findme a truck? but also to linguistically odd forms suchas may you find me a truck? Shatzs rule, mommy says, child does suggests that these children had a simple interactional procedure, something like a narrative script, for linking utterances with intentions. This procedure took at least some account information conveyed bythe social context of the of utterance. Early studies of these problems tended explore the extent which young to to children take social variables suchas age, status, or role of speakersor hearers into account in formulating or responding to a speech Shatz and Gelman act. (1973) found that 4-year-olds used shorter and complex utterances to less 2-year-olds than to adult addressees. James (1978) noted that preschool children were able to modify their directive speech acts according to theage of the hearer, whereas Ledbetter and Dent (1988) demonstrated differential responsiveness to direct and indirect directives by very young children. Bates (1976) found politeness developing by at least age 4 in her study, which included childrens judgments of varied levels of utterance politeness and effectiveness in social contexts. Leonard and Reid (1979) studied childrens judgments of appropriateness of a variety illocutionary acts used in approof priate and inappropriate conversational contexts, and found a steady increase in ability to perform judgments 4 through 6 years of age a wide variety from for of illocutionary acts. Performance also varied not only a function of nonas linguistic verifiability of events referredto in the utterance but also in termsof speakers politeness toward hearer. Becker (1981) found 4- and 5-year-olds reliably able to discriminate speakers status on the basis of features of requests. A somewhat later line of studies, while continuing to examinerelatively fixed social features such 2s age, status,or role of participants, also included attributes that are more interactionally constructed and negotiated, such as intention or communicative goal, in the scope of their investigation. Wllhnson, Wilkinson, Spinelli and Chiang (1984) studied 5- to 8-year-old childrens production of requests as well asjustifications of their judgments appropriof ateness of requests. A school setting not unlike the present studys setting was used, but in addition to a teacher-pupil dyad, a pupil-pupil dyad also was employed as an elicitation context. Childrenwere found to be quite exacting in terms of linguistic form when it came to accepting appropriate performances of requests for action, but were stringent for requests for information, less

Childrens Attributions


implying that they took speakers intentions into account when judging utterance appropriateness. Moreover, in explanations of their judgments, children across theage range studied consistently mentionednot onlyspeaker politeness but also the speakers object in requesting. Some recent studies have examined childrens metapragmatic knowledge in more direct ways. Baroni and Axia (198 9) in a studyof metapragmatic knowledge of a directives politeness as a function of personal familiarity,found that ability to identify the most likely speaker of polite less polite directivesas a and function of participants familiarity increased fromages 5 to 7 , but that these two age groups abilities to explain their attributions terms of pragmatic in rules for politeness remainzd equally rudimentary. The authors raised the interesting question of whether such developmental effects might be attributable to childrens changing abilities to represent explicitly to themselves the multifaceted pragmatic rules governing such conversational interactions. Garton and Pratt (1990) examined 8- to 12-year-old schoolchildrens abilities to rate a range direct and indirect directives in terms of politeness, of effectiveness, and likelihood of use. Although that study found these three dimensions strongly intercorrelated for all the age groups studied, some less marked distinctions did emerge among thedimensions, as a functionof linguistic form of the stimulus request. In a descriptivestudy of six children ranging in case age from 2 1/2 to 4 1/2, Becker (1988a) classified 6 5 metapragmatic utterances spontaneously produced in the home over the courseof a year. Seventeen of the utterances were categorized as strategic manipulationsof the rule systems by means of reference to a potentially conflicting rule (I cant talk; Im dead), implying well-developed ability, at least in the familiarity routine of the and home setting, to anticipate and subvert others intentions. Beckers (1988b) study of five of the same subjects described responsesto a particular class of parental efforts to teach pragmatic skills, indirect and direct correctionsof 10 categories of childrens pragmatic errors. Parents indirect techciques, rather than their direct teaching efforts, yielded the highersuccess rate: 59% and 5%, 3 respectively. Clearly the former represent a complex pragmatic-inferential challenge to the children because, as Becker notes, theymust computeparents communicative intentions in order to determine the point of that class of corrections. Less sanguine descriptions of the extent childrens pragmatic abilities of emerge fromseveral recent investigations. Abbeduto, Nuccio Bibler, Al-Mabuk, Rotto, and Maas (19 9 2)studied 6 - to 11-year-oldsability to employ speakers a general conversational goal indetermining the point of an individual speech act. They devised a role-played customer-merchant scenario was varied in that terms of the definiteness reference in the object of the customers inquiry. of The three levels of definiteness of reference thought on the basis of theory, were or perhaps native speaker intuitions, carry distinct presuppositionsas to the to customers general conversationalgoal. Because only the 11-year-olds and the



adult controls were able to make the necessary semantic discriminations and consequently construct the preferred pragmatic responses, the authors concluded that only those groups took speakers goals into accountwhen age computing illocutionary point. A similar finding emerged from Bernicot and Lavals (1993) study of 4-, 7-, and 11-year-olds understanding of felicity conditions of promises. Whenthe children wereasked to justify their completions of fairly extensive story scenarios that contained promises, the 11-year-old group, but not the younger groups, tended to incorporate information about the speakers intentions, or the joint intentions and desires of speakers and listeners. Such conclusions run counter to those ofReeder and Wakefield (1987), who showed that 3-year-olds couldreliably comprehend the intended meanings of indirect requests,even when linguistic information systematically removed was from the presentation. Those children were shown to have employed contextual information that could reveal a speakers intent. Further, Babelots (1996) study found children young as 17 months demonstrating an expectation that as speakers intentions would play a role in the interpretationof linguistically vague or indirect utterances, These studies suggest that even very young children can make considerable ofcontextual information, it is presented use if in simple and age-appropriate contexts, order to discriminate whether a in speaker intended toconvey a request for information, a request for action, an offer, or an assertion. Wakefield (1998) discovered that children by age 4 were ablereliably to attribute sophisticated degrees of certainty to speakers. Similarly, Hickmann, Champaud, and Bassano (1993) studied S-, 7-, and 9-year-olds knowledge of the rules of use for the French epistemic modal auxiliary croirethink, believe. Their work employed a richly detailedchildand friendly experimental paradigm involving animal and child puppets as eyewitnesses to misdeeds, modalized ornonmodalizedaccusationsof culpability, and both a narrative retellingtask and ajustification interview.All age groups made use of the witnesses epistemic status, thatis, their prior knowledge,intheirjustificationsofthevariousaccusationsthatwere dramatized in the procedure; but they managed this task in increasingly sophisticated ways over the three age levels sampled. Contrary to Abbeduto et al.s (1992) conclusions, the limitations the youngest children studied by of Hickmann et al. seemed to reside in the breadth participants viewpoints the of children were able to take into account, not in their actual ability mapthose to participant featuresonto linguistic operations. The moral to this is largely methodological. It would appear from tale the evidence just reviewed that when researchers devise elicitation procedures or experimental tasks that are themselves developmentally accessible, engaging, and interesting to children, their studies (ceteris paribus) will offer more accurate estimatesof the extent of discourse and pragmatic development. Our pragmatic and discourse studies need draw in less confounded fashionupon to

Childrens Attributions


childrens developing sense of narrative structure, which in turn based upon is their growing experience plausible sequencesof human encounters and the of motives behind such culturally embeddedevents (Donaldson, 1978; ErvinTripp & Kuntay, 199 6). It is just those sorts of settings and procedures that yielded earlier identification of the some the rootsof narrative development of and pragmatic competence in the workBabelot (19 9 6), Becker (19 8 Sa), of Hickmann etal. (1993),and Reeder and Wakefield (1987).

A long-standing debate in applied linguistic educational thought concerns and the relationship of learners explicit knowledge about aspects of a languageto proficiency with that language. It possible to trace someof its roots to an is educational debate about the role the that study of grammar plays in promoting proficiency in English composition and reading for older learners (Elley, Barham, Lamb, & Wyllie, 1976). Freires (1972) more general notion of conscientiza@o for adult learners, or the German pedagogical concept of Sprachbewusstheit are also fundamental notions in this debate. In second language learning, the question was also framed more broadly, and not restricted in scope to literate proficiency. The issue became known as the interface debate and concernedwhether a link between implicit and explicit linguistic knowledge could be opened up instruction and practice (e.g., by Krashen, 1981,vs. Bialystok, 1978, McLaughlin, 1990, or Sharwood Smith, 1981). Not surprisingly that debate to a range of studies of the effects of led formal instruction (as distinct from drill or practice with linguistic forms) on the development of second language proficiency (Ellis, 1984; Long, 1983; Mellow, 199 6). The issue has also been studied in terms of relations between linguistic awareness and reading proficiency (Downing, 1987; Willows & Ryan, 1986; Yaden & Templeton, 198 6) and between reading proficiency first and second in languages and linguistic awareness in the second language (Anthony, 1984; Bialystok & Ryan, 1985.) Apart from one early study Kroll(1978), in which by writing proficiency and audience awareness were investigated, the overwhelming majority of studies about links between linguistic awareness and literate proficiency have examined structural aspects of languagerather than functional or pragmatic aspectsof language in use. We turn now to a study whose purpose to determine the was extent to which young school-aged childrens ability to attribute pragmatic intentions was linked to the early development of expository-descriptive and narrative writing proficiency, respectively. In broader terms, the study provides one way of exploring relationships among fundamental elements human linguistic of interaction and the manifestationof the resulting narrative and descriptive abilities in an educationally important domain,written composition.




line of visual regard

F I G . 7.1.

The communication scenario.



Forty-two English-speaking children from middle-class communities in two western Canadian cities took part in the study. They selected becauseof were their ongoing participation in a larger, university lab-school based investigation of the emergenceand development of language and literacy, initially in the preschool years, but eventually, in the early primary school years, for five years of longitudinal observations(Reeder, Shapiro, Watson, & Goelman, 1996). The study reported here took place in the final year of observations. Children ranged in age from 6 2 to g ; ~ with a mean age of 7;9, and were attending the ; , first, second,and thirdgrades oflocal public and private schools,in which the language of instruction was predominantly English but i several cases, French. n The sample included 20 girls and 22 boys. All of the children were judged by their teachers to possess native-speaker proficiency in English, including several children for whom English was not the home language. None the particof ipants had experienced developmental delays or learningdifficulties that had been drawn to the attention school officials. of

Data Collection

Childrens Attributions


The studys data were collectedin three steps:

Presentation of a puppet-played communicationscenario. Administration of a brief interview designed to probe childrens attributions of the puppet speakers communicative intent and childrens understanding and identification of the source of their attributions. 3. Administration of two writing tasks in two genres, narrative, and expository1.


TheCommunicationScenario. An audiotaped stimulus utterance was presented by means of puppet play in a model schoolyard or schoolroom context designed to predispose a directive interpretation. The standardized directive contextt, illustrated in Fig. 7.1, depicts a Speaker in the role of teacher puppet and a Hearer in role of a pupil puppet. The Speaker the is placed near a plaything, while the child puppet is some distanceaway from the plaything. Stimulus utterances consisted the general form, Would you of like to play on the X? where X was one of a varietyof playground toys positioned beside the Speaker puppet. The Attribution Interview. This consisted of two probes followingthe presentation of the communication scenario. The first dealt with the childs attribution of the Speakers intention, and the second with the basis on which the child believed he she was making the attributions intent. Attribution or of of Speakers intention was probed by asking the child, Why do you think the teacher said that? along with any follow-up questions or clarifications. Participants basis for attribution of Speakers intention was probed by asking them, How do know that? together with follow-upor clarification you any questions. An example of the two-partinterview trialis the following:

E: Why do you think the teacher said that? (probe for attribution of Speakersintention) S: To see what he [child puppet] can do. E: Howdo you know that? (probe for basis of Speaker attribution) S: Cuz thats whatall m y other teachers wanted me to do.
Each child received two such trials, with each trial employing a different plaything inthe presentationof the stimulusscenario. As part of a preliminary study of the attribution interview (Reeder, 19 g 6 ) , a coding taxonomy (shown i Appendix A) was constructed on the basis of n responses to the questionof Attribution of Speakers Intention andresponses coded by two coders. Similarly, the coding scheme appearing in Appendix B was constructed from the responses to the question about the childs Basis for



T A B L E 7.1 Examples of Scoring of Quality of Speaker-Attribution Responses

Score = I

( W h y do y o u think the teacher said that?) Because she wanted her try and make a picture. to (How do o u know that?) y I just know.
Score = 5

( W h y do y o u think the teacher said that?) Maybe becausethe person had done all the other activities except make a picture, or maybe allthe other activities were occupied wanted him make a picture. so she to (How do o u know that?) y All teachers would probably like a good picture, just messy blobof paint or not a something. So she just asked the boy to make a picture.
Scale: 1 - 5. N = 42, M = 2.90, SD= 1.16. Attribution of Speakers Intention. Responses were coded into these categories by the two coders. The descriptive characteristics, distributions, coherence and of childrens responses to the two questions in the Attribution interview were also presented inReeder (19 9 6). Quality of attribution was assessed by two raters, who scored combined the responses to thetwo questions of each Attribution probe on a scale from o to 5. The rating scales values are exemplified in Table 7.1. Raters achieved 6 9% agreement in their initial ratingsand resolved all disagreements by discussion to achieve a consensus rating.
Expository-Descriptive Writing Task. This task consisted of the researcher asking each child to Write a short letter to a friend or a favorite cousin in another city whowill be visitingyou soon, telling all you can about your own room. The taskwas administered in small groups, and although untimed, generally took subjects about10 minutes to complete a draft copy. Written samples were scored for three characteristics: overall writing quality, quantity of output, and structural features associated with syntactic maturity writing in (ODonnell, Griffin, & Norris, 1967;OHare, 1973). Overall writing qualitywas rated by two markers, employing holistic ratings (Greenberg, 1994). were They instructed to work individually rate each sample a o to 10 scale for the and on extent to which each sample appeared have fulfilled the rhetorical purposes to of the task. A nonresponse or a completely unintelligible response received a score of o from each rater, while a pictorialsingle-word or phrasal response or

Childrens Attributions


was scored 1 out of 10 by each rater.The two raters scores were then summed, to yield a total score of 2 0 for each piece of writing. Examples of the use of out the holistic rating of writing quality are providedFigs. 7.2,7.3, and 7.4, and in illustrate low-,mid- andhigh-scoring cases using thesummed scores out of 20 points. Quantity of output rated by counting total numbers of words, was clauses, and sentences produced. Structural features associated with syntactic maturity were number of words per sentence, number of words perclause, and number of clauses per sentence.
Narrative Writing Measure. Children looked at Tomi de Paolas wordless picture book, Pancakesfor Breakfast (1978), and discussed it informally with the researcher. Children were then invitedto retell the story in their own words. On a subsequentvisit, children reviewed the book, then were asked to retell the story in writing, again in their ownwords. The same featureswere scored in narrative writing as in the expository-descriptive writing task: Quality, Quantity, and StructuralFeatures. Quality was assessed by samemethod as the that described for the expository-descriptive writing task, holistic scoring by two raters.

Design and Analysis

Two factorial experiments were designed to evaluate the extent to which specific features of writingwere associated with chronological and quality age of pragmatic attributions. The various features writing noted earlierserved of as dependent measures in a series of analyses of variance with planned comparisons. Chronologicalage and quality of attributions served as grouping variablesasfollows:WritingMeasurex Age Group (Younger, Older) x Attribution Quality (High, Low). Age groups were formed by splitting the population into two equal groups, follows: as
Younger (N= mean = 85.67 months (7;2),= = 5.9, range = 74-93 months, 21, m), mean= 99.67 months (8;4),= = 4.4, range = 94-109 months. Older (N=

Attribution Quality groups were formed by splitting the sample follows: as


(N= mean = 3 . 6 5 , s = .74, range = 3-5, 26), Low (N= 16), mean= 1.69, a= .48, range = 1-2.

The distribution of attribution quality scores among the four groups so constructed was as follows:
Younger, High Attribution Quality (N= 13), mean = 3.46, = .6 6, Younger, Low Attribution Quality(N= 8, mean = 1 . 6 2 , D = .52, Older, High Attribution Quality(N= 13), mean = 3 . 8 5 , D = .go, Older, Low Attribution Quality (N 8), mean = 1.75, ,Q .46. = $=




7.2. Low-qualitywriting sample(score = 2/20).

Children's Attributions


undr m y bed I hav towes and a faow thing eles


7.3. Mid-quality writing sample (score = 4/20).



Ma chanbre et tres petit. Jer 11 petits oursons. Jerun tres grand ours. Jer trks un grand lit. Jer beaucoup de livre. Jer un petit maison et dans la maison vit de lapin. Jer un lit poour mes poupes et beaucoup danimals. Jerun chat U nchieau est beaucoup de poisons. Jer baucoup de vCtemans pour mes poupes. Jer des vile avec beaucoups beaucoups de maisons. Jer Sonie. Jer un beaucoups de casests. Jer lamp. Jer beaucoups de vktemans pour moi. un


7.4. High-qualitywriting sample (score = 20/20).

Children's Attributions


The general formof the hypothesis being evaluated each analysis was: There in should be a greater contrast between 7 and 8 years in quality, quantity, and structural features in both the expository-descriptive and narrative writing samples for the children demonstrating higher quality of pragmatic attributions compared to children in the lower pragmatic attribution group.

Expository-DescriptiveWriting and Pragmatic Attribution Ability

As shown in Table 7.2, older subjects demonstrating superior pragmatic attribution ability scored significantly higher the measure of on overall writing quality than theyounger subjects with superior pragmatic attribution ability. Second, the older group of subjects who demonstrated superior pragmatic attribution ability also scored significantly higher total words produced in on the writing sample than the younger subjects with superior pragmatic attrib tion ability. Third, therewere significant effects of age,but not pragmatic attribution ability, on two additional measuresof writing: words per sentence and words per clause.
Narrative Writing and Pragmatic Attribution Ability

A s shown in Table 7.3, the older group of subjects who demonstrated superio

pragmatic attribution ability scored significantly higher on the measure of overall writing quality than the younger subjects with superior pragmatic attribution ability. There was no similar contrast found between younger and older children in the lower attribution quality group, and attribution group no contrasts within age groups were significant. Second, the older group of subjects who demonstrated superior pragmatic attribution ability also scored significantly higher total sentences and total on clauses produced in the writing sample than theyounger subjects with superior pragmaticattribution ability. The lower attribution groups demonstrated developmental contrasts on no either total sentences produced or total clauses produced. Third, the older group of subjects who demonstratedlower pragmatic attribution ability scored significantly higher on words per sentence, words per clause, clauses per and sentence produced in the writing sample theyounger subjects with than lower pragmatic attribution ability. Finally, there was a significant effect of age, but not pragmatic attributionability, on total words written. Thus for six of the seven dependent measuresassessed, age and attribution quality interacted.Of these six significant interactions, four measures-overall quality, word count, sentence count, and clause count-revealed the lowest scores for the younger, attribution quality group and the highest scores f low the older, high attribution quality group.



Mean Expository-Descriptive Writing Scores by Age and AttributionCroup

Older/ Older/ Younger/ Writing measure
Overall quality

T A B L E 7.2

20.875 3.000 3.250 5.640 7.625

High Low High attributionattributionattribution

17.38 5 8.240 5.783 2.615 2.231 6.615 27.625 11.921 9.144 3.500 9.875 39.923 4.000 12.228 4.769 9.257 13.077


Word count
Sentence count



Clause count

Wordslclause Clauseslsent.

4.940 0.833





1.33 9


*Age x Attribution interactions; **Age effects.


This theoretical approach to relations between pragmatic knowledge and literate proficiency predicted significant contrasts in quality, quantity, age and structural features in both the expository-descriptive and narrative writing samples for the children demonstrating higher quality pragmatic attribuof tions compared to children in the lower pragmatic attribution group. This working hypothesis was partially borne out in the present study. In contrast with the group of subjects showinglower pragmatic attribution skills, the subjects with higher attribution skills demonstrated significant developmental contrasts on the majority the measures of writing ability in both genres of when 7-year- olds were compared with8 -year-olds. Evidently it is not sufficient for children to develop only in formal linguistic terms to succeed in many aspects of writing proficiency: Thereis evidence in the present results for the role of pragmatic competence, possibly of a veryspecific sort, as a component of several aspects of proficiency in the two types writing tasks we observed of here. A general developmental increasein writing ability around age 7 has been well documented inlarge-scale studies of childrens writing abilities (Perera, 1984; Wilkinson, Barnsley, Hanna, & Swan, 1980). What the present results demonstrate is the reliable contribution of skill in the attribution of intentional states in othersto specific aspects of writing development, particularly where overall quality and sheer quantityor fluency of writing is evaluated. The study further demonstrates that is possible to examine the development early it of writing ability in finer grained detail, isolating some extent the dimensions to

Childrens Attributions


T A B L E 7.3 Mean Narrative Writing Scores by Age and Attribution Group

Older/ Older/ Younger/ Writing measure

Overall quality Word count Sentence count Worddsentence Clause count Wordsklause Clauses/sentence

attribution (N = 8)

High Low High attribution attribution attribution

( N = I ~N = ()
34.462 2.769 9.0 56 6.769

8) 51.625 14.575 6.506 2.215 7.875 3.75 0

67.692 12.250 9.846 1.730 7.079 6.231 12.385 .013*



3.000 2.879 .802



4.692 5.493 1.327


.024* .018*

*Age x attribution interactions; *+Age effects.

of overall rhetorical qualityor effectiveness, fluency, and structuralcomplexity. An interesting sourceof variation in some aspects of literate development, when carefully defined, appears be metapragmatic knowledge. Contrary to to Bereiters (1980) sweeping claim that expository writing is not well developed in 8-year-olds in general, this has been shown here to havedeveloped significantlyby about age 8, but not for children. all Pragmatic attributionskill was found tohave underpinned this writing ability when writing was measured in terms of overall rhetorical quality and sheer quantity of output, but not terms of structural features previously in shown to be associated with (ODonnell; Griffin, Norris, 196 7 ) .Why then age & did the significant age-linked contrasts in syntactic complexity (words per sentence, words perclause, clauses per sentence) of narrative writing showup in the less pragmatically sophisticated group of children not in their more but pragmatically sophisticated agemates? The answer may lie in the dissimilar baseline scores between the two younger groups.In effect, the younger learners in the low pragmatic attribution groupneeded to make comparatively larger gains from theirlow baseline levels in order to reach syntactic complexity levels comparable to their high attribution group agemates. They had to come from behind. In contrast, the younger, high attribution group already functionwas ing at relatively higher levels in words per sentence, words perclause, and clauses per sentence, and had less opportunity to reveal developmental contrasts with their older counterparts. longitudinal studyof such groups A might reveal two developmental curves for written syntactic development in the narrative genre, one somewhat in advance the other as a function of of pragmatic attribution ability.



Such developmental variation has educational significance, particularly for the development of academic discourse in the middle later schoolyears by and students with differing proficiencies in the language of school instruction. Cummins (1981) and Collier (1987) observed in large-scale school system surveys of immigrant students the relative speed with which students acquire basic interpersonal communicativeskill (BICS). This contrasted with the longer developmental timetable for mastery of academic discourse with its increased abstractness of reference decreased interpersonal and contextual support. and Cumminss findings offer another way of interpreting the present studys results, whereby childrens prior andsomehow fundamental grasp of the pragmatics of interpersonal communication appears have provided a foundation for to success in the narrative genre, and less so in the expository-descriptive writing genre. It would be expected that ability to attribute intentions to conversational participants would underlie the development of superior narrative writing, given that genres reliance on characters dramatically realized by their conflicting intentions or motives. Our results are generally line with such an account in of development, in that a significant contribution of attribution ability was demonstrated in six of theseven narrative measures employed here. What was not self-evident was that attributionability wouldalso have been linkedto the development of at least two measureswriting in the expository-descriptive of genre. From the standpoint a conceptual of analysis of the task of narration, it would follow that attribution skill would play a role in narrative writing proficiency.On the other hand, attribution ability maybe a less salient element of skill in writing exposition or description, to the extent that we can apply Cumminss and Colliers description of academic discourse proficiency to the present studys descriptive-expository writing task. However, because of cognitive and linguistic demands imposed by the design characteristics academic of discourse and the influence general literate experience, this proficiency might of simply develop later, and hence show evidence of the influence attribution less of ability i its early stages. n either account, the developmental relationship n O between the twogenres of writing can be understood more fully by means of Cumminss and Colliers second language developmental findings. The fundamentalrole played by narrative ability with its heavy provision of interpersonal, pragmatic scaffolding further illuminated by observations is the of Egan (1988) about young school childrens thinking and learning, and by Bruner ( 1 9 9 0 , 1 9 9 6 ) who speaks of the modeof thinking and feeling helps children (indeed people generally) that create a version of the worldin which, psychologically, they envisage a place can for themselves-a personal world. I believe thatstory making, narrative, is what is needed for that.. . (Bruner, 1996, p. 39) .

On the other hand, would we not expect similar developmental patterns why to underlie both narrative and descriptive writing ability?Not only can a fair

Childrens Attributions


case be made for the fundamental played by narrative modes of shaping role experience in various other genres, but it is possible to find evidence of the narrative genre within the structure some good examples of expository of writing. It is probably no accident that scientists often refer to the interpretation of their data as telling a story. Some of the most engaging-albeit popular-recent scientific writing is framed in explicitly narrative terms:I think of the work of Stephen Jay Gould or Oliver Sacks. It is therefore perhaps not so surprising to find that strong metapragmatic understanding is associated not onlywith successful and fluent narrative writing,but also with the better and more fluent writing children undertake in the expository and descriptive genres. Common discourse abilities could underlie ability in narrative and both expository-descriptivegenres.Thesediscourseabilitiescouldmanifest themselves not only cognitively (Stein& Trabasso, 1992) but also in terms of socialization into what Bamberg (chap. this volume) terms how to 3, feel and how to understand the emotional-and we would add intentional- states of others.

The author wishes to acknowledge the assistance of JaneWakefield, who assisted with all aspects of data collection and analysis, and JonShapiro, for his collaboration on the design of the writing tasks. The collection of data was two supported by a grant R. Watson, K. to Reeder, J. Shapiro, and H. Goelman from the Social Sciences & Humanities Research Council of Canada. Analysis ofdata was supported by a grant to the author from the Hampton University of Fund, British Columbia.

Abbeduto, L , . Nuccio Bibler, J, AI-Mabuk, R., Rotto, P., & Maas, F.(1992). Interpreting and . responding to a spoken language: Childrens recognition and use of a speakers goal. Journal of Childhnguage, 19,677-693. Anthony, R.(1984). Metalinguistic awareness and readingin bilingual education: Implications o f a Unpublished Ph.D. dissertation, University o f Toronto. Chilcotin/English study. Babelot, G. ( 1 9 9 6 , July) Comprehension by young children adultssentences according to the of social o f Child situation. Paper presented at the Seventh International Congress for the Study Language, Istanbul, Turkey. Baroni, M.R., & Axia, G. (1989). Childrens meta-pragmatic abilities and the identificationof polite and impolite requests. First Language, 9,285-297. Bates, E.(1976). Language and context: The acquisition ofpragmatics. York Academic Press. New . based Becker, J (1981,April). Preschoolersjudgments ofspeaker status on requests.Paper presented at the BiennialMeeting o f the Society for Research in Child Development, Boston. Becker, J (1988a). I cant talk, Im dead Preschoolers spontaneous metapragmatic comments. . Discourse Processes, 11,457-467. Becker, J (1988b). The success o f parents indirect techniques for teaching their preschoolers . pragmatic skills. First Language, 8,173-182.



Bereiter, C. (1980). Development in writing. In L.W. Gregg & E. R. Steinberg (Eds.), Cognitive processes in writing(pp.73-93). Hillsdale, NJ: LawrenceErlbaum Associates. knowledge Bernicot, J., & Laval, V.(1993, July). Promises in children: Comprehension, metapragmatic and adaptation to the situation. Paper presented at theSixth International Congress for the Study of Child Language, Trieste, Italy. Bialystok, E. (1978). A theoretical model of second language learning. Language Learning, 28, 69-84. Bialystok, E., & Ryan, E. (1985). A metacognitive framework for the development first and secof ond language skills. InD. Forrest-Presley, G. Mackinnon, & T. Waller (Eds.), Meta-cognition, 102-128). New York AcademicPress. cognition, and human performance (pp Bruner, J. (1990). Acts ofmeaning. Cambridge, MA: Harvard UniversityPress. Bruner, J. (1996). The culture ofeducation. Cambridge, MA:Harvard UniversityPress. Collier, V.(1987). Age and rate of acquisition of second language for academic purposes. TESOL Quarterly, 21,617-641. Cummins, J. (1981). Age on arrival and immigrantsecond language learning in Canada: A reassessment. Applied Linguistics,2, 132-149. De Paola, .(1978). Pancakesfor breakfast. T New York HarcourtBrace Javanovich. Donaldson, M. (1978). Childrens minds. Glasgow: Fontana/Collins. . in Downing, J (1987). Changing perspectiveson literacy. In D. Wagner (Ed.), Thefuture ofliteracy a changing world (pp. 27-47). Elmsford, NY: Pergamon Press. Egan, K. (1988). Primary understanding. New York Routledge. Elley, W. P., Barham, I., Lamb, H., & Wyllie, M. (1976). The role of grammar in a secondary school curriculum. Research in the TeachingofEnglish,10(1), 5-21, Ellis, R. (19 8 4). Can syntax taught? A study effects offormal instruction be of the on the acquisition of WH questions by children. Applied Linguistics, 5(2), 138-52. Ervin-Tripp, S., & Gordon, D. (1985). The development of requests. In R. Schiefelbusch (Ed.), Language competence:Assessment and intervention (pp. 61-95). San Diego,C A College Hill. Occasions andstrucErvin-Tripp, S. & Kiintay, A. (1996, July). Conversational narratives ofchildren: for tures. Paper presented at theSeventh International Congress the Studyof Child Language. Istanbul. Freire, P. (1972). Pedagogy ofthe oppressed. Harmondsworth Penguin. Garton, A., & Pratt, C. (1990). Childrens pragmatic judgements of directand indirect requests. First Language, IO, 51-59. Gibbs, R. (1987). Linguistic factors in childrens understanding of idioms. Journal of Child Language, 14,569-586. Greenberg, K. L. (1994). Writing evaluation. In A. C. Purves, L. Papa, & S. Jordan (Eds.), Encyclopedia ofEnglish studies and language arts: Vol.2 (pp. 1307-1309). New York Scholastic. Halliday, M. (1975). Learning how to mean. London: Edward Arnold. Hickmann, M., Champaud, C., & Bassano, D. (1993). Pragmatics and metapragmatics in the development of epistemic modality: Evidence from childrens reports of think-statements. First Language, 359-389. James, S. (1978). Effect of listener age and situation on the politeness of childrens directives. Journal ofPsycholinguisticResearch,7.307-317. Krashen, S. (1981). Second language acquisition and second language learning. Oxford, UK: Pergamon. Kroll, B. (1978). Cognitive egocentrism and the problemof audience awareness. Research in the Teaching ofEnglish, 12,269-281. Ledbetter, S.,& Dent, C.(1988). Young childrens sensitivity direct and indirectrequests structo tures. First Language, 8,227-245. Leonard, L. B., &Reid, L. (1979). Childrens judgments of utterance appropriateness. Journal of Speech andHearingResearch, 22,500-515.


Childrens Attributions


Levinson, S. (1983). Pragmatics. Cambridge, U K Cambridge University Press. Long, M. (1983). Does second language instruction make a difference? A review o f research. TESOL Quarterly, 17,359-382. McLaughlin, B. (1990). Restructuring. AppliedLinguistic, 11,113-128. Mellow, J. D. (1996). A longitudinal study ofthe efects ofinstruction on the development ofarticle use by adult Japanese ESL learners. Unpublished doctoral dissertation, University o f British Columbia, Vancouver, Canada. ODonnell, R. C., Griffin, W. J., & Norris, R. C. (1967). Syntax ofkindergarten and elementary school children: A transformational analysis.Champaign, I L National Council o f Teachers o f English. OHare, F. (1973). Sentence combining: Improvingstudent writing withoutformalgrammar instruction. Urbana, I L National Council o f Teachers of English. Perera, K.(1984). Childrens writing and reading: Analysing classroom language.Oxford, UK: Basil Blackwell. Reeder, K. (1996). Childrens attributions o f speakers pragmatic intentions. In M. Aldridge (Ed.) Child language(pp. 149-164). Clevedon, U K Multilingual Matters. Reeder, K., Shapiro, J., Watson, R., Goelman,H. (Eds.).(lgg6). Literate apprenticeships:The emer& gence oflanguage and literacy in the preschool years. Norwood, NJ: Ablex. . Reeder, K., & Wakefield, P J. (1987). The development of young childrens speech act comprehension: How much language is necessary? Applied Psycholinguistics, 8,1-18. Shatz, M. (1978). On thedevelopment o f communicative understandings: An early strategy for interpreting and responding to messages. Cognitive Psychology, 10, 271-301. Shatz, M., & Gelman, R.(1973). The development o f communication skills: Modifications in the speech o f young children as a function o f listener. Monographs ofthe SocietyforResearch in Child Development,38 (5,Serial No. 152). Sharwood Smith, M. (1981). Consciousness raising and the second language learner.Applied Linguistics, 11,159-168. Stein, N., & Trabasso, T. (1992). The organisation o f emotional experience: Creating links among thinking, language, and intentional action. Cognition and Emotion,6, 225-244. Wakefield, J. (1998). Childrens understanding o f expressions o f possibility and necessity. In A. Aksu-Kog, E. Erguvanli-Taylan, A. S. oszoy, &A.Kiintay(Eds.), Perspectives on language acquisition: Selectedpapersfrom the V I M International Congressfor the Studyofchild Language (pp. 97-107). Istanbul: Bogaziqi University Press. Wllkinson, A., Barnsley, G., Hanna,P,, &Swan, M. (1980). Assessing language development.Oxford, U K Oxford University Press. Wllkinson, L. C., Wllkinson, A. C., Spinelli, F., & Chiang, C. P. (1984). Metalinguistic knowledge o f pragmatic rules in school-age children. Child Development,55, 2130--2140. W~llows, & Ryan, E. (1986). The development of grammatical sensitivity and its relationship to D., early reading achievement. ReadingResearch Quarterly, 21,253-266. Winner, E. (1988). The point o f words: Childrens understanding o f metaphor and irony. Cambridge, M A Harvard University Press. Yaden, D., & Templeton S. (Eds.). (1986). Metalinguistic awareness and beginning literacy. Exeter, NH: Heinemann.




Examples of attribution of Speakers intention

Speaker wants act A

She wants them to draw apicture

She wants her to learn something

Speaker wants to know proposition P


To find out if she could make a picture To see ifshe was interested in playing with the blocks

Speakers role
Because probably shes the art teacher and she would like the little girl to make a picture

She wants to encourage the kids to learn how to use the blocks and stuff

Hearer wantsact A

The little kid was wondering ifhe could make apicture The little boy might want toplay with the blocks

Hearers role
She (pupil) didnt knowhow. She has to learn

Institutional routine

Thats what youre supposed to do at school Maybe because there were too many children at the other stations


Childrens Attributions


Examples of reported bases for attributions

I looked at it and just thought I I justhave a feeling

Role knowledge


Because shes a teacher and shes supposed to teach the kids Well, if youre in school its good to learn how to draw, not just be able to print. .. Because thats what my teacher. her.

..she doesnt like people hanging around

Proxemic or kinesic cues

She (teacher) had the table right beside her

Shes (pupil) just wandering around

Performance of the utterance


By the words on the tape Because she said Can youdraw a picture?

The utterances formal features

Because she was mixing thewords up but I could tell The expression in her voice

Illocutionary act
By the questions sheasks

Pragmatic strategyor feature of discourse

Because the little girl probably doesnt like Art and the Art teachersaying is please because she wants herto Well, its hardto explain. She said it, but said it words not saying it but she in said it in her words. Its hard to explain. She said it in a diflerent way than you usually say it but she wouldntsay it so it would insult the kid.

This Page Intentionally Left Blank

Author Index

Abbeduto, L.,145,146,159 Aksu-koc, A., xi, xvii, xviii Alici, D., xvii, xviii AI-Mabuk, R.,145,146,159 Andersen, E., xvi, xviii, 96,108,110 Anthony, R.,147,159 Argeman, K., 1 6 , 2 7 . Astington, J, xix, xvii Avidor, S., 2 6 , 2 9 Axia, G., 145,159 Bloom, L.,113,118,140,141 Blum-Kulka, S., 73, 9 0 Bowerman, M., 31.32,51 Bretherton, I., 137,140 Brice Heath, S., 7 3 , 7 4 , 9 0 Brossard, M., 117,141 . Bruner, J, 96,110,158,160 Budwig, N.,6 , 7 1 6 Bunkachoo, 93,94,97,99,105,106,

Babelot, G., 146,147,159 Bamberg, M., xi,xiv, x v , xvii, 4 , 2 7 , 5 5 , 60,61,71,72,73,90,159 Barbiere, M., 119140 Barham, I., 147,160 Barnsley, G., 156,161 Baroni, M., 145,159 Bassano, D., 146,147,16 o Basso, E.,58,71 Bates, E., 144,159 Batoreo, H., xi, xii, 31,33,34,35,36,37, 3 8 , 4 1 , 4 2 , 4 3 > 4 4 , 4 5 , 4 6 , 4 751 > Bazanella, C., 4 , 2 7 Becker, J., 144,145,147,159 Beeghly-Smith, M., 137,140 Ben-Haviv, O.,1 9 , 2 7 Bereiter, C., 157,160 Berman, R.,xi, xii, xiii, xix, xviii,1, xiv, 2,3,4,10,11,12,14> 15>16,18,19, 2 0 , 2 3 2 2 5 , 2 6 , 2 7 , 2 8 , 2 9 , 3 L33251, 73,75,86,87,88,90 . Bernicot, J, 1 4 6 , 1 6 0 Berthoud-Papandropoulou,I., 119,140 Bialystok, E., 147,160

Calleri, D., 4 , 2 7 Capatides, J, 113,140 . Cassano, L.,3 6 , 5 1 Champaud, C., 1 4 6 , 1 4 7 , 1 6 0 Chavegatto, V., 3 6 , 5 1 Chiang,C., 96,112,144,161 Choi, S., 31,32, 51 Chvany,C.,zo,z8 Clancy, P., xix, m i , xviii, 9 7, 0,102, IO 105,110 Clark, E., 9 , 7 1 5 Colavita, F., 119,140 Collier,V., 158,160 Cook-Gumperz, J, 9 8 , n o . Corsaro, W., 98,108,110 Crago, M., xix, xviii Cummins, J, 158,160 .

Damrad-Frye, R., , 2 7 4 Davitz, J , 5 6 , 7 1 . Day, C., xix, xviii de Paola,T., 151,16 o




Dent, C., 144,160 Dickenson, D.,74,91 Doi, T , .110 Donaldson, M., 147,160 Donelan, N., 114,140 Downing, J, 147,160 . Dromi,E., zo,23,27,~8 Duarte, I., 31.34,35,51 Dunn, J, 113,114,117,118,137,140,141 .

Grize, 7.-B., 119,141 Gruendel, J, 9,z9 . Guimaraes, A.,36,51 GUO, 119,137,141 J.,

Haden, C., 7 4 9 0 Haight W., 114,117,118,140,141 Halliday, M., 144,16o Hanna, P., 156,161 Harkins, D., 74,90 Harre, R.,59,72 Hayashi, O.,93,111 Heath, S.,xix, xviii Heelas, P ,5 6, 72 . Hendriks, H.,n,z8,31,36,52 Heritage, J, 114,141 . Herman, J., 4,28 Herrera, C., 114,140 Hickmann, M., xiii, xix, xii, 4,5,11,28, 31,32,33,37>49>50> 52,73,90, 146,147,160 Hicks, D.,xii, xix Hoff-Ginsberg, E., 72 Holzman, M,, 96,111 Hood, L.,113,141 Hopper,P.,zo,28,66,72 Hough-Eymi, W., xix, xviii Hymes, D., 96,111

Edwards, D., 9,71 6 Egan, K., 158,160 Eisenberg, A., xix,v , 114,117,118,140, x 141 Elley, W., 147,160 Ellis, R.,147,160 Ervin-Tripp, S., xix, mi, 96,108,110,

Faria, I.H., xii, 31,34,37,51 xi, Favre, C., 119,140 Fivush, R.,74,90 Fletcher,P ,59,71 . Freire, P., 147,160

Garton, A.,145,160 Garvey,C., xix, xviii, 96,111,114,117, 118,12l,l4O,141 Gelman, R.,144,161 Gibbs, R.,143,160 Gillett, G.,5 9,72 Giora, R.,2,28 Girnius-Brown, O.,114,141 Gleason, J.,59,71,96,111 Goddard, C., 31,51,57,71 Goelman, H., 148,159,161 Goodwin Harkness,M., 114,141 Gordon, D., 6,10 8,m , 143,160 9 Greenberg, K.,150,160 Greif,E.,96,111 Griffin, W., 150,157,161

Ide, S., 94,95,111 Iijima, T., 105,111

James,S., 108,111, i44,160 Johnson, C., xi Johnston, J., 32,52

Kahanowitz, S.,n,28 Kail, M., 4,5,28 Karmiloff-Smith,A., 73,87,90



Katzenberger, I., 11,12,14,25,27,28 Keren-Portnoy, T., 119,141 Kern, S., 4, IO,28 Kernan, K.,2,29 Kikuchi, Y., 94,111 Kindaichi, H., 93,111 Kochanska, G.,114,141 Kokuritsu Kokugo Kenkyuujo,109,
11 1

Krashen, S., 147,16o Kroll, B., 147,160 Kuczinski,L., 114,141 Kuno, S., 95,111 Kuntay, A., xii, xix, 147,160 Kupersmidt, J, 26,29 .

Labov,W., 1,20, 26,29 Lamb, H., 147,16 o Lamb, S., 137,141 Lampert, M., 119,137,141 Laugaa, A., 117,141 Laval, V.,146,160 Ledbetter, S., 144,160 Leff, J, 5 6 , ~ . Leonard, L., 144,160 Levin, I., 27 Levine, L.,58,72 Levinson, S., 143,161 Liang, J, 1 ~ 2 8 . Liwag,M., 56,58,72 Long, M., 147,161 LUtZ, C., 57,58,72

Mayer, S., 137,141 Maynard, S., 94,111 McCabe, A., xii, xiii, xix, , XX, 2,3,9, xv 11,18,29,73> 74>90 McLaughlin, B., 147,161 McNew, S., 137,140 Mellow, J, 147,161 . Menig-Peterson, P., 2,29 Miller, P., xix, xv Minami, M., 9,29,73,74,90 Mizutani,N., 93,94,95,97,108,111 Mizutani, O.,93,94,95,97,108,111 Mori, J, 111 . Morioka, K., 97,111 Munn, P,, 113,114,117,118,140 Muraishi, S., 97,103,104,105,111 Murata, K., 97,112 Musatti, T., 137,141

Nakamura, K., xi, mi, xvii, xviii, 9 7, 3,9 110,112 Nakane, C.,93,112 Nelson, K., xvii, xix, xi, 9,29 Nicolopoulou, A., xviii, xx Ninio, A., xiii, xx Niyekawa,A., 94,95,112 Norris, R.,150,157,161 Nuccio Bibler,J., 145,146,159

Ochs, E.,xv, xviii,xx, 58,72,73,90 OConnor,M., 108,111 ODonnell, R.,150,157,161 O'Hare, F., 150,161 Oishi, H., 93,112 Okuyama, M., 94,112 Oz, A., I, 29 Ozcaliskan, S., 6 0 , 72

Maas, F, 145,146,159 . Mackie,V., 97,111 MacWhinney, B., 37,52,59,71 Marchman,V., 73,90 Martin, S.,93,95,111 Martlew, M.,v , xx x Masiello,T , .114,117,118,140,141 Matsumoto,D., 56,72 Mayer, M.,xii,xix, 3,4,29,33,52,74, 90>91

Parush, T , .119,141 Perera, K.,156,161 Perez-Leroux,A.,xvii, xx




Peterson, C., xii, xiii, xix, , x x , 2,3,9, W 11,18,29,73>74>0 9 Piaget, J., xviii, xx Pitcher, E., 1 ~ 1 5 , 29 Polanyi, L., 1,29,73, 90 Pontecorvo, C., 114,141 Potter, J, 6 9 , p . Pradl, G., 9,15,29 Pratt, C.,145,160 Prelinger, E.,11,15,29

Shantz, C., 114,117,141 Shapiro, J, 148,159,161 . Sharwood Smith,M., 147,161 Shatz,M, 25,29,96,112,144,161 Shen,Y., 1,2,20,27,28,29 Shibata, T., 93,111 Silva, E.,51 Silveira,M., 5 1 36, Sinclair, H., 113,114,119,127,136,137, 139,140,141 Slobin, D., xii, xiii, xiv, xix, 2,3,4, xx,

Radke-Yarrow,M., 114,141 Ratner, N., 59,71,96,110 Ravid, D.,20,26,29 Reeder, K., xiv, xvii, 143,146,147, xi, 148,149,150,159,161 Reese, E.,74,90 Reid, L., 144,160 Reilly, J, 2, 29 . Reimann, B., x v , xx Reinhart, T.,1,20,29 Renner, T., 74,90 Rojo-Torres, M., 141 Roland, F., 5,11,28 Rosenberg, J, 108,111 . Rotto, P., 145,146,159 Roy, C.,9 6,110 Rumelhart, D., 1,2,29 Ryan, E., 147,160,161

30,31,32,33*34,5b 52,73>75,86, 87,88,90 Slomkowsky,C., 114,140 Smoczynska, M.,31,33,49,52 Snow,C.,xiii, x v , xx, 73,74,91 Sorsby, A.,x v , xx Souza, R., 6,51 3 Spinelli, F., 96,112,144,161 Stavans,A., xi, xiii, xviii, 73,74, 88, 91 S t e i n , N . , 5 6 , 5 8 , 6 7 , 6 8 , 7 2 , 1 5 9 r161 Sternberg, M., 1,30 Sutton-Smith, B., n,30 Swan, M., 156,16 1

Talbot, J, 61,72 . Talmy,L.,32,33,50,52 Tardif, T , . 117,141 Templeton, S., 147,161 Tesla, C., 114,141 Thompson, S.,6 6,72 Trabasso,T , . 56,58,72,159,161 U Umiker-Sebeok, D., ~ 3 0

Said, E., 1,29 Sanchez-Lopez, I., 4,5,28 Sarbin, T., 70,72 Scheuer,N., 119,140 Schieffelin,B., x v , xviii, x x , 58,72,73, 90 Schlesinger,I., 119,141 Scollon, R., xiii, 73,9 1 xx, Scollon, S.,xiii, xx, 73,91 Sebastian, E., 26,29 Segal, R., 12,29 Seidman, S., 9,29

Van Dijk,T., 2,30 Veneziano, E.,xi, m, xvi, xvii, 113,114, Vygotsky, L.,xviii,xx, 137,141



Wade, E., 5 8 , p Wakefield, J, 146,147,161 . . Walters, J, 108,112 Watson, R.,xv,xx, 148,159,161 Wlerzbicka,A., 31,52,57,58,72 4 Wlgglesworth, G., xi, xiii, xviii, , 3 0 , 73,91 Wllkinson,A., 96,112,144,156,161



Willows, D., 147,161 Winner, E., 143,161

Wlttgenstein, L., 6 0 , 7 2 Wyllie, M., 147,160

Yaden, D., 147,16 1 Yoshioka,Y., 94,112

This Page Intentionally Left Blank

Subject Index

A Action structure,25 Adverbs, 19-20 Affective capacities, xi Age-related analyses, xiii American Psychological Association,

Cognitive-linguistic representations, xv Communicative contexts,x i Communicative intentions,145 Conversational abilities, childs,v , 113 x Conversational discourse, xvi Conversational goal, childs, 145 contextual information in, 146 Conversationally embedded narratives,

Conscientizacao, 147

Anger, xiv,55-71 See also Personal anger narratives; Grammar, of anger; Explanatory discourses, on anger; Attribution ability, 158 xiv, Attribution of intentions, 143 constructed in interactions,
144--145,149--150 See also Pragmatic attributions

Background information, 9 , 1 8 , 2 5 , 3 9 motivation, 8 setting, 2 0 , 2 2 3 Basic interpersonal communicativeskill (BICS), 158 Berger, Andrea,71 Bhatia, Sunil,71

Coulon, Darine,140 Crosscultural differences, xviii Crosscultural narrative investigation Australian-English subjects,73-90 Israeli-Hebrew subjects,73-90 Crosslinguistic study,20 Crosslinguistic variation, xviii

de Paola, Tomi, 151,160 Digressions, xiii,75 See also Narratives, parentkhild, asides; Discourse, xi, xiv, x v , 7 5 0 , Discourse-learning contexts, xiii Discourse-making process, xvi Discourse-organizational capacities, xiii Dynamic actions,32 Dynamic discursive perspective, xi Dynamic negotiation,m i

Cambridge University Press, 0 5 Cat and horse series,11 Cat Story, 7 , 3 9 - 4 0 , 4 3 - 4 6 , 4 8 , 5 0 3 CHILDES system,37 Childrens theoryof mind, 137 Cognitive skills, xi, xii, xvii, 25 Cognitive-linguistic capacities, xviii

Elicitation methods of, xii, xvii,1 6 n,



procedures, 146 setting, 9,21,26 techniques, xi See also Picturebook elicitations Emotion concepts, 56,58-59,69 Emotion talk,x v , 55-60,66 attributing blame in,6-6 8,70 6 elicitation techniquein, 6 8 eliciting empathy in, 66-68,70 See also Anger; Sadness constructions; Emotional development, xiv, 71 Emotions in linguistic forms, 59, 69 language-appropriate labels for, 57 natural perceptionof, 57 English composition, 147 Explaining, origins of in interpersonal relations, 113,115, 137 Explanatory capacity, childs and conflicts, 117,123-124,134,

representation of movement in, 31-50 semantic and syntactic characteristics of,38,45,47,50 and spatial references, 36 theoretical framework 33,40,44 of,

and verbalizations, 119--120 in maternal interactions, 113-140 role of justifications in, 114, 119--121,
also Justifications) types of explanations in, 130-131,138-140 role of oppositions 114-121, in, 124-127,130-140 resolution of, 121--122, 124 types of, 117--11g, 125,128-12g Explanatory discourses on anger, 55,63-64 (seealso Anger; Grammar, of anger; Personal anger narratives) on sadness, 6 4-65 (see also Sadness constructions; Grammar, sadof ness) European Portuguese,37,41,47,49 and language acquisition, 49 locative character of, 40 morphological and lexical markers in, 42,49-50
n.4-127,130--134,136--140 (see

Fight stories elicitation settings in, 9-14 setting elements in, 3, 6-9,13 temporal location,17 transition markers in, 19 Fonds national suisse de la recherche scientifique, 140 Form:function relations, xi, xiii, 15mi, 16,38,60,67 Framing information,3,7 Frog stories,75,91 as crosslinguistic study, 25 elicitation settings in, 9-14 movement in, 33 setting elements in, 3-6,13-14 temporal location, 16-17 and tense shifting,ZO-ZI, 23 transition markers in, 18

Gauthier, Isabelle, 140 Global discourse level, xii Global narrative action structure, 20 Global plot-organization as action structure, 2,1z Grammar, 147 of anger, v , 6 0-6 5 x of sadness, v , 6 0-65 x

Hampton Fund, 159 Hebrew context,3-4, IO,13-14,16, 1g-20,22-25 See also Crosscultural narrative investigation, Israeli-Hebrewsubjects; Hierarchically organized (story), xiv Horse Story, 37-40,43-46,50



Intentional representations, xvii Interlocutor queries, 2 Interpersonal exchange activity, xiii Italian language,4 9

Language proficiency, 147,158 Language-related analyses, xiii Lexical and morphological cues, See also Motion, lexical and morphological markers for; Lexical devices, xiii Linguistic awareness, childs, 147 Linguistic forms, 144,147 Linguistic-pragmatic knowledge,143,
156 49

Japan Society for the Promotion of Science, n o Japanese language acquisition of polite forms, 93-110 (see also Keigo; Polite language; Polite speech) Japanese narrative interactions, 74 Junta Nacional de Investigacao Cientifica e Tecnologica,0 5 Justifications, xi, xiv,v , xvi, 144,146 x See also Explanatory capacity, childs, role of oppositions in;

Linguistic representations, xvii Linguistic structures, xv Linguistic utterances,143 See also Utterances, childs; Linguistics Program of National the Science Foundation,27 Literacy, xi, xii,6 , 7 4 , 8 6 , 1 4 3 , 1 4 7 2 Literacy development, 73 Locative expressions, 31-32 Locative states, 32 Locative structures, 48

Keigo(politeness), 9 3 - 9 4 , 9 6 - 9 8 ,
106-10 9 bikago, 9 3 , 9 5 - 9 6 , 1 0 5 , 1 0 9 kenjoogo,93-95, I O O - ~ O Z , 106 sonkeigo,93-94,100--101,106 teineigo, 93, 95,97,100-109 See also Polite language; Polite

Make-believe stories, xvii, 11-12,16 Metacognitive comments,8 , 2 5 Metalinguistic operations, xvii, xi Metapragmatic awareness, xviii competence, xiv knowledge, 145,157 operations, xi utterances, 145 Metarepresentational operations, xvii, xi Metatextual comments,8 Monologic narrative, 2 Morphology, transparent, 32 Morphosyntactic devices, xiii Motion global devices for, 32 lexical and morphological markers for, 32-34> 38,42,49 local devices for, 32 Motivational elements, xii, , 1 4 6 in narratives,4-6,25

speech Know that type, xvii Know-how ofthe mind, 137

Language in relation to emotion, 6 , 5 8 , 6 0 , 5

in relation to thought, 56,58,70-71 structure, xii,33 Language acquisition, 6 , 3 6 1 and adult grammar,48-49 of mother tongue, 31 and organization of discourse information, 31 and spatial relationships, 31



Narrative ability, xiv, xii, 16,73-75 xi, beginnings, 21,25 construction, 25 development, xi, xiii, 25,143,147 discourse, 2,11 form, 18 functions, xii, 15 genre, xii, 16,73 openings, 15-17 relation o f linguistic form:narrative function in, 26 scene, 1 - 2 schema, 25 structures, xii, 1, 9 supported with pictures, xvii transitions, 15,17-20 writing, 155,157-159 Narratives crosscultural differences in, 73, interactive, 74-75,88 juvenile, 9 parentlchild, 74-75,79,81-82, asides, 77, 84,86 child utterances, 77,84 focusing devices,77,83 non-story-related clauses, 77, story-related clauses, 77, spontaneous, xvii National Science Foundation,110 Natural semantic metalanguage, 31 Nelson, Keith,71
80-83 83-84 84-85 85-86

Personal experience narratives, xiv, xvii, Piaget, Jean,136 Picture-based elicitations, 5,12,14 Picture-based tasks, 26 Picturebook narratives,xii, 2 Pictured storybook, 3,11-12,14 Plot,xiv, 8,17-20,23 Polish language,32-33,49 Polite language linguistic formsof, xvi, 9 6,9 social context for, 96-97,105-106, Polite speech, 143-145 in social contexts,144 Portuguese language,32-34,36 See also European Portuguese Pragmatic attributions,151,155-157 Pragmatic intentions, xiv, 144,147 Pragmatic knowledge,x v , xvii, xviii Pragmatic representations, xvii Prompting, 74
See also Keigo; Polite speech;





Reading, 147 Reynolds, Ayden,71 Roadville and Trackton,74 Role play,xviii, 97-99,101-104,

S Sadness constructions, xiv, 55-71 See also Grammar, of sadness; Explanatory discourses,on sadness; Scene setting,47-8,IO,12,17,22-23 crosscultural differences in, 27 crosslinguistic differencesin, 27 descriptions of, 19-20 transitions in,26 and communicative context,13-14 by elicitation method, 13-14 by narrative genre,13

Oral stories,xii, 1 ~ 1 7

Pancakesfor Breakfast, 151 Personal anger narratives,21,55 Personal conflict narratives, xii



Scene-setting elements, 1-3,9,14,11,15, elicitation context of, 3 linguistic forms of, 3 Scene-setting information, 5 , 1 2 Scene-setting orientation, 9 Shapiro, Jon, 159 Sicard, Michelle,71 Social abilities, childs, v , 113 x Social-cognitive knowledge, xvii Social Sciences & Humanities Research Council of Canada, 159 Spatiotemporal frame,xii, xiii Speakers intentions, 146 Spencer Foundation, n o Spracbbewusstbeit, 147 Start o f story, 2,12 Story grammar,1 See alsoGrammar Story construction,xiii Story openings, xii stereotypic lexical items in, 25 See alsoNarrative openings Story planning abilities, xii Story plotline,1 See alsoPlot Story-setting elements, 5 Story setting functions,14 Story structure,3 8 - 4 ~ 4 6 - 4 9 Storybook, 10 5 open-ended, textless, xiii Storybook reading,73-74 Storytelling, 24,105 conversational style in,xiii performance, 2,14 tasks,xii,xiii, 7 7 , 8 4 , 8 6 - 9 0

Tel Aviv University,27 Temporal and spatial information, 74 Tense shifting, 21-22, 26 and motivation,2 0 from background to foreground,

textual devices for,24 Theory of mind framework, xvii Transition markers, 17-20 Transitional narration, 9

University o f British Columbia, 159 University of California, 27,110 U.S.-Israel Binational Science Foundation, 27 US.-Japan Educational Commission (Fullbright Program), 110 Utterances, childs, xiv, 143 -144

Verb tense, shiftin, 18,19-24 and linguistic compensation,2 0 , 2 4 Verbs, 38

Wakefield, Jane, 159 Writing, xiv,143,147,150-151,155-159

Zero marking,18