The Development of the Mediated Mind Sociocultural Context and Cognitive Development
Edited by
Joan M. Lucariello Boston College
Judith A. Hudson Rutgers University
Robyn Fivush Emory University
Patricia J. Bauer University of Minnesota
LAWRENCE ERLBAUM ASSOCIATES, PUBLISHERS 2004 Mahwah, New Jersey London
This edition published in the Taylor & Francis e-Library, 2008. “To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.” Copyright © 2004 by Lawrence Erlbaum Associates, Inc. All rights reserved. No part of this book may be reproduced in any form, by 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 07430-2262 Cover design by Kathryn Houghtaling Lacey Library of Congress Cataloging-in-Publication Data The Development of the Mediated Mind: Sociocultural context and cognitive development / Lucariello, Joan; Hudson, Judith A.; Fivush, Robyn; and Bauer, Patricia J. p. cm. ISBN 0-8058-4473-2 (cloth: alk. Paper). Copyright information can be obtained by contacting the Library of Congress.
ISBN 1-4106-1042-X Master e-book ISBN
To Katherine: Our Teacher, Mentor, and Friend.
Contents
Contributors
ix
Preface
xi
1 Katherine Nelson’s Vision of the Mediated Mind
1
Judith A. Hudson, Joan Lucariello, Robyn Fivush, and Patricia Bauer
2 Two Kinds of Knowledge Acquisition
13
Jean M. Mandler
3 New Insights Into the Functions, Development, and
33
Origins of Theory of Mind: The Functional Multilinear Socialization (FMS) Model Joan Lucariello
4 Meaning and Use: Children’s Acquisition of the Mental
59
Lexicon Janet Wilde Astington and Joan Peskin
5 Voice and Silence: A Feminist Model of Autobiographical
79
Memory Robyn Fivush
vii
viii
CONTENTS
6 Developments in Early Memory: Multiple Mediators
101
of Foundational Processes Patricia J. Bauer and Melissa M. Burch
7 The Development of Future Thinking: Constructing
127
Future Events in Mother–Child Conversation Judith A. Hudson
8 Narratives, Gossip, and Shared Experience: How
151
and What Young Children Know About the Lives of Others Susan Engel and Alice Li
9 Acquiring Art, Spoken Language, Sign Language, Text,
175
and Other Symbolic Systems: Developmental and Evolutionary Observations From a Dynamic Tricky Mix Theoretical Perspective Keith E. Nelson, Patrick L. Craven, Yue Xuan, and Marnie E. Arkenberg
10 Literacy and the Mediated Mind
223
Bruce D. Homer
11 Katherine Nelson: Contextual Functionalist
239
Jerome Bruner
12 The Virtues of Rigorous Interdisciplinarity
245
Merlin Donald Author Index
257
Subject Index
265
Contributors
Marnie E. Arkenberg Department of Psychology Pennsylvania State University 130 Moore Building University Parke, PA 16802 Janet Wilde Astington Institute of Child Study University of Toronto 45 Walmer Road Toronto ON M5R 2X2
[email protected] Patricia J. Bauer Institute of Child Development University of Minnesota 51 East River Road Minneapolis MN 55455
[email protected]
Melissa M. Burch 845 Arbogast Shoreview, MN 55126 Patrick L. Craven 708 Whitehall Road State College, PA 16802 Merlin Donald Dept of Psychology Queen’s University Humphrey Hall, 50 Arch Street Kingston ON K7L 3N6
[email protected] Susan Engel 1023 Mill River Rd. Great Barrington MA 01230
[email protected]
Jerome S. Bruner 200 Mercer Street New York NY 10012
[email protected]
ix
x Joseph A. Glick Grad. School and University Center City University of New York 365 Fifth Ave. New York NY 10016
[email protected]
CONTRIBUTORS Joan Lucariello Lynch School of Education Boston College Campion Hall 239East 140 Commonweath Chestnut Hill MA 02467
[email protected]
Bruce D. Homer Dept. of Applied Psychology New York University 239 Greene Street, 5th Floor New York NY 10003
[email protected]
Jean M. Mandler Department of Cognitive Science Univ. of California, San Diego La Jolla CA 92093-0515
[email protected]
Robyn Fivush Department of Psychology Emory University 532 N. Kilgo Circle Atlanta GA 30322
[email protected]
Keith E. Nelson Department of Psychology Pennsylvania State University 414 Moore Building University Park PA 16802
[email protected]
Judith A. Hudson Department of Psychology Rutgers University Busch Campus New Brunswick NJ 08903
[email protected]
Joan Peskin Dept. of Human Development OISE-Univ. of Toronto 252 Bloor Street, West Toronto M5S 1V6 Canada
[email protected]
Alice Li Jefferson Medical College 244 South Nine St. 2nd Floor Philadelphia PA 19107
[email protected]
Xue Xuan Heritage Oakes 13B 10 Vairo Boulevard State College, PA 16803
Preface Three years ago, when we heard that Katherine Nelson was retiring from the Graduate Center of the City University of New York after 25 years on the faculty, we knew we had to mark the occasion. Katherine is such a remarkable influence on all of our lives in so many ways. Obviously, no student of developmental psychology doubts the contribution of Katherine’s work on our understanding of the cognitive development of children in social and cultural context. As students of Katherine’s, we also acknowledge the influence of her personal mentoring on our individual professional and intellectual development. How best to provide the personal and professional thanks for Katherine’s many contributions? Our first thank you was a symposium at the 2001 meetings of the Society for Research in Child Development in Minneapolis, Minnesota. In that symposium, the four of us presented papers outlining the ways in which our individual research has been and continues to be influenced by Katherine’s theoretical work in language, concepts and memory. A more personal thank you followed during a party in Katherine’s honor that we hosted at Patricia Bauer’s home. This celebration was extremely well attended by Katherine’s students and colleagues (we can attest to how much food and drink was consumed!). Many of Katherine’s students came forward with poignant and funny stories of the ways in which Katherine has touched their lives. Our second thank you was a special issue of the Journal of Cognition and Development in Katherine’s honor. This was an extremely unusual honor, as the journal does not normally allow this type of special issue. Katherine, however, had been a founding member of the journal and had been the editor of Cognitive Development (the intellectual predecessor of the Journal of Cognition and Development) from 1990-1995. During Katherine’s editorship the journal became the premier outlet for publishing innovative research in cognition and it’s development. xi
xii
PREFACE
Again, Katherine’s intellectual prowess shaped the field. The articles published in the special issue showcased the depth and the diversity of her influence. This edited volume is our third thank you to Katherine. In bringing together many of her former students and current colleagues, the contributions to this volume represent an amazing array of theoretical and research interests. Katherine is a rare occurrence. Her work both draws on and contributes to ongoing debates in language and literacy, conceptual bases of knowledge, and memory, all within an historical, social, and cultural context. Katherine is a broad and integrative thinker and each of us has benefited enormously from our interactions with her, both through her published work and personally. Perhaps it is unusual for us to continue to develop forums for thanking Katherine, but then again Katherine is an unusual person. Given that we have had the privilege of learning so much from her, we are grateful that we can now share Katherine’s vision through this volume. —Joan M. Lucariello Judith A. Hudson Robyn Fivush Patricia J. Bauer January 2004
1 Katherine Nelson’s Vision of the Mediated Mind Judith A. Hudson Rutgers University
Joan M. Lucariello Boston College
Robyn Fivush Emory University
Patricia J. Bauer University of Minnesota
Katherine Nelson’s view of cognitive development as an interactive process, mediated by culture, context, language, and social interaction, has shaped the current field of cognitive development. Her retirement from City University of New York provided the impetus to bring together a collection that reflects on critical issues in current cognitive developmental theory that have been influenced by her theoretical vision. Katherine Nelson has been the field’s leading functionalist, arguing for a view of cognitive development as the result of real-world meaning-making and problem-solving activities. Katherine introduced a functionalist perspective to language development in the early 1970s in her groundbreaking work on individual differences (Nelson, 1973). She saw these differences as emanating from children’s different uses of language. Katherine was the first to bring a functionalist perspective to the study of concept and category formation, arguing that object 1
2
HUDSON, LUCARIELLO, FIVUSH, BAUER
concepts and object categories are based in understanding the uses of objects in people’s lives (Nelson, 1974). In the early 1980s, Katherine established that young children have organized event knowledge (Nelson, 1986). In so doing, she revealed that child cognition was very much about making sense of a sociocultural world. This research dramatically and permanently altered the conception of the preschooler from one of a deficient logician and/or a deficient strategist to that of a cognizant participant in the social world. It also revealed that making sense of sociocultural experience is one of the primary functions of cognition. Katherine has also explored the mediated mind with respect to the sociocultural basis for cognition. She was the first memory researcher to posit a social origin for autobiographical memory (Nelson, 1989). In her latest book (1996), which has the mediated mind as its topic, she viewed ontogenetic development in terms of stages or phases that recapitulate phylogenetic and evolutionary development. The impact of Katherine’s views on the sociocultural basis of cognition and her functionalist perspective on cognitive development are evident in the collection of chapters in this volume. The contributors examine ways in which cognition is embedded in everyday, meaningful activities and the role that social context and cultural symbol systems (e.g., language and text) influence children’s developing concepts and thought. The concept of the mediated mind is examined from a variety of perspectives and includes research in concept development, memory development, language learning, the development of literacy, narrative analysis, and children’s theory of mind. Several themes emerge from this collection of papers. First, cognitive development is viewed as a process of meaning making, with children striving after meaning in their everyday interactions. Thus, contributors examine how children come to make sense of their world and how sense making is embedded in ongoing everyday interactions with the objects and people in their world. Second, making sense of the world occurs within social interactions in which adults mediate children’s cognitive processes. Several contributors examine the ways in which the social interactions in which children engage exert a powerful mediating influence on the skills they develop and the meanings they construct. Third, language is both a tool and a process by which children make sense of the world. Language, as a cognitive phenomenon, must be understood as emerging from children’s developing understanding of the world around them and children’s emerging language abilities change the nature of their interactions with others.
FUNCTIONALISM AND COGNITIVE DEVELOPMENT In 1974, Katherine first proposed that concepts are constructed from a functional core. The idea that early concepts are based on what objects do and, in particular, what people can do with them and not necessarily what they look like, represented a unique approach to thinking about concept formation. In chapter 2, Mandler pro-
1. INTRODUCTION: THE MEDIATED MIND
3
vides a functional view of knowledge acquisition and representation that echoes many of Katherine’s original views. Mandler distinguishes two kinds of knowledge that correspond to different kinds of learning. One kind of knowledge is conceptual (i.e., explicit, declarative) and the other is inaccessible (i.e., sensorimotor, implicit, procedural). Mandler argues that recognition of these two types of knowledge is central to understanding cognitive development, particularly that in infancy. In Mandler’s terms, “Until it is routinely specified exactly what kind of knowledge is being dealt with in any particular kind of learning, it will not be possible to … specify the relevant mechanisms of knowledge acquisition.” Mandler then goes on to develop this thesis in three domains: language acquisition, memory, and categorization. In categorization, she argues that it is crucial to distinguish between “perceptual” and “conceptual” categories. The former are based on the physical resemblance of one object to another and are formed automatically (without conscious awareness). Conceptual categories, in contrast, are based on what things are, what they do, and what can be done with them. They are formed through the conscious process of perceptual analysis. Mandler acknowledges Katherine Nelson’s intellectual priority to this type of argument. It bears more than passing resemblance to Nelson’s argument that, whereas category extension may be based on perceptual features, category intension is based on function. Mandler’s enterprise of distinguishing kinds of knowledge entailed in different kinds of learning casts a very functionalist eye on cognition. Unlike other chapters in the volume, in which there is extended discussion of sociocultural influences on the developments in question, Mandler argues that much of the early representational work she describes goes on before language and, thus, before culture exerts its effect. As such, the representational products that are the subjects of her chapter may be viewed as the “raw materials” on which sociocultural influences operate. Whereas Mandler focuses on the development of children’s concepts about objects, chapters by Lucariello (chap. 3) and by Astington and Peskin (chap. 4) provide functionalist accounts of how children acquire concepts about people, that is, psychological concepts of the mind. Along with many of the contributions to this volume, these chapters provide excellent illustrations of how the enterprise of making sense of the world occurs within social interactions in which adults mediate children’s cognitive processes. These chapters focus on the ways in which theory of mind (ToM), and varying ToMs, is constructed through social interaction of different kinds. In chapter 3, Lucariello presents a new model of the nature, origins, and development of theory of mind. In this model, ToM is understood in terms of its functions in people’s lives and hence is conceived as a differentiated cognitive ability. Social ToM has to do with individuals’ reasoning about the mental and emotional states of others and is used in social interaction. Intrapersonal ToM is reasoning that entails one’s own mental states and hence is involved in learning. Both of these ToMs are thought to develop from sociocultural experience, and to have a
4
HUDSON, LUCARIELLO, FIVUSH, BAUER
distinctive basis therein. Socialization experiences vary with respect to self-concept and language use. Socialization that emphasizes an interdependent self-concept and pragmatic-interpersonal language use is likely to foster social ToM development. Socialization that accentuates an independent self-concept and referential-intrapersonal language use is likely to facilitate intrapersonal ToM development. Accordingly, children whose socialization was predominantly of one form or the other would manifest differential strengths, hence uneven ToM development across social and intrapersonal ToM kinds. Lucariello finds evidence for greater strength in social, than intrapersonal, ToM in low-income children. The socialization experiences of low-income and minority children are typically those that would foster social ToM. Hence, these children were presented with metarepresentational tasks of false belief, distinguishing appearance–reality, and representational change. The false-belief task assesses other’s beliefs, and hence is a social metarepresentation reasoning task. The latter two tasks assess own beliefs, and hence are intrapersonal metarepresentation reasoning tasks. Results show ToM to be differentiated by function. Children performed significantly better on the social ToM task of false belief. Accordingly, these data support the hypothesis of a distinct sociocultural basis for social ToM. This ToM development has its basis in socialization that emphasizes pragmatic-interpersonal language use and an interdependent self-concept. The hypothesis on distinctive sociocultural bases for distinct ToM kinds was supported in additional findings. Those children who produced task-appropriate vocabulary showed significantly better performance on the appearance–reality and representational change tasks than those children who did not. Hence the data showed that, as proposed, intrapersonal ToM reasoning is facilitated by the referential-intrapersonal language mode. Astington and Peskin ask how children acquire language, specifically the mental lexicon associated with ToM in chapter 4. They design and implement a training intervention with low-income kindergarten children. The same picture books are read to both control and experimental groups. However, for the experimental groups, the original text is pasted over with text and with questions rich in explicit cognitive terms. For the control groups, the text is stripped of any metacognitive terms. Interestingly, control groups’ ability to explain action mentalistically improved more than that of the experimental groups. However, Astington and Peskin rightly note that mentalistic concepts were integral to the stories (although the texts themselves contained no metacognitive terms). The findings suggest that exposure to metacognitive terms is only one aspect of the sociocultural world that may foster an understanding of mental states. Intense exposure to storybooks with pictures of tricks, disguises, and hiding acts promotes children’s conceptual understanding. The finding that control groups’ performance on the false-belief explanation task exceeded that of experimental groups indicates that children in the control groups appear to have acquired a deeper understanding of mental states. In making sense of the sociocultural
1. INTRODUCTION: THE MEDIATED MIND
5
world of stories, these children had to be more active in trying to comprehend what the characters knew and thought. This cognitive constructive activity with respect to sociocultural context was also seen in the experimental groups’ acquisition of the mental lexicon. These children subsequently used a wider variety of metacognitive terms in their own storyelling than did children in the control groups. However, at this stage, these children appear to have been using the terms without real understanding of them because their scores on the metacognitive term comprehension test did not improve from the pretest. Although the sociocultural context provides the raw material (the terms), children have to make something of this information. They use their own internal cognitive resources in a process that Nelson (1996) called co-construction. Hence, whereas children’s mentalistic concepts are dependent on the categories and relations among categories lexicalized in their language, children themselves have an important role to play in getting hold of their meaning.
SOCIAL INTERACTION AS A MEDIATOR FOR MEMORY DEVELOPMENT Katherine’s view that the social interactions in which children engage exert a powerful mediating influence on the skills they develop and the meanings they construct has transformed the study of memory development in young children. In 1993, Nelson proposed that autobiographical memory originates in the sharing of past experiences in everyday conversations with parents. By talking about the past with adults, children learn the social significance of sharing memories and acquire the narrative forms for reporting past events. Two chapters in this volume maintain the core notion that talking about past events is formative in memory development, but examine effects of talk about the past in different ways. Fivush, in chapter 5, approaches the question of how memories of personal experiences are modulated through joint reminiscing with others. Fivush proposes that what we choose to tell and not tell emerges from our interactions and from the people with whom we reminisce. Taking both a developmental and feminist perspective, Fivush argues that autobiographical memory must be conceptualized within a framework that examines the ways in which personal experience is given voice and told from the individuals’ “owned” perspective. This contrasts with the ways in which personal experience may be silenced, either through disallowing certain stories to be told or imposing certain perspective on the remembered event. Autobiographical narratives are validated or invalidated by the individual, by the conversational partner, and by the larger community. A model of autobiographical memory is advanced that is based on two dimensions, voice and silence by self or other. The model is illustrated with examples from early mother–child reminiscing. Fivush’s chapter extends the social interaction model of autobiographic memory development in important ways. She illustrates how language is critical to the development of autobiographical memory more generally and, specifically, to the
6
HUDSON, LUCARIELLO, FIVUSH, BAUER
development of validated and invalidated autobiographical narratives. In this model, what is not said or “voiced” is as important as what is said; who decides what can be talked about is a critical element of the specific interactive context and is determined, in part, by sociocultural norms. Thus, Fivush expands the concept of the sociocultural basis for the formation of autobiographical memory by examining the larger context in which particular interactions occur. In chapter 6, Bauer and Burch investigate sociocultural context in the form of maternal scaffolding. They examine whether maternal scaffolding, already known to play an important role in the development of early autobiographical memory and narrative, exerts an influence on the development of other types of memory. Recall for sequences of actions over extended delays is examined in two studies. In the first, the ways in which mothers talk with their 2-year-olds, in teaching them specific action sequences, influence children’s subsequent memory for those sequences. More specifically, when sequences were relatively simple, mothers responded to both their children’s dispositional characteristics and their verbal productivity. This level of maternal scaffolding was related to children’s subsequent engagement and exhaustiveness of recall, but not to the organization of recall. With more difficult action sequences, mothers responded more to the demands of the situation than to their children’s characteristics. This level of scaffolding was only minimally related to subsequent memory. In a second study, Bauer and Burch describe relations between maternal reminiscing style and children’s verbal productions during a deferred imitation task. Expected correlations were obtained between mothers and children within the reminiscing context. Most interesting, however, was that maternal associative talk during reminiscing was related to children’s verbal productivity during deferred imitation. The Bauer and Burch results confirm that children’s developing memory skills are embedded in a culturally mediated context. Although interactive conversations about past events have been the focus of much of the research in this area, these findings indicate that interactions between maternal verbal behavior and children’s mnemonic performance can be found across multiple contexts. This work also examined how child and mother characteristics, as well as task difficulty, contributed to memory performance, indicating there are multiple mediators of the behavior of both mothers and their children. This research extends the conceptions of the effects of sociocultural context on cognitive development by studying a larger range of interactive contexts that contribute to the development of memory in young children.
LANGUAGE AS A SYMBOLIC TOOL AND AN INTERACTIVE PROCESS Katherine’s work has emphasized that language, as a cognitive phenomenon, must be understood as emerging from children’s developing understanding of the world
1. INTRODUCTION: THE MEDIATED MIND
7
around them and children’s emerging language abilities change the nature of their interactions with others. Language is a representational tool for both communicating with others and organizing one’s own cognition. Language also provides the child with entry into a “community of minds” that allows the child to accumulate a vast array of cultural knowledge. These themes are especially evident in several contributions to this volume that focus on various ways in which talk with others shapes children’s cognitive, language, and narrative development. In chapter 7, Hudson provides illustrations of the social origins of conceptualizations of what is virtually universally viewed as a philosophically abstract concept of time. The specific focus is future time, the often unacknowledged complement of the stuff of which memories are made, namely, times past. Hudson argues that young children use their general knowledge of how events unfolded in the past to think about and plan future occurrences of similar events. She further argues that the process of building the future based on the past occurs in the context of conversations with adults. The empirical foundation for the argument is a study of conversations about past and future events held between mothers and their 2½and 4-year-old children. By examining how parents structure conversations about the future, and how children participate in such conversations, Hudson charts the development of children’s conceptualizations of future time. She reports that in talking about events that had not yet happened, mothers referred not only to the future, but grounded their comments in talk about the past. This shows the important link between children’s general event knowledge and their construction of a mental time line on which to record the events of their lives. This analysis illustrates another way in which development of things as disparate as a personal, historical past and abstract concepts of future time, are influenced by the sociocultural context in which children develop. In chapter 8, Engel and Li discuss how children make sense of others through the use of narrative, in this case, narratives about other people or “gossip.” A line of research is described that investigates the kinds of stories children tell about other children from age 4 to 10. In general, as children get older, they tell stories about their friends that are longer, more descriptive, and more evaluative. Their stories also become increasingly more other-focused and more concentrated on the internal landscape of others. Most of the stories were acquired from having directly witnessed or participated in the event, although children also increasingly tell stories they have heard from their friends. The authors make a compelling argument for role of stories about others (gossip) on the development of self and other understanding and the role of narrative as a way of knowing: “Early on we may use stories to know ourselves. Over time, we use them to know others. Moreover, the knowledge of other people that we acquire through their narratives may, in some ways, be more important and powerful, than the knowledge we get of those people through first hand, or direct, experience.” Engel and Li also find evidence for experimenter effects in the elicitation of children’s stories about others. What some might attribute to “noise” in the data,
8
HUDSON, LUCARIELLO, FIVUSH, BAUER
they explore more deeply, and in doing so, acknowledge the fundamentally interactive aspect of telling stories. What gets said and how it is said is, in part, a function of whom children are talking to and how they respond to the story being told. These themes were addressed earlier by Fivush, and remind again of the ways in which specific interactive contexts shape the construction of narrative. Learning language and other cultural symbol systems is the topic of chapter 9 by Nelson, Craven, Xuan and Arkenberg. These symbol systems include text, first language structures, art, and sign language. Learning is considered within a variant of dynamic system theorizing. The authors propose that abstractions in working memory and encoding of these into long-term memory—processes at the heart of learning—are greatly dependent on a converging set of conditions called the “Dynamic Tricky Mix.” Although patterns of text/graphics/sign/ speech make symbol-system challenges processable, equally important is the convergence of favorable variables. These variables include motivations, social structures (e.g. script events), emotional regulation, expectancy, confidence, and self-esteem. When many converging, favorable conditions “dynamically collaborate,” learning may be exceptionally quick and powerful. Dynamic Tricky Mix theory is then applied to several domains of symbolic learning, including first language learning and art. The authors also review recent research and thinking on children with specific language impairment (SLI), from the perspective of Dynamic Tricky Mix theory. They propose that the essential foundations for communicative progress may be in place in terms of attention, memory, and abstraction mechanisms, as well as some core social and linguistic concepts, but the child’s application of these foundational readiness capacities to further communication progress may be delayed until new, productive “mixes” of interactional conditions are introduced to the child. Similar accounts of ways that acquisition rates have been vastly accelerated by taking advantage of existing readiness in language-typical, mathematically delayed autistic and other children are also presented. Finally, the authors draw parallels across evolutionary time. Discussion of the early development of language and symbol systems are used to illustrate the importance of converging conditions and the ways in which developmental progress in symbol use may at points be stalled and then be perturbed by relatively small shifts in a few conditions toward new dynamic patterns and extremely high rates of physical and cultural change. Literacy, even more than spoken language, allows for the possibility of knowledge to be imparted across space and time. In chapter 10, Homer reviews historical perspectives on literacy, highlighting the debate over whether literacy transforms thought or is simply a cultural tool that makes use of preexisting cognitive skills. Relying on Olson’s (1994) arguments that reading provides a model for how to interpret the world and writing provides a model for understanding the structure of language, Homer argues that literacy fundamentally changes cognition. He reviews several recent studies that converge on the conclusion that, through the de-
1. INTRODUCTION: THE MEDIATED MIND
9
velopment of literacy, children gain a different form of representation of language and thought, an awareness of perspective, and a metalinguistic awareness of relations between speech and text. Homer’s views on the ways in which literacy influences mind are clearly consonant with Katherine Nelson’s framework for understanding the ways in which language is transformative in development. Homer also points out that literacy transforms cognition at different levels. Some of the effects of literacy are at the cultural level, for example, the development of new, abstract concepts that can be passed on to children through language. Other transformations occur at the individual level and can be observed as changes in children’s cognitive functioning in memory, scientific reasoning, and metalinguistic awareness that are tied to the acquisition of literacy.
THEORETICAL ROOTS OF THE FUNCTIONALIST APPROACH As discussed earlier, Katherine Nelson’s mediated mind approach to cognition is fundamentally a functionalist perspective. Cognitive development is driven by the search for meaning; cognitive processes and mental structures develop and evolve in the context of real-world problem-solving and meaning-making enterprises. The theoretical foundations of Katherine’s functionalist approach to cognitive development play a central role in chapters 11 and 12 by Bruner and Donald. Bruner’s contribution (chap. 11) aptly discusses Katherine as a scientist of the “contextual functionalist” persuasion. He writes, “To be at peace with herself and with psychology, she has forever been in search of the contexts that give particular human acts their meaning. … She’s been tireless, like Minerva on the hunt, but she’s always known one big thing: to catch your quarry you need to know the setting in which they live.” Bruner takes Katherine’s work with the New York Child Language Group on analysis of the Emily monologues as a prototypical illustration of her contextual functionalist orientation. This work appeared in the famous volume she edited, Narratives from the Crib (in which Bruner has a contribution co-authored with Joan Lucariello). In the “Emmy” work, Katherine’s taste for function and context was atmospheric and set the stage for the fundamental exploration of the functions of monologue and the different contexts in which child monologue has to be understood. Katherine’s views on the development of the mediated mind (Nelson, 1996) were strongly influenced by Donald’s views on the evolution of human cognition. Donald (1991) characterized the human mind as a “hybrid” of biology and culture. He suggested that technological and cultural advances in conjunction with biological evolution created four distinct qualitative shifts in cognition. Associated with each transition was a corresponding development of a new mode of mental representation. Katherine’s volume examined parallel shifts in the development of lan-
10
HUDSON, LUCARIELLO, FIVUSH, BAUER
guage, social functioning, and cognition in young children. She argued that during the preschool years, there are qualitative shifts in the development of a mediated mind, and identified several interrelated functions of language that are responsible for the emergence of the mediated mind (see Homer, chap. 10 in this vol. for a more thorough discussion). In chapter 12, Donald advances a broad functionalist perspective to understanding the mind. Donald provides a provocative discussion of the need to construct interdisciplinary approaches for understanding the human mind. Whereas evolution and development are clearly different constructs, concerned with different time lines and different mechanisms, development is integral to evolution. Any understanding of the human mind must make sense in both evolutionary and developmental contexts. Moreover, in order to provide adequate theoretical explanations of both evolution and development, Donald asserts that researchers must examine how humans live in the real world. In this way, Katherine Nelson provided a clear and coherent statement of the integration of development and evolution. Although Donald argues for the necessity of domain-specific experimentally controlled research on cognition, he notes that this type of research must be understood in conjunction with a larger theoretical vision. In his words, the “interdependencies and complexities inherent in the mental structures of human adults could only be understood better in the light of a comprehensive evolutionary scenario of human emergence.” Accordingly, it is necessary to consider evolutionary history and particularly, the evolution of culture in order to understand human cognition. In both human evolution and development, the individual becomes a part of a distributed cognitive system. What human evolution has allowed for is the development of memory, attention, and metacognitive skills that allow the individual to take advantage of what evolving culture provides.
CONCLUSIONS The significant contribution of this volume is that it addresses all these aspects of the mediated mind. Indeed, a single chapter often strikes more than one note in its treatment of the mediated mind. More noteworthy still are the importance, range, and prescience of the cognitive processes explored in this volume through the “mediated mind” lens. Memory—both autobiographical and event-semantic—theory of mind, mental representation, temporality, narrative, and metalinguistic awareness comprise the chapter topics in this volume. The breadth of topics represented here is a tribute to the impact Katherine Nelson’s vision on many developmental “domains.” The contributors acknowledge and honor the work of Katherine Nelson. Her theory and research paved the way for the advances in understanding a mediated mind that are evident here and that will continue to shape notions of how the human mind develops and evolves within a social, interactive world.
1. INTRODUCTION: THE MEDIATED MIND
11
REFERENCES Donald, M. (1991). Origins of the modern mind: Three stages in the evolution of culture and cognition. Cambridge, MA: Harvard University Press. Nelson, K. (1973). Structure and strategy in learning to talk. Monographs of the Society for Research in Child Development, 38(1–2), 136. Nelson, K. (1974). Concept, word, and sentence: Interrelations in acquisition and development. Psychological Review, 81, 267–285. Nelson, K. (Ed.). (1986). Event knowledge: Structure and function in development. Hillsdale, NJ: Lawrence Erlbaum Associates. Nelson, K. (Ed.). (1989). Narratives from the crib. Cambridge, MA: Harvard University Press. Nelson, K. (1993). The psychological and social origins of autobiographic memory. Psychological Science, 4, 1–8. Nelson, K. (1996). Language in cognitive development: Emergence of the mediated mind. New York: Cambridge University Press. Olson, D. R. (1994). The world on paper. New York: Cambridge University Press.
2 Two Kinds of Knowledge Acquisition Jean M. Mandler University of California, San Diego
Perceptual categories differ from conceptual ones. Perceptual categories are not accessible to conscious manipulation (K. Nelson, 1985)
I came to developmental psychology rather late in life. In graduate school I never read Piaget or had a course in development, and although I studied with Jerry Bruner, that was before he became a developmentalist. Mostly I ran rats and continued to do so for a number of years after getting my degree. So when I became disaffected with animal research and wanted to know about learning in a more human setting, I had a lot of catching up to do. That was about 30 years ago, but I can still remember my first impressions of the developmental literature. Much of it reminded me of the animal discrimination literature that I was trying to escape. Hull and Spence seemed to have an awful lot of influence on the field of children’s learning and development. But, as I delved into the language learning literature, I found really interesting work that came from another tradition entirely. Katherine Nelson’s work was one of these. At the time (the early 1970s), I was associate editor for the Psychological Review. We did not get many developmental Copyright ã 2002 Jean Mandler. This chapter is based on chapter 12 in the book, The Foundations of Mind (J. M. Mandler, 2004).
13
14
MANDLER
articles, but one that did pass my desk was by Katherine, entitled “Concept, Word, and Sentence” (K. Nelson, 1974). In it she promulgated her functional core theory of early concept learning. In her view, the earliest concepts are organized around what objects do in the situations in which children encounter them, with the perceptual information that allows identification of objects being less critical and outside the core meaning. That struck me as insightful and potentially far-reaching in its implications for how the human mind becomes organized. I became, at least temporarily, a Nelsonian. (Then I discovered Piaget, and like many others, was overwhelmed by the elegance and comprehensiveness of his theory, and so became a Piagetian instead. It took me a good many years before the contradictions in Piaget’s theory became so obvious that I could no longer remain among the faithful.) By the time I began doing infancy work, I had more or less forgotten about Katherine’s earlier work on concept formation. But her thinking must have continued to influence me because my own research rapidly led me to conclusions similar to those she had promulgated a decade or more before. The first of these is exemplified by the quotation at the head of this chapter: There is more than one kind of categorization, only one of which is accessible. The second is that the core of the first accessible categories (concepts) is based on what objects do (or is done to them), not what they look like. It is these different kinds of knowledge acquisition I discuss in this tribute to Katherine. In particular, I want to emphasize the importance for further progress in the field of determining exactly which kinds of knowledge acquired in infancy are implicit or procedural in nature and which explicit or declarative. As Piaget pointed out many years ago, one of the most important developmental factors is whether knowledge is accessible conceptual knowledge or inaccessible sensorimotor knowledge. Even if we reject his stage theory of knowledge acquisition it is important not to lose that insight (see Karmiloff-Smith, 1992). Until it is routinely specified exactly what kind of knowledge is being dealt with in any particular kind of learning, it will not be possible to answer questions about domain-specific versus domain-general learning or to specify the relevant mechanisms of knowledge acquisition. This chapter illustrates the importance of the distinction between implicit or procedural knowledge and explicit or declarative knowledge in three different areas of research in infancy, showing in each case how a failure to make the distinction has led either to confusion or unnecessary disputes within the field.1 The three areas are memory, categorization, and language acquisition.
MEMORY I begin with memory for events, because that is the area in which the distinction between implicit and explicit memory has been most studied and for which there is the most evidence, both psychological and neurological. The following are the
2. TWO KINDS OF KNOWLEDGE ACQUISITION
15
fundamental distinctions. There is procedural memory, which comes in several varieties but none of which requires conscious awareness of what was learned or of the occasion of the learning. Classic examples are Piaget’s sensorimotor learning, visual expectations about what will come next in a sequence of events, and kicking to the sight of a mobile that has been previously reinforced for doing so. These kinds of learning and memory are relatively low-level processes, and similar processes are ubiquitous in the animal kingdom. Declarative memory, on the other hand, is a higher-order cognitive process that allows bringing to conscious awareness a previously experienced event or fact learned about the world. This kind of memory is illustrated by recall and recognition. By recognition I refer specifically to the awareness of prior occurrence that enables someone to state that a given stimulus has been experienced before. Declarative memory is packaged rather differently and uses different brain systems from procedural memory. As far as the neural substrate of these different kinds of memory is concerned, a great deal of research remains to be done on infants, but currently it appears that development of visual expectations depends on striatal structures, and acquisition of conditioned responding depends on the cerebellum and some of the deep nuclei in the brainstem (C. A. Nelson, 1997). These kinds of learning begin very early in ontogeny. In contrast, the ability to recall or explicitly recognize something as having been experienced before is thought to depend on the hippocampus and other medial temporal lobe structures, as well as higher cortical association areas, including the prefrontal cortex (Mishkin & Appenzeller, 1987; Squire & Zola-Morgan, 1991). Although the medial temporal lobe develops early (C. A. Nelson, 1997), the neocortex and its reciprocal connections with the hippocampus develop more slowly (Bachevalier & Mishkin, 1994). The best available evidence suggests that in infants these various components of the declarative memory system begin to coalesce in the second half of the first year (Carver & Bauer, 2001; Carver, Bauer, & C. A. Nelson, 2000). Patricia Bauer, Leslie Carver, Laraine McDonough, and I have all studied the development of recall—and in particular, recall of temporally ordered events—using a method adapted from Piaget’s observations of deferred imitation. We act out novel event sequences for babies, using little model props. After modeling, we bring the babies back at a later time, give them the props, and see if they can reproduce the sequences without any further reminders. Some 9-month-olds can do this for event sequences after delays of a month (Carver & Bauer, 2001). By 10 to 11 months, infants can reproduce temporally ordered events after delays of 3 months (Carver & Bauer, 2001; J. M. Mandler & McDonough, 1995). At 11 months, infants are also capable of reproducing events consisting of a single act a full year later (McDonough & J. M. Mandler, 1994), but single acts do not require recalling temporal order and so create an easier task than sequences. Indeed, it is possible that the well-known limitations on young babies’ ability to imitate event sequences may not be a difficulty with imitation per se but due to immaturity of the retrieval mechanisms required to string together more than one act in a novel sequence.
16
MANDLER
We know that deferred imitation assesses declarative memory because amnesics cannot do it. Some years ago, when I presented our deferred imitation data at a conference, a developmental researcher suggested that the distinction between declarative and procedural memory was a false one. Another researcher insisted that to ascribe recall to infants was unwarranted by the data, that it could be accomplished by low-level sensorimotor learning. In the lengthy argument over these issues that ensued, a neuropsychologist present said that he would not really believe that deferred imitation required declarative memory unless it could be shown that amnesics could not do it and he thought it likely that they could.2 I was surprised by this view, which implied that amnesics who could not say what happened yesterday, could nonetheless act it out. That would place the locus of amnesia in the verbal system, which did not seem appropriate. In any case, as it turns out, he was wrong. We tested amnesics on the same deferred imitation tests we had used with infants and they failed (McDonough, J. M. Mandler, McKee, & Squire, 1995). Bauer (in press) also pointed out that if performance on deferred imitation tasks were due to implicit motor learning, then performance should be disrupted by changes in context between learning and test, but no such context dependency is found in deferred imitation (e.g., Barnat, Klein, & Meltzoff, 1996; Bauer & Dow, 1994). It seems clear that at least by 9 months, infants have developed a declarative memory system that contains some long-lasting material. What about younger infants? Have they also developed a declarative memory system? Using a deferred imitation paradigm, Barr, Dowden, and Hayne (1996) were able to get 6-month-olds to reproduce a single act after 24 hours, but this is the earliest age for which there is evidence. Rovee-Collier (1997) made a brave attempt to distinguish implicit and explicit memory in earlier infancy by comparing variables that are thought to affect these two kinds of memory differentially in adults and assessing whether they also differentially affect the two kinds of memory in infants. She used two operant conditioning tasks and measured retention of the learned response with or without “reinstatement.” She claimed these tasks measure explicit and implicit memory respectively. Because even 3-month-olds show the effects of reinstatement, that would mean that recall memory is present at a very young age. Indeed, this was suggested to be the case by Rovee-Collier (e.g., the discussion section of J. M. Mandler, 1990; see J. M. Mandler & McDonough, 1997, for commentary). Unfortunately, it is exceedingly difficult to equate these two conditioning tasks with the classic tasks measuring implicit versus explicit memory in the adult memory literature (e.g., stem completion vs. free recall). In the long run, the only necessary and sufficient condition for explicit memory is conscious awareness. Rovee-Collier (1997) insisted that this requirement is untenable from a developmental perspective, because preverbal infants cannot tell us whether they are aware of the past and therefore their consciousness is “solely a matter of philosophical speculation” (p. 469). As an infancy researcher I sympathize with her
2. TWO KINDS OF KNOWLEDGE ACQUISITION
17
frustration in this regard, but nevertheless consciousness is the ultimate criterion of explicit memory. This is why McDonough and I resorted to the study of amnesics, not because infants are like amnesics (far from it) but because amnesic performance can be used to show that a nonverbal task such as deferred imitation requires conscious awareness of the past. At present, there is no comparable test to show that conscious awareness of the past accompanies dishabituation to a novel stimulus or retention of conditioned foot-kicking. In particular, I can see no justification in calling the retention of a conditioned foot-kicking response—with or without a reinstatement cue—an explicit memory task. It may be that the only way to test Rovee-Collier’s contention that young infants in conditioning tasks are demonstrating declarative memory is to see if amnesic adults show retention of conditioned foot-kicking, or fail as they do on deferred imitation. Similar problems in interpretation have arisen over the years about recognition memory when the infant literature is compared to adult studies (e.g., Fagan, 1973). Because infants respond differently to old and new stimuli, it is often assumed that there is a straightforward continuum in the development of recognition memory from infancy to adulthood. However, recognition is typically measured in infants by a classic measure of procedural memory—namely, habituation to a repeated stimulus followed by dishabituation to a new stimulus— and recognition is measured in adults by a classic measure of declarative memory—namely, by asking them to make a judgment of old or new. This should certainly make us suspicious. Nothing can be concluded about a continuum when such different measures are used at different points. A continuum that does extend from the baby habituation data to adult data is repetition priming, which is a standard measure of implicit memory (Squire & Knowlton, 2000). Virtually from birth, babies habituate to stimuli they have repeatedly seen, which is to say they can be primed. So can adults; for example, words that have been recently studied tend to be read more rapidly and show an increased likelihood to be used spontaneously in stem completion tasks. Normal adults can say that they have seen the stimuli before, whereas amnesic adults, who show normal priming on stem completion, cannot (Graf, Squire, & G. Mandler, 1984). This is not just a phenomenon of brain damage. Even normal adults look like amnesics on stem completion if their conceptual processing is disrupted; if they are instructed to look for vowels rather than to read the words for meaning, they too show priming while being unable to recall the words (Graf, G. Mandler, & Hayden, 1982). Adult priming tasks are like the habituation-dishabituation, or familiarization-preferential looking, tasks used with infants in that they show the effects of prior exposure, independently of conscious awareness. Similarly, the galvanic skin response to familiar stimuli has been used to show a continuum in implicit recognition memory from children to adults in the absence of explicit recognition (Newcombe & Fox, 1994). What we do not have at this point is evidence of explicit recognition memory in young infants. Conceptual interpretation is crucial to declarative memory. But it is not possible to question preverbal infants and so
18
MANDLER
there is ambiguity as to what recognition in habituation-dishabituation studies means. In any given instance is it what has been called “perceptual identification” (Jacoby, 1983) or “primitive recognition” (J. M. Mandler, 1984), which is implicit and does not require conceptual interpretation, or is it an example of explicit declarative memory? As discussed later, it is plausible that infants begin to build a declarative knowledge base quite early in infancy. However, at present there are no experimental techniques to provide positive evidence for declarative memory earlier than 6 months. McKee and Squire (1993) suggested that when a delay is included, the habituation-preferential looking tasks, such as used in infancy, do measure explicit memory. Even young infants show the effects of exposure to visual stimuli, typically with a novelty preference (Fagan, 1973), although sometimes after long delays recognition is expressed by a preference for the familiar stimulus (Bahrick, Hernandez-Reif, & Pickens, 1997). In the only study I know of to test whether declarative memory might be demonstrated by preferential looking after a delay, McKee and Squire (1993) found that elderly amnesic patients with hippocampal damage who were familiarized with a set of pictures looked longer at new pictures after a delay of 2 minutes, but this novelty effect disappeared by 1 hour. Control subjects looked longer at new stimuli at the 1-hour delay, but not after 4 hours. The relatively poor performance of the amnesic subjects suggested to the authors that declarative memory might be involved in this kind of forced-choice preferential looking test. However, the study also showed that the looking task was more difficult than an explicit recognition task in which subjects had to point to the new pictures. This is a surprising result, because usually tests of implicit recognition are easier than explicit tests. Furthermore, the amnesic subjects, although poorer than controls, showed good explicit recognition of novel stimuli at an hour delay, even though they did not show longer looking at them. These unusual patterns make the data extremely difficult to interpret. To further complicate adult–infant comparisons, in young infants novelty preferences may be reflexive or obligatory in nature, which is not the case for older infants or adults. Using ERP measures, C. A. Nelson and Collins (1991) found that 6-month-olds responded as much to the frequency of occurrence of a stimulus as to its familiarity, suggesting that nonspecific novelty detection might play a role in preferential looking measures. Not until 8 months did ERP data differentiate familiar and novel stimuli regardless of frequency of exposure (C. A. Nelson & Collins, 1992). Because the hippocampus may be involved in novelty preferences as well as explicit recognition in infancy, C.A. Nelson (1997) suggested that the form of “explicit” memory that is dependent on the hippocampus early in life differs qualitatively from that observed later in the first year; he called it “pre-explicit” memory. In sum, preferential looking after a delay is a troublesome measure to interpret if the interest is in its implicit or explicit status. Although it can be used with both infants and adults, it is open to question as to whether the same underlying process is involved. Furthermore, the evidence that it taps declarative memory is
2. TWO KINDS OF KNOWLEDGE ACQUISITION
19
dubious. It is unlikely that newborns’ or very young infants’ preferential looking at a novel stimulus after habituation implies explicit awareness of prior occurrence. We are on firmer ground when we use recall as our measure. It is not possible to chart the development of various memory functions if we do not agree that there is more than one kind of memory, or if we fail to (or cannot) use the same kinds of tests at different points in development. Some perceptual tests require only procedural memory, but for tests of declarative memory an explicit conceptualization of the event to be recognized or recalled is required. One must interpret what has been observed in order to recall it later. Such interpretation must begin early in infancy, although it is not known exactly when or how deeply. All too often it is assumed that visual associations themselves automatically confer meaning, whereas for a declarative meaning system to develop, the meaning of what is seen must be constructed (Piaget, 1952). There must be some mechanism to account for infants’ beginning interpretations of the world. I have suggested that some process, such as perceptual analysis, takes place to get this interpretive process started, beginning at least by a few months of age and perhaps earlier (J. M. Mandler, 1988, 1992, 2004). The function of perceptual analysis is to conceptualize what the infant observes. It does this by analyzing and reducing perceptual displays into simple meanings, such as “self-starting,” that form the basis of concepts. This analysis redescribes perceptual information into a simpler and potentially retrievable conceptual form. For example, an infant may not actively process most details of an event in which someone gets up out of a chair and picks the infant up. Perhaps the infant only notices that something starts to move itself and comes to the infant contingently upon the infant’s cries. The meaning of the event for the infant may be something rather like, “A self-starting contingent interactor appeared when I cried.” If that was the interpretation then that is what would be put in the infants’ growing conceptual store and therefore potentially available for recall at a later time (as opposed, e.g., to what the person looked like). It might take several months before more specific conceptualizations begin to be formed, some sufficiently detailed to engender recall memories more comparable to those of adults. It also seems possible that to the extent that the earliest conceptualizations are global and lacking in detail they would tend to be irretrievable later, as in infantile amnesia. In any case, until we know how young infants go about interpreting what they perceive, we will not be able to determine what they will be able to recall at a later time. Ascribing conceptual interpretation to infants is seen by some infancy researchers as attributing unduly precocious capacities to them (Haith & Benson, 1998). But there is no reason to consider conceptual interpretation (or recall based on it) as precocious. It is only precocious within the context of a theory such as Piaget’s (1952) that denies this capacity to infants until a late age. In most current discussions of infant development, it is more widely understood that from early in life infants interpret the world while they look at it. The fact of conceptual interpretation is equally important in understanding categorization, discussed in the next section.
20
MANDLER
CATEGORIZATION Categorization is another area of research that has been hampered by not making a distinction between procedural and declarative representation. In the adult categorization literature there have been many arguments over whether categorization is rule based or based on similarity (see, e.g., vol. 65 of Cognition), and these arguments tend to spill over even into infancy studies. Various theories have been proposed and tested in different laboratories using entirely different stimuli—from dot patterns and geometric forms, through disease symptoms and professions, to “things to take on a fishing trip.” This hodgepodge of stimuli is like comparing apples and oranges (although, of course, apples and oranges are both fruits; the ambiguity about the nature of categories is there in folk beliefs as well as in scientific ones). It is implausible that identical processes are involved as one moves from categorizing purely perceptual patterns to complex conceptual knowledge. As both Katherine Nelson and I have discussed (e.g., J. M. Mandler, 1998, 2000a; K. Nelson, 1985), there are (at least) two kinds of categories formed in infancy. Perceptual categories are one of these. Given a series of stimuli that bear some physical resemblance to each other, such as cats or tables, the perceptual system abstracts the principal components of the presented information, thus forming a prototype or schema that influences recognition of similar stimuli at a later time. This kind of categorization is part of the visual input system, and is modular in the sense that it is automatic and mandatory, and the information it is using is not accessible to consciousness (Moscovitch, Goshen-Gottstein, & Vriezen, 1994). As discussed in the last section, the process is what is known as perceptual identification, or primitive recognition. This kind of category should be theoretically differentiated from conceptual categories. Concepts (which are a type of category) formulate what kinds of things cats or tables are (J. M. Mandler, 2000a). Adults do their conscious thinking, planning, and problem solving with the large repertoire of concepts they have built up over the years. However, infants face the task of getting this repertoire started. As they begin to encounter animals, vehicles, furniture, and so forth, they must form some idea of the meaning of these things. As suggested in the last section, through perceptual analysis they can analyze what objects are doing. So, in addition to forming categories of things that look alike, they begin to interpret what those things are doing or is being done to them. That is, they begin to conceptualize them — for example, as animate and inanimate things, or things that do and things that get done to. These two kinds of categories, perceptual and conceptual, represent procedural and declarative knowledge respectively, and differ in a number of ways. I have discussed these differences at length in J. M. Mandler (2000a), and will recapitulate them here only briefly. First, the two kinds of categories are formed from different kinds of information. Perceptual categorization computes perceptual similarity. At least early in infancy it does so independently of knowledge about function or kind; indeed, it can occur even in the complete absence of meaningfulness. In con-
2. TWO KINDS OF KNOWLEDGE ACQUISITION
21
trast, conceptual categories of objects compute conceptual similarity, which has to do with class membership or kinds. Elephants and robins are similar because of their animate characteristics, not because of their perceptual appearance. The concept of a kind may include a perceptual description, but its initial core is the assignment of the object to a domain, which itself is characterized by a few abstract characteristics. For infants, the important domain characteristics typically evolve around event or role information. For example, animals may be characterized for infants mainly as things that start themselves and react to other objects contingently from a distance. This description is derived, of course, from perceptual information, but is more abstract in that it characterizes roles rather than movement per se. Second, perceptual categories contain more detailed information (at least in infancy) than do conceptual categories. As Quinn and his colleagues showed, 3-month-olds can form a perceptual category of dogs that is differentiated from a perceptual category of cats in a very few trials (Quinn, Eimas, & Rosenkrantz, 1993). A perceptual procedure that can tell dogs from cats in a few brief trials must be operating on a great deal of detailed perceptual information (even if confined to the head region; Quinn & Eimas, 1996) in order to extract a summary representation. Many early concepts, on the other hand, are global, relatively crude, and lacking in detail, as exemplified by infants’ ability to conceptually differentiate animals from vehicles but not most animals from each other (J. M. Mandler & McDonough, 1998). Third, there is a different course of acquisition for the two kinds of categories. As young as 3 months, infants are proficient perceptual categorizers at what is usually called the basic level (e.g., Quinn et al., 1993). For conceptual categories, on the other hand, even 1-year-olds are more proficient at making a few broad distinctions that separate one domain from another than they are at making the finer distinctions required to categorize at the basic level (J. M. Mandler & McDonough, 1998, 2000). Furthermore, because forming perceptual categories is automatic, mandatory, and operates beyond the bounds of consciousness, the processor does not face the bottleneck that consciousness or attentive awareness imposes, namely, a restriction to a serial process that can handle only relatively small amounts of information at a time (G. Mandler, 1985, 2002). This lack of restriction means that a great deal of information can be processed in parallel. In contrast, concept formation is a more focused and limited process. It is serial in nature and does face the bottleneck of consciousness, which means that only limited amounts of information can be processed at a time. Fourth, and following directly from the above, there is a difference in selectivity. Both for learning these different types of categories and for using them in daily intercourse with the environment, conceptual categories are selective in the sense that they rely on the limited amount of information that has been attentively processed. I have suggested, for example, that infants selectively attend to kinds of motion when conceptualizing objects as animate or inanimate, and will use this ev-
22
MANDLER
idence as a kind of “rule” to categorize a given object as either animate or inanimate. In this sense, the conceptual categories of infants are like the rule-based categories of adults. “If it starts itself, it must be animate… .” I once referred to this (to the horror of a reviewer) as an example of infant concepts having necessary and sufficient conditions in contrast to the typically more nuanced concepts of adults. The idea behind this comment was that if infants are aware of almost nothing about the import of perceptual differences among various animals, but only that all of them start themselves and move in a contingent fashion, then the concept of animal is so spare as to consist of little more than a single (and therefore necessary) criterion. In contrast, perceptual categories are for the most part unselective. For example, Mareschal, French, and Quinn’s (2000) analysis of young infants’ discrimination of dogs from cats indicates the large number of features that infants pick up in a few exposures. Fifth, perceptual and conceptual categories serve different functions. Perceptual categorization is used for recognition and object identification. Conceptual categories, on the other hand, are used for recall, as we have seen, but also for making inductive generalizations and in other kinds of thought. In our work studying inductive generalization, McDonough and I use a technique similar to the imitation tasks described in the last section. In this case, there is no delay, but instead of providing the props used by the modeler for the infants to use in their imitations, we give them a choice between various new objects that are more or less physically similar to, or more or less conceptually similar to, the originals. By this generalized imitation technique, we can test how far infants are willing to generalize what they have observed as well as the bases for their generalizations. For example, we model a simple event, such as a dog being given a drink from a cup, and then give the infants a choice between a cat and a car, or between a bird and a car, or even between another dog and a cat, to use to imitate what they have observed (J. M. Mandler & McDonough, 1996, 1998). All the data indicate that infants imitate what they have conceptualized, rather than what they have seen. For example, when we model for 14-month-olds a dog drinking and provide a different dog and a cat or a different dog and a rabbit from which to choose for their imitation, the infants choose randomly between them (J. M. Mandler & McDonough, 1998). They saw a dog being given a drink, but interpreted it as an animal (or more probably a land animal) being given a drink. This does not mean they did not see a dog or cannot tell a dog from a cat—we know from Quinn and his colleagues’ work they can make this identification much earlier. But the situation is like my sitting down on a chair and you imitating me—you probably would not pay much attention to the differences between your chair and mine—they are just chairs. In the case of the infants, the dog and cat are just animals. They are using their conceptual knowledge about the world, not their perceptual categories, to control their inferences. Finally, and crucial for seeing how the differences between perceptual and conceptual categories relate to the procedural-declarative distinction, the infor-
2. TWO KINDS OF KNOWLEDGE ACQUISITION
23
mation in perceptual categories is inaccessible, whereas the contents of concepts are accessible for purposes of thought, problem solving, recall, and so on. Of course, a concept may contain perceptual information, but only if it has been analyzed (see the discussion of learning color terms in the next section). There is virtually no direct information on accessibility in infancy, but it seems reasonable to generalize from adults. There is ample evidence that for adults perceptual categories are impenetrable; for example, people have no access to the information that enables them to categorize a face as male or female (Abdi, Valentin, Edelman, & O’Toole, 1995). As for the converse, that concepts are accessible to infants as well as to adults, the imitation of past events that infants can do from about 6 months of age, discussed in the last section, requires not only conceptualization but also accessibility, as shown by the inability of amnesic adults to do such imitation. There is by now a great deal of evidence indicating that adults categorize on several different bases, in some cases responding to perceptual similarity and in other cases to explicit rules. There is also evidence that different neural circuits are involved in categorization tasks based on explicit rules versus those that are based on memory for exemplars (Smith, Patalano, & Jonides, 1998; Ullman et al., 1997). For example, Smith et al. (1998) did a PET study in which subjects were either in a categorization condition in which they needed to rely on memory for exemplars or in a rule-based condition in which they needed to follow a rule. While they were engaged in these categorization tasks, brain scanning was carried out. The resulting images showed distinct areas of activation involved in the two kinds of categorization, in addition to a number of common areas. In particular, the distinctive areas in the rule condition were in a region of parietal cortex associated with selective attention, and in prefrontal cortex, which is an area associated with applying rules. The common areas and those exclusive to the exemplar memory condition were mostly in the visual cortex. Smith et al. (1998) noted that it might be difficult to assess whether the memory-based results they found were due to exemplar similarity or prototype similarity. This assessment may be important. Conscious memory for specific instances requires declarative memory and is damaged in amnesia, whereas memory for prototypes is an implicit perceptual process that is spared in amnesia (Kolodny, 1994). There is still relatively little hard evidence about the relations between implicit and explicit perceptual memory in these tasks, nor can an easy equation be made between conscious memory for exemplars and exemplar models of categorization (e.g., Nosofsky, 1992). Nevertheless, what has been done to date indicates that different processes and brain regions are involved in different kinds of categorization. Just as many of the arguments in the adult literature fade away as it becomes clear that there is more than one basis for categorization, it is to be hoped the same thing happens in the developmental literature. The failure to recognize that there is more than one basis for categorization has led to possibly unnecessary debate. For
24
MANDLER
example, according to traditional theory, the first categories are perceptual and at the basic level and only later are superordinate categories learned (Rosch et al., 1976). When we showed that infants form several kinds of global (superordinate) categories before achieving basic-level categorization, in order to maintain the traditional view it became necessary to find a perceptual basis to account for our findings. When this was not forthcoming, it was suggested that rather than there being a “basic-to-global” trend in perceptual development, there is a “global-to-basic” sequence instead—but not just for concepts as I had proposed, but for any infant category (Quinn & Johnson, 2000). As someone who had tried to show that there is more than one kind of categorization each with a different developmental course, I described this turn of events as my “winning the battle but losing the war” (J. M. Mandler, 2000b). It was being proclaimed that, yes, global categories in infancy do precede basic-level ones, but this is merely a characteristic of perceptual development and so there is still no need to talk about anything other than perceptual categories in infancy. This is a difficult view to maintain in light of the research discussed in this chapter. Furthermore, if there is more than one kind of categorization and they have different developmental courses, such an “either–or” question as, “Does categorization proceed from a global level to a basic level or from a basic level to a global level?” cannot be answered in a sensible way. On the other hand, once the view that there is more than one kind of categorization is accepted, this kind of controversy disappears. Unfortunately, as a result of the view that there is only one basis for categorization, there has not been a great deal of effort expended on determining the kinds of processes that infants use in various tests of categorization. For example, what determines whether an infant’s attention is being directed by the perceptual appearance of the stimuli being presented, or by the kinds of things the stimuli represent? One problem with the habituation-dishabituation, or familiarization-preferential looking, technique is that because of the often close relation between appearance and kind, in many cases the infants’ responses could be determined by either. Although our work showed that infants treat birds and planes differently in this kind of test despite great perceptual similarity among the exemplars, and therefore were probably responding to conceptual rather than perceptual differences (J. M. Mandler & McDonough, 1993), some ambiguity in the basis of the responding typically remains when members of one domain (e.g., animals) are contrasted with members of another (e.g., vehicles). This is one of the reasons why McDonough and I shifted to use of the generalized imitation test. Here we could not only pit perceptual similarity against conceptual similarity, but show that just as for older children and adults, it is conceptual kind that determines the inductive generalizations that are made, and not physical appearance (J. M. Mandler & McDonough, 1996). The data also show that this is a test of explicit or declarative knowledge, not implicit knowledge. For example, if an experimenter models both a correct and an incorrect activity, such as giving both a dog and a car a drink, infants typically reject the incorrect modeling, choosing only the correct one (J. M. Mandler &
2. TWO KINDS OF KNOWLEDGE ACQUISITION
25
McDonough, 1996). On the other hand, if two correct actions are modeled, such as washing both a dog and a car, infants will imitate both (J. M. Mandler & McDonough, 1998). It is clear that the infants are treating the little scenarios that are acted out as representations of real-world events, and imitate according to what they think is correct, rather than passively mimicking the experimenter’s actions. The data indicate they can demonstrate explicit knowledge in this fashion as early as 9 months of age (McDonough & J. M. Mandler, 1998). The data discussed in this section are basically no different from those discussed in the previous section. Whether it is a question of remembering or categorizing, very different processes are made use of in implicit and explicit tasks.
LANGUAGE ACQUISITION Almost more than any other area, this is a topic in which the distinction between implicit and explicit knowledge is as vital as it is rarely discussed (a notable exception being Karmiloff-Smith, 1992). Consider the work of Marcus, Vijayan, Bandi Rao, and Vishton (1999), showing that 7-month-old infants easily learn a sequential series of syllables, such as la di la, which they then generalize to the same pattern applied to new syllables, such as do si do, and discriminate from different patterns such as do si si. The authors talked about this as a kind of existence proof for the ability of young infants to learn grammatical rules. But no meaning or anything special to language is required for this sort of learning, only the automatic pattern formation and generalization that is typical of procedural learning. Declarative knowledge is not required for this sort of implicit learning, and it can be implemented easily in a pattern-learning computer simulation (Sirois, Buckingham, & Shultz, 2000). Segmentation of speech sounds and developing expectations about their regularities are also present in nonhuman primates, indicating that the mechanisms involved in Marcus et al.’s finding are not specific to the language capacity (Hauser, Newport, & Aslin, 2001). Furthermore, the same kind of rapid pattern learning of nonlinguistic spatial sequences has been shown by Smith (1984) in 5-month-olds. A good deal of learning grammatical patterns probably takes place in this automatic way. Expectations are built up about all kinds of sequences, including about the orders in which familiar words and even word classes appear. This happens without having to pay attention to order per se. The work on learning artificial grammars shows that grammatical patterns can be learned without explicit memory for the lexical items involved. For example, Knowlton, Ramus, and Squire (1992) found a dissociation in amnesic patients between explicit memory for presented instances, which was poor, and judgments of grammaticality, which were as good as for normal subjects. Even in the lexicon, irregular morphological transformations (e.g., “went” as the past tense of “go”) seem to depend on declarative memory, whereas those transformations that are productive (e.g., the past tense
26
MANDLER
-ed) appear to depend on grammatical rules supported by procedural knowledge (Ullman et al., 2001). The rhythms and sequences of speech are exactly the sort of patterns that in general “come for free;” that is, expectations of regularities are formed automatically. It appears that one needs to attend to “went” if one is to recall it, but not to “walked.” Not only has there sometimes been conflation of pattern learning with the acquisition of linguistic meaning, there has also been a tendency in the field of word learning to emphasize referential learning at the expense of meaning. For example, connectionist modeling of language learning often makes the assumption that words are mapped onto perception, missing the representation of meaning that mediates this learning. It may be possible that rote associations can be automatically formed between various sights and labels for them, but as discussed later, it is likely to be rare if it occurs at all. In any case, if the discussion of language only concerns associations between visual and verbal patterns, then the whole issue of the conceptual meaning of language has been set aside. People not only learn the familiar patterns of their language, but also what those patterns mean, and nothing in the patterning of do si si or of determiner noun verb determiner noun tells them the conceptual content that is being expressed. Similarly, a child using the term dog to refer to dogs tells us something about reference, but not necessarily about the sense of the term for the child. As we know, the word dog can be applied correctly and still have a different meaning for the child than for the adult (e.g., Clark, 1983; McDonough, 2002). If we are to understand the course of language acquisition, it is crucial to differentiate implicit perceptual knowledge about the world from explicit conceptual knowledge of what is being communicated. Here is a dramatic example, taken from some interesting work by Susan Carey 20 years ago (Carey, 1982). She introduced 2- and 3-year-olds to a new color word, “chromium,” and studied their learning of this word. The children’s teacher introduced the term to them in one of several contexts, such as asking for one of two differently colored but otherwise identical trays, “Give me the chromium tray, not the red one, the chromium one.” A week or more later, they were tested to see whether they had learned this new color term. All the children knew at least one color term at the start of this study. About half learned the new term from this exposure, although there were many incomplete meanings. Of interest for present purposes was the contrast Carey made between these data and those of Rice (1978). All of Carey’s participants knew at least one color word, whereas Rice studied children of the same age who had no color terms. She tried to teach them “red” and “green.” As many as 2,000 trials over a period of several weeks were required for them to learn. As Carey put it, her studies showed that knowing one or two color words is as good as knowing nine or more as far as achieving a fast mapping for a new color word is concerned (Carey, 1982). Carey asked: The puzzle is this: What is the hump a child must surmount to learn that first color word? … We know that the infant perceives colors and so represents them, and that he
2. TWO KINDS OF KNOWLEDGE ACQUISITION
27
can remember them. … The concept COLOR is definitionally and developmentally primitive by anybody’s account: what is the process of going from COLOR to color word and making the latter available as a lexical organizer? No theory of lexical acquisition has even sketched an answer to this question.” (p. 380)
My answer is that color is primitive only in the sense that encoding it is an automatic procedure, available from early in infancy. It is represented implicitly. However, color is not accessible to the thinking mind until perceptual analysis takes place. It is only after one or more colors has been actively noticed and analyzed (e.g., by contrasting apples and oranges, or apples and tomatoes) that an explicit concept of color is formed. Only when color becomes declarative knowledge can children think about it and home in on the appropriate domain when they hear color terms used. Perceptual analysis identifies the domain for them, and only after that can they learn the correct mappings (“blue” to blue, “red” to red, and so forth), which takes further time (Shatz & Backscheider, 1999). Thus, even in a domain that seems purely perceptual, some explicit conceptual knowledge appears to be required to learn appropriate names.
CONCLUSION This chapter has discussed three areas in infancy research in which it seems to be crucial to know whether procedural or declarative knowledge is involved: memory for events, categorization, and language learning. There surely are more. For example, one of the largely unaddressed issues in current research is the status of the knowledge that infants have about objects—not the difference between animals and vehicles, but objecthood itself. The data that Spelke, Baillargeon and their colleagues collected (e.g., Baillargeon, Kotovsky, & Needham, 1995; Spelke, Breinlinger, Macomber, & Jacobson, 1992) showing what infants know about object permanence, solidity, substance, compressibility, inertia, the effects of gravity, and so forth—what is the status of these kinds of knowing? Are these facts that infants can potentially think or “reason” about? Or are they implicit sensorimotor knowledge that affects infants’ expectations but not their conscious thought processes? I raised this issue in the last Handbook of Child Psychology (J. M. Mandler, 1998), but haven’t yet heard anyone address it. I believe it is vital to further theoretical advancement that we do so. Psychologists have come far in understanding the importance of differentiating implicit and explicit knowledge, in part due to Katherine Nelson’s work. But there is still much to learn, and only when the contribution of implicit and explicit knowledge to the tasks we set infants can be specified will we be able to give a complete picture of the developing mind.
Postscript In contrast to other chapters in this book, there was no mention here of the influence of sociocultural context. That is because infants are to some extent shielded
28
MANDLER
from such influence by their lack of language. Even without language, there are of course some cultural influences that affect the course of categorization, such as the fact that a vehicle category is an early achievement of California infants. However, by and large, the early development of the ability to recall past events, to categorize objects, and to learn the important basics of language such as that it is used for communication, are all governed by universal factors common to infants in all cultures. It is when the foundations have been laid down and the naming practices of the culture begin to teach the infant which details are important that more cultural influence can be seen. Even so, in so far as categorization is concerned, there may be less influence of one’s native tongue on object categorization than on categorization of relations (Gentner, 1982). In two recent studies we found identical categorization of spatial relations by preverbal infants being raised in Englishspeaking and Korean-speaking homes, but major changes in these cultural groups after the onset of spatial language comprehension (Choi, McDonough, Bowerman, & J. M. Mandler, 1999; McDonough, Choi, & J. M. Mandler, 2003). Thus, culture may have important effects on thought but minimally so in infancy.
ENDNOTES 1
The implicit–explicit distinction is not identical to the procedural–declarative distinction, although they are related closely enough that the terms can be used interchangeably in this chapter. The imlicit–explicit distinction is more often used when discussing types of processing and the procedural–declarative distinction is more often used when discussing types of representation (for further discussion, see J. M. Mandler, 1998, 2004). 2 A condensed (and somewhat Bowdlerized) version of this heated debate can be found in the discussion section of J. M. Mandler (1990).
REFERENCES Abdi, H., Valentin, D., Edelman, B., & O’Toole, A. J. (1995). More about the difference between men and women: Evidence from linear neural networks and the principal-component approach. Perception, 24, 539–562. Bachevalier, J., & Mishkin, M. (1994). Effects of neonatal temporal lobe lesions on visual recognition memory in rhesus monkeys. Journal of Neuroscience, 14, 2128–2139. Bahrick, L. E., Hernandez-Reif, M. & Pickens, J. N. (1997). The effect of retrieval cues on visual preferences and memory in infancy: Evidence for a four-phase attention function. Journal of Experimental Child Psychology, 67, 1–20. Baillargeon, R., Kotovsky, L., & Needham, A. (1995). The acquisition of physical knowledge in infancy. In G. Lewis, D. Premack, & D. Sperber (Eds.), Causal understandings in cognition and culture. Oxford, England: Oxford University Press. Barnat, S. B., Klein, P. J., & Meltzoff, A. N. (1996). Deferred imitation across changes in context and object: Memory and generalization in 14-month-old infants. Infant Behavior and Development, 19, 241–251.
2. TWO KINDS OF KNOWLEDGE ACQUISITION
29
Barr, R., Dowden, A., & Hayne, H. (1996). Developmental changes in deferred imitation by 6- to 24-month-old infants. Infant Behavior and Development, 19, 159–170. Bauer, P. J. (in press). New developments in the study of infant memory. In D. M. Teti (Ed.), Handbook of research methods in developmental psychology. Oxford, England: Blackwell. Bauer, P. J., & Dow, G. A. (1994). Episodic memory in 16- and 20-month-old children: Specifics are generalized but not forgotten. Developmental Psychology, 30, 403–417. Carey, S. (1982). Semantic development: The state of the art. In E. Wanner & L. R. Gleitman (Eds.), Language acquisition: The state of the art. Cambridge, England: Cambridge University Press. Carver, L. J., & Bauer, P. J. (2001). The dawning of a past: The emergence of long-term explicit memory in infancy. Journal of Experimental Psychology: General, 130, 726–745. Carver, L. J., Bauer, P. J., & Nelson, C. A. (2000). Associations between infant brain activity and recall memory. Developmental Science, 3, 234–246. Choi, S., McDonough, L., Bowerman, M., & Mandler, J. M. (1999). Early sensitivity to language-specific spatial categories in English and Korean. Cognitive Development, 14, 241–268. Clark, E. V. (1983). Meanings and concepts. In P. Mussen (Series Ed.), & J. H. Flavell & E. M. Markman (Vol. Eds.), Handbook of child psychology: Vol. 3. Cognitive development. New York: Wiley. Fagan, J. F., III. (1973). Infants’ delayed recognition memory and forgetting. Journal of Experimental Child Psychology, 16, 424–450. Gentner, D. (1982). Why nouns are learned before verbs: Linguistic relativity versus natural partitioning. In S. A. Kuczaj II (Ed.), Language development: Vol. 2. Language, thought, and culture. Hillsdale, NJ: Lawrence Erlbaum Associates. Graf, P., Mandler, G., & Haden, P. (1982). Simulating amnesic symptoms in normal subjects. Science, 218, 1243–1244. Graf, P., Squire, L. R., & Mandler, G. (1984). The information that amnesic patients do not forget. Journal of Experimental Psychology: Learning, Memory, and Cognition, 10, 164–178. Haith, M. H. & Benson, J. B. (1998). Infant cognition. In W. Damon (Series Ed.) & D. Kuhn & R. Siegler (Eds.), Cognition, perception, and language: Vol. 2. Handbook of child psychology. New York: Wiley. Hauser, M. D., Newport, E. L., & Aslin, R. N. (2001). Segmentation of the speech stream in a non-human primate: Statistical learning in cotton-top tamarins. Cognition, 78, 353–364. Jacoby, L. L. (1983). Perceptual enhancement: Persistent effects of an experience. Journal of Experimental Psychology: Learning, Memory, and Cognition, 9, 21–38. Karmiloff-Smith, A. (1992). Beyond modularity: A developmental perspective on cognitive science. Cambridge, MA: MIT Press. Knowlton, B. J., Ramus, S. J., & Squire, L. R. (1992). Intact artificial grammar learning in amnesia: Dissociations of classification learning and explicit memory for specific instances. Psychological Science, 3, 172–179. Kolodny, J. A. (1994). Memory processes in classification learning: An investigation of amnesic performance in categorization of dot patterns and artistic styles. Psychological Science, 5, 164–169.
30
MANDLER
Mandler, G. (1985). Cognitive Psychology. Hillsdale, NJ: Lawrence Erlbaum Associates. Mandler, G. (2002). Consciousness recovered. Amsterdam: John Benjamins. Mandler, J. M. (1984). Representation and recall in infancy. In M. Moscovitch (Ed.), Infant memory. New York: Plenum. Mandler, J. M. (1988). How to build a baby: On the development of an accessible representational system. Cognitive Development, 3, 113–136. Mandler, J. M. (1990). Recall of events by preverbal children. In A. Diamond (Ed.), The development and neural bases of higher cognitive functions. New York: New York Academy of Sciences. Mandler, J. M. (1992a). How to build a baby II: Conceptual primitives. Psychological Review, 99, 587–604). Mandler, J. M. (1998). Representation. In W. Damon (Series Ed.) & D. Kuhn & R. Siegler (Eds.), Cognition, perception, and language: Vol. 2. Handbook of child psychology. New York: Wiley. Mandler, J. M. (2000a). Perceptual and conceptual processes in infancy. Journal of Cognition and Development, 1, 3–36. Mandler, J. M. (2000b). What global-before-basic trend? Commentary on perceptually based approaches to early categorization. Infancy, 1, 99–110. Mandler, J. M. (2004). The foundations of mind: The origins of conceptual thought. New York: Oxford University Press. Mandler, J. M., & McDonough, L. (1993). Concept formation in infancy. Cognitive Development, 8, 291–318. Mandler, J. M., & McDonough, L. (1995). Long-term recall in infancy. Journal of Experimental Child Psychology, 59, 457–474. Mandler, J. M., & McDonough, L. (1996). Drinking and driving don’t mix: Inductive generalization in infancy. Cognition, 59, 307–335. Mandler, J. M., & McDonough, L. (1997). Nonverbal recall. In N. L. Stein, P. O. Ornstein, B. Tversky, & C. Brainerd (Eds.), Memory for everyday and emotional events. Hillsdale, NJ: Lawrence Erlbaum Associates. Mandler, J. M., & McDonough, L. (1998). Studies in inductive inference in infancy. Cognitive Psychology, 37, 60–96. Mandler, J. M., & McDonough, L. (2000). Advancing downward to the basic level. Journal of Cognition and Development, 1, 379–403. Marcus, G. F., Vijayan, S., Bandi Rao, S., & Vishton, P. M. (1999). Rule-learning in seven-month-old infants. Science, 283, 77–80. Mareschal, D., French, R. M. & Quinn, P. C. (2000). A connectionist account of asymmetric category learning in early infancy. Developmental Psychology, 36, 635–645. McDonough, L. (2002). Basic-level nouns: First learned but misunderstood. Journal of Child Language, 29, 357–377. McDonough, L., Choi, S., & Mandler, J. M. (2003). Understanding spatial relations: Flexible infants, lexical adults. Cognitive Psychology,46, 229–259. McDonough, L., & Mandler, J. M. (1994). Very long-term recall in infants: Infantile amnesia reconsidered. Memory, 2, 339–352. McDonough, L. & Mandler, J. M. (1998). Inductive generalization in 9- and 11-month olds. Developmental Science, 1, 227–232.
2. TWO KINDS OF KNOWLEDGE ACQUISITION
31
McDonough, L., Mandler, J. M., McKee, R. D., & Squire, L. (1995). The deferred imitation task as a nonverbal measure of declarative memory. Proceedings of the National Academy of Sciences, 92, 7580–7584. McKee, R. D., & Squire, L. R. (1993). Journal of Experimental Psychology: Learning, Memory, & Cognition, 19, 397–404. Mishkin, M., & Appenzeller, T. (1987). The anatomy of memory. Scientific American, 256, 80–89. Moscovitch, M., Goshen-Gottstein, Y., & Vriezen, E. (1994). Memory without conscious recollection: A tutorial review from a neuropsychological perspective. In C. Umiltá & M. Moscovitch (Eds.), Attention and performance XV: Conscious and nonconscious information processing. Cambridge, MA: MIT Press. Nelson, C. A. (1997). The neurobiological basis of early memory development. In N. Cowan (Ed.), Memory development. London: University College London Press. Nelson, C. A., & Collins, P. F. (1991). Event-related potential and looking-time analysis of infants’ responses to familiar and novel events: Implications for visual recognition memory. Developmental Psychology, 27, 50–58. Nelson, C. A., & Collins, P. F. (1992). Neural and behavioral correlates of recognition memory in 4- and 9-month-old infants. Brain and Cognition, 1992, 105–121. Nelson, K. (1974). Concept, word, and sentence: Interrelations in acquisition and development. Psychological Review, 81, 267–285. Nelson, K. (1985). Making sense: The acquisition of shared meaning. Orlando, FL: Academic Press. Newcombe, N., & Fox, N. (1994). Infantile amnesia: Through a glass darkly. Child Development, 65, 31–40. Nosofsky, R. (1992). Exemplar-based approach to relating categorization, identification, and recognition. In F. G. Ashby (Ed.), Multidimensional models of perception and cognition. Hillsdale, NJ: Lawrence Erlbaum Associates. Piaget, J. (1952). The origins of intelligence in the child. New York: International Universities Press. Quinn, P. C., & Eimas, P. D. (1996). Perceptual cues that permit categorical differentiation of animal species by infants. Journal of Experimental Child Psychology, 63, 189–211. Quinn, P. C., Eimas, P. D., & Rosenkrantz, S. L. (1993). Evidence for representations of perceptually similar natural categories by 3-month-old and 4-month-old infants. Perception, 22, 463–475. Quinn, P. C., & Johnson, M. H. (2000). Global-before-basic object categorization in connectionist networks and 2-month-old infants. Infancy, 1, 31–46. Rice, M. (1978). The effect of children’s prior nonverbal color concepts on the learning of color words. Unpublished doctoral dissertation, University of Kansas. Rosch, E., Mervis, C. B., Gray, W., Johnson, D., & Boyes-Braem, P. (1976). Basic objects in natural categories. Cognitive Psychology, 3, 382–439. Rovee-Collier, C. (1997). Dissociations in infant memory: Rethinking the development of implicit and explicit memory. Psychological Review, 104, 467–498. Shatz, M., & Backscheider, A. G. (1999). Toddlers create lexical domains: The case of color. Unpublished manuscript, University of Michigan, Ann Arbor. Sirois, S., Buckingham, D., & Shultz, T. R. (2000). Artificial grammar learning by infants: An auto-associator perspective. Developmental Science, 3, 442–456.
32
MANDLER
Smith, E. E., Patalano, A. L., & Jonides, J. (1998). Alternative strategies of categorization. Cognition, 65, 167–196. Smith, P. H. (1984). Five-month-old recall and utilization of temporal organization. Journal of Experimental Child Psychology, 38, 400–414. Spelke, E. S., Breinlinger, K., Macomber, J., & Jacobson, K. (1992). Origins of knowledge. Psychological Review, 99, 605–632. Squire, L. R., & Knowlton, B. J. (2000). The anatomical bases of memory. In M. S. Gazzaniga (Ed.), The new cognitive neurosciences (2nd ed.). Cambridge, MA: MIT Press. Squire, L., & Zola-Morgan, S. (1991). The medial temporal lobe memory system. Science, 253, 1380–1386. Ullman, M., Corkin, S., Coppola, M., Hickok, G., Growdon, J. H., Koroshetz, W. J., & Pinker, S. (1997). A neural dissociation within language: Evidence that the mental dictionary is part of declarative memory, and that grammatical rules are processed by the procedural system. Journal of Cognitive Neuroscience, 9, 266–276.
3 New Insights Into the Functions, Development, and Origins of Theory of Mind: The Functional Multilinear Socialization (FMS) Model Joan Lucariello Boston College
A much heralded milestone in theory of mind (ToM) development is achievement of a metarepresentational ToM. This ToM is an advanced understanding of mental states that is based in the ability to maintain multiple, contrastive representations of an object or event. This ability affords distinguishing mental states from reality by keeping in mind both representations of reality and representations that contrast with reality. A metarepresentational ToM is evident in behaviors such as false belief (e.g., knowledge of another’s erroneous belief with respect to reality), distinguishing appearance from reality, and representational change (e.g., knowledge of one’s own past false belief). A metarepresentational ToM is attained around age 4 to 5, as indicated by success on tasks that measure these metarepresentational behaviors (e.g., Astington, Harris, & Olson, 1988; J. H. Flavell, E. R. Flavell, & Green, 1983; Gopnik & Astington, 1988; Perner, 1991). 33
34
LUCARIELLO
The importance of this cognitive achievement has given rise to a veritable industry of associated theory and research. However, for all the attention dedicated to this behavior, a functional analysis has been remarkably absent. A functionalist perspective to understanding ToM is adopted here. On this functional analysis, some of the most fundamental assumptions now held about a metarepresentational ToM are overturned. A new account of ToM—its functions, origins, and development—will be offered. This account is termed the Functional Multilinear Socialization (FMS) model of ToM. The FMS model and data that support it are presented in this chapter. On the FMS model, ToM is viewed in terms of its functions in the lives of children and adults. Whereas clearly some have argued that ToM, or “mind reading,” is critically important in the conduct of social interaction, not all ToM functioning is in the service of social interaction. There are other major uses for ToM, which may be classified as intrapersonal. These uses, such as reflection and learning, are identified and discussed. The distinct functions of ToM, social and intrapersonal, will be seen to correspond with, indeed be the basis of, distinct kinds of ToM. Hence, ToM will be defined as a differentiated, nonintegrated cognitive skill. Consistent with the idea of a differentiated ToM, development of ToM, heretofore understood as unilinear, will be cast as multilinear. Moreover, development will be understood as extending into late childhood and as consisting in the emergence of progressively more advanced ToMs. Furthermore, this functional analysis will lead to a new proposal on the origins of ToM and one that posits a distinct sociocultural basis for distinct ToM kinds. Viewing ToM through a functionalist perspective and with a basis in sociocultural context is an endeavor well-placed in a volume dedicated to examining the mediating influence of sociocultural experience on cognition.
THE FUNCTIONAL VIEW: SOCIAL AND INTRAPERSONAL THEORIES OF MIND On the FMS model, a metarepresentational ToM (and the more advanced ToMs that develop subsequently) must be understood in terms of its functions. Metarepresentational reasoning has different functions and these serve as the basis for distinct kinds. One function is social interaction. Understanding and predicting the mental and emotional states of others, especially when these are inconsistent with reality, is critical in the conduct of nimble social interaction. For this function, reasoning about others’ representations is what is important. Reasoning about the contrastive representations or mental states of others may be termed social metarepresentation. False belief, which entails reasoning about another’s contrastive and erroneous beliefs with respect to reality, is a case in point. Hence the false-belief task measures social metarepresentation. The second function for metarepresentational ToM is intrapersonal. Two major intrapersonal uses of ToM include reflection and learning. Intrapersonal-reflec-
3. THE FMS MODEL
35
tion is metarepresentational reasoning that entails one’s own contrastive mental representations about an object or event where no learning is involved. Intrapersonal-reflection is involved in distinguishing appearance from reality. When intrapersonal ToM entails learning it can be termed constructivist. Representational change, and the tasks assessing it, is a case of intrapersonalconstructivist metarepresentational reasoning. Representational change entails learning in the form of knowledge replacement. One maintains contrastive representations of one’s past, incorrect knowledge of an object or event and one’s new, corrected knowledge of such. Reasoning that entails one’s own contrastive representations, as occurs in reflection and learning, may be termed intrapersonal metarepresentation. On the FMS model, social and intrapersonal metarepresentational ToMs do not rest in a single, underlying ToM ability. Accordingly, understandings of own versus others’ representations are not purported to emerge necessarily together at a single ontogenetic time point. Hence the FMS theory runs counter to the current view that ToM is an integrated cognitive skill (e.g., Gopnik & Astington, 1988; Gopnik & Wellman, 1994; Wellman, Cross, & Watson, 2001). On the integrated view, reasoning about own and others’ representations are based in a single and the same ToM ability and these reasonings emerge together ontogenetically. If reasoning about others’ versus own representations do not amount to the same skill cognitively, then differential performance across tasks measuring each would be expected. Such evidence is starting to build. Performance across false-belief (“other” belief) and representational change (“own” belief) tasks is not correlated for 3-year-olds or for 4-year-olds (Moore, Barresi, & Thompson, 1998). Similarly, although intertask correlations were found for nearly every task among eight that measured false belief in preschoolers, a notable exception was the “recall-own-false-belief” task (Cutting & Dunn, 1999). Moreover, an ordering of successful performance has been found, with false belief acquired prior to representational change (Gopnik & Astington, 1988). So too performance across the false-belief and own-belief appearance–reality tasks is not correlated (Meins, Fernyhough, Wainwright, Gupta, Fradley, & Tuckey, 2002). In addition, whereas 5-year-olds commonly pass the other-belief reasoning task of false belief, a “source” task proved very difficult for 4- and 5-year-old children, with the majority of 5-year-olds failing this task (Ruffman, Perner, Naito, Parkin, & Clements, 1998). Source tasks assess children’s understanding of the source of their own representations (e.g., children are asked what they need to do to know an object’s color; the correct answer is “see it” vs. “feel it”). Finally, although children in the 3- to 5-year-old age range were able to explain why another might experience mixed emotions about a given situation, they were unable to provide an occasion on which they themselves experienced mixed emotions (Hughes & Dunn, 1998). All these findings speak to a distinction between “own” and “other” mental state reasoning. In contrast to these findings, the recent meta-analysis conducted by Wellman, Cross, and Watson (2001) documented no differences in performance across self
36
LUCARIELLO
and other reasoning tasks. However, there are three significant limitations to the Wellman et al. (2001) analysis. First, their review ended with studies published prior to January 1998. Hence the more recent studies showing such differences and cited here were not included (e.g., Cutting & Dunn, 1999; Moore et al., 1998; Ruffmann et al., 1998). Second, the studies reviewed were those that limited their assessments to that of false belief (own and other). However, reasoning about other and own mental states can be assessed in many other tasks, such as “source” and appearance–reality tasks (both related to own representations) (Ruffman et al., 1998) and tasks related to emotional states (Hughes & Dunn, 1998). Finally, and perhaps most importantly, the ToM literature, and hence the studies reviewed, has in the main been conducted on samples of middle-income children. Where studies have had low-income participants, SES has rarely, if ever, been an independent variable. Middle-income children may show little diversity across self and other tasks, but this finding cannot automatically be taken to mean such reasoning is one and the same thing. This population might not be expected to vary across such measures, whereas other populations might (see later discussion on socialization origins of ToM). Indeed, the Cutting and Dunn (1999) study that showed reasoning about own and other beliefs not to be correlated included low-income participants. In summary, findings are starting to accrue on differential performance across tasks of other and own belief. Yet, these findings have been incidental. There have been no direct tests of the hypothesis that metarepresentational reasoning differentiates according to others’ or own representations. In addition, prior research has not distinguished reflection and learning/constructivist lines within the intrapersonal track of metarepresentational reasoning. Accordingly, direct test that metarepresentational reasoning differentiates along these three lines is needed.
MULTILINEAR DEVELOPMENTAL COURSE: METAREPRESENTATIONAL TO PERSPECTIVAL TO DIALECTICAL ToMs The FMS model of ToM provides a new proposal about ToM development that accords with its nonintegrated nature. The ToM developmental course is multilinear. ToM development proceeds along “social” and “intrapersonal” tracks. Along these tracks, development consists in a hierarchical progression in terms of the complexity of the relation able to be maintained across representations. This development is captured in the progression from metarepresentational to perspectival to dialectical ToMs. See Table 3.1. A metarepresentational ToM entails maintaining contrastive, alternative, representations or emotions about an object or event. One juxtaposes representations that are alternatives to one another. Hence, with respect to social metarepresentation and the false-belief task, the contrastive, alternative representations are one’s own ver-
37
representations
10+ One and Another reps True and True opposing emotions
Dialectical: contradictive, concurrent, true reps1 Tasks/Behaviors:
1
7+
Perspectival: contrastive, concurrent, true reps1 Tasks/Behaviors: One and Another reps True and True appreciation of ambiguity, mixed emotions
5+
Other’s or Self’s reps Untrue or True false belief
Social: Others’ Age Representations
Metarepresentational: contrastive, alternative reps1 Tasks/Behaviors
ToM Developmental Path (Relation of Representations)
Past and Present reps True and True homonym learning
One and Another reps True and True appreciation of ambiguity, mixed emotions
Past and Present reps True and True counterscript learning
Past or Present reps Untrue or True representation change
One or Another reps ------------------------appearance vs. reality
One and Another reps True and True opposing emotions
Learning/Constructivist
Reflection
Intrapersonal: Own Representations
Theory of Mind (ToM) Kinds
Developmental Path for ToM by Age and ToM-Kind
TABLE 3.1
38
LUCARIELLO
sus that of the other. Moreover, these representations are additionally alternative in that one of them (one’s own) is true with respect to reality and the other (someone else’s) is false. With respect to intrapersonal-reflective metarepresentation, the alternative, contrastive representations maintained in the appearance–reality task are of an object’s appearance versus its reality/identity. Development consists in a move toward a perspectival ToM, which is the ability to maintain contrastive, concurrent, and true (equally correct with respect to reality) representations or emotions about an object or event. On the “social” ToM track, it is evident in reasoning about others’ two equally plausible and contrastive interpretations or emotions (of an object or event). Such is seen in understanding another’s mixed emotions or ambiguous interpretations. The intrapersonal-reflective ToM line entails maintenance of one’s own contrastive, concurrent, true representations about an object or event. Hence, one’s own appreciation of ambiguity or experience of mixed emotions would be emblematic behaviors. Subsequent ToM development leads to the emergence of a dialectical ToM. Here one maintains concurrent and true (equally correct) representations or emotions about an object or event, where these representations are contradictive to one another. A social dialectical ToM would consist in reasoning about others’ two equally plausible, contradictory interpretations (of an object or event), as in their experiencing opposing emotions to the same target (e.g., Sam loves the bird because his best friend gave it to him and he hates the bird because Sam hates birds.) Intrapersonal-reflective dialectical ToM consists in one’s own representations/emotions about an object or event being contradictory (e.g., I love the bird because my best friend gave it to me and I hate the bird because I hate birds.). This same developmental sequence of ToMs occurs with respect to “intrapersonal–constructivist” ToM, that is, where ToM is entailed in learning. On a constructivist theory of knowledge acquisition, learning occurs in the context of current knowledge. Learning is most challenging when it involves accretion of new knowledge that is inconsistent with current knowledge (Gelman & Lucariello, 2002). Much learning, particularly that beyond very early childhood, however, is of this more challenging sort. ToM, in terms of metarepresentational, perspectival, and dialectical ToMs, is the core cognitive process needed to learn in such cases. See Table 3.2. Where learning entails knowledge revision, that is, correcting one’s knowledge in the face of new inputs, a metarepresentational ToM is the core cognitive process. Contrastive, alternative representations of one’s past versus new knowledge must be entertained. Moreover, these contrastive representations are additionally alternative in that the new belief is true (correct) whereas the past belief is false. See Table 3.1. Such learning is assessed in the representational change task. For example, in the unexpected-identity, deceptive object method (e.g., rock-sponge), the pretest question “What is it?” elicits one’s current representation of the object as a sponge. Then the child is asked what he or she thought the object was before
39
Representational change task
Homonym learning
Accommodation
Transcendence
Transcendence
Metarepresentational
Perspectival
Dialectical
Knowledge revision: Current knowledge: False New knowledge: True
Knowledge acquisition: Current knowledge: True New knowledge: True & contrastive to past
Knowledge acquisition: Current knowledge: True New knowledge: True & contradictive to past
Counterscript learning
Task/Behaviors
Functions
Constructivist ToM
Learning-Challenge (simple®complex)
Requisite Cognitive Processes
Learning Challenges by Requisite Cognitive Processes and Emblematic Tasks/Behaviors
TABLE 3.2
40
LUCARIELLO
touching it. The correct answer is “rock.” Learning as knowledge revision also entails the cognitive process of “accommodation,” described by Piaget (1952, 1954). One has a belief that, in light of additional inputs, is incorrect. The belief changes accordingly. Learning does not always entail knowledge revision. In some cases, it entails knowledge acquisition, where no revision of knowledge is involved and where one’s current knowledge must be maintained as is. This would be the case where current knowledge is true and the related to-be-acquired knowledge is also true, but contrastive. See Table 3.1. Learning homonyms would be an example. Children must learn a second, new, and contrastive meaning for a word for which they already have a meaning. See Table 3.2. For such learning, a perspectival ToM, which is the ability to maintain contrastive, concurrent, true representations, is required. In addition, another cognitive process is needed. Rather than accommodation, a process of “transcendence” is required. Transcendence is moving beyond past representations, when these must be maintained (due to being reliable) in acquiring new, related representations that are inconsistent with past representations. Some cases of knowledge acquisition are even more challenging. This is so where current knowledge is true and the related to-be-acquired knowledge is true but also contradictive. See Table 3.1. For such learning, a dialectical ToM, which is the ability to maintain contradictive, concurrent, true representations, is required. Here too, the additional cognitive process of transcendence is needed because past representations must be maintained (due to being reliable) in acquiring the new, related representations that are contradictory. See Table 3.2. Learning counterscripts would be an instance of such learning (see Lucariello, 1994; Lucariello & Mindolovich, 1995, 2002; Lucariello, Mindolovich, & LeDonne, 2003 for a more complete definition of counterscript events). Counterscripts are knowledge representations for unexpected events that exhibit systematic, shared structural features. One class is situationally ironic events. These include “backfired plan” events, wherein your goal-directed actions lead to the opposite outcome. “Role reversal/contradictory” events, wherein bankers are poor and dentists are missing teeth, are also situationally ironic. Counterscripts may be contrasted with “scripts.” These are initial event knowledge structures for highly expected, commonplace events (Bauer & Mandler, 1990, 1992; Bauer, Wenner, Dropik, & Wewerka, 2000; Nelson, 1986; Nelson & Gruendel, 1981). In learning counterscripts, contradictory tiers of event representations are to be maintained in the knowledge base. These tiers are the primary (past) representations, scripts, and the secondary or metarepresentations, counterscripts. In acquiring and processing “counterscript representations” (“not A”), which render how events are “not supposed” to occur, one necessarily understands “script representations” (“A”), which specify how they are “supposed” to occur. Counterscript learning is very difficult, as indicated in story completion (Lucariello & Mindolovich, 1995) and story recall (Lucariello, Mindolovich, & LeDonne, 2003) tasks.
3. THE FMS MODEL
41
Delineating the intrapersonal constructivist ToM line illuminates the role of ToM in learning. As the learning challenge increases from knowledge revision to knowledge acquisition and to such when it entails contrastive then contradictive representations, the intrapersonal ToM required progresses from metarepresentational to perspectival to dialectical, respectively. The FMS model of ToM greatly expands the terrain of ToM development as staked out by current theory. To date, the major proposal has been that a new ToM, an “interpretive” or “perspectival” mind, emerges around age 7 (Carpendale & Chandler, 1996; Chandler, 1987; Chandler & Lalonde, 1996). The interpretive mind entails appreciating the relativity of knowledge across persons, that is, that persons experiencing the same information can come to different beliefs about it. This understanding represents a progression from false-belief understanding (wherein persons have experienced different information about an object or event and hence have different beliefs about it). However, viewing ToM development as the emergence of an “interpretive” mind constrains ToM development to the social track and limits its trajectory to perspectival reasoning. The FMS model incorporates this developmental line while expanding ToM development to include the intrapersonal ToM line. Moreover, it posits a further progression, hence culminating point, for multilinear development in a dialectical ToM.
SOCIALIZATION ORIGINS OF ToM: NO MODULES, SIMULATIONS, OR THEORIES The FMS model of ToM not only differentiates across reasoning about own and others’ representations, but also proposes that each of these forms of reasoning has a distinctive sociocultural basis. Differential exposure to the respective, relevant, sociocultural experiences would lead to uneven development across these forms of reasoning. In some children, social metarepresentation might emerge first, function more effectively, and develop further than intrapersonal metarepresentation. For other children, the reverse might hold. No current theory of ToM origins would predict (or could account for) an uneven developmental course that could take either of these directions. On the FMS model, social ToM originates and develops in socialization experiences that orient the child to the mental states of others. Obviously, social interaction is the fundamental experience wherein it is necessary to read the mind of others. Not surprisingly then, the development of early social ToM in the form of metarepresentational reasoning is related to participation in the very basic, ubiquitous, social-interactive exchanges that typify Western culture. The meta-representation evident in false belief develops more readily in children who have social interactions with siblings (Dunn, Brown, Slomkowski, Tesla, & Youngblade, 1991; Lewis, Freeman, Kyriakidou, Maridaki-Kassotaki, & Berridge, 1996; Perner, Ruffman, & Leekam, 1994; Ruffman et al., 1998) and with friends, with
42
LUCARIELLO
whom they engage in frequent reference to mental states (Hughes & Dunn, 1998), and with adult kin (Lewis et al., 1996). The more complex forms of social ToM, “perspectival reasoning” and “dialectical reasoning,” develop from other and nonubiquitous types of socialization experiences that orient the child to the mental states of others. One of these is an interdependent or collectivist self-concept whereon self is understood fundamentally in terms of relations to others (Markus & Kitayama, 1991). Maintaining relationships with others and ensuring positive social interactions requires knowing how others are feeling, thinking, and likely to act and hence should lead to more readily accessible knowledge of the other (Markus & Kitayama, 1991). Indeed, Chinese children, who presumably have an interdependent self-concept, show greater perspective-taking ability than Australian children, with presumably an independent self-concept (Fu-Xi & Keats, 1989). Reciprocally, 4- and 5-year- old Japanese children (who presumably have an interdependent self-concept) generally fail an own-belief (“source”) task, while performing successfully on the other-belief, false-belief task (Ruffman et al., 1998). Socialization of a pragmatic-interpersonal model of language use would also direct children’s attention to others and their mental states. This model is evident in both the semantics and pragmatics of language. Semantic-syntactic relations, such as vocative (e.g., Mary; hey) and social expression (e.g., thank you; I’m sorry), organize children’s behavior on a social and emotional plane (Blake, 1994). Certain pragmatic functions, such as instrumental-regulatory (attempts to regulate the action of another or use another to achieve an end) and interpersonal or interactive (establishing or maintaining contact with another), are social in nature (Halliday, 1975; Nelson, 1981). Socioemotional functions include the “interpersonal expressive,” which establishes, describes, and manages relationships and influences the quality of behavior (e.g., bye now; that’s mine). Another socioemotional function is the “effective,” which results in material gain and direct changes in behavior (e.g., you can help; I want cookies) (Blake, 1994). Finally, the “self-expressive social” function expresses social customs and reports feelings and opinions to an intended listener (e.g., I don’t want to play ball; hurt foot). Presumably, exposure to and use of language in these ways would facilitate development of social perspectival and social dialectical ToMs. In Western culture, the interdependent self-concept and the pragmaticinterperonal model of language are not as ubiquitous as the common social- interactive experiences said to underlie the development of metarepresentational ToM. Accordingly, social perspectival and social dialectical reasoning might not be expected to be as pervasive in Western culture as is social metarepresentational reasoning. Intrapersonal ToM, in contrast to social ToM, originates and develops in culture-specific socialization practices that orient individuals to their own mental states. These sociocultural experiences include an independent self-concept and a referential-intrapersonal model of language use.
3. THE FMS MODEL
43
Having an independent self-concept means maintaining independence from others by attending to the self and by discovering and expressing one’s unique inner attributes (Markus & Kitayama, 1991). The person is a self-contained motivational and cognitive universe (Geertz, 1975). Accordingly, an independent self-concept should direct the individual’s attention to his or her own mental states (Greenfield & Bruner, 1966; Markus & Kitayama, 1991). Greenfield and Bruner (1966) found self-consciousness (reasoning about one’s own mental states) to result, at least in part, from the independent orientation. Indeed, self-consciousness was lacking in their Wolof subjects in Senegal. These subjects were characterized by an interdependent or collectivist orientation. In their conservation studies, Greenfield and Bruner were unable to use the traditional subjective form of justification questions (e.g., Why do you think there is more water in this glass than in that glass?) because the Wolof had difficulty with it. The question had to be changed to an objective form (e.g., Why is there more water in this glass than in that one?). Relatedly, Delgado-Gaitan (1994) showed that an interdependent self-concept detracts from the development of “critical thinking” processes in children. Reciprocally, an independent self-concept has been proposed to facilitate complex reasoning with respect to one’s own mental states, such as hypothetical and counterfactual reasoning (Markus & Kitayama, 1991). The referential-intrapersonal model of language socialization also serves to direct children’s attention to their own mental states. This socialization takes several routes. It is seen in the conversational model that guides parents’ discourse with their children (see Snow, 1986). This model is based on an “intentionalist” approach to the child. The child is seen as an intentional agent, with mental states, from the outset of life. Children are treated as intentional communicative partners, with speech directed to them from infancy. Furthermore, characteristics of child-directed speech, such as expansions and requests for clarification, focus on the child’s intentional states. Indeed, expression of intentional states has been proposed as the basis for the child’s word learning and early multiword utterances (Bloom, 1993). Moreover, expressions of one’s internal states is seen in more advanced language uses, such as the symbolic (as in pretend play) and hypothetical (e.g., “if…). These uses derive from and focus attention on the mental states of the speaker (Halliday, 1975). Socialization experiences then vary with respect to self-concept and language use. Socialization that emphasizes an interdependent self-concept and pragmatic-interpersonal language use is likely to foster advanced social ToM development, such as social perspectival and social dialectical reasoning. (Socialization consisting in very ubiquitous social-interactive exchanges underlies the less complex social metarepresentational reasoning.) Socialization that accentuates an independent self-concept and referential-intrapersonal language use is likely to foster intrapersonal ToM development. Accordingly, children whose socialization was predominantly of one form or the other would manifest differential strengths, hence uneven, ToM development across social and intrapersonal ToM kinds.
44
LUCARIELLO
As noted, no current account of the origin of ToM would presume or could explain uneven development consisting in differential emergence-time, accuracy, and/or developmental endpoint across social and intrapersonal ToM reasoning. Three of the four current accounts of ToM origins—modular mechanism, theory–theory (TT), sociocultural—advocate an integrated ToM view. They describe reasoning about own and others’ representations as equivalent and deriving from the same theory of mind. Hence these reasonings are said to emerge together and at the same ontogenetic point. The modularity account of theory of mind, Theory-of-Mind Mechanism/Selection Processing (ToMM/SP) proposes that ToM has a specific, innate basis (German & Leslie, 2000, 2001; Leslie & Polizzi, 1998; Leslie & Thaiss, 1992; Scholl & Leslie, 1999, 2001). With respect to specificity, the essential character of ToM is said to be determined by specialized mechanisms deploying specialized representations that do not apply to other cognitive domains and so can be selectively impaired. With respect to innateness, the essential character of ToM—including concepts of belief, desire, and pretense—is thought to be part of our genetic endowment, which is triggered by appropriate environmental factors. The origin of ToM is thought to be a cognitive module that spontaneously and postperceptually processes behaviors that are attended and computes the mental states that contribute to them. In doing so, it imparts an innate concept of belief, which is hence available to the child prior to other abstract concepts acquired through general theory construction. Supplemental processing, termed selection processing, represents general executive processes that operate when the child is to select the correct content of beliefs that are false. The SP is needed because the ToMM automatically attributes beliefs with contents that are true. This prepotent response needs to be sometimes inhibited, as in the case of false belief. On the ToMM/SP account of ToM origins, the innate concept of belief encompasses beliefs of both self and other. Indeed, such a module would not be posited unless such a distinction was presumed cognitively irrelevant. The theory–theory (TT), or conceptual change account of ToM, describes ToM development in terms of an implicit theory that affords generalizations in which beliefs and desires, and so forth, are interrelated (Gopnik, 1993; Gopnik & Wellman, 1994; Perner, 1991; Wellman & Cross, 2001). Perner’s (1991) account discusses children’s ToM development as a move from “situation-theorists” at 2 to 3 years of age to “representational theorists” at 4 to 5 years of age. Whatever the theories attributed to the child on TT accounts, they do not encompass a significant distinction between the mental states of others and self. Indeed, Wellman et al. (2001) state that ToM development is “of an interrelated body of knowledge, based on core mental-state constructs such as ‘beliefs’ and ‘desires,’ that apply to all persons generically, that is, to both self and others” (p. 678). The sociocultural account too makes no differentiation in ToM across reasoning about self and others’ mental states. The contextual variables that affect development of ToM reasoning are thought to do so equally for reasoning about own
3. THE FMS MODEL
45
and others’ representations. These sociocultural variables include the semantics of language (Vinden, 1996), conceptions of self and personhood (Vinden & Astington, 2000), and social interactions (Dunn et al., 1991; Hughes & Dunn, 1998; Lewis et al., 1996; Perner et al., 1994; Ruffman et al., 1998). The fourth account, simulation theory (ST) (Harris, 1992), does differentiate between reasoning about self and others’ representations. However, it assigns primacy to reasoning about one’s own representations. Children are said to improve their theory of mind understanding through a simulation process. Understanding other minds is based in using one’s own experience to simulate that of others. The simulation process begins with children feeding into their own perceptual and or emotional system another person’s currently attended visual target and or emotional stance toward the target online. It develops to attributing the stance that is being simulated to the other person (e.g., “liking X”) then to imagining another person’s intentional stance (e.g., pretend another person sees an invisible object or wants an object they themselves do not want). The fourth developmental step is imagining an intentional stance toward counterfactual targets. On ST, persons start out with more accurate and more advanced reasoning about their own, than others’, mental states. Hence, ST claims uneven ToM development in only one direction. Accordingly, this theory, unlike the present FMS model, could not account for the reverse situation wherein individuals might begin with more accurate and more advanced reasoning about others’ mental states. In summary, all current accounts of ToM origins run counter to the claims of the FMS model. On this model, social and intrapersonal ToMs have distinctive bases in socialization experiences. Their unique origins allow for uneven development, whereby either of these two ToM kinds might emerge earlier, be more accurate, and/or develop further than the other.
SUMMARY CLAIMS OF FMS ToM MODEL A metarepresentational ToM is not an integrated cognitive ability. Rather, it differentiates into distinct kinds based on distinct functions. Social metarepresentation entails reasoning about others’ representations and is used and needed for social interaction. Intrapersonal metarepresentation entails one’s own mental states and is involved in reflection and learning. The development of ToM is multilinear, encompassing development in these three ToM lines. Moreover, development consists of a hierarchical progression in terms of the complexity of the relation able to be maintained across representations. This development is captured in the progression from metarepresentational to perspectival to dialectical reasoning along social and intrapersonal-reflective and intrapersonal-constructivist ToM tracks. Finally, social and intrapersonal ToMs have distinctive bases in sociocultural experiences. Their unique origins afford the prospect of their uneven development. Either ToM might emerge earlier, function more accurately, and/or develop further than the other kind.
46
LUCARIELLO
RESEARCH HYPOTHESES AND DESIGN The research reported here examines the hypothesis that ToM differentiates into reasoning about others’ representations (social ToM) versus self/own representations (intrapersonal ToM) and that the latter differentiates into reflection and constructivist forms. This hypothesis on ToM kinds can be examined, indeed may only be examinable, by relying on the second hypothesis—that social and intrapersonal metarepresentation each originate in distinct sociocultural experiences. Children whose socialization experiences would lead to uneven development across social and intrapersonal metarepresentation should show differential performance across tasks assessing these different kinds of reasoning. Low-income children are selected for study here. Their typical socialization experiences would underlie development of social metarepresentational reasoning (early and later forms) and not support development of intrapersonal metarepresentational reasoning. These children tend to be socialized into the pragmatic-interpersonal model of language use (e.g., Miller, 1982, 1987, 1988). Parents of these children do not rely on the conversational-intentionalist language style in interacting with their children. Often adult speech is not directed to very young children and, when so, it does not typically include expansions and clarifications of the child’s meaning (Brice Heath, 1983). Moreover, socioemotional functions predominate in maternal speech (Blake, 1994). Furthermore, these children receive less exposure to the referential-cognitive model of language use. Quantitatively, they experience less parental speech (Hart & Risley, 1995; Hess & Shipman, 1965). Qualitatively, and with respect to vocabulary specifically, low-income parents show meager use of “sophisticated” words (Weizman & Snow, 2001) and abstract words (Hess & Shipman, 1965) and a less diverse and less hierarchically differentiated vocabulary (Lawrence & Shipley, 1996). Not surprisingly, low-income children’s language uses are consistent with their socialization. These uses have been described as “contextualized” (Snow, 1991). Language largely serves interpersonal and emotional functions (Blake, 1993, 1994; Brice Heath, 1983, 1989; Miller, 1986). As Snow points out, “the skills relevant to interpersonal negotiation are honed through face-to-face conversations in which speakers and hearers may draw upon such resources as shared knowledge, gesture, interactive negotiation of meaning, and listener feedback” (p. 7). Interpersonal functions such as teasing and storytelling are commonly used by low-income children (Brice Heath, 1983; Miller, 1986). Moreover, low-income children tend to rely more heavily on socioemotional functions and socioemotional semantic-syntactic expressions than do middle-income children (Blake, 1994). The socialization of low-income children with respect to self-concept would also more likely lead to development of social, as opposed to intrapersonal, ToM. An interdependent self-concept is characteristic of minority children, many of
3. THE FMS MODEL
47
whom are low income (Delgado-Gaitan, 1994; Goldenberg & Gallimore, 1995; Greenfield, 1994). Accordingly, low-income children are presented with three prototypical metarepresentational ToM tasks within an unexpected-identity task. The false-belief task measures social metarepresentation. The distinguishing appearance–reality and representational change tasks assess intrapersonal metarepresentation in its reflection and constructivist forms, respectively. Performance across these tasks should not be correlated. Moreover, due to their socialization experiences, these children should perform fairly successfully on the other-belief task of false belief, but much less successfully on the own-belief tasks of appearance–reality and representational change. In addition, their vulnerability with respect to intrapersonal metarepresentation may cause their performance to vary across the two intrapersonal metarepresentation tasks. Representational change entails learning, hence constructivist metarepresentation. Distinguishing appearance from reality does not. Accordingly, it should be easier for children to engage in intrapersonal metarepresentational reasoning in the latter task that demands only reflection.
Method Participants Thirty low-income 5 to 6-year-old children participated. They were recruited from a child-care center, for low-income families, in a large eastern city.
Materials Children were first shown and then manipulated a deceptive object. This was a small sponge in the shape of, and painted to look like, a rock.
Procedure Children were all tested individually in a separate, quiet room in the child-care center by the same experimenter. The experimenter had a coding sheet on which children’s responses were recorded. In addition, these responses were audiotaped to check for accuracy of the written record. The method followed that of Gopnik and Astington (1988). Children were told: “We are going to play a game where we show you some things and ask questions about them.” Children were first shown the object in its deceptive form. They saw the rock-sponge at the other side of the table and were told to look at it. The true nature of the object was then revealed. Children were allowed to pick up and squeeze the object. Then children were asked to identify the true nature of the object as a sponge. They were asked the pretest question “What is it?” The correct answer is “sponge.” This pretest question is
48
LUCARIELLO
to ensure that the children represented the object as a sponge, after its deception was revealed. Pretest Question Administration and Coding. Many children had difficulty providing the label “sponge” in response to this pretest question. In response to this difficulty, the experimental procedure of Gopnik and Astington (1988) was modified. Children were given two additional attempts to answer the question. If they did not respond with the label “sponge” to the initial “what is it?” question, then they were asked a second question, “what else is it?” If this failed to elicit the “sponge” response, then they were asked a third and final question, “Is it like anything you have seen or used before?” If they responded in the affirmative, then they were asked the follow-up question, “Then what is it like?” Children’s responses to these three pretest questions were coded into several mutually exclusive categories: I. LABELS A. Normative—these responses were the label “sponge” for the object B. Alternative—these responses were semantically appropriate labels or labels with modifiers for the object (e.g., “pillow,” “rock pillow,” “little mattress,” “soft cotton,” “dishwasher rag,” “rag”). II. NON-LABELS A. Label Misses 1. label hedge—these responses were phrases that marked the object as similar or dissimilar to what it appears to be (e.g., “it’s like a rock,” “not a rock”) or what it really is (e.g., “it’s like a sponge”). 2. incorrect label—these responses were incorrect labels for the object (e.g., “bed,” “mom’s pillow so soft like that”). B. Object Use—these responses were phrases that talked about the object in terms of what one can do with it (e.g., “something to wash dishes with,” “something you clean with,” “scrub with”). III. OTHER A. Featural—these responses named a perceptual characteristic of the object (e.g., “it looks like it’s a real rock,” “soft”). B. Identity Error—these responses wrongly named the identity of the object in terms of its appearance (e.g., “real rock”). C. None—these responses provided no information (e.g., “it’s nothing”). Test Questions’ Administration. After the pretest questions, the object was then returned to its previous deceptive state (placed out of reach of the child). Then the three sets of test questions were administered following Gopnik and Astington (1988). These investigated understanding of false belief, the appearance–reality distinction, and representational change. Order of these three sets of test questions was counterbalanced across children.
3. THE FMS MODEL
49
The representational change question set was: “When you first saw this, before you touched it or squeezed it, what did you think it was: Did you think it was a rock or did you think it was a sponge?” The false-belief question set was: “X hasn’t touched this, [s]he hasn’t squeezed it. If X just sees it over here like this, what will [s]he think it is? Will [s]he think it’s a sponge or will [s]he think it’s a rock?” The appearance–reality distinction question set included the appearance question (“What does this look like? Does it look like a rock or does it look like a sponge?”) and the reality question (“What is this really? Is it really a sponge or is it really a rock?”). Also counterbalanced, within these question sets across children, was the ordering of the words “rock” and “sponge” in the second and final question in each of the three sets. Children who gave only “Other” responses (featural, identity error, none) to the pretest questions did not receive the test questions. This was because their pretest answers gave no indication that they had established a representation for the object. Six of the 30 children gave only “other” responses and hence did not receive the test questions. Of the remaining 24 children, for those who did not provide the “sponge” label in response to the pretest questions, their most label-like response (selected in descending order from among the “Label” and “Non-Label” responses) was inserted into the test questions, replacing the word “sponge” (e.g., “When you first saw this, before you touched it or squeezed it, what did you think it was: Did you think it was a rock or did you think it was a (pillow/like a rock/bed/something to wash with)?” Scoring Responses to Test Questions. For each of the three sets of test questions, each child was assigned a score of “1” if the set was answered correctly (passed) and a score of “0” if it was not. Children were scored as having passed the representational change task, if they correctly reported their initial representation of the object as a rock. Similarly, they were scored as having passed the false-belief task, if they correctly reported the false belief that another child would think the object was a rock. To pass the appearance–reality question, children had to answer both the appearance question correctly (stating that the object looks like a rock) and the reality question correctly (stating that the object really is a sponge or whatever the child had referred to it as in the pretest questions, e.g., “pillow”).
Results The mean proportion of children successful on each of the three tasks—false belief, appearance–reality distinction, representational change—was computed. These means are presented in Figure 3.1. Intertask correlations were computed. These data are presented in Table 3.3. As can be seen, no significant correlations were found across any of the three tasks. Task means were compared by paired (nonindependent) t-tests. Children did best on the false belief task. This was the only of the three tasks on which a clear
50
LUCARIELLO
Mean proportion of children successful
majority of children were successful. Nearly two thirds succeeded on this task. Performance was weakest on the representational change task. Only one third of children succeeded on this task. The mean score on the false-belief task (M = .63; SD = .49) was significantly higher than that for representational change (M = .33; SD = .48), t(23) = 2.29, p <.05. No significant difference in mean scores was found between the false-belief and appearance–reality tasks. Mean scores across the two intrapersonal reasoning tasks of appearance–reality and representational change tasks were also compared. Better performance was found on the appearance–reality task, t(23) 1.74, p < .05, one-tailed, p < .10, two-tailed. See Fig. 3.1. Use of a one-tailed test is justified here because the repre-
FIG. 3.1. Mean proportion of children successful on the three metarepresentational tasks.
TABLE 3.3 Correlations for Metarepresentational ToM Tasks 1. False Belief 2. Appearance–Reality 3. Representational Change Note. N = 24. * p < .05, ** p < .01
1
2
3
—
.32
.18
—
.30 —
3. THE FMS MODEL
51
Mean proportion of children successful
sentational change task was predicted to be more challenging than the appearance–reality task. An additional analysis assessed the role of language, and in particular, referential-intrapersonal language, in task performance. The sample of 24 children was divided into two groups based on their pretest answers. The “Label” group (N = 15) was comprised of children who provided “normative label” (N = 2) and “alternative label” (N = 13) responses. The “Non-Label” group (N = 9) was comprised of children who provided Label Misses (N = 5; 3 = label hedges; 2 = incorrect labels) and Object Use (N = 4) responses. Mean scores for these two groups on each of the three tasks were compared by independent t tests. These means are presented in Fig. 3.2. The Label group performed better than the Non-Label group on all three metarepresentational tasks. The difference between these two groups achieved significance on the two tasks assessing intrapersonal ToM metarepresentational reasoning—the appearance–reality task, t(22) = 2.71, p < .05 (two-tailed), and the representational change task, t(22) = 2.05, p < .05 (two-tailed). The two
FIG. 3.2. Mean proportion of children successful, divided into Label and Nonlabel groups, on the three metarepresentational tasks.
52
LUCARIELLO
groups did not differ significantly on the social ToM metarepresentation task of false belief, t(22) = 1.36.
DISCUSSION The FMS model views ToM in terms of its functions that serve as the basis for distinct kinds. Reading the mental and emotional states of others can be referred to as social ToM. Social ToM is used for social interaction. One’s own mental states are entailed in reflection and learning. When used for these functions, ToM is intrapersonal and differentiates into reflection and constructivist forms, respectively. Accordingly, on the FMS model, ToM is a nonintegrated, differentiated cognitive behavior. These proposals about ToM were studied by reliance on a second hypothesis of the FMS model. This is the claim that social and intrapersonal ToMs each have a distinctive basis in sociocultural experience. In the present research, low-income children served as participants. Their socialization was proposed to foster social ToM development and functioning, while being less facilitative of intrapersonal ToM development and functioning. The data supported all these proposals about ToM. Children’s performance across own- and other-belief tasks was not correlated. These data support the idea of a differentiated ToM. Children’s greatest success was in the social ToM task of false belief. Morover, they performed better on this task than on the intrapersonal learning/constructivist ToM task of representational change, on which they performed poorly. Hence, a pattern of uneven development across reasoning about own and others’ representations was found. These children were stronger in social than intrapersonal metarepresentation. Moreover, their vulnerability with respect to intrapersonal metarepresentation was proposed as possibly affecting their performance across the two intrapersonal metarepresentation tasks. Representational change entails learning, hence constructivist metarepresentation. Distinguishing appearance from reality does not. Accordingly, it would be easier for children to engage in intrapersonal metarepresentational reasoning in the latter “reflection” task. The findings show this. These children were more successful on the appearance–reality than on the representational change task. The present data showing representational change to be more difficult than distinguishing appearance and reality are consistent with Gopnik and Astington (1988). The hypothesis on the distinctive sociocultural origins of social and intrapersonal metarepresentation was additionally examined by assessing the role of referential-intrapersonal language in task performance. Children who could provide Label responses in the pretest have access to the referential-intrapersonal language model in this experimental task. These children performed significantly better on the two intrapersonal metarepresentational tasks of appearance–reality
3. THE FMS MODEL
53
and representational change than children who could not provide label responses. These data support the proposal that the referential-intrapersonal model of language socialization is facilitative of the development and functioning of intrapersonal ToM. The uneven ToM developmental pattern found here of greater strength in social, than intrapersonal, ToM undermines the claims about ToM made by all the major, current ToM accounts. The FMS model is the only account of ToM that affords and explains this pattern of uneven development in ToM reasoning. Its claim that distinct ToMs have a distinct sociocultural basis paves the way for this finding. In contrast, Harris’ (1992) simulation theory predicts uneven development in the opposite direction, with greater strength in intrapersonal ToM. Moreover, uneven development in any direction would not be predicted or accounted for on the modularity view (German & Leslie, 2000, 2001; Leslie & Polizzi, 1998; Leslie & Thaiss, 1992; Scholl & Leslie, 1999, 2001), the theory-theory view (Gopnik, 1993; Gopnik & Wellman, 1994; Perner, 1991; Wellman & Cross, 2001), and the sociocultural view (Dunn et al., 1991; Hughes & Dunn, 1998; Lewis et al., 1996; Perner et al., 1994; Ruffman et al., 1998; Vinden, 1996; Vinden & Astington, 2000). On all these accounts, ToM is an integrated phenomenon, with reasoning about own and others’ representations emerging together and at the same ontogenetic point. The integrated view rests, in part, on the considerable data showing middle-income children to achieve blanket success on these three metarepresentational tasks at roughly the same age of 4 to 5 years (e.g., Astington et al., 1988). The proposal here is that middle-income children’s success across these tasks is not on the same basis. These children presumably succeed on the false-belief task due to their experiences in the very basic, ubiquitous social interactions, said here to underlie an early form of social ToM, metarepresentational reasoning. Success on the own-belief tasks of appearance–reality and representational change, however, is presumably based on their socialization experiences that foster attention to own mental states. Also supporting the integrated view of ToM are data showing performance across own- and other-belief tasks to be correlated, as found in the Wellman et al. (2001) meta-analysis. As noted, however, socioeconomic status was not a variable in that meta-analysis. The present research with low-income children fails to find correlations across own- and other-belief tasks. These data add to those from many studies conducted since that meta-analysis showing lack of correlation across own and other reasoning tasks (Cutting & Dunn, 1999; Meins et al., 2002; Moore et al., 1998; Ruffman et al., 1998) and emotion tasks (Hughes & Dunn, 1998). In summary, the FMS model’s view of metarepresentational behavior as nonintegrated and differentiated was supported by the present findings. Distinctions in ToM across social and intrapersonal reasoning, and within the latter across reflection and constructivist kinds, were seen on tasks assessing each of these
54
LUCARIELLO
forms of reasoning. Future research might explore the hypotheses of the FMS model on the development of ToM. Development is proposed to occur along a multilinear path that corresponds to ToM’s variegated nature. This exploration could take two directions. One could examine whether this differentiation continues, as hypothesized here, along the subsequent developmental trajectory of perspectival and dialectical ToMs. Moreover, one would expect and could examine the possible earlier emergence of this differentiation in reasoning about own and others’ mental states prior to the advent of metarepresentational reasoning.
ACKNOWLEDGMENTS I would like to thank Mary LeDonne for her work on data collection, Catherine Mindolovich for work on data analysis, and Judith Hudson for her insightful comments on an earlier version of this chapter. Requests for reprints should be addressed to: Joan Lucariello, Boston College, Campion 239E, Chestnut Hill, MA 02467.
REFERENCES Astington, J. W., Harris, P. L., & Olson, D. R. (Eds.) (1988). Developing theories of mind. Cambridge, England: Cambridge University Press. Bauer, P. J., & Mandler, J. M. (1990). Remembering what happened next: Very young children’s recall of event sequences. In R. Fivush & J. A. Hudson (Eds.), Knowing and remembering in young children. Emory symposia in cognition Vol. 3, (pp. 9–29). New York: Cambridge University Press. Bauer, P. J., & Mandler, J. M. (1992). Putting the horse before the cart: The use of temporal order in recall of events by one-year-old children. Developmental Psychology, 28, 441–452. Bauer, P. J., Wenner, J. A., Dropik, P. L., & Wewerka, S. S. (2000). Parameters of remembering and forgetting in the transition from infancy to early childhood. Monographs of the Society for Research in Child Development, 65 (4, Serial No. 263). Blake, I. K. (1993). The social–emotional orientation of mother–child communication in African American families. International Journal of Behavioral Development, 16, 443–463. Blake, I. K. (1994). Language development and socialization in young African-American children. In P. M. Greenfield & R. R. Cocking (Eds.), Cross-cultural roots of minority child development (pp.167–195). Hillsdale, NJ: Lawrence Erlbaum Associates. Bloom, L. (1993). The transition from infancy to language: Acquiring the power of expression. Cambridge, MA: Cambridge University Press. Brice Heath, S. (1983). Ways with words. Cambridge, MA: Cambridge University Press. Brice Heath, S. (1989). Oral and literate traditions among Black Americans living in poverty. American Psychologist, 44, 367–373. Carpendale, J. I., & Chandler, M. J. (1996). On the distinction between false belief understanding and subscribing to an interpretive theory of mind. Child Development, 67, 1686–1706.
3. THE FMS MODEL
55
Chandler, M. J. (1987). The Othello effect: Essay on the emergence of skeptical doubt. Human Development, 30, 137–159. Chandler, M. J., & Lalonde, C. (1996). Shifting to an interpretive theory of mind: 5- to 7-year-olds’ changing conceptions of mental life. In A. Sameroff & M. Haith (Eds.), The five to seven year shift: The age of reason and responsibility. The John D. and Catherine T. MacArthur Foundation series on mental health and development (pp. 111–139). Chicago, IL: University of Chicago Press. Cutting, A. L., & Dunn, J. (1999). Theory of mind, emotion understanding, language, and family background: Individual differences and interrelations. Child Development, 70, 853–865. Delgado-Gaitan, C. (1994). Socializing young children in Mexican-American families: An intergenerational perspective. In P. M. Greenfield & R. R. Cocking (Eds.) Cross-cultural roots of minority child development (pp. 55–86). Hillsdale, NJ: Lawrence Erlbaum Associates. Dunn, J., Brown, J., Slomkowski, C., Tesla, C., & Youngblade, L. (1991). Young children’s understanding of other people’s feelings and beliefs: Individual differences and their antecedents. Child Development, 62, 1352–1366. Flavell, J.H., Flavell, E.R., & Green, F.L. (1983). Development of the appearance–reality distinction. Cognitive Psychology, 15, 95–120. Fu-Xi, F., & Keats, D. M. (1989). “The Master and the Wolf”: A study in the development of social perspective taking in Chinese and Australian children. In D. M. Keats, D. Munro, & L. Mann (Eds.), Heterogeneity in cross-cultural psychology (pp. 419–425).Amsterdam: Swets & Zeitlinger B.V. Geertz, C. (1975). On the nature of anthropological understanding. American Scientist, 63, 47–53. Gelman, R., & Lucariello, J. (2002). Role of learning in cognitive development. In H. Pashler (Series Ed.) & C. R. Gallistel (Vol. Ed.), Stevens’ handbook of experimental psychology:Vol. 3. Learning, motivation, and emotion (3rd ed., pp. 395–443). New York: Wiley. German, T., & Leslie, A. (2000). Attending to and learning about mental states. In P. Mitchell & K. Riggs (Eds.), Children’s reasoning and the mind (pp. 229–252). Hove, England: Psychology Press. German, T. P., & Leslie, A. M. (2001). Children’s inferences from knowing to pretending and believing. British Journal of Developmental Psychology, 19, 59–83. Goldenberg, C., & Gallimore, R. (1995). Immigrant Latino parents’ values and beliefs about their children’s education: Continuities and discontinuities across cultures and generations. In M. L. Maehr & P. R. Pintrich (Eds.), Advances in motivation and achievement: Culture, motivation and achievement (Vol. 9, pp. 183–228). Greenwich, Connecticut: Jai Press. Gopnik, A. (1993). How we know our minds: The illusions of first person knowledge of intentionality. Behavioral and Brain Sciences, 16, 1–14. Gopnik, A., & Astington, J.W. (1988). Children’s understanding of representational change and its relation to the understanding of false belief and the appearance–reality distinction.Child Development, 59, 26–37. Gopnik, A., & Wellman, H. M. (1994). The theory theory. In L. Hirschfeld & S. Gelman (Eds.), Domain specificity in cognition and culture (pp. 257–295). New York: Cambridge University Press.
56
LUCARIELLO
Greenfield, P. M. (1994). Independence and interdependence as developmental scripts:Implications for theory, research, and practice. In P. M. Greenfield & R. R. Cocking (Eds.), Cross-cultural roots of minority child development (pp.1–40). Hillsdale, NJ: Lawrence Erlbaum Associates. Greenfield, P.M., & Bruner, J.S. (1966). Culture and cognitive growth. International Journal of Psychology, 1, 89–107. Halliday, M. A. K. (1975). Learning how to mean: Explorations in the development of language. London: Edward Arnold. Harris, P. L. (1992). From simulation to folk psychology: The case for development. Mind and Language, 7, 120–144. Hart, B., & Risley, T. R. (1995). Meaningful differences in the everyday experience of young American children. Baltimore: Paul H. Brookes. Hess, R. D., & Shipman, V.C. (1965). Early experience and the socialization of cognitive modes in children. Child Development, 36, 869–886. Hughes, C., & Dunn, J. (1998). Understanding mind and emotion: Longitudinal associations with mental-state talk between young friends. Developmental Psychology, 34, 1026–1037. Lawrence, V. W., & Shipley, E. F. (1996). Parental speech to middle- and working-class children from two racial groups in three settings. Applied Psycholinguistics, 17, 233–255. Leslie, A. M., & Polizzi, P. (1998). Inhibitory processing in the false belief task: Two conjectures. Developmental Science, 1, 247–254. Leslie, A. M., & Thaiss, L. (1992). Domain specificity in conceptual development: Neuropsychological evidence from autism. Cognition, 43, 225–251. Lewis, C., Freeman, N. H., Kyriakidou, C., Maridaki-Kassotaki, K., & Berridge, D. M. (1996).Social influences on false belief access: Specific sibling influences or general apprenticeship? Child Development, 67, 2930–2947. Lucariello, J. (1994). Situational irony: A concept of events gone awry. Journal of Experimental Psychology: General, 123, 129–145. Lucariello, J., & Mindolovich, C. (1995). The development of complex metarepresentational reasoning: The case of situational irony. Cognitive Development, 10, 551–576. Lucariello, J., & Mindolovich, C. (2002). The best laid plans: Beyond scripts are counterscripts. Journal of Cognition and Development, 3, 91–115. Lucariello, J., Mindolovich, C., & LeDonne, M. (2004). An eye for trouble: Cognitive complexities of acquiring counterscripts for situationally ironic events. Manuscript submitted for publication. Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Review, 98, 224–253. Meins, E., Fernyhough, C., Wainwright, R., Gupta, M., Fradley, E., & Tuckey, M. (2002). Maternal mind-mindedness and attachment security as predictors of theory of mind understanding. Child Development, 73, 1715–1726,. Miller, P. (1982). Wendy, Amy, Beth: Learning language in South Baltimore. Austin: University of Texas Press. Miller, P. (1986). Teasing as language socialization and verbal play in a white working-class community. In B. B. Schieffelin & E. Ochs (Eds.), Language socialization across cultures. Studies in the social and cultural foundations of language, No. 3 (pp. 199–212). New York: Cambridge University Press.
3. THE FMS MODEL
57
Miller, P. (1987). The socialization of anger and aggression. Merrill-Palmer Quarterly, 33, 1–31. Miller, P. (1988). Early talk about the past: The origins of conversational stories of personal experience. Journal of Child Language, 15, 293–315. Moore, C., Barresi, J., & Thompson, C. (1998). The cognitive basis of future-oriented prosocial behavior. Social Development, 7, 198–218. Nelson, K. (1981). Individual differences in language development: Implications for development and language. Developmental Psychology, 17, 170–187. Nelson, K. (Ed.). (1986). Event knowledge: Structure and function in development. Hillsdale, NJ: Lawrence Erlbaum Associates. Nelson, K., & Gruendel, J. (1981). Generalized event representations: Basic building blocks of cognitive development. In M. E. Lamb & A. L. Brown (Eds), Advances in developmental psychology (Vol. 1, pp. 131–158). Hillsdale, NJ: Lawrence Erlbaum Associates. Perner, J. (1991). Understanding the representational mind. Cambridge, MA: MIT Press. Perner, J., Ruffman, T., & Leekam, S. R. (1994). Theory of mind is contagious: You catch it from your sibs. Child Development, 65, 1228–1238. Piaget, J. (1952). The origins of intelligence in children. New York: International University Press. Piaget, J. (1954). The construction of reality in the child. New York: Basic Books. Ruffman, T., Perner, J., Naito, M., Parkin, L., & Clements, W. A. (1998). Older (but not younger) siblings facilitate false belief understanding. Developmental Psychology, 34, 161–174. Scholl, B. J., & Leslie, A. M. (1999). Modularity, development, and “theory of mind.” Mind and Language, 14, 131–153. Scholl, B.J., & Leslie, A. M. (2001). Minds, modules, and meta-analysis. Child Development,72, 696–701. Snow, C. E. (1986). Conversations with children. In P. Fletcher & M. Garman (Eds.), Language acquisition (pp. 69–89). Cambridge, England: Cambridge University Press. Snow, C. E. (1991). The theoretical basis for relationships between language and literacy in development. Journal of Research in Childhood Education, 6, 5–10. Vinden, P. G. (1996). Junin Quechua children’s understanding of mind. Child Development, 67, 1707–1716. Vinden, P. G., & Astington, J. W. (2000). Culture and understanding other minds. In S. Baron-Cohen, H. Tager-Flusberg, & D. J. Cohen (Eds.), Understanding other minds: Perspectives from developmental cognitive neuroscience (pp. 503–519). Oxford, England: Oxford University Press. Weizman, Z. O., & Snow, C. E. (2001). Lexical input as related to children’s vocabulary acquisition: Effects of sophisticated exposure and support for meaning. Developmental Psychology, 37, 265–279. Wellman, H. M., & Cross, D. (2001). Theory of mind and conceptual change: Reply. Child Development, 72, 702–707. Wellman, H. M., Cross, D., & Watson, J. (2001). Meta-analysis of theory-of-mind development: The truth about false belief. Child Development, 72, 655–684.
4 Meaning and Use: Children’s Acquisition of the Mental Lexicon Janet Wilde Astington and Joan Peskin University of Toronto
The little girl had the makings of a poet in her who, being told to be sure of her meaning before she spoke, said: “How can I know what I think till I see what I say?” —Wallas, 1927
Katherine Nelson would no doubt feel sympathy for the little girl whom Wallas quoted—she has the makings of a poet, perhaps, but a psychologist too. Nelson (1996) showed how children’s developing understanding of the world depends on their developing language abilities. In her terms, early pragmatic language— “use before meaning”— leads to the acquisition of “meaning from use” (Nelson, 1996, p. 145). The little girl, it might be said, will know what she thinks as she sees what she says and sees how others react and reply to what she says. Children’s conceptual development is crucially dependent on language. As children learn the words their culture uses, they discover things that would not exist for them without language. Importantly, in Western culture, they discover the mind (Astington, 1993) or, in the more widely used terminology, they acquire a 59
60
ASTINGTON AND PESKIN
theory of mind (Astington, P. L. Harris, & Olson, 1988). During her long and prestigious career, Katherine Nelson influenced many researchers in many different ways, as the chapters in this volume testify. Work on children’s theory of mind has been richly informed by her ideas—perhaps ironically, because she denied that children develop any theory about the mind. Indeed, some researchers are less decided on this question now, thanks to her influence (Astington, 1996). This chapter discusses the importance of language development to theory-of-mind development, in particular, the role played by children’s acquisition of terms that refer to the mind, that is, the mental lexicon. This acquisition is fundamental to children’s understanding of the mind: “In Western cultures, the acquisition of an understanding of beliefs, desires, and intentions is completely tied up with the acquisition of a mentalistic language for talking about these events” (Olson, 1994, p. 251). However, what is most needed, but not yet agreed on, is a clear explanation of how children acquire these terms. This is where Nelson’s theory of experiential semantics—the acquisition of meaning from use—makes a fundamental contribution. There are two quite different perspectives on how children acquire the mental lexicon (Montgomery, 2002). On the one hand, some researchers assume that children are working out word-referent relations. For example, in Bartsch and Wellman’s (1995) view, children’s early language production reflects their theory of mind: Desire and belief are the fundamental concepts in the theory, that is, they are the referents, and the child’s task is to map the word onto the referent. On the other hand, other researchers assume that what children are doing is learning how to use words in context. According to Montgomery (2002), “The word learning task the child faces is not finding a referent, but rather learning the appropriate role a mental verb plays in characteristic situations or events” (p. 365). Nelson (e.g., 1996) is a leading proponent of this latter, contextual viewpoint, as is discussed later. However, description must precede explanation (Carey, 1990). That is, one needs to describe what occurs, before attempting to explain how children acquire the mental lexicon and how this acquisition relates to their developing understanding of mind. First, the chapter discusses the development of children’s theory-of-mind research and Nelson’s view of it. Second, it discusses children’s acquisition of the mental lexicon. Third, it presents theoretical explanations of children’s acquisition of the mental lexicon, including Nelson’s theory of experiential semantics. Fourth, it includes a brief review of a recent study that attempted to help children acquire the mental lexicon (Peskin & Astington, 2002). Finally, it considers what light these findings throw on the issue of theoretical explanations of children’s acquisition of the mental lexicon.
CHILDREN’S THEORY OF MIND Children’s theory of mind has been a lively research topic in development psychology for the past two decades. Premack and Woodruff’s (1978) study, often
4. CHILDREN’S ACQUISITION OF THE MENTAL LEXICON
61
cited as the starting point for investigation in this area, provided the definition developmentalists first used: “In saying that an individual has a theory of mind, we mean that the individual imputes mental states to himself and to others. A system of inferences of this kind is properly viewed as a theory, first, because such states are not directly observable, and second, because the system can be used to make predictions, specifically about the behavior of other organisms” (Premack & Woodruff, 1978, p. 515). That is, children’s theory of mind underlies their ability to interpret other people’s behavior by the attribution of mental states. How is it possible to determine whether a child is actually attributing mental states in predicting another’s behavior? Wimmer and Perner’s (1983) study, another often-cited starting point for children’s theory of mind research, provided the answer in the form of the now famous “false-belief” task. The task arranges that the child and another person have different information about a situation and then asks the child to predict what the other will do in that situation. If the child predicts that the person will act on the basis of his or her knowledge of the situation, not on the basis of the child’s own knowledge, it can be inferred that the child has a theory of mind in Premack and Woodruff’s (1978) sense. This task has been replicated, refined, and extended over the years, yielding clear evidence that children acquire an understanding of false belief sometime toward the end of the preschool years (Wellman, Cross, & Watson, 2001). However, early on, Dunn (1988) pointed to a problem that exists if success on the false-belief task is to be taken as the marker of children’s understanding of people’s minds. Her detailed observations of family interactions showed that even young toddlers seem to have a good understanding of other people, well before they can pass the experimental false-belief tasks. In part, this discrepancy might be resolved by pointing out the difference between experimental and observational situations, and the fact that the experimental tasks focus on belief, whereas in naturalistic situations desires, motives, and emotions are the salient mental states (Astington, 2003). However, this is not the whole solution. It is here that Katherine Nelson’s work makes an enormous contribution. It is important to remember that the child is not just coming to understand the social world from the outside, as it were, but right from birth is a participant within the social world. Throughout her career, Nelson has emphasized the importance of children’s own participation, their experiences in the world, as fundamental to their conceptual development. In her initial work on early word learning, for example, Nelson (1973, 1974) showed that to a large extent meaning is dependent on function, that is, what children can do with objects and how they experience them. Likewise, as children acquire taxonomic knowledge, their categories are at first experientially based, that is, objects in the same category are those that can substitute for one another in an event (Nelson, 1988). Indeed, an important part of Nelson’s work is devoted to showing how experientially based event knowledge is the basis of early cognitive development (Nelson, 1986). And so it is with theory of mind, in Nelson’s view. That is, the development of the ability to interpret behavior mentalistically derives from children’s experience in the social world, not from
62
ASTINGTON AND PESKIN
the child’s construction of a theory of mind (Nelson, Plesa, & Henseler, 1998). It comes from making sense of social interaction as children acquire the language and mores of their particular culture. The emerging ability to use language is an essential part of this development. On this view, it is by virtue of being a linguistic creature and growing up within a certain culture that children acquire their understanding of mind (Astington, 1996). This can be construed as a Whorfian argument, although it is not intended in the radically relativist way often associated with Whorf (1956). Our argument, like that of Nelson (1996), is that as children acquire language they learn to think in cultural ways. This is a Vygotskian perspective that sees children internalizing their culture’s construal of mind through linguistic social interaction (Astington, 1996). On this view, the theory of mind, if it exists at all, is in the folk ways and speech practices of the culture—the child would never come up with the theory except by participating in dialogue with more knowledgeable members of the culture. Furthermore, it is within this dialogue that the mental lexicon plays a vital role.
CHILDREN’S ACQUISITION OF THE MENTAL LEXICON Although Nelson’s theory of experiential semantics applies to the acquisition of all words, the focus in this chapter is on the subset of terms used in making mentalistic interpretations of behavior, that is, the mental lexicon. This is the set of semantic terms that refer to mental states of belief, desire, intention, emotion, and perception. This definition does not include terms for all internal states, such as physiological states (e.g., hungry, tired), and it does not include moral terms (e.g., good, naughty), although some inventories of children’s mentalistic language do include all such words (e.g., Bretherton & Beeghly, 1982; Bretherton, McNew, & Beeghly-Smith, 1981). Indeed, there are many different classification schemes used in the literature (for discussion see Astington & Pelletier, 1996). An important focus in the literature, as well as here, is on cognitive, sometimes called metacognitive, terms. Most of the words in this category that children use and researchers study are verbs (e.g., think, know, guess, remember), but some are nouns (e.g., surprise, guess) and a few are adjectives and adverbs (e.g., sure, thoughtfully). It is not uncommon in the child language literature for the designation “mental” or “mental state” to be used to refer only to the set of cognitive terms, in contrast to desire and feeling state terms (e.g., Brown & Dunn, 1991; Shatz, Wellman, & Silber, 1983). However, in this chapter, “mental terms” is used to designate all mental state terms, and “metacognitive terms” designates the subset of mental terms that refers to belief states. Because verbs, rather than other parts of speech, are the most widely used mental terms, most of the research is focused on verbs. Mental verbs frequently occur as the main verb in a complex sentence that has a subordinate clause as its gram-
4. CHILDREN’S ACQUISITION OF THE MENTAL LEXICON
63
matical object; for example, “Maxi thinks that the chocolate is in the cupboard,” “He wants to have some.” Such constructions allow for a separation of the attitude and content of a person’s mental state: The attitude denotes what kind of mental state it is (belief, desire, etc.) and the content describes what the mental state is about. The construction consists of a main sentence with another sentence embedded in it. In the first example, the embedded sentence “the chocolate is in the cupboard” forms a subordinate clause that acts as the grammatical object of the verb think in the main sentence. Such clauses are known as object complements. They allow for the report of false beliefs; that is, the whole construction can be true even though the embedded sentence is false. Interest in children’s use and understanding of mental terms began in the 1970s, in the context of research on the production and comprehension of complex syntactic constructions, including the object complements just discussed (R. J. Harris, 1975; Limber, 1973). Subsequently, the acquisition of complementation became a particular focus of investigation (e.g., Bloom, Rispola, Gartner, & Hafitz, 1989). Bloom et al. studied four children’s speech production longitudinally from age 2 to 3. They showed that children’s early complement constructions involve a small set of main verbs: primarily the perception verbs see and look and the metacognitive verbs think and know. Later, J. G. de Villiers and P. A. de Villiers (2000) focused on the acquisition of complementation in relation to theory-of-mind development, in particular the development of false-belief understanding, arguing that acquisition of the syntax of complementation precedes and promotes false-belief understanding by providing the appropriate format for representing false beliefs. Recently, however, findings from training studies (Lohmann & Tomasello, 2003) suggest that the syntax of complementation is not of sole importance in the development of false-belief understanding. More general linguistic factors, such as discourse about alternative viewpoints, also contribute to children’s developing understanding of false beliefs. However, in many languages, including English and German (the language in which the Lohmann and Tomasello study was conducted), mental verbs with object complements provide the typical way of talking about alternative viewpoints, and indeed, these two factors combined had the greatest effect in the training study. Thus, acquiring syntactic and semantic understanding of mental verbs is a crucial component in developing a theory of mind. Studies of the development of semantic understanding comprise an important part of the early investigation of children’s acquisition of the mental lexicon. Studies conducted in the later 1970s and early 1980s focused on a small set of verbs (i.e., know, think, guess, remember, forget), investigating children’s ability to demonstrate their understanding by using these related terms appropriately. In general, the studies involved tasks in which characters (or the children themselves) did or did not have information about the location of a hidden object and did or did not succeed in finding it. Such object-locating tasks are interestingly similar to the Wimmer and Perner (1983) false-belief task and, indeed, a
64
ASTINGTON AND PESKIN
false-belief task occurs as a pretest in an early investigation of children’s comprehension of the mental verbs think and know (Johnson & Maratsos, 1977). In the main task, 4-year-olds, but not 3-year-olds, could distinguish between individuals’ knowing, if they had evidence, and thinking, if they did not have evidence, that a hidden object was under a certain box (Johnson & Maratsos, 1977). Similarly, studies on children’s understanding of mental terms marking certainty showed that 4-year-olds, but not 3-year-olds, can distinguish between be sure and think and can appropriately differentiate know from both think and guess (Moore, Bryant, & Furrow, 1989; Moore & Davidge, 1989). Other studies showed that before age 6 children could not distinguish between knowing and correctly guessing a hidden object’s location; that is, they would say that they had known if they chose correctly, even though they had no evidence prior to choosing (Miscione, Marvin, O’Brien, & Greenberg, 1978). Similarly, 4-year-olds claimed that story characters “remembered” if the characters found the hidden object, and “forgot” if they did not find it, regardless of whether the characters had previous knowledge of its location (Wellman & Johnson, 1979). Johnson and Wellman (1980) showed that by age 7 children could distinguish appropriately between remember, know, and guess by taking account of several aspects of the situation, such as prior information and successful search, but younger children focused on single salient aspects, sometimes leading to inappropriate uses, as Miscione et al. (1978) reported for the know/guess contrast. Thus, coming to understand relations in the semantic field of metacognitive terms is a slow process, which begins around age 4 but is not complete for some years, even for the common terms know, think, guess, remember, and forget. Mastering less common metacognitive terms (e.g., infer, hypothesize, etc.) is a task that continues throughout the school years (Astington & Olson, 1990). In marked contrast, studies of children’s production of metacognitive terms show that this begins well before age 4. Bretherton and her colleagues (Bretherton & Beeghly, 1982; Bretherton et al., 1981) were probably the first developmentalists to apply Premack and Woodruff’s (1978) definition of theory of mind to the human species, claiming that older infants’ ability to engage in intentional communication implies that they have a theory of mind, albeit an “implicit and fairly rudimentary” one, which soon becomes an “explicit, verbally expressible theory of mind” by age 2 or 3, when children begin to use mentalistic terms (Bretherton et al., 1981, pp. 340, 356). These they construed broadly as “internal state” terms, divided into six categories: perception, cognition, volition, morality, emotion, and physiology. Their category of “cognition” terms includes verbs referred to here as metacognitive. Bretherton and Beeghly’s data come from mothers’ reports of their 28-month-old children’s language use. Metacognitive terms were produced less frequently by these children than terms in the other five categories. However, even at this young age know was used by 66% of the sample, and think, remember, forget, pretend, and dream were used by 30%. The beginning of this section identified
4. CHILDREN’S ACQUISITION OF THE MENTAL LEXICON
65
know and think as the most common metacognitive terms used before age 3 (Bloom et al., 1989). Bretherton and Beeghly (1982) included all uses in their count of terms—the high percentage of children using know was due to the frequency of “I don’t know” utterances. In a study published around the same time, Shatz et al. (1983) argued that idiomatic and conversational uses of terms should not be counted, but only uses that make clear reference to mental states. Another, more thorough study (Bartsch & Wellman, 1995) similarly used the Shatz et al. coding scheme. But, in that study, only 40% of the metacognitive terms children used were coded as referring to mental states (Bartsch & Wellman, 1995, p. 43). However, as becomes evident later in the chapter, if the interest is in the acquisition of meaning from use, then all uses are potentially of interest and importance. That is the issue taken up in the next section.
THEORETICAL EXPLANATIONS OF CHILDREN’S ACQUISITION OF THE MENTAL LEXICON The preceding section provides a descriptive overview of research on children’s production and comprehension of mental terms. However, such an inventory does not really address the question of what children are doing in acquiring such terms. As mentioned in the introduction, there are at least two different perspectives on this issue (Montgomery, 2002). According to one, children are working out word-referent relations, mapping mental terms onto mental concepts (Bartsch & Wellman, 1995). The other perspective sees children as learning how to use words in context, determining the role that mental terms play in various circumstances. This latter is Nelson’s (e.g., 1996) view, as discussed later. These two different descriptions of what the child is doing (i.e., learning to label referents, or learning to use words in context) provide the beginning of an explanation, or at least point toward what it is that needs to be explained. A major problem for proponents of the word-referent mapping view is that referents of mental terms are abstract. Even though children may use the same strategies as they use in acquiring terms for concrete referents, the fact that the referents are abstract poses an added difficulty. Although his particular concern is with the acquisition of emotion words, Beckwith (1991) made a general proposal for how children might learn terms with abstract reference. The theoretical framework is that of nominalist bootstrapping. The idea is that the acquisition of an abstract term is assisted by linking the term to some perceptible entity or event. That is, the meaning of abstract terms follows on from some meaning formed on the basis of perceptible experience. This explains why metacognitive terms are difficult to acquire: “Nominalist bootstrapping would account for the late appearance of cognitive terms by claiming that children must build paradigms of the appropriate external correlates” (Beckwith, 1991, p. 81).
66
ASTINGTON AND PESKIN
Some support for Beckwith’s view is found in studies of children’s acquisition of the meaning of terms such as know, guess, lie, and promise. As described in the previous section, at first children use the terms know and guess dependent on the state of affairs in the world. If they guess correctly, that is, what they say accords with the way they perceive the world to be, then they say that they knew, even though they had no evidence prior to guessing (Johnson & Wellman, 1980; Miscione et al., 1978). Similarly, they will call mistakes “lies,” even though they are aware that the speaker is not to blame (Wimmer, Gruber, & Perner, 1984). That is, their use of the term lie is determined by the mismatch between what people say and the perceptible state of affairs in the world, not what the individuals know or what their intentions are. Further, they will say that individuals promised if the promised action is performed, but will say they did not promise if the promise is broken (Astington, 1988). That is, their use of the term promise, like lie, is based on a mismatch between what was said and the state of affairs in the world. However, these data, although they show that children may at first focus on tangible correlates of verbs referring to abstract entities, do not speak against the view that children are learning the appropriate role that such verbs play in characteristic situations or events (Montgomery, 2002; Nelson, 1996). Children will often hear the terms used in this way, that is, on the basis of the way things are in the world. For example, how often do people say, “I knew that would happen”—even though they could not have known until it did? Likewise, people say, “You didn’t guess” when someone does guess, but incorrectly. Furthermore, and more important, not all early uses are based on perceptible real-world events. Johnson (1982) pointed out that because preschool children focus on single salient aspects of a situation, such as successfully finding a hidden object even though they had no prior information about its location, they may use an inappropriate term, such as saying they knew where the object was (Johnson & Wellman, 1980). However, if prior information is made salient and then the object is surreptitiously moved from where they saw it so that they do not find it, they will appropriately say that they knew (not guessed) the location even though their search was unsuccessful (Johnson & Wellman, 1980). That is, their use of terms is based on salient features that are not always perceptible features. Other data support the viewpoint that children are learning the appropriate roles that verbs play in situations and events. For example, 40% of 11-year-olds chose the verb hypothesize as the correct term to use in a science context, even though the particular event made it obvious that remember was the more appropriate term (Astington & Olson, 1990). Presumably, their choice was based on the fact that they recognized hypothesize as a “science” word, even though they did not know the precise meaning of the term. That is, for them meaning is dependent on context. Katherine Nelson is the leading proponent of this contextual view of semantic development (Montgomery, 2002).
4. CHILDREN’S ACQUISITION OF THE MENTAL LEXICON
67
Nelson’s Theory of Experiential Semantics: Acquiring Meaning from Use A great strength of Nelson’s theory is that it addresses the issue of how children learn the meaning of words for unobservable abstract entities, such as belief and desire. Much research on early word learning focuses on the acquisition of names for objects and, to a lesser extent, action verbs. However, even at the start of language learning, children acquire terms for a variety of less tangible concepts (Nelson, Hampson, & Kessler Shaw, 1993). Explaining how children acquire the meaning of such terms is more challenging but has a more useful, broader application. It is certainly what is required in order to explain how children acquire the mental lexicon. Building on her earlier work (Nelson, 1986, 1996) showing that experientially acquired event knowledge is the base of children’s early conceptual development, Nelson argued that language plays an integral role in further conceptual development. “Use of language in and of itself conveys many of the values, categories, and mythologies of the culture” (Nelson, 1996, p. 312). That is, children’s concepts are determined by the categories and the relations among the categories lexicalized in their natural language. Acquiring the meaning of words that refer to unobservable abstract entities like mental states is essentially constructing concepts of them. Nelson argued that young children do not possess such concepts before they have words for them. Hearing mental states referred to by others provides the impetus for concept formation. Nelson (1996) emphasized, “Importantly, the concepts referenced by the mental state words may not exist as such for the young child until they are constructed in response to reflections on the use of the terms by others to refer to oneself” (p. 309). Language both structures the concepts and provides the medium in which such structures can be conveyed. Nelson took the Wittgenstein slogan “meaning is use” and argued that meaning must be extracted from use (Levy & Nelson, 1994). At first, the words children use are tightly tied to particular discourse contexts. Children produce these words in specific contexts, modeled on adult usage, before they have any real meaning for the terms, that is, before they have the concept to which the adult term refers. This is what Nelson called “use before meaning.” How then does meaning arise from use? First, adults use terms in contexts relevant to the child’s own interests, and the child uses the same form in the same contexts. This early use is entirely pragmatic, a sort of meaningless repetition, which nonetheless establishes the term in the child’s productive vocabulary. This leads to the child’s noticing the term when adults use it in other contexts and allows the child to compare the ways in which self and others use the term. By this stage, the child has competent productive use of the term but still may fail comprehension tests designed to probe for an understanding of the precise meaning of the term. For example, the previous section on children’s acquisition of the mental lexicon referred to studies that showed chil-
68
ASTINGTON AND PESKIN
dren do not always make precise distinctions between the terms know, guess, remember, and forget until the early school years (e.g., Johnson & Wellman, 1980; Miscione et al., 1978; Wellman & Johnson, 1979), although these terms are part of their productive vocabulary in the preschool years. Such preschool usage is generally appropriate because it is modeled on adult usage. Geach (1957) claimed that the appropriate use of a linguistic term is a sufficient, if not a necessary, condition for possessing an understanding of the corresponding concept, but Nelson would presumably argue against this view. Dunn (1991), discussing very young children’s use of mental terms such as mean to and on purpose, said that these phrases are always used appropriately in her extensive data sets of natural language use. This is what might be expected, if children are using terms in the same contexts and ways that they have heard them used. Inappropriate uses are rare but do occur. For example, Shultz (1980, p. 157) reported the charming story of a 3-year-old, spooning cereal into her baby brother, who then simply dumped a spoonful on the baby’s head and said to her mother, “I didn’t do it on purpose.” For another example, a 3-year-old greeted her grandfather with: “We’re having a surprise party for your birthday! And it’s a secret!” (Astington, 1993, p. 98). Such examples are striking because they provide such clear evidence of use without meaning. Nonetheless, it is true that most early uses fit the situation, which is what would be expected if children are modeling their production on adult usage. The question is how the child progresses from this stage to that of having full understanding of the relational and contrastive meanings of a term, which is what is required for correct performance on comprehension tests. Nelson distinguished three different levels of use and understanding (Levy & Nelson, 1994; Nelson & Kessler Shaw, 2002; Nelson & Lucariello, 1985). At first, terms are used pragmatically in the same situations and discourse contexts as those in which the child has heard others use the term. For mental terms, Nelson called this level “embedded in context.” The second level is “denotation,” which implies “conceptual knowledge of the single term but without any necessary connection to related conceptual terms” (Nelson & Kessler Shaw, 2002, p. 47). The third level is “sense,” in which terms are related to one another within a lexical-semantic system. At the second level, terms can be used symbolically but they are not organized into a system of sense relations. It is the difference between the second and third level that allows for explanation of appropriate productive use in novel contexts by children who at the same time fail comprehension tests of the term. Comprehension tests, as the previous section of the chapter showed, are designed to assess children’s understanding of the contrastive uses of a set of related terms. This understanding depends on an organization of sense relations at the lexical level, independent of actual contexts and situations. Thus, 4-year-olds, who have been appropriately using terms like think, know, guess, and remember from age 2 or 3, and who still fail to choose the correct term in comprehension tests, “may know a good deal about the system of use but little about the system of meanings for commonly used words, whose very
4. CHILDREN’S ACQUISITION OF THE MENTAL LEXICON
69
meaning depends on their relations to other terms in the domain of knowledge and discourse” (Nelson & Kessler Shaw, 2002, p. 53, italics in original). Donaldson (1978, 1992) made a somewhat similar argument in discussing how children move from treating language as embedded in particular contexts to the decontextualized treatment of language. This is what is required, not just in comprehension tests of particular terms, but to understand problems, such as Piagetian class inclusion problems, where it is essential to pay scrupulous attention to the words of the question and not be misled by more obvious extralinguistic features of the situation. It is a shift in attention, Donaldson argued, from what people mean to what words mean, and children’s ability to make this shift is assisted by the acquisition of literacy, which fixes the words on the page. Donaldson’s “disembedding from context” and Nelson’s “sense relations among words” are what is required to have full control of word meanings and perform correctly in comprehension experiments. Thus, Nelson argued that children acquire the mental lexicon by using the terms they hear others use, at first without real understanding. Following this pragmatic use, the terms acquire meaning for the child and the corresponding concepts are constructed. Thus, a very important question concerning children’s acquisition of the mental lexicon is the following: In what contexts and under what circumstances do children hear mental terms used? Nelson herself called for more studies addressed to this question: “Language constitutes a medium through which … categories and structures can be directly imparted. … More studies of the ways in which parents and teachers incorporate talk about mental states in their discourse with children are needed to illuminate this process” (Nelson, 1996, p. 312).
A TRAINING STUDY Our training study was explicitly designed to determine whether it was possible to help children acquire a mental lexicon by using these terms in stories that were read to them and in conversations about the stories (Peskin & Astington, 2002). Findings from longitudinal studies show that parental use of terms in conversation is related to children’s use (Furrow, Moore, Davidge, & Chiasson, 1992; Moore, Furrow, Chiasson, & Patriquin, 1994; Ruffman, Slade, & Crowe, 2002). A particularly important source of mentalistic terms may be in stories and in conversations about stories as parents read to their children. Narratives, even those for young children, are often full of references to people’s mental states, to what different characters think, want, know, remember, and so on (Cassidy et al., 1998; Dyer, Shatz, & Wellman, 2000). There are differences in the extent to which parents use mental terms to talk about people’s thoughts and motives (Gearhart & Hall, 1982). Some parents are more inclined to see children, right from the start, as individuals with minds of their own (Meins & Fernyhough, 1999). They are thus more likely to reason with
70
ASTINGTON AND PESKIN
their child, considering the child’s point of view and explaining different points of view. Such explanations often involve metacognitive terms, such as “He didn’t know you wanted it,” “He thought it was his turn,” and so on. A number of studies have shown that children whose parents talk to them in this way during their preschool years are likely to acquire richer understanding and productive use of the mental lexicon than children whose parents are less inclined to use this style of interaction (Furrow et al., 1992; Moore et al., 1994; Ruffman et al., 2002). Furthermore, it seems likely that children who have such facility with mentalistic language will be at an advantage in the early school years because this language is important in school tasks and classroom communication (Astington & Pelletier, in press). The training study was conducted in schools in low-income neighborhoods where it is less likely that children will have been exposed to high levels of mentalistic talk in the home (Gearhart & Hall, 1982; Hall, 1987; Hall, Scholnick, & Hughes, 1987). It is also less likely that these children will have spent much time having stories read to them at home (Adams, 1994; Teale, 1986); such experience also provides a rich source of mentalistic language (Cassidy et al., 1998; Dyer et al., 2000). The study addresses the following questions: First, if children are exposed to parent and teacher book reading and conversation with dramatically increased numbers of metacognitive terms, then will it help them to acquire a mental lexicon? Second, will it result in greater conceptual understanding of the mental world? In the study (Peskin & Astington, 2002), the same picture books were read to both experimental and control groups, but the original text was pasted over with text and text questions rich in explicit metacognitive terms for the experimental group, and stripped of any metacognitive terms for the control group. It must be noted, however, that young children’s narratives, and the pictures that illustrate them, are full of implicit references to people’s mental states, to what different characters think, want, know, and so on (Cassidy et al., 1998; Dyer et al., 2000). Indeed, whereas Nelson (1996) emphasized semantic terms as a particularly powerful system for acquiring meaning, she acknowledged that it is just one of many cultural systems of meaning making. Explicit metacognitive terms may be an important component in developing an understanding of one’s own and other people’s minds, but there is also a role for the observation of tricks, disguises, hiding, and so on, all of which may be illustrated in children’s stories. In the study, the comparison was therefore between an experimental group that received dramatically increased numbers of explicit metacognitive terms, and a control group that observed implicit references to characters’ beliefs and desires. A total of 4 teachers and 48 children were involved in the intervention. There were approximately equal numbers of boys and girls with a mean age of 4 years and 6 months at the start of the study. The children were equally divided between control and experimental groups, and approximately equally drawn from eight junior kindergarten classes. The four teachers involved each had both a morning and an afternoon class, which made it possible to control for teacher and time-of-day
4. CHILDREN’S ACQUISITION OF THE MENTAL LEXICON
71
effects by having two of the teachers work with an experimental group in the morning and a control group in the afternoon, and reversing this order for the other two teachers. As is typical of the population from which the participants were drawn, approximately 70% of them spoke another language in addition to English at home, but they had sufficient English-language competence to be included in the study, as judged by the classroom teacher and experimenters. The study design involved control and experimental groups with pretesting just before, and posttesting just after, a 4-week intervention period. During the intervention, stories were read to the children at home and at school. Each child was given a set of books to take home and a sticker chart on which to record reading sessions. In school, the classroom teacher read the same books to the whole class group, again recording reading sessions. In addition, the experimenters withdrew the study children from the classroom in groups of two or three and read and discussed the stories with them. Approximately 5% of words in the experimental-group stories were metacognitive terms: know and think occurred very frequently; wonder, figure out, guess, remember, forget, decide, and understand occurred quite frequently; and there were occasional occurrences of explain, pretend, imagine, surprise, expect, and be sure. Furthermore, in order to encourage interaction between the adult reader and child listener, questions were written into the stories for both experimental and control groups. For the experimental group, these questions were designed to encourage both comprehension and production of explicit metacognitive language. The same set of tasks was used for pretesting and posttesting, but with different content at the two times. The tasks were a storytelling task measuring metacognitive term production (McKeough, 1992); a test of comprehension of eight different metacognitive terms (Astington & Pelletier, 2002); false-belief prediction tasks (Perner, Leekam, & Wimmer, 1987); false-belief explanation tasks (Bartsch & Wellman, 1989; P. L. Harris, Johnson, Hutton, Andrews, & Cooke, 1989); and a general language test, the Test of Early Language Development, (TELD–III; Hresko, Reid, & Hammill, 1999). Only children who scored 2 or less out of 4 on the false-belief prediction pretest were included in the intervention. There were no significant differences between the control and experimental groups on any of the pretest measures, including age. Further, there were no significant differences between the two groups in the number of times the stories were read to them at home or at school, by teachers and experimenters. The posttest findings were interesting and not entirely as anticipated. The first question was whether children exposed to parent and teacher book reading and conversation with dramatically increased numbers of metacognitive terms would be helped to acquire a mental lexicon. As expected, the experimental group children, who had heard many metacognitive terms in the stories read to them during the 4-week intervention, produced significantly more types of metacognitive terms in the posttest storytelling task than the control group children did. However,
72
ASTINGTON AND PESKIN
neither group showed any significant improvement on the metacognitive term comprehension task. That is, the group with greater exposure to a variety of metacognitive words acquired appropriate productive use of them, without improving their comprehension of contrastive uses within the set of words. In Nelson’s terms (Nelson & Kessler Shaw, 2002), they had acquired a system of use but had not yet mastered the corresponding system of meanings. The second question asks whether children who were exposed to dramatically increased numbers of metacognitive terms would acquire greater conceptual understanding of the mental world than children who were not given such exposure. Interestingly, there was a significant increase for both groups in their false-belief prediction test scores and no difference between the two groups. Both groups also improved significantly on the false-belief explanation task but, unexpectedly, the control group improved significantly more than the experimental group. The intervention did not affect the general language ability test scores for either group. Thus, the children who had intense exposure to mentalistic language subsequently used this language to a greater extent than children who had not had such exposure. However, their conceptual understanding of mental states, as indicated by false-belief prediction and explanation tasks, did not improve relative to the group of children who had not been exposed to such language in the stories that were read to them. Indeed, both groups showed markedly improved performance on these tasks, perhaps because of the story reading. Although there was no control group of children who did not have stories read to them, the false-belief test performance of the study children improved more than might be expected over such a short period of time without any intervention (Hughes et al., 2000; Mayes, Klin, Tercyak, Cicchetti, & Cohen, 1996). As described, despite the fact that no metacognitive terms were used in the control group’s story texts and questions, the stories themselves included mentalistic concepts, as is typical of preschool storybooks (Cassidy et al., 1998; Dyer et al., 2000). For example, one text was the story of Rosie the hen, who walks around a farm, unaware that she is being followed by a fox, who suffers numerous misadventures (Hutchins, 1968). Table 4.1 shows a sample of text taken from near the beginning and the end of the story, in the two versions. It is evident that although the control group story includes no metacognitive terms, Rosie’s lack of knowledge is made perfectly clear, perhaps by the use of perceptual terms, although it is clear anyway from the illustrations. Indeed, the original story text is comprised of only 32 words. Thus, mentalistic concepts were an integral part of the control group children’s stories, even though the texts contained no metacognitive terms. The findings suggest that exposure to such stories, even without a single explicit metacognitive term in the text, promotes children’s conceptual understanding. As Nelson (1996) pointed out, children live in a sociocultural system that has emerged over time from use within successive interpretive communities. Hearing metacognitive
4. CHILDREN’S ACQUISITION OF THE MENTAL LEXICON
73
TABLE 4.1 Comparison Samples of Text from Control and Experimental Group Stories Control Group Story
Experimental Group Story
Fox is not careful and bumps into a rake. Rose hears the loud BUMP but does she turn around? No, she keeps her eyes right on the road. She doesn’t look left and she doesn’t look right. Watch out, Rosie!
Did you know that Fox would bump into a rake? Rosie heard the loud BUMP but did she figure out that is was hungry Fox behind her? No, she didn’t turn around. She doesn't know that he's behind her. Watch out, Rosie!
"Oow-eeee!" Fox cries as the bees chase after him. Do the bees sting him? Yes, they do, again and again. "Oow-eeee!" Fox cries as he runs away as fast as he can. Does Rosie turn around? No, Rosie just keeps on walking slowly along... And she gets home just in time for dinner.
Does Fox know that the bees are following him? Yes, he knows. “Oow-eeee!” Fox cries again and again. But does Rosie know that Fox has been following her? No, Rosie doesn’t know. She doesn’t even guess. Rosie just keeps on walking until she reaches her home. She is just in time for dinner.
Note: The metacognitive terms were not italicized in the children’s story books.
terms is likely to be only one part of children’s developing understanding of mind. Indeed, in light of the finding that the control group’s performance on the false-belief explanation task exceeded that of the experimental group, it may be that they acquired a deeper understanding because they had to actively construct their own mentalistic interpretation of the stories. The final section of the chapter considers what light the findings throw on the issue of theoretical explanations of children’s acquisition of the mental lexicon and, more generally, what conclusions might be drawn about this issue, especially in light of Nelson’s work.
CONCLUSIONS: ACQUIRING THE MENTAL LEXICON So, is it possible to help children to acquire the mental lexicon? The children in the study who were exposed to a wide variety of metacognitive terms subsequently used a wider variety of such terms, in their own storytelling, than the children who had not heard these terms in the stories that were read to them. Thus, it is possible to help children to acquire productive use of the terms. However, at this stage, the children may have been using the terms without real understanding of them (“use before meaning”) because their scores on the metacognitive term comprehension test did not improve from the pretest. As Nelson’s theory of experiential semantics argued, children use metacognitive terms pragmatically before acquiring semantic understanding of them. Most important, in her theory, the acquisition of this understanding is crucially dependent on the child’s own cognitive experiences and resources. Although the
74
ASTINGTON AND PESKIN
sociocultural context provides the raw material (the terms), children have to make something of this using their own internal cognitive resources, in a process that Nelson (1996) calls co-construction. That is, although children’s mentalistic concepts are dependent on the categories and relations among categories lexicalized in their language, as discussed earlier in the chapter, children themselves have an important role to play in conceptual construction. From this perspective, it is interesting that in the training study, the control group children’s ability to explain action mentalistically improved more than that of the experimental group. Metacognitive terms were deliberately excluded from the texts and the questions in the texts to which these children were exposed. However, as is typical (Cassidy et al., 1998; Dyer et al., 2000), mentalistic concepts were integral to the stories, even though the texts themselves contained no metacognitive terms. The findings suggest that exposure to metacognitive terms is only one aspect of the sociocultural world that fosters an understanding of mental states. Intense exposure to storybooks with no metacognitive terms, but with pictures of tricks, disguises, and hiding, promotes children’s conceptual understanding. Indeed, in light of the finding that the control group’s performance on the false-belief explanation task exceeded that of the experimental group, it may be that they acquired a deeper understanding because, in making sense of the stories, they had to be more active in trying to comprehend what the characters knew and thought. They may, indeed, have had more experience than the experimental group in actively constructing their own mentalistic interpretation. The important point is that children are not passive recipients of mentalistic concepts, but are actively involved in their construction. Certainly, more needs to be said about how exactly these concepts are constructed, but that is a matter for another time.
ACKNOWLEDGEMENTS We are grateful for funding support from the National Academy of Education, the Natural Sciences and Engineering Research Council of Canada, and the Dr. R. G. N. Laidlaw Research Centre, Institute of Child Study, University of Toronto.
REFERENCES Adams, M. J. (1994). Beginning to read: Thinking and learning about print. Cambridge, MA: Bradford Books/MIT Press. Astington, J. W. (1988). Children’s understanding of the speech act of promising. Journal of Child Language, 15, 157–153. Astington, J. W. (1993). The child’s discovery of the mind. Cambridge, MA: Harvard University Press. Astington, J. W. (1996). What is theoretical about the child’s theory of mind? A Vygotskian view of its development. In P. Carruthers & P. K. Smith (Eds.), Theories of theories of mind (pp. 184–199). Cambridge, England: Cambridge University Press.
4. CHILDREN’S ACQUISITION OF THE MENTAL LEXICON
75
Astington, J. W. (2003). Sometimes necessary, never sufficient: False belief understanding and social competence. In B. Repacholi & V. Slaughter (Eds.), Individual differences in theory of mind: Implications for typical and atypical development. Macquarie Monographs in Cognitive Science Series (pp. 13–38). New York: Psychology Press. Astington, J. W., Harris, P. L., & Olson, D. R. (Eds.). (1988). Developing theories of mind. New York: Cambridge University Press. Astington, J. W., & Olson, D. R. (1990). Metacognitive and metalinguistic language: Learning to talk about thought. Applied Psychology: An International Review, 39, 77–87. Astington, J. W., & Pelletier, J. (1996). The language of mind: Its role in learning and teaching. In D. R. Olson & N. Torrance (Eds.), The handbook of education and human development: New models of learning, teaching and schooling (pp. 593–619). Oxford, England: Blackwell. Astington, J. W., & Pelletier, J. (2002). Theory of mind and metacognitive vocabulary development in first and second languages. Unpublished manuscript, University of Toronto. Astington, J. W., & Pelletier, J. (in press). Theory of mind, language, and learning in the early years. In B. D. Homer & C. Tamis-Lemonda (Eds.), The development of social cognition and communication. Mahwah, NJ: Lawrence Erlbaum Associates. Bartsch, K., & Wellman, H. M. (1989). Young children’s attribution of action to beliefs and desires. Child Development, 60, 946–964. Bartsch, K., & Wellman, H. M. (1995). Children talk about the mind. New York: Oxford University Press. Beckwith, R. T. (1991). The language of emotion, the emotions, and nominalist bootstrapping. In D. Frye & C. Moore (Eds.), Children’s theories of mind (pp. 77–95). Hillsdale, NJ: Lawrence Erlbaum Associates. Bloom, L., Rispola, M., Gartner, B., & Hafitz, J. (1989). Acquisition of complementation. Journal of Child Language, 16, 101–120. Bretherton, I., & Beeghly, M. (1982). Talking about internal states: The acquisition of an explicit theory of mind. Developmental Psychology, 18, 906–921. Bretherton, I., McNew, S., & Beeghly-Smith, M. (1981). Early person knowledge as expressed in gestural and verbal communication: When do infants acquire a “theory of mind”? In M. E. Lamb & L. R. Sherod (Eds.), Infant social cognition (pp. 333–373). Hillsdale, NJ: Lawrence Erlbaum Associates. Brown, J. R., & Dunn, J. (1991). “You can cry, mum”: The social and developmental implications of talk about internal states. British Journal of Developmental Psychology, 9, 237–256. Carey, S. (1990). On some relations between the description and the explanation of developmental change. In G. Butterworth & P. Bryant (Eds.), Causes of development (pp. 135–157). London: Harvester Wheatsheaf. Cassidy, K. W., Ball, L. V., Rourke, M. T., Werner, R. S., Feeny, N., Chu, J. Y., Lutz, D. J., & Perkins, A. (1998). Theory of mind concepts in children’s literature. Applied Psycholinguistics, 19, 466–470. de Villiers, J. G., & de Villiers, P. A. (2000). Linguistic determinism and the understanding of false beliefs. In P. Mitchell & K. J. Riggs (Eds.), Children’s reasoning and the mind (pp. 191–228). Hove, England: Psychology Press. Donaldson, M. (1978). Children’s minds. Glasgow, Scotland: Fontana. Donaldson, M. (1992). Human minds. Harmondsworth, England: Penguin.
76
ASTINGTON AND PESKIN
Dunn, J. (1988). The beginnings of social understanding. Cambridge, MA: Harvard University Press. Dunn, J. (1991). Young children’s understanding of other people: Evidence from observations within the family. In D. Frye & C. Moore (Eds.), Children’s theories of mind (pp. 97–114). Hillsdale, NJ: Lawrence Erlbaum Associates. Dyer, J. R., Shatz, M., & Wellman, H. M. (2000). Young children’s storybooks as a source of mental state information. Cognitive Development, 15, 17–37. Furrow, D., Moore, C., Davidge, J., & Chiasson, L. (1992). Mental terms in mothers’ and children’s speech: Similarities and relationships. Journal of Child Language, 19, 617–631. Geach, P. T. (1957). Mental acts. London: Routledge & Kegan Paul. Gearhart, M., & Hall, W. S. (1982). Internal state words: Cultural and situational variation in vocabularly usage. In K. Borman (Ed.), The social life of children in a changing society (pp. 219–252). Hillsdale, NJ: Lawrence Erlbaum Associates. Hall, W. S. (1987). The semantic-pragmatic distinction in the investigation of mental state words: The role of the situation. Discourse Processes, 10, 169–180. Hall, W. S., Scholnick, E. K., & Hughes, A. T. (1987). Contextual constraints on usage of cognitive words. Journal of Psycholinguistic Research, 16, 289–310. Harris, P. L., Johnson, C. N., Hutton, D., Andrews, G., & Cooke, T. (1989). Young children’s theory of mind and emotion. Cognition and Emotion, 3, 379–400. Harris, R. J. (1975). Children’s comprehension of complex sentences. Journal of Experimental Child Psychology, 19, 420–433. Hresko, W. P., Reid, D. K., & Hammill, D. D. (1999). Test of Early Language Development: 3rd ed.Austin, TX: Pro-Ed. Hughes, C., Adlam, A., Happ, F., Jackson, J., Taylor, A., & Caspi, A. (2000). Good test-retest reliability for standard and advanced false-belief tasks across a wide range of abilities. Journal of Child Psychology and Psychiatry and Allied Disciplines, 41, 483–490. Hutchins, P. (1968). Rosie’s walk. New York: Aladdin Paperbacks. Johnson, C. N. (1982). Acquisition of mental verbs and the concept of mind. In S. Kuczaj II (Ed.), Language development: Syntax and semantics (pp. 445–478). Hillsdale, NJ: Lawence Erlbaum Associates. Johnson, C. N., & Maratsos, M. P. (1977). Early comprehension of mental verbs: Think and know. Child Development, 48, 1743–1747. Johnson, C. N., & Wellman, H. M. (1980). Children’s developing understanding of mental verbs: Remember, know and guess. Child Development, 51, 1095–1102. Levy, E., & Nelson, K. (1994). Words in discourse: A dialectical approach to the acquisition of meaning and use. Journal of Child Language, 21, 367–389. Limber, J. (1973). The genesis of complex sentences. In T. E. Moore (Ed.), Cognitive development and the acquisition of language (pp. 169–185). New York: Academic Press. Lohmann, H., & Tomasello, M. (2003). The role of language in the development of false-belief understanding: A training study. Child Development, 74, 1130–1144. Mayes, L. C., Klin, A., Tercyak, K. P., Cicchetti, D. V., & Cohen, D. (1996). Test-retest reliability for false belief tasks. Journal of Child Psychology and Psychiatry, 37, 313–319. McKeough, A. (1992). The structural foundations of children’s narrative and its development. In R. Case (Ed.), The mind’s staircase: Exploring the conceptual underpinnings of
4. CHILDREN’S ACQUISITION OF THE MENTAL LEXICON
77
children’s thought and knowledge (pp. 171–188). Hillsdale, NJ: Lawrence Erlbaum Associates. Meins, E., & Fernyhough, C. (1999). Linguistic acquisitional style and mentalizing development: The role of maternal mind-mindedness. Cognitive Development, 14, 363–380. Miscione, J. L., Marvin, R. S., O’Brien, R. G., & Greenberg, M. T. (1978). A developmental study of preschool children’s understanding of the words “know” and “guess”. Child Development, 49, 1107–1113. Montgomery, D. E. (2002). Mental verbs and semantic development. Journal of Cognition and Development, 3, 357–384. Moore, C., Bryant, D., & Furrow, D. (1989). Mental terms and the development of certainty. Child Development, 60, 167–171. Moore, C., & Davidge, J. (1989). The development of mental terms: Pragmatics or semantics? Journal of Child Language, 16, 633–641. Moore, C., Furrow, D., Chiasson, L., & Patriquin, M. (1994). Developmental relationships between production and comprehension of mental terms. First Language, 14, 1–17. Nelson, K. (1973). Structure and strategy in learning to talk. Monographs of the Society for Research in Child Development, 38(1–2, Serial No. 149). Nelson, K. (1974). Concept, word, and sentence: Interrelations in acquisition and development. Psychological Review, 81, 267–285. Nelson, K. (1986). Event knowledge: Structure and function in development. Hillsdale, NJ: Lawrence Erlbaum Associates. Nelson, K. (1988). Where do taxonomic categories come from? Human Development, 31, 3–10. Nelson, K. (1996). Language in cognitive development: The emergence of the mediated mind. New York: Cambridge University Press. Nelson, K., Hampson, J., & Kessler Shaw, L. (1993). Nouns in early lexicons: Evidence, explanations and implications. Journal of Child Language, 20, 61–84. Nelson, K., & Kessler Shaw, L. (2002). Developing a socially shared symbolic system. In E. Amsel & J. P. Byrnes (Eds.), Language, literacy, and cognitive development: The development and consequences of symbolic communication (pp. 27–57). Mahwah, NJ: Lawrence Erlbaum Associates. Nelson, K., & Lucariello, J. (1985). The development of meaning in first words. In M. D. Barrett (Ed.), Children’s single word speech (pp. 59–86). Chichester, England: Wiley. Nelson, K., Plesa, D., & Henseler, S. (1998). Children’s theory of mind: An experiential interpretation. Human Development, 41, 7–29. Olson, D. R. (1994). The world on paper. Cambridge, England: Cambridge University Press. Perner, J., Leekam, S., & Wimmer, H. (1987). Three-year-olds’ difficulty with false belief: The case for a conceptual deficit. British Journal of Developmental Psychology, 5, 125–137. Peskin, J., & Astington, J. W. (2002, April). Using cognitive terms to enrich story texts for disadvantaged young children. Paper presented at the annual meeting of the American Educational Research Association, New Orleans, LA. Premack, D., & Woodruff, G. (1978). Does the chimpanzee have a theory of mind? Behavioral and Brain Sciences, 1, 515–526. Ruffman, T., Slade, L., & Crowe, E. (2002). The relation between children’s and mothers’ mental state language and theory-of-mind understanding. Child Development, 73, 734–751.
78
ASTINGTON AND PESKIN
Shatz, M., Wellman, H. M., & Silber, S. (1983). The acquisition of mental verbs: A systematic investigation of the first reference to mental state. Cognition, 14, 301–321. Shultz, T. R. (1980). Development of the concept of intention. In W. A. Collins (Ed.), Minnesota symposia on child psychology: Vol. 13. Development of cognition, affect, and social relations (pp. 131–164). Hillsdale, NJ: Lawrence Erlbaum Associates. Teale, W. H. (1986). Home background and young children’s literacy development. In W. H. Teale & E. Sulzby (Eds.), Emergent literacy: Writing and reading (pp. 173–206). Norwood, NJ: Ablex. Wallas, G. (1927). The art of thought. London: Cape. Wellman, H. M., Cross, D., & Watson, J. (2001). Meta-analysis of theory of mind development: The truth about false-belief. Child Development, 72, 655–684. Wellman, H. M., & Johnson, C. N. (1979). Understanding mental processes: A developmental study of remember and forget. Child Development, 50, 79–88. Whorf, B. L. (1956). Language, thought, and reality. Cambridge, MA: MIT Press. Wimmer, H., Gruber, S., & Perner, J. (1984). Young children’s conception of lying: Lexical realism—moral subjectivism. Journal of Experimental Child Psychology, 37, 1–30. Wimmer, H., & Perner, J. (1983). Beliefs about beliefs: Representation and constraining function of wrong beliefs in young children’s understanding of deception. Cognition, 13, 103–128.
5 Voice and Silence: A Feminist Model of Autobiographical Memory Robyn Fivush Emory University
Talk about the past is ubiquitous. Whether we are talking with old friends or new acquaintances, colleagues or family members, everyday conversations are replete with the stories of our lives. But what is it we talk about when we talk about the past? And what is that we do not say? Although the psychological study of autobiographical memory has experienced a resurgence in the past decade, the focus has been on formulating how autobiographical memories are represented and retrieved. The majority of research focuses on accuracy and retention over time (e.g., Conway & Rubin, 1993; Rubin, 1996). Moreover, virtually all the research conceptualizes autobiographical memory as internal to the individual; that is, although autobiographical memory may be used to achieve social and emotional goals (e.g., Conway & Pleydell-Pearce, 2000), it is a system that is organized within individual minds. In this chapter, I present a different approach to autobiographical memory based on Katherine Nelson’s (1989, 1993, 1996, 2001) social-constructionist model. How are memories of personal experiences modulated through joint reminiscing with others? More specifically, how does what individuals choose to tell 79
80
FIVUSH
and not tell emerge from the kinds of interactions in which they engage and the people with whom they reminisce? I approach this question from both developmental and feminist perspectives. To foreshadow, I argue that autobiographical memory must be conceptualized within a framework that examines the ways in which personal experience is given voice, told from the individual’s “owned” perspective, versus the ways in which personal experience may be silenced, either through disallowing certain stories to be told or imposing certain perspective on the remembered event. Essentially, how are autobiographical narratives validated or invalidated by the individual, by the conversational partner and by the larger community? To flesh out this argument, I first describe the basic principles of feminist theory in more detail and define voice and silence within the feminist concepts of place and power. Language is clearly central to this argument, and, therefore, in the second section, I show how language is critical to the development of autobiographical memory more generally and, specifically, to the development of validated and invalidated autobiographical narratives. Following from this, I develop a model of autobiographical memory based on two dimensions, voice and silence by self or other, and this model is illustrated using examples from early mother–child reminiscing. Finally, I return to the issues of language, memory, voice, and silence, and revisit relations between what may be remembered and what may be said.
THE FEMINIST PERSPECTIVE Although there are multiple feminist theories (see Rosser & Miller, 2000), all share the common core assumption that place and power are critical in understanding human culture (Yoder & Kahn, 1992). In order to explicate these concepts, I focus on feminist standpoint theory (Alcoff & Potter, 1993; Haraway, 1988; Harding, 1993) because this theory emerged from, and critiques, specific aspects of the scientific method as used within the social sciences (Fivush, 2000). Importantly, feminist standpoint theory endorses the scientific method and the role of experimentally derived empirical data, but argues for placing empirical data in a more contextualized framework of knowledge and objectivity. More specifically, feminist standpoint theory reformulates some of the basic assumptions of logical positivism, which has guided theory and methodology in psychology for over 100 years (e.g., Heidbreder, 1933; Pepper, 1946).
Logical Positivism Logical positivism makes three fundamental assumptions (Haraway, 1988; Harding, 1993). First, the scientist is assumed to be objective; by virtue of this objectivity, the social scientist can observe human behavior without bias. Thus, the scientist is in a privileged position to define reality (i.e., the “real” objectively de-
5. A FEMINIST MODEL OF AUTOBIOGRAPHICAL MEMORY
81
fined reasons why people behave the way they do). Second, behavior can be reduced into component elements in order to assess cause-and-effect relations. Once these elements are identified, they can be put back together to explain the whole; the whole is no more than the sum of its parts. Third, knowledge is represented independently in the individual mind and the sociocultural worlds in which individuals live are not relevant. Reality (i.e., the real causes of behavior) can be objectively defined without recourse to history, culture, or context; the goal of psychology is to discover basic principles of human behavior that can predict across gender, race, class, and context. Obviously, this is a simplified exposition, and many theories, especially within developmental psychology, have questioned these assumptions. In particular, approaches stemming from Vygotsky’s (1978) sociocultural theory of development have highlighted the contextualized nature of human cognition (see Gauvain, 2001, and Rogoff, 1990, for overviews). Still, these basic assumptions have guided much of the work in developmental psychology within Piagetian and information-processing models, the two dominant developmental models (Miller & Scholnick, 2000).
The Concept of Place In contrast to assumptions underlying logical positivism, feminist theories focus on the interrelated and contextual basis of knowledge (Longino, 1993; L.H. Nelson, 1993). Knowledge is embedded in the way in which social activity is structured and knowledge emerges from social interactions. Knowledge must be considered in terms of who knows, in what situations, and for what purposes. Because knowledge cannot be extricated from sociocultural structures, the observer can never be completely unbiased. An observer is always, by definition, observing from a specific place or perspective. This is the “standpoint” in feminist standpoint theory, and it is defined historically, culturally, individually, and situationally. Historically and culturally, each individual is positioned in a particular time and place, socialized within the specific belief systems of the historical and cultural milieu. This is not to say that individuals can never see beyond these socialized lenses, but that it is difficult and never completely successful. Each individual is a member of a specific gender, race, and class, and is thus defined historically and culturally as a particular kind of person. This definition allows the individual access to certain ways of knowing and denies access to other ways of knowing. For example, being male or female will provide the individual access to particular kinds of activities (e.g., Batman vs. Barbie birthday parties, fraternity initiations vs. sorority teas), and these activities lead to the development of specific kinds of skills over others. Thus, as individuals engage in culturally prescribed activities, they learn to perform in ways appropriate to their “place” in the social structure (Fivush, 1998). Obviously, these kinds of activities change historically (and per-
82
FIVUSH
haps even radically as the last three decades of racial and gender discrimination in access to various activities and institutions have loosened), but they are always present.1 Finally, all behaviors are influenced by the specific situation in which the individual is embedded. Behavior is constructed with other people in particular situations in which multiple goals are negotiated and achieved. Thus place is a dynamic concept; an individual’s historical, cultural, and situational position in an ongoing stream of human activity is always evolving, although each of these levels of place evolve at different rates.2 The concept of place changes the understanding of the scientific method. Although scientists are trained in specific tools that allow for more systematic and objective observation, even scientists remain embedded in a particular historical and cultural context that can never be completely overridden. Thus, scientists must seriously consider the standpoint from which they are observing, and how this might affect their observations. Further, behavior must be conceptualized as dynamic and fluid, and therefore it cannot be reduced to independent cause-and-effect relations, but must be understood transactionally and reciprocally. Finally, knowledge is not in an individual’s head, but in the relation between the individual and the environment.3 Thus, rather than objectivity being defined as an unbiased perspective (the view from nowhere), feminist standpoint theory defines objectivity as the coordination of multiple perspectives; objectivity emerges from diversity of perspectives (the view from everywhere) (Bordo, 1990; Code, 1993; Harding, 1993). The ultimate goal of psychology is not to deduce context-free universal principles of behavior, but rather to specify the conditions under which different individuals will display specific kinds of behaviors.
The Concept of Power Due to the way in which society has come to define specific roles in the social structure, some standpoints are imbued with more authority or power than others. Views from more culturally accepted standpoints are considered the center, whereas views from less accepted standpoints are at the margins. The view from the center is given “voice.” It is the accepted version of a shared socially constructed reality, whereas views from the margins are “silenced.” These stories are either not heard or these perspectives are not validated. In this sense, power gives voice. Within the psychological literature, power is often equated with dominance or status (e.g., Raven, 1992). However, critiques of power stemming from feminist theory have provided a more nuanced understanding of this concept (Griscom, 1992; Yoder & Kahn, 1992). In summarizing and integrating across psychological and feminist conceptualizations of power, Hall and Halberstadt (1996) defined three interrelated dimensions that have contributed to theorizing about power: process, location, and quality. Process refers to whether power is conceptualized as static or dynamic. Location refers to where the power resides, in the individual or in the relationship. Finally, power can be cooperative, in that individuals use
5. A FEMINIST MODEL OF AUTOBIOGRAPHICAL MEMORY
83
their power democratically (power-with), or power is coercive, in that individuals use their power autocratically (power-over). Arguing from a feminist perspective, Griscom (1992) agreed that power is more than coercion or dominance; power can be power over other people, but power can also be power-with others, or power over oneself, in the sense of empowerment. Importantly, power-over is not always “bad” and power-with “good.” There are times within particular relationships or contexts when dominance is appropriate and other times when power should be shared relationally. Dominance and empowerment overlap in complex ways and the appropriate balance between them will evolve over time within relationships and contexts. Power can be something that individuals dictate, abdicate, share, or own. Thus, power is always relational. Power exists between people and emerges from relationships; power is a process that occurs over time. Finally, power must be conceptualized as an intersection of the individual and society; individuals exist within societal power structures and societal power structures are simultaneously created by individuals.
Voice and Silence The way in which voice and silence are conceptualized emerge from place and power (Belenky, Clinchey, Goldberger & Tarule, 1986; Gilligan, 1982). From the feminist concept of place, voice and silence must be seen as dynamic and relational. Voice and silence will emerge within the individual as a function of their historical and cultural place and their individual history of specific interactions with specific others. The ways in which individuals develop voice or silence will have important implications for the development of an autobiographical life story. Experiences that are voiced provide a sense of validation; experiences are accepted as real and the individual’s perspective on the experience is viewed as appropriate. Experiences that are silenced lead to a sense of existential despair; experiences are not heard or the individual’s perspective on the experience is not accepted as appropriate. From the feminist concept of power, how voice emerges over time within specific relationships and whether voice is cooperative or coerced raises additional questions about authority. Who has the authority to author the autobiography? Are individuals allowed their own voice or are particular ways of telling the story imposed on them? And on the other side of the dimension, do individuals choose not to report certain information or are they simply not heard by those they tell? Thus, power may be expressed as voice or as silence depending on who has the authority to give voice or to silence. This review of the feminist concepts of place and power culminates in a two-dimensional model of autobiographical memory, with voice and silence as one dimension and self and other as the second dimension. The two dimensions can be crossed yielding four quadrants, as displayed on Fig. 5.1. Clearly, voice and silence imply language, at least metaphorically, in that what is voiced is said
84
FIVUSH
FIG. 5.1. Voice and silence in autobiographical memory.
and heard, whereas what is silenced is either not told or not heard. Therefore, it is important to examine the role of language in the development of autobiographical memory.
LANGUAGE AND AUTOBIOGRAPHICAL MEMORY The concepts of voice and silence point to the critical role that language plays in modulating consciousness (see Damasio, 1999, Donald, 1991, and Nelson, 1996, for further theoretical arguments). More specifically, in terms of autobiographical memory, language allows for new ways of organizing and evaluating personal experience (Fivush, 1998, 2001; K. Nelson & Fivush, 2002). Language allows individuals to share their past experiences with others, and through this joint reminiscing, children learn the canonicalized narrative forms for representing and reporting their personal past. Although experienced events may be temporally sequenced before the advent of language (e.g., Bauer, 1997), linguistically based narrative forms allow for a deeper and more complex organization of personal memories (Bruner, 1990; Fivush & Haden, 1997). First, narratives provide an experiential context for discrete events. It is through narratives that single events can coalesce into larger themes that encompass life periods (e.g., school years, adolescence, midlife), life tracks (e.g., relationships, career) and more abstract motifs (e.g., regrets, accomplishments,
5. A FEMINIST MODEL OF AUTOBIOGRAPHICAL MEMORY
85
legacies). By placing single events in the larger context of an unfolding life story, young children begin to develop a sense of a true autobiography, a self with a past, present, and future (Habermas & Bluck, 2000; McAdams,1992). Second, and related to the first, narratives, by definition, are evaluative (Bruner, 1990; Labov, 1982). Autobiographical narratives move beyond telling what happened to include information about what the event meant to the individual; autobiographical narratives are about a self that has an emotional stake in the event and its consequences. Finally, language allows individuals to co-construct their personal experiences with others; it is through socially sharing the past that experiences take on personal coherence and meaning. Individuals’ autobiographical lives are created through and with others (Gergen, 1994; Middelton & Edwards, 1990). For young children who are just beginning to use language to reminisce about their past with others, the ways in which adults help them to organize and evaluate their experiences may be a critical filter through which children come to understand what these experiences mean for their understanding of the past and of themselves (Fivush, 2001; Fivush, Haden, & Reese, 1996). There is now a substantial body of research demonstrating that children are learning both the canonical narrative forms and an evaluative stance on their personal past through participating in adult-guided reminiscing (see Nelson & Fivush, 2002, for a review). Parents who engage in more elaborated and narratively coherent reminiscing with their preschool children have children who come to tell more narratively coherent, detailed stories of their own experience later in development (Haden, Haine, & Fivush, 1997; Peterson & McCabe, 1992; Reese, Haden, & Fivush, 1993). And parents who engage in and encourage their children in interpreting and evaluating their experience facilitate their children’s developing use of narrative evaluation as they grow older (Fivush, 1991; Haden et al., 1997). Narrative evaluation provides a subjective perspective on the past; this is what happened, and this is how I think and feel about it. Through reminiscing with others, children come to understand that their perspective on the event may or may not be the same as someone else’s perspective. In this way, children come to understand that they have a unique perspective on what occurred. In a very real sense, it is only when we share experiences with others that they become our own (Fivush, 2001). But language is a two-edged sword. By definition, individuals do not report everything they remember about any given event on any given recall occasion. By foregrounding particular aspects of an event, other aspects are necessarily backgrounded. This is not to argue that autobiographical memory is linguistically represented. Obviously, representations of personally experienced events are quite complex and represented in multiple modalities, including visual, auditory, and kinesthetic sensations, as well as language, and much of what is remembered may never be put into words and never spoken. Language does not determine
86
FIVUSH
memory, but in sharing the past with others, language has a privileged place; it is primarily through language that individuals can communicate their past experiences and what these events mean (Pillemer & White, 1989). But the very act of putting experiences into words validates some aspects of the experience over others, validates a particular interpretation or evaluation of the event over others (K. Nelson, 1996). As the novelist Janet Fitch (1990) explained, “That was the thing about words, they were clear and specific—but when you talked about feelings, words were too stiff, they were this and not that, they couldn’t include all the meanings. In defining, they always left something out” (p. 265).
The Role of the Other If the personal past takes on meaning as it is socially shared with others, then the ways in which others listen to, hear, and interpret the past has implications for what aspects of the past will be validated. Listeners can accept or dismiss, negotiate, cajole, or coerce particular evaluations over others (see Pasupathi, 2001, for a theoretical review). Through this jointly constructed version of what occurred and what it means, some aspects of memories are given voice whereas others are silenced. Moreover, as discussed earlier, from feminist standpoint theory, what is voiced and what is silenced occurs at multiple levels simultaneously—cultural, individual, and situational. At the cultural level, cultures define a canonical life story and how to tell it (e.g., Connerton, 1989). In Western culture, a focus on the self and individual achievement is considered appropriate, whereas in eastern culture, the focus is on the individual’s place in the larger community and that person’s contribution to a moral society (Oyserman & Markus, 1993), and these differences are reflected in autobiographical memory. For example, Asian Indians from rural villages have few and sparsely detailed memories of their childhood or even their recent past (Leichtman, 2001). When asked to recount their personal experience, they respond that they do not remember, and that their memories are unimportant in the context of the larger community. Intriguingly, when reminiscing with their preschool children, Asian parents do not talk as much about the past as do Caucasian parents, they do not talk in as much elaborated detail, and they do not focus on the child to the same extent as Caucasian parents (see Leichtman, Wang, & Pillemer, 2003, for a review). Rather, Asian parents focus on the community and moral behavior to a greater extent than do Caucasian parents (Mullen & Yi, 1995). These different patterns emerge in children’s later independent autobiographical narratives, with Asian children narrating shorter, less detailed, and less self-focused experiences than Caucasian children (Han, Leichtman & Wang, 1998). Thus, the child’s developing skills in recounting the past are modulated such that cultural expectations about self and autobiography shape what information is and is not reported.
5. A FEMINIST MODEL OF AUTOBIOGRAPHICAL MEMORY
87
At the individual level, the specific kinds of experiences that are considered reportable and not reportable depend on where an individual is situated in the larger society. As argued earlier, it matters whether one is male or female, White or Black, rich or poor. Being a member of a particular race, class, and gender defines the individual’s place in the larger culture as well as their everyday interactions. The kinds of activities deemed appropriate and the kinds of interactions in which individuals are expected to engage change as a function of their place in the larger culture. For example, in Western culture, it is more acceptable for females to experience and express emotions than males (Basow, 1992; Fischer, 2000). As adults, women report experiencing and expressing emotions more intensely than do males (Fischer, 2000), and include more emotional information when reporting the personal past (Bauer, Stennes, & Haight, 2003; Davis, 1990). Similarly, parent–daughter reminiscing is substantially more emotion laden than parent–son reminiscing. With preschool daughters, parents talk more about emotion overall, talk about a wider variety of emotional experiences, and evaluate and validate their daughters emotional experience to a greater extent than with sons (see Fivush & Buckner, 2000, 2003, for a review). By the end of the preschool years, girls are reporting their personal past in more emotional terms than are males (Buckner & Fivush, 1998; Reese, Haden, & Fivush, 1996). In this sense, emotions are voiced for females but silenced for males (see Fivush, in press, for a full theoretical discussion). Finally, the specific situation in which individuals are recalling a past event influences what is reported. Who they are telling what story to for what purpose matters, both from the teller’s perspective and the listener’s perspective. From the teller’s perspective, people may choose to disclose some information to some people but not others. For example, Tenney (1989) examined the information that new parents told friends versus family about the birth. When talking with family, new parents focused on the infant’s characteristics and vital statistics. With friends, in contrast, new parents talked about the difficulties of the labor. Thus, the teller focuses on different aspects of the event depending on the intended audience. From the listener’s perspective, more attentive and concerned listeners elicit longer and more coherent narratives than do inattentive and distracted listeners (Pasupathi, Stallworth, & Murdoch, 1998). Of course, the teller and the listener are in a relationship and each mutually influences each other. For example, a large literature on self-disclosure indicates that the gender of both the teller and the listener matters; females disclose to both men and women but males tend to disclose only to female listeners (Snell et al., 1989). Overall, then, in any given recall context, we need to consider the ways in which specific information about the past is allowed to be voiced or is silenced by the culture, by the individual’s place in society, and by the specific situation in which an individual is recalling a specific event with a specific listener. Thus, voice and silence emerge within ongoing interactions in which the teller and listener negotiate or coerce a particular version of the past, as illustrated in Figure 5.1.
88
FIVUSH
VOICE AND SILENCE, SELF AND OTHER As already alluded to, children are learning the forms and functions of talking about the past in early parent-guided reminiscing. How might voice and silence add to an understanding of the development of autobiographical memory? In order to explore the usefulness of the proposed model, I will discuss each quadrant in turn, using examples from previous research on parent–child reminiscing (Fivush & Fromhoff, 1988; Reese et al., 1993). However, it is important to first emphasize several points. First, it is obvious that parents hold power over children, although the way in which this power is expressed may vary widely both across individuals and situations. Thus, relations between power and voice and silence emerge from specific evolving relationships, as is discussed in more detail later. Second, although each quadrant is discussed as a category for the sake of explication, the model conceptualizes voice and silence and self and other as dimensions rather than categories. Specific autobiographical memories can be more or less voiced and this dimension can be modulated more or less by self or other. Third, any specific autobiographical memory will have elements of both voice and silence by both self and other. For purposes of exposition, specific conversations are discussed as illustrative of one side of these dimensions or another, but it should be kept in mind that, in reality, memories are a complex interweaving of voice and silence by self and other. Finally, issues of voice and silence by self and other are not simply a matter of what is said and not said, but the conversational process by which specific aspects and evaluations of the past are validated, imposed, negated, or avoided. It is in the process of sharing experiences with others that each individual comes to have a voice or is silenced.
Self-Voice After the good times were over, as we grew older, we were to tell each other stories about the past, each adding his or her own fragments of pleasurable detail, until the joint memory became something larger than each single memory, and yet became something that each of us possessed fully, as if it were solely our own. (Wilson, 1998, p. 142)
In the self-voiced quadrant, individuals have power and voice over their own autobiographical experience. Although autobiographical memory is still shared in the social interaction, individuals have the authority to describe and evaluate their own experience and these experiences are validated by the listener, as can be seen in this conversational excerpt between a mother and her 5-year-old child talking about visiting a museum of natural history (M stand for Mother, C for Child and “…” indicates some intervening comments) M: What other kinds of dinosaurs were in there? C: Uh, Tyrannosaurus Rex. The first thing we came in, rrraarr!
5. A FEMINIST MODEL OF AUTOBIOGRAPHICAL MEMORY
89
M: (laughing) That’s right. And he was huge, wasn’t he? C: Huge, very huge. They take it, they dug up the bones. You know why? M: …They figured out how big the real ones were and then they made these. C: Nuh uh, they didn’t make those. M: They didn’t? C: Those were real bones. M: It was? C: They figured out how to put ’em out, up together. M: They did?… and they made ’em move, didn’t they? Didn’t they move? C: No. M: They did too move. (laughing) C: No, he did not. It did not have his skin on. M: Oh, that’s right, one of ’em was just bones. C: That was Tyrannosaurus Rex. M: Tyrannosaurus Rex was just his bones. Ok. Several things are notable in this conversation. First, both the mother and child are fully engaged, each responding to the other’s comments. Neither the mother nor child leads the conversation, but each responds to the other and then introduces a new aspect of the event to which the other again responds. The event is fully co-constructed. Moreover, even though it is clearly not the case that each always agrees with the other, there is a real sense of listening and responding to the other. When there is disagreement, the other is not ignored. Rather disagreements are negotiated until agreement is reached. Most important, the child feels comfortable challenging the mother’s version (”No, he did not”) and the mother accepts and validates her child’s version of the event “Oh, that’s right”). Importantly, it is not just the child’s version of the facts of the event that are validated in these conversations; it is also the child’s emotional reaction and evaluation, as shown in this excerpt between a mother and her 5-year-old child talking about a visit to the lake during which the child and her sister fell into the water: C: I remember Lauren, me falling, going into the water. M: …I do too. That was upsetting, wasn’t it? C: (makes crying noise) M: That’s what you did. It kind of scared you, didn’t it? C: I don’t like that! M: I don’t blame you!
90
FIVUSH
When the child recalls the upsetting incident, the mother immediately confirms the memory as shared (“I do too”), and then provides an evaluation of the event to which the child assents. As the child emphatically elaborates on her perspective of what occurred (“I don’t like that!”) the mother completely validates this perspective (“I don’t blame you!”). Conversations in which children are given voice include maternal affirmation and validation of what occurred and how the child felt about it. When there is disagreement, the mother and child negotiate a resolution rather than the mother imposing her version of what happened on the child. Thus, autobiographical memories falling along the self-voice dimension are validated; children learn to own their experience and to have authority in the construction of their own life story.
Other Voice It is our parents … who not only teach us our family history but who set us straight on our own childhood recollections, telling us that this cannot have happened the way we think it did, and that that, on the other hand, did occur just as we remember it. (McCarthy, 1957, p. xx, italics in the original)
In conversations falling into the other-voice quadrant, the mother tends to impose her version of what happened on the child. It is not so much that the mother disagrees with the child’s memory or evaluation of what happened as much as that the mother simply tells the event to the child and the child contributes little to the emerging narrative, as can be seen in this excerpt between a mother and her 4-year-old child about a visit to the zoo: M: Do you remember, we were strolling Baby around and do you remember when we went over near those ducks, what happened to Baby’s binkie (pacifier)? C: It fell in the water with the ducks….(Daddy) washed it under the bridge thing. M: Yeah, he found it under that bridge thing. And remember the ducks tried to get Baby’s binkie? And Daddy got it, but we had to wash it off first, didn’t we? We couldn’t give it to the baby when it had been in that yucky water, could we? And do you remember when we went in that building and Uncle Bob put you on his shoulder and we watched those penguins? And that lady was feeding the penguins? Do you remember that? I remember that. Do you remember when we went to eat, do you remember the special kind of french fries we had? What kind were they? C: (unintelligible) M: But do you remember the shape, what kind of shape those french fries were in? Those french fries were in little animal shapes, remember? We had those animal shape french fries? They were neat.
5. A FEMINIST MODEL OF AUTOBIOGRAPHICAL MEMORY
91
In this conversation, the mother essentially tells the child what happened, including what specific aspects were interesting and why (”But do you remember the shape?…They were neat.”). The child is an engaged listener but not a contributor. In this way, autobiographical memories falling along the other-voice dimensions are imposed; children are not the authors of their own story but have their stories told for them and about them.
Other Silence It didn’t seem like the kind of story that would gather with time, but instead would retract, condense, and turn into one of those things that nobody talked about, and in a year or so it would all be forgotten. (Proulx, 1992, p. 21)
In conversations that fall into the other-silence quadrant, children’s versions of what happened are silenced by the mother. In contrast to self-voiced memories, when children disagree with their parents, the child’s version of what occurred is dismissed. In contrast to other-voiced memories, it is not that the mother tells the story for the child, but that the child’s perspective is negated, as seen in this example of a mother discussing a trip to American Adventures, an amusement park, with her 6-year-old child: M …That was our first time there, and I thought you had— C: (interrupting) No, I don’t, no, it wasn’t my first time there. M: Yes it was. C: You don’t remember. Mom, remember when we went to it, umm, not at Chad’s birthday and not when we met Lauren, some other time. M: Oh, that was when we went to that place in Florida. C: No. M: With the rugs? C: No. M: Okay, well, that’s enough about American Adventures. I want to talk about something else. The mother is sure that this was the child’s fist visit to this amusement park, but the child is convinced that he had been there before. In fact, the child is quite insistent, providing several challenges and specific information to cue the mother’s memory of the other visit. First the mother simply denies the child’s memory, then she assumes it was a different memory (“that place in Florida”) and when the child again insists, the mother simply refuses to continue the discussion. Clearly, this child’s autobiographical memory is negated; it simply did not happen. Further, just as self-voiced memories can validate the child’s evalua-
92
FIVUSH
tion, other- silenced memories can silence evaluations as well as actual facts, as seen in this conversation between a mother and her 4-year-old child about a visit to an amusement park: M: …Was that fun to go on the ferris wheel? C: No. M: It wasn’t fun? You said it was fun. Was it scary? C: Yeah. I didn’t like the swings. M: I know you like to swing. But you just sat there. In this brief excerpt, the mother twice denies her daughter’s evaluation of the event. First, the mother states that the child had fun even when the child denies it, and then when the child says that she does not like to swing, the mother directly contradicts her. Autobiographical memories that fall on the other-silenced dimensions are negated; either the event or the child’s perspective on the event is simply ignored. When this happens, children are not given the authority to tell their own story. Things did not happen the way they thought they did; they did not feel what they thought they felt. In essence, they do not know who they are.
Self Silence One benefit, which I have lost, of a life where many things go unsaid, is that you didn’t have to remember things about yourself that are too bizarre to imagine. What was never given utterance eventually becomes too nebulous to recall. (Smiley, 1992, p. 305)
Finally, there are some memories that are too painful and the individual simply chooses not to remember. Surprisingly, even very young children can consciously make this decision, as seen in this conversation between a mother and her 5-year-old child talking about going to the wake when the child’s preschool teacher died: M: And what was the wake like? C: Well, it had sadly music and it was really sad to talk about. So I don’t want to talk about it. M: Well, let’s talk about it right now and, if you don’t want to ever talk about it again, that’s fine. C: I didn’t want to talk about it at the wake. (very softly) (Several intervening questions and answers) C: But I don’t want to talk about this cause you’re almost gonna make me cry.
5. A FEMINIST MODEL OF AUTOBIOGRAPHICAL MEMORY
93
M: Okay, I won’t. We won’t talk about it anymore. Obviously, this child found this event emotionally difficult and does not want to bring these emotions back to mind by remembering. Rather then trying to resolve these difficult feelings, the child chooses to silence herself; to simply avoid thinking about or talking about this event. Whereas emotionally difficult experiences are the most likely candidates for self-silencing, especially early in development, there are other reasons to self-silence as well, such as impression management for both self and other. In silencing themselves, individuals lose some of the richness of autobiographical memory. Whether it is whole events, or particular perspectives (e.g., specific kinds of emotional reactions) through self-silencing, the child loses part of her past.
LANGUAGE, MEMORY, VOICE AND SILENCE In this chapter I provided a theoretical framework for understanding the development of autobiographical memory from the feminist perspectives of place and power. These concepts led to a two-dimensional model of autobiography based on voice and silence and self and other. Although still preliminary, this model provides a useful heuristic for understanding how children begin to construct an autobiographical life story in collaboration with their parents. Again, all parents and children engage in both voicing and silencing by both self and other, and memories of any given event will most likely have elements of both of these dimensions to a greater or lesser extent. Moreover, as has been emphasized throughout this chapter, issues of place and power are dynamic; the ways in which parents and children reminisce about their shared past will change as a function of the child’s age and cognitive abilities, as well as the specific events that are under discussion. To summarize, to the extent that parents share power with their children, children are given voice; they are empowered and have authority over their life experiences. Children’s perspectives on what happened and how they felt about it are validated. To the extent that parents exert power over their children by imposing certain stories, children may come to tell these stories but not necessarily from their own subjective perspective. They will not have a sense of ownership of these memories; they are not the authors of their own autobiography. To the extent that parents exert power by refusing to hear particular versions of reality, children will be silenced; their perspective on their past will be invalidated. And to the extent that children silence themselves, and refuse to voice certain experiences or emotions, their perspective will not be socially shared and therefore not open to validation or invalidation. It is essential to point out that in the process of parenting, some imposition of appropriate life stories is necessary. In order to be a socially appropriate member of their culture, individuals must learn to tell certain stories in certain ways. There
94
FIVUSH
are cultural prescriptions on what a life should look like and how people should react to certain kinds of events. Thus, other-voice is a necessary aspect of developing a culturally appropriate life story. Similarly, some perspectives may simply be unacceptable for certain people in certain situations. Thus, other-silence may also be a necessary part of developing an appropriate autobiographical narrative. Finally, it may be more important to tell some experiences in particular ways than others. For experiences that may be self-defining, having one’s own voice may matter a great deal more than for experiences that do not serve a self-defining function. Similarly, experiences that are important for defining the family as a cohesive social group may need to be told from a shared family perspective, and thus parents may be more likely to impose a particular shared perspective on narratives of these events than on other events. The issue may be more about the balance of self versus other and voice versus silence in one’s own autobiographical story. It may also be that how parents accept or impose particular versions of reality on their children is critical. The way in which power is exerted or shared within the developing parent–child relationship may be more important than the number of times the child’s voice is heard versus silenced. This brings things back to the issue of maternal reminiscing style discussed earlier. Mothers who are more elaborative may be helping their children develop voice to a greater extent than mothers who are less elaborative. Elaborative mothers both encourage and evaluate their children’s participation in the co-construction of past experiences, as well as providing more richly embellished accounts of what occurred. In contrast, less elaborative mothers tend to simply repeat the same questions over and over, and repetition of questions is especially likely when the child does not respond with the information the mother was looking for (e.g., Fivush & Fromhoff, 1988; Reese et al., 1993). In this way, elaborative mothers may fall more on the voiced side of the dimension, both self and other, whereas less elaborative mothers may fall more toward the silenced end of the dimension. If so, then we might also speculate that children of more elaborative mothers will have more power over the construction of their own life story. Their memories and their perspective on their past are validated and their voice is heard. Elaborative mothers and children co-construct richly detailed and shared memories of their past together. In contrast, children of less elaborative mothers are not allowed their memories or their perspective on what occurred. Given the role of language in the organization and maintenance of a consciously accessible, socially sharable autobiographical memory (K. Nelson & Fivush, 2002), these children may come to have more fragmented, disorganized memories of their past. They may have more difficulty understanding who they are in relation to their own past experiences or in relation to others with whom they have shared these experiences. Finally, consider the issue of place. As already discussed, an individual’s gender, race, and class help define the kinds of experiences that are validated and invalidated. In illustrating the model in this chapter, I have relied on White middle-class mothers reminiscing with their preschool children. Clearly how
5. A FEMINIST MODEL OF AUTOBIOGRAPHICAL MEMORY
95
power is displayed in the parent–child relationship, and the role of particular kinds of events or perspectives in constructing an autobiographical life narrative will be modulated by both the parent’s and the child’s place in the larger social order. There is a growing body of research attesting to differences in parent–child reminiscing as a function of culture and gender (see Fivush & Haden, 2003, for a review), but much less is known about the role of race and class (but see Wiley, Rose, Burger, & Miller, 1998). These remain important questions for future research. Ultimately, this model provides a way of thinking about the development of autobiographical memory that focuses on authority and ownership rather than on accuracy per se. Cultural, individual, and situational factors privilege some experiences over others. Voice and silence extend previous theorizing about the social construction of autobiographical memory first developed by Katherine Nelson (1989, 1993, 1996, 2001). Especially with the feminist concept of power, this model adds an important dimension to the social construction of autobiographical memory. Memories are not simply jointly constructed; some individuals have more power to guide the narrative in a particular direction than others. How this power is displayed in parent–child reminiscing is a critical question that has implications for the way in which individual children come to understand and “own” their experience. If children are developing voice and silence in parent-guided reminiscing, then there is a need to examine the specific ways in which particular memories are validated or invalidated. The construction of an autobiographical life story is a complex developmental process. Each individual develops a voice, but perhaps more provocative, this voice, by definition, implies silence as well. What can be said about the past must always be placed against the backdrop of what may be remembered but is silenced.
ENDNOTES 1
Although feminist theories tend to view race, class, and gender as socially constructed, it is important to point out that there may be biological predispositions to engage in certain kinds of activities over others, especially for gender (see e.g., Maccoby, 1998, for a full exposition of the biological basis of gendered play behavior). 2 Feminist analyses of place share much in common with Soviet activity theory in general (Gauvain, 2001), and Vygotsky’s (1978) sociocultural developmental theory in particular. Vygotsky’s sociocultural theory of development also posits that individual mind emerges from cultural and contextual interactions. This similarity is not surprising as both Vygotsky’s theory and feminist standpoint theory emerge from the philosophical premises of Hegel’s dialectics. 3 Within psychology, this perspective has been argued most strongly by Gibson (1982), who introduced the concept of “affordances” as behaviors that are elicited by the individual–environment interaction. However, this theory focuses on the interactions between the person and the physical environment, whereas feminist theory focuses on the interaction between the person and the socially constructed human environment.
96
FIVUSH
REFERENCES Alcoff, L. & Potter, E. (1993). Feminist epistemologies. New York: Routledge. Basow, S. A. (1992). Gender stereotypes and roles. Belmont, CA: Brooks-Cole. Bauer, P. (1997). Development of memory in early childhood. In N. Cowan (Ed.). The development of memory in childhood (pp. 83–112). Sussex, England: Psychology Press. Bauer, P. J., Stennes, L., & Haight, J. (2003). Representation of the inner self in autobiography: Women’s and men’s use of internal state language in personal narratives. Memory, 11, 27–42. Belenky, M. F., Clinchy, B. M., Goldberger, N. R., & Tarule, J. M. (1986). Women’s ways of knowing: The development of self, voice and mind. New York: Basic Books. Bordo, S. (1990). Feminism, postmodernism and gender skepticism. In L. Nicholson (Ed.), Feminism/postmodernism (pp. 133–156). New York: Routledge. Bruner, J. S. (1990). Acts of meaning. Cambridge MA: Harvard University Press. Buckner, J. P., & Fivush, R. (1998). Gender and self in children’s autobiographical narratives. Applied Cognitive Psychology, 12, 407–429. Code, L. (1993). Take subjectivity into account. In L. Alcoff & E. Potter (Eds.), Feminist epistemologies (pp. 15–48). New York: Routledge. Connerton, P. (1989). How societies remember. Cambridge, England: Cambridge University Press. Conway, M. A., & Pleydell-Pearce, C. W. (2000). The construction of autobiographical memories in the self-memory system. Psychological Review, 107, 261–288. Conway, M. A., & Rubin, D. C. (1993). The structure of autobiographical memory. In A. F. Collins & S. E. Gathercole & M. A. Conway & P. E. Morris (Eds.), Theories of memory (pp. 103–139). Hillsdale, NJ: Lawrence Erlbaum Associates. Damasio, A. (1999). The feeling of what happens: Body and emotion in the making of consciousness. New York: Harcourt. Davis, P. J. (1990). Gender differences in autobiographical memories for childhood emotional experiences. Journal of Personality and Social Psychology, 76, 498–510. Donald, M.(1991) Origins of the modern mind: Three stages in the evolution of culture and cognition. Cambridge, MA: Harvard University Press. Fischer, A. H. (2000). Gender and emotion: Social psychological perspectives. New York: Cambridge University Press. Fitch, J. (1990). White Oleander. New York: Little, Brown. Fivush, R. (1991). The social construction of personal narratives. Merrill-Palmer Quarterly, 37, 59–82. Fivush, R. (1998). Interest, gender and personal narrative: How children construct self-understanding. In A. Karp, A. Renninger, J. Baumeister, & L. Hoffman (Eds.), Interest and gender in education (pp. 58–73). Kiel, Germany: Institute for Science Education. Fivush, R. (2000). Accuracy, authorship and voice: Feminist approaches to autobiographical memory. In P. Miller & E. Scholnick (Eds.) Toward a feminist developmental psychology (pp. 85–106). NY: Cambridge University Press. Fivush, R. (2001). Owning experience: The development of subjective perspective in autobiographical memory. In C. Moore & K. Lemmon (Eds.), The self in time: Developmental perspectives (p. 35–52). Mahwah, NJ: Lawrence Erlbaum Associates.
5. A FEMINIST MODEL OF AUTOBIOGRAPHICAL MEMORY
97
Fivush, R. (in press). The silenced self: Constructing self from memories spoken and unspoken. In D. Beike, J. Lampinen, & D. Begrand (Eds.). The self and memory. Psychology Press. Fivush, R., & Buckner, J. P. (2000). Gender, sadness and depression: Developmental and socio-cultural perspectives. In A. H. Fischer (Ed.), Gender and emotion: Social psychological perspectives (pp. 232–253). Cambridge, England: Cambridge University Press. Fivush, R., & Buckner, J. P. (2003). Constructing gender and identity through autobiographical narratives. In R. Fivush & C. Haden (Eds.), Autobiographical memory and the construction of a narrative self: Developmental and cultural perspectives (pp. 149–168). Mahwah, NJ: Lawrence Erlbaum Associates. Fivush, R., & Fromhoff, F. (1988). Style and structure in mother–child conversations about the past. Discourse Processes, 11, 337–355. Fivush, R., & Haden, C. (1997). Narrating and representing experience: Preschoolers’ developing autobiographical recounts. In P. van den Broek, P. A. Bauer, & T. Bourg (Eds.), Developmental spans in event comprehension and representation: Bridging fictional and actual events (pp. 169–198). Mahwah, NJ: Lawrence Erlbaum Associates. Fivush, R., & Haden, C. (2003). The construction of an autobiographical self: Developmental and cultural perspectives. Mahwah, NJ: Lawrence Erlbaum Associates. Fivush, R., Haden, C., & Reese, E. (1996). Remembering, recounting and reminiscing: The development of autobiographical memory in social context. In D. Rubin (Ed.), Reconstructing our past: An overview of autobiographical memory (pp. 341–359). New York: Cambridge University Press. Gauvain, M. (2001). The social context of cognitive development. New York: Guilford. Gergen, K. J. (1994). Mind, text and society: Self-memory in social context. In U. Neisser & R. Fivush (Eds.), The remembering self: Construction and accuracy in the life narrative (pp. 78–104). New York: Cambridge University Press. Gibson, E. J. (1982). The concept of affordances in development: The renascence of functionalism. In W. A. Andrews (Ed.), The concept of development: The Minnesota Symposium on Child Psychology (Vol. 15, pp. 55–82). Hillsdale, NJ: Lawrence Erlbaum Associates. Gilligan, C. (1982). In a different voice: Psychological theory and women’s development. Cambridge, MA: Harvard University Press. Griscom, J. L. (1992). Women and power: Definitions, dualism and difference. Psychology of Women Quarterly, 16, 389–414. Habermas, T., & Bluck, S. (2000). Getting a life: The emergence of the life story in adolescence. Psychological Bulletin, 126, 748–769. Haden, C., Haine, R., & Fivush, R. (1997). Developing narrative structure in parent–child conversations about the past. Developmental Psychology, 33, 295–307. Hall, J. A., & Halberstadt, A. G. (1996). Subordination and nonverbal sensitivity: A hypothesis in search of support. In M. R. Walsh (Ed.), Women, men and gender: Ongoing debates (pp. 120–133). New Haven, CT: Yale University Press. Han, J. J., Leichtman, M. D., & Wang, Q. (1998). Autobiographical memory in Korean, Chinese, and American children. Developmental Psychology, 34, 701–713. Haraway, D. (1988). Situated knowledges: The science question in feminism and the privilege of partial perspectives. Feminist Studies, 14, 575–599.
98
FIVUSH
Harding, S. (1993). Rethinking standpoint epistemology: What is “strong” objectivity. . In L. Alcoff & E. Potter (Eds.), Feminist epistemologies (pp. 49–82). New York: Routledge. Heidbreder, E. (1933). Seven psychologies. Englewood Cliffs, NJ: Prentice-Hall. Labov, W. (1982). Speech actions and reaction in personal narrative. In D. Tannen (Ed.), Analyzing discourse: Text and talk (pp. 25–54). Washington, DC: Georgetown University Press. Leichtman, M. D. (2001, July). Preschooler’s memory environments and adults’ recollections in India and the U.S. Paper presented at the International Conference on Memory, Valencia, Spain. Leichtman, M.,Wang, Q., & Pillemer, D. P. (2003). In R. Fivush & C. Haden (Eds.), Autobiographical memory and the construction of a narrative self: Developmental and cultural perspectives (pp. 73–98). Mahwah, NJ: Lawrence Erlbaum Associates. Longino, H. E. (1993). Subjects, power and knowledge: Description and prescription in feminist philosophies of science. In L. Alcoff & E. Potter (Eds.), Feminist epistemologies (pp. 101–120). New York: Routledge. McAdams, D. P. (1992). Unity and purpose in human lives: The emergence of identity as a life story. In R. A. Zucker, A.I. Rabin, J. Aronoff, & S. J. Frank (Eds.). Personality structure in the life course (pp. 323–375). New York: Springer. Maccoby, E. (1998). The two sexes: Growing up apart, coming together. Cambrige, MA: Harvard University Press. McCarthy, M. (1957). Memories of a Catholic girlhood. New York: Harcourt Brace. Middleton, D., & Edwards, D. (1990). Conversational remembering: A social Psychological approach. In D. Middleton & D. Edwards (Eds.), Collective remembering (pp. 23–45). London: Sage. Miller, P. J., & Scholnick, E .K. (2000). Toward a feminist developmental psychology. New York: Routledge. Mullen, M., & Yi, S. (1995). The cultural context of talk about the past: Implications for the development of autobiographical memory. Cognitive Development, 10, 407–419. Nelson, K. (Ed.). (1989). Narratives from the crib. Cambridge, MA: Harvard University Press. Nelson, K. (1993). Explaining the emergence of an autobiographical memory in early childhood. In A. Collins, M. Conway, S. Gathercole, & P. Morris (Eds.), Theories of memory (pp. 355–385). Hillsdale, NJ: Lawrence Erlbaum Associates. Nelson, K. (1996). Language in cognitive development: Emergence of the mediated mind. New York: Cambridge University Press. Nelson, K. (2001). From the experiencing I to the continuing me. In C. Moore & K. Skene (Eds.), The self in time: Developmental issues. Mahwah NJ: Lawrence Erlbaum Associates. Nelson, K., & Fivush, R. (2002). The emergence of autobiographical memory: Submitted manuscript. Nelson, L. H. (1993). Epistemological communities. In L. Alcoff & E. Potter (Eds.), Feminist epistemologies (pp. 121–159). New York: Routledge. Oyserman, D., & Markus, H. (1993). The sociocultural self. In J. Suls (Ed.), Psychological perspectives on the self: The self in social perspective (Vol. 4, pp. 187–220). Hillsdale, NJ: Lawrence Erlbaum Associates.
5. A FEMINIST MODEL OF AUTOBIOGRAPHICAL MEMORY
99
Pasupathi, M. (2001). The social construction of the personal past and its implications for adult development. Psychological Buletin, 127, 651–672. Pasupathi, M., Stallworth, L. M., & Murdoch, K. (1998). How what we tell becomes what we know: Listener effects on speakers’ long-term memory for events. Discourse Processes, 26, 1–25. Pepper, S. C. (1946). World hypotheses: A study in evidence. Berkeley: University of California Press. Peterson, C., & McCabe, A. (1992). Parental styles of narrative elicitation: Effect on children’s narrative structure and content. First Language, 12, 299–321. Pillemer, D., & White, S. H. (1989). Childhood events recalled by children and adults. In H. W. Reese (Ed.), Advances in child development and behavior (Vol. 22). New York: Academic. Proulx, E. A. (1992). Postcards. New York: Scribner’s. Raven, B. H. (1992). A power/ interaction model of interpersonal influence: French and Raven thirty years later. Journal of Social Behavior and Personality, 7, 217–244. Reese, E., Haden, C. A., & Fivush, R. (1993). Mother–child conversations about the past: Relationships of style and memory over time. Cognitive Development, 8, 403–430. Reese, E., Haden, C., & Fivush, R. (1996). Mothers, father, daughters sons: Gender differences in reminiscing. Research on Language and Social Interaction, 29, 27–56. Rogoff, B. (1990). Apprenticeship in thinking. New York: Oxford University Press. Rosser, S. V. & Miller, P. J. (2000). Feminist theories: Implications for developmental psychology. In P. Miller & E. Scholnick (Eds.), Toward a feminist developmental psychology (pp. 11–28). New York: Cambridge University Press. Rubin, D. (1996), Reconstructing our past: An overview of autobiographical memory. New York: Cambridge University Press. Snell, W. E., Jr., Miller, R. S., Belk, S. S., Garcia-Falconi, R., & Hernandez-Sanchez, J. E. (1989). Men’s and women’s emotional self-disclosure: The impact of disclosure recipient, culture and the masculine role. Sex Roles, 21, 467–486. Smiley, J. (1992). A thousand acres. New York: Harper Collins. Tenney, Y. J. (1989). Predicting conversational reports of a personal event. Cognitive Science, 13, 213–233. Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes. Cambridge, MA: Harvard University Press. Wiley, A. R., Rose, A., Burger, L. K., & Miller, P. J. (1998). Constructing autonomous selves through narrative practice: A Comparative study of working class and middle class families. Child Development, 69, 833–847. Wilson, B. (1998). Blue windows: A Christian Science childhood. New York: Picador. Yoder, J. D., & Kahn, A. S. (1992). Toward a feminist understanding of woemn and power. Psychology of Women Quarterly, 16, 381–388.
6 Developments in Early Memory: Multiple Mediators of Foundational Processes Patricia J. Bauer and Melissa M. Burch University of Minnesota
During her career, Katherine Nelson has profoundly influenced research and scholarship in many different domains of developmental psychology in general, and cognitive and language development in particular. Indeed, in the two main areas of work in our laboratory—memory development and conceptual development—Katherine’s work has exerted significant shaping force and influence. In the domain of memory in particular, we feel the proximal and distal effects of Katherine’s work on a daily basis. Her 1986 edited volume, Event Knowledge: Structure and Function in Development, made abundantly clear that 3-year-old children have well-ordered representations of everyday events and routines (e.g., going to fast food restaurants). The demonstration sparked systematic examination of preschool-age children’s memory in natural contexts. Simultaneously, it compelled investigations of children younger than age 3: If 3-year-olds already are competent mnemonists, then the capacity to construct event representations must have developed earlier. It was in search of the emergence of the ability to remember the past that our laboratory ventured into the land of the pre- and early verbal child. Throughout most of the life of our laboratory, the focus of research has been on describing normative trends in the development of memory in the first two years of 101
102
BAUER AND BURCH
life, and exploring experimentally-controlled determinants of whether and for how long young children remember specific laboratory events encoded and tested in the context of the elicited-imitation paradigm (i.e., in elicited imitation, props are used to produce single actions or multistep sequences that either immediately, after a delay, or both, the child is encouraged to imitate; the task is viewed as a nonverbal analogue to verbal report). True to the promise of Katherine Nelson’s work with older children, our efforts to explore memory in younger children were richly rewarded. The work made apparent that young children not only remember past events, but their memories are well organized: causal, temporal, and goal-based organization is readily apparent (Bauer, 1992; Bauer & Dow, 1994; Bauer & Travis, 1993; Travis, 1997). At age 9 months, the capacity to retain organized memories over extended delays is seen in roughly 50% of infants tested (e.g., Bauer, Wiebe, Waters, & Bangston, 2001; Carver & Bauer, 1999). By 13 months, long-term recall is reliably observed (e.g., Bauer & Hertsgaard, 1993), and by 20 months, it is both reliable and robust (Bauer, Wenner, Dropik, & Wewerka, 2000). As it is for older children, recall is influenced by the organization of event representations (e.g., Bauer & Dow, 1994; Bauer, Hertsgaard, Dropik, & Daly, 1998) and by the availability of verbal reminders at the time of recall (Bauer, Hertsgaard, & Wewerka, 1995). In short, normative trends in early memory development are relatively well established, as are some of the experimentally-controlled determinants of early memory. Given that we were starting almost from scratch, it was necessary and appropriate that the early work with children younger than age 3 be aimed at describing normative trends. However, virtually exclusive focus on normative developmental trends, and experimentally-controlled determinants thereof, came at the expense of recognition of the social and emotional context in which early mnemonic development takes place. As a result, little is known about how mnemonic processes become “mediated” by the context in which they develop, a process that Katherine Nelson has shown to be significant (Nelson, 1996). In this chapter we explore the mediational impact of one context in particular, namely, that of mother–child interaction. Focus on patterns of mother–child interaction is not new either generally or in the domain of memory. However, the potential significance of mother–child interactions for very early memory development is only beginning to be explored. In this chapter, we take two steps toward amelioration of our ignorance of the influence of maternal behavior on very early memory development. We focus first on interactions between mothers and their 24-month-old children in the context of laboratory memory tasks designed to impose different cognitive demands. We attend to how maternal verbal scaffolding affects children’s nonverbal memory performance, as well as how characteristics of the children, specifically their temperaments and their levels of language proficiency, are related to the behaviors exhibited by the children’s mothers. We then focus on the apparent extension from one mnemonic context to another, of maternal influence on verbal memory expression in the preschool years. In each case, the goal is to understand how social, emotional, and cognitive factors jointly shape children’s mnemonic performance.
6. MEDIATING YOUNG MINDS
103
MATERNAL VERBAL SCAFFOLDING OF 24-MONTH-OLD CHILDREN’S MEMORY OF EVENT SEQUENCES EXPRESSED NONVERBALLY Some of the earliest work on the social context in which preschoolers’ memories develop was conducted by Susan Engel, under the mentorship of Katherine Nelson. Engel (1986) found that the way mothers structured conversations with their preschool-age children affected children’s narratives about the past. This finding has since been replicated and extended: Findings of relations between maternal verbal behavior and preschool-age children’s autobiographical memory narratives are robust.
Relations Between Maternal Language and Preschool-Age Children’s Memory Narratives The literature on preschool-age children reveals relations between variability in maternal language in the context of joint reminiscing and children’s autobiographical or personal memory narratives. In brief, researchers have observed two “styles” of conversation. Mothers who frequently engage in conversations about the past, provide rich descriptive information about previous experiences, and invite their children to “join in” on the construction of stories about the past, are said to use an elaborative style. In contrast, mothers who provide fewer details about past experiences and instead pose specific questions to their children (e.g., “What was the name of the restaurant where we had breakfast?”) are said to use a repetitive, or low elaborative style. Maternal verbal stylistic differences have implications for children’s autobiographical memory reports. Specifically, children of mothers whose language more closely approximates the elaborative style report more about past events than children of mothers whose language more closely resembles the repetitive style (e.g., Fivush & Fromhoff, 1988; Hudson, 1990; Tessler & Nelson, 1994). Relations between maternal language style and children’s memory narratives are observed concurrently and over time. For example, Reese, Haden, and Fivush (1993) found that maternal use of a more elaborative style when children were 40 and 46 months facilitated children’s independent narrative accounts at 58 and 70 months of age.
Maternal Language Variability with Younger Children in the Context of Elicited Imitation Whereas most of the work on relations between maternal language style and children’s memory narratives has been conducted with older children, a small number of studies have been conducted with pre- and early verbal children. For example, Hudson (1990) reported effects of differential degrees of maternal verbal elabora-
104
BAUER AND BURCH
tion on 24- to 30-month-olds’ participation in memory conversations. Farrant and Reese (2000) extended this finding to children as young as 19 months. In Burch and Bauer (2003), we sought to expand this small literature into the context of the elicited-imitation paradigm. Examining possible relations between maternal verbal behavior and children’s performance in the elicited-imitation task allows us to forge a link between the literatures on verbal and nonverbal recall. It also permits us to examine a naturally-occurring behavior—that of variability in maternal language—under the conditions of greater experimental control afforded by a specific laboratory task. In addition to examining relations between maternal verbal behavior and young children’s nonverbal memory performance, we also sought to inform the question of how maternal verbal mediation of mnemonic processes is related to characteristics of the child. In other words, how does what the child brings to the mnemonic context relate to what the mother brings? Farrant and Reese (2000) found that children’s receptive and expressive language skills were related to mother–child conversational variables as early as 19 and 25 months. To examine the possibility of similar relations between language and maternal verbal behavior in the context of a nonverbal task, in Burch and Bauer (2003), we obtained measures of expressive language abilities of the children. We also investigated possible relations between maternal verbal behavior and another salient characteristic of children, namely, their temperaments (i.e., constitutionally based patterns of responding to environmental stimuli: e.g., Gunnar, 1990; Rothbart, Ahadi, & Hershey, 1994; Rothbart & Bates, 1998). In previous research, in the context of elicited-imitation tasks administered by professional experimenters, we have found relations between children’s temperament characteristics and nonverbal recall. Specifically, at age 9 and 13 months, high levels of positive affectivity are related to long-term recall (Bauer, Burch, & Kleinknecht, 2002; Kleinknecht, Rademacher, & Bauer, 2002). At age 20 months, high levels of attentional control are related to long-term recall (Kleinknecht et al., 2002; see Kleinknecht et al. for discussion of the bases for relations between temperament and memory, as well as discussion of the findings). In Burch and Bauer (2003), we tested whether characteristics of children’s temperament are directly related to task performance, as well as whether they are systematically related to variability in maternal language. Study Design. Typically, in the elicited- or deferred-imitation paradigm, to-be-remembered events are modeled by an experimenter and children are encouraged to imitate. However, in Burch and Bauer (2003), to permit examination of possible effects of variability in maternal language on young children’s recall, mothers, rather than experimenters, tested the children. Sixteen mothers and their 24-month-old children participated. Mothers first gave their children a baseline period during which the children were to explore the event-related props. Mothers then demonstrated the test sequences for their children and then elicited their children’s imitation (i.e., immediate recall) of the sequences. One week later, the dyads
6. MEDIATING YOUNG MINDS
105
returned to the laboratory and mothers tested their children’s delayed recall. To prepare them for their task, mothers were exposed to a silent videotape on which the sequences were demonstrated. Beyond that they were to “talk naturally,” mothers were given no instructions regarding how to verbally communicate as they demonstrated and then tested their children’s memories for the sequences. To examine effects of cognitive demand on maternal language and children’s recall, the dyads were presented with sequences of three different lengths: 4 steps, 5 steps, and 6 steps. Based on previous related research, we expected 4-step sequences to be relatively easy for the 24-month-olds (Bauer et al., 2000) and 5-step sequences to be at an optimal level of challenge (Bauer & Travis, 1993). We expected 6-step sequences to represent a level of difficulty at the boundary of or perhaps even beyond the 24-month-olds’ “zone of proximal development,” because not until 30 months do children typically perform at above-chance levels on 6-step sequences (Bauer, Dow, Bittinger, & Wenner, 1998; Bauer & Fivush, 1992). In addition to data on children’s immediate and 1-week delayed recall performance, to assess children’s expressive language abilities, we collected MacArthur Communicative Development Inventories for Toddlers questionnaires (Fenson et al., 1994). The MacArthur is a parent-report instrument on which parents are asked to indicate which of 680 words their children produce. Finally, to assess children’s temperament characteristics, we collected Toddler Behavior Assessment Questionnaires (TBAQ; Goldsmith, 1996) on the children. The TBAQ is a parent-report instrument consisting of 111 items referring to a variety of scenarios describing situations in which young children might be engaged. An example question is, “When your child wanted to eat something sweet before dinner was finished but did not get it, how often did s/he protest by crying loudly?” Parents were asked to indicate how often their children had such a reaction within the last month (scales range from 1, “never,” to 7, “always,” with the opportunity to indicate that the situation does not apply). The 111 items on the questionnaire cluster into five independent subscales: Activity Level, Anger Proneness, Interest and Persistence, Pleasure, and Social Fearfulness. We focus our discussion on the TBAQ subscales of Interest and Persistence and Activity Level because, as observed in Kleinknecht et al. (2002), they were the most predictive (details on the small number of relations involving the other subscales of the TBAQ are available in Burch & Bauer, 2003). Maternal Language Variability. We coded each of the mothers’ utterances into one of the seven mutually exclusive and exhaustive categories described in Table 6.1. The first five categories—namely, elaborations, repetitions, affirmations, negations, and off-task—are those used in studies of maternal style variability in older children (e.g., Reese et al., 1993). The categories of deflections and regulations were added to capture interactions that occurred as a result of the context of the prop-supported elicited-imitation task. Because very few negations were observed, they are not discussed. Although regulations of children’s behavior were
106
BAUER AND BURCH TABLE 6.1
Categories of Maternal Utterances Used with 24-Month-Old Children in the Context of Elicited Imitation Elaborations:
Utterances that served to introduce an object or event or that provided additional information related to an object or event previously introduced. For example, “look at the farm stuff,” and “where does the horse go on the farm?”
Repetitions:
Utterances that were a verbatim or gist repetition of a prior utterance. For example, “what do we do on the farm?”, after having just said “do we do something on the farm?”
Affirmations:
Utterances that served to affirm a child’s behavior or previous utterance. For example, “good job,” or “that’s right.”
Negations:
Utterances that served to deny a child’s behavior or previous utterance. For example, “no, not like that.”
Off-Task:
Utterances not directed to the task, including such things as conversation with the professional staff member directing the sessions, and comments about activities outside the laboratory.
Deflections:
Utterances that served to hold a place in the conversation (without providing any new information) or that directed the turn to the child. For example, “that’s a farm (elaboration), isn’t it (deflection),” and “look at this (regulation)—wow! (deflection).”
Regulations:
Utterances the purpose of which seemingly was to direct the child’s behavior or attention to the task. For example, “sit down,” and “watch me.”
not infrequent, there were few correlations with this category, and for this reason, it is not discussed. And, because off-task utterances are not related to the task, they are not discussed (see Burch & Bauer, 2003, for additional details). Mothers exhibited variability in their language both when teaching the sequences and when testing their 24-month-old children’s memories. As illustrated in Table 6.2, the number of utterances in each category varied considerably, as did the total number of category tokens produced. The number of utterances also varied across sequence step lengths, such that across categories, mothers produced substantially more total category tokens on the 6-step sequences, relative to the 5and 4-step sequences, which did not differ from one another. Nevertheless, mothers’ levels of production of the different utterance types were correlated across levels of difficulty of the events. For example, the number of elaborations used as mothers elicited their children’s immediate recall of the 4-step sequences was correlated with the number of elaborations used to elicit immediate recall of the 5-step and 6-step sequences (rs = .80 and .58, respectively; correlations have 14 degrees of freedom; unless noted, all p values are < .05); the number of elaborations used to elicit recall of the 5-step and 6-step sequences also was correlated (r = .85). In
6. MEDIATING YOUNG MINDS
107
TABLE 6.2 Maternal Language Variability with 24-Month-Old Children in the Context of Elicited Imitation Step Length Category of Maternal Utterance
M
Elaborations
60.90
Repetitions
8.30
Affirmations
34.89
Deflections Total
6-Step (SD)
5-Step
4-Step
M
(SD)
M
(22.42)
38.78
(9.32)
36.73
(10.86)
(5.19)
5.43
(3.97)
4.48
(3.76)
(18.57)
23.20
(12.80)
27.58
(14.54)
(SD)
17.71
(8.03)
14.40
(7.04)
15.06
(5.05)
168.95
(41.70)
124.29
(26.19)
125.73
(36.90)
other words, although mothers varied in the number of tokens produced across levels of difficulty of the task, they nevertheless exhibited stability across sequence lengths. Thus, even though Burch and Bauer (2003) used a nonverbal task with 24-month-olds, the same type of maternal language variability that has been reported in the literature on verbal narrative production was observed with preschool-age and older children.
Relations Between Mothers’ Language and Children’s Recall Performance, Temperament Characteristics, and Language Abilities Having observed variability in measures related to maternal style, we explored whether it was related to children’s immediate or 1-week delayed recall performance. As becomes apparent, the patterns of relations differed as a function of the level of difficulty of the to-be-remembered events. Because relations on the 5-step sequences were a mixture of the patterns observed on the simpler and more difficult sequence lengths, for convenience, we present the relations for the 4-step and 6-step sequence lengths only (details on relations on the 5-step sequences are available in Burch & Bauer, 2003). Four-Step Sequences. On the relatively “easy for age” 4-step sequences, maternal verbal behavior during the spontaneous, child-controlled baseline period was related to the total number of target actions (including repetitions) that the children produced both immediately and after the delay. Specifically, maternal use of elaborations (r = .72), affirmations of their children’s behaviors (r = .62), and overall talkativeness (as measured by the total number of category tokens produced; r = .73) as their children manipulated the event-related objects spontaneously all were correlated with the total number of target actions children produced
108
BAUER AND BURCH
in immediate recall. The same variables were correlated with the total number of target actions produced at delayed recall as well (rs = .51, .85, and .79, respectively). Maternal verbal behavior was also related to children’s engagement during immediate recall of the 4-step sequences. Mothers who produced relatively more verbal elaborations during the immediate recall period had children who produced both a larger number of total target actions and a larger number of different target actions (both rs = .50). Production of a large number of total target actions by the children also was related to maternal repetition of utterances and to maternal affirmations (rs = .55 and .69, respectively). Maternal affirmations during the immediate recall period also had a cross-lagged relation to children’s performance after the delay: Mothers who affirmed their children during immediate recall had children who were more engaged in the task at delayed recall, as measured by the total number of target actions produced (r = .58). The children also did their part to keep the system going: Children who produced a larger number of target actions during immediate recall had mothers who produced both a large number of affirmations and more total category tokens during delayed recall (rs = .67 and .55, respectively). Finally, maternal affirmations during the delayed-recall period were related to production of individual target actions by the children (r = .64). Two patterns in the data from the 4-step sequences are especially noteworthy. First, on these easy sequences, maternal behavior was not in response to the children’s memory performance, per se: Systematic variability in maternal language was apparent in the baseline phase, before any memory behavior was observed. Second, maternal verbal behavior was related to children’s levels of engagement in the task (i.e., to the total number of individual target actions produced, including repetitions), but not to the exhaustiveness of children’s recall or to its organization (as measured by production of different target actions and ordered pairs of actions, respectively). That is, there was only a single correlation between maternal verbal behavior and the diversity of target actions that the children produced; there were no correlations with ordered reproduction of the sequences. These patterns suggest that on event sequences that are well within a child’s developmental level, the mother–child interaction is not related to memory, but to the dyad’s pattern of interaction in the context of new materials or experiences: Mothers who are more verbally engaged have children who are more nonverbally engaged. Examination of the pattern of relations between maternal verbal behavior and children’s temperament characteristics revealed that mothers were especially verbally engaged with children rated as having high levels of Interest and Persistence. Similarly, mothers were especially verbally engaged with children with higher levels of reported productive vocabulary. With respect to the relations with temperament characteristics, in the context of the simplest sequences, mothers who rated their children high on the Interest and Persistence subscale of the TBAQ produced more elaborations (r = .58), more repetitions (r = .55), and more total category tokens (r = .66) during the baseline phase. At immediate recall, they produced more elaborations (r = .49), more repetitions (r = .52), more affirmations (r = .62),
6. MEDIATING YOUNG MINDS
109
and more total category tokens (r = .53). At delayed recall, they produced more affirmations (r = .63). Thus, children who were perceived as typically showing evidence of interest and persistence received in the elicited-imitation context more verbal scaffolding during baseline and immediate recall, and to a lesser extent, during delayed recall. In the context of the 4-step sequences, relations between maternal verbal behavior and children’s temperament characteristics were confined to the Interest and Persistence subscale: Maternal verbal behavior was unrelated to scores on the Activity Level subscale of the TBAQ. Whereas maternal verbal behavior varied as a function of perceived levels of Interest and Persistence among the children, there were not direct relations between the temperament characteristics of Interest and Persistence and children’s performance on the simplest sequences. The only direct relations between children’s temperament characteristics and their performance on the 4-step sequences were with Activity Level. High scores on the Activity Level subscale were correlated with higher levels of production of the individual actions of the sequences in baseline, immediate recall, and delayed recall (rs = .42, .58, and .57, respectively). Thus, a characteristically high level of activity was perhaps beneficial to engagement in the imitation task and this “carried over” to recall performance when task demands were low. With respect to relations with children’s productive vocabularies, there was a positive correlation between expressive vocabulary and children’s immediate ordered recall of the 4-step sequences (r = .50, p = .05). Thus, children with larger productive vocabularies showed more coherent memory performance immediately following modeling. In addition, there were relations between children’s productive vocabularies and maternal verbal behavior. Mothers who reported that their children had larger productive vocabularies produced more elaborations as they modeled the event sequences at the first session (r = .60). Maternal verbal behavior while the children were engaged in immediate and delayed recall of the 4-step event sequences did not differ as a function of children’s language skills. Thus, on the simplest sequences, maternal behavior during initial teaching of the events (modeling at Session 1) varied as a function of their children’s productive language skills. Children with higher levels of language competence received more elaborative encoding support. Consideration of the relations observed among maternal language, children’s recall of the 4-step sequences, their temperament characteristics, and their levels of language development, suggests that children whose parents perceived them as being high in Interest and Persistence and as having large vocabularies were treated to high levels of maternal language during both baseline and recall. In turn, the children were highly engaged in the task, both immediately and after the delay. The association between levels of Interest and Persistence and children’s performance is best characterized as indirect. Rather than between the temperament characteristics of the children and recall performance, the relations were between temperament characteristics of the children and maternal behavior, and between
110
BAUER AND BURCH
maternal behavior and children’s performance. We believe that this was observed because in the case of a relatively low level of challenge such as that posed by 4-step sequences, high levels of self-generated and self-regulated attention are not necessary to ensure engagement in the task. Instead, characteristically high levels of activity are sufficient to see the children through. Nevertheless, even in an “easy” context, control of attention and persistence are beneficial because of the high level of supportive maternal language they engender. In the case of children’s language, there were both direct and indirect relations between reported expressive vocabulary and recall of the 4-step sequences. Thus, higher levels of productive language are beneficial in and of themselves, and they are associated with high level of supportive maternal language. Six-Step Sequences. In contrast to the 4-step sequences, on the very difficult 6-step sequences, maternal verbal behavior in the baseline phase was largely unrelated to children’s behavior in the immediate recall period. The only relation observed was between total category tokens during baseline and children’s engagement at immediate recall, as measured by the total number of target actions produced (r = .49, p = .05). Maternal verbal behavior at the first session was related to children’s temperament characteristics. As they modeled the 6-step sequences, mothers of children with lower scores on the Activity Level subscale produced more deflections (r = –.50) and more total category tokens (r = –.56). One interpretation of these relations is that mothers whose children do not typically spontaneously show high levels of activity saw greater need to engage their children in the more difficult sequences. The “ploy” was effective: Children whose mothers frequently deflected the turn to them as they modeled the event sequences during the first session produced a larger number of different target actions during delayed recall (r = .55). Finally, children with higher levels of reported productive vocabulary received fewer regulations during modeling of the longest sequences (r = -.67). This was the only relation between children’s vocabulary and maternal behavior on 6-step sequences. Whereas there were few concurrent relations between maternal and children’s behavior during the first session, there were a number of concurrent relations during the second session. Specifically, mothers who during delayed recall produced more elaborations, more repetitions, and more affirmations had children who were more engaged in the difficult task, as measured by the total number of target actions produced (rs = .59, .54, and .80, respectively). Maternal verbal behaviors were delivered without regard for children’s temperament or language characteristics. The only relation between children’s temperament and maternal behavior at Session 2 was that mothers of children rated high on the Interest and Persistence subscale provided more affirmations during delayed recall (r = .55). There were no relations between children’s vocabulary and maternal behavior at Session 2. The sparse pattern of correlation with temperament characteristics and vocabulary levels on the more challenging sequences relative to the simpler, 4-step sequences, suggests that as the level of challenge increased, mothers responded more to the “situational” factor of their children’s needs for more scaffolding and
6. MEDIATING YOUNG MINDS
111
support, and less to the “dispositional” factors of whether, under typical circumstances, their children show a high level of interest and persistence, and whether they demonstrate proficiency with language. Indeed, as noted earlier, across the sample, mothers provided more total category tokens on the more difficult 6-step sequences, relative to the easier 5- and 4-step sequences. Although maternal verbal behavior on the most challenging sequences did not vary as a function of characteristics of the children, one characteristic of the children, namely, Interest and Persistence, was related to performance on the most difficult tasks. On the 6-step sequences, higher scores on the Interest and Persistence subscale were associated with high levels of delayed recall of the temporal order of the sequences (r = .50). There were no relations between scores on the Activity Level subscale and performance on the 6-step sequences. Thus, against a backdrop of uniformly high levels of maternal verbal scaffolding (i.e., provided regardless of children’s temperament), children with characteristically higher levels of self-regulation of attention and interest had an advantage over their less self-regulated peers, perhaps because they were better able to capitalize on the scaffolding provided by their mothers. A similar pattern was not apparent with language. Conclusions and Implications. What do these relations imply about the interaction of social, emotional, and cognitive factors in the domain of early memory development? First, they make clear that by age 24 months, the interaction of factors that heretofore has been observed in the domain of verbal expression of memory extends into the domain of nonverbal expression of memory: Variability in maternal verbal behavior relates to differences in children’s nonverbal recall performance. Second, the findings imply that that there are multiple influences on the behavior of both members of the mother–child dyad. Children’s behavior varied as a function of their own temperament characteristics and their mothers’ verbal behaviors; maternal verbal behaviors varied as a function of children’s temperament characteristics and children’s language. The patterns of interaction among the factors further varied as a function of the level of difficulty of the to-be-remembered material. When the material was less challenging, mothers seemed to respond more to the “dispositional” characteristics of their children’s typical responses and behaviors and their expressive language. When the material was more challenging, mothers seemed to respond more to the “situational” characteristics of the task demands: They increased their levels of verbal elaboration, regardless of the characteristics of their children. Third, although on the more difficult sequences there were not direct relations between children’s temperaments and their mothers’ verbal behavior, temperament characteristics nevertheless mattered: Children with characteristically higher levels of Interest and Persistence performed at higher levels on the most difficult sequences, likely because they were able to take advantage of the greater verbal scaffolding provided by their mothers. Thus, what the children brought to the interactional context influenced the “mediation” they received (the level of maternal verbal scaffolding), as well as their ability to capitalize on it.
112
BAUER AND BURCH
RELATIONS BETWEEN MOTHERS’ VERBAL AND CHILDREN’S MNEMONIC BEHAVIORS ACROSS CONTEXTS The concept of a “mediated mind” implies that the social, emotional, and cognitive factors that influence early development have an impact that endures beyond the point in time at which their interactions are observed. In the literature on autobiographical memory development, it is clear that early maternal verbal style influences extend over time: There are both concurrent and cross-lagged relations between maternal verbal behavior and children’s autobiographical narratives. Another implication of the internalization model is that influences on early development will have an impact that extends across contexts. Although not entirely ignored, examinations of maternal style influences across contexts have received less attention than those across time. In this section, we address the relative paucity of relevant research by examining relations between maternal verbal behavior in the mother–child interview context and children’s verbal behavior in an independent mnemonic task. To date, the question of whether the “tutoring” that children receive in mother– child contexts has effects beyond them has been addressed in two ways. First, researchers have asked whether maternal style is consistent across contexts. There is little evidence of such relations. For example, Haden and Fivush (1996) found low correlation between the maternal support provided in mother–child memory conversations and in free play. Similarly, Haden, Reese, and Fivush (1996) found little consistency across mother–child memory conversations and reading of familiar and unfamiliar books. The second approach to the question of whether there are cross-context relations with maternal style variables has been to ask whether maternal style in the context of mother–child reminiscence relates to young children’s independent memory narratives in the context of experimenter–child interviews (e.g., Hudson, 1990). For example, Peterson and McCabe (1994) examined parents’ use of different question types and statements and children’s provision of when and where information in conversations with an experimenter. Consistent with expectations derived from an internalization model, they found predictive relations between parents’ use of Wh-context questions (e.g., “Where did we go yesterday?”) and Yes/no context questions (e.g., “Did we go to McDonald’s?”) at 26–31 and 32–37 months and the amount of when and where information children provided to an experimenter at 38–43 months. However, even at 38–43 months, few concurrent relations were observed, and there were neither concurrent nor predictive relations with other question types or with parental statements. Information about narrative dimensions other than when and where (i.e., who, what, why, and how) was not provided in the report. In a longitudinal study, Farrant and Reese (2000) examined children’s productions across narrative dimensions. They found modest concur-
6. MEDIATING YOUNG MINDS
113
rent correlations between the number of content-filled contributions that children 32 months of age made in independent (experimenter–child) interviews and the number made in shared (mother–child) interviews. However, at the preceding and following time points, namely, 25 months and 40 months (respectively), independent and shared memory contributions were not related. Thus, the extent to which the “mediation” of mind that children experience in the context of mother–child conversations about past events extends to independent mnemonic narrative contexts remains in question. The opportunity to observe relations between mothers’ verbal behaviors and children’s verbal mnemonic behaviors across contexts was afforded by a long-term longitudinal study in our laboratory. As described in Bauer et al. (2000), children originally were enrolled in the study at ages 13, 16, or 20 months, at which time they were exposed to specific, novel, multistep event sequences. The children’s nonverbal recall of the sequences was tested 1, 3, 6, 9, or 12 months later. At age 36–52 months, some of the children returned to the laboratory and participated in a follow-up study in which we examined the verbal accessibility of the children’s memories for the event sequences. For each of 12 event sequences in turn (three that the children originally had only watched; three that they had been permitted to imitate one time prior to imposition of the 1–12 month delays; three that were novel, control sequences at the time delayed recall was tested; and three that were novel, control sequences at age 36–52 months), the props for the sequence were placed on the table in front of the children and the children were encouraged to talk about the props and the event that the props could be used to create. Once they had described the event or appeared to have exhausted the effort, the props were given to the children and they were encouraged to enact the event (see Bauer, Kroupina, Schwade, Dropik, & Wewerka, 1998, and Bauer, Wenner, & Kroupina, 2002, for procedural details). After all 12 events had been tested, the children participated in a mother–child interview. In the interview, children’s mothers asked their children about approximately six events that the dyad had shared together. Thus, in a single session, children participated in an experimenter–child interview in which children’s verbal reports about specific laboratory sequences were “scaffolded” by the props used in the events, and they participated in a mother–child interview in which scaffolding was provided by the children’s mothers. To address the question of whether verbal mnemonic behavior in these two very different contexts is related, we developed a coding scheme that could be used to capture the content of the children’s unique verbal contributions equally well in both contexts. In addition, the coding scheme captured all of the traditional narrative dimensions, except how. The coding categories are described in Table 6.3. For a subset of the children, all of whom had been 20 months of age at the time of original experience of the events, the children’s utterances were coded into one of the seven mutually exclusive categories indicated. The mothers’ utterances were coded into the categories described in Table 6.1.
114
BAUER AND BURCH TABLE 6.3 Categories of Utterances Used by 3-Year-Old Children
Labels:
Utterances that provided the label of the event sequence or the topic of the event discussed. For example, “that’s a gong” and “we went to visit Grandpa,” respectively.
Objects:
References to the objects included in the event sequence or that were present in the event discussed. For example, “the bell” and “he had a ball,” respectively.
Activities:
Utterances that described the actions in the event sequence or the event discussed. For example, “hang up the bell” and “we kicked the ball,” respectively.
Sequences: References to the order in which actions in the event sequences should be completed or in which they unfolded in the event discussed. For example, “then ring it” and “first we went to the party,” respectively. People:
Mention of participants in the event sequence or the event discussed. For example, “I rang it,” and “Billy was there,” respectively.
Locations:
References to the relative location of an object in the event sequence or to rooms, cities, states, etc., in the context of the event discussed. For example, “it goes on top” and “we went to San Francisco,” respectively.
Reasons:
Provision of causal connections that explain or permit actions in the event sequence or the event discussion. For example “hang it up so I can ring it” and “we went outside because I wanted to play,” respectively.
Relations Between Maternal Verbal Behavior and Children’s Verbal Behavior in the Mother–Child Interview Context Because the coding scheme to capture the children’s verbal behavior was unique to this analysis, we first asked whether using it, we replicated the findings of relations between maternal verbal style variables and children’s autobiographical memory narrative productions (e.g., Reese et al., 1993). By every metric, we did. First, not surprisingly, given the robustness of the phenomenon, we observed in our sample the same type of variability in maternal verbal behavior as has been observed in other samples of mothers and their preschool-age children. As reflected in Table 6.4, panel A, there was considerable variability among mothers in their verbal productions in virtually every category. Second, as has been observed in other similar samples, maternal verbal behavior was highly intercorrelated (see Table 6.5). Most notably, mothers who produced a large number of elaborations also produced large numbers of affirmations of their children’s behaviors, they often repeated their children’s utterances, deflected the conversational turn back to their children, and they talked more overall.
6. MEDIATING YOUNG MINDS
115
TABLE 6.4 Means (and Standard Deviations) for Maternal Verbal Productions and Children’s Verbal Productions in the Context of Mother–Child Interviews Panel A: Maternal Verbal Category Elaborations
M
(SD)
15.49
(6.05)
Ratio
3.23
(1.46)
Affirmations
7.01
(3.18)
Repetitions
5.54
(2.88)
Deflections
2.23
(1.45)
Negations
0.71
(0.62)
Regulations
0.51
(0.74)
Associations
0.32
(0.52)
33.09
(12.21)
Labels
0.16
(0.19)
Objects
1.79
(0.96)
Activities
1.24
(0.81)
Sequences
0.10
(0.28)
Reasons
0.06
(0.13)
People
1.88
(1.01)
Locations
0.52
(0.34)
Total talk
5.78
(2.70)
Total talk Panel B: Children’s Verbal Production
Third, variability in maternal verbal behavior was related to variability in children’s behavior. An inspection of Table 6.4, panel B makes it apparent that there was variability in children’s behavior. Although children’s levels of production of tokens in some of the categories (i.e., event labels, sequence, and reasons) was quite low, in the categories of objects, activities, and people, it was higher and variable; there was considerable variability in children’s total talkativeness in the context of the interviews. As reflected in Table 6.6, variability in the categories of objects, activities, people, and total talk among the children in the sample was related to variability in maternal behavior in production of, in particular, elaborations, affirmations, and total talk, and to a lesser extent, repetitions and deflections (the category of regulations is not represented in the table because there were no relations involving this category in this or subsequent analyses). Thus, although the coding categories used to capture the verbal behavior of the children in the sample were unique to this study, as has been observed in other samples, variabil-
116
BAUER AND BURCH TABLE 6.5
Intercorrelations Between Categories of Maternal Verbal Production in the Context of Interviews with 3-Year-Old Children Maternal Verbal Catetory
2
1. Elaborations
3
4
.81*** .57**
5
6
7
8
9
.68***
.95***
.52*
.81***
2. Ratio 3. Affirmations
.50*
4. Repetitions
.47*
5. Deflections
.73*** .45*
.70***
6. Negations 7. Regulations 8. Associations 9. Total Talk Note: Only statistically significant correlations are represented: *p < .05. **p < .01. ***p < .001.
TABLE 6.6 Relations Between Maternal Verbal Productions and Children’s Verbal Productions in the Context of Mother–Child Interviews Children’s Verbal Production (Mother–Child Interview) Maternal Verbal Category
Labels Objects Activities Seq.
Elaborations
.58**
.44*
Affirmations
.81***
.53**
Reason People
.59**
Repetitions Deflections
Location
Total Talk
.70***
.67***
.76***
.81***
.62***
.50*
.44*
.39*
Negations Associations Total Talk
.64***
.49*
Note: Only statistically significant correlations are represented: *p < .05. **p < .01. ***p < .001.
.74***
.71***
6. MEDIATING YOUNG MINDS
117
ity in children’s productions was systematically related to maternal verbal behavior. The strongest and most consistent pattern of relations was observed with the maternal verbal categories of elaborations, affirmations, and total talk.
Relations Across Experimenter–Child and Mother–Child Contexts There are at least three ways to evaluate the question of whether children are consistent in their verbal productions across different contexts. The first is to evaluate children’s mean levels of performance in the two contexts. The second is to evaluate the degree of correlation in children’s behavior in the two contexts. The third is to evaluate the degree of correlation between children’s behavior in the experimenter–child interview and mothers’ verbal behaviors in the mother–child interview. We approach the question in each way, in turn. However, before turning to questions of consistency across contexts, we address the question of degree of consistency within the experimenter–child context. Correlations Across Event Types Within the Experimenter–Child Context. Although within the experimenter–child context, children encountered events at three different levels of familiarity (“old” events experienced three times at 20 months of age—events previously imitated and events previously only watched; events that were new and served as control events at Session 4; events that were new and served as control events at Session 5), they nevertheless were consistent in their levels of production of different category tokens. Specifically, there were significant within-task correlations in five of the seven categories (labels, activities, sequence, people, and locations); children’s total production also was correlated across levels of event familiarity (details are available from the authors). By and large then, in the experimenter–child interview context, children’s verbal productions were internally consistent. Moreover, with the exceptions that children made more references to objects and locations before the props were provided, relative to after they were provided, the number of references that children made within each category was comparable before and after the props were given to them. For this reason, and because utterances indicative of memory are observed only before the props are provided (Bauer, Kroupina et al., 1998; Bauer, Wenner et al., 2002), subsequent analyses and discussion focus on children’s behaviors before the props were delivered to them. Mean Levels of Verbal Production. We conducted preliminary analyses to determine whether children’s verbal productions differed on the events that they had been permitted to imitate and the events that they had only watched. In no case did the levels of performance differ. In subsequent analyses, we used the mean of children’s productions on these two event types as a measure of their production of the different verbal categories on “old” event sequences.
118
BAUER AND BURCH TABLE 6.7
Means (and Standard Deviations) of Children’s Verbal Productions in the Context of Experimenter–Child and Mother–Child Interviews Interview Condition Mother–Child Interview
Event Type Within Experimenter– Child Interview Children’s Verbal Production
“Old”
New Session 4
New Session 5
M
(SD)
M
(SD)
M
(SD)
M
(SD)
Labels
0.32
(0.15)
0.35
(0.26)
0.36
(0.24)
0.16
(0.19)
Objects
1.08
(0.54)
1.06
(0.69)
0.75
(0.58)
1.79
(0.96)
Activities
1.55
(0.67)
1.36
(0.85)
1.05
(0.73)
1.24
(0.81)
Sequences
0.14
(0.29)
0.26
(0.51)
0.11
(0.27)
0.10
(0.28)
People
0.27
(0.47)
0.30
(0.64)
0.09
(0.19)
1.88
(1.01)
Locations
1.10
(0.48)
1.06
(0.82)
0.68
(0.56)
0.52
(0.34)
Total Talk
4.48
(1.65)
4.39
(1.78)
3.05
(1.57)
5.78
(2.70)
Table 6.7 shows descriptive statistics on children’s verbal productions in the experimenter–child interview context. For ease of comparison, children’s behaviors in the mother–child interview context are reproduced in the right-most column. The category of reasons is omitted from the table, because no tokens of this type were observed in the experimenter–child interviews. For each category, we conducted 1-way analyses of variance comparing levels of production on old events, events new at Session 4, events new at Session 5, and in the mother–child interview context; significant effects were further examined using Tukey tests of significant difference (p < .05). Mean levels of production differed across the contexts in four of the six categories, as well as in total talk: Fs(3, 64) > 3.95, ps < .05. The children produced more labels for events new at Session 4 and new at Session 5, relative to in the mother–child context. Children also mentioned location more frequently on the old events and on the events new at Session 4, relative to in the mother–child interview context. Conversely, the children mentioned objects and people more frequently in the mother–child interviews, relative to in the experimenter–child interviews. The children also produced a greater amount of information in the mother–child interview context than in the context of talking about events new at Session 5. Correlations in Children’s Behavior Across Contexts. There were few correlations between children’s behaviors in the experimenter–child and mother–child interview contexts. Children who used more labels to refer to old events and to events that had been new at Session 4 also used more labels in the mother–child interview
6. MEDIATING YOUNG MINDS
119
context: rs(21) = .49 and .47, respectively. Children who more frequently referred to activities while talking about old events in the experimenter–child interview context less frequently mentioned activities in the mother–child interview context: r(21) = –.59. There were no other statistically significant relations across contexts. Although there were no other reliable correlations, there was a suggestive trend in relation between mentions of people (including the children making reference to themselves) in the experimenter–child and mother–child contexts: Children who made more frequent references to people (almost exclusively to themselves) in the context of the events that had been new to them at Session 4 and the events that were new at Session 5 also tended to make more references to people in the mother–child interview context: r(21) = .38 and r(19) = .54, respectively (ps < .10). A speculative interpretation of these trends is that children who “people” their collaborative narratives also are beginning to include an “auto” in their independent narrative reports. Correlations Between Children’s Behavior in the Experimenter–Child Interview and Mothers’ Behavior in the Mother–Child Interview. Although there was not a great deal of consistency in children’s behaviors in the experimenter–child and mother–child interview contexts, there were relations between maternal verbal behaviors in the mother–child context and children’s verbal behaviors in the experimenter–child context. As noted earlier, in the mother–child interviews, children’s verbal productions were related to maternal elaborations, affirmations, and total talk, in particular. Across contexts, there were few correlations between these maternal verbal categories and children’s verbal behavior in the context of the experimenter–child interview: As reflected in Table 6.8, across these three categories of maternal behavior, there were only three significant correlations with children’s behavior. Moreover, none of the observed correlations was with children’s verbal productions about the events with which they were most familiar, namely, the old event sequences. The most consistent pattern of relations across interview contexts was with the maternal verbal category of associations, maternal productions of which in the mother–child interview were correlated with children’s mentions of activities, sequence, people, and total talk in the experimenter–child interview. Maternal use of associations was related to children’s verbal productions at all three levels of familiarity: old events, events new at Session 4, and events new at Session 5. It is noteworthy that whereas maternal associations seemed to facilitate children’s independent verbal reports, they did not facilitate children’s participation in the mother–child interviews: There were no relations between maternal associative talk and children’s verbal productions in the mother–child interview (see Table 6.6). Although this difference in patterns of relations in the two contexts was unexpected, it is readily interpretable. In the context of the mother–child interview, mothers are attempting to get their children to talk about a specific episode of the mother’s choosing. For this purpose, reminding children of episodes other than
120
BAUER AND BURCH TABLE 6.8
Relations Between Maternal Verbal Producitons in the Context of Mother–Child Interviews and Children’s Verbal Productions in the Context of Experimenter– Child Interviews Maternal Verbal Category Elaborations
Children’s Verbal Production (Experimenter–Child Interview) Labels
Objects
Activity
Sequence People Location
Total
N4 .57**
Affirmations
N5 .44*
Deflections
N5 –.51*
Negations
N4 .44*
Associations
O .52*
O .78**
O .42*
O .50*
N4 .59** N5.68*** Total talk
N5 .80***
N5 .57**
N4.45*
Note: The maternal-verbal category of Repetitions and the child-verbal-production category of Reasons are not represented in the table because there were no significant relations involving these utterance types. “O” indicates old event sequences, “N4” indicates event sequences that were new at Session 4, and “N5” indicates event sequences that were new at Session 5. Only statistically significant corrrelations are represented: *p < .05 **p < .01. *** p < .001.
the target is not especially helpful. However, mothers who routinely make reference to how events or episodes are related to one another may be providing their children with the tools needed to make connections between the currently-available cues (in this case, event-related props) and stored memory representations (in this case, the event sequences the props were used to enact). Conclusions and Implications. The goal of this research was to determine whether the social, emotional, and cognitive factors that influence early memory development extend across different mnemonic contexts. In this case, the specific factor under study was maternal verbal scaffolding; the contexts examined were those of mother–child interviews about shared past events and experimenter–child interviews about laboratory event sequences. We expected that children’s previous experiences in mother-supported interactions would mediate their performance in independent prop-supported remembering. In a pattern analogous to that observed in explorations of consistency in maternal verbal style across tasks (e.g., Haden & Fivush, 1996; Haden et al., 1996), little
6. MEDIATING YOUNG MINDS
121
consistency in children’s verbal memory contributions was found across the experimenter–child and mother–child interview contexts. Nevertheless, there was evidence that children’s performance in both contexts was facilitated by support from the mother, even when she was not one of the conversational partners. Within the context of the mother–child conversations about past events, maternal elaborations, affirmations, repetitions, and deflections were predictive of children’s contributions of information about the objects, activities, and people in events from the past. Thus, mothers who wove a more coherent narrative and provided more support for their children during these conversations had children who were more successful in reporting on the main elements of the past event (i.e., the who and the what of the experience). Whereas maternal elaborations, affirmations, repetitions, and deflections predicted children’s contributions during the shared memory conversations, they did not predict children’s independent memory contributions. Nevertheless, there was evidence of “mediation” of children’s minds. The extent to which mothers related past events to other experiences within the mother–child conversation predicted the extent to which children commented on the activities and sequences, and the total amount of information that children provided about the laboratory events. Thus, “tutoring” children to relate experiences to one another facilitated their abilities to verbally report on events previously experienced and to talk about novel events (i.e., events new at Session 5). That different variables were related in different contexts makes an important point about maternal contributions to the development of narrative competence: Effects are not global, but specific. That is, different aspects of maternal behavior support different components that together constitute mnemonic narrative competence in young children. These findings make clear that effects of maternal mediation are specific, even as they reach from the context of mother–child conversations about the past to experimenter–child interviews about laboratory events.
CONCLUDING REMARKS In her long and productive career, Katherine Nelson has made numerous significant contributions to the understanding of development in general and cognitive development in particular. One of her most enduring contributions likely will be the role that she has played in encouraging attention to the social and emotional aspects of cognitive development. Historically, the field of cognitive developmental science has largely ignored these rather “messy” factors in favor of variables that can be manipulated and controlled. However, it is obvious that most of development does not take place in controlled laboratory settings. Rather, it occurs in contexts peopled by social-interaction partners who inject their own sources of variance into each and every interaction with the child. Moreover, children have affective (e.g., temperament) and cognitive (e.g., language) char-
122
BAUER AND BURCH
acteristics that add variance to their task performance as well as to the “performance” of their social-interaction partners. One reaction to this complexity is to ignore it and continue to study what can be manipulated. Another is the reaction advocated by Katherine Nelson, which is to acknowledge the complexity and use it to understand the multiple mediators of foundational cognitive processes, such as memory. The work reported in this chapter is an illustration of what is to be gained by following Katherine Nelson’s lead, once again. In Burch and Bauer (2003), we found that at age 24 months, children’s performance on controlled laboratory tasks was affected by characteristics of their own temperaments, by their own levels of language proficiency, and by their mothers’ verbal behavior. The patterns of interaction of these sources of variance differed as a function of the level of challenge of the mnemonic task. When the going was relatively easy (i.e., on four-step sequences), both direct and indirect relations between characteristics of the children, maternal verbal behavior, and memory performance were observed. As the going got tougher (i.e., on six-step sequences), children’s mothers seemed to respond more to the situation, and less to their children’s dispositions, by providing more verbal support. Nevertheless, characteristics of the children still mattered: Children with typically higher levels of Interest and Persistence were able to take better advantage of the verbal scaffolding provided by their mothers. The entire pattern speaks to multiple mediators of the behavior of both mothers and their children. The significance of the process of socialization of mnemonic behavior was readily apparent in the results of the second study on which we reported. As have been observed many times over, we found relations between maternal elaborations and 3-year-old children’s contributions to conversations about shared past events. We also observed that effects of maternal verbal behavior extended beyond the context in which they typically have been observed, into the more independent mnemonic context of experimenter–child interviews about controlled laboratory events. Suggestively, the maternal variables that related to children’s contributions in the mother–child interviews were not the same as those that predicted children’s contributions in the experimenter–child interviews. When the children were “on their own” in the independent mnemonic task, it was not the familiar variable of maternal elaborations that related to performance. Rather, maternal associations seemed to carry the children through. This finding points to the importance of examining relations between a wide range and diversity of maternal verbal behavior and children’s verbal mnemonic performance across multiple contexts. As noted earlier, the concept of a “mediated mind” implies that social, emotional, and cognitive factors that influence early development have an impact that is felt across contexts and over time. Evidence of Katherine Nelson’s “power” as a mediator of minds is apparent in both of these respects. Her approach has influenced research in many different contexts, and by several cohorts of investigators. In this chapter alone we saw the utility of formulations developed in the context of shared conversations about naturally occurring events for an understanding of
6. MEDIATING YOUNG MINDS
123
children’s behavior in the context of prop-supported interviews about specific laboratory events. That this chapter is the work of an advisor and her former student, neither of whom was ever “officially” tutored by Katherine Nelson, is ample evidence of the influence of these powerful ideas over time and distance. The authors of this chapter and the field at large have been enriched by Katherine Nelson’s commitment to the ideal that cognitive structures are cultural constructions, mediated by many minds.
ACKNOWLEDGMENTS Support for collection of the data reported in this chapter was provided by a grant from the National Institute of Child Health and Human Development (HD-28425) to Patricia J. Bauer. We also thank Patricia Dropik, Christine Leehey, Rebecca Niedfeldt, Jennifer Wenner, and Sandi Wewerka for assistance with data collection; Stephanie Bangston, Mari Strand Cary, Jessica Fredrick, Emily Stark, and Jennie Waters for assistance with coding; and other members of the Cognition in the Transition laboratory group for their help in the various phases of this research. We extend a special note of thanks to the mothers and children who gave of their time and energy to make this work possible.
REFERENCES Bauer, P. J. (1992). Holding it all together: How enabling relations facilitate young children’s event recall. Cognitive Development, 7, 1–28. Bauer, P. J., Burch, M .M., & Kleinknecht, E. E. (2002). Developments in early recall memory: Normative trends and individual differences. In R. Kail (Ed.), Advances in child development and behavior (pp. 103–152). San Diego, CA: Academic Press. Bauer, P. J., & Dow, G. A. A. (1994). Episodic memory in 16- and 20-month-old children: Specifics are generalized, but not forgotten. Developmental Psychology, 30, 403–417. Bauer, P. J., Dow, G. A., Bittinger, K. A., & Wenner, J. A. (1998). Accepting and exempting the unexpected: 30-month-olds’ generalization of event knowledge. Cognitive Development, 13, 421–452. Bauer, P. J., & Fivush, R. (1992). Constructing event representations: Building on a foundation of variation and enabling relations. Cognitive Development, 7, 381–401. Bauer, P. J., & Hertsgaard, L. A. (1993). Increasing steps in recall of events: Factors facilitating immediate and long-term memory in 13.5- and 16.5-month-old children. Child Development, 64, 1204–1223. Bauer, P. J., Hertsgaard, L. A., Dropik, P., & Daly, B. P. (1998). When even arbitrary order becomes important: Developments in reliable temporal sequencing of arbitrarily ordered events. Memory, 6, 165–198. Bauer, P. J., Hertsgaard, L. A., & Wewerka, S. S. (1995). Effects of experience and reminding on long-term recall in infancy: Remembering not to forget. Journal of Experimental Child Psychology, 59, 260–298.
124
BAUER AND BURCH
Bauer, P. J., Kroupina, M. G., Schwade, J. A., Dropik, P., & Wewerka, S. S. (1998). If memory serves, will language? Later verbal accessibility of early memories. Development and Psychopathology, 10, 655–679. Bauer, P. J., & Travis, L. L. (1993). The fabric of an event: Different sources of temporal invariance differentially affect 24-month-olds’ recall. Cognitive Development, 8, 319–341. Bauer, P. J., Wenner, J. A., Dropik, P. L., & Wewerka, S. S. (2000). Parameters of remembering and forgetting in the transition from infancy to early childhood. Monographs of the Society for Research in Child Development, 65 (4, Serial No. 263). Bauer, P. J., Wenner, J. A., & Kroupina, M. G. (2002). Making the past present: Verbal reports of preverbal memories. Journal of Cognition and Development, 3, 21–47. Bauer, P. J., Wiebe, S. A., Waters, J. M., & Bangston, S. K. (2001). Reexposure breeds recall: Effects of experience on 9-month-olds’ ordered recall. Journal of Experimental Child Psychology, 80, 174–200. Burch, M. M., & Bauer, P. J. (2003). Relations between maternal style, children’s recall memory, and temperamental characteristics. Manuscript in preparation. Carver, L. J., & Bauer, P. J. (1999). When the event is more than the sum of its parts: Ninemonth-olds’ long-term ordered recall. Memory, 7, 147–174. Engel, S. (1986). Learning to reminisce: A developmental study of how young children talk about the past. Unpublished doctoral dissertation, City University of New York Graduate Center. Farrant, K., & Reese, E. (2000). Maternal style and children’s participation in reminiscing: Stepping stones in children’s autobiographical memory development. Journal of Cognition and Development, 1, 193–225. Fenson, L., Dale, P. S., Reznick, J. S., Bates, E., Thal, D. J., & Pethick, S. J. (1994). Variability in early communicative development. Monographs of the Society for Research in Child Development, 59 (5, Serial No. 242). Fivush, R., & Fromhoff, F. (1988). Style and structure in mother–child conversations about the past. Discourse Processes, 11, 337–355. Goldsmith, H. H. (1996). Studying temperament via construction of the toddler behavior assessment questionnaire. Child Development, 67, 218–235. Gunnar, M. R. (1990). The psychobiology of infant temperament. In J. Colombo & J. Fagan (Eds.), Individual differences in infancy: Reliability, stability, prediction (pp. 387–409). Hillsdale, NJ: Lawrence Erlbaum Associates. Haden, C. A., & Fivush, R. (1996). Contextual variation in maternal conversational styles. Merrill-Palmer Quarterly, 42, 200–227. Haden, C. A., Reese, E., & Fivush, R. (1996). Mothers’ extratextual comments during storybook reading: Stylistic differences over time and across texts. Discourse Processes, 21, 135–169. Hudson, J. A. (1990). The emergence of autobiographical memory in mother–child conversation. In R. Fivush & J. A. Hudson (Eds.), Knowing and remembering in young children (pp. 166–196). Cambridge, MA: Cambridge University Press. Kleinknecht, E. E., Rademacher, J., & Bauer, P. J. (2002). Pleasure gives way to interest and persistence: Changing relations between temperament and memory at 13 and 20 months. Unpublished manuscript.
6. MEDIATING YOUNG MINDS
125
Nelson, K. (Ed.). (1986). Event knowledge: Structure and function in development. Hillsdale, NJ: Lawrence Erlbaum Associates. Nelson, K. (1996). Language in cognitive development: Emergence of the mediated mind. New York: Cambridge University Press. Peterson, C., & McCabe, A. (1994). A social interactionist account of developing decontextualized narrative skill. Developmental Psychology, 30, 937–948. Reese, E., Haden, C. A., & Fivush, R. (1993). Mother–child conversations about the past: Relationships of style and memory over time. Cognitive Development, 8, 403–430. Rothbart, M. K., Ahadi, S. A., & Hershey, K. L. (1994). Temperament and social behavior in childhood. Merrill-Palmer Quarterly, 40, 21–39. Rothbart, M. K., & Bates, J. E. (1998). Temperament. In W. Damon (Series Ed.) & W. S. E. Damon & N. V. E. Eisenberg (Vol. Eds.), Social, emotional and personality development: Vol. 3. Handbook of child psychology (5th ed., pp. 105–176). New York: Wiley. Tessler, M., & Nelson, K. (1994). Making memories: The influence of joint encoding on later recall by young children. Consciousness and Cognition, 3, 307–326. Travis, L. L. (1997). Goal-based organization of event memory in toddlers. In P. W. van den Broek, P. J. Bauer, & T. Bourg (Eds.), Developmental spans in event comprehension and representation: Bridging fictional and actual events (pp. 111–138). Mahwah, NJ: Lawrence Erlbaum Associates.
7 The Development of Future Thinking: Constructing Future Events in Mother–Child Conversation Judith A. Hudson Rutgers University
Since the publication of Katherine Nelson’s (1986) groundbreaking studies on the structure and development of children’s event knowledge, research has provided a wealth of information about how generalized event representations support children’s cognitive development. Schematically organized event knowledge provides a form of mental representation that children draw on when engaged in cognitive tasks such as recalling events, understanding and constructing stories, communicating with parents and peers, and drawing inferences (Nelson, 1986). According to Nelson, general event knowledge provides a “cognitive context,” an organized knowledge base that children can rely on, both implicitly and explicitly, to guide their recall of events and their reasoning about events. Research has shown that preschool children are able to recall stories, understand temporal sequences, and to draw inferences from text when the material they are thinking about refers to familiar events, but fail to display the same level of ability when asked to remember or reason about unfamiliar situations (see Hudson, 1993, for a review). 127
128
HUDSON
Cognitive developmental research has also been influenced by Nelson’s attention to the social and interactional roots of memory and conceptual development (Nelson, 1993, 1996). When reminiscing about past events, engaging in pretend play, or experiencing a new event, the goal of the activity in which they are engaged, the interpersonal connection between participants, and the language used by the participants to communicate to one another are critical components that structure children’s understanding and participation. Children’s event knowledge contributes to these situations, but is not separate from the overall context. An important development during childhood is the decontextualization of event knowledge, but early in development, event knowledge is embedded in a larger social and cognitive context as children try to make sense of the sequence of everyday events they experience in and to understand the language used to refer to those events, both present and nonpresent. This chapter focuses on these themes, the role of general event knowledge in supporting cognitive processing and the role of social interaction in conceptual development, in examining how children learn to think about future time. It is proposed that young children rely on general event knowledge in thinking about and planning for future occurrences of familiar events. Furthermore, talking with adults about these future occurrences is the natural context in which children learn how to plan for future events and how to conceptualize future time. Research on children’s planning for familiar events is reviewed as well as research on children’s understanding of future time. Data from a recent investigation on mother–child talk about future events is presented. This study examined how mothers talk to preschool children about past and future events. The analyses presented focus on how everyday discussions of future events can contribute to the development of children’s explicit understanding of future time, which is fundamental to the development of planning skills.
THE DEVELOPMENT OF CHILDREN’S SCRIPTS AND PLANS Although general event knowledge automatically guides children’s everyday event comprehension and social interactions, there are also situations in which an explicit awareness of event knowledge is essential. Planning is a task that relies on explicit consideration of event sequences and options (Schonick & Friedman, 1993). Planning can be facilitated by thinking about what usually happens in familiar events and using that knowledge to predict and plan for a future event (Hudson & Fivush, 1991, Hudson, Sosa, & Shapiro, 1997). Thus, well-organized event knowledge should provide young children with a rich knowledge base to draw on when planning for future, familiar events. However, research has shown that although preschool young children have well-organized script knowledge, they do not automatically use that knowledge in planning tasks. In one study, 3-, 4-, and 5-year-olds were asked either to describe what happens in two familiar
7. THE DEVELOPMENT OF FUTURE THINKING
129
events, or to plan aloud for future instantiations of going to the beach or going grocery shopping (Hudson, Shapiro, & Sosa, 1995). Compared to their script reports, 3-year-olds’ event plans were shorter and showed little evidence of planfulness. These differences were striking given that their script reports were adequate and they could simply have provided a script report in response to the request for a plan; a general description of “what happens” could serve both functions. Younger children appeared to have difficulty using their general event knowledge to plan. Research on children’s event planning has also shown that preschool children benefit considerably when they are provided with support or assistance. A study by Hudson and Fivush (1991) investigated preschool children’s ability to plan for future events (a birthday party and breakfast) while shopping for groceries in a pretend grocery store. Three- and 4-year-old children had difficulty keeping a future event goal in mind and using that goal to direct their shopping activities. However, their performance improved when they were provided with external supports for the task, that is, a verbal reminder of the event goal or the opportunity to shop in a store where all the birthday items were placed together and all the breakfast items were placed together in another location. A series of experiments found that preschool children benefited from reminders and prompts in their efforts to carry out plans for familiar events (art projects involving complex events sequences) and that preschool children were better able to construct and carry out event plans for making snacks and art projects in their preschool if they have listened to an adult plan aloud (Hudson et al., 1997). These studies indicate that preschool children find the task of planning for the future more difficult than just thinking about “what happens.” Although they acquire well-organized general event knowledge very early, they do not seem to know how to use this knowledge in the service of planning and they benefit from training and support. For preschool children, the task of using event knowledge to plan future events is not easy or automatic, but they are more successful when their performance is supported or scaffolded by adult models, reminders, and a structured environment.
PLANNING FOR EVERYDAY EVENTS IN NATURALISTIC CONTEXTS The results from these experimental studies of children’s event planning gave rise to many questions about how children plan in naturalistic contexts. Although children benefit from adult support and scaffolding in experimental planning tasks, is there any evidence that adults provide this kind of support in planning for everyday events? Real-world situations that provide children with support in thinking about how general event knowledge can be used to predict future events could provide very important lessons in learning how to plan. But is
130
HUDSON
this actually how children learn to plan in real-world situations? Do parents engage young children in planning for future events? If so, how do parents structure this kind of interaction? Are there particular ways that parents scaffold their children’s involvement in event planning that are more or less effective in teaching children how to plan? Although there has been little investigation of how parents talk to children about future events, there has been extensive research conducted on parent–child talk about past events. A large body of research has shown that naturalistic conversations about past events provide young children with a context for learning how to reminisce (Engel, 1986; Farrant & Reese, 2000; Fivush 1997; Haden, Haine, & Fivush, 1997; Hudson, 1990; Sachs, 1983). The cognitive activity of remembering is accomplished through conversation, with parents structuring the conversation by asking children questions and providing their own memories. By talking to children about the past, parents elicit children’s recall and model a verbal retrieval process. Conversations about the past also teach children the social significance of sharing memories. Nelson (1993) argued that the social sharing of memories in natural conversation leads to the onset of autobiographic memory. Through memory conversations, children learn that their culture values joint verbal remembering as a way of sharing and preserving memories of the past. Continued discussions of the past provide opportunities to learn how to structure memory narratives and reinstate children’s memories for past experience (Hudson, 1990). By analogy, everyday conversations about future events may also serve as a social and linguistic context for learning how to think about future time and how to plan for future events. Conversations about the future are a logical context to investigate children’s developing concepts of future time because temporal concepts are constructed by language. Time does not exist in the world as a physical entity to be discovered through exploration and manipulation, but must be socially constructed through language (Nelson, 1996; Weist, 1989). Indeed, there is significant social and historical variation in the ways that different cultures refer to time (Friedman, 1990). Conceptualizing and talking about the future presents a particularly difficult cognitive challenge for children because, as compared to talking about the past, they do not have stored memory representations of the future that they can access and refer to when talking about what will transpire. Rather, the future has not yet been determined and can only be constructed through language. In addition, the language used to refer to the future is more hypothetical than talk about the here and now and then and there. Thus, thinking about the future is conceptually more difficult than thinking about the past and talking about the future is linguistically more difficult than talking about the past. However, when thinking about future instances of familiar events, children can draw on generalized event representations to predict the future. Although they may not do so spontaneously, parents’ questions may prompt children to use general event knowledge to think about future
7. THE DEVELOPMENT OF FUTURE THINKING
131
events. When examining how parents talk to children about future events, it is important to examine parents’ use of hypothetical language and their references to general event knowledge to understand the degree to which hypothetical thinking and thinking about general event knowledge is explicitly encouraged.
REPRESENTING PAST AND FUTURE TIME Research on children’s understanding of future time supports the view that this is a more difficult task than thinking about the past. One method for studying children’s temporal concepts is to use a sequencing task in which children are asked to sequence real-world events from the past, present, and future along a timeline (Benson, 1997, Friedman 1977, 1990, 2000; Friedman & Brudos, 1988). Research using this methodology has found that by 4 to 5 years, children are able to construct correctly ordered sequences for everyday activities such as eating dinner, going to bed, waking up, and having lunch, as well as for more distant past and future events such as birthdays and holidays. However, 4- and 5-year-olds are more accurate in sequencing activities from the past than from the future. These findings support the view that temporal understanding develops in the context of thinking about routine events, but that thinking about a yet undetermined future is more difficult than reasoning about an already experienced past. Studies by Lucariello and Nelson (1987) and by Benson (1997) indicate that parents talk with young children about the future as much as or even more than they talk about the past. However, these studies focused on children younger than 4 years and were more concerned with the frequency of future talk than with analyzing the content of future talk. More research is needed on how parents structure conversations about future events with children during years 2 to 5 to investigate the role of parent–child conversation in the emergence of children’s concepts of future time. Research has shown that the distinction between past and future time is evident by age 4, yet little is known about how this distinction is acquired. Research also points to some confusion in preschool children’s understanding of temporal proximity in relation past and future time. It is not yet clear if these confusions are evidence of incomplete or fuzzy temporal concepts or confusions about temporal reference. Data on the temporal references that children are exposed to in naturalistic contexts may shed light on these issues. For example, do parents clearly mark discussions of past and future events? Do they use markers other than tense, which may be less salient that specific temporal terms (e.g., We went to the park as compared to yesterday we went …) Are references to temporal proximity either more frequent or more explicit than references to past versus future events? Studying the ways in which adults talk to children about future events at an age when fundamental temporal concepts are acquired can provide insight into how exposure to temporal language may contribute to observed developmental differences in the acquisition of temporal concepts.
132
HUDSON
AN INVESTIGATION OF MOTHER–CHILD CONVERSATIONS ABOUT PAST AND FUTURE EVENTS In comparing conversations about past and future events, the focus was on two major questions. First, how do parents structure conversations about the future? There is considerable research on how parents structure conversations about past events with young children, but do they use the same techniques when discussing the future? The issue here is whether parents employ a general strategy to engage young children in conversation that is used regardless of the topic of discussion or do parents use different strategies depending on the topic or goal of the conversation. This issue is further complicated when different parents may have different goals in mind, even when discussing the same topic (past or future) with their children. Research on parent–child talk about the past has found evidence for two different styles of maternal elicitation that are thought to reflect different goals (Engel, 1986; Fivush & Fromhoff, 1988; Hudson, 1990; McCabe & Peterson, 1991; Reese, Haden, & Fivush, 1993). Parents using an elaborative style tend to pose many questions to their children, and they provide more and more information in their repeated queries. They seem to have the goal of engaging their children in the conversation and use more elaborative questions in an attempt to provide children with more cues about the event. Parents using this style also follow-up on their children’s comments, even when children provide information that does not address the question posed. Their efforts are directed toward keeping the conversation going and maintaining the child’s participation. In contrast, parents using a repetitive style tend to repeat very similar questions without necessarily providing additional information. Their goal appears to be one of eliciting answers to specific questions and the repetition of the same questions is an attempt to direct children’s attention to those questions. Parents using this style are less likely to follow-up on children’s responses that are not relevant to the parent’s questions. Parents may have different goals in talking about the future. As in conversations about the past, some parents may be more interested in maintaining the conversation and use a more elaborative questioning style, whereas other parents may be more focused on eliciting specific information from their children using a more repetitive questioning style. As found in research on parent–child talk about the past, different styles of eliciting children’s contributions to talk about the future may influence the development of temporal concepts and planning skills in young children. An important issue for this investigation was to identify which maternal elicitation variables were related to children’s participation in conversations about the future. The analyses of mothers’ use of temporal language has been reported in Hudson (2002) and the major results are summarized in this chapter. A second important question concerned how children at different ages participate in conversations about the future. What kinds of contributions can they make?
7. THE DEVELOPMENT OF FUTURE THINKING
133
What kinds of questioning techniques are more effective in eliciting children’s contributions? Because thinking about the future necessarily involves hypothetical thinking, can children as young as 2½ years of age provide meaningful information in these conversations? How does general event knowledge support children’s thinking about future events? To investigate this last question, parents were asked to discuss both novel and familiar future (and past) events with their children. When discussing future familiar events, children would be able to draw on general knowledge about the event to predict future occurrences, but they would have less background knowledge to support their thinking about novel events. It was hypothesized that children would participate more fully in conversations about familiar events as compared to conversations about novel events and that parents might employ different strategies for engaging children in discussions about familiar and novel future events. New analyses of children’s contributions and of the relation between maternal and child contributions are presented in this chapter. Fifty-six mothers and children participated: 28 children at 2½ years (15 females, mean age = 30 months, age range = 28–31 months) and 28 children at 4 years (14 females, mean age = 47 months, age range = 46–49 months) . Mothers were asked to initiate conversations with their children about past and future events. Conversations took place in participants’ homes on two occasions within a 2-week period. In each session, either four past events or four future events were discussed (order was counterbalanced). In each conversation, mothers were asked to discuss two novel and two familiar events: Novel events had either never been experienced or had only been experienced once and familiar events had been experienced two or more times.
Coding Maternal Speech Type of Utterance. Each maternal utterance was first coded as one of eight conversational categories (questions, contextual statements, evaluations, prompts, memory placeholders, metacognitive comments, placeholders, clarifications, associative talk) using a coding scheme adapted from Reese et al. (1993).1 The majority of maternal contributions consisted of questions (including general queries, “What happened when we went to Grandma’s?” requests for information, “Who’s gonna be at the party?” and yes–no questions, “Do you want to go to the playground tomorrow?”) and contextual statements (statements that provided new information). Because these contributions represented mothers’ substantive contributions and are generally considered the most revealing measures of maternal elicitation style (Reese et al., 1993), they were further coded as to whether they consisted of an elaboration or a repetition. Elaborations referred to new information about the event under discussion, whereas repetitions referred to information previously discussed.
134
HUDSON
Point of Reference. All questions and contextual statements were coded in terms of the their temporal point of reference (Hudson, 2002): Past episodes—“What did you do the last time you went to the playground?” General event knowledge—“What happens at preschool?” Future action—“What are you going to do at the beach?” Future plans—Includes references to possible actions (“Maybe we’ll see horses there”), predictions (“Do you think Amanda will be there?”), and elicitation of preferences (“Would you like to have a birthday party at McDonald’s?”) Temporal terms. Mother’s speech was also coded for five categories of temporal terms (Hudson, 2002): Episode markers. References to “last year” or “next time” that served to index which episode of a repeated event was referred to. Although sometimes these included references to conventional time units (e.g., “year”), they did not refer to an interval of time, but simply marked which episode was under discussion. Conventional time markers. References to any conventionalized unit of time such as day, week, month, year, Saturday, July, spring, at night, an hour. References to conventional units referring to the immediate 24 hours were coded in a separate category. Conventional time—within 24 hours. Because several studies found that immediate temporal references are more common in speech to young children than distant references, references such as yesterday, today, tomorrow, now, this afternoon, tonight, last night were coded in a separate category of conventional time references. Sequence. Reference to time in terms of a sequence of actions within or between events, such as after, before, next, then, when we first came in, the last thing. Indeterminate interval: These references referred to an interval of time, but did not code the interval in terms of a conventional time unit, for example, in a while, soon, after a long while, while I’m gone, a long time ago. Temporal references were coded across all utterance categories and the frequency of production was calculated for each event under discussion.
Child contributions All comments made by children were coded as one of 10 types of contributions; elaborations, questions, yes or no responses, repetitions, placeholders, evaluations, metacognitive comments, associative talk, off-topic talk, and unintelligible utterances. Children’s substantive contributions to the conversations were largely
7. THE DEVELOPMENT OF FUTURE THINKING
135
in the form of elaborations, that is, statements that provided new information about the event under discussion. All elaborations produced by children were therefore coded for their temporal point of reference.
Results Mothers’ Contributions Temporal Frames of Reference. To examine how mothers structured conversations about the future, we were especially interested in the temporal frames of reference that mothers used (Hudson, 2002). Frequencies of use of each frame of reference are shown in Figure 7.1. Analyses performed on these data showed that there were differences across past and future event conversations in mothers’ use of temporal frames. When talking about the past, mothers largely referred to past episodes and general event knowledge, although these frequencies varied across novel and familiar events. When talking about the future, mothers used more varied frames of reference and referred to general event knowledge as much as they referred to future actions. Mothers referred to general event knowledge more often when discussing familiar events, whereas future reference was used more when discussing novel events. Interestingly, maternal point of reference did not vary significantly by age of the child for either past or future conversations.
FIG. 7.1. Mean frequencies of past, general, future, and hypothetical references in mothers’ combined questions and contextual statements for novel and familiar events in past and future conversations (error bars represent the standard errors).
136
HUDSON
Temporal Terms. The frequency of mention for each category of temporal terms is shown in Figure 7.2. Analyses of mothers’ use of each category of temporal reference yielded significant conversation differences for conventional time references and sequence terms (Hudson, 2002). Mothers used conventional time markers more when discussing future events and used sequence terms significantly more often when discussing past events. Significant age effects were found only for mention of conventional temporal terms; mothers of older children used more conventional terms (M = 2.21, SD = 2.12) than mothers of younger children (M = 1.46, SD = 1.49). Age differences in mothers’ use of conventional time markers are consistent with the research on children’s understanding of temporal terms. Although children understand sequence terms such as first, next, and then by age 3, it is not until approximately age 8 that children fully understand conventional temporal terms such as next year or in 2 weeks. It is not surprising then, that mothers used more conventional time terms when talking to 4-year-olds than when talking to 2½year-olds. However, it is interesting that mothers used conventional time terms more often when discussing future events than when discussing past events. This finding is another indication that mothers’ talk about the future was temporally more complex than their talk about the past.
FIG. 7.2 Mean frequencies for mention of temporal terms in past and future conversations (error bars represent the standard errors).
7. THE DEVELOPMENT OF FUTURE THINKING
137
Children’s Contributions How did children respond? To examine children’s contributions, the number of elaborations they produced across conversations was analyzed. Elaborations were responses that provide new event information and are a measure of children’s substantive contributions to the conversations. As with mothers’ contributions, children’s elaborations were also coded for frame of reference. Unlike mothers’ contributions, however, many of children’s utterances could not be reliably coded for point of reference. Children’s elaborations often consisted of short phrases in response to mothers’ queries that did not include a verb, for example, “What do you wear in the pool?”, “ A bathing suit.” When a phrase included a verb, tense markers were frequently unclear either because of the child’s articulation difficulties or because tense could not be determined from the verb used (e.g., “I put it down.”) An indeterminate category (none) was therefore included in this analysis. In addition, because children very rarely used hypothetical or conditional references, all references to future events were combined into a single future point of reference code without distinguishing utterances with a future hypothetical perspective. Number of Elaborations. First, the total number of elaborations produced by children across past and future conversations was analyzed. Although older children produced more elaborations than younger children, the difference was only significant for future conversations (M = 20.33, SD = 17.05 for older children and M = 10. 68, SD = 8.71 for younger children) but not for past conversations (M = 20.32, SD = 14.58 for older children, and M = 17.09, SD = 11.87 for younger children). In addition, younger children, but not older children, produced significantly fewer elaborations in future conversations than in past conversations. These findings suggest that talking about the future was more difficult for the younger children than talking about the past; they contributed less to these conversations than to conversations about the past and they contributed less to these conversations than did older children. In contrast, the 4-year-olds contributed as much new information to conversations about past and future events. Temporal Frames of Reference. To examine the point of reference children used in past and future conversations, separate analyses were performed for past and future conversations. The number of elaborations produced in each conversation by age and point of reference is shown in Table 7.1. As this table shows, the majority of children’s utterances could not be coded for point of reference. In the analysis of past conversations, children referred to the past more frequently than they used a general point of reference, and they used the general reference more frequently than future reference. In conversations about future events, children used the general point of reference more frequently than the past, but there was no significant difference between the frequencies of general and future time reference (see Table
138
HUDSON TABLE 7.1
Mean Number of Elaborations Produced by Children in Past and Future Conversations by Age and Temporal Frame of Reference Temporal Frame of Reference None
Past
General
Future
Past Conversations Younger children
10.7
4.2
1.7
0.2
Older children
12.4
6.1
1.7
0.2
Mean
11.5
5.1
1.7
0.2
Future Conversations Younger children
7.5
0.4
1.6
1.5
Older children
12.8
0.8
3.8
3.8
Mean
10.1
0.6
2.6
2.3
TABLE 7.2 Mean Number of Elaborations Produced by Children in Future Conversations by Type of Event and Temporal Frame of Reference Temporal Frame of Reference Type of Event
None
Past
Novel
10.3
0.7
2.7
1.3
9.3
0.5
2.5
3.6
Familiar
General
Future
7.1). As shown in Table 7.2, children used the future point of reference more often when discussing routine events than when discussing novel events, but the number of references to the past and to general event information did not vary significantly across novel and familiar events. These results indicate that children used tense markers to distinguish references to different points in time and used the appropriate tense when discussing past and future events. However, reference to general event knowledge, which was relevant to both discussions but not the central topic of either, varied across conversations. While discussing the past, both older and younger children reported general event information at the same rate and reported this type of information less than past information. When discussing future events, older children reported more general event information than younger children and older children reported the same amount of general and future information. This finding supports the notion that general event knowledge was more relevant to discussion about future events and
7. THE DEVELOPMENT OF FUTURE THINKING
139
that 4-year-olds were more competent at talking about the future than were 2½-year-olds. The finding that more references to the future were made when discussing future familiar events as compared to future novel events suggests that children were better able to conceptualize and talk about the future in the context of familiar events. That is, they were able to use general event knowledge to predict what would happen in the future, whereas when discussing future novel event, they were less able to contribute information about future actions. The difference between discussing future novel and familiar events is illustrated in the following example. The child is asked about an event that has been experienced once in the past, but the child has no recollection of the experience: Example 1: Age 4 years, novel event: Going to Hershey Park M: Where do you want to stay when we go? C: A hotel. M: You want to stay in a hotel. Do you think that we should find a place with room service? C: Why? M: To bring us our food. Or do you want to go to restaurants? C: I want room food. M: Do you remember anything else about Hershey Park that we will do again? C: No. M: Do you like going on rides? They have rides there. C: Umm, yeah. M: You like all those rides, huh? C: Yeah, I don’t know nothing else. M: How about the chocolate? You like the chocolate? You want to get some chocolate to bring home? C: Yeah. M: What kind of chocolate is there? C: I don’t know. This example illustrates the difficulties children faced when talking about future, novel events. If the event has never been experienced, then the child can only guess as to what will happen or recall information that was told. If the event has been experienced one time in the past, but the child does not recall that event well, references to the single past episode are likely to produce “I don’t know” re-
140
HUDSON
sponses. In contrast, consider the following examples of discussions about future, familiar events: Example 2: 2½-year-old, familiar event: Church M: Anthony, where do we go on Sunday together? C: Church. M: C: M: C:
Right, we go to church together. And what do we do there? We sing. We sing, uh huh. What else? Do you know what else we do? We pray?
M: Right, we pray and we listen carefully and we are quiet, right? C: Yeah. M: C: M: C: M:
Are we going to church this Sunday? Yeah. Who is going to go? Mommy. Who else?
C: Daddy. Example 3: 4-year-old, familiar event: Grocery store M: Do you know what we have to do tomorrow? What do we do all the time? We have to go to the grocery store. What do we do there? C: We shop. M: Yeah, but tell me how we do that. Come on. C: We get a cart, some food, and go around. M: Yeah, what else? C: M: C: M: C: M:
We get cookies. Sometimes if you guys are good, right. Yeah and sometimes we get milk and eggs. Usually we do. What else? What do we have to do with our groceries? Take them home and eat them for dinner. Well, yeah. But we do, what do we have to do at the store? Do we have to pay? C: Yeah. M: How do we do that?
7. THE DEVELOPMENT OF FUTURE THINKING
141
C: We wait and be quiet. M: Yeah, we have to wait in line and be good. In these two examples, the mothers essentially elicited script reports from their children about what usually happens at church and at the grocery store. This information is relevant to thinking about the next episode of these events and it is evident that the children both have relevant general event information to report in these conversations, although as was typical in future conversations, the older child reported more information than the younger child.
Relations Between Mothers Temporal Talk and Children’s Contributions The next set of analyses focused on the relation between characteristics of mothers’ contributions and the amount of event information children were able to provide. Here the focus was on which maternal conversational variables were associated with greater information reported by children. Again, the measure of interest was the number of elaborations produced by children, that is, the degree to which they were able to provide new information to the conversations, not just repetitions of previously stated information. These analyses were meant to test for the effects of maternal elicitation variables that were independent of the number of questions that mothers asked. Because children’s contributions were almost always a response to a maternal question, it was important to control for variability in the number of questions asked. Preliminary correlations indicated that, as expected, there were significant correlations between the number of questions posed by mothers and the number of elaborations provided by children. Partial correlations were therefore computed between maternal conversational variables and the number of elaborations produced by children, controlling for the total number of questions asked by mothers. Separate correlations were computed for older and younger children, in past and future conversations. Variables that occurred very infrequently (M < 1.00) were excluded from these analyses. As shown in Table 7.3, there were age and conversation differences in the patterns of significant correlations. In discussions about past events, mothers’ use of contextual statements was positively correlated with children’s elaborations. Contextual statements provide more information about the event under discussion in the form of a statement instead of a question. Providing more event information appears to facilitate children’s recall of the events at both ages. However, providing more event information did not facilitate children’s participation in future event conversations to the same degree. Older children, but not younger children, also benefited from mothers’ use of prompts (e.g., “what else?”) when discussing
142
HUDSON TABLE 7.3
Correlations Between Maternal Variables and Total Child Elaborations by Age and Event Conversation, Controlling for Total Number of Mothers’ Questions Future Events Younger Maternal elaboration score
Older
Past Events Younger
Older
–.16
.00
.33
Prompts
.05
.13
–.28
–.09
Contextual statements
.30
.34
.48*
.46*
.29
–.53**
.18
.00
.50**
.42*
Point of reference Past General Future Hypothetical Prediction Preference
.23 –.09 --–.15
.22
.04
.35
--
--
.52**
--
--
.07
--
--
.03
--
--
.42**
--
–.27
--
–.28
.37
.11
Temporal terms Conventional time
.06
Conventional within 24 hrs
.45*
Sequence
.58**
–.11 .24
*p < .05. **p < .01. ***p < .001.
past events. Perhaps this more general memory prompt was less effective for younger children who required more specific memory questions to elicit responses. When discussing future events, older and younger children’s participation was related to different maternal variables. Younger children provided more new information mothers when used temporal terms that referred to the immediate 24-hour period as well as temporal sequence terms. This suggests that younger children were better able to think about future events occurring later in the same day or the next day as compared to events in the more distant future. In addition, they seemed to benefit from explicit references to event sequence. The following excerpt illustrates how a mother of a younger child used sequence terms when discussing an immediate future event: Example 4: 2½-year-old, routine event: Going to the park M: David, do you want to go to the park tomorrow or the next day? What do you like to do at the park?
7. THE DEVELOPMENT OF FUTURE THINKING
143
C: I like to play. M: you like to play at the park and what do you do? D: Play on the slide. … M: They have a lot of slides, right? Which one’s your favorite? C: The bumpy one. M: The bumpy one. Which one are you gonna go on first when you get to the park tomorrow? C: I do the slide that goes under the tunnel. M: The one that’s the tunnel, that the first one you’re gonna go on? And then which one? C: Then I’m going to go on the bumpy one. M: Then you’re gonna go on the bumpy one. So first you’re gonna go on the tunnel slide and then you’re gonna go on the bumpy slide. Is there anything else you’d like to do there? The data in Table 7.3 also show that older children’s elaborations were significantly related to mothers’ references to general event knowledge, hypothetical references, and use of conventional time markers beyond the immediate 24 hours, but were negatively correlated with mothers’ references to past events. One interpretation of these findings is that 4-year-olds, as compared to younger children, were better able to conceptualize future events as hypothetical entities occurring in an extended future time frame. They responded with new information when asked to think hypothetically about future events and when asked about their general event knowledge (see also Example 3). Mothers’ references to both hypothetical events and to general event knowledge are shown in the following examples: Example 5: 4-year-old, novel event: Trip to the shore M: C: M: C: M: C: M:
And you know where else we may go this summer? Where? We may go to the shore? Which shore? The New Jersey shore. Have you ever been to a shore? Yeah, at North Carolina. That’s a shore. You are right, well this will be a shore in New Jersey. We may go. We will do the same type of stuff. What did we do in North Carolina?
C: I don’t know. M: Well what do you think you would want to do if we went down to the shore?
144
HUDSON C: I would like to get some water and some flowers and plant them. … M: You are so cute. What else would you like to do down there? C: Uh, I would like to throw some big sticks into the water. M: Some big sticks into the water? C: Yeah. M: Are there any big sticks down there at the shore do you think? C: We can bring some sticks. M: We can bring some sticks, yeah. What else do you think you’d like to do? Example 6: Age 4 years, novel event: Mommy and Daddy going away for the weekend M: You know something, this weekend, Mommy and Daddy are going to go away, okay, and someone is going to come and baby-sit. Do you know who it is? It’s Nicole. She used to baby-sit for you when you were little, only one year old. C: When I was a baby? M: When you were a baby. What do you think you’ll do with Nicole? Do you know? C: I don’t know. M: You don’t know. Take a guess. Go ahead. C: Jump on the trampoline. M: Jump on the trampoline. … What if she says, “Tanya, no don’t do that.” What would you do? C: I wouldn’t do it. M: You won’t do it? How about bedtime? What would you do a bedtime? C: Go to sleep. M: Did you say go to sleep? What will Nicole say? C: Goodnight. M: Goodnight. That’s it? She’s going to send you upstairs and say goodnight? C: And shut off the lights.
These examples also illustrate how eliciting children’s views about what they would do or would like to do using hypothetical language was more effective in eliciting a response than querying children about specific past episodes. As illustrated in Example 1, asking specific questions often yielded an “I don’t know” response, whereas asking about what happens in general or what could happen was
7. THE DEVELOPMENT OF FUTURE THINKING
145
more likely to elicit an elaborative response. However, this type of elicitation was more effective for 4-year-olds than for 2½-year-olds.
IMPLICATIONS FOR CHILDREN’S UNDERSTANDING OF THE FUTURE The analyses of mother–child conversations about past and future events have yielded two important generalizations. First, mothers structure conversations about past and future events in different ways; how events are talked about depends to a great degree on whether the participants are reminiscing about the past or planning for the future. Second, there are marked age differences in how mothers and children discuss future events. These may reflect developmental differences in the children’s temporal understanding in general, and in their ability to conceptualize future events in particular. One striking difference in how mothers structured conversations about the past and future was the variety of types of temporal reference that were used in discussing the future as compared to the past. In future conversations, mothers talked about both future actions and general event knowledge and used more hypothetical references. Even talk about the past, although less frequent than other temporal references, was nevertheless included on an average of four times per conversation. The temporal frame switching from past to general to future can be seen in Examples 1, 3, and 6. One implication of this finding is that talk about the future both requires more understanding of time and also provides a natural context for learning time concepts. That is, the greater temporal references used in future conversations as compared to past conversations, indicates that mothers’ language in talking about the future was conceptually more difficult for children to understand. However, mothers’ references to temporal concepts such as tomorrow, in two weeks, every year, on Saturday, and in the summer time in the context of discussing future everyday events provides a context for children to understand what these temporal terms refer to and how to use language to construct a mental time line. As discussed earlier, mothers frequently switched temporal frames when talking to children about future events and used more temporal terms in discussing future events as compared to past events. To understand what their mothers are referring to, children must differentiate specific events from general event knowledge and understand how these events are ordered along a time line that starts before the present and extends into the future, such as the hypothetical time line displayed in Fig. 7.3. It is noteworthy that younger children contributed less to conversations about future events than did older children, suggesting that they may have had difficulty accessing such a mental time line. Although 4-year-olds were able to participate in past and future conversations with equal ease, 2½-year-olds may have found the future conversations more difficult to understand.
146
HUDSON
FIG. 7.3. Diagram of a mental time line showing how temporal terms are used to point to different locations in the near and distant past and future. The event box at the top represents how knowledge about the sequence of actions within a single event is represented in a GER (generalized event representation). Note that sequence terms can refer to events within a narrow temporal frame, such as the sequence of events in a GER, as well as a much larger time line reaching into the distant past and future. Sequence terms refer to the temporal ordering of events and do not point to specific points in time. However, conventional temporal terms are used to point to specific temporal locations of events along the mental time line. Indeterminate interval terms refer to imprecise durations of time, and not specific units of time or particular temporal locations.
Children at 2½ and 4 years of age were also quite different with respect to how they responded to different ways of structuring future conversations. Although younger children contributed less information overall to future conversations as compared to past conversations, their contributions were positively correlated with mothers’ references to events occurring within the immediate 24-hour time frame. This is consistent with the view that temporal proximity is an important construct in young children’s conceptions of time and suggests that reference to more proximal future events facilitates younger children’s contributions to future event discussions. In contrast, 4-year-old children’s elaborative responses were related to mothers’ use of general and hypothetical time frames and to conven-
7. THE DEVELOPMENT OF FUTURE THINKING
147
tional time references, Mothers’ increased use of conventional time terms when talking to older children appears to be an appropriate adaptation. Research has generally shown that understanding of conventional time terms occurs much later than understanding of sequence terms and the finding that older children’s information responses were related to mothers’ use of conventional time terminology indicates that mothers were keeping pace with children’s developing understanding of temporal language. This interpretation implies that mothers use of temporal language does not contribute to children’s emerging temporal concepts, but rather that children’s conceptual development is what affects how much of mothers’ temporal language is understood. However, there are other indications that mothers’ language can contribute to the development of temporal concepts. For example, research has shown that temporal proximity is a salient concept for preschool children, perhaps even more salient than the distinction between past and future. In these conversations, mothers used terminology referring to temporally proximal events (today, tomorrow, yesterday) with equal frequency across past and future events. However, when talking about future events, mothers frequently mentioned past events and referred to general event knowledge. Such frame switching would make it more difficult to differentiate future from past events. Clearly, the next phase of research in this area must focus on studies that allow for better interpretation of the causal direction of the effects observed here.
WHERE DO WE GO FROM HERE? Examining how mothers talk to children about both past and future events indicates that conversations about events are rich sources of information about temporal concepts. There is evidence that variations in mothers’ use of temporal terms and temporal frames of reference are associated with variations in children’s contributions and these relations vary by age of the child and by type of conversation, either past or future. These variations in relations between mothers’ use of temporal language and children’s contributions to conversations are important to investigate further because mothers’ use of temporal language did not vary solely by child age. Mothers used similar levels of temporal language overall when talking with 2½ and 4 year old children, but different maternal language variables facilitated older and younger children’s involvement in the conversations that may correspond to children’s level of temporal understanding. It is also clear that conversations about past and future events are structured differently by mothers and that children participate in different ways in these two types of conversations. Whereas mothers use of prompts, reminders, and elaborations may facilitate children’s participation in memory conversations, these devices were not related to child participation in future conversations. Re-
148
HUDSON
searchers examining parent–child interactions in memory conversations have speculated that the underlying goal of joint reminiscing, that is, the social and affiliative goal of sharing memories together, is what drives the structure and flow of memory conversations. Similarly, a better understanding of the realworld functions of future event conversations can help to understand mothers’ and children’s participation in future event talk. For example, was mothers’ reference to past episodes and general script knowledge in future conversations a function of the not-well-defined task of simply talking about a future event ? Would the same degree of past and general reference in real-world situations be present when mothers were actually preparing children in advance for an upcoming event? Another topic that warrants further investigation is the exploration of individual differences in the ways that mothers elicit children’s participation in future conversations. Recent research has suggested that variations in maternal elicitation style in past conversation are related to other variables such as temperament and attachment style (Reese, 2002). The results from this investigation indicate that mothers structure past and future conversations differently and the same variables (e.g., elaboration ratios) may operate very differently across conversations. Nevertheless, there may be important variables such as the degree to which mothers elicit children’s desires and preferences in future talk that are associated with other greater child involvement. Hypothetical language and attempts to elicit children’s input into decision making by querying children about their wishes and desires may be ways that parents can scaffold children’s participation in constructing plans for future events (Hudson, 2000, 2002). The results from this investigation provide some support for this idea; mothers’ use of hypothetical language was associated with more contributions by 4-year-olds. The degree to which mothers incorporated children’s comments into their comments can also be examined as another measure of how much children’s contributions were acknowledged by mothers. Ultimately, mothers’ efforts to include children in the planning process may be more important in shaping children’s conceptions of future time than the specific temporal terms that they use. Finally, researchers need to investigate how hypothetical and planning talk in future conversations relates to the development of planning skills in children. Is there a relation between how parents talk about the future with their children and how well children are able to plan for future events? These data indicate that hypothetical talk is not frequent in conversations with parents and preschool children, but does it increase with child age? Longitudinal studies of parent–child talk about future events and separate assessments of children’s ability to plan for future events can address these questions. Research on narrative and co-constructed event narratives demonstrates that how individuals talk about events has a profound affect on how they think about events, other people, and themselves. Examining how children talk about the future provides a window onto their conceptions of the future that include concepts
7. THE DEVELOPMENT OF FUTURE THINKING
149
of temporal relations, thinking about hypothetical concepts of the enduring self, and projections of future relations.
ENDNOTE 1
See Hudson (2002) for more details regarding the coding of the conversations.
REFERENCES Benson, J. B. (1997). The development of planning: It’s about time. In S. L. Friedman & E. K. Scholnick (Eds.), The developmental psychology of planning: Why, how, and when do we plan? (pp. 43–75). Hillsdale, NJ: Lawrence Erlbaum Associates. Engel, S. (1986). Learning to reminisce: A developmental study of how young children talk about the past. Unpublished doctoral dissertation, City University of New York Graduate Center, New York. Farrant, K., & Reese, E. (2000). Maternal style and children’s participation in reminiscing: Stepping stones in children’s autobiographical memory development. Journal of Cognition and Development, 1, 193–225. Fivush, R. (1997). Event memory in early childhood. In N. Cowan (Ed.), The development of memory in childhood (pp. 139–161). Hove, East Sussex: Psychology Press. Fivush, R., & Fromhoff, F. (1988). Style and structure in mother–child conversations about the past. Discourse Processes, 11, 337–355. Friedman, W. J. (1977). The development of children’s knowledge of cyclic aspects of time. Child Development, 48, 1593–1599. Friedman, W. J. (1990). Children representations of the pattern of daily activities. Child Development, 61, 1399–1412. Friedman, W. J. (2000). The development of children’s knowledge of the times of future events. Child Development, 71, 913–932. Friedman, W. J., & Brudos, S. L. (1988). On routines and routines: The early development of spatial and temporal representations. Cognitive Development, 3, 167–182. Haden, C. A. (1998). Reminiscing with different children: Relating maternal stylistic consistency and sibling similarity in talk about the past. Developmental Psychology, 34, 1366–1371. Haden, C. A., & Fivush, R. (1996). Contextual variation in maternal conversational styles. Merrill-Palmer Quarterly, 43, 200–227. Haden, C. A., Haine, R. A., & Fivush, R. (1997). Developing narrative structure in parent–child reminiscing across the preschool years. Developmental Psychology, 33, 295–307. Hudson, J. A. (1990). The emergence of autobiographic memory in mother–child conversation. In R. Fivush & J. A. Hudson (Eds.), Knowing and remembering in young children (pp. 166–196). New York: Cambridge University Press.
150
HUDSON
Hudson, J. A. (1993). Understanding events: The development of script knowledge. In M. Bennett (Ed.), The Child as psychologist: An introduction to the development of social cognition (pp. 142–167). London: Simon & Schuster. Hudson, J. A. (2000). The remembered and anticipated self: mother–child talk about past and future events. In K. S. Skene & C. Moore (Eds)., The development of the temporally extended self in preschool children: Theory and research (pp. 55–74). Mahwah, NJ: Lawrence Erlbaum Associates. Hudson, J. A. (2002). “Do you know what we’re going to do this summer?”: Mothers’ talk to young children about future events. Journal of Cognition and Development, 1, Hudson, J. A., & Fivush, R. (1991). Planning in the preschool years: The emergence of plans from general event knowledge. Cognitive Development, 6, 393–415. Hudson, J. A., Shapiro, L. R., & Sosa, B. B. (1995). Planning in the real world: Preschool children’s scripts and plans for familiar events. Child Development, 66(4). Hudson, J. A., Sosa, B. B., & Shapiro, L. R. (1997). Scripts and plans: The development of children’s event knowledge and event planning. In S. L. Friedman & E. K. Scholnick (Eds.), Why, how, and when do we plan? The developmental psychology of planning (pp. 77–102). Hillsdale, NJ: Lawrence Erlbaum Associates. Lucariello, J. & Nelson, K. (1987). Remembering and planning talk between mothers and children. Discourse Processes, 10, 219–235. McCabe, A., & Peterson, C. (1991). Getting the story: A longitudinal study of parental styles in eliciting narratives and developing narrative skill. In A. McCabe & C. Peterson (Eds.), Developing narrative structure (pp. 217–253). Hillsdale, NJ: Lawrence Erlbaum Associates. Nelson, K. (1986). Event knowledge: Structure and function in development. Hillsdale, NJ: Lawrence Erlbaum Associates. Nelson, K. (1993). The psychological and social origins of autobiographic memory. Psychological Science, 4, 1–8. Nelson, K. (1996). Language in cognitive development: The emergence of the mediated mind. Cambridge, England: Cambridge University Press. Reese, E. (2002). Social factors in the development of autobiographical memory: The state of the art. Social Development, 11, 124–142. Reese, E., Haden, C. A., & Fivush, R. (1993). mother–child conversations about the past: Relationships of style and memory over time. Cognitive Development, 8, 403–430. Sachs, J. (1983). Talking about the there and then: The emergence of displaced reference in parent–child discourse. In K. E. Nelson (Ed.), Children’s language (Vol. 4, pp. 3–28). Hillsdale, NJ: Lawrence Erlbaum Associates. Scholnick, E. K., & Friedman, S. L. (1993). Planning in context:? Developmental and situational considerations. International Journal of Behavioral Development, 16, 145–167. Weist, R. M. (1989). Time concepts in language and thought: Filling the Piagetian void from 2 to 5 years. In I. Levin & D. Sakae (Eds.), Time and human cognition: A life-span perspective. North-Holland: Elsevier.
8 Narratives, Gossip, and Shared Experience: How and What Young Children Know About the Lives of Others Susan Engel and Alice Li Williams College
“My earliest memory is dipped in red.” So begins the autobiography of Elias Canetti, the 1981 Nobel Prize winner for literature (Canetti 1979). Deeply personal and idiosyncratic, this kind of memory fragment has a lot to tell us about how people recall their lives. We have learned a great deal in the past 20 years about how and why people remember their past. In at least some of this work, psychologists have tried to trace the developmental pathways that connect young children’s casual and conversational recollections of their experience to the finely crafted, sometimes searing memories recounted in authors’ autobiographies. So what causes many people to tell stories about their past when they are little? How does the process emerge, and what individual and cultural forces determine people’s proclivity and ability to recollect their past as they get older? Why do some people (e.g., Elias Cannetti) seem to retain or rediscover vivid forms of autobiographical 151
152
ENGEL AND LI
narrative, whereas so many, at least in this culture, lose whatever narrative sparkle they had? And what difference, if any, might this sparkle make in the life experience of an ordinary adult? The novelist draws on personal memory, and perhaps represents the extraordinary version of an ordinary, universal, activity—telling stories of personal experience. But an individual’s own memories are only part of the story. The second, less famous, line of Leo Tolstoy’s Anna Karenina goes like this: “Everything was confusion in the Oblonksy Household.” Tolstoy’s writing, exemplified in this sentence, represents the asymptote of another, equally essential human process—telling stories as a way of thinking and communicating about other people’s experience (Tolstoy, 1978). Not only do humans tell stories to create, communicate and know themselves, and understand how the world works, they also tell them to know other people. When and how do we listen to other people’s stories, and how might this contribute to our knowledge and understanding of another aspect of the world, for example, the world of other people’s consciousness? This chapter describes what has been learned about children’s autobiographical narratives and how understanding grew in large part out of the framework Katherine Nelson has been articulating over the past 20 years. It discusses a newer line of inquiry to emerge from this framework—when and how children talk about other people’s lives. Somewhere along the line, children may listen to the stories their friends tell about themselves, and at the same time they may begin to use the narrative form to think about the experience of other people. But just as autobiographical conversations long precede a fuller sense or use of autobiography, gossip (i.e., the stories heard and told about others) is a relatively late appearing phenomenon. If someone happens to walk into a day-care center for toddlers they will see great industry, but hear little gossip. Children may be building blocks, draping themselves with neckties and scarves, swishing in the water table, smacking each other over the head with plastic dinosaurs, or even answering questions about why another little girl is sobbing and has a red patch on her arm from being smacked. Next door in the 3- and 4-year-old room, there may be some version of these same activities. But there will also be a great deal of conversation. Some will involve heavy negotiations over the exact nature of the game (e.g., “no, we are super guys, and this will be the planet”). Some of it may involve a discussion of rules (e.g., “You can’t have that one. Sarah said I could use that one”). But a good deal of it will involve the telling of personal narratives, or at least the expression of narrative fragments (e.g., “I used to have one like this. I used ta , I used ta, my mom gave me a big one just like this”). In fact, a room full of 3- and 4-year-olds is a room full of personal narratives. But what, if anything, does each child take in from the swarm of stories and narrative fragments wafting through the verbal atmosphere? The discussion begins by tracing the origins of this idea in the work of Katherine Nelson and some of her students, and shows how one path out of her laboratory led from scripts to gossip. In 1980, Nelson had just completed the first set of stud-
8. NARRATIVES, GOSSIP, AND SHARED EXPERIENCE
153
ies showing that young children are able to articulate and draw on knowledge about the skeleton of routine events before they are able to describe specific memories or experiences they have had (Nelson, 1986). For example, a child who simply shrugs and cannot answer when asked what he had for breakfast that morning can answer the question “What do you have for breakfast?” with the following: “Usually I have Cheerios and milk, but sometimes when it’s cold mommy makes oatmeal.” Along with this came the discovery that these skeletons usually take the form of scripts: They specify people, places and actions, and organize that information around salient goals (going out to dinner, going to day care, having a birthday party, bedtime etc.). In Nelson’s work, the role of the other became more salient over time. This came about for several reasons. For one thing, the idea that the social meaning of routines was important is implicit in the theory itself: If routines such as birthday parties and bedtime rituals form the bedrock of early mental representation, the significance of the social nature of these routines cannot be ignored. Early on, Nelson’s theory emphasized the good fit between events in the world, and the child’s representation of those events. This had a somewhat Gibsonian flavor to it. The way the child thought matched the way the world was organized. A term often used in those early days to describe the mechanism by which the child’s mental representation fit onto the events in the world, was “map,” suggesting some kind of neat fit between the representation and the represented. However, it became clear over time that social meanings were central to the process Nelson was describing. Moreover, some of the research her students were engaged in showed that when children talked about events, a great deal of it emerged in conversations with the people around them. The powerful role the other plays in the process of reminiscing was first made clear by accident. The first author, a student of Katherine’s, was engaged in a study that was supposed to identify some of the ways in which children first used words free of their initial context. In order to watch this process closely, tape recordings were taken of a 16-month-old girl getting up and having breakfast with her mother several times a week, for a period of months. The goal was to capture word use, but the tapes also captured a surprising amount of talk about the past and the future. This led to a closer look at the interactions that supported and influenced the child’s increasing ability to narrate experience. It led to the belief that not only were children learning to tell stories and talk about their personal past at an early age, but that this was also a profoundly interactive process. In fact, early on, Engel (1986) argued that talking about the past was an intersubjective process in which it was hard and possibly meaningless to disentangle what was “in the adult’s mind” from what was “in the child’s mind” (Engel 1986). That work traced the process by which children constructed descriptions of their past experiences in conversation with others. The data showed that even before young children can name events from their past, or describe the past in a way that conveys perspective, sequence, or interpretation, they can participate in such a
154
ENGEL AND LI
description. Moreover, the role of the other seems to be important in not only triggering, but shaping the kind of reminiscing in which a child engages. The early work on reminiscing came at a time when the importance of autobiographical memory was on the ascendance. Psychologists, and developmental psychologists in particular, were zeroing in on the many functions of autobiographical memory, and beginning to trace its emergence in young children. It seemed that young children used personal memories to solve all kinds of emotional and cognitive puzzles, about themselves and the nature of the world.
CHILDREN’S AUTOBIOGRAPHICAL NARRATIVES— MAKING SENSE OF THE WORLD AND OF THE SELF Children begin telling stories about their own experiences when they are as young as 18 months of age (Engel, 1995). Research has shown that children can re-enact small bits of experiences when they are as young as 9 months (Carver & Bauer, 1999; Carver, Bauer, & Nelson, 2000), and by 13 months they can accurately recall events even after a long period of time (Bauer, 1996). Then, by age 2½, they can organize their recollections in a story form, providing an organized and coherent account when asked about their past experience (Fivush et al., 1987). Through the process of creating these stories, they gradually make sense of the language, people, events, and places in the world, and how the self fits into that world. Indeed, children use language in the form of general scripts and specific narratives, to construct stories about events that have happened, are currently happening, or that will happen. Stories can play the dual role of reflecting individuals’ experience and helping them to make sense of the experience. For example, by “representing familiar events in canonical causal-temporal sequences of actions,” organized in terms of central goals, objects, and roles that people play (Nelson, 1993, p. 2), scripts provide children with a narrative “skeleton” into which they can order and understand new events in their lives. After children have established a sufficient body of script knowledge, they are then able to use that knowledge as a background from which to remember or reconstruct memory for a specific and/or novel event (Nelson, 1993). Narratives also allow children to sequence, to identify specific components of the past (place, actors, actions), and to begin to locate a subjective self within the past. Four dimensions of narratives facilitate the understanding of experiences: sequencing, canonicality, intentionality, and perspective taking (Bruner & Lucariello, 1989). Through sequencing, conjunctive, temporal, and causal order are introduced into the experience. When describing an event, people make a distinction between what is ordinary and what is unusual—thus canonicality. The storyteller also indicate the actor’s intentions and feelings—thus intentionality. Finally, the narrator assumes a stance, or perspective, toward what is represented in the narrative. In the process of translating the experience into a story, these nar-
8. NARRATIVES, GOSSIP, AND SHARED EXPERIENCE
155
rative features help to facilitate the ordering and understanding of experiences. Bruner and Lucariello first outlined the way in which the now famous Emily, at 16 months, sorted out the usual from the unusual through her nightly bedtime monologues about her day. One important feature of their argument was the idea that in sorting out these characteristics of experience, children (and presumably adults) are also sorting out characteristics of stories (they use the term canonical and noncanonical). Thus, they and others have argued that understanding self in the world is accomplished through the understanding and use of stories. Narrative is essential, in other words, to human understanding. Not only do autobiographical scripts and narratives provide children with mental representations of how the world works, but the process of constructing these stories also allows children to discover and develop a sense of self in the world. The storytelling process, whether to ourselves or to others, is an important self-defining activity. By recounting, contemplating, and interpreting their experiences, children can learn to reflect and think about different aspects of their identity, and eventually construct a self that has continuity through time and space (Labov & Waletzky, 1997; Neisser, 1988).
THE ROLE OF OTHERS From the beginning, the development of autobiographical narratives takes place in social situations. During the initial nonverbal stage of development, the emergence of self in stories is made possible with the help of parents. For example, infants listen to the stories that their parents tell them about themselves. Through these stories, children learn about their past and how that past fits into the bigger social context. With increasing verbal and cognitive abilities, children become more active participants in the co-construction of their autobiographical stories with their parents or other family members. As children grow older and begin to spend more time in day-care and preschool settings, they are eager to share those personal stories with their peers and people outside of the family. They use stories to guide them through play and social interactions. For example, narrative knowledge helps children to assign roles and create plots in their pretend play (Engel, 1995). Stories about personal experiences are also powerful tools for managing social relationships. They initiate, break, and maintain intimacy through the stories they tell to accompany their symbolic play (Anderson, 1993). Moreover, the storytelling process in child–child interactions provides further opportunities for children to develop and define their identities, because the stories themselves, especially those about personal experience, are important vehicles for self-representation (Ely, 1997). Children construct actual and possible selves through selective memory of past experiences and imagination. Through factual and fantastical narratives, they present themselves in ways they want to be seen by
156
ENGEL AND LI
their peers, and children are highly attentive to each other’s stories. The storytelling practice with friends allows young children to build up a repertoire of stories that highlight different aspects of the self. These are then building blocks for constructing a sense of identity, a sense of who they are. But do these stories also give children a sense of who their friends are? Whereas much work has been done investigating ways in which independent and co-constructed storytelling helps to shape children’s sense of self, less is known about how these narratives might also contribute to children’s emerging conceptualization of other people. For young children, the spheres delineating the self and others are initially vague and fuzzy. Only gradually do they begin to differentiate the two as separate wholes. Keller, Ford, and Meacham (1978) found that when preschoolers (3-, 4-, and 5-year-olds) are asked to describe themselves, they most frequently define themselves in terms of activities (e.g., “I can pick up things”) and rarely in interpersonal relationships (e.g., “I have nice friends”). Not surprisingly, when it comes to describing other people, young children generally focus on more physical than psychological aspects of a person, often centering on activities and momentary moods. However, as they enter into the school years, the importance of psychological attributes increases (Secord & Peevers, 1974). As children begin to branch outward in their social world, peer relationships become increasingly more salient and important. Friends serve important functions not served by parents, such as opportunities for learning social skills (e.g., negotiation), facilitating social comparisons, and fostering a sense of group belonging (Rubin, 1980). The latter two are central in shaping the children’s sense of identity. Some understanding of friendship, therefore, is necessary to this line of inquiry.
THE NATURE OF FRIENDSHIP IN YOUNG CHILDREN Selman’s structured clinical interviews with children led him to posit the development of friendship as a series of transformations (1981): Stage 0: Young children, from about age 3 to 5, typically conceptualize friends as “momentary physical playmates.” Children at this stage characteristically only consider the physical attributes, physical accessibility, and activities of playmates, rather than psychological attributes, such as interests or personal needs. Stage 1: Children at this stage, around ages 6 to 8, typically view friendship as “one-way assistance,” in the sense that friends accomplish activities that the self wants. Knowing at this stage thus involves being aware of one’s likes and dislikes. Stage 2: Children advance to this stage of “fair-weather cooperation” around age 9 to 11, and begin to see friendship as a two-way street, rather than just one-way. However, there is no underlying continuity that can maintain the friendship dur-
8. NARRATIVES, GOSSIP, AND SHARED EXPERIENCE
157
ing times of conflict and adjustment. The reciprocity is only focused on specific incidents, rather than on the friendship as an enduring social relationship. Stage 3: By around 12, children typically reach the stage of “intimate and mutually shared relationships.” At this time, friends are regarded as providers of emotional support and intimacy. There is also an underlying continuity that can sustain the friendship if conflicts occur. Children also realize at this stage that close friends need to be psychologically compatible—share similar interests and mutually agreeable personalities (Rubin, 1980). This stepwise progression in children’s concept of friendship is paralleled by three dimensions of social understanding (Rubin, 1980). First, there is a shift from the children’s initial egocentric view of friendship to their later ability to take on other people’s point of view. Second, there is progression from seeing people as concrete, physical entities to abstract, psychological entities. Children become increasingly more likely to provide psychological explanations for their friends’ behavior and this psychologizing gets more sophisticated with age. Finally, there is a shift in the child’s conception of friendship from viewing it as a transitory interaction to a more permanent and lasting social relationship. Overall, the common thread across the three is that there is a general progression from focusing on the concrete, observable characteristics of people to other more abstract and inferred traits (Rubin, 1980). In addition, it is not a surprise that peer interaction is an important context for children to develop these strands of social understanding. Dunn (1996) discussed specific social processes that allow children to do so. These include repeated experiences of sharing and negotiating in a pretend world with playmates; of managing arguments and conflicts, in which the child is faced with points of view that are different from his; and of conversing about various aspects of the social world, such as talks about mental states and narratives. Surprisingly, we know very little about children’s narratives about others. We do know that children are not only enthusiastic about stories about their own lives, but are also fascinated by stories about other people’s encounters. Indeed, as children age and begin expanding their social network, they make increasing references to others in their narratives. For example, in The Beginning of Social Understanding, Dunn (1988) stated that during childrens’ third year their narratives about others increase dramatically, and he argued that this parallels children’s increased curiosity about other people. Miller et al. (1992) also demonstrated that compared to younger children, the 5-year-olds in their study portrayed more of their personal experiences in conversations as interpersonal, made more reference to peers, and expressed slightly more modes of self–other relations (e.g., joint activity and social comparisons). When asked to describe others, younger children tend to use more undifferentiating terms, such as possession and location items than do older children. As children progress from being self-focused to more other focused, they gradually employ more differentiating
158
ENGEL AND LI
(i.e., personal characteristics like interests, likes, beliefs, values, feelings, etc.) and dispositional terms (i.e., describe the person in terms of traits with implications for his behavior in a range of situations) (Secord & Peevers, 1974). Preece tape-recorded naturally occurring narratives produced by three children (ages 5–7) as they rode to and from school for 18 months (1987). Out of the 16 forms of narratives she identified, two of them relate to narratives about other people: “anecdotes of vicarious experience” and “tattle-tale.” Together they account for 25% of the total number of narratives identified. Anecdotes of vicarious experience tend to depict out-of-the-ordinary events, transgressions, and misadventures, and were used more often as the subjects got older. Dunn (1988, 1994) showed that as children move from toddlerhood to preschool age, their conversations contain more references to feelings. However, children’s narratives about others are not merely windows onto their level of emotional understanding. As Bruner so compellingly argued, personal narratives create two levels of meaning: the external and internal worlds. The external landscape refers to the sequence of events and the action in the story. The internal landscape conveys thoughts, motivations, fantasies, and emotions. We already know something about how children’s narratives depict inner and outer landscapes of their own lives. Also something is known about how older children navigate these levels of meaning in the stories they might hear (Feldman, 1989). But enough is not yet known about how children exchange stories with one another as a way of learning about other people’s experience. Steedman (1982) analyzed a story written by a group of schoolgirls, as well as the tape-recorded sessions during which they wrote the story. Her analyses showed that the girls use the storytelling process as well as the text to negotiate social understanding between one another, and also as a way of exchanging ideas and knowledge about the larger social world. The girls in Steedman’s study were 8 and 9 years old. What about younger children? Where and how do stories first emerge as a means of learning about the lives and experiences of others?
STUDYING BORROWED EXPERIENCE: A FIRST ATTEMPT In order to begin to answer this question, 3 and 4 year olds were recorded in their day-care centers to see if children do in fact tell one another anything about their lives. The investigation looked in three places: the snack table, the block corner, and the art center. It found that whereas children narrate their play and negotiate roles in the block area, and discuss their activity and their art in the drawing area, it is over food that they tell one another about their lives. Data showed that 3-year-olds exchange virtually no information with one another about their lives. In contrast to the 3-year-olds, who rarely spoke to one another about their lives in the past or at home, 4-year-olds regularly sent out trial balloons, or brief utterances about life at home (e.g., two preschool children are
8. NARRATIVES, GOSSIP, AND SHARED EXPERIENCE
159
playing on the edges of the block corner, and one says to the other, “My baby sister, she, last night she barfed!” He keeps his eye on the second child for a moment, as if waiting to see if his playmate will pick up this line of conversation). Often these trial balloons elicited little or no response from peers. In some cases, a 5-year-old or an adult would eventually step in and ask the kind of question that would allow the 4-year-old to elaborate. It seemed clear that the 4-year-olds were trying to use autobiographical information as a way to engage their peers. By the time children are 5 they engage in genuine conversation (with adults and other 5-years-olds) about their lives. They talk about things they have done (e.g., “we went to Disney world”), they describe behaviors of family members (e.g., “My brother always jumps on me”) and they talk about plans for the future (e.g., “We’re gonna get a new house”). It was only among the 5-year-old peers that children actually asked one another about their lives at home. For example, one 5-year-old pointed to a burn on another 5-year-old’s lip. Child 1—How did it get burned? Child 2—At home. Child 1—Did it hurt when you got it? Child 2—(nods). Autobiographical exchanges, regardless of age, ocurred almost exclusively at the art and lunch table. There were virtually no references to the children’s lives in free play settings. Children also made autobiographical references for a variety of social and affective purposes. These purposes include gaining intimacy, achieving dominance, entertaining, and presenting a self. These last two are of particular interest in the current context. For example, during a conversation between four 4-year-olds at a lunch table, they have been discussing how gross it is when food falls on the floor, in response to one child dropping a piece of melon on the floor. One little boy suddenly bursts out, “Ya know what? Ya know what? My sister eats off the floor!” He is trying to liven up the discussion and get some attention by putting his own experience up for consideration. Consider also the following example of a 5-year-old girl addressing a group of children at her snack table: “My mom has lollipops at her office. She brings ’em home for me sometimes. And when I was little, I’d ask her to bring me home a lollypop and when she got back, the first things I’d say would be mommy, did you bring me home a lollypop? I used to say it all the time.” There is much you might say about what these children do for themselves when they offer such autobiographical pieces to their friends. But it seems equally interesting to think about what these utterances might mean, or begin to mean to the listeners, the other children. They provide opportunities, perhaps bootstraps, for children to think about the experience of others, and to recognize that other people have experience—extended, private, and displaced.
160
ENGEL AND LI
These data indicated that children do tell friends about their lives, but often these reports are tentative and remain undeveloped. Children do not seem to pick up on one another’s information, or scaffold in the way the teller needs. Although this is not surprising, given what is known about other aspects of the preschooler’s development, it is important in understanding the fate of autobiographical information. Children at this age may be telling, but they are not necessarily hearing.
NEW QUESTIONS, NEW RESEARCH In order to find out if young children actually talk about one another’s lives, children were tape-recorded talking to one another, in day care and school. It was extraordinarily difficult to catch them gossiping. Children clammed up on this topic the minute researchers were in the room. If they were gossiping about one another, they were not going to do it in front of a strange adult. The next step was to ask children to talk about the lives of their friends. There is abundant evidence showing that the audience has significant influences on children’s narratives. For example, parents exhibit different styles of eliciting narratives from their children. These differences have an impact on the verbal style and/or ability of their children (McCabe & Peterson, 1991). Furthermore, children themselves often give each other feedback about what makes a story good (i.e., what topics are exciting to hear about, what is the most interesting way to begin a story that would grab the audience’s attention, where and when to dwell on and elaborate (or not) certain points of the story, etc.; Engel, 1995; Brice Heath, 1983). Finally, other researchers, such as Pellegrini and Galda (1990) examined the effect of various experimenters’ questions on a child’s narrative responses. The importance of interlocuter led to a further look at the effect of experimenter; would different interviewers elicit different types of stories about others? It seemed that the way the experimenter probes for the stories should influence the kinds of stories elicited. Two different college students conducted the study, asking 4-, 7- and 10year-olds to talk about their friends. The subjects in this study were 43 predominantly middle-class White children, of which 21 were female, from a rural community in Massachusetts. The interviews were conducted in a quiet space in the children’s school, and each child was interviewed alone. The setting was informal, the tone conversational. One of the two female experimenters asked children the following questions: Can you tell me who are your good friends? Who is your best friend? Can you tell me some stories about him/her? The experimenter followed up with “Can you tell me more about it?” (referring to whatever the child started to talk about). Throughout each story, the interviewer would offer neutral and supportive comments, such as “Tell me more,” “Really?”, “Yeah?”, or “Then what happened?”, to show interest in what the child was saying and to facilitate the storytelling process (McCabe, 1997). We were particularly interested in how children acquired autobiographical and narrative infor-
8. NARRATIVES, GOSSIP, AND SHARED EXPERIENCE
161
mation about their friends. Thus, after the end of each story, the experimenter asked, “How did you know about this?” to find out where the child learned about the story.
WHAT WE’VE LEARNED When asked to tell stories about friends, the children in this study provided two kinds of biographical information: stories—either scripts about routine events, or narratives about specific past or ongoing events and simple descriptions about the friend (giving discrete pieces of information not in the context of a story). What follows is a description of the scripts and the narratives the children produced. The transcripts were coded for a range of information, including type (scripts vs. stories), length, detail, level and kind of organization, and the representation of subjectivity (narrators and/or characters in the story). This investigation was interested in developmental differences between the 4-, 7-, and 10-year-olds, as well as the relation between story type and the other measures. It also looked at the effects of interviewer.
SCRIPTS VERSUS STORIES The older children in the study offered more stories about their friends than the younger children. Both total number of scripts and narratives increase as a function of age. Four-year-olds provided, on average, 2.14 scripts and 1.36 narratives; 7-year-olds provided 3.94 scripts and 4.75 narratives; and 10-year-olds provided 5.38 scripts and 5.62 narratives. Even though the numbers of scripts and narratives used at each age are not significantly different from each other, a look at the means suggests that scripts about friends appear earlier than narratives do, and the use of scripts continue to be important in narrative constructions of other people’s lives, as demonstrated by the equal frequency of scripts and narratives at age 10. What is the relation between scripts and narratives about friends? It seems that as children get older, they fill in, modify, refine, and reshape the scripts using more specific narrative experiences. In other words, there is constant construction, deconstruction, and reconstruction of scripts as a framework to accommodate the new and more complex experiences and insights involving others. The study also found an experimenter effect in the number of narratives obtained. Although children do produce more narratives as they get older, this number varies depending on the interviewer (one experimenter elicited an average of 5.04 scripts per subject, and the second an average of 2.33 per subject). However, the experimenter did not affect the number of scripts elicited (4.0 scripts per subject vs. 3.50 scripts per subject). This provocative result indicates that somehow the formation of scripts is more impervious to external experimenter influences. It seems to suggest that scripts are a less socially influenced form through which
162
ENGEL AND LI
children make sense of experience. This finding provides indirect evidence for the notion that scripts are less socially constructed than narratives, or at least less responsive to the kinds of momentary, situational influences present when children tell stories to the people around them. In order to find out just what it is that might change in children’s stories about other people, the study compared the three age groups along the following dimensions: length, descriptiveness, and evaluative content. Length was assessed by counting number of prepositional comments within a narrative (defined as an account that included specific time and place, actions, characters, and a point of view). Descriptiveness refers, in these data, to the number of description words included in the narrative (e.g., “One time she said she played football with the boys and they threw a ball at her stomach and she threw up because it hit her stomach. And the school has to put this yellow sign around the barf so no one would step into it”). The idea of evaluation is taken, of course, from Labov and Waletsky, and refers to words or phrases that direct the listener/reader to an interpretation about how the story should be taken. A 10-year-old girl, for instance tells a story about a friend who never wants to stay inside and talk during recess, and the narrator says, “She’s okay. But sometimes she gets really annoying.” In terms of the length, descriptiveness, and the number of evaluative comments used in the stories, there were no significant age main effects for the scripts, although there was a slight increase in all three categories. In contrast, the increases in these categories for narratives were much more salient and are all statistically significant. These measures were calculated as absolute counts, but also as proportions of the overall length of the narratives. There are good reasons to take both measures equally seriously, but that is another discussion. The more stringent measure is the proportional one, which is presented next and in Table 8.1. First, it is not surprising that as children age, their narratives about their friends would become longer, more descriptive, and include more evaluations. The nonproportional and proportional data reveal that the stories not only included more descriptions and evaluations because they are longer, but that they do actually become more descriptive and evaluative with age. It has been found that children employ evaluative devices in their personal narratives when they are as young as age 2 (Miller & Sperry, 1988) and the number and type of evaluative techniques increase with age (Hudson & Shapiro, 1991). As children’s cognitive Table 8.1 Length and Structure of Children’s Narratives 4-year-olds Narrative length
1.05
7-year olds
10-year-olds
2.12
5.26
Number of decriptive words
.068
.23
.25
Number of evaluative words
.034
.13
.25
8. NARRATIVES, GOSSIP, AND SHARED EXPERIENCE
163
and linguistic abilities and social understanding advance with age, such skills would give them the tools to more fully understand their friends, and in different contexts. As the line between the self and others becomes more differentiated, children can also increasingly gain the ability to evaluate their friends, the role that they play in the stories, and the stories themselves. Second, consistent with the previous idea that narratives are somehow more likely to be subjected to experimenter influences, the interviewer is found to affect the descriptiveness of the narratives. Although it is true that narratives do become increasingly more embellished and enriched as children get older, the extent of such embellishment, however, is mediated by the audience to whom the child is trying to tell the story. Again the person that the child is interacting with can influence the amount of information, the type of information, the number of stories, and the quality of the stories told by children.
COHERENCE AND STRUCTURE In terms of the structure of the stories, children tell increasingly more coherent and complete narratives about their friends as they get older. Older children are more likely to tell stories that build up chronologically to a high point and end with a resolution. Story structure refers to the level of coherence of a narrative. This measure was originally adapted from Peterson’s (1994) classification of structure. Each narrative was coded in the following way: 0—Narrative is confused or disoriented. “Once M let me come over to her house. And … Mom said she has to go to her house and she put my shoes on for me. And then I walked home but I live very close to M …but not to R, cuz I miss her.”—a 4-year-old girl. “Once … one time H came to my house for a long time. And since I haven’t seen her for a long time, my mom invited H.” (Interviewer: “What happened after she came over?”) “Well she didn’t come over yet. My mom might call her.”—a 4 year-old girl. 1—Chronological. Narrative describes successive events that are sequentially and logically ordered (no high point). “We were on the bus and S was being funny. We were like making silly things. We did that for the whole bus ride, then we got off and went to the museum.”—a 7-year-old boy. “Once she came to my house and we hid some cookies. We played together. And then she went home.”—a 4-year-old girl.
164
ENGEL AND LI
2—Ending at the high point. Narrative builds chronologically to a high point and ends; there is no wrap up or resolution. “In first grade last year, we were doing this project. And suddenly, N would start to say things like ‘flush me down the toilet! Flush me down the toilet!’ And I have a friend, S, who said to N.M, ‘He means you!’ And N was like, ‘HUH!’ ”—a 7-year-old boy. “And once, the teacher said to S, ‘Don’t eat the snow.’ And S said, ‘Why?’ Then the teacher said, ‘It might have cat pee pee on it!’ Hahahahaha!” (laughing)—a 4-year-old boy. 3—Ending with resolution. Narrative builds chronologically to a high point; then resolves. “And C told me she has lots of crystals and also in kindergarten, there was another little girl named E. She’s not very nice; I have to say. She and I were fighting over one, which was a geo. It was a very, very peculiar looking one. It had swirls of blue, purple, and gray, and black. I thought it was beautiful. E loved it and it would break. And I have to admit, I got a lot of scrapes that day but it was worth getting the geo. Because I told E that there’s a bigger one over there. And she was like, ‘Where?’ And I got the geo. And she was like, ‘No, there was not.’ And we both went for the geo. And C was like, ‘E, let K have the geo!’ But E didn’t let me, and C and I tried to push E away and both of us got so much blood dripping down on us. But nobody had to go to the emergency room because Mrs. C (the teacher) cleaned up the blood and stuck like twenty band-As on our legs.”—a 7-year-old girl. However, one has to be careful not to over rely on this kind of coding scheme. It gives the sense that stories by younger children are somehow not up to par with those by older children. Narrative elements can be viewed as contributions to the ongoing conversation and social interaction (Engel, 1995). This characterization of narratives is a reminder that not only does the cognitive context in which children’s narratives are produced need to be considered, but so must the social context of narrative production. Not all the stories elicited are a “Number 3” (the highest score given for the most complete story), and not all stories need to be a Number 3 in real life. Assuming that children tell stories to fulfill certain social functions and to communicate certain points, as long as these functions are served and the message is conveyed, not all stories have to be structurally complete. Although coding criteria are useful for tracking developmental changes in structure, complexity, and the logic of a child’s narrative, it is important not to lose sight of whether all of these have relevance to social situation in which the narrative was created. For example, the interviewer asked C (a 7-year-old girl) to tell a story about C’s friends, M and I: “M and I are really good figure skaters. So when they came to my party they did all these weird tricks. They passed like every single level of skating.
8. NARRATIVES, GOSSIP, AND SHARED EXPERIENCE
165
So we were all like, wow.” This episode would only be coded as a Number 2 story because it ends on a climax instead of some resolution. However, in this case, if the child’s goal is to convey to the experimenter her amazement at her friend’s skating ability, and her story fulfills that function, she might not need to reach a resolution. Furthermore, whether or not a child expands a story also depends on the listener. This explains and is reflected by the experimenter difference found for story structure. One experimenter tended to elicit narratives with higher coherence scores. This is not surprising in that different interviewers probably probed children to complete their stories to different degrees. The extent to which an experimenter delves and encourages a child to elaborate the story would depend on what the former perceives as the communicative function of that particular story. Therefore, the structure of a narrative cannot be taken out of its social context. As Hudson and Shapiro (1991) summed it up, “Different narratives serve different communicative functions and the structure that define narrative genre are a result of those communicative functions. By altering the communicative function of narrative production, we necessarily alter the coherence of the narrative produced.”
Stories of Shared Experience and Stories About Friends’ Experience In terms of the type of experiences described, overall, children gave approximately equal numbers of stories of joint and “alone” experiences. However, stories (both scripts and narratives) about “alone” experiences—events solely experienced by the friend—increase significantly as a function of age. They become increasingly more salient in children’s stories about others as the children get older compared to stories of joint experiences. In contrast, there is no such incline for stories about joint experiences. There were no age differences in the number of narratives depicting joint experiences, and there was actually a significant decrease in scripts describing joint events. This result indicates that real face-to-face interactions, such as joint experiences, are generally important across ages 4–10, serving as essential contexts in which children of this age come to develop an understanding of other people. These experiences (which could include play, conversations, and conflicts) are educational settings from which children can learn more about their friends’ personalities, emotions, and how to negotiate and deal with their friends. The result further shows that as children age, their narrative construction about others becomes more other-focused, as indicated by the increasing number of stories depicting experiences solely experienced by the friend. The stories were categorized as either depicting a joint experience, directly experienced by both the narrator and the friend, or an alone experience, experienced solely by the friend. For example, the following narrative by a 7-year-old boy is a good example of a joint experience: “S is funny. We were on the bus and S was being really funny. We were like making silly things. We did that for the whole bus
166
ENGEL AND LI
ride, then we got off, and went to the museum.” Note that in recounting a joint experience he talks about the friend, but almost completely in terms of his own perspective. The landscape of action pertains to both of them, the landscape of consciousness belongs to the teller. This next narrative by a 10-year-old boy is an example of an “alone” experience: “He told me last week that his brother tripped his leg so he couldn’t run as fast as he usually could [in the race.] So he ran, still the fastest of anybody in the school, but slower, not his usual speed.” This example focuses much more purely on the other child, putting that child in a context that is relevant for him, not the teller. However, it contains no perspective, no sense of what the other child felt or thought. This is consistent with the finding that children give increasingly more other-oriented descriptions as they age when asked to give descriptions of friends (Secord & Peevers, 1974). This is the category in which no personal involvement, or the degree to which an individual involves himself when describing others, is expressed. Thus, as children’s stories about their friend become more complex, their degree of personal involvement in these stories also diminishes. Children are more able to adopt an impersonal stance when telling stories about others as this second sphere of reality (lives of other) becomes less fuzzy and more differentiated from their own.
LANDSCAPE OF FRIENDS As children become more other focused in their understanding of their friends, they are also increasingly more attuned to the internal landscape of their friends. Indeed, as was hypothesized, as they age children tell more stories that depict the internal landscape or a mix of internal and external landscapes (the effect is stronger for narratives than scripts). In contrast, there was no age difference in the frequency at which children provide scripts or narratives that depict the external landscape. For each story, the predominant landscape was coded as external or internal. If neither predominates, the story was coded as mixed. External landscape describes the sequence of events and the actions in the story. For example, consider a narrative by a 10-year-old boy: “This one time, me and W were in the playground and we had gone past where we could go. And we had gone into the woods back there. And it was past recess time … and we got into trouble.” Internal landscape depicts the friend’s feelings and insights into the friend’s personality, motivations, and thought. For example, consider a narrative by a 10-year-old girl: “She likes to read. She likes books and she likes Harry Potter. And she just read the BFG for our book report but she hated it cuz the teacher told us that the author doesn’t like children so he makes bad stuff happen to them.” Out of all the stories elicited, approximately two thirds were such stories, signifying that the landscape of action (external) is often the focus as children come to understand their friends. Eventually, as children become more cognitively, socially, and emotionally sophisticated, they can then increasingly talk about their
8. NARRATIVES, GOSSIP, AND SHARED EXPERIENCE
167
friends’ internal states. This finding is consistent with literature on children’s social understanding of others that children become more focused on psychological attributes as they get older. There is an increasing interest and awareness of other people’s subjectivity. The extent to which children focus on the internal versus external landscapes in their stories about friends is, however, subject to experimenter influence. One interviewer elicited more internal and mixed landscape stories than the other did, indicating that the listener can shape the balance of external and internal landscapes depicted in stories by children between age 4 and 10. This could partly be because the external landscape is still the dominant mode of understanding for children at this age, it is difficult for them to naturally tell stories depicting internal landscapes. If the study were conducted with children from a greater age span, then such age-related increase in the focus of other people’s internal states may be more prominent and less subjected to experimenter influences. In any case, such experimenter difference suggests that adults can lead children to talk about and focus more on the internal details of others. Thus, even though children do naturally have an increased awareness about others’ psychological attributes as they become older, individual differences in the extent of this awareness can stem from the degree to which they were “trained” to focus on such details about other people. Children obtained most of their stories from direct experiences; however, as they get older, they also increasingly tell stories that they heard from their friends. The great majority of the stories were obtained from direct experience. Some were gathered indirectly (i.e., told by the friend, or what Preece called vicarious experience). Although no age differences were found in the number of stories from Source 1, those gathered from Source 2 increased with age. Very few stories were from Source 3, or gossip. This finding again suggests that children build up much of their understanding about their friends through spending time together, experiencing the events together (Source 1). The ability to learn about someone through oral biographies (i.e., Source 2) or gossip (i.e., Source 3) emerges later. As children become older, their ability to learn about their friends is no longer restricted to having to be physically involved; they can begin to understand the friend from a more abstract point of view. As they age and experience a greater variety of events, their experiences in other situations can also transcend across contexts to help them to understand their friends’ life experiences.
WHAT CHILDREN KNOW ABOUT THEIR FRIENDS BUT DO NOT PUT INTO STORIES There is also a certain amount of information that children seem to know about their friends that does not make its way into a narrative form. In the case of nonnarrative information; seven categories of information about friends were identified: intrapersonal, physical, object-oriented, event-oriented, biographical,
168
ENGEL AND LI
interpersonal, and friend versus self. Overall, older children gave more details about their friends than did younger children. However, such increase is not linear across the three ages. There is an especially sharp increase in the amount of information from the 7- to 10-year-olds compared to the 4- to 7-year-olds. This indicates that there might be some sort of a major qualitative change in the child’s cognitive and social understanding of others between ages 7 and 10.
Information The number of discrete pieces of information was counted in each interview. The comments were sorted into seven categories: Intrapersonal—describes the friend’s internal, psychological attributes. (e.g., “She’s smart.” “He’s nice.” “She’s funny and crazy.”) Physical—describes the friend’s physical attributes. (e.g., “He has darkish hair.” “He’s tall.” “He wears glasses.”) Object-oriented—describes the friend’s possession. (e.g., “She has a bicycle.” “She has two cats.” “He lives in a big house.”) Event-oriented—describes an event that the friend has experienced in one sentence/utterance. (e.g., “She had a big birthday party.” “We went swimming together on Sunday.” “He transferred to another school at the end of last year.”) Biographical—describes the friend’s family background and interests. (e.g., “She has two sisters.” “Her parents divorced when she was little.” “He likes to play video games.” “He’s a real fast runner.”) Interpersonal—describes the friend’s social life. (e.g., “She has a lot of friends.” “He likes to play with girls.”) Friends versus self—describes the friend in relation to the speaker. (e.g., “He’s shorter than me.” “I love her so so so much.” “The first time I met Kate was actually here in the day care.”) Even though there is an increase across most categories, the increase in the total pieces of information can be mostly attributed to the dramatic incline in the number of intrapersonal and biographical details given. Not only do these two types of information increase as a function of age, the proportional data suggest that they might also become more salient and significant in the repertoire of knowledge that children have about their friends. Intrapersonal details refer to the child’s insights regarding the friend’s psychological attributes and global dispositions. This finding is consistent with literature demonstrating that as children get older, they become increasingly attuned to the inner lives of the people around them. Their focus shifts from the concrete physical details to the more abstract psychological attrib-
8. NARRATIVES, GOSSIP, AND SHARED EXPERIENCE
169
utes. Furthermore, biographical information also increases sharply from age 4 to 10, especially from age 7 to 10. Such information includes items that characterize the friend’s interests and family background. The finding suggests that children seem to pay increasingly more attention to details that shape and help to explain the type of person their friend is. These results indicate that even though the discrete pieces of information were not woven into a story per se, they can be used to construct a life story of the friend. The explosion of information before age 10 strongly supports the idea that by that time, children already possess the elements that would allow them to tell the life story of their friend. However, these details are not naturally organized into a narrative form until later in development.
THE ROLE OF THE AUDIENCE Younger and older children make different use of their interlocuter. A number of interesting age ´ interviewer effects were found. For instance, as children get older, one experimenter tends to elicit more total pieces of information (offered independently of stories) than the other. In terms of the number of narratives depicting the external landscape, one interviewer had an increase from age 4 to 7 but a decrease from age 7 to 10, and the other interviewer had increases all across the age span. In terms of the number of narratives focusing on the internal landscape, all such stories were obtained by one interviewer only and the number increases with age. There are also some fascinating gender ´ interviewer effects. For example, one interviewer tends to elicit more interpersonal information (details about the friend’s social life) from males than females, whereas the opposite is true for the other interviewer. One interviewer also tends to elicit longer scripts from boys than from girls, while the opposite is true for the other interviewer. Or, in terms of the number of evaluation terms used in narratives, both interviewers elicited more such comments from girls than from boys, but the gap is bigger for one interviewer than the other. There were other interesting gender differences as well.
GIRLS AND BOYS Girls were found to employ more evaluative terms in their narratives about their friends (girls offered an average of .18 evaluative terms for their narratives, whereas boys included an average of .09 for theirs). This indicates that girls are more likely than boys to contribute their own thoughts and feelings when talking about their friends. This result is related to research showing parents are more likely to use an elaborative style when talking about the past with daughters than with sons. Conversations with girls are longer than that with boys and parents asks more questions (Reese & Fivush, 1993). Because girls are often encouraged by their parents to tell more detailed personal stories, they are perhaps also more likely to assess how they feel or think about the friend, the story, and how they fit
170
ENGEL AND LI
into the picture when they tell stories about their friends than boys do. In research focusing on characteristics of narrative style associated with gender, it has been found that mothers’ talk about emotions when their children were 18 months old predicted the children’s use of emotional language at age 24 months (Dunn, Bretherton, & Munn, 1987). Structurally, girls also tend to have fuller, longer, more coherent, and more detailed autobiographical narratives (Buckner & Fivush, 1998), and are more likely to include structural narrative elements such as turning points (Engel & Sprengelmeyer, 2001). Together, all of these interaction effects suggest that certain age- and gender-dependent characteristics of children’s stories about their friends can be molded and shaped by the person to whom the child is telling the story. The quality, the content, and the focus of these stories can change depending on the audience who is listening to the story. It would be fascinating to further investigate the kinds of stylistic differences in interviewing that could affect such differences in the stories children tell about their friends.
CONCLUSIONS What have researchers learned about the emergence of gossip? When asked to tell stories about their friends, children between age 4 and 10 provided two kinds of biographical materials: stories (including scripts describing routine events, or narratives documenting specific past or present events) and discrete pieces of information. The number of stories increases with age. In general, as children get older, they tell stories about their friends that are longer, more descriptive, and more evaluative. Their stories also become increasingly more other focused and more concentrated on the internal landscape of others, although stories of joint experiences and external landscape remain important during this age span. Most stories were learned from having directly witnessed or participated in the event, although children also increasingly tell stories that they have heard from their friends. Most of these effects are more salient for narratives than for scripts, suggesting that scripts are less subjected to developmental changes. Interviewer is found to have a profound effect on the type and number of stories elicited about friends. For the most part, the age and gender main effects still remain strong even after removing the experimenter variable, however, the latter plays a significant role in affecting the amount and type of information, the number of scripts, and the descriptiveness, content, and structure of the narratives obtained. Finally, for information about the friend that is not in the context of a story, such information increase with age and much of this increase can be attributed to an increase in intrapersonal and biographical details. As children age, their stories about other people also become increasingly more other focused—both in terms of their friends’ lives and their friend’s subjectivity—and less predicated on mutual experiences and external actions. This study is
8. NARRATIVES, GOSSIP, AND SHARED EXPERIENCE
171
only the beginning step in exploring children’s stories about the people around them. Not enough is known about when and how children spontaneously engage in gossip, or how this activity first manifests itself in the flow of everyday activity and conversation. One interesting distinction that continues to percolate if not haunt the research on narrative concerns the difference between fragments of narrative that unfold in the course of everyday conversations, and the kinds of stories and recollections manifested in texts. There are important differences between what is said and how it is said when creating text and what is said and how it is said when in the flow of conversation. This difference is made more complicated by the fact that young children often do not write easily. Some earlier work tackled this problem by inviting young children to tell stories, but making the textual nature of the task salient to them by showing them what they have said as a transcribed text, and inviting them to revise it on a subsequent occasion. It was found that children can take an autobiographical stance toward their life story (or stories as the case may be) when they are as young as 6 years old.The study described herein framed it as a conversational task rather than a textual one. However, the issue of stance or frame continues to be important. A child’s use of the narrative form to think about other people may vary as a function of whether they are chatting, telling stories as monologues, or writing about the other person. This is probably equally true of how and what they internalize from what they hear of other children’s life stories and personal recollection. It would seem, however, that these data support the idea that at some point, fairly early on (by age seven anyway), children begin to construct narratives about other people. One question is how and whether they use these stories to think more closely about those people, or whether those narratives just become another version of the kinds of narratives they tell about themselves. But it seems clear at this point that understanding the order of the day, and some of the external or logical causes of events, is only part of what makes personal narratives and autobiographical remembering so powerful a mental process. The narrative framework also provides a way of knowing other people’s stories, which is one essential way of knowing other people. Bloom argued that Shakespeare was the first writer to invent a person with words. It seems that by watching children telling stories about other people, researchers can find the roots of this astounding skill lying semi dormant in most people—the ability to know the other by creating that person in words. Early on individuals may use stories to know themselves. Over time, people use them to know others. Moreover, the knowledge of other people that is acquired through their narratives may, in some ways, be more important and powerful than the knowledge individuals get of those people through firsthand, or direct, experience. That is, the stories people tell about others, and the knowledge they have of others based on their stories, may be quite different. Katherine Nelson’s emphasis on the essential role that language plays in allowing children to think about things
172
ENGEL AND LI
underscores the usefulness of this kind of distinction. Language about a person is different from actions with the person. Moreover, language (in this case narrative language) as a way of knowing someone may take two forms: the narratives constructed about the person, and the narratives heard about the person. Either way, language gives access to unique aspects of the person. Gossip gives people what a meeting may not. It seems plausible that when children learn about their friends, they not only add dimension to their mental landscape, they also add an essential aspect of being human—learning about the consciousness of other people. In the data described here, children first tell about other people only when they can use their own consciousness (shared experience). The ability, or tendency, to tell stories that focus on the other person’s internal experience is a relatively late accomplishment. Finally, the sense that individuals can actually create a known other out of words, or know the person through the stories heard about them (gossip or novels), seems to arrive remarkably late on the scene. When Tolstoy introduces the Oblonsky family, readers know right away that the Oblonskys feel and think, and that the reader will be drawn into the way the characters feel and think. The novel begins with the internal landscape of others. Researchers have ahead of them the delectable task of detailing just how it is that anybody comes to apprehend the internal landscape of others. This new line of work, linking stories of personal experience to stories about others, is one pathway along which science can learn how it is that people ever have a meeting of the minds.
REFERENCES Anderson, A. (1993). We knew each other when we were childs: Friendship and collaborative narration in preschool. Honors thesis, Williams College, Williamstown, MA. Bauer, P. J. (1996). What do infants recall of their lives? Memory for specific events by oneto two-year-olds. American Psychologist 51(1): 29–41. Bloom, H. (1988). Shakespeare: The invention of the human. New York: Putnam. Brice Heath, S. (1983). Ways with words. Cambridge: Cambridge University Press. Bruner, J. (1986) Actual minds, possible worlds. Cambridge, MA: Harvard University Press. Bruner, J., & Lucariello, J. (1989). Monologue as narrative recreation of the world. In K. Nelson (Ed.) Narratives from the crib. Cambridge, MA: Harvard University Press. Buckner, J. P., & Fivush, R. (1998). Gender and self in children’s autobiographical narratives. Applied Cognitive Psychology, 12, 407–429. Canetti, E. (1979). The tongue set free. New York: Farrar,Strauss,& Giroux Carver, L.J., & Bauer, P.J. (1999). When the event is more than the sum of its parts: Long-term recall of event sequences by 9-month-old infants. Memory, 7, 147–174. Carver, L.J., Bauer, P. J., & Nelson, C. A. (2000). Associations between infant brain activity and recall memory. Developmental Science, 3(2), 234–246. Dunn, J. (1988). The beginnings of social understanding. Cambridge, MA: Harvard University Press.
8. NARRATIVES, GOSSIP, AND SHARED EXPERIENCE
173
Dunn, J. (1994). Understanding others and the social world: Current issues in developmental research and their relation to preschool experiences and practice. Journal of Applied Developmental Psychology, 15, 571–583. Dunn, J. (1996). The Emanuel Miller memorial lecture 1995-Children’s relationships: Bridging the divide between cognitive and social development. Journal of Child Psychology, 37(5), 507–518. Dunn, J., Bretherton, I., & Munn, P. (1987). Conversations about feeling states between mothers and their young children. Developmental Psychology, 23, 132–139. Ely, R. (1997). Everything including talk and why you hafta listen. Journal of Narrative and Life History, 7(1–4), 351–357. Engel, S. (1986, April) Mother child conversations about the past, Paper presented at the biennial meetings of Human Development, Nashville, TN. Engel, S. (1995). The stories children tell. New York: Freeman. Engel, S., & Sprengelmeyer, A. (2001, April) The plot thickens: Children’s autobiographies and parents’ biographies of their children. Poster presentation, biennial meetings of the Society for Research in Child Development, Albuquerque, NM. Feldman,C.(1989, November) Genres as mental models. Paper presented at Instituto de Filosofia, Naples. Fivush, R. (1991). The social construction of personal narratives. Merrill-Palmer Quarterly, 37(1), 59–82. Fivush, R., Gray, J. T., & Fromhoff, F. A. (1987). Two-year-olds talk about the past. Cognitive Development, 2, 393–409. Hudson, J. A., & Shapiro, L. R. (1991). From knowing to telling: The development of children’s scripts, stories, and personal narratives. In A. McCabe & C. Peterson (Ed.), Developing narrative structure (pp. 89–136). Hillsdale, NJ: Lawrence Erlbaum Associates. Keller, A., Ford, L. H., & Meacham, J. A. (1978). Dimensions of self-concept in preschool children. Developmental Psychology, 14(5), 483–489. Labov, W., & Waletzky, J. (1997). Narrative analysis: Oral versions of personal experience. Journal of Narrative and Life History, 7(1–4), 3–38. McCabe, A. (1997). Developmental and cross-cultural aspects of children’s narration. In M. G. W. Bamberg et. al. (Eds.), Narrative development: Six approaches. (pp. 137-174). Mahwah, NJ: Lawrence Erlbaum Associates. McCabe, A., & Peterson, C. (1991). Getting the story: A longitudinal study of parental styles in eliciting narratives and developing narrative skill. In A. McCabe & C. Peterson (Ed.). Developing narrative structure (pp. 217–253). Hillsdale, NJ: Lawrence Erlbaum Associates. Miller, P. J., Mintz, J., Hoogstra, L., Fung, H., & Potts, R, (1992). The narrated self: Young children’s construction of self in relation to others in conversational stories of personal Experience. Merrill-Palmer Quarterly, 38(1), 45–67. Miller, P. J., & Sperry, L. (1988). Early talk about the past: The origins of conversational stories of personal experience. Child Language, 15, 293–315. Nelson, K. (1986). Event knowledge: Structure and function in development. Hillsdale, NJ: Lawrence Erlbaum Associates. Nelson, K. (1993). Events, narratives, memory: What develops? In C. Nelson (Ed.), Memory and affect in development: The Minnesota Symposia on Child Psychology (Vol. 26, pp. 1–24). Hillsdale, NJ: Lawrence Erlbaum Associates.
174
ENGEL AND LI
Pellegrini, A.D., & Galda, L. (1990). The joint construction of stories by preschool children and an experimenter. In B.K. Britton & A.D. Pellegrini (Eds.), Narrative thought and narrative language. Hillsdale, NJ: Lawrence Erlbaum Associates. Peterson, C. (1994). Narrative skills and social class. Canadian Journal of Education, 19(3), 251–269. Preece, A. (1987). The range of narrative forms conversationally produced by young children. Child Language, 14, 353–373. Reese, E., & Fivush, R. (1993). Parental style for talking about the past. Developmental Psychology, 29, 596–606. Rubin, Z. (1980). Children’s friendships. Cambridge, MA: Harvard University Press. Secord, P., & Peevers, B. H. (1974). The development of person concept. In T. Mischel (Ed.), Understanding other persons (pp. 117–142). Oxford: Blackwell. Selman, R. L. (1981). The child as a friendship philosopher. In S. R. Asher & J. M. Gottman (Eds.), The development of children’s friendships (pp. 242–272). Cambridge, England: Cambridge University Press. Steedman, C. (1982) The tidy house: Little girls writing. London: Virago Press. Wells, G. (1986). The meaning makers. Portsmouth, NH: Heinemann. Tolstoy, L. (1978) Anna Karenina (C. Garnett, trans.). New York: Bobbs-Merrill.
9 Acquiring Art, Spoken Language, Sign Language, Text, and Other Symbolic Systems: Developmental and Evolutionary Observations From a Dynamic Tricky Mix Theoretical Perspective Keith E. Nelson Patrick L. Craven Yue Xuan Marnie E. Arkenberg Penn State University
The power of symbolic systems to mediate and organize people’s experience of the world is central to modern cultures. For many years, Katherine Nelson’s work has touched on developmental processes underlying acquisition of symbolic systems, especially spoken language. This work has substantially influenced interest in and
175
176
NELSON ET AL.
empirical research on how multiple factors dynamically converge to support children’s progress in complex symbol systems. More broadly, over the past 20 years, there has been a strong expansion of studies on children’s narrative, vocabulary, syntax, and literacy development that strive to capture the embeddedness of children’s learning in meaningful, socially rich, scaffolding exchanges with those more expert in these systems—adults in most studies, but with some attention as well to siblings and peers. Implicit in much of this work is that children’s progress in other symbolic systems—notably, art, music, dance, and multidomain productions—may similarly depend on the “who and how” of the child’s interactions with other more expert symbol users. This chapter argues that the writing of Nelson and her colleagues converges very well with other investigators in identifying key theoretical processes, remarkable individual differences, and many rich patterns of child–other interactions that have the potential to contribute to specific developmental advances by the child.
THE DYNAMIC TRICKY MIX THEORY One organizing frame is the theoretical model we term the Dynamic Tricky Mix theory. Using this particular dynamic systems model helps to differentiate (among many potentially supportive patterns) interactive patterns that fail to support learning, those that support modest rates of learning, and those that scaffold the highest rates of learning. Progress in making such differentiations is reviewed, and then leveraged to suggest research pathways to further understanding. Comparisons are sought among what is known about children’s acquisition of art skills, sign language, multiple subdomains of spoken language, and to a brief extent for written language and mathematics. Comparative discussion also covers language-delayed children, autistic children, and typically developing chilren. Finally, all these issues are explored for a quite different period of human development—physical evolution and cultural evolution across the periods from 300,000 to 2,000 years ago. Dynamic Tricky Mix theory emphasizes that multiple complex conditions need to “cooperate” and converge to support learning (K. E. Nelson, 1991a, 2001; K. E. Nelson, Heiman, & Tjus, 1997; Nelson, Perkins, & Lepper, 2003; K. E. Nelson & Welsh, 1998). This convergence is a dynamic process, requiring coordination in real-time episodes (cf. Damasio, 1999; Peltzer-Karpf, & Zangl, 2001; Thelen & Smith, 1994). A comparison to the complex dynamic systems contributing to tornado production is in order. Once formed, tornadoes are easily tracked phenomena even though it is tough to predict exactly when and where potentially favorable conditions will converge. Similarly, in many instances, effective dynamic convergence in learning episodes may be “found” through observation even though it may be difficult to predict in advance when and where the “tricky” convergences will occur. As Fig. 9.1a suggests, children will LEARN only when there is some above-threshold level of dynamic convergence of different kinds of learning con-
9. A DYNAMIC TRICKY MIX PERSPECTIVE
177
FIG. 9.1a. Dynamic Tricky Mix Model.
ditions. The LEARN acronym is just one way of organizing these partly overlapping categories of conditions: Launching, Enhancing, Adjustment, Readiness, and Networks.
Launching Conditions Launching conditions (L) require some involvement in a relevant activity that includes new challenges. These conditions also include at least minimal attention and abstraction and storage in long-term memory of some partial information about the challenges. For example, children learning to read might encode into long-term memory some partial representation of text structures that are challenges for their current developmental levels; some information about the placement of who could be encoded, as for relative clauses such as “The girl who has new purple shoes won the race.” Similarly, children learning a new lexical item, such as purple or fiffin, may at first encode a simple association between some part of the spoken word and some part of the accompanying context. Active involvement with the context that carries challenges does not happen automatically. The mere desire for engagement with others may serve as a stepping stone to language learning. In fact, Slomkowski, K. Nelson, Dunn, and Plomin (1992) showed that extroversion may contribute to language learning. Longitudinal relations between temperament and vocabulary knowledge are sug-
178
NELSON ET AL.
gestive of the contribution of extroversion to language acquisition. For these children, the intrinsic desire to be a part of the larger social system may serve as an essential foundation for learning. For children whose desire to engage with others, for whatever reason, is less strong, optimal learning may only occur with help from the larger social system, as in playful episodes initiated by a parent, grandparent, or older sibling. Note also that some launching conditions, such as relatively high attentional capacity, and temperamental variables may be relatively stable over time, but they are surely subject to influence by factors such as mood and expectancy. In addition, what motivates a child to engage in a task may vary from day to day, hour to hour, or minute to minute. Continual adjustment in response to the child by the larger social system and by the child in regard to the larger context supports paths toward the avenue of optimal learning.
Enhancing Conditions Learning of complex symbol systems depends heavily on enhancers (E; e.g., paraphrases, salience enhancers, challenge-carrying recasts, rapid interplay of multiple information channels in multimedia) that minimize competing processing demands and that specifically scaffold comparison and abstraction processes in working memory. Other examples of enhancers are adult emotional availability (Pressman, Pip-Siegel, Yoshinaga, Jubicek, & Emde, 1998), adult and child strategies that fit learning goals as well as the current context of learning, and attributions to the child for effort and success (Dweck, 1990, 2000) Learning of new communicative skills may further be enhanced by settings and strategies that bring out the child’s focused attention, persistence, planfulness, and monitoring of plans (cf. Bransford, Brown, & Cocking, 2000; Perkins, 1993). More broadly, each culture finds its ways of embedding language and other symbolic systems within purposeful social activity. Language incorporates a multitude of specific language structures but is not reducible to these; instead, it is inherently social in nature. Thus, the learning of language is dependent on and enhanced by others’ guidance toward the acquisition of shared meaning. This sociability of language is at the heart of Katherine Nelson’s conception of the socially shared symbolic system: The meanings of the symbols (words, grammar, narratives) contained within a language are social constructions that emerge within a social matrix. Language acquisition is essentially a social practice that occurs within a socially shared symbolic system (K. Nelson, 1986, 1998). Acquisition may obey similar processes through different cultural circumstances, with the particular persons (e.g., siblings, aunts, parents) and the particular interactions that help provide the convergent mixes necessary for acquisition varying substantially both within and across cultures (cf. Allen & Crago, 1996). Parents, older siblings, and other social partners whose own positive social and emotional adjustment are regularly part of social transactions can be powerful contributors to a child’s development in symbolic domains, because they can help induce the
9. A DYNAMIC TRICKY MIX PERSPECTIVE
179
child’s current online processing efficiency and because they may influence longer term adjustment skills in the child. Acquisition will fail, as is the case with some autistic children, or falter (Baker & Cantwell, 1982) when there is a lack of establishment of the fundamental social processes that launch the child’s engagement in language and provide enhancers of continued engagement and effective processing of language.
Adjustment Conditions Crucial adjustment (A) processes in the learner are influenced by developmental status and by partner, setting, and many other variables. Some adjustment steps are almost always required to promote processing of relevant changes (Damasio, 1994; K. E. Nelson, 1987, 1989, 1998, 2000). As part of such adjustment processes, the child will apply new strategies, will find pleasure in different aspects of the challenges, and will be attuned to different aspects of the situation continually throughout the learning experience. When a preschool child repeatedly tends to show poor social, emotional, and motivational adjustment during learning episodes, there are risks for development of relatively low cognitive and social skills (Cole, Michel, & Teti, 1994). In contrast, when such regulation goes well, positive expectancies of an interesting learning activity, low anxiety, high self-esteem, appropriate emotional states, and friendship and emotional bonding have an opportunity to dynamically converge with other learning conditions. In fortunate interactive cycles across time, improving adjustment conditions in the child learner accompanied by similarly positive behaviors in the child’s partners may help to ratchet up the likelihood of overall favorable mixes and thus the probability of learning.
Readiness Conditions Readiness conditions (R) incorporate background aspects of preparedness for new challenges. The child’s characteristic temperament, attentional resources, intelligence, effects of recent stressors, memory capacity, pattern detection capacities, and potentially levels of any (to-be-discovered) language-specific in-built representations or mechanisms are examples.
Network Conditions Neural networks (N) must be intact and appropriately active during learning episodes. As learning proceeds toward stable structures with flexibility across contexts, new structures require thorough integration into representational networks. Convergent learning conditions favoring such integration include relatively low immediate demands for learner response, many enhancers, smooth adjustments (see conditions above), and some sustained active processing. To this end, the pro-
180
NELSON ET AL.
cesses building increasingly effective neural networks are also dynamic system processes with similar competitive aspects operating over very different time scales in evolution (Elman et al., 1996; Lieberman, 2000; Nakisa & Plunkett, 1998). Both in individual brains and in species evolution these dynamic processes allow for—but do not guarantee—emergence over time of qualitatively new organizational levels (Gould, 1989). These conditions are revisited throughout this chapter with particular attention to the potential roles of different kinds of event sequences—from very local, 15-second episodes, to those within an hour or so, to longer term cross-situation recurrences and scripts—and to individual differences among children.
CHILDREN’S SYMBOLIC DOMAINS AND CONVERGENT MIXES:ASSUMPTIONS OF AND EVIDENCE FOR DYNAMIC TRICKY MIXES From the Dynamic Tricky Mix theoretical perspective, multiple predictions for the acquisition of new structural levels in all symbolic domains follow. These general predictions are examined, and then children’s acquisition of particular symbolic domains are discussed. Many interactional episodes of apparent relevance to learning first language, literacy, or any other symbolic domain will have no effect on acquisition. This prediction follows from the necessity for any learning of achieving a threshold level of LEARN convergence in the midst of ongoing interaction. Being “close” in the sense of some conditions being favorable will not support learning unless, as in certain chemistry reactions, a fuller set of well-timed conditions co-occur and interact. High variability of learning rates will be expected for individuals across relevant learning contexts and also between individuals even when contexts appear quite similar. This follows because threshold levels of convergence are required for any learning to occur, and because for above-threshold levels the rates of learning will accelerate as dynamic mixes become increasingly favorable on multiple converging components. The variability predictions also are based on the observation that there are manifold patterns in which learner Readiness characteristics and all other components will either succeed or fail to converge dynamically within the online temporal windows of interaction events. These windows frequently will be only 5 to 20 seconds long, and in many other instances (e.g. early dyadic narratives) will be in the range of from 1 to 5 minutes. Sustained rapid growth by a child will be relatively rare, for all the reasons so far articulated on trickiness of mixes, but it will occur under predictable sets of circumstances—precisely when positive dynamic mixes are themselves sustained and very highly positive. Examples of individual children achieving very high and sustained growth rates with the support of exceptional dynamic mix conditions have been reported for early literacy (Soderbergh, 1977), sign language and
9. A DYNAMIC TRICKY MIX PERSPECTIVE
181
speech bilingualism (Maxwell, 1983; P. M. Prinz & E. A. Prinz, 1981), and art skills in the preschool years (Golomb, 1995; Ho, 1989). Again, however, note that when a child is in a potential learning activity, the LEARN components need to converge at a threshold level in real time, as in 10- to 60-second interactive episodes with a social partner and/or a book or video or computer, in order for any learning to take place. Beyond threshold, as the dynamic mixes become more and positive in their convergence online, learning rates are expected to increasingly accelerate. When the mix is very strongly positive in terms of all components, challenges are so readily and deeply processed that learning from single or otherwise quite rare events is likely. When the Dynamic Tricky Mix remains highly positive over many consecutive episodes, a condition of deep enjoyment and absorption in the activity of learning called “flow” may occur (M. Csikszentmihalyi & I. Csikszentmihaly, 1988). Figures 9.1b to 9.1e illustrate how across different occasions convergent mixes of learning conditions may generally improve but may also on occasion stall. By a hypothesized 10th episode in a series of increasingly successful learning episodes, as in Fig. 9.1e, high positivity in all LEARN categories is present and these converge to produce very high rates of learning. Children sometimes learn readily from a single occasion, establishing representations in long-term memory that guide appropriate generalization and application of new symbolic structures to new contexts occurring days or weeks later. This kind of deep learning from infrequent, “rare” events is possible because some
FIG. 9.1b. Unfocused Chaos.
182
FIG. 9.1c. Positive Spirals, Episode #1.
FIG. 9.1d. Positive spirals, Espisode #2, Generally Increasing Positivity.
NELSON ET AL.
9. A DYNAMIC TRICKY MIX PERSPECTIVE
183
FIG. 9.1e. Positive Spirals, Episode #10, Very High Positivity Building on Prior Episodes.
rare occasions provide high positivity in all relevant components of a Dynamic Tricky Mix. Demonstrations from experimental work with controlled input for young children have been especially clear for 1½- to 3-year-olds learning new nouns such as “fiffin,” “lobster,” or “lubnik” (K. E. Nelson, 1982, 2000; K. E. Nelson & Bonvillian, 1978). Naturalistic work also suggests that this kind of learning takes place for many young children (e.g., Bates, Bretherton, & Snyder, 1988; L. Bloom, 1991; Carey, 1978; Leopold, 1939). Children typically appear Ready in terms of abstraction mechanisms by about 18 months of age to learn in this fashion, but the field is open for appropriate tests of whether younger children may also be prepared. As with related phenomena of fairly rapid learning covered in the next section, the individual child’s focused cognitive engagement with the situation and the adjustment of social–emotional processes must dynamically converge in the brief window of time when word-to-be-learned and referential event co-occur in a highly processable frame. Many variants on cultural patterns that serve as successful enhancers of symbolic skill growth are expected because there are few enhancers that are necessary for all children. At the same time, there will be some enhancers that are necessary for children with one set of Readiness characteristics and other enhancers that are necessary for children with another set of Readiness characteristics. The requirement of online complex nonlinear convergences for learning by individuals may lead to some theoretical and practical frustrations—the math, literacy, or language
184
NELSON ET AL.
example presented to a child and transcribed as data by a researcher may have no relevance to learning because it receives no “uptake” by the child. Understanding of when learning does and does not take place requires more differentiated analyses of multiple learning conditions. Children with different combinations of temperament and cognitive characteristics for example may respond very differently to potential catalysts and may require quite different enhancers to achieve a threshold of growth for a particular domain and stage. High within-group variability along these lines can be expected for children without any disability as well as for children who are deaf, language delayed, developmentally delayed, autistic, or with some motor disabilities (cf. von Tetchner & Martinsen, 2000). Likewise, the sources of powerful mediation toward effective mixes may sometimes come from the much-studied interactions between mothers and children, but at other times will arise from independent book reading, exploration of art books and other media, child–child interactions, and overheard and overseen interactions by the child of adult–adult and child–child interactions. Multiple biological differences will contribute to pace and patterns in symbolic system acquisition, but these will play out in very complex fashion depending on the way in which these aspects of Readiness for a child accompany different dynamic mixes overall. As just one example, children with Down Syndrome have serious memory and abstract limitations compared with more typically developing children. The implications of these in-built biological differences, however, have changed dramatically in the last 20 years. Now the child–adult interactions are richer in language complexity, enhancers, discourse challenges, and expectations of substantial progress by the child in language. Consequently, language and instruction based on language have improved greatly compared for the children of the current generation versus the prior. More broadly, for children in general there are tremendous variations in attention, processing at all levels, and temperament that would appear to put larger numbers of children at high risk for difficulties in language acquisition than the number that do develop language disabilities. Taking a longer historical view, for ordinary children today as compared with 500 years ago, it is far more likely that the children with biological-based risks for language learning difficulties will receive substantial “cushions” of positive interactional mixes and will go far toward full mastery of a complex language. This point, and a much further reach into humans’ past cultural and biological foundations, is elaborated in later sections. “Butterfly Effects” will be common. That is, initial conditions and conditions fairly early (but not initial) in acquisition of complex domains often will have large consequences for the nature of the domain organization much later in development. If a child in initial stages of language acquisition receives very highly positive dynamic mixes, rich in challenges and converging social and emotional adjustments and enhancers, and if that child is also early in brain maturation compared to peers, an early onset of extensive language may then influence many con-
9. A DYNAMIC TRICKY MIX PERSPECTIVE
185
current aspects of building varied cognitive representations that in turn will mutually influence later language progress and a relative centrality of language in the child’s awareness, self-concept, and cultural identity. Similarly, at the ages of 5 and 6 years, the child’s level of language and organization of language may reflect the child’s “butterfly profile” evident before age 2 years. The child’s current representations of the world echo early communicative propensities to be, for example, a “nominal relier”—e.g., a child who is concerned with basic level object names, such as dog, cookie, sock, or a social butterfly who has few, or none, such nominals in their first 20-50 words but rather many social-expressive chunks such as do-it- again, thank-you, and see-you-later (K. Nelson, 1973; K. E. Nelson et al., 1985). The following sections presesnt applications of Dynamic Tricky Mix theory, including passive sentence acquisition, varied lexical and syntactic acquisitions, narrative development, and art development.
Lessons from Children’s Acquisition of Passives In order to acquire mastery of the overall syntax of any language, a child must succeed at acquiring many different individual structures. Among a set of interesting complex structures that have received considerable attention are passive sentences. For passive sentence acquisition, an acquisition model has been delineated within Dynamic Tricky Mix theory. This model incorporates a comparison/contrastive component highly similar to that in Nelson’s Rare Event Learning Mechanism, providing an emphasis on the processing steps involved in noticing new challenges, abstracting structure, storing gestalt exemplars as well as structural descriptions, and retrieving and comparing exemplars that occurred in input at widely separated times (cf. K. E. Nelson, 1977, 1980, 1981, 1987, 1989, 2000, 2001; cf. also Bohannon & Stanowicz, 1989). In Dynamic Tricky Mix theory, these processing steps contributing to acquisition always must be seen as dependent on and as co-occurring with the broader set of online convergent dynamic processes that are jointly emotional, social, motivational, and executive in nature. A fortunate, tricky mix must converge if any progress in passive acquisition is to occur. As more becomes known about such mixes, increasing the frequency and positivity of the mixes should lead to accelerations of passive acquisition rates under well-specified positive mixes. Seven stages in the acquisition of passives are laid out in Table 9.1. First consider the common aspects of development, regardless of language, culture, and varied interactional patterns within a culture. A key phenomenon in syntax acquisition is Launching some representation of the structure. This is not automatically given either by biological representations of syntax that anyone has so far documented or by simple infusion of exemplars in “input” within hearing of a child. Stage 1 in passive acquisition begins with the first new representations in long-term memory that bear in any way on passives, often with isolated gestalt,
186 New LT Storage
3 Gestalt sentences 2 isolated by-clauses 5 Gestalt sentences, some produced 5 isolated by-clauses 3 sentences with s.d. 4 favored verbs, with multiple sentences
X X
X Yes. Get favored over Be
Partial, LT set & new Fuller structural d. New exemplars improving Hot spot verbs favor
4. Exemplars & descriptions expand
Highly active
Highly active
Partial
Partial structural d.
LT set analyzed
3. Active internal analysis
3 Gestalt sentences 2 isolated by-clauses
1
X
X
Similar Total # LT Exemplars; Structure illustrative changes Map
X
X
Verb Analysis
By-clause Yes discrepant
X
Hot Spot
Gestalt exemplars Isolated by-clauses Isolated lexical adjectival passives
X
Structural Description (SD)
More exemplars
Discrepancy Gestalt detected
Comparison/ Contrast
2. Beyond Catch-22, exemplars stored
1. Initial launch
Stages
Dynamic Tricky Mix Acquisition States for Passives
TABLE 9.1
187
Highly active
Now inactive
LT set & new Full structural d. Full New exemplars. Consolidate, speeding reorganize networks
LT set & new Full structural d.
7. Consolidation ofpassive
Full
Highly active
6. Speed of retrieval& consolidation of passive
Nearly full
Highly active
Hot Spot
LT set & new New structural exemplars
Nearly full
Structural Description (SD)
5. Structural descriptions refined
New LT Storage
LT set & new New structural exemplars
Comparison/ Contrast
5. Structural descriptions refined
Stages
Yes, extensive
Yes, extensive
Full, most complex
Modest, fairly complex
Yes, Partial improving Be Passives, most kids
20 sentences with s.d. Note: no access now to Gestalt sentences or isolated by-clauses of earlier steps
10 Gestalt sentences, some produced 10 sentences with s.d.
8 Gestalt sentences, some produced 10 sentences with s.d. Verbs clustered by similarity, related examples of variation
8 Gestalt sentences, some produced 10 sentences with s.d. Verbs clustered by similarity, related examples of variation
Similar Total # LT exemplars; Structure illustrative changes Map
Partial Yes, improving Be Passives, most kids
Verb Analysis
Dynamic Tricky Mix Acquisition States for Passives
TABLE 9.1 (cont.)
188
NELSON ET AL.
unanalyzed sentences. These are somehow registered as salient and/or discrepant from other sentences and the structural syntactic representations already in long-term memory. Why some children begin in different ways is not well understood, but by Stage 2 in English there are idiosyncratic differences in which gestalt sentences are available in long-term memory plus a special focusing of attention and abstraction called a Hot Spot in acquisition. Within this hot spot children soon have some separate representation of passive by-clauses and identification/tagging of such clauses as discrepant from other syntactic representations (e.g., of by the house, or by his book). Acquiring a structure for which at first Catch-22 circumstances seem to apply—high structural complexity and no initial representation—becomes feasible because each step toward acquisition is a partial one and at each stage richer long-term data have been stored so that new interactions and exemplars now support pattern detection and abstraction based on a richer and richer set of exemplars and prior analyses, not just the data of the most recent conversational sampling. By Stage 3, the beginnings of structural descriptions occur along with more verbs treated in isolation as “verb islands” (Tomasello, 1995). The hot spot of enhanced salience and enhanced retrieval priming for anything related to passives is very evident at Stages 4 and 5 and so is the continued build-up in long-term memory of gestalt sentences, isolated by-clauses, and increased numbers of verbs placed in passive context. For many, but not all, children, get passives are favored over be passives at Stage 4 but both are represented by Stage 5. At Stages 5 and 6, both verb analyses and structural detail on passives improve. At Stage 6, considerable consolidation of structural descriptions and related networks occurs, and speeding of processing of passives occurs. By Stage 7, former gestalts and isolated by-clauses are excluded from systematic representations. However, within organized networks there is redundant information that includes passive structural descriptions, storage of verbs in a manner that reflects their similarities, and particular sentences with their full structural descriptions. Comparisons and retrieval during passive sentence comprehension and production accordingly can be facilitated by different levels of abstraction and different concrete instantiations of the passive. From Dynamic Tricky Mix theory it is strongly predicted that interesting individual variations on acquisition pathways will occur when individuals encounter very different sets of input exemplars, different enhancing dialogues, and different social and emotional qualities in interactions that carry input exemplars. One replicated finding is that rates of passive acquisition can be accelerated considerably by new focused conversational presentations of passives within recasting dialogues. An example would be to recast “Buffalo bump wagon” or “The buffalo bumped the wagon” (child’s utterance) as the adult reply “Yes, the wagon was bumped by the buffalo.” As this example illustrates, and as argued consistently by Nelson and colleagues (e.g., K. E. Nelson, 1977, 1980, 2000; K. E. Nelson & Welsh, 1998), “recasts” provide relevant child-system/challenge comparison possibilities
9. A DYNAMIC TRICKY MIX PERSPECTIVE
189
regardless of whether the child produces an error in their “platform” utterance. Across 2 to 20 sessions of dialogue treatment, with each session carrying around 20 to 25 recasts into passives, children initially lacking the passive have shown quite rapid passive acquisition. This has been shown in language-typical children (N. D. Baker & K. E. Nelson, 1984; K. E. Nelson, S. M. Camarata, Welsh, Butkovsky, & M. Camarata, 1996) and in children with specific language impairment (SLI; S. Camarata, K. E. Nelson, & M. Camarata, 1994; K. E. Nelson et al., 1996; K. E. Nelson, Craven, Xuan, Arkenberg, & Lauck, 2002). The most recent intervention work has identified important characteristics of adult clinicians’ passive recast patterns that are related to the acquisition pathways shown by children with SLI. Most notably, in contrast to the input in English reported for naturalistic samples, the clinicians’ passives were most often be (rather than got) passives and most often irreversibles such as The bicycle was ridden by the boy (rather than reversible, as in The cat was chased by the dog). Sensitivity to these input particulars was shown by earlier, relative to naturalistic, acquisition by these children in intervention of both irreversible passives and be passives, as Tables 9.2 and 9.3 demonstrate. According to dynamic tricky mix theory, a child must be ready for acquisition in terms of already-established linguistic and cognitive levels and in terms of intact-enough attentional, memory, and pattern detection and abstraction mechanisms. The SLI children in the studies just reviewed demonstrated such adequate TABLE 9.2 Input in Intervention Shifts SLI Children’s Use of be and get be
get
All Passives
Input examples of passives, treatment
93%
7%
100%
10 SLI children, 5–7 yr, treated
77%
23%
100%
Language normal children, 5–8 yr. (Marchman et al., 1991)
10%
90%
100%
TABLE 9.3 Input and Acquisition of Four SLI Children, Treatment Input Influences Reversibility Reversible
Nonreversible All Passives
Input examples of passives, treatment
31%
69%
100%
4 SLI children, 6–7 yr, treated
11%
89%
100%
Language normal children, 5–7 yr (Horgan, 1978)
53%
47%
100%
190
NELSON ET AL.
Readiness, despite their slow overall progress in language acquisition. But often it is easy to underestimate a child’s readiness. In the case of readiness for passives, most earlier work concentrating on Indo-European languages such as English, French, and German led to the impression that typical 8-year-olds were ready for passive acquisition but that 1½- to 4-year-olds usually are not. For example, Sinclair (1968) from a Piagetian perspective suggested that reversible mental operations of the sort seen in Concete Operations (e.g., conservation of liquid quantities) were a necessary foundation for passives. Naturalistic work in English, German, and Hebrew all converged in indicating that passives typically emerged after age 4 and that acquisition frequently is an extended process over multiple years with completion often as late as 9–11 years of age (Horgan, 1978; Marchman, Bates, Burkhardt, & Good, 1991). But most naturalistic samples in these languages fit within a Dynamic Tricky Mix account of very poor convergent mixes for passives—infrequent adult passives and with most of those exemplars not patterned into facilitative dialogue sequences such as child-platform/ adult-recast. In sharp contrast, when either naturalistic circumstances or focused intervention provides relatively frequent passives within positive, converging social-emotional interactions then passive acquisition readiness has been demonstrated easily in 1½- to 4-year-olds. In English, the growth recasting intervention studies with language-normal children fit this conclusion (N. D. Baker & K. E. Nelson, 1984; K. E. Nelson et al., 1996). Strong converging evidence is shown by work on early acquisition in K’iche’, Zulu, Sesotho, and Inutikut (Allen & Crago, 1996; Demuth, 1990). In Inutikut it is especially clear that passives are richly embedded in positive social-emotional dialogue situations, that there are varieties of those dialogues, and that part of the converging mix that contributes to saliency and acquisition of passives by the children is that passives are highly useful ways of communicating certain messages that are difficult to otherwise convey.
Varied Language Acquisition Phenomena There are many demonstrations that learning in 1½- to 5-year-olds of new lexical and syntactic structures can be triggered by a relatively modest set of input exemplars embedded in meaningful, positive social interactive occasions spaced over 3 days to 6 months. At the same time, however, there are tremendous individual differences in which new symbolic structures are acquired from seemingly similar experimental presentations. Moreover, the acquisition pathways for those structures undergoing acquisition show extensive individual differences in such characteristics as how quickly and appropriately generalizations are made to new sentence and new object/action/event contexts. This section explores just a few ways in which dynamic mix perspectives help illuminate these phenomena and how findings from particular investigative procedures need to be seen in the broader context of work in the field.
9. A DYNAMIC TRICKY MIX PERSPECTIVE
191
Consider the work of Tomasello and his colleagues on made-up (“nonce”) verbs and nouns and observed differences between 2-, 3- and 4-year-olds. Across 6 to10 sessions in about 2 to 3 weeks, exposures to new verbs (dakking) and nouns (wug) are interwoven in repeated demonstrations (e.g., 10) that also vary the particular context in which the novel word is embedded. Strong individual differences are seen in each of a series of studies. Tomasello and colleagues (Akhtar & Tomasello, 1997; Tomasello, 1995; Tomasello, Akhtar, Dodson, & Rekau, 1997) stressed that children with overall language levels of 2-year-olds, as compared with those with 3to-4-year-old language levels, are less capable of rapid learning of new lexical items and that when they do learn they generalize less readily, suggesting more “item-based” representations rather than representations placed in a systematic set of already-established lexical and syntactic structures. Verbs, because of presumed greater complexity of their relations with contexts and other syntactic structures, are seen as harder to acquire than nouns. From the Dynamic Tricky Mix view, many of these findings are predictable from the particular kinds of encounters children experienced. In large part, verb islands and other item- based sorts of learning patterns may rest more on the particular exemplars and convergent language and social framing contexts than on the stage of cognitive and language development for the child. Only a few frames for teaching (e.g., “Look! Gopping, gopping!”) and eliciting new productions (e.g., “What are these?” “What did I do?”) were employed, so there can be little confidence that for children at a particular language level the procedures for verbs were as “well-mixed” as those for nouns. From the Dynamic Tricky Mix perspective, it is essential to see that the younger children may be quite capable of more rapid and more extensive, more generalized learning if only a rich and different mix of conditions is provided. There are many diverse contextual frames under which children may be facilitated in their initial abstractions, in their tentative deployment of such abstractions for verbs or passives or nouns or any structure (cf. Strohner & K. E. Nelson, 1974), and in further inferences and abstractions derived from long-term data converging with exemplars presented online in particular discourse situations (cf. Bowerman, 1985; Carey, 1978; Clark, 1971; Maratsos & Chalkley, 1980; Maratsos & Deak, 1995; K. Nelson, 1988). Thus, it is highly probable that the complexity of learning verbs as compared to nouns varies depending on child and context. Also consider investigations in which nouns and/or verbs have been studied in their acquisition by 18- to 28-month-olds across several months rather 2 weeks (e.g. L. Bloom, Tinker, & Marguilis, 1993; Mervis & Pani, 1980; K. E. Nelson, 1982; K. E. Nelson & Bonvillian, 1978). Looking only at learning and generalization that has taken place in the first 2 weeks, as in the studies by Tomasello and colleagues, it becomes evident that it is only an occasional child who has already shown such progress and the progress is shown for only a minority of the words under investigation. But the capacity of most of these children to do such learning is shown by the learning that emerges for each child who is seen
192
NELSON ET AL.
across several months—again, with important by-child differences in which words are acquired. All of these results are comprehensible if a young child learner is Ready in terms of mechanisms, but learning about particular words on particular occasions and across occasions depends on the particulars of online momentary convergence of many favorable conditions. A great many conditions need to converge for word learning. These include abilities to detect and abstract speech elements (Nittrouer, Manning, & Meyer, 1993; Jusczyk, 1993; Jusczyk, Frederici, Wessels, Svenkerud, & Jusczyk, 1993), sensitivity and awareness to phonological units (Bowey, 1996), the ability to hold phonological representations in queue (Gathercole & Baddeley, 1990; Gathercole, Hitch, Service, & Martin, 1997; Gathercole, Service, Hitch, Adams, & Martin, 1999), the ability to inhibit proponent responses, and more. These various abilities ebb and flow throughout linguistic learning (Arkenberg & Ryalls, 2003), resulting in differences in the mix of conditions for different children at different points of time and in different situations. Furthermore, Network representations may drastically differ for children of different ages (reflecting different levels of linguistic experience). Related arguments are that the overall relative long-term stability in environmental conditions for mammalian species members has resulted in neural systems that are essentially “prepared” for evolutionarily expected input (Greenough, Black, & Wallace, 1987). Although much of the evidence for this type of experience-expectant system resides in visual perception, an analogy can surely be made for auditory systems as well. Similar arguments concern the evolution of memory systems that anticipate the need for selective retrieval not only of highly related structures or events, but also the retrieval of inconsistent or contrasting information that helps to set boundary conditions on new abstractions and on the deployment in particular online contexts of particular language or other symbolic structures (Klein, Cosmides, Tooby, & Chance, 2002). Returning to the consideration of syntactic structures acquired under experimental intervention covering several months, as with the work on passives discussed earlier, very similar kinds of individual differences and contributing processes can be seen. In syntax acquisition, a converging pattern is seen across different laboratories. When intervention procedures include convergently mixed combinations of challenging syntax (passives, past tense, relative clauses, etc.) and interested, positive adults placing such challenges in recasting of children’s utterances (as in the passive acquisition covered earlier), acquisition rates have been strongly accelerated. Evidence that the mix of conditions just described underlies the accelerated rates, rather than merely modeling with increased frequency the target structures, comes from direct comparisons with more standard nonconversational treatment procedures. Moreover, these results are seen in both language-typical children (N. D. Baker & K. E. Nelson, 1984; K. E. Nelson, 1977; K. E. Nelson et al., 1996) and children with Specific Language Impairment (S. Camarata et al., 1994; Fey, Cleave, & Long, 1997; Fey, Cleave, Long, & Hughes, 1993; K. E. Nelson, 2001; K. E. Nelson et al., 1996; Robertson & Weismer, 1999).
9. A DYNAMIC TRICKY MIX PERSPECTIVE
193
As Dynamic Tricky Mix argues, children at similar language ages but different overall cognitive levels will bring different cognitive Readiness into the online mix of conditions for syntax acquisition in these studies. This viewpoint helps explain the otherwise surprising finding that 6-year-old SLI children, with their well-documented low rates of prior progress in language, learn at rates that match or exceed those of language-matched younger children with no history of language delay. If social-emotional adjustment factors truly contribute to dynamic converges online that affect learning of varied kinds of new communicative challenges, then this should be a measurable phenomenon. So far there are just a few such studies that have helped to account for children’s rates of language progress and serve to illustrate new methodological steps stimulated by the theory. In each of the studies, videotapes of two early sessions of intervention were analyzed to determine the child’s “enjoyable engagement” or “social-emotional-cognitive” engagement. As Dynamic Tricky Mix theorizing predicts, children’s higher enjoyment/engagement scores early in intervention were predictive of larger learning gains across several months of intervention. These developmental gains were shown in syntax for children with SLI in Haley, S. Camarata, and K. E. Nelson (1994), in syntax for language-typical children in Newby (1994) and K. E. Nelson and Welsh (1998), and in reading levels and language levels for autistic children (Heimann, K. E. Nelson, Tjus, & Gillberg, 1995; Tjus, Heimann, & K. E. Nelson, 1998, 2001). In related research that is naturalistic and longitudinal rather than interventionist, Hart and Risley (1995) found that positive “Feedback Tone” (including responsive recasts and positive affective tone) by the parents of children at 1 to 3 years of age predicted child language level at age 3 and at age 9. Similarly, Nicely, Tamis-LeMonda, and Bornstein (1999) demonstrated that children’s more rapid language development in the period from 9 to 21 months is associated with high levels of maternal attunement (matching) to infant affect at 9 months. In each of these studies with some measurement of affective patterns, part of the dynamic mix contributions to language learning may have rested on positive and well-attuned affect of parent to child on particular learning occasions. But another important contributor may be the positive kinds of cycles illustrated earlier in Figs. 9.1a to 9.1e, with a build-up within the same dyad that continues interactions across many occasions of more and more positivity and attunement at the emotional, social, and expectancy levels. So far, empirical work on either lexical or syntactic or narrative structures in children has not provided differentiated evidence on what adaptations of dynamic mixes are needed to reach thresholds of acquisition for children at different cognitive and language levels between about 6 months and 6 years of age. However, it is almost certain that there will be some enhancers that are necessary for children with one set of Readiness characteristics and other enhancers that are necessary for children with another set of Readiness characteristics (cf. Bransford et al., 2000; Kaiser, Yoder, & Keetz, 1992; K. E. Nelson, 1991a, 2000, 2001). Similarly, beyond the thresholds of launching acquisition the mix of conditions that may move
194
NELSON ET AL.
a 2-year-old toward really high acquisition rates may prove substantially different as compared with similarly optimal conditions for a 4- or 6-year-old.
Lessons from Acceleration in Narrative Many arguments have been presented that improved, decontextualized narrative abilities before school entry should enhance children’s success in literacy activities at school (e.g., Bruner, 1986; Fey, Catts, & Larrivee, 1995; Fivush, 1991; Gillam, McFadden, & van Kleeck, 1995; Peterson & McCabe, 1991; Snow & Dickinson, 1990). In one study, training mothers toward increased quantity and quality (including many recast elaborations) of narrative discussions across one year with their 4-year-olds led the children one year after the intervention (the children now averaged 5 years 8 months) ended to hold narrative skill advantages. Specifically, seven children receiving the narrative intervention as compared with seven nonintervention children produced more sophisticated narratives in terms of spatial and temporal context and uniqueness of information (Peterson, Jesso, & McCabe, 1999). However, most prior studies on training preschool children in narrative skill have been limited in their low specificity of procedures and outcomes, as well as by few children as participants. Similarly, without yet showing which among many potential contributors are causal, a variety of studies indicate that more literacy activities by parents themselves and by parents with children, more extended decontextualized narrative sequences in adult-child dialogues, and related variables are often associated with higher narrative skills and better reading achievement when children are at the kindergarten to grade 2 levels (e.g., Blachman, 1997; Snow & Dickinson, 1990). High potential remains for larger, well-specified studies in this domain, particularly because strongly positive dynamic mixes of enjoyment, social engagement, and multiple kinds of challenges—vocabulary, syntax, narrative structure, and pragmatic conversational skills—all can be incorporated in explorations of varied narratives. Naturalistic work makes clear that within the same stretches of 10 minutes or so of dialogue, a child may encounter challenges and scaffolds for progress in any of multiple domains, including self-concept, memory, emotion understanding, scripts, narratives, vocabulary, syntax, phonology, and pragmatics. Valuable explorations along these lines with identification of many potential enhancers of language acquisition progress have been provided by, among others, K. Nelson (1998), Fivush (1994), Fivush and Slackman, 1986, Hudson (1990, 2002), Hudson and Shapiro (1991), and Snow, Perlmann, and Nathan (1987).
Lessons from Acceleration of Art Skill Acquisition From a dynamic systems framework emphasizing real-time convergence of social, emotional, motivational, and cognitive conditions, similar interventions with enriched mixes should support rapid learning in art and language. To date there is
9. A DYNAMIC TRICKY MIX PERSPECTIVE
195
just one study that jointly sought acceleration in art and language through positively mixed learning conditions (Ninkovic & K. E. Nelson, 2001). Skilled artists “conversationally” introduced for 5- to 7-year-olds challenging new grammatical structures (passives, relative clauses, etc.) and new drawing structures (occlusion, perspective, shading etc.) in the context of ongoing favorable emotion regulation and social engagement. For example, skilled artists not only scaffolded artistic advances, but also provided recasts targeting more advanced linguistic structures in an effort to promote acquisition of these structures. So, for example, children who drew a picture of a pig were shown in an adult’s art “recast” that objects could be drawn in front of others to indicate that the pig was behind something. In addition, these children were provided with linguistic recasts. Thus, a child who simply said “See the pig” received more structurally complex utterances such as, “Yes, that is the pig who is hiding.” Challenges were playful, emotionally positive continuations of what the child had already expressed in drawing or language (video analysis confirmed these patterns). Learning over base rates was accelerated strongly, and each of the children learned each of the art and language structures targeted. As predicted, creativity on general measures also advanced (cf. Torrance & Ball, 1984), based presumably on gains in skills and on strengthened general expectations that new challenges could be met in flexible and playful fashion. Over many positive episodes in this 5-month afterschool program, it is clear that children not only can learn significant new symbolic skills but that their characteristics as learners may change in ways that will increase the probability that effective dynamic mixes will be created in new contexts (cf. Dweck, 2000). A series of other studies also demonstrates that in the unusual instances in which there are highly challenge-enriched (art challenges) and socially positive interactions with artists, young children typically respond with strong acceleration of art skill acquisition (Aronsson, 1997; Caldwell, 1997; Golomb, 1992, 1995; K. E. Nelson & Pemberton, 2003; Paley, 1995; Pemberton & K. E. Nelson, 1987; Thomas & Silk, 1990; B. Wilson & M. Wilson, 1982). Two examples of how preschool children’s art complexity levels changed when given recasting sequences in art, a child drawing followed by a more advanced adult drawing on the same topic continuing in “conversational” cycles in art, are given in Figs. 9.2 and 9.3 (K. E. Nelson & Pemberton, 2003; Pemberton & K. E. Nelson, 1987). These rapid gains in intervention studies of 10 hours or less reveal a high Readiness for acquiring new symbolic representational skills in art when an artist interacts with interest, positive emotional tone, and challenges with a child of 3 to 5 years who contributes to the conversational communicative cycles of interaction at their own entry level of art. For children who go on to become truly expert, the levels of symbol system complexity in art become akin to the complexities of language and music, and such high-level art supports a vast and flexible range of communicative expression distinct from that achievable in language or music. The Readiness of child learners in the preschool years 1 to 5 is in some ways best illustrated by the most exceptional rather than the most typical cases. This dis-
196
NELSON ET AL.
FIG. 9.2. Pretest drawing on left, post-intervention on right (from Pemberton & Nelson, 1987).
FIG. 9.3a. Pretest drawing before dialogues with artist.
9. A DYNAMIC TRICKY MIX PERSPECTIVE
197
FIG. 9.3b. Posttest drawing after dialogues with artist (from Nelson & Pemberton, 2003).
cussion considers just a few such exceptions in the light of Dynamic Tricky Mix theory. The beautiful artwork of Stephen Wiltshire has been published in such books as Floating Cities (Wiltshire, 1991). Two examples are shown in Figs. 9.4 and 9.5. His detailed creation from memory of ornate buildings is particularly striking. Stephen’s beginnings symbolically are exceptional because his preschool and school age language skills were extremely limited, his art flourished during the same period, and he struggled to move beyond severe autism. With exceptional exposure to art Launching challenges accompanied by rich socioemotional enhancers, through adolescence and into his twenties Stephen’s art continued to advance. Language and social skills did gradually improve during the same period, through many attempts at rich scaffolding that still leave open how far Stephen will proceed in these domains. Stephen’s career is a very well-documented case of how far artistic development can go when language and social skills remain at fairly early stages (Sacks, 1995).
198
FIG. 9.4. Stephen Wiltshire drawing of Venice scene.
FIG. 9.5. Stephen Wiltshire sketch of boats in harbor.
NELSON ET AL.
9. A DYNAMIC TRICKY MIX PERSPECTIVE
199
Lessons from Acquisition of Multiple Symbolic Systems From the individual child’s point of view, it may not seem particularly “tricky” to acquire 2 or 3 languages to fluency by age 6 years. From the Dynamic Tricky Mix perspective, “all” that is required is that the child have emotionally and socially engaging conversations in each language, that a full range of language structures be included in such conversations, that processing enhancers (e.g., conversational recasts, dramatic stories with recast or recast-like sequences) are also part of the mix, and that the child strongly engage the social communicative interactions. Only 2 to 10 hours per week of such rich mixes are likely to suffice for acquiring each language. Indeed, when such conditions have been documented the acquisition of multiple languages in the preschool period is usually observed (cf. de Houwer, 1995; Fillmore, 1989). For adults, rapid progress in acquiring a second or third language is unusual but appears sensitive to the same availability (usually rare) of rich dynamic mixes, including strong processing enhancers such as recasting patterns on a regular basis (Nichols, Lightbown, & Spada, 2001). The flexible Readiness of children at 1 to 5 years of age when given rich dynamic mixes of learning conditions is also illustrated by acquisition of language in mixed modes. For example, spoken language and text and sign language emerge together under rich conditions (Maxwell, 1983). Text and spoken language proceed together at ages 2 to 3 for some children (e.g., Soderbergh, 1977). And other well-documented cases of bilingual acquisition of sign language and spoken language also are available (P. M. Prinz & E. A. Prinz, 1981; Wilbur, 1987). In the preschool period, mastery of language accompanied by acquisition of world-class skills in art have been seen in the case of one Chinese girl who, by age 2, became part of her artist father’s community of artists studying at his studio. Challenges to launch her progress in art thus could come from her father’s paintings, from his student’s paintings, and from the artistic dialogue among this community. High positivity in her father’s emotional stance and ways of encouraging her art were also part of the mix. This girl, Yani, sent her paintings on international exhibit by age 8 and continued on the path to become an accomplished artist in adulthood (Ho, 1989). Substituting for the moment language for art, it is not at all unusual to find that children encounter sufficiently positive dynamic mixes for acquisition of first languages in the preschool period and continue on to be fluent first language users as adults. What is unusual, as with Yani, is to see anything like the richness of dynamic mixes for art in the preschool that approach the richness of what most children encounter in language. Complex, expert art, music, dance, and mathematics do involve high-level communication. Theories of children’s development of symbolic systems thus need to take into account the Readiness of children to acquire very complex art and other complex symbolic systems in the preschool when the dynamic mix of conditions approaches for these systems the richness of convergent learning conditions typical in the case of spoken language. The question of whether an individual child
200
NELSON ET AL.
is more biologically ready to acquire one symbolic system than another can only be given a plausible answer if the child has encountered nearly identical richness of learning conditions for the systems being compared.
DYNAMIC MIXES AND DEVELOPMENTAL PROGRESS ACROSS INDIVIDUAL LIFESPANS AND ACROSS EVOLUTIONARY TIME In Language in Cognitive Development, K. Nelson (1998) provided a fascinating discussion of how ordinary children in their development may parallel or fail to parallel the kinds of long-term changes across evolutionary time that Donald proposed. Nelson stressed several similarities: (a) Early stages of event representation are nonsymbolic, but may be quite complex; (b) a second stage of symbolic, but nonlinguistic, representation follows; (c) “full” syntactically governed language greatly expands and mediates narrative in language and also mediates a vast expansion of perspective taking, reflection, analysis, planning, and organizational abilities; (d) complex and often formal theories and narratives depend on an extended period of further learning which often extends over middle childhood and adolescence and sometimes extends over most of the modern lifespan. Differences Nelson stressed are that modern human infants learn from culturally embedded social interactions that richly scaffold in interesting ways the child’s progress in all stages of nonsymbolic and symbolic representation, that these social interactive conditions are likely vastly different than those in place during pre-hominid and early hominid groups, and that there are many ways in which the developing representations of event, action, nonlinguistic symbols, and symbolic language are intertwined rather than neatly separated into stages. Nelson also argued that it is important to learn more about the specific online embeddedness of individual children in the process of acquiring new levels of representation in language and other domains. This point is highly compatible with the Dynamic Tricky Mix theoretical perspective that has been applied to children with disabilities as well as to ordinary children, and one that serves as a launching pad for comparisons of developmental changes across individual lifespans to those across evolutionary time. In the comparative discussion below, the similarities and differences drawn from Nelson’s and Donald’s accounts and a range of related issues are addressed. Throughout, there is considerable emphasis upon similarities in dynamic processes, but this should not be taken to indicate that differences do not need to be examined between modern individual development and cultural change and the longer span changes in physical, genetic makeup and in cultural symbol systems and their transmission. As just one example, one bundle of developmental and societal issues is raised when a modern 2-year-old encounters poor-quality day care that is thin in challenges as well as converging social-emotional positive conditions, and a different bundle of in-
9. A DYNAMIC TRICKY MIX PERSPECTIVE
201
dividual, species, interspecies, and cultural issues is raised when two genetically distinct early hominid groups face similar rapid shifts in climate, food, predators, and disease (cf. Cosmides & Toobey, 1987; Deacon, 1997; Johanson & Blake, 1996). First, compare a subset of modern children with Specific Language Impairment (SLI) at ages 2 to 4 years to a subset of Homo Erectus in the approximate period of 500,000 to 400,000 years ago. In each chosen subset, assume that for an extended period—say 5 months for the SLI children chosen, and 5,000 years for the Homo Erectus subset—there is no advance whatsoever in symbolic systems. What are the changes in the mix of conditions that would lead in the next period of the same length to acquisition of significant new levels of symbolic communication? Challenges to representational complexity need to occur and to be combined with contexts that minimize attention to distracting stimuli and that incorporate social and emotional and mnemonic scaffolding of sustained attention and processing. It is unknown whether the factors that promoted good adjustment and effective processing online for middle to late Homo Erectus are more like conditions that function in that fashion for 1-, 2-, or 3-year-old children or for bonobos, gorillas, or chimpanzees who in special enriched interactional mixes have made progress in acquiring varied symbolic systems (cf. King, 1999). Equally for individual Homo Erectus and the modern child with SLI, progress toward more complex representations mentally and more complex communicative expressions will depend on some online episodes in which these kinds of convergences occur within frames of from 5 to 30 seconds or so. Positive dynamic mixes sustained over some longer episodes of several minutes to several hours will also be relevant for a smaller number of acquisitions. Finally, successful individual learning episodes need to recur for most acquisitions—spaced over weeks and months and years. Moreover, very rapid and sustained progress will depend on the kinds of increasingly positive sets of conditions illustrated in Figs. 9.1a to 9.1e. Then, of course, gains by individuals in the evolutionary context will be contributors to shifts over generations only if there is successful cultural transmission and/or an impact on survival or reproductive rates for individuals with varying levels of achievement in symbol system use and/or invention. Note, however, as Gottlieb (2002) argued, that for mice and women (and men) shifts in behavior during a lifetime can contribute to what happens in the next generation even without genetic change, and such behavior shifts also may lead to consequences at the levels of natural selection and genetic mutation and recombination. Despite these substantial similarities in the kinds of convergent processes at work in acquisition, some striking differences in how development proceeds are apparent as well. Enhancers for learning and invention available from social partners to scaffold processing seem almost certainly to be less frequent and less powerful in a prelinguistic Homo Erectus community than in a modern literate society. Adjustment processes, emotionally and socially, would no doubt involve some significant differences too.
202
NELSON ET AL.
A further, fundamental difference applies. Modern 1-year-olds and 3-year-olds rarely perish if they fall a bit behind their peers in the pace of their early symbolic or later language-based symbolic development. But under certain mixes of circumstances, being ahead of other Homo Erectus in symbolic communication skill may insure survival rather than death. Individuals who are better able to share and respond appropriately to such simple messages as don’t eat, turn left, stop, follow (this) path, and meet there (let alone, Run to the tree now to avoid being eaten by the crocodile) may live longer and be more likely to pass on their genes. This is a scenario, of course, for selective survival of the most “symbolically fit” brains and the shift across generations toward more complex brains that will support learning of increasingly complex symbolic systems (cf. Deacon, 1997; Donald, 1991), and that once launched can lead to co-evolution of cultural development, display and transmission of ever-increasing communicative complexity, and brains that support such complexity. From the Dynamic Tricky Mix perspective, however, it is essential to see how fragile the progress of acquiring symbolic systems turns out to be both in modern preschool periods and in early hominid evolution. The saving grace in both contexts is that despite the difficulty of bringing all needed components for significant developmental change together online with good timing and convergence, there are many distinct patterns that do achieve this. So, overall, one can expect that only a small minority of social interactive episodes will directly support developmental progress, but that successful episodes can be co-constructed by the participants, like successful cooking achievements, in many particularized recipes. Now move the scenario along in time. Compare the same window of several months of development for SLI children to early Homo Sapiens during the period that Donald (1991) described as Mythic. During this period in evolution, it is suspected that different communities displayed vastly different levels of symbolic communication. The Challenges potentially to be learned by children and by adults thus include particular within-group symbolic systems, particular out-group communication alternatives, and the social-interactive strategies (which also are challenges to be learned by the next generation) that support both transmission of existing communication systems and the expansion of those systems. Again, when learning succeeds for any of these challenges, convergent positive dynamic mixes are required at the level of individual learners and when these are well above threshold quite rapid learning will occur. In at least a few locales, a rich record suggests by about 3,500 years ago the devotion of considerable financial resources and planfulness to increase the probabilities that a small minority of children (perhaps only 1% of a community) would encounter symbol-system challenges and social scaffolding of learning symbolic systems, as in the Egyptian community of Deir-el-Medina where many sorts of artists, hieroglyphic carvers, and artisans lived in comfortable lodgings with plentiful supplies of paint, carving tools, stone, wood, papyrus, metal, and so forth (R. Hagen & R. Hagen, 1999). Along with priests, kings, queens, and others receiving the highest levels of educa-
9. A DYNAMIC TRICKY MIX PERSPECTIVE
203
tion, these artists and their children almost certainly were using symbols to some extent on the reflective, systematized levels of symbols, including external records of events and ideas that Donald (1991) places in the final Theoretic stage of development. In societies like Egypt and Mesopotamia in the period from 4,000 to 3,000 years ago, and in Mesoamerica beginning around 2,500 years ago, there were elites with mastery of considerable complexities in spoken language, forms of written language, poetry, narrative, painting, sculpture, dance, music, myth, mathematics, religion, drama, and architecture. For the less elite of these societies, symbol use may have been far more restricted even in spoken language and mental lives may have been primarily at the Mythic, oral narrative level. No one knows at what point it became commonplace for the vast majority of children in every culture to achieve high levels (but still with plenty of individual variability in final language levels achieved) of spoken language complexity. Considerable mystery surrounds the status of language at around 30,000 years ago. What is clear, in contrast, is that many different symbolic systems were in evidence and that incontestable art creations survived. Art in the form of paintings, with clues therein to the occurrence as well of music and dance and ritual. Art in the form of sculpture. Art in the form of rock carvings from multiple continents. Flutes and other musical instruments were also available. What is known (cf. Tattersall, 1999) about such art and other symbolic productions suggests some strong parallels with modern human infants and toddlers as portrayed by Katherine Nelson. It appears plausible that for modern human forbearers before highly complex language emerges, there are rich mixes of varied ways of using symbol-action sequences. With variations across communities, these may have included not only drawing and painting but some mime and drama, sculpting, art as painted on the body and designed into clothes and grave decorations, dancing, singing, and perhaps other music, words, and “multimedia” mixes. This stress is similar to Donald’s up to a point. But the Dynamic Tricky Mix perspective suggests even more variability, more unevenness, in within-community and between-community symbolic status and also in which symbolic profiles (how much art, words, dance, music, etc ) were displayed. Donald (1991) found “uncontestable” the claim that spoken language has been part of modern human behavior for all of the last 50,000 years. However, for existing traditional societies that are very “stone age” in certain patterns of living, there is no hard evidence that their levels of languages complexity are similar to those existing 50,000 or even 20,000 years ago. Many other inferences have been drawn concerning the levels of language that would be “required” for certain levels of hunting, or tool use, or varied levels of cooperative behavior, but to date wildly different inferences have been made by different scientists (e.g., Johanson & Blake, 1996; Johanson & Shreeve, 1989; Jolly, 1999; Stringer & McKie, 1996; Tattersall, 1999). Moreover, there is no hard evidence of the vocabulary, syntax, and narratives of any hominids living more than about 10,000 years ago. Going further back in time, the available artifacts and fossil remains both become increasingly sparse and inconclusive. Donald’s hypothesis
204
NELSON ET AL.
that a stage of prenarrative mimetic (imitative mimes and other reenactments) representation was a precursor (from late Homo erectus, at about 300,000 years ago) to true language and narrative representation has some plausibility, but it too leaves indeterminate the particular modes of such mimesis. Overall, then, an open mind seems appropriate concerning the ways in which dance, art, music, spoken language, sign language, and symbolic dramas and rituals may have been interwoven at any point during the period from –300,000 years to –10,000 years ago. Whatever the mix of symbol systems in development, whenever survival rates favor those with more complex symbolic profiles, then genetic shifts across generations may build toward better and better brain foundations for the teachability and learnability of varied kinds of symbol production, symbol comprehension, or both. Moreover, even within a generation there may be expansions of the symbol systems and the strategies for successfully teaching them. Reduction or elimination of other hominid species by disease or by aggression from another hominid species may in some instances help conditions of easy hunting, foraging, or travel that will mix in a dramatically positive fashion with other converging conditions favoring symbol complexity advances. Climate changes, including some fairly dramatic ones in the course of a single hominid generation, may have favored both overall flexibility and flexibility in symbolic communication (cf. Deacon, 1997; K. Nelson, 1988; Donald, 1991; Stringer & McKie, 1996). Very rapid joint advances in the percentage of symbol users in a hominid community and in the complexity of the symbol systems will be expected only when the tricky convergence of highly positive dynamic mix conditions occurs generation after generation. More typically, there will be long plateaus in which a fairly stable level of symbol use prevails. In the view taken here, a tricky, precarious set of converging conditions is required for significant steps forward in acquiring symbolic systems by an individual in its lifetime or by a community or by a species. A speculative comparison of how available time might be used can now be offered. There are available strong experimental demonstrations (many briefly reviewed earlier in this chapter) of children’s acquisition of substantial new symbolic levels after 16 hours or so of rich dynamic convergences in art, sign language, spoken language, and text. Assume that a child in just those 16 accelerated, very positive hours of learning acquired 1% of a symbolic domain. Then that 1% gain would have been achieved in .005% (.00005) of one rough estimate of a lifetime of 320,000 alert hours (e.g., 75 yr ´ 364 days ´ 12 highly alert hr). Similarly, from a punctuated equilibrium view of human evolution of both physical and cultural-symbolic characteristics, assume that certain 1% changes happen to be concentrated in particular dynamically convergent, rapid growth periods of just 1,000 years. These high acceleration periods contrast with the more typical average evolutionary rate—across no-growth, slow-growth, and high-growth periods—of just 1/10 of 1% (.001) change (one “Darwin”) in any significant characteristic in 1 million years. Note that the hypothetical fast pace here in 1,000 fast growth years, represents a pace of 10,000
9. A DYNAMIC TRICKY MIX PERSPECTIVE
205
Darwins (1,000% change in a million years). Clues that this may not be too far-fetched come from analyses in the last 10 years of the same finch species Darwin studied and other organisms such as tropical fish, but in the face of dramatic environmental changes and survival differentials (Grant, 1986; Weiner, 1994). Measurable changes in the survivors of droughts and other environmental shifts indicate genetic changes in 1 to 10 generations, and with a pace that wildly punctuates the genetic stability when the environment is stable—40,000 to 60,000 Darwins!
Sum of Parallels Between Modern Children and Hominid Evolution The previous comparisons strongly suggest that despite many contextual differences, some similar basic processes are at work in changes in hominid species characteristic biology and culture and changes in the symbolic skills across childhood for children in the 20th and 21st centuries. Plateaus with little or no change alternate with periods of moderate to rapid change. Change periods are created in large part by dynamic, real-time convergences in the lives of individuals that bring together challenges, readiness for the challenges, online emotional and social adjustment processes and strategic tools that support processing and learning of challenges, enhancers in terms of sheer amounts of time spent in interactions with varied symbol-using partners, contextual qualitative enhancers that also feed positively into attention and learning, and maintained neural networks that sustain efficient performance of already-acquired symbolic skills as well as further readiness for new learning. Time scales, of course, differ depending on which steps forward are at issue. But across many thousands of years in evolution and many weeks in a child’s life, an illusion of steady change may be created because progress has been measured at widely separated times and no “capture” is made empirically of the many potentially productive episodes that do not actually contribute to the developmental advances. Another angle of description on these phenomena is from exceptional but naturalistic cases of rich immersion and/or instruction totaling about 1,600 hours of excellent, deeply engaging dynamic mixes—sufficient in a great many cases for the achievement of complex, expert (still with room to grow, though) levels regardless of whether the symbolic domain is music, mathematics, literacy, sign language, spoken first language, second language, dance, drama, drawing, painting, or sculpture (B. S. Bloom, 1985; M. Csikszentmilhalyi & I. Csikszentmilhalyi, 1988; Ericsson, 1996). Huge variability in both the biological makeup and the culturally embedded experiences of modern-day children also provide a significant window into processes relevant to human evolution. Among those variations are learning rate variations (even when conditions are highly similar), acquisition pathway differences, and differential sensitivities to potentially enhancing/debilitating environ-
206
NELSON ET AL.
mental circumstances (cf. Fenson, Dale, Reznick, Bates, & Thal, 1994). Variability of individuals increases the chances that even if contexts for learning vary dramatically, some individuals will be able to learn even in those contexts with no learnability for the typical individual. Across evolutionary and individual history, the characteristics of the learner may change as a result of the contexts encountered (cf. Gottlieb, 2002). For individual children in the period between ages 1 and 5, if highly positive and supportive mixes of learning conditions concerning narrative, text, and first language vocabulary and syntax keep recurring, the child’s Readiness by about 5 years for new learning contexts including those of school may be very richly enhanced. The child may learn readily in new contexts in part because the child contributes to the online convergences of conditions—most notably by positive expectancies, well-established emotional-social regulation strategies, and highly practiced, redundant, and frequently reorganized networks of already established knowledge to compare with new challenges of any kind (Bransford et al., 2000; Dweck, 1990, 2000; K. E. Nelson, 2000; K. E. Nelson & Welsh, 1998; K. E. Nelson, Welsh, S. Camarata, Heimann, & Tjus, 2001). Part of this pattern of increased readiness incorporates large individual differences between children because of their past cycles of interaction and their in-built biological profiles. But for most children when they establish the kinds of increasingly sophisticated “true language” representations in syntax, vocabulary, and narrative that Donald and Nelson discussed under the Mythic Stage and the Theoretic Stage, new learning of additional communicative structures and new learning in general will mix in new kinds of learning-to-learn Readiness along with all the other familiar components of Dynamic Tricky Mixes.
REAL CHILDREN VERSUS IDEAL CHILDREN AND THE NATURE OF LANGUAGE ACQUISITION Around 300 BC there were no published accounts of the details of children’s conversations with others and their progress toward language mastery, so Platonic notions of ideal forms innately prepared in children’s minds had no grounding in language acquisition stages or in actual interactional observations. The same lack of grounding characterizes Chomsky’s speculations in 1964 to 1968 (e.g., Chomsky, 1968) that an innately given language (later “organ”) device underlies language acquisition in children. But, of course, the revolution in children’s language studies has grown slowly across the period 1964 to current work, and there was very little in 1964 to guide anyone’s theories. Nevertheless, since 1964, Chomsky and other writers who have endorsed his basic innate language organ/device have taken one of two stances toward the increasingly rich, crosslinguistic set of empirical studies on children’s interactional communicative patterns and children’s pathways toward language mastery, with Chomsky himself (e.g., 1986) actively avoiding examination of such evidence (cf. Hyams, 1986;
9. A DYNAMIC TRICKY MIX PERSPECTIVE
207
Lightfoot, 1989), or instead selectively reviewing observations on actual children and their conversational experience so that only the small minority of convenient, friendly observations are cited to “demonstrate” the strength of Chomsky’s position (e.g., Pinker, 1994). Within this literature defending Chomsky’s position, five kinds of Ideal forms are assumed. “The child” is assumed to be one entity, in contrast to substantial biological variations covered by countless theorists (cf. Elman et al., 1996; Gottlieb, 2002; K. Nelson, 1981; K. E. Nelson, 1980, 1991a) and subsumed in Dynamic Tricky Mix theory under Readiness and Networks. “The input” is assumed to be for all children essentially the same and so impoverished, so chaotic, and so inconsistent as to force one to conclude that a child could only come to use language if it was innate. “The language organ” is assumed to be one organ (perhaps with some separate modules) distinct from the rest of the brain in some fashion. Language “acquisition” is seen as one process of triggering and selecting particular innate representations that fit with the child’s language community, in contrast to the many variations on language acquisition pathways within culture that have been observed. “Language” is seen as formal, fixed linguistic structures (paradoxically changing in nature with the ebb and flow of competing linguistic theories) in the head, rather than as language embodied in social interactional goals and contexts. Within the broader scientific community, there are many efforts that challenge all of these idealizations of the Chomskian approach. Scientific investigations can be seen to jointly seek accurate characterizations of the biological mechanisms contributing to learning language and other symbolic systems, understanding of both similarities and differences between spoken and signed languages and other symbolic domains, individual differences in social communicative interactions during childhood, individual differences among children in their makeup and in their pathways through language acquisition, and understanding of language itself. Many findings and theoretical conceptualizations arising in the context of these broader investigations have been touched on earlier in this chapter. Taking these together with a diverse set of scholarly writings in developmental psychology and cognitive science and child language (e.g., Bohannon, Padgett, K. E. Nelson, & Mark, 1996; Bonvillian & Folven, 1993; Bowerman, 1985; Elman et al., 1996; Mahoney, 1991; Maratsos & Chalkley, 1989; Maratsos & Deak, 1995; K. Nelson, 1998; Slobin, 1985), it is clear that there has been much progress in accounting for many aspects of the very powerful general learning and abstraction mechanisms the child brings to the specific tasks of learning spoken languages and sign languages, and moreover that there is much progress in explaining how the complexities of a vast and often abstract language system is constructed by children. In 1964–1968, it was merely questionable for Chomsky to claim that “the input” was chaotic and that there were no means of accounting for language acquisition without very powerful innate language-specific structures, but in 2002 it would run counter to a deep and broad literature to claim either that specific facilitators typically are lacking in conversational input or that innate mechanisms
208
NELSON ET AL.
are required by default because no other acquisition accounts could plausibly deal with the complexity of language acquisition. Next we take a closer look at exceptional children with Specific Language Impairment (SLI) or autism. When a child fails to progress in language acquisition for substantial periods, as is the case for most autistic children and some children with Specific Language Impairment, then the explanation from the Chomskian perspective must be in the genes, in a somehow “broken” language organ, rather than in any special deficiencies in an already impoverished idealized notion of “the input” available to children. Indeed, at an operational level the ways in which most children with severe language delay have been treated until recently is highly consistent with Chomskian notions—children are given very little treatment that incorporates adult broad fluency in language within engaging conversational episodes. Thus, self-fulfilling negative cycles are set up. Children with broken language organs are not expected to get better, do not get conversationally enriched treatment, and when they do not improve their language levels much across many years further negative expectations are set up for the next generation of language-delayed children. Longitudinal studies based on earlier ways of treating language delay show, unfortunately, that children with Specific Language Impairment were indeed likely to have later problems in adolescence and adulthood with language production and comprehension, literacy, and academic achievement (Aram, Edelman, & Nation, 1984; Records, Tomblin, & Freese, 1992; Strothard et al., 1998). However, by now there are many exceptions to this pattern. These exceptions should be taken seriously. Consider documented cases when a child has not been making progress in language for many years, but then receives enriched language input as part of socially and emotionally engaging conversational interactions for about 2 or 3 hours per week over the course of 6 months to a year and shows excellent progress in acquiring language. Because arranging—naturalistically or in interventions—highly supportive conversational exchanges for language acquisition is very tricky and complex, and because children bring important individual differences in their temperament, emotions, and biological preparedness for learning complex systems, fitting stalled-then-fast acquisition patterns to Dynamic Tricky Mix theory is straightforward. Chomskian models bog down in the face of the evidence on when children stall and when they move forward in language acquisition. If the stalling at say age 4 or 5 by a child with SLI or autism is attributed to a broken or damaged language organ, then when some children not only show progress but end up acquiring a full language system (Bonvillian & K. E. Nelson, 1978; R. Koegel & L. Koegel, 1995) a Chomskian theorist could “accommodate” by reversing field and claiming an intact language acquisition device after all. But then there is a circularity rather than theoretical explanation, because the only evidence of the language organ appears to be success at language acquisition and success at language acquisition is attributed to the presumed language organ. A different approach to the biological contributions to language acquisition is
9. A DYNAMIC TRICKY MIX PERSPECTIVE
209
to recognize the great many ways in which human children differ in their profiles of strength and weakness in attention, memory, pattern abstraction, sequential planning of complex motor sequences, perception, and potentially language-specific mechanisms of some sort, and then to convergently seek evidence on how these individual patterns of Readiness in children play out under different mixes of social–cultural–interactive learning conditions to produce individual pathways in language acquisition. In this regard, make no mistake, multiple theories (including Dynamic Tricky Mix theory) attribute very powerful biologically in-built acquisition processes to children by typical maturational ages of from 6 to 12 months, and as maturation and learning proceed, the available processes for further acquisition become far greater in their power during the period from 1 to 4 years.
CONCLUSIONS Painting, sculpture, dance, spoken languages, sign languages, text, mathematics, music, and multimedia—all are modern symbolic systems that are highly complex. Their remarkable complexity far exceeds any documented symbolic system 20,000 years ago. Children today are able to learn these complex communicative systems only because they bring very powerful cognitive and social learning mechanisms into any communicative interactions and because every culture provides ordinary children extraordinary quantities of social communicative episodes of some sort accompanied by sporadic richness of scaffolding specifically relevant to supporting the child’s developmental progress in symbols. For most children, most of the time, language or other symbol system acquisition is not taking place. From the Dynamic Tricky Mix theoretical perspective presented here, learning rises above zero only when challenges are accompanied in real-time by dynamic convergence of positive motivational, social, emotional, and processing enhancers. When all key conditions become increasingly positive in such dynamic convergences, the rates of acquisition by children correspondingly increase as well. Accordingly, it is “tricky” and unusual to see a child show sustained high rates of learning over many years, but it is predictable whenever there are many recurrent social episodes in which there are highly positive convergences of learning conditions within and outside the child. Acquiring systematic literacy skills or systematic fluency in a first language is a gradual process with the required acquisition of very many different structures and skills. This chapter lays out considerable detail about the acquisition of one syntactic structure, the passive. Children who would not be expected from the literature of 20 years ago to learn the passive easily, very young children under 30 months of age, and children with the significant language delays of SLI also with quite young language ages, have been shown to be quite Ready cognitively and linguistically. Newly described convergent dynamic mixes that support such surprising acquisition of the passive are provided by natural cultural variation sup-
210
NELSON ET AL.
porting young children’s acquisition in such languages as Inuktitut and Sesotho and also by recast conversational intervention in English for children with SLI. Thus, the most effective mixes of conditions for supporting early and rapid passive acquisition are similar to those that appear highly favorable for acquisition of social scripts, narrative structure, and a broad variety of syntactic and pragmatic skills. All of these involve the convergent mixing together of clear-cut complexity challenges to the child within small episodes where the child and others co-construct shared meanings in a positive emotional-social context that also provides occasional scaffolding enhancers such as recasting and paraphrasing, subtle questioning and prompting, and sensitive topic extensions. A sense of humor is often part of these dynamic mixes, and it certainly is in the case of excellent (judged by their students objective gains) tutors in mathematics, who rely on “indirect” ways of layering in challenges that have much in common with what has been described herein for facilitative language-learning contexts (Lepper, Woolverton, Mumme, & Gurtner, 1993). The theoretical account in this chapter is primarily from the angle of the child as a learning system, supported or not supported by some convergence of enhancers (recast and narrative sequences, e.g.) and certain developmentally appropriate challenges and many other resources provided by social interactive partners. As Nelson (1998) advocated in her discussions of the “mediated mind,” the primary account is of “the interplay of individual experience and the engulfing cultural communicative society” (p. 327). But in reality there are many kinds of systems with mutual influence over developmental time. One can move to the level of the child’s parent or the child’s clinician approaching interactions with the child. The adult not only brings certain communicative strategies but a whole system of strategies for interaction and for self-regulation, a certain temperament, and a specific constellation of knowledge and skills. Similarly, the child–adult dyad can be treated as a system in other analyses. Consider once again the detailed analyses of passive acquisition in relation to shifts in what happens within dyads on the level of particular passive exemplars embedded in particular discourse and social-emotional-cultural patterns. When the mix of conditions within the dyad includes richness of challenges and richness of adjustment and processing enhancers, early and fast passive acquisition is seen by the child. But there likely are child to adult causal influences operating here too, as when a child who shows early passive comprehension or produces simple early passives (evidence of a hot spot in acquisition) may influence the adult to provide still richer quantities of passives and richer qualities of passive scaffolds. Similarly, there are multiple levels of system analyses to apply to issues of match or mismatch between a child’s characteristic ways of interacting socially and approaching challenges as compared with a particular adult’s ways of being in the world and approaching the enculturation of a child generally and also specifically as regards the child’s acquisition and use of language and other symbolic systems.
9. A DYNAMIC TRICKY MIX PERSPECTIVE
211
Shifts across many months and years in child learner and potential adult scaffolder may also be an essential part of children’s varied success at mastering symbolic system. Positive expectancies of interesting interaction, high attentiveness, positive emotion, and smooth emotion regulation all are considered to contribute to positive and effective overall mixes of conditions for particular children on particular occasions of encountering challenges. But these contributing factors are also influenced over many occasions and months by the kinds of “engulfing” cultural interactions the child regularly experiences. Over time, some children will tend to bring more and more favorable conditions as learners and social partners, so that—other things being equal on new potential learning occasions—they will be more likely to experience dynamic convergence mixes leading to learning. This kind of positive cycling toward more and more positivity and learning could proceed in a way that generalizes across most learning contexts and social partners, or it could instead be confined to new interactions with particular social partners who have shared many prior positively toned and learning-effective episodes. Regardless of the particular patterns that unfold, common processes will apply, with learners and their social partners in transactional fashion influencing each other across time and thus helping to set up Readinesses in the learner and Enhancers in their partners that will be essential contributors to whatever dynamic convergences emerge on new occasions (Dweck, 2000; K. E. Nelson, 2000, 2001; Nicely et al., 1999; Sameroff & Fiese, 1989). Illustrations of these complex patterns over time have been given for passives in discourse in this chapter and in a range of prior work for syntactic structures and lexical acquisition and narratives (e.g., Bamberg, 2001; Bohannon et al., 1996; Fivush, 1991, 1994; Kaiser et al., 1992; K. E. Nelson & K. Nelson, 1978; K. E. Nelson & Welsh, 1998; K. E. Nelson et al., 1995; K. E. Nelson et al., 2001). Sustained rapid advances by a child, a culture, or a species will be relatively rare for all the reasons so far articulated on trickiness of mixes, but it will occur under predictable sets of circumstances—precisely when positive dynamic mixes are themselves sustained and very highly positive. Important educational implications are that small but well-chosen perturbations have the potential to have very substantial positive effects on learner progress (cf. Bransford et al., 2000; Perkins, 1993). Research in the future should better elucidate the dynamic mixes associated with zero, slow, moderate, and turbo rates of developmental advances (cf. Campione & Brown, 1987; K. E. Nelson, 2001; K. E. Nelson, Perkins, & Lepper, 2003). To achieve this, more studies are needed that measure in an online dynamic fashion what Launching challenges are encountered, what processing Enhancers are present, what levels of emotional Adjustment and social engagement and expectancies are in place, what Readiness characteristics and Network characteristics are activated in the learner by the context, and what sorts of learning or stalling are emergent. In this search for refined developmental accounts, it should be remembered that art and music and mathematics are no less species-specific in natu-
212
NELSON ET AL.
ral contexts than language. When these other symbolic systems reach maximum complexity and fluency even if only in a small percentage of modern humans, the dynamic mixes that support such high-level acquisition need to be analyzed and compared with those that support successful and varied pathways to high language fluency. And in all such work on developmental theory care must be taken to recognize and in some ways work around the limits of any convenient abstractions such as “word,” “sentence,” “drawing,” “concerto,” “narrative,” “language,” “deferred imitation,” “relative clause,” “first derivative,” and so on. None of these are things in the brain, but rather are meaningful only as emergent phenomena in real-time in dynamic fashion as part of the ongoing life of creatures with complex levels of consciousness, a body sense, goals, emotions, motives, expectancies, perceptions, and vast networks of representations available to parallel retrieval and reorganization (L. Bloom & Tinker, 2001; Damasio, 1994, 1999; Edelman & Tononi, 2000; Thelen & Smith, 1994). Striking flexibility in acquiring diverse symbol systems is characteristic of modern humans, in part demonstrated by the full sweep of children’s acquisitions considered in this chapter but also by a broader literature on exceptional children including those considered gifted and those considered autistic savants. Consider art, music, mathematical calendar and other calculations, spoken language. sign language, and architectural design. Each of these symbol systems may reach high levels of complexity and fluency without the other symbol systems being in place (Dawson, 1989; Gould, 1997; K. E. Nelson, 1991b; Obler & Menn, 1984; Sloboda, 1996). Dual mastery of exceptional combinations of these symbol systems also occurs, as in cases of deaf artists fluent in sign language but not in spoken language, or language-delayed children who become excellent in mathematics, art, or music before they reach even moderate complexity levels in spoken language. By contrast, in modern literate and industrial-technological societies the typical sequences of spoken language mastery followed (with just moderate overlap with late language acquisition stages) by acquisition of written language and optional mastery of music and/or mathematical systems or, quite rarely, significant art skill, barely scratch the surface of most children’s Readiness for flexible symbol system acquisition. Put simply, typical sequences depend on typical patterns of dynamic mixes and the many interesting exceptions concerning what symbol systems are acquired depend on exceptional sorts of dynamic mixes. Circularity is escaped because both natural and planned “experiments” indicate that when exceptional rich challenges and supportive learning conditions are provided, then art, sign language, mathematics, and literacy often prove highly learnable in children with no prior history of such achievements. Parallels with the long view back into evolution once again are intriguing. For the moment, imagine that the accumulated graphic records of Stephen Wiltshire’s accomplished productions in art were discovered but with absolutely no record of Stephen’s life circumstances or his fluency levels in spoken language, written lan-
9. A DYNAMIC TRICKY MIX PERSPECTIVE
213
guage, or any other symbolic system. What inferences would be drawn of the language levels that would be “required” for such graphic complexity ? If found in 2003 in an old farmhouse attic, typical 2003 language levels would no doubt be presumed. But shift the context of discovering Stephen’s art to the caves of Spain and France, below bison and lions and hunters and voluptuous women painted 25,000 or more years ago on the cave roofs and walls, then considerable dissociation of art mastery levels and language mastery levels would be “legal” or plausible to most scholars. Considering more broadly, then, the status of potential symbol users 25,000 to 250,000 years ago, this chapter has argued that these early periods of symbol development reflected some of the same basic processes observable today across a full range of exceptional as well as typical development and interventionist as well as naturalistic investigations. Thus, it is highly probable that given a Ready-enough brain for complex symbols, the course of symbol-system development could follow many mixed modalities, that different dynamic mixes of conditions existed in different social groups and ecological and temporal niches, and that progress in each of the symbol systems (or mixed symbol systems) showed dramatic irregularities.
ACKNOWLEDGMENTS Background research and reflection underpinning the ideas and data presented in this chapter were provided by grants to the first author from the National Science Foundation (BNS-8013767), National Institutes of Health (MH 19826h, HD 06254, R01-NS26437, and R01DC00508 and 1P50 DC0382-01898), the U.S. Department of Education (G008302959, G008430079, G0008300361), and the Hasbro Children’s Foundation.
REFERENCES Akhtar, N., & Tomasello, M. (1997). Young children’s productivity with word order and verb morphology. Developmental Psychology, 33, 952–965. Allen, S., & Crago, M. (1996). Journal of Child Language, 23, 129–155. Aram, D., Edelman, B., & Nation, J. (1984). Preschoolers with language disorder: 10 years later. Journal of Speech and Hearing Research, 27, 232–244. Arkenberg, M. E., & Ryalls, B. O. (2003). Individual differences in children’s word learning: Dynamic mixes of underlying conditions. Unpublished manuscript, Penn State University. Aronsson, K. (1997). Barns vårldar—barns bilder. Stockholm: Bokførlaget Natur och Kultur. Baker, L., & Cantwell, D. P. (1982). Language acquisition, cognitive development, and emotional disorder in childhood. In K. E. Nelson (Ed.), Children’s language (Vol. 3, pp. 286–321). Hillsdale, NJ: Lawrence Erlbaum Associates. Baker, N. D., & Nelson, K. E. (1984). Recasting and related conversational techniques for triggering syntactic advances by young children. First Language, 5, 3–22.
214
NELSON ET AL.
Bamberg, M. (2001). Why young American English-speaking children confuse anger and sadness: A study of grammar in practice. In K. E. Nelson, A. Koc, & C. E. Jonson (Eds.), Children’s language (Vol. 10, pp. 55–72). Mahwah, NJ: Lawrence Erlbaum Associates. Bates, E., Bretherton, I., & Snyder, L. (1988). From first words to grammar: Individual differences and dissociable mechanisms. New York: Cambridge University Press. Blachman, B. (1997). Foundations of reading acquisition and dyslexia: Implications for early intervention. Mahwah, NJ: Lawrence Erlbaum Associates Bloom, B. S. (1985). Developing talent in young people. New York: Ballantine. Bloom, L. (1991). Language development from two to three. New York: Cambridge University Press. Bloom, L., & Tinker, E. (2001). The intentionality model and language acquisition: Engagement, effort, and the essential tension in development. Monographs of the Society for Research in Child Development, 66(4). Bloom, L., Tinker, E., & Margulies, C. (1993). The words children learn: Evidence against a noun bias in early vocabularies. Cognitive Development, 8, 431–450. Bohannon, J. N., Padgett, R. J., Nelson, K. E., & Mark, M. (1996). Useful evidence on negative evidence. Developmental Psychology, 33, 551–555. Bohannon, N., & Stanowicz, L. (1989). Bidirectional effects of imitation and repetition in conversation: A synthesis within a cognitive model. In G. E. Speidel & K. E. Nelson (Eds.), The many faces of imitation in language learning (pp. 121–150). New York: Springer-Verlag. Bonvillian, J. D., & Folven, R. (1993) Sign language acquisition: Developmental aspects. In M. Marschark & D. Clark (Eds.), Psychological perspectives on deafness (pp. 229–268). Hillsdale, NJ: Lawrence Erlbaum Associates. Bonvillian, J. D., & Nelson, K. E. (1978). Development of sign languages in autistic children and other language-handicapped individuals. In P. Siple (Ed.), Understanding language through sign language research. New York: Academic Press. Bowerman, M. (1985). Beyond communicative adequacy: From piecemeal knowledge to an integrated system in the child’s acquisition of language. In K. E. Nelson (Ed.), Children’s language (Vol. 5, pp. 1257–1320). Hillsdale, NJ: Lawrence Erlbaum Associates. Bowey, J. A. (1996). On the association between phonological memory and receptive vocabulary in five-year-olds. Journal of Experimental Child Psychology, 63, 44–78. Bransford, J.D., Brown, A. L., & Cocking, R. R. (2000) How people learn: Brain, mind, experience, and school. Washington, D C: National Academy Press. Bruner, J. S.(1986). Actual minds, possible words. Cambridge, MA: Harvard University Press. Caldwell, L. B. (1997). Bringing Reggio Emilia home. New York: Teachers College Press. Camarata, S., Nelson, K. E., & Camarata, M. (1994). A comparison of conversation based to imitation based procedures for training grammatical structures in specifically language impaired children. Journal of Speech and Hearing Research, 37, 1414–1423. Campione, J., & Brown, A. L. (1987). Linking dynamic assessment with school achievement. In C. S. Lidz (Ed.), Dynamic assessment: An interactional approach to evaluating learning potential (pp. 82–114). New York: Guilford. Carey, S. (1978). The child as word learner. In M. Halle, J. Bresnan, & G. A. Mutter (Eds.), Linguistic theory and psychological reality (pp. 264–293). Cambridge, MA: MIT Press. Chomsky, N. (1968). Language and mind. New York: Harcourt, Brace, Jovanovich.
9. A DYNAMIC TRICKY MIX PERSPECTIVE
215
Chomsky, N. (1986). Knowledge of language: Its nature origin and use. New York: Praeger. Clark, E. V. (1981). Lexical innovations: How children learn to create new words. In W. Deutsch (Ed.), The child’s construction of language (pp. 299–328). London: Academic Press. Cole, P. M., Michel, M. K., & Teti, L. O. (1994). The development of emotion regulation and dysregulation: A clinical perspective. In N. A. Fox (Ed.), The development of emotion regulation: Biological and behavioral considerations. Monographs of the Society for Research in Child Development, 59 (Serial No. 240), 73–100. Cosmides, L., & Tooby, J. (1987). From evolution to behavior: Evolutionary psychology as the missing link. In J. Dupre´ (Ed.), The latest on the best: Essays on evolution and optimality (pp. 163–228). Cambridge, MA: MIT Press. Csikszentmihalyi, M., & Csikszentmihalyi, I. (1988). Optimal experience. New York: Harper Collins. Damasio, A. (1994). Descartes’ error: Emotion, reason and the human brain. New York: Putnam. Damasio, A. (1999). The feeling of what happens: Body and emotion in the making of consciousness. New York: Harcourt Brace. Dawson, G. (1989). Autism: Nature, diagnosis & treatment. New York: Guilford. Deacon, T. W. (1997). The symbolic species: The co-evolution of language and the brain. New York: Norton. de Houwer, A. (1995). Bilingual acquisition. In P. Fletcher & B. MacWhinney (Eds.), The handbook of child language (pp.219–250). London: Blackwell. Demuth, C. (1990). Subject, topic, and Sesotho passive. Journal of Child Language, 17, 67–84. Donald, M. (1991). Origins of the modern mind. Cambridge, MA: Harvard University Press. Dweck, C. S. (1990). Motivation. In A. Lesgold & R. Glaser (Eds.), Foundations for a psychology of education (pp. 87–136). Hillsdale, NJ: Lawrence Erlbaum Associates. Dweck, C. S. (2000). Self-theories. Philadelphia: Taylor & Francis. Edelman, G. M., & Tononi, G. (2000). A universe of consciousness. New York: Basic Books. Elman, J. L., Bates, E. A., Johnson, M. H., Karmiloff-Smith, A., Parisi, D., & Plunkett, K. (1996). Rethinking innateness: A connectionist perspective on development. Cambridge, MA: MIT Press. Ericsson, K. A. (1996). The acquisition of expert performance: An introduction to some of the issues. In K. A. Ericsson (Ed.), The road to excellence (pp. 1–50). Hillsdale, NJ: Lawrence Erlbaum Associates. Fenson, L., Dale, P.A., Reznick, J. S., Bates, E., & Thal, D. (1994). Variability in early communicative development. Monographs of the Society for Research in Child Development, 58 (5, Serial number 242). Fey, M. M. E., Catts, H., & Larrivee, L. (1995). Preparing preschoolers for the academic and social challenges of school. In M. M. E. Fey, J. Windsor, & S. Warren (Eds.), Language intervention: Preschool through the elementary years (pp. 3–37). Baltimore, MD: Brookes. Fey, M. M. E., Cleave, P., & Long, S. (1997). Two models of grammar facilitation in children with language impairments: Phase 2. Journal of Speech and Hearing Research, 40, 5–19. Fey, M. M. E., Cleave, P., Long, S., & Hughes, D. (1993). Two approaches to the facilitation of grammar in children with language impairment: An experimental evaluation. Journal of Speech and Hearing Research, 36, 141–157.
216
NELSON ET AL.
Fillmore, L. W. (1989). Teachability and second language acquisition. In M. Rice & R. Schiefelbusch (Eds.), The teachability of language (pp. 311–332). Baltimore: Brookes. Fivush, R. (1991). The social construction of personal narratives. Merrill-Palmer Quarterly, 37, 59–81. Fivush, R. (1994). Constructing narrative, emotion, and self in parent–child conversations about the past. In U. Neisser & R. Fivush (Eds.), The remembering self: Construction and accuracy in the self-narrative (pp. 136–157). New York: Cambridge University Press. Fivush, R., & Slackman, E. (1986). The acquisition and development of scripts. In K. Nelson (Ed.), Event knowledge: Structure and function in development (pp. 71–96). Hillsdale, NJ: Lawrence Erlbaum Associates. Gathercole, S. E., & Baddeley, A. D. (1990). The role of phonological memory in vocabulary acquisition: A study of young children learning arbitrary names of toys. British Journal of Psychology, 81, 439–454. Gathercole, S. E., Service, E., Hitch, G. J., & Martin, A. (1997). Phonological short-term memory and new word learning in children. Developmental Psychology, 33, 966–979. Gathercole, S. E., Service, G. J., Hitch, G. J., Adams, A., & Martin, A. J. (1999). Phonological short-term memory and vocabulary development: Further evidence on the nature of the relationship. Applied Cognitive Psychology, 13, 65–77. Gillam, R., McFadden, T. U., & van Kleeck, A. (1995). Improving narrative abilities. In M. Fey, J. Windsor, & S. Warren (Eds.), Language intervention: Preschool through the elementary years (pp. 145–182). Baltimore: Brookes. Golomb, C. (1992). The child’s creation of a pictorial world. Berkeley, CA: University of California Press. Golomb, C. (1995). (Ed.). The development of artistically gifted children. Hillsdale, NJ: Lawrence Erlbaum Associates. Gottlieb, G. (2002). Developmental-behavioral initiation of evolutionary change. Psychological Review, 109, 211–218. Gould, S. J. (1989). Wonderful life. New York: Norton. Gould, S. J. (1997). Questioning the millennium. New York: Crown. Grant, P. R. (1986). Ecology and evolution of Darwin’s finches. Princeton, NJ: Princeton University Press. Greenough, W. T., Black, J.E., & Wallace, C. S. (1987). Experience and brain development. Child development, 58, 539–559. Hagen, R., & Hagen, R. (1999). Egypt: People, gods, pharaohs. Cologne: Taschen Verlag. Haley, K., Camarata, S., & Nelson, K. E. (1994). Positive and negative social valence in children with specific language impairment during imitation based and conversation based language intervention. Journal of Speech and Hearing Research, 37, 141–148. Hart, B., & Risley, T. R. (1995). Meaningful differences in the everyday experience of young American children. Baltimore: Brookes. Heimann, M., Nelson, K.E., Tjus, T., & Gillberg, C. (1995). Increasing reading and communication skills in children with autism through an interactive multimedia computer program. Journal of Autism and Developmental Disorders, 25, 459–480. Ho, W.-C. (1989). Yani, the brush of innocence. New York: Hudson Hills Press. Horgan, D. (1978). The development of the full passive. Journal of Child language, 5, 65–80.
9. A DYNAMIC TRICKY MIX PERSPECTIVE
217
Hudson, J. A. (1990). The emergence of autobiographic memory in mother–child conversation. In R. A. Fivush & J. A. Hudson (Eds.), Knowing and remembering in young children (pp. 166–196). New York: Cambridge University Press. Hudson, J. A. (2002). “Do you know what we’re going to do this summer?”: Mothers’ talk to preschool children about future events. Journal of Cognition and Development, 3, 49–72. Hudson, J. A., & Shapiro, L. R. (1991). From knowing to telling: The development of children’s scripts, stories, and personal narratives. In A. McCabe & C. Peterson (Eds.), Developing narrative structure (pp. 89–136). Hillsdale, NJ: Lawrence Erlbaum Associates. Hyams, N. M. (1986). Language acquisition and the theory of parameters. Dordrecht, Holland: D. Reidel. Johanson, D., & Blake, E. (1996). From Lucy to language. New York: Nevraumont. Johanson, D., & Shreve, J. (1989). Lucy’s child: The discovery of a human ancestor. New York: Morrow. Jolly, A. (1999). Lucy’s legacy: Sex and intelligence in human evolution. Cambridge, MA: Harvard University Press. Jusczyk, P. W. (1993). From general to language-specific capabilities: The WRAPSA model of how speech perception develops. Journal of Phonetics, 21, 3–28. Jusczyk, P. W., Frederici, A. D., Wessels, J., Svenkerud, V. Y., & Jusczyk, A. M. (1993). Infants’ sensitivity to the sound patterns of native language words. Journal of Memory and Language, 33, 630–645. Kaiser, A., Yoder, P. J., & Keetz, A. (1992). Evaluating milieu teaching. In S. F. Warren & J. Reichle (Eds.), Causes and effects in communication and language intervention (pp. 9–47). Baltimore: Brookes. King, B. J. (Ed.). (1999). The origins of language: What nonhuman primates can tell us. Sante Fe, NM: School of American Research Press. Klein, S. B., Cosmides, L., Tooby, J., & Chance, S. (2002). Decisions and the evolution of memory: Multiple systems, multiple functions. Psychological Review,109, 306–329. Koegel, R., & Koegel, L. (1995). Teaching children with autism. Baltimore: Brookes. Leopold, W. F. (1939). Speech development of a bilingual child: A Linguist’s record. Vocabulary growth in the first two years. Evanston, IL: Northwestern University Press. Lepper, M., Woolverton, M., Mumme, D. L., & Gurtner, J. L. (1993). Motivational techniques of expert human tutors: Lessons for the design of computer-based tutors. In S. P. Lajoie & S. J. Derry (Eds.), Computers as cognitive tools (pp. 75–105). Hillsdale, NJ: Lawrence Erlbaum Associates. Lieberman, P. (2000). Human language and our reptilian brain. Cambridge, MA: Harvard University Press. Lightfoot, D. (1989). The child’s trigger experience: Degree-0 learnability. Behavioral and Brain Sciences, 12, 321–375. Mahoney, M. (1991). Human change processes. New York: Basic Books. Maratsos, M., & Deak, G. (1995). Hedgehogs, foxes, and the acquisition of verb meanings. In M. Tomasello & W. E. Merriman (Eds. ). Beyond names for things: Children’s acquisition of verbs (pp. 377–404). Mahwah, NJ: Lawrence Erlbaum Associates. Maratsos, M. P., & Chalkley, M. (1980). The internal language of children’s syntax: The ontogenesis and representation of syntactic categories. In K. E. Nelson (Ed.), Children’s language (Vol. 2, pp. 127–214). Hillsdale, NJ: Lawrence Erlbaum Associates.
218
NELSON ET AL.
Marchman, V. A., Bates, E., Burkhardt, A., & Good, A. B. (1991). Functional constraints on the acquisition of the passive: Toward a model of the competence to perform. First Language, 11, 65–92. Maxwell, M. (1983). Language acquisition in a deaf child of deaf parents: Speech, sign variations, and print variations. In K. E. Nelson (Ed.), Children’s language (Vol. 4, pp. 283–314). Hillsdale, NJ: Lawrence Erlbaum Associates. Mervis, C. B., & Pani, J. R. (1980). Acquisition of basic object categories. Cognitive Psychology, 12, 496–522. Nakisa, R. C., & Plunkett, K. (1998). Evolution of a rapidly learned representation for speech. Language and Cognitive Processes, 13, 105–127. Nelson, K. (1973). Structure and strategy in learning to talk. Monographs of the Society for Research in Child Development, 38(149). Nelson, K. (1981). Individual differences in language development: Implications for development and language. Developmental Psychology, 17, 170–187. Nelson, K. (1986). Event knowledge: Structure and function in development. Hillsdale, NJ: Lawrence Erlbaum Associates. Nelson, K. (1988). Constraints on word learning? Cognitive Development, 3, 221–246. Nelson, K. (1998). Language in cognitive development: Emergence of the mediated mind. Cambridge, England: Cambridge University Press. Nelson, K. (2002). Developing a socially shared symbolic system. In E. Amsel & J. P. Byrnes (Eds.), Language, literacy, and cognitive development: The development and consequences of symbolic communication (pp. 27–58). Mahwah, NJ: Lawrence Erlbaum Associates. Nelson, K. E. (1977). Facilitating children’s syntax acquisition. Developmental Psychology, 13, 101–107. Nelson, K. E. (1980). Theories of the child’s acquisition of syntax: A look at rare events and at necessary, catalytic, and irrelevant components of mother–child conversation. Annals of the New York Academy of Sciences, 345, 45–67. Nelson, K. E. (1981). Toward a rare-event cognitive comparison theory of syntax acquisition. In P. S. Dale & D. Ingram (Eds.), Child language: An international perspective (pp. 229–240). Baltimore, MD: University Park Press. Nelson, K. E. (1982). Experimental gambits in the service of language acquisition theory. In S.A. Kuczaj (Ed.), Language development, syntax and semantics. Hillsdale, NJ: Lawrence Erlbaum Associates. Nelson, K. E. (1987). Some observations from the perspective of the rare event cognitive comparison theory of language acquisition. In K. E. Nelson (Ed.), Children’s language (Vol. 6, pp. 289–331). Hillsdale, NJ: Lawrence Erlbaum Associates. Nelson, K. E. (1989). Strategies for first language teaching. In M. Rice & R. Schiefelbusch (Eds.), The teachability of language (pp. 263–310). Baltimore: Brookes. Nelson, K. E. (1991a). On differentiated language learning models and differentiated interventions. In N. Krasnegor, D. Rumbaugh, & R. Schiefelbusch (Eds.), Language acquisition: Biological and behavioral determinants (pp. 399–428). Hillsdale, N. J.: Lawrence Erlbaum Associates. Nelson, K. E. (1991b). Varied domains of development: A tale of LAD, MAD, SAD, DAD, and RARE and surprising events in our RELMS. In F. S. Kessel, M. H. Bornstein, & A. J. Sameroff (Eds.), Contemporary constructions of the child: Essays in honor of William Kessen (pp. 123–142). Hillsdale, NJ: Lawrence Erlbaum Associates.
9. A DYNAMIC TRICKY MIX PERSPECTIVE
219
Nelson, K. E. (2000). Methods for stimulating and measuring lexical and syntactic advances: Why Fiffins and lobsters can tag along with other recast friends. In L. Menn & N. B. Ratner (Eds.), Methods for studying language production. Hillsdale, NJ: Erlbaum Nelson, K. E. (2001). Dynamic tricky mix theory suggests multiple analyzed pathways (MAPS) as an intervention approach for children with autism and other language delays. In S. von Tetzchner & J. Clibbens (Eds.), Understanding the theoretical and methodological bases of augmentative and alternative communication (pp. 141–159). Toronto: International Society for Augmentative and Alternative Communication. Nelson, K. E., Baker, N. A., Denninger, M., Bonvillian, J., & Kaplan, B. (1985). Cookie versus do-it-again: Imitative-referential and personal-social-syntactic- initiating language styles in young children. Linguistics, 23, 433–454. Nelson, K. E., & Bonvillian, J. D. (1978). Early language development: Conceptual growth and related processes between 2 and 4½ years of age. In K. E. Nelson (Ed.), Children’s language (Vol. 1, pp. 467–556). New York: Gardner Press. Nelson, K. E., Camarata, S. M., Welsh, J., Butkovsky, L., & Camarata, M. (1996). Effects of imitative and conversational recasting treatment on the acquisition of grammar in children with Specific Language Impairment and younger language-normal children. Journal of Speech and Hearing Research, 39, 850–859. Nelson, K. E., Craven, P. L., Xuan, Y., Arkenberg, M.E., & Lauck, G. (2002, July). Acquisition of passive sentence structures by children during intervention and under naturalistic conditions. Paper presented at the joint conference of the International Association for the Study of Child Language and the Symposium on Child Language Disorders, Madison, WI. Nelson, K. E., Heimann, M., & Tjus, T. (1997). Theoretical and applied insights from multimedia facilitation of communication skills in children with autism, deaf children, and children with motor or learning disabilities. In L. B. Adamson & M. A. Romski (Eds.), Research on communication and language disorders: Contributions to theories of language development (pp. 296–325). Baltimore: Brookes. Nelson, K. E., & Nelson, K. (1978). Cognitive pendulums and their linguistic realization. In K. E. Nelson (Ed.), Children’s language (Vol. 1, pp. 223–286). Hillsdale, N.J.: Lawrence Erlbaum Associates. Nelson, K. E., & Pemberton, E. (2003). Child art. Unpublished manuscript, Penn State University. Nelson, K. E., Perkins, D., & Lepper, M. (2003). Accelerating learning: Why significant learning is rare and what to do about it. Unpublished manuscript. Penn State University, University Park, PA. Nelson, K. E., & Welsh, J. A. (1998). Progress in multiple language domains by deaf children and hearing children: Discussions with a Rare Event Transactional Model. In R. Paul (Ed.), The speech/language connection (pp. 179–225). Baltimore: Brookes. Nelson, K. E., Welsh, J., Camarata, S. M., Butkovsky, L., & Camarata, M. (1995). Available input and available language learning mechanisms for specifically language-delayed and language-normal children. First Language, 15, 1–17. Nelson, K. E., Welsh, J., Camarata, S., Heimann, & Tjus, T. (2001). A rare event transactional dynamic model of tricky mix conditions contributing to language acquisition and varied communicative delays. In K. E. Nelson, A. Koc, & C. Johnson (Eds.), Children’s language (Vol. 11, pp. 165–196). Hillsdale, NJ: Lawrence Erlbaum Associates.
220
NELSON ET AL.
Newby, K. (1994). The relationship between social valence and target acquisition in two types of language intervention. Honors thesis, Penn State University, University Park, PA. Nicely, P., Tamis-Lemonda, C. S., & Bornstein, M. H. (1999). Mothers’ attuned responses to infant affect expressivity promote earlier achievement of language milestones. Infant Behavior and Development, 22, 557–568. Nichols, H., Lightbown, P. M, & Spada, N. (2001) Recasts as feedback to language learners. Language Learning, 51, 719–758. Ninkovic, M., & Nelson, K. E. (2001, October). Accelerating development in art, language, and creativity. Paper presented to the Cognitive Development Society, Virginia Beach, VA. Nittrouer, S. Manning, C., & Meyer, G. (1993). The perceptual weighting of acoustic cues changes with linguistic experience. Journal of the Acoustical Society of America, 94, S1865. Obler, L., & Menn, L. (1988). The exceptional brain: Neuropsychology of talent and exceptional abilities. New York: Guilford. Paley, N. (1995). Finding art’s place. New York: Routledge Peltzer-Karpf, A., & Zangl, R. (2001). Figure-ground segregation in visual and linguistic development: A dynamic systems account. In K. E. Nelson, A. Koc, & C. Johnson (Eds.), Children’s language (Vol. 11, pp. 197–226). Hillsdale, NJ: Lawrence Erlbaum Associates. Pemberton, E., & Nelson, K. E. (1987). Using interactive graphic challenges to foster young children’s drawing ability. Visual Arts Research, 13, 29–41. Perkins, D. N. (1993). Person-plus: a distributed view of thinking and learning. In G. Salomon (Ed.), Distributed cognitions (pp. 88–110). Cambridge, England: Cambridge University Press. Peterson, C., & McCabe, A. (1991). Linking children’s connective use and narrative macrostructure. In A. McCabe & C. Peterson (Eds.), Developing narrative structure (pp. 29–54). Hillsdale, NJ; Lawrence Erlbaum Associates. Peterson, C., Jesso, J. B., & McCabe, A. (1999). Encouraging narrative in preschoolers: an intervention study. Journal of Child Language, 26, 49–67. Pinker, S. (1994). The language instinct. Cambridge, MA: MIT Press.. Pressman, L. J., Pipp-Siegel, S., Yoshinaga-Itano, C., Jubicek, L. F., & Emde, R. N. (1998, April). The relation of hearing status and emotional availability to child language gain. Paper presented to the International Conference on Infant Studies. Atlanta. Prinz, P. M., & Prinz, E. A. (1981). Acquisition of ASL and spoken English by a hearing child of a deaf mother and a hearing father: Phase II, early combinatorial patterns. Sign Language Studies, 30, 78–88. Prinz, P. M., & Strong, M. (1998). ASL Proficiency and English literacy within a bilingual deaf education model of instruction. Topics in Language Disorders, 18, 47–60. Records, N., Tomblin, J., & Freese, P. (1992). The quality of life of young adults with histories of specific language impairment. American Journal of Speech-Language Pathology, 1, 44–53. Robertson, S. B., & Weismer, S. E. (1999). Effects of treatment on linguistic and social skills in toddlers with delayed language development. Journal of Speech, Language, and Hearing Research, 42, 1234–1248. Sacks, O. (1995). An anthropologist on Mars. New York: Knopf
9. A DYNAMIC TRICKY MIX PERSPECTIVE
221
Sameroff, A., & Fiese, B. (1989). Transactional regulation and early intervention. In S. J. Meisels & J. Shonkoff (Eds.), Early intervention: A handbook of theory, practice, and analysis (pp. 187–244). New York: Cambridge University Press. Sinclair, J. (1968). Developmental psycholinguistics. In D. Elkind & J. H. Flavell (Eds.), Studies in cognitive development. New York: Oxford University Press. Slobin, D. I. (1985). Crosslinguistic evidence for the language-making capacity. In D. I. Slobin (Ed.), The crosslinguistic study of language acquisition (pp. 1157–1256). Hillsdale, NJ: Lawrence Erlbaum Associates. Sloboda, J. (1996). The acquisition of musical performance expertise: Deconstructing the “Talent” account of individual difference in musical expressivity. In K. A. Ericsson (Ed.), The road to excellence (pp. 107–126). Mahwah, NJ: Lawrence Erlbaum Associates. Slomkowski, C. L., Nelson, K., Dunn, J., & Plomin, R. (1992). Temperament and language: Relations from toddlerhood to middle childhood. Developmental Psychology, 28, 1090–1095. Snow, C. E., & Dickinson, D. K.(1990). Social sources of narrative skills at home and at school. First Language, 10, 87–103. Snow, C. E., Perlmann, R., & Nathan, D. (1987). Why routines are different: Toward a multiple-factors model of the relation between input and language acquisition. In K. E. Nelson (Ed.), Children’s language (Vol. 6, pp. 65–98). Hillsdale, NJ: Lawrence Erlbaum Associates. Soderbergh, R. (1977). Reading in early childhood: A linguistic study of a preschool child’s gradual acquisition of reading ability. Washington, DC: Georgetown University Press. Stringer, C., & McKie, R. (1996). African exodus: The origins of modern humanity. New York: Holt. Strohner, H., & Nelson, K. E. (1974). The young child’s development of sentence comprehension: Influence of event probability, nonverbal context, syntactic form, and strategies. Child Development, 45, 567–576. Strothard, S. E., Snowling, M. J., Bishop, D. V. M., Chipchase, B. B., & Kaplan, C. A. (1998). Language-impaired preschoolers: A follow-up into adolescence. Journal of Speech and Hearing Research, 40, 407–418. Tattersall, I. (1999). The last Neanderthal: The rise, success and mysterious extinction of our closest human relatives. New York: Nevraumont. Thelen, E., & Smith, L. B. (1994). A dynamic systems approach to the development of cognition and action. Cambridge, MA: MIT Press. Thomas, G., & Silk, A. M. J. (1990). An introduction to the psychology of children’s drawings. New York: Harvester Wheatsheaf. Tjus, T., Heimann, M., & Nelson, K.E. (1998). Gains in literacy through the use of a specially developed multimedia computer strategy: Positive findings from 13 children with autism. Autism: The International Journal of Research and Practice, 2, 139–156. Tjus, T., Heimann, M., & Nelson, K.E. (2001). Interaction patterns between children and their teachers when using a specific multimedia and communication strategy: Observations from children with autism and mixed intellectual disabilities. Autism, 5, 175–187. Tomasello, M. (1995). Pragmatic contexts for early verb learning. In M. Tomasello & W. E. Merriman (Eds. ), Beyond names for things: Children’s acquisition of verbs (pp. 115–146). Hillsdale, NJ: Lawrence Erlbaum Associates.
222
NELSON ET AL.
Tomasello, M., Akhtar, N., Dodson, K., & Rekau, L. (1997). Differential productivity in young children’s use of nouns and verbs. Journal of Child Language, 24, 373–387. Torrance, E. P. & Ball, O. (1984). Torrance tests of creative thinking. Bensenville, IL: Scholastic Testing Service. von Tetchner, S.., & Marinsen, H. (2000). Introduction to augmentative and alternative communication (2nd ed). London: Whurr. Weiner, J. (1994). The beak of the finch. New York: Knopf. Weizman, Z. O., & Snow, C. E. (2001). Lexical input as related to children’s vocabulary acquisition: Effects of sophisticated exposure and support for meaning. Developmental Psychology, 37, 265–279. Wilbur, R. (1987). American Sign Language; Linguistic and applied dimensions (2nd ed.). Austin, TX: Pro-Ed. Wilson, B., & Wilson, M. (1982). Teaching children to draw. Englewood Cliffs, NJ: Prentice-Hall. Wiltshire, S. (1991). Floating cities. New York: Summit Books.
10 Literacy and the Mediated Mind Bruce D. Homer New York University
The purpose of the science of cognitive psychology and cognitive development … is to study how human infants, children, and adults meet, master, and use cultural knowledge and cultural forms. —Katherine Nelson, 1996, p. 23.
Throughout her career, Katherine Nelson has attempted to expand the scope of developmental psychology beyond research on the isolated mind of the individual child. For example, her early work on language acquisition (e.g., Nelson, 1973) stressed the importance of function and social context in children’s early concept formation. Nelson’s later work, exemplified by her 1996 book, Language in Cognitive Development: The Emergence of the Mediated Mind, focused on the role of culture in cognitive development. In this book, Nelson argued that cognitive development is a process involving a system of different strands—including biology, psychology, and social processes—that contribute to the developing mind. Fundamental to this characterization of development is the process of cultural mediation. Nelson argued that from the very outset, children’s knowledge of the world is mediated by virtue of being situated in a cultural environment. Culture mediates de223
224
HOMER
velopment in two main ways. First, children’s specific experiences in the world are determined by their social and cultural environment. Second, it is from culture that children acquire the representational systems that ultimately become the “medium, mediator and tools” of thought. As the title of her book suggests, Nelson’s research focused primarily on the mediating effects of language in cognitive development. Although Nelson did claim that language is the most important representational system, she also put forth that other representational systems, such as writing and mathematics, play crucial roles in children’s cognitive development. This chapter provides a brief summary of Nelson’s developmental theory, focusing on the mediating role of language in the developmental process. It then explores how literacy, similar to spoken language, mediates children’s cognitive development.
NELSON’S EXPERIENTIAL VIEW OF COGNITIVE DEVELOPMENT Nelson (1996) described her theory as experiential in nature. Central to the experiential view is the notion that cognitive development is driven by children’s experiences in the world. Experience is also essential in the foundational theories of both Piaget (1929) and Vygotsky (1978), however, these theorists differ in exactly which experiences they stress as being central. According to Piaget’s cognitivist approach, it is experience at the individual level that is the most critical for development: Children’s interactions with their physical environments provide them with the experiences that they require in order to form and modify their mental representations of the world. For Piaget, cognitive development is characterized as the systematic and progressive modification of these mental representations. In contrast, according to Vygotsky’s sociocultural approach, it is experience in the social realm that is the most critical for cognitive development: Through interpersonal interactions, children encounter and learn to use cultural tools and knowledge. For Vygotsky, cognitive development is characterized as this acquisition of cultural tools and knowledge. According to Nelson (1996), cognitive development can only be fully understood by examining children’s experiences at both the personal and interpersonal levels. To do this, Nelson drew on both Piagetian and Vygotskian theories, but went beyond a simple “interactionism.” Similar to Piaget, Nelson characterized the developing child as an acting and interacting individual who builds representations of the world. Children’s representations are constructed based on their experiences in the world, according to principles that are “built in” to the cognitive system. It is not just the physical world that children are attempting to understand, but also (or even more so) the social world. Nelson (1996) stated that “the primary cognitive task of the human child is to make sense of his or her situated place in the world in order to take a skillful part in its activities” (p. 5). This emphasis on activi-
10. LITERACY AND THE MEDIATED MIND
225
ties has its theoretical roots in Vygotskian theory. Nelson merged the two foundational theories by maintaining the individual mind while at the same time stressing the crucial role of the social environment. Similar to Piaget, Nelson posited qualitative changes in the type of mental representations and activities that children are capable of throughout development. However, whereas Piaget theorized that changes in the representational systems are predetermined and endogenous to the child, Nelson suggested that the major transitions in cognitive development are associated with children’s acquisition of cultural representational systems, most importantly language. This, of course, is similar to Vygotsky’s claims that language and other cultural tools are central to children’s cognitive development. The difference, however, is that Nelson maintains an emphasis on the individual mind, albeit a mind inextricably linked to culture.
STAGES OF DEVELOPMENT Nelson’s (1996) theory of cognitive development has been influenced by the phylogenetic theory of Donald (1991). Donald characterized the human mind as a “hybrid” of biology and culture. He suggested that technological and cultural advances in conjunction with biological evolution created qualitative shifts in human cognition. According to Donald, there have been four distinct phases in the evolution of human culture, and consequently, of the human mind. Associated with each transition is a corresponding development of a new mode of mental representation. Although Nelson did not suggest that there is a simple recapitulation of these phases in ontogeny, she did suggest that there are some interesting parallels.
Donald’s Stages in the Evolution of the Human Mind Donald (1991) identified the most basic phase of culture as episodic. In episodic culture, the mind is only able to represent the world through event perceptions, which means that events are the “atoms” of existence. In episodic culture, there is no representation—thought is always in the present. The next phase is that of mimetic culture. It was through the addition of mimetic skills, which Donald defined as the ability to reenact or represent an event through action, that the transition to this second phase occurred. One of the most important implications of the addition of mimetic representations is that, in contrast to episodic representations, which are tied to a specific event at a particular place and time, mimetic representations are liberated from temporal constraints. This allows mimetic representations to serve a dual purpose: Similar to language, they are used both for individual cognition and for social communication. The next cultural transition came about through the development of language, which led to the emergence of narrative culture. Donald (1991) postulated that language was a public, collective invention. It was the social communication made
226
HOMER
possible by mimetic abilities that allowed language to be developed. In addition to cultural changes brought about by its communicative function, language also altered the nature of mental representations by allowing for the formation of verbally encoded mental models. According to Donald, lower level representational functions are not lost with the development of new representational abilities, but instead they become “encapsulated” by the more advanced representations. Encapsulation means that the lower level cognitive structures continue to function, but their output can be represented in the higher level representations. With language, therefore, it was possible for mental models to be formed that synthesized across time and experience. Through language, people began to formulate verbally encoded mental models in the form of myths and stories (i.e., narratives) that were used to understand the world. This means that individuals’ conceptualization of the world is closely linked to language, particularly in the form of stories. This lead Donald (1991) to claim that “thoughts and word cannot be dissociated from one another” (p. 253). According to Donald’s (1991) phylogenetic theory, the final cognitive transition leads to theoretic culture. This last transition is brought about through the invention of visual symbolic systems that, Donald argued, allowed for the development of two phenomena that radically altered human culture: external memory and theory construction. Donald pointed out that whereas oral cultures rely almost exclusively on individual biological memory, cultures with visual symbolic systems rely much more on external memory devices (e.g., books, maps, diagrams, etc.). He suggested that this reliance on external storage devices must “fundamentally change” the architecture and capabilities of the human mind. The most important of these changes is the development of what has been called analytic or paradigmatic thought . Donald claimed that the advent of analytic thought led to new products that are typically nonexistent in cultures without graphic symbols, the most advanced of which is the formal theory. In theoretic culture, formal theories become the dominant method of predicting and explaining the world.
Nelson’s Ontogenetic Stages of Development Nelson (1996) did not suggest that ontological development involves an exact recapitulation of the phylogenetic stages proposed by Donald. She acknowledged the obvious problems with the analogy, including the fact that children are developing within an existing culture that supports their development and provides a predetermined end-goal for children’s development (i.e., the acquisition of the specific skills and knowledge required to fully participate in that particular culture). However, Nelson did suggest that Donald’s theory can provide an “interesting starting point” for understanding the importance of cultural acquisition in children’s cognitive development. For example, Nelson (1996) indicated that several of the themes put forth by Donald (1991) in his explanation of the evolution of
10. LITERACY AND THE MEDIATED MIND
227
human cognition are also key for understanding the development of children’s cognition. Perhaps most important is Donald’s claim that stages in the evolution of human cognition were due to changes in representational ability. Nelson argued that Donald’s focus on the importance of representational ability and mental model building has important developmental implications. She suggested that the developing child can be characterized as “moving through a series of representational potentials” (Nelson, 1996, p. 75) that correspond roughly to the transitions that Donald proposed for human evolution. Drawing on Donald’s (1991) phylogenetic stages, Nelson (1996) identified similar stages in the development of children’s cognition. In Nelson’s model of cognitive development, transitions between stages correspond to children acquiring new modes of representation from their culture. She argued that infants begin life with the ability to form event representations, an ability that infants have in common with other primate minds and that corresponds to episodic culture. During the second year of life, children learn to form mimetic representations. For children, this involves play, imitation, and intentional communication through movement and basic language. A major shift is seen around age 4 when, for the first time, children begin to use language in a truly symbolic manner. This allows children to use linguistic representations both for communication with others and for their own cognition. At this point, language begins to shape and form children’s cognitive functioning. The final transition comes about in early adolescence as children acquire external representational systems, such as literacy. It is through the acquisition of these external representations, that children acquire the ability to engage in more abstract, “scientific” cognition. In her work, Nelson focused primarily on the transitions in early childhood that are associated with language acquisition. More specifically, she examined the interplay between language, social functioning, and cognition. (For recent reviews of this work, see Nelson, in press, and Tomasello, 2002.) Nelson (1996) argued that during the preschool years, language is the major catalyst of developmental change. During this period there is a progression from an unmediated to a mediated mind, and language is the medium through which the mediated mind is created. To understand exactly how language mediates cognitive development, Nelson (1996, in press) asserted that there is a need to understand the functions of language. Drawing on the philosophical and psycholinguistic literatures, Nelson identified several interrelated functions of language that are responsible for the emergence of the mediated mind.
LANGUAGE AND THE MEDIATED MIND In the broadest sense, it is the capacity of language to represent knowledge that is responsible for its mediating effects on cognition. The representational properties of language initially effect the external, communicative functions of children: Lan-
228
HOMER
guage allows for the sharing of complex information between individual minds. Eventually, however, language also comes to serve internal, cognitive functions. Ultimately, people do much of their thinking in language. The mediated mind comes about because of both the communicative and cognitive functions of language. By sharing thoughts and ideas between people through language, knowledge can extend far beyond individual experiences. Through linguistic communication, individuals can benefit from the experiences of others. For children, with the acquisition of language comes access to the vast array of knowledge and information that is part of their culture. This means that children do not have to touch the hot stove for themselves; rather, they can be told by Mother, “Don’t touch, hot!” It also means that children do not have to rediscover for themselves that vast store of knowledge accumulated by their culture. Through language, children can learn the information that is important for surviving in their culture. They can be told what is good to eat, when is the best time for catching fish, how to order a meal at McDonald’s, and so on. Linguistic communication is the primary way through which culture “enters the mind” of the child (Nelson, 1996). The ability of language to communicate thoughts allows for thinking to become a collaborative activity. With language, ideas can be placed into the “public space.” This means that not only can ideas be communicated from one individual to another, but they can also be jointly constructed and modified by a “community of minds.” With children, this process is used to shape and modify their concepts. In her theory of language acquisition, Nelson (1996) argued that young children’s concepts can be quite different from adult concepts, even though they may be using the same words. Gradually, through linguistic interactions, children’s concepts are shaped to become similar to adult concepts. Part of children’s language learning, therefore, is learning how their culture represents the world. Another important function of language is what Nelson (1996) called its “stabilizing” effect. Thought expressed in language must be segmented into discrete units (i.e., morphemes or words) and organized into a linear form that conforms to syntactic and grammatical rules. This has two main consequences. First, only certain aspects of a situation can be encoded and represented by language. Linguistic encoding therefore forces individuals to “make choices about what aspects to represent” (Nelson, 1996, p. 122). Second, there is a structure imposed by language onto thought. Whereas thought is often rapid and fleeting, when represented in language, it is slowed down and therefore becomes more explicit, conscious, and objective. According to Dennett (1994), language creates a new class of “objects-to-contemplate.” Through words, individuals can recall and contemplate experiences from another place and time. In fact, language enables individuals to formulate ideas and concepts that could not be derived from unmediated direct experience. Abstract concepts such as “justice” and “morality” are not possible without the representational capacity of language. Certain concepts, such as temporal units (e.g., hours, days, years), exist only because they are socially constructed through
10. LITERACY AND THE MEDIATED MIND
229
language (Nelson, 1996). It is through words that concepts can be formed that generalize across place and time. As children acquire language, they are also acquiring these abstract concepts. Finally, language presents children with new tools of thought for planning, problem solving, and evaluation. For example, Vygotsky (1978) identified the primary function of “private speech” as being the regulation of actions. According to Vygotsky, this function first appears in communicative speech (e.g., “Don’t touch!”) and secondly in private speech as the child internalizes this planning and regulating function of speech. Narrative is another example of a “tool of thought” made possible by language. Donald (1991) argued that narrative is the “natural product” of language. There is now considerable evidence of the vital role played by narrative in cognitive development. Much of this work has examined the relation between narrative and autobiographical memory (e.g., Brockmeier & Carbaugh, 2001; Fivush, 2001; Han, Leichtman, & Wang, 1998). Interesting cross-cultural research on this area has been conducted by Wang, Leichtman, and their colleagues. For example, Wang, Leichtman, and Davies (2000) examined stylistic and content differences in the narratives that were constructed when Chinese and American mothers were asked to recall a past event with their 3-year-olds. Of particular interest was the degree to which mothers elaborated on the stories told by their children, a factor shown to have important implications for the development of children in American cultures (e.g., Fivush & Fromhoff, 1997; Leichtman, Pillemer, Wang, Koreishi, & Han, 2000). Wang et al. (2000) found that the American mothers were more likely to supplement their children’s responses with additional information, and tended to ask questions about their children’s personal thoughts and feelings. In contrast, the Chinese mothers were less likely to embellish their children’s stories and were more likely to repeat factual questions. The Chinese mothers tended to focus more on moral rules and behavioral standards rather than on the individual thoughts and feelings of their children. Corresponding cultural differences were found in the 3-year-olds’ contributions to the joint narratives. In a related study, Wang and Leichtman (2000) investigated differences in Chinese and American 6-year-olds’ narratives. Children from both countries were asked to tell a series of stories and recall a number of emotional memories. Wang and Leichtman found that in their narratives, the Chinese children were more likely to emphasize social harmony and moral correctness, whereas the American children were more likely to emphasize their autonomy and sense of self. These findings provide an interesting parallel to the joint narratives in the Wang et al. study with 3-year-olds and suggest that cultural narrative structures help shape autobiographical remembering. (For a review, see Wang & Brockmeier, 2002.) To briefly summarize, there are several functions of language that cause it to mediate children’s cognitive functions. Language provides a means for the transmission of thoughts. It allows people to share ideas, gives access to cultural knowl-
230
HOMER
edge, and provides a means for socially shared cognition. Language also “slows down” and “stabilizes” thought, which makes thought more explicit, conscious, and objective. Through language, abstractions and generalizations can be made, which allows for the formation of ideas that are not possible without language: Language provides new “objects-to-contemplate.” Finally, language allows for the creation of new tools of thought, such as narratives, that alter planning and problem solving. Through these functions, language has a great effect on cognition. However, both Donald (1991) and Nelson (1996) proposed that an additional transformation occurs with the invention (historically) or acquisition (developmentally) of external symbolic systems, such as literacy. The next section adopts Nelson’s functional approach to examine the cognitive effects of the external symbolic system of literacy.
LITERACY AND THE MEDIATED MIND Nelson (1996) claimed that even with the acquisition of language, there are “strong limitations” to cognitive functioning that remain in place until after children acquire culturally designed external symbolic systems. Arguably, the most important of these external symbolic systems is written language. Even though there is almost universal acceptance that literacy is important, there is still considerable controversy regarding the cognitive implications of learning to read and write.
Historical Perspectives on the Cognitive Effects of Literacy Debate on the cognitive consequences of literacy is almost as old as writing itself. In Plato’s Phaedrus, for example, Socrates is critical of writing and considers it inferior to speech. Among Socrates’ critiques is that writing will lead to a deterioration of memory because people will become dependent on text to remember things. Since Plato’s time, however, most theorists have tended to characterize literacy as augmenting cognition. In fact, in the recent past it was often asserted that literacy, in particular the invention of the alphabetic script, is an essential step in the “evolution” of societies. For example, Levy-Bruhl (1923) claimed that writing is essential for the “civilized” mind, and Rousseau (1754–1791/1996) suggested that the different ways of writing correspond to different “stages” in the way that people gathered into nations: whereas “savages” depict objects, and “barbarians” make signs for words and propositions, “civilized” people use an alphabet. More recently, these “great divide” theories of literacy, which claim that writing—particularly alphabetic writing—is essential for advanced thought and subsequently label over one half of humanity as inferior, have fallen out of favor. They have been criticized as being ethnocentric (Scribner & Cole, 1981) and “romantic” (Harris, 1989). These criticisms led Scribner and Cole to argue that literacy is simply a cultural tool that requires the use of certain preexisting cognitive skills. Ac-
10. LITERACY AND THE MEDIATED MIND
231
cording to Scribner and Cole, the only cognitive effects of literacy are due to practice effects, and the increases in other, more general cognitive abilities that have often been attributed to literacy (e.g., logical reasoning) are in fact the result of explicit training in schools. Although few contemporary theorists would suggest that literacy is necessary for having a “civilized” mind, it is still possible that the cognitive ramifications of literacy are greater than simply providing an opportunity to practice certain mental skills. Certainly Donald (1991) suggested that literacy and other external symbolic systems have significant cognitive ramifications. According to Donald, changes in representational abilities were responsible for transitions into new levels of human cognition and culture. The transition from narrative culture (i.e., language) to theoretic culture (i.e., literacy and other external symbolic systems) must therefore have its roots in the differences between speech-based and text-based representation. The most general difference between text and speech is that they exploit different modalities: Whereas speech is aural, writing is visual. McLuhan (1962) thought that this shift in modalities alone could alter cognition. He suggested that as cultures moved from being primarily aural (i.e., speech oriented) to primarily visual (i.e., text oriented), the balance of the brain was changed. Literate humans, he claimed, traded an “ear for an eye” and in so doing, altered the way they perceive and think about the world. There has been little empirical support for this modality theory of literacy. A more promising approach has been based on the fact that literacy is a different medium with different properties, such as the relative permanence of textual representations compared to speech. The cognitive effects of literacy may have arisen through learning to use this new medium, and through exploiting the new genres of communication that literacy allows . Donald (1991) argued that the most important function of literacy (and other external representational systems) is to provide a source of “external memory.” By external memory, Donald made an exact analogy to biological memory, suggesting that memory capacity is expanded by being able to store and retrieve knowledge through writing. Although it serves the same basic function as biological memory, there are some definite advantages to external memory. By using external symbolic systems, memory storage becomes practically limitless in its capacity. Because a physical artifact is created, external memory also becomes much more robust and precise than biological memory. According to Donald, the proliferation of external symbolic systems and increased reliance on external memory led to the mind becoming a hybrid of biology and technology. In modern, literate cultures, while engaging in complex thought, humans “virtually always employ external symbolic material … as their true ‘working memory’” (Donald, 1991, p. 329). Donald further postulated that it is the increased capacity provided by external memory that led to the development of analytic thought, the most important aspect of which is the formal theory. Olson (1994) discussed the ways in which the development of writing systems transformed human cognition. There are two main claims to Olson’s theory. The
232
HOMER
first is that literacy, especially reading, provided a model for the way in which people interpret the world. As evidence of this claim, Olson referred to the British empiricists, such as Bacon and Boyle, who were influenced by theories of reading. They believed that “the Book of Nature” should be read in the same way that text should be read for meaning: The text is available for all to read, but meaning must be interpreted, and similarly, “facts” can be observed by all, but “causes” must be inferred from the facts. The second claim of Olson’s theory is that writing provides a model for understanding the structure of language. He suggested becoming aware of the particular aspects of speech represented or codified by a culture’s script. For example, the advent of writing systems that represented verbal form rather than meaning allowed for the differentiation between what is said and what is meant: What is said are the words written on the paper, whereas what is meant is how those words are to be taken. With the development of each script, language is thought of in terms of what is represented in that script. For example, with syllabic scripts like Vai, language is thought of in terms of syllables; with alphabetic scripts like Greek or English, language is thought of in terms of phoneme-like sound units that correspond to letters. Olson (1991, 1994) also made the point that writing allowed for the development of specialized modes of communication, such as the expository essay. Through developing and learning to use these specialized forms, a unique set of intellectual competencies arose. The medium in which people store and organize information alters how they ultimately understand this information. Although Olson focused primarily on cultural and historical implications of the model theory, he did suggest that children undergo a similar process in development.
Developmental Perspectives on the Cognitive Effects of Literacy How does literacy mediate the cognitive development of children? For a number of reasons, this is not an easy question to answer. In her discussion of the role of language in cognitive development, Nelson (1996) acknowledged that children are “never truly prelinguistic” (p. 329). By this she meant not only that the human brain has evolved with special capacities for processing speech sounds, but also that from birth on, children are exposed to talk. Infants are born into a linguistic culture and cannot help but be affected by it even before they fully acquire language. A similar claim can be made in regard to literacy. In fact, when trying to identify the cognitive effects of literacy, it is even more problematic to refer to the “preliterate” child. Well before they themselves are able to read and write, children are exposed to literacy and to literate modes of discourse, which must affect their cognitive development. In a historic context, Illich (1991) identified a similar issue when he argued for the acknowledgment of “lay literates.” By this term, Illich was referring to the illiterate laity of the 12th century whose “mental space,” he argued, had been altered through exposure to the literate clergy. Similarly, the
10. LITERACY AND THE MEDIATED MIND
233
“preliterate” child’s mind has been affected by virtue of being born into a literate society. Certain concepts and modes of discourse that may be a product of literacy can be passed on to preliterate children through spoken language. Nonetheless, it is reasonable to assume that there are some observable effects of literacy during children’s cognitive development. A useful strategy for identifying these cognitive effects is to adopt the functional approach used by Nelson (1996) in her examination of the cognitive effects of language. In general, the effects that Nelson identified were related to the way in which language represents and allows for the communication of information. As children acquired linguistic capacities, their cognition is altered. Similarly, children’s cognition will be altered as they acquire various functions of literacy. The primary function of external symbolic systems is to serve as external memory (Donald, 1991). It would be expected, therefore, that with the acquisition of this function, the workings of memory will be altered. In a series of experiments, Eskritt, Lee, and Donald (2001) examined university students’ use of external symbols to aid their performance in a memory task. In the task, participants were shown an array of cards and asked to remember matching pairs. The participants were allowed to make notes to help them remember. The cards were then turned face down and the participants were asked to find the matching pairs of cards. Eskritt et al. found that when participants’ notes were removed, they performed worse on the task than participants who had not been allowed to make notes at all. This finding suggests that adults do use external symbolic representations as “external memory” (and partially confirms Plato’s original concern about reliance on text to remember things). In a follow-up study with children performing a similar memory task, Eskritt and Lee (2002) found that it was not until grade 5 (an age by which most children are literate) that children were able to produce notes that could be used to aid their memories. Taken together, the studies support Donald’s claim regarding the integration of external representation into memory functioning, and suggest that literacy does mediate memory. Arguably, the most important function of language identified by Nelson (1996) is to bridge the gap between individual minds. Through speech, thoughts can be communicated from one person to another. Speech has a limitation, however. Barring the use of modern audio and telecommunication technologies, communication through speech must occur immediately with someone who is within hearing range. In contrast, literacy provides a more lasting record of the thoughts being communicated. This means that communication through writing can not only bridge the gap between people, but also the gaps between space and time. Ideas and thoughts that are written down can be communicated to someone who is not physically present—who may not even be born yet. This “permanency” of written communication affects the way that ideas are placed into “public spaces.” With speech, ideas can be shared and socially constructed. With text, the sphere of the “public space” goes beyond the immediate social environment. Someone can read Piaget’s writing and respond directly to his ideas, adding their own thoughts. It is
234
HOMER
through writing that people are able to “stand on the shoulders of giants” to create new knowledge. This means that children can acquire knowledge not only from people in their immediate environments, but also from the past and from other places. The function of schooling is largely to teach children how to use and organize knowledge that is stored in text and other modes of external symbolic storage. Another important ramification of the greater permanency of written text is that it allows individuals to revisit their own thoughts. In part, this is what Donald (1991) suggested when he referred to external memory, but the implications are greater than just providing a more stable version of biological memory. Through writing, individuals are able to examine their own thoughts from a third-person perspective: They can be both author and audience. This allows them to edit their words and to revise their ideas. It also allowed for the invention of new modes of communication, such as the essay and scientific report. These new genres of communication are tools of thought and mediate cognition (Olson, 1994). It would be expected, therefore, that children’s cognition would be affected as they became fluent in certain literate genres. Some empirical support for this thesis comes from research by Keys (1994, 1995), who examined the effects of learning to write laboratory reports on children’s scientific reasoning skills. Scientific reasoning is quite difficult for children—even undergraduate students often have difficulty differentiating between theory and evidence . In Key’s study, the scientific reasoning skills of ninth-grade students were found to improve significantly as a result of their experiences learning to write a proper lab report. Another function of language that has parallels in writing is “stabilization” of thought. Nelson (1994) argued that language “stabilizes” thought by forcing it to be segmented into discrete units and organized into a linear form. Nelson and others suggested that it is because of this stabilization that thought enters into conscious awareness. With literacy, there is a further stabilizing of thought, but there is also a stabilization of language. This is what Olson (1994) is suggesting with his script-as-model theory. As already mentioned, Olson theorized that writing provides a model for reflecting on spoken language. Writing is responsible for bringing aspects of language into consciousness by providing the set of categories that is used to reflect on and analyze speech. According to Olson (1994), as children become literate, they are also learning how to think about language: “Learning to read is learning to hear speech in a new way” (p. 85). There is a growing body of research to support the claim that literacy mediates the development of metalinguistic awareness (for a recent review of this literature, see Homer, 2002). Read, Zhang, Nie, and Ding (1986), for example, found a link between phonemic awareness and alphabetic writing. In their study, Read at al. asked Chinese adults to add and delete consonants in spoken Chinese words. (An English example of this is saying “fish” without the “/f/.”) Some of the participants had been taught to read and write only in Chinese characters, and others had also learned the alphabetic script, Pinyin. Read et al. found that only the participants with prior exposure to the alphabetic script were able to segment words into pho-
10. LITERACY AND THE MEDIATED MIND
235
nemes. This was found even for participants whose exposure to Pinyin had occurred many years previous to the study and who could no longer read or write in Pinyin. Further support for the link between alphabetic and phonological knowledge comes from work by Vernon (1993), who found that only children who have begun to relate letters to subsyllabic units (i.e., phonemes) are able to succeed on a phoneme deletion task. Homer and Olson (1999), testing the script-as-model hypothesis, examined the role of literacy in children’s development of the metalinguistic concept of “word.” Children, from age 4 to 6, were given a metalinguistic awareness task in which they were required to segment speech into words. In this task, the children were told a story, during which the experimenter would occasionally pause and ask the children to repeat the last word that they had just heard. The children also received a number of text-based tasks that assessed their understanding of “word” as a unit of text. For example, the children were shown and read a piece of text and then asked to circle one of the words in the text. The results of the study indicated that children who were able to “repeat the last word” in the speech-based task, understood “word” as a unit of text. This finding suggests that children’s metalinguistic understanding of word develops as children attempt to relate written text to speech. In a related study (Homer, Xu, Lee, & Olson, 2003), literacy and metalinguistic awareness in 4- to 6-year-old children from two cultures with very different scripts were investigated. English-speaking Canadian and Mandarin-speaking Chinese children were given tasks that assessed their metalinguistic understanding of the concepts of word (the most salient feature of English script) and character/syllable (the most salient feature of Chinese script). (Syllables are the closest equivalent to Chinese characters in English; DeFrancis, 1984.) The children were given speech-based tasks that asked them to segment spoken language into words and syllables/characters, and text-based tasks that asked them to identify words and syllables/characters in writing. As predicted by the literacy-as-model theory, children were first aware of the metalinguistic concept that is the most apparent in their culture’s writing system. In addition, awareness of this concept in speech was found to require an understanding of the concept as a unit of text. Overall, the results supported the hypothesis that literacy contributes significantly to children’s development of metalinguistic awareness. Ongoing research (Homer, Xue, & Xu, 2003) is examining this effect in children enrolled in a Chinese-English bilingual school. Preliminary results from this work suggest that for the bilingual children, there is the same effect of literacy on metalinguistic awareness in each language. However, there is little evidence of any crosslinguistic transfer of metalinguistic awareness. To briefly summarize, there are a number of ways in which literacy mediates cognitive functioning. Some of the effects of literacy, such as the development of new, abstract concepts, occur at the cultural level and can be passed on to children through language. Other effects, however, occur at the individual level and are therefore observable as changes in children’s cognitive functioning that are tied to
236
HOMER
the acquisition of literacy. Evidence of these latter effects of literacy has been found for children in a number of domains, including memory, scientific reasoning, and metalinguistic awareness.
CONCLUSIONS This chapter began with a declaration by Nelson (1996) that the goal of developmental psychology should be to understand how children “meet, master, and use cultural knowledge and tools.” Nelson claimed that by understanding how children acquire and use cultural tools, it will be possible to understand how the mind works. Through Nelson’s own work and the work of others, there is now a rather good indication of the ways in which language enables new modes of representation and communication, and thereby fosters the emergence of the mediated mind. This chapter has attempted to outline some of the ways in which literacy, similar to language, also enables new modes of representation and communication. Although there is a need for more research on effects of literacy on cognitive development, there can be little doubt that literacy plays a vital role in the development of the mediated mind.
REFERENCES Brockmeier, J., & Carbaugh, D. (Eds.). (2001). Narrative and identity: Studies in autobiography, self, and culture. Philadelphia: John Benjamins. Bruner, J. (1986). Actual minds, possible worlds. Cambridge, MA: Harvard University Press. DeFrancis, J. (1984). The Chinese language: Fact and fantasy. Honolulu: University of Hawaii Press. Dennett, D. (1994). Language and intelligence. In J. Khalfa (Ed.), What is intelligence (pp. 161–178). Cambridge, MA: Cambridge University Press. Donald, M. (1991). Origins of the modern mind: Three stages in the evolution of culture and cognition. Cambridge, MA: Harvard University Press. Eskritt, M., & Lee, K. (2002). “Remember where you last saw that card”: Children’s production of external symbols as memory aids. Developmental Psychology, 38(2), 254–266. Eskritt, M., Lee, K., & Donald, M. (2001). The influence of symbolic literacy on memory: Testing Plato’s hypothesis. Canadian Journal of Experimental Psychology, 55(1), 39–50. Fivush, R. (2001). Owning experience: The development of subjective perspective in autobiographical memory. In C. Moore & K. Lemmon (Eds.), The self in time: Developmental perspectives (pp. 35–52). Mahwah, NJ: Lawrence Erlbaum Associates. Fivush, R., & Fromhoff, F. (1997). Narrating and representing experience: Preschooler’s developing autobiographical recounts. In P. Van den Broek, P. A. Bauer & T. Bourg (Eds.), Developmental spans in event comprehension and representation: Bridging fictional and actual events. Mahwah, NJ: Lawrence Erlbaum Associates.
10. LITERACY AND THE MEDIATED MIND
237
Fivush, R., & Haden, C. (1997). Narrating and representing experience: Preschoolers’ developing autobiographical recounts. In P. van den Broek, P. A. Bauer, & T. Bourg (Eds.), Developmental spans in event comprehension and representation: Bridging fictional and acutal events (pp. 169-198). Mahwah, NJ: Lawrence Erlbaum Associates. Han, J. J., Leichtman, M. D., & Wang, Q. (1998). Autobiographical memory in Korean, Chinese, and American children. Developmental Psychology, 34(4), 701–713. Harris, R. (1989). How does writing restructure thought? Language and Communication, 9(2/3), 99–106. Homer, B. D. (2002). Literacy and metalinguistic thought: Development through knowledge construction and cultural mediation. In J. Brockmeier, M. Wang & D. R. Olson (Eds.), Literacy, narrative & culture. (pp. 266–287). Surrey, England: Curzon Press. Homer, B. D., & Olson, D. R. (1999). The role of literacy in children’s concept of word. Written Language and Literacy, 2(1), 113–140. Homer, B. D., Xu, F., Lee, K., & Olson, D. R. (2003). The role of literacy in Canadian and Chinese children’s metalinguistic awareness. Manuscript submitted for publication. Homer, B. D., Xue, Q., & Xu, F. (2003). Literacy and metalinguistic awareness in Chinese-English bilingual children. Manuscript in preparation. Illich, I. (1991). A plea for research on lay literacy. In D. R. Olson & N. Torrance (Eds.), Literacy and orality (pp. 28–46). New York: Cambridge University Press. Keys, C. W. (1994). The development of scientific reasoning skills in conjunction with collaborative writing assignments: An interpretive study of six ninth-grade students. Journal of Research in Science Teaching, 31(9), 1003–1022. Keys, C. W. (1995). An interpretive study of students’ use of scientific reasoning during a collaborative report writing intervention in ninth grade general science. Science Education, 79(4), 415–435. Kuhn, D., Amsel, E., O’Loughlin, M., Schauble, L., Leadbeater, B., & Yotive, W. (1988). The development of scientific thinking skills. San Diego, CA: Academic Press. Leichtman, M. D., Pillemer, D. B., Wang, Q., Koreishi, A., & Han, J. J. (2000). When baby Maisy came to school: The effects of maternal interviews on preschoolers’ event memories. Cognitive Developement, 15(1), 99–114. Levy-Bruhl, L. (1923). Primitive mentality. London: George Allen & Unwin. McLuhan, M. (1962). The Guttenburge galaxy. Toronto: Universtity of Toronto Press. Nelson, K. (1973). Structure and strategy in learning to talk. Monographs of the Society for Research in Child Development, 38(1–2), 136. Nelson, K. (1994). Long-term retention of memory for preverbal experience: Evidence and implications. Memory, 2(4), 467–475. Nelson, K. (1996). Language in cognitive development: Emergence of the mediated mind. New York: Cambridge University Press. Nelson, K. (in press). Cognitive functions of language in early childhood. In B. D. Homer & C. Tamis-LeMonda (Eds.), The development of social cognition and communication. Mahwah, NJ: Lawrence Erlbaum Associates. Nelson, K., & Shaw, L. K. (2002). Developing a socially shared symbolic system. In E. Amsel & J. P. Byrnes (Eds.), Language, literacy, and cognitive development: The development and consequences of symbolic communication (pp. 27–57). Mahwah, NJ: Lawrence Erlbaum Associates.
238
HOMER
Olson, D. R. (1991). Literacy as metalinguistc activity. In D. R. Olson & N. Torrance (Eds.), Literacy and orality (pp. 251–270). New York: Cambridge University Press. Olson, D. R. (1994). The world on paper. New York: Cambridge University Press. Piaget, J. (1929). The child’s conception of the world (J. Tomlinson & A. Tomlinson, Trans.). London: Routledge & Kegan Paul. Read, C. A., Zhang, Y., Nie, H., & Ding, B. (1986). The ability to manipulate speech sounds depends on knowing alphabetic reading. Cognition, 24, 31–44. Rousseau, J.-J. (1966). Essay on the origin of languages. In J. H. Moran & A. Gode (Eds.), On the origin of language: Two essays by Jean-Jacques Rousseau and Johann Gottfried Herder (pp. 5–74). New York: Frederick Unger. (Original work published 1754–1791) Scribner, S., & Cole, M. (1981). The psychology of literacy. Cambridge, MA: Harvard University Press. Tomasello, M. (2002). Things are what they do: Katherine Nelson’s functional approach to language and cognition. Journal of Cognition and Development, 3(1), 5–19. Vernon, S. (1993). Initial sound/letter correspondences in children’s early written productions. Journal of Research in Childhood Education, 8, 12–22. Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes. Cambridge, MA: Harvard University Press. Vygotsky, L. S. (1986). Thought and language (A. Kazulin, Trans.). Cambridge, MA: MIT Press. Wang, Q., & Brockmeier, J. (2002). Autobiographical remembering as cultural practice: Understanding the interplay between memory, self and culture. Culture & Psychology, 8(1), 45–64. Wang, Q., & Leichtman, M. D. (2000). Same beginnings, different stories: A comparison of American and Chinese. Child Development, 71(5), 1329–1346. Wang, Q., Leichtman, M. D., & Davies, K. I. (2000). Sharing memories and telling stories: American and Chinese mothers and mothers and their 3-year-olds. Memory, 8(3), 159–178.
11 Katherine Nelson: Contextual Functionalist Jerome Bruner New York University School of Law
Katherine Nelson started breaking away from the narrow research norms of psychology before ever she knew they existed. She never wanted to become, never became, and still does not want to be the standard psychologist, pinning down one variable at a time at the .001 level, with the constricted results published in some refereed journal of multi-adjectival “psychology.” Temperamentally, she resists and has always resisted hunt-and-peck research, never satisfied until she gets things in context. She never succumbed even when, somewhat inadvertently, she became a graduate student at UCLA, where hunt-and-peck one-variable-at-a-time used to be a plank in the department’s research program. And she has continued to resist it ever since—to the great benefit of the emerging science of psychology. To be at peace with herself and with psychology, she has forever been in search of the contexts that give particular human acts their meaning. Her research questions have always been about the function or role played by some action or thought process in the lives of situated human beings. Her work on human development is deeply in the functionalist tradition, all about function in context. How, for example, do particular acts fit into larger, longer sequences, like those Schank-Abelson (1977) “scenarios” that she put to work in her research in the 1980s? She has been tireless, like Minerva on the hunt, but she has always known one big thing: to catch your quarry, you need to know the setting in which they live. 239
240
BRUNER
If I had to label Katherine, I think I’d label her a “functional contextualist,” or a “contextual functionalist.” But, rather than talking generalities about the quintessential Katherine, I want instead to illustrate my point by reference to just one of her research ventures—one in which I and a few other independent-minded colleagues joined together under her leadership. I have in mind the landmark study of young Emily—“Emmy” for short—the 2-year-old whose rich and quirky monologues we all worked on together for several years—indeed, from 1983 to 1989. Then we published a joint book about “narratives from the crib” (Nelson, 1989). The research group was a small, strong-minded little gang, including Katherine, Carol Fleisher Feldman, Daniel Stern, John Dore, and myself, plus four just finished or about to finish doctoral students (Elena Levy, Julie Gerhardt, Rita Watson, and Joan Lucariello). We sweated together over transcripts of Emmy’s after-lights-out bedtime monologues, trying to make sense not only of Emmy’s rambling monologues, but also of each other’s efforts to do so. All of this took place in a series of informal but frequent seminars during the academic years 1982–1983 and 1983–1984 and in a few pick-up seminars in the years following when we began discussing each other’s write-ups of the work. Several others, who participated but who were not directly involved in the nitty-gritty of data analysis, occasionally joined. These occasional visitors contributed substantive ideas, of course, but mostly they helped establish a kind of “linguistic ideology.” These visitors were a high-powered lot: Alison Gopnick, Bill Hirst, Catherine Snow, and Susan Sugarman, none of whom were reluctant to express their views! Those were lively, hard-working seminars, and a genuine learning experience. Indeed, we soon got to referring to ourselves rather grandly as the “New York Language Acquisition Group,” or NYLAG, pronounced Nylag (rhyming with spy flag). It did not take long for group identity to set in at the meetings on the sixth floor of the flagship building of the Graduate Center of the City University of New York. It was located at the midtown heart of the city on 42nd Street between Fiftth and Sixth Avenues. It overlooked the august and revered New York Public Library, New York’s living testament to the great 19th-century architect, Stanford White, whose building seemed an arm’s length away through Katherine’s office windows. Odd, but that view from the windows kept reminding me (us?) that we were up to something, that we had better get to it. The NYLAG gang did its job, wrote its book, and is now scattered all over the place. It had a powerful effect on, and changed, us all. I sometimes think of the subsequent impact of that work group on the world, with its participants now at 10 different universities in 4 different states, and in 3 different countries. The number of man- and woman-hours that went into the transcription, analysis, discussion, and writing-up of Emmy’s crib talk, although impossible to reckon, must be in the thousands. Katherine brought that group together, kept those strong-minded people focused on a common task, and was our conscience as well as our den mother! Her way of thinking about things left a trace on every one of us, indeed put its print on every one of the many hours spent on the job. It is not that she bullied us
11. KATHERINE NELSON: CONTEXTUAL FUNCTIONALIST
241
into thinking her way (although she did keep everyone to deadlines, mirabile dictu). Rather, her perspective, her way of thinking, was atmospheric and ever-present. Let me say a word about that way of thought, about the atmosphere it created. Here we were, the gang of us, suddenly in possession of a treasury of tapes of a young child’s bedtime monologues, spoken when she was all alone, in bed with the lights out (recorded on cassettes from a tape recorder placed under her bed by her parents). We, the researchers, had the daunting task of making sense of this rich sample of isolated, presumably out-of-context, toddler speech. Indeed, we did not even have firsthand knowledge of what Emmy was presumably talking about, no protocols of what she had been up to during the day that might have precipitated all that interesting chatter. Katherine would occasionally get the parents to clarify odd episodes, but that was it. If ever there was a corpus of speech in need of contextualization, this was it. This was the ultimate challenge for context-hungry Katherine! Indeed, I’ve sometimes suspected that it was precisely the isolated nature of Emmy’s corpus of monologues that impelled Katherine. How is it possible to make sense of those rich protocols? But we are getting ahead of the story. It is revealing how Emmy’s tapes came to be. Katherine explains in the opening paragraphs of the book that “the collection of data for this project was inspired in a particular way by Ruth Weir’s study of her son Anthony’s crib monologues. Weir concentrated on the linguistic forms of Anthony’s talk, but his references to daily events were at least as provocative as his practice of language forms” (Nelson, 1989, p. v; see Weir, 1962). Indeed, what originally led Katherine to get Emmy’s parents to collect those tapes was precisely that they might provide a “condensed summary of the day,” that might allow for a better understanding of how a little kid constructed her memory of those “daily events.” Doubtless, her interest in obtaining such a corpus of monologues was fueled by Katherine’s earlier preoccupation with episodic memory (which interest, by the way, was spiced by her skepticism about standard accounts of infantile amnesia, a subject on which she was still busily publishing when the Emmy study got started). But once Katherine started listening to Emmy’s tapes, her taste for function-and-context began reasserting itself. The first thing that struck her about the tapes, for example, was the difference “between Emily’s long and richly textured monologues and her short and spare contributions to pre-bed dialogues with her parents” (Nelson, 1989, p. vi). She began asking her usual questions. What functions does monologue serve? Is it for language practice, for constructing cognitive representation, for achieving self-regulation, for role playing, or for something else? And are crib dialogues, in her words, “a function looking for a form”? Do they get routinized over time to conform to parental speech, come to have a recipient function as well as an inner one? Questions, questions, questions! In Carol Feldman’s and my first conversation with Katherine about the project, her ratio of interrogatives to declaratives must have been five-to-one! I think that’s what hooked us.
242
BRUNER
Episodic memory may have kindled the Emmy project, but the fire quickly spread. Very soon it raged around the classic questions of the functions and uses of speech—inner and outer, expressive and communicative, integrative and differentiative, representational and transformative. Katherine’s role in those seminars (and in the exchanges of draft chapters that eventually followed) was to keep the full picture in mind. And she can be a very persistent lady! All of us, needless to say, had our own hobby horses to ride (or broncos to bust). Her job—whether she planned it that way, whether it was thrust on her by happenstance, or whether it was her temperament speaking up—was to keep our noisy seminar sessions from becoming captive to any single hobby horse or bucking bronco! I even recall saying to her once in exasperation, “Oh Katherine, stop being so goddamned eclectic.” She was not being eclectic, however, but just mindful of the different contexts in which monologues had to be understood. She even hit on a master stroke for keeping the whole stable of hobby horses galloping. There was in New York at that time a lively, informal organization called the New York Child Language Group (NYCLG). It met semi-annually, and all of us belonged to it. Its membership included sacerdotal Chomskians, dedicated functionalists like Lois Bloom and her students, speech-act theorists, word counters, even a Bloomfieldian or two. There was even one member who could always be relied on to ask a question about how some particular presentation related to the ideas of Max Wertheimer! There was not only variety in those NYCLG semi-annuals, but a high intellectual decibel level. Well, just as the Emmy seminars on 42nd Street were reaching their first contentious crescendo—which hobby horse should be the “real” pace setter—Katherine suggested trying out our various points of view at the December 1983 meeting of the NYCLG. Everyone agreed, and it was a very lively meeting. None of the hobby horses stole the show! Katherine (1989) wrote about the meeting in these revealing words: The response that greeted these papers indicated that the data were of considerable interest to a wide audience, and encouraged us to extend our analysis. Moreover, we realized that although each of us approached the data from a somewhat different viewpoint there were common themes running through our work that could be brought together to shed light on what had come to seem a basic (if not the crucial) question: Why did Emily engage in monologue? (p. viii).
That NYCLG meeting got us seriously to work: How to develop our respective points of view while, at the same time, addressing the many-sided general question of “function?” It was that resolve that made the book what it is, and made it face up to Katherine’s ever-recurring barrage of questions. Not surprising that our book ended with three connected parts—like Caesar’s Gaul, in tres partes divisa est, all under a common rule. The common rule for the book was FUNCTION: What were the uses of monologue? What did it accomplish? How was it structured to fill its functions? Where did it lead? The tripartism
11. KATHERINE NELSON: CONTEXTUAL FUNCTIONALIST
243
of the book was, of course, of Katherine’s inspiration: three hobby horses, but all reigned to the same wagon. The book is worth a revisit with all that in mind. Part I is entitled “Constructing a World”—the master Function. Its three chapters deal with several crucial ways in which monologue serves to fashion or construct reality. Katherine’s opening chapter brings her back to how monologue turns the things of everyday life into a form fit for remembering. Here are the last two sentencs of her chapter: “What I have sketched here, then, is a picture of a child building a mental model of the world that represents reality as she understands it on the basis of her own direct experience of events and the verbally presented representations of her parents. … This model of the world is a continuously evolving knowledge system with remarkable power” (Nelson, 1989, p. 63). Emmy evolves a way to “know” the world in her own light, yet to relate it to the grown-up world represented in the speech of her parents. What better instrument than monologue? The next two chapters provide variations on this same theme. The one by me and Joan Lucariello (and I shall never forget the bouts of relief-producing hilarity triggered by our puzzling quest!) was to explore why Emmy so favored the narrative or storytelling form in her monologues. We came to the conclusion, on the basis of our endless poring over the transcripts, that narrative provided a primitive, possibly innate form for structuring “what in adult parlance, we refer to as action, cognition, and feeling” (Bruner & Lucariello, 1989, p. 75). Narrative, in a word, involved some sort of initial expectation about how things ordinarily are, some upset of this expectation, and an account of efforts to restore (or replace) the original expecatation. We even conjectured that the demands of narrative structuring goaded Emmy into exploring the case grammar needed for realizing the requirements of narrative recounting. In the end, Emmy achieves “a triumph of ordinariness, a kind of routinization in story” (Bruner & Lucariello, 1989, p. 96). [That exercise, I must say, set me off on a new line of work in the years after.] Carol Feldman’s following chapter examines how Emmy’s narratives provide her with a means for posing and domesticating problems: “She develops linguistic procedures for framing matters that she does not know something crucial about” and the principal means for doing so is the story form, a basic instrument for problem posing. It is the child’s route into interpretive thinking. I need not go into as much detail about the two remaining parts of the book, for they carry on with the same mission. Part II is concerned with how the young Emmy constructs the language of narrative. Levy looks at Emmy’s monologue making as a step toward text formation. Gerhardt poses the interesting conjecture that monologue may be a child’s first step on the way toward appreciating genre. And Dore sees Emmy’s dialogue as a “reenvoicement” of dialogue, a way of transforming social speech into a more viable form of inner-directed speech. Finally, part III addresses the intriguing question of the role of monologue in the construction of a Self. There are chapters by Watson on the self-regulative task
244
BRUNER
of relating dialogue to monologue, by Katherine on the role of monologue in extending the Self over time through the recapture of the past, and by Stern on the role of monologue in Emmy’s reworking of potentially threatening major life events. Stern’s attention is pinpointed on three critical events in Emmy’s young life: (a) a change in her bedroom at 21 months in preparation for (b) the birth of her brother at 23 months, and (c) a long plane trip to visit her grandmother at 29 months, and (d) entering nursery school at 31 months. He explores the implications of a point made by Bruner and Lucariello in their chapter that the “engine of narrative” is some sort of trouble or violation of the expected, and relates Emmy’s narratives of those four critical events in her life to Freud’s classic idea about “disequilibrium in psychosomatic systems.” Like the other chapters in the book, Stern’s illustrates the range of psychological issues to which the Emmy project related itself. And it is Katherine we have to thank for that range. That was what she stood for: Let the hobby horses pull together, in a common direction, but let them also show their own paces. That’s what struck me after re-reading the book in one gulp again, cover to cover, the first time since it appeared over a decade ago, and with lively memories of the work-in-progress that started two decades ago. What a tribute to her gift for keeping in mind the richness of the contexts in which we carry out the particularities of our day-to-day lives—and to her gift for gently keeping others in mind of it as well.
REFERENCES Bruner, J., & Lucariello, J. (1989). Monologue as narrative recreation of the worldly. In K. Nelson (Ed.), Narratives from the crib. Cambridge, MA: Harvard University Press. Nelson, K. (1989). Narratives from the crib. Cambridge, MA: Harvard University Press. Schank, R. C., & Abelson, R. P. (1977). Scripts, plans, goals, and understanding. Hillsdale, NJ: Lawrence Erlbaum Associates. Weir, R. (1962). Language in the crib. The Hague, Netherlands: Mouton.
12 The Virtues of Rigorous Interdisciplinarity Merlin Donald Queen’s University
Evolution is one thing, and development is quite another. They involve a different kind of dynamic, and their empirical methodologies are far apart. But, on the other hand, they also have important areas of overlap, especially in the domain of theory. This is due to the fact that both evolution and development use time as an organizational and exploratory principle. Only temporal analysis can reveal the details of emergent structure and dynamic processes. This is especially important in the case of human cognition, which defies reduction to a static model. Cognitive processes unfold on two time scales, the first (development) measured within fractions of a single lifetime, and the other (evolution) measured in multiples of many lifetimes. Evolution is ruled by mechanisms that are distinct from those that govern development; the former act on entire populations, whereas the latter act on individual organisms. However, there are direct linkages between evolution and development. Developmental processes are subject to natural selection, and development must be an integral component of any comprehensive theory of evolution. Evolutionary theories must be compatible with developmental facts, because it is the developing organism in the real world that undergoes evolutionary change. There is a feedback loop from development to evolution. Baldwin (1896) argued that developmental plasticity might acquire fitness value under certain cir245
246
DONALD
cumstances, and become subject to natural selection in its own right. This was the basis of what became known as the “Baldwin Effect.” Baldwin was a Darwinian, concerned with the possibility that the activities of organisms might actually affect their own evolution. He understood that the Lamarckian notion of inheriting acquired characteristics (e.g., the notion that increased exercise of the vocal tract might lead to the evolution of an improved vocal tract) could not work. But he realized that learning might provide a potential feedback loop from development to evolution, because learning allows certain species to modify their evolutionary environments. In the case of human beings, whose main adaptive “environment” is human culture itself, Baldwinian logic seems to apply particularly well. Cultural evolution profits from the collective cognitive power of the group, or network. In the case of hominids, cultural evolution was probably very slow and limited, as it is in primates. It was essentially a secondary byproduct of cognitive evolution, never gathering any momentum as an independent force. But cultural evolution became more important as hominid cognition became more closely interlinked with the challenges that culture provided. As competition between hominids became more important than coping with the environment, culture eventually took the lead in hominid evolution, seeking out and rewarding (with better reproductive success) those genotypes best equipped to handle its accelerating demands. These demands weighed more and more heavily on developmental plasticity, and on abstract capacities and skills that were transmitted and cultivated in communities of mind, or cultures. This amounts to an elaboration of the classic Baldwin Effect. As hominid culture became more complex, cognitive adaptability to this increasingly unpredictable aspect of the environment became more crucial for survival and reproduction. A coevolutionary spiral had been triggered. As hominid culture turned more toward innovation and change, it placed yet more importance on plasticity and flexibility. This spiraling coevolution of culture and cognition might explain the extraordinary expansion of the hominid brain during the past half million years: Hominids were becoming symbiotically dependent on culture, and this produced more selection pressure favoring brains that were well designed for living in close-knit cultural groups. As this process unfolded, the nature of the cultural demands driving the process continued to change, placing new selection pressures on the hominid genotype. This tension culminated in the arrival of modern humans, and the emergence of high-speed linguistic communication. Deacon (2003) deconstructed the Baldwin Effect and proposed a much more biologically detailed version of this idea. Deacon noted the well-documented phenomenon that certain genes may become “redundant” after a major environmental change. Say there is a drastic climate change, and a species suddenly finds that its environment provides fruit, where it formerly did not. It now has a ready supply of Vitamin C, formerly a scarce commodity. Prior to the climate change, the organism had to manufacture its own Vitamin C by means of an innate mechanism, but now the vitamin is provided by the environment, and the need for this innate mechanism has evaporated. Formerly, an absence of these Vitamin C-producing genes
12. THE VIRTUES OF RIGOROUS INTERDISCIPLINARITY
247
would have been very maladaptive, probably fatal. Now, with an abundant supply of Vitamin C available, these genes are “redundant.” As a consequence, they are no longer subject to selection pressure, and there is no constant correction on their precise reproduction. Thus, they become subject to reproduction errors. They wander, and become “unmasked.” Over time, this unmasking process leaves an entire region of the genome open to wider and wider variation. Any trait associated with that region becomes subject to unmasking. Cultural evolution can accelerate this effect, especially in human culture, because it can introduce changes into the environment at a much more rapid rate, and these changes can accumulate, as in the evolution of technology. Cultural innovation can effectively make certain genes redundant, and help unmask areas of the human genome. Such a mechanism could be potentially revolutionary in its genetic effects, changing the adaptive landscape rapidly. This could have changed the hominid genome much faster than, say, random mutations or classical gradualistic evolution. Deacon argued that human language must have originated in this way. The same logic applies to a much wider arena than language: human culture itself. Cultural evolution set the stage for language, and language itself emerged out of the network, that is, out of many minds engaged in a collaborative culture. The brain was adapting to this communicative culture as it evolved. Thus, the cultural environment might have contributed mightily to the final evolutionary push that created modern humans. Human beings are now a “hybrid” species, tethered to biology on the one hand, and to cultural change on the other. However interesting and exciting this idea may be, it still awaits detailed verification and elaboration. But the benefits of considering developmental and evolutionary logic as part of the same theoretical enterprise go far beyond the specific case of the Baldwin Effect and cultural evolution. They are but one illustration of the benefits to be gained from applying interdisciplinary thinking to cognitive science.
THE CONVERGENCE OF ONTOGENETIC AND PHYLOGENETIC FACTORS Katherine Nelson is unusual in having examined in detail both cognitive development and cognitive evolution. Her work has much to offer anyone interested in how the complex relations between evolution and development work together to shape the adult human mind. Nelson set out her agenda with remarkable clarity and candor, examining the convergences of ontogenetic and phylogenetic processes in cognition. This is not to imply that she holds recapitulatory views. She made it very clear from the start that this was not the case. But there are major areas of agreement between developmental theory and the emerging field of human cognitive evolution. The latter may lack the vast empirical enterprise that nourishes developmental theory, but has amassed a comparative database on humans and their closest primate relatives, and can also draw on an enormous amount of func-
248
DONALD
tional anatomical evidence. Nelson (1996) draws on both of these literatures, with considerable success. Language in Cognitive Development (Nelson, 1996) is one of the best examples of the theoretical benefits of interdisciplinary convergence in cognitive science. Interdisciplinarity is not always a good idea. Attempts at putting together ideas drawn from disparate areas of research do not always work well. In fact, too often the label “interdisciplinary” simply implies that a work so labeled lacks any discipline whatever. But done well, and in the right historical context, interdisciplinary convergence is a very powerful theoretical strategy. In the case of human cognition, it may be argued that the human mind is too complicated for any single field to master, and in this case, interdisciplinarity is not a luxury but a necessity. The question is how and where to find a method powerful enough to approach this particular kind of interdisciplinary problem. The mind is a super-complex system. The rules that govern the dynamic self-assembly of the human mind are not going to be solved by simple models or straightforward technological advances, such as more powerful microscopes. Human cognitive systems have all the complexity of living systems, plus the complications added by the need to plug into a cultural network. This implies that cognitive theory must account for more than the individual mind. It must also account for the emergence of the cognitive community itself, which is a complex network. The network, after all, helps form the individual mind, and ultimately generates and preserves cultural representations. The biological side of a human mind may be confined within the boundaries of an organic being, but the collective side of cognition is distributed across many minds, and is supported further by a variety of communication and memory technologies that are subject to cultural evolution and have no fixed form. Humanity’s cognitive “mechanism” is thus both an organic and a distributed process. Science is only beginning to comprehend the scale of the conceptual and methodological challenges involved in the study of dynamic systems of this complexity. Thus, the really difficult research question in this field is: What is the question? Researchers really do not know how to approach the topic, and the dilemma is not technical, but rather theoretical. They are close to having the technical capacity to track every neuron, and label every chemical transaction in the brain simultaneously, and this exposes an Achilles’ heel: Researchers do not know what to ask of this technology. Given such powerful methods of investigation, what should they look for? The answer is not at all clear. However, one thing is clear. The answer will not come from blindly collecting more data, or by automatically applying whatever technology the engineers come up with next. Cognitive theory must develop a strategic approach on its own level, and construct useful hypotheses, not only by following each technological trend as it appears (although this is inevitable), but also by absorbing all the substantive material that might prove relevant and formulating a better theoretical framework. This is how astronomy and biology progressed. At many points in the history of
12. THE VIRTUES OF RIGOROUS INTERDISCIPLINARITY
249
these sciences, theory was far ahead of their empirical databases. It was a vital source of guidance in deciding what kind of evidence to seek. One should not expect to arrive at the cognitive equivalent of a Copernican or Darwinian synthesis in cognitive science after the fact, that is, by collecting data and then deriving a theory from it. That has virtually never been the case. Theory has led the way, making its own breakthroughs, focusing investigation, and enabling empirical researchers to ask the right questions. To achieve this, theory did not follow a narrow path. Integration and convergence have been the order of the day. The basic principles of theoretical convergence are conceptually simple, but difficult to implement effectively. Convergence involves seeking evidence that points to a common conclusion across several disciplines. Thus, for instance, a researcher might look across all possible sources of knowledge about memory in order to build a better theory of memory. Questions of “scientific taste” often come up in doing this. Is it good taste to search too widely, or prematurely, or too narrowly? How do researchers decide where to draw the line? This is admittedly difficult, and hard to judge until after the fact. The most successful applications of convergent evidence are judged ultimately by the canon of parsimony, which demands the simplest and most general solution. The most efficient path to such a solution is to map out any new territory by placing it in the major coordinates of a wider theoretical world, to define the space within which new evidence will have both location and meaning. This principle applies to cognitive theory. When lines of evidence are brought together from several different cognitive research areas and point to the same conclusion, time is saved because theories in each area are prevented from wandering off in directions that contradict evidence from other areas. Theories are thus made stronger and wider. A spatial metaphor might help here. Early mapmakers could infer a great deal by finding convergence points that would link together two databases that were far apart (say, two provinces of ancient Greece). The basic coordinates of early maps were established by fitting together initially disconnected observations from many places, and looking for convergent information that would allow them to be placed on a single map. The same idea applies to nonspatial models of the world. The theory of evolution was constructed by applying the principle of convergence on a very large scale. At first, it relied on rather thin evidence, but the overall theory proved quite robust because it drew together so many disparate bits of information from fields as far apart as geology and anatomy. In the 20th century, the theory was confirmed and extended by its convergence with findings from many other disciplines, including morphology, taxonomy, ecology, geography, mineralogy, cell biology, and finally, genetics. Darwin was trying to simplify extreme complexity, and he succeeded. The domain of cognition involves, if anything, even more complexity. Cognitive theory has to start with Darwinian theory. Mind emanates from living things; and all the complexities of evolutionary biology are present in the problem. But, in the case of
250
DONALD
the human mind, the problem also encompasses the additional complexities of culture, which leads directly, among other things, to the problem of anthropocentrism. Darwin started with a set of specific observations that showed the functional convergences of anatomy, environmental conditions, and specific behavioral traits. He knew that such coincidences were extremely unlikely to occur unless there was a compelling underlying process to explain them. The same principle applies to the human mind. In this case, researchers should be seeking convergences, not only with the human sciences, but with the natural sciences as well. One reason for the robust nature of convergent theoretical syntheses is that they are not paradigm bound. Single research areas within disciplines are often dominated by a few narrow paradigms, and theories based exclusively on such a narrow evidential base do not easily generalize to the larger world of other paradigms. This is particularly true in the field of cognitive research. Classical laboratory studies of, say, short-term memory, phonology, or speech recognition were limited in their applicability to theories of mental development because the paradigms were so particular to each field. They could not easily be extended to the complex world in which children lived and developed, and their methods often overlooked the real problems with which a growing child must deal on a day-to-day basis. This should not be taken as a criticism of specialized research. Specialization is here to stay. Cognitive science is a complex business, and it requires highly trained specialists, because its methodologies are very difficult to master. But it is not only the collection of data that demands methodological rigor and special training. Theory also demands it, and the discipline of theory building does not follow simply from mastering the demands of design and data collection. It follows from mastering the specific skills of the theoretician. These are different skills and require a distinct kind of preparation. In physics, this fact of life was formally recognized long ago; as a result, theoretical physics involves a different kind of training, in a different academic program, from experimental physics. But in cognitive science this distinction is not made, and theoreticians are apparently expected to emerge as if by magic, without any special training. Moreover, in cognitive research, laboratory workers in particular can become so deeply committed to a specific paradigm that they see such “invasions” of their turf as a threat to be resisted, rather than as a positive source of new insight to be embraced. As a result, their ideas may remain vulnerable to the possibility that evidence from another field might eventually invalidate them. A good case in point was Behaviorism’s theoretical collapse in the face of persuasive ethological evidence. For 50 years, the laboratory methods of Behaviorism dominated most North American departments of psychology, and relied on one set of approved paradigms, while rejecting important sources of evidence on the same subject matter, the animal mind, simply because they were drawn from other fields. The result was disastrous and wasteful. Convergent theory building across fields would have worked better. It would have avoided many of the naïve propositions that emanated from that field.
12. THE VIRTUES OF RIGOROUS INTERDISCIPLINARITY
251
Theories constructed from a wider survey of evidence are less likely to fall victim to the same kind of error. Of course, this is true only if theorists take pains to acquire sufficient expertise in all the empirical fields from which an argument is being constructed. But the expression “sufficient expertise” does not imply that theorists must always have a technical mastery of every field they survey. Rather, they must have enough knowledge of the strengths and weaknesses of various methodologies to be able to evaluate properly the claims emanating from a given area. Multidisciplinarity is far from a new idea. It proved its effectiveness as a theoretical strategy long ago, as far back as the 18th century. At that time, theoreticians in a variety of fields drew their evidence from many different sources, and systematically compiled a massive database from which they extracted patterns that were otherwise not in evidence. This was true of the chemical table of elements, and of the first great geological classification systems, which integrated a massive amount of evidence from the study of fossils, rocks, environmental change, climate, mining, and biology. These great scientific achievements involved a vast exercise in theoretical synthesis across bodies of evidence that were obtained with many different methods. The approach was a convergent one, and the result was robust theory, which later led to finer and finer differentiation of both theory and method. This kind of synthesis has always been one of the pillars of science. However, this kind of integrative thinking has become very difficult to practice in the professional environment of experimental psychology. This is partly due to a tremendous increase in the number of paradigms, but also to a certain ideological resistance to psychological theory as an end in its own right. Cognitive research tends to cluster itself into small panels and groups that are self-defined by their methodologies, rather than by their place in a larger theoretical world. Thus, there are “brain imaging scientists,” “human short-term memory researchers,” “theory of mind researchers,” “simulation science,” and so on, defined largely by paradigms. This is fine, as long as theory escapes any attempt to confine it narrowly to the data of any single paradigm. Unfortunately, this is not easy to achieve when researchers are trained as they are. Despite this trend, out of scientific necessity, theory in cognitive science has become more interdisciplinary. The message implicit in Darwin’s early multidisciplinarity is the following: When pursuing the theoretical Big Picture, a quick and effective way to eliminate potential theories and hypotheses is to look for failures of convergence. Use the data drawn from one field as a verification check on the interpretation of data from another. For instance, reject out of hand any psychological theory that does not make physiological or biological sense, and vice versa. And reject any evolutionary interpretation that does not converge with what is known of development. Above all, in the case of living organisms, this caveat implies that science should not ignore how animals (including humans) live in the real world. Researchers should try to account carefully for the impact of the real world on the evolution and development of humanity as a species. This is the normal condition
252
DONALD
of developmental theory, but it is rare elsewhere in cognitive science, because it immerses the theorist in a vast and at times impenetrable morass of data from many fields. But there is no choice. There could be no better rationale for insisting that every cognitive scientist have extensive training in theories of mental development. The alternative is to pursue blindly some line of cognitive research or simulation that may prove, in the long run, to be uninteresting or even misleading. The early history of experimental psychology should have provided cognitive science with many unforgettable examples of this danger, but this seems to have been forgotten. All those thousands of papers in psychophysics and animal behavior, with their equations, esoteric terminology, and aspirations to the precision of particle physics were left mouldering on library shelves because their proponents failed to take into account what science already knew of ethology, physiology, and cognition. A significant segment of modern cognitive research is leaving itself vulnerable to the same disease. Developmental theory is one of the cures. It forces scientists to synthesize across paradigms. It should be a required part of every cognitive science curriculum. But it is not, nor, incredibly, is the theory of evolution.
ACHIEVING BALANCE How does a budding interdisciplinary thinker decide where to draw the line in constructing a synthesis? How much is too much? It is not possible to include absolutely every field that might be relevant to a subject as broad as the nature of the human mind. There are no good written guidelines to aid aspiring theorists, and they must depend largely on good examples. Katherine Nelson’s theoretical work is a canonical example of how to balance the demands of width and depth. Nelson started with a complete mastery of her own field (or fields—her developmental theories represent a significant multidisciplinary synthesis within the subdisciplines of developmental psychology) when she decided to enter the treacherous waters of human evolution. This she did carefully, and with panache. Her key insight was that cognitive processes that were scaffolded during development might have a similar vertical scaffolding during their evolution, and vice versa, even though the underlying mechanisms of change were very different. She showed how the patterns of the developing mind were revealed in its unfolding, and then how those patterns resonated with what is thought to have occurred on the evolutionary time scale. This perspective afforded her an opportunity to find things that were otherwise blurred or hidden in noise. And this exercise paid off. The key to her success was that she had a firm agenda, and applied a rigorous standard of verification. I tried a somewhat similar exercise in my home discipline of neuropsychology, where there was a similar skepticism about the value of integrative theory. In my
12. THE VIRTUES OF RIGOROUS INTERDISCIPLINARITY
253
early studies of case histories documenting the cognitive sequelae of brain injury, I found that the breakdown of the adult mind after brain damage often reveals a confusing mix of cognitive disorders, all simultaneously present (these tend to be the cases that are not mentioned in the literature). It is extremely difficult, if not impossible, to factor out from such a disordered mind the discrete “modules” of the mind solely on the basis of clinical observations and tests. The underlying cognitive organization of the brain is hidden below the observable behavioral surface, because the adult clinical case presents symptoms that are the end result of a very complex developmental (and evolutionary) history. This imposes a frustrating limitation on what can be said about neuropsychological evidence taken in isolation. Theories of modularity proliferated in early neuropsychology, but nobody could agree on which modules were real and which were imaginary. The only escape from this frustrating theoretical trap was to recognize the necessity of going outside the field, examining other evidence on modularity, and looking for convergences between disciplines (Donald, 1991). The same problem was even more acute in cognitive neuroscience. Methodological innovation during the 1970s and 1980s was focused on event-related potentials (ERPs), with a special emphasis on their relation to language and selective attention. As long as researchers in this field stayed close to their data, their results were plagued with ambiguities, due partly to the limited cognitive paradigms that could be used with such recording methods, and partly to the limitations inherent in surface electrical recording. It was extremely difficult to build theories of any depth or breadth based exclusively on such evidence. The only way to reduce the number of plausible theories was to look for convergences between evidence collected with these methods with data from other fields. Given the nature of the questions being asked (e.g., What is the internal architecture of the language brain?), this required not just a synthesis of adjacent areas of neuroscience and cognitive science, but a much more ambitious program seeking a broader synthesis that reached out to biology and anthropology. Surely, the best ruling paradigm for such a broad synthesis should be an evolutionary one. The interdependencies and complexities inherent in the mental structures of human adults could only be understood better in the light of a comprehensive evolutionary scenario of human emergence. Nelson decided to take a similar theoretical course in her own field, for somewhat similar reasons. A qualitatively different perspective on the essentially static paradigms of cognitive neuroscience could be sought by adding evolutionary time as a major variable and narrowing down theoretical alternatives by using convergence as an eliminative device. The longer term objective of such an exercise would be similar to Nelson’s: theoretical unity. Just as physics must seek to make the theory of the very large (relativity) compatible with the theory of the very small (quantum mechanics), so psychology might make the theory of slow-moving cognitive dynamics (evolution) compatible with the theory of fast-moving cognitive dynamics (development).
254
DONALD
THE ORIGINS OF COGNITION IN THE INTERACTIONS OF BRAIN, MIND, AND CULTURE Evolutionary and developmental processes are both part of the same dynamic cognitive system. Many of the most abstract and uniquely human abilities of the adult mind are neither inborn nor automatic, but rather are the result of a hybrid brain-culture developmental process that interweaves the phenotype with cultural evolution. In effect, as cultural evolution has gathered momentum, the balance of power in cognitive development has shifted to culture, and the evolving human brain has tracked the changing cognitive demands imposed by it. Human beings are thus cognitive hybrids, tethered to both biology and cultural environment. Another way of phrasing this is that humans were the first species to evolve a truly “distributed” cognitive system, that is, a system in which thought and memory are carried out in a community of minds. In a network, individuals are joined to a larger cognitive architecture that can have powers (e.g., deep memory resources and diverse expertise) that are not available to single individuals. Networks can also serve as generators of novel and powerful cognitive tools (e.g., languages, instruments, and symbolic notations). Languages in particular are network-level phenomena. They come into existence only on the network level. They are negotiated, like treaties. They are inherently conventional systems that are created only in groups. Language is the most salient and distinctive mental capability of the human species. But in its origins, language is more like an ecosystem than the individual organisms that make up ecosystems. Lexicons and grammars emerge at the group level, not in isolated individuals. They exist in the spaces between people, and regulate their cognitive transactions. This has enormous consequences for the kinds of developmental theories that can reasonably be constructed. If the network comes first in development, then humans must have had networks, or proto-networks, long before they had languages. This observation led to my “culture-first” theory of human cognitive evolution (Donald, 1991). A sophisticated form of high-speed communication, like language, could not have evolved unless some kind of cultural network was already in place. That network was undoubtedly based on unique human communicative skills such as gesture, mime (including role playing), and imitation. These establish the invisible mimetic dimension of human culture, which apes find so difficult to understand, and which still provides a very efficient platform for the development of language. These processes are the precursors of language in children. It is fair to say that the human brain cannot realize its design potential outside of culture. It is designed to serve as a component in a distributed system, rather than to operate solipsistically, as a stand-alone device. This has many implications for the kind of theory researchers should be constructing. If humans had evolved as self-contained creatures capable of solving the world entirely on their own, they would have a very different design from the exiting one. For one thing, the re-
12. THE VIRTUES OF RIGOROUS INTERDISCIPLINARITY
255
sources needed to understand the human world would have had to be built-in. But they are not. Symbolic systems and codes, and the basic habits of rational thought are assimilated from culture. Thus, the developing human mind follows a strategy that is radically different from that of any other species. In order to achieve its mature form, it is dependent on information that is held in something external to both genome and brain: culture. Culture is an integral component of humanity’s evolving distributed cognitive system (Donald, 2001). This vast and complex system encompasses both the human genome and various external sources of replicative information. Culture is equipped with a variety of memory media, and follows its own independent course of evolution. Culture can store and transmit many kinds of specialized programming, and even impose specific cognitive architectures on the brain, such as those of reading, writing, and speaking English. It achieves this through a process of deep enculturation that modifies the developmental process. Culture thus amounts to a massive offline network resource that recreates in the brains of individuals a virtual operating system that they will use in thinking, judging, and remembering all their lives. Human brains have adapted to the presence of this unique enculturation process for hundreds of thousands of years. This explains why human brains are set up the way they are, and why the human species has evolved more epigenetic plasticity, amodal integrative ability, and uncommitted memory capacity, rather than a novel armamentarium of new mental modules. In this context, it is obvious why the human brain is so malleable, to a degree unknown in other species. The developing human brain “assumes” that it will receive basic operating instructions from a culture—any culture. Those will enable it to develop fully-human cognitive capabilities. Of course, culture is unpredictable, but it is only from culture that the mind can acquire a mature adult shape. The adult form of a human mind is highly uncertain, and dependent on the particularities of culture. An adequate model of the human cognitive process will have to explain the entire cultural ecosystem within which the human mind lives and reproduces itself, as well as the internal organismic processes that are joined to the ecosystem. This idea affects the theoretical framework in which the human mind must be seen. The hominid brain did not need to evolve a language module, any more than it needed a logical thinking module, or an artistic intuition module, or a mathematics module. Rather, it needed the things that responded to the cognitive demands of culture: a wider and deeper working memory system, much better attentional skills, and the various metacognitive skills essential for navigating the complexities of culture. The major evolutionary demands imposed by a burgeoning culture insisted on widely applicable, amodal capacities, not rigid encapsulated modules. Human life has much in common with animal life, but the overwhelming influence of culture in development gives the mental lives of human beings an extra dimension. Human sensory and basic perceptual capacities are almost identical to those of apes and monkeys. But culture cannot be read by those mechanisms. Cul-
256
DONALD
ture is invisible to any mind not equipped to decipher its complex codes, and especially to track intentionality in other minds. The capacities that enable human children to decipher such things are broad metacognitive and integrative skills that are normally identified with conscious processing. The remarkable proficiency of human beings in this regard sets them apart from other species. Human children can cope with the invisible “second world” of culture only because they can find, track, locate, and remember aspects of the human social world that completely escape their closest ape cousins. The codes of human culture amount to an encryption device that conceals the meaning of much that people do from other species. What underlies the extraordinary cognitive capacities that enable human children to read these codes so easily? This is the central question that cognitive science must answer. And to answer it, researchers must become more interdisciplinary in their thinking. An ecologically valid ethology of the human mind requires immersion in cultural studies, as much as it demands training in the natural sciences. The complexity and subtlety of the human mind is evident only in its cultures, just as the complexity of many other species is evident only in their natural habitats. To build an adequate model of this process, scientists have to pay close attention to the work of theorists such as Nelson. Developmentalists are perhaps the only cognitive scientists who acknowledge fully the three central facts of the field: the human mind is a dynamic process, whose realization is harnessed to the creative engine of culture, and whose structure is solidly planted in brain physiology. Most other subdisciplines tend to acknowledge only two of these, and some only one. But there will not be a convincing theory until all three are accounted for. This presents a challenge that science has never had to face before, but in facing it, scientists will almost certainly redefine human nature.
REFERENCES Baldwin, J. M. (1896). A new factor in evolution. American Naturalist, 30:441-451, 536-53. Deacon, T. (2003). Multilevel selection in a complex adaptive system: The problem of language origins. B. Weber and D. Depew (Eds.) Evolution and learning: The Baldwin Effect reconsidered. Cambridge, MA: MIT Press. Donald, M. (1991). Origins of the modern mind:Tthree stages in the evolution of culture and cognition. Cambridge, MA: Harvard University Press. Donald, M. (2001). A mind so rare: The evolution of human consciousness. New York: Norton. Nelson, K. (1996). Language in cognitive development: Emergence of the mediated mind. Cambridge, England: Cambridge University Press.
Author Index A Abdi, H., 23, 28 Abelson, R.P., 239, 244 Adams, A., 192, 216 Adams, M.J., 70, 74 Adlam, A., 71, 76 Ahadi, S.A., 104, 125 Akhtar, N., 191, 213, 222 Alcoff, L., 80, 96, 98 Allen, S., 178, 190, 213 Amsel, E., 237 Amsterdam, A., 244 Anderson, A., 155, 172 Andrews, G., 71, 76 Appenzeller, T., 15, 31 Aram, D., 208, 213 Arkenberg, M.E., 8, 189, 219 Aronsson, K., 195, 213 Aslin, R.N., 25, 29 Astington, J.W., 3, 4, 33, 35, 45, 54, 55, 57, 59, 60, 61, 62, 64, 66, 69, 70, 71, 74, 75, 77 B Bachevalier, J., 15, 16, 28 Backscheider, A.G., 27, 31 Baddeley, A.D., 192, 216 Bahrick, L.E., 18, 28 Baillargeon, R., 27, 28 Baker, L., 179, 213 Baker, N.A., 185, 219 Baker, N.D., 189, 190, 192, 213 Baldwin, J.M., 245, 246, 256 Ball, L.V., 69, 70, 71, 74, 75 Ball, O., 195, 222 Bamberg, M., 211, 213 Bandi Rao, S., 25, 30 Bangston, S.K., 102, 124
Barnat, S.B., 16, 28 Barresi, J., 35, 36, 53, 57 Barr, R., 16, 29 Bartsch, K., 60, 65, 71, 75 Basow, S.A., 87, 96 Bates, E.A., 105, 124, 180, 183, 190, 205, 207, 214, 215, 218 Bates, J.E., 104, 125 Bauer, P.J., 6, 15, 16, 29, 40, 54, 84, 87, 96, 97, 102, 104, 105, 106, 107, 113, 122, 123, 124, 125, 154, 172 Beckwith, R.T., 65, 75 Beeghly, M., 62, 64, 65, 75 Belenky, M.F., 83, 96 Belk, S.S., 87, 99 Benson, J.B., 19, 29, 131, 149 Benson, M., 172 Berridge, D.M., 41, 45, 53, 56 Bishop, D.V.M., 208, 221 Bittinger, K.A., 105, 123 Blachman, B., 194, 214 Black, J.E., 192, 216 Blake, E., 201, 203, 217 Blake, I.K., 42, 46, 54 Bloom, B.S., 205, 214 Bloom, L., 43, 54, 63, 65, 75, 171, 172, 183, 191, 214, 242 Bluck, S., 85, 97 Bohannon, N., 185, 207, 211, 214 Bonvillian, J.D., 183, 185, 191, 207, 208, 214, 219 Bordo, S., 82, 96 Bornstein, M.H., 193, 211, 220 Bourg, T., 97, 125 Bowerman, M., 28, 29, 191, 207, 214 Boyes-Braem, P., 24, 31 Bransford, J.D., 178, 193, 206, 211, 214 Breinlinger, K., 27, 32
257
258 Bretherton, I, 62, 64, 65, 75, 170, 173, 183, 214 Bretherton, I, 62, 64, 65, 75, 170, 173, 183, 214 Brice Heath, S., 46, 54, 160, 172 Brockmeier, J., 229, 236 Brown, A.L., 47, 178, 193, 206, 211, 214 Brown, J.R., 41, 45, 53, 55, 62, 75 Brudos, S.L., 131, 149 Bruner, J.S., 9, 43, 56, 84, 85, 96, 154, 155, 158, 172, 194, 214, 236, 240, 243, 244, 244 Bryant, D., 64, 77 Buckingham, D., 25, 32 Buckner, J.P., 87, 97, 170, 172 Burch, M.M., 6, 104, 105, 106, 107, 122, 123, 124 Burkhardt, A., 190, 218 Butkovsky, L., 189, 190, 192, 211, 219 C Caldwell, L.B., 195, 214 Camarata, M., 189, 190, 192, 193, 211, 214, 216, 219 Camarata, S.M., 189, 190, 192, 193, 206, 211, 214, 216, 219 Campione, J., 211, 214 Canetti, E., 151, 172 Cantwell, D.P., 179, 213 Carbaugh, D., 229, 236 Carey, S., 26, 27, 29, 60, 75, 183, 191, 214 Carpendale, J.I., 41, 54 Carver, L.J., 15, 29, 102, 124, 154, 172 Caspi, A., 71, 76 Cassidy, K.W., 69, 70, 71, 74, 75 Catts, H., 194, 215 Chalkley, M., 191, 207, 217 Chance, S., 192, 217 Chandler, M.J., 41, 54 Chiasson, L., 69, 70, 76, 77 Chipchase, B.B., 208, 221 Choi, S., 28, 29, 30 Chomsky, N., 206, 214 Chu, J.Y., 69, 70, 71, 74, 75 Cicchetti, D.V., 71, 76 Clark, E.V., 26, 29, 191, 215 Cleave, P., 192, 215 Clements, W.A., 35, 36, 41, 42, 45, 53, 57 Clinchy, B.M., 83, 96 Cocking, R.R., 54, 55, 178, 193, 206, 211, 214 Code, L., 82, 96 Cohen, D., 71, 76 Cole, M., 230, 231, 238 Cole, P.M., 179, 215 Collins, A.F., 96 Collins, P.F., 18, 31
AUTHOR INDEX Connerton, P., 86, 96 Conway, M.A., 79, 96 Cooke, T., 71, 76 Coppola, M., 23, 32 Corkin, S., 23, 32 Cosmides, L., 192, 201, 215, 217 Cowan, N., 96, 149 Crago, M., 178, 190, 213 Craven, P.L., 8, 189, 219 Cross, D., 35, 36, 44, 53, 57, 61, 78 Crowe, E., 69, 70, 77 Csikszentmihalyi, I., 181, 205, 215 Csikszentmihalyi, M., 181, 205, 215 Cuhn, D., 29, 30 Cutting, A.L., 35, 36, 55 D Daiute, C., 173 Dale, P.A., 206, 215 Dale, P.S., 105, 124 Daly, B.P., 102, 123 Damasio, A., 84, 96, 176, 179, 212, 215 Damon, W., 29, 30, 173 Davidge, J., 69, 70, 76 Davis, P.J., 87, 96 Dawson, G., 212, 215 Deacon, T.W., 201, 202, 204, 215, 246, 256 Deak, G., 191, 207, 217 DeFrancis, J., 236 de Houwer, A., 199, 215 Delgado-Gaitan, C., 43, 47, 55 Demuth, C., 190, 215 Dennett, D., 228, 236 Denninger, M., 185, 219 de Villiers, J.G., 63, 75 deVilliers, P.A., 63, 75 Dickinson, D.K., 194, 221 Ding, B., 234, 237 Dodson, K., 191, 222 Donald, M., 9, 10, 11, 84, 96, 200, 201, 203, 215, 225, 226, 227, 229, 231, 233, 234, 236, 253, 254, 255, 256 Donaldson, M., 69, 75 Dowden, A., 16, 29 Dow, G.A., 16, 29, 102, 105, 123 Dropik, P.L., 40, 54, 102, 105, 112, 113, 123 Dunn, J., 35, 36, 42, 45, 53, 55, 56, 61, 62, 68, 75, 76, 157, 158, 170, 173, 177, 221 Dweck, C.S., 178, 195, 206, 211, 215 Dyer, J.R., 69, 70, 71, 74, 76 E Edelman, B., 23, 28, 208, 213 Edelman, G.M., 212, 215 Edwards, D., 85, 98 Eimas, P.D., 21,31
AUTHOR INDEX Elman, J.L., 180, 207, 215 Ely, R., 155, 173 Emde, R.N., 178, 220 Engel, S., 7, 8, 103, 124, 130, 132, 149, 153, 154, 155, 160, 164, 170, 173 Ericsson, K.A., 205, 215, 221 Eskritt, M., 233, 236 F Fagan, J.F., 17, 18, 29 Farrant, K., 104, 112, 124, 130, 149 Feeny, N., 69, 70, 71, 74, 75 Feldman, C., 158, 173, 240, 241, 243 Fenson, L., 105, 124, 206, 215 Fernyhough, C., 35, 53, 56, 69, 77 Fey, M.M.E., 192, 194, 215, 216 Fiese, B., 211, 221 Fillmore, L.W., 199, 216 Fischer, A.H., 87, 96, 97 Fitch, J., 86, 96 Fivush, R., 5, 6, 8, 80, 84, 85, 87, 88, 94, 95, 96, 97, 103, 105, 112, 114, 120, 124, 125, 128, 129, 130, 132, 133, 149, 150, 154, 169, 170, 173, 174, 194, 211, 216, 229, 236 Flavell, E.R., 33, 55 Flavell, J.H., 29, 33, 55 Fletcher, P., 57 Folven, R., 207, 214 Ford, L.H., 156, 173 Fox, N., 17, 31 Fradley, E., 35, 53, 56 Frederici, A.D., 192, 217 Freeman, N.H., 41, 45, 53, 56 Freese, P., 208, 220 French, R.M., 22, 30 Friedman, S.L., 128, 131, 149, 150 Fromhoff, F.A., 88, 94, 97, 132, 149, 154, 173 Frye, D., 75, 76 Fung, H., 157, 174 Furrow, D., 64, 69, 70, 76, 77 Fu-Xi, F., 42, 55 G Galda, L., 160, 174 Gallimore, R., 47, 55 Garcia-Falconi, R., 87, 99 Garman, M., 57 Gartner, B., 63, 65, 75 Gathercole, S.E., 192, 216 Gauvain, M., 81, 95, 97 Geach, P.T., 68, 76 Gearhart, M., 69, 70, 76 Geertz, C., 43, 55 Gelman, R., 38, 55 Gentner, D., 28, 29
259 Gergen, K.J., 85, 97 German, T., 44, 53, 55 Gibson, E.J., 95, 97 Gillam, R., 194, 216 Gillberg, C., 193, 216 Gilligan, C., 83, 97 Goldberger, N.R., 83, 96 Goldenberg, C., 47, 55 Goldsmith, H.H., 105, 124 Golomb, C., 181, 195, 216 Good, A.B., 190, 218 Gopnik, A., 33, 35, 44, 53, 55, 240 Goshen-Gottstein, Y., 20, 31 Gottlieb, G., 201, 206, 207, 216 Gould, S.J., 180, 212, 216 Graf, P., 17, 29 Grant, P.R., 205, 216 Gray, J.T., 154, 173 Gray, W., 24, 31 Greenberg, M.T., 64, 66, 68, 77 Greenfield, P.M., 43, 47, 54, 55, 56 Green, F.L., 33, 55 Greenough, W.T., 192, 216 Griscom, J.L., 82, 83, 97 Growdon, J.H., 23, 32 Gruber, S., 66, 78 Gruendel, J., 40, 42, 57 Gunnar, M.R., 104, 124 Gupta, M., 35, 53, 56 Gurtner, J.L., 210, 217 H Habermas, T., 85, 97 Haden, C.A., 94, 95, 97, 99, 103, 105, 112, 114, 120, 124, 125, 130, 132, 133, 149, 150, 236 Haden, P., 17, 29, 84, 85, 87, 88, 97 Hafitz, J., 63, 65, 75 Hagen, R., 202, 216 Haight, J., 87, 96 Haine, R.A., 85, 97, 130, 149 Haith, M.H., 19, 29 Halberstadt, A.G., 82, 97 Haley, K., 193, 216 Halliday, M.A.K., 42, 43, 56 Hall, J.A., 82, 97 Hall, W.S., 69, 70, 76 Hammill, D.D., 71, 76 Hampson, J., 67, 77 Han, J.J., 229, 237 Happ, F., 71, 76 Haraway, D., 80, 97 Harding, S., 80, 82, 98 Harris, P.L., 33, 45, 46, 53, 54, 56, 60, 70, 71, 75, 76 Harris, R.J., 63, 76, 230, 237
260 Hart, B., 46, 56, 193, 216 Hart, D., 173 Hauser, M.D., 25, 29 Hayne, H., 16, 29 Heidbreder, E., 80, 98 Heimann, M., 176, 193, 206, 211, 219, 221 Henseler, S., 62, 77 Hernandez-Reif, M., 18, 28 Hernandez-Sanchez, J.E., 87, 99 Hershey, K.L., 104, 125 Hertsgaard, L.A., 102, 123 Hess, R.D., 46, 56 Hickok, G., 23, 32 Hitch, G.J., 192, 216 Homer, B.D., 8, 9, 10, 234, 235, 237 Hoogstra, L., 157, 174 Horgan, D., 190, 216 Ho, W.C., 181, 199, 216 Hresko, W.P., 71, 76 Hudson, J.A., 7, 103, 112, 124, 127, 128, 129, 130, 132, 134, 135, 136, 148, 149, 150, 162, 165, 173, 194, 217 Hughes, A.T., 70, 76 Hughes, C., 35, 36, 42, 45, 53, 56, 71, 76 Hughes, D., 192, 215 Hutchins, P., 71, 76 Hutton, D., 71, 76 Hyams, N.M., 206, 217 I Illich, I., 232, 237 J Jackson, J., 71, 76 Jacobson, K., 27, 32 Jacoby, L.L., 18, 29 Jesso, J.B., 194, 220 Johanson, D., 201, 203, 217 Johnson, C.N., 64, 66, 68, 71, 76, 78 Johnson, D., 24, 31 Johnson, M.H., 24, 31, 180, 207, 215 Jolly, A., 203, 217 Jonides, J., 23, 32 Jubicek, L.F., 178, 220 Jusczyk, A.M., 192, 217 Jusczyk, P.W., 192, 217 K Kahn, A.S., 80, 82, 99, 193, 217 Kail, R., 123 Kaiser, A., 193, 211, 217 Kaplan, B., 185, 219 Kaplan, C.A., 208, 221 Karmiloff-Smith, A., 14, 25, 29, 180, 207, 215 Keats, D.M., 42, 55
AUTHOR INDEX Keetz, A., 193, 211, 217 Keller, A., 156, 173 Kessler Shaw, L., 67, 68, 69, 71, 77 Keys, C.W., 234, 237 King, B.J., 201, 217 Kitayama, S., 42, 43, 56 Kleinknecht, E.E., 104, 105, 123, 124 Klein, P.J., 16, 28, 29 Klein, S.B., 192, 217 Klin, A., 71, 76 Knowlton, B.J., 17, 25, 29, 32 Koegel, L., 208, 217 Koegel, R., 208, 217 Kolodny, J.A., 23, 29 Korosheptz, W.J., 23, 32 Kroupina, M.G., 113, 123, 124 Kuczaj, S.A., 76, 218 Kuhn, D., 237 Kyriakidou, C., 41, 45, 53, 56 L Labov, W., 85, 98, 155, 162, 173 Lalonde, C., 41, 54 Lamb, M.E., 47 Larrivee, L., 194, 215 Lauck, G., 189, 219 Lawrence, V.W., 46, 56 Leadbeuater, B., 237 LeDonne, M., 40, 56 Lee, K., 233, 235, 236, 237 Leekam, S.R., 41, 45, 53, 57, 71, 77 Leichtman, M.D., 86, 98, 229, 237 Leopold, W.F., 183, 217 Lepper, M., 176, 210, 217, 219 Leslie, A.M., 44, 53, 55, 56, 57 Levy-Bruhl, L., 230, 237 Levy, E., 67, 68, 76, 240, 243 Lewis, C., 41, 45, 53, 56 Lieberman, P., 180, 217 Lightbown, P.M., 199, 220 Lightfoot, D., 207, 217 Limber, J., 63, 76 Lohmann, H., 63, 76 Longino, H.E., 81, 98 Long, S., 192, 215 Lucariello, J., 3, 4, 9, 38, 40, 55, 56, 131, 150, 154, 155, 172, 240, 243, 244, 244 Lutz, D.J., 69, 70, 71, 74, 75 M Maccoby, E., 95, 98 Macomber, J., 27, 32 Maguire, M.C., 173 Mahoney, M., 207, 217
AUTHOR INDEX Mandler, G., 17, 21, 29, 30 Mandler, J.M., 3, 15, 16, 17, 18, 19, 20, 21, 22, 24, 25, 27, 28, 29, 30, 31, 40, 56 Manning, C., 192, 220 Maratsos, M.P., 64, 76, 191, 207, 217 Marchman, V.A., 190, 218 Marcus, G.F., 25, 30 Mareschal, D., 22, 30 Margulies, C., 191, 214 Maridaki-Kassotaki, K., 41, 45, 53, 56 Marinsen, H., 184, 222 Mark, M., 207, 211, 214 Markman, E.M., 29 Markus, H.R., 42, 43, 56, 86, 98 Martin, A.J., 192, 216 Marvin, R.S., 64, 66, 68, 77 Maxwell, M., 181, 199, 218 Mayes, L.C., 71, 76 McAdams, D.P., 85, 98 McCabe, A., 85, 99, 112, 125, 132, 150, 160, 173, 194, 217, 220 McCarthy, M., 90, 98 McDonough, L., 15, 16, 21, 22, 24, 25, 27, 28, 29, 30, 31 McFadden, T.U., 194, 216 McKee, R.D., 16, 18, 31 McKeough, A., 71, 76 McKie, R., 203, 204, 221 McLuhan, M., 231, 237 McNew, S., 62, 64, 75 Meacham, J.A., 156, 173 Meins, E., 35, 53, 56, 69, 77 Meltzoff, A.N., 16, 28 Menn, L., 212, 214 Merriman, W.E., 221 Mervis, C.B., 24, 31, 191, 218 Meyer, G., 192, 220 Michel, M.K., 179, 215 Middleton, D., 85, 98 Miller, P.J., 46, 56, 57, 80, 81, 96, 98, 99, 157, 162, 174 Miller, R.S., 87, 99 Mindolovich, C., 40, 56 Mintz, J., 157, 174 Miscione, J.L., 64, 66, 68, 77 Mishkin, M., 15, 16, 28, 31 Mitchell, P., 55, 75 Montgomery, D.E., 60, 65, 66, 77 Moore, C., 35, 36, 53, 57, 64, 69, 70, 75, 76, 77 Moscovitch, M., 20, 31 Mumme, D.L., 210, 217 Munn, P., 170, 173 Murdoch, K., 87, 99 Mussen, P., 29
261 N Naito, M., 35, 36, 41, 42, 45, 53, 57 Nakisa, R.C., 180, 218 Nathan, D., 194, 221 Nation, J., 208, 213 Needham, A., 27, 28 Neisser, U., 97, 174, 216 Nelson, C.A., 15, 18, 29, 31, 154, 172 Nelson, K., 1, 2, 5, 8, 9, 10, 11, 13, 14, 20, 27, 31, 40, 42, 57, 59, 60, 61, 62, 65, 66, 67, 68, 69, 71, 77, 79, 84, 85, 86, 94, 95, 98, 101, 102, 103, 121, 123, 124, 125, 127, 128, 130, 131, 150, 152, 153, 154, 171, 172, 173, 174, 175, 177, 178, 185, 191, 194, 200, 203, 204, 207, 210, 211, 218, 219, 221, 223, 224, 225, 226, 227, 228, 229, 230, 232, 233, 234, 237, 239, 240, 241, 242, 243, 244, 247, 248, 252, 253, 256 Nelson, K.E., 176, 179, 183, 185, 188, 189, 190, 191, 192, 193, 194, 195, 206, 207, 208, 211, 212, 214, 216, 218, 219, 220, 221 Nelson, L.H., 81, 98 Newby, K., 193, 220 Newcombe, N., 17, 31 Newport, E.L., 25, 29 Nicely, P., 193, 211, 220 Nichols, H., 199, 220 Nie, H., 234, 237 Ninkovic, M., 195, 220 Nittrouer, S., 192, 220 Nosofsky, R., 23, 31 O Obler, L., 212, 220 O’Brien, R.G., 64, 66, 68, 77 O’Loughlin, M., 237 Olson, D.R., 8, 11, 33, 54, 60, 64, 66, 75, 77, 231, 232, 234, 235, 237 O’Toole, A.J., 23, 28 Oyserman, D., 85, 98 P Padgett, R.J., 207, 211, 214 Paley, N., 195, 220 Pani, J.R., 191, 218 Parisi, D., 180, 207, 215 Parkin, L., 35, 36, 41, 42, 45, 53, 57 Pasupathi, M., 86, 87, 99 Patalano, A.L., 23, 32 Patriquin, M., 69, 70, 77 Peevers, B.H., 156, 158, 166, 174 Pellegrini, A.D., 160, 174 Pelletier, J., 62, 71, 75
262 Peltzer-Karpf, A., 176, 220 Pemberton, E., 195, 219, 220 Pepper, S.C., 80, 99 Perkins, A., 69, 70, 71, 74, 75 Perkins, D.N., 176, 211, 219, 220 Perlmann, R., 194, 221 Perner, J., 33, 35, 36, 41, 42, 44, 45, 53, 57, 61, 63, 66, 71, 77, 78 Peskin, J., 3, 4, 60, 69, 70, 77 Peterson, C., 85, 99, 112, 125, 132, 150, 160, 163, 173, 194, 220 Pethick, S.J., 105, 124 Piaget, J., 13, 14, 15, 19, 31, 40, 57, 224, 225, 237 Pickens, J.N., 18, 28 Pillemer, D., 86, 99 Pinker, S., 23, 32, 207, 220 Pipp-Siegel, S., 178, 220 Plesa, D., 62, 77 Pleydell-Pearce, C.W., 79, 96 Plomin, R., 177, 221 Plunkett, K., 180, 207, 215, 218 Polizzi, P., 44, 53, 56 Potter, E., 80, 96, 98 Potts, R., 157, 174 Preece, A., 158, 167, 174 Premack, D., 60, 61, 64, 77 Pressman, L.J., 178, 220 Prinz, E.A., 181, 199, 220 Prinz, P.M., 181, 199, 220 Proulx, E.A., 91, 99 Q Quinn, P.C., 21, 22, 24, 30, 31 R Rademacher, J., 104, 105, 124 Ramus, S.J., 25, 29 Raven, B.H., 82, 99 Read, C.A., 234, 237 Records, N., 208, 220 Reese, E., 85, 87, 88, 94, 97, 103, 104, 105, 112, 114, 120, 124, 125, 130, 132, 133, 149, 150, 169, 174 Reid, D.K., 71, 76 Rekau, L., 191, 222 Reznick, J.S., 105, 124, 205, 215 Rice, M., 26, 31, 216, 218 Riggs, K.J., 55, 75 Risley, T.R., 46, 56, 193, 216 Rispola, M., 63, 65, 75 Robertson, S.B., 192, 220 Rogoff, B., 81, 99 Rosch, E., 24, 31 Rosenkrantz, S.L., 21, 31 Rosser, S.V., 80, 99
AUTHOR INDEX Rothbart, M.K., 104, 125 Rourke, M.T., 69, 70, 71, 74, 75 Rousseau, J.J., 230, 238 Rovee-Collier, C., 16, 17, 31 Rubin, D.C., 79, 96, 99 Rubin, .Z, 156, 157, 174 Ruffman, T., 35, 36, 41, 42, 45, 53, 57, 69, 70, 77 S Sachs, J., 130, 150 Sacks, O., 197, 220 Sameroff, A., 211, 221 Schank, R.C., 239, 244 Schauble, L., 237 Schiefelbusch, R., 216, 218 Scholl, B.J., 44, 53, 57 Scholnick, E.K., 70, 76, 81, 96, 98, 99, 128, 150 Schwade, J.A., 113, 123 Scribner, S., 230, 231, 238 Secord, P., 156, 158, 166, 174 Selman, R.L., 156, 174 Service, E., 192, 216 Shapiro, L.R., 128, 129, 150, 162, 165, 173, 194, 217 Shatz, M., 27, 31, 62, 65, 69, 70, 71, 74, 76, 77 Shaw, L.K., 178, 237 Shipley, E.F., 46, 56 Shipman, V.C., 46, 56 Shreve, J., 203, 217 Shultz, T.R., 25, 32, 68, 78 Siegler, R., 29, 30 Silber, S., 62, 65, 77 Silk, A.M.J., 195, 221 Sinclair, J., 190, 221 Sirois, S., 25, 32 Slackman, E., 194, 216 Slade, L., 69, 70, 77 Slobin, D.I., 207, 221 Sloboda, J., 212, 221 Slomkowski, C., 41, 45, 53, 55, 177, 221 Smiley, J., 92, 99 Smith, E.E., 23, 32 Smith, L.B., 176, 212, 221 Smith, P.H., 25, 32 Snell, W.E., Jr., 87, 99 Snow, C.E., 43, 46, 57, 194, 221, 240 Snowling, M.J., 208, 221 Snyder, L., 183, 214 Soderbergh, R., 180, 199, 221 Sosa, B.B., 128, 129, 150 Spada, N., 199, 220 Spelke, E.S., 27, 32 Sperry, L., 162, 174 Sprengelmeyer, A., 170, 173 Squire, L.R., 15, 16, 17, 18, 25, 29, 31, 32
AUTHOR INDEX Stallworth, L.M., 87, 99 Stanowicz, L., 185, 214 Steedman, C., 158, 174 Stennes, L., 87, 96 Stringer, C., 203, 204, 221 Strohner, H., 191, 221 Strothard, S.E., 208, 221 Svenkerud, V.Y., 192, 217 T Tamis-Lemonda, C.S., 193, 211, 220 Tarule, J.M., 83, 96 Tattersall, I., 203, 221 Taylor, A., 71, 76 Teale, W.H., 70, 78 Tenney, Y.J., 87, 99 Tercyak, K.P., 71, 76 Tesla, C., 41, 45, 53, 55 Tessler, M., 103, 125 Teti, L.O., 179, 215 Thaiss, L., 44, 56 Thal, D.J., 105, 124, 205, 215 Thelen, E., 176, 212, 221 Thomas, G., 195, 221 Thompson, C., 35, 36, 53, 57 Tijus, T., 176, 193, 216, 219, 221 Tinker, E., 191, 212, 214 Tjus, T., 193, 206, 211, 219, 221 Tolstoy, L., 152, 172, 174 Tomasello, M., 63, 76, 188, 191, 217, 221, 222, 227, 238 Tomblin, J., 208, 220 Tononi, G., 212, 215 Tooby, J., 192, 201, 215, 217 Torrance, E.P., 195, 222 Travis, L.L., 102, 105, 124, 125 Trosborg, A., 150 Tuckey, M., 35, 53, 56 U Ullman, M., 23, 26, 32 V Valentin, D., 23, 28 van den Vroek, P.W., 97, 125 van Kleeck, A., 194, 216 Vernon, S., 235, 238 Vijayan, S., 25, 30 Vinden, P.G., 45, 53, 57 Vishton, P.M., 25, 30 von Tetchner, S., 184, 222 Vriezen, E., 20, 31 Vygotsky, L.S., 81, 95, 99, 224, 225, 229, 238 W Wainwright, R., 35, 53, 56
263 Waletzky, J., 155, 162, 173 Wallace, C.S., 192, 216 Wallas, G., 59, 78 Wang, Q., 229, 237 Warren, S., 215, 216 Waters, J.M., 102, 124 Watson, J., 35, 36, 44, 57, 61, 78, 243, 244 Weiner, J., 205, 222 Weir, R., 241 Weismer, S.E., 192, 220 Weist, R.M., 130, 150 Weizman, Z .O., 46, 57 Wellman, H.M., 35, 36, 44, 53, 55, 57, 60, 61, 62, 64, 65, 66, 68, 69, 70, 71, 74, 75, 76, 78 Welsh, J.A., 176, 188, 189, 190, 192, 193, 206, 211, 219 Wenner, J.A., 40, 54, 102, 105, 112, 113, 123, 124 Werner, R.S., 69, 70, 71, 74, 75 Wessels, J., 192, 217 Wewerka, S.S., 40, 54, 102, 105, 112, 113, 123 White, S.H., 86, 99, 240 Whorf, B.L., 62, 78 Wiebe, S.A., 102, 124 Wilbur, R., 199, 222 Wilson, B., 88, 99, 195, 222 Wilson, M., 195, 222 Wiltshire, S., 197, 198, 212, 222 Wimmer, H., 61, 63, 66, 71, 77, 78 Windsor, J., 215, 216 Woodruff, G., 60, 61, 64, 77 Woolverton, M., 210, 217 Xuan, Y., 8, 189, 219 Xue, Q., 235, 237 Xu, F., 235, 237 X Yoder, J.D., 80, 82, 99 Yoder, P.J., 193, 211, 217 Yoshinaga-Itano, C., 178, 220 Yotive, W., 237 Youngblade, L., 41, 45, 53, 55 Z angl, R., 176, 220 Z hang, Y., 234, 237 Z ola-Morgan, S., 15, 32 Z
Subject Index A Accelerating learning: Why significant learning is rare and what to do about it, 176 Accommodation, 40 Acts of meaning, 84, 85, 96 Actual minds, possible words, 194 Advances in child development and behavior, 86, 104 Advances in developmental psychology, 40, 42 Advances in motivation and achievement: Culture, motivation and achievement, 47 Affordances, 95 African exodus: The origins of modern humanity, 203, 204, 221 American Journal of Speech-Language Pathology, 208 American Naturalist, 245, 246 American Psychologist, 154 American Scientist, 43 American Sign Language; Linguistic and applied dimensions, 199 Amnesia, infantile, 19 Amnesics, 16, 17, 18 Analyzing discourse: Text and talk, 85 Animal research, 13, 25 Anna Karenina, 152 Annals of the New York Academy of Sciences, 185, 188, 207 An anthropologist on Mars, 197 Appearance v. reality, 4, 36, 38, 50, 52 Applied Cognitive Psychology, 87,170, 192 Applied Psycholinguistics, 46, 69, 70, 71, 74 Applied Psychology: An International Review, 64, 66
Apprenticeship in thinking, 81 Art cultural symbol as, 8 learning through, 194, 195, 196, 197 skills acquisition in, 199 The art of thought, 59, 78 Attention and performance XV: Conscious and nonconscious information processing, 20 Audience, 169 Autism, see also Specific Language Impairment effect of, 197 interaction and, 8 Autism, 193 Autism: Nature, diagnosis & treatment, 212 Autism: The International Journal of Research and Practice, 193 Autobiographical memory, see Memory, autobiographical Autobiographical memory and the construction of a narrative self: Developmental and cultural perspectives, 86, 87 B Baldwin Effect, 246, see also Evolution The beak of the finch, 205 The beginnings of social understanding, 61, 76, 157, 158 Beginning to read: Thinking and learning about print, 70 Behavior correlations in, 118, 119, 120 interpretation of, 61 verbal relationship to, 114, 115, 116, 117, 118
265
266 Behavioral and Brain Sciences, 44, 53, 60, 61, 64, 207 Behaviorism, 250 The best laid plans: Beyond scripts are counterscripts, 40 Beyond modularity: A developmental perspective on cognitive science, 14, 25 Beyond names for things: Children’s acquisition of verbs, 188, 191, 207 Blue windows: A Christian Science childhood, 88 Book of Nature, 232 Brain and Cognition, 1992, 18 Brain scanning, 23 Bringing Reggio Emilia home, 195 British Journal of Developmental Psychology, 44, 53, 62, 71 British Journal of Psychology, 192 Bruner, Jerome, 13, 240, 244 Butterfly Effect, 184, 185 C Canadian Journal of Experimental Psychology, 233 Canonicality, 154 Categorization forms for, 20 global-to-basic in, 24 perceptual/conceptual forms of, 3, 14, 20, 21, 22, 24 study of, 1, 2, 14, 20, 23, 24 Causal understandings in cognition and culture, 27 Causes and effects in communication and language intervention, 193, 211 Causes of development, 60 Change, representational, 4, 50, 179 Child Development, 17, 33, 35, 36, 41, 44–46, 53, 61, 64, 66, 68–70, 95, 102, 105, 129, 130, 131, 191, 192, 229 Child Language, 158, 162, 167 Child language: An international perspective, 185 Children’s friendships, 156, 157 Children’s language, 130, 176, 179, 181, 185, 183, 191, 194, 206, 207, 211 Children’s minds, 69, 75 Children’s reasoning and the mind, 44, 53, 63 Children’s single word speech, 68 Children’s theories of mind, 65, 68 Children talk about the mind, 60, 65, 71 The child’s construction of language, 191 The child’s creation of a pictorial world, 195 The child’s conception of the world, 224 The child’s discovery of the mind, 59, 60 the Child
SUBJECT INDEX autism in, 28 development of, 2 language’s effect on, 2 mediated mind of, 10 mnemonics and, 6, 122 mother’s interaction with, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111 The child as psychologist: An introduction to the development of social cognition, 127, 150 Co-construction, 5, 74 Cognition, 20, 22-25, 35, 44, 53, 61-63, 65, 234 Cognition and Emotion, 71 Cognition, perception, and language: Vol. 2. Handbook of child psychology, 19, 20, 27 Cognitive development evolutionary/developmental processes in, 254 functions of, 2, 242, 243 as interactive process, 1, 224, 227 theory, 1 through language, 1, 2, 3, 8, 9, 223, 224, 227, 229, 230, 248 literacy and, 8, 230, 231, 232, 233, 234, 235, 236 mediated approach to, 9 research in, 245, 246, 247, 248, 249, 250, 251, 252 scaffolding for, 6, 105–115, 120, 121, 129, 160, 176, 200, 201, 210, 211, 252 stages of, 2, 3, 9, 223, 224, 227 terms of, 62 theory of, 248, 249 writing systems for, 231, 232 Cognitive Development, 19, 24, 40, 69-71, 74, 85, 86, 88, 102, 105, 114, 128, 129, 131–133, 154, 191, 204, 229 Cognitive development and the acquisition of language, 63 Cognitive evolution, 247 Cognitive Psychology, 21, 22, 24, 25, 28, 33, 191 Cognitive Science, 87 The cognitive basis of future-oriented prosocial behavior, 35 Collective remembering, 85 Computers as cognitive tools, 210 Concept formation of, 2, 3, 20 general imitation test for, 24 perception v., 24 study of, 1, 2, 24 understanding of, 74
SUBJECT INDEX The concept of development: The Minnesota Symposium on Child Psychology, 95 Conceptual analysis, perceptual v., 20, 21, 22 Concept, word, and sentence: Interrelations in acquisition and development, 27 Conditioning, 16, 17 Confidence, 8 Consciousness and Cognition, 103 Consciousness recovered, 21 The construction of an autobiographical self: Developmental and cultural perspectives, 95 The construction of reality in the child, 40 Constructivism, 35 Contemporary construction of the child: Essays in honor of William Kessen, 212 Context effect of, 1 function in, 239, 241, 242 importance of, 239 Contextual functionalism, 9, 239, 240, 241, 242, 243, 244 Counterscripts, 40 Critical thinking, 43 Cross-cultural roots of minority child development, 42, 43, 46, 47 The crosslinguistic study of language acquisition, 207 Culture acquisition of, 62, 228 change in, 8 cognitive development in, 223, 224, 254, 255 effect of, 1, 151 evolution of, 8, 10, 225, 226, 246, 254, 255, 256 knowledge of, 7 language in, 67 place/power in, 80, 93 symbols/tools of, 2, 178, 199, 236 unmasking in, 247 Culture & Psychology, 229 D Darwin, Charles, 246, 249, 250, 251 Deferred imitation, 15, 16, 17 Descartes’ error: Emotion, reason and the human brain, 179, 212 Developing narrative structure, 32, 160, 162, 165, 194 Developing talent in young people, 205 Developing theories of mind, 33, 53, 60 Developmentalism, 13 Developmental Psychology, 16, 18, 22, 35, 36, 40–42, 45, 46, 53, 62, 64, 65, 85, 86,
267 102, 112, 130, 156, 170, 177, 169, 185, 188, 191, 192, 207, 211, 229, 233 Developmental Science, 15, 21, 25,44, 53, 154 Developmental spans in event comprehension and representation: Bridging fictional and actual events, 84, 102 The developmental psychology of planning: Why, how, and when do we plan?, 131 Development and Psychopathology, 113 The development of children’s friendships, 156 Development of cognition, affect, and social relations, 68 The development of emotion regulation: Biological and behavioral considerations. Monographs of the Society for Research in Child Development, 179 The development of the temporally extended self in preschool children: Theory and research, 148 The development and neural bases of higher cognitive functions, 30 The development of artistically gifted children, 181, 195 The development of complex metarepresentational reasoning: The case of situational irony, 40 The development of memory in childhood, 84, 130 Discourse Processes, 70, 87, 88, 94, 103, 112, 120, 131, 132 Dishabituation, v. habituation, 24 Distributed cognitions, 211 Dore, John, 240, 243 A dynamic systems approach to the development of cognition and action, 176, 212 Dynamic assessment: An interactional approach to evaluating learning potential, 211 Dynamic Tricky Mix, see also Language; Specific language impairment acquisition states for passives in, 185, 186, 187, 188, 189, 190 conditions of, 8 learning through, 180, 181, 183, 190, 191, 192, 193, 207, 209 model of, 176, 209 E Early intervention: A handbook of theory, practice, and analysis, 211 Early sensitivity to language-specific spatial categories in English and Korean, 28 Ecology and evolution of Darwin’s finches, 205
268 Egypt: People, gods, pharaohs, 202 The emergence of autobiographical memory, 84, 85, 94 Emergent literacy: Writing and Reading, 70 Emily monologues, 9 Emotion, 8 Enhancers for learning, 178, 179, 183, 201 ERP, see Event-related potentials Event knowledge: Structure and function in development, 40, 61, 67, 101, 127, 153, 178, 194 Event learning in later childhood: Acquiring counterscripts, 40 Event-related potentials (ERP), 18, 253 Events memory of, 14, 15, 127, 128, 130 planning for future, 127–141, 148 Evolution Baldwin Effect for, 246 development v., 10, 245, 246, 252 history of, 10 The exceptional brain: Neuropsychology of talent and exceptional abilities, 212 Expectancy, 8 Experience interpretation of, 155 organization of, 175 as vicarious, 158, 167 Experiential semantics, 60, 61, 62, 66, 67, 73 F False-belief, see also Language income level and, 49, 50, 52 linguistics and, 63, 71, 72 object complements as, 63 social metarepresentation and, 34, 38, 52 task of, 4, 61, 63, 64, 71 ToMM/SP for, 44 understanding of, 33, 35, 41, 63 Familiarization, v. preferential looking, 24 The feeling of what happens: Body and emotion in the making of consciousness, 84,176, 212 Feminism, 80, 93 Feminism/postmodernism, 82 Feminist epistemologies, 80–82 Feminist standpoint theory, 80, 81 Feminist Studies, 80 Finding art’s place, 195 First Language, 69, 70, 85, 189, 190, 192, 194, 211 The five to seven year shift: The age of reason and responsibility, 41 Floating cities, 197 FMS, see Functional Multilinear Socialization model Foundations for a psychology of education, 178
SUBJECT INDEX Foundations of reading acquisition and dyslexia: Implications for early intervention, 194 The foundations of mind: The origins of conceptual thought, 19 Freud, Sigmund, 244 Friends external/internal landscape of, 158, 166 nature of, 156, 157, 158 progression in, 156, 157 story information about, 167, 168, 169 From first words to grammar: Individual differences and dissociable mechanisms, 183 From Lucy to language, 201, 203 Functionalism analysis through, 34 core of, 2, 3, 14 perspective of, 1, 9 Functional Multilinear Socialization model (FMS), 33, 34, 35, 36, 41, 52 G Gender and emotion: Social psychological perspectives, 87 Gender stereotypes and roles, 87 Gerhardt, Julie, 240, 243 Gopnick, Alison, 240 Gossip, 7, 167, 170, 172, see also Narrative The Guttenburge galaxy, 231, 237 H Habituation stimulus and, 17 dishabituation v. , 24 Handbook of Child Psychology, 27 Handbook of child psychology: Vol. 3. Cognitive development, 26 Handbook of research methods in developmental psychology, 16 The handbook of child language, 199 The handbook of education and human development: New models of learning, teaching and schooling, 62 Heterogeneity in cross-cultural psychology, 42 Hirst, Bill, 240 Homo Erectus, 201, 204 Homo Sapiens, 202 Hot spot, 188 How people learn: Brain, mind, experience, and school, 178, 193, 206, 211 How societies remember, 86 Human change processes, 207 Human Development, 41, 61, 62 Human language and our reptilian brain, 180 Human minds, 69
SUBJECT INDEX I Imitation, conceptualization as, 22 In a different voice: Psychological theory and women’s development, 83 Individual differences in infancy: Reliability, stability, prediction, 104
Infancy, 24 Infant Behavior and Development, 16, 193, 211 Infant memory , 18 Infants, see also Memory global-to-basic response in, 24 research with, 1–4, 27, 101, 102, 103 response of, 17, 18, 69, 70 ToM in, 64 verbal/behavior relationship in, 114, 115, 116, 117, 118 Infant social cognition, 62, 64 Inference, 22, 61, 127, 232 Intentionality, 154 Interaction, see also Language; Memory; Narrative between mother-child, 102–111 as social, 1, 2, 121 Interdisciplinary convergence, 248– 251, 253, 256 Interest and gender in education, 82, 84 International Journal of Behavioral Development, 46, 128 International Journal of Psychology, 43 Interpretation of world literacy for, 232, 235 role of, 19 Intrapersonal-reflection, metarepresentational reasoning as, 35, 52 An introduction to the psychology of children’s drawings, 195 Introduction to augmentative and alternative communication, 184 J Journal of Applied Developmental Psychology, 158 Journal of Autism and Developmental Disorders, 193 Journal of Child Language, 27, 46, 63, 65–70, 157, 178 190, 194 Journal of Child Psychology and Psychiatry, 71 Journal of Child Psychology and Psychiatry and Allied Disciplines, 71 Journal of Cognition and Development, 20, 21, 60, 6 5 , 6 6 , 10 4 , 11 2 , 11 3 , 1 3 0 , 132,134,135,136, 149, 194, 227, Journal of Cognitive Neuroscience, 23 Journal of Experimental Child Psychology, 15, 17, 18, 21, 25, 63, 66, 102 Journal of Experimental Psychology: General, 15, 40
269 Journal of Experimental Psychology: Learning, Memory, & Cognition, 17, 18 Journal of Memory and Language, 192 Journal of Narrative and Life History, 155, 162 Journal of Neuroscience, 15, 16 Journal of Personality and Social Psychology, 87 Journal of Phonetics, 192 Journal of Psycholinguistic Research, 70 Journal of Research in Childhood Education, 43, 46, 235 Journal of Research in Science Teaching, 234 Journal of Social Behavior and Personality, 82 Journal of Speech and Hearing Research, 189, 190, 192, 193, 208 Journal of Speech, Language, and Hearing Research, 192 Journal of the Acoustical Society of America, 192 K Knowing and remembering in young children, 103, 112, 130, 132, 194 Knowing and remembering in young children. Emory symposia in cognition, Vol.3, 40 Knowledge, see also Theory of mind acquisition of, 3, 7, 13, 14, 40, 233, 234 conceptual v. sensorimotor, 14, 16, 27 of events, 14, 15, 127, 128, 130 experience as, 61 feminist theories and, 81 memory and, 3 organization of, 2 prior learning for, 38, 41 procedural v. declarative, 14, 16, 17 replacement/revision of, 35, 38, 39, 40 social interaction and, 81, 121 types of, 3, 24, 25 Knowledge of language: Its nature, origin and use, 206 L Label responses, 48, 51, 52 Language, see also Specific language impairment; Speech acquisition of, 3, 4, 14, 25, 26, 27, 185, 199, 206, 207, 223, 227 autobiographical memory and, 84, 85, 86 cognitive development through, 1, 2, 3, 8, 9, 223, 224, 227, 229, 230, 248 complementation in, 63 conceptual development dependent on, 59, 228 culture acquisition through, 67, 228 development of, 2, 254 effect of, 1, 6, 7
270 future planning with, 127–141, 148 implicit v. explicit, 25, 26, 27 intentionalist approach for, 43 maternal variables in, 6, 105–116, 132–136, 141–145, 147, 148, 160 mental terms as, 62, 63, 64 metacognitive terms as, 62, 63, 64, 73 narrative in, 7, 8, 69, 70, 103, 154, 163, 165, 168, 171, 229 passives in, 185, 186, 187, 188, 189, 190 personality leading to, 177, 178 social interaction of, 42, 62 sociocultural variables of, 45, 46, 62 stabilizing effect of, 228, 230, 234 structure of, 8 styles of, 103, 132, 133, 148 symbolism of, 2 system of use v. meaning in, 68, 69 ToM and development of, 4, 59, 60 voice/silence in, 82, 83, 84, 87–95 word-referent mapping in, 65 Language acquisition, 43 Language acquisition and the theory of parameters, 206 Language acquisition: Biological and behavioral determinants, 176, 193, 207 Language acquisition: The state of the art, 26, 27 Language and Cognitive Processes, 180 Language and Communication, 230 Language and mind, 206 Language development from two to three, 183 Language development: Syntax and semantics, 66, 181, 191 Language development: Vol. 2. Language, thought, and culture, 28 Language in cognitive development: The emergence of the mediated mind, 2, 5, 59, 60, 62, 65, 66, 67, 69, 70, 71, 74, 77, 79, 84, 85, 95, 98, 102, 124, 128, 130, 150, 178, 194, 200, 207, 210, 218, 223, 224, 225, 226, 227, 228, 229, 230, 232, 233, 237, 248, 256 Language intervention: Preschool through the elementary years, 194 Language Learning, 199 Language, literacy, and cognitive development: The development and consequences of symbolic communication, 68, 69, 71 Language socialization across cultures. Studies in the social and cultural foundations of language, No. 3 , 46 The language instinct, 207 Language, thought, and reality, 62 The last Neanderthal: The rise, success and mysterious extinction of our closest human relatives, 203
SUBJECT INDEX The latest on the best: Essays on evolution and optimality, 201 LEARN acronym, 177, 178, 179, 180, 181, 182, 183, 211, 213 Learning, see also Dynamic Tricky Mix; Sequencing; Theory of mind challenges for, 39, 41, 179 conditions for, 179, 180, 181 enhancers for, 178, 179, 183, 201 intrapersonal ToM in, 3 prior knowledge in, 38 rates for, 180 by sensorimotor, 14, 16, 27 types of, 3, 25 of words by Network representations, 192 Learning how to mean: Explorations in the development of language, 42, 43 Learning to reminisce: A developmental study of how young children talk about the past, 103, 130, 132 Levy, Elena, 240, 243 Lexicon, mental, 4 Linguistics, 185 Linguistic theory and psychological reality, 183, 191 Literacy cognitive development and, 2, 8, 230–236 perspectives on, 8 Literacy and orality, 232 Literacy, narrative & culture, 234 Logical positivism, 80, 81 Lucariello, Joan, 240, 244 Lucy’s child: The discovery of a human ancestor, 203 Lucy’s legacy: Sex and intelligence in human evolution, 203 M MacArthur Communicative Development Inventories for Toddlers, 105 Making sense: The acquisition of shared meaning, 20 Making stories: Law, literature, life, 244 Meaningful differences in the everyday experience of young American children, 46, 193 Mediated mind concept of, 2, 112, 122 development of, 9 exploration of, 2, 10 literacy and, 230, 231, 232, 233, 234, 235, 236 Memories of a Catholic girlhood, 90, 98 Memory autobiographical narrative and, 84, 85, 151
271 for events, 14, 15, 127, 128, 130 knowledge and, 3 map for, 153 mnemonics and, 6, 122 neurology and, 15, 23 organization of, 153 perspective of, 5 as pre-explicit, 18 procedural v. declarative, 15, 16, 17 and recognition, 15, 17, 18 research in, 2, 14, 127, 128 social interaction for, 128, 153 types of, 19 working v. long-term, 8 in young children, 101, 102, 152, 153, 155 Memory, 15, 87, 102, 154, 229 Memory and affect in development: The Minnesota symposia on Child Psychology, 154 Memory, autobiographical accuracy/retention of, 79 development of, 6, 79, 103, 104 evaluation of, 86 independent/shared, 112, 113 language and, 84, 85, 86 maternal scaffolding in, 6, 105–115, 120, 121, 129, 160, 176, 200 narrative for, 84, 85 origin of, 5 problem solving with, 154 self/other for, 83, 88, 89, 90, 91, 92 social-constructionist model for, 79, 128 voice/silence for, 82, 83, 84, 87–95 Memory development, 15, 18 Memory for everyday and emotional events, 16 Mental acts, 68 Mental lexicon, 62, 64, 65, 66, 67, 68, 73, see also Language Mental terms, see Language Merrill-Palmer Quarterly, 46, 85, 104, 112, 120, 157, 194, 211 Meta-analysis of theory-of-mind development: The truth about false belief, 35, 36 Metacognitive terms, see Language Metarepresentation, 4, 36, 38, 39, 46 Methods for studying language production, 179, 183, 185, 188, 193, 206, 211 A Mind so rare: The evolution of human consciousness, 255 Mind and language, 45, 53 Mind, human, 2, 9, 14 Minding the law, 244 Mind in society: The development of higher psychological processes, 81, 95, 224, 229 Mind reading, see Theory of mind (ToM)
SUBJECT INDEX The mind’s staircase: Exploring the conceptual underpinnings of children’s thought and knowledge, 70 Mnemonics, 6, 122, see also Memory Modeling, 15 Monographs of the Society for Research in Child Development, 40, 61, 102, 105, 113, 185, 206, 212 Mother autobiographical memory, maternal scaffolding and, 6, 105–115, 120, 121, 129, 160, 176, 200 child’s relationship with, 102–111 language, maternal variables and, 6, 105–116, 132–136, 141–145, 147, 148, 160 Multidiciplinarity, 251, see also Interdisciplinary convergence Multidimensional models of perception and cognition, 23 N Narrative construction of, 8, 103 evaluation of, 2, 85, 162, 194 external/internal landscape of, 158, 166 gender differences in, 169 information in, 168 in language, 7, 8, 69, 70, 103, 154, 163, 165, 168, 171, 229 perspective of, 154, 165 structure of, 163 text v. oral, 171 use of, 7, 69, 70, 171, 243 validation of, 5, 6 Narrative and identity: Studies in autobiography, self, and culture, 229 Narrative, autobiographical development of, 155 external/internal landscape of, 158 memory and, 84, 85, 151 physical/psychological aspects of, 156 scaffolding for, 159, 160, 176, 200 scripts v. stories in, 161, 162, 163 self-concept through, 156 setting for, 158, 159, 164 sharing of, 155 Narrative development: Six approaches , 160 Narratives from the crib, 2, 9, 11, 79, 95, 154, 243 Narrative thought and narrative language, 160 Nelson’s Rare Event Learning Mechanism, 185 Networks, 254
SUBJECT INDEX
272
Psychological Science, 23, 25, 26, 128, 130 The new cognitive neurosiences, 17 Psychology Novels, see Gossip concept of place/power in, 81, 82, 83 NYCLG, see New York Child Language Group goal of, 81 NYLAG, see New York Language Acquisition Psychology of Women Quarterly, 80, 82, 83 Group The psychology of literacy, 230 O Q Objects, 2, 3 Questioning the millennium, 212, 216 Observation, 176 On the origin of language: Two essays by R Jean-Jacques Rousseau and Johann Rare Event Learning Mechanism, 185 Gottfried Herder, 230 Reading in early childhood: A linguistic study Ontogenetic development, 2, 247, 248, 249, 250, of a preschool child’s gradual acquisi251, 252 tion of reading ability, 199 Optimal experience, 181, 205 Reading, 8, 70 Origins of the modern mind: Three stages in the Reality, v. appearance, 4, 36, 38, 50, 52 evolution of culture and cognition, 84, Recall, see also Memory 200, 201, 203, 225, 226, 227, 229, 230, development of, 15 231, 233, 234, 253, 254 measurement of, 19 The origins of intelligence in children, 40 stem completion v., 16 The origins of intelligence in the child, 19 Recognition and memory, 15, 17, 18 The origins of language: What nonhuman priReconstructing our past: An overview of automates can tell us, 201 biographical memory, 79, 85 Reinstatement, 16, 17 P Relations between maternal style, children’s rePatterning, 25, see also Sequencing call memory, and temperamental charPerception, 21, 23 acteristics, 104, 105, 106, 107 Perceptual analysis The relation of hearing status and emotional conceptual v., 20 availability to child language gain, function/process of, 3, 5, 19, 20, 22 178 Perceptual identification, 18, 20 The remembering self: Construction and accuPersonality structure in the life course, 85 racy in the life narrative, 85 Perspective, 9 The remembering self: Construction and accuPET study, 23 racy in the self-narrative, 194, 211 Phaedrus (Plato), 230 Reminiscence, 154 Phylogenetic theory, 226, 227, 247–252 Repetition priming, 17 Piaget, J., 13, 14, 15, 19, 40, 224, 225 Research Pinyin, 234, 235 animals for, 13, 25 Plato, 230 interdisciplinary convergence for, 250, 251, Pleasure gives way to interest and persistence: 253 Changing relations between temperamemory for, 2 ment and memory at 13 and 20 months, Research in Childhood Education, 1, 46, 104, 105 223 Postcards, 91 Research on communication and language disPower, 82, 83, 93 orders: Contributions to theories of Preferential v. familiarization looking, 24 language development, 176 Primitive mentality, 230 Research on Language and Social InteracPrimitive recognition, 18, 20 tion, 87 Proceedings of the National Academy of Sciences, Rethinking innateness: A connectionist 1 perspective on development, 180, Psychological Bulletin, 85, 86 207 Psychological perspectives on deafness, 207 The road to excellence, 205, 212 Psychological perspectives on the self: The self in Rosie’s walk, 71 social perspective, 86 Rote association, 26 Psychological Review, 2, 13, 14, 16, 17, 19, 27, 42, Routines, social meaning of, 153 43, 61, 79, 192, 201, 206, 207
SUBJECT INDEX S Scaffolding sociocultural context as, 6, 105–115, 120, 121, 129, 160, 176, 200, 201, 210, 211, 252 Science, 15, 17, 25 Science Education, 234 Scientific American, 15 Scientific method, 80, 82 Scripts, 161, 162, 163, 232, 235, see also Counterscripts; Narratives Scripts, plans, goals, and understanding, 239 Selection processing (SP), 44 Selectivity, 21, 22 Self-concept through autobiographical narrative, 156 income level for, 46 monologue for, 243, 244 socialization and, 4, 42, 43, 46, 156 Self-esteem, 8 Self-theories, 178, 195, 206, 211 The self and memory, 85, 87 The self in time: Developmental issues, 79, 95 The self in time: Developmental perspectives, 84, 85, 229 Semantics, see Experiential semantics; Language Sensorimotor skills, 14, 16, 27 Sequencing, see also Narrative action of, 6, 15, 25, 131 through narrative, 154 Seven psychologies, 80, 98 Sex Roles, 87 Shakespeare, William, 171 Sign language, 8, 180, 181 Sign Language Studies, 181, 199 Simulation theory (ST), 45 SLI, see Specific language impairment Snow, Catherine, 240 Social Development,, 35, 36, 53, 148 Social, emotional and personality development: Vol. 3. Handbook of child psychology, 104 Social interaction cognitive development and, 121, 128 feminist theories for, 81, 82 metarepresentational reasoning for, 34, 41 race, class, gender for, 87, 95 self/other relations in, 42, 157, 158 ToM for, 52 Social metarepresentation false-belief and, 34, 38 reasoning for, 34, 41 The social context of cognitive development, 81, 95 The social life of children in a changing society, 69, 70 Sociocultural experience
273 categorization and, 27, 28 cognitive development in, 1, 2, 3, 8, 9, 62, 223, 224 development of, 95 friends in, 156, 157, 158 linguistics in, 72, 73, 74 memory and, 5 reasoning for, 41 symbolic systems within, 72, 73, 74, 178, 184, 199 ToM reasoning for, 44, 45, 52, 72 utilization of, 2, 4, 61 voice/silence in, 82, 83, 84, 87–95 Socioemotional functions, 42 Socrates, 230 SP, see Selection processing Specific language impairment (SLI) autism as, 197, 208 research into, 8, 201 Speech as bilingual, 180, 181 as communication, 233 rhythms/sequences of, 25, 26 text and, 9, 231 Speech development of a lingual child: A linguist’s record. Vocabulary growth in the first two years, 183 The speech/language connection, 176, 179, 188, 193, 206, 211 ST, see Simulation theory Stem completion, v. free recall, 16 Stern, Daniel, 240, 244 Stevens’ handbook of experimental psychology: Vol. 3. Learning, motivation, and emotion, 38 Stimuli, response to, 17, 18 The stories children tell, 154, 160, 164 Story, see Narrative; Narrative, autobiographical Studies in cognitive development, 190 Sugarman, Susan, 240 Sugarman, Susan, 240 Symbolic systems acquisition of, 175, 184, 199, 204, 207, 212, 230 types of, 8, 176, 199, 203, 209, 226 The symbolic species: The co-evolution of language and the brain, 201, 202, 204, 246 T Task implicit v. explicit, 20–25, 27 metarepresentational performance for, 51 motivation for, 178
SUBJECT INDEX TBAQ, see Toddler Behavior Assessment Questionnaire The teachability of language, 199 The teachability of language, 179, 185 Teaching children to draw, 195 Teaching children with autism, 208, 217 Test of Early Language Development (TELD-III), 71 Test of Early Language Development: 3rd ed., 71 Text formation of, 243 speech and, 9, 231 symbolism of, 2, 8 Theories of memory, 79, 95 Theories of theories of mind, 62 Theory of mind (ToM) construction of, 3, 4, 35, 39 definition of, 34 development of, 33, 37, 38, 43, 44, 53 dialectical analysis for, 36, 39, 40 FMS model for, 33, 34, 35, 36, 41, 52 functionalist analysis for, 34 income and, 4, 47, 48, 53, 70 infant acquisition of, 64 language development and, 59, 60 literature for, 36 metarepresentation for, 33, 34, 36, 38, 39, 45 origin of, 34, 44, 45 perspectival analysis for, 36, 39 social, 3, 4, 34, 42, 46, 52, 62, 72 ST for, 45 ToMM/SP for, 44 Theory of mind (ToM), intrapersonal development of, 4, 38, 42, 46 language and, 4 reflection/learning as, 34, 35, 36, 38, 52 Theory-of-Mind Mechanism/Selection Processing (ToMM/SP), 44 Theory-theory (TT), 44 A thousand acres, 91 The tidy house: Little girls writing, 158 Time concept of, 7, 145, 146, 147, 228, 229 markers for, 134, 143, 145 Time, actions and cognition: Towards bridging the gap, 149 Time and human cognition: A life-span perspective, 130 Toddler Behavior Assessment Questionnaire (TBAQ), 105 Tolstoy, Leo, 152, 172 ToM, see Theory of mind ToMM/SP, see Theory-of-Mind Mechanism/Selection Processing The tongue set free, 171 Torrance tests of creative thinking, 191, 195
274 Towards a feminist developmental psychology, 80, 81 Transcendence, 40 The transition from infancy to language: Acquiring the power of expression, 43 TT, see Theory-theory The two sexes: Growing up apart, coming together, 95 U Understanding language through sign language research, 208 Understanding other minds: Perspectives from developmental cognitive neuroscience, 45, 53 Understanding other persons, 156, 158, 166 Understanding the representational mind, 33, 44, 53 Understanding the theoretical and methodological bases of augmentative and alternative communication, 176, 185, 211 A universe of consciousness, 212 Unmasking, 247, see also Evolution Using cognitive terms to enrich story texts for disadvantaged young children, 60 V Validation of narrative, 5, 6 Visual Arts Research, 195 W Watson, Rita, 240, 243 Ways with words, 46, 54, 160 We knew each other when we were childs: Friendship and collaborative narration in preschool, 155 Wendy, Amy, Beth: Learning language in South Baltimore, 46 Wertheimer, Max, 242 What is intelligence, 228 White Oleander , 86 White, Stanford, 240 Why, how, and when do we plan? The developmental psychology of planning, 128, 129 Women, men and gender: Ongoing debates, 82 Women’s ways of knowing: The development of self, voice and mind, 83 Wonderful life, 180 World hypotheses: A study in evidence, 80 The world on paper, 8, 60, 231, 232, 234 Writing, 8, 70, 231, 232 Y Yani, the brush of innocence, 181, 199
