Words, Thoughts, and Theories

Words, Thoughts, and Theories

Hardcover(First Edition)

$36.90 $45.00 Save 18% Current price is $36.9, Original price is $45. You Save 18%.

Product Details

ISBN-13: 9780262071758
Publisher: MIT Press
Publication date: 12/01/1996
Series: Learning, Development, and Conceptual Change Series
Edition description: First Edition
Pages: 284
Product dimensions: 6.33(w) x 9.24(h) x 0.91(d)

About the Author

Jeffrey Rosenschein is the Sam and Will Strauss Professor of Computer Science at the Hebrew University of Jerusalem.

Read an Excerpt

One influential approach to cognitive development began almost eighty years ago in Paris. With a treble "formation" in biology, logic and psychoanalysis, Jean Piaget conducted his first investigations with children. Charged with the task of standardizing a set of reasoning tasks, Piaget was delighted to discover that various concepts were neither innate nor open to explicit teaching, but spontaneously constructed by the child. Piaget's claims have been under sustained scrutiny, especially in the past thirty years, and the clamour of that empirical attack has been sufficient to convince many lay observers that Piagetian theory is dead. Insiders would probably admit, however, that several of his ideas are alive and well, if not universally accepted: that children are constantly engaged in reconstructing the conceptual vessel that they sail in; that, in so doing, they move from a less adequate conception of the world to a more adequate conception; that the main engine of cognitive development is an endogenous, constructive process that cannot be hurried; and that attempts to teach children cannot by-pass that constructive process because children invariably use it to make sense of — or distort — whatever instruction they receive. Despite the continuing power of these ideas, the past thirty years of research on cognitive development have seen various efforts to recast the Piagetian programme. First, Piaget assumed that the infant began life with a set of sensory-motor reactions but no conception of the objects, persons, or space that elicited those reactions. Given the plethora of striking findings with very young infants, few investigators these days would be willing to acceptsuch radical naivete. Second, Piaget was inclined to think of the child as concurrently engaged in various interconnected intellectual projects concerning space, time, weight, and so forth — making progress in each at roughly the same rate. Contemporary research favours instead the notion of "domain-specificity" — the possibility that the child's grasp of, for example, the principles of everyday physics need have little connection with his or her grasp of everyday psychology, so that a given child may be precocious in one and slow in the other. Third, although Piaget was intrigued by analogies between the young child's conception of say speed or time, and similar notions to be found in the history of science, he emphasized analogies of content rather than process. Indeed, he was very far from supposing that the intellectual groping of the pre-schooler has much to do with the research activity of the scientist. Several current theorists, by contrast, press for a much more pervasive analogy at the level of process. According to these investigators, the process of conceptual change in childhood is deeply akin to the process of theory construction, revision and replacement among scientists, notwithstanding differences of method. Alison Gopnik and Andrew Meltzoff acknowledge their debt to Piaget, and also emphasize, to varying degrees, these three revisions: the competence of the young infant, the notion of domain-specificity, and the child as scientist. Their account of cognitive development runs roughly as follows. At birth, the baby is equipped with a set of "dedicated" interpretative devices or modules. Armed with these devices, the toddler constructs a variety of mini-theories, one dedicated to the behaviour of inanimate objects, another to animate agents and so forth. These mini-theories come under increasing strain when the child is faced with observational anomalies. As in the history of science, the anomalies are handled by ad hoc postulates tacked on to the existing theory. As time goes on, the new postulate and its implications are digested more fully to the point where they can displace or trigger a reorganization of existing ideas. Gopnik and Meltzoff apply this vision to children's understanding of three domains: objects and their appearances; human agency; and conceptual kinds or categories. Gopnik and Meltzoff make two further claims about the developing relation between language and thought, both connected to their view of the child as scientist. First, they argue that a theoretical shift in, for example, the way that the child conceives of objects will have an impact on the child's use of language. Thus, as in science, a theoretical shift has repercussions for linguistic practice. Second, exposure to key theoretical terms triggers a conceptual search for their meaning. By implication, the language to which children are exposed offers an endless series of take home exercises that children need to work through. Hence, according to Gopnik and Meltzoff, the traffic between conceptual development and language is bi-directional. It does not flow only from cognition to language, as Piaget typically argued. In setting out their arguments regarding each domain, Gopnik and Meltzoff review the available evidence in a creative fashion. Anyone looking for an overview of much of the most exciting research in the past ten to fifteen years on early conceptual development would do well to start with their book. Their co-ordination of two lines of evidence that are normally kept separate — children's understanding of the permanence and the appearance of objects — is especially intriguing. It allows them to discuss not just how the child conceives of an object in impersonal space but also whether the child understands when and how an object is seen by other observers. This conjunction also implies — contrary to strong versions of the domain-specificity thesis — that the child's grasp of naive physics cannot be neatly separated from his or her grasp of naive psychology. More important for an assessment of the likely impact of their book is an evaluation of their over-arching analogy: the child as scientist. Is the analogy well founded, and will it advance the field? First, we can acknowledge, along with Gopnik and Meltzoff, that one rival theoretical position has major problems by comparison. Some investigators, impressed with the competence of the young infant, have proposed that certain types of core or modular knowledge last a lifetime. For example, it has been argued that infants have a module that enables them to understand the behaviour of medium-sized, three-dimensional solid objects — no constructive process being needed. Similarly, it has been argued that infants have a module dedicated to the understanding of propositional attitudes, such as beliefs and desires. The problem for the modularity thesis is that infants and young children display persistent errors in their dealings with both objects and attitudes, errors that eventually get sorted out. Admittedly, modular approaches have a variety of answers to these early limitations, but the answers typically involve some non-modular process to explain either a gap between competence and performance or a delay in the activation of the relevant module. Given that error correction is abundant in the course of cognitive development, the upshot is that any modular account has to saddle itself with an unsatisfactory mix of modularity and non-modularity or modularity and pre modularity to even make a start at explaining developmental change. By contrast, the child-as-scientist analogy is perfectly content with error correction — indeed, that is one of its major explanatory targets. Gopnik and Meltzoff are also right to insist that the children are not simple empiricists. They do make empirical generalizations, but these generalizations are the product of underlying theoretical commitments about the nature of reality. For example, when infants watch an object move behind a screen, they not only predict its reappearance at the far end of the screen, they also "commit" themselves to the enduring existence o the object while it is invisible. Proof of that commitment is their puzzlement at various types of mysterious disappearance or conjuring trick that imply the non-endurance of the object. Gopnik and Meltzoff argue that empiricist accounts, including the computer-based models of neural networks proposed by connectionists, are unable to account for the characteristic features of theory change: the emergence of such ontological commitments, and their elaboration or replacement in the course of cognitive development. However, this is to underestimate contemporary connectionist models. For example, with respect to object permanence, a couple of recent models have shown how a network that predicts the reappearance of an object from behind a screen with increasing accuracy, also "commits" itself to the enduring existence of that object during its invisibility. Moreover, attempts to model children's understanding of a see-saw or balance beam have successfully reproduced the way that children focus first on weight, and then gradually acknowledge the role of position on the arm. Indeed, Gopnik and Meltzoff ignore the connectionist threat to their own stance. It is important for Gopnik and Meltzoff that there be a clear divide between the predictions that children derive from their theory-based, ontological commitments and the a-theoretical, empirical generalizations which feed into such theories. In connectionist accounts, by contrast, those learning algorithms that can arrive at empirical generalizations — for example, about what sounds are permitted to follow one another in English — are not obviously different from the algorithms that go on to generate first one "commitment" and then another. In short, on connectionist accounts, the child as empiricist, and the child as theorizer, are two facets of the same type of learner. Still, notwithstanding the threat from connectionism, Gopnik and Meltzoff make a provocative case for taking the analogy between child and scientist seriously. Part of their claim is that the analogy is probably better stated the other way round. It is not so much that children are like scientists, but rather that scientists are like children, because — unlike most adults — scientists retain a conceptual stance that is characteristic of childhood: they remain open to contradiction and revision in the face of contradictory evidence. This assertion is part of a larger agenda advocated by Gopnik and Meltzoff, namely that although the traditional vehicle for the study of theoretical change has been the history and philosophy of science, a more natural and ubiquitous vehicle is the study of theoretical change in children. In setting out this agenda, they draw an intriguing analogy. Prior to Darwin, horticulturists had developed some understanding of selective breeding Guided by an understanding of this process of "artificial" selection, Darwin postulated a process of natural selection, IN the wake of Darwinian theory, the study of taxonomy became less preoccupied with the demarcation of a supposedly fixed set of species, and more concerned with questions about the origin and transformation of species. Moreover, the process of artificial selection was gradually seen for what it was: a late-appearing exploitation of a mechanism that had operated blindly and unremittingly throughout evolution. Gopnik and Meltzoff conceive of the same prospect for the study of conceptual change. Science is best seen as a recent upstart — a more or less deliberate exploitation of some cognitive machinery that has been around for about 50,000 years, and is best exemplified by cognitive development in children. On this view, the study of the history of science will gradually be seen as the intellectual equivalent of studying horticulture; the mainstream study of conceptual change will no longer focus on science but on conceptual change in childhood. Are there any major obstacles to this bracing prospect? The main target of Gopnik and Meltzoff s book is cognitive development in the human child. Yet cognitive development in non-human primates displays several of the key features that they see as symptomatic of a theoretically oriented creature. To take one example, in some classic early experiments on search in the late 1920s, Otto Tinklepaugh established that monkeys would retrieve a piece of banana that they had seen placed in a box, even after a delay. In a variant of the standard task, Tinklepaugh surreptitiously removed the banana during the wait period, substituted a much less desirable piece of lettuce, and watched what happened. The puzzlement and anger of the animals when they discovered that the box contained only lettuce testified to their belief in object permanence. More generally, the development of search in non-human primates displays the sequence of errors and improvements that is found among human infants. Using Gopnik and Meltzoff's criteria, then, it is not just human children but also non-human primates that are theoretically minded. On one reading of Gopnik and Meltzoff's thesis, this conclusion is unproblematic. There is no particular reason for assuming that only human beings are capable of theory-like change in their thinking. Moreover, the empirical study of such cognitive change might benefit from a comparative dimension. Thus, we might conclude that institutionalized science fuses a natural theoretical capacity that is found in the young of various species with other specifically human practices and beliefs. By implication, if we want to understand conceptual change. we had better study it in its most natural form — in young primates, human and non-human. If we focus on conceptual change in science, we may be distracted by various cultural artefacts that sometimes bolster but also obscure the natural cognitive process. However, one might also draw a different conclusion from the primate evidence. Arguably, institutionalized science is underpinned by cognitive capacities that are distinctive among human beings, particularly children. On this argument, although primates may display those features of the theoretical mode that are emphasized by Gopnik and Meltzoff, there are likely to be other distinctively human features of that mode. Here is one speculative suggestion: unlike non-human primates, human children have a strong natural disposition to attribute observable actions — including simple acts such as reaching for an object — to hidden forces, notably beliefs, desires or intentions. More specifically, this early common-sense appreciation of psychological states involves a different kind of explanatory move from anything found in naive physics. The proposed causal powers — namely mental states — are ontologically distinct from what they are supposed to explain — namely actions and utterances. In naive physics, by contrast, the postulate of a permanent, enduring object is ontologically continuous with what it is supposed to explain, namely the object that reappears after a temporary disappearance. The disposition to postulate distinctive hidden powers might lead you to invoke gravity, alchemical forces, a deus ex machina, or all three. This line of argument implies that the study of cognitive development will offer a new perspective not just on science, but also on magic and religion,. In short, the alternative to Gopnik and Meltzoff may be even more bracing than the agenda that they have in mind. Either way, they are to be applauded for writing a book which is infectiously ambitious, even as they keep their feet on the ground.The Times Literary Supplement June 13, 1997

Table of Contents

Series Foreword xi
Preface and Acknowledgments xiii
The Other Socratic Method
Socrates' Problem
Augustine's Problem
A Road Map
I The Theory Theory 11(62)
The Scientist as Child
But Surely It Can't Really Be a Theory?
A Cognitive View of Science
Naturalistic Epistemology and Development: An Evolutionary Speculation
Science as Horticulture
Objections: Phenomenology
Objections: Sociology
Objections: Timing and Convergence
Objections: Magic
Empirical Advances
What Is a Theory?
Structural Features of Theories
Functional Features of Theories
Dynamic Features of Theories
Theories in Childhood
Theories as Representations
Theories, Modules, and Empirical Generalizations
Theories and Development
Modules and Development
Modularity in Peripheral and Central Processing
Empirical Generalizations: Scripts, Narratives, and Nets
Interactions among Theories, Modules, and Empirical Generalizations
Nonconceptual Development: Information Processing and Social Construction
II Evidence for the Theory Theory 73(114)
The Child's Theory of Appearances
The Adult Theory
The Initial Theory
The Paradox of Invisible Objects
An Alternative: A Theory-Change Account
The Nine-Month-Old's Theory
The A-Not-B Error as an Auxilliary Hypothesis
The Eighteen-Month-Old's Theory
Other Evidence for the Theory Theory
Semantic Development: "Gone" as a Theoretical Term
Later Developments: From Object Permanence to Perspective Taking
Later Semantic Developments: "Gone" and "See"
The Child's Theory of Action
The Adult Theory
The Initial Theory
The Nine-Month-Old's Theory
The Eighteen-Month-Old's Theory
Other Evidence for the Theory Theory
Semantic Development: "No," "Uh-oh," and "There"
Later Developments: From Actions to Desires
Later Semantic Developments: "Want"
The Child's Theory of Kinds
The Adult Theory
Categories and Kinds
The Initial Theory
The Nine-Month-Old's Theory
The Eighteen-Month-Old's Theory
Other Evidence for the Theory Theory
Semantic Development: The Naming Spurt
Later Developments
III Theories and Language 187(38)
Language and Thought
A Theory-Theory View
Methodological Issues: Specificity and Correlation
Developmental Relations between Language and Cognition
Theories and Constraints
Crosslinguistic Studies
Individual-Difference Studies
The Darwinian Conclusion
Who's Afraid of Semantic Holism?
A Developmental Cognitive Science
Computational and Neurological Mechanisms
After Piaget
Sailing in Neurath's Boat
Notes 225(4)
References 229(22)
Index 251

What People are Saying About This

John Campbell

Words, Thoughts, and Theories is a big, powerful book, argued with enormous vigour and lucidity. There has so far been no general statement of the 'theory theory' appraoch, no attempt to show that there is a single picture of development that can be applied in many areas. This book provides exactly what is needed. It will be compulsory reading for developmental psychologists, but ought to be of interest to cognitive psychologists generally, and to the increasing number of philosophers with an interest in developmental psychology, as well as providing some quite new questions for philosophers of science. The lucid style and clarity of organization should make it widely accessible. It is a genuinely seminal work.

Susan Gelman

This is an important and ambitious book that provides a compelling theoretical framework for understanding cognitive development. The claims are striking and turn upside-down some of our most cherished assumptions about children's thinking. It is written in a most delightful style: witty, accessible, even fun.

John H. Flavell

This is one of the ablest, most persuasive expositions of the 'theory theory' appraoch to cognitive development the field has yet seen. The authors' strong background in philosophy, developmental theory, and psychological research with both infants and young children equip them admirably to write such a book. they make a convincing case that this theory can both account for many existing developmental facts and also guide the search for new ones.

Simon Baron-Cohen

The book is astonishing in its scope and clarity. It successfully integrates philosophy, cognitive development, and cognitive science in away that has rarely if ever been done. The idea that childrendevelop theories which evolve and reorganize into newer and more powerful theories, like mini-scientists, is of course not new; but in Gopnik and Meltzoff's hands it received a thorough treatment, across such a wide rangeof domains.

Carolyn B. Mervis

This is an important and exciting book. Gopnik and Meltzoff argue that cognitive development should be thought of as theory change. While this claim has been advanced before, solid data to support the claim across different aspects of cognitive development have not been offered. In this book, Gopnik and Meltzoff offer a carefully reasoned, balanced, argument for their position that the early development of children's theories of object permanence/appearance, action, and kinds follows a path analygous to that of the development of scientific theories. Words, Thoughts, and Theories should be mandatory reading for anyone interested in cognitive development or cognitive change.

Journal of Consciousness Studie - Shaun Gallagher

Beyond the good science that they contribute to their own idea... it is surprising and wonderful how Gopnik and Meltzoff transcend their own field to demonstratethe relevance of their research to other disciplines.

From the Publisher

"Beyond the good science that they contribute to their ownidea... it is surprising and wonderful how Gopnik and Meltzofftranscend their own field to demonstratethe relevance of theirresearch to other disciplines." Shaun Gallagher , Journal ofConsciousness Studies

"The book is astonishing in its scope and clarity. It successfullyintegrates philosophy, cognitive development, and cognitive science in away that has rarely if ever been done.

The idea that childrendevelop theories which evolve and reorganize into newer and more powerfultheories, like mini-scientists, is of course not new; but in Gopnik andMeltzoff's hands it received a thorough treatment, across such a wide rangeof domains." Simon Baron-Cohen, Lecturer inPsychopathology, Departments ofExperimental Psychology and Psychiatry, University ofCambridge


The book is astonishing in its scope and clarity. It successfully integrates philosophy, cognitive development, and cognitive science in away that has rarely if ever been done. The idea that childrendevelop theories which evolve and reorganize into newer and more powerful theories, like mini-scientists, is of course not new; but in Gopnik and Meltzoff's hands it received a thorough treatment, across such a wide rangeof domains.

Simon Baron-Cohen, Lecturer in Psychopathology, Departments of Experimental Psychology and Psychiatry, University of Cambridge

Customer Reviews

Most Helpful Customer Reviews

See All Customer Reviews