Perceptual and Cognitive Development

Rochel Gelman, Terry Kit-Fong Au

• Publisher: Academic Press
• Year: 1996
• ISBN:
  9780080538624
• Topic: Psychology
• Subtopic: Neuropsychology
• Index: Psychology

Perceptual and Cognitive Development illustrates how the developmental approach yields fundamental contributions to our understanding of perception and cognition as a whole. The book discusses how to relate developmental, comparative, and neurological considerations to early learning and development, and it presents fundamental problems in cognition and language, such as the acquisition of a coherent, organized, and shared understanding of concepts and language. Discussions of learning, memory, attention, and problem solving are embedded within specific accounts of the neurological status of developing minds and the nature of knowledge.

• Research advances and theoretical reorientations are updated in the Second Edition; the revision focuses more attention on the cognitive and biological sciences and neuroscience
• Illustrates how the developmental approach can yield fundamental contributions to our understanding of perception and cognition as a whole
• Discussions of learning, memory, and attention permeate individual chapters

Cognition and Language

Chapter 4

Concepts and Theories

Susan A. Gelman

I INTRODUCTION

Concepts are fundamental to all of human experience. Naming objects, recognizing novel instances, generalizing from the known to the unknown, making inferences, and learning new information all make use of concepts. Thus, for decades psychologists have been deeply interested in the structure of concepts and their development (Bruner, Goodnow, & Austin, 1956; Bruner, Olver, & Greenfield, 1966; Inhelder & Piaget, 1964; James, 1890). In recent years, the study of concepts has undergone an important shift: whereas concepts were traditionally treated as isolated, atomic units, it is now recognized that they are interrelated and influenced by larger knowledge systems or theories (Murphy & Medin, 1985). This shift has altered current thinking about the very nature of concepts and their organization.

This chapter reviews recent arguments and evidence demonstrating the links between theories and concepts. Theories do not exist apart from concepts, and (more surprising) in many domains concepts do not exist apart from theories (Barrett, Abdi, Murphy, & Gallagher, 1993; Carey, 1985; S. A. Gelman, Coley, & Gottfried, 1994; Keil, 1989; McCauley, 1987; Murphy, 1993; Murphy & Medin, 1985). Thus, a theory of biology includes abstract concepts such as “alive” or “inheritance” and more particular concepts for individual species such as “dog” or “pterodactyl.” At the same time, concepts—both concrete and abstract—are imbued with theory. For example, a child’s concept of “alive” will reflect whether she has a mature theory of biology or instead thinks that all things that move (including clouds and bicycles) have intentions and desires (Carey, 1985).

What do we mean by “concept” or “theory”? Both are mental representations that give order to experience. Clark (1983, p. 789) defines concept as “a set of properties that are associated with each other in memory and thus form a unit.” This unit can refer to either an individual (e.g., Socrates) or a category (e.g., birds), and is often encoded in a single word (e.g., animal, mine, mama). In contrast, theories are larger structures that include interrelated sets of beliefs, causal explanations, and predictions (e.g., a theory of mind; see Taylor, this volume). Thus, theories “are sets of interconnected relations, and concepts are the units connected by these relations” (Murphy, 1993, p. 177). Although traditionally scholars have questioned whether young children are capable of forming either kind of construct (Inhelder & Piaget, 1964; Piaget, 1929), a growing body of evidence suggests that children form a rich and complex set of both concepts and tacit theories, beginning in the first year of life.

The structure of the chapter is as follows. First I discuss concepts, raising the question of which aspects change developmentally and which are fundamentally stable throughout life. Then I turn to theories, illustrating with children’s beliefs within the particular domain of biology. In the final section, I consider mutual influences of theories on concepts and of concepts on theories.

II CONCEPTS

A Conceptual Functions

Concepts¹ serve many important functions: to organize information efficiently in memory, identify novel entities, solve problems, form analogies, enable inferences that extend knowledge beyond what is already known, embody and pass down ideological beliefs, create figurative images (with metaphor and poetry), and so on (S. A. Gelman & Medin, 1993).

At least two of the basic functions of concepts are present in children even before they begin to speak: that of ordering information efficiently, and that of fostering inferences about the unknown. Efficiency is arguably needed for survival (Bruner et al., 1956; Rosch, Mervis, Gray, Johnson, & Boyes-Braem, 1976). If children were unable to categorize, their experiences would be overflowing with individual beings, objects, properties, sensations, and events too numerous to hold in memory. Fortunately, a child’s world is not so chaotic. From earliest infancy children form categories that are strikingly similar to those of adults. In their first few months of life, children form categories of speech sounds (Eimas, Siqueland, Jusczyk, & Vigorito, 1971; Mehler et al., 1988; Kuhl, 1985; Werker, 1989), faces (Cohen & Strauss, 1979; Fagan & Singer, 1979), emotional expressions (Kestenbaum & Nelson, 1990), colors (Teller & Bornstein, 1987), objects (Baillargeon, 1993; Baillargeon, Spelke, & Wasserman, 1985), and cross-modal mappings (Meltzoff & Borton, 1979; Lewkowicz & Turkewitz, 1981; Rose & Orlian, 1991), all of which preserve distinctions that adults find relevant and important. (See Spelke & Hermer, this volume, for review.) Similarly, preverbal children categorize animal species, both real and artificial (Cohen & Younger, 1983; Quinn & Eimas, 1986; Quinn, Eimas, & Rosenkrantz, 1993), on a variety of tasks including habituation, paired preference, and sequential touching (Mandler, Bauer, & McDonough, 1991). By 18 months of age, most children have begun a naming explosion, adding roughly nine new words each day to their vocabulary (Carey, 1978). Assuming that most new words encode concepts, this suggests that 1- and 2-year-old children are adept at concept acquisition.

Less work has documented the inductive function of categories, although it is equally important. Without the inferential capacity of categories, it would be necessary to determine by trial and error how to interact with every new object—to decide, for example, whether an object was suitable for wearing, reading, sitting upon, and so on. On seeing an apple one would not know whether it was edible—regardless of how many apples one had eaten in the past—because it would be an individual whose characteristics need to be discovered anew.

It is not surprising, then, that the capacity to use categories as the basis of inferences also appears to be available to very young children. Rovee-Collier and colleagues found that under appropriate conditions, 3-month-olds can form a novel category-based inference (Hayne, Rovee-Collier, & Perris, 1987). After a baby learns that one mobile shakes in response to his kicking, he infers that other mobiles also shake in response to kicking, particularly if they are also similar in other respects. Baldwin, Markman, and Melartin (1993) demonstrated that infants between 9 and 16 months of age can make specific inductions about novel actions from one category member to another. Baldwin et al. presented infants with a novel toy that produced a nonobvious and unanticipated effect (e.g., a can that wailed when tilted or shook). The subjects were then shown a second toy that closely resembled the original toy but failed to produce the interesting effect (e.g., a wailing can that had been altered so that it made no sound). In their exploratory play, the infants attempted to produce the original effect (e.g., tilting and shaking the silent can), and showed increased and persistent exploration when the second toy failed to meet their expectations. Control conditions demonstrated that children did not show the same degree of interest or exploration of the second toy when the original toy did not produce the interesting effect; nor did they show increased attention or category-specific actions when the second toy was a different kind of object altogether from the first (e.g., if the first was a wailing can and the second was a toy horn). Thus, 9-month-olds drew category-specific inferences about nonobvious object properties, based on only a single example.

B Developmental Dichotomies?

Although category functions of efficiency and induction appear early in infancy and may be unchanging developmentally, the content of concepts and their organization into larger systems are open to developmental change. Just how much change is involved has been a question of intense interest and debate. On many traditional accounts, concepts undergo a fundamental, qualitative shift with development. That is, children and adults are often said to be at opposite end points of various dichotomies, moving from perceptual to conceptual (Bruner et al., 1966), from thematic to taxonomic (Vygotsky, 1962), from syntagmatic to paradigmatic (see Nelson, 1977, for review), from concrete to abstract (Piaget, 1951), or from similarity to theories (Quine, 1977).

These developmental dichotomies are intuitively appealing, in part because each promises to capture a broad array of changes in a simple model. Furthermore, in many contexts children do behave in ways that would support one or another of these distinctions.² For example, in the well-known conservation error studied by Piaget, children below age 6 or 7 years of age report that an irrelevant transformation leads to a change in quantity (e.g., the volume of liquid in a container increases when the water is poured into a taller, skinnier container; Piaget, 1952). Children appear to focus on one salient but misleading dimension (e.g., height), forgoing a deeper conceptual analysis. Throughout the past several decades there have been many demonstrations that “the preschool-age child is prone to accept things as they seem to be, in terms of their outer, perceptual, phenomenal, ‘on-the-surface’ characteristics” (Flavell, 1977, p. 79).

However, as an account of what children are capable of doing, developmental dichotomies are inadequate (Bauer & Mandler, 1989; R. Gelman, 1978; Gibson & Spelke, 1983; Markman & Hutchinson, 1984; Nelson, 1977). With appropriately sensitive tasks, children sometimes can display competencies that do not emerge in their everyday performance (S. A. Gelman, 1994). For example, Smiley and Brown (1979) found that although first grade children preferred to group pictures thematically (e.g., putting a spider with a web), they were able to sort taxonomically (e.g., putting a spider with a grasshopper) when asked whether there was a different way of answering. The children even provided sensible justifications for their taxonomic choices.

The past twenty years have yielded a greater understanding of children’s concepts not only because of methodological advances, with the development of more sensitive techniques for assessing children’s knowledge (R. Gelman & Baillargeon, 1983), but also because of changing assumptions about the nature of concepts. According to most developmental dichotomy positions, children’s concepts are uniform across domains and across contexts (e.g., uniformly “concrete” or “perceptual”), and adults’ concepts have an abstract, logical structure that is lacking in children. More recently, both of these assumptions have been questioned. Specifically, there is now compelling evidence to suggest that: (1) children’s concepts are domain specific and context dependent; (2) the internal structure of children’s categories often resembles that of adults’ categories; (3) for both children and adults, categories are arranged into hierarchical systems that include abstra...

Table of contents

• Cover image
• Title page
• Table of Contents
• Copyright page
• Contributors
• Foreword
• Preface
• Perception and Cognition
• Cognition and Language
• Models of Development
• Index