Culture, Thought, and Development

Larry Nucci, Geoffrey B. Saxe, Elliot Turiel, Larry Nucci, Geoffrey B. Saxe, Elliot Turiel

• Publisher: Psychology Press
• Year: 2000
• Print ISBN: 9780805830095
• eBook ISBN: 9781135676988
• Topic: Psychology
• Subtopic: Developmental Psychology
• Index: Psychology

In this volume, the reader will find a host of fresh perspectives. Authors seek to reconceptualize problems, offering new frames for understanding relations between culture and human development.

Contributors include scholars from the disciplines of philosophy, law, theology, anthropology, developmental psychology, neuro- and evolutionary psychology, linguistics, cognitive science, and physics. To help organize the discussions, the volume is divided into three parts. Each part reflects an arena of current scholarly activity related to the analysis of culture, cognition, and development.

The editors cast a wide but carefully crafted net in assembling contributions to this volume. Though the contributors span a wide range of disciplines, features common to the work include both clear departures from the polemics of nature-nurture debates and a clear focus on interacting systems in individuals' activities, leading to novel developmental processes. All accounts are efforts to mark new and productive paths for exploring intrinsic relations between culture and development.

Part I
EPISTEMOLOGICAL ISSUES

Chapter 1
Would You Say Developmental Psychology Was a Science? The Cultural Paradigm of Mind

Joseph Margolis

Temple University

ANTHROPOMORPHIZING PRELINGUISITIC THOUGHT

We are, I think, at an extraordinary place in our theorizing about the human mind—a fortiori, about any mind, no matter how different it may be from ours. Piaget is the great initiator of our best contemporary efforts to understand the development of intelligence in children from the earliest prelinguistic and “preoperational” stages (which are not the same, as Piaget makes clear) to the dawning of linguistic competence and the maturation, within its terms, of the fullest forms of rational and logical order that we know. About that continuum, I have a great many qualifications to insist on, most of which, I daresay, Piaget would have opposed; although it is my honest belief—if you do not mind the impertinence—that he would have been well advised to come to terms with most of them. There! I cannot displease you more by the details of my own theory than by this frontal confession: I feel sure you will allow me to continue as a self-acknowledged deviationist without portfolio; there cannot be much harm in it because it is hardly meant to misrepresent Piaget’s own doctrine.

What I think is extraordinary about our theorizing about the mind’s development, working at the end of the 20th century, is that we are poised rather uncertainly between the poles of Piaget’s original inquiries about its earliest stages and the recent flowering of new inquiries into artificial intelligence and the machine simulation of intelligence. Now, it is at that precise point that I intrude a suggestion for the largest possible departure—the largest I can imagine—from Piaget’s very sane and sensible observations about prelinguistic children. Here, I claim, the description of prelinguistic thought in children, not at all unlike the description of the minds of sublinguistic animals, machines, and even the biology of the brain and body thought to be implicated in what we are pleased to call the development of the mind, is always anthropomorphized and cannot but be anthropomorphized, in the sense that such descriptions are modeled on the paradigm of mental life—the reflexive, self-conscious, linguistically apt capacity of encultured human selves (ourselves) in the way of reporting what they are conscious or aware of.

I do not mean by this that there is no nonlinguistic thinking or consciousness; or that there is none among the linguistically apt; or that the prelinguistic cannot intelligibly be said to exhibit processes very different from what obtain among linguistically apt adults or in the stages of early language acquisition; or that there are no significant idiosyncratic variations at both the prelinguistic and linguistic stages; or that there are no important species-specific differences between prelinguistic human and sublinguistic animal intelligence; or that the successful machine simulation of the human mind rightly counts as an accurate description of human mental processes. No. I mean only that, in observing and describing the intelligence of prelinguistic children (and the rest), we cannot escape the endogenous constraint of having to subtract from, deform, or qualify, what, reflexively, we theorize obtains at the languaged and “operational” level at which we describe ourselves—when we describe the minds of children or other creatures or machines or, indeed, the biology or subconscious processes of the mind.

In a word, psychology rightly proceeds top down, although most of the prevailing models—neurocomputational or simply materialist—favor bottom up strategies. (See Sellars, 1963; Churchland, 1989.) I do not say psychology is necessarily top down, only that every known bottom-up exemplar fails to make its case. Top-down conceptions win faute de mieux. At any rate, that is the essential focus of dispute in the whole of psychology at the end of the 20th century. Piaget, I should say, was a top-down theorist who, however, reversed the priorities between children and adults: his principal work does not concede that the development of the child’s intelligence must be seen, top down, from the reporting vantage of our reflexive description of ourselves. That is the nerve of Piaget’s structuralism (see Piaget, 1971a).

To be sure, this is not an innocent or modest adjustment; it is not meant to be. There are enormous, wide-ranging consequences that follow from this single confession. For instance, with this argument, a modular theory of mind proves impossible: I mean one such as the innatist doctrine Chomsky (1980a) and Fodor (1983) proposed (in somewhat different ways). I should say that Chomsky’s notion of a universal grammar cannot rightly be modular, can only be “homuncular” (or modular in the homuncular way); although saying that hardly disallows some forms of innatism. I may as well say that I have the gravest doubts about Chomsky’s universal grammar—which, of course, cannot be what it is said to be if strict modularism fails. Cognitive intelligence cannot be completely modular in any one creature; that would be a contradiction in terms. On the anthropomorphizing thesis, it cannot but fail, because the posit of a modular grammar will require positing a whole series of associated modules, from the conjoint functioning of which the seemingly molar functioning of an integrated human subject would first emerge or be derived. By homuncular, by contrast, I mean precisely the subfunctioning of any would-be module (or nonmodular process) as a relationally defined part of some molar functioning, without reference to which it (the subfunctional function) designates nothing. I believe it to be a very strong intuition of Piaget’s that we cannot map the development of the child’s intelligence except by internalizing in constructivist terms the mental import of the child’s overt bodily activity. (See Piaget & Inhelder, 1969.)

With that much I wholly agree. I go on to insist, however, that the modeling of the child’s molar intelligence is itself anthropomorphized in terms of the paradigm of “selfconscious” mind: our reflexive reporting capacity. Even at the adult level, of course, it is clear (in the reporting sense) that the human mind is not a mere language machine. Beyond that, I permit myself two inferences against certain fairly well-known views in the philosophical literature regarding the nature of mind. Nagel (1979), for one, is very well known for having raised the question, “What is it like to be a bat?” (that is, seen from the “inside”). Nagel intended to draw attention to the puzzle implicit in admitting minds in the first place (which, he believed, could not be adequately characterized in physical terms) and the impossibility of characterizing the mind of the bat conformably (because we are not bats). Nagel, I should say, was mistaken in his second claim: bats do not know what it is like to be a bat; only humans do! That is the lesson “anthropomorphizing” affords. Only human adults know what it is like to be a prelinguistic child. Piaget (I think) ultimately agrees, although this seems much clearer in his later than in his earlier work. It is certainly not compatible with his structuralism, strictly construed.

Again, if this picture be correct, then, as far as we understand machine intelligence, it must be the case that human intelligence cannot be adequately simulated by an intelligent machine because, of course, machine intelligence is thoroughly modular and human intelligence cannot be. If that is so, then computational models of the mind as different as those advanced by Dennett and Churchland (familiar, I am supposing, to devel-opmental psychologists) must be utterly mistaken, although that is not to say that we cannot admit modular subfunctions keyed to parts of our neurophysiology (parts of the cerebellum, for instance) within the homuncular model that I have already acknowledged may be needed. It goes without saying that, apart from simulation, the human mind cannot be modeled neurocomputationally. (See Dennett, 1969; Churchland, 1989; Edelman, 1987).

You may, of course, contest this strong verdict—to the effect that the human mind cannot be adequately simulated or analyzed in machine terms. I defer to some extent to your objection. I do believe, however, that the verdict is the correct one, but it would take us too far afield to draw out the necessary argument. The matter depends on what we make of the molar powers of encultured selves (or encultured minds). I myself make a great deal of that, too much in fact for any mere intelligent machine to acquire, unless it already begins to be an android of some sort—capable of being encultured in the way human infants are.

Let me mention, then, two insuperable difficulties to fix our ideas against the reductive and eliminative strategies: for one, the well-known “frame” problem; for another, the problem of explaining reference and predication in mental terms. As I see matters, these are very closely related problems. I shall touch on the second only, in order to make matters clearer—but not yet. In any case, there is no algorithmic solution to either problem, which is precisely why machine simulation and analysis are limited in the way they are. If machines functioned in a molar way akin to the human mode of functioning, then (on the argument) they could not function computationally. If their functioning were completely modular and computational, they could not but be subaltern (homuncular) modes of functioning, and they could never adequately model human functioning. If that be so, then strong nativisms, such as Chomsky’s and Fodor’s, must fail because they too are modular, which is to say, modeled on the structure of machine simulation. The only possible way to turn the tables would be to provide a completely modular—hence, a computationally conceivable—analysis of the cognitive powers of the human mind. But if no (psychological or cognitive) module were cognitively penetrable from another would-be module internal to the same creature, the model proposed would instantly fail to capture what was most characteristic of human cognition functioning at the molar level. You see the heroic difficulty the bottom-up theorists embrace.

Part of my argument, I must concede, is directed against Piaget. Although Piaget rightly opposes Chomsky’s innatism, chiefly (as I understand it) because Chomsky does not distinguish with sufficient care between the prelinguistic and preoperational aspects of the child’s development (about which I thoroughly agree), Piaget himself is not altogether convincing in his efforts to explain and justify the necessary concatenation of the constructed stages of prelinguistic intelligence. If you permit the observation, I fear you will also have to concede that Chomsky will have very cleverly caught Piaget, in their well-known debate, when Chomsky (1980b) questioned Piaget about the invariant sequence of the structured stages of childhood development. The weakness of Piaget’s argument is partly due to conceptual difficulties in his account of genetic structuralism, particularly with respect to the relationship between observational findings and structuralist interpretations (see Piaget, 1970), and partly due to the fact that there is no empirical way to construe the structuralist (the necessary) sequence of mental development except in innatist terms (that is, where environmental factors merely trigger the interior evolution of innate modules).

If Piaget had held to something like the homuncular model, he would have been obliged to concede an ampler and more variable development of the self’s conceptual powers; conceding that, he would have drawn back from a strict structuralism and admitted the nonmodular (hence, significantly interactive) powers of molar selves within their environing worlds. I think this cannot be very far from Bruner’s objection to Piaget (Bruner, 1983, ch. 8), for instance. Certainly it bears on whether, and when, a child can escape “egocentrism.”(See Boden, 1979, ch. 3; Bruner, 1975; Vygotsky, 1962.)

My own objection is a little simpler. The very effort to map the developmental intelligence of the child proceeds in the reverse direction from that development; that is, it must be anthropomorphized, described, top down, in terms of the child’s gradual “transformation” into a linguistically apt “self” (the functional “site,” let us say, of such reflexively apt functioning). This is not to urge the exclusive importance of the purely linguistic, but emphasizes, instead, the peculiarity of developmental psychology viewed as a science.

If it is possible to begin to grasp the viewpoint of another, even at the sensorimotor level, as Bruner supposes, then Piaget must be seriously mistaken in his treatment of the relationship between structuralist and empirical work. It is just such a consideration that supports my remarks on modular and homuncular explanations and on the ubiquity of anthropomorphizing. What I am saying is this. First, it may, contrary to what Piaget supposes, be possible for children at the sensorimotor stage to begin to understand the viewpoint of another—which makes sense only if the modularist orientation is mistaken. Second, even the egocentric development of the child is, to some extent, an artifact of our anthropomorphizing, which goes against the strict autonomy of the structuralist method, particularly the reading of the evolution of intelligence. Third, the very formation (or “construction”) or “morphogenesis” of the child’s intelligence entails forms of collective enculturation, whether linguistic or prelinguistic, whether sensorimotor or more advanced, that require a less tendentious methodology than Piaget allows. In short, Piaget’s strict structuralist analyses of intelligence determine what we may admit, empirically, in the study of the child; whereas the anthropomorphizing model concedes, quite frankly, that there is no empirical access to the child’s cognitive development save through the anticipatory descriptions uniquely accessible to a society of already competent selves.

If one calls the modular model into question, one calls into question as well the evolutionary autonomy of the sequence of interactional processes between organism and environment—at least as far as intelligence goes—extending even to the pertinent developmental stages of the embryo. Piaget’s strenuous interest in bringing his research to bear on the resolution of the Lamarckian/Darwinian quarrel may have obscured for him the ultimate disanalogy between the development of mind and the development of body. (See Piaget, 1971c; Waddington, 1957.) For, if one admits the sui generis nature of cultural formation, the would-be “ontogenetic” variations of the would-be “phylogenetic” evolution of the mind may mask the entirely different historical and culturally constructive processes by which a biologically gifted child is transformed into a linguistically and socially apt self.

THE PARADIGM OF THE MENTAL

Certainly, Piaget’s paramount contribution (which I cannot pretend to assess with authority) is centered in the interplay between biological and cultural processes of “structuring” the intelligence of the prelinguistic and preoperational child. There can be no question that Piaget’s bias in favor of the self-regulating processes of epigenesis has alerted him in an unusually astute way to certain essential features of the earliest (sensorimotor) intelligence of the pre-encultured child. (I use the term bias deliberately here, but not with malice or derogatory intent.) Nevertheless, there are at least two fundamental objections to be lodged against Piaget’s general treatment of the development of intelligence. One I have already explored: namely, the inescapably anthropomorphizing control of the description and explanation of the prelinguistic child’s development. The other answers to the complexities of how we should understand cognition itself and why it is that the paradigm of the mental and the intelligent cannot but be formulated in terms of the paradigm of cognitive states.

The model of consciousness, I should say, is self-consciousness; and self-consciousness is, paradigmatically, the competence of human selves to report (linguistically) what they are conscious or aware of. Actually, it is the second theme that drives the first; that is, it is the epistemic paradigm of the mental that ensures the inevitability of anthropomorphizing the description of the intelligence of prelinguistic children. For, if you feature the cognitive side of the mental, you realize that you cannot address the mental life of children, animals, or machines except in terms of a linguistic model. Only if the cognitive could be reduced behaviorally, neurophysiologically, or computationally, could the irreducibility (hence the anthropomorphizing) of the cognitive be retired. Piaget obviously does not believe that that is possible—and, if I may say so, neither do I.

But it is one thing to admit irreducibility; it is quite another to insist on the necessary —structuralist—ingredients of knowledge itself (for instance, the alleged necessities of logic). Piaget (1971a, 1971b) makes a great deal of the inherent inadequacy of the contingent, descriptive, or empirical features of cognition as opposed to the underlying, necessary and necessarily linked, structuralist order of cognitive development. But if you feature the culturally constructed nature of the self, formed by the child’s internalizing the language and collective practices of its environing society, then the import of treating the mental in terms of what is paradigmatically cognitive (which I take to be well-nigh ineluctable and which Piaget plainly favors) is at once that there cannot be a principled disjunction between the empirically descriptive and the supposedly necessary structuralist dimensions of cognition itself. If what is paradigmatic of the “minds of selves” is culturally constructed and if the description of the mental is anthropomorphized in terms of that paradigm (wherever the mental is itself other than encultured or enlanguaged), then (a) there are no strict necessities of the mental (in the epistemologist’s sense); (b) the enculturing and cultural features of the mental cannot be captured by any extension of the biological or biologically grounded evolution of organisms or minds; and (c) whatever we posit as the sensorimotor, prelinguistic, and “intermediary” or “intuitive” stages of a child’s cognitive development cannot fail to be an artifact of our anthropomorphizing tertia.

I do not deny that Piaget makes concessions in this general direction, and I have no interest in entering the turf quarrels between the Piagetians and the non-Piagetians. But there is a decisive weakness in Piaget’s account of the social life of the child, which, rightly understood, leads in the direction of correcting (if I may say so) his general theory. You will find a tactful sense of Piaget’s tendency to scant the social—in friendly efforts toward rehabilitating his doctrine, in the face of a long history of objections. For example, in an effort to counter the often unfavorable reading of Piaget’s reference to the social dimension of learning, particularly after he formulated a full picture of sensorimotor development, Lourenço and Machado (1996, p. 150) emphasized that, according to Piaget, “cognitive structures and operations come from the subject’s own coordination and self-regulation”; and that, in spite of that, Piaget characteristically insists that, were it not for the social milieu of learning, “the individual would never achieve complete conservation and reversibility.”

Now, I think there cannot be any doubt that when he speaks of the social context in which the child develops, prior to acquiring language and in originally acquiring language and operational competences “egocentrically,” Piaget believes the child’s development may well be facilitated (“triggered”) socially, although the learning process is not itself (for that reason) essentially social—in the distinct sense of being inherently socialized, encultured, or characterizable as the mastery of competences that are collective or consensual and not reducible to any “compound” of prelinguistic competences. (See Piaget, 1950.) That, as I say, is what Chomsky astutely noticed.

My own theory insists on the primacy of the cultural—in epistemic terms. The paradigm of the mental (I argue) is and must be reflexively specified and hence is characterized in terms of our reporting—our cognitive—competence. But that same competence (what, minimally, we mean by self-consciousness) is an artifact of enculturation, the acquisition, in infancy, of a natural language and the emergence of a “self.” A self, then, is simply (as matters now stand, because we know no extraterrestrials) any member of Homo sapiens transformed or “second-natured” by the acquisition of a collectively shared language and culture—that is, a language and culture already mastered by an aggregate of apt adult selves. (See Margolis, 1997.) But in saying that, I do not deny at all that prelinguistic infants and nonlinguistic animals exhibit distinctive forms of intelligence and learning. I say only that what they exhibit is attributable under the reflexive paradigm.

What this shows is that Piaget has no clear way to demonstrate or dispute whether or to what extent the incipient enculturation of the human infant—quite apart from social facilitation—is an essential ingredient in any and every stage in the process of cognitive development. The argument is not (or not simply) that Piaget scants the social, but rather ...

Table of contents

• Cover Page
• Title Page
• Copyright Page
• Preface
• Part I Epistemological Issues
• Part II Personal, Social, and Affective Development
• Part III The Development of Physical and Spatial Knowledge