The Cultural Origins of HUMAN COGNITION
Harvard Univ. Press 2000
Keywords: - innate cognitive modules - cognitive skills of language and complex mathematics are like chess: they are the products of both historical and ontogenetic developments working with a variety of preexisting human cognitive skills, some of which are shared with other primates and some of which are uniquely human - my account for how a single human cognitive adaption could result in all of the many differences in human and nonhuman primate cognition is that this single adaptation made possible an evolutionarily new set of processes, that is, processes of sociogenesis - symbolic forms of communication to emerge - human cognitive ontogeny - imitation - The new and powerful forms of social cognition that result open up the cultural line of human development in the sense that children are now in a position to participate with other persons in joint attentional activities and so to understand and attempt to reproduce their intentional actions involving various kinds of material and symbolic artifacts. And, indeed, this tendency to imitatively learn the actions of other persons is a very strong one - Children's mastery of one very special cultural artifact-language-has transforming effects on their cognition. Language does not create new cognitive processes out of nothing, of course, but when children interact with other persons intersubjectively and adopt their communicative conventions, this social process creates a new form of cognitive representation - all, the tired old philosophical categories of nature versus nurture, innate versus learned, even genes versus environment are just not up to the task - they are too static and categorical - if our goal is a dynamic Darwinian account of human cognition in its evolutionary, historical, and ontogenetic dimensions
In general, in my opinion, many theorists are much too quick to explain uniquely human cognitive skills in terms of specific genetic adaptations - typically without any genetic research, it should be added. It is a popular procedure mainly because it is so quick, easy, and unlikely to be immediately refuted by empirical evidence. But another important reason for many theorists' tendency to posit innate cognitive modules as a method of first resort is a lack of appreciation of the workings of human cultural-historical processes, that is, processes of sociogenesis, both in the sense of their direct generative powers and in the sense of their indirect effects in creating a new type of ontogenetic niche for human cognitive development. And, importantly, historical processes work on a completely different time scale than evolutionary processes (Donald, 1991).
Let us take as an example the game of chess. The children who learn to play this game do so in interaction with mature players, and some of them develop quite sophisticated cognitive skills in the context of this game, many of which would seem to be domain specific in the extreme. A cognitive psychologist can only marvel at the complex plannings and imaginations required for orchestrating a king side attack in which the opposing king's pawn protectors are first eliminated by means of a bishop sacrifice, and then the king's movements are restricted and the attack is consummated with the coordination of knight, rook, and queen. Despite the cognitive complexities involved, and despite the domain specificity of the cognitive skills involved, I have never seen anyone posit an innate chessplaying module. The reason is that chess is a very recent product of human history, and there are even books with pictures that trace its historical development. Chess was originally a simpler game, but as the players came to some mutual understandings of things that would make the game better, they modified rules or added new ones undl they produced the modern gamefor which children today can, over the course of a few years of play and practice, develop quite impressive cognitive skills. Of course chess does not create in children basic cognitive skills such as memory, planning, spatial reasoning, categorizationthe game could evolve only because human beings already possessed these skillsbut it does channel basic cognitive processes in new directions, helping to create some new and very specialized cognitive skills as a result.
My contention is simply that cognitive skills of language and complex mathematics are like chess: they are the products of both historical and ontogenetic developments working with a variety of preexisting human cognitive skills, some of which are shared with other primates and some of which are uniquely human. These are easiest to see in the case of mathematics because-and this is somewhat like chess -
(a) we can trace much of the historical development of modern mathematics in the last 2,000 years,
(b) in many cultures the only mathematical operations used are very simple counting procedures (and their arithmetic variants), and
(c) within cultures using cplex mathematics many individuals only learn some simple procedures.
These facts thus constrain the possibilities, so that modularity theorists can posit as a mathematics module something that contains only the most basic of quantitative concepts.
In the case of language, however,
(a) we know very little of its history (only the relatively recent history of the few languages that have been written down),
(b) all cultures have complex languages, and
(c) all typically developing children within a culture acquire basically equivalent linguistic skills.
These facts make it clear that language is different from mathematics and chess, but they do not specify the reason for this difference. It may just be that language, for whatever reason, began its historical development first - early in the evolution of modern humans some 200,000 years ago - and so reached something near its current level of complexity before modern languages began to diverge from this prototype. If we may use ontogeny as any guide to cognitive complexity, modern children begin using natural languages with much sophistication well before they master complex mathematics or chess strategies. Perhaps the reason that language is cognitively primary is that it is such a direct manifestation of the human symbolic ability, which itself derives so directly from the joint attentional and communicative activities that the understanding of others as intentional agents engenders. The point is thus that language is special, but not so special.
And so my account for how a single human cognitive adaption could result in all of the many differences in human and nonhuman primate cognition is that this single adaptation made possible an evolutionarily new set of processes, that is, processes of sociogenesis, that have done much of the actual work and on a much faster time scale than evolution. Perhaps this single novelty changed the way human beings interacted with one another, and with much effort over much historical time these new ways of interacting transformed such basic primate phenomena as communication, dominance, exchange, and exploration into the human cultural institutions of language, government, money, and science - without any additional genetic events. The transformations in the different domains of human activity as a result of this new adaptation clearly were not instantaneous. For example, human beings were already communicating with one another in complex ways when they began to understand one another as intentional agents, and so it took some time, perhaps many generations, for this new understanding of others to make itself felt and thus for symbolic forms of communication to emerge. The same would have held true for the other domains of activity such as various forms of cooperation and social learning - as this new kind of social understanding gradually enabled new kinds of social interactions and artifacts.
Table 7.1 presents an oversimplified and certainly not exhaustive listing of some domains of human activity and how they might have been transformed by the uniquely human adaptation of social cognition as it worked itself into various social interactive processes over many generations of human history.
Ideally we should know much more than we do about the process of sociogenesis in different domains of activity in human history. Cultural psychologists, who should be concerned with this problem, have mostly not spent great effort in empirical investigations of the historical processes by means of which particular cultural institutions in particular cultures have taken shape - for example, processes of grammaticization in the history of particular languages or processes of collaborative invention in the history of the mathematical skills characteristic of a particular culture. Perhaps the most enlightening investigations of these processes are studies by intellectual historians concerned with such things as the history of technology, the history of science and mathematics, and language history (see Chapter 2). But these scholars are mostly not concerned with cognitive or other psychological processes per se, and so the information to be gleaned by psychologists is decidedly indirect. And there are perhaps some relevant facts to be gleaned from studies of cooperation in which two partners who are naive to a problem domain manage to collaboratively invent some new artifact or strategy - in a manner analogous to processes of cultural creation in historical time (see Ashley and Tomasello, 1998).
In all, we may underscore the power of sociogenesis by proposing a variation on our recurrent theme of a wild child on a desert island. In this case, let us suppose that a giant X-ray comes down from outer space and makes all human beings over one year of age profoundly autistic - so much so that they cannot intentionally communicate with one another or with the infants (although, miraculously, they are able to provide the infants with sustenance and protection). So the one-year-olds are left to their own devices to interact with one another (Lord of the Flies style), with the hulking infrastructure of modern technology rusting in the background (Mad Max style). The question is: How long would it take for the children to re
so on? I am certain that there are scholars who think it would take place almost immediately, especially in the case of language, but I believe that this is a naive view that seriously underestimates the historical work that has gone into these institutions as they have ratcheted up in complexity over many generations historically. (And studies of children creating gestural signs in interaction with language
Ontogeny is a very different process for different animal species. For some species it is important that their young be almost fully functional from the time they encounter the outside world, to maximize their chances of surviving to the age of reproduction, whereas for other species a long ontogeny, with much individual learning, is the life-history strategy of choice. Learning is thus a product of evolution - one of its strategies, if we may anthropomorphize the process a bit - as are culture and cultural learning as special cases of the "extended ontogeny" evolutionary strategy. There is thus no question of opposing nature versus nurture; nurture is just one of the many forms that nature may take. The question for developmentalists is therefore only how the process takes place, how the different factors play their different roles at different points in development. At birth human infants are poised to become fully functioning adult human beings: they have the genes they need and they are living in a prestructured cultural world ready to facilitate their development and actively teach them things as well. But they are not at that point adults; there is still more work to be done.
It is important to note that human cognitive ontogeny is not a replay of chimpanzee ontogeny with a "terminal addition" on the end. As I argued in Chapter 3, human cognitive ontogeny is unique from very early on, perhaps from birth, as human neonates do various things that demonstrate a special form of identification with conspecifics (e.g., neonatal mimicking and protoconversations). This is the uniqueness from which all else flows, as it enables infants to exploit a novel source of information about other persons: the analogy to the self. At around nine months of age, analogizing self and other persons enables infants to attribute to other persons the same kinds of intentionality in which they themselves are just beginning to engage (and they may also analogize to the self, somewhat inappropriately, in their causal reasoning about why inanimate objects behave as they do). The new and powerful forms of social cognition that result open up the cultural line of human development in the sense that children are now in a position to participate with other persons in joint attentional activities and so to understand and attempt to reproduce their intentional actions involving various kinds of material and symbolic artifacts. And, indeed, this tendency to imitatively learn the actions of other persons is a very strong one, as young children sometimes imitate adult actions with objects even when they would do better to ignore them, and in language acquisition they have a long period where they essentially reproduce exactly the relational structure of the adult utterances they are hearing. This is the cultural line of development at its strongest, and it is the reason why four-year-olds in different cultures are so different from one another in terms of the specific behaviors in which they engage. But throughout this early period, and even more strongly later, children are also making individual judgments, decisions, categorizations, analogies, and evaluations-more or less from the individual line of development-and these interact in interesting ways with children's tendencies in the cultural line of development to do what the other persons aronnd them are doing.
Children's mastery of one very special cultural artifact-language-has transforming effects on their cognition. Language does not create new cognitive processes out of nothing, of course, but when children interact with other persons intersubjectively and adopt their communicative conventions, this social process creates a new form of cognitive representation - one that has no counterpart in other animal species. The novelty is that linguistic symbols are both intersubjective and perspectival. The intersubjective nature of human linguistic symbols means that they are socially "shared" in a way that animal signals are not, and this forms the pragmatic matrix within which many inferences about the communicative intentions of other people may be made-why they chose one symbol rather than another that they also share with the listener, for example. The perspectival nature of linguistic symbols means that as children learn to use words and linguistic constructions in the manner of adults, they come to see that the exact same phenomenon may be construed in many different ways for different communicative purposes depending on many factors in the communicative context. The linguistic representations thus formed are free of the immediate perceptual context not just in the sense that with these symbols children can communicate about things removed in space and time, but also brain," that this radically new and powerful form of cognitive representation emanates not from any new storage facilities or computing power inside the human brain, but rather from the new forms of social interaction, enabled by new forms of social cognition, that take place between individuals inside human cultures.
Language is also structured to symbolize in various complex ways events and their participants, and this is instrumental in leading children to "slice and dice" their experience of events in many complex ways. Abstract linguistic constructions may then be used to view experiential scenes in terms of one another in various analogical and metaphorical ways. Narratives add more complexity still, as they string together simple events in ways that invite causal and intentional analysis, and indeed explicitly symbolized causal or intentional marking, to make them coherent. And extended discourse and other kinds of social interactions with adults lead children into even more esoteric cognitive spaces, as they enable them to understand conflicting perspectives on things that must be reconciled in some way. Finally, the kind of interaction in which adults comment on children's cognitive activities, or instruct them explicitly, leads children to take an outsider's perspective on their own cognition in acts of metacognition, self-regulation, and representational redescription, resulting in more systematic cognitive structures in dialogical formats. Whether or not different languages do these things differently, as in classic arguments about "linguistic determinism," learning a language or some comparable form of symbolic communication-as opposed to not learning one at all - seems to be an essential ingredient in human intersubjectivity and perspectival cognition, event representation, and metacognition.
I believe that this is what all of the thinkers quoted at the beginning of this chapter, each in his own way and with different specifics than are in the current argument, were attempting to articulate when they made their various claims to the effect that human thinking is essentially operating with symbols. Human beings can of course think without symbols if by thinking we mean perceiving, remembering, categorizing, and acting intelligently in the world in ways similar to other primates (Piaget, 1970; Tomasello and Call,1997). But the uniquely human forms of thinking- for example, those in which I am engaged as I formulate this argument and attempt to anticipate the dialogic responses it will elicit from other thinkers (and perhaps my response to those responses)- do not just depend on, but in fact derive from, perhaps even are constituted by, the interactive discourse that takes place through the medium of intersubjective and perspectival linguistic symbols, constructions, and discourse patterns. And it is not unimportant that an individual can gain mastery in the use of such symbols and their concomitant ways of thinking only over a period of several years of virtually continuous interaction with mature symbol users.
And so, like evolution and history, ontogeny really matters. Human beings have evolved in such a way that their normal cognitive ontogeny depends on a certain kind of cultural environment for its realization. The importance of biological inheritance in the ontogenetic process is underscored by the problems of children with autism, who do not have in its full-fledged form the human biological adaptation for identifying with other persons, and so do not end up as normally functioning cultural agents. The importance of cultural inheritance in the ontogenetic process is underscored by the many cognitive differences that exist among the peoples of different cultures and by the unfortunate cases of neglected or abused children brought up in culturally deficient circumstances, but it is highlighted even more if we imagine the cognitive development of children growing up without any culture or language at all. A child raised on a desert island without human companions would not come out as Rousseau envisioned, a "natural" human being free of the constraints of society, but rather would come out as Geertz envisioned, something of a monster, something other than a fnlly human intentional and moral agent.
Focus on Process
We are, as Wittgenstein (1953) and Vygotsky (1978) saw so clearly, fish in the water of culture. As adults investigating and reflecting on human existence, we cannot take off our cultural glasses to view the world aculturally-and so compare it to the world as we perceive it culturally. Human beings live in a world of language, mathematics, money, government, education, science, and religion-cultural institutions composed of cultural conventions. The sound "tree" stands for what it does because, and only because, we think it does; men and women are married because, and only because, we think they are; I can obtain a car in exchange for a piece of paper because, and only because, we think the paper is worth as much as the car (Searle, 1996). These kinds of social institutions and conventions are created and maintained by certain ways of interacting and thinking among groups of human beings. Other animal species simply do not interact and think in these ways.
But the human cultural world is not thereby free of the biological world, and indeed human culture is a very recent evolutionary product, having existed in all likelihood for only a few hundred thousand years. The fact that culture is a product of evolution does not mean that each one of its specific features has its own dedicated genetic underpinnings; there has not been enough time for that. A more plausible scenario is that all human cultural institutions rest on the biologically inherited social-cognitive ability of all human individuals to create and use social conventions and symbols. However, these social conventions and symbols do not wave a magic wand and turn nonhuman primate cognition into human cognition on the spot. Modern adult cognition of the human kind is the product not only of genetic events taking place over many millions of years in evolutionary time but also of cultural events taking place over many tens of thousands of years in historical time and personal events taking place over many tens of thousands of hours in ontogenetic time. The desire to avoid the hard empirical work necessary to follow out these intermediate processes that occur between the human genotype and phenotype is a strong one, and it leads to the kinds of facile genetic determinism that pervade large parts of the social, behavioral, and cognitive sciences today. Genes are an essential part of the story of human cognitive evolution, perhaps even from some points of view the most important part of the story since they are what got the ball rolling. But they are not the whole story, and the ball has rolled a long way since it got started.
In all, the tired old philosophical categories of nature versus nurture, innate versus learned, even genes versus environment are just not up to the task - they are too static and categorical - if our goal is a dynamic Darwinian account of human cognition in its evolutionary, historical, and ontogenetic dimensions.
HOME BOE SAL TEXTE