4. Language and Metacognition / Theory of Mind
A. How does cognitive architecture inform and constrain language acquisition? Does it make sense to talk of a “language of the mind”, or a “universal grammar”? What do these concepts actually mean, and what implications do they have? Do the detailed features of the language that we speak also have an impact on our cognitive architecture, as Sapir and Whorf believed?
We know that human cognitive architecture must constrain language acquisition in some way, because the linguistic output of children is underdetermined by their linguistic inputs: they do not simply parrot well-formed sentences that they have already heard, but are able to construct novel well-formed sentences for themselves from an early age. The question of exactly what constraints exist in the cognitive architecture is controversial, however. Chomsky has recently retreated from his earlier position that certain grammatical universals were encoded in the brain, and now takes the minimalist view that the only truly unique feature of human language may be recursion – the ability of syntactic constructions to nest inside one another, allowing a theoretically infinite range of constructions to be produced – and even recursion may be shared with mathematical reasoning. Along with Hauser and Fitch, he has argued that most or all other elements of the cognitive architecture which contribute to language – the sensory-motor and conceptual-intentional systems – are basically shared with nonhuman animals. Other leading figures such as Pinker and Jackendoff disagree, viewing language as enabled by a system of domain-specific adaptations in the human brain. However, it is noticeable that even they do not talk much of syntactic universals, but spend more time discussing areas of language such as phonology and morphology.
Pinker has written in the past (especially in his 1994 book The Language Instinct) of a “language of the mind” or a “language of thought”, but like “universal grammar”, this is a phrase that does not seem to come up as much as it used to. The basic idea was that information was encoded in a sort of private, universal language within the mind/brain, before being translated into a natural language for transmission to other people. Perhaps with the growing influence of connectionist models on psychology and the realisation that the brain does not work like a digital computer, this metaphor lost a lot of its appeal. In any case it is noticeable that Pinker and Jackendoff now write of language being adapted primarily for communication with other individuals, with any internal uses being simply beneficial side-effects. The concept of universal grammar now seems to have been more or less abandoned, in part as linguists started trying to apply the theory to more and more languages and found it harder and harder to fit it to all of them. Linguists are now more likely to think in terms of phylogenetic family resemblances between languages, and perhaps in an epidemiological way in terms of features that are more or less likely to spread. But it may still be productive to think of cognitive universals that either constrain the ways in which we tend to speak, or influence the things which we need to talk about. Recursion may be one example of this: there is one example of a language (Pirahá) which supposedly doesn’t have it, but this is very controversial. Once recursion is invented it seems unlikely that it would ever be lost, because it is just so damn useful – think how clunky the sort of language you get in early reading books is: e.g., “Here is Peter. Here is Jane.” instead of “Here are Peter and Jane”.
Accepting the fundamental diversity of human languages leads naturally into discussions of linguistic relativity: the influence of particular languages on thought processes. After the initial exuberance of the Sapir-Whorf hypothesis, the idea of linguistic relativity fell out of fashion in the 1970s and ‘80s; but as Lucy shows, this was not really scientifically warranted, as there is plenty of empirical evidence to suggest that language does indeed influence cognitive processes (albeit perhaps not in as radical a form as Whorf had supposed). I found that Lucy’s article raised disturbing questions for me regarding the extent to which scientific exploration is dictated by fashion. In this course, we are focusing on human universals—as is the current fashion in cognitive anthropology—but it may well be that in future years, cognitive anthropologists will renew the focus on cultural differences in human psychology (while hopefully accepting that there is a universal, species-typical cognitive architecture underlying these).
B. Evaluate Tomasello’s shared intentionality hypothesis of the evolution of language in terms of what we know about nonhuman primate theory of mind, and the relationship between language and theory of mind / metacognition. Is theory of mind / metacognition enough to drive human distinctiveness? Is language needed to drive human distinctiveness?
In a nutshell, Tomasello’s hypothesis is that humans are adapted to share information with other humans, and are able to model the intentional planning of others (not just their goals, as chimpanzees are). He argues that both shared intentionality – i.e., taking another person’s point of view into account – and the cooperative instinct are necessary for language to work effectively. Once you have both of these ingredients, language is actually kind of trivial: it’s basically like a really elaborate form of pointing or pantomiming.
So in Tomasello’s view theory of mind is not actually enough to drive human uniqueness: you could have animals which were very well adapted for life in a Machiavellian, ultra-competitive kind of society, but which would not be able to evolve language because they would not have any motivation to share information with one another (and what would be the point of sharing information in such a society?—nobody would trust you anyway). But complex theory of mind is one of the crucial, and uniquely human, ingredients which allowed language to evolve.
Tomasello has done most of his research with older infants and quite young toddlers, and one area which he hasn’t really followed through as yet is the later development of language and theory of mind. It seems plausible that language might feed back into the development of theory of mind and metacognition, enabling more complex forms of mental operations to take place; and that these developments in turn might enable the development of still more complex forms of language—language that is not simply concerned with the here and now, but with past, future, possible and impossible worlds. This sort of language and mentalising, it could be argued, is what has truly made humans unique, because it allows for cultural coordination and technological innovation on a grand scale—the scale of civilisations.