What is Innateness?

by Michael Lopresto

Forest_PathWhen it comes to explaining human cognition and human uniqueness, everyone seems to think that nature and nurture constitute a false dichotomy. Both nature and nurture work together harmoniously to contribute to the cognitive traits that make humans profoundly different to every other animal on the Earth. Unlike every other animal on Earth, humans are uniquely flexible; we have inhabited every kind of environment, engaged in intergenerational social learning, cooperated with those outside of our immediate group, accurately described things we'll never directly observe, and much more. Humans are cognitively flexible, behaviourally flexible, communicatively flexible and representationally flexible. Representational systems employed by humans are open-ended and unprecedented in the animal kingdom: natural languages like English and Chinese, artificial languages like predicate logic, formal languages like those in mathematics, pictures, diagrams, weather maps are all but a few of the representational systems employed by humans (not to mention mental representations, which are likely to be analogues of the aforementioned systems).

One of the central questions of cognitive science is explaining how humans acquire cognitive traits, including ones that contribute to human uniqueness. Is the trait for language innate or learned? Is the trait for mental time travel (the ability to experience one's past or future) innate or learned? Is the trait for moral reasoning innate or learned? And so forth.

Nativists are those who say that lots of cognitive traits are innate, and empiricists are those who say that very few cognitive traits are innate. The nativism/empiricism distinction is not to be confused with the rationalism/empiricism debate of early modern philosophy. That debate was primarily over epistemology, while the contemporary debate is primarily over psychology. However, questions of epistemology and psychology were systematically conflated, as Kant and others pointed out, and we ought to be careful not to conflate the same questions now. Even so, there are fairly clear links between the two questions. The rationalists of early modern philosophy, like Descartes and Leibniz, argued that a great many cognitive traits were innate, and the empiricists of that era, like Locke and Hume, argued that very few cognitive traits were innate. (Although those philosophers spoke in terms of “innate knowledge” and “innate ideas”—phrases that certainly need careful interpretation).

However, the question “Is cognitive trait X innate or learned?” presupposes that the concepts INNATE and LEARNED are somewhat well defined.[*] (I take it for granted that the concept COGNITIVE TRAIT is uncontroversial, i.e. phenotypic traits relating to things like thinking, inference, perception, intelligence, and so forth.) Our question certainly doesn't presuppose that for any cognitive trait it's all or nothing; totally innate or totally learned, or even totally acquired through environmental interaction.

Firstly, it's trivially true that any innate cognitive traits will be environmentally supported via nutrition, metabolism and so forth. That doesn't make any putative innate cognitive trait any less innate. Secondly, for any cognitive trait, whether it be our capacity for language or our remarkable ability to attribute mental states to others purely on the basis of behavioural evidence, there may well be an answer to the question of acquisition where the notion of either innateness or learning is doing the explanatory work, whilst the other is mere background information. For example, the fact that genetic activity contributes to our mindreading ability may be no more informative than the fact that the relevant brain regions are made up of atoms. In other words, it may be trivially true that genetic activity contributes to our mindreading ability, and citing that fact will only be explanatory in a superficial sense.

So how should we unpack the notions of innateness and learning? The word “learning” refers to a few different cognitive phenomena, but we have an intuitive grasp on things that count as learning and things that don't. I learned how to speak English and do arithmetic, but I never learned how to have hands or grow hair. And it's a conceptual truth that if a trait is innate then it's not learned. For example, if our ability to distinguish faces from non-faces is innate, then we never learned how to do it. Rather, that ability was acquired through growth, maturation, or some other biological process. The problematic concept is INNATE. How do we give it an adequate definition for constructing theories in cognitive science? As Stephen Stich pointed out in the Introduction to his (1975) collection Innate Ideas:

The controversy is easy enough to summarize: Some philosophers, as well as linguists, psychologists, and others, allege that human beings have innate knowledge or innate ideas. Others deny it. But what is it to have innate knowledge or an innate idea? There is a pattern running through much of the debate in this area. Advocates of the doctrines of innate ideas and innate knowledge commonly take the notion of innateness itself to be unproblematic. They explain it with a few near synonyms, “inborn” or “unlearned,” or with a metaphor or allegory, and leave it at that. The doctrine's opponents often begin by puzzling over just what the doctrine could possibly mean. They go on to construct a variety of accounts, arguing against each in turn. The advocate's rejoinder, as often as not, is that he has been misunderstood. Thus, in approaching the debate over the innateness doctrine, we would do well to ponder what we are saying of someone when we say that he knows something or has an idea innately. (1975, p. 1, emphasis in original)

This insight is particularly astute because it's so often true, even 40 years later. Nativists often take the concept INNATE to be unproblematic, while empiricists often take the concept to be fundamentally confused. For example, the linguist Dan Everett, a thoroughgoing empiricist, is fond of arguing that absolutely no precise or coherent definition of innateness has ever been given. Nativists like Chomsky will much more often retort that they've been misunderstood, instead of actually arguing against their critics. (Descartes was careful to give a dispositional analysis of INNATE, of which I'll say more shortly.) However, it would be a mistake to say that what distinguishes nativism and empiricism is that nativism presupposes the coherence of INNATE, while empiricism rejects its coherence. Empiricism rests on the coherence of INNATE just as much as nativism does because empiricism only makes sense to the extent that it posits at least one innate cognitive mechanism, even if that mechanism is to learn new mechanisms. A mind with no innate learning mechanisms would never learn anything.

There are three main approaches to analysing the concept INNATE in cognitive science (there is likely a pluralistic analysis for the concept in science generally, as INNATE will refer to different things in immunology etc. from what it will refer to in cognitive science). The first is the invariance approach, which says that a cognitive trait is innate if and only if it is robustly acquired across an appropriate range of environments, while those that are highly variant are not innate. This was arguably Descartes' view of innateness, and sophisticated versions of the invariance account are held today by most philosophers of biology, including Andre Ariew, Elliot Sober and Elizabeth O'Neill. However, the invariance account suffers from several pitfalls. Firstly, some cognitive traits are highly invariant but are nonetheless learned. For example, the beliefs that the sky is blue and that water is wet are both highly invariant, but are clearly learned through experience. Piaget argued that many concepts were highly invariant, but also learned. It would be bizarre to categorise Piaget as a nativist. Secondly, the invariance account can't make sense of the fact that some cognitive traits could be highly variant, and yet still be innate. So on a Chomskyan view, a native Chinese speaker acquires the grammar for her language. But this grammar is very different to the one acquired by a native French speaker, for example. And yet, the acquisition process is equally innately specified for the both the Chinese speaker and the French speaker, according to Chomsky. On Jerry Fodor's view of concept acquisition, most concepts are not learned, but are instead the result of a brute-causal process called “triggering”, that occurred when one is presented with the right stimulus. So the concept of, say, BOOK, is triggered when one is first presented with the appropriate stimulus at the appropriate age, but the acquisition of BOOK is nonetheless innate, but highly variant as many humans throughout space and time never acquire the concept BOOK. The point of this objection to the invariance view is definitely not that the Chomskyan view of language acquisition, or the Fodorian view of concept acquisition, is true. Rather, these views are coherent, and therefore a satisfactory account of innateness ought to make sense of these possibilities. The invariance account just doesn't seem to have the resources.

The second is the primitivist approach, developed in its most sophisticated form by Richard Samuels, which says that a cognitive trait is innate if and only if it is not acquired through a psychological process (like learning, perception or inference) and is the result of the normal course of development. So the primitivist account is consistent with our conceptual constraint that if a cognitive trait is innate, then it's not learned. Furthermore, it can explain why it is that innate traits can be highly variant: some psychological primitives can be triggered by various environmental contingencies. Why the normalcy condition? Because there will be cognitive traits that are not acquired through a psychological process but are nonetheless not innate, like acquired sociopathy, which can be caused by insult to the brain. The normalcy condition rules out these sorts of counterexamples. So, as Samuels says, a cognitive trait is innate according to primitivism (for organism O) if and only if it satisfies two conditions:

1. Primitiveness Condition: It is psychologically primitive in the sense that (a) it is posited by some correct psychological theory, but (b) no correct psychological explanation of its acquisition exists (not just in this historical moment, but in principle).

2. Normalcy Condition: It is acquired by O in the normal course of development.

The third is the eliminativist approach, which says that INNATE is a fundamentally confused concept and should therefore be completely dispensed with in scientific practice. The use of INNATE in scientific practice is likely to be just an artefact of scientists illegitimately deploying a concept from folk biology. The philosophers Fiona Cowie and Paul Griffiths, and the biologist Patrick Bateson, among others, have argued for this position. The argument goes something like this. The folk or vernacular concept INNATE can refer to a number of different properties:

1. Having an (adaptive) evolutionary explanation

2. Being insensitive to variation in “extrinsic” factors in development

3. Being present at birth or inborn

4. Being “universal” in a couple of senses

a. Being pancultural (i.e. present in all human cultures)

b. Being monomorphic (i.e., a trait takes only one form among members of a species)

5. Not being acquired by learning (Paul Griffiths 1997, What Emotions Really Are)

Scientists may (and often do) begin by pointing out that an organism possesses one of the above properties, and will then infer (illegitimately) that the organism possesses one or more of the other properties. The folk concept INNATE leads to faulty generalisations and inferences. Just because a cognitive trait is present at birth, like a neonate being able to recognise her mother's voice, doesn't mean that the trait was not acquired by learning, as this learning could have occurred in utero (and likely did). The use of psychoactive drugs to create altered states of consciousness is pancultural, but probably does not have an adaptive evolutionary explanation, as hominin environments were likely too heterogeneous with respect to psychoactive substances to support an increase in fitness for its consumers through building a resistance to neurotoxins naturally present in some hominin environments.

However, as Richard Samuels points out, just because some people are confused about INNATE, doesn't mean that the concept itself is confused. What seems to be going on is that people often confuse properties that are evidencefor innateness, with properties that constitute innateness. Perhaps it's a good opportunity for philosophers to clarify the concept INNATE, so that scientists can use it more fruitfully in their work? In my view, this is exactly what Samuels does, in getting away from the folk concept of INNATE and providing an analysis that accounts for the actual explanatory purchase that the concept has in cognitive science. Furthermore, unlike the invariance account, the primitivist account gives a mechanism for the acquisition of innate cognitive traits, i.e. nonpsychological processes like growth and maturation. The primitivist account also gives an explanation of exactly what the nature/nurture debate turns on: nativists posit lots of psychological primitives, whereas empiricists post very few psychological primitives. Finally, primitivism makes sense of the various arguments that are thought to be relevant to establishing the truth of innateness claims about the mind, like the poverty of stimulus argument. The poverty of stimulus argument says that there is not enough information in a child's environment to enable her to learn a language. Her linguistic evidence dramatically underdetermines her grammar. Therefore, language is not acquired through a psychological process like learning, but is instead acquired through a nonpsychological process like growth. All these considerations together show that innateness is likely to be psychological primitiveness, and not invariance, and most certainly not something that's just fundamentally confused.



[*] I use capitals to name concepts.

Like what you're reading? Don't keep it to yourself!
Share on Facebook
Facebook
Tweet about this on Twitter
Twitter
Share on Reddit
Reddit
Share on LinkedIn
Linkedin
Email this to someone
email