Question How do philosophers address the nature-nurture controversy?

Accepted:
October 15, 2005

Comments

I'm not sure that philosophers have a uniform answer to this question, and as you may have guessed, before even proposing one they tend to spend a lot of time trying to get clear on just what the question is supposed to be. If the question is: To what extent is human behavior explicable in terms of our biological endowment, and to what extent is our behavior explicable in terms of environmental influences?--if that is the question, then one point to notice is that some current views in the philosophy of biology would deny that the question makes a lot of sense. For example, Paul Griffiths in such places as _What Emotions Really Are_ defends a "general systems theory" according to which it is part of our biological heritage to be inherently plastic, and thereby inherently capable of being importantly molded by environmental influences. That would be one thing separating our species from many (though certainly not all) others. If this view is correct, then it is hard to see what the point is of trying to figure out whether a given bit of behavior is due to nature or nurture--in far too many cases it will be due to both. The point, then, is that we need to get clearer on what would count as a purely "natural" or genetic explanation of behavior, and what would not, before even having a grip on what the question is asking.

I take it that thenature/nurture controversy is the controversy concerning how much of ourcognition is innate, rather than acquired. One thing that philosophers havetried to do is get clear about what ‘innate’ might mean in this context. ‘Innate’ is not a term that is used incontemporary biology. The first idea that pops to mind is that a trait isinnate if an only if it is present at birth. But this won’t do because prenatallearning is possible. And some apparently innate characteristics, such as facialhair, appear only after birth. Thus presence at birth is neither sufficient nornecessary for innateness. Terms like ‘genetically determined’ and ‘geneticallyspecified’ won’t do either. No trait is 100% genetically determined, since theenvironment always has a causal role to play in development. And ‘geneticallyspecified’ is just a metaphor: genes do not in any literal sense specifyphenotypic traits. (For discussion, see Richard Samuels (2004) "Innateness and CognitiveScience", Trends in Cognitive Sciences Vol. 8 issue 3, availablehere: http://www.kcl.ac.uk/kis/schools/hums/philosophy/staff/r_samuels_papers/...).

Yet cognitivepsychologists and philosophers do engage in debates about whether certaincognitive traits are innate. For example,Noam Chomsky famously argued that humans have a rich body ofinnate knowledge about human languages (see e.g. Chomsky’s (1980) Rules andRepresentations); Jerry Fodor famously agued that many of our ordinaryconcepts are innate (see e.g. Fodor’s (1975) The Language of Thought); and psychologists and philosophers argue overwhether we have innate representations of ‘folk psychology’ (roughly speaking, innate beliefs about how minds work) and ‘folkphysics’ (roughly speaking, innate beliefs about the behaviour of physicalobjects). (For discussion, see Susan Carey and Elizabeth Spelke ‘DomainSpecific Knowledge and Conceptual Change’ in Hirshfeld and Gelman, eds, (1994) Mapping the Mind). All these claims arehotly debated. So what might people be arguing about?

I wouldsuggest that there are two distinct notions of innateness that are useful andinteresting. One of them was articulated by the biologist C. H. Waddington (Waddington, C.H. (1940) Organizers and Genes). Thisis ‘canalization’: a trait is canalized to the extent that itsdevelopment is causally insensitive to environmental and genetic variation. So,for example, the development of toes is highly canalized, since it isrelatively insensitive to variations across genes or environment. Developmentof skin colour is less canalized, since it is causally sensitive to e.g. sunlight.

The otheruseful notion of innateness was articulated by Richard Samuels(see the article mentioned above). It is a notion that applies specifically tocognitive traits and doesn’t generalize to others. According to this notion, acognitive trait is innate if and only if (a) it is psychologically primitive,where this means that it is not acquired by a process of learning or any otherpsychological process – thus it is not acquired by e.g. induction, deduction orperception, and (b) it is acquired in some normal manner. Clause (b) is thereto rule out, e.g., a case in which alien scientists helpfully program a map ofthe Solar System into your brain – this representation would not be acquired bya psychological process, but it would, of course, still be acquired rather thaninnate.

Thus whenpsychologists and philosophers argue over whether this or that cognitive traitis innate, one might usefully interpret them as discussing the extent to whichthe trait is developmentally canalized or one might usefully interpret them asdiscussing whether it is psychologically primitive. I leave it to the reader toconsider how the two notions interrelate.

Fordiscussion of all of the above, see ‘Poverty of Stimulus Arguments’ on mywebsite.

Let me add some comments to Mitch Green's and Gabriel Segal's. (And a quick plug: you might want to check out my entry on "Nativism" in the new edition of the Macmillan Encyclopedia of Philosophy, due out soon.)

Two quick points, and then a longer one.

First: Showing that a trait has a "biological basis" is not the same thing as showing that it is "natural" in any sense that can be opposed to "nurtural" (is that a word? It should be.) Unless you are a dualist, you shouldn't be surprised to find that psychological states are correlated with, depend upon, or are flatout identical with biological states. (Indeed, you shouldn't be surprised even if you are a dualist, but that's another story.) But that means that any acquired trait will have some biological effect. Showing, therefore, that the brains of musically accomplished individuals are different from people who aren't hardly shows (as one NPR story reported, I swear to God) that musical talent is innate. Similarly, from the mere fact that, say men's brains are different from women's brains it follows only that men and women are different, not that they were born different, or that they are necessarily different.

Second short point: there is nothing inherently good about nature. This should be painfully evident to any human being even dimly aware of the world around us, the earthquake in Pakistan being only the most recent example of nature at its most natural. And yet, there is a persistent implication in much talk of the "natural" that the course of nature is generally the best. "If you want right guidance," says Jean-Jacques Rousseau, "always follow the dictates of nature." He says this just before announcing that it follows from woman's nature that her purpose is to please and serve men. And Kant says that logical thinking is unnatural, and therefore not recommended for, women. A woman who tries to philosophize, Kant says "may as well grow a beard." But I don't want to leave the impression that it's only when philosophers are trying to rationalize sexist domination that they presume nature is good. Many philosophers have tried to draw normative conclusions from (alleged) facts about human nature. I take the strong line that facts about human nature have no normative implications at all. (See my essay, "Natures and Norms" in Ethics, October 2000, Vol. 111, no. 1)

The long point concerns confusion about the notion of "heritability". This is the term that is most often used when some new result about the "genetic basis" of this or that trait is reported in the media. Heritability is a measure of the degree to which variation within a certain population can be explained by genetic variation as opposed to variation in some environmental factor, and not, as I suspect many people think, a measure of how "strongly" caused, or controlled a trait is by the genes, of how insensitive it is to environmental factors. The best way to see the difference is through the following thought experiment, due to biologist Richard Lewontin.

Imagine, first, that you have a handful of corn seeds, each one genetically different from every other. Now plant them all in the same soil, and provide exactly the same amount of water, warmth and light. Chances are that the resulting plants will differ in height. If so, the explanation for the differences is entirely genetic: because, for this population, the environmental factors affecting growth were all the same, any differences in growth must be attributable to differences in the plants' genotypes. The heritability of the phenotypic property for this population is 100%.

But that doesn't mean at all that the environment has nothing to do, causally, with how tall the plants will grow. For consider a second thought experiment. This time, take a set of genetically identical seeds, but grow each one in different conditions: vary, for example, the amount of water you give to each plant. Chances are, again, that there will be variation in height -- probably at least as much variation as there was among the plants in the first experiment. But now the explanation for the variation will be entirely environmental -- it has to be, because there is no genetic variation among the plants. The heritability of height for this population, in this range of environments, will accordingly be low.

What these thought experiments show is that there is no sense to the unrelativized question, "how heritable is height?" It only makes sense to ask, how heritable is height for this population in that particular set of environments. "Heritability", in other words, is not a measure of canalization, as Gabriel Segal explains it. (For further details, read Lewontin's excellent essay, "The Analysis of Variance and the Analysis of Causes" -- American Journal of Human Genetics, 26, 400-411, reprinted in The IQ Controversy, the excellent anthology edited by philosophers Ned Block and Gerald Dworkin.)

The parade case to show that high heritability doesn't entail immutability is the case of phenylketonuria, a serious form of mental retardation. The condition results from an inability to properly metabolize the amino acid phenylalanine, which itself results from a genetic defect. Until the chemistry of the condition was understood, the phenotypic trait, the mental retardation, was, essentially, 100% heritable for all populations -- the difference between those who had the condition and those who didn't was all genetic. But once the chemistry was understood, researchers were able to prescribe a diet -- one that sharply restricted the amount of phenylalanine in the food -- that prevented the toxic buildup of that element in the child's body, and thus prevented the retardation. The heritability of phenylketonuria, accordingly, dropped. Now part of the explanation for why some people have the condition and others don't is environmental; two individuals with the same genetic defect can differ with respect to this illness because one has had the good fortune to be given the special diet, and the other, sadly, has not.

This case makes another important point: it may not be at all obvious what environmental factors may make a difference to some phenotypic trait.

The most useful way to think of our biological nature, then, I think, is in terms of what biologists call a "norm of reaction." This is a function that, ideally, tells you what phenotypic traits a given genotype will produce in various environments. It's possible to actually calculate norms of reaction in cases where it's technologically feasible and morally permissible to create large numbers of genetic clones, and where something is known about the environmental factors that are involved in the expression of various traits. So we can plot norms of reaction for height as a function of rainfall in corn plants. Obviously, the human case fits none of these conditions, so norms of reaction for individual genotypes are not on the horizon. Not to worry -- if there's a a trait we are interested in manipulating, we don't necessarily need to know anything about how genes contribute to producing it.

The notion of heritability gives us, I think, another way of understanding the claims of cognitive scientists that this or that trait is "innate." We can think of innateness in this context as heritability taken across species: the explanation for the difference between apes and human children with respect to the acquisition of language is almost entirely genetic. (And not, as earlier behaviorist theories would have had it, largely environmental.)

I argue for this general way of thinking about human nature in my essay "The Role of 'Human Nature' in Feminist Theory" in Janet Kourany's edited volume, Philosophy in a Feminist Voice.

Question How do philosophers address the nature-nurture controversy? Accepted:October 15, 2005

Question How do philosophers address the nature-nurture controversy?

Accepted:
October 15, 2005