Sunday 20 June 2010

Naturalism and dysfunction

At the Sorbonne conference yesterday, I gave a presentation (here) to an audience which included Jerome Wakefield. My presentation, in a nutshell, was that the Wittgensteinian argument I’d deployed against Millikan in my 1998 book only worked against one of two reductionist aims. But that that is OK since that is the relevant aim for the philosophy of disorder.

Taking there to be a distinction between Fodor’s aims and his actual arguments in Psychosemantics as an example, I contrasted the logic of his practice with the logic of the practice of Millikan. Fodor’s aims are made clear in this passage.

I suppose that sooner or later the physicists will complete the catalogue they’ve been compiling of the ultimate and irreducible properties of things. When they do, the likes of spin, charm and charge will perhaps appear upon their list. But aboutness surely won’t; intentionality simply doesn’t go that deep. It’s hard to see... how one can be a Realist about intentionality without also being, to some extent or other, a Reductionist. If the semantic and intentional are real properties of things, it must be in virtue of their identity with... properties that are neither intentional nor semantic. If aboutness is real, it must be really something else. [Fodor 1987: 97]

The promise of the programme is that intentionality itself will be naturalised through reduction. If not, then it would be convicted of being unreal (since the third position: appearing on the physicists’ basic list of properties isn’t a runner).In fact, however, what actually happens in Psychosemantics is more basic. A causal theory of reference is turbocharged by the asymmetric dependence theory to try to make space for the possibility of false thought (which is as far as he goes towards providing an account that is sufficient for the normativity of mental content) and then tied to a language of thought: syntactically characterised inner representations. But none of this goes any way towards reducing the nature of content as such, of the normative liasons between concepts and contents.

Thus what he actually attempts is to show that it is not mysterious (that it is natural) that creatures like us can think thoughts or can, if you will forgive me, respond to the space of reasons. Thus the question is: Given the space of reasons, how is it possible for creatures like us to respond to them?

Millikan, by contrast, aims to do something more ambitious with her evolutionary theory. Assuming a tool which is already more developed to account for the normativity of content (because, in turn, she has in one sense a less ambitious aim: to reduce intentionality to biology not physics) she has a check list which is then more ambitious.

First, it is not simply that a general evolutionary theoretical explanation can be given for the possession of intentional mental states (of why it is advantageous to be able to represent the world). It is rather that each particular type of content can be explained in this way. Thus the explanations cannot be question-begging: the selective advantages conferred must be characterisable in non-intentional terms. The meaning must drop out of the evolutionary theory.

But second, more relevant to my concerns, and perhaps motivated by the thought that once she is in for the penny of all of the first point she might as well be in for the pound of this also: she aims to naturalise the space of reasons itself (not just our ability to respond to it). She will naturalise conceptual liaisons themselves. Given a teleological account, logic itself will become ‘the first of the natural sciences’ [Millikan 1984: 11]. So her key question is: Given our biological natures, how is it possible for creatures like us to respond to what we take to be the space of reasons, whatever it is.

We can think of the differences in accord with the Euthyphro ‘paradox’. We all agree that: For any act x: x is pious if and only if x is loved by the gods. But: Is the pious loved by the gods because it is pious, or is it pious because it is loved by the gods? Fodor’s and Millikan’s projects take opposing views. In effect, Fodor derives engineering constraints on the gods given that we know that they are able to track piety, antecedently understood. Millikan, by contrast, aims to explain piety by describing the engineering of the gods.

There is, however, a familiar objection to Millikan. A teleological account of function is a form of interpretational theory. Past behaviour is a set of signs to be interpreted. Like the interpretation of signs, such behaviour is consistent with an unlimited number of possible functions or rules including both continuations that seem natural and logical and an unlimited number of other ‘bent’ rules that deviate in unnatural ways.

What ensures the determinacy of biological function – what selects just one of the rules – is an explanation of the presence of a trait couched in intentional terms which interprets what the trait is for. But finite past behaviour can be explained as exemplifying many different or ‘bent’ functions or rules, all of which would have been equally successful in the past but which diverge in the future.

Millikan’s reply to this is:
[The ‘bent’ rule] is not a rule the hoverfly has a biological purpose to follow. For it is not because their behaviour coincided with that rule that the hoverfly’s ancestors managed to catch females, and hence to proliferate. In saying that, I don’t have any particular theory of the nature of explanation up my sleeve. But surely, on any reasonable account, a complexity that can simply be dropped from the explanans without affecting the tightness of the relation of explanans to explanandum is not a functioning part of the explanation. [Millikan 1993: 221]

But this reply only works if the simplicity of an explanation can be assessed in a non-question-begging way. The problem is that the explanation of the survival value of the trait corresponding to a particular mental content has to be given without presupposing its content. And, of course, the content of the proper function is not just a matter of looking to behavioural dispositions but selecting a function which best explains them.

I am now less convinced than I used to be about the universal applicability of this argument against a reductionism based on biological functions. It depends, also, on the aim of the reduction. So if the question is: Given the space of reasons, how can we respond to them?(construed as an invitation to armchair engineering) then the Wittgensteinian objection has no force because the explanation presupposes merely conceptual normativity which was always simply presupposed in the reductionist question. But if the aim is to naturalise the conceptual space of meaning itself then the objection looks to hold because a key question is begged about that space.

What then the reductionist aim of the invocation of biological function in the philosophy of illness, disease or disorder? Two options can be articulated by translating from Fodor’s and Millikan’s questions in the philosophy of content. There, I suggested that Fodor’s question is: Given the space of reasons, how is it possible for creatures like us to respond to them? Translated into medicine we get something like: Given the conceptual space of illness, how is it possible for creatures like us to suffer it? And that doesn’t seem a question worth giving an a priori answer to.

Millikan’s question was: Given our biological natures, how is it possible for creatures like us to respond to what we take to be the space of reasons, whatever it is? Translated into medicine, we get something like: Given our biological natures, how is it possible for creatures like us to suffer what we take to be illness, whatever it is.

Millikan’s looks the better model question for the philosophy of medicine. It makes questioning the nature of the concept of illness itself central, jnot just something presupposed. But of so, because it shares the task of naturalising the normativity of pathology, Wittgenstein’s objection is a serious objection. (That is, a biological teleological account cannot rule out wildly divergent accounts of the functions in play, functions which explain the presence of traits.) And if so, we need a better version of naturalism for the philosophy of medicine.

I didn’t understand Jerry’s reply (to which I wasn’t allowed by the circumstances of the session to reply). It seemed to have two elements (but I may simply not have got the first part at least of what he was saying). He first suggested that my argument turned on a Quinean notion of the indeterminacy of translation. Quine (by implication: like Wittgenstein) had established that meaning is indeterministic. I took the implication of this to be that nothing much follows from that about meaning.

But second – and this argument I thought I did follow – Millikan’s project does require determinacy of meaning which a Quinean argument undermines. But the functions involved in disorder do not require such determinacy. So a Quinean argument for the former does not apply to the latter.

There would still be something a bit odd about this, though. If one believed that Quine was right about meaning one could not also his views as an argument against a reduction of meaning via function by arguing that the latter wasn’t deterministic. Quine’s account of the nature of meaning would make the target of the reductionist analysis indeterministic. So if the analysis was also indeterminsitic, so much to the good for the reduction via function.

So I think what he must have been saying was that Quine was wrong about meaning but Quinean arguments about indeterminacy had something right about them. (And in fact he said a few times last week that he’s a bit of a Searlean: meaning has to do with consciousness and is irreducible.) Had Quine argued with Millikan he would have argued for the indeterminacy of her analysis of functions. Given that meaning is deterministic, that would have been a problem for her. But the kind of function involved in functional accounts of disorder can tolerate such indeterminacy so that would not be a problem for his own analysis.

I’m not a persuaded by this reply because the translation from Wittgenstein to Quine is misleading. Quine accepted a degree of indeterminacy in his positive account of meaning. He thinks that the evidence that fixes meaning only goes so far. This of course is because Quine builds in an assumption that the evidence has to be physicalistically described. Against that background, meaning would be indeterministic. But what justifies that restriction? (Answer: Quine’s scientism.)

Wittgenstein does not think that meaning is indeterministic. The contexts which play a role in constraining it are described in intentional terms. What Jerry takes to be a parallel between Quine and Wittgenstein is not part of the latter’s positive account but rather a reductio ad absurdum of reductionism. More significant, however, is that the negative argument does not deliver merely a domesticated indeterminacy but rather no shaping of content in the future at all.