Friday, 1 April 2011

pigs, positions, preferences: the zeugmatics of causal theories

It has - in my experience - become common for philosophers to try to persuade me of the fallacious character of what I previously took to be an a priori piece of understanding regarding my knowledge of my own mind (that reasons for action are avowable infallibly - which is to say that what we sincerely offer as reasons provides what here count as 'my reasons') by making reference to the empirical findings of Nisbett and Wilson.

A classic example is the case of why I make the choices I do. You present me with a line of differently coloured or patterned pairs of jaunty socks, and ask me to pick one. I take the pair on the right. You ask me why I took this one, and I tell you something about the charming 'little piglet' motif embroidered on them.

But then it turns out that I am no exception from a general empirical finding: that people in such situations are more likely to pick the socks (or whatever) on the right. "Aha", you say, "You thought you were picking those socks because you liked the pig motif, but really it was because of their position". The disturbing conclusion is supposed to be that we don't know our own minds - we don't know (as the gloss, which already builds in some optional philosophical 'commitments', will have it) which causal processes lie behind our very own actions.

The problem with the typical gloss is, to my mind, that it encourages a conflation which for shorthand could be described as an undiscriminating amalgamation of reasons and causes. It conflates the causal processes which lead me to perform certain actions (a bias for the right guides my choices) with the teleological ends of these actions (the reason I chose the socks because they had a pig on them) - bringing both of these under a global concept of 'reason for action' which is obfuscatory as regards the different logics of the different forms of explanantia. (The tendency I've commonly met with is to bring these two forms of explanation together by placing the teleological ends as the contents of desires and to see these desires as entering competitively into the matrix of causes which guide our actions.)

This, it seems to me, just won't do. Let us accept that desires have a dispositional form. If I like little piggy socks then, everything else being equal, I will avail myself of some. Desires, that is, are not individuated simply through our avowals ('I like piggy socks') but also through our actions (getting myself some piggy socks). Let us accept too that I am sometimes or often unaware of the efficient causes of my actions (thus I just didn't know that I am more likely to pick things on the right of a line). Finally let us accept that we may be prone to offering post-hoc rationalisations or justifications of our choices by references to appealing features of the objects chosen. It still doesn't follow, I want to claim, that it is perspicuous to present the findings of the experiments with a sentence like "You thought you chose them because of the pig symbol, but really you chose them because of their position".

One way to bring out what is wrong with this is to note that we wouldn't think it correct to describe a tendency to take objects on the right as a desire for objects on the right. Even to describe it as a preference seems more obfuscatory than clarificatory. To talk here of 'unconscious desires/preferences' seems further to conflate an epistemological with a logical concern.

How then should we describe what has happened? Well, we must acknowledge that our choice was (partly) determined by position. We must acknowledge too that, had another pair of socks with a little sheep motif been on the right, then we may have misleadingly talked of our appreciation of the sheep motif as the reason for our choice. And accept the implication from this that we tend to post-hoc rationalise our choices - to invoke desires where such an invocation is otiose.

But at no point need we fall into supposing that there is some general sense of 'explain' or 'reason' which pits positional effects against piggy appreciations. Instead there are two separate matters at play in such experimental results: (1) The limited character of our knowledge of the efficient determinants of our own actions (this is what is interesting in the findings of Nisbett and Wilson), and (2) the tendency of two key criteria (avowal and behavioural disposition), which jointly give meaning to the language game of desire, to come apart in practice (with avowal unwarrantedly outstripping behaviour through egregious self-confidence). When the social psychologist defeats the self-ascription of reason in avowal he is not to be understood as making a move within the same language-game - he is not saying x where the agent says not x. Rather he is giving us reasons to suspend our belief that here we do meet with a reason-giving language game - and reasons to consider, instead, that we meet here not with reasons but with mere rationalisations. For utterances which self-ascribe reasons only count as giving reasons if the agent is, as we must assume by default that he is, self-possessed. Yet what the social psychologist (or Freud and Bernheim, in their post-hypnotically active suggestion to the patient to open the umbrella for the doctor before he leaves the room to go out, which patient when asked 'why on earth did you do that?' provides but a daft rationalisation) engineers is a situation in which the agent lacks self-possession. A situation in which, if you like, they are instead possessed instead by the hypnotist or by the glamorous array of little piggy socks on show.

My confidence regarding my desires (2) may be overinflated. My lack of knowledge regarding the determinants of my actions (1) may be deflating. Nevertheless my overconfidence in my desires (2) is not to be understood as a pretence to the same kind of self-knowledge as would be held by someone who knows better than I the determinants of their actions (1). To say "You thought you chose them because of the pig symbol, but really you chose them because of their position" is not itself logically wrong-headed, but then neither is Dickens' zeugma: "Miss Bolo went home in a flood of tears and a sedan chair". Unlike the scientific naturalist, however, Dickens remained content to play; we are not asked by him to take seriously the idea that on another occasion, whilst we thought Miss Bolo went home in a sedan chair, actually she went home in a flood of tears.