Philosophy Blog

Life at the thought-face.

Wednesday, October 16, 2002

Since writing the last post I have begun to wonder whether I have fallen into the trap of asking bad questions which create confusion out of clarity. Think of the parallel question 'Why be moral?'. Think about this long enough and you will realize that there is a big difference between giving an answer which serves to justify my existing disposition to good behaviour and one which would persuade an amoralist to change her mind.

If someone seriously doubted that they should, at least in some contexts, conform their thinking to good patterns of inference, then there would be no persuading them otherwise. Is my worry anything more than that? I am not sure.

.: posted by Tom Stoneham 3:10 PM

Why be rational?

We can distinguish within moral discourse between terms which are evaluative, such as good and bad, and terms which are normative or prescriptive, such as required or permitted. And there is a position, often attributed to Hume, which says that the only way to move from an evaluation to a prescription, from the judgement that doing X is would be cruel, to the judgement that one ought not to do X, is via a desire to do what is good and avoid what is bad. However, there is a great deal to be said for the opposing view that moral evaluations are essentially normative, that we cannot make sense of judging that an act is good without also judging that it should be done.

Is there a similar structural feature in theoretical reasoning? We have ways of assessing patterns of reasoning as good or bad, by reference to logic, probability theory, scientific methodology etc. We also talk as if someone who believes p and that if p then q, ought to believe q. So in talking about reasoning we make both evaluative and normative judgements. Setting aside any worries there might be about the ground of the evaluative judgements, we still face the puzzle about how to move from the evaluative to the normative. Logic is obviously evaluative, but equally obviously not normative. Logic and allied disciplines give us a means of partitioning inferences, conceived of not as psychological transitions but as relations between propositions, into the good and the bad, but it should (!) be puzzling how that has any bearing on what we should believe.

Here is a typical example of the slide (from Dororthy Edgington):

"Logic does not tell you what to believe, but rather that some beliefs rule out others: some combinations of belief are consistent, other combinations are not."

Surely logic tells us that some combinations of propositions are inconsistent, and thus that some combinations of belief consitiute belief in an inconsistent set of propositions. But how does the fact that one proposition is inconsistent with another make the belief in one proposition 'rule out' (causally prevent??) the belief in the other? That is my problem.

Suppose logic tells us that the inference from p to q is a good one. I believe p and I am aware of the validity of if p then q. One reason it does not follow that I should believe q is that a perfectly reasonable response would be to change my mind about p. Sometimes when we realize the logical consequences of our beliefs, we thereby realize that we were mistaken. But that can easily be met by making the normative consequence disjunctive: I ought to either reject p or accept q, which is equivalent to saying that I ought not to (continue to) accept p and deny q. But we are still left with a transition from the evaluative to the normative which needs explaining.

So far I can only think of three options:

1. Logic is not merely evaluative but also normative. Since logic is ostensibly about the relations between propositions, but the normative claims we are searching for are about the relations between mental states, this view is faced with a stark choice between psychologism logic or logicism about psychology. That is, we could say that logic is in fact about the relations between mental states, or we could say that mental states are essentially logical. Psychologism about logic is infamous. Logicism about psychology is rarely made explicit (the notable exception being Colin McGinn), but it is not clear that it does the trick. First off, when we say that q follows from p, there is no implication of a causal relation. But logicism about psychology has to face the fact that the relations between mental states are causal, so that in a logical mind the belief that p causes the belief that q. Secondly, if logicism is to get the consequence that someone who believes p should not also deny q, rather than the merely evaluative claim that someone who believes p and not-q is not functioning properly, the relation of logical consequence needs to be given a normative reading along the lines of if p is true then q ought to be true (and everything that ought be true is necessarily true?). This is pretty implausible.

2. We need to be Humeans about rationality. So the only way to get from 'it would be logically bad to believe p and deny q' to 'someone who believes p ought not to deny q' is via a desire to think what is logically good. There are two problems with this. To see them, let us consider how the process would work. First I recognize that p entails q. I conclude that, if I want to be logically good, I should not believe p and not-q. I do want to be logically good. So I should not believe p and not-q. The content of this normative claim is conditional: if I am going to get what I want, I should not believe p and not-q. The first problem is that this normative claim seems to weak. If I want and ice-cream and the ice-cream is in the freezer, then I should go to the freezer. Maybe I don't bother, then I have not got what I wanted. But it is a fact of life that for one reason or another we do not always get what we want, so I can live with that. The normative force of such claims is very weak, but the normative force of the claim that, since p entails q, one should not believe p and not-q is supposed to be much stronger: it is no defence against the charge of irrationality that one could not be bothered! The second problem is that while the combination of the conditional and my desire do entail that I should not believe p and not-q, they only give me a reason not to believe that if I accept the logical consequences of things I believe, but that is exactly what is at issue.

3. Constraining one's thoughts by reference to good patterns of inference is part of what it is to be a person / have a mind. The idea here is that a person just is someone who recognizes evaluative claims about good and bad inferences as creating normative constraints upon their thoughts. Perhaps, even more strongly, it might be claimed that mental states such as beliefs are necessarily such that their typical transitions follow patterns which are logically good. However, like the similar position of logicism about psychology, that threatens to lose the normative element: it does not follow from the fact that if p entails q, the belief that p and not-q is, in ideal circumstances, impossible, that one ought not to believe that p and not-q. Whichever version we were to adopt, the vacuity of this proposal should be apparent when we consider the analogous proposal in ethics: a creature which did not find moral evaluations as having normative force would not be a person. Even if true, this does not explain how evaluations have normative force for me (assuming I am a person).

One paragraph discussing each of these options is a pretty cursory treatment, so there may be more to be said for them. And there may be other options which I have not considered. But for the time-being, the whole thing looks very puzzling indeed.

.: posted by Tom Stoneham 11:27 AM

Tuesday, October 15, 2002

Education

Well, term will have started at Universities everywhere by now. I am not teaching this term, but recently came across a good quotation which, when taken out of context, seems very relevant to that perennial question 'What good is a philosophy degree?':

Education is what is left when what has been learnt has been forgotten.

This is such a good aphorism that there is not even any need to explain it. The education acquired through the study of philosophy might be achievable through the study of some other subject, so the reason to choose philosophy is that what has to be learnt (and later forgotten) in order to gain the education is of sufficient interest to motivate the student.

Why did I say that it was a good quotation 'when taken out of context'? Because it comes from BF Skinner, the notorious behaviourist. Skinner was, I gather, really just a methodological behaviourist, in that he did not deny the existence of mental states, but merely said that they were irrelevant to scientific psychology. However, given the explanatory aspirations of a scientific psychology, this amounts to the philosophically significant doctrine that we can understand and explain human actions without reference to mental states. The basic idea is that human, and animal, actions are merely responses to external stimuli, and the responses are by and large learned. So learning is the creation of stimulus-response pairs. But in the quotation, he is using 'learn' in the everyday sense of being able to repeat information, and reserving 'education' for acquired stimulus-response pairs.

.: posted by Tom Stoneham 12:00 PM

Thursday, October 10, 2002

Moral Dilemmas

The moral dilemmas which most of us face in our lifes are pretty insignificant. No one will die and any hurt or offence will probably be forgotten or forgiven within a year. And yet these minor dilemmas are major problems at the time they occur.

My current dilemma is over two incompatible invitations. This is not a mere social dilemma for two reasons. Firstly, I am not only deciding for myself, but also for my daughter: by accepting one of the invitations on behalf of the family, I would cause her to upset her best friend. Secondly, both invitations to a greater or lesser extent involve the obligations of friendship, and those obligations have an ineliminable ethical dimension.

At first sight it might seem that we could apply utilitarian reasoning here, because the adverse onsequences are of the same kind: disappointment. Surely we can weigh up the likely intensity and duration of the disappointment caused by rejecting each invitation and identify a best choice?

The problems with this approach are immediate. For example, do you measure intensity of disappointment by the subjective feeling or by how easily the feeling can be overtaken by new pleasures? How do you take account of previous disappointments? How do you aggregate over people? What if one of the people involved has become so jaded from experience that they rarely feel disappointment at being let down? But even if there were answers to be found to these questions, there remains the more fundamental problem that a result like 'Action A will cause 10% more disappointment than action B' does not seem to resolve the dilemma in the way that 'Action A will cause great disappointment but action B will be happily accepted by all' does.

We could instead start looking for principles that might resolve the dilemma. For example, one might thnk that the obligations of childrens' friendships are trivial compared to the obligations of lifelong adult friendship. But on the other hand, such adult friendships are only made because of the way things worked out in childhood, and a parent who puts too little value on their child's friendships may prevent that child from learning to make the deep bonds which structure a worthwhile adult life. After all, an adult friendship worth keeping will survive a large measure of disappointment, but maybe not a child's friendship. Equally one might question the appropriateness of impartiality in these cases, for should one not care more about a friend's disappointment than his wife's? Or you own child than someone else's?

The more one thnks about this sort of case, the more apparent it becomes that there is nothing, no calculation, no principle, no argument, which will solve the problem. The decision is so hard because the normal model of working out which is the right course of action does not seem to apply here. The problem with this model is that it concentrates on choices and decisions that can be seen as trying to conform to some pre-existing goal or norm (it need not be objective). But moral dilemmas, however trivial, do not appear to work like that, they require a different kind of decision. Somewhere in The Salterton Trilogy, Robertson Davies notes that the idiom 'making one's mind up' also has a literal reading: sometimes our decisions have a creative function. When faced with a moral dilemma one is choosing not so much a course of action as the type of person one wants to be. Of course, one decision does not a personality make. Rather, in making such a decision soeone is not casting an opinion on which is the right course of action, but showing that she wants to be the type of person who puts, say, the obligations of parenthood over friendship, or loyalty over an easy life.

That is the kernel of truth in virtue ethics.