Wednesday, January 27, 2010

Is-Ought Muddles

There has recently been some discussion in the blogosphere relevant to the is-ought problem, in particular at "Rationally Speaking":

Hume's Guillotine

On Morality, a Response to Julia

Chris Schoen also had some comments:

Oughtism

Horse's Mouths

One of the problems with thinking through this subject is that it is dominated by the slogan that you can't get an 'ought' from an 'is'; and the problem with doing philosophy by slogans is that slogans pick up baggage over time. This is especially true here. I've argued that Hume's argument on this point is usually taken out of its original context, which is an attack on moral rationalism, and I've also argued that Hume is right to think this is a problem for rationalists although not for sentimentalists like himself. But whenever I argue this, whether here or elsewhere, I find that I am suddenly involved not in one argument but in half a dozen, and this is because when people say that you can't get an 'ought' from an 'is', they can mean half a dozen completely different things. Strictly speaking, the original slogan was about relations: you can't get an 'ought' relation from an 'is' relation. This problem can be evaded by denying the rationalist claim that 'ought' is a relation; and this is precisely why Hume's moral theory is not endangered by his own argument. However, if you take it out of context and interpret only vaguely some of the phrases in Hume's actual argument, it becomes tempting to use the same slogan to claim that you can't deduce an 'ought' statement from an 'is' statement. Taken at face value this is thoroughly and obviously false, since any 'ought' statement can be reformulated as an 'is' statement. For instance, I can formulate 'We ought to give to the poor' as, quite trivially, 'We are such that we ought to give to the poor'. The second is an 'is' statement; and from it you can derive the 'ought' statement. So when people can't seriously mean the Statement interpretation; that would make it a logical problem, and a purely artificial logical problem based on a complete confusion, at that. People do, in fact, sometimes talk as if it were a logical problem; but obviously if there is a problem relating 'ought' and 'is', however those are understood, it must be due to some features of our substantive account of what they are, not a mere matter of logic. And there are plenty of things that make this obvious; it is simply not reasonable to claim that medical science has nothing to say about what you ought to do in order to be happy, or that physics has nothing to say about what you ought to do in order to go to the moon.

So we could perhaps set the Statement interpretation aside; people often talk as if it were the interpretation they meant, but if it were this would merely be a sign that they are logically confused. What people usually try to do is to take a dyad and deny that either member of the dyad is reducible to the other. There are several interpretations of this sort, but the most common is Fact-Value, when people say "ought and is" what they usually mean is "fact and value". They mean you can't get values from facts. Now, this is very ambiguous. It could mean that from facts it is impossible to tell whether something is valued. Taken this way it is obviously false (if something is valued by someone, that is a fact), but astonishingly you can find people who talk as if this were what they meant. But we can perhaps set this Literalistic Fact-Value interpretation aside as merely suggested by verbal sloppiness rather than real intent. More interesting as an interpretation, and probably what most people who talk about the opposition of Fact and Value are actually groping toward, is the claim that your recognizing something as fact does not guarantee that it is a value for you. Finally we hit something plausible. Whether it is right, however, depends crucially on how you classify things as Facts or Values.

There is a further problem with the term 'value' itself, which can mean either 'the property of a thing that is the reason a thing is valued' or 'things valued'. Happy puppies are things that most of us value, but we value them because they are cute, cuddly, lovey-dovey, and so forth. People genuinely value their children; they value them for many reasons: parental love, reciprocation of filial love, and so forth. But there are plenty of facts about both of these things, and recognizing some of these things may well guarantee that it is a value for you. A similar problem arises with 'fact'. A happy puppy or a child is not a fact; it's a thing about which factual statements may be true or false, but facts are not things, and have a propositional character. Rain is not a fact, but it may be a fact that it is raining. Dogs are not facts, but it may be a fact that they are chasing you. Or, if we are using the term in such a way that we can use it of both, we are introducing yet another ambiguity into this ever-increasing swampy muddle that is the 'is-ought problem'.

So some people generalize it. The problem is not Second Person but First Person Plural: there is no fact such that it guarantees that something ought to be a value for all of us. This has the advantage of getting us back to oughts, and thus of not involving any sort of lie when someone says that the problem is due to the 'oughts'. But, again, this simply depends on how you are classifying facts and values; and the same issue recurs again. It also shows what we should have already known: 'oughts' and 'values' are different things. Even putting the problem in this First Person Plural way requires recognizing that they are not the same; otherwise we are committing ourselves to an infinite regress. Why can't we get values from facts? Because it is a fact that no fact requires that values be valued. And why is that? Because it is a fact that it is a fact that no fact requires that valuing values be valued. This is thoroughly absurd. So 'ought' and 'value' are not the same thing, and should not be treated as if they were.

Problems like these will, in fact, arise at any level, because both facts and values are purely derivative; neither of them is the primitive it is treated as being. The reason these two, 'fact' and 'value', became attached to the is-ought slogan was due to logical positivism, which attempted to put forward a purely emotive theory of morals that would sharply distinguish morals from sciences and both from mere nonsense. The problem is that there are plenty of facts about emotions, from at least some of which you would have to be able to derive some moral statements if the emotivist theory were true. Logical positivism, here as elsewhere, was never able to make itself self-consistent. But very few people are tempted by logical positivism these days. Why do we find people still proceeding as if the logical positivists were right on this point, when they weren't even coherent, and almost nobody is a logical positivist, anyway?

There is reason to think that no interpretation of the problem (beyond Hume's original and narrow Relation interpretation) can be rationally sustained, or is even coherent. James Chastek once joked that we can't derive an 'ought' from an 'is'; therefore we ought not to try. The joke, of course, which apparently would be entirely lost on some people, is that the statement is self-refuting: it can only be true if one can derive an 'ought' from an 'is'. But it is hardly plausible that someone claiming that we can't derive an 'ought' from an 'is' is claiming that this has no ramifications for how we ought to reason, no matter how the 'is' and 'ought' are interpreted. Something suspicious is afoot. If 'ought' cannot be derived from an 'is', and this is a fact, then we can't derive from it any conclusion about whether we ought to try deriving an 'ought' from an 'is', nor about whether we ought to act as if we could. This is absurd. But if we mean instead that 'ought' ought not to be derived from an 'is' (whether it can be or not) then it becomes difficult to understand why anyone would believe this at all.

When we throw away the slogan and look at the actual problems people are addressing when they raise it, we find that a common problem, the one raised in the "Rationally Speaking" exchange above, is the question of whether "it is possible to use scientific facts to justify selecting one particular set" of moral principles. This is more interesting than anything involving the slogan, and it simply doesn't need the slogan. And it shows what the real problem is. It is how what is (and only that to the extent that scientific inquiry considers what is) relates to what is good. We know -- or we do if we don't confuse ourselves with incoherent slogans about ought and is or fact and value -- that standard scientific practices can genuinely inform us about some goods: health, for instance. It does not follow from this that there is nothing to be known about such goods except what standard scientific practices can determine. Even if it did, it would not directly follow that there are no goods beyond those about which standard scientific practices can give us complete information. For instance, medical science can inform you about what constitutes this real and genuine good, a healthy body, and what you ought to do to get it. It does not directly follow from this that medical science, or even any more fundamental or extensive natural science, can tell you that a healthy body is a better good than, say, the production of beautiful art. Rimbaud, in order to write better poetry, deliberately deranged his senses, and seriously endangered his health, with drugs and alcohol and crazy living. Is there any experiment that can be made, or measurement that can be taken, or mathematical theory that can be tested, that would prove that this was, overall, a good or bad thing to do? If there is, it certainly can't be assumed on the basis of the fact that scientific inquiry can shed light on goods like health and human taste for beautiful language. The one does not require the other.

The fact of the matter is that we have good reason to think that scientific inquiry of the sort we find in physics, chemistry, and biology can only shed a limited light on the goods that can be and are considered in human reasoning, and in anything like their current forms they can only shed a limited light on a limited number of these goods. The first follows from the fact that standard scientific practices require an immense amount of abstraction from details, but details, sometimes weird details, are clearly relevant to the proper understanding of some goods, for instance, those that we especially love, like children and spouses and friends. Not only that, but details are relevant to how these goods relate to other goods. Practical reason, unlike theoretical reason, has to deal directly with the actual circumstances of particular cases. And the second is seen easily enough in the fact that trying to solve serious ethical problems using only currently known facts from physics, chemistry, and biology, is thoroughly futile -- we inevitably make assumptions going beyond these facts. It is one thing to argue that physics, chemistry, and biology can help us a bit in narrowing down which moral principles, out of all possible candidates, are right. This is a fairly modest claim, and has much to be said for it. It is another thing to suggest that we can narrow the candidates down to one solely on the basis of physics, chemistry, and biology; we obviously cannot do this with the physics, chemistry, and biology we have now, and it is irrational to make claims about the world solely on the basis of what you imagine science will somehow discover at some point in the future. We'd need a very substantive argument for something like this; and substantive argument is precisely what we never find, just definition and fiat.

So when all the muddles are taken away, and all the absurdities are thrown out, we are left with a very modest position, that science can shed some light on the moral life. There is good reason, contrary to the thrust of the slogans, to think that is true; and there is good reason, contrary to some others, to think that the 'some' should be emphasized. The speculative mean is sometimes as golden as the ethical one.