"Logical-Linguistic Method in Moral Philosophy"

The full title of Oskari Kuusela's paper in Ethics in the Wake of Wittgenstein is "Logical-Linguistic Method in Moral Philosophy: Resolving Problems From Iris Murdoch and Bernard Williams With Wittgenstein." It's good but also long, and I'm not sure I have understood all of it. Which is a reason to blog about it and try to straighten out my thoughts. A general problem for anyone trying to do Wittgensteinian ethics is avoiding the twin dangers of not being Wittgensteinian and not doing ethics, so I'll look out for that too.

Kuusela's goal in the paper is to show how a Wittgensteinian approach to moral philosophy can avoid the problems that Iris Murdoch and Bernard Williams identify in much analytic work in ethics, while also showing explicitly and in detail how moral philosophy can be done without becoming merely empirical. The problems that Murdoch points out are:

1. narrowness
2. ahistoricity
3. imposing a false unity on morality
4. pretending to be neutral while failing to be so     

The complaint about narrowness has to do with the concepts that philosophers look at when they bring the methods of linguistic analysis to moral philosophy. If we assume a Kantian view of ethics, for instance, then we might focus on notions of duty and obligation. If we are consequentialists then we might focus on the idea of goodness. And if we think that the only serious options in normative theory are either Kantianism or consequentialism then we might focus on all of these, but not on what Williams calls thicker concepts such as rudeness or courage. Not to mention less obviously ethical concepts that might nevertheless be used in ethical thinking. The narrowness in question is meant (I take it) to be within meta-ethics, since Murdoch is talking about people who aim to be neutral and are "anxious not to moralise" (Existentialists and Mystics, p. 74). But the distinction between meta-ethics and normative theory seems hard to maintain here. It would indeed be a mistake for a would-be neutral describer of moral language to reduce it to a single formula. This would (almost certainly) be a bad description, because it would (I believe) be oversimplified. But Kantians and utilitarians typically do believe that ethics can be reduced to a single formula. And this isn't (usually, necessarily) a result of describing language-use simplistically. If they are mistaken, their mistake is moral. (That is, Kantians don't claim that everyone is a Kantian and that their use of moral language shows this to be the case. What they claim is that everyone should be a Kantian.) So the complaint about narrowness, as I think Murdoch would accept, is partly evaluative. It more or less assumes the falseness of moral theories such as Kantianism and utilitarianism.

Something similar can be said about the ahistoricity complaint. Linguistic concepts are social and historical, so analyses of concepts that ignore differences over time or place can be regarded as faulty. Murdoch (p. 66) mentions the complaint that "as soon as you regard your moral system as a sort of fact, and not as a set of values which only exist through your own choices, your moral conduct will degenerate." This, as she notes, is a moral complaint. It is far from being a neutral fact that all philosophers must accept on pain of irrationality. If we take ourselves to be investigating concepts as used in various times and places then we had better not take an ahistorical approach. But if we are concerned with only the timeless truths of morality then we need not, and probably should not, worry about historical changes. 

Some of what goes for complaint number one also goes for number three. If we really want to describe ethical (or ethically significant) uses of language then we should certainly avoid imposing a false uniformity on these uses. But whether any alleged uniformity really is false is likely to be debatable. And moral philosophers who do not intend to do merely empirical, descriptive work are likely to be most interested in (alleged) underlying similarities. Everyone will agree that false unity should not be imposed. Not everyone will agree, though, on whether correct ethical views will contain some unity. Nor will they agree on what the nature of this unity might be. 

Similarly, pretending to be neutral while failing to be so is uncontroversially a bad thing. But is being neutral a good aim or not? Murdoch doesn't seem to think it is good. Wittgenstein recommends description (which sounds neutral) and not advancing theses, but we shouldn't jump too quickly to taking this as insistence on neutrality. His method seems more dialectical than exact neutrality, involving various points of view, not a view from nowhere. Still, there does seem to be a difference on this point between Murdoch and Wittgenstein. Murdoch says:

Kuusela, slightly differently, says that: "although we might not be able to completely avoid bringing our own moral views into philosophical accounts of morality, we ought at least to do so consciously and openly" (p. 42). He sounds a little reluctant, and I think he could be more open about it, although there is no evidence that he is trying to hide anything. (For instance, my sense is that he rejects utilitarianism, Kantianism, and Thomism, but I don't think he ever explicitly says so. Nor that he is rejecting them on moral grounds. I'll say more about this below.)

In Kuusela's view, Murdoch is right, but not as explicit as she might be about how moral philosophy should be done. She also, he thinks, does not clearly explain how what she recommends would be different from an empirical investigation such as those conducted by sociologists, anthropologists, and psychologists. And if we merely describe different ethical ways of thinking and behaving then the specter of relativism looms.

I'm not sure about any of this. Murdoch wrote quite a lot of moral philosophy, so we can see from that how she thinks moral philosophers should proceed. And the passage above about pictures that half describe and half persuade sounds a lot like descriptions of what Kant and Mill do in their best known work in moral philosophy. So maybe that is the kind of thing she thinks we ought to be doing. The "half persuasion" part does not sound purely empirical, so there goes that worry. And the relativism concern seems minor given that it is presumably possible to describe different systems or cultures without judging, as the relativist does, that all are equally good. Kuusela himself notes (p. 41) that philosophers since Aristotle have been pointing this out. His solution to the problem is not wrong, so I don't mean to overdo any criticism here. But the solution is well known, and he spends little time spelling it out, which makes the problem seem rather insignificant. In other words, I'm not completely persuaded that there is much of a problem to be solved in the first place. Nevertheless, Kuusela is right that there at the very least might seem to be not just one but three potential problems here, and it is not a bad idea to try to address these concerns. My complaint, so far as I have one, is really only that I agree with Kuusela so much that his conclusions don't surprise me as much as they might some other people. That's not much of a complaint.  

The solution he proposes to the shortcomings he sees in Murdoch (and Williams) involves using models for purposes of comparison and clarification. Thus:

the Wittgensteinian method outlined makes possible the reinterpretation of [Kantian and utilitarian ethical] theories as clarificatory devices employed to explicate specific aspects of morality, but without any need to accept that either of them could by itself explain what is essential to morality. (p. 36)

I agree that theories like these can be seen as clarifying aspects, but not the whole, of how we think about ethics. Compare "'What Is Ethical Cannot Be Taught'—Moral Theories as Descriptions of Moral Grammar" byAnne-Marie Søndergaard Christensen (in Wittgenstein's Moral Thought), in which she argues, among other things, for "a conception of moral theories as descriptions of morally relevant uses of language, moral grammars". What, though, about the "need to accept that either of them [i.e., Kantianism and utilitarianism] could by itself explain what is essential to morality"? Some people, presumably, genuinely believe that utilitarianism captures the essence of (true) morality, while others think something similar about Kantianism. Probably some people think that some moral theory of this kind must be true. But I just don't know how many people really feel as though either utilitarianism or Kantianism must be true, and feel this as a dilemma in the way that free will vs. determinism, say, can feel like a dilemma. Kuusela writes (p. 37) that:

If the suggestion is that we all feel as though either utilitarianism or Kantianism must be the whole truth about ethics, but that it is hard to decide which is true because each has strong points in its favor, then this feels like a false premise. But Kuusela makes no such claim, so he can't be convicted of that. On the other hand, if he isn't saying that, then his claim seems a bit weak. There are many ethical theories to choose from, including pluralism and anti-theory, so the stalemate has gone a long way to being dissolved already. Nor do many people, I would think, feel forced to stretch a theory to do what it struggles to do. As with what he says about relativism, perhaps the problem here is simply that I agree so much with Kuusela that his claim doesn't strike me as especially new. One thing he does that is new is to provide a philosophical justification, instead of agreeing with bits of one theory and bits of another in an ad hoc way, for regarding each theory as partly right. Perhaps this is really the main point he wants to make.  

Finally, he offers some other criticisms of utilitarianism and Kantianism, e.g., "it is still a mistake to confuse a theory [...] with reality" and "it cannot be assumed that justification must always refer to either the motive or consequences" (both from p. 37). This is true. But it isn't the same kind of mistake to believe that a particular theory happens to describe reality accurately, surely. That is, one can believe in utilitarianism, say, without confusion. And while assuming that the motive alone is the only thing ever to justify an action would be a mistake, one surely could believe this. It might be bad to believe it, but it doesn't seem to have to involve the kind of confusion that Wittgenstein aims to help us with.

In short, I think that most, if not all, of what Kuusela says is right. I would like to see more about just how short the alleged shortcomings in Murdoch's and Williams' work are, about just how much we need to get into Wittgenstein's work to reach the conclusions about ethics that Kuusela reaches, and about precisely what these conclusions are and imply. But the paper is already long, and going into all that would make it even longer, by quite a bit. Which goes to show that there really is a lot to talk about here, which is what Kuusela's and De Mesel's book aims to show. So it's a successful first essay.