I’d expected, on the basis of an earlier tangential encounter, to be rubbed the wrong way by later Wittgenstein. As I keep on through Monk’s biography, though, I find myself reminded of something I think Robert Nozick tried to do. The title of his second book, Philosophical Explanations, refers to his idea of a “second mode of philosophy, not directed to arguments and proofs: it seeks explanations.” An “explanation” in Nozick’s sense doesn’t seek to prove, as against a skeptic, that motion occurs, or that there are subjective experiences. Instead, it tries to show how it might be that something we are sure enough of (things move; we have experiences) is the case in the face of considerations which would appear to exclude the possibility (Zeno’s arguments, someone like Daniel Dennett).
What called this to mind was (later) Wittgenstein’s insistence that there were no real philosophical “problems,” but only linguistic “puzzles.” Philosophy, he thought, does not “find things out” as, say, the natural sciences do: it only clarifies our concepts. We might cast Nozick’s explanations as Wittgensteinian attempts to “help the fly out of the bottle” — we don’t expect to prove that there are no experiences, or that there are. Arguably, what we’re doing when they seek these “explanations” is clarify what we mean by apparently conflicting propositions (Wittgenstein might say: deciding how we are prepared to use them), such that they don’t in fact conflict.
I don’t recall anything in Nozick indicating that he would have accepted this spin on his project. It may even seem that he has the very opposite of Wittgenstein’s humble vision of the role of philosophy: in Invariances he characterizes superstring theory as a branch of metaphysics. But I think it’s mainly put that way for flourish: what I take him to have meant is that science is not strictly separate from philosophy. Or, to bring out the Wittgensteinian flavor, philosophy is not its own field of inquiry: it is not a way of finding things out, but a method of organizing the things we find out. Wittgenstein saw that as a chief distinction between the sciences and philosophy, but in a sense it also indicates how they are fused. What we learn about neuroscience or quantum physics might jar with our notions about mind or matter or causality. That we recognize this as a conflict entails that some part of the empirical result we’re getting is mapped closely enough by our concept (or “fits” our use of a term), but some other part does not. In the case of a logical conflict, it may seem both that the concept must map onto the observed behavior, and that it cannot.
Now, Wittgenstein was super-deferential to ordinary language: linguistic puzzles, he seemed to say, arose only in the face of perverse philosophical impulses, from attempts to rip words from their day-to-day context. Nozick seems more willing to revise those ordinary concepts in the face of tension (even if we only see the tension because of those impulses), but is following what we might think of as a Wittgensteinian line in, for example, his final thoughts on ethics, in Invariances. (Ignoring, for a moment, that Wittgenstein didn’t think philosophy had anything to say about ethics.) He begins that chapter with a look to usage. That is, we all know that people do things we describe as “moral deliberation” and “trying to follow an ethical code” and the like, so Nozick looks at what we know about the world in an attempt to clarify what’s going on in those cases.
Now, there’s an opposite approach, which, just to fix ideas, we can think of as Kantian. On that approach, we have a concept (maybe an a priori concept) of what it is to act under a moral obligation, and it can be an open question whether it’s really possible for us to ever do this. A big chunk of the Groundwork of the Metaphysics of Morals is devoted not so much to establish the content of the Categorical Imperative, which Kant seems to take almost for granted, but rather trying to establish that it’s possible for us to be motivated by (to really act on rather than merely in conformity with) a command of pure reason. Nozick’s inquiry flips this. He takes as his starting point the large set of behavioral instances we are all (presumably in ordinary language) willing to call “trying to behave ethically,” whether or not we agree with the content of the ethics in question. To bring out the difference, consider his response to the problem many Kantians wrestle with, of how moral rules are “binding.” It’s supposed to be constitutive of a moral rule, on that view, that it’s something we have a kind of overriding reason to follow. Now we seem to have this deep problem about how it could be that we always have a reason to do such-and-such. Nozick seemed to think this just showed that Kantians were being too demanding — asking too much of the concept of “morality.”
Still, once he’s offered an explanation of ethical behavior as the function of a specialized, evolved cognitive module, Nozick wants to be able to say something about why we ought to generally do what’s right — why to take that module as authoritative, and not some other evolved impulse. He offers some remarks about the function of the ethical module in consciousness, which Nozick takes as a sort of essential human characteristic, something constitutive of the person. Now, if the connection between consciousness and the ethical module is supposed to be given as a reason we have to behave morally, it’s not clear how that’s going to get off the ground. It’s rather evidently not the case that we literally go unconscious if we don’t try to adhere to some ethics, and there’s nothing about the historical or physiological connection that we could present to a full-blown moral skeptic and expect him to reply: “Ah, yes, at last I see why I should(n’t) do x, y, and z!”
There’s a different, Ludwiggier way to take it, though. It could be taken as a more biologically thorough version of Wittgenstein’s elaboration of rule following as a “way of life,” dealing specifically with moral life. There is, perhaps consistent with Nozick’s aversion for what he called “commiting philosophy on people,” no thought of arguing a genuine skeptic, one who really does stand outside any moral community and feel no ethical pull, into seeing why he must be bound by moral rules. There is a suggestion, recalling Gauthier, of why a strictly self-interested person would very often find it worth his while to adopt the moral point of view, but there are certainly no guarantees, and at any rate, someone might be equally skeptical about whether self-interest must be a binding motivator. (Unless one uses “self-interest” in that tautological sense, in which case the arguments for a self-interested morality don’t go through, since the tautological notion of “self-interested behavior” becomes content free.) It is an account, but not a justification, in the sense that we might give an account, but not a justification, of the rules of logic. As with ethics, if one were really a skeptic about the validity of logical rules of inference as a set, there would be nothing non-question begging one could say to make the skeptic see why they’re right. Any further justification would require some minimal inference or deployment of a rational principle, at the very least at the gap to the conclusion: “and therefore we’re entitled to infer in this way.” This doesn’t undermine our confidence in the rules, it shows that there is something ill-formed about the demand: give me a reason to accept any reasons! Are moral reasons like this? They seem to generate the same problem. What would a non-question begging value statement look like? We could try to reduce one kind of practical reason (moral) to a subset of another (egoistic), but these attempts are seldom all that convincing, and the question can just be raised again about the subvening category of reasons. And anyway, reductions of those sort don’t respect the independent status of ethics that comports with our usage. Our attitude towards immorality is not the same is our attitude towards imprudence, just frieghted with the extra annoyance and fear associated with robbery or assault.
As I’m sure some smarter person than I has already noted (and, I suppose, Wittgenstein himself was trying to get across in the little he would say about ethics), that means that ethics is not primarily about accepting some set of propositions — of believing some set of axioms to be true. Then you’d need a “rule to interpret a rule.” But the parallelism with Wittgenstein’s examples won’t do entirely. We can accept “language games” as matters of shared practice in part because we aren’t even sure what it means to claim that an entire linguistic community is “misusing” the language in some way (e.g. by having departed from an archaic form). But we do very much want to say that it’s possible for a whole moral community to be wrong on a point of ethics. Anyway, this is all quite a lot of crude rambling, and there’s much more of it to be done, but it’s late. Maybe I’ll work this out a bit more when I’ve slogged through some more of the secondary literature on Wittgenstein and rule following.