1. METAPHYSICS

Simon Blackburn has characterized his central concern in metaethics as follows:

If I understand him correctly, Scanlon's aim in this book is to urge that this is a largely imaginary problem. Reasons are real but they are not part of the natural world. He seeks to secure a respectable metaphysical modesty, and hence a lack of vulnerability to charges of queerness, for this realism by adopting what he calls a ‘domain-centered account of ontological questions’ according to which,

Such domains, of which the normative is one, science another, have standards which are internal to them. The correct answers to questions within them are determined by those standards. Normative talk, realistically construed in this way, thus fails to be ontologically disreputable in the way talk of witches is ontologically disreputable. For witch talk trespasses on the scientific domain and commits itself to suppositions within it that are known, by the standards internal to it, to be false (21). There is no domain-independent way of understanding terms like ‘real’, ‘exists’ or ‘world’ available to make sense of asking whether reasons exist, are real, are part of the world, other than by inviting appeal to the standards of the normative domain. (I’m not quite sure to what domain we should assign this very general, higher-level talk about domains.)

Witch-talk can be rejected because the beliefs it expresses conflict with what our best science tells us is true. Scanlon's view would only be, he tells us,

Of course normative claims and scientific claims may often readily conflict. Suppose Nicky thinks that abortion in circumstances C is permissible only if the foetus to be aborted lacks natural property N and also that abortion is permissible in those circumstances. Now she comes to believe, in the light of new scientific data, that the foetus does have natural property N. There is a conflict and something has to give.

There are interesting general questions about what should give in such circumstances. A certain orthodoxy – by no means unchallenged – accords science – often with various ‘if's and ‘but's – a sort of priority. (A modern locus classicus is Weber Reference Weber, Shils and Finch1949. For more recent discussion see Kitcher Reference Kitcher2011.) Nature doesn't much care about the moral concerns of us humans and we should not allow our moral preoccupations to determine or influence the answers we accept to scientific questions. To suppose otherwise, to tolerate the subordination of science to our moral sensitivities, involves, orthodoxy tells us, a failure of objectivity, a certain reproachable wishfulness that threatens the sort of corruption of scientific practice exemplified by Lysenkoism. If we accept this orthodoxy and seek successfully to reject or revise normative claims that conflict with science, we seem to avoid conflict between these two domains only by according one of them a certain priority. One worry then for Scanlon is that it is hard to see how this sort of priority could be either explained or evaluated without engaging in just the kind of ‘placing’ Blackburn promotes and Scanlon repudiates. Just what is it about the scientific domain, and just what is it about the normative domain, that makes it appropriate to understand the priority relations between them in this kind of way?

That worry – and I think it is a real worry – noted, it needs stressing that Scanlon does not quite claim that normative and scientific claims don't conflict in ways most such examples would readily tell against. He restricts this claim to the case of what he calls pure normative claims (20–22, 37–42). Most normative claims are not pure but mixed, entanglements of fact and value, typified by so-called thick concepts. Pure normative claims do not involve or presuppose any non-normative claims. It is these latter pure claims that he takes not to conflict with the claims of science.

By way of illustration (my own, not Scanlon’s) we might think of a simple example of an alleged conflict between the normative and the empirical. Some people hold the controversial but widespread normative belief that we should, as a political community, espouse radically egalitarian social policies. But some economists defend the also controversial but also widespread view that such policies would be economically ruinous. Given a plausible supplementary premise about the normative ineligibility of the ruinous, we might face another conflict between scientific and normative judgements. If the economists are right, the egalitarians may have to give ground. A neat move here is to look for ways of finessing egalitarianism so it that does not give such hostages to empirical fortune. And perhaps a neat way of doing just that is offered by Rawls’ Difference Principle that tolerates only such social and economic inequalities as benefit the least advantaged. If economic ruin is very plausibly supposed to be bad for everyone, including the worst off, the Difference Principle will direct us to tolerate as much inequality as is needed to avoid such ruinous outcomes. We might think of this as a move from mixed egalitarian principles which presuppose the falsity of certain economic doctrines to a pure one which does not.

But no. The Difference Principle is itself deeply empirically presupposition-laden. In section 26 of A Theory of Justice, Rawls considers the kind of putative counterexample whereby it might prohibit some astronomical benefit to some relatively advantaged group at some quite microscopic cost to the least advantaged. Rawls’ response to this concern is to rehearse some empirical reasoning intended to convince the reader that such a possibility is unlikely enough to be legitimately ignored. He goes on to note: ‘We should also observe that the difference principle not only assumes the operation of other principles, but it presupposes as well a certain theory of social institutions’ (Rawls Reference Rawls1972: 158). So the difference principle is not pure at all. At this point we might start to wonder, Is anything? For Rawls himself the answer seems to have been No as we see when he proceeds to contrast his own view with that of ‘some philosophers’ who

Are there such pure principles? Perhaps. Scanlon characterizes pure principles at one point as principles where any dependence of the normative on the non-normative has been ‘subjunctivized away’ (40). I think this idea can be unpacked roughly this way. Suppose, for example, that the Difference Principle is indeed dependent on various e.g. social scientific presuppositions in the way Rawls supposes. Let P be a long conjunction of all such presuppositions. Then we can formulate what we might call the Subjunctivised Difference Principle:

In circumstances where P we should do as the Difference Principle directs.

If we can do this, and maybe we always can, there is at least in principle a way of arriving at the pure normative claims on which any mixed claims depend, although the ‘in principle’ is starting to tell a little and we might reasonably doubt if we can ever know or state any true pure normative claims.

There is still scope to doubt if there are any at all. At least it might depend on other theoretical beliefs we have about the normative domain. For a certain kind of utilitarian it might seem straightforward. What we should do is determined by certain extremely general principles about the overriding importance of maximizing the satisfaction of the preferences of any such creatures as there may be to which anything deserving the name of ‘preferences’ can properly be ascribed. Everything true about what we should do then follows from that together with the vast number of non-normative truths that fully describe our circumstances. For a certain kind of neo-Aristotelian virtue theorist, on the other hand, the facts of ethics for us humans are deeply rooted in the contingencies of human nature and the human condition and there is really nothing significant for us to say about the normative that fully abstracts away from all these. If the latter view is right, there may be limits to what we can hope intelligibly to subjunctivize away.

2. EPISTEMOLOGY

David Enoch (Reference Enoch2011: 5.3.1) has a nice objection to Scanlon's view. What about what he calls counter-reasons? Imagine some community engaged in counter-normative talk where they make various claims that, were they claims about reasons, would strike us as grotesquely false (that something will cause me pain is a counter-reason to do it) but these grotesque falsehoods are warranted by the standards internal to counter-normative discourse and claims about counter-reasons are accorded by members of this imaginary community just the kind of practical-action guiding significance accorded to claims about reasons by us. Don't we here want to say that these folk are wrong where that has a sense that is domain-independent?

A few feet from me now is an electric socket. Simply by pressing a finger to it I could put myself in the way of a moment of very considerable pain. (I tried this experiment as a very stupid small child and know whereof I speak.) The counter-reasoner, given this information, says, ‘Do it!’, the reasoner, ‘Don't do it!’ These are not competing claims about reasons but competing imperatives, competing determinations of, echoing Gibbard (Reference Gibbard2003), what to do. The conflicting items are prescriptions, not assertions, but they conflict all right.

Scanlon will say ‘Don't do it’. For we have, he insists, no reason to do what we have counter-reason to do. Taken as normative claims, the judgements by which the counter-reasoners guide their actions are wrong and we discover this by engaging in normative reasoning: when we think about what reasons there are, we see that counter-reasons are not reasons (29–30). And yet a troubling symmetry persists. When the counter-reasoners, we might well imagine, engage in counter-reasoning, they are persuaded that the thing to do is what counter-reason requires of them. Who says we’re right?

We discover normative truths, says Scanlon, the same way we discover mathematical truths, ‘simply by thinking about things in the right way’ (71). As the fourth lecture unfolds, this is spelled out by endorsing the familiar method of reflective equilibrium as a method of discovery in both the normative and mathematical domains. Here we begin with a set of considered judgements that seem clearly right to us and then attempt to frame more general principles that will explain them, continuously making revisions and adjustments at both ends as we work towards the ideal end state of reflective equilibrium.

Thomas Kelly and Sarah McGrath (Kelly and McGrath Reference Kelly and McGrath2010) worry that there is no reason to suppose this is a method of discovery unless we have some sound independent warrant for the judgements with which we start. Otherwise there is no good reason to place our trust in the judgements that emerge at the end. Garbage in, garbage out. So that, in Scanlon's words ‘all the justificatory work is done at this initial stage, before the process of seeking equilibrium has begun’ (83).

Scanlon isn't too worried by such concerns. The justifying force of the method does indeed depend in the credibility of the considered judgements it takes as inputs but this is not, he stresses, the same as the credence you or I or Scanlon happens to attach to it.

What conditions might those be? Scanlon doesn't tell a story here. We can readily enough invoke the sort of filters specified by Rawls.

That is a bit minimal and it is hard to see how we could clearly determine that even these conditions were conducive to arriving at credible judgements without some independent standard by which to assess their credibility. I might say that the credible judgements are the ones that agree with my, or perhaps on a good day, our (for some specification of ‘us’) own judgements and that the good conditions are the conditions in which people turn out most reliably to agree with these judgements. But that would be a bit unsatisfying, to put it mildly. Or with my or our own in some ideal state, reflective equilibrium attained. But that is really no better, given Kelly and McGrath's point. Garbage in . . . . So the worry remains that when the counter-reasoners think about normative matters under these same conditions they might arrive at considered judgements very different from ours. And that what emerges in reflective equilibrium for them looks grotesquely false to us.

In the course of an interesting discussion of expressivism, Scanlon endorses Andy Egan's well known Smugness Objection to that view (Egan Reference Egan2007):

I won't here consider the merits of this view though I argue elsewhere that the expressivist can answer Egan's objection (Lenman Reference Lenman2014). The thought I want to close on is that certain kinds of realist might be open to a similar worry about what Julia Markovits, making a similar point in a slightly different context, calls epistemic humility (Markovits Reference Markovits2014: 54–8). Scanlon's moral epistemology ends up resting, just as the expressivist's is alleged to by Egan, on a deep self-confidence about one's own normative perspective. Of course it is recognized as open, from within, to criticism and reflective scrutiny and there are substantive arguments to be had with those one finds oneself differing from. But what of those – real or imaginary – who may be beyond the reach of our reasoning (and we of theirs) because the place from which they start is too far away? I don't see how there is room in Scanlon's picture for the possibility that they might just be right. So I suspect he may be just as open to an Egan-style complaint about epistemic humility as he and Egan take the expressivist to be.

There is a lot more going on here than I have space to discuss. This book also has illuminating and extremely interesting things to say about supervenience, constructivism, particularism, normative strength and other matters.

Tim Scanlon has surely thought harder and more deeply about moral and normative questions than perhaps any other philosopher now alive. As the foregoing rather hurried sketches of arguments may suggest, I do not agree with Scanlon about what reasons are. But whatever reasons are, anyone who cares about moral philosophy has some very good reasons to read this important, rich and endlessly rewarding book.