This marks the 10th of 11 virtual meetings on Derek Parfit’s Climbing the Mountain. First, though, a programming note: next week, we’ll be discussing chapter 12 of the latest version of Parfit’s manuscript, available here. As it turns out, though, there’s a 13th chapter in the new manuscript, and we hadn’t planned on that with the rotating schedule we set up for posting this summer. So here’s the plan for now: next Thursday, Doug will post a summary/discussion of the latest version’s chapter 12, and then the following week we’ll simply set up a post to have an open discussion on the latest version’s chapter 13. And if a new version of the manuscript comes out in the meantime, well, we give up.
There are three main discussions in this chapter. The first focuses on objections to Rawlsian contractualism, the second focuses on how what Parfit calls Kantian contractualism avoids the objections to Rawlsian contractualism, and the third explores how certain interpretations of, and revisions to, Scanlonian contractualism can render it a very plausible way to defend the Deontic Beliefs Restriction without abandoning appeal to our moral intuitions as worthless. I won’t spend a lot of time on the first two issues: I think Parfit’s mostly right, but at any rate he’s simply reiterating (in many cases) objections made by others, albeit in succinct and very clear ways. I mostly, then, want to try and figure out his defense of a version of Scanlonian contractualism in the final section.
Under the Rational Agreement Formula of contractualism, “Everyone ought to follow the principles to whose universal acceptance it would be rational in self-interested terms for everyone to agree” (200). However, it’s either unclear what’s implied by this formula, or it gives unfair advantage to those with the greater bargaining power. Rawls thus attempts to answer these objections by appeal to the veil of ignorance (VI), such that no one knows the particular facts about themselves or their circumstances. Now according to Rawls, this version of contractualism provides an argument against all forms of utilitarianism, but Parfit argues that this isn’t true, nor does the view support acceptable non-utilitarian principles. On the one hand, if the idea behind the VI is simply to make us impartial, then you could get this by either giving the relevant contractors no knowledge of the probabilities of ending up in one position or another, as Rawls does, or you could give your contractors an equal chance of ending up in any position. Following Nagel, Parfit argues that taking the latter option could actually support some form of utilitarianism (viz., the utilitarian average principle), and that Rawls’ reasons for preferring the former are implausible.
On the other hand, even if Rawls admits that the overall view might yield utilitarianism, he can still maintain that it doesn’t have to, that it can also yield the non-utilitarian principles favored by so many. This is also false, argues Parfit, insofar as Maximin yields certain indefensible moral conclusions in certain cases, and in others it simply yields no clear verdicts (e.g., promise-keeping, lying, imposing risks on others, etc.). In addition, we think that in ordinary moral life, there are all sorts of relevant considerations other than the size and number of the resulting benefits and burdens in moral reasoning, but Rawlsian principles don’t include any of these non-utilitarian considerations – it’s just all about self-interested calculations among the representative parties.
Of course, I wonder how fair all of this is, particularly in light of the fact that Rawls was merely presenting a theory of distributive justice solely for the basic structure of society, and not a thoroughgoing normative ethical theory (a point made even more clearly in Political Liberalism). True, some remarks in TJ seem to indicate otherwise, but these might safely be ignored as inconsistent with what’s become evident as the larger, purely political, project.
At any rate, the Kantian Contractualist Formula (KCF) is alleged to be better than Rawls’ formula at achieving Rawls’ own aims in rejecting the Rational Agreement Formula (RAF). KCF (“Everyone ought to follow the principles whose universal acceptance everyone could rationally will”) and RAF both require unanimity, but KCF doesn’t rely on the notion of agreement. Given that fact, it avoids the problem of unequal bargaining power. It also assures impartiality, without the crude “frontal lobotomy” (212) approach of Rawls’ VI. Unanimity isn’t guaranteed, on KCF, so “it would be morally more significant if unanimity could be achieved. That would be true if there are some principles that, even with full information, everyone could rationally choose” (212). Achieving this will actually depend on what we think about reasons and rationality, whether or not, for example, reasons depend on desires or on values. KCF succeeds only if desire-based theories and Rational Egoism are false. Furthermore, moral reasoning on KCF may appeal to all sorts of non-utilitarian considerations, so it can certainly support non-utilitarian principles. Of course, for KCF to succeed, it has to provide clear verdicts sufficiently often, producing one and only one relevant principle that everyone would have sufficient reason to choose. Parfit thinks this condition could be met, though, demonstrating how the Principle of Equal Shares does so: in cases where some quantity of unowned goods can be shared between different people, and no one has any special claim to these goods, and if equally distributed the goods would produce the greatest sum of benefits (everyone receiving equal benefits from the equal distribution), then everyone could rationally choose the principle giving them equal shares, and this is the only principle that everyone could rationally choose.
I fail to understand how this single demonstration provides an argument for what’s needed here, though, namely, an argument for why KCF would meet this sort of condition sufficiently often. Indeed, this is sort of a precious case: how many times do we find ourselves in a situation with unowned goods, with no one having a special claim to the goods, and so forth? Perhaps I’ve just missed the role this is supposed to play in some larger argument, but from what I could tell, there’s a real gap here.
Finally, consider Scanlon’s formula (SF): everyone ought to follow the principles that no one could reasonably reject. Here, Parfit wants to defend Scanlon from the charge of emptiness. Since Scanlon uses the term “reasonable” in a partly moral sense, goes the complaint, “everyone could claim that the moral principles which they accept could not be reasonably rejected” (214), such that SF would make no difference to our moral thinking. But Scanlon explicitly appeals to the Deontic Beliefs Restriction (DBR), which says that when applying some moral formula, we can’t appeal to our intuitive beliefs about which acts are wrong. What, then, to do? One option is to reject DBR. This won’t do, though: the best method of determining the correct moral principles should include appeal to all our beliefs, moral beliefs as well. Another option is to go strictly constructivist, such that we construct all of the moral facts through some procedure, with the possible implication that all of our currently-held moral beliefs are wrong. This, too, won’t do. The best method, then, is going to be that of reflective equilibrium, which includes reference to our intuitive beliefs about wrongness. We need, then, to preserve DBR, but how can we do so if we also deploy the method of reflective equilibrium?
The answer depends on a distinction between two senses in which some property of an act makes that act wrong. In one sense, it’s trivially the case that wrongness is just the property that makes some act wrong in a non-causal fashion. In the other, more important sense, when acts have some other property, e.g., the property of being a lying promise, it’s these other properties that non-causally make these acts wrong. So the property of being a lying promise isn’t the same property as wrongness; rather, the fact that some act is a lying promise is what non-causally makes that act have the different property of being wrong.
Scanlon’s claim has been that his contractualism gives an account of what it is for acts to be wrong, but Parfit argues that he’s better off claiming instead that contractualist formulas describe, not wrongness itself, but one of the properties that itself non-causally makes an act wrong. So if some act is a lying promise, then this property would make it one that couldn’t pass the non-rejectible principles test, in which case both these facts (the fact that it’s a lying promise and the fact that it can’t pass the test) are what make the act wrong. If Scanlon’s formula thus took this form, he could (a) defend DBR, and (b) preserve our moral beliefs as valuable to moral theorizing. The way to do so would be to insist that:
It is only while we are asking what these contractualist formulas imply that we should not appeal to our beliefs about the wrongness of any of the acts that we are considering. We can appeal to these beliefs at a later stage, when we are deciding whether we ought to accept these formulas. (222-223)
As I understand it, then, this revision would undermine two popular objections seen in recent years to contractualism. On the one hand, it would undermine the aforementioned “emptiness” objection, which is that the contractualist formula (“an act is wrong just in case it fails to pass the non-rejectability test”), if “wrongness” just means “fails to pass the non-rejectability test,” simply means “an act fails to pass the non-rejectability test just in case it fails to pass the non-rejectability test.” If, however, the Scanlonian formula is taken to give a higher-order account of wrongness, then what the formula means is simply that an act that fails the non-rejectability test is made to have another (lower-order) property of being wrong in some other non-contractualist sense.
Second (and Parfit doesn’t discuss this), it would seem to undermine the “irrelevance” objection (which came up last week in comments), which is essentially that the contractualist formula doesn’t seem to be in close enough contact to actual acts to make them wrong in the sense familiar to moral deliberation. When I realize that some action will harm a loved one, it strikes me as wrong solely insofar as it would harm that person, and this seems on its face to have nothing to do with non-rejectability. On Parfit’s revision, however, the fact that it would harm the loved one is what makes this act one that couldn’t pass the non-rejectability test. But that certainly doesn’t deny that “harms a loved one” isn’t itself a wrong-making property of an act. And the same would go for “creates uncompensated disvalue,” or “fails to maximize utility,” where these would also fail to pass the non-rejectability test. Notice, though, that Parfit never talks here in terms of what is the fundamental wrong-making feature of actions; instead, he just says that the contractualist formula is a higher-order wrong-making property, “under which all other such properties or facts can be subsumed, or gathered” (222). This move can thus render the contractualist formula the supreme principle of morality, without also making it the fundamental criterion of wrongness. It gives us unity without monism.
Now this is all quite subtle and tricky, and the only thing I’m sure about is that I haven’t done it justice. I look forward, then, to your comments.