I think that’s right. I might think that there are many ways one can work here – bottom-up and top-down. So, you might start from the the different versions of the substantial moral principles that accept the deontic distinctions to different degrees and assess which ones of them could be reasonably rejectable piecemeal. Once this process is done, you could ask whether the set of principles recognises deontic distinctions in general, to what degree and in how many contexts. If the set does recognise these distinctions to some significant degree, you might think that this is a contractualist vindication for them.

But, at the same time, you could start from the different substantial sets of principles that are so far neutral about the distinctions. You could then modify the sets (whilst keeping their other substance constant) with different sensitivities to the deontic distinctions and compare the reasons to reject the sets. Again, it might turn out that some degree of deontic constraints is recognised.

Of course, these are two different processes but once carried out their results should converge if there is *the* set of principles that is not reasonably rejectable.

Your point is a good one. I considered something along the lines you suggest for justifying the relevance of certain distinctions, but wasn’t persuaded it would work, for the following reason: it seems to me that if general distinctions like intending/foreseeing were to be justified in contractualist terms, they would have to be justified with respect to a particular principle for the regulation of deliberaiton in a particular type of situation. So I don’t see how the general salience of this type of distinction is to be justified. Rather, it would have to be justified as relevant in certain contexts, and its relevance would be explained by certain other considerations. In this respect, the kinds of distinctions Kamm is concerned with strike me as quite different than, for instance, something like fairness, whose salience as a reason for rejection is readily illumined in contractualist terms. But this isn’t a decisive case for thinking that distinctions like intending/foreseeing can’t be justified on contractualist grounds.

“Her general argument is that the advertised view is false; contractualism presupposes, without argument, the justified relevance of many of the kinds of key distinctions that non-consequentialists take to be important—like intending/foreseeing, harming as a means vs. a side affect, and harming vs. not aiding—but whose relevance they accept as in need of justification.”

I think there is a way in which Scanlon could accept that he uses these deontic distinctions (if he did) but not without argument or justification. He does often emphasise the piecemeal nature of contractualist deliberation. We point our attention to one moral principle whilst holding others fixed as also possible grounds for objections. In this way, it is not a problem that he uses deontic distinctions in assessing other principles.

What he will have to do in some point is to assess the reasonable rejectability of the deontic distinctions whilst holding other principles fixed. This would be to compare worlds in which different distinctions or none at all are appreciated and looking at what kind of standpoints are created in the long run to different individuals. I actually think that some forms of the distinctions could hold their ground and not be reasonably rejected. It’s true that this isn’t done in WWO but it is a long book already.

I said my piece on the Gap Question in my ‘Contractualist Replies to the Redundancy Objections’ in Theoria. Anyway, looks like I’ve got more reading to do.

I guess I’m in the minority here in thinking that is far from obvious that Kamm is right about contractualism’s reliance on further, undefended or analyzed, non-consequentialist distinctions, and in fact believe that Kamm is wrong about this. Your first point, Matthew, is about this, pressing the question of “why is it the case that one should be entitled to means of saving one’s life, provided no one has a prior claim to it”.

My suggestion, put more broadly, is that one thing persons could have reason to care about is, for instance, having some kind of right to control that which they secure through some form of relevant effort. That is, a principle that did not make this kind of control over what a person has secured possible is one that could be reasonably rejected, as the kind of control in question is something individuals reasonably care about being able to have in certain circumstances. Now, I didn’t try and work out what the precise principle would be in the life jacket case, and in particular, what its justification might be, but I don’t see anything suspect in the general kind of justification I’ve gestured at. The kind of reason in question is something that is perfectly acceptable in contractualism, and is part of the force of claiming that reasons for rejection don’t all have to do with welfare. Scanlon cites fairness as a relevant example of a non-welfarist consideration, and there are many more examples in the chapter on responsibility of WWO.
A better discussion of the point would have to work out the details of the relevant principle, but I don’t see any reason yet for thinking that the principle is going to invoke, as part of its justification, the kinds of distinctions Kamm takes to be relevant.

On your second point, concerning aggregation: why can’t we allow aggregation in the assessment of how much there is to blame a person for? All contractualism claims is that principles assigning blameworthiness do not take into account, as part of their justifications, aggregative considerations. The theory isn’t committed to the claim that aggregative considerations cannot be relevant for understanding other aspects of moral assessment of an individual’s conduct. Which isn’t to say I think that aggregating in assessing how much there is to blame is the right way to go, but just to say it isn’t precluded.

Gerald, thanks for your comments. I do think the value of mutual recognition is relevant for assessing RR, and in particular, I think its important for justifying the salience of various kinds of symbolic and non-welfare implications of proposed principles the appeal to which can justify RR a proposed principle. I didn’t discuss what Kamm says about objective value and the value of mutual recogition because I would have had to get entangled in a long discussion of Scanlon’s account of morality’s authority. But you’re right: some attention to the value of mutual recognition, and the role it plays, will no doubt further help dispel the sense that contractualism helps itself to certain non-consequentialist distinctions.

Maybe I shouldn’t be satisfied with what Kamm says about the gap question, but I felt I never understood what Philip is getting at in his Analysis article. So maybe I can ask you to say more about why you think Kamm’s analysis is inadequate? Her point, as I understand it, is that if we just look to the wrong making properties, we leave out a crucial fact that the person has a claim against you not to be treated by you in that way. Explaining why it is that the relevant wrong making properties do have a claim on us for our attention will have to be, then, an explanation of why it is that these considerations are considerations that one has reason to be attentive to insofar as one is concerned with not wronging others (or standing in a relationship to others of mutual recognition). I like this picture, but you don’t?

Oh, I also promised to put in a plug for the Kamm Fest at Rutgers Camden Law School at the end of February. Featuring McMahan, Scanlon, Shriffin, Kagan, and Gideon Rosen, it will certainly be an outstanding occasion.

I agree with both of Matthew’s points, however. It seems fairly undeniable to me that Scanlon is implicitly relying on a number of non-consequentialist distinctions in his endorsement of principles that survive the reasonable rejectability (RR) test. On one view, that doesn’t sink his contractualism, but indicates rather that it is very much work in progress. And perhaps that doesn’t matter so much: which philosophical system doesn’t have loose ends to clear up?

But Kamm’s argument is more threatening than this suggests. For the tenor of her criticism is that her brand of intuitive non-consequentialism is the only way to go (see esp. pp. 473-4). At this rate, we will be left with a squashed lump of clay (namely, the non-consequentialist intuitive data), at the cost of having lost the statue itself (namely, contractualism). So what does Scanlon need to do to avoid this outcome?

One thing he might resist is Kamm’s suggestion that RR is concerned solely with the objective value of human persons – those beings to whom justifications can be given and understood. Scanlon operates with a more heteronomous picture, as Kamm points out (p. 460) – he cites, as an example, the ideal of living in unity with our fellow creatures. Scanlon can’t afford to lose this cast of justification. And I don’t think he needs to – yet. It depends on whether a consultation of those ideals will prove to be more fruitful in the arrival at RR-surviving principles than simply a recitation of our objective value. I wish I had more to say about how those ideals will concretely help. I can’t help suspecting they do help, though.

Now let me turn to the ‘Gap Question’ (p. 462), which has been much discussed in the literature. To put it another way, why doesn’t Scanlon pass the buck for wrongness as he does for goodness? Rahul seems content to along with Kamm’s revisionary suggestions here, but I’m not fully convinced.

Philip Stratton-Lake wrote a helpful article about this a few years ago (‘Scanlon’s contractualism and the redundancy objection’, Analysis, 63.1, January 2003, pp. 70-76). In acting on moral reasons, conscientious agents will tend to be moved by an act’s wrong-making features, rather than by the mere fact that the act is wrong, or that it falls under a principle that could be reasonably rejected. But, as Stratton-Lake points out, that still leaves room for a distinction between an analysis of the property of wrongness itself and an analysis of wrong-making features. Of course, it looks like over-counting to continue to insist that an act’s being wrong is directly reason-giving in addition to the reason-givingness supplied by the act’s wrong-making features. Stratton-Lake suggests, accordingly, that Scanlon should just surrender the claim that wrongness is directly reason-giving.

Perhaps Scanlon can afford to do that. But I suspect he can perhaps hold out for something slightly different. For it’s not just the case that X’s wrong-making features provide a complete explanation of why I have a reason not to do X. For we can raise the further question: what makes it the case that X’s wrong-making features are reason-giving, given the competition for our moral attention and moral resources? We need to appeal to a further ideal, or higher-order constraint on what justifications pass muster. RR is obviously going to be central to Scanlon’s answer to this further question. But then, it seems to me, he needn’t relegate RR to an account of wronging, rather than an account of wrongness.

Now compare buck-passing about goodness. What matters to X’s being good are the goodness-making features which provide us with reasons for responding to X in certain ways (promoting, upholding, or honouring X). Consider the question: what makes it the case that X’s goodness-making features are reason-giving? Either we’re going to go in a Moorean direction at this point, by citing special non-natural properties, or else we’re simply going to shrug our shoulders. I therefore think there’s a difference between goodness and wrongness here which can justify Scanlon’s refusal to pass the buck for wrongness though not for goodness.

1. You said,

“In the shipwreck case, for instance, couldn’t it be argued that the assurance that comes with being entitled to means of saving ones life, provided no one has a prior claim to it, be something that those at risk in cases like shipwreck have reason to care about? If so, it is arguable that a principle for cases like shipwreck that did not allow for the possibility of someone who comes to have the means to save their own life being entitled to it could be reasonably rejected. It isn’t obvious, then, that there aren’t ways of explaining Scanlon’s conclusions about the shipwreck case that don’t rely on the distinctions Kamm takes to be relevant.”

Does Kamm need to deny that there may be other ways of explaining the results of the Shipwreck Case? It seems that all she wants to say is that the explanations (including yours) focus on how costs to people come about, and that these explanations are in need of a justification, which contractualism may not provide, at least directly. For example, take your principle of “being entitled to means of saving ones life, provided no one has a prior claim to it.” It seems that Kamm would not deny that this principle could explain the results of the Shipwreck Case. What concerns her, I think, is that it is not obvious that this principle is directly derivable from contractualism. That is, why is it the case that one should be entitled to means of saving one’s life, provided no one has a prior claim to it? Rightly or wrongly, she wants to say that Scanlon helps himself to these kinds of principles, when Scanlon should be providing justifications for them.

2. You said,

“Kamm argues that Scanlon’s tie-breaker argument has difficulty accounting for the fact that saving the smaller number becomes a greater wrong as the number in the larger group grows. Though she offers a clever solution consistent with Scanlon’s model for responding to this objection, I didn’t see why one couldn’t claim that what drives our intuitions about this is not the fact that the wrongdoing is more serious, but that as the number left to die grows, there is simply more to blame the blameworthy for.”

Take your solution, couldn’t Kamm respond that, on the tie-breaking model, one would still need to explain why “there is . . . more to blame the blameworthy for” without aggregating? If so, she might argue that her solution is still necessary.