“Explanation is about what makes something true, justification about what makes it reasonable for us to believe that it’s true—and the two are entirely different. Cohen, I believe, is talking only about the former—about moral metaphysics, not moral epistemology. And his views about moral metaphysics leave him perfectly free to adopt reflective equilibrium as his moral epistemology.”

That Cohen is perfectly free to adopt reflective equilibrium as his moral epistemology doesn’t mean that he in fact (coherently) does, because he may fail to make the perfectly sensible distinction between epistemology and metaphysics that Tom does.

Jon Mandle #7

“I referenced, but didn’t discuss, Cohen’s note 19 on p.243. In that note he says: “Nor need I deny that, to employ Geoffrey Sayre-McCord’s description of the ‘method of reflective equilibrium,’ ‘the process of developing an acceptable moral theory is a matter of shifting back and forth among the various moral judgments one is initially inclined to make and the more or less abstract theoretical principles one is examining …’‘” After the quote, he adds: “What I would deny is an expanded description that adds factual beliefs to the mix.”

Cohen seems to think that the metaphysical claim about explanation applies to the epistemological issue of justification, since he thinks that reflective equilibrium must exclude facts. This is why, presumably, as Jon points out in the post, his language about the precise content of his thesis is so sloppy in the chapter (I admit, I’ve only skim read the chapter, having read F&P some time ago; although the appendix about God is quite interesting). The question then become whether the polemical uses which Cohen puts the thesis to are ones that a thesis which seems to have no defensible – unless someone wants to defend them – epistemological implications can be put to. Is the original position machine supposed to be giving us a metaphysical explanation, or an epistemological justification of the principles it generates? Would it make sense for, say, a view like the later Rawls’ which is supposed to be abjuring from deep metaphysical claims to regard the original position as explaining what grounded, rather than justified, the principles it generated? Does it follow, as Cohen claims, that because Rawls says that the difference principle wouldn’t cause too much inequality that an equality principle is prior to the difference principle, rather than the other way round (as in, ‘do things which help people pursue their projects’; why?; it means people ought to keep their promises)? Are we ever going to be really interested in what metaphysically explains a principle as opposed to what justifies it epistemologically?

The hypothetical setup is basically this: suppose Rich has 2 units of corn, while Poor has none. Suppose further that eating corn is not a matter of life or death, but that those who have no corn are forced to eat something horrible instead. To have one unit of corn is to have enough to eat, while to have two units of corn is to have so much that eating the second unit would make you sick. So the marginal utility of consuming the first unit is high, and the marginal utility of consuming the second unit is very low. The standard utilitarian-egalitarian would conclude that in order to maximize utility, we should take the second unit from Rich and give it to Poor; but this doesn’t follow.

It doesn’t follow because what has not been taken into account is the possibility of production. It is precisely because there is a very low marginal utility of consuming the second unit of corn that Joe Rich prefers to plant it, rather than eat it (having been already satiated by the first unit he has eaten.) If you do take the corn away from Rich and give it to Poor, you have a situation where all the corn is eaten, leaving none for planting – which means that there is no corn for anyone to eat for the indefinite future (and, needless to say, this is an outcome with lower long run utility than just leaving both units in the hands of Rich.)

It is of course still possible in the real world that utility could be maximized by widespread redistribution, but it certainly doesn’t follow with anything like as much ease as people often think it does.

Well of course you can ask!

I understand the confusion: I don’t mean, of course, simply that it is physically possible to ask. In light of my examples and arguments above, I think that it is open for me to ask in the sense of being an open question whether suffering in some factual context actually counts as something bad. If this is true, then it really doesn’t look like “suffering is bad” is any kind of ultimate principle, of justice or any other kind. If suffering only counts as bad in some contexts, but not in others, then whether “suffering is bad” is true depends in part on the facts of the case. I believe this means that it cannot be ultimate in Cohen’s sense.

But here’s the point: I think that it is always open to me to ask, “Why is suffering here bad?” or “Why should it give me a reason?” When I do this, I’m asking for justification.

Well of course you can ask! (Your formulations here aren’t always helpful, by the way, since “Why should it give me a reason?” might have a different focus to the one we’re interested in here.)

By highlighting that in some circumstances suffering is not always bad, and the different point that “suffering is bad” is not always even a pro tanto reason, I mean to show that it is not an unjustified justifier.

Now, someone may offer suffering as a reason to do or avoid doing, and I may accept this reason without comment or concern. But here’s the point: I think that it is always open to me to ask, “Why is suffering here bad?” or “Why should it give me a reason?” When I do this, I’m asking for justification. (Looking at what Chris says in 35, I want to acknowledge that this request may indeed sometimes seem weird, but I think the weirdness is just that in some cases it seems obvious that suffering is bad. But this does not mean that a justification cannot be given – it doesn’t mean that there is no reason, as with the uncaused cause there is no cause.)

In fact, doesn’t Cohen’s position entail that there aren’t any unjustified justifiers, other than the ultimate principles, whatever they are (and whether they are principles of justice or something else)? Cohen holds that any principle that is non-ultimate can be explained/endorsed/affirmed/justified/grounded in some other principle, all the way up to the non-fact-dependent ultimate principles. So, unless “Suffering is bad” is an ultimate principle of justice, then I don’t think it will count even for Cohen as what I’m thinking of as an unjustified justifier.

Your position seems to be this: that justice is basically a matter of institutions and that we ought to make these in a way that maximizes the good, where “maximizing the good” is given content by your equality-adjusted utilitarianism. So we “make” justice in the sense that all there is to justice is the devising of institutions, and where this may vary from place to place and from time to time, depending on the facts.

That’s not a view I agree with, but it is certainly one I recognise as respectable. So you clearly disagree with the whole idea that “justice” has any kind of free-floating, pre-institutional content. Again, fair enough, though I’d just register the suspicion that your views about equality may commit you to more than you suppose!

What goaded me (and Tom, I think) was the suggestion both from you and from some of the orthodox Rawlsians in the discussion, that people who think of justice as an independent value in the way Cohen does, are thereby committed to some kind of metaphysical weirdness that you are not. We may or may not be committed to metaphysical weirdness (I’d rather not be, but I’ll bite the bullet if I have to), but everyone who grants authority to any kind of objective normative standard (does’t have to be justice, doesn’t even have to be part of morality) is in the same boat.

My interpretation of Bentham obviously entails the stuff you mention. OTOH, I’m happy to give arguments in support of them if they are challenged, so I don’t think it’s right to describe them as presuppositions. Which do you have a problem with?

On a different tack, I started by saying that the view I hold seems to me to be more consistent with the statement “justice is something we make” than with “justice is part of the world as it exists”. Tom disputed that – I’d be interested in your views.

Finally, sorry for derailing the thread to some extent. In my defence, the points raised were suggested to me by my reading of Cohen and the discussion here, even if I have gone off at a bit of a tangent.

I don’t think that you are getting the point that Tom is making, namely that you are taking a standard as the right one and that you (must) think that people who take a different view are making a mistake. Something (some standard or principle) is supplying your interpretation of “best consequences” (and, for that matter, explaining why equality is of value). Suppose you respond by saying, as I suspect you’re tempted to, that your view is based in what people want, or choose, or prefer, you still have to provide an explanation of why a state of affairs in which people get those things is better than one in which they don’t. Perhaps you shrug and find the demand for some justification of that weird (in which case see “unjustified justifier” and “suffering is bad” above) or perhaps you gesture to some further, deeper principle …

On equality, I can’t see how you can’t see that your interpretation of Bentham above presupposes all kinds of stuff (equality as default, no intrisic differences in moral worth among humans, possibly hypothethical differences in moral worth between humans and some other creatures justifying weighting differences).

(Attempting a restatement):
(a) The view that justice is a thing-in-itself, or part of the way the world is, seems to me to be inconsistent with consequentialism (in my interpretation, the view that the social institutions that should be described as “just” are those which we collectively judge as likely to have the best consequences).
(b) In making a collective judgement about the best consequences, we are, in some sense, making justice
(c ) Independently of the above, I’m specifically advocating procedures in which all count equally, and in which more equal allocations are preferred to less on the grounds of contingent facts about humans

As I say, I could easily be getting all this wrong. But I don’t see how anything I’ve written lumps me in with “Utilitarians, economists, etc. [who] like to claim their view has preferred status over alternatives,”, unless you mean “preferred by me”.

Granting that this is true, so what? Diminishing marginal utility* applies equally to lifetime consumption streams as to consumption at a point in time. Taking account of production doesn’t change anything. It’s still true that transferring half a grain from Jane to Joe will increase aggregate lifetime utility for the two of them.

*Note that I’m using this formulation since that’s the one that has been challenged. We can do the same thing in Rawlsian terms if we want.

I don’t think “then some form of egalitarianism in the provision of goods and services follows from diminishing marginal utility” is true at all. Even with the assumptions that everyone has a similar utility function, and redistribution is costless, and there are no adverse incentive effects caused by redistribution, the conclusion doesn’t follow, as I think is pretty clearly shown by David Schmidtz in his paper Diminishing Marginal Utility and Egalitarian Redistribution.

I’ve been told off on here before for pointing this out and not including the argument, so here goes: in a world where there is production as well as consumption, DMU might be precisely what causes people to invest rather than consume, which increases the total stock of goods produced, and thereby increases utility. To follow Schmidtz’s example, it may well be true that Jane Poor gets more utility from consuming an extra unit of corn than Joe Rich, but it is this very fact that makes Joe Rich more likely to plant (rather than consume) his second unit of corn than Jane. Taking it from Joe and giving it to Jane would mean that there is less corn to be enjoyed as a result. I think you fall into the trap Schmidtz mentions, along with a lot of other smart people, of thinking that even though the DMU argument yields strongly egalitarian conclusions only in a world without production, some suitably weakened egalitarian conclusions still hold in a world with production. But this is not necessarily the case.

Chris B,

What I’m not clear about is whether the principle that decides whether welfare should trump justice or not is itself a principle of justice.

It seems as though Cohen has ruled out the possibility that it is, by all his talk of ‘all-things considered’ judgements, which he explicitly distinguishes from considerations of justice. So yeah, it does seem as though there must be a higher level normative principle (or set of principles) that resolves any conflict between the important considerations (justice i.e. equality for Cohen, Pareto, welfare, liberty, etc) that we want to take into account. But it also seems as though these higher level principles have a better claim to being principles of justice than what Cohen has put forward as principles of justice, because they (the higher level principles) are the fundamental principles that we use to decide, crudely speaking, who should get what.

That’s not a factory tour at all—you’re buying ready-made.

1) A ‘summary description of desirable outcomes’ makes claims about which outcomes are desirable—claims it apparently treats as independently true, i.e. as having independent existence.

2) And ‘consequentialism plays the role of unmoved mover here—I don’t derive it from anything else’ makes precisely my point. You treat consequentialism as underivatively true, in just the way Cohen treats his principles of justice. And there’s no reason to think consequentialism or any other view has preferred status here. Utilitarians, economists, etc. like to claim their view has preferred status over alternatives, but it’s a ploy that doesn’t fool anyone. We’re all in the same boat.

Tom, I won’t claim that “make” is exactly right, but here’s how my “something like utilitarianism” is produced .

First, I think, like Bob Goodin, that utilitarianism and its variants are best thought of as public philosophies, not as guides to individual ethics. I think (more so following the discussion we’ve had of Cohen) that something similar is true of “justice”; it’s a summary description of desirable outcomes that can be pursued in political and social action, not an abstract principle with an independent existence.

As regards utilitarianism, I support egalitarian (in the sense I’ll describe) versions of consequentialism of which utilitarianism is the oldest. I guess consequentialism plays the role of unmoved mover here – I don’t derive it from anything else. My support for egalitarianism starts with (my interpretation of) an argument by Bentham. Starting from the default position that, in the evaluation of political actions, everyone should count equally, Bentham argues that no one person or group is better than the rest in a way that would justify (to the rest) a claim that they should be given special weight in evaluating outcomes. This claim is contingent – if there really were a class of beings categorically different from ordinary humans in some way recognised by all, it wouldn’t apply.

That gets as far as “consequences for everyone should count equally”, and then some form of egalitarianism in the provision of goods and services follows from diminishing marginal utility (or some variant for different ways of evaluating welfare), as expressed in Rawls or Harsanyi style reflective equilibrium.

That concludes your factory tour.