I agree that there is a pretty good interpretive case that the later Rawls is up to much more than I have given him credit here. But I also think there is an equally good interpretive case that Rawls (and the Rawlsian project more generally) is guilty of the things I claim. But the real significance of the ideal/non-ideal distinction, at least for me, doesn’t concern figuring out what Rawls was up to (though that is an interesting question as well). It concerns how we go about constructing, and assessing, our normative theories. And so echoing Loren’s approval of Harry’s final point, I agree that we need a sense of both the ingredients and proportions.

Cheers,
Colin

That said, I do wonder if there’s just a hint of condescension lurking in this, the second use of “rescuing” in a draft paper title from Cohen?

“Rescuing” seems strongly to suggest that someone or something needs to be rescued, and that others in a position to give aid are either impotent to do so or fail to recognize the need. Would I be forgiven for suspecting that Cohen’s word choice implies that he (rather than, say, a muddled constructivist or bona fide conservative) is the man for the job?

I find much that is interesting and fruitful in Colin’s work, and I know that Rawls gave us the ideal/nonideal distinction, but I’m not sure the distinction is really all that useful, especially for understanding what Rawls himself was up to, or what we can do in the real world with his approach. Nor am I persuaded that the really critical problem with purported ideal and first-order theorists (Nozick, Dworkin, Rawls, GA Cohen?) is that they ‘ignore too much’ of the messy real world.

For Rawls there seems to be a pretty good interpretive case that his later concerns with reasonable disagreement and public justification (roughly, second-order concerns) are entirely consistent with the constructivist elements of Theory of Justice. True, Rawls gives priority to liberty, but his discussion and deployment of reflective equilibrium (narrow and wide) seems to suggest a far richer approach to reasoning about justice and the real world than you seem willing to grant him, Colin.

More pointedly, without something that looks like “ideal theory” philosophical analysis, I wonder if we can persuasively endorse one nonideal approach and set of prescriptions from competing approaches and recommendations within the same rough issue space?

Finally (and murkily, and perhaps contradicting my point directly above), I’m really not sure what all the fuss is about. If some metaethics and normative ethics types want to learn some history and do some social science to explore questions of justice and legitimacy in the real world, thus complicating their concepts and arguments, then fantastic! If some cognitive and social scientists and historical-interpretive political theorists want to get up to speed on, and contribute to, abstract debates in, say, epistemology, ethics, and (“ideal theory”) political philosophy, then that’s great too! Indeed, this describes a real-world trend.

There are culture shocks and the inevitable faux pas here and there, but by and large there’s much interesting work going on along a spectrum from abstract conceptual analysis and argument, on the one hand, and straight-up practical study and recommendations about what we should do in the real world, on the other. All of us here could together create a vast list of colleagues working somewhere between abstract philosophy and practical theorizing—some are in philosophy departments, others in social science or humanities departments, some in interdisciplinary centers, and some are not in academic employment at all, but instead are doing cool stuff in the real world, yet are strongly informed by theoretical debates they themselves occasionally contribute to.

I guess that, given this fruitful diversity of backgrounds and approaches we’ve seen emerging of late, I’m not sure what’s gained by pushing so hard to fit some folks into the “ideal theory” box with an eye to insisting on their fatal simplifications vis-a-vis the real world. That distinction certainly doesn’t help explain why I find your work on genetics and justice so interesting: my endorsement (everyone, read Colin’s stuff) does not turn on the fact that you consider yourself working in nonideal theory. Nor does my unease with, say, Cohen’s critiques of Rawlsian constructivism have to do with Cohen doing ideal theory.

I agree with you that conditions can vary significantly. I think the most the theorist (even the non-ideal theorist) can hope for is to give us a sense of what the “big picture” (of the moral and political landscape) looks like in particular cases (e.g. justice in affluent, unequal liberal democracies, etc.). So some abstraction is needed, and there will inevitably be some indeterminacy, etc. Thus I think what we (i.e. normative theorists) really ought to be concerned with is what justice requires “many-things-considered”. See my post here:

http://colinfarrelly.blogspot.com/2007/07/what-justice-requires-many-things.html

So for me the key issue becomes- what can we afford to ignore or bracket in our normative theorizing. And I think the ideal theorist makes a fatal mistake by ignoring too much (thus giving us an impotent theory). Granted the non-ideal theorist will face potential problems as well. But I would rather see us invest a greater portion of our energies into tackling those concerns.

And Paul, Yes- I would consider Dworkin (like Rawls) to be a first-order theorist. Granted Rawls is more nuanced in this respect than say Nozick. Nozick is more obviously a first-order theorist- liberty wins the day (end of story). For Rawls the story is- liberty is first, equality of opportunity second and then the difference principle third.

Cheers,
Colin

So basically the central concern with such normative theorizing is to win the philosophical debate concerning the primacy of one’s favored concept- be it equality, liberty, democracy, sufficiency, priority, etc. Whereas a second-order theory is more concerned with determining what would constitute a reasonable balance between the competing values that arise in the debates at the first-order level.

I suspect my use of these terms don’t map neatly onto your distinction between justice and legitimacy.

I agree: these terms don’t map neatly onto the distinction between justice and legitimacy (as I set it out above). But I guess then I find the terms somewhat misleading. I suppose I would prefer something like “myopic” or “fetishistic” instead of your “first-order,” since it is quite possible to reject a single-minded concern with “distributive equality” or “freedom as non-interference” or “efficiency as optimal GDP” and still engage in a sort of “ideal” or “first-best” theorizing that, while useful for the purpose I identify in my last comment above, is not sufficient to guide action in our imperfect world. And even if it fails to hit its mark, I guess I see Rawls’s TJ as decidedly second-order (in your term) theorizing. It certainly makes no fetish out of any particular value in the way that many in the “Equality of What” debate have done when they say, at the outset, “I admit that my discussion touches on only one aspect of justice. It will certainly be necessary, even at the level of ideal theory, to balance distributive equality (of resources, welfare, opportunity for advantage, etc.) against other values, such as efficiency, non-interference, and the individual prerogatives of those whose grew up with an unfair bundle of resources.”

And I don’t think Rawls is as unique in this regard. Consider Dworkin, who in “Equality of Resources” says that his envy test may always fail, since we cannot realistically redress all deficits in “internal resources”. My guess is that you would count this as problematically first-order reasoning, since it simply explores the nature of an ideal of justice that is one amongst many, and indeed one that cannot be fully satisfied and must be weighed against those considerations that stand in the way its full realization. But then, in chs. 8 & 9 of Sovereign Virtue, Dworkin offers arguments for a different envy test, which can in fact be satisfied even when there remain deficits in internal resources. Yet instead of lamenting, as he did in ch. 2, that his ideal of equality of resources can never be realized, he offers the following diagnosis: the fact that it can never be realized in light of the costs it would impose is itself a reason to rethink equality of resources, as defended in ch. 2, as an ideal of justice. The demands of justice must be sensitive not only to the ideals of distributive equality, but also to the costs of implementing it, and an accurate account of what justice is will respond to both of these, rather than elevating one to an ideal consideration while counting the other as a mere practical constraint that impedes the realization of justice in this world.

While I’m not sure that Dworkin or Rawls gets it right in the end, I am hesitant to call either a first-order theorist. And since you suggest that much of TJ is first-order, I suppose you would count Dworkin as first-order too. But this seems to overlook sometimes subtle yet still central moves and themes in their work. And once we’re on the lookout for such themes in the work of others, I wonder just how much first-order theorizing there really is out there. From what I can tell, most political philosophy these days is decidedly second-order, in your sense.

As such, it is possible to suggest that we shouldn’t be providing recipes of this kind – we should be providing an account of which ingredients we should use, and a more or less detailed account of what the relationship between / priority of the ingredients should be. We could make this very detailed and run the risk it won’t be helpful in diverse circumstances (non-ideal theory?), or we could make it very generally-applicable, and run the risk that it won’t provide us with all the details we need in practice (ideal theory?). The cake analogy doesn’t produce an argument for one or the other, it seems to me.

In his brief methodological remarks, Jerry runs together two claims: that philosophers qua philosophers aren’t interested at all in how values weigh against each other, and that they aren’t interested in exactly how they weigh against each other. (For the latter see, e.g. the remark about the person who mentions decor in restaurants.)

The first claim seems to me, descriptively, inaccurate. I can show you lots of philosophical discussions of how, in general, equality weighs against utility or virtue against pleasure. But it may be true that philosophers don’t usually go beyond fairly abstract comparisons to attempt exact ones, and Jerry could be right in so far as he makes just that weaker claim.

Moreover, the claim about interest could be rephrased as one about ability, i.e. that philosophers qua philosophers are no better than anyone else at making exact comparative judgements. They may better at what Jerry’s paper does, i.e. looking at ongoing ethical discussions and extracting from them and making explicit a type of value-consideration that people are using without being fully aware of that fact. They may also be better at formulating abstract possibilities for how values compare with each other and maybe choosing between them. (So they can say that 99% sugar is out.) But they may be no better—they may even be worse!—at making exact comparisons. That may require a kind of context-sensitive judgement that they have no more capacity for than others. And what they’re no better at qua philosophers, they should be hesitant about doing qua philosophers.

I certainly think some humility can be called for in philosophical discussions of concrete ethical issues. Past a point those issues require a kind of judgement that philosophical training need not give you more of and may even leave you with less of.

99% sugar, please!

oh well.

as for the rest of it, i’ll need to read the second half, but the first half looks a lot like the standard conservative dodge of saying “i cannot justify my preferences in any above-board, articulate manner, so i’m going to erect a theory that makes my underhanded, inarticulate inability into a new kind of justification—and claim that it’s a morally superior kind, to boot.”

i’m hoping it gets better.

Professor GA Cohen, (and Harry B too), in writing, in tales from people who know him, come across as thoughtful, kind and gentle person, who is sometimes not interested in doing things he, nevertheless, respects. (Harry, you engage more widely, here at least.) But I see no claims about the worth of disciplines where he distinguishes politics or PT from political philosophy/ethics. So he prefers gardening to cooking – we actually need both.

I actually borrow the term “first-best conceptualism” from Adrian Vermeule (Judging Under Uncertainty) who uses it in the context of legal philosophy, though I think it captures a similar range of issues in political philosophy (indeed Gutmann and Thompson use the term “first-order theory” to capture the same thing in Why Deliberative Democacy). So basically the central concern with such normative theorizing is to win the philosophical debate concerning the primacy of one’s favored concept- be it equality, liberty, democracy, sufficiency, priority, etc. Whereas a second-order theory is more concerned with determining what would constitute a reasonable balance between the competing values that arise in the debates at the first-order level.

I suspect my use of these terms don’t map neatly onto your distinction between justice and legitimacy. But none-the-less, I think a second-order theory can address the concern you have regarding civil disobedience, etc. Indeed, it probably is better suited in this respect than a first-order theory. Rawls’s discussion of civil disobedience in A Theory of Justice is actually a great example. For Rawls presents civil disobedience as a middle position we can pursue when a conflict arises between the duty to obey the law (in a nearly just society) and the duty to protect one’s rights and liberties. And he tries to reconcile these two things by stipulating that the injustices in question must be substantial, that one first try normal political appeals, your actions must be public, etc. So Rawls’s discussion of these things is a great example of second-order analysis of these issues. His primary concern is to reconcile these competing demands of justice rather than give absolute priority to one or the other. I just wish the bulk of A Theory of Justice followed the example Rawls outlined in the chapter on civil disobedience.

Cheers,
Colin