Tuesday, May 31, 2005
Seidman on reflection and rational endorsement
The reflective structure of her consciousness gives an agent the capacity to turn her attention not just onto the world, but also onto the beliefs, desires, and other mental states which the world causes her to have. And the reflective structure of her consciousness gives an agent the capacity to distance herself from these mental states, and to ask whether she ought to endorse them -- whether she has good reason to harbor them, and to allow them to inform her deliberaation about what to do or what to believe. This is an instance of what Korsgaard calls "the Normative Question". The problem which the capacity to ask the Normative Question poses, Korsgaard argues, is the problem of mustering the resources to answer this question in a rationally satisfying way -- the problem, as Korsgaard sees it, of finding the grounds to justify allowing some mental state to inform one's deliberation. If we allow our deliberation to be shaped by mental states which we cannot justify, Korsgaard claims, then we are less than fully rational.
Seidman's project in his paper is to show that Korsgaard has placed the justificatory bar too high. As the passage above makes clear, Korsgaard maintains that an agent has failed to be rational whenever the agent cares about something without having good reason for doing so. Seidman argues that this standard should be weakened to say that an agent violates rational norms just in case the agent continues to care despite having good reason not to.
As the tenor of my recent posts may indicate, I'm inclined to side with Seidman on the large question here. Korsgaard's view, it seems to me, alienates the agent from her empirical self and in so doing commits us to a misdescription of much of what is important in human life. That said, I think Seidman's gloss of Korsgaard contains a serious omission and that this omission prevents him from really engaging Korsgaard's thought. The key here is that for Korsgaard the difficulty isn't just that we might turn out to be 'less than fully rational', but that insofar as we fall short of rationality we fail to constitute ourselves as agents. In underselling the depth of Korsgaard's concern, Seidman avoids accepting the burden of showing that his alternative standard resolves Korsgaard's underlying worry. As such, the argument he presents doesn't really amount to a (direct) challenge to her view.
But that's all scholarship, and if politics is applesauce then scholarship is no better than horseradish. In the remainder of this post I'll look at Seidman's argument on its own terms without worrying whether it amounts to an effective rejoinder to Korsgaard.
In talking about practical rationality, Seidman's most basic analytic commitment is to a particular conceptual claim about what he calls 'concerns.' The claim is that "an agent's concerns dispose her to regard certain features of her environment as giving her reasons for action." So understood, concerns are the ground for an agent's first order reasons and generally, to use Seidman's words, 'comprise the background against which she deliberates.' The problem, as Seidman sees it, is to show that there is no rational failure involved in having a concern which has not been ratified by a process of rational deliberation. He takes this to be a question about when concerns require justification and his central contention is that they don't require justification unless some kind of challenge has been raised.
Seidman's general strategy in establishing this contention is to argue "that for someone who harbours a given concern, and who has no pressing reason to set it aside or give it up, whatever external justifications are possible are quite beside the point." The idea appears to be to undermine the notion that rational justification is necessary by showing that there is no role for rational justification to play in these cases.
Before continuing, let me note that the locution 'external reasons' is unfortunate here. The phrase appears to commit Seidman to the substantive thesis that justificatory reasons must necessarily be external reasons. Since all internalists hold that justificatory reasons are internal and some internalists hold that the very idea of an external reason is incoherent, such a thesis would be extremely controversial, and yet Seidman doesn't present any arguments for it. In fact, he explicitly declares that he intends, "to remain agnostic as to what sorts of considerations can be levied in legitimate rational criticism of a concern, and what sorts of considerations can be appealed to in order to justify allowing some concern to inform one's deliberation." In light of this, it seems to me that the most charitable way of reading Seidman is to suppose that when he uses the phrase 'external reasons' he means to be referring to justificatory reasons, whatever those turn out to be.
For the purposes of the argument, the relevant concerns are those which ground some of the agent's practical reasons for action, but which were not accepted by the agent on the basis of rational deliberation. My own preference for aspirin is a paradigm of such a concern -- it was, in a sense, given to me rather than arising as the result of a deliberative process. In such cases, Seidman argues, we can often produce justificatory reasons which would be sufficient to justify opting into the concern if we didn't already have it. Those justicatory reasons, however, don't track the actual reasons we would give in favor of the concern given that we already subscribe to it. Moreover, Seidman argues, our actual reasons will typically presuppose commitment to the concern at issue and will, as such, fail to be justificatory reasons for it. The upshot, Seidman concludes, is that rational deliberation can't play any role in grounding concerns in such cases.
I won't try to develop a full critique of this line of argument here, but let me briefly suggest a couple lines of objection.
One family of objections to Seidman's view focuses on his apparent commitment to the claim that the reasons which can justify rational endorsement of an immanent concern must be drawn from the stock of reasons which would justify opting into the concern. Many philosophers have held pretty much the opposite view, arguing that values and the reasons which support them cannot be fully perceived by someone who isn't already committed to the values. The most striking historical example of this is probably Aristotle's contention that the question of the rational basis of the ethical life can only be pursued by those who have themselves acquired virtue, but it is a well established and plausible view.
A related objection has to do with Seidman's claim that reasons which are derived from a prior commitment to a value can't be used in a justification of that value. Presumably Seidman's thought is that justification will be blocked on account of circularity. Well, in the first place, not all circles are vicious so even if Seidman is right about the circularity he has more work to do to show that there really is a threat to justification here. Second and more importantly, it isn't so clear that prior commitment must lead to circularity. Hume famously argued that the particular pleasure of moral goodness is only available to the virtuous. Since that pleasure is only available to those who are antecedantly committed to the value of moral goodness, it is only in virtue of their prior commitment that the pleasure can function as a reason. What's important to notice is that the connection between the commitment and the reason here is empirical rather than logical and that, as such, no circle arises. One can make similar points with regard to the epistemology of value. Prior commitment to a value exposes the individual to the operation of the value and in so doing grounds judgments about the functioning of the value, judgments which would not be available to someone who did not share such commitment. Again, the connection between commitment and reasons is real without being rational and no circle arises.
1 If, on the other hand, Seidman is talking about genuine external reasons then internalists will agree with his intermediate conclusion, but not on helpful grounds. They will agree, that is, that external reasons are beside the point 'for someone who harbours a given concern, and who has no pressing reason to set it aside or give it up', but this claim will be accepted because the internalist holds that external reasons are always beside the point.
2 Seidman is here treading on some of the same ground covered by Evan Tiffany. Some of my analysis of Tiffany's view can be found here.