As argued before, Rawls is very sensitive to the significance and roles of epistemic uncertainty in the original position. (In this he is a good student of Frank Knight (recall Arrow's response).) So it should come as little surprise that despite superficial features to the contrary ("full awareness of the relevent facts"), Rawls recognizes various features of uncertainty in his discussion (inspired by Royce) of rational plans of life which extend over the whole of one's life (A Theory of Justice, 358ff). One way this fact is signaled, is Rawls's embrace of Herbert Simon's satisficing in making subplans permissible (367) and (echoing L.J. Savage's treatment of uncertainty) adopting the move from objective to "subjectively rational plan[s]" (366). A nice pay-off of this latter move is that it blocks (many) paternalist impulses and respects autonomy.
Moreover, Rawls recognizes that (i) external circumstances are not always favorable to rational plans--the "achievement of happiness...always presumes a degree of good fortune" (358-9); (ii) in the initial planning stage "only the outlines of [the hierarchically structured sub-] aims and interest can be foreseen, the operative parts of the subplans that provide for them are finally decided upon independently as we go along." (361) That is, there is epistemic uncertainty about both the context in which we wish to execute our plans (that is out of our control) and uncertainty about a lot of details of and the means to implement our plans. While in (i) and (ii) uncertainty is external to our own control, it should be noted that in (ii) part of the uncertainty is also about how we develop as we execute and fulfill our plans. (I return to this feature below.)
Even so, a core commitment of Rawls's approach to a rational plan of life is that "we can choose now which desires we shall have at a later time" (364) as well as the attitudes we might have toward those desires. (Again, Rawls allows some uncertainty about the exact content of the plan, which is why he also sometimes talks of a "class of plans.") That is to say, our earlier self can command how later selves are manifested even if the details of this manifestation are unknown in advance. What allows the earlier self to legislate, as it were, to the later self is the existence of "permanent aims and interests," (361) relative to stable (and favorable) background conditions. In Rawls's hands such planning is explicitly analogous to, but (because of the Knightian suspicion that one has no knowledge of “the likelihoods of the possible circumstances” (134)) not identical with, the kind of utility maximizing models we find in formal epistemology.
This is not the place to articulate the many roles that rational life planning has in Rawls's scheme. Much of the technical debate has involved the kind of risk aversion in the plan(s) we should adopt in the original position. There has, of course, been criticism of the desirability of a rational plan (see this excellent article by Charles Larmore). But to the best of my knowledge there has been no suggestion that the conditions that a rational plan are supposed to meet may not be fulfillable even under fortunate conditions.
L.A. Paul's work on Transformative Experience (TE; here's the book), which is self-consciously articulated as a challenge to first-person rational choice models, also undercuts the viability of the very idea of a rational plan that is held together by -- Rawls's terms -- "permanent aims and interests." Assuming that regular readers of this blog only need to be reminded of what the formal features of TE are, here's a useful introduction to TE:
An epistemically transformative experience is an experience that teaches you something you could not have learned without having that kind of experience. Having that experience gives you new abilities to imagine, recognize, and cognitively model possible future experiences of that kind. A personally transformative experience changes you in some deep and personally fundamental way, for example, by changing your core personal preferences or by changing the way you understand your desires, your defining intrinsic properties, or your self-perspective. A transformative experience, then, is an experience that is both epistemically and personally transformative. Transformative choices and transformative decisions are choices and decisions that centrally involve transformative experiences...[I]n cases of transformative choice, the rationality of an approach to life where we think of ourselves as authoritatively controlling our choices by projecting ourselves forward and considering possible subjective futures is undermined by our cognitive and epistemic limitations.
For the sake of present argument, I stipulate that L.A. Paul's position on TE is not just coherent, but will survive scrutiny (obviously, a big if (see here for my original, critical response.). One consequence of Paul's approach is that rather than merely disclosing or manifesting ourselves through our actions, we may really discover we are not who we wished to be, and could not be so. (See here for some musings on my part.)
Recall from above that a Rawlsian life-plan is subjectively rational. What's crucial for present purposes is that the very possibility of a transformative experience should undercut our confidence that genuine rational life planning is possible in all respects with regard (and this is key) to one's ""more important aspirations being fulfilled" (Rawls, TJ, p. 359). What Paul shows is that authoritatively legislating to future selves from a first-person perspective is a fragile enterprise with no insurance available.+
The best one can really do in good faith, once one has been reminded of Paul's lessons on TE, and here I quote Frank Knight (1922) is: "For the time being, an individual acts (more or less) as if his conduct were directed to the realization of some end more or less ascertainable, but at best provisional and vague. The person himself is usually aware that it is not really final, not really an "end"; it is only the end of the particular act, and not the ultimate end of that."* (475) (Recall.) But this position is too thin even for Rawls's purposes.
+This position came up in discussion with the late Ned McClennen when he taught Michael Bratman's views on planning once, but we lacked Paul's way of conceiving the situation.
*Knight also thought that even in the economic sphere stability of desires over time could not be taken for granted: the "wants which impel economic activity and which it is directed toward satisfying are the products of the economic process itself" (1922: 457)
Comments