Alan Thomas ‘Critique and Utopia’: Amartya Sen’s Critique of Rawls’s ‘Transcendental Institutionalism’ Friday November 30 15:15

‘Critique and Utopia’: Amartya Sen’s Critique of Rawls’s ‘Transcendental Institutionalism’

15:15 – 16:00


Part of the workshop ‘Fairness and Norms’ that accompanies the 2013 Descartes Lectures



This paper evaluates Sen’s critique of Rawls in The Idea of Justice. It is argued that Sen and Rawls are simply at cross-purposes over the predicates “ideal” and “non- ideal”. Sen’s critique has two components: one diagnostic and one normative. The diagnostic component claims that a “transcendental institutionalist”, such as Rawls, assesses institutions to the exclusion of individual motive; furthermore, he does so from the standpoint of an ideal or “perfect” theory of justice. The normative component claims that the theorist of justice ought, rather, to compare two specific social states via the method of pair-wise comparison without the assumptions of “ideal theory”. The linchpin of the overall argument that connects the two components is this: assume that pair-wise comparisons can be made only if one presupposes the underlying ideal of that social state than which none other is better. This assumption would be analogous to a “chain fallacy”: the scope confusion involved in arguing from “every chain has an end” to “there is an end to all chains”. Sen attributes this fallacy to Rawls: the putative assumption that the method of pairwise comparison is a local instance of the application of the content of a “perfect” theory of justice. Any local and specific appeal to the “more just than” relation between two social states depends on the idea of the “identification of a possibly unavailable perfect situation that could not be transcended”. [Sen, 2009, p. 9] His conclusion is that dropping the goal of formulating a theory of “perfect” justice allows one to proceed with the method of pairwise comparison alone. The goals of Sen’s approach to justice are “enhancing” justice or “removing” injustice with no role for utopian speculation.

This paper demonstrates that Sen’s critique both rests on flawed assumptions and is normatively unacceptable in its own right (independently of issues about Rawls exegesis). Sen’s account reveals a mistake about the use of the terms “ideal” versus “nonideal” in Rawls’s own use. These terms refer to the assumption that a theory is implemented with full compliance or with less than full compliance; it is hard to see how comparisons of social states in terms of justice could dispense with this distinction. Rawls adds further distinctions: between non-compliance that is deliberate and voluntary or non-compliance that it the product of natural limitations or unfortunate circumstance. Nothing here implies that “ideal” is a predicate of the content of Rawls’s two principles; the word “perfect” is used for the relation between those principles and the institutional arrangements that express them. Sen has undoubtedly identified an issue here: Rawls takes it to be important to a theory of justice that it reflexively explain the possibility of its generating support for itself over time. There is a question as to whether these functional criteria are intrinsic, or extrinsic, to Rawls’s process of “construction”. If these criteria are intrinsic to the process, then G A Cohen complained that Rawls was tainting the content of justice. But, on the extrinsic reading that I favour, when one asks the question of whether the theory is feasible under ideal conditions (where full compliance is assumed) or non-ideal conditions (where it is not), this issue is wholly distinct from the truth or falsity of the principles of justice thus applied. Rawls’s functional criteria are not determinative for the truth of the two principles; one can agree with that point while not agreeing with Cohen that this leads to Platonism. So it is not appropriate for Sen freely to substitute, as he does, the word “perfect” for the word “ideal” in his characterization of the content of the two principles that make up justice as fairness.

This use of “perfect” instead of “ideal” maneuvers Rawls into a position where his view can be characterized by Sen as committed not to a transcendental account of justice, but to a transcendent one, specifically, a form of Platonism of justice of the kind explicitly endorsed by Rawls’s critic, Cohen. If “ideal theory” is not “perfect theory” then Sen’s putative methodological innovation can be explained as one aspect of what Rawls would call imperfect compliance theory. The method of pairwise comparison is not a method that Rawls cannot endorse. So it cannot be true that the contrast between “transcendental institutionalism” and “realization-focused comparison” is “quite momentous”. [Sen, 2011, p. 7] Furthermore, there are Rawlsian reasons to reject Sen’s alternative of working within non-ideal theory alone without the normative goal of formulating a realistically utopian view. A. John Simmons has argued that Sen’s restrictive comparativist methodology has untoward consequences in setting ourselves appropriate normative goals in the theory of justice. As Simmons notes, all of Rawls’s adjustments in non- ideal theory are with the overall and integrated goal of pursuing justice as a whole. I examine this problem in the specific case of the claim that we ought to retreat from welfare state capitalism and endorse a property-owning democracy instead. I contrast a Rawlsian view on that issue with one recently put forward by Ingrid Robeyns, where Robeyns puts Sen’s methodology to use. It is argued that Robeyns strategy illustrates the pitfalls of restricting oneself to comparativist claims alone in the theory of justice.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s