In the introduction to Liberalism and the Limits of Justice, Michael Sandel, a well-known critic of liberal political theory, calls attention to two different ways in which philosophers have understood the right as prior to the good. In the first, “moral” sense, this idea describes a first-order ethic that stipulates certain absolute duties and prohibitions that override other moral and practical concerns, even the overall welfare of a political community. In the second, “foundational” sense, it describes “a form of justification in which first principles are derived in a way that does not presuppose any final human purposes or ends, nor any determinate conception of the human good” (3). Now, some liberal theorists, like J. S. Mill, the famous nineteenth-century utilitarian, affirm the first sense of the right as prior to the good but not the second; for Mill, justice is merely a matter of social utility, and utility is metaethically prior to the (more or less) absolute rights and duties that tend to promote it. Conversely, for Immanuel Kant and liberal idealists of his ilk, the two senses in which the right is prior to the good are intimately connected: certain duties and prohibitions must take unqualified precedence over other moral and practical concerns, since first principles of right are derived a priori and not from unstable and epistemically dubious empirical circumstances. One of the central assertions in Liberalism and the Limits of Justice is that John Rawls, the most influential liberal political theorist of the twentieth century, inherits and modifies this Kantian connection between the two senses in which the right is prior to the good, which, Sandel contends, presupposes a certain conception of the self as prior to its ends. That is, a subject for whom justice is primary in the ways Rawls insists must be a subject “unencumbered” by constitutive purposes and values, even if Rawls himself disavows the Kantian conception of the noumenal self and its basis in transcendental idealism. It is this portrait of the “unencumbered” liberal self that Sandel calls into question.
To be sure, Rawls aims to disjoin the primacy of justice (in the moral and foundational senses) from its Kantian metaphysical basis so as to render his theory of justice more plausible; he in no way seeks to defend transcendental idealism as a requisite component of justice as fairness. To this end, he appropriates from Hume the “circumstances of justice,” i.e. the empirical conditions that make human cooperation both possible and necessary, in his description of the original position, which effectively replaces the noumenal self of transcendental idealism. In this way, Rawls thinks that he can preserve the moral force of the primacy of justice within the scope of an empirical theory. Yet this attempt to situate his theory somewhere between Kant and Hume, Sandel claims, is unsatisfactory: the “notion of right which Rawls seeks to recapture derives its force from a moral metaphysic that rules out precisely the appeal to [empirical] human circumstances on which Hume’s account of the virtue of justice is based” (36). For Kant, any appeal to empirical circumstances in the attempt to formulate moral laws is conditional and, for that reason, unreliable, as the principles derived from such circumstances are inextricably linked with those circumstances, and if they should alter or at some point not obtain, then the principles would be rendered null. If the moral law is universal, Kant insists, it must be deduced a priori. Sandel also detects a set of objections to Rawls from the Humean side: Rawls wants to defend the primacy of justice as the first virtue of social institutions, yet there seem to be some institutions, like the family, in which it is plausible that justice should not be the first virtue. As Hume himself maintains, justice has a remedial character that presupposes another set of virtues of at least a comparable and perhaps even superior order (like love or benevolence), such that the Humean empiricist is likely to hold that “justice is the first virtue of social institutions not absolutely, as truth is to theories, but only conditionally,” i.e. when the circumstances of justice obtain (31). In sum, Rawls’s attempt to replace Kantian metaphysics with a more modest Humean empiricism as the basis for his account of the original position leads to several inconsistencies and contradictions.
Beyond these inconsistencies, Sandel detects a far more fundamental problem with the original position. The key to his critique lies in Rawls’s notion of reflective equilibrium, a dialectical method of justification in which we sometimes adjust the conditions of the contractual circumstances of the original position in view of our considered convictions about what is just, and sometimes withdraw our considered convictions and conform them to the principles of justice once we find the premises of the original position reasonable. If the principles derived from the original position are accountable to our considered convictions about what is just, by what standard, Sandel asks, are we to find the premises of the original position plausible? For Sandel, it must be the case that “the independent yet provisional standard by which the reasonableness of our descriptive assumptions [that structure the original position] is assessed” is provided by “the nature of the moral subject as we understand it,” or put differently, how we understand ourselves as moral persons (48). In other words, if we take the double-barreled approach of reflective equilibrium seriously, the original position is not merely a handy heuristic device to be dispensed with once the principles of justice have been worked out, but a clue to who we are as moral subjects.
With this clue in mind, Sandel attempts to reconstruct Rawls’s conception of the moral subject and identifies two salient features: the first is plurality, or the antecedent individuation of distinct persons with separate systems of ends. By plurality, Sandel (and by extension Rawls) means simply that we are distinct individuals before we form relationships and cooperate with others for the sake of mutual benefits; plurality therefore describes “the terms of relation between the self and the other that must obtain for justice to be primary” (53). The second salient feature of Rawls’s moral subject, inferred from the parties’ “mutual disinterest” in the original position, is that the self is a subject of possession, i.e. a self that is prior to and distanced from ends it comes to possess by voluntary choice. Mutual disinterest, Sandel astutely notes, “is not an assumption about what motivates people [i.e. that they are solely interested in their own ends], but an assumption about the nature of subjects who possess motivations”—it simply “holds that all interests must be the interests of some subject” (54). Such a self may or may not have communitarian aims, it may or may not pursue exclusively private ends not shared with anyone else—in either case the choice is its own. That the self can make such a choice implies, however, that the moral subject Rawls has in mind exists independently of the interests it possesses; as prior to its ends, “its bounds [are] fixed once and for all such that they are impermeable, invulnerable to transformation by experience” (57). Sandel contrasts this Rawlsian conception of the “unencumbered” moral subject with an intersubjective conception of the “encumbered” self for whom the primary question is not, “What ends should I choose,” but rather, “Who am I?” This is because the encumbered self is “crowded by the claims and pressures of various possible purposes and ends,” all of which help constitute my identity, and in relation to which I must discern what is me from what is mine (57-9). Such a self is not prior to its ends but discovers itself amidst ends that already encroach upon it; its task of self-command is therefore not one of will, but of intellect.
In what amounts to the basic thesis of the book, Sandel contends that Rawls’s conception of the unencumbered moral subject cannot support his theory of justice. To demonstrate this, he appropriates Robert Nozick’s critique of the difference principle. For Nozick, the difference principle (i.e. the first part of the second principle of justice), which permits social and economic inequalities only insofar as they benefit the least well-off in society, uses persons as mere means and not also as ends in themselves. This is because the difference principle, on Rawls’s own interpretation of it, treats the arbitrary (from a moral point of view) distribution of individuals’ natural talents as a common asset, on account of which it only permits inequalities linked to the use of these talents insofar as such inequalities incentivize activity that benefits everyone, especially the least well-off. For Nozick, to treat individuals’ natural assets as common property quite blatantly instrumentalizes them—it uses them as mere means—and blurs the distinction between persons that Rawls is so keen to uphold. Yet Rawls, Sandel observes, can mobilize his conception of the moral subject to rebuff Nozick’s critique: if the self is prior to its ends, then it is likewise prior to its attributes, and the difference principle only makes use of these attributes as means to others’ welfare. While this defense successfully defeats Nozick’s objection, it nevertheless leads Rawls to posit a radically de-situated moral subject, one that closely resembles the Kantian self of transcendental idealism that justice as fairness takes such pains to avoid. Still, Sandel points to an “alternate defense available” to Rawls that “ties the notion of common assets to the possibility of a common subject of possession” (79). Here, Sandel reintroduces his intersubjective conception of the self with which he contrasted the Rawlsian moral subject earlier: if the difference principle is not to use persons as mere means to others’ ends, “it can only be possible under circumstances where the subject of possession is a ‘we’ rather than an ‘I,’ which circumstances imply in turn the existence of a community in the constitutive sense” (80). In short, the difference principle, so central to Rawls’s theory of justice, can only be justified with recourse to the kind of encumbered self, constitutively linked to a particular community, that Rawls himself rejects.
In a similar move, Sandel demonstrates why the original position can only adequately justify the principles of justice “at considerable cost to certain voluntarist and individualist assumptions” central to Rawls’s theory of the person (105). To demonstrate why, Sandel first observes that, from a Rawlsian perspective, an actual, real-life contract must presuppose an antecedent principle to justify its results: it is not simply that a contract to which both parties voluntarily consent is necessarily justified, for we can always ask the further question, “but is what they have consented to fair?” An actual contract, because it is a case of only imperfect procedural justice, presupposes a moral backdrop in view of which its requirements can be assessed. If this is true, then the obvious question arises as to how the original position, as a modified version of the social contract, does not, like ordinary contracts, similarly presuppose a moral backdrop. For Rawls, the original position, as an exercise in pure (as opposed to imperfect) procedural justice, is meant to produce principles of justice in view of which all subsequent contracts or promises can be assessed. He believes that it can provide the ultimate sanction to the principles of justice comparable to the role pure practical reason plays in Kantian ethics. Yet in order to provide this ultimate sanction, Sandel insists, Rawls implies that the principles to which the parties consent in the original position are just not because they are those reached as a result of a fair procedure, but because they are the right principles. In fact, Rawls describes the contractual situation in such a way that it ensures that the right principles are chosen, and the rightness of their choice, over and above the fact of their choice, is what justifies the principles of justice. As Sandel explains, “the secret to the original position—and the key to its justificatory force—lies not in what they do there but rather in what they apprehend there. What matters is not what they choose but what they see, not what they decide but what they discover” (132). If this is true, then “the voluntarist aspect of the enterprise is not as spacious as would appear” (127): procedure does not seem prior to principle as Rawls maintains, and the self presupposed by the original position is not one prior to and distanced from ends freely chosen, but one already situated in moral space, capable of reflection on pre-contractual principles of justice.
Apart from the fact that the unencumbered subject of possession presupposed by Rawls cannot support his theory of justice as fairness, Sandel finds Rawls’s conception of the person unpersuasive for several other reasons as well. In the penultimate chapter of the book, for example, he explains why such a self is capable of only a thin and ultimately trivial sort of reflection; if who we essentially are is “barren of constituent traits,” detached from our attributes, values, and ends, then there is no real self “for reflection to survey or apprehend” (160-61). In another, related critique, Sandel observes that heteronomous desires entirely determine the unencumbered self’s conception of the good, whatever it may be. “But if my conception of the good is simply the product of my immediate wants and desires,” he writes, “there is no reason to suppose that the critical standpoint it provides is any more worthy or valid than the desires it seeks to assess” (165). Provided that some conception of the good does not violate the principles of justice, any conception will do; indeed, as Rawls himself admits in A Theory of Justice, “that we have one conception of the good rather than another is not relevant from a moral standpoint” (537). Given this impoverished theory of the good, the unencumbered self reasons with respect to its private aims in a way that is instrumental, even utilitarian; it matches means to ends, but does not choose its ends on the basis of any considered evaluation of their respective merits.
Ultimately, these external critiques of the unencumbered self complement Sandel’s internal critiques to form a robust and systematic objection to justice as fairness that has exercised tremendous influence since its publication. Still, Sandel leaves several questions unanswered. Is Liberalism and the Limits of Justice a full-blown rejection of Rawlsian liberalism, or could Rawls incorporate Sandel’s intersubjective conception of the self into his theory without a complete overhaul of justice as fairness? Does Sandel endorse some modified form of political liberalism, or does he think the liberal philosophical project is doomed to internal contradictions and untenable presuppositions? Does Sandel constructively offer some form of common-good liberalism substantively different from, say, the utilitarian liberalism of J. S. Mill? If so, we need more to differentiate his brand of liberalism from the communitarian theories he distances himself from in the preface to the second edition.