Empiricist Objections to A Theory of Justice


“Justice,” John Rawls writes, “is the first virtue of social institutions, as truth is of systems of thought,” in his A Theory of Justice (p. 2).  Truth and justice, he claims, are “the first virtues of human activities,” and thus “are uncompromising” (p. 4).  The principles of justice are determined from an Archimedean point, a point from which to assess the basic structure of society, known as the “original position” to Rawls. “The choice which rational men would make in this hypothetical situation of equal liberty, assuming for the present that this choice problem has a solution, determines the principles of justice” (p. 11).  By contrast, Michael Sandel, in his Liberalism and the Limits of Justice, illustrates that on an empiricist interpretation of the original position, “justice is the first virtue of social institutions not absolutely, as truth is to theories, but only conditionally, as physical courage is to a war zone” (Sandel, p. 31). This challenge to Rawls’ idea of the primacy of justice is fueled by Rawls’ admission of certain empirical truths that must be true if the kind of cooperative ventures necessary to justice are to be “possible and necessary” (Rawls, p. 109).

What then, are these circumstances, truths known to those in Rawls’ original position? While nothing is known by any individual in the original position about his conception of the good or his standing in society, the circumstances of justice are known (p. 111). The circumstances of justice can be either objective or subjective. The objective circumstances are those such as human vulnerability and relative scarcity of goods. The subjective circumstances, those with which Sandel’s objection is more interested, are centered on the idea that while persons have roughly similar interests, there exist competing plans of life that lead them to conflicting claims on the available natural and social resources (p. 110). “Unless these circumstances existed there would be no occasion for the virtue of justice,” Rawls states (p. 110).

            Given that the virtue of justice is admittedly dependent upon empirical preconditions, Sandel objects, the empiricist sees no clear way in which the priority of justice can be unconditionally affirmed (Sandel, p. 30). The virtue of justice can be no more universal than the empirical conditions upon which it is dependent; and argument, then, that the circumstances of justices prevail in all societies, Sandel, argues, is needed to establish that the primacy of the virtue of justice is universal (p. 30). Furthermore, Sandel argues, “if the virtue of justice is measured by the morally diminished conditions that are its prerequisite, then the absence of these conditions … must embody a rival virtue of at least commensurate priority, the one that is engaged in so far as justice is not engaged” (p. 32).  By “morally diminished conditions,” Sandel is referring to the fallen conditions which justice remedies (p. 32). As an analogy, he states that a society which most highly values peace and tranquility might come to value courage most highly during a time when war is necessary, as courage remedies them from destruction. Of course, the country would place even greater value on not facing the threat of destruction or the loss of life that comes along with the absence of war.

Discord, in particular, is the morally diminished condition remedied by justice: “[J]ustice can be primary only for those societies beset by sufficient discord to make the accommodation of conflicting interests and aims the overriding moral and political consideration” (p. 31).  The absence of discord, must, then, be at least as valuable a virtue as justice. All other things remaining equal, in fact, any rival virtue which remedies discord more effectively than justice, under this view, is preferable to justice. Sandel cites Hume, whom Rawls says he largely follows (Rawls, p. 109), as support of the suggestion that benevolence, for instance, can function as a nobler virtue than justice, rendering justice useless (Sandel, p. 32).

Sandel lists several examples of social institutions in which it is not clear that “the accommodation of conflicting interests and aims” is of overriding importance, or of greater remedy than other values: families, tribes, neighborhoods, cities, towns, universities, trade unions, cultural or religious communities being some examples “with more or less clearly defined common identities and shared purposes, precisely those attributes whose presence signifies the relative absence of the circumstances of justice.”

In summary, the empiricist argument proceeds as follows: Justice can be the primary virtue of social institutions only in those institutions in which the circumstances of justice obtain. However, in some institutions, we observe, these circumstances are not prevalent.  Also, because the value of justice derives from its use in accommodation of discord, it can be no more valued than the absence of discord (or any more effective remedy to discord). Therefore, just is not the primary virtue of social institutions.

A number of responses could be made in defense of Rawls’ position. Though one might object that Rawls’ work intended only to analyze the big institutions of a large-scale society (as opposed to small institutions like marriage), this is not the case. Rawls lists the monogamous family, for instance, as an example of an institution. 

A different might be to say that while perhaps the conflict of interest does not seem to exist within the institution itself, the circumstances of justice do obtain in the larger society in which that institution exists. A married family is not, generally speaking, an entire society, because a society is “a closed system” (Rawls, p. 7).  The objectives of the institution may differ from the aims of the society, where the circumstances of justice do hold. For instance, suppose a business unanimously desires to dump chemicals into a creek, while the rest of the neighborhood unanimously want it left unpolluted.  So while the circumstances of justice do not necessarily hold within the institution if we treat it as a closed system, it needs be remembered that that institution is a part of a larger society in which the circumstances of justice do obtain. Thus Rawls could say that justice needs to be applied to the institution’s rules for its interaction with the larger society, the grander institution or set of institutions of which it is a component. The Jones family, for instance, might have a common will to plunder the home of the Smith family, and enslave them: within the Jones family, there is no discord or disagreement about this. However, perhaps there will be some discord within their neighborhood when they actually do attempt to plunder and enslave the Smith household. Rawls can suggest that it’s obvious that at some level, families, tribes, neighborhoods, etc., do experience conflict, if not internal conflict but conflict encountered in their necessary interactions with the rest of society, and for that reason they do not truly escape the boundaries of the principles of justice.

The empiricist may rejoin with something along the lines of, “But suppose that there was a closed system at some point in history.” It would not be too hard to imagine that, say, two lovers became stranded on a desert island. And indeed this would be an institution in a society in which perhaps conflict of interest is not prevalent at any level in their dealings with others. Perhaps one might concede to such an objection, “Technically that is possible. However, these are extremely abstract examples that will have no significance to us (as we do not live in, nor encounter, such societies).” An empiricist might, then, attempt come up with an example that is not fully self-contained, but close enough to being self-contained and free of conflict, either internal or with the outside world, that the circumstances of justice are not prevalent, e.g., a convent or monastery.

Rawls may simply stick to his guns regarding institutions such as marriage, tribes, and religions, saying that they are indeed to be rejected if they do not abide by the principles of justice. He could point out all the unhappy points about religious wars, tribal living, the inequities brought about by the monogamous family, and the like, saying something along the lines of, “But these examples just further go to demonstrate that a just society is always better [or some adjective to the effect of ‘superior’ here] than these.” An empiricist may respond by asking what is meant by “better” (or whatever description/adjective is used) and then determine whether or not this adjective has some pre-determined conception of the good loaded into it, which could not have been known to those legislating from the original position.  If there is some implicit concept of the good, though, then a person will not be able to derive this statement from the original position, so the point would be moot. Besides, the empiricist might also say, “That’s not exactly what I mean. I’m arguing that the circumstances of justice do not obtain under these institutions.” (This could lead back to the sort of arguments described above.)  In either event, it seems that this line of response could lead back to two other, larger arguments, rather than providing any definite closure.

To the idea of a value greater than justice—the absence of discord, for instance—Rawls may simply make a claim that there for all the institutions that have been tried (be they religious, ethnic, etc.), none have been more successful or will be more successful in remedying discord than justice. This is not a trivial claim, but is perhaps highly subject to sociological review and support. Furthermore, it could be argued that even if a plan does remedy discord better than justice, it may do so at some unacceptable cost. (For instance, the Borg on Star Trek resolve discord by destroying individuality and assimilating people into a collective interest.)

In the end, the fact that the kinds of institutions listed by Sandel, such as marriage and religions, are not entirely contained from conflict calls into question whether or not they very meaningfully stand outside the circumstances of justice. Rawls, on the other hand, would be hard-pressed to prove that every self-contained society that has existed or will ever exist in the universe, though, is or was characterized by the circumstances of justice, although he may dismiss such examples as so abstract as to be practically out of this world (indeed, in another closed system).  To the objection that there is a virtue of at least commensurate value to justice, Rawls may likely do best to simply provide the best sociological support for the idea that justice is superior at remedying discord, or that any rival value of greater than or equal effectiveness in remedying discord unacceptably undermines another value.


- Ryan Renn