Rawls is famous for two things: his attempt to ground morality in rationality and his conception of justice as fairness. His work has been resounding on both fronts, the first constituting the justificatory framework for the second. Yet from the beginning, the outcome has been more doctrinaire than the method should have allowed with design details promising objectivity. This article goes to that beginning, or to a reasonable proxy for it, in the "Outline of a Decision Procedure for Ethics," with the aim of exposing and examining the discrepancy where it originates. The goal is not to prey on the earliest version of an initiative later undergoing revision but to identify and investigate the inception of a systematic bias that is retained rather than revised in subsequent iterations. The primary finding is that justice as fairness is too dogmatic an outcome for the decision procedure proposed. The corresponding principles (of justice as fairness) are consistent with the proposed procedure only relative to a specific system already presumed valid. A secondary finding, supported by the same considerations as the first, is that the decision procedure itself, regardless of whether it works well with the particular principles in question, may not be as useful as it appears, turning out to justify only, or mostly, what is widely accepted anyway.
We all like a good story. And if it is told just right, it does not even have to be true. The crystal ball, the deck of cards, the coffee grounds--they are all extraneous to the fortune told. The divination itself does not come from them. The oracle may, as it happens, be right on target, but it is the show that promotes persuasion. Without the razzle-dazzle, what the diviner has is just another story, true or false, right or wrong, good or bad. The purpose of this article is to determine where Rawls gets his story, how credible it is, and whether it works.
This is not an assault on Rawls. He tells a good story. But it does recall the allegorical introduction above. Essay after essay, book after book, everything comes out just right, that is, his own moral outlook turns out to be validated by an ethical justification model of his own design. There would be nothing strange in this if it were not for Rawls's assurance that the model guarantees objectivity. As it is, given the claim to objectivity, what is suspicious is that an objective approach should be so biased toward a particular outcome, or if one prefers, toward a particular kind of outcome.
A tempting response is that a good model would naturally be biased toward the right answer. Another is that ignoring or weeding out wrong answers is not a violation of objectivity. There is indeed nothing wrong with giving the same answer every time if it is the right answer. But that is precisely what is in question.
To restate the problem, this time in more formal terms than in the opening allegory, Rawls operates with a decision procedure for ethics that keeps corroborating the same moral outlook, a liberal one, whereas the objectivity he claims for the procedure might reasonably have been expected to be consistent with a wider range of moral, social, or political perspectives, or perhaps with a single position equidistant to polar extremes. (1) The aim of this article is to ascertain whether this is a real problem and whether it may be overcome to salvage the analytic apparatus Rawls recommends for ethical justification.
Departure from objectivity is, of course, always a real problem. And this is the conclusion of this article as well. But its focus is more specific. It is concerned with two related questions. The first is about internal consistency: (1) Does the ethical justification model developed by Rawls justify the moral principles advocated by Rawls? The second is about universal relevance: (2) Can the ethical justification model developed by Rawls justify any moral principles?
Covering both questions, the overarching reason for suspicion, to be confirmed or rejected here, is that certain methodological constraints, conceived as anchors to objectivity (freedom from bias) and safeguards against relativism, may not just undermine the justification of the particular principles presented by Rawls, thus making him dogmatic in endorsing them, but also preclude the justification of any moral principles whatsoever, or at least any useful or substantive ones (the kind directly relevant to the adjudication of moral problems). In the first case, the orientation toward objectivity and the precautions against relativism contradict the principles adopted. In the second case, they engender sterility: the justification of generalizations that are already widely accepted.
This is not, even incidentally, a rejection of the principles espoused by Rawls. More accurately, it is not a critique of those principles independently of their relation to the justification model employed to validate them. The principles themselves are neither surpassed by alternatives nor in urgent need of justification. I myself would gladly defend them on that point, that is, from a standalone perspective separated from the question of justification by Rawls, but I will not do so, not here anyway, as this is an altogether different matter. I will, however, question their justification through the decision procedure developed and proposed by Rawls.
Let me acknowledge at once that both the model and the output have evolved over time. I do not deny this. What I deny is that the evolution has made my concerns irrelevant. Granted, ethical justification has always been a work in progress for Rawls. But my impression is that the problems identified in this article were never met at any point in that process because they were never even on the agenda. This impression is not a hunch but an observation, or to be precise, a judgment formulated on the basis of a series of observations. While negative claims are notoriously difficult to prove, section 5 of the present article is a representative effort. The aim there is not to establish beyond a doubt that Rawls later says nothing that could possibly be interpreted as an improvement upon any of the difficulties in his earlier works but to demonstrate a relevant methodological continuity between the earlier and later works that corroborates the timeless relevance of my concerns here.
To be sure, the model's evolution can be traced through various paradigms in succession: the considered moral judgments of competent moral judges; the strategic outlook of contractors working out a reflective equilibrium from the original position; the various modes of public reason (or public justification) converging toward an overlapping consensus (grounded in a reasonable moral psychology). The corresponding normative output undergoes development as well, ranging from the relatively detailed set of principles presented early on to the more elegant and now famous twin principles explicating Rawls's conception of justice in social institutions.
Regular recalibration on the part of Rawls requires meticulous attention on the part of the critic. The broader the scope of consideration, the better. But that is a project for a book, at least one, probably more. A feasible alternative to comprehensive coverage is to work with A Theory of Justice (1971/1999, hereafter "Theory"). (2) That might be the obvious choice, but that work has undergone enough critical scrutiny to save generations of scholars the trouble of looking for something new in it. Besides, it remains in the middle, drawing on earlier material and anticipating later development, but representing neither perspective adequately. Fresher insight can be had by going back to the beginning, which cannot be too far from Rawls's "Outline of a Decision Procedure for Ethics" (1951, hereafter "Outline"). (3)
Starting from the beginning should prove enlightening (section 2) so long as observations and inferences (sections 3-4) are filtered through an awareness of subsequent departures and turning points (section 5). Proceeding on that basis, this article examines, first whether Rawls's decision procedure for ethics justifies the moral principles he purports to be justified by that model, second whether that procedure facilitates the adjudication of actual moral problems (perennial or emerging) or remains an exercise in formalism with little or no practical import or serviceable implications.
The conclusion regarding the first question is that the validation connection between the model and the output is not as strong, or even as clear, as Rawls would have us believe, and that Rawls is, in this respect, inevitably dogmatic in his support of the principles he advocates. The conclusion regarding the second question is that design elements intended to ensure objectivity and thus to avoid relativism weaken the credibility of the model as a decision procedure relevant to moral reality, or to be more specific, as a decision procedure capable of justifying the kind of moral principles that would be useful in the adjudication of moral problems.
This, in effect, is to leave Rawls with a vacuous decision procedure and dogmatic principles. And that, in turn, may seem to underestimate his contributions. But it is not as harsh as all that. While there is indeed some tension between this criticism and Rawls's reputation, there is also room enough for both. Nothing is ever gained from reacting to mediocrity. And no harm can come from exposing a problem in excellence. Criticizing Rawls detracts no more from his contributions than centuries of criticism have from Plato's.
Here is the immediate problem motivating Rawls to seek a solution later becoming the focus of his life's work:
Does there exist a reasonable decision procedure which is sufficiently strong, at least in some cases, to determine the manner in which competing interests should be adjudicated, and, in...