Rawls's Political Liberalism: Reconstructing Stability, Not Abandoning It — Epoche C1
The argument that had to be rebuilt Part III of John Rawls's A Theory of Justice (1971) contains an argument that his later work withdrew, and the withdrawal is the subject of this essay. The argument concerns stability , which in Rawls's technical use is not the absence of unrest but a specific property of a conception of justice: a conception is stable when a society governed by it tends to generate in its citizens the very sense of justice that sustains it, so that compliance does not depend on coercion or on luck. A conception that requires policing to survive has failed a test Rawls thought any acceptable theory must pass. Two further terms are needed throughout. A comprehensive doctrine is a general view about what is valuable in a human life and about what the world is like — a religion, a philosophical ethics such as utilitarianism, a secular humanism. A political conception of justice is a much narrower thing: a set of principles for the basic structure of society alone, worked up from ideas already latent in the public culture of a democracy, and presented without deriving it from any comprehensive doctrine. Rawls calls such a conception freestanding . The received account of Rawls's development says that Political Liberalism (1993) abandons the stability project, trading a deep moral consensus for a truce among groups that cannot defeat one another. That reading gets the direction of travel wrong. Rawls kept both the question and the standard of success; what he replaced was the argument that was supposed to meet the standard, and he replaced it because he had come to think it unsound. What follows sets out the failed argument, the reason it failed, and the exact shape of what took its place — together with three points at which the usual compressed summary of that replacement, this essay's earlier version included, states it inaccurately. What stability meant in A Theory of Justice The Part III argument has two stages, and the compressed account of it conflates them. The first stage is a moral psychology. Rawls argues that citizens growing up under institutions that are in fact just will come to acquire a sense of justice, through a sequence he describes as the morality of authority, then of association, then of principles: a child first responds to the love of those who care for it, then to the norms of the groups it joins, and finally to the principles themselves once it grasps how they support the arrangements it has come to value. This stage yields a disposition, not yet a reason for keeping it. The second stage supplies the reason, and Rawls calls it the argument from congruence . The question it answers is a familiar sceptical one: even granting that a just society produces citizens with a sense of justice, is acting on that sense good for the person who has it ? Might a rational citizen not conclude that her sense of justice is a constraint on her good rather than a part of it? Rawls's answer is that in a well-ordered society the right and the good coincide — that acting from the sense of justice belongs to a person's own good, because it expresses her nature as a free and equal rational being. The support for that last step is the Kantian interpretation of justice as fairness advanced earlier in the book, on which choosing the principles from the original position expresses our nature as autonomous agents. The correction to be made here concerns what congruence is between. It is not, as the argument is often summarised, a fit between citizens' several comprehensive doctrines and the principles of justice. It is a fit between each citizen's good , rationally understood, and her sense of justice — and the fit is secured by attributing to everyone a single comprehensive view of the person, the Kantian one. That is what makes the argument vulnerable, and identifying the vulnerability correctly is what makes Rawls's later move intelligible rather than arbitrary. Why the argument failed, on Rawls's own account Rawls states the diagnosis himself in the Introduction to Political Liberalism : the account of stability in Part III is unrealistic and must be recast, because it supposes that in the well-ordered society all citizens affirm justice as fairness as part of one comprehensive philosophical doctrine. Two considerations make that supposition untenable, and they do different work. The first is the fact of reasonable pluralism : a democratic society under free institutions will permanently contain a plurality of comprehensive doctrines that are incompatible with one another yet reasonable. The word 'reasonable' is doing real work and must not be read as a compliment. Rawls grounds it in the burdens of judgement , six sources of disagreement that afflict conscientious reasoners without any fault on anyone's part: the empirical evidence bearing on a case is conflicting and complex, and hard to assess; even where the relevant kinds of consideration are agreed, their relative weight may be disputed; our concepts are vague and generate hard cases, so judgement must be exercised within a range where reasonable people differ; how we assess evidence and weigh values is shaped by our total experience, and our experiences always differ; different kinds of normative consideration, of different force, often bear on both sides of a question, making an overall assessment difficult; any set of institutions can realise only some of the values that might be realised, so a selection and an ordering must be made. The list is what converts a sociological observation into a philosophical result. If disagreement arose only from error, ignorance or self-interest, one could hope for convergence under free discussion. The burdens of judgement show that the exercise of reason itself, under the conditions of freedom, produces divergence. Pluralism is therefore not a defect to be corrected but the expected outcome of the institutions the theory is meant to justify. The second consideration is the one