When it was first published, Anderson reviewed John Rawls' 'Political Liberalism' in Dissent under the title 'Designing Consensus', and some version or other of that review is in his recent collection, Spectrum. Anderson's critique is a litany of disappointments. He seems to view Rawls' 'Theory of Justice' as, although unsatisfactory, a valuable contribution to normative political theory, and strikes the pose of a chastising and particularly contemptous teacher as he catalogues the ways in which Rawls has failed to live up to this early promise. So, as is inevitable, I'm going to struggle to avoid doing the same in attempting to analyse his errors.
There were, according to Anderson, four problems with 'Theory of Justice': the device of the original position was circular; the priority of liberty over material equality was unjustified; the difference principle is somewhat ambiguous; it has nothing to say about international relations. Although Anderson misses the mark somewhat with the third and fourth, since the point of Rawls' project at this point was to provide a set of principles to guide the construction of just regimes in states rather than define a just tax code or deal with issues of international justice, the first and second, it does have to be said, are staples of critiques of Rawls. According to Anderson, rather than attempt to learn from the criticism of his betters, Rawls petulantly decided that none of them were to be addressed in 'Political Liberalism', and that he would make his own choices about what he would do with his philosophical project. In this vein, Rawls structured 'Political Liberalism' around a retreat from the metaphysical into the political, a withdrawal which grounds itself in what he calls the 'public culture of a democracy', a strategy which Anderson passes this comment on:
What was a latent and subtle circularity in A Theory of Justice becomes a more gross and explicit one in Political Liberalism. For Rawls simultaneously appeals to the natural outlook of a democratic society to found his conception of the person, and to his conception of the person to found the basic structure of a democratic society... In a vicious circle, public arrangements are deduced from personal capacities defined as adapted to public arrangements.
Anderson is not actually totally alone in making something like this critique. The almost perverse misunderstandings of Rawls in Joseph Raz's 'The Morality of Freedom' spring most immediately to mind, but there are other culprits as well. The problem with it is that it quite fundamentally fails to grasp the thrust of Rawls' retreat into a much more restricted set of claims, a thrust which is both epistemological and ontological. This withdrawal is motivated by the success of critiques which focussed on the first of Anderson's four problems with 'Theory of Justice', the circularity of the device of the original position. It was felt that the assumptions which structured the original position were those of full-blown ethical liberals, and that this was objectionable, because Rawls made little attempt to vindicate those assumptions, meaning that those who weren't full-blown ethical liberals could legitimately object to them. Rawls' attempt to correct this, which I think he correctly sees as being the source of almost all the criticism of 'Theory of Justice' as a project - rather than of the set of principles generated by that project - is the retreat of 'Political Liberalism'.
The clue is in the title. The project of 'Political Liberalism' depends on an understanding of the political as a sphere with - at least as far as the Atlantic democracies are concerned, but perhaps elsewhere: Rawls' restriction implies no more than that, and certainly not an exclusion -distinctive epistemological requirements. This is liberalism applied to politics, and politics alone - more specifically, politics as construed in the idea of democracy. It is interested in the question of, as it itself puts it:
[how] is it possible that there may exist over time a stable and just society of free and equal citizens profoundly divided by reasonable though incompatible religious, philosophical and moral doctrines?
The centrality of disagreement to this question is key, I think, to Anderson's misunderstanding. He reads it as a historical coincidence, and pours scorn on Rawls' obsession with what is referred to as the 'fact of reasonable pluralism'. But the 'fact of reasonable pluralism' is central: it is what makes it a mistake to demand that those who do not share Rawls' liberal sympathies simply submit to the liberal assumptions of 'Theory of Justice'. They disagree, and in most cases they are reasonable to do so. Indeed, unless they disagreed, it would not be a problem to demand that they submitted, since they would do so voluntarily. Yet, if they did not disagree, the principles on which they should structure their society would be rather different, because the freedoms which Rawls continues to place at the centre of his theory would be unneccessary: those who do not disagree do not require the freedom to act differently from each, to deviate from the plan, which is, after all, in the total absence of disagreement, surely correct.
Rawls' references to the 'public culture of democracy' are references to this: unless people have a need to structure the rules which govern the institutions in which they live out their lives, the appeal of democracy becomes rather opaque. The fact of reasonable pluralism, by positing a situation in which genuine plurality exists, makes that need clear. What Anderson refers to as Rawls' conception of the person is grounded in this need, created by difference, since it is a conception of the person which will lead to reasonable disagreement. Unsurprisingly, that need, and the set of issues around it, then ought to structure the institutions under which they live. A democratic state will have to take into consideration why democracy matters when making decisions about how to live its collective life. The circle does not have to be vicious.
The ontological and epsitemological retreats that Rawls makes in the face of this understanding of what provides the moral force of democracy are precisely those which Anderson finds it so easy to misrepresent and then denigrate, specifically, an unwillingness to demand adherence to a comprehensive or metaphysical conception as a precondition of consent to the arrangements he suggests. Because the moral force of democracy depends on the existence of substantive difference, to make such a demand would run directly counter to its own moral motivation. Rawls has to abjure from the allegedly comprehensive doctrines of 'Theory of Justice', or risk undermining the motivation of his whole project, which of course explains the obsession with skirting round the issue of truth and the conditions of public reason.
Anderson, for whatever reasons, is unable to grasp this, despite Rawls' persistent worrying away at the problem of what might constitute conditions under which a state was neither doing too much nor too little to be said to be respecting people as free and equal in this sense throughout 'Political Liberalism'. But then, as we have already seen, Anderson regards liberal and social democracies as the site of oppression. How deep that hostility goes is not quite clear, although his apparently approving quotation of the admittedly rhetorically impressive, but essentially philosophically bankrupt, passage at the beginning of Alisdair MacIntyre's 'After Virtue', would suggest a basic failure to understand the distinction between politics and ethics. What Rawls understands, but Anderson seems not to, is that politics is a question of how we are to live together, whilst ethics is a question of how I should live. That the problem of politics refers to a we is decisive, since that we requires difference, and to avoid the collapse of the political into the ethical, that difference must be respected. We are brought back to the horseshoe: the question of whether or not one is prepared to pay that respect or not, after all, shapes it.