Jack Knight’s and James Johnson’s book is fascinating, interesting and compelling. It is not the kind of book on which I could write deep or far-reaching criticisms, so I fear that I will have to limit myself here to quibbling about what could perhaps be seen as details – and that is their criticism of Rawlsian-style normative political theory.
Knight and Johnson understand Rawlsian ‘ideal theory’ as being characterized by the fact that “nearly everyone complies with, and abides by, the principles of justice” (quoting Rawls). They criticize Rawlsian ideal theory since “participants in these normative debates commonly assume that their relevant audience has already committed to the nonviolent resolution of their disagreements and differences” (Knight and Johnson p. 275). They also criticize Rawlsian ideal theory for its lack of analysis of institutions, for assuming that “matters of implementation are of secondary concern”, and for its belief that “the concern for institutions is derivative or parasitic on the specifications of ideals” (p. 15)
I have a lot of sympathy and respect for their project, and some sympathy for their criticism of (Rawlsian) ideal theory (witness my discussion of priorities in ideal and nonideal theory here at Crooked Timber a while ago). However, I am worried that Knight and Johnson are throwing away the baby with the bathwater. Is it not possible to suggest that Rawlsian-style normative theory is valuable if we understand its meta-theoretical status a little differently? Rather than understanding ideal theory as sketching a utopian world where all comply with the normative ideals and which is taken to be the one and only truth the theorist believes in and wants all of us to believe in, we could understand such theories rather as examining the moral values and normative principles which would be feasible if all would endorse them (hence the full compliance clause), and which could serve as an ideal which we can use in our democratic deliberations on which normative principles our institutions should embody.
Much actual democratic debate proceeds without being aware of the normative claims made, and without seeing the values embodied in particular institutional proposals (there are plenty of examples around – such as the turn to commodifying higher education in Europe right now, which is wrongly seen by many as a ‘technocratic’ and value-free proposal but in fact has deep normative consequences). Ideal theories such as those in mainstream normative political philosophy are indeed (as Knight and Johnson rightly point out) most of the time (but not always!) limited to discussions about these values and normative principles, but the depth of those discussions make clear that these are complex notions – which real political debate often does not recognize. Using a metaphor, ideal theory is not a proposal for a law, but rather fuel and inspiration for real-life political debate. Different types of vehicles need different types of fuel – but some for some vehicles this will be the best type of fuel.
So my suggestion would be that one could read ideal theory more charitably, by not assuming that it tries to give the one and only right answer to questions about how we should organize society, but rather by focusing on three other functions it has. First, studying ideal normative political theory has the important role of increasing the normative skills and capacities of participants in real democratic debates. Second, ideal theory analyses the complexities of the values which economic and social institutions and policies entail in great detail. Thirdly, ideal theory sketches institutional proposals (such as Rawls’s property-owning democracy, or Philippe Van Parijs’s unconditional basic income) which participants in real-democratic debates could propose in political debates as an experiment which they want to try out (which in the case of basic income was done by the village Otjivero in Namibia).
An additional defense of ideal theory would be to regard it as one task that needs to be performed in an intellectual division of labor. If we want our theoretical work to have sufficient depth, some of us will indeed spend 15 years working on an account of the notion of justice, leaving the institutional chapter of that project to others (is it problematic that Rawls has merely hinted that property-owning democracy would be an institutional design which we should consider, and which he takes to be best in line with his principles of justice, if other political scholars such as Martin O’Neill and Thad Williamson are the ones who develop that institutional account in more detail?).
All I’ve said in no way diminishes the claim that an institutional turn in normative political philosophy is much needed, and that normative political theorists should take pragmatism more seriously (although I expect that not all will be willing to go this consequentionalist and nonfoundationalist route). My claim is more limited, namely that even those who endorse a pragmatist project like the one proposed in The Priority of Democracy need not have to reject mainstream normative political theory – as long as the status of those theories is seen in a different light than described by Knight and Johnson.