Idealising the subject of justice

There are good kinds of ideal theory and bad kinds of ideal theory. In her 2009 article Laura Valentini argues Rawls employs the former in his early work A Theory of Justice and the latter in his subsequent work The Law of Peoples. The problem in The Law of Peoples is that Rawls “idealises the subject of justice” by assuming all states in the international community (the subject of justice) will be liberal democracies. This assumption feeds prescriptions that simply will not work – and will, if anything, exacerbate injustice – for non-liberal or impoverished regimes. Yet Rawls also employs ideal theory in A Theory of Justice to model individuals in the original position behind a veil of ignorance. This, however, is considered an acceptable idealisation because it does not idealise the subject of justice. The subject of justice is the basic structure being chosen not the original position from whence the individuals are choosing. I look to defend this account of acceptable and unacceptable idealisation from an objection that needs to be addressed to fill out the broader analytical picture. Individuals in the original position will still be choosing a basic structure that is presumably not the actual basic structure. The problem here is that political liberals assume the basic structure being chosen is the subject of justice (pace Cohen, 1997). So while the original position construction is not in itself bad idealisation, it would appear it is still inevitable that there will be such idealisation at some point in the process. We will inevitably be idealizing the subject of justice when we prescribe an ideal basic structure. Does this imply all considerations of constitutional policymaking from the perspective of (at least) political liberalism is doomed to failure? My strategy for avoiding this conclusion is to, first, outline a distinction between a realistic account of the subject of justice and a non-realistic one. Bad idealisation is idealisation of the subject of justice that uses unrealistic ideals; good idealisation is idealisation of the subject of justice that uses realistic ideals. (I should flag though that I am not yet sure how this distinction relates to political realism – if it does at all.) By a realistic ideal I mean one that is empirically possible i.e. not metaphysically impossible (see Kripke, 1971). The second strategy that flows from this is to be open to stipulating multiple ideals. That is to say, open to accepting different basic structures in different circumstances. I have explicitly defended the jargon-heavy concept of empirical possibility for political philosophy at previous meetings of the ECPR Methods Group – and I will do so again here. What I am trying to get at is just how many different problems in the literature it appears to solve.