Ideal Theory and the Separateness of Persons

John Rawls famously sees ideal theory as a necessary precursor for non-ideal theory. Amartya Sen, on the contrary, argues that we simply have no use for knowledge of the ideal in non-ideal theory, since it will do nothing to inform political decisions between feasible options. Such knowledge is neither necessary nor sufficient, Sen argues, for comparing the justice of different options. In response to this charge, defenders of ideal theory have argued for the importance of having an ideal to point us in the direction of large benefits in the long-term (Simmons 2010). This response has in turn been challenged on the grounds that an understanding of how decisions will take society closer towards a distant ideal depends on a causal theory that neither exists nor is forthcoming (Wiens 2012). In this paper, I make no attempt to defend the ideal of perfect justice as a long-term target. I instead propose a different way of challenging Sen’s view that an ideal of a perfectly just society is neither necessary nor sufficient for comparing different arrangements. This is the view that ideal theory is an essential part in a model of justice (as fairness) meant to constrain the pursuit of a better society. It restricts what institutions can demand of each individual for the sake of societal improvements. Ideal theory performs this function by stipulating how much each individual contributes under ideal conditions of everyone’s strict compliance with the principles of justice. When we then turn to non-ideal theory, we can demand no more of each individual than what she or he would contribute under full compliance in a perfectly just society. Insofar as we accept this non-consequentialist constraint on the pursuit of the good, ideal theory is necessary for comparing feasible alternatives. One arrangement, x, might be more desirable than another, y, but y might nonetheless be more just insofar as x, and not y, would demand more of some individuals than they would contribute under ideal conditions of full compliance. Ideal theory thus formulates a constraint meant to respect people as separate persons. At least from a non-consequentialist perspective, this is a promising way of showing how knowledge of the ideal of justice is necessary for comparing the justice of different feasible arrangements.