First, some links to the individual posts in this series.
- Liberty or liberties?
- Restricting liberty for liberty’s sake
- Restricting liberty for harm’s sake
- Choosing (more) liberty
- The priority of liberty (same post as previous)
These five points can be boiled down to three key questions: the quantification of liberty; the joint possibility of individual liberties, posing the need for protective as well as permissive rights; and the human preference for liberty over (for example) material wealth or peace and quiet. On the first of these I’d say that Rawls carries the day, at least in the sense that Hart’s challenge prompted some valuable elaboration and clarification of his model. On the other two – which are inter-related, at least in Rawls’s presentation of them – I’m less sure. Rawls’s argument is airtight – and he appears to escape the charge of designing a model world for model citizens – but I’m not entirely convinced; I think Hart’s scepticism may be the X that marks the spot of quite a deep equivocation.
Quantifying liberty (posts 1 and 2)
The quantification of liberty appears to pose problems for Rawls in two respects: allocation and comparison. Suppose that the basic liberties can each be considered as contributing a quantum of fungible ‘Liberty-stuff’ to an overall total. In that case, there is no reason to take the actual list of basic liberties as definite, and it may be that a model which maximised the amount of Liberty-stuff allocated to all citizens would take us in directions that Rawls wouldn’t want to go (e.g. away from the institution of private property). If, on the other hand, the basic liberties are seen as individually and discretely valuable – because what they provide is not fungible Liberty-stuff but different and distinguishable types of freedom – then there seems to be no basis on which to strike a balance between them. This second possibility becomes more pressing if we consider Rawls’s dictum that a liberty should only be restricted for the sake of another liberty: if liberties are incommensurable, how can this be achieved except by random selection?
Rawls addressed both these points in the 1982 lectures by introducing the notion of ‘significance’, and in the process making it clearer that liberties function in his model as a means to an end. He wrote: “a liberty is more or less significant depending on whether it is more or less essentially involved in, or is a more or less necessary institutional means to protect, the full and informed and effective exercise of the moral powers” – these in turn being the powers to co-operate reasonably and to seek the good rationally. The basic liberties are liberties which are in fact conducive to reasonable co-operation and rational deliberation; the issue of comparison is resolved by considering the conditions created by the exercise of a particular liberty. The question of allocation (and fungibility) is not banished as easily; however, what Rawls can offer is a strong presumption that the basic liberties he lists do have at least some ‘significance’ in his sense, and the criteria with which an alternative list would need to be justified.
Compossibility (posts 3 and 4)
The two questions of whether restrictions on liberty could be justified in order to prevent harm, and of whether the choice of greater liberty for all would be rational, both turn on the Kantian question of the conditions for the joint possibility (or compossibility) of individual wills freely exercised. We know from experience that societies can function on the basis of a shifting balance of freedom and coercion; the question is whether there is a coherent and non-arbitrary solution to the problem, a framework of basic rights or liberties which will tend to produce social harmony out of the free independent actions of individuals. Hart’s comments suggest that, in his view, Rawls has assumed that his model has this virtue but has not demonstrated it.
Rawls’s answer to this point is, once again, to invoke the ‘significance’ of the basic liberties and their exercise. The point, in other words, is not to maximise liberty or liberties, but to create those conditions which are best created by the exercise of the basic liberties. It follows that a balance between the basic liberties, and hence the limitation of particular (less significant) liberties, is an integral part of the model. The basic liberties are both self-limiting and mutually limiting: to the extent that a liberty furthers “the full and informed and effective exercise of the moral powers”, to that extent it will tend to be protected over and against liberties whose exercise is less significant.
I find this argument convincing but unpersuasive; to put it another way, it seems to answer the question within its own framework but at the cost of making that framework less attractive – and, perhaps, distancing it from the world in which the question was asked. It may be significant that the question of harm is one on which (at least according to Hart’s account) Rawls is all over the place: starting from the presumption that a liberty should only be curtailed for the sake of another liberty, we can cover the idea of infringing liberty to protect from harm by invoking the liberty-reducing effects of harm, an association between citizenship and the exercise of liberties, and ideas of a duty of care to animals and the natural world, but it ends up looking like a bit of a hack. (The hackwork is mostly mine, but the gap it fills seems to have been left by Rawls.)
The preference for liberty (post 4)
Lastly, Hart poses two questions which can both be taken as calling into question Rawls’s assumption of a preference for liberty (as distinct from, say, material wealth or a quiet life). Hart suggests that Rawls has tacitly built his model society on a liberal model of active civic virtue, thereby resolving all such questions in favour of the – undeniably virtuous but potentially strenuous and unrewarding – pursuit of reasonable co-operation and rational deliberation.
The charge is made lightly but it is potentially devastating, reducing Rawls’s model to a utopian vision of how people would live if only they were good. Rawls rejects it, quietly but firmly; while he concedes that his presuppositions are liberal, they find expression not in the characters of the subjects populating his model but in their nature. Specifically, it is in their nature to work together with other individuals (and hence to value reciprocity and fairness) and to value some states of affairs more highly than others (and hence to value morality and logic). The only qualities Rawls reads into the subjects in his model are the capacities to co-operate reasonably, where ‘reasonableness’ includes a sense of justice, and to deliberate rationally, where ‘rationality’ includes an idea of the good. Everything else in the model follows – which is to say, everything in the model follows from some facts about people as they are.
Again, I find myself convinced but unpersuaded. One reason why I’ve harped on Rawls’s particular definitions of the ‘reasonable’ and ‘rational’ is that they’re easily overlooked, but make Rawls’s model much easier to understand if they’re taken into account – cf. Hart’s puzzlement over how the model would work on assumptions of ‘self-interested rationality’. But another reason is more critical. Certainly there are such things as co-operative reasonableness and rationality oriented towards an idea of the good, and practices informed by them; an account of society based exclusively on individual self-interest would be not so much impoverished as false. But bracketing out means/end rationality and reasonableness in the pursuit of one’s own desires seems like an equal and opposite distortion. The question is not whether Rawls’s moral powers are a human reality but whether they could ever do the work he wants them to – and saying that they could do if their exercise were given priority over less moral pursuits is begging the question.
I sense that Hart saw a deep equivocation here, between a model which could exist (in the sense that it rests on valid assumptions about human nature) and one which could exist (in the sense that the model itself represents an imaginable society). It may be that Rawls only saw himself as developing the first of these; however, to the extent that such an abstract standard can be a driver for reforms to the society we have, it must surely be possible to envisage reforms which would represent steps towards it, even if they were fated never to reach it. And, if Rawls’s model is supposed to represent something approachable (even if not attainable), we’re back to the original question: why are his subjects so nice? The answer seems obvious – it’s because they’ve chosen to prioritise conditions favouring the exercise of the two moral powers – but this only defers the question: in the light of most of human history to date, why have they chosen to do that? If we’re going to have a society founded on a complex balance of basic liberties – and it sounds like a good idea – where are we going to get the people who want one? And, if we haven’t got those people, maybe we should be working on something for the people we have got – a theory of justice as between flawed, lazy, selfish and intermittently deceitful people, for example.