As noted in the previous post, in his 1973 paper on John Rawls’s A Theory of Justice, Hart put forward five queries. They can be summarised under the following headings.
- Liberty or liberties? In the first part of this series, we saw that Rawls’s model of multiple ‘basic liberties’ is problematic: either they’re commensurable (which suggests that they may be fungible and hence that one or more of them can be dispensed with) or they’re incommensurable (in which case it’s debatable whether they have any common property of ‘liberty-ness’).
- Restricting liberty for liberty’s sake Rawls argues that the only justification for limiting a liberty is an overall extension of liberties. What issues does this raise in terms of resolving potential conflicts between liberties?
- Restricting liberty for harm’s sake Rawls appears not to grant that liberties should sometimes be limited for harm-related as well as liberty-related reasons. Is this sustainable?
- Choosing (more) liberty Rawls argues that subjects in the original position would, in their own interest, tend to choose more rather than less extensive liberties. Given the potential adverse effects of liberties extended to the whole of society, is this valid?
- The priority of liberty Following on from the previous point: Rawls appears to believe that, all else being equal, subjects in the original position would choose a quantum of liberty over a quantum of material benefit. Is this an unstated presupposition on Rawls’ part? How is our perception of his model affected if this is granted?
Now for part 2.
Restricting liberty for liberty’s sake
Rawls argues (in Hart’s words) that “basic liberties may be limited only for the sake of liberty”. Restriction, or the unequal distribution, of basic liberties can only be justified if the adjustment yields “a greater equal liberty” or “the best total system of equal liberty” (the last two phrases are Rawls’s). In simple cases what this means may be fairly straightforward; an example is the imposition of rules of debate, which both curtails the liberty to speak at will and protects the liberty to speak at length, for a net expansion of effective freedom of speech. Similarly, public order laws and military conscription can (sometimes) be justified as present curtailments of citizens’ liberties to prevent greater future inroads on the same liberties. In more complex cases, when rival and – ex hypothesi – incommensurable liberties are at stake, the adjustment between competing liberties should be made from the standpoint of “the representative equal citizen”, on the basis of what “it would be rational for him [sic] to prefer”.
Hart finds both these formulations problematic, suggesting that criteria of value will necessarily be involved in both. Beginning with the simple case, he argues that what rules of debate help to secure “is not a greater or more extensive liberty, but a liberty to do something which is more valuable for any rational person than the activities forbidden by the rules”. I think this is an important point, which could be pressed further. If I interrupt a speaker at a public meeting by shouting obscenities, I’m not ‘speaking’ – or, by extension, exercising freedom of speech – in the same sense that the speaker is: there’s a qualitative difference between using one’s voice to express ideas and using it to prevent ideas being expressed. More difficult examples are available, and the argument could be taken further: there is also a qualitative difference between using speech to create the conditions for an informed dialogue and using it to shut dialogue down, for instance by giving amplified and officially-sanctioned expression to settled government policy. But the principle is the same: one does not protect freedom of speech by limiting freedom of speech, as Rawls suggested. The freedom to speak in such a way as to develop one’s ideas without interruption is protected by limiting the freedom to interrupt; the freedom to engage in dialogue is protected (perhaps) by limiting the freedom to fill the space available with a privileged monologue.
I would also go further than Hart in critiquing Rawls’s examples of military conscription and public order legislation, which Hart passes over with a nod to their plausibility. To be precise, he uses formulations like “may be justified” and “might be plausibly said”; Hart the lawyer concedes that a case can be made out, so Hart the philosopher doesn’t need to get involved. But it’s arguable that this concession itself rests on an equivocation. It certainly is the case that infringements on civil liberties can be justified on the grounds that subversive organisations, if they achieved their aims, would infringe those same civil liberties more extensively. It can be done – it’s syntactically possible and rhetorically quite acceptable – and it often has been: if you don’t like Special Branch tapping a few people’s phones, imagine if it was the Stasi tapping everybody’s phone! (This specific illustration may have been overtaken by developments.) But this is a question of rhetoric and not of measurement – not least because the future potential curtailment of liberties can, by definition, not be measured. We can see this point more clearly if we look at Rawls’s formulation of the circumstances in which conscription might be justifiable – “if it is demanded for the defense of liberty[sic] itself, including here not only the liberties of the citizens of the society in question, but also those of persons in other societies”. Rhetorically this formulation plainly works well: it advances the plausible proposition that the loss of liberties in conscription can in some circumstances be justified on the basis of the liberties secured or defended thereby. But if we ask it to do a more demanding job – not to state the proposition but to ground it, by making it possible to compare one set of liberties to the other in a measurable way – the difficulties are obvious: the set of liberties to be defended is not only a potential future attribute of a political situation but an attribute of the situation of a different group of people. Ironically, what seems to lurk behind Rawls’s formulations is an idea of fungible Liberty-stuff, whose production can be restricted now so as to produce it in greater volumes at a later date.
In short, the ‘sameness’ of the liberties being curtailed and expanded, suspended and secured in the ‘simple’ case is more apparent than real: if we curtail a liberty for the sake of the overall balance of liberties (which, for Rawls, is the only justification for doing so) then we are always in the position of adjusting rival and incommensurable liberties. This makes the device of adjusting competing liberties from the standpoint of “the representative equal citizen” even more important. Hart expresses bafflement on this point: if we take it that two citizens may reasonably differ in the importance they attach to different values, it must follow that two citizens may reasonably differ in the priority they give to two competing liberties – not only in particular cases (disagreements which could conceivably be resolved by an appeal to a larger scheme of liberties) but in principle. As Hart points out, there is nothing irrational about valuing the liberty to roam above the liberty to enjoy property free from trespass, or vice versa; and examples could be multiplied. Once again it would appear that political disagreements are being elevated to the level of principle and then forestalled (or precluded), albeit in this case without any specification of how this would be done.
With regard to the “representative equal citizen”, Rawls’s 1982 lectures are of little help; he acknowledges Hart’s scepticism on this point but does not address it directly. It is worth noting that Rawls’s reference to the rational preferences of the representative equal citizen appears to rest on a very specific definition of rationality; Rawls defines the “two moral powers” as “the capacity for a sense of right and justice (the capacity to honor fair terms of cooperation and thus to be reasonable), and the capacity for a conception of the good (and thus to be rational)” (emphasis added). The only substantial reference to the “representative equal citizen” appears after Rawls’s definition of a “fully adequate scheme” of liberties as one in which the basic liberties are adjusted “so as to allow the adequate development and the full and informed exercise of both moral powers”. Such a scheme, Rawls argues, “coheres with that of adjusting the scheme of liberties in accordance with the rational interests of the representative equal citizen”. Other than noting that the term ‘rational’ is, presumably, used here to denote the capacity for a conception of the good (rather than simply referring to means/end rationality), it is difficult to gain much enlightenment from this. Either the conflict between rival liberties can be resolved in principle (in which case let’s get on and see how we can do it), or it can’t (in which case we are leaving a lot of important questions to be settled politically – and it’s not clear what philosophical work the basic liberties are doing). Rawls appears to be putting forward a middle position, in which conflicts between liberties can be resolved at the level of principle but we don’t know how. If, as Rawls seems to be suggesting, the key factor in making the resolution philosophically possible is the nature of the adjudicator – the “representative equal citizen” with her Good-oriented rationality – then we don’t seem to be saying much more than that people would get on much better if they were nice.
However, the 1982 lectures do enable us to resolve this question differently – or, perhaps, to sidestep it altogether. As the reference to an independently-justified “fully adequate scheme” suggests, the judgment of the “representative equal citizen” may be a device we can dispense with; perhaps there is something about the basic liberties themselves which makes it possible to resolve conflicts between them (without thereby specifying how the conflicts would be resolved). This line of argument seems to be closer to Rawls’s thinking, at least by the time of the 1982 lectures. In that text he introduces the criterion of “significance”: “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”.
This is a particularly interesting formulation. Not only does it abandon the austere “only restrict liberty for liberty’s sake” position which Hart had found so problematic; it also suggests a metric of sorts for comparing different liberties, without the need to resort to ideas of fungible Liberty-stuff. The good, for societies, is the general, informed and effective exercise by individuals of the powers to co-operate reasonably and seek what they consider to be the good rationally. The basic liberties, having been defined as preconditions for this social good, should be maximised to the extent that doing so tends to promote it – and adjusted relative to one another to the extent that these adjustments promote it.
I think this is an important – nay, significant – addition, which goes a long way towards fleshing out the idea of a resolvable conflict between basic liberties. However, some of the credit for it should probably go to Hart, whose criticisms are not so much answered by it as conceded; arguably they were unanswerable within Rawls’s framework as it stood.