Chandran Kukathas Oxford University Press, Oxford, 2007, 292pp., £19.00, ISBN: 978-0199219209

In his new book, Chandran Kukathas makes two fundamental claims. The first is that justice is neither a necessary nor a sufficient condition for legitimate authority. The second is that in a pluralistic world, legitimate political authority will be minimal, in that its role will be restricted to the preservation only of the freedom of individuals to associate with whomsoever they please. The Liberal Archipelago is lively, provocative defence of these two claims and the liberal theory that Kukathas builds around them.

Kukathas offers an ambitious defence of his two claims, beginning with a theory of fundamental human nature and interests. What makes humans distinctive, he claims, is our recognition of the motive of principle – as opposed to interest or affection – as having normative primacy. We are creatures of conscience, even when we fail in fact to act as we think we should. It follows that ‘[i]f there are any basic human interests, that interest [sic] is at a minimum, an interest in living in accordance with the demands of conscience’ (p. 55). Respect for this basic human interest requires freedom of association and mutual toleration; the good society is one in which these are secure. The highest authority in the good society, then, will ensure that they are; but it will do no more, because to do any more may involve forcing individuals to lead lives that conflict with the demands of their consciences. In particular, the highest authority will not pursue the value of justice, pace most liberal theorists, for justice is no less controversial than any comprehensive moral conception. (This is where Rawls goes wrong, as Kukathas reads him.) So the highest authority must be minimal in the sense I described above.

Within the good society, meanwhile, there will be many different communities that may pursue justice, or indeed salvation, or any other value. So long as the freedom of individuals to associate with one another as they please is secure, these communities may distribute property, operate hierarchies and engage in illiberal (or other) practices as they please. In these matters, their practices are authoritative simply to the extent that individuals accept them as such; and individuals do that by failing to dissociate from them. Thus, whether or not an authority is just or good is irrelevant to its legitimacy: the highest political authority is legitimate to the extent that it protects freedom of association, whereas other authorities – and there may be many, overlapping authorities in society – are legitimate to the extent that they are accepted by those over whom they claim jurisdiction.

Kukathas defends this view against important objections from two directions in particular. On one side, he faces objections from liberals concerned that his view pays only lip service to the value of freedom of association and that it offers no basis for social unity. On the other side, he faces objections from advocates of the politics of recognition, concerned that he is insufficiently attentive to the claims of minority communities. (A large proportion of the book discusses multiculturalism, and Kymlicka's multiculturalist liberalism in particular.) On both counts, readers may find themselves unsatisfied by Kukathas's responses. Consider by way of example Kukathas's response to the concern about freedom of association. The concern is that failure to dissociate from a community ruled by an oppressive, abusive or unjust authority does not necessarily indicate acceptance of that authority. It is a familiar truth that the costs of dissociation may be too great for an individual to be genuinely able to leave a community, even if no one would in fact force her to stay. Liberals may therefore be inclined to strengthen the value of freedom of association by ensuring (through tax-funded provisions) that exit is not excessively costly. Kukathas's reply is simply that this is to favour the state's authority rather than the community's in a ‘stand-off that cannot be resolved’ (p. 147), for once we allow any liberal state intervention, there will be no stopping point short of taking children from their parents to give them to more worthy couples. But, he claims, we should in fact favour the community's authority, because ‘this may still be less dangerous than conferring greater powers upon the state on the assumption that it can only do good’ (ibid.). However, this last claim is weak. Since Kukathas is not sceptical about the possibility that we could know a state's actions to be just and the community's to be unjust, why not advocate just intervention while acknowledging the dangers of excessive state power? Moreover, the no-stopping-point argument seems invalid: a natural stopping point would be set by minimal conditions for voluntariness of association (such as a certain awareness of alternatives, education to some specified level and the like). More must surely be said about both these points.

There are other tensions and arguments in need of fuller development in The Liberal Archipelago (although for some of that development, Kukathas does direct his readers to other published works of his). Not enough is said about the distinction between legitimacy and justice; Kukathas offers, but then effectively abandons, a Kantian justification for freedom of conscience, which is never clearly related to the more prominent human interest justification; and the book seems to me to perch uncomfortably on the fence between pragmatic and principled arguments for the kind of liberalism that Kukathas favours. But many of its virtues outweigh these shortcomings. Kukathas has a wonderful range of examples drawn from both life and literature, a clear style and a nice sense of humour. He often captures the essence of his opponents’ positions with admirable pithiness and an eye for the heart of an argument. The early chapters in particular provide a textbook example of orderly liberal theory-building from assumptions to political principles. Readers of The Liberal Archipelago will find much to take issue with; but it is a credit to Kukathas's clarity and engaging style that they will have little trouble identifying precisely where they and he differ.