In recent years a growing scepticism about multiculturalism both in theory and in practice has become evident. Such scepticism has come not only from those that one would expect to be ill-disposed to the recognition of the collective rights of minorities, but also from those – such as liberal egalitarians – who one might have believed would find multiculturalism congenial. The scepticism has addressed the practical political consequences of multiculturalist policies, and the theoretical warrant for such policies. Brian Barry's somewhat intemperate and irritable counterblast, Culture and Equality (2001), sets a certain kind of benchmark for principled liberal doubts about multiculturalism. However there have been other expressions of philosophical dissatisfaction such as Susan Moller Okin's feminist worries about the fate of women within multiculturally protected group, or Jeremy Waldron's affirmation, against an insistence upon the importance of difference, of the truth of liberal cosmopolitanism.

Such diverse scepticism has developed apace with a proliferation of multiculturalist writing that also amounts to an intellectual industry. Additionally, this abundance of work has been marked by a significant diversity of outlooks. Moreover this diversity would not always appear to derive from any obvious bedrock commonality of animating beliefs. For instance, it is very hard to see in what sense both Will Kymlicka and Chandran Kukathas are kindred spirits, either as liberals or as multiculturalists. One further feature of changes in multiculturalist theory as it has developed is worth noting. It has in recent years moved from general statements of its normative justification to a concern to address particular cases and to make finer distinctions in its practical elaboration. Thus, for instance, there has been a commendable attempt to distinguish carefully between the kinds of groups that might merit the protection of multicultural protection. In consequence the defenders of multiculturalism have offered more nuanced, more politically practicable, and more context sensitive defences than the early justifications suggested.

Against this interesting background further informed contributions to the debate about multiculturalism would normally be most welcome. However Paul Gilbert's book is a disappointment. The writing is clear and accessible, the philosophical analysis is at times acute, the examples chosen are occasionally provocative, and the overall commitments are perspicuous. However it has a very old-fashioned feel to it. Although the opening chapter name checks all the right writers and the usual suspects – Will Kymlicka, Charles Taylor, Yael Tamir, Bhikhu Parekh, Iris Marion Young, James Tully, and Nancy Frasier – what follows could just have easily written without the benefit of consulting their work. And when Gilbert does come to consider the particular views of individual writers he is ungenerous in his treatment of them and gets those views quite badly wrong.

Of that more anon, but in the first instance it might help broadly and crudely to summarise his views. This is as follows: the politics of multiculturalism appeals to the fact that cultural identities matter (and matter greatly) to people. However this is fundamentally mistaken. Cultural identity is neither of the same significance across all instances, nor in any of them does it have a deep-lying character. Rather a sense of a shared cultural identity is a response to circumstances, most often it would seem on his account to the provocations of others on the outside of the group in question. So a politics of identity gets it badly wrong – in both intellectual and resultant normative terms – about identity. It is not that a cultural identity can properly motivate a certain kind of distinctive political demand; rather it is the case that such an identity is itself the political product of certain kinds of circumstance. To reinforce his negative and deflationist view of cultural identity Gilbert sees a historical parallel in the earlier intellectually second-rate defences of nationality and nationalism.

So much by way of a statement of the overall thesis which Gilbert defends and amplifies in various ways. Some of what follows its initial statement can be dispensed with as offering little or nothing to the central claim; or doing so only if one is prepared to accept further, and contentious, theoretical assertions. For instance, Gilbert offers a psychoanalytic account of individual and collective identity that will appeal only to those who are already favourably disposed to Freudianism. Thus, having summarised Habermas's hermeneutic characterisation of psychoanalysis, Gilbert continues: ‘Yet it is, arguably, only if psychoanalysis functions in this… interpretative way that it can motivate a genuine politics’ (132). The argument (of the ‘arguably’) is presumably with others in the psychoanalytic camp; those outside it would not even begin to entertain the claim that psychoanalysis can motivate a genuine politics.

There is a chapter devoted to the typology of identities which is ingenious but utterly devoid of factual support or indeed compelling examples. There is a chapter on the ‘art ‘of identity which turns on the idea that there is a shared ‘Irishness’, appreciatively shared only by those within the group. That idea is summarily and rightly ‘destabilised’ (Gilbert's term) if not outright knocked out by the observation in the chapter's final two pages that there is no such thing as a single, homogeneous Irish cultural identity.

Finally, there is a chapter on the embodiment of cultural difference which starts from the example of a French woman wearing the burqa, and pursues questions of how one might or might not wear one's identity on one's skin. There are interesting things said about identity and embodiment. But this chapter is worrying for its use of the particular example. The front book cover shows a couple walking near the Eiffel Tower, the woman, evidently Muslim, wearing a veil. The same photograph is repeated on the back cover and the text immediately below claims that ‘cultural identities are increasingly invoked in support of political claims, and these claims commonly lead to acrimony and violence.’ It would be hard to resist the thought that the claims of Muslim identy lead to violence. The case of a Muslim identity is invoked at a number of points during the book. These usages strike me as often oversimplifying and naïve. In one place (121) Gilbert says that the attempt to impose uniformity on the variety of possible modes of one and the same religion ‘constitutes contemporary Islamism’. What exactly is ‘Islamism’ and is it not the sin of Gilbert's own work that it attempts to impose uniformity on the ways in which Muslims have sought to understand, practice, defend and assert their identities? That sin is of course compounded by the association of a Muslim identity or ‘Islamism’ (whatever that is) with violence.

The principal theoretical failing of Gilbert's book is that it attacks a straw man. It criticises an account of the normative justification of multiculturalist politics that all those who seek to provide such a justification are often at considerable pains to distance themselves from. Among the mythical assumptions of multiculturalist philosophy that it would relatively easy to find extensive textual rebuttals of, but which Gilbert apparently attributes to this philosophy, are that it is concerned with cultures narrowly construed, that it views all cultures, understood simply as cultures, as worthy of protection, and that the principal reason to be a multiculturalist is that culture provides its members with a deep identity that matters to its members.

Let me take a little time to acquit Will Kymlicka, the arch and canonical liberal multiculturalist, of at least some of the charges made against him by Gilbert. On page 176 Kymlicka is accused of conflating ‘culture’ with ‘community’. But Kymlicka is clear throughout his work that the ‘culture’ of multiculturalism is a term of art with which he, like others, is unhappy – not least for its connotations. On page 18 of his Multicultural Citizenship he defines a culture as ‘an intergenerational community, more or less institutionally complete, occupying a given territory or homeland, sharing a distinct language and history'. This is not an implicit conflation of distinct ideas but a cautious, careful and very open definition of his topic. On page 67 Gilbert asserts that for Kymlicka if a culture, as a ‘context of choice’, is ‘unsettled by sudden exposure to other norms or by proscription of their own then the result is supposedly confusion or anomie’. Where does Kymlicka say anything of the kind? On pages 167–8 of his Liberalism, Community, and Culture he is in fact explicit that the view of ‘culture’ to which he adheres is not one in which ‘changes in the norms [and] values’ of one's community … would amount to loss of one's culture.' He immediately continues that he uses culture in the sense in which ‘the cultural community continues to exist even when its members are free to modify the character of its culture, should they find its traditional ways of life no longer worth while.’ On pages 174–5 Gilbert argues, seemingly against Kymlicka, that if the value of culture lies in its protection of autonomy, then this can only be true of some and not all cultures. Yet in Chapter 9 of Liberalism, Community, and Culture and in Chapters 3 and 8 of the later Multicultural Citizenship Kymlicka is careful to spell out at length the limits of liberal toleration of group rights, and to do so precisely in terms of what may not be done to individual members by some cultures and in some ways.

Matthew Festeinstein's Negotiating Diversity is an excellent recent critical survey of multicultural writing, and a good place to go for a sense of the richness of the debates about its subject. Gilbert's book simply misses its target.