My lecture is about a surprisingly new subject.footnote We have become so used to terms like ‘collective identity’, ‘identity groups, ‘identity politics’, or, for that matter ‘ethnicity’, that it is hard to remember how recently they have surfaced as part of the current vocabulary, or jargon, of political discourse. For instance, if you look at the International Encyclopedia of the Social Sciences, which was published in 1968—that is to say written in the middle 1960s—you will find no entry under identity except one about psychosocial identity, by Erik Erikson, who was concerned chiefly with such things as the so-called ‘identity crisis’ of adolescents who are trying to discover what they are, and a general piece on voters’ identification. And as for ethnicity, in the Oxford English Dictionary of the early 1970s it still occurs only as a rare word indicating ‘heathendom and heathen superstition’ and documented by quotations from the eighteenth century.

In short, we are dealing with terms and concepts which really come into use only in the 1960s. Their emergence is most easily followed in the usa, partly because it has always been a society unusually interested in monitoring its social and psychological temperature, blood-pressure and other symptoms, and mainly because the most obvious form of identity politics—but not the only one—namely ethnicity, has always been central to American politics since it became a country of mass immigration from all parts of Europe. Roughly, the new ethnicity makes its first public appearance with Glazer and Moynihan’s Beyond the Melting Pot in 1963 and becomes a militant programme with Michael Novak’s The Rise of the Unmeltable Ethnics in 1972. The first, I don’t have to tell you, was the work of a Jewish professor and an Irishman, now the senior Democratic senator for New York; the second came from a Catholic of Slovak origin. For the moment we need not bother too much about why all this happened in the 1960s, but let me remind you that—in the style-setting usa at least—this decade also saw the emergence of two other variants of identity politics: the modern (that is, post suffragist) women’s movement and the gay movement.

I am not saying that before the 1960s nobody asked themselves questions about their public identity. In situations of uncertainty they sometimes did; for instance in the industrial belt of Lorraine in France, whose official language and nationality changed five times in a century, and whose rural life changed to an industrial, semi-urban one, while their frontiers were redrawn seven times in the past century and a half. No wonder people said: ‘Berliners know they’re Berliners, Parisians know they are Parisians, but who are we?’ Or, to quote another interview, ‘I come from Lorraine, my culture is German, my nationality is French, and I think in our provincial dialect’.footnote1 Actually, these things only led to genuine identity problems when people were prevented from having the multiple, combined, identities which are natural to most of us. Or, even more so, when they are detached ‘from the past and all common cultural practices’.footnote2 However, until the 1960s these problems of uncertain identity were confined to special border zones of politics. They were not yet central.

They appear to have become much more central since the 1960s. Why? There are no doubt particular reasons in the politics and institutions of this or that country—for instance, in the peculiar procedures imposed on the usa by its Constitution—for example, the civil rights judgments of the 1950s, which were first applied to blacks and then extended to women, providing a model for other identity groups. It may follow, especially in countries where parties compete for votes, that constituting oneself into such an identity group may provide concrete political advantages: for instance, positive discrimination in favour of the members of such groups, quotas in jobs and so forth. This is also the case in the usa, but not only there. For instance, in India, where the government is committed to creating social equality, it may actually pay to classify yourself as low caste or belonging to an aboriginal tribal group, in order to enjoy the extra access to jobs guaranteed to such groups.

But in my view the emergence of identity politics is a consequence of the extraordinarily rapid and profound upheavals and transformations of human society in the third quarter of this century, which I have tried to describe and to understand in the second part of my history of the ‘Short Twentieth Century’, The Age of Extremes. This is not my view alone. The American sociologist Daniel Bell, for instance, argued in 1975 that ‘The breakup of the traditional authority structures and the previous affective social units—historically nation and class...make the ethnic attachment more salient’.footnote3

In fact, we know that both the nation-state and the old class-based political parties and movements have been weakened as a result of these transformations. More than this, we have been living—we are living—through a gigantic ‘cultural revolution’, an ‘extraordinary dissolution of traditional social norms, textures and values, which left so many inhabitants of the developed world orphaned and bereft.’ If I may go on quoting myself, ‘Never was the word “community” used more indiscriminately and emptily than in the decades when communities in the sociological sense become hard to find in real life’.footnote4 Men and women look for groups to which they can belong, certainly and forever, in a world in which all else is moving and shifting, in which nothing else is certain. And they find it in an identity group. Hence the strange paradox, which the brilliant, and incidentally, Caribbean Harvard sociologist Orlando Patterson has identified: people choose to belong to an identity group, but ‘it is a choice predicated on the strongly held, intensely conceived belief that the individual has absolutely no choice but to belong to that specific group.’footnote5 That it is a choice can sometimes be demonstrated. The number of Americans reporting themselves as ‘American Indian’ or ‘Native American’ almost quadrupled between 1960 and 1990, from about half a million to about two millions, which is far more than could be explained by normal demography; and incidentally, since 70 per cent of ‘Native Americans’ marry outside their race, exactly who is a ‘Native American’ ethnically, is far from clear.footnote6

So what do we understand by this collective ‘identity’, this sentiment of belonging to a primary group, which is its basis? I draw your attention to four points.