Academic journal article
By Jones, Frank L.; Smith, Philip
Journal of Sociology , Vol. 37, No. 1
The comparative analysis of national identities
It is widely acknowledged that the study of national identity and nationalism has emerged as a significant area for contemporary sociological research. Less often recognized, however, are the characteristics of the field. It is a domain of inquiry primarily driven by historical, theoretical and qualitative research agendas. A straw poll asking about the major figures in the field would doubtless nominate scholars like Anthony Smith, Eric Hobsbawm and Benedict Anderson. There are no comparable figures in the area of quantitative empirical scholarship, which has so far lagged behind.
There are some serious implications arising from this imbalance. One is that we have very little idea about how similar national identities are in comparative terms, either in intensity or form. The methodological dictates of existing approaches tend towards analyses that emphasize the uniqueness of each national identity in terms of specific origins, content and structure. Of course, ultimately every experience is unique and needs to be understood in terms of its historical and socio-cultural context. But at the same time sociology as a generalizing discipline demands the construction of more abstract models. For this reason historically informed attempts to develop typologies and generic accounts, such as the distinction between ethnic and civic nationalism (Smith, 1991: 11-12), have been welcome contributions to sociological debate. Even here, however, there is a tendency to emphasize divergence rather than commonality.
Because the elaboration of theoretical models of national identity so often involves the construction of ideal types, an unintended consequence can be a picture of the world divided among nations, each informed by mutually exclusive models of citizenship and belonging. One thinks here of the various studies contrasting France and Germany, or of concepts like multicultural nations, settler states, ethnic nations and so on (e.g. Bryant, 1997; Greenfeld, 1992; Smith, 1995). These present an image of the globe that resembles a jigsaw puzzle. Each nation is a piece, and each of these is painted in just a single colour taken from a palette of limited concepts: orange for multicultural, blue for ethnic nationalism, and so on. Typifying knowledge is thereby attained but at the cost of gross simplification. If radical difference is reaffirmed, it must be constructed in terms of the distances separating the paradigms through which identity is represented. What is needed, we suggest, is an approach that pays due attention to nuanced differences in comparative discussions of national identity but also transcends the detail of idiographic historical experience. In other words, we need to balance our thinking by attending to commonality as well as to divergence.(1)
A possible way forward is to think through a third, and we believe more enduring, problem with dominant trends in the field. Historical and qualitative research methods typically rely either on documentary evidence (for example, immigration policy, the media and political speeches) or, less often, on key historical events (for example, national rituals and civil wars). There is always a problem in making inferences from this level of analysis to the beliefs of ordinary citizens. There is no compelling reason why models developed from such sources should correspond to prevailing beliefs and attitudes at the level of individual consciousness. The meanings of `citizenship' in everyday life, for example, bear no necessary relationship to their meanings in legal statutes, Weberian theory or multicultural policies. For most people, citizenship is just something you need in order to get a passport or a job. When it comes to theories about national identity, the work of Anthony Smith, Leah Greenfeld and others shows that `ethnic' and `civic' national identities can be analytically derived from the historical record (Greenfeld, 1992; Smith, 1995). …