by Oivind Fuglerud, 2004
In approaching the different responses to the engulfing violence above I avoided framing this question in terms of ‘migration choice’. From an anthropologist’s point of view the application of choice models in migration research has provided limited insights. We should, I believe, in analyses of displacement and migration put less emphasis on causal connections and more on the understandings and discourses that organise actors’ understanding of their own actions. As demonstrated by studies of the relationship between internal migration and suicide (Kearney and Miller 1987; Daniel 1989), in Sri Lanka the issues of migration and dis-placement cannot be separated from the larger theme of discursively con-structed identities.
In this chapter I have outlined two responses to a situation of prevailing violence in Sri Lanka; the Tamil migration as refugees to Western countries and a Muslim community’s withdrawal into its own defended territory. In my understanding this difference illustrates the point made by Hägerstrand (1969) thirty years ago, that migration – except for the most extreme cases of forced migration – is one out of several ‘time-space strategies’ available to social actors, the others including staying, com-muting, circulating etc. Instead of considering staying as the expected outcome of time-embedded processes, and moving the phenomenon always to be explained, one point of entry into the study of migration,‘ forced’ or not, is to identify determinants of the different time-space strategies which people choose. This process of identification should include not only economic goals and constraints but the perceptions and meanings through which actors understand the various elements in their life world and related to which they strive to secure a livelihood