When is a migration a diaspora?
The checklist, and what replaced it. William Safran (1991) proposed criteria for a diaspora: dispersal from an original centre; a collective memory or myth of the homeland; partial alienation in the country of settlement; the homeland held as a place of eventual return; a commitment to maintaining it; and a group identity defined by that relationship. Robin Cohen (Global Diasporas, 1997) loosened the checklist into ideal types — victim, labour, trade, imperial and deterritorialised diasporas — and insisted that real histories mix them. A single family can be a trade diaspora in one generation and a victim diaspora in the next.
Routes, not roots. Paul Gilroy (The Black Atlantic, 1993) and Stuart Hall (1990) argued that diasporic identity is not a diluted copy of an authentic original waiting at home. It is made in transit and in mixture — a becoming rather than a possession — which is why the second generation is not simply less connected than the first. Rogers Brubaker (2005) added the discipline: treat diaspora as a stance, project and claim that people make, not as a bounded group you can count.
How anthropologists study it
Follow, do not sample. George Marcus (1995) argued that a connected world needs multi-sited ethnography: follow the people, follow the thing, follow the money, follow the story. None of the migrations on this page can be studied from either end alone — the village and the city are one field site with an ocean in the middle, and the ties arcing overhead are what makes them one.
The habit to watch for. Andreas Wimmer and Nina Glick Schiller named it methodological nationalism: assuming the nation-state is the natural container of a society, so that anyone living across two of them registers as an anomaly needing explanation. Drop the assumption and these stop being exceptional lives — they are ordinary lives lived in a transnational social field. Watch the vocabulary too: expatriate, migrant and alien often describe the same movement with very different amounts of respect attached.
A rite of passage that never finishes. It is tempting to read migration through van Gennep’s three phases — separation, liminality, incorporation — and the first two fit well. But Victor Turner’s liminal phase is supposed to end. For a worker on a visa that can never become citizenship, or a family in a camp for twenty years, it does not. The unfinished middle is precisely what the ethnography has to describe.