52 Thomas D. Hall
change, or disappear are shaped by their position within the world-system: core, semipe-
riphery, or periphery. One application of such an approach can be seen in Stein’s claim
that Uruk constituted a world-system (1999). He suggests that Uruk colonists in south-
ern Anatolia had minimal influence on the local residents. This is as one would expect,
since the outsiders were a distinct minority at a great distance from their homeland.
Furthermore, he implies that the colonists were there at the sufferance of the locals,
who selectively adopted only those foreign elements they found particularly useful. This
assessment rings true. This is a region where ethnic identities and the political-economy
interact intensely.
In pre-capitalist world-systems, the congruence between position in the world-system
and geographical location typically was stronger than in contemporary times, but often
changed significantly over time (Smith 2005). This raises the question of timescale. When
a social relation changes relatively very slowly, as often occurs in the ancient world, it does,
indeed, seem constant. However, when viewed from a longer timescale, it will often be
seen to be changing continuously (Hall 1998, 2004), as in the case of Greece and Egypt,
for example. Contemporary globalization processes in the global world-system have pro-
duced such rapid changes in identities that their impermanence and fluidity are readily
apparent. Hence, there is widespread contemporary concern with the social construction
of race, ethnicity, nationality, and sexual identity. Closely connected to the question of
timescale is the recursive relation between subparts and the overall system. An analysis
that examinesonlysystem-to-subpart relations, oronlysubpart-to-system relations, misses
many of the complex processes of change and stability. In short, a major contribution of
world-systems analysis is its insistence that all social analyses must simultaneously attend
to historical and inter-relational contexts.
Ethnicity and other forms of identity are quite complex (Caliendo and McIlwain
2011). Fortunately, readers will find an excellent overview of ethnicity and identity
in Bernard Knapp’s and Johannes Siapkas’s chapters (Chapters 3 and 5, respectively).
I want to add two points that complement and extend Knapp’s discussion. First,
the concept of ethnicity must be historicized. The referents of ethnicity—in both
contemporary and ancient settings—are not constant, but in fact vary considerably.
Timescale considerations crosscut this variation. This adds some insight to the debates
about instrumental or socially constructed versus primordial concepts of ethnicity. The
primordial approach views ethnicity as something deeply embedded, nearly immutable,
and socially prior to all other social relations. The instrumental or social construction
approach views ethnicity as built from social interactions. This view is an extension of
Fredrik Barth’s (1969) ethnographic demonstration that ethnicity is fluid and mutable.
In some instances, individuals or families may change their ethnic identity once or more
in one lifetime. The two concepts are often seen as complete opposites. I argue they
are poles of a continuum. A key difference is timescale (Hall 2004). When the social
processes that construct ethnicity take place over many generations, ethnicity doesseem
to be primordial. When the processes occur over years or decades, social construction
becomes more obvious. As a rule of thumb, in ancient times these processes were
slow, whereas in the recent centuries they have been more rapid (with exceptions
in either setting). The problem, or “error,” with the primordial view is that when
it is taken as “normal,” it renders ethnic change unexpected and problematic rather
than typical.