38 A. Bernard Knapp
intimately connected to a particular place, whether at the village, community, or territo-
rial level. This “sense of place” (Feld and Basso 1996) is deep and enduring: villages or
communities become places of memory, where long-standing identities are confirmed or
new ones imagined (Mills 2004: 11). Even so, identity cannot be “possessed” by social
groups or individuals, nor can it be seen solely as the end result of belonging to a com-
munity. Rather, it is an unstable, often transitory relation ofdifference. An archaeological
approach focusing on differences—in symbolism, representation, boundaries, etc.—may
help to single out material practices shared between individuals or social groups, and thus
may provide insights into people’s social identities.
Given the complex and fragmented data sets with which archaeologists must work,
they tend to treat identity in one-dimensional terms. They also tend to assume that visual
representations (e.g., on painted pottery or frescoes, figurines) depict people displaying
their identity. Such representations, however, might be concerned more with ideological
or other constructs of identity than with actual lived experience (Pollock and Bernbeck
2000; Dipaolo Loren 2001). The way people dress and adorn their bodies also may
form a significant aspect of representing their identity (but cf. Dipaolo Loren 2003).
Other, shared social practices—visible materially as bodily ornaments, symbols, tools, or
utensils—may be actively involved in signifying ethnicity or in creating and expressing
social identities.
Questions of identity are fundamental to the material and social practices that link
personal experience to collective action; in other words, identity is linked closely to
the ways people perceive themselves as belonging to certain groups, and not to others
(Diaz-Andreu et al. 2005: 1). Nonetheless, if social identity is situational and negotiated,
then every social encounter has the potential to produce different material refractions of
identity (Mills 2004: 6). Thus, it would seem that both ethnicity and identity continue
to be difficult for archaeologists to pin down: as Joffe (2003: 88) exclaimed, “Capturing
ancient identities is like trying to drive nails though blobs of mercury,” but at least he
acknowledged that “...we can corral most of the blobs.” Let us see how all these caveats
and conditions play out in a protohistoric case study.
Ethnicity and Archaeology in the
Eastern Mediterranean
The Mediterranean region provides several classic examples of ethnicities as perceived
in historical documents, from the epic ventures of the Odyssey or the Aeneadto
episodes featuring Phoenician, Greek, or Roman colonists and merchants (see Bonnet,
Chapter 22, in this volume). Although “ethnic” groups are mentioned in historical
documents, or portrayed in stereotyped representations on painted pottery, monuments,
or wall paintings, it is quite difficult to link such groups to material culture distributions.
For one thing, it is typically rulers or bureaucrats—not the people whose identity is
involved—who provide such “ethnic attributions” (Brumfiel 1994: 96). For another,
although archaeologists also often use collective terms such as “Phoenician” or “Etr-
uscan,” presumably based at least in part on the perceived use of a common language,
we cannot assume that the speakers of a single language constitute a single ethnic group