120 Gary Reger
in the interests of all groups was the desire to trade; one space where the rules were
worked out was in the sexual relationships between colonial men and indigenous women.
This congeries of characteristics lends support to the view of the existence of “many colo-
nial locales in history where complex local arrangements have been made and endured
sometimes for centuries. These middle grounds were by definition places of creative
hybridization” (Woolf 2009: 209). Be that as it may, caution is called for. For the mid-
dle ground, as so understood, is notipso factoa space for the production of a hybridized
ethnicidentity. Participants in the world of the middle ground do not necessarily undergo
a qualitative shift in their ethnicity; rather, they learn, and participate in the creation
of, a set of rules allowing them to interact successfully for well-defined purposes with
the other. Viewed through this lens, the middle ground may appear rather as a space
where ethnic boundaries are protected and strengthened precisely by figuring out ways
to interact without contamination (see the clear statement in Malkin 2011: 39, 45–8,
with explicit rejection of the notion of hybridity). Perhaps we might regard such a mid-
dle ground as just a grand elaboration of the “silent bartering” that Herodotus and
other Greco-Roman writers report as the barest way for the civilized and the barbarian
to trade.
In the Bedroom
Parents are crucial to agronomic hybridization, and so it is no surprise that a possible
context for the emergence of a hybrid ethnicity occurs in inter-marriage between dif-
ferent ethnic groups. If exogamy, as has been said, is the best way to destroy ethnic
boundaries (Whittaker 2009: 189, paraphrasing Barth), then surely it might also play a
role in hybridizing them. In the Greco-Roman world, this practice has a claim to being
especially widespread in colonial situations. Surely the most familiar story is the kid-
napping of the Sabine women by the Romans, but the theme can be found repeatedly
in colonization/foundation stories (see, for instance, Dougherty 1993: 67–8). Histori-
cal evidence for ethnic intermarriage has been sought in the practices of Roman troops
stationed throughout the empire in the first through third centuriesCE. Troops were
often stationed at great distances from their homelands, so that access to women of
the same ethnicity for long-term, childbearing relations was extremely difficult, if not
impossible. (It should be remembered that the Roman prohibition on soldiers’ mar-
rying, lifted finally under Septimius Severus, did nothing to stop troops from estab-
lishing relationships resulting in children but not formal marriage; see Phang 2001.)
It has thus been argued that soldiers were likely to have found partners among local,
non-Roman women; their children, then, would be good candidates for a pool of people
among whom a hybrid ethnic identity might emerge. Examples might include Carteia
in Spain and Lugdunum Convenarum in France, both said to be populated by chil-
dren of Roman soldiers and local women (Woolf 2011: 18). However, in the case of
North Africa, such evidence as we have for such relationships—typically commemorative
epitaphs—shows virtually no such mixing; in fact, the chief pool of women to whom sol-
diers had recourse were the daughters of their comrades or predecessors (Cherry 1998:
101–40). Perhaps local women were more often exploited as short-term sexual slaves, as