218 Nino Luraghi
Balkan Ghosts, as in Robert Kaplan’s influential and misguided 1993 book (cf. Mueller
2000). The fact that surveys conducted by Yugoslavian sociologists in the 1980s revealed
lower levels of prejudice and hostility between Serbs and Croats than those character-
istically found in studies of ethnic attitudes in the United States comes as a sobering
reminder for those who get easily hypnotized by this rhetoric (references to the data in
Mann 2005: 360).
The success of the term “ethnic cleansing” is more than a mere terminological curios-
ity. In post–World War II parlance, the eviction of minorities could not, as before, be
described as “racial cleansing,” because the whole etymologic area of “race” has become
taboo, and for very good reasons. However, this is only one half of the story. While it
is clearly the case that the vocabulary of ethnicity can be a euphemistic replacement for
that of race, it is more than just that. First of all, the term “ethnicity” has the advantage
over “race” of not implying naturalistic or biological claims that our dominant cultural
discourse considers implausible. However, even more importantly, the term “ethnicity”
has traditionally been open to the defining use of cultural difference, and is therefore
a more flexible framework than race—a flexible framework for violence, among other
things. The fact that “ethnic identity” has become an accepted category in public dis-
course, with a reputable scholarly pedigree, has the paradoxical consequence of making
available a new system of inclusive and exclusive boundaries, which in turn lend plausi-
bility to the concept of ethnic conflict, which, as Michael Mann has shown, is a creation
of the twentieth century (Mann 2005; see also Wimmer 2002).
Of course, it would be ridiculous even to contemplate the notion that the study of
ethnic identity may in any sense be held responsible for the so-called ethnic conflicts
of the late twentieth century. However, a case may be made, and has been made in
various ways by various scholars, that even the most sophisticated approach to ethnic
identity runs the risk of reifying a cultural phenomenon, thereby affirming the per-
spective of some historical agents as if it were an objective fact. One factor that plays
a role in making scholars less sensitive to the potential creation of factoids by ethnic
rhetoric has to do with the habit, especially common in North America for historical
reasons, of regarding ethnicity as the underpinning of the rightful claims of underpriv-
ileged groups. In this framework, any attempt at deconstructing ethnicity pointing to
its constructed nature and to its frequent lack of roots in the past, as in the instrumen-
talist paradigm, can be seen as undermining those claims and promoting the agenda
of the dominant group. One familiar example is provided by Susan Alcock’s study of
Greek identity in the Roman Empire and of Messenian identity under Spartan dom-
ination (Alcock 2002), with its emphasis, on the one hand, on the implicitly opposi-
tional nature of the cultivation of Greek cultural heritage (see already Alcock 1993),
and on the other on the early archaic roots of Messenian identity. In both cases, the
rejection of an instrumentalist approach, which would emphasize the importance of the
contexts in which both phenomena are actually observed rather than their presumed
distant origins, appears to be a misguided expression of sympathy for what are per-
ceived to be oppressed human communities. One is left to wonder why the supposed
Spartan oppression of the once-free Messenians should provoke more outrage than the