Ethnicity and Geography 303declare them Libyans, since they do not want to have to observe Egyptian rites, in par-
ticular abstaining from beef. The oracle declares that, living on land watered by the Nile,
they must live as Egyptians. If the land inhabited by Egyptians is Egypt, then the peo-
ple who live in Egypt are Egyptians; however, at the same time, geographic distinctions
determined by purely topographical or geophysical features will inevitably conflict with
divisions based on cultural considerations. In this discussion, Herodotus slyly acknowl-
edges that a simplistic correlation of ethnic and geographic definition must fail to account
for aspects of culture that correlate with ethnicity.
The tendency to differentiate culturally distinct groups primarily by location remained a
central feature of geographical writing in succeeding centuries. The earliest independent
works in the corpus of geographical writers—works such as thePeriplousof Hanno, prob-
ably a garbled translation from a Punic original, and the fourth-centuryPeriplouswrongly
attributed to Skylax of Karyanda—develop the spatial conception of the world pioneered
by Hecataeus, while at the same time including occasional ethnographic description as a
means of ethnic identification. The fourth-century historian Ephorus, in the geograph-
ical books of his universal history, mapped the peoples of the world along the same
lines as his predecessors, producing a schematic representation of the world as a rect-
angle, bounded by Skythians to the north, Ethiopians to the south, Kelts in the west,
and Indians in the East (FrGrHist 70 F30a, b, Str. 1.2.28, Kosmas Indic.Topog. Chr. II:
148)—although, at the same time, he showed particular interest in the founding of cities,
migrations, and the founders of colonies (T 18a, Polybius in Str. 10.3.5), suggesting a
particular interest in origins and relationships between peoples. Similar to Hecataeus and
Herodotus, the later geographers tend to emphasize aspects of the way of life of the
peoples they describe that distinguish them from themselves and mark similarities with
putatively related groups. Such works implicitly seek to correlate spatial relationships to
ethnic relations. Later works of geographical description tended to include ethnogra-
phy and origin-stories, while focusing more on determining the exact physical location
of peoples and places. At the same time, a sub-genre emerged, of more or less purely
ethnographic writing, with works titledbarbarika nomima, or monographs such as the
PersikaandIndikaof Ktesias. Such works blended catalogues of customs with natural ̄
history and human history, but without the geographical framework of Hecataeus.
The desire to account for the link between lands and the peoples who inhabit them
led Greek writers to the development of several explanatory mechanisms for this con-
nection, beyond the genealogical one. One, predicated on the close association of ethnic
genealogies to “land genealogies,” was the claim of autochthony. The claim, that a par-
ticular people was “born from the earth” in the land they inhabit, was by no means a
universal explanation; it was developed and deployed in certain contexts, often for spe-
cific rhetorical or political purposes (see Roy, Chapter 16 in this volume; also Rosivach
1987; Loraux 1990; Pelling 2008–9). At the same time, even those who did not accept
the mythical dimension of autochthony began to advocate a role for the land and environ-
ment in the shaping of ethnic character. This idea has been attributed to Herodotus: in
the passage on the Egyptians discussed earlier, he posits a definitional role for geography,
although he recognizes that questions of cultural identification can complicate the rela-
tionship between a land and its people. The clearest expression in theHistoriesof the
idea that lands and their environments shape ethnic character comes at the very end of
the work, in the little scene in which Artayctes suggests to Cyrus that the Persians move