archaeologists recorded evidence for an ancient presence at 336 sites in the
Çankırı and Karabük provinces of Turkey. Most of the evidence dated from the
Hellenistic and Roman periods, but twenty-six of the sites showed a presence in
the Early Bronze Age.^49 Before 3000 BCthe region seems to have been inhabited
mostly by hunters and gatherers. Project Paphlagonia found not a single Neolithic
(10,000 to 6000 BC) site and only six Chalcolithic (6000–3000 BC) sites, all from
the middle or later phases of that period.
Unlike the southern shore of the Black Sea, the northern shore obviously
attracted Neolithic settlers. The location of the settlements shows that the
colonization must have come by sea. That the settlers came from the west is
suggested by their fit within the Cucuteni-Trypillian archaeological culture, which
proceeded into the steppe from the eastern foothills of the Carpathian mountains.
The steppe itself was not very appealing: rarely receiving more than 300 mm of
precipitation a year, the grass was deeply rooted and the sod too tough to hoe
(oxen were not yet draft animals in the sixth and fifth millennia BC). The river
valleys were more welcoming, offering limited agriculture as well as grazing for
domestic animals, some trees, and of course plenty of fish. Neolithic settlers
accordingly made their way up the valleys of the Dniester, Southern Buh (or Bug)
and the Dnieper. In the Dniester and the Southern Buh the Neolithic material
remains are characteristic of the Cucuteni-Trypillian culture. The material record
from the Dnieper valley includes pottery very similar to that of the Cucuteni-
Trypillian culture but has conventionally been assigned to the Sredny Stog culture,
named for an island in the river some 150 miles upstream from the Black Sea.
Dereivka is the most informative of the Sredny Stog sites.
River valleys beyond the Dnieper were much more difficult to reach from the
Bosporus and the Balkans. For boats sailing eastward along the north shore of the
Black Sea the Crimean peninsula was a major obstacle. The peninsula itself was
mostly grassland steppe (the earliest evidence for food production there dates to
the middle of the fourth millennium BC),^50 with mountain ranges along the
southeastern coast, and circumnavigation was long and arduous. But there is good
reason to think that colonists did round the Crimea and proceed eastward.
One destination that could be considered, if we bypass the Kerch Strait and the
Sea of Azov and sail to the eastern shore of the Black Sea, is southern Caucasia.
Thomas Gamkrelidze and Vyacheslav Ivanov made a detailed argument that it
was in the Kura and Araxes valleys during the Neolithic period that PIE evolved
(their argument depended in part on correspondences between PIE and both
Proto-Kartvelian and Proto-Semitic).^51 Although 30 years ago I was persuaded
that south Caucasia was the PIE Urheimat,^52 I no longer am. Because terms for
wheels and wagons were inherited from PIE by all branches of the Indo-European
language family, linguists since early in the nineteenth century have recognized
that wheeled vehicles must have been important to the PIE speakers. In the last
30 years evidence has continued to emerge for the very early presence and
prominence of wheeled vehicles in northern Caucasia, but the same has not been
true for southern Caucasia.^53
Origins and spread of Proto-Indo-European 13