mouthpiece between its teeth: if the horse clamped down on the ends of the jointed
canons it pushed the joint up into the roof of its mouth, an unpleasant and painful
sensation.
Anatolian chariot drivers were slow to adopt the jointed snaffle, although early
examples have been found in southern Caucasia, and the earliest examples in
temperate Europe may not be earlier than 1000 BC.^94 Apparently it reached the
Eurasian steppe at about the same time, and in this horse country it was revolu -
tionary. On the steppe very few bits are attested for the second millennium BC,
but a great many for the first. In Hüttel’s catalogue fewer than thirty bits are attested
for the steppe in the period ending with the tenth century BC, and none of them
are metal.^95 For the next three centuries, in contrast, Ute-Louise Dietz catalogued
from the same region 645 bits, 90 percent of them either bronze or iron and most
of these jointed snaffles.^96 From the “Skythian period” of the steppe hundreds of
iron jointed snaffles, most of them badly oxidized, have been recovered but never
published. My conservative estimate was that for every bit from the steppe attested
for the second millennium BCwe have fifty from the first millennium BC.^97
On the steppe the use of a jointed snaffle bit spread quickly both east and west,
and secure horseback riding spread with it: Iron Age riders sat forward, close to
the withers. The jointed snaffle bit contributed to the transition of steppe society
from the coexistence of settled and nomadic communities to unmixed nomadism.^98
By 700 BCthe jointed snaffle—whether of bronze or of iron—was the standard
bit from the Rhine river to the Altai mountains, and it remains the standard over
most of the world today. Soon after they became good riders, men on the steppe
learned how to draw a bow on horseback and the consequences were of historic
importance. In Barton Hacker’s summary, “archers on horseback were something
new—and transformative. Horse archery and equestrian techniques spread quickly
and widely through the steppes, as did the full-blown pastoral nomadism with which
they were closely associated.”^99
Gimbutas’ thesis that men in the Eurasian steppe were good riders already in
the fifth and fourth millennia BCfails on several grounds, and its survival now
depends on Anthony’s interpretation of the beveling of seven horse pre-molars
from Neolithic sites in northern Kazakhstan. Both Gimbutas’ original thesis and
the revision of it come up against not only the most direct archaeological evidence
for bridling components but also a wide variety of historical evidence—from the
documented world of the Near East—that both the driving and the awkward riding
of horses began shortly before 2000 BC, and that men did not become good riders
until ca. 1000 BC. The most recent argument that driving preceded good riding
on the steppe we owe in part to Anthony’s own initiative. We now know, thanks
to the carbon dates from Krivoe Ozero, that light, spoke-wheeled carts drawn by
horses were built east of the Urals ca. 2000 BC, and that these proto-chariots soon
spread through the steppe as far west as the upper Dnieper.
Because the prehistory of the steppe is by definition undocumented by written
records, and is rarely illustrated in pictorial evidence, it is unusually pliable and
open to diverse reconstructions. We are fortunate that on the subject of “the taming
of horses” our ample archaeological evidence from the steppe is supplemented
The Kurgan theory and taming of horses 49