a chariot and the archer who drew his composite bow on a rapidly moving
vehicle. There had been warriors in the third and early second millennium BC,
but neither in the Near East nor anywhere else were they an elite class. For the
rest of the Bronze Age the new military class was to be near, or at, the top of a
ranked society.
Chariot warfare and a warrior class
In a long series of articles and books Georges Dumézil argued that in Proto-Indo-
European society men were divided into three classes: priests, warriors and
producers (artisans and farmers).^2 Dumézil found such a division especially in
ancient India and Italy, but also in early Keltic and Germanic societies. The
assumption that a tripartite society was characteristic of the Proto-Indo-European
speakers, however, was supported by no archaeological evidence, as Colin
Renfrew objected in his critique of Dumézil’s interpretation of Indo-European
mythologies.^3 Although Renfrew put Proto-Indo-European back in the seventh
millennium BC, he had no doubt that even if one accepts the “short” chronology
and puts the PIE society late in the fourth millennium BCthere is no archaeological
evidence whatever—in any of the possible PIE homelands—for either the priestly
or the warrior class that Dumézil posited. Renfrew was right, so far as I can see,
that the tripartite scheme cannot fit the PIE speakers, regardless of where or when
one places them.
In the middle centuries of the second millennium BC, however, the existence
of a warrior class in several Indo-European subgroups is very clear. For Aryan
India it is attested by literary evidence, for considerable parts of temperate Europe
by archaeological evidence, and for Mycenaean Greece by all sorts of evidence:
documentary, archaeological and literary. There is also ample archaeological
evidence for a warrior class in northern Italy in the middle centuries of the second
millennium BC. Although Dumézil ignored the time, the place and the circum -
stances of the warrior class’s appearance, he was right to emphasize—as also did
Émile Benveniste and other Indo-Europeanists before and since—the existence
of an elite warrior class in many prehistoric Indo-European societies.
As I have argued in the preceding chapter, before ca. 1750 BCmost of western
Eurasia had no warrior class and certainly no military elite. Clearly there was none
in the Greek mainland or in temperate Europe, where weapons designed for battle
were not yet in use. Neither was there a warrior class—to say nothing of a military
elite—in the Near East during and before the Age of Hammurabi. Siege warfare
required the service of thousands of laborers, whom royal officials conscripted
either from the kingdom’s ilkum-servants or from its general population. When
battles needed to be fought they were fought largely by rural youths, many of whom
the kings hired from nomadic pastoralist tribes. Although these hand-to-hand
warriors were supported by grants of land or by daily provisions, they were far
from an elite. They were simply doing for a king what they had learned to do in
order to protect their flocks and herds, and with the same weapons. The only hints
of an elite military class in western Eurasia prior to ca. 1750 BCare the spears
110 Chariot warfare and militarism