The three functions of Dumezil's tripartate system are: magico-religious, warrior, and herder-cultivator.
The priests are the providers of sovereignty, maintainers of the magico-religious function, and bringers of law (and, potentially, order). Most often, according to Dumezil, the deities associated with this area are paired, one religious deity and one legal deity:
|religious deity||legal deity|
I've thrown in that last one to show that even in America, you can find IE functions.
As a cultural example, in Gaul you find the Druids, who, according to Caesar's description in his commentaries on the Gallic War, dealt in the training of elite students, healing, advisors, astrologers, and law-givers. They fulfilled the basic functions of things religious and law-related.
The second function, that of the warriors, is generally associated or concerned with force in some way, either as
aggressors or as defenders. They protect and they conquer people. Most deities associated with this group are singular:
I think Dumezil is wrong on ascribing this to Thor, personally, as he is more of a third functional deity to me, though he does have warrior aspects. I feel Odin would be more appropriate to this function.
A cultural example of this from Gaul would be the knights (equites) that Caesar mentions in his commentaries, who are charged with keeping the people safe from outsiders.
The third function is the herder-cultivators, or the producers. Concerned with maintaining fertility and sustenance, these are the general people, the farmers and herdsmen that make up the bulk of the population, deriving their protection and support from the first two functions, offering the fruit of their labours in return for that support. Most often, these deities are a pair and a third, according to
A cultural example from Gaul would be the people Caesar forgets about: the farmers and general producers.
Knowing what we do about the functions and the cultures involved, we have to ask ourselves if this is central to our understanding of Indo-European cultures. I argue that it is not central, but is very useful.
We can glean important information from the study of this tripartation system (Mallory, 139), but this is emphatically *not* history. It relies on a sociological reading of the myths, which strikes me with suspicion, because such readings rely on reduction of the myths to a Jungian or Campbellian set of ideas, which can be dangerous. Tripartaion, then, can give us clues about where to seek things in the historical record, but it cannot provide that record. "Dumezil himself has insisted that his Indo-European civilization is one 'of the spirit', and that it need not be tied down to the real Proto-Indo-European world." We should take this theory as just that: a theory. (Mallory, 142)
Mythological evidence must be held up to archeological proof before it is presented as fact, and the central danger in making tripartation central to our Indo-European studies is that we will describe something as "fact" before we have ever managed to prove that it is.