One of the most immediately obvious things to any student of comparative mythology is the similarity
between the different folk beliefs of various cultures, specifically the ancient Indo-European peoples. This can take the
form of general motifs (for example, the war in heaven story, which is common to many mythologies) or the specific
appropriation of deities and legends by one culture at the expense of another (e.g. Apollo entering the Roman
pantheon, name unchanged, by way of the Greeks). In 1929, a prominent scholar of Indo-European history
and religion named Georges Dumézil put forth the theory that one of the hallmarks present in all forms of Indo-European
civilization was a division of society into three distinct classes: those who fight, those who pray, and those who toil. The
name given to his idea is the trifunctional hypothesis.
According to Dumézil, this is one of the most remotely ancient identifying features of the proto-
Indo-Europeans, alternately called Aryans or the more politically correct Arya (this seems to have been a name by which
these people referred to themselves, still present in placenames like "Iran" and "Eire;" it more properly refers
to the Indo-Iranian subgroup and will be used in that context). This was a thorough and strict division, supposedly
reinforced through religion. Additionally, it was understood that the first two groups, warriors and priests, were the
traditional rulers over the third groups, the agricultural workers. This sovereignty could be a joint effort or it could be
one class dominating the other two, but Dumézil was explicit that warriors and priests had separate functions in rulership
and were therefore represented by two different types of deities and traditions.
Dual Sovereignty
In his book Mitra-Varuna, Dumézil outlined the trappings of this dichotomous form of sovereignty,
looking at prototypical figures he called the magician-king and the jurist-priest. The magician-king is wild, unruly, and
very much connected with the spiritual world, while the jurist-priest is temperate, reserved, and concerned with earthly
matters. This is similar to the Apollonian/Dionysian dichotomy proposed by Friedrich Nietzsche, with Apollo
representing a more rational, logical worldview and Dionysus being indicative of pathos and ecstasy. The magician-king
and the jurist-priest are opposites but are inextricably tied together.
The Aryan deity Varuna (perhaps a linguistic forerunner of the Greek god Ouranos) represents the earliest known example
of the magician-king, violent and typified by an inflexible adherence to the letter of the law; indeed, his name may derive
from a proto-Indo-European word for "binding." The magician-king is seen as the progenitor of society, but it is the duty of
the jurist-priest to codify and maintain it. The jurist-priest is represented by Mitra, whose name derives from a PIE word
meaning "friend;" under Varuna, the law is no more advanced than debts and an eye for an eye while Mitra relaxes this
somewhat by initiating the notion of contracts and establishing a relatively even playing field among people entering into
legal agreements (rather than being an arbiter of punishment as Varuna). Indeed, despite their differences, the two gods are
frequently referenced in tandem as Mitra-Varuna in the context of taking oaths as coequal guarantors of the legitimacy of
said oaths.
Dumézil compares Varuna and Mitra with other figures like the Germanic deities Odin (mystical) and Tyr (prudent), the
early Roman kings Romulus (unpredictable founder) and Numa (conservative legislator), and the Irish-Celtic gods Lugus
(name derived from proto-Celtic for "oath") and Nodens (the derivation of this name is uncertain). A strange motif of
unclear meaning involving eyes and hands also plays a role in this concept and extends the metaphor to other figures and
stories. While there is nothing particularly significant involving the eyes and hands of Mitra and Varuna (or indeed Romulus
and Numa), Odin and Lugus are both one-eyed gods and Tyr and Nodens are both one-handed. Tyr is probably the oldest of the
Germanic deities, his name being an abbreviated form of Tiwaz, which derives from the same PIE word that gave us Zeus,
Dios, deus, and theos. Tyr is regarded as the father of the Germanic pantheon while Nodens was the original ruler of the
Tuatha de Dannan, a legendary race of people who conquered Ireland from the Fir Bolg. Additionally, two early Roman
heroes, Horatius Cocles and Mucius Scaevola, are respectively one-eyed and one-handed. Zeus seized power from his father
Cronos with the assistance of Cyclopes and strange hundred-handed creatures (it is worth noting as well that
Cronos governed the world without laws while Zeus provided order; there is a legend that Romulus was killed by leading
aristocrats due to his arbitrary rulership and replaced with Numa to bring stability). In the cases of Cocles and Lugus,
their eyes were lost in battle while both Scaevola and Tyr were required to sacrifice their hands to uphold oaths. These
early Romans are typically regarded as being at least partially historical, but it seems clear that these ancient legends
were superimposed upon the real figures for the purposes of narrative.
Society and Ritual
It is rather difficult to say with any certainty what ancient proto-Indo-European society was really like.
Their language, such as it was, exists only in hypothetical reconstructions based on expert analyses of other known
Indo-European languages. Not that knowing it would do us much good, since it wasn't a written language (as far as we know,
anyway). Even after the original culture spread from its (still unknown) geographical center, the majority of the texts that
we have from various Indo-European peoples are religious in nature and are not really of much use in determining the way
society functioned. Dumézil's most famous student, the Estonian-American scholar Jaan Puhvel, compared the utility of this
to trying to figure out the nuances of Christian theology and society based on the songs in hymnal books.
While he might be overstating his case, the essential point is still valid. Like the language problem, the best we can do is
to compare more historically recent examples of Indo-European society and extrapolate from that the nature of PIE societal
division.
The easiest place to start is probably the Indian caste system. Dumézil, Puhvel, and others hypothesize
that the original three castes were the Brahmins, the Kshatriyas, and Vaishyas, with others such as the Shudras being
added later to accommodate demographic shifts. These three were, respectively, priests, warriors, and workers. It is supposed
that ancient India was a hotbed of racial struggle, involving native Dravidians and invading/occupying Aryans from Iran.
There is evidence both for and against this notion, usually referred to as the Aryan Invasion Theory. The name pretty much
says it all; these Aryans were able to essentially conquer the Dravidians and establish the caste system which placed
themselves at the top and the natives -- who did not follow or neatly fit into the established order -- at the bottom as
Shudras and Untouchables. The mixing of indigenous and invading genes over the last several millennia makes DNA testing
essentially irrelevant in this, so we'll probably never really know for sure if it was an "invasion" or just a migration and
whether or not the caste system was in place before or after the Aryans came.
What we do know, however, is that over time, the caste system became quite rigid with sovereignty being exercised
exclusively by warriors and priests, whatever their racial extraction. Although it seems not to have been determined by birth
at first, inter-caste mixing was not acceptable either. The Vedas say that Brahmins ought to cease all religious
instruction at the sight of soldiers and not to resume it until after they have left; they aren't even allowed to get on
horses, given their association with war. Interestingly, the same prohibitions existed much later for Roman priests, known
as flamines. Dumézil posited a linguistic relationship between "brahmin" and "flamen," which makes sense when you consider
that his most famous book dealing with this subject is titled Flamen-Brahman.
The Romans also had something of a caste system in place, although it was a bit more fluid than the Indian one seems to
have been. Romans were divided primarily into patricians, plebeians, and slaves, with other subdivisions
existing and coming and going over time. Patricians were aristocrats, plebeians were workers, and slaves were, well, slaves.
Within the patricians existed a group called Equites, sometimes called Equestrians or Knights. Originally, the only people
who could be soldiers were those who held property, namely patricians. These warrior patricians evolved into the Senatorial
class and were prohibited from engaging in commerce, which meant that Equestrians, plebeians, and freed slaves were the
only ones who could work for money. The priests of pre-Republican Rome came exclusively from the patrician class as well,
which echoes back to the concept of dual sovereignty. It is interesting that slaves in Rome were primarily from conquered
nations, similar to the manner in which ancient Aryans were alleged to have placed native Dravidians at the bottom of their
own caste system. Even in the medieval era, Western European society was generally categorized into bellatores, oratores,
and laboratores, literally those who fight, those who preach, and those who work.
Indo-European religion fills in a lot of the other blanks here. A standard Persian prayer exhorts Ahura-Mazda (a dyad of
deities like Mitra-Varuna) to protect the empire from "hostile armies, a poor harvest year, and from false wisdom," showing
the importance of the threefold division. An optimistic saying in the Rigveda says that a kshatriya will overcome adversity
through arms, a vaishya by his money, and a brahmin by his offerings. In Germanic lands, a king's success was measured by his
combat prowess (sigrsäll), the country's harvest (arsäll), and his personal runic knowledge.
Starting in the 1950s, very well-preserved bodies dating back between 1,500 and 2,000 years ago began to be discovered
in bogs across the British Isles, Germany, and Scandinavia. Somewhat prosaically, these are called "bog bodies." An
interesting feature of these bog bodies is that in addition to their amazing state of preservation (skin and hair are still
on them and undigested food is still present in some of the stomachs), all of these people seemed to have been murdered in
multiple, redundant ways: usually through strangulation, blunt head trauma, and a slash to the throat. A clue to solving this
mystery comes to us from the Roman poet Lucan, who speaks of a Gallic practice of human sacrifice to Teutates, Esus,
and Taranis. Teutates was a god of war, Esus was a god of the people, and Taranis was a god of the heavens. It is theorized
that each required a separate type of sacrifice and instead of sacrificing three people one time, the Celts sacrificed only
one person three times, evolving into what is called the triple death ritual. Whether the sacrifices were voluntary or
involuntary is unknown to us, although the victims with few exceptions mostly seem to have been young men in their 20s or
30s. Ergot has been found in some of their stomachs, suggesting that they were in some sort of hallucinogenic state just
prior to their deaths. The triple death motif appears prominently in Germano-Celtic lore, with figures such as Merlin and
Odin enduring three types of deaths.
Conclusion
I personally find the argument that proto-Indo-European society was built around three classes convincing
given the evidence. However, the one thing that should almost go without saying is that most societies throughout history,
whether Indo-European or not, were based to some extent on the same tripartite division. Certainly, all three classes are
features of basically every society. The stark divisions existed in France until the French Revolution, most famously
through the concept of the Three Estates. In any case, the trifunctional hypothesis serves primarily to show that the more
different people are, the more alike they really are.