Arguments against Dumézil

I have said on countless occasions that Dumézil and his theories are not very popular among scholars nowadays. I have read arguments against his tripartite system that were sometimes convincing, sometimes not, but no scholar who disgards the hypthesis of Dumézil presents a workable alternative. Currently I am reading the book The War Of The Gods by Jarich Oosten. The book is of 1985 so the criticism is not just of today. The book is subtitled The Social Code in Indo-European Mythology. Some of you might now know what sort of book this is, but only when I started reading it, I learned that this book is written from the (cultural) anthropological viewpoint. Not completely my thing it seems, but the author writes fairly clearly and he takes a couple of pages to say something about Dumézil and his tripartite system which is worthy to think over. I will quote the mentioned book extensively.

Dumézil has invented the traditional approach to the pantheon. The functions of the gods cannot be derived from their nature, rather their nature has had to be inferred from their functions. Thus the pantheon is no longer conceived of as a collectivity of gods, but as a structured complex of functions. The organisation of the pantheon is explained by his theory of tripartition. Although names of gods will vary, and different gods will succeed eachother in course of time, the structure of the pantheon will remain essentially the same. There will always be sovereign gods, war gods, and fertility gods. This conception of the pantheon as a complex structure was of great importance of the antropological study of Indo-European religions, and while Dumézil’s theory of the tripartition as the fundamental ordering principle of the pantheon cannot be accepted, his conception of the pantheon as a complex structure is very fruitfull.

Most gods represented many functions at the same time, and this constitutes a major problem for Dumézil’s classification. We have seen that sovereign gods were usually war gods, war gods fertility gods (Indra, Mars, Thor, etc.), and fertility gods were often related to kingship (the Áshvins, the Vanir, etc.). It could be argued that each god had a particular function that was most characteristic of him, and could be classified accordingly. This is clearly a risky procedure. We would have to develop criteria to distinguish between primary and secondary functions, and it is the combination of the functions that is often crucial. Sovereign gods had to be war gods and this informs us about the nature of sovereignty in Indo-European cultures. These gods came to power because they were able to defeat their enemies. The fact that many war gods were at the same time fertility gods tells us something about the intrinsic relation between fertility and war, death and life in Indo-European religions, and we should not disregard the fertility function because we want to classify these gods as war gods. It is significant that many goddesses like Freyja in Scandinavian mythology and Morrigan in Celtic mythology were at the same time goddesses of war and of fertility. Women, however, play no part in Dumézil’s tripartition, and as a consequence it tells us nothing about the nature of the opposition male-female in Indo-European religions.

Dumézil’s model obscures the interweaving of functions. The relation between fertility and kingship is of great importance for our understanding of the sacral connotations of Indo-European king and their relation to the fertility of the land.

Oosten has a point here. Dumézil talked himself out of this by saying that in Scandinavian myth, the functions moved downwards with a half, but if Oosten is correct (and at first sight he seems to be) this does not only go for the Scandinavian pantheon. Oosten continues with saying that:

In short, Dumézil’s tripartition tends to simplify the complex structure of the Indo-European pantheon. It is better to classify the pantheon of the gods in such a way that the manifold functions of the gods can be examined without any arbitrary editing of their complex organization and functions.

And there comes the antropological approach:

The participants themselves did not classify their pantheon primarily in terms of the functions of the gods, but in terms of their kinship relations.

Oosten says that this is the reason for the extensive genealogies that we can find in myths and that the way the genealogical link is presented tells us something about the person giving the link. Snorri saw Odin as the highest of the gods and therefor Thor was a ‘lesser God’ and thus his son, rather than a brother.
So far, so good, but Oosten continues with arguments that I (with my Dumézilian perspective) cannot completely follow:

The supreme gods were great warriors. They wielded magical weapons that were made for them by mythical smiths. Thus Zeus and Indra wielded the thunderbolt, Odin the magical spear Gungir, etc. Although other war gods existed, the sovereign gods were often thought to be superior in strength and skill. They had proven their strength in battles with mythical monsters and in violent conflicts with their ancestors. Indra killed the snake Vrtra and his father Tvastr, Odin and his brothers killed the giant Ymir, Lug killed his grandfather Balor with his magical sling, Zeus his father Kronos, etc.

In Dumézil’s scheme, Indra and Thor are no sovereign Gods, but Gods of the second function. The hammer which Thor uses to slay giants is something different alltogether from the spear with Odin uses to ‘sanctify the battle field’; Odin does not use it to kill. The ‘original offering’ of Ymir to create the world, is something very different from Indra’s fight with the dragon. Oosten poses arguments without paying enough attention to his cause. Even without Dumézil scheme, one can still not equate Thor’s hammer with Odin’s spear or the killing of Ymir with the slaying of Vrtra (the latter is more like Thor’s fights with the Midgardsomr).
Then follows a whole range of information about “father gods” which does not conflict with Dumézil’s scheme in my opinion, however Oosten seems to talk mainly about the Odin/Varuna type of sovereign.
Furtheron Oosten says:

Thor was considered the god the god of the farmers and he possessed many features that were attributed to the sovereign gods in other Indo-European religion (possession of the thunderbolt, drinking of hugh quantities of the sacred potion, etc.). Although Tyr is usually thought to have been the predecessor of Odin, Thor is an equally suitable candidate forthis position.

That is a challenging remark, but faulty in my opinion. The thunder bolt of power fits well in the second function and where did Oosten find that Thor drinks huge quantities of mead? Beer, sure, but mead? Besides, several pages earlier says that old Norse tivar is an old term for “God”, Týr thus is such a *deiwos, an ancient sky-god, perhaps even (one of) the first, later replaced in popularity by Odin who on his term had to make place for Thor (instead of the other way around) among the Icelandic settlers.

Nope, Oosten gives a few good arguments against the hypothesis of Dumézil, but I find him not entirely convincing. This is partially caused by the fact that (so far) he has presented no alternative, but also because of the way he argues.

Sure, if we use Dumézil’s structure too strictly, we run into a lot of different problems; and sure, our ancestors undoubtely did not see their pantheon in the structure that Dumézil ages and ages later found in it through comparison with family-mythologies; and sure for many peoples the sovereign Gods were not the most important, but it is not for nothing that Dumézil said that the structure reflected on society so a farmer would more likely prefer a ‘third function God’ and a warrior a ‘second function God’.

Dumézil’s hypothesis is not a holy law that we have to force everything into, but personally I still find it the most workable structure for investigations in the field. It shakes and rattles since Dumézil posed it, but in my opinion it still stands. It is not for nothing that to this very day, there are scholars and non-scholars who use it.

It might be refreshing to read some other ideas, so perhaps it is good that the book of Oosten has a different viewpoint of my usual literature, but I wonder if he will come up with something credible.