Thursday, 30 October 2014

Ethnological Method

Applying the comparative method in historical linguistics is relatively straightforward and academics habitually apply it to any language they come across at some point or other. But the comparative ethnology attached to linguistic constructions isn't so straightforward or quite so reflexively applied, presumably because 'ethnology' is a word reminiscent of the nineteenth century and because there is no academic department devoted solely or even primarily to such things.
What I mean by this is that the study of the poetics, myths, houses, boats, weapons, foodstuffs, recipes, tools, and so forth, associated with different language families is somewhat under-developed and connecting archaeology and linguistics isn't often a rigorous process. While there are quite a few excellent introductions to historical linguistics on the market (Campbell; Crowley; Trask), there aren't so many on the use of historical linguistics or language families in general studies of prehistoric topics. That's a pity, because with so many well-established language families around these days, it ought to be a great time for more synthetic prehistory.

One of the best books on Indo-European that I've read is Martin West's Indo-European Poetry and Myth. I've mentioned it before, and I tend to revisit it around this time of year, inspired by the strange and atavistic festivals of mid-autumn. It always gives up new things to ponder and enjoy, primarily because it's so packed with information. What's especially good is that West has clearly thought a great deal about the method he applies in the book; while it seems a tad scattershot, there's a real rhythm in the madness, and he tells us so at the beginning.

He says (p.19):
As soon as [some Indo-Europeanists] find a parallel between two individual traditions, say between Greek and Indian myth, they at once claim it as a reflex of 'Indo-European', without regard either to the groupings of the Indo-European dialects or to the possibilities of horizontal transmission. Greater sophistication is needed.
 This kind of ethnology is a field with plenty of nuts - Afrocentrists, neo-Nazis, ancient astronaut types, Hindu nationalists, you name it. Greater sophistication is a must.

West's plan - quite a common sense one, I'd say - is to use the sub-divisions of language families as the basis of comparison: comparisons between Germanic traditions won't take us back to proto-Indo-European itself, but only, perhaps, to proto-Germanic. But it's a bit more complex than that. Showing a simplified diagram of the Indo-European family with four levels (proto-Indo-European as Level 1; Anatolian and Mature Indo-European as Level 2; Western, Northern-Central, and Eastern as Level 3; and the families of Italic, Celtic, Germanic, Baltic, Slavic, Greek, Armenian, and 'Aryan' taking up the fourth), West remarks (p.20):
It is to be noted that these levels are stemmatic, not synchronic. A significant parallel between Homer and the Rigveda should take us back to Level 3 (and according to the argument presented earlier) to about 2300 BCE, the time of the Sixth Dynasty in Egypt; one between Italic and Celtic will likewise take us to Level 3, but perhaps only to 1300 BCE, contemporary with Mycenaean Greece or early Vedic India. [...] To get back to the deepest level, to PIE, we shall require a comparison involving Anatolian.
 This stemmatic approach makes archaeological comparison a bit tricky (it confounds dating a bit), but it's probably the way to go for things like myth and ritual, for which we rely largely on textual information. Later:
Most of the time... the evidence will take us no further than Level 2 or 3.
 And moreover (p.20-21):
Sometimes I am aware, or the reader may discover ahead of me, that certain things which I attribute to Indo-European (PIE or MIE) are also to be found in Semitic or other non-Indo-European cultures. That does not lessen the value of the results, as my object is to identify whatever is Indo-European, not just what is distinctively or exclusively Indo-European. [...] Nor is the reconstructive method invalidated by objections on the lines 'the parallel motifs that you note in this and that source need not imply a common Indo-European prototype, because they occur all over the world'. If a motif is indeed universal, all the more likely that it was also Indo-European.
Other considerations seem fairly sensible - parallels between traditions have to be 'specific and detailed enough to indicate a historical connection', and efforts have to be made to discount horizontal transfer, just as they are in historical linguistics.
[Linguists] are usually able, on phonological or morphological grounds, to identify elements that a language has acquired by horizontal transmission and not by inheritance, for example Iranian loanwords in Armenian or Celtic ones in German. It is not so easy in the case of myths and motifs, unles they are tied to specific names.
 He proceeds to give examples showing the difficulty of eliminating this kind of horizontal transmission - for instance, archaeological and historical evidence of connections between speakers of Indo-Aryan and Hellenic languages in the Mycenaean period, or contacts between Hittite and Hellenic speakers. This makes this kind of ethnological approach inherently difficult to pull off and any answers derived from it inherently speculative and inconclusive.
When we have parallels that extend all the way from India or Iran to the Celtic world, their probative value may be rated particularly high, because horizontal transmission seems virtually ruled out. But even then we must be cautious. The heroic traditions of both India and Ireland portray warriors using horse-drawn chariots. We might be tempted to infer that this was an Indo-European style of warfare... But it cannot be so, for archaeology tells us that the war-chariot with spoked wheels - the only type of vehicle light enough for horses to pull at speed - first appeared east of the southern Urals around 2100-2000 BCE, long after the Indo-European cultural continuum had broken up. (p.23)
The Trundholm sun chariot - it forms the cover of Indo-European Poetry and Myth, but as the author notes it doesn't go back to an Indo-European precedent. Wiki.
West's solution to this is, as he does throughout the book, to throw all the spaghetti on the wall to see how much might stick. He routinely puts weak arguments next to strong ones. He doesn't shy away from more speculative arguments or similarities that might be better explained through means other than direct transmission from proto-Indo-European precedent, but rather approaches these problems head on, noting where something might have been introduced from elsewhere (eg Semitic, Dravidian, etc). This requires erudition, knowledge of several languages, and an awareness of the archaeology, and that makes this kind of thing rather difficult. Perhaps that's why there haven't been so many books on the process.

Anyway, his is an excellent book and the impressionistic method he outlines is a good starting point for grappling with language families and history, and students of Austronesian, Arawakan, Mixe-Zoquean, and so forth, would do well to give it a read - or, better yet, to own a copy and dip into it now and then. More than anything, West shows how hard it is to demonstrate shared heritage, and that's important when arguments over shared heritage form the backbone of several nationalist and fascist movements.
If ideas and myths could spread so far and so fast over lands that had been Indo-Europeanized long before, how can we ever know if we are getting back to an original common heritage? Perhaps we cannot. Perhaps we must content ourselves with identifying 'isomyths', elements shared by a particular pair or a particular constellation of peoples, acknowledging that they may date only from a comparatively late phase in the long history of the diaspora.


  1. This is one of the best pieces you've written. I appreciate the distinction between comparative linguistics and comparative ethnology, or the lack thereof. Much work needs to be done in clarifying antecedents of culture traits, peoples, etc.

    Are you related to Martin West?

    1. Thank you - I just threw this together after reading that particular section. There's much more to say, of course.

      Not related as far as I'm aware!


You can post anonymously if you really want to, but I would appreciate it if you could provide some means of identifying who you are, if only for the purpose of knowing who has written what.