In Search Of The Indo-Europeans: Language, Archaeology and Myth (Mallory)

[ewx]

Owis Ekwosk^we

In Search Of The Indo-Europeans: Language, Archaeology and Myth, J.P.Mallory, ISBN 0500276161

I've been interested in linguistics and the idea that a large number of modern languages can be traced back to common ancestors in the distant past for some time, but mostly this interest has been indulged in individual articles and conversations rather than in full-sized books. Obviously this was something due for correction.

The Indo-European languages are a collection of languages with a number of shared or related features which are now widely held to have evolved from a hypothetical “ancestor” language (though with borrowings from unrelated languages along the way) generally referred to as Proto-Indo-European. Examples of such languages include most European languages, Iranian languages such as Persian and Kurdish, and Indian languages such as Hindi and Bengali.

Early in the book is a table of numbers in many of these languages; there are wide differences, but there are enough similarities that there is obviously a case to answer. Through the book other points of similarity between smaller groups of languages are exhibited. The main thrust of the book is not in demonstrating the relatedness of the various languages involved, or describing the partially 'reconstructed' Proto-Indo-European language, however, but in the search for the time and place that the speakers of PIE lived and how the descendants of their language spread so far.

Given that this was thousands of years ago it might seem like an impossible task. Various techniques are mentioned. For instance, if most Indo-European languages have related words for “horse” then we can deduce that a word for horse must have been present in PIE, so its speakers must have at least known what a horse was. Another is to compare cultural traits found in modern or known historical societies with an Indo-European language with archaeological finds and see where they match and where they do not.

The conclusion offered, after an interesting wander around the world, is that the original Indo-Europeans probably inhabited the plains north of the Black Sea and Caspian Sea around six thousand years ago. It is much less certain about the means by which the Indo-European languages spread, though perhaps this is inevitable - in an illiterate society a language change won't show up directly in the archaeological record.

Mallory also mentions some alternative theories about the location of the Indo-European homeland, though mainly to rubbish them. While having a strongly held viewpoint is perfectly reasonable these sections felt rather petty and out of place compared to the rest of the book; this reader felt ill-equipped to tell whether the criticisms were valid, while someone who was already familiar with the field would presumably be familiar with the relevant controversy. I think the book would have lost little of real value had it not had the argument or even any mention of the alternative theories.

Comments

[songster.livejournal.com]

"Ewok eggs" ?

[ewx.livejournal.com]

Reconstructed form of ‘the sheep and the horses’ (http://www.geocities.com/Athens/Forum/3807/features/language.html).

[ex-lark-asc.livejournal.com]

That sounds *very* Latin.. how direct is the connection between PIE and Latin thought to be?

[vyvyan.livejournal.com]

Latin is a daughter of Proto-Italic, which is a daughter of PIE. In that sense, the connection is no more direct than between PIE and any other second-order descendant of PIE. Latin is much closer in time to PIE than modern IE languages are, so it looks much more similar (it has undergone fewer changes) - other IE languages attested from the same time as Classical Latin (e.g. Greek, Sanskrit, Gothic, little bits of Gaulish) also look pretty similar to PIE (and to Latin). Moreover, Latin and Italic (and Italian within Romance) are fairly conservative members of IE (unlike, say, Albanian, Parsi or Irish, which have undergone huge amounts of change).

[burkesworks.livejournal.com]

The conclusion offered, after an interesting wander around the world, is that the original Indo-Europeans probably inhabited the plains north of the Black Sea and Caspian Sea around six thousand years ago

Maybe Mallory has a point; I remember reading a few years ago that there is a theory that the closest extant languages to the original Proto-Indo-European are in fact Lithuanian and Latvian. Again, a subject that has never failed to fascinate me.

[vyvyan.livejournal.com]

I read that a while ago; it's pretty good on explaining the archaeological data, though the linguistic side is sometimes a bit dubious and there are other plausible interpretations which get ignored (Mallory is an archaeologist, not a historical linguist, after all). For instance, the point you mention about how we conclude that the Proto-Indo-Europeans lived in an area with horses - all the horse cognates in IE languages tell us is that PIE speakers were familiar with something which they called *ekwos, not that it was a horse as we know it: the meaning could have drifted over time. (In the case of horses, this is unlikely since I think pretty much all the descendents have the meaning "horse", but with other words, attested in fewer daughter branches, the problem can be serious; many forms reconstructable for PIE have very uncertain meanings.)

Mallory, IIRC, basically presents the status quo view on PIE origins (Marija Gimbutas' Kurgan model). I think it's reasonable for him to mention other proposals - it is after all a subject which has been controversial amongst linguists pretty much since the dawn of modern linguistics! (Which I am counting as the start of the 19th century, when IE began to be seriously researched, and principles such as regular sound change were observed.) However, another model which has been much discussed since the late 80s places PIE in Anatolia, around 8000 years ago. This is explored as part of a wider theory in Colin Renfrew's Language and Archaeology: The Puzzle of Indo-European Origins, which I would definitely recommend as a counterbalance to Mallory. (Renfrew is also an archaeologist rather than a linguist, but he recognises his limitations.)

My own research, incidentally, supported an earlier date for PIE than is traditional (more like 6000BC than 4000BC), with the first split being between Hittite and everything else, which would make an Anatolian origin much more plausible (c.f. Africa as the source of modern humans, and the site of the first three or four basic splits in the tree of human genetic diversity). So these days I tend more in Renfrew's direction than Mallory's (also, Renfrew is a great chap to talk to in person - enormously charismatic!).

[ewx.livejournal.com]

It's not the mentioning of other theories that I objected to at all - it's the treatment they get, which is very dismissive. Renfrew's book is already on my wishlist, and it's very likely I'll read it at some point.

[vyvyan.livejournal.com]

Oh, OK, I must have misunderstood what you meant by I think the book would have lost little of real value had it not had the argument or even any mention of the alternative theories. I think I would be disappointed by a book which purported to be about Indo-European origins, but which didn't mention alternative models (even to dismiss them).

The impression I got when I was immersed in this a few years ago was that most historical linguists (traditionalists, anyway) loved Mallory's account, while archaeologists seemed to be moving in Renfrew's direction.

[ewx.livejournal.com]

The intended interpretation was that I thought that treatment of alternative models was so bad that the book would have been better than it is now if it hadn't mentioned them at all. It would be better still if it mentioned them in a less negative way. (I thought the rest of the book was fine.)

[ewx.livejournal.com]

I've ordered a copy of Professor Renfrew's book.

In Search of the Indo-Europeans

[lethargic-man.livejournal.com]

It's a few years since I read this book^*, but I the impression I remember getting was that the book was part of an ongoing rather heated debate between Mallory and Renfrew, though I agree it did seem unnecessarily petty.

Amusingly, whilst Mallory is critical of the Encyclopaedia Britannica's lack of information on the subject, more recent editions of of the Britannica refer back to this book in the bibliography!

There was an article (http://www.nature.com/nsu/031124/031124-6.html) livredor pointed me to in Nature a couple of years ago (December 2003, I think) arguing from rates of linguistic change that the Proto-Indo-Europeans lived in Anatolia 9000 years ago - which is, of course, what Mallory spends a lot of the book arguing against. (Though the article acknowledges the Kurgan migration would still play a part in the picture; unlike Renfrew as portrayed by Mallory, they do pay attention to the archaeological evidence.) livredor, "lj user="rysmiel"> and I spent some time discussing this (though none of us have read the full article, only the summary of it then available on the Web (and now even that doesn't seem to be freely available)). It's a puzzling conundrum, paralleling the similar discrepancy between dates for human evolution from palaeoanthropological evidence and mtDNA rates of change.