PIE and Me

If you do serious thinking or studying in Proto-Indo-European linguistics, etymonline is not made for you.

It's embarrassing to have to warn people off like that, but it has to be done. They ignore the signs all over the lawn. If you're really a student of linguistics, and the crudeness of the presentation of PIE on etymonline doesn't warn you off it, I don't know what will.

The PIE currently in etymonline represents various stage of giving up on the whole thing, except in the broadest, most general way. The most it aims to tell you is something like:

"linguists (many of them) think (or sometime have thought) that this word and that word had a common prehistoric ancestor. And here's a set of letters in an accessible alphabet that looks something like what that never-written ancestor-word might look like if the prehistoric folk had had only these 26 letters and knew what to do with them."

Which is eye-rolling stupid. At the beginning, I didn't put PIE in etymonline. I do history; PIE has asterisks. I say "potato" and they say "pəˈteɪtəʊ." PIE scholarship is like baking: you mix ingredients and, voila, get a pie. History is the butcher shop, the slabs of meat are in front of you, or they aren't.

But the people using etymonline kept pestering me for it, so I buckled. Pokorny was still the bible, and an updated version of his roots, without stumbling about in German, was available in the Calvert Watkins "American Heritage" book of PIE roots.

That list of roots was the closest thing ever to a "popular" PIE: it was included in the regular AH dictionary, too, and you could buy it separately at chain book stores in the flyover states. That list, checked against and supplemented by whatever was in the old print OED, formed the basis of the PIE entries in the earliest etymonline.

But it was at the end of its run. From about 2010 users of the site began to draw my attention to new IE dictionaries for Germanic, Latin, and Greek (the books identified in the site text as "Boutkan," "de Vaan," and "Beekes"). Those books brought the scholarship up to that date.

A great many of the reconstructed etymologies in the three new books differed from those in Watkins. The newer books gave the reconstructed roots in forms other than the old ones, used different superscripted characters, gave other cognates, and not infrequently fell to arguing among themselves.

I get it, I think. The initial reconstructions (Pokorny era) were tightly focused on tracing threads through the ancient IE languages to reweave a PIE root. A following generation of scholars makes it its business to notice that other, non-IE, languages must have been spoken in Europe before the IE hordes arrived. People in new places often pick up the local names for things even while they're exterminating the locals.

So the next generation of linguists pulls up the carpet and suddenly sees pre-IE words poking up into IE languages like indigenous words on the American map.

Thus many of the old answers flip. And there's another generation of scholars rising. You can watch the same tug-of-wars in every wing of the humanities building. But with PIE there's no box of tangible evidence you can go to and pass around. PIE seems to me a branch of philosophy. Or the stuff of a new discipline born from linguistics, archaeology and genetics.

Yet it also seems always just a step away from circularity (... the laws are based on the evidence, which is discovered and extracted by means of the laws ...). A 3-D chess match with strict rules and invisible pieces. PIE debates remind me of watching my son and his friends in the back yard trying to have Pokemon battles when there's no real Pokemon.

It is not, at any rate, history. Etymonline is meant to tell you as much as can be found about the history of the words. History (read backward) ends at the first written word, and stands watching as PIE marches blithely on into the quicksand.

And never mind, for now, the temerity of modern academics reconstructing "why did they call it that?" There seems to be a sometimes-presumption that having spent 10 years in grad school studying PIE conjugation makes one an expert on the thought processes of post-Ice Age warrior-bands that likely sprinkled the children with blood of horses slaughtered after ritually mating with a queen.

I trust myself to do history, I don't trust myself to do the high-wire thinking involved in PIE linguistics.

It would be pointless to try to rewrite etymonline every year to chase PIE scholarship's tail; it seems to be chasing its own just fine. Hence, however, the deplorable state of PIE on etymonline. The PIE root entries are probably most out of date. They will show a lot of no-longer-current material from Watkins and Pokorny. If the information in the word-entry and that in the PIE root entry clash, trust the word entry.

Now, the rest of you PIE kids, get off my lawn.