I have not been convinced that evidence of a Proto-World language is recoverable, but I have been sympathetic to the idea that it is possible, and extremely critical of the way more traditional linguists have gone about criticising Merritt Ruhlen and Joseph Greenberg. When you have to go all Mean Girls* right off the bat, as my linguistics podcast did today, I am immediately suspicious. When I was in school we were assured that it was impossible to detect relationships between languages at a depth of more than 5000 years, and now it's grudgingly up to 10,000. There are also small bits of evidence for demonstrable relationships up to 20,000, such as the purported connection between Ket in the Yenisien Valley of Eastern Russia and some Native American languages, and some odd things that go even deeper. Also the evidence from other fields, such as genetics, has been supporting said "impossible" connections in biological fact, making linguistic connections at least more possible.
The Aspie quality that makes some historical linguists good at their job is not without its downside. If you can't make them believe it incontrovertibly, they smugly believe it must therefore not be true at all. Sigh.
It is a signal-to-noise problem, in that one side says they believe they can hear someone broadcasting on the wireless, while others say they are imagining things. Yet as equipment and technology get better, we might hope to accurately discern signal better. Also, there are a bunch of Russian linguists working with disparate families we have been unfamiliar with who have been claiming associations at deep time-depths for years, and I don't have any reason to believe they were less competent than our linguists.
Yet in defense of the critics, we have to ask that even if we can detect the signal, based on focusing on Swadesh words and particular sounds (see my first back-link for some of that), what have we got, really? In the Americas, Greenberg and those who follow him believe that there is one big Amerind family that includes most of the continent and all the Proto-Algonkian, Proto-Iroquoian, and Proto-Everything, plus some minority Na-Dene languages (Navajo, Hopi, and many Canadian) and the Inuit languages. Most linguists would break that into many more families. They say there's not enough signal to make solid assertions. And at this point we have to ask even if Greenberg, et. al are right, what's it mean? These are similarities so faint as to be unnoticeable. If there are connections between tribes and movements to be traced, perhaps we had better leave that to the archaeologists and geneticists at this point.
****
Fun fact: many Germanic and Slavic languages have a word for hoarding things that is based on the word "hamster." Yet it is not so in English, which is a Germanic language. Why not? Well, hamsters are not native to the British Isles, and only go as far as the Continent. Which is why instead of hamstering things, we squirrel them away.
*I saw a clip from "Mean Girls" for the first time in the review of The Dawn of Everything at ACX that I posted on a couple of weeks ago. While the characters were exaggerated for art's sake, I was impressed that they seemed like a great many of the girls I went to high school (and college) with - and more of the boys, once I thought about it. We are all Mean Girls. In the case above though, the podcasters were both female, adult academics, so maybe I just haven't the faintest idea what the real percentages are.
For a novel (stalled at 1/3) I thought it might be fun for an exceptionally old creature to speak PIE. Ummm. The reconstructed vocabulary is rather sketchy, and it wasn't clear to me what sort of grammar it had.
ReplyDeleteI don't know how the specialists distinguish between an original PIE word transmogrified by the daughter languages, and the daughter languages borrowing a word from a group with far-roving traders.
It took me a while which word based on "hamster" you had in mind, but yes, indeed: chomikować :D
ReplyDeleteSeems like that single Amerind language family theory could be tied to the old and maybe still prevailing belief that there was a single migration event that populated the Americas. Maybe as more support for multiple migrations into the Americas develops the language groupings will follow.
ReplyDeleteWe know of the three waves reflected in the language groups I mentioned above. The first, main one that peopled both continents; then the Athabaskan, Navajo one and the Inuit.
ReplyDeleteThough there is archaeological evidence for earlier settlers, there doesn't seem to be anything in language and only the faintest traces deep in the Amazon of earlier genetics. That group, which may be Pacific coast migrants who hewed to the shorelines all the way down, doesn't seem to have left much impact. It is possible, certainly, that there were multiple waves of peoples who died out. Settling new territory is hard.
The argument is over subsequent splits after the Beringians moved south around 15K years ago. they had moved into Beringia, which was largely savanna, about 23 kya. Writing slows language change and they didn't have any. Even with writing, look how hard it is to understand English from 1000 years ago, though we can identify looks of connections. So the languages changed over the millennia, and traditional linguists would shrug that they might all have come from a small group of related languages, but there is no way of telling now. They have a point. But the most common words, from the Swadesh list, show more stability and might provide stronger signal.