Another week of reading McWhorter's The Power of Babel. The most fascinating part I read this week was about pidgins and creoles. He tells the stories of a large number of these, each one interesting on its own, and each adding to a picture of the general pattern that lead to their development. I've always been interested in creoles, and this just adds to that sense of how rich language is, regardless of the actual language.
(As a brief side note, since several of the creoles discussed originated in Suriname, I also spent some time learning more about the three little countries along the northern coast of South American where neither Spanish nor Portuguese is the colonial-era-inherited language. It's so easy to forget about them, thinking that all of South America is the rest of the countries. I've even seen in trivia-type books for kids that all countries in South America have Spanish or Portuguese as their national languages. The three Guianas have their own very different history, which never even merited a mention in my college history classes, and unique geography, including some of the world's tallest and largest waterfalls. So it's too bad they get ignored.)
I also mentioned last week that the epilogue of McWhertor's book addresses the question of the original Ur-language, from which all others are descended. First, he affirms that there likely was one language. I've seen some arguments that language could have developed at various times among different human populations, but he finds this unlikely. Also unlikely to him, however, is that we could ever know anything about the actual words of that language. The sheer time scale means that even with the most conservative language change imaginable, every sound will have changed into something unrecognizable. Where there do seem to be similar words across proto-languages (the prominent proponents of the idea are Greenberg, Ruhlen, and Bengtson), the evidence is pretty suspect--often involving too small samples, questionable recreations of even those proto-languages, and a stretch to find words only tangentially related by meaning.
Alas--it would be cool to believe that tik was the original word for finger.
What McWhertor does argue, though, is that creoles can offer us a glimpse into the underlying structure of the original language. For example, even when the dominant languages that gave rise to a pidgin, and from there to a creole, have a subject-object-verb order, the creolized language almost always has a subject-verb-object order. It probably also was not a tone-based language, and probably did not have elaborate inflections. And you could probably change word class (ie you could verb nouns) without having to change the word (though even then you might have run afoul of the language police...). I've seen elsewhere an argument that clicks are also likely a more recent evolution. I'm not seeing that addressed in this epilogue as I skim over it again, but I seem to remember something he said suggesting the opposite, that clicks might well have been part of the original sounds of language.
We'll never know with any certainty, but it's certainly fun to imagine how that first language might have been.