Yes, the EHL linguists are correct in confining themselves to basic vocabulary (pronouns, numbers, function words, body-part names, and so on and so forth), however their methodology is not strict enough and can hardly be called scientific.

For instance, Greek and Hawaiian are not related. A thorough analysis of their grammar will make this quite clear. However, they do (by shear chance) share a number of common forms: ‘meli’ means honey in both languages and ‘aeto’ and ‘aetos’ each mean ‘eagle’ in Hawaiian and Greek, respectively. There is a great and impressive list of comparable and exact forms like those above and, by EHL’s methods (often called mass comparison, but renamed as “multilateral comparison” by Joseph Greenberg), we might suspect a genetic link between the two.

What we need to stress, rather than [possibly and often likely chance] resemblances, are systematic correspondences. For instance, examine these correspondences between English, Latin, and Greek:

father, pater, pater.
foot, ped-, pod-.
for, pro, para.

None of the English forms share a great resemblance to either Latin or Greek, but a larger study of both vocabulary and other Germanic languages reveals what is known as the First Germanic Consonant Shift.

Among other things, Proto-Indo-European [p] underwent lenition and became Proto-Germanic [f] (probably first via a transition to an aspirated unvoiced bilabial plosive [pʰ] and then to the fricative, but this is in question).

There is a lack of resemblance, but there is a systematic, statistically proven correspondence. Via this method, commonly known as the comparative method, we can reconstruct languages back 6,000-8,000 years, but beyond that, any semblance of genetic relation becomes clouded by the background noise of chance resemblances.

To attempt to identify a single linguistic ancestor of all the world’s languages and with that, to ignorantly assume that all of these languages descend from only one proto-language is at best unscientific and is at worst [the more likely candidate] brash foolishness.

Your thoughts on micro languages are dead on. As it is in Africa right now there are thousands of regional languages and dialects which ultimately spawned off others. It could be very probable that this was the case hundreds of thousands of years ago as well. But one thing the EHL drives home is that throughout human prehistory there have been population bottlenecks. During these times, populations dwindled in size and languages theoretically will become more consolidated as well as other cultural entities, such as technology, art, music, etc.

What the EHL is doing is trying to find a point where languages were more similar than different. Seems very difficult, even impossible… but again they are trying. I appreciate the effort, furthermore, I appreciate their website and the data they have made public. It’s pretty interesting, well organized, and sets an example for other fields of academics to make their data available. That is very commendable and shows their willingness to be open.

What I want to ask and open this thread up to, is how can we study languages in any other method than how the EHL is doing it? We’ve addressed some of the pitfalls… so are there other ways of studying language evolution? Is it possible? Any ideas?

But I don’t agree that language evolved with EMH at 60k bp – or that there was one original language that splintered into many as a result of OoA – I think people were speaking for hundreds of thousands of years before that, and that there were many micro languages spoken regionally – transmission of a single language between many isolated groups of archaic humans would have been nigh impossible, but nothing to stop regions having their own languages and dialects etc.

I also think that language evolved more than once in different parts of the world, such as Asia as well as Africa, back in the Lower Palaeolithic – anyway thanks for the link, something definitely worth looking into in greater detail, imo.

Tim