Wednesday, May 8, 2013

Study Purporting To Link Seven Language Families At 15 kya Time Depth Based On Iffy Data

On the web site of the Proceedings of the National Academy of Sciences, in the "Early Edition" section, is an article by Mark Pagel, Quentin D. Atkinson, Andreea S. Calude, and Andrew Meade: "Ultraconserved words point to deep language ancestry across Eurasia". The authors claim that a set of 23 especially frequent words can be used to establish genetic relationships of languages that go way, way back — too far back for successful application of the standard historical linguistics methodology for establishing language families, the Comparative Method. The idea is that, once you've determined that these 23 words are super-stable (because they're used so often), you don't need systematic sound/meaning correspondences at all; finding resemblances among these words across several language families is enough to prove that the languages are related, descended with modification from a single parent language (a.k.a. proto-language).

From Language Log.

As the trenchant analysis in the linked post explains, the study is deeply flawed because it is based on dubious choices of cognates in the seven proto-languages which were compared. In many cases, there are several cognates for the allegedly ultraconservated word and there is no consensus on which is correct in the proto-language, but the authors simply choose one by whim.

The data also cast serious doubt on the hypothesis that these words are truly ultra-conserved. For example, only a quarter of the ultraconserved Indo-European cognates survive in English, despite a time depth from proto-Indo-European to English estimated by more reliable means on the order of 6,000 years. The survival rates of these "ultraconserved" words are even lower in some of the other language families examined.

The hypothesis that the seven families discussed in the paper are part of a single macro-language family is not widely accepted among linguists, and isn't a good fit to genetic data for the populations who speak them either, as explained in comments to this post at Dienekes' Anthropology blog.

There is lots of very good research in the area of historical linguistics, but this, like a number of other papers with Quentin D. Atkinson as one of the authors, isn't one of them.

1 comment:

Maju said...

Thanks for the critical review, Andrew, I was totally puzzled by the conclusions myself, so I'm glad that some linguists have already made the work of criticizing it, with such a negative result.