The article entitled “Mysterious Indo-European homeland may have been in the steppes of Ukraine and Russia” (http://news.sciencemag.org/archaeology/2015/02/mysterious-indo-european-homeland-may-have-been-steppes-ukraine-and-russia) reveals the results of as yet unpublished research which circulates on the web.
My more than thirty-years long experience with Romance linguistics leads me to believe that the study shows interesting (wrong?) presuppositions about the connections between DNA and language as they relate to Indo-Europeans.
The following three points vitiate the interpretation of the scientists’ data:
1) This statement “the genetic makeup of today’s Europeans is more complicated than anyone expected” underlines the original shaky ground for equating DNA and language(s). Sciences based on biology have as yet to make a strong connection to culture. Linguistically speaking, language transmission and change are as complicated – if not more – as the DNA transmission and change.
2) Any introduction to Romance linguistics shows that this statement is unscientific: “linguists cannot be sure if the Latin attested to in written documents really was the direct ancestor of later Romance languages”. In fact, linguists are sure that written Latin is not the direct ancestor of Romance languages. Their statement also assumes that there is one linguistic ancestor for all Romance languages: this assumption is theoretically correct, but assumes geographical, cultural, historical, social uniformity where there was none.
3) Why is the original homeland of the Indo-Europeans so important? Is there even such a location as the “original homeland”? It is also interesting that the article talks about a Ukrainian-Russian area (north of the Black Sea) which these days receives military attention. Needless to say, migrations may be culturally predictable, but their impact on the DNA is yet to be determined.
Of course, the researchers behind this study conclude that “the ultimate question of the Proto-Indo-European homeland is unresolved by our data” and that more ancient DNA may finally tie our linguistic history with our genes. It is unclear, though, how exactly this may be done.