Language and genes: Say what?

This article talks about efforts to correlate language and genes in Africa. The project is pretty familiar and the language-gene connection has plenty of controversial angles, but something's gone horribly wrong in the press coverage here:

Christopher Ehret, a noted specialist in African historical linguistics at UCLA and a member of Tishkoff's team, said his analysis of tribal languages revealed striking patterns of migration across Africa.

"When people move, they borrow words from the people where they settle," he said. Those new words inserted into older languages, he said, can tell us when the newcomers arrived.

For example, Ehret said, the "click" language still spoken among people as varied as the San of South Africa, the Pygmy tribes of Central and West Africa and the Hadze people far to the east may well be the original spoken language of all humans - and the genes of those distant click speakers indicate they share a common ancestry, the scientists noted.

There isn't a click language, or even a single family of click languages, but a number of distinct families. For example, I think the genetic affiliation of Hadze is regarded as uncertain. The deal of making it sound like the original human language is still being spoken out there somewhere is several steps over the edge.

Who knows what's going on here, but you gotta figure Ehret was grossly misquoted.