As reported previously on Babel’s Dawn, the draft version of the Neanderthal genome was presented in Chicago last week (press release here) and it confirms the earlier finding that Neanderthal’s share the same FOXP2 gene found in humans. FOXP2 is the most important gene known to support language. Without it in its human form, rapid speech and the ability to learn speech sounds is limited. The gene’s presence in Neanderthals is very strong evidence that the human lineage was speaking fluently before the Neanderthals split from the line that became Homo sapiens. The original dating of the appearance of the FOXP2 gene in its human form put it between 200 and 100 thousand years ago. Many arguments about the recency of language have claimed authority based on that date, and now find their cards are very weak.
The FOXP2 gene provides a means to date the evolution of the human brain and the emergence of fully human speech capabilities. … Enard et al, (2002) [abstract here], using the techniques of molecular genetics, estimate that the human form appeared fairly recently, sometime in the last 100,000 years—in the time frame associated with the emergence of anatomically modern H. sapiens. …Natural selection acting on the mutations that yielded [FOXP2’s] human form would have enabled rapid, encoded speech, in turn enhancing the selective value of the mutations that shaped the modern human vocal tract. These events, which led to the emergence of fully modern speech, language, and cognition, appear to have occurred sometime in the period between 90,000 and 50,000 BP/ [p. 52]
The paper was published with comments from other scholars. Robin Melrose responded:
Lieberman has made a valuable contribution to the debate on language evolution by claiming that human speech in its present form cannot have developed until some 50,000 to 90,000 years ago with the emergence of the human form of the FOXP2 gene. [p. 55]
Tobias Riede was much more skeptical of Lieberman’s conclusion and went straight for a problem with his analysis of FOXP2 and its role in vocal learning, but he could not refute the date and was forced to nibble at the edges of Lieberman’s argument. Lieberman, in his rebuttal, stressed the date quite heavily.
If the very recent date cannot stand now, how far back does the human form of the FOXP2 gene really extend? In November 2006 this blog reported on a paper presented at a conference in Stellenbosch, South Africa claiming that the original dating effort on FOXP2 had been grossly in error and the true date of the human version of the gene was 1.8 to 1.9 million years ago. (See: FOXP2 May Be Older). The report concluded:
This blog is in no position to consider the technical argument made, but if the date stands up it will push the discussion back to the beginnings of Homo rather than at its end, where many people put it today.
I have emailed the paper’s main author, Karl Diller, to ask for an update on his work, but have not yet had a response. However, all dates on this gene are likely to be taken with several grains of salt without multiple, independent confirmations. We have three scenarios:
- FOXP2 is Recent: Language evolves without benefit of capacity for fine control of output or ability to learn to make cultural sounds. That capacity comes late and is followed by a very quick evolution of the physical ability to master rapid speech. Neanderthals do not have this trait and do not evolve the more rapid speech. This scenario, long favored by many, can no longer stand, now that the Neanderthal is known to have had the same gene. An alternate tack would be to say that FOXP2 is old, but its language effect is recent. The claim cannot be proven, but perhaps more appealing to those who must fall back on it is the fact that it cannot be disproven either.
- FOXP2 began with Homo ergaster/eructus: Lineage gets early capacity for fine control of output and the ability to learn cultural sounds. The rest of the speech characteristics then evolve over a long period of time. By the time of the Neanderthal split most language traits have evolved and the two lines have similar language capacities.This scenario is supported by the work done by Diller and Cann.
- FOXP2 began shortly before the Neanderthal/Sapiens Lines split: Lineage develops some basic language traits, then develops the ability to learn and make rapid speech. The lineage then splits, presumably two different sets of language traits then evolve. This scenario is the most conservative interpretation of the fact that Neanderthals and humans share the same FOXP2 gene.
I have my own opinion about which line makes the best sense, but I don’t feel like expressing it today. A week like this tells us we are making slow progress, real progress, in knowing something of our ancestry. We had three basic scenarios, now we have whittled it down to two. As the police like to say: we’ve ruled out a suspect.