NEWS RELEASE, 2/16/99
UC Berkeley graduate students poke
hole in highly touted theory that Neanderthals could
BERKELEY-- Three graduate students at the University of California, Berkeley, have thrown cold water on last year's highly publicized claim that Neanderthals could talk.
Close human relatives known for their protruding brow and squat, heavy bodies, Neanderthals apparently branched off from our early ancestors some 400,000 years ago and disappeared about 30,000 years ago.
The original claim that Neanderthals could use language was made by a team of Duke University researchers based on an analysis of the size of the hypoglossal canal, a hole on either side of the base of the skull that accommodates the nerves innervating the tongue and controlling its movement.
In their April 1998 report in the Proceedings of the National Academy of Sciences, the Duke researchers asserted a correlation between speech and the size of the hole, assuming that larger holes mean larger nerves and more complex tongue function. They then argued that the hypoglossal canal in Neanderthals and early Homo sapiens is close in size to those of modern humans, suggesting that they used language as early as 400,000 years ago.
The UC Berkeley students refute that assertion in an article in the Feb. 16, 1999, issue of the Proceedings of the National Academy of Sciences. (For photos, click here.)
"If this hypothesis were right, then humans should have a bigger hypoglossal canal than monkeys, since monkeys don't talk," argued David DeGusta, a graduate student in the Laboratory for Human Evolutionary Studies of the Department of Integrative Biology at UC Berkeley. "But a lot of monkeys and apes have hypoglossal canals within the size range of humans."
Correcting for mouth size, like the Duke researchers did, the UC Berkeley group found 15 non-human primate species whose average hypoglossal canal size is larger than the modern human average. For example, relative to mouth size, the average gibbon - an ape - has a hypoglossal canal twice as large as the average modern human.
"More than half of the monkeys we measured have hypoglossal canals that are in the modern human size range, both absolutely and relative to mouth size," he added. "So if you find a Neanderthal with a canal also in the modern human size range, did it have human-like vocal abilities or monkey-like vocal abilities? Based on the hypoglossal canal, you can't say."
After examining skulls from many species of non-human primates, modern humans and million-year-old human ancestors, the authors found no correlation between canal size and language ability, leading them to conclude that the size of the hypoglossal canal simply does not reflect vocal capability or language usage.
According to F. Clark Howell, professor of anthropology at UC Berkeley and the National Academy of Sciences member who forwarded the paper to the journal, "This paper is a disclaimer. It says we can't use their (the Duke researchers') criterion to suggest language capability. It doesn't work."
The students conclude in their article that "the date of origin for human language and the speech capabilities of Neanderthals remain open questions."
DeGusta was very intrigued by the initial study, authored by Richard F. Kay, Matt Cartmill and Michelle Balow of the Department of Biological Anthropology and Anatomy at Duke University Medical Center.
"It struck us as a really intriguing idea, a really new idea, that had the potential to give us insight into how speech developed," DeGusta said.
Upon looking at skulls in the laboratory, however, the correlation between canal size and language didn't seem to hold. So he and coauthors W. Henry Gilbert, also of UC Berkeley's Department of Integrative Biology, and Scott P. Turner of the campus's Department of Anthropology, spent last summer looking at a slew of modern primate skulls.
They measured the size of the hypoglossal canal in 75 individual skulls representing 32 different species of non-human primate, including Old and New World monkeys, prosimians (lemurs) and the apes - chimpanzees, gorillas, orangutans and gibbons. Most of the skulls came from the Museum of Vertebrate Zoology at UC Berkeley.
For comparison they also measured the hypoglossal canal size of 104 modern human skulls from different museum collections. The Duke researchers had studied 48 modern human skulls.
What the UC Berkeley researchers found was a large overlap between the sizes of the canals in humans and non-human primates. Even when the researchers corrected for the size of the tongue, as indicated by the size of the palate, the overlap remained.
"We found 40 different skulls of non-human primates that had hypoglossal canals both absolutely and relatively within the modern human size range," DeGusta said.
In addition, they found that the two canals in each skull are themselves often different sizes. In one modern human skull, the area of one hole was twice that of the other.
The authors also looked at five human cadavers and measured the size of the hypoglossal canal and the size of the nerve running through it, to see if a larger hole size means a larger nerve. They also counted the total number of axons or nerve fibers running through the nerve, to see if a correlation existed there as well.
But they found no correlation between canal size and nerve cross-sectional area or number of axons. Among the set of five skulls, one of the smallest nerves actually ran through the largest canal.
"We can only conclude that the functional basis for linking canal size with speech is questionable," DeGusta said.
Finally, they looked at the canal size in two ancient human ancestors, 3.2 million-year-old fossils of Australopithecus afarensis - the same species as the famous "Lucy" skeleton - and a 2.4 million-year-old specimen of Australopithecus boisei.
All four skulls examined had canals within the size range of modern humans. One
Australopithecus afarensis specimen had a canal that was bigger than 70 percent of the modern humans they measured.
"We think it's very unlikely that Australopithecus spoke, because they didn't make tools, they had a very small brain and they had no evidence of symbolic behavior," DeGusta said. "However, using the Duke criteria, you would have to say that Australopithecus had language too, and that language arose 3.2 million years ago, not 400,000 years ago."
The Duke researchers had looked at the canals of a 2.5 million-year-old relative, Australopithecus africanus, and found that, on average, they were smaller than the average canal size in modern humans. They then used this size difference to make a distinction between speaking and non-speaking species.
However, DeGusta noted that many modern humans have smaller canals than the A. africanus specimens. The actual size of the canal in an individual is most important, not the average size in a human population, he argued.
"An individual's mechanical ability to speak can only be influenced by their own canal size," DeGusta says. "Their nerves run through their canals, not through the average-sized canal."
"In any case," he added, "we don't have populations of fossil humans with hypoglossal canals, we only have scattered specimens that preserve the canal. You can't take the average of three individuals and assume that you have the average size of the whole species."
Altogether, the three lines of evidence assembled by the UC Berkeley researchers argue that "the size of the hypoglossal canal is not a reliable indicator of speech," the authors wrote.
"Any one of these objections makes the original hypothesis suspect," DeGusta said. "It's really hard to explain away all three."
"The original group had not done adequate anatomical
controls and checks, while these young researchers have
conducted an expanded study of the relevant anatomy, with an
enlarged sample and a larger number of controls, not to
mention their study of the anatomy of living human beings,"
Howell said. "They have convincingly shown that this line of
inference just doesn't pan out - there's too much overlap
with apes, and they don't speak.
Send comments to: email@example.com