participate in the conversation, if only to maintain a position in this intellectual niche that is of such commanding
interest to the larger scientific public. Some linguists have indeed risen to the challenge, as will be seen below. The
present chapter too is offered in this spirit: I a mnot sure how seriously I want to take it, but as long as there is a
debate, it is worth taking part.
Beyond the sociological issues, proposals about language evolution face two major difficulties. One is a question of
data. There is no direct evidence for early forms of language until the advent of writing about 5,000 years ago, and by
thenwe are dealingwith fullymodern language. Languages may change and“evolve”in the sense ofculturalevolution,
but as far as can be determined, this is in the context of a fullybiologicallyevolved language capacity. For the priorfive
million years, we can make only very indirectinferences based on thenature ofartifactssuch as toolsand pictures, and
on equivocal hints about the structure of the brain and the vocal tract.
Indeed, the latter have over time proven less telling than originally thought. For instance, one of the early pieces of
evidence(Lieberman 1984) concernedthefactthat theNeanderthallarynx, likethatof apes, is situated muchhigherin
the vocal tract than that of modern humans, a position not conducive to producing the modern human variety of
speech sounds. (Darwin pointed out this difference between humans and other primates, but Lieberman actually
worked out the acoustics.) On the other hand, Fitch (2000) shows that, although the larynx of monkeys and goats is
positioned much like that of Neanderthals, it descends substantially during the animal's vocalizations—to something
much closer to the modern human position. There is no reason not to assume the same was true of Neanderthals, in
which case certain aspects of Neanderthal acoustics would have more closely approached the modern standard than
Lieberman claimed.
Fitch also reviews fossil evidence fro mbrain endocasts (which can reveal he mispheric differences), fro mfossil hyoid
bones (the attachment point for many vocal tract muscles), and from the size of the canal in the base of the skull for
the hypoglossal nerve that controls the tongue. Though each of these has been offered as evidence for or against
speech in hominids, Fitch concludes that recent results have rendered all this evidence rather equivocal. And even
though Lieberman's arguments about the more limited acoustic possibilities of the upper Neanderthal vocal tract
(tongue position and so forth) have not been challenged, this still tells us little about whether Neanderthalsspoke, and,
more important, about what they had to say. In short, there is virtually nothing in the paleontological record that can
yield strong evidence about when and in which stages the language capacity evolved.
Moreover, although there are numerous systems of animal communication,