conclusion, and thus it begins to dissociate linguistics from biology. There are two reasons why Chomsky might be
wary of an evolutionary argument (he never states them explicitly, so I am speculating here—see Newmeyer 1998a).
First, in Chomsky (1972b) he examines and rightly rejects many proposals for antecedents of human language in
animal communication: the gap between human languages and other natural systems of communication is vast.
Second, UniversalGrammar onhisconceptionissotightlyorganizedthatitsincrementaldevelopmentthroughnatural
selection looks on the face of it unlikely. As will be seen in Chapter 8, the conception to be worked out here is
somewhat more amenable to an evolutionary approach. Thus it should be possibleto keep evolutionin the argument,
and to treat Universal Grammar biologically just like any other cognitive specialization in the natural world.^45
4.9 Evidence outside linguistic structure for Universal Grammar/Language Acquisition Device
Over the years, a large number of phenomena have accumulated that are taken to provide evidence for a human
cognitive specialization for language acquisition. It is worth briefly enumerating them, with a few comments. Fuller
discussions of most of them appear in Pinker (1994b) and Jackendoff (1994).
4.9.1 Species-specicity
It is an ancient observation that only humans speak. This distinction is undeniable, even if we reject the traditional
conflation of this trait (e.g. by Descartes) with the possession of a soul, free will, and a moral capacity. We can also
accept the existence of numerous systems of communication in the animal world (Hauser 1996) without denying the
uniqueness of language.
The question is, to what is this species-specificity due? After all, there are
94 PSYCHOLOGICAL AND BIOLOGICAL FOUNDATIONS
(^45) One sometimes encounters proposals thatwhat evolvedwas nothumans butlanguage. For instance,Andy Clark (1996) proposes that“language is...an artifact, onewhich
has itselfevolvedso as to be easilyacquired by younghumans (perhaps exploiting processingbiases inherentin theyoung).”But whatcan it mean for language to evolveby
“itself”?Thenoisesare notsubjectto natural selection; only theorganis mis. That is, for language to evolvewe must suppose thatorganis ms evolvedthatwere equippedto
carry out these particular sorts of sound-to-meaning mappings. The“processing biases inherent in the young”likewise had to evolve so that children could learn to make
sense of these noises in the environment. Unlike artifacts such as bows and arrows, language was not designed by people; people evolvedso as to be able to have it, just as
bats evolved so as to be able to echolocate and elephants evolved so as to be able to pick things up with their noses. See Ch. 8 for further discussion.