3.5 Four challenges for cognitive neuroscience^27
Just as linguistic structures like Fig. 1.1 are functional characterizations that require neural instantiation, so the
functional regularities that we state as rules of grammar must be neurally instantiated. To repeat a point from Chapter
2: although a great deal is known about functional localization of various aspects of language in the brain, I think it is
fair to say that nothing at all is known about how neurons instantiate the details of rules of grammar. In fact, we don't
even haveany idea ofhowa singlespeechsound such as /p/—muchless a category likeNP—is instantiated in neural
firings or synaptic connections. The rest of this chapter will lay out four challenges that linguistic combinatoriality and
rules of language present to theories of brain function—challenges that to my knowledge have not been widely
recognized in the cognitive neuroscience community.
3.5.1 The massiveness of the binding problem
Consider what happens in the perception or production of our familiar sentence (23).
(23) The little star's beside a big star.Much of the neuroscienceof language has been concerned with how words stored in long-ter m me mory are activated
(“light up”) in the course of sentence perception and production (e.g. Caramazza and Miozzo 1997; Pulvermüller
- But activation of words alone is not sufficient to account for the understanding of sentences. If understanding
(23) consisted only of activating the words, the sentence in (24a), not to mention the complete nonsense in (24b),
would“light up”the same words and hence be understood the same.
(24) a.The big star's beside a little star,
b. Beside a the big little star star's.Clearly a sentence is more than a collection of words: the word meanings are structured into the meaning of the
sentence by means of semantic relations among them. These semantic relations are to some degree signaled by the
syntactic structure of the sentence, which in turn is correlated with the linear order of the phonological words.
Thus all of the structure modeled in Fig. 1.1 must be functionally present—at once—in order for this sentence to be
grasped. Introspectively, the whole seems to be maintained at least for some brief period after the completion of the
sentence.
58 PSYCHOLOGICAL AND BIOLOGICAL FOUNDATIONS
(^27) This section is based on unpublished work done in collaboration with Maja Matarić. It also draws a great deal on the work of Gary Marcus (1998; 2001).