162 CHAPTER 5
Syntactic Structures(1957). In this book, Chomsky argued that phrase-struc-
ture grammar was inadequate, and he proposed an alternative that proved to be
so powerful that it revolutionized linguistics.
Many books have explored the significance ofSyntactic Structuresand re-
lated works, and what follows must, necessarily, be just a short summary.
Syntactic Structuresargued that phrase-structure grammar could never be a
viable intellectual enterprise, offered a new grammar to replace it, reasserted
rationalism as thesine qua nonof linguistics, and established language study
firmly as a branch of psychology. It vitalized the emerging field of cognitive
psychology, gave birth to a new area of language study called psycho-
linguistics, influenced philosophers working in the philosophy of language,
and gave English teachers a new tool for helping students become better writ-
ers. Given such intellectual influence, Chomsky has been, with good reason,
widely hailed as one of the more important thinkers of the 20th century.
Among modern intellectuals cited by other writers, Chomsky ranks eigth
(Harris, 1993, p. 79).
That Chomsky found phrase-structure grammar lacking is an understate-
ment. But why? Chomsky had several criticisms of phrase-structure grammar;
perhaps the two most important involved describing and explaining language.
Phrase-structure grammar focused onlanguagesrather thanlanguage.
Structuralists studied a given language in order to record as many features of it
as possible, building a corpus, or body, of utterances that formed the foundation
of the grammar for that language. These utterances were sentences and expres-
sions that native speakers actually used, what are calledattested utterances.
The corpus was made up only of attested utterances, and the grammar was con-
structed so that it described them.
Chomsky argued that this whole approach was misguided. Basing grammar
on attested utterances cannot lead, he claimed, to an adequate description of a
given language for the simple reason that it is based on a finite set of utter-
ances/sentences, whereas any language is potentially infinite. From this per-
spective, no matter how large the corpus, it never can constitute a significant
portion of the language.
A related problem is that the resulting grammar may describe the corpus, but it
does not describe all the grammatical sentences of the language. It fails to account
for the fact that language is inherently creative, with few sentences ever being re-
peated exactly from one situation to the next. That is, phrase-structure grammar
can describe just attested utterances; it cannot describe the infinite number of
grammatical sentences that may have been uttered before the corpus was com-
piled, that have yet to be uttered, or that never will be uttered but are potential utter-
ances. Even though phrase-structure rules such as those we developed in the