some kind of verb, and similarly with the four labels in (7b) that seem to say
my student days, to the effect that it was supposed to stand for 'main verb').
SYNTAX IN MID-20TH CENTURY AMERICAN LINGUISTICS 417
Fortunately, the cavalier attitude of generative grammarians towards syn-
tactic terminology did not last long. Important steps were taken in Chomsky
(1965) towards requiring that syntactic category labels have substance, no-
tably the separation of subcategorization features from categories in a nar-
rower sense, and the requirement that syntactic category names correspond to
categorial rather than relational notions, thus excluding the 'Comp' of (7a),
which meant "complement", but allowing the 'Comp' of more recent syntac-
tic work, which means "complementizer", a class of elements that mark com-
plements. Among generative grammarians whose work in the 1960s displays
concern for adopting'syntactic terminology that is related in a systematic way
to a coherent system of concepts, I would single out Paul M. Postal, John
Robert Ross, and Peter S. Rosenbaum.
The neglect of parts of speech is a major respect in which early generative
grammar deviated sharply from traditional grammar, notwithstanding numer-
ous statements by Chomsky to the effect that generative grammar in a sense
rehabilitated traditional grammar. One obvious reason for this lack of
concern for parts of speech is that in the work of earlier grammarians,
descriptivists as well as 'traditional grammarians', morphology had been a
major criterion for identifying parts of speech (in the case of Hockett 1958,
the sole criterion), whereas in early generative grammar, morphology was
assimilated to either syntax or phonology when it was recognized at all.
10
I do not mean to suggest that the notion 'part of speech' figures signifi-
cantly in more recent generative grammar. Tart of speech' seemed to have
taken a foothold in the version of 'X-bar syntax' practiced in the early 1970s,
in which the 'X' was generally a part of speech. However, subsequently, X-
bar syntax has developed in ways that commit it to rejection of the idea that
the part of speech of the head of a constituent is the X of its X-bar catego-
rization. In current orthodox X-bar syntactic theory, so-called functional
heads of phrasal constituents have proliferated, and the elements of the func-
tional categories are held not to belong to any 'lexical category'. Thus, 'func-
tional elements' such as determiners, complementizers, and negative ele-
ments cannot have the status of nouns, verbs, adjectives, or prepositions (let
alone 'adverbs', which the Chomskyan features ±N and ±V do not provide
for); accordingly, analyses in which determiners, or negative elements, or
complementizers differ from one another in part of speech are excluded.
1 0
Janda & Kathman (1992:153) have remarked that "morphology might almost be charac-
terized as the Poland of grammar [...] it does periodically shrink and occasionally even dis-
appear as the result of the partitions carried out by the surrounding forces of its powerful
neighbors".
418
JAMES D. McCAWLEYt
I close by observing that the authors that I have cited illustrate three very
different attitudes towards linguistic terminology, which do not correlate at
all with their linguistic ideologies. Some fear terminology as a source of er-
ror, most notably, Fries. Some regard terminology as an innocuous but tire-
some bother, as illustrated by Bloch & Trager, Bloomfield, and Chomsky.
And still others regard terminology as a valuable aid to insight, at least if
chosen properly, an attitude displayed by Gleason and by several authors
who were cited above but not discussed in detail, including Ross, Postal, and
my all-time greatest hero in linguistics, Otto Jespersen, whose terminological
innovations such as 'tag question', 'cleft sentence', 'extraposition' have
greatly facilitated the doing and the teaching of syntax. My attitude towards
terminology is, of course, that of Jespersen.
11
Dostları ilə paylaş: