Generative Semantics
Generative Semantics
Generative Semantics
distinctions
of
logical
category,
supplemented by lexical exception
features (e.g., verbs and adjectives
would both belong to the category
predicate, usually confusingly called
V by GSists, with adjectives differing
from verbs in bearing a feature
licensing
the
application
of
a
transformation that inserts a copula),
while other generative grammarians
took syntactic categories to have at
most a tangential relation to semantic
categories.
d. The linguistic level or levels relevant
to the choice
of the lexical material of a sentence.
One who
holds that there is no level of syntactic
deep structure
as distinct from semantic structure is
forced to
recognize syntactic structures whose
ultimate
units
are
semantic
rather
than
morphological in
nature, such as a syntactic structure
[Brutus DO
SOMETHINGx (X CAUSE (BECOME (NOT
(Caesar ALIVE))))] underlying Brutus
killed
Caesar. (Here and below, capitalization
is used as
an informal way of representing
semantic units
corresponding roughly to the words in
question.)
GS-ists
accordingly
proposed
transformations that
combined
semantic
units
into
complexes that
could potentially underlie lexical items,
e.g., predicateraising (proposed in McCawley, 1968)
adjoined a predicate to the immediately
superordinate
predicate, thus allowing the derivation
of
such complexes as NOT-ALIVE, BECOMENOTALIVE
(die), BECOME-NOT (cease), and
CAUSE-BECOME-NOT-ALIVE or CAUSEdie
(kill). Intermediate derivational stages
involving
GS Policies
Research
on
the
Conduct
of
system.
c. Adoption of a static conception of
linguistic rules:
rules were thought of not in terms of
the popular
metaphor of assembling linguistic
structures
and converting structures on one level
into
corresponding structures on another,
but as derivational
constraints, that is, as specifications of
what a structure may or may not
contain and
of how a structure on one level may not
contain
and of how a structure on one level may
or must
differ from the corresponding structure
on another
level. This difference in the conception
of rules
resulted in difference with regard to
what theoretical
notions posed conceptual problems
(Laudan,
1976) for each approach; thus GS-ists
readily accepted
rules that specified relations among
nonadjacent
levels of structure (what Lakoff, 1970b
dubbed global rules), a notion that was
unproblematic
from their conceptual vantage point but
outlandish from the vantage point of
the operation
metaphor for rules, while rejecting the
idea of
ordering among rules, a notion that was
unproblematic
for those who accepted the operation
metaphor
but was difficult to make coherent with
the
GS conception of rules as derivational
constraints.
d. Disdain for those concerns of
Chomskyan generative
grammarians that had little connection
with
linguistic facts or with detailed linguistic
description,
such as mathematical models and
speculation
The History of GS
Bar Theory.
Bibliography
Bach E & Harms R T (eds.) (1968).
Universals in linguistic
theory. New York: Holt, Rinehart and
Winston.
Binnick R I (1971). Come and bring. Lin 2,
260267.
Brentari D, Larson G N&MacLeod L A (eds.)
(1992). The
joy of grammar: a festschrift in honor of
James D.
McCawley. Philadelphia: Benjamins.
Chomsky N A (1965). Aspects of the theory
of syntax.
Cambridge, MA: MIT Press.
Chomsky N A (1972). Studies on semantics
in generative
grammar. The Hague: Mouton.
Fodor J A (1970). Three reasons for not
deriving kill from
cause to die. Lin 1, 429438.
Grice H P (1967). Logic and conversation.
In Grice H P
(ed.) (1989) Studies in the way of words.
Cambridge,
MA: Harvard University Press.
Gruber J S (1965). Studies in lexical
relations. In Gruber
J S (ed.) (1976) Lexical structures in syntax
and
semantics. North-Holland: Amsterdam.
Harris R (1990). The generative heresy. Troy,
NY: Rensselaer
Polytechnic Institute, Unpublished Ph.D.
diss.