Page 360 - DENG504_LINGUISTICS

Linguistics

Notes are then mapped (via the transformational subcomponent) into surface structures. The meaning of
each sentence is derived from its deep structure by means of the semantic rules of interpetation. The
phonetic realization of each sentence is derived from its surface structure by means of the phonological
rules. The phonological and semantic components are, as stated earlier, ‘interpretive’ because they
interpret the output of the base component.
Important changes in the theory are apparent ‘in the representation of lexical categories and the
relationship between the lexicon and the rest of the grammar, the introduction of separate phonological
and semantic components, and the reformation of transformational component of the grammar. As a
result of these formations applying to the concept of kernel sentences, and of a set of optional
transformations applying to the underlying forms of kernel sentences to produce more complex
sentence, was abandoned.’
The base component of a Generative grammar consists of PS rules + lexicon. It operates on the basis
of three types of rules—categorization, formative and category symbols. Both formatives and category
symbols are vocabulary symbols. For instance, to generate a sentence—*Honesty may admire the
man—formatives will be the, man, etc. category symbols will be S, NP, V, etc. The formative,
furthermore, can be divided into lexical items (honesty, man) and grammatical items (Perfect,
Progressive, etc., except possibly for the, none of these are represented in the simplified example
given).
Among the rules of the categories component of an Aspect-type grammar, we might find the following
(See Chomsky, 1965 : 85) :
1. NP (Det) N (S)
2. N—[+ N + common]
3. [+ Common]—[± Count]
4. [± Common]—[± Animate]
5. [+ Animate]—[± Human]
6. [+ Count]—][± Abstract]
Rule (1) is a phrase-structure rule of the kind we have discussed in the previous sections : it rewrites
the category that appears to the left of the arrow as the string of (one or more) categories that appear
to the right of the arrow. Such rules are called branching rules in Chomsky (1965). Rules (2)-(6) are
subcategorization rules; and their function is to develop the category N into a set of features. The
sub-categorization rules given here can be interpreted as follows : every member of the category
‘Noun’ has the property (or feature) of being either ‘common’ or ‘non-common’ (‘plus Common’ or
‘minus common’); all categories that have the property ‘plus common’ must be either ‘plus Count’
(i.e. countable) or ‘minus Count’; and so on. The set of features that results from the application of the
subcategorization rules is called a ‘complex symbol’ (abbreviated elsewhere in Chomsky, 1965, as
CS). One such complex symbol might be [+ N,—Count, + Abstract]; another be night [+ N, + Count,
+ Animate, + Human].
Given the formalization of the syntactic properties relevant to the subclassification of nouns, we can
organize the lexicon appropriately, with entries of the following form :
1. honesty : [+ N, -Count, + Abstract]
2. man : [+ N,+ Count : + Common, + Animate, + Human].
These entries may be read as : the lexical item honesty is an uncountable, abstract noun and the
lexical item man is a countable, common, animate, human noun. Now, assuming for the sake of the
argument that the deviance of *Honesty may admire the man is correctly accounted for by saying
that the verb admire occurs only with human nouns in the subject position, we can fomulate a selection
rule to this effect in term of the feature [ + [ + Human], Aux.....]. So Chomsky treated selectional
restriction as a matter for syntax, rather than semantics. He remarks, “it should not be taken for
granted, necessarily, that syntactic and semantic considerations can be sharply distinguished.”
Alternately, the sub-categorization rules may be eliminated from the system of rewriting rules entirely
and be assigned to the lexicon.

354 LOVELY PROFESSIONAL UNIVERSITY