FFF CONFERENCE CTF07

Britta Sauereisen - The Interpretation of Adjective-Noun-Combinations on the Basis of a Cell Assembly Theory

The meaning of adjective-noun combinations is derived compositionally. In this process, the adjective can affect the meaning of the noun (1), and vice versa (2).

 

(1)    der (alte) Kanzler
            the (old, i. e. former) chancellor
(2)    der rote Ferrari – die rote Zwiebel
            the red Ferrari – the red onion


This fact can easily be accounted for with a cell assembly theory (Hebb 1949, Pulvermüller 1996) if one assumes that a cell assembly contains all polysemous meanings of a given lexical item. In the compositional process of adjective and noun, the parts that fit together in the cell assembly are enforced while the ‘non-fitting’ parts are inhibited. The same principle holds for adjective-noun compounds such as der Altkanzler (which has the interpretation given in (1)) which are not adequately accounted for with a generative approach (Chomsky 1995, a. o.). However, a problem shared by both generative and cognitive theory is the restriction on the readings available in the predicative use of adjectives (3,4):


(3)    der alte Freund
         the old friend ‘of long duration’ or ‘of old age’
(4)    der Freund ist alt
         the friend is old ‘of old age’


While (3) has both the reading of old in the sense of ‘of long duration’ and ‘of old age’, (4) only has the ‘old age’ reading. This entails that the interpretation of a single lexeme is crucially affected by the syntactic configuration in which it occurs, which is unexpected in a strictly cognitive-associative approach.


(5)    die Freundschaft ist alt
         the friendship is old ‘of long duration’


However, the conclusion that the use of different syntactic configurations corresponds to different lexemes, thus potentially accounting for (3) and (4), is not legitimate, since (5) shows that adjectives in their predicative meaning are not obligatorily restricted to a single reading. To account for these data, I assume that a syntax module exists that permits free association processes. These processes take place within a syntactic frame (template, e. g. NP). At the boundaries of these templates, however, the corresponding cell assembly values are fixed (here: friend = person), which can only be overruled with great cognitive effort. This explains why (4) and (5) have only one reading left, the one that is a property of the referent in the NP (person in (4), and friendship in (5)). I therefore conclude that there has to be an independent syntactic module besides the cognitive, memory-based interpretation system.

 

References

Chomsky (1995). The Minimalist Program.

 

Hebb (1949). The organization of behavior: A neuropsychological theory.

 

Kamp (1975). Two theories about adjectives.

 

Larson (1999). Semantics of adjectival modification.

 

Partee (1994). Lexical semantics andcompositionality.

 

Pulvermüller (1996). Neurobiologie der Sprache.