CNL Lunch

Silvia Gennari

Representational complexity of verb meanings

Thursday October 11th, 12:30pm, 3416 Marie Mount Hall

This paper investigates the effect of verb meanings on lexical access and sentence processing. We show that stative verbs, which have simple meanings, (e.g. believe, loathe, exist, belong, cherish, deserve) are recognized and read faster than eventive verbs, which have complex meanings (e.g. discover, arrive, emerge, sell, eat, borrow). These findings follow from theories of verb meanings that argue for representational differences. States and events are the most general verb types. Stative verbs denote facts and have no internal temporal or causal structure. Their truth at a given interval i entails their truth at every instant within i (temporal homogeneity). Events, in contrast, lack this entailment, as they typically entail a change from an initial to a resulting sub-component state (die = being alive/being dead). They thus have complex internal causal structure and lack temporal homogeneity.

Most lexical semantic theories (cf. [1,2,3]) explain this distinction by proposing semantic representations that differ in complexity or internal composition. Such differences suggest that verb types could be processed differently. In particular, if the meaning complexity correlates with lexical access time [6] and with reading time during sentence comprehension [7], this predicts that stative verbs must be accessed or read faster than eventive ones. This prediction contrasts with atomic theories of meanings (cf. [4, 5]), where word meanings are taken as indivisible wholes, and therefore, predict no effect of internal complexity.

To test these hypotheses, we conducted two experiments. The first experiment was a visual lexical decision task (n=48; 64 words; 32 per category). Verbs were matched for frequency and word length. Imageability ratings were collected independently and used as a covariate, since the higher the imageability rating the faster the reaction time (RTs). Analysis of RTs (with verb-type/imageability as factors) revealed a main effect of verb types (p =.0001 across subjects and items) and no interaction with imageability.

The second experiment was a self-paced reading study (50 verbs per category contained in stative or eventive sentences respectively). Verb types were matched according to several criteria: (a) frequency and word length, (b) total number of syntactic frames and thematic roles, and (c) number of syntactic arguments and thematic roles for most frequent frames. An Analysis of reading times with sentence position and verb type as factors revealed a main effect of position (p=.0001), and a main effect of verb type (p= .01) but no interaction. This result indicates that processing an event sentence involves more processing cost than processing a state sentence, as predicted on the basis of semantic complexity. In addition, an effect of verb type also obtained at the 4th and 5th positions, i.e., the verb and the following article or preposition. The converging evidence from these studies supports the prediction that internal causal complexity of verb meanings (events) take longer to process than simpler ones (states). The results thus are consistent with lexical semantic theories that postulate differences in verb representational complexity.

References

[1] Dowty, D. 1979 Word Meaning and Montague Grammar, Reidel, Dordrecht.

[2] Krifka, M. 1989 Nominal reference, Temporal Constitution & Quantification in event semantics, in Bartsch et al., Semantics and Contextual Expressions, Foris.

[3] Verkuyl, H. 1989, Aspectual classes and aspectual composition. Linguistics and Philosophy, 12,39-94.

[4] Fodor, J. A., three reasons for not deriving "kill" from "cause to die". Linguistic Inquiry, 1.4, 429-438.

[5] Fodor, J.D., Fodor, J.A. & Garrett, M. 1975. The psychological unreality of semantic representations, Linguistic Inquiry, 6,4, 515-535

[6] Balota, D. 1994 Visual word recognition: The journey from features to meaning, in Gernsbacher, et al. Handbook of psycholinguistics, Acad. Press.

[7] Rainer, K. & Duffy, S. 1986 Lexical complexity and fixation times in reading: Effects of word frequency, verb complexity, and lexical ambiguity. Memory & Cognition, Vol 14(3), May.