Processing Verbs in German Idioms: Evidence Against the Configuration Hypothesis

Abstract

This study investigated whether verbs in figurative language activate different types of associations than do verbs in literal language. In a sentence-priming experiment, we compared sentences featuring verbs in idiomatic phrases with control sentences in which the same verbs were meant literally. Participants made lexical decisions about nouns that were associated with either the verb's literal meaning, with the figurative meaning of the phrase, or were unassociated nouns. Verbs in sentence-final position were highly predictable in both types of sentences and hence the phrasal meaning was rendered as figurative or literal before the sentence-final word. Our results showed that in literal sentences, the activation of literal associations was much stronger than that of figurative associations, whereas in idiomatic sentences, associations with the literal meaning of the verb were activated to the same degree as were associations with the figurative meaning. This contrasts with previous psycholinguistic research suggesting that literal associations should not be activated as soon as the phrase has been recognized as idiom. We argue that the literal verb meaning is essential in both figurative and nonfigurative language and present a model of idiom recognition that integrates our findings.

Notes

¹Exceptions, that is, the so-called “right-dislocation” are possible if the sentence is marked.

²The experiment was advertised through the “Language Experiments Portal” at URL http://www.language-experiments.org/.

³In order to examine whether the phrasal meaning of literal sentences is processed differently than the phrasal meaning of idiomatic sentences, literal and idiomatic sentences were compared by means of their baseline conditions (i.e. the unrelated conditions). The one-way ANOVA with repeated measures factor phrase (literal/idiomatic) was significant in the participants analysis, F1(1, 48) = 6.18, p = .0164, but not in the items analysis, F2(1, 56) = 2.04, p = .1588, indicating that responses for unrelated nouns were slightly faster following idiomatic sentences (625 ms) than after literal sentences (636 ms).

This effectwas further examined separately for unrelated nouns of verb associations and for those of figurative associations, since unrelated nouns matched to verb associations were of higher frequency and fewer letters than those matched to figurative associations. Regarding the baseline of verb associations, the effectwas significant in the participants analysis, F1(1, 48)=6.59,p=.0135, but not in the items analysis, F2(1, 56)=2.35,p=.1306, indicating that responses for unrelatednounswereslightly faster following idiomatic sentences (596 ms) than after literal sentences (611 ms).However, regarding the baseline of figurative associations, the effectwas not significant (F1 and F2 < 1), confirming that the phrasal meaning of literal sentences (661 ms) was processed equally fast as that of idiomatic sentences (654 ms).