engleski

Syntactic processing of linguistic and musical stimuli: Is there a cognitive overlap?

Language and music, among other features, are believed to share similarities in structural processing of linguistic and musical stimuli. To test whether this holds true, we designed a word-by-word self-paced reading task, where linguistic and musical stimuli were presented simultaneously. By simultaneous presentation of the stimuli, we aimed to examine if their process of structural integration relied on the same resource networks, as stated by the Shared Syntactic Integration Resource Hypothesis (Patel, 2003). Each segment of a single sentence, that is, each word, was accompanied with a different chord of a single harmonic progression. While linguistic stimuli consisted of sentences that contained either subject-extracted or object-extracted relative clauses, musical stimuli were varied in terms of which key was a single chord played in. In this way, a chord was either ''in'' or ''out'' of key with a key dominant for that particular harmonic progression. Presentation of each of the sentence was followed by a comprehension question requiring a ''yes'' or ''no'' answer. If there is a shared dependency when it comes to structural integration of linguistic and musical stimuli as proposed by Patel (2003), then any element of a sentence that makes it harder to create a hierarchical structure, in this case, an object-extracted clause, should be particularly noticeable when paired with a harmoniously unexpected chord. This should be observable in longer reading times and more comprehension errors. In a 2x2 experiment, one critical item consisted of 4 stimuli, one for each condition. There were 4 times as many critical items as conditions, which means a total of 16 critical items. We tracked participant's (N=77) response times and comprehension accuracy. Results indicate that participants had faster reading times when presented with ''in-key'' chords than they did when presented with ''out-of-key'' chords (F(1, 76) = 9.55, p < .01, ηp2 = 0.11). They were also more accurate in comprehension of sentences containing subject-extracted relative clauses than they were in comprehension of sentences containing object-extracted relative clauses (F(1, 76) = 9.36, p < .01, ηp2 = 0.11), a phenomena well-known in the literature (Gibson, 2000). Also, we have not observed longer reading times and more comprehension errors when object-extracted clauses were paired with harmoniously unexpected chords. These results indicate that structural integration of linguistic and musical stimuli does not necessarily rely on the same resource networks as stated by Patel (2003).

linguistic syntax ; musical syntax ; syntactic processing ; structural integration

nije evidentirano