In search of a syntactic relationship between rhythm and language

Abstract

Linguistic and musical syntax have been shown to rely on shared resources (Patel, 2003). In this study, we investigate the effect of an accompanying rhythm, another hierarchical domain, on sentence processing accuracy and speed. Two experiments were conducted to inspect the possible existence of shared processing of linguistic and rhythmic syntax. In the first experiment, sentences were manipulated to yield two levels of complexity: subject and object relative clauses. Rhythm was manipulated by using structured rhythms: sL, ssL, sssL, and sLL pulse sequences where “s” stands for a shorter inter-stimulus interval and L for a longer one. The s:L duration ratio was 1:2 (150ms:300ms). An isochronous (sssss..) control condition was also included. In the critical condition, the upbeat (s) that is heard right at the relative clause was presented 50 ms earlier. The dependent measure was participants’ correct response rates and response times on the subsequent comprehension questions. The second experiment was identical to the first experiment with the exception that distorted rhythms were replaced with random beats to eliminate any structure in that condition. It was hypothesized that the rhythmic deviation in Experiment 1 would create competition of syntactical processing resources and hence decrease linguistic comprehension performance. In Experiment 2, we expected the random but not the structured rhythm condition to debilitate semantic processing of the sentences due to a complete "disconnectedness" between rhythmic and linguistic syntax. All the findings were null. There was no main effect of sentence and rhythm and interaction between the two in either of the experiments. This may indicate that rhythm and harmony rely on a different mechanism.