| Summary: | 碩士 === 國立中正大學 === 語言學研究所 === 93 === Abstract
This thesis explores the formal parallels between language, specifically prosody, and music, in support of Borchgrevink (1982), Lerdahl and Jackendoff (1982;1983), Jackendoff (1989), Pinker (1997) and Sundberg and Lindblom (1991). We apply Optimality Theory (OT) to music, specifically Chinese folk songs, to show more specific resemblances between constraints for music and prosody.
We first introduce Lerdahl and Jackendoff’s (1983) Generative Theory of Tonal Music (GTTM), which helps formalize parallels between language and music, and use it to analyze Chinese folk songs. While doing the analysis, we find that the rules that GTTM proposes are vague and should be ranked, suggesting the possibility of a constraint-based approach, i.e. Optimality Theory.
Then we apply OT to Chinese folk songs. We argue in favor of the position of Gilbers and Schreuder (2002) and Van der Werf and Hendriks (2004) that OT gives a better description of musical structure than GTTM does. Since OT has violable and ranked constraints, the above weaknesses of GTTM can be avoided. In addition, since the constraints we propose for music closely match constraints for prosody, our OT analysis reveals more specific resemblances between constraints for music and prosody than had previously been noted. We have found language and music both use alignment constraints to mark boundaries, have BINARITY and TERNARITY constraints for feet (“groups” in music), and WEIGHT-BY-POSITION and WEIGHT-TO-STRESS for stress.
Since Lerdahl and Jackendoff (1983) claim that as in language, there are universals across different musical genres, we compare Chinese folk songs with other musical genres, specifically Western classical music and waltzes. We have found that different musical genres use the same constraints and that some constraints even have the same ranking in different musical genres. Our comparison of Chinese folk songs to waltz music also tells us that musical grammars can differ in their constraint rankings. Therefore we conclude that as in linguistic grammars, constraints for music are universal and grammars of musical genres only differ in rankings.
In order to see how the constraints we propose for music actually work and provide an empirical basis for those constraints and their ranking, we also have run a judgment experiment on musical grouping, which is adopted from Van der Werf and Hendriks (2004). With an intention to find similarities and differences across musical genres, we have two groups of participants with experience in two different musical genres, specifically Chinese classical music and Western classical music. The result of the experiment supports the reliability of our constraints and their ranking, and unexpectedly reveals the universality across different musical genres.
|
|---|
