Is it appropriate to speak of " musical language " ? In 1973, in a series of lectures at Harvard, composer Leonard Bernstein called on researchers to propose a musical grammar comparable to Noam Chomsky's " generative grammar ". In response, composer-musicologist Fred Lerdahl joined forces with linguist Ray Jackendoff to propose A Generative Theory of Tonal Music (1983). The theory postulates the existence of four structures, all hierarchical :
- Grouping: motifs, musical phrases, etc.
- Metric structure: pulsation and its multiples.
- Time-span reduction (TSR), which builds on the previous two levels to form a time tree with several levels of embedding.
- Prolongational reduction (PR), based on the increase or reduction of tonal tension.
These last two structures are supposed to be organized in binary trees and, in this sense, strongly resemble the structures postulated by linguists. Katz and Pesetsky go so far as to propose an extreme view: the lexicons of musical and linguistic structures differ, but the syntactic structures are the same.
