Figure 1 Schematic representation of the theory by Lerdahl 1989.

Many of these components are psychologically plausible. First, we know that grouping is a major characteristic of perception (Fraisse 1974) and the experiments carried out by Deliège (1987) confirm the psychological validity of different grouping rules involved in the model. Second, several experiments have shown the listener’s ability to abstract metrical structure (Povel 1981, Essens & Povel 1985, Sloboda & Parker 1985), and many others have pointed out a very sophisticated implicit knowledge of tonal hierarchies by the listener (Krumhansl 1979, Krumhansl & Kessler 1982, Bharucha & Krumhansl 1983). Other data indicates that the listener manages to organise musical events in a hierarchical way (Deutsch 1980, Stoffer 1985, Serafine 1989). At the very least, the possibility of extracting a link between different variations and a theme, suggests that a level of coding exists where the musical phrase is represented by its underlying network of tension and relaxation (Bigand 1990a, 1990b, 1990c).