selections, and for the combinations of music and film selections. Figure 6 Schematic representation of the simple associative model of meaning of film which integrates affective and denotative meanings from both musical and visual (scene) information from the film.

It is suggested that the principles demonstrated in the simplest experiment (the bouncing ball) are applicable to the more complex experiments which also showed effects of music on the interpretation of the visual information. Thus, all of these examples are consistent with a model that assumes that both music and film generate representations of specific denotation and affect (see Figure 6). These may be weighted according to their salience and clarity and then combined for the final outcome meaning. The associationist framework can help to provide the empirical data needed to fine tune this theory. The data, however, may or may not be most parsimoniously accounted for in terms of associationism. Related and more recent cognitive models may prove to be better candidates, but at present we do not have the appropriate data to judge this. This paper, therefore, does not argue that musical phenomena ultimately will be explained by an associationist theory. The intent of the argument for associationism is instead pragmatic and methodological. If the argument has been successful, the reader will agree that associationism provides a good foundation for exploiting the insights that both film music and film music theorists offer us about complex musicognitive processes.6