Conclusion

Our study investigated tonal expectations in moderately and weakly musically experienced listeners, and emphasizes the ability of these listeners to develop expectations in agreement with the tonal framework predicted by music theory, even when the cues available to extract differentiated tonal frameworks were rendered minimal. In this view, observing priming effects of a medium (Experiment 1) or small (Experiment 2) effect size should not be detrimental to the psychological importance of the observed effect. As underlined by Prentice and Miller (1992), a large effect size (i.e., d > .80 following Cohen) is not the only way to demonstrate that an effect is important; the importance of an effect should also be seen as a function of how minimal a manipulation of the independent variable can be to still produce an effect (see Prentice & Miller, 1992 for a review of the psychological relevance of small effect sizes). Observing cognitive expectations for nonmusician listeners despite the weakness of the tonal manipulation and the control of both tone repetition and spectral richness of the sound suggests the psychological relevance of the observed tonal relatedness effect.

Our study provides an argument in the debate on the modular nature of syntactic processing in music. Manipulations of tonal relatedness have often been compared to syntactic manipulations, in analogy to language (see Patel, 2003 for a review). However, in contrast to language, musical syntax andpsychoacoustic properties of soundare strongly entwined. The psychoacoustic properties of sound have had a fundamental influence on the development of the Western tonal system, leading to the sensory relatedness of syntactically related musical events. For the experimental investigation of music processing, the lack of control (or insufficient control) of sensory components in musical stimuli results in a joint manipulation of cognitive and sensory components, thus preventing conclusions about purely syntactic processing. Our study removed as much sensory components as possible: in contrast to previous studies on musical expectations, our study went beyond the control of tone repetition and included the control of the spectral complexity of the sound. The overall outcome of experimental results and simulations with the sensory model pleads for the existence of musical syntax processes that are independent from the processes linked to psychoacoustic features. As stated in Bigand, Tillmann, and Poulin-Charronnat (2006), the next step in assessing the respective role of syntactic and psychoacoustic processes should be to search for distinct neural signatures of syntactic and psychoacoustic processes. Material controlled as in our present study will be useful in this respect and, ideally, it should further be crossed with comparable material that systematically manipulates the psychoacoustic components while controlling syntactic ones.