Experiment 3 studied the influence of tonal relatedness on pitch perception with a comparison judgment task using the sensitivity of a rating-scale paradigm. Since the task was a same/different comparison of two tones, no internal reference pitch was needed and a possibly confounding influence of tonal relatedness on mistuning/wrongness judgments should also have been avoided14. Furthermore, the same/different task allowed us to test discrimination for small mistunings. A 9-cent mistuning (0.5%) is close to the discrimination threshold for non-musicians (0.86% according to Micheyl, Delhommeau, Perrot, & Oxenham, 2006).
For the 9-cent mistuning, area scores revealed an influence of the tonal relatedness of the to-be-compared tones: pitch discrimination was better when the target was tonally related than when it was less related. This result strengthens the results of Experiments 1 and 2 showing that tonal expectations influence pitch perception. Though the observed effect size of tonal relatedness on pitch discrimination was in the “small effects” range (Cohen’s classification), the psychological relevance of an effect also depends on how minimal a manipulation of the independent variable can be and still produce the effect (Prentice & Miller, 1992). In our study, the change in tonal relatedness was rather subtle (tonic versus subdominant, as opposed to, for example, tonic versus leading tone) and instantiated by a one-note difference in related and less-related contexts.
The influence of tonal relatedness on pitch discrimination was observed only for the finer mistuning, but not for the 17-cent mistuning. This finding suggests that the influence of tonal expectations on pitch discrimination might come only to influence performance when the to-be-made comparisons are sufficiently fine-grained. Even if performance level is not at ceiling for the 17-cent mistuning, the required comparison might be too coarse and thus the task might be too easy, so that facilitated processing based on tonal expectations has no apparent effect.
It is worth emphasizing that participants could make the same/different judgments while ignoring the preceding context since the beginning melodic context is irrelevant for the task. This strategy would have been particularly possible because no internal reference pitch was needed (in contrast to tasks in Experiments 1 and 2). Nevertheless, fine pitch judgments were influenced by the tonal function of the target tones. This result indicates that listeners process the tonal hierarchy of the context and automatically develop tonal expectations. Automatic musical expectations have been previously shown for chord processing with the harmonic priming paradigm (Bigand, Tillmann, Poulin, D’Adamo, & Madurell, 2001; Justus & Bharucha, 2001; Tillmann & Bigand, 2004; Tillmann, Janata, Birk & Bharucha, 2003).
4. In contrast to Experiment 1, no significant response bias was observed, even if hits and false alarms were slightly more numerous for 17-cent targets in the related than in the less-related conditions. This difference between the experiments might be due to the same/different task (avoiding a possible confound between mistuning and tonal function) and to the error feedback that might have imposed a fixed criterion, thus preventing a bias shift.