Disentangling cognitive and sensory expectations

When investigating the processing of tonal expectations, it is important to keep in mind that those expectancies are generated both by bottom-up processes (sensory expectations) and top-down processes (cognitive expectations), and that only the latter involve listeners’ implicit knowledge about tonal structure. Using strong violations leads to a confound between sensory and cognitive expectations: As stated by Koelsch (2007), an ERP evoked by a C# major chord presented in a C major context could not be attributed only to top-down processes because this chord would also bring acoustical incongruency causing a violation of sensory expectations. In a first attempt to disentangle cognitive and sensory processes of tonal expectations, Regnault, Bigand and Besson (2001) compared ERPs elicited by a tonal violation and by an acoustic violation. Eight-chord sequences had their six first chords manipulated so that the final chord was either a tonic (tonally expected) or a subdominant (less tonally expected), albeit being acoustically the same chord in both conditions. These sequences had been found previously to elicit stronger expectations for the final chord when it was the tonic than when it was the subdominant, in a behavioural study using the priming paradigm (Bigand & Pineau, 1997). This tonal manipulation was crossed with an acoustical manipulation: the final chord being either kept consonant or rendered dissonant. Different ERP components were elicited by the tonal violation and by the acoustical violation: The acoustical violation elicited an LPC between 300 and 800 ms, whereas the tonal violation elicited an earlier P3 (centrofrontal, between 200 and 300 ms). This result questions the interpretation of the LPC (and P600 and P3b) components found by Besson et Faïta (1995), Janata (1995) and Patel et al. (1998). In those three studies, musical events of the more incongruous condition were also less acoustically similar to their context, therefore the positivities found in these studies may be confounded by sensory processing of dissonance. More recently, Koelsch et al. (2007; Koelsch & Sammler, 2008) discussed that their previously used unexpected Neapolitan sixth chord confounded cognitive and sensory processes, and replaced it by a double dominant (i.e., the dominant of the dominant, e.g. D-F#-A in C major) or by a supertonic (i.e., the chord built on the second scale tone, e.g. D-F-A in C major). This change, together with the particular arrangement of the 4-chord contexts, resulted in two models of acoustic congruency (the short-term memory model by Leman, 2000 and the pitch commonalities by Parncutt, 1989) predicting stronger congruency between the contexts and the two unexpected chords (double-dominant and supertonic) than between the contexts and the tonic chords. An ERAN was still elicited by tonally unexpected chords, thus pleading for its association with cognitive processing of tonal violations.

ERPs studies disentangling cognitive and sensory expectations have only investigated chord processing and not tone processing. Focusing on tone processing would give more control on the experimental material because tone sequences are less complex than chord sequence. In addition, since sequences of single tones convey less tonal information than chord sequences, observing ERPs associated with tonal relations would highlight the strength of cognitive tonal processes. Finally, investigating tonal expectations in tone sequences would establish a link with the investigation of pitch processing and the influence of top-down processes on pitch processing. ERPs studies have shown pitch processing to be sensitive to top-down processes linked to attention (Kauramäki, Jääskeläinen, & Sams, 2007) or musical expertise (Besson et al., 2007; Tervaniemi et al., 2005).