Musical expectations within chord sequences: Facilitation due to tonal stability without closure effects.

Abstract

Musical priming studies have shown that musical event processing is facilitated for the tonally related, stable tonic (chord, tone) in comparison with less-related, less stable events. However, target events have always been in the final position of the musical sequences, position at which the tonic is the most expected event as it brings closure. Priming data thus contain a confound between tonal stability and end-sequence wrap-up processes, comparable with those reported for sentence processing. To investigate musical expectations without this confound, our study omitted the advantage of the tonic linked to the final position and placed related and less-related targets at various positions within 8-chord sequences. To indicate to-be-processed targets, visual information was synchronized with the presentation of each chord. Data showed higher accuracy and faster correct response times for stable tonic over less-stable dominant targets. The here introduced musical priming paradigm contributes to our understanding of listeners' knowledge about tonal hierarchy and provides a new tool for testing musical integration and event processing in musical materials.Keywords: priming, music cognition, nonmusicians, implicit knowledgeSupplemental materials: http://dx.doi.org/10.1037/a0030454.suppIn context, listeners develop expectations about future events, and these expectations-when they are fulfilled-lead to facili- tated (i.e., faster and more accurate) processing. Facilitation effects have been studied with the behavioral priming paradigm for lan- guage (e.g., McNamara, 2005) and music (e.g., Tillmann, 2005), thus investigating the influence of the perceivers' knowledge on structure processing, the development of expectations, and their influence on event processing.For music, the relation between the prime context and the to-be- processed target event is manipulated based on tonal structures. Tones and chords differ in their degree of tonal stability and functional importance in a given key. For chords, for example, the tonic is at the top of the tonal hierarchy, followed by the dominant, then the sub- dominant, the other in-key chords, and, finally, the out-of-key chords. In musical priming studies, the prime context (e.g., a chord sequence) instills a given key and the tonal stability of the target (e.g., a chord) is manipulated: it is the stable tonic chord in the expected condition, and a less stable in-key chord, such as the dominant or subdominant, in the unexpected condition (Bigand & Pineau, 1997; Tillmann, Janata, Birk, & Bharucha, 2008). The prime context is thought to activate listeners' knowledge about keys and tonal hierarchies, lead- ing to the development of stronger expectations for the stable tonic chord than for less-stable chords, thus resulting in facilitated process- ing of tonic targets. To assess processing speed, target chords are rendered acoustically dissonant in half of the trials, and participants make speeded judgments of the sensory consonance/dissonance of the targets.1 Faster correct judgments were made for targets higher in the tonal hierarchy (e.g., for tonic chords than for dominant chords). Musical priming studies have investigated tonal expectations in both harmonic and melodic sequences (see Bigand & Poulin-Charronnat, 2006, for a review; Marmel, Tillmann, & Dowling, 2008). Priming effects provide evidence that listeners, even without explicit musical training (i.e., referred to as nonmusicians), have acquired knowledge about the Western tonal musical system by mere exposure to musical pieces in everyday life (Krumhansl, 1990).However, an important criticism to previous priming studies is that targets were presented as the final event of the sequences. The expected tonic created a cadential resting point at the end of the sequences, and, consequently, it was strongly expected not only because of its tonal stability but also because it provides a feeling of closure, which was not the case for the unexpected less-stable event. …