engleski

Aural perception of bitonality compared with an analysis using the scalar dissonance model

The reception of the term polytonality gave rise to a contested debate, resulting in inconsistent uses of the term, incompatible analytical approaches, and a lack of an appropriate definition. nevertheless, the notion of polytonality implies a combination or superimposition of tonally res- onant layers, whose interpretation depends on the conceptual framework used to explain their relationship. Martins (2017) proposed an analytical apparatus named the scalar dissonance aimed at modeling contrapuntal relations between distinct layers of polytonal texture as interactions between scales. The two essential properties of the scalar dissonance model are mismatch (abbr. MISM: ‘friction’ or ‘roughness’ between layers, measured by the number of notes intersecting conflicting steps in another layer, divided by the number of layers conceived) and porosity (abbr. PORO: “channeling” or “pivot” notes between the layers, measured by their number). The principal goal of this exploratory study is to examine the relationship between the perceived degree of bitonality (as the simplest kind of polytonality) with analytical data (MISM and PORO) obtained using the scalar dissonance model, which are expected to show a positive correlation. The study was conducted among 146 participants, students of music at BA and MA levels. The statistical model (composed of variables such as age, sex, major subject, duration of music education, year of study, the most recent grade achieved in Aural Training courses, handedness and the possession of absolute pitch) was tested against the results of a series of listening experiments consisting of 25 pairs of bitonal stimuli (composed samples and excerpts of selected pieces by Milhaud, Bartók and Casella) recorded on the piano. The participants compared the perceived degree of bitonality for each pair. The results of the listening experiments were also compared against the findings of an analysis conducted using the scalar dissonance model. The overall results of the comparative analysis were interpreted in light of the basic principles of aural perception (Bregman, 1990 ; Huron, 2016). The statistical model describing the linear relationship between the performance in listening exercises and the general variables shows high significance (p = 1.767 x 10-7). The optimal model (p = 1.444 x 10-8) consisted of the following variables within a given significance level (α = 0.1): major subject (p = 2.051 x 10-5), handedness (p = 0.0002529), the most recent grade achieved in aural Training (p = 0.002319), sex (p = 0.0606) and age (p = 0.02195). The possession of absolute pitch, duration of music education and year of study do not show significance. Most answers have confirmed our expectations, i.e. they correlate positively with the scalar dissonance model. The significant predictors of the expected answers (i.e. those that match analytical data obtained using the scalar dissonance model) included criteria of cognitive nature (e.g. major subject or the most recent grade achieved in aural Training), as well as handedness, sex and age.

bitonality, scalar dissonance, aural perception, stream segregation, auditory scene analysis

nije evidentirano