摘要：Background in music theory and analysis. In traditional music analysis, the instrument is
often overlooked while its effect on the formation of structurally relevant aspects of harmony
and counterpoint is regarded as inconsiderable. The principles of harmony and counterpoint are
believed to have a key influence on the formation of the pitch (class) content of a classical
work. It does not follow that the instrument would have no effect on it. However, these aspects
of work’s pitch content have not yet been systematically studied, partly due to the lack of the
adequate methodology.
Background in statistics. Currently we do have a range of new developments in the theory and
practice of multivariate statistical data analysis. This essentially models reality where each
situation, product, or decision involves more than a single variable. However, it was not just
until recently that a new powerful multivariate tool was developed for taking apart separate and
interactive effects of multiple factors. In music, it enables for the first time to take apart the
separate and interactive effects that musical instrument, style, etc. may have on work’s pitch
content.
Aims. The aim of the paper is to highlight the characteristic features of pitch content of soloist
parts of selected violin and cello compositions of Brahms, Saint-Saëns, Hindemith, and
Shostakovich which are caused by the use of the selected instrument.
Main contribution. The MIDI files were converted into dataset. The multivariate effects of the
studied factors on the pitch content were analyzed by statistical program PRIMER 6 &
PERMANOVA. The Euclidean similarity measure was used to construct the similarity
matrices. The statistical differences between the factor levels were assessed by three-way
PERMANOVA with “century”, “composer” (nested in century) and “instrument” as factors.
When a factor and/or interaction was identified as significant (P < 0.05), post-hoc
PERMANOVA pair-wise tests were conducted to detect which levels were responsible for
significant differences. The contribution of different structural elements to the differences was
calculated by SIMPER procedure.
Implications. The analyses showed that the effect of an instrument on pitch content primarily
lies in the extent of the variability of selected elements. Secondly, the historical style seems to
have a slight effect on the use of the instrument. Thirdly, composers take the aspect of an
instrument into account but this does not affect their style.
