Features, Independence and Interaction of

A feature is a distinct and measurable property of a phenomenon or stimulus being encountered. Everything individuals encounter in everyday life has multiple features. A shirt has a certain shape, color, and texture; a drink has a certain temperature, taste, and viscosity, and so on. Music is no exception—it combines information from pitch, time, timbre, and loudness (at the least). The brain is thought to have specialized systems for detecting and analyzing these various features independently, but it must also reintegrate the resulting information to create a coherent musical experience rather than a jumble of features. These twin processes—feature analysis and reintegration—are indirectly reflected in how individuals experience and perceive music.

In some contexts, different features seem to have distinct or independent effects on individuals' experiences. Intuitively, a musician's ability to classify a familiar melodic interval, such as a perfect fifth, should be determined only by the pitches involved independently of irrelevant features of sound such as loudness or timbre. But there is evidence that even the experience of melodic intervals—including familiar ones—is influenced not only by pitch but by attributes of sound such as timbre. Such interactions between features underscore that individuals do not fully understand the processes by which features are analyzed and reintegrated. However, insight into these mechanisms can be gained by examining such interactions in varying contexts.

To investigate this issue one first needs a clear understanding of what it means for features to be independent or interactive. W. R. Garner's seminal work in cognitive psychology provided the concept of separability and integrality—that some features of a stimulus can be processed separately such that changing one will not affect another, but other features simply cannot be separated in perceptual processing and will always interfere with one another. For example, a change in the brightness of a color cannot be reliably separated from a change in its saturation by the human visual system. But separating position in space and color is not a problem.

[Page 460]In music, several researchers have looked for evidence of independence and interaction of two core features of music: pitch and time. The pattern of pitches and durations is what defines the identity of a piece of music, whereas changes in timbre or loudness are typically experienced as secondary characteristics. Therefore, much research has focused on pitch-time integration, most notably drawing from Mari-Riess Jones's theory of dynamic attending. According to Jones, attention is inherently temporally based, and pitch and time are not independent but inextricably linked by combining into a joint accent structure. For example, rhythmic context affects individuals' ability to detect changes to melodic structure. Other researchers, however, disagree, instead contending that pitch and time are processed independently in certain musical contexts. For example, an additive combination of the effects of pitch and time predicted ratings of phrase completion quite well (suggesting independent contributions), whereas an interaction term added no additional explanatory value. Indeed, independence has been observed for a variety of paradigms, including pitch change detection, illusory conjunctions in short-term memory, and the study of double dissociations in neuropsychology. Such evidence supports the idea that pitch and time are processed in a modular fashion.

...