Melody Processing

Grouping

To isolate a melodic line from polyphonic music, one needs to form a coherent perceptual stream of the melody. This process is known as stream segregation, and several principles governing it—as explained by Al Bregman from the 1970s to the 1990s—relate to timbre and temporal and frequency proximity of the successive tones. For instance, if the notes of the melody are scrambled across octaves, melody becomes virtually impossible to recognize because the pitch continuity is broken. Similarly, if two well-known melodies are interleaved with each other within the same register, they are impossible to recognize, but when the other is transposed away, they can be easily recognized, as demonstrated by Jay Dowling in 1973.

Another low-level process that allows one to make sense of tone sequences is grouping of phrases and motives, separated by boundaries. Such boundaries are formed by rests, longer durations, melodic leaps, or relation of the tones to the key (Panel B of Figure 1). Grouping is often marked by other stylistic devices, based on harmonic or melodic conventions. Listeners expect certain regularities at grouping boundaries, and regular grouping structures facilitate processing melodies, evident in empirical studies using recall and recognition tasks.

Figure 1The melody line of the first four bars of W. A. Mozart's Piano Sonata no. 16 in C major K, 545 in score format (A). The following panels demonstrate (B) grouping, (C) metrical structure, (D) contour, (E) tonal hierarchy, (F) probability of pitch transitions, and (G) implied harmony of this melody.

Grouping structures at a higher level creates musical forms (ABA, AABB), which is also a device that influences melody processing. When the appropriate and familiar form is recognized, the repetition of the musical material facilitates the processing of subsequent melodic materials.

Pulse and Meter

Part of the process of parsing a coherent melody perceptually together is to relate the individual tones to an underlying pulse. This refers to a [Page 685]perceptual phenomenon that is inferred from the events in the melody. The listener sets up an isochronous sequence of pulses to track the passage of time within the musical material. Pulse trains are hierarchically organized into different meters by grouping repeated beats at different time scales (one for 4/4 meter is shown in Panel C of Figure 1). The basic period of this pulse train, often called the “tactus level,” is around 60 to 150 beats per minute. Latching onto the meter allows the listener to selectively allocate attention and synchronize perception with the production of music. The result of this for perception of melody is that the events occurring at the dominant hierarchical locations are perceptually more salient.

...