Annotation for Dawe, Lloyd A., Platt, John R., and Racine, Ronald J.
Inference of Metrical Structure from Perception of Iterative Pulses Within Time Spans Defined by Chord Changes
Annotation (by Bill Tilghman):
- This study presents a series of three experiments designed to
examine the relative contributions of different kinds of phenomenal
accent to the perception of meter. For the
purposes of these experiments, three kinds of phenomenal accent
were defined: harmonic (associated with a chord change), temporal
(associated with the initiation of an event of relatively long
duration), and melodic (associated with large leaps and contour
changes). Each of the three experiments systematically pitted two
of the accent types against each other to determine their
relative importance: Experiment 1, harmonic vs. temporal;
Experiment 2, harmonic vs. melodic; Experiment 3, temporal vs.
melodic. In addition, separate data were collected and compared
for "musicians" (more than 3 years formal training) and
"nonmusicians" (less than 2 years training). In the two
experiments in which harmonic accents were present, both
musicians and nonmusicians consistently based their choices of
meter on harmonic change. When no harmonic accents were present
(Experiment 3), musicians and nonmusicians differed in the way
they used temporal and melodic accents to determine meter.
Nonmusicians tended to base their decisions more on temporal
accents than on melodic accents, choosing the meter implied by
temporal accents when the two types of accent were in conflict.
Musicians, on the other hand, seemed to use both melodic and
temporal accents, choosing which ever type of accent implied
duple meter when the two types of accent were in conflict. The
slight but significant preference for duple meter among musicians
(evident in the data from the other experiments as well) is
interpreted by the authors as an indication of musicians'
knowledge of the statistical bias in favor of duple meter in