Information Technology Reference
In-Depth Information
to develop theories of how music is perceived,
represented and remembered by the human
mind. Dowling conducted experiments which
have focused on exploring the roles
pitch con-
tour
and
pitch interval
play as basic elements
of the human memory model for music. Pitch
contour refers to a measure of melody in which
the relative pitch values of successive notes are
compared using some metric. Several formal
explorations of contour measures have been
made by Friedman (1985), Marvin and Laprade
(1987) and Polansky and Bassein (1992), among
others. The most straightforward representation,
often called ternary pitch contour, examines ev-
ery pair of consecutive notes and compares the
pitch values of a given pair; the transition from
one note to the next is recorded as up, down or
same, depending on the relative pitches. A pitch
interval representation similarly examines every
pair of consecutive notes but records the transition
as the actual pitch distance from a given note to
the one which follows it. This value is normally
reported in semitones using the Western musi-
cal scale. Other pitch contour representations
can be devised, dividing the general “up” and
“down” designations into two or more possible
value ranges and, in effect, giving the measure
improved resolution over ternary pitch contour.
In fact, a pitch interval representation can be
seen as an extended case of applying additional
resolution to a pitch contour scheme. Figure 1
gives an example of how a short musical phrase
is represented using pitch intervals and ternary
pitch contour.
Dowling's study with Fujitani (1971) identi-
fied pitch contour and pitch interval measures as
relevant. Studies exploring the effect of using
familiar vs. novel melodies (Dowling, 1978;
Dowling & Fujitani, 1971), and a more recent
study on novel melodies presented repeatedly
over time (Dowling, Kwak & Andrews, 1995),
revealed that most people move from a contour
representation towards an interval representation
as their familiarity with a song increases. Bartlett
and Dowling (1980) pointed to earlier studies
(Attneave & Olson, 1971; Cuddy & Cohen, 1976;
Dowling, 1978) and concluded that “with unfa-
miliar melodies, subjects seemed to have little
trouble reproducing or recognizing the melodic
contour, but they had a great deal of trouble with
the exact-pitch intervals among the notes” (p. 501).
Another study (Dowling, 1986) showed that for
recognizing transpositions of novel melodies,
novices seemed to rely on interval representa-
tion, while moderately experienced subjects use
a method involving contour and a representation
of musical scale and professionals use more so-
phisticated strategies than those tested.
Edworthy (1985) explored the relative abili-
ties of people with musical training to recognize
pitch contour versus pitch intervals in unfamiliar
musical phrases. In her study ten subjects listened
to pairs of melodies which varied in length be-
tween 3 and 15 notes; the second of each pair was
transposed to a different musical key, and in 75%
of the trials, the melody was changed. For the
interval tests, a single pitch interval in the phrase
was altered by between one and four semitones;
Figure 1. An example of pitch contour and pitch interval representations for a musical phrase (Adapted
from Minuet in G, J.S. Bach)
PV:
67
60
62
64
65
67
60
60
69
65
67
69
71
72
60
60
PI:
-7
+2
+2
+1
+2
-7
0
+9
-4
+2
+2
+2
+1
-12
0
PC:
-
+
+
+
+
-
=
+
-
+
+
+
+
-
=
(PV = MIDI pitch values; PI = pitch intervals in semitones; PC = ternary pitch contour)
Search WWH ::
Custom Search