Transcript Document

Perceptual learning on ecologically typical and atypical musical timbres
Kristin Cox, Megan Loveland, & Nestor Matthews
Department of Psychology, Denison University, Granville OH 43023 USA
Background and Purpose
Melody X
Melody Y
Backw ard Melodies
Note Forw ard
Note Backw ard
Pow er (Note Forw ard)
Pow er (Note Backw ard)
Our findings replicate Schellenberg and Trehub
(2003), showing that pitch memory is accurate for
familiar melodies. The present findings also show that
pitch memory is above chance for ecologically atypical
timbres (backward-played songs) despite a strong
forward superiority effect for familiar tunes. This effect
appears to be limited to melodies, as the results from
Experiment 2 show no significant difference between
the forward and backward conditions for single notes.
When using novel melodies, results show direction
specific training effects; either forward or backward
superiority effects can be generated with training.
Furthermore, participants appear to use a melodyspecific strategy as evidenced in Experiment 3 by d’
values greater than zero for each overlapping delta cent
value in the two melodies.
1
1.8
y = 0.0063x 1.0016
R2 = 0.9766
1.6
0.8
1.4
0.6
0.4
0.2
1.2
1
y = 0.018x 0.7838
R2 = 0.9955
0.8
0.6
0.4
0.2
0
0
Lyrics
No Lyrics
0
Lyric Condition
50
100
150
200
250
300
Delta Cents
Experiment 3
The Bottom Line
Shared Frequencies
1.6
1.4
1.2
Pitch Memory (d')
Exp.1 – Participants listened to original version
excerpts of 4 popular tunes: “Mrs. Robinson,”
“Imagine,” “I’ll Be There for You” and “Livin’ La
Vida Loca”. Participants were then presented 160
pitch-altered excerpts to compare to the original.
Excerpts were also played forward and backward as
well as with and without lyrics.
Exp. 2 – Participants made sharper/flatter judgments
about two separate piano tones3 played forward and
backward, and in 50 cent intervals higher and lower
than the original with a maximum of 250 cents.
Exp. 3 –Melodies (see below) incorporating the same
tonalities as exp. 2 were created, and altered according
to the same cent shifts as well as forward/backward.
All participants were exposed to pre and post sessions,
but split into 2 statistically similar groups for training,
listening to either forward or backward excerpts only.
The task was the same as in exp. 2.
Discussion
Experiment 2
Pitch Memory (d')
Method
Forw ard Melodies
Pitch Memory (d')
Recent discoveries indicate that non-musicians have
stunningly accurate memory for the original musical
key of popular tunes1. Musical transients are also
salient in detecting differences in timbre2- a
subjective quality of pitch. The purpose of the
present study is to test these phenomenon with single
piano notes and novel melodies.
Experiment 1
1
0.8
0.6
0.4
0.2
0
Pre-test
Post-Test
Session
Non-musicians show a strong forward-superiority
effect when listening to familiar tunes, but can acquire
a similar backward-superiority effect through training.
Remarkably, performance is above chance levels for
ecologically atypical timbres in both familiar and
unfamiliar tunes. Results also show that individuals
use a melody-specific strategy instead of a frequencyspecific strategy to form memory for original key.
References
1.Schellenberg & Trehub (2003) PMID: 12741751
2. Thayer, R.J. (1974). The effect of the attack transient on aural recognition of
instrumental timbres. Psychology of Music, 2, 39-52.
3.Pantev, Oostenveld, Engelien, Ross, Roberts, Hoke (1998) PMID: 9572139