Preference testing between JI and EQT, 1/4 MT and EQT - 300 teenage musicians at FSU

Hello -

I was writing my daughter about work I had done testing about
300 music students - mostly teenagers - for their preferences between
just and 12 equal temperament versions of synthesized musical phrases -
mostly four part - and in other testing between otherwise similar
quarter comma mean tone and 12 equal temperament piano pieces or
briefer phrases. After I'd written the E-mail to her, it occurred
to me to post the part about the testing I'd done here. The work
is still in progress, and the i's haven't all been dotted nor the
t's all crossed. Although my sample sizes tended to be small, some
tests for statistical significance - e.g., existence of a correlation
between pitch acuity and degree of preference for just intonation for
the 12 and 13 year olds came out significant at the - I believe -
one percent level. Quite frequently, preference for just or equal
temperament
versions depended upon which of the two versions had a wider melodic
interval at some point of emphasis in the phrase - e.g. a phrase
containing a prominent melodic drop of a minor third would tend to
be preferred more in the just version (m3 wider in just), while
a phrase with an emphasized step up from a fifth to a major sixth
above a root note would have a tendency to be preferred in the
equal temperament version as sounding more expressive or emotional.
The test results taught me a lot, and I hope to publicize them in
a tidied up form. If someone were planning to embark on a major
study in the area of tuning preferences, the experiences derived
from the work I did at Florida State might be useful towards
designing the study.

One serious problem with testing groups of listeners in a
classroom setting is that there is the possibility that one individual's
subtle reactions of liking or dislike for a particular example
version might be picked up by others - perhaps subliminally -
and influence their reactions to the examples. I was not present
in person during the testing, but those who gave the tests at
Florida State were aware of this possible interfering factor and
tried to conduct the tests in such a way that the listeners
understood that they were to avoid talking or doing anything
which could disturb the others taking the test.

Here's what I wrote to my daughter:

In 1997 and 1998 I gave - sent cassettes and then in 1998
CDs - listening tests to about 300 music "campers" and incoming
students.

(They had summer music camps at Florida State
in which high school students from all over Florida would come and
spend two weeks there taking an intensive series of individual
and group lessons in their musical instrument and take part in
a musical performance together with other campers at the end
of their time at FSU)

In the tests I asked about their preferences between
different tuning versions of brief musical phrases. You would
have to see it to believe it! The reactions of different listeners
to the comparisons were so different! What one person thought
sounded beautiful another thought sounded terrible! Then I've found
that most people, whatever preferences they have, find it hard to
believe that many others respond to sounds very differently
than they do.

I believe that both "genes" and musical listening experience
are very important as to the way we respond to different musical
sounds and different kinds of music. Some groups I tested were
adolescents ranging in age from 12 to 18, and there were some
interesting relationships. As you would expect, there was some
increase in pitch acuity with age - especially between about 12
and 15 - very slight change after that. I found that among the
youngest ones - 12 years old - those who had the greatest acuity
for detecting pitch differences had a strong tendency to also
prefer the just intonation versions of comparison examples, while
those with the least pitch acuity heavily favored the equal
temperament versions. There was the same trend among the 13 year
olds, but it wasn't as pronounced, and I didn't see much of
any clear cut trend in this direction among the older ones. The
older ones generally had had more musical training than the
younger ones, and I believe that as equal temperament is used
as a standard in most music education (brass is an exception),
as the young musicians get more training and experience in
equal temperament, an innate inclination towards just intonation
of those with such a high level of pitch acuity as to make the
different tunings really sound a lot different to them becomes
increasingly replaced by a learned preference for equal temperament
as a result of more musical training in equal temperament music.
But the story is more complicated than that - all sorts of variation!
In some cases, at the beginning of the test a listener would consistently
prefer one version and then suddenly switch and prefer the opposite
version.

I hope this post contributes constructively.

Dave Hill, Borrego Springs, CA