Cultures using ETs

>Modulation via transposition is central to African xylophone playing.

Hence it is not surprising to find approximations to 5-tET and 7-tET in
Africa. In this case as well as that of Thailand, it is not suprising to
find "large" deviations from theoretical equal temperament. The
tolerance for melodic "errors" is rather wide -- try tuning several
octaves of 5-tET or 7-tET by ear and see how well you do. Use inharmonic
timbres, and remember that many cultures enjoy a "dissonant" octave
(sharp or flat) which preserves the identity of the two notes. The point
is that the concept is one of equal intervals; like Aristoxenus'
concept, this doesn't have to be accompanied by specifying
string-lengths as roots of 2 or whatever; satisfying the musician's ear
is all that counts, and with only 5 or 7 notes per octave, even 30-cent
deviations from true equality can be understood. 5 and 7 are most
frequent because they approximate fourths and fifths as well as the
octave; these are the only rational intervals that can have any meaning
in a melodic context with inharmonic timbres, and again 30-cent
deviations can be accepted in this context.

With long-duration JI simultaneities on harmonic timbres, you can
actually hear 1 or 2 cent deviations from ideal ratios, so the standards
for tuning deviations must be quite different. ET in the west evolved
from tunings which better approximated just intervals, but 16-cent
deviations are now accepted. These 16-cent deviations are more
disturbing to the rational intervals, and more damaging to the equality
of the tuning if rectified, than the 30-cent deviations are in the
systems above.

While African xylophones have inharmonic timbres, their voices do not,
and I've heard just 7-limit intervals in African ensemble singing that
sounded non-Western enough for me to surmise that this was their
traditional form of singing (unlike the stuff Paul Simon made famous). I
don't have any recordings to back this up, though; anyone care to
recommend any?