Categorical Perception Stuff from XA

For posterity's sake, some stuff we were discussing on XA:

The Categorical Experiments - these are retunings of a Bach fugue from
7-EDO up to 5-EDO

http://soundcloud.com/mikebattagliamusic/sets/the-categorical-experiments/

The purpose of this example was to allow the generator to sweep across
the tuning spectrum from meantone to superpyth and beyond, and see
what happens as the ratios in a tuning change but the scalar structure
is left (mostly) intact. Much of the discussion on this happened in
the XA chat room and -NOT- on the board, so I'm archiving it here for
the record.

For the most part, the consensus was that while the intonation did
change noticeably throughout, and while different intonations imparted
their own unique colorations on the sound (sometimes heard as
pleasant, sometimes unpleasant), there is at least some aspect of this
piece that remained consistent for as long as the scalar structure
remained intelligible, presumably the aspect that has to do with
Rothenberg's "rank order matrix" concept.

A few people reported hearing the piece become less intelligible at
the "extremes," for example towards 5-EDO, where "s" became so small
so as to be almost indistinguishable, or towards 7-EDO, where L-s
became so small so as to be almost indistinguishable. Most people
reported still hearing some semblance of diatonic intelligibility even
in 7-EDO. A very small few (myself, John Moriarty, Pete K, Ryan
Avella) reported hearing some sporadic semblance of diatonic
intelligibility even in 5-EDO, sometimes reporting that this
perception increased after listening to the other examples many times
first.

There were a few people in XA chat, whom have always approached
xenharmonic music from more of a scalar angle, that didn't understand
what the point of this experiment was, or why it was supposed to be
interesting that playing the diatonic scale in different tunings made
it still sound like a diatonic scale. One interpretation of this
objection is that these people didn't perceive the shifting ratios as
particularly interesting as long as they were being used to intone the
same scalar, tonal, and overall categorical structure Western
listeners are most accustomed to. Others suggested that many of the
more "discordant" intonations might sound quite palatable if they were
being played over a piece of music "intended" to use those tunings by
the composer. This suggested a followup experiment: to retune one of
the fugues from the well-tempered clavier that would be played in a
major key which tends to have wider thirds for most well-temperaments,
and see if it sounds more "natural" in superpyth than meantone.

Many people seemed to interpret this experiment as a test to see which
tuning people would like the most. The response on XA was a bit varied
in terms of "preferences" - some people preferred around POTE meantone
(Gene), a few preferred the sharper side of things a bit (me, Andrew),
others said they liked tunings even flatter than POTE meantone (Keenan
Pepper). Some said they really liked the really sharp 72-EDO one due
to their being a bit more "xen." Individuals seemed to vary in their
tolerance to discordance of intonation, with Gene mostly hating it and
others finding some of the more discordant mappings to be somewhat
novel.

In general, the people who have spent more time playing around with JI
tended to report a greater perceived shift between things like "major"
and "supermajor," and tended to be more sensitive to the intonation. I
suggest that this may reflect the creation of additional (perhaps
partial) discontinuities in portions of the interval spectrum which
are categorized together for most western listeners. Andrew Heathwaite
was the most vocal person in this group, but he also said that he
heard the harmonic qualities of "supermajor" and "major" as being more
similar than the harmonic qualities of "major" and "minor," perhaps
suggesting a graded categorization of musical intervals. This suggests
that prototype theory may be a useful tool in modeling categorical
perception: http://en.wikipedia.org/wiki/Prototype_theory

That's it for now.

-Mike