AW.: RE: quartertones and complexity

<< From: "Paul H. Erlich" <PErlich@Acadian-Asset.com>

Maybe Ferneyhough was using his ears? >>

I suspect that he was mainly using his eyes. F. isn't a performer himself,
and his recent use of the computer is limited to some notational and
algorithmic assistance, not audio output, so he doesn't have the means to
hear what he is composing except in his imagination. His statements on pitch
practice are somewhat contradictory: on the one hand he wants to hear
tempered pitches as themselves and not approximations of just intervals, or
the other, he is not particularly fussy about precise intonation. Thus, his
quartertone notation may or may not be prescribing 24tet. Presented with the
unruly set of pitches used in his _String Trio_, I was really curious to know
what he might be after, in both the case where the 24tet is played precisely
and the case where it is only approximate.

I recall that the criteria F. used to select the source chords in a piano
piece was something like their "ungainlyness". So he does operate with an
implicit, if primitive, theory of consonance, and an empirical one at that,
but not one that appears to have been formalized in any way. While his
complex rhythmic notation is intended to create a form of rubato, I had hoped
that his use of microtones was intended to do more than to create more than a
"rubato with pitches". I have no objection to that idea, but find the means
F. uses -- the over-precise notation -- to be out of proportion.

There is an interesting paradox concerning the complexity of a set of
pitches. The complexity could be measured in terms of the number of ways in
which the pitches might be associated with one another (i.e. the number of
connections on an interval lattice), yet the larger the number of those
connections, the shorter the algorithm will be for constructing the entire
set, which is another measure of complexity. Thus, pitch sets like an
Eikosany or the 12tet aggregate can be constructed in rather short algorithms
although -- and because -- they are both extremely rich in internal
relationships. A set like that found in F.'s _String Trio_ requires a fairly
long algorithm to construct, yet is less rich in internal relationships.
Since "complexity" is practically F.'s trade mark, the issue has to be
explored.