DESCRIPTION:
Performance by Wilhem Latchoumia at CCRMA. October 15, 2013
Part of a US tour organized by GRAME (www.grame.fr), supported by FACE Council.
Info and program notes: https://ccrma.stanford.edu/events/grame-presents-wilhem-latchoumia-piano-electronics

The work "La rhétorique des doigts" (fingers' rhetoric) was born of
discussions between us, Jérôme Dorival and Yann Orlarey, in
1992. We wanted to make a piano piece in which all stages of
the composition would have been computer generated, starting
from the basic assumption that the fingers of the pianist would
be able, by themselves, to develop a true rhetoric. The work's
title is a nod to a collection of the seventeenth century, the
"Rhetoric of the Gods" composed by the lutenist Denis Gaultier
(1603-1672), including 56 splendid manuscript of lute pieces.
The music is written in the form of tablature, meaning that the
notes of traditional music theory are replaced by fingering
indications: when this finger is placed in such a case (between
two frets) on the neck of the lute, we obtain this note.

We can therefore say that even if a lute tablature gives an
indication of a move to make, the result is nevertheless a
musical note. Our project of a «fingers choreography» was
part of that line. It involved a very precise description of the
possibilities of the fingers, individually and as a whole. We
described the basic principles for the operation of the fingers,
hands, forearms, arms etc.. and then transformed these principles
into computer programs, using CLCE (Common Lisp Composition
Environment) that was then in use at Grame. For example a
pianist has only ten fingers, which excludes, in principle, he plays
more than ten notes (except in certain cases where a finger can
press two keys simultaneously) or if the thumb and little finger
play notes simultaneously removed (an octave or more), this
implies that the index can not play a key adjacent to the thumb.
A simple look at some classical piano works shows this evidence
(though sometimes violated). We discovered then the physiological
limits, the impossibilities of the finger position relative to each
other, the difficulties of any kind were more than we had imagined,
and therefore program lines stretched.

The next step was to create a «reservoir of notes» and a «reservoir
of rhythmic values» (including silences) where the program was «to
choose» at random. We could then listen in the studio, from a
sampler, the first notes of the work.
Our job was then, more specifically, to «select» the most
interesting generated materials and to compose them in sequence
and so on. It is clear that changes in «reservoirs» of notes and
rhythms could radically change the musical result, and the entire
computer device called a learning «game», guided by the ear and
the aesthetic choices. We were somewhat in the position of a
gardener who has planted a shrub that looks and grow,
«correcting» the daily growth. Once the work was finished, it is still
the computer that wrote the score,but it is a human pianist, Bruno
Robilliard, who played the work for the first time, which was our
ultimate goal since the project began.