Cornelius Cardew: Autumn '60 for orchestra (1960).
Let's toss our notions about an "orchestra", a "score", an individual instrumental "part", as well as the division of labor between composers and performers (and, among performers, between players and conductors) up in the air and listen to what happens when they come down again in new configurations: Autumn '60 is a delightful example of such an exercise.
Autumn '60 is "for orchestra", but the identity of that orchestra is variable. Although most performances have probably been for chamber groups, the score is as usable, potentially, by a single soloist on a lonely woodblock as by an orchestra of late romantic scale and proportions.
The score, which all players receive, features a system with two staves, the upper consisting of one to seven conventional markings — clefs, notes (which are limited to f, gb, ab, bb, and db), articulations, ornaments, instrument names, dynamics, etc. — assigned to the quarter note beats of simple meters, from which individual players are to extract their own parts and enter into the lower staff.
The process of part extraction appears, at first, to be rather open, but it gradually reveals itself to be in many ways quite constrained, in particular through the instruction to "ignore any two of the indications for any particular beat" as well as those combinations of notation which lead to an instruction to "play nothing", each of which might serve to create a kind of negative space around the material identity of the piece.*
The degree of cooperation in part preparation between the players is not specified; indeed, the parts may be extracted in complete independence, and further independence is introduced by the interventions of a conductor, who also uses the score to prepare a part, in which details of form, tempo, rubato, ensemble composition and texture, and even additional degrees of performer liberty may be introduced.
There were certainly more radically indeterminate pieces in the late 1950's and early 1960's, but I find that Cardew was perhaps uniquely sensitive, here and in his Octet '61, to the wealth of possibilities in subtly moving the traditional lines of demarcation identified above. Need I add that Autumn '60 is also great fun to play? Rehearsing, in particular, is a necessarily and satisfyingly engaging process, a reminder of how collaborative an enterprise music can be.
* In the pair of performances of Autumn '60 in which I took part, there was a marked sense of a tonal identity, arising methought, from an interplay between asserted and negated elements not at all unlike that found either in classical tonality or the use of complementary sets in some atonal music; this is a nice reminder that one asserts a tonal center, for example, as much through its presence as through its strategic and tactical absence.