From an interview with Robert Carl, author of Terry Riley’s In C (Oxford Press, 2009):
I think there’s a subtle but a real difference, though, between the repertoire and In C. In C has an open instrumentation and an open duration. And as a consequence of the kind of accordion structure it has, where it can expand or contract, the relationships between the modules are also different from performance to performance, even though the sequence always remains the same. Every time someone decides to perform it, every time someone decides to record it, it’s a new version, maybe a new realization. Maybe we should be using the word realization instead of interpretation, because interpretation suggests a 19th-century ideal of a score which is a fixed artifact that one is supposed to realize as close as possible to the text. But with In C, you can only get so close and then like a magnet you bounce off it.
The distinction is an interesting one, and from reading the interview, it’s clear that Carl is no dogmatist. He’s not making a stark distinction here that he’ll then defend to the hilt. He’s pointing out, in his word, the subtle distinction between what is understood to be “repertoire” and what Riley’s In C proved to be. This is a distinction between works whose depiction in standard musical notation is fairly fixed, and those works, such as Riley’s, that are fluid, in that they depend as much if not more on instructions/procedures as on musical notes. The score for In C, for example, if half notation (53 short melodic segments) and half “Performing Directions” (including such ambiguous koans as “If for some reason a pattern can’t be played, the performer should omit it and go on”).
In complimenting several renditions of In C that he thinks are particularly successful, Carl says, “For me, I like the works that take off from it and actually make truly new pieces, but that’s just my taste.”