In a single richly suggestive word, "song," Sessions sums up all the factors—melodic, harmonic, rhythmic, textural, dynamic, articulative—that contribute to what I have called musical line: "Each one of these various aspects derives its functions from the total and indivisible musical flow - the song. . . . [M]usic can be genuinely organized only on this integral basis, and . . . an attempt to organize its so-called elements as separate factors is, at the very best, to pursue abstraction, and, (...) at the worst, to confuse genuine order with something which is essentially chaotic."1 Analysis, whose functions as a valuable tool for the training of composer and performer Sessions has so well explicated and demonstrated, is now all too often called on to justify and to further this essentially unmusical, or at best nonmusical, pursuit of abstraction. Herein lies the explanation for the increasing doubt of the general usefulness of the discipline that Sessions has lately evidenced.2 For the creation and analysis of art are two distinct activities, confused at the artist's peril. ". . . [A]nalysis cannot reveal anything whatever except the structural aspects of a completed work . . . Discoveries after the fact are necessarily verbalized in terms of preexistent contexts; it hears forward, as it were, in terms of the contexts.3 · 1. "Song and Pattern in Music Today," The Score 17 : 77-78.· 2. See, e.g., "Song and Pattern," p. 78, and "To the Editor," Perspectives of New Music 5 : 92-93.· 3. Questions about Music , pp. 109-110. Edward T. Cone, composer and professor of music at Princeton University, has written Musical Form and Musical Performance and The Composer's Voice, edited Berlioz's Fantastic Symphony, and coedited Perspectives on American Composers and Perspectives on Schoenberg and Stravinsky. In a slightly different form, this essay was delivered as an address at Amherst College on the occasion of a music festival honoring Roger Sessions. (shrink)