I have become completely dissatisfied with [serialism's] narrow terms. I found the palette of constant chromaticism increasingly constricting, nor could I accept any longer the limited range of gestures that always seemed to channel the music into some form or other of expressionism. The over-intense manner of serialism and its tendency to inhibit physical pulse and rhythm led me to question a style which made it virtually impossible to express serenity, tranquility, grace, wit, energy.George Rochberg, liner notes for the 1973 recording of his Third String Quartet issued on Nonesuch Records LP (italics mine).
I happen to like Rochberg's 12-tone music (2nd Symphony, Piano Trio) and I have no problem with his decision to abandon serialism (by which he means only serial organization of pitch, not integral serialism of Babbitt, Boulez, and Stockhausen) in favor of (mostly) tonal composition. But his attempted justification of this decision strikes me as borderline incoherent.
To begin with, Rochberg equates "constant chromaticism" with 12-tone music, which is simply false because chromaticism also pertains to tonal music whose expression of key is highly ambiguous or obscure. Such tonal music (e.g., Roger Sessions' 2nd symphony) may sound 'atonal', while 12-tone music whose basic tone rows are designed to create strong illusions of tonal centers may sound 'tonal', albeit ambiguous with respect to its key (e.g., Benjamin Frankel's symphonies, Interlude from Krenek's opera Karl V, Dallapiccola's Variations for Orchestra).
Second, Rochberg's claim that 12-tone music is incapable of expressing psychological (emotional) states traditionally associated with tonal music is absurd. Keeping in mind that our perceptions of such states in music are highly subjective (what sounds 'witty' and 'graceful' to Boris may sound 'banal' and 'trite' to Natasha), I get as much 'wit' and 'grace' from Krenek's 5th Symphony and his works for string orchestra as I get from Haydn. If I want 'tranquility' and 'serenity', I can find them as easily in the delicately shimmering and haunting sonorities of Gunther Schuller's Symphony (1965) or Boulez' Pli selon pli as I find them in the works of Debussy. As for 'energy', I get more of it from one short movement of a Roger Sessions symphony than from all of Bruckner's elephantine Scherzos combined. And when it comes to 'love', 'passion', 'sensuousness', 'sadness', and 'tragedy', no tonal music can match the impact with which these emotions have been expressed in 12-tone music (e.g., Berg's Lulu).
Finally, there is Rochberg's implied - and at best highly dubious - suggestion that some form of return to tonality is the only viable alternative for composers unhappy with (perceived) emotional limitations of serialism. Elliott Carter's music is neither tonal nor serial, yet, when Carter feels like it, he can infuse his music with intense, sweeping romantic passion and poignant, bittersweet longing (e.g., episodes for massed strings in Boston Concerto, Interventions, and Three Occasions); or with wit, playfulness, and happiness as pronounced as in the sunniest music of Haydn and Mozart (e.g., Oboe Concerto).
Rochberg's diatribe against 12-tone music would not have bothered me much if it weren't for the fact that he spoke not only as a composer, but also as a professor of composition - an academic appointment he held at the University of Pennsylvania for many years. With such incoherent and misleading denunciations of 12-tone music coming from professors of composition at Ivy League universities, it is no surprise that an average concert-goer in America still thinks of Schoenberg's Violin Concerto as a musical equivalent of a root canal...