Pierre Boulez WEBERN Complete Works

The complete recording of Anton Webern's output just released by Deutsche Grammophon does very much more than refute old and sterile arguments against serialism. But that it does, and overwhelmingly. Webern -- always a close observer of his teacher Schoenberg, though, as an essentially lyrical composer, thoroughly independent -- took up Schoenberg's new serial technique in 1924 and never looked back. For the rest of his life, 21 years, he went on delighting in the opportunities serialism presented for making patterns: patterns like those of the snowflakes, flowers and crystals he admired in the Austrian mountains, patterns that would support his music's exquisite strains of melody and be supported by them.

He was making patterns with the past, too. Serialism reinvigorated for him the standard forms, especially variations and canon. And it brought him closer to the old masters. Starting with his Symphony (1927-28), most of his movements are canons of one kind or another, often allowing an expressive gesture to be answered, balanced and pinpointed by a copy moving in the other direction, a lift by a sigh, along the course of a regularly waving rhythm.

Canonic and variation forms were also outcomes of a quest for integration, for creating music in which a motif of three or four notes would be constantly present in different colorings, registrations and rhythmic placements. Hence the paradox that the music feels, expressively, so fragile that it might fall apart if one put a finger on it and yet, structurally, it is tightly made and reinforced in every direction. It is at once tender and tough.

As for links with predecessors, that same symphony, although it has just two movements, is scored for chamber orchestra and lasts less than 10 minutes, has something of the grand melodic sweep Webern admired in Bruckner. The concerto that came soon after is a modern ''Brandenburg,'' and Bach is invoked again in the two cantatas on spiritual themes that came near the end of Webern's life.

At the same time, Webern outfaced his nostalgia by resolute adherence to the new means he had devised for himself, with cues from Schoenberg, and by steady exploration of their possibilities. He never worried that his music, in essential respects, sounded quite unlike anything that had come before or was being written around him. He just went on, with exemplary persistence. He had no way of knowing that the vacuum in which he worked would rapidly be filled after his death, not least by Stravinsky, who learned a lot from his music in the 1950's, but also by many younger composers.

Among the eager Webernians then was Pierre Boulez, who returns to be the mastermind of the new recordings, just as he was 30 years ago for a set made by CBS, now available on CD from Sony Classical. But there are differences. One is that the new box (Deutsche Grammophon 457 637-2; six CD's) is twice as large, including many works Webern withheld from publication.

Some of these are juvenilia, imparting the unsurprising news that the composer at 16 was a talented, hopeful, somewhat incompetent beginner. His later rejects, though, include wonderful pieces, especially among the songs and instrumental movements he wrote in 1913 and 1914. During that period he gave thought to a sequence of orchestral pieces, some with solo soprano, rather in the manner of a distilled Mahler symphony. There might have been a similar string quartet with voice. Much later, though, Webern decided to issue sets of purely instrumental movements: the Six Bagatelles (Op. 9) for string quartet and the Five Pieces (Op. 10) for orchestra.

This left out of account not only the song movements -- two with orchestra and the one with quartet are breathtaking -- but also quite a number of orchestral movements. Mr. Boulez includes five, and two extra bagatelles.

A CD player can be programmed to present, say, Opus 9 followed by the two unselected bagatelles and, to end, the song with quartet, which not only provides a passionate slow finale but also leaves a clue to the music's expressive core, in a sense of grief and loss. Similarly, one can reconstruct a vocal symphony, which would have to include an alarming brassy piece (No. 3 among the additional orchestral movements) and the magical setting of a Stefan George poem, with its delicious spot for voice and percussion, and its penultimate gesture of a huge rise from the singer on the word ''holy.''

These and many other pieces sound, here, marvelous to perform. All the string quartets and trios are played by the Emerson String Quartet, which, strong and expressive, makes every little miniature sound big. Nearly all the songs, and the soprano parts in the cantatas, are sung by Christiane Oelze, for whom the music seems to have been waiting. Defying gravity, Ms. Oelze moves with ease through the enormous pitch intervals Webern loved and makes them beautiful and true, keen points in the continuing phrase and markers of exaltation or anguish.

Her contributions include, happily, all the songs with piano, which again embrace remarkable items Webern did not publish: the Five Dehmel Lieder of 1906-8, right on the bright moonlit borders of atonality; and 4 George songs from the next year, in addition to the 10 published as Opuses 3 and 4. Webern changed his mind about the planning of these cycles, eventually deciding that each should have an introduction followed by four songs in which the singing persona's feelings are reflected in nature (Op. 3) or in a tragic relationship (Op. 4). The numbers thus omitted are well worth hearing, especially when sung so well -- and played so well, by Eric Schneider.

Among other exceptional pianists at work here are Pierre-Laurent Aimard, in the quartet with saxophone and the concerto, and Krystian Zimerman, in the Variations (Op. 27) and two other pieces. Mr. Zimerman gives a beautiful account of the variations: the finale, highly effective, has wide-spanning melodies, often violent and gentle in the same breath, searching in a musical space that comes to be defined by chord resonances in the background.

But of course the performer who figures most prominently and comprehensively is Mr. Boulez, as conductor. To an astonishing degree, his tempos are close to those of his earlier recordings. Yet consistency of timing is deceptive, for within identical spans a lot has changed. Mr. Boulez is working here with the Berlin Philharmonic and the Ensemble Intercontemporain: musicians who have a much fuller appreciation than anyone did three decades ago of Webern's flow, dependent on subtle phrasing and a chamber-musical responsiveness of one player or section to another.

Where, for instance, the arrangement Webern made of the six-part ricercar from Bach's ''Musical Offering'' almost fell apart in the 1969 London Symphony recording, the new version is secure and even imposing. The earlier performances of the original works often sound scrappy and preliminary, and though the first clarinet in the symphony achieves some suavity in rather torn textures, the same part emerges in the later recording far more gorgeous and sensitive.

Inevitably, there are losses as well. The spikiness Mr. Boulez found in this repertory when he was in his mid-40's was not just a result of unconsidered playing: it came from a conviction that the music was fierce and that it mattered. Witness, for example, the swing between aggressiveness and recuperation in the second movement of the symphony, or the way the choral women in the third movement of the Second Cantata seem to sing with teeth bared, like Valkyries. These moments are more beautiful in the later recordings but not necessarily more exciting.

There are also performances in the earlier set that will not be supplanted, like the account of the two Rilke poems, with Heather Harper, or the many appearances of another soprano, Halina Lukomska, whose flaming voice is so different from Ms. Oelze's but equally apt.

Something else has happened to Webern during the last three decades: we know far more about his life, and about his opinions, which were not all edifying. His pursuit of purity in his music -- of complete homogeneity and integration, of absolute precision in the minutest detail -- and the high value he placed on German culture led him, crazily and dismayingly, to consider that his ideals were shared by the Third Reich.

He was not an anti-Semite. Indeed, he helped conceal Jews in Vienna. But he seems to have thought that Hitler was some kind of agent of spiritual regeneration, and that the spreading Nazi conquests of 1939 to 1941 were all to the benefit of the nations overcome: this even though the annexation of Austria in 1938 had put an end to his activity as a conductor and to any hopes he had of hearing his music again, other than in neutral Switzerland. The Nazis closed his public career. And yet, privately, he applauded them.

Knowing this, we might want to listen to the Variations for orchestra of 1940 a little more carefully and a little more critically: to pay less attention, perhaps, to the coherence and symmetry the music so ostentatiously exhibits than to the delicacy, strangeness and variety of its component parts, and even to insist, contra Webern, on multiplicity and ambiguity as essential elements in his art.

More useful, too, than dismissing him for his foolish views would be to learn from his example of magnificent but, in crucial respects, misaligned idealism. Snowflakes and flowers are all very well, and we need them, but their rules arCDe not those of politics. (Paul Griffiths / The New York Times)

CD 1 / CD 2 / CD 3 / CD 4 / CD 5 / CD 6