Less than half the songs in this almost complete edition (only a couple of early pieces written specially for bass voice are omitted) were published and given opus numbers by Webern himself. Well over half date from early in his career, and a high proportion of them are tonal, or at least have strong tonal implications. Yet it would be a great mistake to assume that the majority are uncharacteristic juvenilia. The very earliest of them, indeed, written before Webernhs 16th birthday (his first meeting with Schoenberg as yet far in the future; he had still to hear his first Mahler symphony), is quite astonishingly typical of him: in its brevity and economy (a mere 22 bars, lasting barely a minute; very few notes, and not one of them superfluous), its quietness (it begins pp, retreats to ppp and ends ''as gently as possible'') and its extreme refinement of expression—its subject is not spring, nor the end of winter, but the transition between them, when ''the land no longer lies in deep sleep but a light slumber, in which the early blackbird's song is already awakening lovely dreams of morning''. No less prophetic are the beautiful logic of the song's construction from its opening four-note idea and, strangest of all, the fact that Webern has used the few banks lines that remain at the end of his sheet of manuscript paper to draft an arrangement of the piece for voice, oboe, two horns and harp (a characteristic ensemble for a mature Webern work, but where can he have seen or heard a model for it in 1899?). It is as though all the essential features of his style were there from the beginning, to be polished and perfected but not significantly altered by the momentous encounter with Schoenberg.
But if none of these songs is in anything but the literal sense 'early', only a few sound really late. The songs of Op. 4 (dating from the end of the years of study with Schoenberg) sound conspicuously less 'advanced' than their immediate predecessors of Op. 3 (one of the Op. 4 set amazes us by lasting—and earning, with its dark gravity of expression and its unhurried melodic arches—a full 4'16''; we are reminded that it was at just this time that Webern was considering writing an opera), and a good deal less so than his instrumental works of this period. The Lied itself seems to have insisted upon these vocally grateful phrases, these chords that may not confirm but certainly do not deny tonal inference. As late as Op. 12 he can still, in song at least, write a skeletal but far from serial-sounding quasi-folk-tune to accompany a naive folk text or, to illustrate one of Hans Bethge's adaptations of Chinese verse, evoke a distant flute with a vivid onomatopoeia thatMahler would have enjoyed. Only in the Hildegard Jone settings of Opp. 23 and 25 do we encounter the intricate purity and bracing rigour of the late cantatas, and yet these qualities are audibly implicit in all that went before.
Although the insights this set offers into the development and essential consistensy of Webern's language are fascinating and instructive, it is not, I suspect, for listening at a sitting. The expression is often very concentrated and epigrammatic (you need to raise the stylus and think about each song before proceeding) and many of them are rather slow. Hearing the same voice in all of them does not help. Dorow and her pianist have an assured grasp of Webern's style, and her intimacy of manner is often just right, but the resulting sound is sometimes shallow and small-scaled: I could have done with a bigger voice or a darker one occasionally (two of the published sets specify a medium voice, not a high one). The very close and unatmospheric acoustic, with the voice seemingly not quite in focus, also tends to miniaturize the music somewhat. But it is a set that I am grateful for, none the less, and shall return to it often.'