Harnoncourt, like Stefano Mollo for Abbado, has researched Schubert's own manuscripts, and noted that many unauthentic amendments have found their way into the printed editions of the symphonies. Some of Harnoncourt's corrections are already familiar from the Abbado set, such as the removal of the eight bars later added to the Fourth Symphony's first movement exposition; but the differences between Harnoncourt's 'findings' and Mollo's are puzzling (Harnoncourt does not incorporate Abbado's corrections in the slow movement and Scherzo of the Ninth Symphony). A few of Harnoncourt's corrections are listed in his booklet, though nowhere near as many as appear in DG's notes for Abbado, which is a pity: it would be helpful to know whether the sources of the many startling moments in Harnoncourt's cycle are textual or interpretative; for instance, is the Ninth Symphony's last chord, here extended with a long diminuendo, Schubert or Harnoncourt?
That dying away of Schubert's last symphonic sound is a surprising conclusion for a finale—unlike Abbado's, a whirling, spinning vivace—borne aloft on astonishingly precise articulation of its rhythms and accents, and a springy delivery of the triplets. Characteristics, of course, one has come to expect from an Harnoncourt performance. Still, what a joy to hear this Allegro, and those of most of the earlier symphonies, seized with such bright and light-toned enthusiasm. Here is urgent, virile and vehement playing, never over-forceful, over-emphatic or burdened with excessive weight. What came as a surprise was the consistent drawing out of these scores' potential for sadness and restlessness. Harnoncourt does not set apart the first six symphonies as merely diverting (out-and-out charm is seldom part of Harnoncourt's Schubertian vocabulary): their bittersweet ambiguities and apparent affectations of anxiety here acquire a greater significance, and the cycle, as a whole, a greater continuity.
Up to a point, the darker, more serious Schubert that emerges here, derives from the type of sound Harnoncourt fashions from his orchestra; not least, the lean string tone and incisive brass. And maybe, up to a point, from the corrections: Harnoncourt refers to the manuscripts as often being ''harsher and more abrupt in tone [than the printed editions], juxtaposing extreme dynamic contrasts'', though you can't help feeling that contrasts in general have been given a helping hand. Trios are mostly much slower than the urgent minuets/scherzos that frame them (with pauses in between the two). And Schubert's less vigorous moments are very noticeable as such, and are inflected with varying degrees of melancholy—it is uncanny how the string playing, in particular, often suggests a feeling of isolation (along with the sparing vibrato is an equally sparing use of that enlivening facility: staccato). Even the Second Symphony's perky Trio where, incidentally, Harnoncourt has woodwind turns not trills, seems tinged with a certain regret. Abbado's players sound a great deal more game. As they do in the finale of the Sixth Symphony; the only movement where Harnoncourt's approach is ruthlessly extreme: this begins as a deadpan moderato (the marking is Allegro moderato), and as levels rise, so, sharply, does the tempo; if Schubert intended this movement as a tribute to Rossini, here it seems a rather blatant satire.
Tovey's comment on the Fifth Symphony—''like a delightful child overawed into perfect behaviour... by sheer delight in giving pleasure''—here seems very wide of the mark. Norrington, before Harnoncourt, demonstrated this apparently sunny symphony's capacity for a driven Sturm und Drang restlessness; Harnoncourt's is considerably less driven, but Harnoncourt's 'child' is reluctant to join in: the high flute at the vigorous end of the first movement's first subject could almost be a cry for help. As for the finale's second theme ''running along merrily'' (Tovey), I'm tempted to observe that it's partially lamed by Harnoncourt's preceding stinging fortissimos; the music's minor-key episodes taking their toll on the attempts at merriment as the movement progresses to its hardly convincing major-key conclusion. And why not? Significantly, Schubert marks only a single forte at the finish.
The Unfinished Symphony's first movement is a stark, harrowing experience (yet it remains a well-tempered musical one: gestures are never exaggerated); the opening is as cold as the grave itself; the second subject knows its song is short-lived (curiously, the transition to the second subject seems to have shed a bar, and the last two bars of that cello theme start with quavers, not crotchets). In both movements, the elucidation and balance of texture can only be described as masterly: just listen to the trombones casting shadows in both codas.
This, then, is as seriously pondered, coherent and penetrating a view of the complete cycle as we have had. Whether or not you feel Harnoncourt focuses too much on Schubert's darker side (a feature I have probably overstated in this review), you have to marvel at his ability to realize his vision. And if this great orchestra were not convinced by that vision, I doubt that it would have co-operated as wholeheartedly. The recorded sound offers that inimitable Concertgebouw blend of the utmost clarity and wide open spaces. Abbado's set walked away with a 1989 Gramophone Award; had Harnoncourt's set been available at the same time, then...