It is difficult to conceive of a masterpiece in any other form than it is. The impression the listener receives from Sibelius's Fifth Symphony—or The Rite of Spring or La mer — must convey what I think Schoenberg called the illusion of spontaneous vision. It is as if the artist had caught a glimpse of something that has been going on all the time and that he has stretched out and effortlessly captured it. One of Sibelius's letters written to his friend Axel Carpelan in the autumn of 1914 puts it perfectly: "God opens his door for a moment, and his orchestra is playing Sym. 5". But of course it is not at all like that as we know from the Beethoven sketchbooks and from the fact that Sibelius worked for seven years (1912-19) before the Fifth Symphony reached its final form. Sibelius was nothing if not self-critical and a number of his works underwent their birthpangs in public, including En saga, the Lemminkäinen Legends and, of course, the Violin Concerto. The main theme came to him much earlier than 1903 and he recognized it for what it was, an inspired idea which remained unchanged. After its first performance in Helsinki in 1904, in which as Tawaststjerna put it, "a red-faced and perspiring Viktor Novácek fought a losing battle with a solo part that bristled with even greater difficulties... than in the definitive score", Sibelius decided to overhaul it. He realized the necessity to purify it, to remove unnecessary detail or ornament that impedes the realization of a cogent structure. Tawaststjerna goes into great detail about its genesis and the puzzling way in which Sibelius behaved towards Willy Burmester, who was to have given its premiere première (Sibelius Vol. I; London: 1976, pages 270–87). In its finished form, it was given in Berlin with Karl Halir as soloist and Richard Strauss conducting.

Listening to Sibelius's first thoughts played with great virtuosity and excellent taste by Leonidas Kavakos and the excellent Lahti orchestra is an absorbing experience. One is first brought up with a start by an incisive orchestral rhythmic figure at what would be fig. 1 of the 1905 score (track 1, 1'23") from whereon the orchestra plays a more assertive role in the proceedings. In the unaccompanied cadenza 21 bars later (2'11") there is some rhythmic support while to the next idea on cellos and bassoon (fig. 2), the soloist contributes decoration. And then at seven bars before fig. 3 (3'41") a delightful new idea appears which almost looks forward to the light colourings of the later Humoresques. Although it is a great pity that it had to go, there is no doubt that the structural coherence of the movement gains by its loss both here and on its reappearance (14'03"). It is the ability to sacrifice good ideas in the interest of structural coherence that is the hallmark of a good composer. There are other changes Sibelius must have regretted though not, I suspect, the second long and unaccompanied cadenza which bristles with fascinating difficulties but whose removal greatly improves the overall shape of the movement. There are some hair-raising difficulties early on in the finale and I rather regret the disappearance of the delightful idea starting at about 1'30". The fewest changes are in the slow movement which remains at the same length. As in the case of the Fifth Symphony, where the revision is far more extensive than it is here, the finished work tells us a great deal about the quality of Sibelius's artistic judgement, and that, of course, is what makes him such a great composer.

This disc offers an invaluable insight into the workings of Sibelius's mind, and I have to say that even in its own right, the 1904 version has many incidental beauties to delight us. Kavakos and the Lahti orchestra play splendidly throughout and the familiar concerto which was struggling to get out of the 1903–04 version emerges equally safely in their hands. The BIS team have put us greatly in their debt by making the two versions available for study side by side and I hope they will go on to do the same for the Fifth Symphony.