Prokofiev was always loath to waste good ideas and both the Third and Fourth Symphonies derive from other sources; the Third reworks ideas from the opera, The fiery angel on which he had finished work in 1925, and the Fourth from his ballet, The prodigal son of 1929. Plans to put the opera into production were never realized; hence his determination to rescue some of the material from obscurity by using it in a symphony. The Third Symphony has successfully led a life of its own: Prokofiev was always at pains to stress that it was a self-contained work and, unlike Hindemith in Mathis der Maler, did not link the two works together with a title. The Symphony No. 4 does not stand on its own two feet in quite the same way, and its balletic origins are obvious, both in terms of the melodic substance and its organization. Prokofiev himself recognized this and drastically revised the score later in life. The extent of this overhaul, made immediately after the completion of the Sixth Symphony, can be gauged by the fact that the 1930 version takes 23'08'' and the revision 37'12''. The first two movements are much expanded; indeed, in playing time the first is doubled, and the Andante is half as long again. The orchestration is richer and among other things a piano, which Prokofiev had used in the Fifth and Sixth Symphonies, is added. Not all of Prokofiev's afterthoughts are improvements. The Sinfonia Concertante, Op. 125, for example, is not always superior to the Cello Concerto, whose material it reworks, and the revision of the Fourth Symphony is not (to my mind at least) in every respect an improvement on the original. In any event, the couplings here make excellent sense: those who want the original version of the Fourth Symphony have it coupled with No. 3, while the Classical Symphony, which most collectors have anyway and will not want to duplicate, is coupled with the revision.
Neeme Jarvi's account of the Third Symphony can more than hold its own with the opposition. Chandos accommodate it on one side, as did Philips in their version by the Concertgebouw Orchestra under Kondrashin though theirs derives from a broadcast and is far less successful as a recording. Moreover, extraneous noises which are inevitably a hazard in the concert hall, rule this version out of court. Riccardo Chailly's DG studio recording with the German Youth Orchestra is remarkably accomplished but does not displace the classic Abbado/Decca version, either as sound or as a performance. Perhaps the Decca recording despite its relative age has marginally greater transparency: the part-writing registers more effectively there, whereas the Chandos recording has a somewhat more reverberant acoustic. Neeme Jarvi is particularly successful in the Andante, which derives from the last act of the opera, and succeeds in conveying its sense of magic and mystery. The opening idea is austere but lyrical, and the music to which it gives rise is extraordinarily rich in fantasy and clothed in an orchestral texture of great refinement and delicacy. In the scherzo, said to have been inspired by the finale of Chopin's B flat minor Sonata, he secures a very good response from the SNO strings. The writing for divisi strings is extremely demanding as well as highly original and in its whirlwind of activity it evokes the strange, supernatural and 'possessed' atmosphere of the opera.
Along with Symphony No. 2, the Fourth is the least often recorded of the Prokofiev cycle: Georges Sebastien on Urania (not generally available in the UK) and Martinon on Turnabout (TV370525, 6/74—nla) are the only others I can think of who have given us the 1930 edition. If you came to this version first, the 1947 revision seems very padded-out, particularly in the slow movement and the scherzo. On the other hand, I suspect that if one had made the acquaintance of the expanded version first, the original would seem to be more like a ballet suite than a symphony and with insufficient tonal contrast. Neither has the degree of organic cohesion or symphonic drama of Nos. 5 and 6, but both have delightful and characteristic ideas. The scherzo, drawn from the music for the Temptress in the ballet, is particularly felicitous. Prokofiev at his second best is infinitely superior to many of his contemporaries at their very best. Jarvi succeeds in making out a more eloquent case for the revision than either Rozhdestvensky and the Moscow Radio Orchestra (HMV SLS844, 5/73—nla) or Weller and the LPO (Decca SXL6908, 10/79—nla).
Neeme Jarvi gives an exhilarating account of the Classical Symphony, one of the best on record. The slow movement has real douceur and the finale is wonderfully high spirited, so much so that my thoughts even turned to Koussevitzky's pioneering records on which my generation was nurtured (HMV D1857/8, 10/31), and the Scottish National give an uncommonly good account of themselves. There are some fine versions of this enchanting score on the market, among which I much like Malko and the Philharmonia (HMV) and Marriner (Decca) but there are others of excellence, and Jarvi's figures high among them. Both records, then, are to be strongly recommended and though Nos. 3 and 4 will be an obvious first priority for many collectors, I was more persuaded by Op. 112 than ever before.'