There seem to be two schools of thought among conductors about how long the first movement of the Franck Symphony should last. Plasson (EMI) and the older generation of Beecham (also EMI) and Monteux (RCA) take under 18 minutes over it; Chailly agrees with Karajan (EMI) on a timing of well over 20 minutes. Like Karajan's his is a big performance, and not only in duration: the orchestral sound is massive (though by no means congested: the details of Franck's woodwind scoring are perfectly clear throughout), the crescendos are huge, the proportions of the architecture monumental. His is the slowest account of this movement that I have heard but there is always a perceptible sense of movement, often a most impressive one. He takes almost as much care as Karajan does over dynamic markings, but to rather more austere effect, solemn rather than dramatic. His approach to the other movements is consistent with this: there is more earnestness than grace to his (also slowish) Allegretto, and his finale is soberly serene, without much hint of dramatic urgency, let alone jubilance. At the opposite end of the spectrum is Beecham, his first movement about three and a half minutes shorter than Chailly's, his orchestra sounding almost unkempt by comparison (Beecham knew how French brass players love to blare in Franck's more purple pages, and he gave them their heads; the rather scrawny recording emphasizes this) but with such vitality and fire (and in the coda such powerful, almost tragic drama—one is reminded of the first movement coda in Beethoven's Ninth even) that we seem to be listening to a different symphony. Which of course we are: Beecham emphasizes the Gallic elements in Franck (the ardour of the Allegretto, the exultancy of the finale), Chailly the Teutonic; both are present and both are important, but which is to preponderate?
Monteux's approach is also Gallic, of course, with a tension to the first movement that is close to Beecham's and a tender expressiveness to the Allegretto, but he is let down by a coarse recording that makes both outer movements sound raw and strenuous. Plasson belongs to the Gallic group, and is better recorded than either Monteux or Beecham, but in their company his performance seems under-inflected and characterless. On the Teutonic wing Chailly sounds like Karajan without the extremes: he avoids any hint of brutality in the massive tuttis or of heaviness in the finale, but he also lacks Karajan's refinement of phrasing, his nervous energy. Should you wish for compromises Bernstein (DG) strikes me as a pretty happy medium (signalled by the fact that he breaks the 'first movement rule' adumbrated above: beginning slower than anyone, but with very strong contrasts of tempo, his timing is just over 19 minutes). There is an occasional touch of histrionic vulgarity to his reading, and it is not quite the French vulgarity that Beecham enjoys so much, but for my money his is the reading in which the Gallic and the Teutonic in Franck are most in balance.
As to couplings, Beecham and Monteux offer nothing at all, Bernstein only a rather over-blown account of Saint-Saens's Le rouet d'Omphale. Of those who sensibly and quite generously couple the Symphony with the Variations, Chailly has the most striking soloist. Bolet gives a hugely grand performance, decidedly slow in the later variations, but with a marvellous clarity of part-writing and a great richness of piano sonority (the recording makes his piano sound about ten yards wide). Collard, with Plasson, is more poetically sensitive, more sprightly, less commanding, and he is recorded in a more natural perspective. Weissenberg has his showy and delicate moments, but it is not only his recessed recording that makes him sound the junior partner to Karajan in this rather drilled reading. Even so, if you insist on this coupling Karajan it must be (recently transferred to CD on EMI's Studio label CD CDM7 69008-2—to be reviewed later); if not, and you want a modern recording also with the option of CD, Bernstein. But I must say that if I were buying just one recording of the Symphony, and one that convinced me that the first movement needs to be 17 minutes long (let alone 20) it would be Beecham's.'