Extremely few people will ever have heard Clovis et Clotilde, the cantata with which the 18-year-old Bizet won the Prix de Rome. It was given a single performance on the day of the awards in 1857 but remained unpublished, and was resuscitated—by Jean-Claude Casadesus and the artists in the present recording—only in 1988, on the occasion of the 150th anniversary of Bizet's birth. It deals with the conversion to Christianity of Clovis, king of the Franks—to the delight of his wife Clotilde and her father, an archbishop—after winning the battle of Tolbiac in the fifth century. One has, somewhat wryly, to admire Erato's even-handed appraisal of the work: in the booklet the commentator remarks that ''All in all, there are few prophetic flashes of genius to be seen in it'', while the conductor writes that he ''was over-whelmed to find a lyrical, sensitive, glowing and youthfully vigorous work that beyond the slightest doubt heralded the arrival of a future genius''. Incumbent as it is for an interpreter to champion a work he is performing (and patently, Casadesus's reading is deeply committed), I can't help feeling that advocacy has clouded his critical judgement. There are, certainly, good things in this youthful work—some pleasant melodic lines, a Gounodesque prayer, an oboe obbligato in the second verse of an otherwise unremarkable Romance for Clotilde—but much is conventional (perhaps Bizet was being prudent in avoiding what might upset the Prix de Rome judges?), and the love duet of the eponymous pair is frankly hammy. Garino, both lyrical and heroic, is much the most pleasing of the soloists; the steadiness of Martinovic's line throughout is marred by a throbbing vibrato; and Caballe, very much the prima donna, had obviously not given enough thought to the relation of verbal-meaning stress to phrase-shape, so that we get, for example, ''par la victoire'' and ''pareil a l'ouragan''.
Roma, which Bizet thought of as a symphony but was published (after his death) as a concert suite, was conceived at the end of his time at the Villa Medici, though he continued to tinker with it for the best part of a decade. Admittedly uneven in quality, it nevertheless contains too much skilful writing and colour to be ignored, though he was curiously misguided about its good and bad points (he was gleeful, for example, about the middle of the Andante and the finale's second subject being identical, despite that theme's weak, sanctimonious, sub-Gounod character). The Lille orchestra's performance, basically praiseworthy, suffers from very weak violin tone—not, I think, due so much to the players as to disadvantaged microphone placing. This is already noticeable in the first movement at the return of the initial theme, where the sliding strings are barely audible at first; it becomes more serious in the scherzo, which needs more 'presence' if it is to sparkle as it should; and in the finale the violins are completely submerged at the move to E flat (00' 24'' in) and the parallel place in the recapitulation. Like Gardelli on Orfeo/Harmonia Mundi (but not Fremaux on EMI), Casadesus follows the score in giving the nightingale-like arabesques at the end of the Andante to the violin section as a whole (which executes the floridities very neatly), and there is a finely controlled soft ending to the movement; but the carnival finale is disappointingly tame—Fremaux's rowdily energetic reading is much more fun—and for all the notice that is taken of the cross-accents in its march theme Bizet might have saved himself the trouble of writing them. Though the Fremaux version of Roma dates from 1974, it is still the most recommendable (see also CH on page 42); but the main interest of this disc lies in the first recording of the cantata.'