Maurice Ravel (1875-1937)
Daphnis et Chloé (1912) [54]
Sinfonia of London and Chorus/John Wilson
rec. 2022, Church of St Augustine, Kilburn, London
Chandos CHSA5327 SACD [54]

Ravel’s Daphnis et Chloé as a complete ballet score has been a staple of the record catalogues since the earlier days of stereo, and has been singularly fortunate in the quality of many of the performances that have been enshrined on disc. One of the earliest of these was a celebrated Decca LP conducted by Pierre Monteux, who gave the stage première of the ballet and was given a superlatively engineered recording which has remained in the catalogue till this very day more than sixty years later. Which makes it all the more surprising when we learn that the original score was riddled with errors, attributable in part to the cavalier treatment of the music by the French publishers but also to the composer himself. Apparently during the original rehearsals Ravel would frequently give instructions for amendments to his scoring, accentuation and so on, as indeed would be expected in a score of this complexity; these alterations would then be incorporated into the individual orchestral parts, but were never incorporated into the full score or even noted. Monteux and his successors prepared extensive lists of these misprints and other mishaps, which were incorporated into their own performances; but it was not until the pandemic of 2020-21 that John Wilson actually sat down to codify these myriad discrepancies (apparently there were over 1000 of them!) and provide a score which reflected the composer’s intentions as evidenced in the first performances he supervised for Diaghilev’s Ballets Russes in 1912.

It might appear that the sheer number of these amendments would be immediately apparent to the listener, but in point of fact that is not the case. In the first place I suspect that many of the earlier recordings of the score will have incorporated some if not all of Ravel’s alterations. In the second place the sheer complexity of the orchestral writing will have served to conceal any actual inaccuracies in the notes themselves, let alone discrepancies in accentuation and other matters. The celebrated Lever du jour which opens the final scene of the ballet is probably the best archetype of impressionist textures in the orchestral repertory, with rapid washes of arpeggios in strings and woodwinds providing an aural equivalent of the dawn mist through which the sun slowly rises, and where pinpoint accuracy from any individual player is subordinate to their contribution to the overall effect; this style of writing, going back to Wagner’s Magic Fire Music, almost militates against any attempt to clarify the individual notes themselves.Some conductors, notably Pierre Boulez, have attempted to provide exactly that sort of clarity in their recordings of the score, with very mixed success. In his earlier Cleveland LP the CBS engineers aid and abet the conductor with some very peculiar microphone placements which produce totally unrealistic balances between sections of the orchestra, and these only serve to highlight some decidedly rough-and-ready playing from some soloists challenged by Boulez’s sometimes eccentrically fast speeds. When Boulez returned to the score for DG some years later, the recorded balances were much better judged but the speeds remained problematic.

Mind you, there might well be some concerns regarding precisely what speeds Ravel himself intended. Monteux, who presumably took his instructions from the composer, not only gives us one of the speediest traversals of the sunrise in the final scene, but then adopts a speed for the final orgiastic Danse finale which is well below Ravel’s metronome mark (and indeed the tempo adopted by most of his successors). Now the latter may well have been a response to the fallible playing of the London Symphony Orchestra, at the time going through a particularly bad patch in terms of orchestral standards; but the faster speed for the Lever du jour does have a positive advantage with the growing string melody forging onwards to its magnificent climax rather than climbing upwards with the more ponderous grandiosity that can afflict conductors who adopt a more measured approach. In the same way the very opening of the ballet, with its request for a very slow tempo followed by a gradual acceleration, is in itself somewhat contradicted by the fact that the note values in the score already achieve this accretion of speed without any added metronomic impetus.

That opening also serves to highlight a very particular problem with Ravel’s score, whatever revisions are made to it. And that is the contribution and nature of the chorus. Now it seems to me that the composer’s actual intentions are clear: that the chorus is to be identified with the corps du ballet on stage. At the beginning of the score they are marked to be off stage, behind the scenes; then, as the dancers enter, the voices are marked as drawing closer until they are stated to be “on stage”. Now even in concert performances this is of course hardly susceptible of realisation; and in theatrical circumstances it is totally impracticable. Ravel compounds the problem by his failure to specify the vowels his choir should vocalise to; and he seems at various times to expect smaller or larger numbers of voices, quite apart from the extraordinary passage between the first and second scenes when the chorus sing unaccompanied with offstage brass fanfares (also marked as growing closer as the music progresses). In staged performances the usual procedure is to employ a smallish number of vocalists who sing from the orchestra pit; but this of course sacrifices any attempt to portray voices at a distance in the opening scene and elsewhere. In the concert hall conductors can deploy a larger number of singers seated in serried rows behind the orchestra; but then there is a danger that the voices can overwhelm the orchestra at climaxes – especially since Ravel writes for sopranos and tenors in their highest registers – when they should clearly form part of the overall texture. It is perhaps not so surprising that Diaghilev commissioned Ravel to provide an alternative version of the score dispensing with the chorus altogether; Ravel objected strenuously to what he regarded as a judgement made purely on economic and financial grounds, but there is perhaps a case to be made for occasional performances adopting this expedient especially since Ravel did supply a rather beautiful orchestral interlude to substitute for his choral transition into the second scene.

There remain, too, some other minor considerations which can cause problems. In the Lever du jour Ravel asks for some woodwind phrases imitating birdsong during the dawn chorus to be played by solo players on stage. Again, this immediately causes problems; if the instrumentalists are actually situated in the middle of the dancers, this can play havoc with the choreography, but if they are placed behind the scenes then their chances of being clearly heard are reduced to near zero. In the same way the wind machine, which Ravel employs almost as an abstract constituent of the percussion section rather than as an atmospheric effect, needs to be at once clearly audible and to sound from within the orchestra; some conductors seem to regard it almost as an excrescence, reducing it to near inaudibility. All of these difficulties, of course, can be readily overcome with a recording that employs microphone placement imaginatively and effectively, and engineers that are prepared to engage with the composer’s expressed wishes.

So, how does John Wilson’s actual recording measure up to these exacting standards? Well, in the first place, the actual orchestral sound is certainly clarified and the impressionist wash of sound is more translucent and pointillistic. There is more of the element of classical Greece, as exemplified for example by Satie’s Gymnopédies and Debussy’s Danseuses de Delphes, than we have grown to expect in this score. Even the passages for open natural horn arpeggios and flute harmonics, notoriously difficult to keep in tune, sound poised and delicate rather than rustic. Wilson, unlike Boulez, resists the temptation to create excitement by whipping up the tempo – for example in the dance for the pirates – and on the other hand he avoids the danger that the very opening might become becalmed by moving steadily forward as the music becomes more defined. He uses a small chorus, around the size that might fit into an orchestra pit during a ballet performance, but specially chosen for strength of tone and security of pitch without the danger that the voices might overwhelm the orchestra at times when they are clearly to be regarded as forming part of the overall texture (as in the final pages). Like Monteux, he keeps the music moving forward during the Lever du jour, but the recording engineers make sure that the solo woodwind lines onstage are heard in a distinct and different acoustic. His wind machine is occasionally a little too discreet for my taste, and seems to be somewhat detached and insufficiently awestruck during the intervention of Pan at the end of Scene Two; but that is a very minor fault.

One should mention the more recent version by Roth issued some ten years ago, which sought to engage with the score through the use of period instruments; but in its concentration on the actual orchestral sound the dramatic elements were perhaps short-changed, and some of the internal balances seem less than convincing. Many of these same issues were addressed in Dutoit’s CD recording issued some forty years ago, which like the new Wilson recording clearly sought to set new standards for a work that – as I hope I have made clear – is probably incapable of a comprehensively ideal rendition. Monteux’s cautious slowness in the final dance, whatever may have been its cause, is a problem for me; and I have never heard a performance that is satisfactory in every respect. But this is a major contribution to the catalogues; and at least in the future no conductor will have to put up with a score that misrepresents the composer’s intentions. I am delighted to see that the newly engraved score is being published (an extract is included in the booklet). I reviewed this SACD on my normal system with much enjoyment, and I suspect that on enhanced equipment it will sound even better. Booklet notes come in English, French and German, and include a detailed scenario of the stage action linked to plentiful listed cues. Highly recommended, despite the lack of a coupling which makes for rather short measure by modern standards.

Since writing this survey, my attention has been drawn to a video review by David Hurwitz on Classics Today in which he quite intemperately attacks this release, describing it as “the worst ever.” Now, that description is a very large claim – I have certainly heard some extremely bad recordings and performances of Daphnis which would better qualify for that title – and the ill-tempered Hurwitz almost totally ignores many of the problems which I hope I have identified with this score, and Wilson’s attempts to address them – whether or not you would regard these as successful. I am much gratified by Wilson’s engagement with the score and its difficulties, and find his solutions to a very large degree fulfilled.

Paul Corfield Godfrey

Help us financially by purchasing from