o Machaut Messe de Nostre Dame. Ballades — Plourez dames; Nes qu'on porroit. Rondeaux — Sans cuer dolens; Dis et sept cing; Puis qu'enoubli. Le lay de bonne esperance. Oxford Camerata / Jeremy Sununerly.

Naxos CD CD 8 553833 (78 minutes: DDD). Texts and translations included. Recorded in association with BBC Radio 3 and France Musique.

Machaut Messe de Nostre Dame (including plainchant appropriate for the Feast of Purification of the BVM). Ensemble Organum / Marcel Peres.

Harmonia Mundi 0 0 HMC90 1590 (57 minutes: DDD). Text and translation included. Also includes a sampler CD containing a 'portrait' of the Ensemble Organum.

Messe de Nostre Dame — selected comparisons:

Taverner Consort, Parrott (8/88) (EMI) CDC7 47949-2 Hilliard Ens, Hillier (2/90) (HYPE) CDA66358

I seem to be reviewing Mass recordings in pairs at the moment. This particular work needs no introduction, however. Machaut's Mass is unquestionably the most famous composition of the fourteenth century, and these two discs only add to an already substantial discography. Both have undoubted selling-points, and besides one could hardly imagine more contrasted approaches — though in line with performances by the Taverner Consort and the Hilliard Ensemble, both the Ensemble Organum and the Oxford Camerata employ voices alone. The former (like the Taverner Consort) present the work in a liturgical context with very fine plainchant interpretations, and the latter (like the Hilliard) complement the Mass with a selection of Machaut's songs.

Summerly's reading has the advantage of having been recorded in the very building where the Mass may first have been heard more than 600 years ago — Rheims Cathedral. How much of an advantage that is depends largely (I suspect) on the listener's turn of mind; in any case, the claim in the booklet-notes that this is the closest we will get to actually hearing what Machaut heard leaves me rather sceptical. Never mind; this is a clean, wellbalanced rendition, using solo voices, like the Taverner Consort: but (unlike that group) with countertenors, rather than high tenors, on the top lines. The tone-quality here is more restrained than

Parrott's. Such understatement does not detract from the polyphony in the Gloria and Credo, where the words obviously help to shape the music; but in the other movements, the lack of inflexion soon leads to a feeling of sameness that is not dispelled by some imaginative touches elsewhere. There is neither quite the sharpness of Parrott's account (still my preferred choice), nor the polish of the Hilliard's reading, nor finally the adventurousness of that of Peres.

For Peres Machaut is profoundly weird. It would exceed the scope of this review to detail its many eccentricities, but the main one must be the vocal colour adopted by Peres's singers. In his last few recordings, he has worked with traditional Corsican singers whose highly distinctive timbre and approach to ornamentation he has taken as the basis for his explorations. How far the ensembles have evolved as a result can be gauged by listening to their recording of the Tournai Mass, done some six years ago. There the singers (led by those magical Catalans Josep Benet and Josep Cabre) were classically trained, and though their sound was distinctive enough, it could hardly be described as bizarre. In this particular case, I suspect, a few listeners may have to overcome a powerful urge to switch off pretty early on.

Which would be a great shame. Notwithstanding the obvious misgivings one might have (approach to musica ficia, to ornamentation, to plainsong intonation, and problems of ensemble), Peres's reading makes a point that is so often conveniently ignored: we have no idea what Machaut's singers actually sounded like, or how they produced the sound in their throats. Peter Phillips once made that point, envisaging the possibility that we might find the 'authentic' sound unbearable. As I have got used (slowly) to Organum's sound, I have been reminded how far Machaut's world is from our own. This recording questions a fundamental and untestable assumption about medieval polyphony. As such, it is an intriguing alternative to other all-vocal performances, even if there are too many other imponderables to warrant an unconditional recommendation. FF