François-Joseph Gossec : Requiem. Bernadette Degelin (soprano I), Greta de Reyghere (soprano II), Howard Crook (tenor), Kurt Widmer (bass), Maastricht Kamerchor Conservatorium, Musica Polyphonica, Louis Devos. Erato ECD 75359 (1988), barcode 3269657535922 (also found under barcode 089088535925)
Recorded April 1986, Eglise Sainte Marguerite, Béguinage de Lier (Belgium)
Erato 2292-45284-2 (1989 ?), barcode 022924528422:
Apex 8573 89234 2 (2001), barcode 685738923427):
Erato 2564 69709 9 (2007), barcode 825646970995:
Also reissued in a compilation set (4 CDs), Erato 2292-45703-2 (1991?), barcode 022924570421: Grands Requiems (with Mozart, Campra, Charpentier):
Mozart? Berlioz? Verdi? Fauré? No: Gossec
Originally posted on Amazon.com, 6 September 2011
I remember. It must have been in the late 1980s or early 1990s (probably, in fact, on the occasion of the celebrations of the bicentennial of the French Revolution), a live performance of Gossec’s Requiem was given in Paris – it what church exactly I’m not sure. I knew and admired the Requiem, first from hearing it and taping it off the French radio, and then from this Erato recording of Louis Devos, first released in 1988. I had a trio of friends, very musical, one even one of the major young French cellists of then and today, and I strongly encouraged them to come with me, promising them that they were in for a great discovery, music as beautiful as any Mozart Mass or Requiem. They were skeptical before – and scorningly derisive after. They laughed at me, and at Gossec. In turn, I was shocked. Was it I that was deluding myself as to the beauties of the music, seeing gold where there was only muck? Or them who, for whatever reason, were simply unable to hear?
Years later, with much more listening under my belt – or, rather, in my ears – returning to the Requiem, I am ready to staunchly assert: they were simply unable to hear, and it was a tragic but classic case of listening to the reputation, and not to the music. In 1989 still, it was possible, even for performing musicians, to believe, consciously or unconsciously, that posterity knew better, and that if Gossec wasn’t a recognized composer, and if Harnoncourt or Gardiner didn’t conduct it, then it necessarily meant that his music coulnd’t be good. So much has been unearthed, recorded and rediscovered since, I should hope that we now know better: posterity is a fickle and oftentimes arbitrary mistress (they’re still my friends, though).
François-Joseph Gossec’s Requiem (“Missa pro defunctis” or “Messe des Morts”), composed in 1760 when the composer was only 26, signaled his breakthrough to public success. The influence of Gossec’s Requiem on Mozart’s has often been noted (just compare the Dies Irae, track 7, and Mozart’s Rex Tremendae), but listen to the Tuba Mirum, track 8, with its ominous antiphonal trombones placed at a distance: now I know where Berlioz got the idea from in his own Tuba Mirum from the Requiem: it is tantamount to plagiarism! The melismata of the “Te Decet Hymnus” (track 3) recall Mozart’s Mass in C, and the violence of “Mors stupebit et natura” (track 9) falls somewhere between Haydn’s Nelson-Mass, Beethoven’s Glückliches Fahrt and again Berlioz’ Requiem. And, why not? I hear adumbrations of Verdi’s Requiem in the lyricism of the Recordare (track 11), and of Fauré in the appeased and, respectively, ethereal and pastoral quality of the “Requiem Aeternam” (rack 2) or “Pie Jesu Domine” (track 22).
The – literally – tremulating voices on “Quantus tremor” is a masterstroke (“What trembling there will be / When the Judge shall come”, at 1:39 into the Dies Irae, track 7). The plangent Recordare (with trio of soprano, tenor and baritone, track 11), with its ostinato, chaconne-like basso depicting (my interpretation) tears dropping, is the kind of music that can draw tears from me: if you know Marin Marais’ extraordinary Sonnerie de Sainte Geneviève du Mont de Paris – this is exactly the same. The soprano duet on Lacrimosa, track 15, is as beautiful as those of Pergolesi’s Stabat Mater, or any written by Mozart in his C-major mass, many years before he composed it, and the touching tenor-baritone duet with chorus “Pie Jesu Domine” (track 22) is another one of those sublime moments. The two tenor-baritone duets (the other is the “Cedant hostes”, track 20), to which the Recordare trio with soprano can be added, are indeed one of the originalities of Gossec’s Requiem. I can’t describe all the beauties and wonderful touches of imagination of the music. This Requiem is an extraordinary work, worthy of standing on the same plane as Mozart’s Requiem and Mass in C or Haydn’s Nelson-Mass.
Not that the Requiem is a much-performed and recorded piece – nothing near Mozart’s or Haydn’s great choral masterpieces – but if number of recordings gives any clue, it is Gossec’s most popular work, with five recordings that I’m aware of. The first was still in the LP days, by Jacques Houtmann conducting the Liege Symphony Orchestra; it has been reissued to CD by Koch-Schwann, CD 312041 K2, barcode 9002723130416. [Addendum from September 2017: I had it on LP, and since originally writing this review in 2011 I found the CD edition at a reasonable price, but I haven’t yet played it]. Capriccio released only in 1992 a version recorded in 1980, with the dean of Cologne’s “period-practice” ensembles, Cappella Coloniensis under Herbert Schernus, label number 10 616. So it is really this recording by Louis Devos, first published on CD by Erato in 1988, that became the entry point in Gossec and reference for the work. Later came Diego Fasolis (recorded live in 1998, published by Naxos in 2001, Naxos 8.554750-51) and Jean-Claude Malgoire (in 2003 on the French label K617, label number 152, barcode 3383510001529). The ensuing comparison is based only on Schernus and Fasolis, the Malgoire and CD edition of Houtmann entered my collection only after I had completed this review.
First thing to note: the differences in the editions performed are striking. The liner notes to the Devos (authored by himself) and Schernus CDs mention using Gossec’s autograph – obviously the same copy in both cases: it is kept at the Royal Library in Brussels – beside the one known printed score (from 1780, say all the scholarly sources; Devos however dates it from 1774), noting (according to Capriccio) that it is “in many respects far more plain than the version known hitherto, and also that the instruments are used more consistently in their timbres”. It seems, but is not entirely clear, that it is Gossec’s original version, later revised for publication, but Devos doesn’t clarify the issue by calling it “a new version of the Requiem itself, revised, abridged and reorchestrated by Gossec’s own hand”. In fact I am very puzzled when Devos further writes that “every bar was different and the final fugue completely crossed-out and re-written” (“complètement rayée et réécrite”, my own translation, the one provided in the booklet is ambiguous), which seems to indicate either that Gossec re-wrote a pre-existing version (itself in manuscript form, making that autograph a palimpsest?), which seems contradictory with the notion that this would be the original version, or that Gossec had second thoughts after jotting down a first version of the final fugue – a more likely interpretation.
Anyway, to be told that both Devos and the Capriccio version used the autograph for their performance only obscures the issue, since we are not told how exactly they used it and what exactly are the differences: did they perform the autograph or did they freely choose between autograph and printed score? And, when they performed from the autograph, are the differences only details not audible without the score, or more substantial? Those questions wouldn’t be raised if Devos and Schernus performed the same or a very similar version. But the differences between them are much greater than between Devos and Fasolis, and I suspect that those sections missing in Schernus’ version are so not because of any observance of the manuscript, but simply as a result of the producers’ decision in order to fit the work on a single CD. Silly thinking, since, with a TT of 66 minutes, the CD could have easily fit 10 more minutes (indeed Devos’ is 74). And if indeed based on autograph, then Devos was right NOT to follow it, because their absence is a great loss. On the other hand, some of the differences involve not cuts, but Schernus (not Devos or Fasolis) performing indeed entirely different, and much simpler settings, and thus seem to indicate that, in those numbers, Devos (like Fasolis) did perform the final and more elaborate printed version. Only on one occasion, in the orchestral introduction, it is Devos who plays a drastically shorter version, where Schernus and Fasolis play an extended one. And here again I suspect he does so not out of any manuscript (or did Schernus here opt to follow the printed score?) but to fit the Requiem on a single CD. I’ve spotted a few more minor and almost imperceptible cuts in the Devos version (in “Te Decet Hymnus” track 3, “Et Lux Perpetua” track 6, “Cedant hostes” track 20 at 2:20), and again my hint is that they were motivated by the same consideration. More regrettable is his omission of the orchestral introduction to “Vado et non revertar” (“I go and shall not return”), the first number of the Offertorium (track 18). Its weeping strings are a beautiful effect. Strangely, only Fasolis’ chorus (but neither Devos’ nor Schernus’) sings the quasi a-cappella line “voca me cum benedictis” on the two introductory chords of “Oro supplex” (track 14): it is very short, a single phrase. As announced by Devos’ presentation, it is not exactly the same fugue on “Et lux perpetua” as Fasolis that he plays at the end (track 22), although it is comprised of the same building blocks.
Schernus’ version is ruled out then by an abundance of factors, the first being these bleeding cuts (the beautiful soprano aria “Exaudi”, track 4 of Devos and Fasolis, as well as the fugal “Et Lux Perpetua” that closes the first part, after the repeat of the “Requiem aeternam”, track 6 of Devos and Fasolis) and its selection of some other, simpler and, frankly, less interesting settings (a short and plangent baritone recitative for “Inter oves” instead of an extended and rather joyous Mozartean aria for soprano, track 12, a short and dramatic “Pie Jesu” with an abruptly concluding “Amen” at the end of the Dies Irae section instead of the touchingly angelic and appeased “Pie Jesu” and the great and mighty Bach-like fugue on Amen, track 17, the short and abrupt concluding “Requiem aeternam” instead a more beautifully appeased and “Faurean” setting of the same words introducing another great fugue on “Et lux perpetua”, track 24). And these subtractions are not made up by Schernus’ inclusion of a penultimate and substantial trio on “Lux aeterna” that the two other versions don’t perform, or of a more extended repeat of the Requiem Aeternam at the end of the first section than what Devos and Fasolis play (track 5). But Schernus is also ruled out by his numerous interpretive shortcomings, of which the most glaring is the substitution, to Gossec’s required tenor of, alternately, the alto and soprano. For more details on this, see my extended review of Schernus’ recording. It changes entirely the character of the music, is far less original than what Gossec wrote (could even Cray XK6 compute how many soprano arias were written in the 18th Century?), and sometimes lends the music an operatic character that is out-of-place. In fact, I prefer Devos’ choice of a tenor even in those arias or duets where, apparently, the score indeed calls for an alto (or maybe a counter-tenor), like “Te Decet Hymnus” (track 3) and “Quid sum miser” (track 10, where Fasolis also uses tenor although he, unlike Devos, did have an alto at his disposal). And here again, the quality of Schernus’ baritone, Alessandro Corbelli, isn’t enough to compensate. Speaking about emendations, strangely, Fasolis turns the Recordare trio (track 9) into a quartet. The trio is made of 6 stanzas and is in fact a succession of solos, each soloist singing one in turn, with the three then joining to repeat the 3rd and 6th. While the odd stanzas are attributed to soprano and baritone, I’m not sure to whom Gossec attributes two and four, alto (Schernus and Fasolis) or tenor (Devos). But anyway, Fasolis then has tenor sing in the ensemble on the repeats of 3 and 6, which doesn’t seem to make sense: why would Gossec have turned this logically architectured trio (and that’s how the Capriccio CD track listing describes it) into a quartet?
With Schernus out of the picture, of the two remaining versions, Devos is without hesitation my favorite. Fasolis may be more complete, but, as indicated, Devos’ cuts are most of the times minor and almost imperceptible. The one he does in the Introduzione is more radical, but, frankly, given the funeral tempo Fasolis’ adopts there, Devos’ terseness (and with a similarly funeral approach) is much preferable. Drawing out the movement to 5 minutes, Fasolis way outstays his welcome and is off to a very boring start, Devos’ 1:20 sounds exactly like an introduction and, on its own, it sounds perfectly right. So my main quibble with Devos’ cuts is the loss of the 1:15 introduction to the accompanied recitative “Vado et non revertar”, with its movingly “weeping” strings, track 18 (that’s how long it takes with Fasolis). The music returns inside the recit, but it isn’t the same.
But these few cuts, independent of the fact that they’ve enabled Erato to fit the Requiem on a single disc where Fasolis takes one and a half (completed by a mediocre version of Gossec’s last symphony) are more than made up by Devos’ interpretive superiority over Fasolis.
Although Fasolis’ sonics are more pungent and vivid (his few timpani outbursts in particular are really explosive, as the striking death knell that introduces the whole Requiem – curiously omitted by Schernus on Capriccio – or those in “Mors stupebit”, track 9, and the woodwinds – oboes and bassoons – are very expressive), Devos’ period-instrument strings (the ensemble plays a semi-tone lower) sounds leaner, more transparent, a bit acid but with much more character (and nowhere more than in the introduction to the Dies Irae, track 5). Fasolis’ modern strings are thicker and, Fasolis’ often romantically drawn-out tempos not helping, more sentimental (try the Introduzione, track 1, or the “Quid sum miser”, track 8). They are also distant and lacking intimacy in the Lacrimosa relative to the voices: with Devos, the strings sound like an equal dialoguing partner, not just an accompanying and supporting one (track 15). But in the same movement, Fasolis’ oboes are very vivid and expressive. Unless you can crank up the volume, Devos’ chorus sounds at times a bit distant and thus looses some of its impact (as in the extraordinary “Mors stupebit”, track 9), but it is also more transparent than Fasolis’. Fasolis’ chorus is more present, full and surrounding (Requiem aeternam, track 2), but that very presence lends it a romantic silk which doesn’t sound entirely right, and sometimes turns even to thickness and a degree of heaviness compared to Devos (as in “Mors stupebit”). Nobody can fail to be moved by Fasolis’ Faurean “Pie Jesu”, track 17, slightly more animated and thus dramatic than Devos’, but it really sounds like Fauré (e.g. very romantic) rather than like Gossec. With Devos the balance of chorus and orchestra is just right.
I also prefer Devos’ vocal trio. Devos’ soprano, Bernadette Degelin, is the purest- and most angelic-sounding, especially felicitous in “Exaudi” (track 4) and “Inter Oves” (track 12), although the voice, using vibrato very sparingly, may sound thin to those more accustomed to beefier operatic sopranos. In Exaudi Fasolis’ Roberta Invernizzi sounds a bit matronly in comparison. Invernizzi is fine in the Recordare (track 11), although a bit too operatically expressive in her second verse, Ingemisco (“I groan as an offender / My face blushes with guilt”). Heard on her own, Invernizzi is beautiful in “Inter Oves” as well, and her operatic timbre is perfectly in situation there. Devos’ soprano II Greta de Reyghere sings only, in duet, in the Lacrimosa (track 15). I find it especially welcome to hear two very homogeneous timbres there, rather than contrasting soprano and alto, although it is quite beautiful also with Invernizzi and Arruabarrena under Fasolis.
If the quality of the tenor is your main discriminating yardstick, you’re not going to find much fodder here. Howard Crook is shared by Devos and Fasolis, and he is fine with both, although there is more youthful clarity in 1988, and the timbre is darker in 1998. Crook is also slightly more dramatic and operatic in 1998 in the accompanied recit. “Vado et non revertar”, and there are differences of detail in the ensuing aria – Fasolis’ tempo is slightly more animated than Devos’ -, but both are fine.
In the mighty and ominous “Tuba mirum”, Fasolis’ Baritone Claude Darbelly doesn’t have the vocal beef to pull it off, and he substitutes with some nasal and vulgar intonations. It is a Sarastro type of voice that this ominous music would call for, although Gossec wrote it for a higher tessitura than that (it descends very low as well). But Darbelly is better in his ensembles, the Recordare Trio (track 11), the “Cedant hostes” and “Pie Jesu” duets (CD 2 tracks 3 and 5), possibly because he is the lower and supporting voice rather than the exposed soloist (but it is not the case in the Recordare Trio, which alternates solos). Devos’ Kurt Widmer doesn’t have a very big voice, but it is firm enough, though not without traces of strain in the “Tuba mirum”. But truth is, Schernus’ Alessandro Corbelli on Capriccio is the best of all three, with more nobility and valiance than true might, but a beautiful and perfectly sustained voice throughout.
Some of Fasolis’ choices of tempo are disputable. I’ve mentioned his impossibly drawn-out Introduzione. Fasolis’ choral fugue in “Et Lux Perpetua”, track 6, is too smoothed-out and lacks angles. Although Devos’ chorus may not sound as smooth and grand, it has the right attacks and Devos is more dynamic. Likewise with the Requiem’s concluding Fugue (a variant of the previous one): Fasolis elicits a fine, brooding atmosphere in the Requiem aeternam introduction, and his final fugue starts well because, pitched a semi-tone higher, he sends his sopranos in the stratosphere; but again his chorus is thicker than Devos’ and his strings excessively romantic. Devos is again more energetic and incisive here, and ultimately much preferable. Fasolis’ elephantine tempo in the Sanctus (CD 2 track 4, Devos’ track 21) is not so much majestic (“maestoso” is Gossec’s indication) as funeral, betraying the meaning of the words: it is a pageant to God, “holy, holy, holy, Lord God of Sabaoth”. Devos has the required sense of exultation.
It’s not always so one-sided, though. With vastly different tempo choices the character of the “Agnus Dei” is radically different between Devos (track 23) and Fasolis (CD 2 track 6), more dramatic and almost an exhortation with the former, a plangent and subdued prayer with the latter. I like both approaches. Devos and Fasolis are both plangent and sentimental in “Quid Sum Miser” (track 10), and here it is Capriccio’s Schernus who is swifter, but the sense of soothing that his approach elicits is not entirely appropriate to the words (“who am I, a wretch, to speak”). In “Exaudi” (track 4), Devos is more plangent and funeral than Fasolis; in terms of what is expressed, Fasolis may be right: the words after all are not funeral, they are “Grant my prayer, to you all flesh shall come”. But then, the central bridge passage with flutes is incomparably beautiful at Devos’ reined-in tempo (1:50), and comes to not much with Fasolis (1:27). In the overbearing “Tuba mirum”, with its Berliozian antiphonal trombones (track 8), the sonic perspective is fine and ominous with Fasolis, but he takes the opening fanfare at a very deliberate tempo that threatens to make it loose all shape. But the rest is vivid and pungent. Devos’ sonic perspective and opening tempo are right, and again his leaner strings have more character than with Fasolis, but in the development his tempo is slightly cautious. Devos adopts a tempo that’s a bit static and solemn in the Lacrimosa (track 15). Although the duet is very reminiscent of Pergolesi’s Stabat Mater, the words aren’t about the sorrows of the Virgin Mary at the foot of the Cross, but the prediction that tears will be shed the day the guilty rise from the ashes to face judgment, and you don’t need to weep to express that; the music is expressive enough that one doesn’t need to milk it. Fasolis is just right here. But I much prefer Devos’ swifter tempo in Judicandus (track 16) to Fasolis’ solemnized one. Fasolis’ Pie Jesu (track 17) is incomparably beautiful, thanks to the smoothness of his chorus, and the final fugue on Amen is great – but Devos’ is fine too. Although, as mentioned, he regrettably omits the orchestral introduction to “Vado et non revertar” (track 18), Devos’ tempo is very held back, turning the music into a dirge, and Fasolis’ is even slower. Not an invalid option, but I prefer Schernus’ swifter tempo here, which conveys better a sense of emotional turmoil. But Devos and Fasolis animate better the ensuing accompanied recitative.
Devos has the magnificent Recordare just right (track 11), with its hypnotizing chaconne-like ostinato basso underpinning plangent, brooding whiffs a melodies from violins and bassoons, but Fasolis isn’t bad either. In “Confutatis maledictis” (track 13), Fasolis is, like Devos, suitably brisk but his strings are relatively smoothed-out and don’t reach the degree of violence of Devos’. His chorus does, though.
Ultimately then, though Fasolis may have some striking individual moments where he is as good or even better than Devos, overall it is Devos’ leaner and more characterful period instrument ensemble that is the most convincing. I’m happy to keep both, but Devos is for listening, Fasolis is only for further reference.
Now, I am wondering, maybe I should give a second chance to my friends, and send them a copy of Devos’ recording. One should never relinquish hope and faith that mankind can improve its hearing.