Les 19 Grands Hymnes Révolutionnaires. Edwige Perfetti, Tibère Raffali, Gilles Cachemaille, dir. Claude Pichaureau. Musifrance Radio France C 524 or 245 006-2, Erato Musifrance 2292-450006 2 (1990)

Les 19 Grands Hymnes Révolutionnaires (works of Rouget de l’Isle, François-Joseph Gossec, Giuseppe-Maria Cambini, Nicolas Dalayrac, Frédéric Duvernoy, Etienne-Nicolas Méhul, Charles-Simon Catel, François-René Gebauer, Etienne Ozi, Luigi Cherubini). Edwige Perfetti, Tibère Raffali, Gilles Cachemaille, Choeur de l’armée française / Chorus of the French Army (dir. Serge Zapolski), Chorale à Coeur joie La Gondoire (dir. Daniel Catenne), Chorale populaire de Paris (dir. Jean-Claude Chambard), Orchestre d’Harmonie des Gardiens de la Paix de Paris / Wind Band of the Paris Police, Claude Pichaureau. Musifrance Radio France C 524 or 245 006-2, Erato Musifrance 2292-450006 2 (1990), barcode 022924500626



Recorded February & November 1987, February & March 1988 at Eglise Notre-Dame-du-Liban, Paris

Track listing and credits:

Opportunity lost – in search of the authentic Marseillaise
Originally posted on Amazon.com, 31 January 2013

Had I known that this apparently inconspicuous disc of songs, hymns and military music from the French Revolution was going to send me on hours of research… I would have done it anyway. But be warned, it turns out this is going to be less a review than a scholarly essay on the sources of The Marseillaise and other revolutionary songs. Now, for the hasty: if you are looking for a collection of such songs, you’ll find no better than this (except if you can locate the 3-CD set from which it is excerpted, “Musique & Revolution”, Erato  245 005-2 or 2292-45005-2, barcode 022924500527 ), even if it is far from satisfactory in terms of authenticity, e.g. playing those pieces as they were originally heard in the 1790 decade. Note however that the Erato CD doesn’t include the texts and not even detailed descriptions of each piece, which won’t be a problem if you understand French because the singing is always intelligible, but might be otherwise (texts are included in the 3-CD set).

So, this was an interesting program of songs and official hymns, interspersed with a few numbers of military music (Marching Drills) used at official ceremonies and celebrations during the French Revolution, masterminded by the musicologist Frédéric Robert, with the Wind Band of the Paris Police (Orchestre d’harmonie des Gardiens de la Paix de Paris) conducted by Claude Pichaureau, and published a year after the celebrations of the bicentennial in 1989. Among its competitors there would be the recordings of Michel Plasson on EMI CDC 7 49470 2, “Révolution française” (which shared the Marseillaise and Méhul’s Chant du depart), Jean-Claude Malgoire on French CBS CDCBS 45607, “Chantons la révolution” (Marseillaise, Chant du Départ and Gossec’s Marche lugubre) and John Wallace on Nimbus NI 5175, barcode 083603517526 or 0710357517525 (Marseillaise, Marche lugubre along with two other hymns not contained in Erato’s compendium – the disc’s main course is in fact Berlioz’ Symphonie funèbre et triomphale). I don’t have the recording of Orchestre de la Garde Républicaine under Roger Boutry, on EMI CDC 7 49473 2  barcode 077774947329, which shares The Marseillaise, Chant du Départ and the “Pas de Manoeuvre” of Duvernoy that is on track 6 of Erato: the truly horrendous rendition of the Marseillaise (heard on YouTube) by popular singer Mireille Matthieu over Boutry’s overblown support will forever be a strong disincentive.

There were more overlaps with a CD belatedly published by Adda in 1991, “Révolution Aux Choeurs”, Adda 590 018, played by the combined forces of the Paris Opera and Soviet Army Choruses and the Wind Band of the city of Le Havre under Philippe Langlet. In addition to the ubiquitous Marseillaise and Chant du Départ, it also had Gossec’s rarer “Peuple, éveille-toi!” (“People, Awake”) and “Chant du 14 Juillet” that are on Erato (tracks 3 and 8), as well as the same “Pas de Manoeuvre” (Marching Drill) of Duvernoy and Gebauer that are on tracks 6 and 14. But the main overlaps were with a double CD on Accord 200282, “Bicentenaire de La Revolution Francaise évoquée par la musique et le texte” (1988), barcode 3229262002820, the reissue of a 3-LP box originally published by Guilde Internationale du Disque in 1962, M2262, “La Révolution française : La grande époque de l’histoire de France, 1787-1799 : évoquée par le texte, l’image et la musique”, in which Pichaureau’s musicologist, Frédéric Robert, was already involved. It shared La Marseillaise, Gossec’s Marche lugubre and “Peuple, éveille-toi”, Méhul’s “Chant du Départ” and “Chant funèbre à la Mémoire de Féraud” (Erato’s track 17), Dalayrac’s “Les Canons ou la réponse au salpêtre” (track 5), Cambini’s “Ode sur Bara et Viala” (13) and “Les Rois, les grands, les prêtres” (15), Catel’s “Ode sur le vaiseau `Le Vengeur'” (11) and another Pas de Manoeuvre of Gebauer, the one on Erato’s track 12. But the Guilde/Accord set was first and foremost a short history of the French Revolution presented by the popular TV historian Alain Decaux, with the songs serving only as illustrations.

The main liability of Pichaureau is that in most cases he doesn’t offer “authentic”/period orchestrations based on the original sources that would enable you to hear that music as it was originally heard in the 1790s, but in truth he offers more than most his competitors. The old Guilde LPs reissued by Accord are a special case. Except for The Marseillaise which credits Berlioz’ orchestration (but it is a very free interpretation of it), the Accord reissue is unspecific as to the orchestrations used but, based on the notes describing each work, it seems that at least in some cases the 1961 team returned to the original sources. Plasson is the furthest away since he uses orchestrations for full romantic symphony orchestra (of Berlioz in the Marseillaise and an uncredited one in Chant du depart, and 20th century ones for most other compositions on his rectial), and from what I can hear online of Boutry he also uses modern, corrupt and grandiloquent arrangements for large wind band and percussions (apparently, in La Marseillaise at least, his own).

In many cases (9 out of its 19 tracks) the Pichaureau/Erato disc credits the arrangements for wind band of Désiré Dondeyne, the long-time conductor of the Paris Police Wind Band (including The Marseillaise, Méhul’s “Chant du Départ” and Gossec’s Marche lugubre), and so does Adda for Langlet. The orchestrations used by Malgoire are uncredited, but in the case of Gossec’s Marche lugubre I suspect he (and Wallace as well) also uses Dondeyne’s orchestration. On the other hand Malgoire’s Marseillaise is an interesting case: it was apparently orchestrated specifically for the recording, and turns out to be based on the very first publication of Rouget de Lisle’s melody, made under the composer’s supervision shortly after the composition (from the end of April 1792). See my review for more details on this. As for the orchestration of the Marseillaise used by Wallace on the Nimbus CD, it is also a modern one done by annotator John Humphries – but more on that anon.

In six other instances Pichaureau uses what are dubbed “reconstitutions” by his musicologist Frédéric Robert (and Robert even adds his “reconstitutions” to two of the Dondeyne orchestrations, those of the Marseillaise and Chant du Départ), and in two cases his own orchestration (and one of these again with Robert’s “reconstitution”). We are not told what these “reconstitutions” exactly are, except for “Chant du Départ” where Robert claims to have returned to Méhul’s original scoring. Catel’s “Hymn sur la reprise de Toulon”/”Hymn on the Capture of Toulon”, track 10, and Gossec’s “Hymn to Voltaire” (track 18) are three-part a-cappella men’s choruses, that were published at the end of the 19th Century in a volume of revolutionary songs (see more about that below), and the credits under another track (Gossec’s “Peuple, éveille-toi”, track 3) indicate that Robert and Pichaureau used that volume; though it is not credited for the two choruses, it is permissible to think that it was also Robert’s source. But if so, what then was his “reconstitution”? Just copying the parts out of the book?

In truth, I find it incomprehensible that with sources so essential to the history and very identity of French culture, the same kind of “urtext” work as on the scores of, say, Lully, Rameau and all the 17th and 18th Century French composers has not been done, at least on The Marseillaise. I had been puzzled not to find, despite hours of online research, any trace of a modern, or even ancient publication of Gossec’s famous and oft-mentioned orchestration of Rouget de Lisle’s song. Now come on, we are talking not about just any obscure and forgotten revolutionary song, we are talking about La Marseillaise! How more central to French culture and identity can you be? So how could Gossec’s score not be published? Well, more hours of online research led to this.

Constant Pierre and Gallica, founts of knowledge

Gallica, the online website of Bibliothèque Nationale de France, is an invaluable treasury of documentation and I kneel to them in gratitude. They have made available online hundreds of facsimiles of sheet music from the era, many from the “Magasin de Musique à l’usage des fêtes nationales”, the print of the association of composers of the National Guard Orchestra (whose school of music was the forerunner of the Conservatoire) that had a monopoly on the publication of songs and hymns used for national celebrations. But, even more invaluable, Gallica also offers online a collection published in 1899 by an archivist of the Paris Conservatoire, Constant Pierre: “Musique des fêtes et cérémonies de la Révolution française” (“Music for the Celebrations and Ceremonies of the French Revolution”). This admirable man spent years locating and collecting all the original sources, but for reasons of practicality he published them in his collection only with a reduction for piano, trying to retain as much as he could the shape of the original orchestrations (at least when these were available, usually in the form of instrumental parts rather than full scores). In an earlier volume, published in 1893, “Musique exécutée aux fêtes nationales de la Révolution française” (“Music Played at the National Celebrations of the French Revolution”), Pierre had published a few complete scores, but frustratingly few: ony four, including Gossec’s “Marche lugubre” (on track 3) and “Peuple, éveille-toi” (track 4). It is the same volume that includes the a-capella men’s chorus that I mentioned above. This volume is more difficult to access online, as Gallica hasn’t made it available. But I’ve found it, on the website of (I think) the Library of the Belgrade University in Serbia. These are the cases that make me hate globalization and technological progress a little less. It is only in .jpg form and can be read only sequentially (which means you have to go through 77 pages if you want to jump to page 78), which is very tedious, but I’m not complaining. Finally, in a complementary volume from 1904, also available on Gallica, “Les Hymnes et Chansons de la Révolution avec notices historiques, analytiques et bibliographiques / The Hymns and Songs of the Revolution with historical, analytical and bibliographical descriptions”, Pierre listed the locations of all theses sources, most of them at the library of the Conservatoire, some at Bibliothèque nationale or at the library of the Opera.

Other possibly than the compositions included in the 1893 publication, it is not clear whether Dondeyne simply reorchestrated Pierre’s piano reductions (possibly using his indications about the original scoring contained in his complementary volume when he provided them), or actually went back to the original sources used by Pierre, and likewise for Robert and Pichaureau when it is their orchestrations that are used, but I strongly suspect the former. In the case of the Marseillaise, for reasons stated further down, it is evident that Dondeyne used Pierre’s piano reduction, and likewise with John Humphries who did the rescoring for Wallace on the Nimbus disc, despite his attempts to dodge admitting so. Dalayrac’s “Les Canons ou la réponse au salpêtre”/”The Cannons or the Reply to Saltpeter” (track 5) is one of the two pieces where conductor Pichaureau is credited as orchestrator. There, Pierre provides a score with piano reduction and gives the reference to a score with parts kept at the Conservatoire’s library and at Bibliothèque nationale, but fails to give any indication of the original instrumentation. But in the 1961 recording reissued on CD by Accord, the (uncredited) annotator said that “we have respected the lineup of the original edition (2 clarinets, 2 bassoons) – or at least of the edition that came down to us, as we were unable to establish for sure if the accompaniment included, as was the use, two horns as well”. Pichaureau’s orchestration is much larger than that, even accounting for the possible two horns. I prefer the leaner orchestration. The song dates from 1794 and was Dalayrac’s response to another song, “The Republican Saltpeter”. To its celebration of saltpeter (one of the main constituents of gunpowder), Dalayrac’s lyricist Coupigny responded with a praise of the cannons. As late as 1798 Cherubini put the original song to music, in view of its publication, and although it botches the chronology, too bad that it too isn’t included on the Erato CD.

I will however grant Dondeyne and Robert benefit of a doubt when what is performed is a piece for which, for some reason, Pierre did not provide a score: what track 9 faultily titles “L’Offrande à la liberté” and attributes to Gossec, is actually a “pot-pourri” or instrumental medley for winds derived from Gossec’s lyric scene Offrande à la liberté (see further about that), but with no indication that it was Gossec himself who wrote it,
and containing the tunes of La Marseillaise and the two popular songs “Ah ça ira” and “La Carmagnole” (see it on Gallica). In that case, Pierre had no reason to provide a score since it isn’t a song. But he must have had some kind of grudge against Cambini, because he also fails to do so with three of his compositions: his “Pas de charge républicain” (“Republican charge”) track 4 (a song and not a military march as the title would seem to imply), the “ronde patriotique”/”patriotic rounddance” “Les Rois, les grands, les prêtres” (“The Kings, the Mighty, the Priests”, track 15, for which Pierre doesn’t even indicate the instruments, other than “orchestre d’harmonie”, wind band), and “Ode sur Bara et Viala” , track 13. Bara and Viala were two young teen-agers who had fallen heroically to the ennemies of the Republic; the score’s exact title is in fact “Ode for the Two Young Heroes Barra [misspelt with two Rs] and Viala, at the time of their translation to the French Pantheon”, and curiously the old Guilde/Accord recording had exactly the same mistaken title as Erato. I wonder if Dondeyne’s arrangements already existed in 1961 and Guilde used them, but it leaves open the question of what these arrangements are exactly in relation to the original sources. All the original scores are duly listed in the catalogs of Bibliothèque nationale, and Gallica has made available online the instrumental parts of the “Pot-Pourri” after Gossec and only the vocal scores of “Pas de Charge” and “Les Rois… “.

Anyway, whether based on Pierre’s piano reductions or the original material, sure, Dondeyne’s, Robert’s, Pichaureau’s or Wallace’s orchestrations for wind band may sound “authentic” and “period” enough and the same with those of the old Guilde recordings, certainly far more than Plasson’s full romantic symphony orchestra or Boutry’s large Wind Band of the Republican Guards, and they give you an idea of how the contemporary “orchestres d’harmonie” might have sounded. Oftentimes, playing by ear, Dondeyne’s and Robert’s orchestrations seem close enough to either the scores, when accessible, or the instrumentation as given by Pierre. Catel’s “Ode on the Warship `The Avenger'” (track 11) is, strangely, the only one for which the track listing credits no one for the orchestration, but I do hear the clarinets, horns and bassoons indicated by Pierre (the typical orchestration for military band in the early 1790s). Gossec’s “Chant du 14 juillet” (track 8, “reconstitution” Robert) seems faithful enough to the orchestration indicated by Pierre, and Pichaureau commendably plays it complete, all three stanzas (and at an unhurried tempo which makes it sound like a chorus from Sarastro’s priests, which is not unfitting given that it is a hymn to the sun), and likewise for Méhul’s “Funeral Lament to the Memory of Representative Féraud”, track 17 (Féraud had been killed by the mob and his head brandished on a pique). In the Marching Drills of Duvernoy (track 6, orchestration Dondeyne) and Gebauer (tracks 12, “reconstitution” Robert) I think I hear all the instruments indicated by Pierre and no others. But in Gebauer’s 2nd Marching Drill, track 14, orchestration Dondeyne, I think I hear oboes that are not in Pierre’s indications, and in Ozi’s Marching Drill (track 16), the second piece orchestrated by Pichaureau (with an added “reconstitution” by Robert), I’m not sure I hear Ozi’s piccolo flutes, and I think I hear more clarinets than Ozi’s two, and what’s for sure is that Pichaureau has a drum that is not indicated by Pierre in the scoring.

Same with Méhul’s “Chant du Départ” (track 7). Here Dondeyne is credited, but with a “reconstruction” by Robert, and in the liner notes Robert says that “it appears here in the original instrumentation of its very author”. Well, I certainly hear Méhul’s piercing clarinets, trumpets, timpani, but I think I also hear horns, that were NOT part of Méhul’s instrumentation according to Pierre. The old Guilde LPs claimed to have used “a symphonic arrangement written at the time, based on the original score for wind band, in which violins and violas replace the clarinets”. Seems strange to have done that if they could have used the original orchestration, and anyway Pierre does not list any such arrangement. It sounds much better with Robert’s piercing clarinets. But I also hear horns in the Guilde version, making me wonder if Pichaureau and Robert didn’t use the same edition (Dondeyne’s?) but simply reinstated the original clarinets in lieu of the strings, but kept the horns that shouldn’t have been there. In the “Pot-Pourri” after Gossec (orchestration by Dondeyne), whose original parts are available on Gallica, I hear what I see, except for one thing: the scoring has 2 clarinets obligato, and 2 clarinets “ripieno” sustaining the first two (usually unissono) but silent in moments of solo flourishes. I wouldn’t stake my hand that Pichaureau doesn’t have them, but if he does, they are not truly perceptible, and whatever effect the arranger might have had in mind is lost.

And I am leaving aside for the moment the issue with the tubas.

Certainly, these are subtleties of orchestral coloring that will be of little interest to most listeners. But it’s not just a matter of instrumentation. Pichaureau and Dondeyne’s straying from a truly “authentic” rendition of these scores is at times far more substantial, and the two afore-mentioned Ode of Catel and Chant du Départ of Méhul are good cases in point. In Catel’s Ode, Pichaureau not only sings only half the verses (which is no problem since the music is by nature strophically repetitive) and even repeats one (so that he can end on the lines “vive la liberté”, which is brilliant but not the song’s original ending): more problematically, he puts entire paid to the strophic structure of the song, singing the wrong verses over the music that was supposed to accompany it (let me not burden you with details here). It’s not as bad as hammering cylinders into square holes, but it is a little bit like it; Catel’s poet, Lebrun, was certainly no Racine, but he was better than Pichaureau’s jumbling makes him. Robert also adds a short introductory ritornello (taken in fact from the last bars of the song), which according to Pierre’s vocal score was not in the original song. The old version reissued on Accord didn’t have the introductory ritornello, but for botching the relation text-music is was even worse.

Now, I’m not saying this is really important. After all, those are not compositions of Bach or Rameau, and the gist and flavor of the music does come out. But it is more of a problem with Méhul’s “Chant du Départ”. Written in 1794, it was an important revolutionary song, as popular as the Marseillaise if not even more, dubbed indeed “the Second Marseillaise” and it almost made it as France’s national anthem. The text of Joseph-Marie Chénier’s (André’s brother, and whose fate was less fateful) and the score of Méhul’s contain 7 verses sung by solo voice, followed by refrain sung by chorus. But the interesting point is that each verse is supposed to be ascribed to a different character, a well as the ensuing chorus: a representative of the people and a chorus of warriors, a mother and chorus of mothers, two elderly men and chorus, a child and chorus, a wife-chorus, a maiden-chorus, and three warriors followed by general chorus, so evidently, the song was meant to be sung by different soloists and varied sections from the choir, including children’s choir. But Pichaureau doesn’t. He cuts the stanzas of the elderly men, the wives and the maidens and shares the rest between his tenor and his mezzo (suitably matronly for the mother, but much too much for the child) and each time and indifferently has the refrains sung by mixed chorus. Again, the old Guilde version was even worse, retaining only two stanzas (including the one of the two elderly men that everybody else cuts), both sung by the same tenor and mixed chorus. Plasson, although he too cuts the stanzas of the elderly men and maidens and uses an (uncredited) orchestration for full romantic symphony orchestra, is much truer to the original intent here.

Cherubini’s extraordinary “Funeral Hymn on the Death of General Hoche” (track 19) is a special case. First, it is mis-titled. Cherubini’s original work was made of two parts, a Funeral March that was played during the procession, and a Funeral Hymn sung at the mausoleum; Pichaureau plays only the March, not the Hymn (and too bad, really). It was scored for a large wind band, and was so successful that it was given in stage form in various operas, in a reorchestration including strings and replacing the tam-tams and side drums by timpani and bass drum. The piano reduction provided by Pierre was based on the latter. The scores he listed for the original version for wind band were kept at the Royal Library in Berlin, having been sold in Germany with Cherubini’s estate. What was kept in the Library of the Conservatoire in Paris was the manuscript of the orchestral version.

I doubt very much that Dondeyne went to Berlin to consult the original version, so again, he must have used Pierre’s piano reduction, rescoring for winds what was written for symphony orchestra. But even if so, cheers to him, his reorchestration is superb, and he must also be praised for reinstating the little interlude with clarinet solo that came after the imposing introduction and before the funeral march proper, and that had been cut in the orchestral version. Pierre hadn’t included it in his piano reduction but provided it in his later collection of descriptions.

But now come the tubas.

Serpents, cornu and buccina

The issue of “authenticity”, or allowing you to hear it as it sounded to the audiences in the 1790s, isn’t just a matter of establishing correct and faithful editions. In the case of Gossec’s “Marche Lugubre” and “Peuple, éveille-toi” (tracsk 3 and 4), these exist: they are contained in Constant Pierre’s 1893 volume, the one with the frustratingly few orchestral scores. And since Erato credits Pierre for “Peuple, éveille-toi”, it is clear that Pichaureau and Robert had access to it – making it, apparently, even more puzzling that Dondeyne’s edition should be used for the Marche Lugubre and not Pierre’s.

But even when the CD claims using the Pierre edition, it is very doubtful that Pichaureau and Robert did, and here is why. First of all, modern tubas didn’t exist in the 1790s. All the scores, and all the instrumentations indicated by Pierre in his 1904 listing, use the instrument that was called a “serpent” (“the bass wind instrument, descended from the cornett, and a distant ancestor of the tuba”, says Wikipedia, and always used as a reinforcement to the bassoons). So substituting modern tubas to the serpent may be practical, but it is no more authentic than using them rather than the Ophicléide in Berlioz’ Symphonie fantastique (the Ophicléide was indeed the “missing link” between serpent and tuba). In fact it is far less, since Berlioz welcomed the advent of the tuba when it was invented, which evidently the long-dead composers of the French Revolution couldn’t do. But that’s not all.

Both the Marche lugubre and “Peuple éveille-toi” were played on 11 July 1791, on the occasion of the translation of the remains of Voltaire to the Pantheon (along with the “Hymn to Voltaire”, not in the version for a-cappella men’s chorus played on track 18 which is a later version, but one with winds). Marche lugubre is an extraordinary funeral march that served many times during the French Revolution (and they had plenty of occasions to use it), and the prototype for all those funeral marches that will culminate in Berlioz’ Symphonie funèbre et triomphale, written as late as 1840 but obviously indebted to those revolutionary marches for wind band. Marche lugubre had been composed earlier but, as Pierre explains at length, that ceremony of the translation of Voltaire was the first time that it was made use of two newly reinvented instruments, the “tuba corva” and the “buccin”, built especially for the occasion and modeled after the descriptions on the Column of Trajan in Rome of the ancient Roman “Cornu” and “Buccina”, and Gossec added them to his March as well as using them in “Peuple, éveille-toi”. According to Wikipedia, the buccina “was originally designed as a tube measuring some 11 to 12 feet in length, of narrow cylindrical bore, and played by means of a cup-shaped mouthpiece”, and the “cornu” was an ancient Roman brass instrument about 3 meters long in the shape of a letter G. According to contemporary testimonies quoted by Pierre, the “tuba corva” made the sound of six serpents (important when you are playing on the Champ-de-Mars or Place de la Bastille for audiences of thousands) and the buccin of four “half-horns” (despite my research, I have not been able to find out what “half-horns” were, maybe simply a reference to the fact that horns then were always played in pairs, alto and basso). With the tam-tam used in Marche lugubre, they hugely impressed the attendance.

But you bet Pichaureau doesn’t use them – and that’s presumably why he uses the Dondeyne edition in Marche lugubre: I suppose it rescores it for modern instruments, so what’s puzzling is that Pichaureau didn’t use, or Erato doesn’t credit Dondeyne’s version for “Peuple, éveille-toi” as well: that’s the version used by Langlet in the competing Adda disc. But wouldn’t it be great to hear those pieces with the original instruments? In fact, when Pierre wrote his books, a century after the facts, they were all but extinct, he was able to locate only one “tuba corva” in a private collection (each played a tone, so Gossec’s harmony required three. The buccin could produce three tones using natural harmonies). You’d need those instruments to be custom-rebuilt for any truly “period” performance of those pieces. And so what? Verdi had it done with far older Egyptian instruments for Aida.

And then, there is the Marseillaise.

In search of Gossec’s authentic Marseillaise

Constant Pierre devotes 50 (!!) pages of explanations on the sources and variants of La Marseillaise. Rouget’s original melody, printed in Strasburg within a month of its composition in April 1792 (it was known then as “The War Song for the Army of the Rhine”), was slightly different from the one universally know today (and Malgoire’s version is interesting for orchestrating precisely that original melody). The orchestration customarily played today at official ceremonies is derived from the one that was established as late as 1887. The earliest known orchestration, and the one that would be the most “authentic” is the one made by Gossec for his lyric scene “Offrande à la liberté” (“Offering to Freedom” – and not “THE Offering to Freedom” or “L’Offrande à la liberté” as it is now called by whoever refers to it – which, granted, is not many people), first played as “Hymne à la liberté”/”Hymn to Freedom”, and in which the Marseillaise is simply called “The March of the Marseillais”. It was premiered at the Paris Opera on October 2, 1792, which is a mere six months after the song’s inception: the success of Rouget’s song was so blazing that it shot from Strasburg to Marseilles and from there to Paris within a few months. It was then known, indifferently, as the Air, or March, or Hymn, or Song of the Marseillais. It is from that lyric scene that the “Pot-Pourri” for winds was derived. A score of the complete “Offrande à la liberté” with a piano reduction by Louis Jadin was published at the time (also available on Gallica), but as Pierre explains (as a general rule, not specifically about “Offrande”), his own piano reductions follow more faithfully the orchestral originals than contemporary piano reductions. Inexplicably Pierre doesn’t give the instrumentation used in the lyric scene itself and is unclear whether the manuscript score and/or parts remain, but sure enough: I’ve checked the online catalog of the Library of the National Conservatoire in Paris, and they list it: “Offrande à la liberté : scène composée de l’air veillons au salut de l’Empire et de la marche des Marseillois avec récitatif, choeurs et accompagnement à grand orchestre”, with a 21-page score and 17 parts. It is also listed by the catalog of Bibliothèque Nationale de France. “Grand Orchestre”: this is not a wind band, this is the Paris Opera orchestra! And it’s been there for two centuries, gathering dust on the shelves of the library? I simply cannot believe that before or since 1989 no one went to Bibliothèque Nationale or the Conservatoire and copied all that to make a modern edition. [and an Addendum from September 2017: and there you go! It was published in a modern edition in 2016 by Centre de Musique Baroque de Versailles. Lauds to them. Now we need a recording! Marc Minkowski, are you awake?]

A year later, Gossec made a second orchestration of the Marseillaise, for the celebration of the first anniversary of the King’s imprisonment on 10 August 1793, this one for wind band. Although it was, for the occasion, written to other words than those of Rouget (by a citizen Varon whose first name history seems to have lost all traces of), it too would certainly provide the basis of another “authentic” performance. However, looking at the score with piano reduction provided by Pierre for that version, it is evident that it isn’t the one orchestrated by Dondeyne. First, Gossec wrote his outdoors version for three-part men’s chorus and not for solo voice and chorus as the opera version (and to other words!), so at best Dondeyne’s version would be a mix of both. But, more substantial still, the wind version is introduced by a long ritornello on the Marseillaise theme, where Gossec had opened the orchestral version with a short fanfare, and it is the latter that Dondeyne puts to music. This is confirmed by other discrepancies between Gossec’s two versions as transcribed in Pierre’s piano reduction.

But here, when reading carefully the fine print of Pierre’s explanations, it becomes a real bag of snakes. He mentions a late publication, from 1798, by the print of the Conservatoire, of a score for solo, chorus and instrumental parts in an arrangement by Gossec for wind band accompaniment (“accompagnement pour musique d’harmonie”), and confesses to not understanding why “the hymn officially adopted as the national anthem wasn’t printed earlier by the organization in charge of distributing the patriotic hymns to the provinces and the armies” (Pierre also mentions a set of parts in manuscript form, kept at the Library of the Conservatoire and that, says Pierre, served for the official ceremonies before that belated publication). Rather than following the 1793 version for wind band with its introductory ritornello, it was the same version as the 1792 opera version, but reduced to three stanzas (Rouget’s 1, 5 and 6) which were, according to Pierre, those sung at the official ceremonies, and with only, if I understand his claim, a small melodic variant on the words “le jour de (gloire est arrivé)” – and I suppose Pierre is referring here only to the melodic shape of the lyrics, since there were (other than the number of stanzas) a number of differences with the Jadin score of the 1792 version, some of them substantial (if Jadin’s score is to be believed, the last stanza was introduced by a long orchestral ritornello). The 1798 version was published under the title “Hymne à la liberté, by Rouget de Lisle, arranged for large chorus by Gossec”. The earlier set of manuscript part that Pierre also mentions had the title “Hymne à la liberté, dit [aka] des Marseillais”.

And it is the piano reduction of that 1798 version with its three stanzas and its title “Hymne à la liberté” that Pierre provides besides the one for the 1793 wind band version, rather than one for the original 1792 opera version – but, for some reason, with the (more widely accepted) melodic variant of the 1792 version. Pierre’s title for his score seems to be a conflation of those of the printed parts and the earlier manuscript parts, since he adds “dit des Marseillais, by J. Rouget de Lisle, arranged for large chorus and orchestra by Gossec”. Pierre doesn’t give the exact instrumentation either, just mentioning “orchestre d’harmonie” (but he does give the instruments of the earlier manuscript parts, different from those of the 1793 version for winds), and as for location of the sources, indicates “author’s collection” and adds “as of now, the copy mentioned is the only existing one”.

Pierre is curiously enigmatic and unusually imprecise about the whole thing, and one could wonder what had become of that 1798 score and parts after his death. In a private correspondence, Frédéric Robert has indicated that they were held at the Music and Theatre Library of Sweden, in the collection bequeathed by collector Daniel Fryklund. And sure enough, I was able to purchase a CD-Rom with scans. The orchestration is slightly different from the one given by Pierre for the previous manuscript parts: 2 flutes (piccolo and big, instead of 1 piccolo and 2 flutes) 2 clarinets, 2 bassoons, 2 horns, trumpets (none in the parts of the manuscript version; in the material received from Sweden, there is only the part for one trumpet, but it says “1re trompette”, which implies that there were two and the sheet for second trumpet is lost), 3 trombones (instead of “serpent”), double bass, timpani (none in the manuscript part). Also, and puzzling, there is no “melodic variant”; either I misunderstand Pierre (he has a way of “coding” the various melodic variants that makes it difficult to decipher), or he made a mistake in indicating one. Dondeyne’s version definitely and unmistakably follows that 1798 version, but is it Gossec’s orchestation? Apparently not. First, Dondeyne’s score, as published by Editions Robert Martin, although it purports to be the “version Gossec”, shows an orchestration much larger and thicker than Gossec could ever have dreamed of: piccolo, flute, oboe, bassoon, 6 clarinets, 4 saxophones, 2 horns, 2 cornets, 2 trumpets, 2 trombones, 2 tubas, contra-bassoon, Double bass, timp. Second, judging by ear, it isn’t Dondeyne’s orchestration, at least not the published one, that Pichaureau performs, and what he does perform is certainly closer to Gossec’s conception, although I’m not sure it is entirely Gossec: the trumpets don’t come out very distinctly but I think they are there, covered by piercing clarinets and trombones, but the double-bass is reinforced by a low brass instrument, tuba or basso trombone, that is not in Gossec, and I certainly don’t hear the timpani part. Also, the third verse is sung by a solo soprano, rather than soprano (“dessus”) and tenor (“haute contre”) as the score indicates. As for Wallace’s Marseillaise on Nimbus, it is based on Pierre’s piano reduction, despite arranger John Humphries evasiveness about it – but see my review for the details about this.

Still today, whatever orchestration they use, more or less based on 1798 or reconstructed ones, most recordings of the Marseillaise (Pichaureau, Langlet, Wallace, Boutry among them) sing those three stanzas, which is a shame as it doesn’t give a complete view of Rouget’s original conception, but is also absurd since the fifth contains topical references (“the accomplices of Bouillé”) that are meaningless today. Interestingly, both the 1792 opera version and the 1798 version with three stanzas has the last one (the sixth in 1792 and third in 1798) played at a suddenly slowed-down larghetto tempo (for which Jadin added the indication “religious chant”). Pierre comments on this only for the 1798 version, but it was already there in 1792 at the opera and seems to have been Gossec’s invention: there is nothing in Rouget’s original conception that seems to imply it, and it certainly wasn’t indicated in his original score. Although the stanza starts with the words “sacred love of the fatherland”, the religious atmosphere sounds very contradictory with the general meaning of the lyrics there, which is as martial and aggressive as the rest: “…lead and sustain our vengeful arms”. This is not exactly Brahms’ Rhapsody for alto. Nonetheless the religious approach must have rapidly become part of the canon, since Berlioz keeps the tradition in 1830, and when the Marseillaise was revived in the late 1870s (it had been banned from Napoleon I to Napoleon III and after), it was still there (I’ve found testimonies of this in poet and critic Theodore de Banville’s critical essays, also on Gallica).

The bottom line is that it is not clear whether Dondeyne’s orchestration is based on Pierre’s piano reduction or on those printed parts now in Sweden, but in any case it isn’t entirey “authentic”. The same is true with Humphries’ for John Wallace – which, nonetheless, is brilliant (Wallace’s British chorus is more of a problem, “Revolution at tea-time”). But I find it truly incredible that nobody seized the opportunity of 1789 to publish a conductor’s score and parts from the authentic 1798 Gossec orchestration, and that the material is still lying idle on the shelves of the Music and Theatre Library of Sweden.

No better than this

And, with all that said, remains that an even more “authentic” Marseillaise remains to be heard, the one that they DID hear on the stage of the Paris Opera in October 1792. The sources are, apparently, accessible, and again it is incomprehensible that 1989 was not the occasion to fund and publish modern and available editions. Opportunity lost and I fear that it won’t be found again until 2089.

So we are left with what we have and it will have to serve until then. Among all those CDs, Pichaureau is certainly valuable for being, with a CD generously timing 72:47 minutes, the most systematic and complete in its exploration of those revolutionary songs (addendum 2017: save the the complete, 3-CD set, which I finally managed to find, years after originally posting this review), the old Guilde disc being “hors concours” because of its narration and its aged sonics. Wallace’s program is only a complement to Berlioz’ Symphonie funèbre, Malgoire’s is mixed with opera excerpts, Plasson’s disc is filled with a substantial and pre-Berliozian Chant National for three antiphonal chorus and orchestras of Méhul, and plays all the pieces in reorchestrations for full symphony orchestra. Langlet offers 15 minutes less of music, and his interpretation of the four hymns is generally appalingly bad; his program is more interesting for the various and rare Marches, Overtures, Symphonnies for wind band that he plays (also from the various ceremonies during the revolutionary era), but even there his interpretation is usually heavy-handed and sluggish.

And, after venting all my frustrations about the opportunity lost for truly authentic renditions, I am even freer to say that Pichaureau’s collection is very enjoyable. Obviously the music isn’t always great – those songs were not written to be sung by professional singers but by the people, much of it consists of simple and catchy melodies in repeated stanza/refrain form, and the “Marching Drill” music, as performed by Pichaureau, in a very playful manner, wouldn’t be entirely out of place in an opera of Auber or Gretry, with young ladies in tutu lifting up their legs – fancy that for military battledress, it would put an end to all wars.

But some of it is, indeed, great. The Marseillaise and Chant du Départ are, evidently, beyond any judgment – history HAS judged, and according to numerous contemporary testimonies the Marseillaise was an essential factor in many victories of the French popular armies in the 1790s, either by whipping them to renewed and frenetic courage, or by impressing the ennemy troups. So I don’t know if it is mighty good music, but it certainly is good mighty music. But the undisputable masterpieces are Gossec’s Marche lugubre (track 2) and Cherubini’s Funeral Hymn (March) on the Death of General Hoche (track 19). Méhul’s “Funeral Lament to the Memory of Féraud” (track 17) is also a superb funeral cantata. I’ve already mentioned Gosssec’s “Chant du 14 juillet” (track 8) sounding like a chorus of Sarastro’s priests. I would have liked for Pichaureau and Robert to play the version with military band of Gossec’s “Hymn to Voltaire” (scoring is clarinets horns and bassoons, as in Catel’s “Ode on the Warship…”), but the a-cappella version is quite beautiful, sounding like appeased Mendelssohn (track 18). That only verse 1 & 4 out of the hymn’s seven are sung is not a problem, due to the song’s strophic and repetitive nature.

Interpretively, Pichaureau is mostly very good. Where his program duplicates that of the old Guilde/Accord CDs, his singers are better and his interpretations usually preferable. His Marche lugubre is impressive, given a truly burdened, funeral and lugubrious atmosphere, on the opposite pole from John Wallace’s blasting and devastating revolt against death (and twice as slow), but equally convincing. In Gossec’s “Peuple éveille-toi” he is unfortunately heavy-handed and sluggish, solemn rather than enthusiastic and here Langlet on Adda and even the old Guilde with its piercing clarinets are preferable, but it is the exact opposite in the “Chant du 14 juillet”, taken by Langlet at an impossibly funeral tempo and where Pichaureau sounds just right. Duvernoy’s Marching Drill is slightly more lively with Pichaureau than with Langlet but not decisively so, and Gebauer’s is played more playfully by Pichaureau (music for an equestrian parade of prancing horses by the Republican Guard), more militarily by Langlet; despite saucy little flutes, the other Gebauer Marching Drill is sluggishly played by the winds of the Air Force conducted by Capitaine Paul Liesenfelt on the old Guilde discs, and here Pichaureau is evidently preferable. In the orchestral introduction of Le Chant du Départ Pichaureau’s brass sound disappointingly dulled and lacking impact compared to Langlet’s, as if surrounded by a thick layer of foam, but on the other hand his woodwinds are deliciously piercing. In the Marseillaise his chorus sounds untidy, which isn’t a problem for a chorus of revolutionary “sans-culottes”, but also undernourrished. These guys would have been badly outnumbered at the battle of Valmy – but hey, they were, and thanks to the Marseillaise they won anyway.

So, as I said in introduction, despite all my issues with authenticiy, as of today, other than the complete 3-CD set from which this compilation is excerpted if you can find it, you will find no better than this if you are interested in that kind of repertoire. Note however that one drawback of the Erato disc is that its booklet doesn’t include the texts, not even detailed descriptions of each work, only a general presentation by Robert. It’s not so much a problem if you understand French, the singers are always intellegible.

Now I’ll wait until 2089 to hear all these pieces played as they were three centuries before, using modern editions of the original scores and reconstructed serpents, tubae corvae and buccins. Hey, performers, that leaves you more than 70 years to learn how to play them perfectly. Start NOW!

Additional resources
See my Gossec composer page for more on Gossec and more Gossec

A highly useful article from Centre de Musique Baroque de Versailles by Gossec specialist Claude Role on the music written by the composer for the ceremonies and occasions of the French revolutionary era, “François-Joseph Gossec: Musique de la Révolution” (link will open .pdf document in new tab – document can be freely downloaded from Philidor, Centre de Musique Baroque’s online library, but it is impossible to provide a weblink to it, so I’ve taken the liberty to download it here).

The book of Marie Mauron on The Marseillaise, Librairie Académique Perrin, 1968, is often quoted as a recommended source on the topic, but it isn’t very erudite, more in the nature of story telling, and I haven’t found it very useful.



Erato’s annotator and musicologist Frédéric Robert is considered the French specialist of the Marseillaise. In 1989, Imprimerie Nationale and Les Nouvelles Editions du Pavillon published a lavish book on the subject by him, “La Marseillaise”, ISBN 2-11-081031-9  – again, it is in French only. It is very useful, the iconography is splendid, it includes many documents on the history of the hymn, a discography, a filmography. But Robert says almost nothing on the matter of Gossec’s orchestrations. I am a bit uncomfortable with Robert’s vagueness here, and his insistence at presenting as “Gossec’s orchestration” what research shows is in fact Dondeyne’s orchestration of Constant Pierre’s piano reduction.

Comments are welcome