French Vocal Berlioz on Record

(Perspectives on authentic French Berlioz, particularly 3 Weimar Cellinis 
and Munch’s Paris 1953 recording of Roméo et Juliette)


Yoël L. Arbeitman

© 2014 Yoël L. Arbeitman

1. Berlioz, whether attending in a concert hall/opera house any of his non-liturgical, large scale vocal works or listening to/watching them at home, often brings me, in addition to the special pleasures, some degree of misery. Quite simply put, the revival and the recording of Berlioz took place too late, after the distinctly French singing style – in language and in text-expression – , in most cases, essentially had ceased to exist (or, perhaps in the best cases, had severely diminished). For the concept of "globalized style" and striking examples, see London Green in Opera Quarterly 19:541 and especially 543. See my review article "Hector Berlioz: Early Vocal Recordings" (this site) for single arias from a much earlier period, when no "complete" vocal Berlioz was recorded. For me French works not only require the language and the style (and, as part of this, there needs to be cathexis) but, as a third sine qua non, the singer’s Fach needs to be suitable. On June 9, 1892 the Opéra-Comique had its premiere of Les Troyens à Carthage. Edmond Clément was Hylas and Léon David was Iopas (see the notes by Victor Giraud for the Romophone CD set "Clément: the complete Pathé recordings and David: the complete recordings"). In John Steane’s notes to the Romophone CD set "Clément: the complete Odéon and Victor recordings" we read that "[Clément] is a useful archetype, as good and pure an example of the French lyric tenor as any. As heard in his recordings, Clément’s voice has no admixture of baritone or of heroic timbre, and he belongs unmistakenly to the national school". He goes on, but obviously one cannot describe that kind of French singing; it must be heard. To our immense loss, neither Clément nor David left any souvenir of the 1892 Troyens. Paul Lewis (notes to Hector Berlioz: Early Vocal Recordings [HB:EVR], see 8. below) reminds us that Maurice Renaud, who sang Chorèbe to Delna’s Cassandre in the Paris 1899 La Prise de Troie, did not record anything of Chorèbe. However the Marston set "Maurice Renaud: the complete Gramophone Recordings 1901-1908" does include his seven Mephistophelean interpretations, namely "Voici des roses" x 3, "Maintenant, chantons à cette belle...Devant la maison" x 2, and "Vrai dieu,...Une puce gentille" x 1. Only the third recording of the second aria is done without its recitative. The Didon of those performances, Marie Delna, did happily leave a 1903 recording of "Chers Tyriens" (see now the Marston CD set of her complete published recordings), while Clément left a 1919 recording of "L’Absence". Additionally, each of these singers’ many non-Berlioz recordings do let us imagine their Berlioz including the Troyens in which they partook. There are however singers, who sang to great acclaim as reported by contemporary reviews, whose recordings do not indicate such ability. Such is the matter with Charles Rousselière who performed Énée to Félia Litvinne’s Didon under Edouard Colonne in 1905. Yet his 1907 "Nature immense" and his other 1903, 1905, 1907, and late 1920s recordings fail to convince.

2. As the Berlioz revival came so late and predominantly through a Scottish and English tradition (and English language), there are essentially for the majority of his major vocal works (outside of Damnation) no complete or even condensed recordings equivalent to the glorious and authentic recordings that do exist for e.g. La Favorite, Faust, Samson et Dalila, Werther, Manon, Pelléas et Mélisande, Pénélope (with Monmart and Jouatte), even Louise (abridged), Carmen (to be noted are the 1911 François Ruhlmann and the 1950 André Cluytens which use Bizet’s dialogue rather than Ernest Guiraud’s recitative—see 6.3.1.- 6.3.2. below), and various all but forgotten works. To be fair to "the other side", it should be noted that this 1941 recording of Pelléas (as well as the 1927 excerpts noted in 3. below) have casting "more Catholic than the Pope’s" in that Debussy’s first two Mélisandes were resp. the Scottish Mary Garden and the English Maggie Teyte, both almost surely on the basis of the perceived cathexis that each infused into her representation. The reader will excuse me if I moderately contradict my opening categorical statement of this paragraph in the process of this review. As for Les Troyens, there are four precious authentic French recordings to consider. Of these four, the Thomas Beecham and the Hermann Scherchen offer more or less complete performances of Part II (Les Troyens à Carthage), the former alone offering Part I as well (La Prise de Troie). The Henri Tomasi offers excerpts of Parts I and II and Willem van Otterloo offers but excerpts of Part II. Certainly the Cassandre/Didon of Beecham’s 1947 Troyens, Marisa Ferrer (1896-1972) is a God-send from another time. She had performed Didon with both Paul Franz and Georges Thill (1929) and with José de Trévi (1931), and Cassandre with the last named (1932 and 1938), the tenor whose abridged recording as Faust in Damnation (1930) under Coppola is for me definitive. Ferrer was 51 at the time of the Beecham recording; she remained magnificent at this age. In neither the Beecham nor the Tomasi performance does she do the repeat in Didon’s entrance "Chers Tyriens, oui, vos nobles travaux..."; this is almost certainly because this cut was normative in the performances of the 1920s and 1930s. For the information on these Troyens performances I am dependent on Louise Goldberg’s appendix "Select List of Performances" in Ian Kemp’s Hector Berlioz, Les Troyens, Cambridge opera handbooks, 1988 (this appendix is an abridgment from her 1973 dissertation. The Franz, Thill matter for 1929 is somewhat confusing as the Énées seem to have alternated as indeed did the Didon, Ferrer with two other singers, a not unusual occurrence of any performance run). Yvonne Corke as Beecham’s Anna has a timbre almost the same as that of Ferrer and this makes for a very different experience in the sisterly Recit and Duet from that of the Tomasi performance (see forthwith). Charles Cambon is both Chorèbe and Narbal. Different singers assume the role of Ascagne in Parts I and II; Maria Branèze of Part II offers a far more effective Ascagne than Irene Joachim does in Part I. The Beecham recording is issued on (Malibran CDRG 162; Somm-Beecham 26-8; Mike Richter CD-ROM "Sir Thomas Beecham", ae103; and nowadays, of course on Youtube). The most recent Somm issue claims to be a definitive, "authorized" transfer; it alone has corrected placement of the "Chasse royale et orage", as Beecham played it, and complete cast. Most regrettably their issue sounds like raw material with no remastering of the historic tapes. Their listing of (Charles) Gambon needs correction to Cambon. An absurdity of this issue is their following the seemingly unbreakable British way of entitling the CD set "The Trojans". Richter’s alone (made by a mutual friend, derives from three sources, patched together to obtain a good result) has the full broadcast announcement, following extended applause, at the opera’s end. Alas, the Énée of Beecham’s recording falls somewhat short, not as to French language and style (of which he is a fine example), but as to Fach. Jean Giraudeau was a lovely, (somewhat) light lyric tenor. We are blessed that a recording survives of Ferrer in a Nov. 22, 1951 abridged Les Troyens (Goldberg lists only a Part I; I have only a Part II; both were performed). This Les Troyens à Carthage runs 86’39" and offers major parts of Acts III, IV, and V with a cast that one dreams of: Didon: Marisa Ferrer, Énée: Georges Jouatte, Anna: Hélène Bouvier (of the noted 1946 Samson et Dalila recording), Narbal: Charles Cambon, Hylas: Joseph Peyron (cf. 7. below re Roméo et Juliette), Ascagne: Nadibe Sautereau, Orchestre de la Radio-lyrique, cond. Henri Tomasi, at the Salle Erard (Paris). It is impossible to overstate the wonders of such performers in this work. With Peyron’s Hylas - esp. in the very ending, the dormition - we are in the presence of vocal magic. Even the unnamed Mercure supremely does his task. The nexus between the lovers’ duet and Mercure’s twofold striking of Énée’s shield/ threefold annunciation of Didon’s doom has rarely, if ever, sounded so chillingly effective. The golden-voiced (unnamed; could be Peyron taking these lines?) Iopas sadly only gets to deliver his lines in No. 46 (perhaps also participates in some ensembles), but his poem is omitted. Ferrer and Bouvier offer the sisterly duet (No. 24) as a true heavenly blend. Bouvier is a true French contralto and the contrast with Ferrer is most effective. If Jouatte alone falls here and there a tad short of either heft or tessitura, he remains a very fine Énée and far preferable to most. Tomasi guides the whole with a loving hand (to this conductor we also owe the unique French mezzo recording [1935, abridged, featuring Alice Raveau] of Orphée [see Roland Graeme "Orphée et Eurydice, Christoph Willibald Gluck, ed. Berlioz", Opera Quarterly 19:555-560]). Beecham’s Énée, Giraudeau, also performed as the Énée in Scherchen’s May 29, 1952 live recording of Les Troyens à Carthage (Tahra TAH 143-440), where he is paired, once again, with an extraordinary Didon, the Armenian mezzo Arda Mandikian (born in Smyrna/Izmir, found refuge in Greece; engaged 1950 by Jack A. Westrup to sing Didon in the production of Troyens by the Oxford Univ. Opera Club, excerpts of which were released on 78 rpms which I have not heard). This little remembered singer again displays her fit for Didon in abridged (sic) excerpts of Les Troyens à Carthage from a June 24, 1952 concert broadcast from The Hague under Willem van Otterloo. Van Otterloo’s very engrossing reading of Berlioz has been released on Gala (100.630), where it occupies c.1? CDs (approx.100 mins.) of a 3 CD set, the rest being devoted to Gala’s own abridgment of the June 20-21, 1985 Melbourne performance of The Trojans, which Gala labels as "abridged version" although the complete (as given) performance was previously released on a small, private label. For this English language performance Gala oddly (and wrongly) lists Sinon rather than Hélénus in Nos. 8 & 9. Returning to van Otterloo’s performance, abridged excerpts as it does consist of, it is highly worthy of recommendation for conductor and for at least some of the singers, some of whom are French, others not. The cast consists of Didon: Arda Mandikian, Énée: Jean Voyer, Anna: Inès Chabal, Iopas: Paul Derenne, Ascagne: Mattiwilda Dobbs, Panthée: Guus Hoekman, Narbal: Jean Claverie, Hylas: Michel Sénéchal, Radio Filharmonisch Orkest, cond. Willem van Otterloo. Voyer is an unusual Énée, truly a heroic tenor of powerful, sometimes overwhelming presence, somewhat reminiscent of Ramón Vinay. I find him quite impressive here although I am certain that some others will find him not suited to the role, heroic tenor, but not a French fort ténor. Derenne’s Iopas is very engaging; his presentation of the poet works well; he employs head voice –in the fourth repeat of "[tu] fais" in the final stanza which begins with "féconde Cères"–to good effect. The Ascagne of this American soprano is excellent and very French in her use of the language. Sénéchal’s Hylas is good enough. Very little of the roles of Anna or Narbal is present. Other than Mandikian and Giraudeau, Scherchen’s fine cast members are mostly unknown to me. As opposed to the Beecham casting, in the sisterly Recit and Duet, Scherchen’s Anna, Jeannine Collard, is a recognizably darker mezzo than her sister (not a contralto however as we have in the Tomasi excerpts). Scherchen plays the sisters off against one another in a sui generis manner, one that offers its own special enchantments. Scherchen’s Narbal, Xavier Depraz, fine singer/actor that he is, expresses little when one has just heard Charles Cambon (not a fair way to listen, I’ll grant) and does not seem well suited to Narbal. While Ferrer only recorded her interpretation relatively late in her career, Georgette Frozier Marrot, who performed Didon in French venues other than Paris (at the same period that Ferrer was so doing in Paris with De Trévi, Thill or Franz), recorded on March 11, 1929 Didon’s two arias in a deeply moving sacramental representation of the queen’s life, office, and career (HB:EVR). Her Enées were apparently René Lapelleteri and Victor Forti (the reference in Goldberg, p.219, to "Mmes Frozier, Marot [sic]... " most likely refers to this singer and not to two separate singers in two roles).

2.1. Even as all performed and recorded Cellinis long suffered from Weimaritis and indeed (but a lesser suffering) from the several stages of the Paris forced revisions and, a century later, from the "improvements" of English scholars (see 6.3.1. below), so too almost all Troyens performances, recorded or not, continue to suffer from Royer’s forcing a Troyens Paris II on Berlioz. MacDonald’s restoration of the essential Sinon scene has been observed in a few recorded performances, first and foremost in the video (now on private DVDs) of the Sept. 29, 1987 Lyons Berlioz Festival performance conducted by Serge Baudo, followed by the unimpressive October, 1993 Dutoit recording, and most recently, by the broadcast of Nelson’s Oct. 28, 2010 performance from Amsterdam. Macdonald’s reconstitution of the lost orchestration was first performed Nov. 27, 1986 at Opera North in Leeds. The 1987 Lyons production manifests that the forced mutilation totally alters the reality of Act I and perhaps of Part I as a whole. With an enticing, albeit almost totally non-Francophone cast, it is highly recommended for (1) the reality of its Sinon and Priam and (2) the presence of yet another dramatis persona: Hector Berlioz, conceiving his gradual creation of Troyens, is the sole presence which we, the audience, follow while the orchestra does the Chasse royale et orage; and (3) the one totally Francophone being is the unique Hylas of Antoine Normand. The restored Paris I of Troyens needs finally to be accepted as the composer intended it. Troyens, Paris II (Hélènus as a poor substitute for Sinon) needs at long last to be acknowledged as the unhappy concession it was. La Prise de Troie without the Sinon pericope represents as mutilated a form as does Weimar Cellini. Nevertheless, for both operas, we have many appealing recordings which do follow mutilated versions, forced on the composer and herein we accept such for their interpretive values (far more so in the case of Cellini).

2.2. It would seem necessary to make note of Guy Chauvet here. In the period with which the entirety of the present review is concerned, he was essentially the only Francophone tenor who had Énée (and Faust of Damnation) as roles he often sang. I have heard five complete, abridged or excerpts performances of his Énée (1964 Buenas Aires [Bruder abridgment], 1965 EMI [CDM 7 63480 2] studio excerpts, the live 1976 Vienna abridged performance [Gala GL 100.609], the1977 complete live Part II Cincinnati May Festival, and Part I only of the Nov. 27, 1983 Hamburg performance in an in-house recording; Cassandra: Karan Armstrong, Chorèbe Mikael Melbye—rest of cast unknown to me—cond. Henry Lewis). None of these five affected me emotionally or intrigued me analytically. To me he seemed a sloppy and somewhat vulgar Énée; perhaps a singer whose potential did not fulfill itself. I therefore find it surprising that Kenneth Meltzer on the ClassicalCDReview site could say in Sept., 2004, concerning the Malibran Chauvet arias CD, "And if all of these attributes inspire comparison to Chauvet’s great predecessor, Georges Thill, the singing on this disc justifies such comparisons". I have not heard the CD in question which includes Énée’s great aria and two arias from Damnation. Of the numerous Énées I have heard and/or seen (in French-, Pseudo-French-, English-, German-, Czech- or Swedish-language performances), while some are reasonably good and others are dreadful and many (or most) are indifferent, only one has ever expressed Énée or Aeneas for me besides Thill and that is Josef Traxel in German (1960 abridged radio performance of Die Trojaner under Hans Müller-Kray). There are fine recorded performances of German language Die Trojaner and English language The Trojans; these I appreciate on their own terms, far more than I do "mock French/Pseudo-French" and the like. As to "the real Énée’s" own Trojan tongue see Yoël L. Arbeitman 1986, "Trojan, Luwian, and the Mass Media, 1985 (C.E.)", Diachronica III.2:283 – 291. Let us leave Troy – it is a vast territory – and proceed now.

3. Happily in recordings of La Damnation de Faust we do indeed have three just about perfect recordings in the noted abridged (79’54") 1930 Piero Coppola (José de Trévi, Mireille Berthon [who also in 1930 was Gounod’s Marguerite in Henri Busser’s classic recording of Faust], Charles Panzéra, Louis Morturier [1930 was also the year that Morturier recorded his abridged Herod’s aria from L’Enfance], Chorale de St. Gervais [Le Flem], Orchestre des Concerts de Pasdeloup [various CD issues; the private transfer from three sets of shellacs I now have by W.G. is far better than Pearl’s]). We—the subset of Berlioz lovers to whom French language and style are vital—owe much gratitude to Coppola for this; equally for his 1927 c.50 mins. of Pelléas (with Panzéra, Yvonne Brothier, and Vanni-Marcoux), his 1927 complete Carmen (with Brothier and De Trévi inter alios), and his 1933 French duets from Lohengrin (Marjorie Lawrence [Ortrud], Brothier [Elsa], Martial Singher (Telramund]); the 1942 Jean Fournet (Georges Jouatte, Mona Laurena, Paul Cabanel, André Pactat, Chorale Émile Passani, Grand Orchestre de Radio-Paris [Dante Lys 319-321 – also containing Fournet’s 1943-44 recording of La Grande messe de morts, both works transferred from Columbia 78 rpms – with important notes and artistic direction of the 1998 issue by Philippe Morin; cf. end of 7. below]); and the 1948 broadcast under Albert Wolff (Georges Jouatte, Suzanne Juyol, André Pernet, Charles Cambon, Orchestre et Choeurs Radio-lyrique). I have it for years in a private transfer (done by a friend) from the French Radio archives which, at that time, sold a tape of anything on demand, but only in person; it was subsequently (2003) issued on Malibran, an issue which I have not heard. I would unhesitatingly say that Mme. Juyol is the unrivaled Marguerite recorded in either an abridged or a complete performance. In the 1950s the 1942-44 Berlioz/Fournet recordings were transferred to LP and were available for a while, but then disappeared from circulation. We cannot take leave of French Damnations without making note of the July 14, 1960 Dutch radio broadcast from Tivoli, Utrecht, under Henk Spruit (Jean Giraudeau, Rita Gorr, Pierre Mollet, David Hollestelle, Groot Omroepkoor, Omroeporkest). Giraudeau’s Faust is sui generis, one that is elegant, self-examining, meditative, and caressing each vowel and consonant of every syllable with love, causing us, the listeners, to inhale his French as one savors fine poetry and wine. It is this that causes one of my most concerned friends to admire his Énée, an interpretation that doesn’t work for me, whereas his Faust does. To me, listening to his Faust, Giraudeau is antipodal to Chauvet in everything that matters. Giraudeau and Mollet play off one another to good effect. Rita Gorr’s huge and steel-like voice hardly succeeds in portraying the young, naïve, readily seducible virgin; rather her voice here portrays maturity and an almost matronly woman. She portrays Marguerite with far more sensitivity in the 1961 Angel LP of Damnation excerpts (Nicolai Gedda, Gérard Souzay, Orchestre du Théâtre National de l’Opéra de Paris, André Cluytens; re-issued as part of a 10 CD collection of excerpt recordings of French operas [EMI 5 73089-2]; the Damnation excerpts are now downloadable from But then again the context may make the difference as these excerpts are separate arias for each of the soloists who never sing together. Certainly the yoking together of Giraudeau and Gorr was a poor choice. Giraudeau’s French world of singing is more properly yoked with Solange Michel’s. Gorr –well – she was of a different world. She did record Samson et Dalila with Jon Vickers after all. Gorr was ideally yoked (in an in-house abridged recording of La Favorite) with Tony Poncet, a totally different style of French singing. In 2003 EMI Classics issued a 10 CD compilation box entitled "Le Chant Français: 1948-1965, les années Pathé". The set includes six Berlioz arias. However here I only wish to comment on two of these, both of Marguerite’s arias: CD 2 begins with Régine Crespin’s "D’amour l’ardente flamme" (cond. Otto Ackermann, 1963 or 1965, confused documentation) and ends with Solange Michel’s "Recit and Le roi de Thulé" (cond. André Cluytens, Feb. 23, 1953). The book-ending of this CD by these two performances expresses all that needs to be said about internationalized vs. French style. Mme. Michel’s recording is worth the price of the 10 CD set (the other Damnation item in the set is Michel Dens’ 1956 "Sérénade"). Roland Graeme, in an admirable review article of 16 recordings of Damnation that are on CD in Opera Quarterly’s Berlioz Bicentennial issue (vol. 19:431-445) comments with detailed perspicaciousness on each of them while also adducing several LP issues not on CD. I am in agreement with most of what he says. However he more or less dismisses the Coppola and, of the 1942 Fournet, he says that he hasn’t heard the "1948 Columbia", while the 1950s Decca and Vega LP excerpts he refers to (with no details) are indeed resp. the 1954 Fistoulari (see 7. below for my comments) and the 10" French LP, Pléiade P3082: La Damnation de Faust, excerpts with Ninon Vallin and Guy Fouché, Orchestre des Concerts Pasdeloup conducted by Pierre Cruchon, March 1, 1955 (see my 2008 review article on historical Berlioz, this site; Fouché at his prime and Vallin past hers [she first sang the role in 1918] produce magic). Twenty-one years after his ground breaking 1942 recording, Fournet conducted a broadcast performance of Damnation (Chauvet, Crespin, Roux, Peter van der Bilt. Groot Omroepkoor, Radio Philharmonisch Orkest, March 23, 1963). Here we have perhaps a perfect representation of the period when, amongst Francophone singers and conductor, French style had waned, globalized style had fully waxed. This performance has been issued on Bella Voce (BLV 107.202) where (1) the tenor is mis-identified as Guy Fouché who indeed recorded the significant excerpts with Vallin (see just above) and (2) the two bonuses (Marguerite’s solos) given are mis-labeled as Crespin (no date, no conductor), but are actually Rita Gorr. In the performance proper, the three Francophone singers are of much interest. For Crespin, to my ears, it is the performance in which she most displays an understanding of Marguerite and has suppressed her diva-ness. A very fine portrait. Roux displays more a rough side than any Gallic suavity and sardonicity inherent in the character and Chauvet, by trying to force his lyrical instrument again and again to be a dramatic tenor, a spinto +++, fails to produce any consistent beauty or insight. Alone in the trio, in those very lyrical lines for the tenor (against Méphistophélès’ vehemence), does he produce anything pleasant. To hear what Chauvet seems to have unsuccessfully been aiming at, listen to the June 23,1956 broadcast of the duo and trio from Damnation with Gustave Botiaux, Jacqueline Lucazeau, Paul Cabanel (a triad of glorious French singers), Orchestre Radio-lyrique, cond Pierrre Delvaux (available inter alia on Mike Richter’s CD-ROM "Opera from Paris - Performances from the French Radio - ae 202" and, of course these days, together with much other Botiaux on Youtube). On June 1, 1983 Austrian radio broadcast a Damnation from the Vienna Festival, one with ¾ Francophone cast, Orchestre de Lyon, conducted by Serge Baudo. At age 55 Alain Vanzo had matured into a lyrico-spinto and sang Faust with the heft and ring that are alone absent in his Cellinis of 14 and 10 years earlier (see 6.2 & 6.3 below). Vanzo expresses every self-delusion, every aching desire of the old philosopher. The Brander of Jean-Marie Frémeau excites. Jean-Philippe Lafont’s Méphistophélès lacks the deepness and darkness to present the seductive evil. He comes off as merely a local guy who is taking an acquaintance bar-hopping and girl-chasing. Felicity Palmer gets Marguerite’s naïve, baneful lust just right in fine French. The performance omits the "choeur des soldats", 2nd part of scene 15. Serge Baudo generally keeps an excited pace with his orchestra. The entertainment in Auerbach’s Cellar gives near perfect Gallic sardonic smirks at rats, fleas, humans, devils, etc. in spite of the less than interesting Méphistophélès. There are brief bios of Chauvet, Fouché, Vanzo, and Botiaux and others on French Wikipedia and nowadays there is always Youtube for forgotten great singers of the past. Conductorially Fournet’s performance is methodical, consistent, metronomical, and plodding. It lacks excitement, but is a very interesting document. Finally here are two relevant quotes from Graeme’s article: "One striking feature of [Damnation’s] discography is the relative absence of native French-speaking solo singers: in this respect, La Damnation de Faust has become thoroughly internationalized" (p.436) and "Yet the selection of studio versions at full price on CD is, in fact, not all that impressive" (p.445). Agreed, but I suggest that we change what recordings we are looking at. There are of course numerous other broadcast Damnations beyond those here surveyed but it is beyond the scope of this review article to adduce them, let alone discuss them.

As concerns Huit scènes de Faust, the interpretive singing of Jean-Paul Fouchécourt as Méphistophélès, in and of itself, is good reason to strongly suggest the 2000 Erato CD (8573-80234-2) conducted by Yutaka Sado (Brander is Frédéric Caton, while Marguerite is performed by a very non-French Angelika Kirchschlager). Here, in addition to the "Chanson" and "Sérénade" of Méphistophélès, No 2, "Paysans sous les tilleuls" is given to Fouchécourt, with the chorus only repeating the refrain. I am not acquainted with Dutoit’s recording which, apparently, uses no Francophone singers at all. Rather absurdly in the 2003 EMI CD (7243 5 57433 2 9) "Berlioz: Roberto Alagna", where this tenor undertakes to record arias of every character which Berlioz wrote for tenor (he does forget Hylas and there is room on the CD for him), in the section of five tracks devoted to Damnation, Méphistophélès’ "Sérénade" from Huit scènes is placed between "Merci, doux crépuscule" and the duo "Grands dieux … Ange adoré", thus having Alagna sing both Faust and Méphistophélès in a conflated *La Damnation de Huit. Just compare Fouchécourt and Alagna in the "Sérénade" and one might reasonably conclude that being Francophone does not, in and of itself, make for French singing. Fouchécourt proves that French singing still exists. It is interesting (I have no further information on this) that in Sado’s complete presentation of Huit scènes, as Méphistophélès Fouchécourt concludes the work with the "Sérénade" proper "Devant la maison", while on the aria CD, Alagna’s singing of this "Sérénade" is preceded by c. 14 seconds of the equivalent of the Recit to this aria in Damnation, here spoken (parlé) by an actor, Gérard Depardieu. On neither recording does the line of Valentin appear (see French libretto this site). On Jan. 19, 2013, upon finishing an earlier draft of the present article, by chance I became aware of the 6 CD Danacord set (DACOCD 691-696) Great Singers & Musicians in Copenhagen, 1931-1939 (Historical Live Recordings from the Archive of Danmarks Radio), CD 2 (692), tracks 3-6 of which contain the 11’23" excerpts from Damnation, put on very temporary recording material by Danmarks Radio for "delayed presentation" in their Thursday concert series, but surreptitiously preserved for eternity by Danmarks Radio’s head of technical services, Frederik Heegard, on copper-plated matrix stampers, acting against all legal regulations. These Nov. 2, 1933 excerpts, performed by Georges Thill, André Pernet, Danish Radio Choir, Danmarks Radio Symphony Orchestra, cond. Egisto Tango, consist of "Villes entourées", "Merci, doux crépuscule", "Esprits des flammes", and "Maintenant, chantons à cette belle...Devant la maison" (the latter through the 2nd "ha!" and with chorus). While a very misspelled "Villes entourées" is listed, the Danish chorus actually sings only the joint "Choeur de soldats et chanson des étudiants" ensemble. As a lover of Berlioz and adorer of M. Thill for approximately ¼ a century, I have been ignorant of these excerpts until – with the greatest of joy – I heard them on Jan. 31, 2013. They and all the material of the CD set were first issued on LP in 1983. Thus we have a 2nd aria from Damnation with Thill in addition to his well-known two studio recordings of "Nature immense" (1925 acoustic, 1927 electric), the only Damnation aria recorded by a fair number of Golden Age Francophone tenors. The fact that these excerpts include "Merci, doux crépuscule" and not another "nature immense" makes them all the more invaluable. We also have these two Mephistophelean arias with Pernet to compare to his participation in the complete 1942 Fournet recording. Miracles do happen. For those not interested in the whole preserved set, the Damnation tracks are available by download from and presumably likewise in other countries.

4. As for L’Enfance du Christ, Pathé recorded a vocally as near a perfect French cast as we will ever find, under André Cluytens in 1950 or 1951 (first released in 1954). To my knowledge, it has not been issued on CD and this is often the case. Michel Roux is a most convincing Hérode. La Vierge Marie is Hélène Bouvier (1905-1978), the Anna of the Tomasi Troyens (see 2. above). Joseph is Louis Noguéra, Le Père de famille is the amazing Henri Médus, and Le récitant is none other than Jean Giraudeau, and others (Centurion and Polydorus not specified); Choeurs Raymond Saint-Paul, Orchestre de la Société des Concerts du Conservatoire. Allow me to say in the most sincere and loving manner that there is something near "nauseating deliciousness" about French sung as the great singers of earlier periods did sing it. The same applies to the Russian language in the Soviet recordings from the 1930s, 1940s, and 1950s of Russian operas. But this recording of L’Enfance has not only such soloists, but indeed a conductor who brings to this work, a work which can be somber and somniferous, a Berliozian understanding of what makes up such a miniature item by this composer and proceeds to work a special magic on it. He understands that here the orchestra, as ever in Berlioz, never accompanies the vocal soloists, but duets with them but on a separate plane and the two planes produce that Berliozian effect. So much of the music is purely orchestral and this Cluytens lets us hear in a special way. L’Enfance, miniature Berlioz that it is, is fully entitled to the standards of all great Berloz works. In the end we have here a recording of a Berlioz vocal work which is fit that we lay alongside those special real, non Pseudo- French performances of Berlioz vocal recordings. Only one other recording of a vocal Berlioz work will we find later on to add to this special class of recordings which included three Damnations so far.

5. In 1969 French Radio broadcast two performances conducted by Pierre-Michel Le Conte (presumably to commemorate the centennial of Berlioz’s death), respectively of Béatrice et Bénédict (complete minus any dialogue, equivalent to the first Davis recording [1963] of this work, give or take a few words) and a Weimar Benvenuto Cellini. Unfortunately I don’t know the exact dates and which came first. At the conclusion of the performance of the former, the announcer days that this complete performance of Béatrice et Bénédict is being presented to you "Dans le cadre de la semaine Berlioz" ("within the scope of Berlioz Week"). The former ranks as the uniquely French performance on record of Berlioz’s last opera. Indeed to my ears the triad of Béatrice, Héro, and Ursule (namely Berthe Monmart [who, in a 1965 broadcast of an abridged concert Troyens from Paris under Le Conte, sang both Cassandre and Didon; I have not come across this broadcast; see Goldberg and cf. further 6. below, "second performance"], Claudine Collart, and the vocally gorgeous Renée Donsoit) is nearly too good to be believable; but there they are. The rather small parts of Don Pedro and Claudio are finely done resp. by Jacques Mars and Pierre Germain, while Louis Noguéra’s (just noted as the Joseph in L’Enfance) Somarone is quite fetching indeed, Choeurs de l’ORTF, Orchestre Radio-lyrique. My only reservation is that the Bénédict of Michel Sénéchal, while unobjectionable, to me is not on the level of the three ladies or of the performance as a whole.

6. We shall consider Le Conte’s Cellini in chronological order along with two other, more or less French performances of the Weimar version, ranging from 1964-1973. For convenience, references to loci in the opera will also be given to the corresponding acts and scenes of the Paris I version. These three performances have in common the consistently enchanting Teresa of Andrée Esposito (the finest French [dramatic, coloratura] soprano of her time, born in Algeria, studied under Noguéra); two of them have as the title character "two different tenors" named Alain Vanzo, while two others have the very welcome presence of the French bass André Vessières in a role (not the same role). Inasmuch as one of these performances included Gedda’s first recorded (performed?) Cellini, we shall also refer to other Gedda performances of the role. It needs to be noted – before anything – that Vanzo represents a world of singing so different from that of Gedda or anyone else who has assumed the role that no fair comparison could be made. Vanzo in either performance offers a very French pulchritudinous expression of Cellini’s character. Additionally he imbues his portrayal with passion. These obtain whatever problems I may find with the interpretation.

6.1. The first of these three performances is that of Nov. 12, 1964 (probably a house recording) with the Orchestre de la Suisse Romande and the Choeur du Grand Théâtre de Genève, conducted by Ernest Louis de Froment. The cast consists of Benvenuto Cellini – Nicolai Gedda, Balducci – André Vessières, Fieramosca – Jacques Doucet, Le Cardinal Salviati (in all these Weimar performances in place of Le Pape Clément VII) – Nicola Ghiuselev, Bernardino – Georg (Giorgio) Pappas (Greek bass, primary career in German speaking countries), Francesco – Pierre Voillier, Pompeo – Olaf de Wyzanowsky, Cabaretier – Hughes Cuénod (as in Davis’ July, 1972 studio recording of the Paris II-dialogue version; obviously ditto for Gedda), Teresa – Andrée Esposito, Ascanio – Michel Hamel, un officier – Jean Angot. This recording was issued in 1965 by the EJS live/ broadcast performance LP label. There are several oddities here which require noting before anything else. (1) Almost anyone who has listened to many Troyens/Cellinis has had the thought as to how a tenor Ascagne/ Ascanio (rather than a mezzo) would come across. Indeed most opera listeners have had such a thought about some or all (specifically adolescent) pants roles (think of Jemmy, Cherubino, Siebel, Adriano [there is a recorded performance of Rienzi with tenor Josef Traxel as Adriano] or Octavian and of course Debussy’s Ysolt is far younger than adolescent). It is a role of an adolescent male written for the female voice (and thus a different matter from what we, today, do with castrato roles). Our Ascanio is sung by a somewhat nasal, very French tenor who is most engaging. It is therefore a great pity that he is given so little of this central role to perform: (a) primarily he is allowed the "Cette somme est due... mes amis"; here and within the whole Inn scene he continues to be good, however he keeps on shocking each time I had to return to these loci during this review, shocking me by his very tenorhood/maleness; (b) the duet-prayer with Teresa however is not written in unison for the Teresa soprano and the Ascanio mezzo, but rather with Ascanio doing the selfsame lines a tad behind Teresa. In this joint prayer for their beloved Cellini, Hamel is too restrained with the result that only, when the listener makes the effort, does he succeed in perceiving the tenor’s participation; and one of the chief delights of the opera, "Mais, qu’ai-je donc?", is simply deleted. Additionally, instead of the male adolescent apprentice (for a suitable mezzo), we hear another adult dramatis persona, as though he is not pushing his master, but rather challenging him. Ascanio, esp. for this 2nd aria just noted, is the most bravura part in all of Berlioz. It would take a tenor such as Juan Diego Flórez or Jean-Paul Fouchécourt to do justice to him, and Hamel, fine tenor that he is, is not that kind of tenor. In the opposite direction, I am happy that I have never heard the contralto option for Hylas. (2) The title character is sung by a non-French singer as is the Cardinal. The Bulgarian bass, at the beginning of his career, in which he was famous for the great Italian and Russian (and a few French) roles he performed, gives us a truly dramatic and convincing, if not French, Cardinal Salviati. The Swedish tenor, who throughout his career was associated with French roles, gives us a strange Cellini, one that is overly dramatic or melodramatic, one that sounds more proper to Verismo, at times sounding like Pagliacci (esp. both the Recit and Romance and the Recit and Aria). At other times his presentation of Cellini here brings Otello to mind ("Un autre fondre ma statue! Dieu!" sounding like "Si, pel ciel"). This is quite surprising, as Gedda in his Sept., 1961 studio singles recordings, with L’Orchestre national de la radiodiffusion française, cond. Georges Prêtre, of both Cellini’s Recit and Romance and his Recit and Aria ("Le Chant Français", CD8; see 3. above) sings as the very model of appropriate style (although never sounds more French than he is). However his other live Cellini in the Dec.15, 1966, Covent Garden performance of the Paris II-dialogue version, conducted by Pritchard (Gala GL 100.618) shows somewhat this same totally unbridled Gedda and both the 1964 and the 1966 show a Gedda who is additionally more baritonal. In the 1972 Davis Philips recording, he is again a model of appropriate style as in the 1961 studio singles, but esp. in the Recit and Romance the self-bridling is done with apparent effort. With Gedda and Ghiuselev both non-French and Gedda lacking style here, one may ask what is so French about this performance that I need to consider it here. The Teresa (as in the following two performances), the fascinating Ascanio (of whom we don’t get enough to form an opinion as to how a tenor Ascanio fares), Vessières’ Balducci which simply is the most convincing I have heard; his delivery of the opera’s opening rhymed verses makes me hear what I never have heard in another performance, and the delivery of Fieramosca’s aria by Doucet is nonpareil.

6.2. The second performance to be considered is Le Conte’s 1969, already referred to. My copy begins with a cast announcement in Italian: Benvenuto Cellini – Alain Vanzo; Balducci – Pierre Thau (as also in the May 8, 1973 Ozawa performance [Opera d’Oro OPD 1373]); Fieramosca – Matteo Manuguerra (French baritone born in Tunisia); Le Cardinal Salviati – André Vessières; Bernadino – Stanislaus Staskiewicz; Francesco – Pedro Proença (this is the name I hear on the cast announcement), Pompeo – Albrecht Klora; Teresa – Andrée Esposito; Cabaretier – Jean Mollien; Ascanio – Lyne Dourian (who also sang Cassandre while Monmart sang Didon in a 1965 presentation at the Opéra of Troyens; each mezzo alternated with another, as likewise the Énée during the run; details in Goldberg and cf. 5. above; there is a fair lot of Dourian on Youtube including La mort de Cléopâtre); Choeurs et Orchestre de l’ORTF, chef des choeurs – Jean-Paul Kreder, edition Choudens, live and before an audience. After years or decades of my listening to Weimar versions simply with knowing the Paris II libretto, it was an invaluable tool, as I proceeded in this review, that Monir Tayeb and Michel Austin have put the Weimar libretto on-line. It is on this that I rely for the Weimar versions, although often what is sung diverges in a word (or more) here and there as well as each version deleting certain parts, presumably to meet time contingencies. As to Vanzo (Weimar [Act I], No. 3 = Paris Act I, 4, No. 4 in the libretto to the Nelson [Virgin Classsics 7243 5 45706 2 9] recording of 2004, presumably the same libretto as given on the Berlioz Website for Paris I [henceforth used to reference the Paris version I, in particular; for the Paris version, I shall ignore tableau, but provide act, scene, and number]), in the trio ensemble, at the repeat of "(Si loin de vous, triste et bannie,) mon âme doit perdre l’espoir" Vanzo’s recourse to "voix mixte appuyée" produced a squeal that to my (perhaps untutored) ears is jarring. Otherwise the duet/trio is lovely. In his 1973 performance (see 6.3 below) Vanzo either employs this phonation much more delicately or the nature of the radio recording obscures it, as Teresa’s singing is far more prominent there. Finally Gedda in his his 1966 broadcast Covent Garden performance nails the phrase with chest voice alone, while doing it fairly well even in the 1972 studio recording. Whether Vanzo’s "squeal" is a misfire or the result of an intentional use of this phonation, it is odd that –of all the tenors I have heard in this role –Kunde alone (Dec., 2003 Nelson recording) produces exactly the same tone here (in his live Nov., 2002 Zurich performance [Paris I, Gardiner] he handles it all with more ease). In the Recit and Romance (Weimar [Act II] No. 5, = Paris Act I, 8, no. 7; in Nelson, for Paris I, relegated to the Appendix) Vanzo’s lyricism is engaging, but one cannot but be struck by how underpowered and overly careful our hero’s presentation is, not so heroic. In (Weimar [Act III], No. 12 = Paris Act II, 2, No. 15) "ma dague en main" he does display his tenorial heroism and in his exasperated response to the ecclesiastical decision to reassign the statue’s casting: "Un autre fondre ma statue!" (Weimar [Act III], No. 14, = Paris Act II, 6, No. 19) he gives us perfect French near-madness (vs. Gedda’s veristic presentation, see 6.1 above) and in the Finale (Weimar [Act III], No. 16 = Paris Act II, 20, No. 31) his prayer "Seigneur, use de ton pouvoir!" leaves us convinced that Heaven will be as convinced as we are. Nevertheless, to my ears, Vanzo in this performance, taken as whole, although most stylish, might seem to lack the necessary heroic aspect, not totally in all places, but most blatantly in the Weimar Act II Recit and Romance (see a few lines up) where he gives us the impossible, a flaccid Cellini. After that he comes into a true expression of the sculptor-hero. I must emphasize that, for the most part, his singing, esp. when in duet with Esposito, rises to the level of the finest French use of textual diction. For our Balducci, Thau gives a dramatic performance. Manuguerra’s performance of Fieramosca is appealing in his interacting with Pompeo; however, while his projection of Fieromosca’s pompous and self-important aria works very nicely, it is not the supreme interpretation within the three performances I have selected to compare. "Not the supreme" is hardly a criticism, merely a comparison. About Vessières’ presentation of Cardinal Salviati I can only say the very thing I said (above, 6.1) concerning his presentation of Balducci in the 1964 performance. Given a choice of a composite ideal performance, I would have to select Vessières for the Cardinal (but see my more avaricious ideal casting in "third performance" below 6.3).

6.2.1. Mollien’s presentation of the Cabaretier is a most successful one, one that eschews the caricature of Cuénod’s in the 1964, discussed above. Whereas Cuénod (and others who use this depiction of the hard-working Inn-keeper) indulges in squeezing the last drop of (self-)mockery, vocally showing the poor man as he was seen and heard by Cellini and his gang of inebriated, impoverished, bohemian, anti-social mates, Mollien’s Cabaretier displays spunk. No doubt the Cuénod type of depiction bases itself (1) not only on the instructions of the libretto "..., à la voix nasillarde" ("in a whining voice") (Weimar [Act II], No.6 = Paris Act I, 9, No. 8 [the problematic expansion of this line of instructions in Weimar is a discussion for another day]), (2) but also on the reaction: "Non, jamais les trompettes/ Du jugement dernier/ Ne sauraient effrayer/ Plus que la voix fatale...Et la liste infernale/ De ce cabaretier" ("No, never shall the trumpets of the last judgment be able to frighten more than the deadly voice.... and the infernal list of this Inn-Keeper") and (3) Cellini’s attempt at mocking him in his imitation to the Inn-keeper’s: "Il faut... Il faut... Il faut payer votre mémoire" by his own "J’acquitte ton mémoire". While for this line even Cuénod in the first performance, considered above, 6.1, has presented it with little of the nasal whining, Gedda’s uptake thereof was actually rather unpleasant, an expression of contempt for the Inn-keeper’s very being. In the 1966 Pritchard performance of the Paris II version (with a different Cabaretier) the response of Gedda’s Cellini is a tad uglier. Although I have generally avoided even listening to the Davis studio recording of 1972 during the course of this article, here Davis’ hand in this whole Inn scene is masterful and shows how much his sense of Berlioz was intuitive. Davis handles Cuénod’s Cabaretier and the Chorus with the utmost musicality. Even Gedda’s Cellini here does his imitation with some sarcasm, not with the vicious contempt of the 1964 and 1966 performances. Returning to our three Weimar recordings, Mollien presents the line like a justly indignant Inn-keeper, and Vanzo in his uptake thereof only slightly exaggerates the whining utterance, but in a mellow manner, showing Cellini and his buddies far less inebriated and hostile in this more nobly conceived and cast performance. One might indeed say that Le Conte had decided to ignore, within conductorial prerogative, the explicit instructions of the score, more firm in Weimar on this point, the "whining of the Inn-Keeper", and so, by letting Mollien become a fellow human being rather than a stereotype of anything, he also had Vanzo imitate the Inn-keeper with only the slightest hint of a beautifully delivered, vocally appealing "whine(?)". We thus have the total Inn scene with unadulterated vocal appeal and, while our chorus sings outstandingly clear French, Vanzo and his "ami" Ascanio and their "tormenter" M. le Cabaretier, each offers the highest level of musical pulchritude in a scene that often tends toward ethical and thus vocal ugliness. If Le Conte has been unfaithful here, I – for one – forgive him. The third performance, to be considered next, 6.3, somewhat exaggerates the Inn-keeper, to which Vanzo exaggerates this delusional way of perceiving the Inn-keeper. Dresse’s is a middling whiner and Vanzo’s imitation is "unpretty". What "our" gang of ruffians see the Inn-Keeper as is the very jaws of avarice. The image (in Damnation) which Méphistophélès brought Faust to see amongst Brander and his companions in Auerbach’s Wine-cellar may well be reflected in the perception that M. le Cabaretier receives of Cellini and his companions. Therefore, here with our Cabaretier, as also with Fieramosca’s aria (Weimar [Act II], No.7 = Paris, Act I, 11, No.10), it is important, when Berlioz has written an outrageous character, that the performer not add to that outrageousness. Overdoing it becomes too much to be effective, a case of less being more. The same applies to Somarone’s aria in Béatrice et Bénédict, the only other outrageous character who comes to mind (arguably, Brander [with his companions] belongs here). This Inn scene, with the appealing interaction of le Cabaretier and Cellini, combined with Dourian’s expression of "Cette somme" and "mes amis" leaves us a remarkable record of what can be. As to Ascanio’s other star appearance, one totally unique in Berlioz and used to express the mordant nature of the imitations with Ascanio’s view of the philosophical implications of what is taking place (never in Berlioz do such vocal somersaults occur for their own sake, as they do so often in both Baroque opera and later in Rossinian opera), some words of preface are necessary. Ascanio’s 2nd great aria and generally the show-stopper of the opera "Tra, la, la, la, la... Mais qu’-ai-je donc?" (Weimar [Act III], No. 11, = Paris, Act II, 7, 22) requires analysis in that its text has been created anew for Weimar and this has consequences for the performances under consideration. The text for either version is to be divided into Parts 1 and 2, the former is further subdivided into 1a and 1b, beginning resp. with "mais qu’ai-je donc?" and "c’est donc ce soir que...". Part 2 begins with "Ah!, ah!, ah!, ah!, la bonne scène". The texts of both 1b and of 2 in the Weimar version have had their bold spice, dangerous wit, and daring salt leeched out, with the substitute ingredient of insipidity poured in, quite simply to fill the space and keep the music, while avoiding blasphemy and making sense of the Weimar mis-arrangement of the scenes – with this imitation aria now before the negotiations between papal authority, represented by the Cardinal, and artistic creativity, personified in Benvenuto. The rewritten 1b dispenses with "baptize", "child", "church", and "Pope as godfather"; 2 substitutes Ascanio’s imitation of Balducci, Cellini, Teresa, and Fieramosca, while Pope – Cellini imitations with their Latin "primo... concedo" and "secondo... concedo" are done away with. As noted, in the 1964 performance (with tenor Ascanio) the aria as a whole is deleted. In the 1973 performance (to be discussed next) Rossi sings only Parts 1a and 1b in its ABA format. Here, in this 1969 performance Dourian sings the entirety of the new text, ignoring its insipidity, giving us ABACA (C = Part 2). When the rewritten aria is performed with targeted singing, with the performer fully cognizant of its purpose, it is a fairly acceptable bow to necessity. Here the rich and purposeful voice of Dourian undertakes the burden of this characterization, puts meaningful stress on a syllable where it is effective, infuses reflection where it is appropriate, while the conductor with the singer paces the whole of this characterization of Ascanio in this aria, such that one must hear unfailing pacing executed by performers who know their characters. Combine this with the following prayer duet and we get a glimpse into a magical entity. It is for such recorded scenes that I feel the need to keep the Weimar Cellini (as historical record), while I find the mutilation of the great work of 1834-38 near criminal. Leaving aside many details from my notes during numerous re-listenings, I conclude with the triad in the Finale, Salviati’s: "J’ te pardonne, ô Cellini!", Cellini’s "Ô ma Teresa" with Teresa’s "Ô Cellini" where Vessières, Vanzo, and Esposito all work their magic for us. With any caveats noted, the many marvels in this performance should keep it as a reference of magical performance in Berlioz’s most difficult work to bring off.

6.2.2. I need here to come to the very moment after the Cabaretier leaves or escapes this crowd (Weimar [Act II], No. 6 middle = Paris Act I, 10, No. 8 middle) in the three Weimar performances under consideration here (1964, 1969, and 1973): it hit my ears that they share a peculiarity: in the lines of Cellini "Mais que notre vengeance/ Frappe ce juif mesquin" ("But let our vengeance/ strike this wretched Jew"). Oddly in their English translation, published in both the trilingual libretto in Davis studio recording of Paris II and in the bilingual libretto to the Nelson Paris I, David Cairns and Liza Hobbs render "But let’s have our revenge on that shabby Jew Balducci"; this explicit adding of the treasurer’s name is not done by the (unnoted) German translator in the libretto to the Davis recording or by Daniela Wiesendanger in her German translation in the booklet for the 2002 Zurich performance –see 2.3.1. below. And indeed in the Davis LSO 1999 broadcast it is precisely "Frappe ce Juif Balducci" that the tenor, Giuseppe Sabbatini, actually sings. In the 1964 performance Gedda certainly seems to my ears to do the line as "Frappe ce mesquin" ("[our revenge] strike that niggardly [one]" rather than "strike that niggardly Jew"); in the 1969 performance, if Vanzo is pronouncing "juif" at all, it is just about swallowed while the preceding and following words, "ce...mesquin" are clearly discernible; in the 1973 performance Vanzo does these two lines (and indeed the entirety of the four line utterance) at such a speed that it is difficult to follow his words at all. After re-listening to the lines in these three performance again and again every so often during seven months and reaching no firm decision on all three, I have now compared two other Weimar performances: in the 1973 Ozawa performance Bonisolli gives equal weight to all the constituent words; in the 1999 Paris II Gergiev Amsterdam broadcast Merritt does the same. Then I tried the Paris II of the 1966 Pritchard live and the Davis studio performances: in the former one cannot tell; in the latter of course all is crystal clear, every word is enunciated. In view of the Weimar instructions "discussion for another day" (6.2. 1. above) and the peculiarity of the Cairns/Hobbs English rendering, this matter in our three performances does lead somewhere. In this passage "mesquin" offers itself as a rhyme for "faquin" which is what Cellini accuses his antagonist as treating him as. Previously (Weimar [Act I, No.3]) Teresa had referred to Fieramosca as "un tel faquin!" (= Paris Davis libretto) and in the Paris I Act I, 3, No.4 on the Berlioz website it was Cellini who so referred to Fieramosca; in the Nelson libretto we have rather Cellini saying "un tel rival!" in place of that and ditto for the Paris I French-German libretto for the Zurich performance. Those places that use "faquin" in Paris seem to be affected by conflation with the Weimar libretto. One might be forgiven for hazarding a guess that at the time of our three broadcasts (1964, 1969, 1973), not far removed from the elimination of European Jewry, it was unpleasant to highlight the phrase about the wretched Jew who rewards the untalented but deprives the genius.

6.3. However for our third performance under consideration here, there is an (in-)house 1973 recording (in my copy, with fine sound except at the very beginning of the singing, right after the overture, where Balducci sounds somewhat distant) from the Opéra de Marseille conducted by Jacques Bazire with some of the same cast, Vanzo, Esposito, Thau, but Fieramosca – Robert Massard (as in the 1966 Paris II-dialogue version noted above re Gedda, and the 1972 Davis studio of that version), Le Cardinal Salviati – Gérard Serkoyan (French bass, Armenian from Istanbul), Bernardino – Bernard Lloyd, Francesco – Bruno Constantini, Pompeo – Raimund Albertini, Cabaretier – Francis Dresse, Ascanio – Antoinette Rossi (who Ms. Rossi was, this superb singer, is a matter on which I have been unable to find any information outside of a note from a very knowledgeable Frenchman telling me that she sang only bit parts; being based in Monte Carlo, she sang such parts also in Marseille when Ducreux took his Monte Carlo productions to Marseille. This does not sound like the singer I hear here; so, perhaps the cast listing is in error; We shall refer to the Ascanio by this listed name). Although I find Vessières’ Cardinal supreme (as I found his Balducci in the 1964) and while Mollien’s Cabaretier is a refreshing change, the 1973 performance, as a whole, has more excitement and propulsion than the 1969 and, happily, the style of the hero has significantly changed. Vanzo here is a different person. One must both praise the conductor and appreciate Vanzo’s transformation. However one cannot say that any of the three performances (as a whole) is definitive. With a combination of elements of the three performances under consideration, we come pretty close to what we would deem the classic French Cellini, certainly singing interpretively on a different plane from what we get on either the Davis or the Nelson studio recordings of the various Paris versions. Avariciously I would select for an ideal Cellini of this period Vessières’ Cardinal and Mollien’s Cabaretier from the 1969 cast, Vanzo’s Cellini and Dourian’s or Rossi’s Ascanio from the 1973 cast, Vessières’ Balducci and Doucet’s Fieramosca from the 1964 set, and have it all conducted by Jacques Bazire from the 1973 performance. Sadly, to my knowledge, there are no French tenor aria recordings from the golden age period for Cellini such as there are for Troyens (Thill), Damnation (quite a few), and even for L’Enfance (Planel). Some time before the Nazis stopped it in 1933, amongst the radio opera performances featuring Josef Schmidt was a German language Benvenuto Cellini. This would be a treasure. Obviously it has not survived. I have become confident that we do know the ideal Cellini singer. In the notes included in the Lebendige Vergangenheit Léonce Escalaïs (89527) CD Leo Riemens observes that this tenor "was a worthy successor to a Alphone Nourrit and a Gilbert Duprez. The records are so valuable because here we can hear how an Arnold (Guillaume Tell), Eléazar (La Juive), Sigurd or Jean de Leyden (Le Prophète) had once sounded". Now Escalaïs (1859-1941) did not have the occasion to undertake Cellini which Duprez created. All facts taken together, it is reasonable to assume that this is how Cellini was intended to sound. Such a singer as the heroic French tenor Jean Nequeçaur in his early 1930s recording (Malibran MR 580) of "Nature immense" would not do as he offers a Faust with all necessary heroic force, but one totally lacking in passion. Yet further light on what an ideal French Cellini was is provided by the 1930 excepts recording of Gounod’s Faust featuring tenor René Lapelleterie (cf. 2. above) who, according to the website Operissimo, in 1914 achieved a "sensational success" as Cellini at the Théâtre des Champs Élysées. In her notes to the combined issue of the 1930 complete and excerpts Faust, Heidi Waleson observes that (compared to César Vezzani of the complete recording) Lapelleterie "takes a more lyric, gentle approach. With the slightly baritonal quality of his tenor and high notes that seem to just float away,...." (Andante 3996-3999). Indeed his is a gentler voice than I might expect for Cellini.

6.3.1 As concerns all these three Weimar performances under consideration, let us take note of what David Cairns in his succinct and vitally illuminating essay "The Romantic Cult of the Artist-Hero" (in the booklet to the 1972 Davis studio recording on Philips) does say. Cairns’ essay includes two diagrammatic charts, the second of which, "List of Changes [between the modified Paris version, used in this recording, and the Weimar version]", is included in the booklet to the cassette release, but not in the booklet to the CD release. The cassette release booklet also includes brief bios of the artists of the recording. However the cassette release booklet is English only (except for the libretto proper in French and English) while all the notes of CD release booklet are English, French, German, and Italian. I have no idea about the booklet to the original LP release. All of this is irrelevant to the souls who now buy these Davis recordings totally noteless, as has become more and more the way with re-releases. Cairns writes: "In this recording the big difference from Weimar [1856 publication] is – as in Covent Garden – the use of spoken dialogue...." Later he speaks of the Weimar version having "unidiomatic recitative and arioso passages and putting them back into dialogue". Now, as we all know (or should know) by now the entire "restoration" of dialogues in place of récitatifs, propagated by the huge authority of the Davis recording with its Cairns essay was authoritative for some decades and accepted without question by mere music/opera/Berlioz lovers (whether the issue was debated in musicological journals I have no idea). Then this entire "restoration" was quite simply done away with by the 1996 New Berlioz Edition, edited by Macdonald. It is therefore doubly unfortunate that (1) persons today purchasing the Philips Davis Berlioz sets which are libretto-less as well as note-less, will lack these notes and (2) the Nelson recording of 2003, issued in 2004, makes no mention of or allusion to this significant history in its English language essay by Macdonald, given also in German translation. Thus, the Nelson booklet contains (1) Macdonald’s essay "Hector Berlioz: Benvenuto Cellini" in English and German, (2) a small note by Cairns in English only, (3) Macdonald’s synopsis in English and German, (4) Nelson’s own notes on the recording in English, French, and German, AND (5) Christian Wasselin’s essay "Benvenuto Cellini ou l’opéra du feu" [sic, no Italics in the French essay’s title] and (6) Wasselin’s synopsis, both in French only. Inter alia Wasselin says such things as: "Or, c’est en s’appuyant sur ce projet avorté, et sur la première conception de l’ouverage (un opéra comique)..., que des Anglais passionés, à la fin des années 1950...ont introduit des dialogues en lieu et place des récitatifs...". It seems the ineluctable conclusion that the English transformations of Benvenuto Cellini should not be revealed to Anglophones and others who do not read French, all of which smacks of scandal. This is a very essential history of Paris Cellini in performance and on record and the way in which one or two generations of Berlioz lovers knew this opera, the Paris version. The essay that mentioned the "passionate Englishmen" would be permitted in French, but not rendered into German or English. To add insult to injury, while the trilingual French, English, German libretto in the influential Davis studio recording bore the headings DIALOGUE, DIALOGUE, DIALOG over the resp. texts, the Nelson recording only bears the heading récitatif over the French text and nothing over the English text where "recitative" should appear, as e.g. in the French, German libretto in Opernhaus Zürich for Gardiner’s performance of this opera (Paris I, just as the Nelson recording) on Nov. 30, 2002 where resp. Récitatif and Rezitativ are given (sincere thanks to the person who supplied me with the libretto and house recording for scholarly purposes). In the noted essay Cairns further writes: "’Benvenuto Cellini’ is in some ways the most original of all Berlioz’s works. Even for him, its rhythmic variety and complexity and its audacity of orchestral colour are exceptional." I can personally add that for me, of all six Berlioz works of which I have discussed recorded performances in this article, Cellini has caused me the most time and the most difficulty, entailing the greatest number of re-visits to the recordings. With all the other works my reactions and the causes for those reactions have been immediate.

6.3.2 In the sad history of the mangling of French opera (-comique), dialogue only is turned into recitative (of any sort), as e.g. Carmen, Les Contes d’Hoffmann, Cherubini’s Médée (best known with Lachner’s originally German recitatives, with the whole translated into Italian, Medea). Benvenuto Cellini does not belong in that group. Berlioz, himself, whatever his original intentions, composed it with recitative.

6.4. In all three performances under review, in the aria "Sur les monts", the tenor does lines 1-4, ignores the four lines following "voyageur", goes directly to "Puis le soir", and concludes with "content". Then he does the first repeat of lines 1-8 with its variants on lines 5, 6, and 8, lines which were ignored in the verse’s first utterance. Next he does the second repeat of lines 1-4, concluding the aria with the short, variant repeat as coda. To repeat with variants that which wasn’t sung in the first place is indeed odd. However Gedda does the same in his 1961 studio aria recording, while he does the integral aria in both his Paris version live 1966 performance and in the 1972 Philips studio recording. However the matter isn’t simply Weimar vs. Paris. Indeed Franco Bonisolli in the live May 8, 1973 Weimar performance (OPD-1373) does the integral aria. So, with my limited resources, I remain mystified by the way the aria is handled by both Gedda and Vanzo in our Weimar performances. So much for 1964-1973 Weimar Cellinis. On Sept. 23, 1987 Georges Prêtre conducted the Choeur de radio France (Marcel Tranchant, chef de choeur) and Orchestre national de France in a concert version of Weimar Cellini on the occasion of the reopening of the Théâtre des Champs Élysées after its having been closed for renovations. So, we have a very French cultural occasion, with the French chorus and orchestra, conducted by the well-known French conductor and we get a cast whose principals are all USA singers of no particular French style. Chris Merritt as Cellini, a tenor who did long become identified with this role, was the very antithesis of the type of Cellini we grew used to in Vanzo and perhaps even in Gedda (on a lesser level); Barbara Hendricks as Teresa, and Florence Quivar as Ascanio: these last two are good singers, but hardly French. One must ask whether Francophone singers who could do these roles had disappeared from France, an impression one also gathers from many commercial recordings of French opera of that period. What happened between 1973 and 1987 to make homogenized, internationalized, and foreign singers participate in a major French cultural occasion performing a great French opera? For performances in exile, such as those of Munch and of Monteux (see next section), this is par for the course, but Prêtre was on native ground, yet perhaps France herself had entered exile. And then on Feb. 22, 1992 the same orchestra and chorus which gave the 1964 Weimar Cellini (see 6.1 above), now not under de Froment, but under John Nelson, broadcast a Paris II Cellini with a largely Anglophone cast, Australian and American with Chris Merritt who here, amidst a very uninvolved performance on the whole, distinguished himself again by a fort ténor execution of Cellini, where he gave us the tenderest soft notes and the strongest extremes that one expects of this hero. Merritt’s Cellini I have heard in both the 1987 Florence Paris II Fedose(y)ev telecast DVD-R, the 1987 Prêtre Weimar broadcast, and the just noted 1992 Nelson Paris II broadcast. Shockingly in the superb televised performance, in the trio the lines from Fiermosca’s "Ah, si j’osai parler tout haut" to Cellini’s "ah, mourir, chère belle" (thus including Cellini’s repeat of "mon âme doit perdre son espoir" in duet with Teresa’s lines) are omitted, a line which is amongst this tenor’s glories. This often intriguing tenor’s assumption of the role in Gergiev’s 1999 broadcast would best be served by destroying all copies (an impossibility, of course). What Berlioz lover can disagree with Macdonald’s statement (essay, Nelson recording booklet): "In general he [Berlioz] placed the dramatic action firmly in the pit.... There are many scenes in the opera...when orchestral activity dominates the proceedings"? Yet, we can never dispense with engaged singers bringing their gifts to bear. The contemporaneity of Vanzo and Gedda assuming this role is a mark of the crisis in French opera. There were to be no more like Vanzo while it was Gedda who went on in 1972 to be Davis’ Cellini in what would be the only available recording of the work in the Paris II version for 30 years and the recording from which most of us learned the work. Gedda also becomes the tenor (as later Domingo was to become) in many commercial recordings of French opera, recordings that took the language and style out of French opera while Vanzo is most known by his radio broadcasts of French opera (so-mis-called "pirates"), whereby French opera had its final expression before these roles became the property of Vickers, Gedda, Domingo, Merritt, Heppner, Kunde, and Giordani (2003 Met Cellini and 2013 Met Énée). None is Gallic; yet to some listeners one or more bring other virtues to Berlioz interpretation, virtues that to these listeners are as or more valuable than idiomatic speech/style. Equally, to some listeners Giraudeau or Vanzo (just to adduce two prime examples), as Gallic as each is, may lack certain virtues necessary for interpretation. For me Vanzo infuses his Gallic Cellini with nobility and vulnerability; but as necessary as I find Vanzo, Merritt’s high ringing notes give something absent in Vanzo’s reading, and one could only wish that, in addition to Vanzo, a French tenor with some of Merritt’s qualities had also left us a recording. One thinks in this period of Gustave Botiaux (3. above for his glorious singing of the Damnation excerpt). One also imagines if the more mature Vanzo (cf. 3. above) had recorded the role. Since Gedda has been central to this entire Cellini discussion, the notes to the 1995 Supraphon CD “Twin” (SU 0022-2 201) issue of Janáček’s Zápisník Zmizelého (The Diary of One Who Disappeared), containing both the interpretations of Beno Blachut (1956) and Gedda (1984) –each with a Czech contralto partner, each with a Czech female chamber chorus and both having the same Czech pianist –cries out to be quoted:

The current worldwide trend to sing Janáček in the original language has produced an increasing number of international singers presenting their own Czech accounts of this work [in contrast to the erstwhile use of vernacular translation as e.g. Max Brod’s German, used by tenor Haefliger & co. in Kubelik’s 1963 recording; YLA]. In 1984 it claimed the attention of … the Swedish tenor … Nicolai Gedda. His reading of the work, modelled upon verismo operatic style…. Gedda’s mastery of Czech commands respect, … Gedda’s interpretation … was adopted by his [Czech contralto] partner in the recording.

7. The precipitating cause of this review of authentic French Berlioz was my first experience (on July 31, 2012) of Charles Munch’s (or Münch, if you prefer) June 25, 1953 live radio performance at the Théâtre des Champs Élysées of Roméo et Juliette with a "cast" of singers and chorus of supreme Francophone qualities (I am not sure whether one may speak of a "cast" in the case of a symphony). This performance was issued by Cascavalle CDs (VEL 3312 "Charles MÜNCH/Hector BERLIOZ") together with Munch’s July 5, 1949 studio recording of Symphonie fantastique, performed at the same site (both works with L’Orchestre national), the latter transferred from the Columbia 78 rpm issue. Munch had made his well-known studio recordings of Roméo et Juliette, first (mono) some months before this Paris performance, on Feb. 22-23, 1953, and his second (stereo) on April 23-24, 1961, for RCA, both with the Boston Symphony Orchestra (henceforth BSO) and each with its own mixture of singers of various nationalities, various New England and university glee clubs and choral societies, none of them French in any sense. These very American (where all nationalities mingle) recordings were long regnant and Munch was long the source where at least Americans found their non-operatic Berlioz (Berlioz opera essentially had not reached the American public yet). In 1953 Munch had given a concert of Roméo et Juliette with John McCollum, Jennie Tourel and Nicola Moscona with the BSO the second weekend of Berkshire Music Festival concerts. I am grateful to Ms. Barbara Perkel, Assistant Archivist of the BSO for filling in details both here and elsewhere for me. I know of no broadcast recording, although one may exist and it would be a privilege to hear Jennie Tourel in Berlioz. She only left recordings of (1) "L’Absence", e.g. in her "Standard Hour" May 3, 1946 radio concert under Monteux (issued on "Jennie Tourel in Opera & Concert 1944-1946", Eklipse CD EKR 55) and in a broadcast April 19, 1970 concert performance (issued on Voxbox CDX 5126), the single song from Les Nuits that everyone seems to record, and (2) the Oct. 9, 1961 recording of La mort de Cléopâtre under Bernstein (Sony SMK 60696, on 7 tracks together with the Oct. 23, 1961, William Lincer, viola, under Bernstein, Harold en Italie). N.B.: it is this same recording that has been issued on Pearl (GEM 0198), on a single track as a 1950 recording. The very Columbia LP issue number given in the Pearl booklet proves its provenance as does listening. There is also preserved an Aug. 20, 1961 Munch broadcast of Roméo &Juliette from Tanglewood with a trio of USA singers, Florence Kopleff, mezzo, John McCollum, tenor, and Donald Gramm, bass (the mezzo being the Marie and the bass his Brander in his studio recordings resp. of L’Enfance and Damnation), a performance I have had for a decade in a private transfer. This live performance is an impressive one. The mezzo is beautiful and fully understands her purpose, a common purpose which in this performance Munch had all three vocal soloists understand (this very fine mezzo participated in Abravanel’s Apr., 1967 studio recording of Mahler’s second symphony). McCollum gives a dramatic expression of the nuances of the text and he sounds quite idiomatic and Gramm does a fine job indeed of Père Laurence, the chorus sings fine. Alas, this fine performance has been issued on the Memories CD label (MR2185-86), with numerous individual tracks with full description on the jewel case back cover. I bought it because I prefer individual tracks to whole acts and in the hope of no LP clicks (I assume my private copy is a transfer from a live performance LP issue). Indeed the sonics on the Memories issue are superb. There is one big problem. For Part II the listing gives a track 8 "Nuit sereine...les jeunes Capulets, sortant de la fête, passent en chantant", 2’13" , followed by a track 9 "scène d’amour" 12’17". In fact tracks 8 and 9 together constitute "scène d’amour" and the choral segment listed for track 8, "...en chantant" (Part II.2a "Double Choeur": "Ohé! Capulets, bon soir!"), has simply been omitted. On my private transfer copy the "scène d’amour" comes to 14’31" and is preceded by by the choral segment, running 3’16", both on a single track running 17’57" which also includes c. 10 seconds of audience rustling at the conclusion. This renders the Memories issue just about worthless. In 2004 BMG issued a Munch/BSO/Berlioz 10 CD box (82876-60393-2) including the two recordings of Roméo et Juliette and of Symphonie fantastique (mono and stereo again) and the single recordings of L’Enfance du Christ (whose cast does include two French singers, one Italian singer, and two American singers), La Damnation de Faust (with a mixed cast: French soprano and baritone, Suzanne Danco and Martial Singher [Marguerite and Méphistophélès], USA tenor, bass, and [stentorian] bass2, David Poleri, Donald Gramm, and McHenry Boatwright [Faust, Brander, and the earthly epilogue voice] with the Harvard Glee Club and the Radcliffe Choral Society). Let us keep in mind that this Munch recording was the first commercial recording of Damnation since the 1942 Fournet. On Singher see 7.1. below. I need to grant that Suzanne Danco’s French is not only syllable clear, but that her Marguerite, in its frightened "virginity", can be very affecting. She transmits the Recit "Que l’air est étouffant" such that this poor, deluded child’s fear truly pierces this listener’s soul. The viola solo of Joseph de Pasquale is also a major asset as is Munch’s pacing of the Recit and aria. She gives effective touches to certain loci as in the scene 13 duo "Je...t’attendais!" and "je meurs...". I only miss the mezzo warmth that Solange Michel (recorded, more or less, at the same time) offers. Danco’s instrument is a less warm one, but one that is no less appealing, as she continues from the duo through the trio and into Marguerite’s other aria "D’amour l’ardente flamme". In the end her Marguerite is as lovely as her 1950 Nuits d’été. However Marguerite was not written for such a non-dramatic soprano voice unadulterated with any touch of mezzo timbre. I have long felt that the tenor, David Poleri, was a poor choice and makes an inadequate Faust. He lacks both the tessitura, the style, and the heroic tone that the character demands and his (and indeed this recording as a whole) is not the way I, myself, would want to hear this dramatic legend and less want anyone else to meet it. The recording falls far short of very good, let alone the excellence of the three 1930-1948 recordings. Graeme (see 3. above for reference, p. 434) wrote: "Thus, for all practical purposes, the complete work’s commercial discography begins with the 1954 RCA recording conducted by Charles Munch". I would say that for all musical purposes the work’s discography should run from the 1930 (abridged) Coppola through the 1948 Wolff. Relistening to this Munch studio Damnation (contrast 7.1. below) has been a less than fully pleasant experience, as opposed to listening to most of Munch’s performances of Roméo et Juliette (even the 1961 studio recording, at least previously, never caused any discomfit). The set also contains the Requiem, Harold en Italie, the well-known Nuits d’été with de los Angeles, four overtures (with that to Béatrice et Bénédict in a mono and a stereo recording), and the "Chasse royale et orage", all of these of course studio recordings. First some positive words on the recording of L’Enfance: the work is cast from strength with all first rate soloists, but with the usual two singers for four roles (as often done in casting of this work) in the case of Valletti and Tozzi (I strongly prefer single role casting, as e.g. in Gardiner’s Jan., 1987 studio recording on Erato); Florence Kopleff is the Marie. Until I heard her participation in Munch’s live 1961 Roméo et Juliette (on this performance see below concerning a summary of the Munch performances of this work), I had (and still would have) no meaningful memory of her voice’s individuality from this Marie. Tozzi does fine as both Hérode and Le père de famille (as opposed to totally wrong in the 1961studio Roméo et Juliette). Yet he is not amongst the most impressive Hérodes. The highlight of the work is Valletti’s performances as (a) Le récitant, where inter alia he does the "Les pélerins étant venus" with utmost beauty (note my comment in 6.3, near end, above, concerning Planel’s recording and see my review referred to in 8. below) and as (b) Le centurion, he and Lucien Olivier as Polydorus steal the show. The chorus is rather unspecific and plus requiring the elevation of the volume whenever it makes an appearance. Munch’s reading of the work as a whole is a somber affair, lacking the vivification of what, in my view, requires a special injection of life, the vivification which Cluytens brought to this difficult work. Let me state that the singers Munch chose for his 1953 studio recording of Roméo et Juliette (Margaret Roggero, Leslie Chabay, Yi-Kwei-Sze) are totally on a different level of suitability from those he chose for his 1961 studio recording (Rosalind Elias, Cesare Valletti, Giorgio Tozzi). For the 1953 studio recording his singers understand what function their "roles" have in this symphonic telling of this story, a story that so overwhelmed Berlioz. All three vocal soloists are admirable, if not French. On the other hand, the singers in the 1961 studio recording, singers much more famous then and today, are so inappropriate as to be gross. The cantique for mezzo [alto] "premiers transports" is one of the most perfect things Berlioz even penned; it is no less heart-rending than Didon’s farewell at the end of Troyens. It is simultaneously wedding hymn and funeral hymn. In the very words of this cantique: "si pure extase /Que ses paroles sont des pleurs!". When performed with respect, it is almost unendurable. The 1961 mezzo, Elias, has no concept of or interest in the sublimity of the strophes she performs. She simply does her piece as another diva aria. The 1961 bass, Tozzi, has no concept of or concern with what it means to be Père Laurence. C. M. G., in his notes to the IICC (CD819) Wagnerismo writes: "...Italian singers and singers schooled primarily in Italian opera sang Wagner very much the way they sang Italian repertoire". He was writing about singers singing Wagner in the Italian language. But here we have Tozzi (and Elias) singing Berlioz in French and still singing very much the way they sang Italian repertoire. Precisely so does Nicola Moscona, the bass in Toscanini’s Feb. 9 & 16, 1947 Berlioz broadcast (complete Roméo et Juliette, orchestral and vocal excerpts from Damnation with baritone Mack Harrell; Guild GHCD 2218/20) perform Père Laurence. The style Moscona and Toscanini used in their famed Dec. 2, 1945 broadcast of the Prologue to Boïto’s Mefistofele was the style they used here. The 1953 mezzo, Maragret Roggero, does appreciate the nature of her strophes, Leslie Chabay (Laszlo Czabay) is a fine tenor, while Yi-Kwei Sze is an impressive and magnificent Père Laurence. Of the vocal soloists in the 1961 stereo recording, only Cesare Valletti is appropriate. Nearer our own time Lorraine Hunt Lieberson, in a live broadcast performance of Roméo et Juliette (Dec. 4, 2004, with two other fine American soloists, Matthew Polenzani and Julien Robbins, Tanglewood Festival Chorus [cond. John Olivier], BSO, conducted by James Levine, whose reading fails to approach the subtitles of the work) knew to express the full sublimity of these strophes. Again, in this variable we are concerned not with native French but rather with soloist awareness and the sensitivity that comes from awareness.

7.1. There also survives a broadcast of Munch’s live Damnation from the Berkshire Festival of 1960. The BSO master tape for that broadcast was destroyed in the WGBH fire of 1961, along with many other master tapes. All copies derive from the radio. Even the best copies I have heard are uneven in their sonic representation. Different copies of this broadcast that I have heard uniformly begin Part III with scene 10 (immediately following full announcement and audience rustling), omitting the "retreat" music and Faust’s aria, "Merci, doux crépuscule!". I assume that the performance or the broadcast omitted that music. Its Faust is John McCollum, noted above for his role in the live 1961 Roméo et Juliette. In this American tenor Munch had a treasure. His French language and style once again appear very fine and indeed it is he and Munch’s orchestra that make this performance a triumph and its inadequate preservation a shame. Eleanor Steber is a grand and worthy Marguerite, when we accept her on her own terms, who, how, what, when her career and style were in its own time and place (rather than dismiss her as another non-French style Marguerite). Martial Singher is our Méphisto, as he also was in Munch’s studio Damnation. Our Brander is David Laurent. Singher was by now long resident in the USA. For all that, he offers an impressive characterization of Méphistophélès. The Festival Chorus and the BSO give us an almost frenetically dramatic performance, one which perhaps esp. in the duo and trio may deny the lyrical for the dramatic. Granted that it is only Méphistophélès’ "Crazy glue" which binds Faust and Marguerite; yet, in some more lyrical readings we – the listeners – can imagine that we perceive love. These live performances give testimony to the degree to which Munch was Berlioz’s apostle to the USA. Munch’s combined USA vocal Berlioz, studio and broadcast, with their odd mixtures of Francophone, "French-friendly" singers, and others quite unaware of and unconcerned with the style, give mixed results. He took what, who was available and sometimes we have a "hit" and ofttimes not. Steber and Singher preceded this participation in Damnation with performances in late Dec., 1959 and mid Jan., 1960 as Cassandre and Chorèbe under Thomas Beecham (essentially prepared by Robert Lawrence) in Washington, D.C. and under Robert Lawrence in New York City performances of Les Troyens, its first USA complete performance(s) (the NYC Lawrence released on VAIA 1006-3). Those two were busy Berlioz performers and/or recreators that year. In the Lawrence Troyens Singher is somewhat less convincing than in the broadcast Damnation. He had been Chorèbe in Paris in 1932 when de Trévi was Énée (according to the "Liste chronologique des Artistes du Chant de l’Opéra de Paris" [on the web]) and in 1939 both Berlioz’s Mèphistophèlés and Chorèbe again. According to Jean Ziegler (notes to the Malibran issue of the Beecham Troyens), Le Théâtre antique d’Orange produced Les Troyens in 1931 with Singher (Chorèbe), Ferrer (Didon), Jane Cros (Cassandre), and de Trévi (Énée), and in 1938 La prise de Troie with Singher, Ferrer, and de Trévi. Singher was never a sublime French stylist or the possessor of the most beautiful baritonal instrument. He did serve French opera well in a long and distinguished career including the studio and the broadcast Munch Damnations.

Having had for so long such mixed reactions, mixed emotions, concerning so many of Munch’s studio recordings of Berlioz’s major vocal non-liturgical works, what made me acquire yet another Munch Roméo et Juliette? This is a question the answer to which is very easy. About a year earlier I had added, to my small treasure collection of authentic French Damnations, excerpts conducted by Anatole Fistoulari from 1954 sung by the beyond description Irma Kolassi and the fine Raoul Jobin. These excerpts are available (and "packaged" with Werther excerpts by the same forces) only by download from Naxos Historical, but not in the USA. Roméo et Juliette is a dramatic-oratorio-symphony-with-operatic-ending. Julian Rushton ("The musical structure" in Ian Kemp [ed.], Hector Berlioz Les Troyens, 1988:120) puts it succinctly: "Mixing structures came naturally to Berlioz and is indispensable to such works as Roméo et Juliette and La Damnation de Faust...". The categorization of Roméo et Juliette is rehearsed admirably by David Cairns in his Berlioz biography, vol.2: 190-206. I will add that it not only builds on Beethoven 9, but it leads the path, not merely to Mahler 2, 3, and 4, but all the more to Mahler 8, Part II in that Berlioz has a single dramatis persona in Père Laurence and Mahler was to have a number of dramatis personae in his eighth symphony’s second part, first and foremost Doctor Marianus. From Mahler it led to all those 20th century symphonies with soloists and/or chorus, none of which could approach the achievements of Beethoven, Berlioz, and Mahler. To refer to Roméo et Juliette as anything other than Berlioz’s symphony is to deny Berlioz’s revolutionary effect on the symphony, as also with Symphonie fantastique. Quite strangely, in his notes to the EMI issue (CDC7 49764 4) of Sawallisch’s recorded live performance of Mendelssohn’s Sinfonie Nr. 2, op. 52 "Lobgesang", Karl Schumann writes (p.4): "Erst an der Jahrhundertwende [1900] brach man nachhaltig das Tabu [of repeating Beethoven’s 1824 concept of combining symphony and singing voices], der jung Sibelius in der legendhaften Kullervo-Sinfonie (1892) und Mahler in der Lied- und Chorstrophen der zweiten Sinfonie (1895). Zwischen Beethoven und den beiden Spätromantikern wagte einzig Felix Mendelssohn-Bartholdy den Versuch". After this statement that between Beethoven and Sibelius/ Mahler, alone Mendelssohn dared the attempt to break the taboo against the repeating of what Beethoven had done, Karl Schumann proceeds to say that Mendelssohn’s op. 52 was misleadingly counted as the composer’s Second Symphony while it actually was a Sinfonie-Kantate written for the Leipzig Gutenberg Festival on June 25, 1840, commemorating the 400th anniversary of the invention of printing. Without discussing either Sibelius’ youthful work or Mendelssohn’s here, I find it of value to call attention to the author’s total forgetting about Berlioz. Nowadays, unfortunately, this Berlioz symphony is most often given in its orchestral movements alone. Such has not been the case with Beethoven 9 or Mahler 8 (in the latter not readily doable) or Mahler 2, 3 or 4. This is an important matter for others to consider. Valery Gergiev did, however, lead the LSO in a complete performance with Olga Borodina, mezzo-soprano, Kenneth Tarver, tenor, and Yevgeny Nikitin, bass on Nov. 13, 2013 in the Barbican, London. In spite of the admirable efforts of both tenor and bass (both of whom cared deeply for their respective characters and the French words), the performance should be forgotten. Although I have mostly made observations on the singers in Berlioz’s works, all the conductors previously discussed ranged from very good indeed to superb. No singers can save a conductor who apparently simply doesn’t care, the antithesis of the cathexis previously adduced here.

A performance of Roméo et Juliette with Kolassi (who performed no opera on stage, but only in the concert hall, on radio, and recordings and, for the most part, recorded mélodies), Joseph Peyron (one of the loveliest French lyric tenors, who performed a very moving Hylas in an abridged broadcast of Les Troyens, cond. by Tomasi; cf. 2. above), and the baritone Lucien Lovano (who can be heard in six opera broadcasts on Mike Richter’s CD-ROM "Opera from Paris", 3. above) could only offer extraordinary expectations, assuming that the soloists, the choruses of the RTF, directed by Yvonne Gouverné, L’Orchestre national, and the conductor were all having a good day. In the event "extraordinary" became too mild an adjective.

Philippe Morin, in the conclusion of his short notes for the Cascavalle set sums up the matter just as I see it, but more politely. Referencing both of the Munch’s BSO recordings of Roméo et Juliette, La Damnation de Faust, L’Enfance du Christ, and La Grande messe des morts, he says: "However remarkable these recordings were, one must listen to the recording of the concert on this album which was recorded on 25 June 1953. Münch (sic) had just recorded, in Boston, in February, the first complete recording of this masterpiece. This 25 June 1953 performance at the Théâtre des Champs Élysées, he played with soloists, choirs, orchestra all speaking the same language, understanding the terrible tragedy of Shakespeare through Berlioz’s language and there the miracle occurred. This was French romantic drama and Münch was there at his peak with his French musicians; such irresistible brilliance, the overwhelming lyricism, fevered passion, and dreamy sensitivity. Everything is present from which to make a reference recording". I have mostly let M. Morin use his own English translation except where (1) a major disaster happened in his translation, totally confusing the reader and (2) where he seems to overly make his English less emotional than his French, resulting in blandness. In both cases I have translated from the French (as well as substituting an English word occasionally). I suggest that this miracle has a name: it is home vs. exile and here "home" means inter alia idiomatic (with the greatest care for expression) French singing (with cathexis), something that was not always to occur just because a performance is given in France or with Francophone singers, chorus, and conductor. So perhaps Munch was no less aware of the spiritual distress I referred to, as affecting me, in his Damnation recording than I was. Let us keep in mind the vocal recordings of Munch in occupied Paris. They all have supreme French singers, as one would expect. This Roméo et Juliette performance is a miracle. The singing is on the highest level of ideal French singing and the role of Père Laurence was certainly written for Lucien Lovano, if I dare say so. Such singing, together with his French orchestra, could not but affect Munch. Munch’s speaking through the orchestra here is as essential as any vocal participation and, just listen to the "Scène d’amour" and the magic of this transformative work seizes one, something that fails to happen in many performances. I share M. Morin’s perspective that the singers, soloists and chorus, are so essential to the fabric of this symphony, that such singers make the performance magic. Yet, as just noted re Levine’s, the whole needs sensitive conducting and emotionally invested singers. As to whether Munch offers the unique orchestral reading of this work, I, myself, do not feel as ready as M. Morin to know that (having heard but a limited number of performances). Concerning this "Scène d’amour", Berlioz notes that, while numerous composers have set such scenes as a tenor-soprano/mezzo duet, he could more meaningfully set it for orchestra. Ironically, Taneyev did the opposite with Tchaikovsky’s 1869 Romeo and Juliet Fantasy Overture when he reset it (1893) as a tenor-soprano love duet, based on sketches that Tchaikovsky had abandoned (incidentally setting a different scene from what Berlioz had set; while Berlioz sets Shakespeare II/2, Tchaikovsky had in mind III/5). Berlioz was of course right as, while in Troyens the love scene, as set for orchestra begins Act IV and then a love scene duet for tenor-mezzo concludes the same act, at this stage of my listening to Berlioz, I have come to the startling (for me) realization that the love music in Roméo et Juliette is beyond those in Troyens. Whether the orchestral love scene in Troyens is rather consummation music than love music is outside present purposes. Munch’s Paris mezzo is supreme as is his bass and their participation is extremely relevant. The appropriate comparandum is not the two love scenes of Troyens, but rather the Andromache and Astyanax "mourning for Hector" scene in that opera. Hugh MacDonald ("Berlioz and Les Troyens", p.21, in the booklet for Dutoit’s Decca recording) writes: "Neither of them sings, yet there is no more moving music of mourning than this. Berlioz knew instinctively that for the most heartrending scenes he could do more with instruments than with voices. The symbolism of Andromache’s silence...great emotion being bestowed on an instrument (here a clarinet), not on the voice, confirm Berlioz’s faith in wordless music, as if Les Troyens still upheld the symphonic ideas of Roméo et Juliette". Berlioz knew when he wanted the human voice and when he wanted the orchestral voice. He was master of both.

Something very similar to what M. Morin is arguing for Munch here happened in 1929-1930 in the life of Pierre Monteux (1875-1964). William Malloch, in his notes to his CD transfer of the great French conductor’s 1930 recordings of Symphonie fantastique, overture to Benvenuto Cellini, and overture to Les Troyens à Carthage (Music & Arts CD-762), quotes a 1942 interview where Monteux spoke of his first recordings with the Orchestre symphonique de Paris, namely the Jan., 1929 Stravinsky Sacre du Printemps and the June, 1930 Berlioz. Piero Coppola (cf. 3. above) was responsible for the technical setting up of the recording matters. "I had received the invitation to come to Paris and make records while I was conducting the Concertgebouw Orchestra in Amsterdam. The idea filled me with much emotion because I had not conducted a French orchestra for many years, and the thought of going back to my native city to make this recording [spec. the Stravinsky, but the emotion persists in these Berlioz works; YLA] thrilled me no end". The result, many will agree, are some of the greatest Berlioz readings on record. This experience is near identical to that of Munch (1891-1968) in the 1953 Paris Roméo et Juliette except for the fact that the city and orchestra alone are responsible for Monteux’s 1929-30 magic, while Munch had place, orchestra, soloists, and chorus to transport him in 1953. Sadly all of these elements are lacking in Monteux’s June, 1962 studio recording of Roméo et Juliette (Westminster: The Legacy CDs [289 471 242 2], stereo) and his live March 8, 1962 broadcast performance of Damnation (BBCLegends CDs [4006-7], mono), both with the London Symphony Orchestra and Chorus, the former with American and English soloists, the latter with a very mixed bag of soloists, both sharing a tenor who is barely suitable in the former and totally unsuitable in the latter and no one having French style with the exception of the Méphistophélès (once again) of Michel Roux. To conclude, I should like to say that now, with this Munch Paris Roméo et Juliette, we have another Berlioz work, one in addition to Damnation, where I believe or feel that there is a perfect recording. I rejoice. There are of course some other ("mainstream") performances of all the Berlioz works that I, like anyone else, enjoy to a great degree, but there is usually compromise on some of my aesthetic criteria. I do need to recapitulate, for the sake of some friends who find me overly "puristic", that, of the four Munch Roméo et Juliette performances that I have heard on record, I find (1) the USA soloists in the Feb. 22-23, 1953 studio performance and (2) the USA soloists in the Aug. 20, 1961 live Tanglewood (Mass) performance excellent in their adopting the style of the work, a French Berliozian style, (3) the USA mezzo and bass in the April 23-24, 1961 studio performance sadly unaware of or unconcerned with the style of the work (whether their fault, Munch’s fault, the record producer’s [who wanted a new, stereo, studio, star-studied record] fault [which then Munch allowed, thus being complicit] is impossible to determine), and (4) the supreme French soloists in the June 25, 1953 live Paris performance celestial. (5) There always could have been a live or studio Paris performance with mediocre performers or at least ones not concerned with the style of the work. This is in fact more likely to have occurred at a slightly later date, as so many recordings and broadcasts show. Many of the singers involved in any of the works discussed in the present review (except Kolassi and Rossi) continually participated in French radio broadcasts of the 1940s, 1950s, 1960s, and 1970s, (broadcasts which have happily been preserved) of today mostly forgotten operas (and highly idiomatic performances of some fairly well-known operas, some of these often released on such labels as Le chant du monde, Opera d’Oro and Gala, etc.) and are all well-known names to those who listen to such recordings (in the computer age, many of these have become more widely accessible). So, now in additions to a few recordings of French, superb Damnations, we have added Cluytens’ L’Enfance to the category of perfect recordings of vocal works, and herewith obviously we have a perfect Roméo et Juliette. While, of the six non-liturgical, major, vocal works of Berlioz discussed here, Roméo et Juliette is the only symphony, it too, without expressive singers (the more criteria they meet, the better), would be simply another Symphonie fantastique. Thus, as I conversed with Hugh Macdonald about Benvenuto Cellini (end section 6. above), just as much or more, I conclude concerning this symphony with chorus, three soloists (one of whom is a dramatis persona), "Yet we can never dispense with engaged singers". The three operas remain without any substantially complete recording of perfection according to the criteria of this reviewer, although the cited Béatrice et Bénédict comes close enough. Troyens and the Paris Cellini will forever remain without abridged or complete perfect recordings, French both in language and in style, because that time and that style are over and shall never return. For those who need French and stylish singing, Troyens (not it alone, obviously Halévy’s La Juive and, for those who love Meyerbeer, all his works), we shall have to be find our true contentment in the few historical excerpts and, in the case of Troyens, the abridged Tomasi broadcast, and only more generally in current complete performances, even the best of which fall far short of the mark.

8. Both Charles Panzéra and Georges Thill, in addition to their noted contributions to the French Berlioz discography, have each left a French language recording of Schubert’s "Erlkönig" ("Le roi des aulnes") with "Berlioz’s sensitive orchestral accompaniment" (notes of Keith Hardwick, 1997, to the CD edition of Schubert Lieder on Record [EMI 7243 5 66154 2 7]). So indeed did Georgette Frozier-Marrot (Malibran MR 597). Cf. 2. above. Thill’s recording is of the voice of the father, while bass and boy soprano present the erl king and the boy resp., obviously an inauthentic way to present Schubert’s Lied.

9. Finally allow me to take advantage of the present occasion to thank Paul Lewis for his correction to my 2008 review (this site) of historical Berlioz recordings, where he notes the 1909-1910 set of Damnation excerpts which includes Berthe Auguez de Montalant singing "D’amour l’ardente flamme". This soprano, whose 1909 recording of "Premiers transports" from Roméo et Juliette appears in Symposium’s HB:EVR, has now had her 1909 recording of the Romance from this potted Damnation (8’10", two 78 rpm sides) appear in the 2009 3 CD EMI Classics set "Les Urnes de l’Opéra" (50999 206267 2 3). One only hopes for the release of the entire set of the fourteen 78 rpm sides that made up this first attempt at an abridged (and "potted") recorded Damnation, not just for "historical reasons", but because it contains other gems beyond Marguerite’s Romance. I have only heard from the excerpts: Journet doing "Esprits des flammes" and the trio "Allons, il est trop tard" with the above named Marguerite, the fascinating tenor Léon Beyle, and baritone Henri Dangès. Thus we already have two Méphistophèléses. The Journet is certainly released; it’s his February 28, 1910 recording, most recently in the Marston set "Marcel Journet The Complete Solo Gramophone Recordings: 1909-1933". Since I don’t know what else is included in this "potted" Damnation, I can only say that from the three excerpts I do have, the trio leads one to hope.

We are most grateful to Dr Arbeitman for his valuable contribution to this site. 

The Hector Berlioz Website was created by Monir Tayeb and Michel Austin on 18 July 1997; this page created on 17 August 2014.

Back to Original Contributions page
Back to Home Page

Retour à la page Contributions Originales
Retour à la Page d’accueil