« Weekend review roundup: guitarist Jason Vieaux, Juilliard String Quartet | Main | Domingo's other opera company gets financial rescue from Los Angeles County »

December 8, 2009

Met Opera's new "Hoffmann" production short on musicological scholarship

Although not nearly as controversial as its season-opening production of "Tosca," the Metropolitan Opera's new staging of "The Tales of Hoffmann" is generating diverse views on the singing (the distinctive sound of tenor Joseph Calleja, who has the title role, will probably always divide listeners) and the production concept by Broadway vet Bartlett Sher.

An additional element, the question of what edition of the score is being used, may not get all audience members as worked up, but it ought to, considering the extraordinary scholarship on Offenbach and "Hoffmann" over the years. The leading force for changing old perceptions and performance practices regarding this opera has long been Marylander musicologist Michael Kaye, who helped bring to light the composer's original intentions and a lot more about Offenbach and this opera.

That the Met chose to go with an outdated version of the score has, understandably, not gone down well with Kaye, who wrote a response to the new production that he shared with me. For those of you heading to New York to catch a performance of the Met's "Hoffmann," or to movie theaters for the HD broadcast on Dec. 19, or sitting home on the 19th to hear it on the radio, I think that Kaye's observations are well worth keeping in mind. Here's what he has to say:


I have devoted nearly three decades to establishing the landmark edition of Offenbach’s “Les contes d’Hoffmann,” which is now a co-edition with Jean-Christophe Keck, being published by a trio consortium of Schott Musik, Boosey&Hawkes, and Bote&Bock. For more than a year I have known that the MET will ignore our long-standing work on the opera in their new production.

They are promoting it’s new production of HOFFMANN with some serious misinformation about Offenbach and it would be great if you would set the record straight.

I respect and acknowledge the fact that a stage director requires an artistic freedom to interpret the works he or she is charged with producing. However, In the recent press release from the MET about HOFFMANN, stage director Bartlett "Sher says, referring to the early German romantic polymath whose stories are used for the opera’s episodic plot. ‘I’m more interested in why Offenbach, who had been a very popular operetta composer, was seeking to write a serious work to gain acceptance. Why, so late in his career, did he feel this need to be accepted? That led me to consider Offenbach’s sense of being Jewish and an outsider.

Whatever group he was in, he always appears as an outsider who never feels like he belongs, never feels like he’s connected.’ The ambiguities and split identities of the characters figure in Sher’s vision of the piece. ‘For any artist, ambition and paranoia are both always present. The door keeps opening and there are many Hoffmanns, identities that keep overlapping. I think the real artistic dilemma for Offenbach is the tension between the cover [sic] and the internal state, and that’s what I hope to try to show.’"

That statement indicates that Mr. Sher has very little understanding of Offenbach at the time he wrote HOFFMANN, or of E.T.A. Hoffmann himself and how Barbier and Offenbach synthesized the essence of E.T.A. Hoffmann's life, works, and literary style in “Le contes d'Hoffmann.” Neither Mr. Sher nor his designer were interested in receiving copies of the most important source materials for HOFMANN that I offered to send them (the final pages of the co-edition with the latest discoveries are still in unpublished proofs for formal publication).

The idea that Offenbach was looking for "acceptance" is really misguided. Yes, he wanted to write for the Paris Opéra and did so, but having composed more than 100 smash hits for the stage; being dubbed the darling of the entire Third Empire Paris; designated as “the Mozart of the Champs-Élisée by no less than Rossini; and able to return Wagner's hatred for Jewish artists with sarcasm and humor mocking the "composer of the future" with salvos in music and onstage is information available from the oldest biographies of Offenbach.

As for James Levine's statement quoted in the press release: "Maestro Levine says of the musical version, ‘The music is so inspired, and I think we have made effective choices in the absence of an authentic, fully realized original version, using a great deal of the information that has come to light over the years.’” – that is total balderdash, inaccurate, and I'm really concerned that people might believe him!

I also don't understand why maestro Levine would permit the MET’s press department to make statements that negate the existence of totally complete manuscript sources for the opera (much more than sketches, including the complete score of the way the opera was first performed in Paris and, in particular, the full manuscript of the Giulietta Act – including the final scene of that act – published for the first time in my editon. Many of those manuscripts, previously unknown to other editors of the score, were fully orchestrated and rehearsed at the Opéra Comique before Léon Carvalho (impresario of that theater and stage director of the premiere) decided to eliminate the Giulietta Act from the opera.

I think it is admirable that Maestro Levine can prepare new scores by Gunther Schuller and Elliot Carter, and (his recent serious health issues aside) shocking that for years maestro Levine has refused to restudy HOFFMANN. Apart from the affront to scholarship, it also deprives the MET artists and their audiences of evaluating and experiencing Offenbach’s own achievements for his masterpiece that we have tried so diligently to reflect in the HOFFMANN edition. Perhaps in future revivals of the new production they can revise their performing version to include the authentic music by Offenbach that will not be heard at the MET this season.

Michael Kaye

PS: In March, the Zurich Opera will mount a new production of HOFFMANN, with Vittorio Grigolo singing the title role for the first time in his career. There, as with many other European opera companies, they have chosen to base their performing version on our co-edition.

UPDATE 12/11/09: Some of the comments posted on this blog entry seem to think that I (or Michael Kaye) was trying to discourage people from seeing the Met's new production. Not at all. Just wanted folks to know that they are not getting the advantage of Offenbach/"Hoffmann" scholarship, something the august Met might have been expected to provide, especially with regard to what Offenbach and librettist Jules Barbier actually wrote for the Giulietta Act. Michael Kaye also has provided some additional information that answers some of the comments posted here:

-- multiple versions of the opera are possible through the possibilities afforded by integral edition: versions with the original spoken dialogue and versions with the recitatives of Ernest Guiraud modified to be compatible with the authentic Offenbach material cut from the world premiere of the opera.

-- Performed uncut, the entire Giulietta act runs only around 37 minutes!

-- When judiciously cut, a theatrically effective version of ALL the dialogues for the entire opera takes 15 minutes to perform.

-- Oeser added more than 30 minutes to the score consisting of music Offenbach never intended for HOFFMANN.

-- For producers desiring to do so, now there are ways to include the apocryphal "Scintille diamant" and Sextet with Chorus in a different context than the traditional version of the score.


Posted by Tim Smith at 12:35 PM | | Comments (24)


After reading this preachy screed, what is one to think? Should we decide not to attend the Met "Hoffman" because of this alleged inattention to "scholarship"? Should we choose not to attend a performance of Mozart's Requiem because the performers choose to use something other than Mozart's "original" score? What do we make of the various versions of "Boris Godunov"? There are times when a little knowledge counts as TMI. This sounds like one of them.

An uncut Hoffmann would be longer than many Wagner operas!!! - bring it on, right?

Of course, there are recordings that contain a lot of the music ommited by the MET, though perhaps not everything that mr. Kaye published - due to the fact that they were made before the score was published (the Cambreling and Tate versions come into mind.)

Plus, as I pointed out before, Hoffman is not even Offenbach's first opera!

This may not be the impression that Mr. Kaye intended, but if I were reading this "cold", the tone of his statement strikes me as one of sour grapes. It may be justified, but it frankly comes off as unappealing nonetheless. I happen to know one of the singers in the Met production, and I am planning to see the HD-transmission, scholarly concerns aside.

Besides Zurich, is Mr. Kaye's version to be used in the Santa Fe Opera production scheduled for next summer?

As a singer, I appreciate knowing what the composer actually wrote, and not what various directors, producers, and conductors cut out for whatever reasons took their fancies.

As a dramatic coloratura soprano, I really appreciate the idea that I could sing all the heroines in this opera, and not just Olympia and Antonia, because Giulietta is supposed to be a coloratura soprano.

Yet we have a mezzo doing the role! No wonder there were issues with the Barcarolle at the MET opening night (according to the Times review); two mezzos muddying up the waters when Offenbach never wrote such a thing.

Intellectual laziness should be a quality that is never seen at such an august house as the MET. And since someone else has gone to the trouble to do all the work, the laziness could even pass unnoticed, if the MET would use Kaye's edition. Really, what good reason do they have for not doing so?

Bravo, Michael Kaye! I will wait to form an opinion of the new "Hoffmann" after seeing the 12/19 HD transmission, but it's refreshing to know some of the 'truth behind the curtain', i.e., the Met didn't want to use Mr. Kaye's scholarship. "Hoffmann" is one of the most unusual pieces in the standard rep and can sustain a varied outlook pertaining to staging. We shall see. P.S.> Is there a way to get Mr. Kaye's comments on a website?

Thanks for commenting. I don't have an answer, alas, to your question at the moment. TIM

This is hardly a question of sour grapes but a legitimate demand for accuracy and respect for scholarship. It is hard to believe that an institution as formidable as the Met would chose to disregard the findings of a respected musicologist who has brought to light the true intent of Offenbach, and go off on a path of Offenbach's need for "acceptance." There is something decidedly wrong with this picture.

It's hard to understand why the Met would chose to go down this path and it does appear to be "intellectual laziness" and disrespect for recognized scholarship that brought to light Offenbach's intent for the opera.

I find this information very disturbing. The MET is wasting millions on very poor "Director Driven" productions of operas. These are some of the worst stagings I have ever seen and demonstrate the directors' ignorance of the material.

I was very impressed by Sher's production of South Pacific, but his staging of August Wilson's Joe Turner last season was terrible. No one dares to criticize him.

It seems to me that in what is supposed to be one of the world renowned opera houses, one should be able to expect both the best scholarship as well as the best voices, music and staging. It is not always possible to present excellence in all of these regards but if we cannot expect it from the MET at the outset of a production, it is a sad moment indeed.

First off, bravo Mr. Smith for a well-written article, a knowledgeable, level-headed, and very human piece of writing that dares to address a topic which most critics ignore. It is clear from the fact that Smith actually took the time to correspond with Mr. Kaye himself that that he is willing to go the extra mile, and is not or will not post rubbish based on half-truths and scholarship which dates back to 1971 (I’ve seen this on many recent websites regarding the new Met “Hoffmann”). His piece tells the truth and needs no commentary.
What I do wish to comment on are the comments themselves, which show a clear lack of understanding in a matter which is very complex, to say the least. I want to preface this reply by stating that I have never met Mr. Kaye, and so possess no personal bias in this issue. I have, however, been very familiar with Kaye’s work on “Hoffmann” for a little over two decades now – through performances, recordings, and by reading his articles in various musical journals. I am also deeply involved, at present, in researching both the history of this work and its sources (I’ve spent the past year on E.T.A. Hoffmann alone) for a forthcoming book on the work. Most people are painfully unaware of this work’s tangled history and of recent scholarship, so I hope to set the record straight.
What Mr. Kaye is expressing here is not “sour grapes” as someone has earlier commented. The issues at stake are far greater than personal slight or personal gain, and it pains me that anyone could view this on such a superficial level. I would ask the reader to put themselves in Mr. Kaye’s shoes. He has spent the past thirty years on painstakingly pasting together, from various manuscripts which have been scattered all over the world, a performing edition of Offenbach’s posthumous masterpiece which clearly reflects Offenbach’s final intentions at the time of his death. And he has achieved this brilliantly. And yet he appears to have met with hostility and indifference at many junctures on this long journey. They, of course, stem from many things: from unwillingness to part from what is comfortable and familiar; from professional jealousy (at least that’s how I see it – Kaye himself has never said so); and from downright laziness. What Kaye is pushing for is not personal fame or advancement, but for Offenbach’s opera to finally be heard (after one hundred years and multiple versions – all of which have manipulated and tampered with Offenbach’s original ideas) in its original form, and for the listening public to make its own decision.
There is much misinformation that has been told about “Hoffmann” over the almost one-hundred and thirty years of its existence, and the fact that many of this is still believed has not helped in gaining Kaye’s edition the wider exposure that it deserves. First, it should be emphasized that Kaye’s edition, unlike Bonynge, Felsenstein, Oeser, is not an new “take” on the work, not a “version” that tries to solve problems which, at one time, seemed insolvable. What the edition does is present the work as Offenbach saw it when he died. We can do that now, because there is so much authentic, dated material available – much of it in Offenbach’s hand. We can re-construct the opera as it was rehearsed at the Opera-Comique (Giulietta act intact) before impresario Leon Carvalho began implementing his devastating cuts. These cuts not only cheated us out of some glorious music, they disfigured the opera’s structure, symmetry and dramaturgy, resulting in a version which, despite its beauty, was nearly incomprehensible as drama.
Details on these cuts and changes and why they were made (almost none of them were made for artistic reasons) are far too complex to go into here. The Giulietta act was cut entirely at the premiere and was not restored in the early published editions of the opera until the second and third incarnations of what Kaye refers to as “Guiraud’s second redaction” (which was not included in stage productions of the opera until several years after the world premiere.) However, the version of the Giulietta act cobbled together by Guiraud was heavily cut, anti-climactic, and was placed before the Antonia act and was performed in that position for more than a century.
Fritz Oeser’s version of the act was based on sketches that dated from around 1879, which lead him to make the prematurely assumptive (and, as it turns out, false) conclusion that Offenbach had barely begun working on the act at the time of his death. In order to fill it out, he borrowed over thirty minutes of music from Offenbach’s earlier “grand romantic opera” Die Rheinnixen, drastically re-wrote the libretto, and composed some recitatives of his own, and supplied his own orchestrations.
Some new myths have arisen after discovery of the original material stem from many sources. One is the fact that so many new “versions” of the opera have appeared over the years that people are skeptical, tired, and confused by the quantity of new information. Kaye’s is not a “new version” but simply a restoration – there is no “new” music in the edition, only what Offenbach originally composed. Secondly, the length of Offenbach’s original has been grossly exaggerated. I doubt that anyone who has stated that the Kaye edition is of Wagnerian proportions are truly familiar with it. One person on another website even stated that the edition is longer than “Parsifal” which like saying that “Mary Poppins” is longer than “War and Peace” – it simply isn’t so. It is longer than the “traditional” Choudens version that is true. But it is a good deal shorter than Oeser’s edition. The “real” Giulietta act does not run longer than 35 minutes – about ten minutes longer than Choudens (Oeser’s Giulietta act is about an hour in length) – and that is little to ask when there is so much to be gained in those extra ten minutes. All-in-all, even if one includes the Guiraud recitatives, the opera does not run longer than “Faust” or “Boris Godunov” and even some uncut Handel operas – and you very rarely hear complaints about the lengths of these two works (except from people who don’t like them!) And, in fact, many of these myths of length stem from Oeser’s impossibly long version – a version which doesn’t represent Offenbach any more “authentically” than Choudens does. Many people think that Oeser and Kaye’s work is interchangeable, as if the Kaye version were only a slightly revised re-working of Oeser (nothing could be further from the truth). This is because many opera houses act as if time stopped in 1976 1979 (as many opera commentators do) and are either unaware of or apathetic about the HUGE amount of scholarship accomplished in the intervening years – scholarship that has invalidated Oeser’s conclusions.
In my opinion, Offenbach’s opera is a masterpiece of music drama and deserves to be heard as the composer intended it. It is not an unfinished work. Offenbach completed the Giulietta act: it is whole, coherent, makes dramatic sense, and includes a brilliant finale that not only balances out the finales to the Olympia and Antonia acts, but also packs quite a punch. What Kaye, and those who support his work, want is not for Kaye to be glorified, but for people to hear this act at last as well as the other beautiful aspects of the edition that can yield several satisfying versions of the opera. It is almost a miracle that the pieces of this version, now re-collated in the Kaye edition, have survived. It is a grotesque injustice that it took so long for it to be heard. And it is only fitting that at the dawn of a new century in this opera’s history the public is given a chance to hear it and evaluate it.
Charlie Richards,
Los Angeles, California

The reversal of the Antonia and Giulietta acts that started a few years ago was quite shocking to me and changed my view of this opera fundamentally. The new Met production messes with the Muse's role, which also is new to me. But such is my love for this opera, I would be willing, even eager, to be shocked again by the composer's original music. Hoffmann simply is good enough to sustain these multiple changes.

I saw the Sher production and think it is visually tacky and inconsistent. I'm quite willing for someone to try again, but the Met is now stuck with this mess. Such is Mr. Levine's influence, I doubt any changes will happen for some time. I have no idea why he refused the opportunity to hew to an authentic Hoffmann score. But then, I also have no idea why he was so happy with Renata Scotto's screechy singing or why he ousted Jon Vickers decades ago. These political and creative decisions are beyond my ken as a mere opera goer who is not a professional musician.

Surely, if this authentic score has the merit you describe, more and more opera houses will adopt it. And eventually, the Met will be forced to do so as well.

Without denigrating Michael Kaye's scholarship, I'd like to know where he came to the idea that any producer is somehow bound to heel in to his findings -- however "authentic" they may be. A producer's first obligation is to the audience, not the scholars nor even a work's creators. Today's producer has the same obligation Offenbach had -- to provide his audience a good show. As regards some notion of "authenticity," need we only re-create the music? Oughtn't we re-create the theater of Offenbach's day, for example, its scenic and lighting equipment and traditions? Should we be censured for turning out the auditorium lights during performances? Can we re-create the equally important half of the initial transaction -- the Parisian audience of circa 1880 with all its expectations and experiences musical and cultural? All too often, "authenticity" becomes a club that punctilious intellectuals use to beat the creativity out of venturesome artists. If the Met chooses one day to mount a production using Michael Kaye's edition -- bravo. I'd love to hear it. If the Met chooses to revert to the early Choudens edition -- bravo. And if it chooses to hew to its present path -- bravo once again. Perhaps, Mr. Kaye will have to endure what so many artists learn all too well: that even if one spends a lifetime at a given task there's the sad but very real possibility that no one will ever care.

Michael Kaye's discovery of the authentic Hoffman is a Nabokovian saga that some music journalist would do well to take up.

Apologies if I've got this wrong but If the Michael Kaye edition is the one that graced Louis Erlo's Lyon Opera disaster of a production then in my opinion the Met are well rid of it. Too many drastic cuts and some wonderful music lost. Though to be honest if you get the Erlo dvd you might wish it was even shorter. Now the Danny Kaye version I would pay to see...

Nick S. does not have his facts straight. The Opéra de Lyon video of
their production that re-opened the opera house in Lyon is not at all
representative of the Michael Kaye edition. It is a bizarre concoction
of the stage director Louis Erlo and his dramaturg Michel Vittoz - not
Michael Kaye. Even the title "... Des contes d'Hoffmann" indicates
that it something is different here. It means "...SOME Tales of
Hoffmann" and the credits on the video and the live telecast from Lyon
clearly state "conceived by the Opéra de Lyon / Adaptation based on
Michael Kaye's edition of Les contes d'Hoffmann." It is described on
the packaging as "Louis Erlo's inspired [?] reworking" of the opera! But as
a derivative work, it sure isn't anything Offenbach wrote. For example, Erlo has Franz sing his aria ("Jour et nuit") in the middle of the Giulietta act (needless to say, this is NOT positioned this way in Kaye's edition)! It was also Erlo's defective idea that because Offenbach
died that the orchestra should just stop playing during the
Hoffmann-Giulietta duet. I don't understand why Nagano agreed to that., as the jewels in the crown of Kaye's edition is the fully restored Giulietta act, containing the gambling scene (including Giulietta's coloratura aria "L'amour luit dit: 'la belle'") and Offenbach's original finale - both cut from Erlo's production (as is the entire fifth act or epilogue). You may want to see my Amazon review ( for my further comments on this misguided production.
Once again this version is all about Erlo - it is not Kaye's and it is certainly does not represent Offenbach's vision.
However, the year after they made their strange stage production, the
Opéra de Lyons made the beautiful audio CD recording of the grand
opera version of the score in Michael Kaye's edition. THIS is the performance to listen to if you want to become familiar with Kaye's edition.

Those of us who have spent any rime living in New York learned long ago not to expect anything from the Met except glorifications of Mr Levine's genius. Clearly he seems to feel that his genius excuses or justifies anything and that ideas that do not come directly from himself or the fortunate few who surround him are not worthy of the light of day. That does not mean that anything coming from the Met is without merit it simply means that the quality of the thinking behind it varies depending on the murky internal politics of its primary movers and shakers. On the face of it Carvalho was similarly burdened by his own genius.

It should be pointed out that some of those leaving comments seem to infer that somehow Offenbach's score does not deserve the same meticulous treatment that has been afforded Bizet's Carmen or even Verdi's Don Carlos. I find those comments disingenuous since I find it more offensive to restore music that the composers themselves discarded because they discovered in rehearsal that they had written too much or that the material simply did not serve the purpose the composer intended and therefore had to be scrapped. This was never the case with Hoffman. Offenbach was four months dead and could not object to the musical mishmash that Carvalho invented -- in some cases out of whole cloth -- with no regard for the composer's score, much less wishes.

I saw the Met's HD production yesterday at a theatre here in Southern California and for the most part enjoyed the new Hoffmann. One easily gets caught up in the "newness" of the production and the wonderful voices to be able to contemplate anything else. It was only after leaving the theatre and reflecting on my own impressions that I began to wonder: "What ever happened to Pitichinaccio?" Based on that thought, I began to do a little reading and research and came upon Mr. Smith's article. Upon reading the words of Mr. Kaye and many other commentators, it would seem that the mighty Met and James Levine have not served Offenbach's legacy well. The Giulietta Act seemed incomplete and lacked something felt in the previous two. Too bad.

Can anyone illuminate exactly what scores are used in the Met production? I recorded it and have listened to the prologue with the Choudens score in hand. There is the insertion of the muse material right after the opening, and then a few bars of a cappella chorus are cut later on, but otherwise it seems to hew pretty close to Choudens (in the Kalmus republication). Has the Met simply used Choudens with some glossing from the Oeser? What was the source of the muse material in the prologue, for example? Any information would be very gratifying. Thank you?

Recently, this issue was discussed on the internet forum, Opera-L and I wrote (the reference was to the MET's "Carmen", but it is hard to discuss "Carmen" scholarship without discussing "Hoffmann" as the two are interconnected for many reasons):

Yes, Ernest Guiraud composed recitatives for the entirety of “Carmen”, third act included, but it is not at all surprising that, once again, the MET is mixing and matching from different editions rather than adhering to one score for this new production. I’ve even been told that for some operas the MET orchestra plays from public domain parts while the conductor is working from a costly critical edition and transferring information to the public domain scores. I remember seeing a 1997 MET telecast of “Carmen” (the Zefirelli production) and noting that Guiraud’s recits were sung in some cases, and dialogue spoken in other passages, so this type of conflation is nothing new for the MET.

Guiraud has taken a lot of heat from purists in regard to his recitatives for “Carmen”, but I think it is best that we remember that without them the opera probably would never have reached its international audience, nor would it have entered the standard repertoire. Notwithstanding, “Die Zauberflöte”, “Fidelio” and “Manon” are just a few operas relying on passages of spoken dialogue to carry the drama forward, and they tend to be performed outside their native countries. (Transaltions of spoken dialogue for operettas is a different matter altogether).

Guiraud also composed connecting scenes of recitatives for Offenbach’s “Les contes d’Hoffmann”, but that situation is far more complex; Guiraud’s duties for “Hoffmann” went far beyond simply substituting sung recitative for the original spoken dialogue. In fact, although Guiraud edited and prepared the first edition of the score published by Choudens, that edition did not include any of his recitatives; those recitatives did not appear in the score until Choudens’ second edition was published. However, in writing recitatives for both these operas Guiraud did a highly commendable job. One must remember that Guiraud was a fine composer in his own right, and his recitatives for “Carmen” are very well constructed and function harmoniously with the rest of the score. Furthermore, his “Hoffmann” recitatives are, in my opinion, often truly beautiful. One notable example is the lovely scene for Hoffmann and Antonia following the Hoffmann/Miracle/Crespel trio, beginning “Ne plus chanter! Hélas! comment obtenir d’elle un pareil sacrifice?”. As we learn from listening to Michael Kaye’s edition of “Les contes d’Hoffmann”, some of the traditional Guiraud recitatives were derived from authentic Offenbach material that was cut from the opera’s premiere; that is, in itself, fascinating to experience in the theatre.

That being said, there is also an excellent case for restoring spoken dialogue in both of these works, especially if we want to arrive at a text which is closer to the composer’s original conception. But it is a clear mistake to mix the two genres, as the MET persists in doing. The results of that tend to be jarring and even bewildering. Please let’s not bicker about poor pronunciation and acoustics – the MET no longer performs Bodansky’s recitatives for “Fidelio” or Berlioz’s recitatives for “Der Freischuetz”, and when they do the Mozarts Singspiels they use variations of the dialogues.

It is an interesting coincidence that this season the MET is presenting two new productions of two late nineteenth century French operas, both of which have complicated textual histories, both of which were published after their respective composers’ deaths, and both of which are linked together by the figure of Guiraud. And in both instances, from a musicological, theatrical, and historical standpoint, the MET has failed them dismally, using performing versions that are hashed together like hack work from corrupt editions, and failing to acknowledge and realize onstage the results of scholarship, authenticity, and dramaturgy that have been made regarding these two works in the last thirty years! Having already been taken to task for this in the press and in comments from opera lovers here and abroad this season, one hopes that the august Metropolitan Opera will take steps (even if it means extra rehearsals) to revise and improve their latest productions of Carmen and Hoffmann, rather than being content to revive that which has already been deemed to be less that the best possible scenarios for these masterpieces.

Many thanks for passing along your very absorbing analysis. TIM

Several year sag I have prepared two documentary programmes under the title "In search of true Hoffmann" because it was quite obvious that it is still not clear what can be regarded as the "authentic" critical redaction of Offenbachs unfinished masterpiece.
The first question what was real Offenbachs wish and what was imposed on him. In general here are few answers:
1. Offenbachs original plan was to compose the serious phantastic opera written for a Hoffmann baritone and four female voices, and only Olympia as the coloratura
2. The concept for the opera was already finished, as a good part of the musical numbers as early as 1879 when a private concert was held in Offenbachs house in Paris.
3 In that time Offenbach was already very ill man, working under heavy pressure because he had too many compositions to work on.
4.. Offenbach was well known for his habit for reutilization of parts or motives of his earlier compositions,
5. In order to speed the completion of the opera. Offenbach had remembered his previous romantic opera Die Rheinnixen which was a monumental flop back in 1864. Offenbach used the music material of Die Rheinnixen as a skeleton for Act IV of Hoffmann:
Rheinnixen ouverture became the Barcarole; Konrads drinking song became the Bachantic song poof Hoffmann; Choir of the fairies became the gambling scene quartet. It must be noted that some other bits and small sections were also reutilized for original concept of Giulietta s act.
6. Since the previous plans for the premiere of Hoffman had flopped, Offenbach was forced to completely rewrite the already composed and partly orchestrated opera in order to fulfill all of the wildest ideas of Carvalho who took over the production
7. The main changes imposed was the transposing of baritone Hoffmann to tenor, and the transposing the roles of Giulietta, Antonia and Stella to colorature soprano, in order to suite the wishes of Carvalhos favourite soprano In the middle of this process Offenbach died leaving the open door for the Carvalhos butchering.
8. it immediately became obvious that it is not possible to finish the score of act 4 and 5 in short time, thus the act 4 was ommited and only much later reintroduced in the heavy truncated version and in the wrong place.
9.In seventies De Almeida had dixcovered 1660 pages of various manuscripts which enabled Oeser to identify the original material for the acts 1-3 and 5 and the new material which enabled him to restore the originally planned act 4 with the all qoutiations from Die Rheinnixen.
10 In eighties Kaye discovered new material which lead to the reconstruction of the second version of act IV as imposed by Carvalho with coloratura as Giulietta and with the aditional material, but with much less material from Die Rheinnixen, and little more material for the Act 5
11. The one of the crucial arguments for Keye was the discovery of the libretto which was approved by the censorship in 1881. This can not be judged as the final Offenbach wish , only as the implementation of Carvalhos wishes
12. Both editions have their place in the regular repertoiry, because Oeser critical redaction represent the original Offenbach intention, while the Kaye/Keck represent what he was forced to do under Carvalhos order.

The "Hoffmann" Debate

Having seen both the live Hoffmann at the Met on Dec 16th and the broadcast on Dec 19th (at an Ingolstadt cinema) I discovered this most interesting discussion which deserves to be continued. It contains contributions by scholars and amateurs (in the true sense of the word) of this wonderful opera. As soon as I find the time I will respond to some of the aspects mentioned here.
One argument, however, I wish to get rid of now:
Why is the Kaye-Keck version played so rarely, altough it is the best version available?
The answer is very simple: money.
The Guiraud-Choudens version, admittedly the least desirable one, is free of royalties and was chosen by Covent Garden, among others and by one of my two favorites of 2009, at the Wroclaw/Breslau opera. (For details see My other 2009 favorite was Dmitri Bertman´s revival for the Vanemuine in Tartu. (, though still without photos.
The Oeser-version represented a great step forward, since it widened the role of the Muse and extended the finale considerably.
The best available version is no doubt the Kaye-Keck version. But why is it performed so rarely (the most recent being at the Praha Národní Dívadlo with Zürich coming at the end of this week)?
I am surprised that among the learned contributors to this debate no one mentioned the enormous royalties to be paid to the Schott-Boosey-Hawkes group. Up to 17 per cent of the box office intake (depending on the type of theater) will be charged for the rights to perform this version. Will the money go to E.T.A. Hoffmann, Michel Carré, Jules Paul Barbier and Jacques Offenbach?
Now assume that the Met has about 3900 seats which are sold. Assume an average ticket price of 100 dollars. Assume that the Met is granted the most favorable rate by Schott, i.e. around 13 per cent of the box office. This will amount to 50,000 dollars - per night.
I have talked to a number of opera managers, dramaturgs and press officers, and they all said that they would have loved to perform the Kaye-Keck version, but could not afford it. The average cost for the Oeser version seems to be clearly under ten per cent, and thus more affordable to many theaters fighting to make ends meet and still wanting to present an acceptable version of this operatic masterpiece to their audiences.
In practice the differences between the Oeser and the Kaye-Keck version are not as great as between the Guiraud-Choudens and the Oeser versions. So the Oeser version, though deficient, is a viable compromise and the most performed of the three.
Should the Schott group revise their royalty tactics, no doubt the Kaye-Keck version will become the champion.

In the meantime I invite all friends of this fantastic opera who have some knowledge of German to visit my homepage (or www.hoffmannserzä, if you have an "ä" on your keyboard) and join the debate under "Diskussion" by sending an email to me. ( or I am not a scholar, but love this opera, which I have seen in 34 different productions in the past three years, with two more coming up within the next two weeks, Zürich and Giessen). I hope that the Prague report will be online before I travel to Zürich next Saturday.
No doubt that I prefer a well produced, well sung and played Guiraud-Choudens Hoffmann to a boring Kaye-Keck specimen.

Many thanks for adding so interestingly to the discussion here. TIM

This is a response to Mr. Stanislav Zivkov’s post on this blog regarding the authentic version/edition of Les contes d’Hoffmann. Although I can appreciate Mr. Zivkov’s taking the time to address this most complicated topic, I feel it is necessary to clarify and correct a few of his statements. Since he has set out his points numerically, I will address them in that order, but first, concerning his premise that “it is still not clear what can be regarded as the ‘authentic’ critical redaction of Offenbach’s unfinished masterpiece,” I must say that there certainly is and that has always been the edition of Michael Kaye, which is now a co-edition with Michael Kaye and Jean-Christophe Keck, the world’s recognized leading experts on the subject and Offenbach’s life and works.
1. “Offenbach’s original plan was to compose the serious phantastic [sic] opera written for a Hoffmann baritone and four female voices, and only Olympia as the coloratura”

This is only partially true. Offenbach did intend the role of Hoffmann to be sung by a baritone (Jacques-Joseph-André Bouhy, who also created the role of Escamillo in Bizet’s Carmen), but the roles of Olympia, Antonia, Giulietta and Stella were always intended to be sung by one soprano, not four. The artist cast to create these four roles was Marie Heilbron. After the failure of the Théâtre de la Gaîté (where the opera was supposed to have had its premiere), and the change of venue to the Opéra-Comique (Salle Favart), these roles were re-cast and the music extensively re-written. But, as stated, Offenbach always intended for one artist to portray the four heroines (three of whom are, in actuality, only facets of the fourth – Stella). This had also been done when Barbier and Carré play Les contes d’Hoffmann, upon which the opera’s libretto is based, had first been performed in 1851.

5. In order to speed the completion of the opera. Offenbach had remembered his previous romantic opera Die Rheinnixen which was a monumental flop back in 1864. Offenbach used the music material of Die Rheinnixen as a skeleton for Act IV of Hoffmann: Rheinnixen ouverture became the Barcarole; Konrads drinking song became the Bachantic song poof Hoffmann; Choir of the fairies became the gambling scene quartet. It must be noted that some other bits and small sections were also reutilized for original concept of Giulietta’s act.

Here, again, we have overstatement and factual error, confused by Fritz Oeser’s adaptation of the Giulietta act, published in 1977. Offenbach borrowed two numbers from Die Rheinnixen: the elf chorus “ Komm’ zu uns und sing’ und tanze” (Act 3, No. 18 - also quoted in the opera’s overture) which became the famous “Barcarolle” and Conrad’s drinking song “Das Land hochgepriesen es sei!” (Act 1, No. 7) which became Hoffmann’s Chant Bachiques (Drinking Song) “Amis, l’amour tender et rêveur”. That is the full extent to which borrowings from the earlier work entered into Offenbach’s plans for Hoffmann at any time during the opera’s conceptual and compositional history. In 1976, when preparing his highly personal arrangement of Hoffmann, Fritz Oeser found himself faced with a dilemma. The bulk of new, previously unknown material from which Oeser was working dated from the earliest phase of composition—the 1879 Hauskonzert given in Offenbach’s own home in Paris. At that time, a good deal of the material for the authentic Giulietta act, eventually rehearsed at the Opéra-Comique before Carvalho decided to scrap the entire act, that we now know Offenbach composed, and which is published for the first time in the Kaye-Keck edition, had not yet been written. That lead Oeser to the erroneous conclusion (which he states in the preface to the vocal score of his edition) that Offenbach had never completed the Giulietta act, which we now know this is untrue. In order to fill out the act, Oeser went back to Die Rheinnixen and added more than 30 minutes of music from the earlier opera, which Offenbach had NEVER intended to use in Hoffmann. This included the third act finale from Die Rheinnixen (No. 23, beginning “Komm’, die nächt’ge Kühle”), which Oeser inserted in the gambling scene as an unwarranted “Quartet with Chorus”. Oeser also made Schlemil the focus of this scene. In Offenbach’s final, authentic, version as we have it in Kaye/Keck, the gambling scene’s primary purpose is to allow Giulietta to seduce Hoffmann, but Oeser felt it necessary to invent a whole new seduction scene for Hoffmann and Giulietta, once again using music from Rheinnixen (in this case Armgard’s ballad “Dort, wo hundertjähr’ge Eichen,” which Oeser transformed into Giulietta’s “Qui connaît donc la souffrance). Oeser also borrowed bits and pieces of recitative from Rheinnixen as well, completely re-writing the libretto for the Giulietta act in the process. What’s worse, Oeser pawned all of this off as Offenbach Hoffmann! But that is not easy to discover unless one reads Oeser’s commentary to his edition, in which he admits that his Oeser did not try to disguise the fact that his additions and changes were, as Lewis Carroll’s White Knight put it, his “own invention”.

So there is really no doubt at all that this music was added to Les contes d’Hoffmann in Oeser’s edition and does not exist in any of the original sources for the opera. However, despite the fact that Oeser was forthright regarding his manipulations, many (including, apparently, Mr. Zivkov) erroneously believe that Oeser’s Giulietta act represents Offenbach’s original intentions. It does not. Even the one number that does originate from Offenbach’s sketches (the first version of Giulietta’s aria “l’Amour luit dit: ‘la Belle’” was heavily manipulated and falsified by Oeser, who added a chorus, re-wrote the text, and erroneously transcribed Offenbach’s manuscript of that music originally for the baritone Hoffmann).

6. “Since the previous plans for the premiere of Hoffman had flopped, Offenbach was forced to completely rewrite the already composed and partly orchestrated opera in order to fulfill all of the wildest ideas of Carvalho who took over the production”

The revisions made to the score for the Opéra-Comique mainly had to do with cast changes. Adele Isaac now took on the four heroines, and Offenbach re-tailored the roles to take advantage of her remarkable vocal range. Hoffmann was re-cast as a tenor for the Opéra-Comique’s star tenor Alexandre Talazac, and so this character’s music had to be altered accordingly. But all other revisions (and we must remember that by the time the venue had changed, Offenbach’s score was far from complete and hardly set in stone) appear to have been Offenbach’s own decisions, and not “wild ideas” thrust upon him by Carvalho. Carvalho did make sweeping and devastating changes to the score after Offenbach’s death, but this is not what Mr. Zivkov seems to be referring to. It would be unwise to think of Offenbach’s 1879 score as a more pristine or purer version than the score rehearsed at the Opéra-Comique’s dress rehearsal (before Carvalho made his cuts), as the latter version reflects many changes which Offenbach himself implemented, signed and stamped.

8. “it immediately became obvious that it is not possible to finish the score of act 4 and 5 in short time, thus the act 4 was omitted and only much later reintroduced in the heavy truncated version and in the wrong place.”

A much quoted piece of lore, but untrue. The Giulietta act was complete, signed, sealed and ready to go, and in fact was rehearsed in full as it appears in Kaye/Keck, as stated above, and as Offenbach’s own manuscripts make crystal clear. The Giulietta act was omitted for reasons of length and unrelated personal problems of the director (which is also probably why Guiraud’s first redaction of the act, which appeared for the first time in the second Choudens edition, is so much shorter than Offenbach’s original).

10. “In eighties Kaye discovered new material which lead to the reconstruction of the second version of act IV as imposed by Carvalho with coloratura as Giulietta and with the aditional material, but with much less material from Die Rheinnixen, and little more material for
the Act 5”

As my previous comments make clear, the version of the Giulietta act found in Kaye/Keck represent Offenbach’s final wishes and do not reflect any impositions made by Carvalho (with the exception of the cast changes). Once again, Mr. Zivkov seems to assume that Oeser’s version represents Offenbach’s 1879 version of the act, which Oeser himself admits it does not. Act V (also known as the “Epilogue”) is the only one of the five acts that Offenbach left somewhat incomplete upon his death. Michael Kaye had access to manuscripts unavailable to Oeser and Kaye’s discoveries also include a great deal of material for the fifth and final act, including an introductory chorus, a double chorus, a solo for Stella, and several alternative versions of the finale, all of which make for several different very satisfactory re-constructions and ways to end the opera.

11. “The one of the crucial arguments for Keye [sic] was the discovery of the libretto which was approved by the censorship in 1881. This can not be judged as the final Offenbach wish , only as the implementation of Carvalhos [sic] wishes”

The Censor libretto, recovered AFTER Michael Kaye completed the first drafts of his editorial work on the opera, confirms everything that Kaye had done in the first phases of his Hoffmann studies, i.e. Offenbach’s musical manuscripts (primary source material) follow the Censor libretto (secondary source material). With the exception of the incomplete last Act, I believe we can safely assume that this libretto reflects Offenbach’s final wishes, without making conjectures about what he might have done if he had lived to revise the opera at a later time. Most importantly, the Censor libretto includes the text of Offenbach’s original finale of the Giulietta act, one of the very last pieces that Offenbach completed for the opera before his death. The ending to the Giulietta act found in Oeser stems from the version found in the early Choudens editions and later further modified by Jules Barbier’s son, Pierre, for the 1905 production of Raoul Gunsbourg in Monte Carlo. Again, Mr. Zivkov has made this statement under the impression that Oeser’s Giulietta act, very little of which stems from Offenbach or Jules Barbier, was Offenbach’s original plan, which was then changed by Carvalho, an impression which careful study of all the materials (including Oeser’s own commentaries) clearly does not bear out.

12. “Both editions [Oeser and Kaye] have their place in the regular repertoiry [sic], because Oeser critical redaction represent the original Offenbach intention, while the Kaye/Keck represent what he was forced to do under Carvalhos order.”

That conclusion is boldly false, and my previous comments (I hope) have proven, this is not the case. The version of the Giulietta act found in Kaye/Keck represents Offenbach’s final say on the matter; careful study of the materials leaves little, if any, room for debate.

We need to distinguish between Offenbach's ORIGINAL intent and what his FINAL intent would have been if he had lived to see the premier and worked on subsequent productions. He was a Man of the Theater and accustomed to making changes to meet external requirements - and changing his intent during the process. So we can't know what his FINAL intent would have been if he'd lived. Leaving it up to today's directors to decide what "version" best represents THEIR vision of what Offenbach's FINAL intent would have been.

Post a comment

All comments must be approved by the blog author. Please do not resubmit comments if they do not immediately appear. You are not required to use your full name when posting, but you should use a real e-mail address. Comments may be republished in print, but we will not publish your e-mail address. Our full Terms of Service are available here.

Verification (needed to reduce spam):

About Tim Smith
Born and raised in Washington, D.C., I couldn't help but develop a keen interest in politics, but music, theater and visual art also proved great attractions. Music became my main focus after high school. I thought about being a cocktail pianist, but I hated taking requests, so I studied music history instead, earning a B.A. in that field from Eisenhower College (Seneca Falls, N.Y.) and an M.A. from Occidental College (Los Angeles). I then landed in journalism. After freelancing for the Washington Post and others, I was classical music critic for the Sun-Sentinel in South Florida, where I also contributed to NPR. I've written for the New York Times, BBC Music Magazine and other publications, and I'm a longtime contributor to Opera News. My book, The NPR Curious Listener's Guide to Classical Music (Perigee, 2002), can be found on the most discerning remainder racks.

I joined the Baltimore Sun as classical music critic in 2000 and, in 2009, also became theater critic, giving me the opportunity to annoy a whole new audience. In 2010, my original Clef Notes blog expanded to encompass a theatrical component -- how could I resist calling it Drama Queens? I hope you'll find both sides of this blog coin worth exploring and reacting to; your own comments are always welcome and valued (well, most of them, at least).

Think of this as your open-all-hours, cyber green room, where there's always a performer or performance to discuss, some news to digest, or maybe just a little good gossip to share.
Note: Tim Smith now writes about the fine arts at This blog will be kept in place as an archive for an indefinite period. Please visit the new location to get the latest Mid-Atlantic arts coverage.
View the Artsmash blog

Baltimore Sun coverage
Baltimore Symphony Orchestra
Marin Alsop
Famous faces in classical music
Sign up for FREE entertainment alerts
Get free Sun alerts sent to your mobile phone.*
Get free Baltimore Sun mobile alerts
Sign up for nightlife text alerts

Returning user? Update preferences.
Sign up for more Sun text alerts
*Standard message and data rates apply. Click here for Frequently Asked Questions.
  • Weekend Watch newsletter
Plan your weekend with's best events, restaurant and movie reviews, TV picks and more delivered to you every Thursday for free.
See a sample | Sign up

Most Recent Comments
Stay connected