Mozart: Martern aller arten (Die Entfuhrung aus dem Serail)

Lilli Lehman, Maria Ivogun, Miliza Korjus, Joan Sutherland, Edita Gruberova

Meyerbeer: Shadow song (Dinorah)

Maria Barrientos, Luisa Tetrazzini, Lily Pons, Maria Callas, Deborah Cook (complete), Natalie Dessay

Verdi: Sempre libera (La Traviata)

Nellie Melba, Luisa Tetrazzini, Licia Albanese, Magda Olivero, Anna Moffo, Joan Sutherland

When it comes to today's operatic singing and recordings one of the comments I hear most often is that the "great old deads" (artists that pre-date the Long Playing record) were so much better.

But is that really the case?

By taking three arias from the soprano literature and doing an objective analysis of recordings made during various eras, it becomes apparent that this issue is not that clear cut.  There are, however, certain qualifications.

As with anything to do with an art form, there are no hard and fast rules when it comes to interpretations.  A number of factors must be brought into the equation for the answer to be as accurate as possible.  With singing you are dealing with a number of outside influences including audience and political taste in artistic matters at the time, emphasis in vocal studios, and the predominate modes of accepted artistic expression at the time.  One must then take into consideration the recording techniques when the recordings were made.  This last factor often colors today's listener's ability to enjoy an early artist.  Fortunately, with some works of Verdi and Puccini we often are able to study the recorded efforts of actual role creators.  But even this remarkable situation must be approached with caution .

I have found it interesting that one can see strong parallels between earlier and more recent singers - and not necessarily to the detriment of those of today.  Naturally all voices are unique entities and remain a law unto themselves, but think about the similarities between the singers below:

• Luisa Tetrazzini-Joan Sutherland
• Miliza Korjus - Luciana Serra
• Frieda Leider - Hildegard Behrens
• Pol Placon - Samuel Ramey
• Conchita Supervia - Cecilia Bartoli
• Marcella Sembrich - Beverly Sills
• Enrico Caruso - Placido Domingo
• Tito Schipa - Luciano Pavarotti
• Emmy Destinn - Gwenyth Jones

Of course I don't mean to suggest that these singers are identical, but there are some fascinating similarities.  These may have to do with their timbre, their technique, or their manner of singing.  All this actually proves is that a certain type of excellence does re-appear with some frequency during the course of vocal history - subject, of course, to modifications brought about by changes in vocal studios, artistic preferences, and the world situation.

Today, the state of recording is quite different than it was in 1910.  Aside from sound quality, the most important difference (and the one that really concerns us here) is the matter of time restrictions.  In earlier years,(1900-1950) artists were hampered by generally having a maximum of four minutes in which to put forth their interpretation of an aria.  (Assuming that they are trying to fit that aria on one side of a 78 rpm disk.) This resulted in often brutally cut versions of arias that, when on stage, they performed in possibly twice the amount of time.  To judge an early artist without taking such things into consideration is not only unfair, it is completely unrealistic.  How can one determine if, for instance, Emmy Destinn sang "D'amore sull ali rose" (Il Trovatore) slower on stage than in the studio?  Technically we can't.  But, by carefully listening to her recording of that aria, noting her reserves of breath, the ease of her production and the particular feel and phrasing that she exhibits during the recording, one can almost (and I stress almost) reconstruct the slower, more interpretive tempo that she might have used for the stage.  This is, of course only conjecture.  And to be honest this exercise is a bit futile.  It can, however, be illuminating.

In matters of technique we are on more solid ground.  78 rpm recordings are very much like today's live performances.  Although an artist could reject a version (or take) of an aria after it was recorded, a new one had to be immediately made (to replace it) if they intended to release a version of that music.  This is why recordings of such artists as Lilli Lehmann, Luisa Tetrazzini, Maria Galvany and Marcella Sembrich are such remarkable documents.  Without splicing, editing or any other extraneous devices, their recordings exhibit a technical accuracy that is often astounding.

Are they better than more recent singers Sutherland, Sills or Gruberova?  Not according to the plethora of live performances available today.  Their approach to music and their artistic priorities were, however, quite different.  You can find both good and bad singers in all eras of singing.  One often comes across recordings from the 1900s and wonders why certain singers were allowed to make recordings.  I can guarantee you that in the year 2095, listeners and collectors will find recordings made by some of today's operatic artists and wonder the same thing.  It is a matter of taste - which does tend to alter with some rapidity - and a unique layering of factors that all combine to produce what is considered the art of singing during a specific period of time.  A recent example of such a vacillation of taste was the "bel canto" movement of the 50s and 60s - now replaced by a surge in the autonomical power of the conductor.  As if minutely mirroring the situation during the reign of the castrati, after 1975 one finds conductors pruning scores of extraneous notes, taking away various liberties from singers and presenting re-studied, often antiseptically clean but boringly uninteresting, performances.  There is an inherent problem with such a movement.  Opera's existence depends on the singers who interpret the roles and a singer cannot interpret operatic music with individuality if they are not encouraged to use their imagination.  You can have a great singer and a poor conductor and still have an acceptable performance.  But reverse the situation and you have an operatic void.  When you take away a singer's ability to infuse personal interpretation - whether it be in the form of ornamentation or stylistic subtleties of rhythm and tempo or vocal inflections, you rob the music of its ability to breath and survive.  Anyone who doubts this can find easy proof by comparing recordings of Verdi's Rigoletto as conducted by Fausto Cleva on BMG CD with Roberta Peters, Jussi Bjoerling, and Robert Merrill, with that conducted by Riccardo Muti with June Anderson, Luciano Pavarotti and Leo Nucci.  Muti brings out elements and colors in the orchestral fabric that one rarely hears and offers a crisply accurate version of the work which somehow still reeks of the textbook.  Cleva, an admittedly less polished conductor provides a taut, living drama.  He does this by allowing the singers to contribute their own concepts - while maintaining a general rein over things.  Whether one agrees with Cleva's (or the singer's) concepts or not, it remains a fact that a theatrical whole is being presented whereas with Muti's scrupulous version there is the aura of the recording studio.  Even worse is the insinuating feeling of "Look how careful we are not to corrupt Verdi's music."

No matter what purists will tell you, the thing that continues to keep operatic music alive is a singer's imagination.  This is a situation that is in serious trouble today.  With the tightened control of the conductor and "his" interpretation, singers are becoming increasingly more inhibited in their ability to overtly express themselves.  Then one must take into account cutbacks in governmental funding and the lack of job opportunities - a combination which makes singers tend to do whatever it takes to "keep" a job once they have it.  That includes discarding their artistic concepts for the wishes of conductors or directors.  How many times have you bought a new operatic recording, listened to it, and thought: "that was well done - all the notes are there, the phrasing is accurate and the singing stylish, but something is missing." What is missing is that indefinable spark that can be only provided by the singer - creative imagination.  It is that single quality that makes one sit up and take notice, that makes a phrase sung by one singer stay with you to the point that when you hear other artists sing the same music you listen in vain for that same illuminating touch.  That is not genius, it is artistic creativity based upon a free use of imaginative coloration.

I remember a performance of Gotterdammerung at the Metropolitan Opera in the late 1980s.  Hildegard Behrens - an admittedly light-voiced Brunhilde - sang certain phrases of the role in such a manner that, for me, they leaped from the score with remarkable clarity and individuality.  So it is with any great artist.  But a great artist is a different creature from a great singer.  The great singer will produce glorious sounds that often seem unconnected to interpretation.  A great artist may make questionable sounds but will, through the originality of their gift and temperament provide a fully-rounded aural portrait of a character.  Luciano Pavarotti is a great singer.  He is not a great artist.  Conversely, Marta Modl was a great artist but not a great singer.  There are those of course who fit into both arenas.  They include Feodor Chaliapin, Maria Callas, Alfredo Kraus and Placido Domingo.

A conductor can be the most remarkable find of the century but if the singers he chooses to surround himself with are drab or artistically arid, he will not be able to present a representative performance.  I have always felt that to be one of the ironic aspects of a few of Arturo Toscanini's operatic recordings.  Although some of the singers are what one would consider first-string, they are often paired with less than suitable partners and often in roles that are not congenial which makes the resulting recordings lopsided.

The question of who sings "Martern aller Arten" better is not a simple answer.  One can discuss the technical aspects of recordings such as the cleanliness of coloratura passages, accurate placement of high notes or the smoothness of voices, but there are invisible aspects that also come into play: stylistic proprieties and the type and emphasis of vocal training at the time of the recording.  For example, the fact that Lilli Lehmann ends her version with a stupendous, if self-gratifying, high C offends many listeners.  But who will deny that after hearing such a display one does not look for a comparable effect of finality in other versions.  That single high note represents all that is good and bad in singing.  Good in that it is spectacularly produced by a woman approaching 60 and a testament to her technique and individuality.  Bad in that it is an unwritten license of Mozart's composition and (at least now) considered in questionable taste.  But is it?  The high C occurs on the word "Tod" (death).  It is an important word in the text and one that has been repeatedly used.  As a device for illumination and emphasis the high note is perfect.  It finishes the aria with a definite, ringing finality.  The question of whether the note belongs or not transcends matters of taste or style.  It is a subjective decision that each listener and generation must make.

With such arias as Amina's ecstatic finale to La Sonnambula, the addition of extraneous high notes becomes less of an issue than their quality.  Throughout the years certain variants and interpolations have become standard in the Italian repertoire of this era.  Verismo operas are less likely to have interpolations, although some do occur - as witness the unwritten high C during the Riddle scene (Turandot) that Calafs are fond of interpolating, or the peniultimate high C that soprano Santuzza's occasionally interpolate at the end of the duet with Alfio (Cavalleria Rustincana).  Marisa Galvany, Julia Varady and Leonie Rysanek have all used that variant.  Ironically, although it is true that in ottocento works it is the soprani that are most often responsible for top note interpolations, in the Verismo repertoire it is the tenors.

What all this means is that there is no simple yes or no answer to the question were singers better back then?  The answer lies in the correlating of all the factors that went into making their recordings and comparing them against all the factors it now takes to make recordings.  We all know that nowadays love duets can be recorded without either singer setting eyes on the other.  They can be different cities or even countries.  Like skilled surgeons, editors know exactly how and where to make silent splices so that the listener only hears the most perfect rendition possible.  Although the practice is deplorable, in many cases today's artists are so well prepared that this minor detail doesn't phase them.  They simply create the illusion that the other singer is with them.  This is often so successful that the listener never suspects such a deception is being carried out.  The accompanying booklets for the recordings of Joan Sutherland's second Lucia and Beverly Sills' Roberto Devereux state that both singers were so concerned about interpretive continuity in the respective final scenas of those recordings that they sang them in one take.  Ironically, marketing often use such comments to impress the listener.  But, after all, isn't that what Sills and Sutherland would have done in the theatre?  Recording a long scena in one take should not be a remarkable feat but by1973 it had become so.

Another factor that must be taken in consideration is the preferred type of voice for a role.  This changes not only from era to era but also from country to country.  Today, in Germany, the singer of Donna Elivra in Mozart's Don Giovanni often has a heavier voice than the singer of Donna Anna - the exact opposite of what is preferred in America.  It is not a matter of which country is right but rather of aural preference.  Just because one usually hears a certain type of voice sing a role does not mean that it cannot be sung equally well by a voice of completely different size and weight.  The success of such ventures depends entirely on the singer's technique and suitability to the role.  Despite the differences in their voices and approach, it would be difficult to choose between Maria Ivogün's "Martern aller arten" and Joan Sutherland's.  Although Ivogün is more light-weight that Sutherland, there is a feisty quality to her singing that is as appealing as the heavy thrust of Sutherland's brilliant coloratura.  They are completely different and yet each is immensely satisfying.

Every era of singing can boast its own great singers and technicians just as it must admit its failures -ludicrous artists that somehow become popular but that with hindsight are actually uninspired and completely forgettable.

Mozart: Martern aller Arten (Die Entführung aus dem Serail) Lilli Lehmann, Maria Ivogün, Miliza Korjus, Joan Sutherland, Edita Gruberova

Lilli Lehmann's 1911 recording of this aria is fascinating for a number of reasons.  Primarily, however, its importance lies in its depiction of Lehmann's ability to sing florid works.  Known for Wagnerian heroines, this recording displays her outstanding technique and the complete knowledge she had of her instrument's capabilities.  The fact that she was nearing sixty only serves to underline all these factors.  It matters little that the aria is cut to ribbons or that there are some questionably coarse vocal attacks.  Like Khephren's sphinx, Lehmann's recording stands as a testament of an artistic craft no longer practiced.  But it is no longer practiced because the priorities of vocal studios have altered - not because there are no singers capable of emulating her effort.

Maria Ivogün's 1919 rendition is an example of the merits of an entirely different type of voice tackling this complicated music.  Light and airy but full of careful passion her voice glitters while Lehmann's throbs.  Because of Ivogün's great musicianship and grand, if subtle interpretation, her performance is just as rewarding as any on disc.  Miliza Korjus, she of the impeccable if icy technique, offers a bravura performance of immaculate coloratura.  Despite her usual penchant for displaying her highest notes, she eschews the (until then) traditional interpolated final high C.  Korjus provides a different yet still valid slant on Mozart's music - special attention to agogic accents coupled with cool and cleanly articulated vocalism.

We skip a number of decades to find a young Joan Sutherland providing the modern equivalent to the Lehmann rendition - a voice of great size and dramatic capabilities yet completely trained in all florid devices.  Perhaps of all the versions, Sutherland offers the best all-round performance.  Certainly one would have to look far and wide to find a more beautiful sustained high C as she exhibits near the end of the piece.  No dramatic or vocal issue is left unaddressed.  The more recent Edita Gruberova combines the best qualities of the Hungarian Ivogün and the dramatic intuitiveness of Lehmann with a "modern" school of interpretive vocalism.  Amazingly facile, Gruberova sails through the music with assurance and inventive aplomb.

In the case of Mozart's music it would seem that the modern artists are superior to the older, more revered artists because they are better able to provide not only technical precision and polish but also that rare quality of artistic dramaticism.

Meyerbeer: Ombre legere "The Shadow Song" (Dinorah) Maria Barrientos, Luisa Tetrazzini, Lily Pons, Maria Callas, Deborah Cook, Natalie Dessay

Dinorah's fleet Shadow Song has been a favorite with prima donne since its premiere in 1859.  Although not much artistically has changed concerning this piece, its presentation has.  By that I mean that cuts customarily used during the early days of recording have been opened and most often the scene is now given in its entirety.  There is a difference in the vocal approach to the aria as well.  This was caused by the virtual dissapearance of the soprano legere during the last three decades.  Ironically, the complete recording of Dinorah, made by Opera Rara, had as its heroine the American soprano, Deborah Cook, who was one of the last of the legere - light of timbre, sweetly delicate above high C - a throwback to the days of Maria Ivogun and Gwen Catley and Amelita Galli-Curci.

Early versions of this aria, which are almost all abridged, differ little except in the ornaments that the singer decided to incorporate.  Although most artists use a duet with flute for the cadenza, it is Maria Barrientos, the lightest voice of all who provides the most realistic vision of madness as she sings her roulades alone but with varying dynamics as if talking to herself.  A clever and powerful effect.

Luisa Tetrazzini's 1907 disc is remarkable for the incredible elan with which she sings the florid music and the inventiveness of her cadenza with flute the conclusion of which she accomplishes in one breath - a brilliant effect.  I find it interesting that in many ways Tetrazzini combines the cool authority of a technician with the warm subtlety of an interpreter.

Lily Pons' rendition (she made two) shows the unfortunate decline in the art of florid singing that was apparent during the Second World War years.  The close intricacy of this aria's fioriture never suited Pons' technical battery and although she made two stabs at recording it, neither is really attractive.  It is only when Pons is singing above the staff or flinging out her staccati that one's interest picks up.

Maria Callas's recording is a classic if only for the reason that it was the first time on disc one heard such a darkly dramatic instrument employed in such music.  Not to mention the first time one was to hear such a dark, sizable voice blend with a flute!  Her 1954 recording signals the beginning of a more modern approach to this music where accent is more on interpretive matters rather than mind boggling virtuosity.  It matters little that Callas' version of this aria is sung in Italian instead of French.  What does matter is that her diction is superb and that she cleanly phrases Dinorah's scales, interpreting the aria with unusual coquetry.  One must remember, when listening to his recording that such singing as this was considered unthinkable.  Being able to hear Callas' voice traversing such arias as the Mad Scene from Lucia or the Bell Song from Lakmé had a staggering impact on not only her immediate listeners but the operatic world in general.

With Deborah Cook we travel back to the days of Barrientos for a lightly fleet, delicate, well-interpreted complete modern version with its accent on bravura.  It is only marred by the decision to finish the aria on the third of the tonic (high F) rather than Db.  Although the note works, and Cook makes a lovely sound, it is unnecessary and takes away from the climax.

Natalie Dessay offers a different high note interpolation - one used by Miliza Korjus in 1936 - a sustained, penultimate A flat above high C.  Although such range is enviable and the daring not inconsiderate, I prefer the finish without it.  Otherwise Dessay's version is typical of the modern singer: pretty, technically assured if a bit cautious (little sense of vocal abandon or immediacy) and without any sense of character at all.  Natalie Dessay is one of the finest singers in front of international audiences but her singing reflects the problems with modern, universal vocal training.  Everything is in place and lovely but completely forgettable.  To prove this point, play one of the earlier recordings - Tetrazzini, Barrientos, Korjus and then play Dessay.  In this instance the older recordings seem to out do the more modern efforts.

Verdi: Sempre libera (La Traviata) Nellie Melba, Luisa Tetrazzini, Licia Albanese, Magda Olivero, Anna Moffo, Joan Sutherland

With Violetta's florid outburst at the end of Act I we find many recordings, most of which are quite fine.  The voices that sing Violetta's music today are heavier - yet still very agile.  Now that the optional penultimate E flat is no longer considered mandatory and left at either the digression of the singer or conductor, the aria is open to even more singers than before.  Naturally, because of this, one also comes across more mediocre performances as well.  Technically this cabaletta is quite difficult, the tessitura is wide-ranging and - especially near the end - quite high.  The degree of technical facility when one compares various eras of recording has remained rather high - especially in those singers that specialize in this role.  It is very obvious that when it comes to this particular aria that there is no voice type that perfectly suits this role.  It is really a case of individual suitability.  Because of this one can find excellent recordings by such diverse singers as Maria Ivogün, Eleanor Steber, Magda Olivero, Edita Gruberova and Anna Netrebko.  Although obviously of different vocal weight and emphasis, all are excellent performances.  Today one runs across such differences much less.  The lightest voice usually heard in this music is a lyric-coloratura.  Even Roberta Peters, probably the lightest (or highest voice) to assume this role in recent memory, did not do so until some twenty years after her debut.  Although such light-voiced artists as Lily Pons and Mado Robin did sing the role it was not one of ethers successes.  The recordings of this cabaletta under scrutiny here are interestingly diverse.  Again, the more recent sopranos offer better performances both stylistically and technically.

Nellie Melba - famed for her flawless technique and exquisite, silvery timbre gives a rather coarse, vulgar account (at least in regards to vocal attack).  She does, however, offer some remarkably rapid coloratura.  Although one might blame time limitations for some of the sloppy work on this disc, that does not account for the grossly over-exaggerated vocal attacks in the lower registers.  In this case, though I suspect that when in front of an audience Melba's tempi were much more gracious and less frenetic.  Melba was an egotist, but she was not a fool.

Luisa Tetrazzini is even more impressive although at times she is careless and tends to breath wherever she needs to - often in the middle of phrases.  Her athletic elan and the immediacy of the high E flat she interpolates are, however, unforgettable.

By the mid-40s, when Licia Albanese was one of the reigning divas at the Metropolitan Opera, the role of Violetta had changed hands to voices of more substance than Melba and Tetrazzini.  Albanese shows the technical fallout that occured due to this switch.  Although her timbre is not unattractive, it always tended to sound overly mature.  But no matter what, this cabaletta pushes her technical abilities to their extreme.  Coloratura is a mish-mash of hit and miss and nowhere can one find the elan of Tetrazzini or the line of Melba.  Once past this music, however, Albanese was better able to serve Verdi's music and perhaps not surprisingly, Violetta was one of her famous roles.  In strictly technical terms, however, this cabaletta shows how half formed Albanese's technique was.  The fact that one sings Verismo heroines should not mean that you can't sing clean fioriture.  The queen of verismo, Magda Olivero, proved that point with great dignity and vocal abandon.  Her rendition, commercially recorded while still in her youth, successfully blends an earthy verismo roughness with a belcanto line creating an odd style but some intense singing.  It is one of the most striking renditions one can hear.  As if that were not enough, Olivero offers a superb high E flat at the end of the scene.

With Anna Moffo we find a more well-rounded portrait of technical precision married to vocal weight and color.  Moffo made a number of international debuts as Violetta and the role was probably her most famous and for good reason.  Her darkly hued, warmly feminine timbre and her natural, easy florid precision would be outstanding in any era of vocalism.  Although Moffo's top E flat is thinner than her lower regions she takes it with such fearless abandon that one is grateful she included it.  Contemporary with Anna Moffo was Joan Sutherland, who sang Violetta for almost 20 years.  Her approach to Verdi's music was unusual in that her voice was capable not only of easily encompassing the more dramatic moments of later acts but also spinning out voluminous coloratura passages and sailing to the top E flat at the end of Act I.  Like Moffo, in strictly vocal terms, Sutherland's recording is one of the most satisfactory you can hear.

The end result is that one should never dismiss any era of singing too easily.  There are good and bad points about each.  Naturally there will be certain timbres that do not appeal to you, the listener or perhaps the manner in which they present the music does not appeal to you.  But take the time to really get to know the voice and subsequently the singer, and then make a decision based on a more informed, rounded portrait of the artist within her era.

(November 10, 2004)