A while back (too long ago for SEO purposes), I posted the first half of a piece about the prima donna assoluta del mondo, Maria Callas. I promised a total of ten heresies about that most famous and so often misunderstood singers. Herewith are the other four.

7. The second Tosca recording is better than the first

The received wisdom is that the recording of Tosca that Callas made under the baton of Victor de Sabata in 1953 is one of the greatest recordings of a complete opera ever made.

It isn’t.

The recording belongs in the class of star-conductor recordings of standard repertory operas that call upon the listener to admire passing details and revel in not-usually-heard instrumental voices, never mind the forest for the ear-caressing trees, and never mind the drama, which suffers enormously when the conductor spends the entire performance staring at his own navel.

If Tosca is to succeed, it needs to be played to the melodramatic hilt of the knife the heroine uses to stab Scarpia. The de Sabata performance is hailed by John Ardoin in his seminal guide to Callas’ recordings, The Callas Legacy, as a tasteful and highbrow Tosca. But who wants a tasteful and highbrow Tosca? Not only has the melodrama been sapped from the score; so has the fun. If you take away the histrionics, there’s really not a whole lot left to the opera. As music, Tosca is far, far less interesting than Puccini’s more adventurous scores such as Madama Butterfly and La fanciulla del West.

Yes, certainly, Callas is in steadier voice under de Sabata than she would be in the 1964 remake under Georges Prêtre, but, already, in 1953, that steadiness is the result of certain amount of holding back the instrument (a comparison to the Mexico performances of 1950 and 1952 is instructive here.) That may, however, be partly the result of the influence of Walter Legge, whose recordings always give me the impression that the producer thought high notes were in bad taste.

To my ears, the show in the first Callas Tosca is stolen by Giuseppe di Stefano, whose charm and personality come across far more vividly than Callas’. In 1953, di Stefano had yet to ruin his own God-given voice, and his “E lucevan le stelle” is one for the record books, even if it’s played like a clarinet concerto with vocal obbligato.

Move ahead 11 years to the Prêtre recording and you get something that sounds a great deal more like Tosca. The vocal chips fall where they choose, but they’re given a chance to communicate the drama. The Italians would call the performance scatenato (literally: unchained), and it’s all the more enjoyable because of it. Even Tito Gobbi seems to be having a better time, and, if he’s perhaps not quite as beguiling as di Stefano, Carlo Bergonzi is a model Cavaradossi in his own right.

Although Tosca was Callas’ first leading role in Greece as well as the staple role of the 1964 comeback, she sang it infrequently during the main part of her career. One reason may be that she was not a natural Puccini soprano in the way Tebaldi and Olivero were. The lay of verismo lines never came entirely naturally to her, partly because she wasn’t singing in her native tongue, and partly because she had a great deal of difficulty sounding convincing in parlando passages. (As her often stilted interviews show, Callas was far more comfortable singing than speaking.) That said, her most convincing Tosca may be her earliest surviving recording of the role, the 1950 Mexico broadcast. It’s definitely scatenato and fun, and the young Callas’ high Cs are thrilling.

8. She would probably have liked to be able to sing better than she did

The single most baleful influence Callas has had on opera is the idea that ugly, unsteady singing is somehow more “dramatic” and “artistic” than producing a steady and, if possible, lovely tone. According to this thinking, Callas was a great artist because she wobbled, or at least because she was willing to sacrifice good singing to some kind of extra-musical expression.

That’s nonsense: you can’t have opera without singing. Yes, Callas interpreted the drama of her roles and had a voice of many colors, but the last thing she did was sing badly on purpose in order to be “dramatic”. Her art was too centered on the music for that. Indeed, I’d suggest that Callas would very happily have been able to sing better than she did during most of her career: consider the exuberance with which she throws her voice around in the Mexico broadcasts. Nowhere is it written that you have to sing badly to be dramatic, even if that’s the lesson so many fools have taken away from the Callas phenomenon.

The evidence shows that Callas spent most of her career watching her voice slip away from her, and it can’t have been a pleasant experience. The comparisons to Tebaldi and her “perfect” instrument must have been onerous: no opera singer wants to hear that she doesn’t sing well or has an ugly voice. And one should never confuse producing an intentionally ugly tone (as Callas did in her early Medeas, for example) with singing badly: you can make a histrionically apposite sound without it wobbling.

As the career progressed and the voice declined, Callas’ nerves must have taken a serious battering every time she had to go out in front of thousands of people and not know whether the notes would come out or not. Making an operatic career on a flawed vocal mechanism that was unraveling before your very ears must have been inordinately nerve-racking. That may well explain a good number of Callas’ neuroses and why, for all intents and purposes, she retired at the age of 36.

While one could argue that some of Callas’ artistry was the result of compensating for a voice that was increasingly unruly, that’s not borne out by the earlier performances that show her singing extremely well and yet doing great things interpretatively. Can one really doubt that she wished she had been able to sing in 1958 the way she had in 1950?

9. She wasn’t an “actress”

Callas didn’t invent operatic acting. Moreover, she most certainly didn’t invent the kind of operatic “acting” to which she inadvertently gave birth. Natalie Dessay, whose hyperactive Lucia would have horrified Callas, was one of many singers who claimed to have been inspired by Callas without having understood the first thing about her art. Mostly – and this is something people persist in not understanding about her – Callas didn’t conceive of acting as something distinct from the music she was singing. Reports confirm that she did very effective things onstage, but, with Callas, the drama always came back to the music.

Eyewitness accounts of Callas on stage often refer to the stillness of her portrayals: she clearly knew when to stand in one place and was able to command the stage without needing to move around. Although she unquestionably had the kind of stage presence you either got or you ain’t, there was also a practical reason for the stillness: the extremely short-sighted Callas had difficulty seeing where she was when she was onstage.

That she wasn’t an actress in the sense of something divorced from the music is borne out by the Pasolini Medea film. Called upon to act without singing, Callas flopped.

The strongest proof that Callas was not an “actress” in the sense of “acting” independently from the music is that her recordings continue to be best sellers and continue to capture the imagination of new generations of listeners. The emotion, the drama, the feeling and the character are all there in the voice. Whatever she was like onstage, I think it fair to say that hearing Callas was more important than seeing her. If you had to see her to get her artistry to understand it, EMI wouldn’t have the sales figures it does and YouTube wouldn’t be awash in bootleg recordings of Callas in live performance.

10. No one in her right mind should emulate her

I’ve said it already, but it bears repeating in this context: Callas ruined her voice. She sang the wrong repertory, sang on vocal capital and arguably never had the voice built up properly to begin with – all mistakes that singers who ruin their voices make. There are singers whose careers were even briefer than Callas’ (the case of her “successor” Elena Suliotis comes to mind) but Callas’ twelve years before the public were definitely brief by operatic standards.

The misconceptions that abound about Callas, especially the fallacy that bad singing is dramatically preferable to good singing and that a wobbly sound is therefore somehow better than a beautiful one, have led to the early demise of countless voices whose owners followed slavishly in their misconception of Callas’ footsteps. Given Callas’ position in the history of opera, there probably always will be Callas imitators, however much copying her interferes with vocal longevity.

The fact is that, of all the opera singers in the world, the last one to take technical advice from is Maria Callas. There are lessons to be learned from her recordings – questions of phrasing and expression and musicality – but these often are the ones young singers fail to take away from listening to Callas. (Indeed, one often wonders how carefully Callas’ imitators even listen to the recordings.) But using Callas’ recordings as a class in vocal technique is a recipe for disaster. The rule for young singers should be this simple: whatever Callas did, don’t do it.

The reader must, of course, decide whether my ten hypotheses are heretical as I thought they were when I first set fingers to keyboard. My criticisms of Callas may sound harsh, but her legacy is rarely taken with anything resembling a grain of salt. Now that I’ve provided a two-pound box of kosher salt to go with the legend, I suppose I also should put in a word in Callas’ favor: I do enjoy her recordings, although not as much as when I first discovered them as a teenager. I listened to some of them again preparing this article, and, yes, while the de Sabata Tosca left me cold, the second Mexico Aïda remains one of the great performances of the opera, with Callas turning in an absolutely thrilling interpretation of the title role. She was, however briefly, capable of some genuinely marvelous things.

Luckily, the Mexico performances are all there on YouTube to be listened to for free and to show what she sounded like when she wasn’t struggling with the vocal problems that led to the premature end of her career.

The sad irony is that so many people who claim to worship Callas fail to get her. She could have had a very positive influence on the future history of opera, but she inspired a mistaken orthodoxy of “acting” and “theater” that fails to capitalize on her genius. Future Aïdas can learn a great deal from her 1951 and 1964 recordings of “Ritorna Vincitor” – how to pace the drama as the monologue develops, where to apply a little more voice, where to apply a little less, in short, how to interpret the music and what the possibilities for musical interpretation are. Instead, we keep getting Lucias who “artistically” struggle with the music and fail totally to capture even the moments of genius Callas could bring even to a role unsuited to her voice.

I often cite Callas’ début as Gioconda in the Arena di Verona in 1947 as one of the two dates from which one can reckon the decline of opera. That’s not Callas’ fault: opera went into decline because of the ways in which people misunderstood her artistry, not because of anything she did. But people persist in taking the wrong things away from her legacy, and, in Callas’ name, leading opera further and further away from that which was central to her artistry, the music. We, deputizing for Callas herself, can only shake our heads at the results.