As I have acknowledged, Callas had some wonderful singing still ahead of her. Indeed, she had greater success in her second Metropolitan season (1957-58) than she’d had in the first. She sang Violetta, which I was unable to hear, and she repeated Tosca, which I did. And it gave her less difficulty than the Norma or Lucia I’d heard. She seemed to have it well in hand, and I recall no wobbles. But I noted in the same EMI Lucia review I referenced above that she was ” . . . not very exciting in the climactic passages of a Puccini opera, because her voice is simply not large enough or warm enough to dominate the orchestra.” From 1956 forward, there is general agreement among those who heard her that the voice was in retreat in terms of volume, that it was becoming less reliably steady, and that the hooded, or veiled, cast on the timbre was increasingly invasive. In the first volume (1960) of the Bulletin of the Institute of Verdi Studies, devoted to Un Ballo in maschera, the Italian critic Giuseppe Pugliese, who surely had heard Callas often in the 1950s, wrote “Whatever may be said or written . . . about the vocal powers of Callas, in this recording [made in September of 1956, the month before her New York debut] they are no longer what they were a few years ago . . . The vocal breadth, the overwhelming richness, the fabulous extension of her voice have gradually diminished, leaving almost unaltered the incomparable charm of vocal colour but limiting its interpretative possibilities . . .” Voice is as voice does, and she changed its way of doing—not, as many singers do, gradually, and in the direction of heavier repertory, but quite suddenly as such things go, and toward a lighter repertory. I can’t think of a comparable case.

Here is what I believe about the trajectory of Maria Callas’ voice. I cannot “prove” any of it—these are conjectures, opinions. But then, as the sagacious William Weaver said, “Everything about Callas is an opinion.”

I believe her voice was “meant to be” that of a spinto-to-dramatic soprano, which through a combination of natural gifts and training had unusual mobility and a high extension. Its unique, haunting quality and her unsurpassed stylistic and dramatic instincts were natural to her, aspects of genius that clambered for expression via whatever her technique would allow. In early usage, the full calibre of the voice emerged, but not in perfect alignment with its other assets. Had that been recognized early on, and the appropriate steps taken, its lower-middle range could have been stronger, the middle octave leaner and brighter, the top freer. Most of the roles she sang would still have been available to her, but would have sounded rather different—cleaner and more direct, and probably less mysterious and complex. The bred-in-the-bone vocal personality might thus have seemed less fascinating, but the actual range of interpretive options would actually have broadened, and the voice should have shown longer endurance.