There are no known recordings of the first Salome, Marie Wittich (Dresden, 1905), and I know of no first-hand descriptions of her voice and technique. In fact, of the principals of the Dresden premiere, the only singer to have left a recorded legacy is the Herod, Karl Burrian, who also sang the part in the Met premiere in 1907. The Salome on that occasion—a one-off, owing to easily offended Anglo-American pieties^(I)—was the Swedish-American soprano Olive Fremstad. Her family had emigrated from Stockholm to Minnesota when she was twelve, but she’d returned to Europe for advanced training (with Lilli Lehmann) and performance opportunities. She initially sang mezzo-soprano and contralto roles (Cologne, Munich, the Bayreuth Festival), but her voice, perhaps due to Lehmann’s influence, gradually migrated upward. She sang an astonishing variety of parts (Jürgen Kesting states that by the end of her Munich years, in 1903, she had clocked in with between sixty and seventy), ^(II)among which Carmen was especially celebrated. She took that role only a few times at the Met, but among her regular assignments from early on were Venus and Kundry, so we gather that seductiveness was a part of her stage persona when called upon. Everything we read about her, including the critical reception of her Salome, suggests a capacity for erotic expression and a willingness to pursue it. Though her first Isolde was still a season off (she sang it on the occasion of Gustav Mahler’s Met debut), it seems fair to say that her combination of qualities came as close to realization of Strauss’s ideal as can be expected. Fremstad’s recordings are acousticals dating from the last years of her operatic career, 1911-14. They include nothing from Salome, but they do show a beautifully shaped Liebestod and a “Hojo-to-ho!” graced by a perfect trill, on-the-button swoops to the high C’s, and a freely sustained top B. In the excerpts that deal more directly with the passaggio connection, the Séguidille from Carmen and “O don fatale,” she reveals a clear, clean, easily accessed chest connection, never deliberately exploited or intensified in the old Italian manner, but also never requiring a vowel modification or positional shift from and into the midrange.

This registrational matter of a strong lower range with clear word articulation, in balance with the required full, freely soaring upper range is of course only one factor, though a crucial one, in the vocal setup needed for a full  interpretation of the  role of Salome. Strauss’s vocal settings in opera embrace a then-new precision in the notation of “line readings,” an insistence on rhythmic exactitude and the inflection of small pitch variations to make sure that every utterance comes out just as he heard it in his mind’s ear, and in close co-ordination with the details of orchestral commentary. Changes in tempo and meter and in refinements of loud-soft variations, changes of key and contrasts in instrumentation, note-by-note articulations and affective instructions, at times crowd in on one another in the space of two or three bars, especially in the writing for Salome and Herod. It amounts to an unprecedented micromanagement of the singer’s options, and while it was part of a general trend in the operatic presentation of character, in line with the shift in the nature of stage reality in theatre both spoken and sung, Strauss was very much in its vanguard. Most of it strikes us as “right,” not only effective but natural, because of Strauss’s keen rhetorical ear—yes, that’s the way a good actor of that time and culture would say it. Thus, the setting is sometimes left in a quite conversational mode, almost like the old recitativo secco, though with light, busy accompaniment, while at others it reaches into a more extended version of that, and at yet others is lifted into the transcendental region we recognize as Strauss’s most glorious pages—all this in a through-composed context, notated and articulated with great precision. He also makes use of rhetorical devices that, though they seem to call for a similar literalness of inflection, actually leave a leeway for fulfillment that depends on close singer/conductor collaboration for realization. This is not the place (nor I the analyst) for a thorough examination of all these relationships, but I’ll trace one here, from the singer’s p.o.v., by way of example. Taking a stroll into the weeds: