During my long career in music, I never met Robert Craft, who died last November at age 92—yet he played no small role in shaping my musical consciousness. As a conductor, Craft recorded the craggiest works of the 20th century, including almost the entire oeuvres of Igor Stravinsky, Arnold Schoenberg, Anton Webern, and Edgard Varèse. As a writer and public intellectual, he weighed in trenchantly not just on musical issues, but on literature, film, and painting as well.

To this résumé must be added the vocation of alter ego, because from 1948 to 1971, Craft was inseparable from the music and household of Igor Stravinsky. The exact workings of their partnership, however, remain a matter of controversy. Not exactly an amanuensis, Craft was a constant companion, artistic consultant, coauthor, coconductor, ghostwriter (for both the composer and his wife), and, after Stravinsky’s death, keeper of the flame. To some extent, he was also a cocomposer, or so he claimed. As early as 1952, he persuaded Stravinsky to rescore a movement of his Cantata—no small matter, considering that Stravinsky is generally considered one of the greatest masters of orchestration. The Cantata also marked Stravinsky’s first move in the direction of serial composition, a change for which Craft took full credit.

Perhaps because we have fetishized the idea of genius, it is hard to picture the exact workings of conjoined creativity, even though the Craft-Stravinsky partnership was not unique. When an interviewer asked Billy Strayhorn to describe his similarly contested collaboration with Duke Ellington, Strayhorn pondered the question, repeated it, then replied, half-ironically, with another: “What would you say I do?” By contrast, Craft responded to questions about his connection to Stravinsky with a mountain of books that extended their collaboration semi-posthumously for another 44 years, and are replete with variously restated accounts of their working and personal relationships that will require an army of musicologists to sort out, parse, and verify.

Here’s a small sample of the challenge. In Stravinsky: Chronicle of a Friendship, published in 1994, Craft rejected the usual notion that he was Stravinsky’s Boswell, arguing instead for a much grander partnership in the spirit of the journal-keeping Goncourt brothers: “I would lean over his shoulder as he wrote, each of us acting as the other’s intercessory.” Craft similarly claimed in Stravinsky: Glimpses of a Life (1992) that the series of conversations between him and the composer, which began to appear in book form in 1958, were “the only published writings attributed to Stravinsky that were largely by him.” Yet Craft also stated in Glimpses that Stravinsky’s English wasn’t fluent enough for a sustained dialogue—an astounding admission, because the Conversation volumes present the composer as a contemporary Dr. Johnson with a masterful command of English, an Oxbridge don’s breadth of cultural reference, and a curmudgeon’s zero tolerance for fools, critics, and most other conductors. In short, “largely by him” was a fiction; the actual words were Craft’s. If you read the articles, letters to the editor, and testy rejoinders that Craft wrote for The New York Review of Books long after Stravinsky died, you’ll find that their style is indistinguishable from the distinctive, combative one that Craft says was entirely Stravinsky’s own.

Craft made similar claims about the Goncourt effect in Stravinsky’s work as a composer, especially when Craft nudged him toward the 12-tone technique of his erstwhile nemesis, Arnold Schoenberg. This unexpected, late-life stylistic conversion puzzled and alienated many of Stravinsky’s old friends and acolytes; but it brought triumph with the score for the ballet Agon, completed in 1957 when Stravinsky was 75 (and Craft, 34). The youthfully inventive score sprang from a performance, conducted by Craft and attended by Stravinsky, of Schoenberg’s Serenade, Op. 24, in 1952. The confrontation with music that Stravinsky had assumed, in ignorance, to be inimical to his own was one lesson in a larger re-education program that the young conductor systematically laid out for his apparent mentor. Stravinsky, who often based his compositions on other people’s music, now seemed eager to absorb all the music, from Gesualdo and Schütz to Varèse and Stockhausen, that Craft placed in his path.

For me, Craft’s relationship with Stravinsky is reminiscent less of the collaboration between the Goncourt brothers than the relationship between Charles Kinbote and the imaginary poet John Shade in Vladimir Nabokov’s Pale Fire. In his annotations of Shade’s poem “Pale Fire,” Kinbote is a collaborator and fabricator every bit as detail-obsessed and boundary-resistant as Craft. Kinbote’s annotations devour Shade’s poem, making it impossible to decide which of the two imaginary authors is, fictionally speaking, real. The analogy with Kinbote turns sinister when applied to Craft’s role in Stravinsky’s actual life. “Cher père, Chère Véra,” the central chapter of Glimpses of a Life, documents the long legal battle with Stravinsky’s children over the composer’s estate, a dispute in which Craft, allied with Stravinsky’s childless second wife Vera, was (to say the least) an interested party. The reader plowing through this heap of dirty laundry with guilty pleasure can only wonder whether the letters, testimonies, and documents presented, summarized, paraphrased, and in many cases translated by Craft were as fictitious as Kinbote’s distant northern land of Zembla. Craft’s most extensive literary dispute was with Stephen Walsh, the British author of a mammoth Stravinsky biography, the second volume of which is largely devoted to questioning Craft’s accounts of his work with Stravinsky and his motives.

* * *

In retrospect, I count myself somewhat fortunate in having never met Craft or tangled with him in print, though I’ll admit that at times I’ve envied not just his life with Stravinsky but also the grand-hotel accommodations and high-table conversations that apparently went with it. Rather than attempting to answer the many questions raised in Craft’s books, I come, instead, to praise what may be his more important legacy: his recordings—particularly the early ones. As with the prose, the recordings will require careful accounting by future scholars. Many of those made after Stravinsky’s death were reissued and repackaged by several labels, making it hard to determine what was new and what recycled. Even more obscure is Craft’s role in the monumental Stravinsky Conducts Stravinsky series, recorded for Columbia Records in the 1950s and ’60s, a period when the composer—never a great conductor—was increasingly frail, and when it was well-known that Craft would prepare rehearsals before Stravinsky mounted the podium. Because the aim of the project was to preserve the composer’s definitive intentions toward his work, and because these recordings have such a distinctive sound, the question of who was leading the band still looms large. In later years, Craft rerecorded most of Stravinsky’s music, often with tempos quite different from those heard on the earlier series. But whose mistakes was he correcting?

The recordings that had the greatest impact on me were the ones Craft made in the 1950s, in particular the complete music of Anton Webern; major works of Berg, Schoenberg, and Varèse; and hot-off-the-presses music from the European avant-garde: Boulez’s Le marteau sans maître, Stockhausen’s Zeitmasse. These recordings, most of them unprecedented, altered the dimensions of my early-adolescent musical universe. A 12-year-old’s mind (reconstructed at a great distance) is hardly an accurate reflection of grand cultural forces, but for what it’s worth, hearing these recordings was as important in dismantling the certainties of my childhood as would be, a few years later, my reading of Paul Goodman’s Compulsory Miseducation or watching Stanley Kubrick’s Dr. Strangelove.

I had developed a precocious, nerdy passion for 20th-century music that began with a recording of Debussy’s La Mer that I heard when I was 5, and was affirmed a few years later when I encountered Copland’s Billy the Kid; but much of the “new” music that came my way (via high-school band or WQXR) in the period sounded old, even to me. A lot of it was stodgy Americana written in a narrow stylistic range from Howard Hanson on the right to Paul Hindemith on the left. In retrospect, I see that I grew up at a time when the Popular Front idiom of the 1930s had become the official music of the Eisenhower years. Some of the most successful—and even most politically potent—examples of the style arrived a decade or two after the Great Depression. Copland’s Appalachian Spring, Barber’s Knoxville: Summer of 1915, Menotti’s Amahl and the Night Visitors, and Broadway musicals like The Pajama Game, The Most Happy Fella, and even West Side Story were all, in their way, belated flowerings of an older leftist sensibility, transplanted to and repurposed in an alien environment. I can now also see how, perhaps under the pressures of McCarthyism, many of these works repackaged political issues as psychological ones. But at the time, they just seemed like the music of my parents’ generation, comfortable (if you were growing up in a New York liberal household) and somehow stale.

While most of my generation would find its own musical vitality in rock and roll or the folk revival, I got my kicks, thanks to Robert Craft, from atonal expressionism. Even though the recording seems to have vanished without a trace, I still recall the jolts of the opening bars from his recording of Schoenberg’s Septet, Op. 29, which Craft conducted with a combination of Viennese Schmäh and un-Viennese honky-tonk aggression. Suddenly, I felt that I’d left the sodden orthodoxies of my Rodgers and Hammerstein childhood far behind. I found a similar electrical charge in Craft’s recordings of Schoenberg’s Serenade, Op. 24, and his Variations for Orchestra, Op. 31. These two pieces reminded me of Stravinsky’s Agon, which had become my favorite piece of music as soon as the recording appeared in 1957; I didn’t know at the time that Stravinsky had studied both scores when he composed his ballets, or that the recordings of the Schoenberg pieces were made at the same studio sessions that had produced Agon. All of these LPs shared a dry acoustic and a punchy, unpolished aggressiveness that made the music sound new—even though the Schoenberg pieces were over 30 years old. Some of the most exciting music that Craft recorded, however, was even older: Schoenberg’s Five Pieces for Orchestra from 1909, or Berg’s Altenberg Lieder from 1912, or just about any piece on the pathbreaking account of the complete Webern.

When I began college a few years later, I learned that Craft’s revelatory recordings of the Second Viennese School were part of a larger reassessment of 20th-century music that had begun in Paris in the late 1940s. Both in Europe and America, educated opinion now believed that Schoenberg, not Stravinsky, was the central figure of modern music; the academic consensus, moreover, was that Schoenberg’s discoveries were just the dawn of a new era, the beginning of an international idiom that would replace the older harmonic language. To his credit, I don’t think Craft, despite all the hyperbole he mustered to promote this repertory, ever fostered these simplistic notions, which reduced a half-century of music to a narrow one-way street; for example, unlike many of the Europeans, Craft didn’t disconnect the “progressive” works of his favorite composers from their “regressive” neoclassical works. For him, these works not only foreshadowed a golden age to come; they created their own golden age, imperishable gems somehow embedded in the rubble of dark times.

How do these recordings—many of them easy to find on YouTube—hold up? Does it matter? Most of the works Craft recorded in the 1950s had almost no prior performance history; Craft, moreover, usually worked with pickup groups (often the top musicians of Hollywood or New York) instead of established orchestras. Knowing all this now, I’ve gone back to the old recordings expecting the worst, and in general have been pleasantly surprised to find that the thrill is still there, even when I find another recording that I prefer. Despite years as a conductor, Craft was notoriously primitive in his technique; I’ve seen some of New York’s finest instrumentalists (all of whom told me how much they liked working with him) mimic his caveman-like downbeats; and yet the recordings convey a love for the music that rises far above accuracy. Craft rarely talked about the emotional content of the music: You would hardly glean from his liner notes that Webern was a devout Catholic, for example. And yet his performances—more, I would say, than those of Boulez—often reveal the expressive essence of the music, whether it is the unaccountably joyful sound of Webern’s Concerto, Op. 24 (composed in Austria in 1934), or the more understandable horror of Schoenberg’s 1909 Premonitions.

As an explorer of unplayed or underplayed repertory, Craft set a high bar for future performers. He was one of the first champions of Schoenberg’s Accompaniment to a Cinematic Scene, Op. 34—orchestral music for an imaginary film that premiered in Berlin in 1930, on the brink of a disaster palpably foretold in the music. It is one of Schoenberg’s greatest works, and like other great works of music, its full impact only appears as the fruit of multiple interpretations. On Craft’s first recording, from 1960, you feel like you’re trapped in a dark movie palace with Dr. Caligari. A more recent recording of the piece by Hans Zender and the Saarbrücken Radio Symphony Orchestra suggests that the imaginary horror movie was the 20th century itself, as well as the beginning of the 21st). Without Craft’s exploratory and explanatory work, much of the most important music of the last 100 years would have been dismissed as difficult and inaccessible when, in fact, it was beautiful and true. For this we owe Craft, as irascible as he seems to have been, much thanks.