The Weill Project blog: Lost in the Stars

Lost in the Stars was Weill's last completed work for theater. It was a long time in the making. Although based on a novel by South African author Alan Paton 🔗 and set in South Africa, it is as much about race relations in the U.S. as about South Africa, and the long road to the creation of the play shows how white people with anti-racist intent can still be unconsciously rather racist.

Weirdly, the road to Lost in the Stars begins roughly a decade before Paton wrote and published Cry, the Beloved Country 🔗, the novel on which the story is based. It began as an effort to dramatize another book entirely, Harry Stillwell Edwards' 🔗 novella Eneas Africanus (1920). In 1938, after working together on Knickerbocker Holiday, Weill and Maxwell Anderson started adapting Eneas Africanus. Edwards' story is set during Reconstruction 🔗, and its central character is a former slave trying, as a point of honor, to return a family heirloom to his former owner. By all plausible accounts, it's a pretty racist book, even if its author probably didn't see it that way, and apparently at first neither did Weill or Anderson. Originally, they imagined Paul Robeson 🔗 (!) in the title role. Robeson basically said, "Hell, no!", though his wife Eslanda who wrote the rejection letter on his behalf toned it down a little. Eslanda's letter objects to this not just as a role for Robeson, but as a work in general. Essentially, the Robesons see the central character as an Uncle Tom, and suggest that Anderson and Weill try writing about "a negro… who has something to do with reality."

Chastened—well, half-chastened—by Robeson's rebuke, they began to rework Eneas Africanus as Ulysses Africanus. The central character changed from being a stable-manager to someone who "takes over a minstrel show for which he directs and performs in a musical version of Homer's Odyssey." This draft borrowed far more from Homer and Virgil than the novella had. The central character makes a metaphorical journey to the Underworld that (in Stephen Hinton's words) "symbolizes a rebirth of the individual"; this rebirth is abetted by another freed slave named Nicodemus (the apocryphal Gospel of Nicodemus tells of the Harrowing of Hell). He is forced (Hinton's words again here) "to confront the bigoted culture in which he has lived up to that point and in the process call into question his own (and his race's) belief system," which is to say that he begins to reject slave Christianity. Going almost 180 degrees from Robeson, they were now hoping to star Bill "Bojangles" Robinson 🔗, whose career had begun in minstrel shows and whose tap-dancing had made him the most highly-paid African American star of his era. To put this in present-day terms, going from Robeson to Robinson was about like going from Samuel L. Jackson to Flavor Flav, pretty much the opposite direction from the evolution of the script.

Anderson and Weill got as far as a script and numerous songs, but Anderson was increasingly focused on his play Key Largo, and Robinson was open-endedly performing in the smash hit Hot Mikado 🔗, first on Broadway and then at the 1939 New York World's Fair. Also, rights to Edwards' novella proved to be too expensive, and World War II was looming, which (as Hinton points out) was not great for what had become a play about racial tensions and questioning religious belief. Anderson and Weill were sitting on an unusable script and a handful of good songs, most notably "Lost in the Stars". The project had led Weill to study the minstrel show tradition, which he eventually put to good use in Love   Life.

The song "Lost in the Stars" had originally been in African-American dialect, but that had changed by the time it was ever recorded or published. Weill's wife Lotte Lenya did the first recording in 1943, followed by Walter Huston in 1944 as a B-side for "September Song" from Knickerbocker Holiday, and Frank Sinatra (1946, released 1949). The Sinatra recording was originally intended to come out at the same time the sheet music was published, but that didn't happen; it was finally released three years later when the song had become the title song of a play. Lenya's 1943 recording was part of a 6-song album that also included another song from the abortive Ulysses Africanus—"Lover Man", later reworked into "Trouble Man" for the Lost in the Stars—plus two Brecht/Weill songs and two of Weill's songs from his time in Paris after first fleeing Germany.

On and off over the next decade Weill and Anderson kicked ideas around for what they might do with these songs. Something science-fictional, with the "stars" being literal. A play that would begin in the kitchen of an apartment on Central Park South, "with," Anderson wrote, "a negro singing 'Lost in the Stars' at the opening as he's washing the dishes—and a group of youngsters [implicitly, white youngsters] who don't know what to do with their lives." When Anderson was involved in the early work on Street Scene before being replaced by Langston Hughes, he wanted to use "Lost in the Stars" and "The Little Gray House" there.

The only real constant? As Anderson had written to Robeson when they first reached out to him, the song fit "the story of a man in a chaotic world in search of his own manhood and his rules of conduct." According to Hinton, Anderson had a strong attachment to stars as an image. Shortly before teaming up with Weill for Knickerbocker Holiday, Anderson had written the 1937 play Star-Wagon, in which the inventor of a time machine tries to go back 35 years and fix his own past, with predictably bad results. Stars also figure prominently in Anderson's poem "Epilogue," a version of which is inscribed on his gravestone.

In 1939, Anderson published an essay, "The Essence of Tragedy"; you can find it online here. 🔗 Excerpting from that: "a play should lead up to a central crisis, …a discovery by the leading character which has an indelible effect on his thought and emotion and completely alters his course of action. The hero … must not be a perfect man… [because] when he makes the discovery… he must change for the better. The fault can be a very simple one—a mere unawareness, for example—but if he has no fault he cannot change for the better, only for the worse. … In a tragedy he suffers death itself as a consequence of his fault or his attempt to correct it, but before he dies he has become a nobler person because of his recognition of his fault and the consequent alteration of his course of action. In a serious play which does not end in death, he discovers that fault during the course of the action, and he does what he can to rectify it at the end. … From the point of view of the playwright, then, the essence of tragedy or even a ser[i]ous play, is the spiritual awakening or regeneration of the hero."

Alan Paton's Cry, the Beloved Country doesn't fit this bill perfectly, but it comes close. It is, in Anderson's words, a book about "lost men clinging to odds and ends of faith in the darkness of our modern earth." By seizing on this book as the structure in which to rework some of the Ulysses Africanus material, Anderson brought the story into the present day, came up with a much more legitimately anti-racist story, but also put the story at one remove from the U.S.

The four main characters of both the novel and the play are a pair of fathers and their respective sons. Stephen Kumalo, the central character, is a black preacher; his estranged son Absalom has become a petty criminal. In a botched burglary, Absalom kills Arthur Jarvis, a white anti-Apartheid activist. Through his father's influence, Absalom truly repents, confesses, and is condemned to die. Arthur's father, James Jarvis, is a rather racist landowner, but in the end, influenced by his son Arthur's writings, James changes his views and ultimately befriends Stephen, realizing they have both lost sons. There are other plot threads; the Wikipedia article on Cry, the Beloved Country 🔗 does a fair job of summarizing them, if you're interested.

As with Street Scene a few years earlier, the "musical tragedy" Lost in the Stars combined Broadway elements with, in Weill's words, "very serious, tragic, quite un-Broadway-ish music of operatic dimensions." Most, though not all, of the more Broadway-ish music came from the residuum of Ulysses Africanus, whereas the "serious" music was written specifically for the adaptation of Paton's novel.

Weill made the interesting decision to "musicalize" only the black characters. This could be seen as a racist move that "others" them, but I would say that for a late 1940s, overwhelmingly white, Broadway audience, the black South African characters were going to be "othered" in any case. Just as arguably, musicalizing gives these characters more emotional power and certainly it makes them more memorable compared to the "un-musicalized" white characters. A Greek chorus of Zulus comments on the action, which certainly brings their point of view forward and puts them in a position of some authority. Weill's music for the play is a conscious mix of idioms: outside of his his own earlier Broadway tunes that made their way into the play he draws on blues and jazz, and on European classical music, but also on Anglican Church music, and he even spent some time studying Zulu music in an attempt to bring in a legitimately African element rather than Euro-American cliches of drum-centered African music.

The play begins with the song "The Hills of Ixopo". Rather than an overture, there is just a four bar intro, in which the brass instruments twice repeat a phrase with exactly the rhythm of Paton's original title, "Cry, the Beloved Country". The lyrics of the opening song are almost verbatim from Paton. This fidelity does not continue through the rest of the play, though, and the drift away from Paton doubtless became even greater as director Rouben Mamoulian 🔗 became so involved in rewrites that Foster Hirsch has characterized him as "the show's third author." Mamoulian insisted on adding at least some humor into the play, resulting in the addition of the satiric song "Big Mole" about diamond drilling. More importantly, he expanded the role of Stephen because he had a genuine star in Todd Duncan 🔗, who had been the original Porgy in the Gershwins' Porgy and Bess and who had more recently been the first African-American to sing with New York City Opera 🔗. Several passages that had originally been sung by the Greek chorus were modified to be first-person songs for Stephen. Also, a previously instrumental passage became a song "Thousands of Miles", allowing the audience a glimpse at an earlier, happier Stephen. All of this suggests that Mamoulian had the clout to address exactly the issues that Agnes de Mille 🔗 had cited in turning down and earlier offer to direct the piece: an "aura of unrelieved tragedy" and the "static" aspect of so much coming from chorus rather than characters.

Paton was not pleased with the changes to his story. Some of these were inevitable. Arthur Jarvis's quoted writings are, in Hinton's words, the "moral and political center" of Paton's novel, but the only traces of these in the play are a few short passages that make their way into the dialogue. Under Mamoulian's influence, the character of Stephen becomes much more central than he was in the novel, but also the moving of certain lines from the Greek chorus to Stephen gave Stephen thoughts and ideas that Paton considered completely out of character for this rural black South African preacher. Above all—and this seems to be Weill and Anderson, not Mamoulian—in the play, Stephen has a real loss of faith, whereas in the novel Stephen wrestles with existential questions, but settles for (Hinton again here) "faith and resignation." In the play, Paton saw "my humble hero in a role he could never have taken." Further, the play's James Jarvis undergoes more of a change than the novel's, resulting in a more optimistic ending. Anderson explicitly said that his reading of Paton was influenced by the ideas of Arnold Toynbee: "brotherhood, amity, tolerance, … understanding that crosses all boundaries." Not that Paton had anything against any of those values, but they add up to something more cheerful than is to be found in his novel. Paton portrays a tragedy; he does not propose a program.

The play opened October 30, 1949 and ran a reasonably respectable 281 performances. Brooks Atkinson of the New York Times gave it an overwhelmingly favorable review, though definitely one that might suggest a monument more than an entertainment: "the kind of theatre we all respect and admire has returned to Broadway." While regretting that the novel's sweeping narrative had to become "skeletonized" to fit the confines of a stage play, he felt that Weill's score had salvaged this situation, "at once the most popular and the most evocative score he has written." He was not alone in his positive view, though there was also some dissent. Howard Taubman, also of the Times in a broad piece on New York theater found the music "too commonplace to evoke the atmosphere of South Africa" and found the title song "embarrassing", "maudlin", and "torchy." (I'll agree on "torchy.") Another Times man, music critic Olin Downes liked the operatic aspects but didn't like the more "Broadway" aspects. In private correspondence, he wrote to Weill, "I am still waiting for the day when you get exactly the subject which you can treat without the faintest consideration of public taste or expediency of any sort." Weill, responding, rejected the concept of writing "without the faintest consideration of public taste," explaining how important it was to him that his work remain accessible to "a typical American audience." Also, compared to Downes, Weill saw the "American popular song" as a more valid artistic form in its own right.

There's a lot to tease out here about genre, about the artistic status of popular song, and about the adaptation of a serious novel into a musical. This play was one of several in which Weill was consciously experimenting toward creating "Broadway opera", a form of the Broadway musical that would have the depth and musical merit of classical opera while retaining a wide audience. Looking back, he didn't so much create a new genre as lay down a marker: Weill showed that the Broadway musical could be not just spectacle but art, and could tackle difficult subjects. There were a few who had insight into that at the time, or shortly thereafter. Virgil Thomson viewed Lost in the Stars as a latter-day Singspiel, an early form of German opera similar to the English ballad opera. Singspiel combined spoken dialogue with a variety of musical forms, ranging from street ballads to arias. Weill himself had played on the term in the title of his first work with Bertolt Brecht, the Mahagonny-Songspiel (1927). Stephen Hinton sees the concept of Singspiel as the unity among Weill's varied theatrical works, both European and American. Even a piece like Street Scene that obeys the classical unities 🔗 of time and place doesn't quite obey unity of action or of tone (the "Ice Cream Sextet" in the middle of a piece about jealousy and murder?). In Stephen Hinton's words, Weill always embraced "dramaturgical counterpoint" and "contamination," thriving on the resulting tension.

Some of this may be much easier to see now, in the 21st Century, where popular music has made its way into the academy. It was probably a lot harder to see in the mid-20th Century, when austere high modernism was at its peak in the academy and in critical circles. Prior to about 1980, it was almost a given for music critics to see Weill's American works as pandering to lowbrow American tastes; less so for theater critics, who could see how well they worked as part of a play. When you think of how dismissive highbrow critics were of jazz and rock until each was well past its respective peak as America's popular music, it's no surprise that Weill's attitude that "good light music is… more valuable than bad serious music." (1936) and "I write for today. I don't give a damn about writing for posterity," (1940) was not going to endear him to the likes of Theodor Adorno 🔗. Having now spent over a year myself living very closely with Weill's music, and especially with arranging some of his songs for the guitar, I can say pretty comfortably that his American music has as much complexity as his earlier theatrical music in Germany and France, even if some of it is written in American popular music genres. The high modernist critics—and also people like Harold Clurman 🔗 of the Group Theatre, who had worked with Weill on Johnny Johnson (1936), but did not like where Weill went afterward—put a high value on authorial voice and "personal language" and failed to find that in American popular song. To me, that says they weren't really inclined to listen.

Which brings us back to the question of adapting a serious novel into a musical play. Brooks Atkinson was doubtless correct in saying that this inevitably meant "skeletonizing" the work; the same usually happens when a serious novel is adapted as a film. The question, to my mind, isn't whether a musical play can "capture the essence" of the book: it probably can't. The question is whether the book can somehow provide a framework out of which to build a good piece of musical theater. Unlike Paton, Weill and Anderson were not really grappling with South African reality. They were dealing with human tragedy, striving for something like universalism, making an allegorical comment on race in America, and trying at least to respect South African cultures as they did so. Eleanor Roosevelt remarked on the play's omission of most of what was specific to South Africa: "perhaps reading the book, which gives a more complete picture of the problem, will make it difficult for some people to enjoy the play." Hinton asks whether the play's element of hope is simply naive, and whether its universalism amounts to sugar-coating, and somewhat answers himself saying that, much as the play steers "a middle course" on genre, it steers a middle course on "social reality" vs. "universalized human experience."

Upon completing Lost in the Stars, Weill and Anderson turned to the project of writing a musicalized Huckleberry Finn. They made some progress on the piece, including five songs, before Kurt Weill died entirely unexpectedly of a heart attack April 3, 1950, while Lost in the Stars was still running on Broadway. When its Broadway run completed in July 1950, the play toured to California and the Midwest; the tour ended after bad reviews in Chicago. New York City Opera revived it in 1958. It was performed a few times in Germany in the early 1960s, as Germans first started to become interested in Weill's American work. A 1972 Broadway revival with what Stephen Hinton describes as a "radically altered score" flopped, and ran for only 39 performances. There was a 1974 film; I've seen only excerpts, but from what I've seen, the less said the better.

Lost in the Stars continues to be staged periodically. There have been at least two limited-run revivals in New York: The Off-Broadway York Theatre Company in 1988 and the New York City Center Encores! series in 2011. The most recent performance I'm aware of was by the CAP UCLA, the Los Angeles Chamber Orchestra, and Saratoga International Theater Institute in 2017. From what I can see of reviews of any of these later productions, there seems to be consensus that Anderson's script is a bit dated, but Weill's music holds up well.

[This essay is obviously very indebted to Stephen Hinton's Weill's Musical Theater: Stages of Reform (University of California, 2012). I've also made use of a few related Wikipedia articles and of several articles in the New York Times, including Brooks Atkinson's Nov. 6, 1949 review.]