The song quoted above, with lyrics by the American playwright Maxwell Anderson and music by the German émigré composer Kurt Weill, was composed in 1939, as part of a larger collaboration they undertook shortly after the completion and successful production of their first Broadway musical, Knickerbocker Holiday (1938) (Hirsch 2002:177; Rabel 2007:550; Sanders 1980:337). For their new project, Anderson and Weill had in mind a ‘Negro’-play with music, for which the Gershwin brothers’ critically and commercially acclaimed opera Porgy and Bess would serve as a model for emulation. Extensive work related to the song ‘Lost in the Stars’ resulted in a repeatedly abandoned and returned-to large-scale project called Ulysses Africanus, with evidence of the resuscitation and abandonment of this planned work spread throughout the following decade.

Before presenting this evidence, it is important to mention that the substantial literary and musical thematic material of Ulysses Africanus developed over this period of time surfaced in a full-length ‘musical tragedy’ called Lost in the Stars, which premiered at the end of October 1949, and was based on Alan Paton’s best-selling novel Cry, the Beloved Country (published in 1948). So, even though Anderson and Weill ostensibly based their work Lost in the Stars on Paton’s Cry, the Beloved Country, the former turns out to be an adaptation of work that was initiated roughly a decade before the latter had even been written or published.

Although the anachronism of an adaptation that precedes the conception of its original might be viewed as problematic, it seems natural enough that Lost in the Stars should carry over some of the narrative and ideological baggage of its pre-Cry, the Beloved Country incarnations – Ulysses Africanus foremost, but also Eneas Africanus (the novella upon which Ulysses Africanus was based). What is extraordinary is that it should also draw on the concept for a spaceship musical called Lost in the Stars that Anderson and Weill attempted to realise in 1947 (Zychowicz 1994:85).

Even the most cursory glance at Western literature makes it evident that borrowing stories, and then retelling them in various ways, is habitual among cultural practitioners (Hutcheon 2006:4). In this article’s context of focussing on the Ulysses myth, one can point out, as the critic George Steiner (1997:23) has, that ‘“Odysseys” are legion from Homer to Joyce and Derek Walcott’. Therefore, there is nothing unusual about Anderson and Weill’s adaptation of Paton’s novel into an Ulyssean musical tragedy. What is perhaps unusual, though, is the way that they went about it – how their adaptation was initiated so long before the purported source novel of Lost in the Stars had been written. For they ultimately drew on multiple texts to inform their adaptation of Cry, the Beloved Country and could therefore not strictly remain ‘faithful’ to their ‘original’ Patonesque source material. But at the same time Lost in the Stars becomes precisely the kind of adaptation that Linda Hutcheon (2006:21–22) calls ‘palimpsestuously intertextual’.

The discourse on adaptations, usually referred to as adaptation theory, is concerned with works that have been reworked across different media. In particular, the issue of ‘fidelity’ rates among the most frequently discussed in the relevant discourse, so that pejorative descriptions of violation, perversion and the like often inform routine and inadequate responses to the adaptations of originary texts (Hutcheon 2006:2, 6–7, 20; Naremore 2000:8–9). Similarly, authors of original works are often unhappy with the adaptations of these works, Alan Paton being no exception in the instance of Lost in the Stars. Even though he was able to remind himself that the ‘making of a book and the making of a play were two separate creative acts’, Paton (1988:20) was nevertheless unnerved by an apparent change of idiom in the musical adaptation of his novel.

The authors of original works are, understandably, within their rights to evaluate the appropriation of their intellectual property with a strict eye, but they also sell their intellectual property to someone else, thus usually foregoing creative control over the resulting adaptations. One of the apparent conditions of modern adaptation theory is an understanding not to privilege adapted texts over their adaptations, on the mere supposition that adapted texts are pristine originals and adaptations derivative and inferior (Hutcheon 2006:xiii). This means that a ‘fidelity criticism’ that always harkens back to the idealised original work is insufficient for the critique of these interlinked texts. This prevailing line of reasoning in adaptation theory echoes the ‘challenges to dominant post-Romantic notions of originality, uniqueness, and autonomy’ present in the semiotic and post-structuralist theories of Barthes and Kristeva (Hutcheon 2006:21). Theoretical borrowings like Barthes’s ‘death of the author’ allow the foregrounding of intertextual readings where the original is not subscribed to as an ideal.

The overriding reason for rejecting ‘fidelity’ as the final arbiter in the criticism of adaptations is an acknowledgement that stories work differently in different media and that the gains and losses made in adaptations are a result of this phenomenon (Hutcheon 2006:16). The process of adaptation thus requires the elimination or changing of parts of the original to make stories viable and effective in their new medium. Anderson did this in Lost in the Stars by removing characters from the story and by inserting new episodes that clarify the consequent missing plot elements (Matlaw 1975:268). Dismissing his work outright for its lack of faithfulness to Cry, the Beloved Country would be unproductive, as the following analysis, by Myron Matlaw, of the relevant adaptation process suggests.

Comparing Lost in the Stars and Cry, the Beloved Country, Matlaw (1975:260–261) has argued for a careful consideration of their different genres, reminding readers of the ‘obvious but all-too-often forgotten generic distinctions’ between works of (printed) novels and (staged) plays. Evaluating the novel and musical play within the specific parameters of their distinct genres, and not simply dismissing Lost in the Stars as a ‘feeble replica of a powerful novel’, is the first step in appreciating the different ways in which both works strive for similar emotive effect (Matlaw 1975:262). Firstly, the two mediums have dissimilar creative processes. Cry, the Beloved Country has only one author, whereas Lost in the Stars was created collaboratively not only through the efforts of a writer and composer but also through the auxiliary efforts of the director, producers, actors, stage technicians, scenic designers, costume designers, make-up artists and orchestra (Matlaw 1975:262). Secondly, the two mediums engage their audiences in distinct ways. Reading a novel is an interior act customarily performed in solitude, at a pace and even in a sequence controlled by the reader, whereas watching a Broadway musical performance is an exterior act that takes place in a communal setting and proceeds at a pace and in a linear sequence determined by the show’s creators (Matlaw 1975:262). Thirdly, the mediums differ in length. Writing Cry, the Beloved Country in the form of a novel, Paton did not have to express its story in clear and concise language, while Anderson and Weill’s success with Lost in the Stars was largely contingent on the concise expression of its action and dialogue.

Any adaptor who would want to dramatise Cry, the Beloved Country would, from the outset, be confronted by the novel’s disjointed narrative. Its chapters are short and nearly always fragmented, and they do not necessarily follow in sequence. Many chapters are purely lyrical descriptions of social and political conditions in South Africa, didactic in their nature, with very little purpose in advancing the action of the plot. This fragmentation also exists on a macro-level, for Paton conceived of Cry, the Beloved Country in three broad sections: Book One, Book Two and Book Three. The timeframe of the first two books overlaps considerably, the first being written from the black protagonist’s perspective and the second from the white protagonist’s perspective. The third book deals with the story’s denouement, which plots the reconciliation of the black and white protagonist, over a rather lengthy 50 pages. Though by no means intractable, these narrative problems require planning and restructuring if they are to be translated effectively for the stage. Anderson decided to telescope the action of Book One and Book Two in his adaptation of Cry, the Beloved Country, alternating scenes from the two books in a continuous flow of action (which, incidentally, is the same technique used in the two film adaptations of Cry, the Beloved Country). To cope with the problems of the micro-narrative, specifically the descriptive chapters that are not integral to the action, but still contain some of the novel’s most moving prose passages, he devised a Greek chorus that would comment on the stage action (Avery 1977:221).

Paton’s stylised prose sets the emotional tone more than anything else in his book (Matlaw 1975:262–263). He softens many important action scenes in the story by only alluding to them. As an example, the violent murder at the centre of the story is conveyed in a newspaper article instead of being described directly. In contrast, theatre thrives on direct action. Things need to happen for the sake of rapid plot advancement. Allusions are too indirect. Instead, ‘speech and action, taking the place of the written word, must move more rapidly, simply, and clearly’ (Matlaw 1975:267).⁴ Moreover, the conventional format of Broadway musicals would have presented Anderson and Weill with a further set of limitations. The work could not be too operatic or too serious, lest it estranged its potential audience. Commercial viability necessitated the insertion of comedy, sexual appeal, romance and sentimentality into a novel that had none of these elements (Matlaw 1975:267). Only by inserting generic musical theatre numbers, that effectively broke the flow of Paton’s original story, could these moods be incorporated.

Ultimately, Matlaw’s analysis argues that Anderson ‘strove not only to dramatize Paton’s story but also to communicate Paton’s attitudes, to recreate the effects Paton had sought, and to evoke comparable responses’ (Matlaw 1975:267). While acknowledging Matlaw’s valuable insights, this article nevertheless argues the reverse: By superimposing a pre-existing Ulyssean framework onto Cry, the Beloved Country, Anderson and Weill specifically undermined Paton’s religious attitudes. Their use of Cry, the Beloved Country as the container for an Ulyssean story that they could not get off the ground is one of the means with which to criticise their efforts.

Maxwell Anderson had a life-long preoccupation with tragic drama, so much so that at a time when verse tragedies were completely out of fashion, he wrote plays such as Elizabeth the Queen, Anne of the Thousand Days and Mary of Scotland in verse. Anderson and Weill’s choice of tragedy as a dramatic vehicle for Lost in the Stars nevertheless is surprising, owing to the mention of a planned ‘lighthearted’ approach for Ulysses Africanus and the spaceship musical in the already mentioned discussions with Robeson and Wharton. Yet it makes conventional sense that an Ulyssean story should be tragic – it being one of the foremost stories of ancient Greek culture which, through countless adaptations, has become one of the pivotal narratives of Western civilisation. In Ulysses’s voyage, we find an allegory of the human need for homecoming through the traps and pitfalls of life; but this homecoming, made extremely difficult by Poseidon, also speaks of the irrationality of human suffering at the hands of the gods.

It makes further sense that Anderson fashioned Lost in the Stars as a musical tragedy when considering the poet Stephen Watson’s (1982:31) excoriating reading of Cry, the Beloved Country – which he characterised, using J.M. Coetzee’s interpretation of the novel, along with other unrelated criticism by George Steiner, as ‘[t]he predominant example of religious tragedy in South Africa’. Watson’s inclusion of Steiner in his classifying the work as a ‘religious tragedy’ evokes the question of God’s presence or absence.⁵ The following commentary on Anderson’s fetish for writing tragedies in which God abandons the world suggests that Paton’s novel was a natural choice for adaptation, yet the theme of divine abandonment that Anderson worked into his adaptation irked Paton enormously when he viewed Lost in the Stars.

Anderson’s interest in tragedy was in all likelihood connected to the semantics of abandonment highlighted earlier in this article, language usage central in much of his writing. In many of his plays, and in his literary criticism, he makes use of astrological imagery to express a trope of spiritual abandonment, usually indicated with specific images of ‘loss’ and ‘stars’. Thus, Anderson describes the human race as groping around helpless on their small planet without a Creator, who has run away. Many of Anderson’s most important characters experience anxiety because of their sense of divine abandonment, of God having absconded his role as caretaker of individuals and of humankind. And so we are all lost in the stars.⁶

In Act 2 scene 6 of Lost in the Stars, the priest Stephen Kumalo is uncertain about continuing his work in Ndotsheni as a preacher, out of shame that his son Absalom is a convicted murderer (Anderson 1951):

This dramatic crisis of conscience undermines the carefully planned surge of hope and salvation that Paton achieves in the third book of Cry, the Beloved Country. The corresponding scene in the novel portrays a God whose presence is immanent throughout creation: ‘And now for all the people of Africa, the beloved country. Nkosi Sikelel’ iAfrika, God save Africa’ (Paton 1949:252). In the novel, Kumalo has long overcome his uncertainties about his station, his parish and his faith. The only remaining uncertainty he has is in the very human condition of not knowing the future:

Cry, the Beloved Country was intended as ‘a story of comfort in desolation’, a description that is the novel’s subtitle (Paton 1949:3). ‘It is quite true that Stephen Kumalo, when he learned that his son had killed a man, suffered a sense of God’s desolation’, Paton (1988:20–21) admitted. But it is not the overall theme of the book, whereas it definitely is the overall theme of the musical adaptation. The ultimate religious message contained in Cry, the Beloved Country portrays a Christian God who not only cares for the world transcendentally, but who is also immanently present in people’s lives (Foley 2009:58). The very idea of a ‘Shepherd’ abandoning his flock flies in the face of Paton’s Anglo-Christian worldview and of the ever-present literal trust in the Christian God that pervades Cry, the Beloved Country.

Anderson used the idea of a broken promise in the biblical garden of Eden and turned the doctrine of original sin on its head: God was to blame for breaking an original pledge, and God chose to leave paradise and go into exile. As in The Brothers Karamazov, God is held accountable for the faults in creation. According to Anderson, the original contract was made when, supposedly at the time of creation, the earth, along with the stars, ran through God’s fingers and fell off to one side, which prompted him to hunt ‘through the wide night air/For the little dark star on the wind down there’ (Anderson 1951:60). It was then, when God found this little star, that he promised ‘to take special care/So it wouldn’t get lost again’. Stephen Kumalo sings about the breach of this contract in the title song ‘Lost in the Stars’: ‘And sometimes it seems maybe God’s gone away. Forgetting the promise that we heard him say –’ (Anderson 1951:60).

Atheists deny the existence of God. Agnostics question it. Anderson, in his turn, seems convinced that there is a God, but he is just not sure if that God is entirely interested in human affairs. Anderson’s God is a voyager who has lost interest in humanity. He is Ulysses, gone off in search of new adventures while humanity sits waiting for Godot. This idea of divine apathy is unusual but not unique. Matthew Arnold heard a melancholy, long, withdrawing roar at ‘Dover Beach’ in 1867, the sea of faith retreating, ‘to the breath/Of the night-wind, down the vast edges drear/And naked shingles of the world’ (Arnold, in Simpson 1972:221). Nietzsche, though universally accredited for declaring the death of God, also transmitted the idea of a God who had run away. In Also sprach Zarathustra, God is implored to return (Armstrong 1993:420). The idea is also intrinsic to the Old Testament book of Job, echoes of which were felt in the Shoah (Wiesel 1960:73). Even the Son of God sobbed at being forsaken on the cross.

Anderson’s religious abandonment anxieties probably date from a traumatic experience in his childhood when he, alone out of his immediate family, was left behind on his grandmother’s farm in a one-horse town called Atlantic, Pennsylvania.⁷ Gilda Anderson, Maxwell’s third wife, told his biographer that ‘from this incident date[d] a certain dreadful, recurring dream of his lasting well into adulthood, in which he always figured as a child being abandoned’ (Shivers 1983:20). Anderson’s biographer has traced the origin of two works by Anderson to this traumatic period in his life. The first is a novel, published pseudonymously, called Morning, Winter and Night, and the second is a play called The Star-Wagon.⁸ Anderson’s daughter, Hesper, has also described the enormous impact this abandonment had on her father and has furthermore intimated that it bears a connection with his play The Star-Wagon (Anderson 2000:175). The present article, however, suggests that it also bears a connection to Anderson’s repetitive reference to ‘loss’ and ‘stars’ throughout his works, and most especially in Lost in the Stars.

Anderson’s Ulyssean homecoming, through Cry, the Beloved Country, was already confirmed in his first reading of the novel. In chapter 7 of Book Two, the character James Jarvis, confronted with the murder of his son Arthur, reads one of Arthur’s political writings on South African race relations, called ‘Private Essay on the Evolution of a South African’. In the essay Arthur, like many awakening South African liberals, had criticised his parents for teaching him nothing of the realities of South Africa. James Jarvis reads that, in spite of this, his son had chosen to devote all of his time, energy and talents to the betterment of South African society:

Anderson would have noticed this navigation by stars. Even more, one sentence later in the essay, Arthur explains his resolve to never ignore morality for expedience:

Like Maxwell Anderson, the character Arthur Jarvis is ‘lost in the stars’. Paton had no way of knowing that Anderson would interpret his novel according to an existing paradigm that contrasted so strongly with his own belief in the immanent presence of the Christian God in worldly affairs. Thus, he missed a warning in the very first letter that Anderson wrote to him, which only becomes clear in retrospect, of where Anderson wanted to steer Paton’s novel:

Anderson and Weill did not understand Paton’s specifically South African ‘tribe’ and thus chose to turn it into a spiritual metaphor of the breaking of all tribes at the hands of God. Their interpretation of Cry, the Beloved Country’s universality was not unique. The feverish reception that the novel had in New York, where it was first published, is evidence already of a universalised reception history. But critical readings of the novel have also stressed this point. Edward Callan, Paton’s first literary assessor, has argued that the novel achieves universal value ‘by incorporating into the actualities of South Africa’s physical and social setting a fundamental theme of social disintegration and moral restoration’ (Callan 1968:49). This theme, he concluded, was ‘worked out through two complementary, or counterpointed, actions: Stephen Kumalo’s physical search for his son Absalom, and James Jarvis’s intellectual search for the spirit of his son Arthur’. These two Ulyssean narratives are the catalysts for the story and also serve as inspiration for its adaptation into Lost in the Stars. Like Tennyson and Whitman’s Ulyssean voyagers, both the protagonists in Cry, the Beloved Country sail forth to seek and find, and like Ulysses their course eventually leads them back home, carrying the burdens of material loss and moral gain.

In any dramatic adaptation of Cry, the Beloved Country, aspects of the original novel would inevitably become ‘lost in translation’. This is certainly true of Lost in the Stars. Paton’s voice had to give way to another person’s language and another person’s music (Matlaw 1975:272). Yet Paton allowed Cry, the Beloved Country to be adapted many times – into a Broadway musical tragedy by Anderson and Weill, into a feature film by the Korda brothers and into a play by Felicia Komai (Callan 1968:22). The feature film adaptation was made in more or less the same timeframe as the musical tragedy, and after seeing it (in 1951) Paton was satisfied that it was a more or less ‘faithful’ rendition of his novel (Alexander 1995:268). He, therefore, had no objection to Felicia Komai’s request, in 1954, to adapt Cry, the Beloved Country into a verse drama; verse drama was the kind of literature that Anderson specialised in.

Paton loathed Lost in the Stars because it turned out that he and Anderson subscribed to mutually exclusive spiritual paradigms. Paton attended the world premiere performance of Lost in the Stars – with his then mistress, the anti-apartheid activist Mary Benson (1989:54) – and did his best ‘to join in the rejoicing, but [his] heart wasn’t in it’ (Paton 1988).⁹

Paton never believed that a Christian should be free from suffering, because Christ also suffered. A line from Cry, the Beloved Country is illustrative. ‘And I come to believe that he suffered, not to save us from suffering, but to teach us how to bear suffering, for he knew that there is no life without suffering’ (Paton 1949:208). Anderson, in his turn, understood human suffering as the breach of an original contract between man and God. But, as Steiner (1997:160) says of a similar device in The Brothers Karamazov: ‘there is, to be sure, not a shred of evidence that any such ticket [promise] was issued to man in the first place’. When Stephen Kumalo sings in ‘Lost in the Stars’ that ‘God stated and promised he’d take special care’, it is purely literary assumption on Anderson’s part, written in reaction to the fire and brimstone environment of his own childhood. Paton thought that the song ‘Lost in the Stars’ was far too sophisticated to be sung by his humble protagonist. It also pained him that this song ‘belonged to the death-of-God genre, or to put it more accurately, to the desertion-of-God genre’ (Paton 1988:20).

Paton was not a purist. If he were, he would not have allowed any subsequent adaptations of his work. Apart from his religious objections, he seems to have had no further problems with Lost in the Stars, and even commented on the beauty of its music. Because of Anderson and Weill’s efforts, we can now read Paton’s original novel with a Ulyssean subtext that was not articulated as strongly in the original. The adaptation loses much of Paton’s prose, but it also contains the last music that Weill composed before his death. The argument is to be made that, with an informed and creative director, the characters of Lost in the Stars could come alive much more vividly on a stage than they ever would in the interior of a reader’s mind. This, of course, depends on personal tastes and on the aesthetics of reading as opposed to those of watching.

Having seen Lost in the Stars in a live production at Cape Town’s Artscape Theatre, late in November 2011, I was struck by the two-dimensional melodramatic depiction of black identity and black suffering that was realised out of Anderson and Weill’s blueprints (score and libretto) for the work. In the light of Essie Robeson’s warning to Maxwell Anderson in 1939, when he set out creating Ulysses Africanus, Uncle Tom, Aunt Jemima, Ol’ Mammy and Jack Johnson were all on stage that night. Nevertheless, the audience reception was very enthusiastic, about half of the audience giving a standing ovation at the end of the performance.

That these negative racial stereotypes could still be so enduring and be transmitted from America’s Civil War era into South Africa’s rainbow nation is a negative testament to the persistence of adaptation. Rather than criticise Lost in the Stars as an ‘unfaithful’ adaptation of Cry, the Beloved Country, one could critique its perpetuation of these stereotypes. The fact that Lost in the Stars began as Ulysses Africanus, which was itself an adaptation of a racist-toned novella set in the American Civil War era, is highly suspect. That such a work could find a performance opportunity in the new South African cultural scene, and in Desmond Tutu’s rainbow nation, is bizarre. It indicates a superficial engagement with and celebration of South Africa’s struggle history. There should be no solace in peopling a stage with black people while leaving racist stereotypes intact as they are adapted interculturally. Had Maxwell Anderson’s Ulysses not got lost in Alan Paton’s beloved country, or more properly been forced into that literary landscape, the operatic or musical tragedy version of Cry, the Beloved Country might present the narrative of black suffering and protest in a less clichéd manner.