Deep red sun climbing from a still sea, the wet shingle ablaze with reflections. Walked into the garden floating on a sea of pearls. The garden casts mysterious shadows. Not a breath of wind.

In Derek Jarman's The Garden (1990) two nearby but apparently unrelated images help indicate the genre of this film. Jarman has a name for these images, which were improvised on the first day of shooting: "emblemata." These are pictures that invite you to read, to interpret. First image: a man in kingly costume looks directly into the camera; he returns in a few moments to the center of the screen and bites deliberately into an apple. We note that the king holds a long feather instead of a mace. A minute later in the montage, a second image: three boys pillow-fighting on a bed, the feathers flying up in the air. By rhyming twice on feathers, these images lead us toward the "tar and feathering" scene, in which the gay couple at the center of the film is tormented and mocked in a restaurant. These feathers, of the king and of the pillows, aim us toward that scene by preparing us with images rather than verbal logic. So much is perfectly clear. But these sets of images, although brief, contain more than little bridges of cinematic continuity, are more than a mere motif or riff on feathers. These images make pointed allusions to a particular history of film, a history that we need to notice and take seriously. Until now, however, this history has not been written in any very deliberate manner. It will be a main goal of this book to elucidate that history—the history of what I call lyric film.

To what, then, do the images allude? The king, frontally posed and later biting into a fruit, is strongly reminiscent of the poet-troubadour in Sergei Paradjanov's The Color of Pomegranates. The pillow-fighting recalls "lyric" moments in great French impressionist cinema, namely scenes in Abel Gance's Napoleon (1927) and Jean Vigo's Zero de conduite (1933). In other words, Jarman self-consciously aligns himself with a tradition of lyric film. The video box of The Garden, produced by Fox Lorber, is full of critical quotations that recognize the film's generic provenance: "A lavish film-poem about the mortification of the flesh!" (Village Voice). These video box quotations—"Derek Jarman's lyrical and visionary movie" (Chicago Reader)—serve not only to celebrate and to entice, but also to warn the prospective video audience that something rather nonnarrative awaits them. Although Jarman's lyricism in film after film is, undoubtedly, quite evident, we also realize that the word "lyrical" is a loosely and commonly used term of critical approbation. We might well say "poetic" when we mean no more than "beautiful"; such a comment may have nothing to do with poetry or genre or structure. But Jarman's films need to be understood in a much more specific and concrete way, insofar as they are related to poetry and poetic films.

Jarman's antipathy toward narrative and narrative cinema is repeatedly expressed with great clarity in his published journals and scripts. As he writes in the War Requiem text, "I've had to move carefully, given my reputation as an enemy of narrative film, since the recording company obviously wants a conventional narrative." On the last page of his script for Caravaggio, he maintains that the structure of that film is, for him, unique: "This film is the first in which I have developed acting parts and bowed to narrative." Jarman's films are intentionally set down against narrative, therefore, but in that intention he is by no means alone. Both European and non-Western cinema are usually less narrative than Hollywood film—the Hollywood of "classical Hollywood narrative." And avant-garde cinema is, broadly speaking, less narrative still.

What is individual about Jarman's conception of film, and central to the tradition to which he belongs, is the degree to which he thinks of this contrast to narrative as poetry. The word "poetry," out of any context, is a remarkably vague and loaded term, but it is used by workers in the field of poetic cinema, I would argue, with alert knowingness and a sharp sense of history. Directors such as Jarman, Andrei Tarkovsky, and Pier Paolo Pasolini, as we shall see, do not just talk about poetry in their polemical tracts and journal entries—and they talk about poetry constantly—they include recognizable poems and poet-figures in their films. These directors repeatedly encourage us to read and interpret their films not as we would read a novel, but as we would read a poem.

Ludwig Wittgenstein—and Jarman's Wittgenstein as well—speaks of "language games": we need to identify what language game we are playing in order to communicate. "Do not forget that a poem," writes Wittgenstein, "although it is composed in the language of information, is not used in the language game of giving information." If cinema is a kind of language, then we need to identify the kind of language that Jarman uses; we need to identify the kind of language game that we find ourselves in when we find ourselves next to words like "poetry." By itself, "poetry" has little meaning. Even as a contrast to "narrative," it is still rather empty, too abstract. But when used by a community of artists, the word begins to take hold, and we can come to grasp it in its relation to other expressive practices and discourses.

Not that all artists, of course, will understand the same thing by "poetry"! Artists do not all speak the same language, any more so than film critics or talking heads on television. In 1953, for instance, there was a famous roundtable discussion on "Poetry and the Film" in which Maya Deren described the contrast between poetic film and conventional film as the contrast between vertical and horizontal, but fellow symposium participants Arthur Miller and Dylan Thomas refused to understand her distinction. Deren felt severely put upon: "I mainly wish to say that I'm a little flabbergasted at the fact that people who have handled words with such dexterity as Mr. Thomas and Mr. Miller and Mr. [Parker] Tyler, should have difficulty with such a simple idea as the 'vertical' and the 'horizontal.'" Deren is one of many film artists to have spoken of poetry and poetic structure in cinema less rigorously, perhaps, than literary theorists, yet more self-consciously and deliberately than other artists. Whereas Miller and Thomas claimed not to know quite what she was talking about, directors like Tarkovsky and Jarman surely would have; they would have been able to play her language game.

Like Tarkovsky, indeed, Jarman not only writes extensively and self-consciously about his cinema, he also writes often of poetry. In the context of his films, what does Jarman mean by poetry? In addition to his films, Jarman was well known for his painting, and sometimes regarded himself, indeed, primarily as a painter. Yet he also published a book of poems early in his career, which we shall study in the next chapter; and all of his journals and books contain numerous poems by himself and others. So poetry in part means poetry in lines, poetry recognizable as such. As we shall see, Jarman's interest in Benjamin Britten's War Requiem was due more to the poems by Wilfred Owen than to the music by Britten. Poetry for Jarman is certainly related to a literary sense of the term, the sense in which William Blake (quoted in Imagining October) and John Keats (whose collected works we see next to Owen's body in War Requiem) are poets.

Poetry for Jarman also does imply a contrast to narrative, similar to Deren's neoformalist contrast of "vertical" poetry to "horizontal" narrative. Jarman's poetic impulses in this regard are manifest and intentional, as noted above, and are well attended to by critical audiences of Jarman's films; see, for example, Tracy Biga's essay "The Principle of Non-Narration in the Films of Derek Jarman." When Jarman says that "The Last of England works with image and sound, a language which is nearer to poetry than to prose," he means that the film's images speak figuratively rather than declaratively and that the film proceeds by association rather than through cause and effect. Jarman uses "poetry" as an oppositional term to contrast the structure and shape of his films with that of linear, more regularly narrative kinds of cinema.

Yet this contrast of poetry to narrative may well seem tenuous, not just to disruption-prone poets like Dylan Thomas, but to more philosophically minded literary theorists as well. In one of the most searching critiques of Jarman's work, Leo Bersani and Ulysse Dutoit ask "Can film be non-narrative?" and question Jarman's own understanding of the category of narrative. Their sophisticated analysis of Jarman in a recent volume on Jarman's Caravaggio will receive a more detailed discussion in later chapters. But their question—a theoretical question—can be answered now, in part, by providing a history of lyric film.

Thus we might first identify their question as an academic question, academic insofar as it does not wish to play the language game in which Jarman places himself. In the 1960s, Christian Metz asked exactly the same question when he wondered whether Pasolini, in his essay "The Cinema of Poetry," understood what narrative was. In his own essay, Metz asserts that a nonnarrative or poetic cinema is a theoretical impossibility. Pasolini, it may be said, explicitly invited such theorists as Metz and Umberto Eco to argue with him in theoretical terms by setting out his essays in the language of academic structuralism. Jarman's loose, artistic language of poetry refers not to semiotics, but, I would argue, to the history of lyric cinema. Can we not attempt to join him in that field? "Nonnarrative" in Jarman's use points not to the works of Alain Resnais or Jean-Luc Godard, where Bersani and Dutoit turn in their discussion of Jarman, but rather to those of Jean Cocteau, Jean Genet, Pasolini, and Tarkovsky. To understand poetry in Jarman's films requires familiarizing oneself with the films and problems in this tradition.

And what, or where, is this tradition? Although the history of cinema is brief compared to that of poetry or architecture, its genres and traditions are of considerable complexity and are still being worked out. Tradition itself as a category or term in the history of art is not unproblematic. Yet in a pragmatic sense we can still speak of tradition and traditions; we can speak this way insofar as the term is critically useful. Jarman, as even he conceived of himself at times, might well be seen as a traditional artist, and he was always self-conscious of his place in history and of his use of history.13 Lyric film is not a genre in the way that western or film noir is, it is true; but it does name a set of affiliations and issues that reach outward toward Jarman.

A few, infrequent films are overtly interested in poets: The Barretts of Wimpole Street (Sidney Franklin, 1934), based on a play about the courtship of Elizabeth Barrett and Robert Browning; Total Eclipse (Agnieszka Holland, 1996), with Leonardo DiCaprio as Rimbaud; Il Postino (Michael Radford, 1995), featuring Pablo Neruda in exile in Italy; and, of course, Dead Poets Society (Peter Weir, 1989), in which Robin Williams promotes poetry as a way of life. These films have poets in them, but the poets are not represented as being much different from other kinds of artists or exceptional persons; a film like Dead Poets Society sets the illusion of liberating creativity among the hoariest confines of Hollywood convention and plot.

Much closer to the sensibility in which we are interested are works inspired by poetry, like Wim Wenders's Wings of Desire (1987), which he says is founded on Rainer Maria Rilke's Duino Elegies, and Atom Egoyan's The Sweet Hereafter (1997), which cuts Russell Banks's novel out of chronological order and powerfully adds Robert Browning's "The Pied Piper of Hamelin." Yet neither Wenders nor Egoyan works with poetry repeatedly; these are relatively singular manifestations of poetry in film.

And so it is a different story I will tell here, one more central to the case of Jarman's cinema. The directors that are most relevant here work repeatedly, even obsessively, with poems, poets, and lyric structure. In "The Cinema of Poetry" (1965), Pasolini argues that the films of Michelangelo Antonioni and Godard are more poetic than silent film. Yet each new generation organizes history in different ways, according to different landmarks. To my mind, Jarman's cinematic practice shows that lyric cinema has been around since the very beginning. (This was also part of Christian Metz's argument with Pasolini.) In this chapter I will examine some of the key practitioners of lyric film and identify some of the recurrent elements of lyric film. Lyric film is a splendidly arguable and movable genre, and the story I tell here would no doubt look rather different if I were not aiming toward Derek Jarman. In this telling, the tradition of lyric cinema goes back at least to D. W. Griffith, and so it is there, now, that we need to turn.

I. D. W. Griffith and the Creation of Another World

This is, I realize, mainly subjective; but it suggests to me the clearest and deepest aspect of Griffith's genius: he was a great primitive poet, a man capable, as only great and primitive artists can be, of intuitively perceiving and perfecting the tremendous magical images that underlie the memory and imagination of entire peoples.

James Agee

Pasolini and Jarman, I suggest, take part in a cinematic tradition that emerges at the very origin of film history, a tradition that meditates deliberately on "poetry" in cinema. Surprisingly, this impulse exists even in the works of the "inventor" of narrative cinema, D. W. Griffith. Although Pasolini and Jarman might seem at first glance to work some light-years from this forefather, this late Victorian, as it were, the attraction to social outcasts and the creation of other worlds form a nexus where poetry and film join in the works of all three directors. As I hope to show, there are relatively clear links between a film like Griffith's Broken Blossoms and later works such as Pasolini's Teorema and Jarman's The Garden. Each director invokes poetry in the name of social responsibility and social transformation. I would claim, furthermore, that Griffith's deployment of poetry is much more self-conscious than has usually been assumed and that Griffith deserves an originating role in the tradition of film poetry that this book seeks to describe.

Although Tom Gunning has shown that a cinema of attractions (the films of Georges Méliès, spectacle, pantomime, magic) offsets a cinema of narration at the origin of film history, narrative quickly becomes the major form of cinematic expression in America. Griffith, at the influential center of narrative form, adapted story after story and novel after novel for his pictures; Gunning lists 1909 Biograph versions of George Eliot's Silas Marner, James Fenimore Cooper's The Last of the Mohicans, and Charles Dickens's The Cricket on the Hearth. Griffith's films look much more like narrative stories than lyrical poems, certainly, but they do not exclude poetry out of hand. On the contrary, Griffith also filmed narrative poems such as Pippa Passes (1909, based on a long poem by Robert Browning), The Golden Supper (1910, based on an Alfred, Lord Tennyson poem), Enoch Arden (1911, Tennyson again) and The Sands of Dee (1912, based on a poem by Charles Kingsley). A title card for The Avenging Conscience (1914) says that the film is "suggested by Edgar Allan Poe's story of 'the Telltale Heart' [sic] and by certain of his poems of the affections." At the beginning of Judith of Bethulia (1913), a card tells us that the film is based on the "poetical tragedy" by Thomas Bailey Aldrich. Poetry is not just one interest among many or a one-time concern; it is arguably at the heart of his cinema throughout the entirety of his career.

Poetry was always for Griffith a high art, a status that more and more he sought for his films. It is no coincidence that the first film review ever published in the New York Times was for Griffith's version of Pippa Passes. Despite, therefore, Griffith's propensity for narrative drive and novelistic adaptation, poetry also shares in his artistic and cultural distinction. The poetry quotation, or the prestige-bearing epigraph, was a key part of Griffith's ambitions for film d'art. The poetic quotation is a carryover from Victorian fiction, whereby a novel like George Eliot's Middlemarch asserts its own high status—despite its "provincial" subject matter and prose form—by providing, at the beginning of chapters, epigraphs from such cultural luminaries as Shakespeare, Cervantes, and Milton. The use of inserted poetry quotations to signal cultural status—a practice that is by no means simple or transparent—continues in Pasolini's first film, Accattone, where he provides a preliminary citation from Dante as epigraphic commentary on the neorealistic content of this film.

Yet poetry is more than just another high art, a noble fiction with a compelling plot; it also appears in relatively complicated ways with respect to narrative. The contrast between poetry and narrative may be open to dispute by theorists, but major artists will continually attempt to work out, cinematically, elements of this vexed generic contrast. In works like Intolerance (1916) and Broken Blossoms (1920), which we shall examine in some detail now, Griffith seems to exhibit the contrast between poetry and narrative that will later be thought through by Deren and Jarman and disputed by Metz. Griffith is not as deliberate as Deren or Pasolini, but I do see his practice both as more self-conscious than it is usually credited with being and as bearing important affinities with the aesthetic and political practice of both Pasolini and Jarman.

Let us begin by noticing how Griffith deploys poetry in his masterpiece, Intolerance. Intolerance is certainly intended as both high art and spectacular film; it competes with European spectacles such as Cabiria (1914) and concludes by rising not only into breathtaking montage, but also, at last, into metaphysical allegory. Toward the end, to help underscore the plight of prisoners, Griffith cites lines from Oscar Wilde's " The Ballad of Reading Gaol":

And wondered if each one of us
Would end the self-same way,
For none can tell to what red Hell
His sightless soul may stray.

The ambitious, four-part structure of Intolerance (Babylon, the story of Christ, sixteenth-century Huguenots, and a modern-day setting) shows Griffith's clear pretensions to visual and literary prestige, and the Wilde quote can only further substantiate the artistic dimensions of the film.

As the poet and early film critic Vachel Lindsay helps us to see, the cultural prestige to which Griffith most obviously aspired was that of Walt Whitman. Lillian Gish's observation that Griffith was never without his Leaves of Grass has often been noted. Miriam Hansen has given the most recent and convincing explanation of Whitman's relationship to the structure and ambitions of Intolerance. For our purposes, let us note that the four narratives of the film are linked by the image of a mother rocking a cradle, over which Whitman's words "out of the cradle endlessly rocking" sometimes appear. The Wilde quotation cited above is inspirational and somewhat schoolbookish, but the Whitman quotation is central to the overall structure of the film. The Whitman phrase stands for repetition and rhythm, just as the film keeps returning to the mother and her cradle. Poetry thus stands in between the narratives, in an abstract, timeless space of sheer repetition. The Whitman phrase is thus clearly more than a nice quote or an amiable snatch of literature. It is a phrase that comments self-consciously on focus and pace; the lines and accompanying image form the point from which the centrifugal forces of the four narratives spiral out. The image pauses narrative in a generalized, motionless space with as much deliberation, contrast, and clarity as a diagram by Roman Jakobson.

This pausing and rhythm is even more characteristic of Broken Blossoms and speaks further to Griffith's intentional use of self-descriptive poetic language. In Griffith: First Artist of the Movies, Martin Williams calls Broken Blossoms a "film poem," and this characterization again has substance, is more than mere appreciation. Like the scenes in Intolerance with Whitman's lines imposed over the intimate, interior spaces of mother and cradle, this film is noticeably quiet, unrushed, and closed-in. From the brutal world of the violent boxer, Battling Burrows (Donald Crisp), Lucy (Lillian Gish) escapes to the loving and glowing world of the Yellow Man (Richard Barthelmess). The ethereal soft focus of these scenes, in fact, proved a striking technical innovation at the time and strongly influenced soft-style photography in later films. "Neither before nor later," writes Richard Schickel, "did Griffith do such a fully realized mood piece." Schickel calls Broken Blossoms a "transitional film, not for him, but for movies, as a whole . . . the first memorable European film made by an American." Martin Williams and Richard Schickel, two of Griffith's finest critics, well recognize the intimate, lingering poetic difference between Broken Blossoms and Griffith's more rambunctiously narrative or spectacular films, but they do not press this identification far enough.

Broken Blossoms is an early "film poem," and Griffith seems to know that it is. Yet Schickel, although finding some element of future artistic talent in a conventional early poem like Griffith's "The Wild Duck" (published in 1907, and reproduced in full by Schickel), cannot abide the exuberantly poetic intertitles of Broken Blossoms.

And then there is the matter of the titles. Even the most passionate of Broken Blossoms defenders are uncomfortable with them. "Oh, lily flowers and plum blossoms! Oh, silver streams and dim-starred skies!" cries the heart of the Yellow Man as he bathes the face of his poor tortured love. And that says nothing about "all the tears of the ages" rushing over his heart when she dies.

Compared to the grace and beauty of the pictures, the language of the intertitles does seem remarkably cloying, even ignorant of its own literary effect. But I think that we might cultivate a more forgiving attitude toward these titles by considering that in these poetic outbursts Griffith is signaling the generic form of his film in a way consonant with later critical identifications (i.e., a "film poem").

Note how deliberately and carefully Griffith adapts and highlights his source. In the Thomas Burke story from Limehouse Nights that provides the origin for Broken Blossoms, the Yellow Man is, although unconsciously, a poet. The poetic outbursts in the film come almost directly from the story. In the film, it is implied that these rapturous poetic lines are from the Yellow Man's point of view, an implication aided by the little Oriental paintings on the backgrounds of the intertitle cards. These poems therefore are not really omniscient Griffith raptures, but rather outpourings from the mysterious and enthusiastic spirit of the Yellow Man. We are told that the Yellow Man's actions are conducted "with perhaps a whiff of the lilied pipe still in his brain." Such an editorial comment deidealizes the source of the poetry and allows, in fact, a skeptical response. Griffith uses the Yellow Man's enthusiasms to characterize the less-than-relentless narrative of his film. The intertitles name the language as poetry by self-consciously speaking of the "lyrical moon" and by calling Lucy "a poem" (both phrases, incidentally, from Burke). Griffith uses the intertitles, along with the soft focus and the relatively actionless narrative, to identify Broken Blossoms as a film poem. The flower is a conventional poetic image, and Broken Blossoms repeatedly uses the image of the flower to develop its own atmosphere of lyric beauty and tragedy. Thus, though we may agree that the intertitle poetry is, without a doubt, conventional, it needs also to be noticed to what degree the use of poetry is both self-referential (the poems name the film) and arguably deidealized (as poetry is located in the opium-smoking Yellow Man).

Griffith's poetry in Broken Blossoms may be defended, then, for its self-consciousness, but it is considerably more difficult to defend his use or appropriation of the Yellow Man as an origin for that poetry. Griffith's reliance on a stereotypically effeminate Yellow Man as a source of poetry and beauty is an important issue, one related to how Pasolini will later be drawn to India and Africa for sources of poetry offering alternatives to capitalist Italy. The glowing, intimate world of Lucy and the Yellow Man is intended by Griffith precisely as a beautiful alternative to the violent, patriarchal world of Battling Burrows. The appearance of poetry in that context becomes a political gesture, although its politics may well resonate differently with audiences now.

The story deploys stock melodramatic figures, and both the passive, innocent girl and the exotic Chinese man are stereotypes that remain difficult to watch, difficult to accept. In a severe but important reading of Broken Blossoms, Gina Marchetti finds that the "beauty and poetry" of the film serve only "to mask its perverse roots" in the realm of "pedophilic fantasy." In this view, the Yellow Man gazing down at Lucy in his bed is caught up not in a dream of poetry, but a dream of child molestation. I think that Marchetti overreads the lust emanating from the Yellow Man, but it would be impossible to contend that the scenes of Lucy in Cheng's bedroom are devoid of erotic drama.

Here, defending Griffith against his use of stereotypes is probably a vain effort. Instead, we might at least see Griffith's poetry and politics as part of a cinematic tradition that attempts to create another world of poetry and cinema to hold against a vision of capitalism and machinery. Pasolini was accused of stereotyping the third world, and he had seemingly much less excuse. Both Griffith and Pasolini, unlike as they are, seek alternatives to the brutalities of industrial capitalism, and in works like Broken Blossoms and Appunti per un Orestiade Africana (Notes for an African Orestes) problematically represent that alternative in the form of an exotic other. Griffith more or less admitted the difficulties of such a representation by concluding Broken Blossoms with the deaths of all three main characters.

In the end, it may come to seem that there is no acceptable alternative to industrialization, hateful as it is; thus, the prevalence of emptiness and desert in both Antonioni and Pasolini and the recurrent ruined landscapes in Jarman. It is only by fatal accident that the careers of Pasolini and Jarman ended where they did, but there is still something of utmost importance about what each director's last film implies about the possibilities of alternative worlds and representation. Pasolini concluded with Salò, one of the most gruesomely unwatchable films ever made, whereas Jarman concluded with Blue, an unwavering screen of blue. These two films are wholly different from each other in effect, but are similar exhibitions of a deep hopelessness with regard to the possibilities of visualizing another world.

II. Soviet Poetic Film from Dovzhenko to Tarkovsky

In "The Modern Cinema and Narrativity" (1966), Christian Metz repeatedly questions critics and theorists who perceive a decline in narrative in modern films, represented by the work of such directors as Resnais, Godard, Antonioni. With a deflating skepticism that we have come to associate with Michel Foucault, Metz asserts that these various descriptions of and claims for the "breakdown of narrativity" are at their foundation part of a "great libertarian myth" (p. 186); that is, the supposed "freedom" of antinarrative is gained only by categorically mistaking what narrative actually is. Metz accuses the mythologizers and critics not only of poor semiology—that is, not knowing what narrative, and thus antinarrative, is—but also of poor cinema history. Thus, Metz questions, for example, the association of a "new cinema" with "poetic cinema" in Pasolini's "A Cinema of Poetry" by querying Pasolini's semiotic terms and, more vigorously, his sense of cinema history. Is it not the case, argues Metz, that Pasolini and his contemporaries have it all the wrong way around? Hence, instead of imagining a history in which contemporary films break down or disrupt an inherited tradition of narrative, is it not the case that film today seems even closer to fiction and that previous eras were in fact much more open to poetry than we are now? Wondering about moments of lyricism in film, Metz recalls Abel Gance's Napoleon and La Roue, the slow motion in the dormitory scene in Vigo's Zero de conduite, the "accelerated filming in the scene with the black coach in Nosferatu, and the incredible aerial travelling shot in the beginning of Murnau's Faustus" (p. 206). In Metz's semiotic view, a film "cannot be a poem"; in Metz's historical view, "the period in which one believed that a film could be a poem is that of the old cinema rather than that of the new cinema" (p. 206).

As Metz suggests, the silent film in many ways seems much more redolent of poetry and lyricism than the films of Resnais and Godard. The examples that Metz collects come mostly from German expressionism and French impressionism. The periodical criticism associated with French impressionism, by writers such as Léon Moussinac and Jean Epstein, often treats cinema in poetic terms and categories; these writers, as Metz helps us to recall, did in fact believe that a film could be a poem. Double exposures and dissolves, along with the use of curious phrases and calligraphy in the intertitles, give rise to many lyrical effects in early French films, some of which even feature poems and poets. Similarly, as German expressionism sought to represent the complexities of internal psychology and expression through distortions of exterior physical reality, early German films, above all the work of the masterful Murnau, often look far more poetic than prosaic or realistic. Tom Gunning and Miriam Hansen, two of the finest scholars of early cinema, have recently investigated ways that the cinematic language of silent film is inherently symbolic—"hieroglyphic" in Hansen's terminology (the term taken from the poet-critic Vachel Lindsay)—and therefore may merit a hermeneutic approach formerly reserved for poetry.37

Jarman was interested in hieroglyphics and Egyptian motifs, as well as secret codes and forms of the occult, from John Dee to Carl Jung. Jarman's films often operate more like silent films than sound films in that the dialogue is often, by contemporary standards, drastically reduced. But for the more limited purposes of this introductory chapter, I need to put early French and German film largely in brackets and focus on a foreshortened tradition of directors in Soviet film—Alexander Dovzhenko, Sergei Paradjanov, and Andrei Tarkovsky.

Although Jarman ultimately made a short film in direct reference to Sergei Eisenstein (Imagining October), I would place Jarman in closer affiliation with this set of Soviet filmmakers. These directors more explicitly, consistently, and rigorously employed the language of poetry. The personal yet overtly political impulses of these Soviet films share similarities with the urgency felt in Pasolini and Jarman, and both Jarman and Paradjanov find in Pasolini a source of inspiration and energy. Thus rather than dividing film history into old narrative and new poetry (as in Pasolini), or even into old poetry and new narrative (which would be one way of reading Metz), we might attribute less importance to early and late, silent and sound, and rather see constellations of film poets, from impressionism to modernism, from classical cinema to the avant-garde.

Let me note also that this trio is connected more figuratively to Jarman than are the other directors in the poetic tradition that I describe in this chapter. Jarman often commented on Pasolini, Cocteau, and Genet; a tradition or context that includes their work does not need to be argued into existence, only studied. But I am not aware that Jarman ever mentions either Dovzhenko or Paradjanov in his extensive prose, although he does refer to Tarkovsky. I think that the parallels which can be drawn between Paradjanov's work and Jarman's are remarkable and critically rewarding, but one can prove absolutely only that both Jarman and Paradjanov found a central font of creative inspiration in Pasolini—that is unquestionably where the branches of this tree join.

One of the key figures in cinematic poetry is the Ukrainian director Alexander Dovzhenko, and one of the most highly regarded of all cinematic poems is Dovzhenko's Earth (1930). As the title of an English translation of his selected works—The Poet as Filmmaker—suggests, Dovzhenko often considered his own films in poetic terms. In his autobiography, Dovzhenko writes:

The few films that I did complete I made with love and sincerity. In these films lies the primary meaning of my life. They are meant to be poetic films, and contemporary life, with the common man at its center, is their chief subject.

The poetic sensibility and structure that Dovzhenko intended has been felt from the very beginning. In The Rise of the American Film (1938), Lewis Jacobs called Earth "a luminous contribution to the realm of lyric cinema" and considers that "Dovzhenko, perhaps more than anyone else, can be called . . . the first poet of the movies." In Film Culture Reader, an important collection of articles and documents treating American avant-garde film of the 1960s, Ken Kelman's essay on Earth moves toward comparing Dovzhenko's nonnarrative impulses with those of directors such as Stan Brakhage, Maya Deren, and Gregory Markopoulos—that is, the creators of what P. Adams Sitney calls "lyric cinema." And Dovzhenko's poetic practice is a recurrent reference point for Tarkovsky, as evidenced in this celebratory paragraph:

Think of Mandelstam, think of Pasternak, Chaplin, Dovzhenko, Mizoguchi, and you'll realize what tremendous emotional power is carried by these exalted figures who soar above the earth, in whom the artist appears not just as an explorer of life, but as one who creates great spiritual treasures and that special beauty which is subject only to poetry. Such an artist can discern the lines of the poetic design of being. He is capable of going beyond the limitations of coherent logic, and conveying the deep complexity and truth of the impalpable connections and hidden phenomenon of life.

For our purposes, as we aim toward Jarman, Dovzhenko's crucial importance in the history of poetic film resides in his implicit investigation of the relationship of the lyrical to the collective. Ordinarily, lyric is regarded as characterized by the subjective and the psychological (classically exemplified by such figures as Sappho and Catullus), but ancient poetry elsewhere does not distinguish so readily between lyric individuality and lyric collectivity (as in Pindar and Horace). It is in the twentieth century that we see most clearly the deliberate adoption of radical stances of subjectivity in poetry of all varieties. Yet at the same time, poetry breaks from convention and even communication in the name of politics. Avant-garde movements often speak in deliberately arcane ways, again in the name of political community. Just so, Pasolini and Jarman repeatedly negotiate the expressive territory between "I" and community, between the individual, suffering speaker and political activism. The subjectivity of poetry in a charged political realm will be an especially crucial problematic in both Pasolini and Jarman.

In Dovzhenko's Earth, the relationship between subjective lyricism and collective expression is theorized by the relationship between the death of Basil (Vasyl), the communist, and the triumphal conclusion of the film. In a scenario for Earth (written retrospectively in 1952), Dovzhenko writes that the nightfall following the harvest contains poetry in itself—a poetry that changes the world:

The apple tree, the willows, the pots on the fences, the old elm tree—every object had become unfamiliar, taken on quite a different nocturnal shape and begun to live a life of its own. It was as if a poet had taken ordinary, everyday words, arranged them into celebratory lines, and transformed them into poetry full of new and exciting meanings.

In such a night, couples are seen one after another, sleeping quietly or simply together. And then, from out of this quiet, we see Basil walking down a darkened lane. He begins to dance with joy, kicking up bright, moonlit dust. Finally, at the end of this extended, deliberate sequence, lovely by any accounting, Basil is shot, and drops to the ground. We see at once the tragic contrast between Basil's premature, sudden death and the peaceful, natural death of Grandfather Semyon amidst the apples, with which the film began.

Yet, equally, we are meant to see the relationship between Basil's nocturnal, dreamy, solitary dance and the song sung by the marching community at the end of the film. Basil's private moonlit dance comes to a sudden, murdered halt, and out of his death comes the image of bright, open song and poised, deliberate march. The difference between the moonlit dance and the sunlit march is the difference between private and public, lyric and ode. In a solitary, nocturnal dream, the world is transformed and magical, as Dovzhenko tells us, and the world is charged with communist hope at the end of the film as well. But Dovzhenko's montage at the conclusion of Earth analytically explains what is left out of moonlit, lyric poetry in order to obtain enlightened song.

Earth's final montage cuts from the singing marchers to three other images: Father Gersaim in an "empty room" (as the scenario says); the ceaselessly running Khoma (killer of Basil); and the anguished, naked form of Basil's wife, Natalka, beating the walls in protest and mourning. This montage analytically separates spheres of transcendence and heightened emotion from the labor songs of the commune. Basil's beautiful, lyrical dance gives way to procession and march, which has nothing to do, the montage implies, with the empty sky of religion ("the empty heavens to which the spiritual strength of people had been directed for centuries in the shape of prayers, supplications and sighs"), the feverish madness of acting against community, or the fervor of subjective eroticism and passion. The montage works to describe what is at stake in the funeral procession and to refine our sense of what sort of changed world this is.

Dovzhenko's Earth gives the impression of lyricism through both content and style: through a virtually plotless narrative joined by dream more than cause and effect; through the ethereal pictures of sky, fruit, and rain; and through its rhythmic patience and deliberation. Indeed, the film is notable for its formality as much as for its atmosphere, light, and manner of proceeding. The formality of Earth is all the more remarkable, perhaps, given its rustic subject matter, and is a main contributor to the film's poetic effects. Poetry has often been defined as speech more formal than ordinary speech, and even though Dovzhenko, Pasolini, and Jarman eschew convention and, often, narrative, they do not fail to employ all manner of artifice and formality. In Dovzhenko the key stylistic formalism might be called the "turning head," where a posed face turns to one direction or another. In a splendid discussion of Earth, Barthélemy Amengual speaks of Dovzhenko's "living statues" and compares their stillness to the photographs in Chris Marker's La Jetée. This rhythmic, formal stillness and turning is central to Dovzhenko's lyrical style; it is related to the formal tableaux in Paradjanov and to Pasolini's famous frontal style of photography. The artifice, poise, and deliberation of the mise-en-scène and editing throughout Earth culminate with the logic of poetry in the collective song of the funeral march, which is both enthusiastic and ordered, both heartfelt and formalized.

This description brings us directly, then, to another remarkable creator of cinematic poetry, the Armenian director Sergei Paradjanov. Dovzhenko himself signed Paradjanov's diploma from the Moscow Film Institute in 1952, and in answer to the question "Why do you make movies?" Paradjanov replied, "To honor the grave of Tarkovsky." Although he outlived Tarkovsky, Paradjanov was born eight years earlier, in 1924, and so I will treat him second in this triad of Soviet film-poets; in reality, their careers overlap at many points. Because his works are in Armenian, and so are less accessible than those in Russian, there is scarcely any scholarship on Paradjanov in any language. For the narrow purposes of this introduction, I will focus on The Color of Pomegranates (1968), a version of the life of the Armenian poet Sayat Nova, and on Paradjanov's cinematic relationship with Pasolini. As noted earlier, I do not see that Jarman cites Paradjanov in his journals, but as I also observed at the beginning of this chapter, I believe that Jarman "quotes" The Color of Pomegranates in The Garden and that Jarman's attitude toward artifice and form is illuminated by comparison with the work of Paradjanov.

Patrick Cazals's Cahiers du cinéma book on Paradjanov contains many interviews with and quotations from the director; from these we can attempt to elucidate the sources of inspiration that Paradjanov drew from Pasolini.48 Cazals himself names a "golden triangle" of directors for Paradjanov ("triangle d'or paradjanovien"): "Pasolini, Tarkovsky, Fellini." (p. 92). Pasolini is a repeated point of cinematic reference for Paradjanov, who summarizes their relationship most generously in this homage (all translations from Cazals's book are my own):

Each time I view Pasolini, I am struck by the same standards. His principles, his spiritual attitude with regards to the Bible, he elevates to the rank of a mission. . . . he makes me discover, astonished, aspects of the world, from antiquity, Rome, Arabia, or simply contemporary life. (p. 137)

Pasolini's exotic settings in works like Medea or The Arabian Nights are undoubtedly related to the poetic projects in Paradjanov's The Color of Pomegranates or Ashik Kerib. Cazals also links Paradjanov's use of asynchronized sound to Pasolini; in both cases, the noticeable unrelatedness of sound to image heightens artifice and defeats realism (p. 83). Most important of all, perhaps, is the link to painterly composition. Paradjanov is a brilliant practitioner of the art of collage, as the illustrations in Cazals's book (including a fragment from the collage "homage à Pasolini" [p. 42]) show. Painting is used not only as an abstract or postmodern reference (as in Antonioni's Red Desert or Godard's Passion), but also as a formal means of lyricizing film. Cazals compares Pasolini's frontal cinematography with Paradjanov's and further suggests that an affinity for tableaux joins the two men (p. 45). Like the halt of narrative in the abstract space of Griffith's rocking cradle, painterly cinema pauses and disrupts narrative. It is no coincidence that poetic filmmakers often focus their attention on painting —Tarkovsky in Andrei Rublev and Derek Jarman in Caravaggio.

The Color of Pomegranates proceeds as a series of tableaux framed by an unmoving camera. The style of the film crosses back and forth between silent cinema and sound: the camera does not move, there is scarcely any dialogue as such, and there are frequent intertitles. Yet at the same time, the screen is saturated with glorious color, and the sound track is equally full of Oriental instrumentation or monastic chanting. Each picture is carefully framed and still, yet at the same time full of life and movement. Water flows, the pages of books flap in the wind, a high-stepping horse struts by. The formalist aspects of Dovzhenko's Earth, where statuesque faces slowly turn, are taken over here and multiplied. Each frame plays out a contrast between formal opposites: linear and arabesque, solid and fluid, horizontal and vertical, covering and revealing, round and flat, empty and full. The images illustrate the words of poetry that are scripted—and scripted beautifully—onto title cards, but clearly the poetry (and the poet, for that matter) is less the subject of the film than an opportunity to visualize objects and people in a poetic manner. Like Dovzhenko's Earth, the film works mythically rather than realistically. Artifice and exoticism are clearly meant to help us transcend the everyday world and see deeper into the flowing essences of life.

The poet's vision is pure and perfectly framed, but passive—offering itself rather than forcing itself. The film repeatedly shows us images of victims, chickens and sheep above all. At one point, the poet digs a grave in a church filled with sheep, and at the end, headless chickens flutter around the prostrate body of the poet. Poetry is not mythologized as salvific or monumentalizing; instead, there is such elegy and suffering that Patrick Cazals writes, "The procession of tableaux appears as an abridgement of a sublime life grasped at the threshold of death, the slow hemorrhage of a suicidal poet" (p. 110). Paradjanov is less activist in his politics than Pasolini or Jarman (although he was perceived as sufficiently nationalist to warrant repeated imprisonment), yet this "suicidal poet" is their kin. As we shall see, twentieth-century cinematic poetry most often turns against itself. Tellingly, the final image of the angel of resurrection is in profile. She is beautiful, and yet she also looks away from us, to one side. We shall return to images that recall the collages and formal gestures of Paradjanov throughout this book.

Andrei Tarkovsky is the most articulate spokesman for the theory of cinematic poetry in the Soviet tradition. Jarman's desire to elaborate discursively and extensively about the meanings and intentions of his films is most clearly reminiscent of texts by Cocteau and Pasolini, but not so far away from the somewhat less voluble Tarkovsky. Tarkovsky's spiritual inclinations may seem to contrast harshly with Jarman's political and sexual concerns, but we will find ourselves at key moments in Jarman's career turning to Tarkovsky for reference. As we shall see, the mirror is a central image for both Jarman and Tarkovsky, serving repeatedly in each case as an almost archetypal figure for the focusing and refracting power of the camera. That Jarman knew Tarkovsky's biography, prose, and films I extrapolate from a complaining remark in Kicking the Pricks: "If Tarkovsky had had the misfortune to be born in Great Britain, I doubt if he would have been able to make a single film; in the Soviet Union he worked with difficulty, but he worked." Jarman would have, I think, been drawn to the language of poetry in which Tarkovsky casts his philosophy of film as well as to the practical difficulties that the Russian director faced when making his films.

Tarkovsky's ideas about cinema are seriously and intellectually thought out, and the most important statements are collected in Sculpting in Time. His aesthetics, which often recur to the poetic, have been well summarized elsewhere, and so, for the most part, I will discuss concrete examples from his practice. A characteristic piece of Tarkovsky's prose goes as follows:

But to return to our theme: I find poetic links, the logic of poetry in cinema, extraordinarily pleasing. They seem to me perfectly appropriate to the potential of cinema as the most truthful and poetic of art forms. Certainly I am more at home with them than with traditional theatrical writing which links image through the linear, rigidly logical development of plot. That sort of fussily correct way of linking events usually involves arbitrarily forcing them into sequence in obedience to some abstract notion of order. And even when this is not so, even when the plot is governed by the characters, one finds that the links which hold it together rest on a facile interpretation of life's complexities.

Tarkovsky reads poetic cinema as both more contingent and more truthful than linear, cause-and-effect narrative, which once again overlaps conceptually with Deren's idea of "vertical" poetry versus "horizontal" narration. Tarkovsky does not wish to essentialize poetry, so he shortly qualifies his description: "When I speak of poetry I am not thinking of it as a genre. Poetry is an awareness of the world, a particular way of relating to reality." Cocteau, Pasolini, and Jarman would all agree, in fact, that poetic cinema is not just poetry, not lineated, metrical writing, so we should not limit our ideas of lyric cinema to conventional, literary forms of poetry. Nonetheless, all of these directors revise our sense of what poetry and poetic cinema are by continually referring to recognizable poems. Their lyric cinema tells us that poetry is not what we think it is, but they also repeatedly use recognizable poems and poets in their work. Thus Tarkovsky's films not only include many dream sequences and follow the logic of dreams, but they also, like Jarman's films, include many interpolations of actual poems. Although the role of poetry in Tarkovsky's Stalker (1979) and Nostalghia (1983) could just as readily be discussed, for the purposes of this chapter, I will limit myself to Mirror (1974), a film that will come up as an important reference for the autobiographical element in Jarman's later films.

Tarkovsky's Mirror is autobiographical to the point of obscurity, and in its personal obliquity shows some affinity with a film like Jarman's The Garden. Just as Jarman films his actual garden, in the midst of an industrial-fringed prairie, Tarkovsky rebuilds the actual house in which he grew up and contemplates the relationship of that inner space to the psychological and social world around it. In his prose, Tarkovsky always attacks overtly polemical and political films, and in this regard he seems to be an aesthetic contrast to Pasolini and Jarman. But few twentieth-century aesthetics, as we have seen, work entirely in formalist or "art for art's sake" terms, least of all in Russia, and Tarkovsky's personal reminiscences are always linked to wider social, national, and even global concerns. Although seemingly cryptic autobiography, The Mirror is as insistent, even tendentious, in its social commentary as his parable of nuclear holocaust, The Sacrifice (1986).

Mirror features four poems by Arseny Tarkovsky, the director's father. These poems self-consciously signal the poetic structure of the film, much like the intertitles of Griffith's Broken Blossoms or Paradjanov's quotations of Sayat Nova. These poems also heighten the autobiographical nature of the film, and are in fact spoken by Arseny himself. In The Cinema of Andrei Tarkovsky, Mark Le Fanu writes that "the tone of Arseny's poems is, like Pushkin's, that of thrilling, lofty affirmation." After quoting from Arseny—"I am one of those who haul the nets / When a shoal of immortality comes in"—Le Fanu says further, "It is a splendid definition of the poet and film-maker."55 Le Fanu characterizes Tarkovsky as a metaphysical opposite of Godard: a spiritual creator contrasted with an ironic one, a believer in the creative power of the word compared to an agnostic.

Yet Tarkovsky's Mirror is about broken homes and a broken world, and I would suggest that Tarkovsky substantially deidealizes his (absent) father's poetry. Arseny's first poem, "First Meetings," is indeed, for the most part, glorious affirmation:

Every moment that we were together
Was a celebration, like Epiphany,
In all the world the two of us alone.
. . . . . . . . . . . . . . . . . . .
You awoke and you transfigured
The words that people utter every day,
And speech was filled to overflowing
With ringing power;
. . . . . . . . . .
Ordinary objects were at once transfigured,
Everything; the jug, the basin—when
Placed between us like a sentinel
Stood water, laminary and firm.
We were led, not knowing whither,
Like mirages before us there receded
Cities built by miracle,
Wild mint was laying itself beneath our feet,
Birds travelling by the same route as ourselves,
And in the river fishes swam upstream;
And the sky unrolled itself before our eyes.

All this affirms, indeed, a transfiguring, poetic vision of the changed world of love. But the poem ends:

When fate was following in our tracks
Like a madman with a razor in his hand.

Arseny's poem furiously doubles back on itself at the close, and Tarkovsky has the poem read over the lonely, reflective mother, her children quietly with her, and a notion of her crying at the end. The poem captures a happy memory with a brutal, sudden closure, but the director makes everything in the poem palpably absent, pathetic, estranged. This effect, which emphasizes the razor more than the mint, is also reproduced when the next poem is read, as the mother walks down a hall in the printing plant. Here we realize that her hurried, crazed search for a typo in the latest edition of the newspaper acts out a displacement of broken love, and this realization is confirmed when one of her fellow workers starts to lecture her about her personal life. The love poem again marks an elegiac memory, but not an affirmative prophecy. In general it seems to me that Tarkovsky's pictures do not illustrate or substantiate the poems as much as they interact complicatedly with them.

Arseny's third poem provides the most affirmative claim and forms the basis of Le Fanu's characterization.

I don't believe forebodings, nor do omens
Frighten me. I do not run from slander
Nor from poison. On earth there is no death.
All are immortal. All is immortal. No need
To be afraid of death at seventeen
Nor yet at seventy. Reality and light
Exist, but neither death nor darkness.
All of us are on the sea-shore now,
And I am one of those who haul the nets
When a shoal of immortality comes in.

Yet Tarkovsky projects beneath these words documentary footage of the Red Army crossing Lake Sivash. As immortality is extolled, we see soldiers tramping through mud. Tarkovsky knew that many of these soldiers would die on this mission, and there is not a very obvious sign of immortality anywhere around. Of all the visual backdrops that Tarkovsky could have chosen for his father's poem, this one seems as bleak as possible, as far from an affirmative heroism as possible.

And when the poem continues, it draws the same bridges between past and present as the film does.

Live in the house—the house will stand.
I will call up any century,
Go into it and build myself a house.
That is why your children are beside me
And your wives, all seated at one table,
One table for great-grandfather and grandson.
The future is accomplished here and now[.]

Arseny's poetic house disputes linear time and gathers together families beneath its vast, welcoming roof. Tarkovsky's film also refuses chronological sequence and joins together past and present (most remarkably at the end, when the two little boys from the beginning of the film are seen walking with the "old mother"—a historically impossible scene). Film and poetry each have the power to invoke absent figures and to juxtapose disparate regions of time, but Tarkovsky's gliding, fallen world seems fundamentally elegiac rather than celebratory. Tarkovsky does not idealize the powerfully synthetic capability of poetry, since the world, both personally and historically, is too ruined to fix. The film's final picture of the "old mother" with the children imagines a breathtaking revenge against time, but it is also transparent fantasy, and the camera retreats away from them into dark woods. At the end of Ivan's Childhood (1962), we enter a dream in which Ivan frolics with his sister on a paradisaical beach, but this too is sheer wish fulfillment, since Ivan has been killed by the Nazis.

As critics have noted, Mirror begins with a kind of set piece of poetic vocation, when the stuttering student suddenly finds his voice: "I can speak!" Yet at the end, the narrator seems to be dying of strep throat. Mirror is lyric cinema, in Tarkovsky's terms and in our own, but as a parable about poetic vocation or the power of artistic creativity, Mirror is complex, not idealized. Mirror invokes poetry as an enabling contrast to prosaic modes of proceeding, but it does not do so triumphantly or self-righteously. This nonidealized, unselfrighteous turn to poetry has much in common with Jarman's own assumption of the poetic.

Critics have noted on various occasions that both Paradjanov and Tarkovsky use over and over again the image of water, in particular the image of running water. Rain and dripping water enable various kinds of visual and sonic effects, no doubt, and critics often explicate these watery images with reference to nature; as homages to Dovzhenko (as the dream of apples and rain in Ivan's Childhood surely is); or as related to oceanic maternity. In Paradjanov's The Color of Pomegranates, I take the liquid movement of water and other fluids to be setting up a relatively formal contrast to the symmetrical and stationary elements in his collages, as formal contrasts to the unbending and unmoving picture frame. In Tarkovsky's Mirror, I take the running water to emphasize the elegiac, gravitational pull of the earth: we feel, there, both chance and inevitability. When the camera tracks through the rooms of the house toward the burning barn in an early scene, "a bottle falls off the table for no apparent reason." The fence breaks beneath the mother and doctor in the first scene; the children give spilled milk to the cat; the barn burns. The slow, fluid tracking shots and the imagery of falling, falling toward the earth, are emblematic of Tarkovsky's magical and also fatalistic cinematic poetry.

Usually, Tarkovsky's cinema is seen as martyring itself on the altar of truth in the face of unsympathetic authorities and distributors, and Tarkovsky's prose certainly enforces this idealizing image. But in my turn I would want to underscore the degree to which Tarkovsky's poetic cinema is not burned at the stake by uncomprehending law, but rather burns itself at the stake as a witness to the destruction of the twentieth century and as an embodiment of the laws of twentieth-century poetry. Once again, this is poetry that rejects serial narrative, but does not reject it scornfully; it brings unlikes together, but does so in mourning, rather than in salvific unity. Through the course of this book, we shall return to the examples set by Paradjanov and Tarkovsky for the illumination they may shed on the cinema of Derek Jarman.

III. Egoyan, Cocteau, and the Poetic Accident

A great film is an accident, a banana skin under the feet of dogma.

Cocteau

If there is an apparently "aesthetic" director of cinematic poetry, then it is not Paradjanov or Tarkovsky but Jean Cocteau. Cocteau happily speaks in terms of "poetry" and "beauty" through all his writings, and he seems to have been even less politically motivated in his mid-century France than Oscar Wilde in his late Victorian England. Yet like Wilde's plays, Cocteau's films have been retrospectively read into a tradition of gay aesthetics, and like Pasolini, Cocteau becomes a recurrent point of reference for Jarman. It is the accidental that is a main element in Cocteau's varying descriptions of the poetic, and he explores this aspect quite deliberately (the paradox of the intentional accident runs throughout his discussions) in both prose and film. In this third section I will focus on Cocteau's fable of poetry, Orpheus (1949), and continue to meditate in some detail on the notion of the poetic accident. In the chapter's fourth and concluding section, I will then examine Cocteau's Testament of Orpheus and its relationship to other lyric cinema by homosexual directors such as Pasolini and Jean Genet.

Because poetry for Cocteau is not premeditated (he always complains against those who self-consciously attempt poetic effects), he welcomes the accidental: "So I hope, with all my heart, that Hollywood will associate a minor branch to the great enterprise of film-making, one that will not be protected against routine by any insurance, but invite accidents to happen." The accidental defies premeditation, but ultimately comes to serve as a revelatory sign, or a bridge to truth, just as the raindrops in Tarkovsky trace a pattern that is both accidental and inevitable. Poetic cinema materializes in images the truthful arbitrariness of its progression—through falling bottles and spilled milk in Tarkovsky and through, on a larger scale, fatal accidents in Cocteau. It is no coincidence that Kieslowski's Trois Couleurs—often characterized as "lyrical" and "poetic"—begins with a car accident in Blue. As seen above all in the coincidences and contingencies of colors, the metaphysical aesthetic of Kieslowski's film is held together by supernatural chance. The thematic embodiment of poetic accident is exemplified also in Atom Egoyan's The Sweet Hereafter, to which we need briefly to turn.

The Sweet Hereafter, an atmospheric, otherworldly film, is adapted from Russell Banks's 1991 novel. The novel shows us a community's response to the tragic crash of a school bus off an icy road into a river. The book moves through five monologues, spoken successively by Dolores Driscoll, the bus driver; Billy Ansel, a Vietnam vet and widower who loses his two children in the event; Mitchell Stephens, a New York City lawyer who seeks damages for the accident; Nichole Burnell, a teenage girl who is crippled in the accident; and lastly, Dolores Driscoll once again. The modern perspectivalism effected by the lack of an omniscient narrator is matched by the novel's bleak resolution: the lawyer can find no cause, no explanation for the accident. The hardheaded lawyer, Mitchell Stephens, returns empty-handed. The accident merely happened, therefore; there is no great metaphysical conclusion or consolation to be found.

Yet even as the novel metaphysically and existentially rejects the lawyer's determined search for causality ("There are no accidents," says Mitchell Stephens), the logic of the novel does not coincide with the usual mechanisms of fictive explanation. There are no deep explanations for accidents, the novel says, bleakly, but the novel itself, in spite of its potentially relativistic perspectivalism or decentering, still obeys conventional contracts between writer and reader. Thus, characters perform their actions and come to their own conclusions for reasons that are clearly articulated within the novel itself. For example, a key obstacle to the lawyer's pursuit of explanation (and money) is the testimony of Nichole Burnell. She lies, claiming Dolores was speeding (no one wants to sue Dolores), in order to revenge herself on her sexually predatory father. The motivation for her lie and the consequences of her actions, as with other motivations and consequences in the novel, are made perfectly intelligible to us. Thus, for all of its angst or formal modernizing, there truly are no narratological accidents in the novel by Russell Banks.

Egoyan's cinematic version of The Sweet Hereafter, then, not only reduces and eliminates aspects of the novel's plot and character—usually necessary when condensing the thickness of a novel to ninety minutes of screen time—but it fundamentally changes the generic mode of the story's representation. One need not be overly essentializing or idealizing about generic form to claim that Egoyan's cinematic version has transformed novelistic causality to poetic accident. Egoyan now frequently cuts back and forth chronologically, as Tarkovsky does in Mirror, so that characters are seen well and whole after the bus crash. The most poignant example of this comes at the end when Nicole (her name is spelled thus in the film) reads a bedtime book to Billy Ansel's two children; they fall asleep, and she walks away to look out a window, which fills with light. Since the children will die and Nicole will never walk again, the scene is a fantasy—although technically a flashback—a fantasy that takes place in heaven, in another world where all is well. Unlike Banks's title, which is sheer, unforgiving irony, as critics noted in their book reviews, Egoyan applies the same title to an authentically reimagined space, the changed world of poetry—the sweet hereafter. If Egoyan's title is not brutally ironic, neither is it sentimentally utopian. The last scene of The Sweet Hereafter recalls, in fact, the last scene of Tarkovsky's Mirror. Both are scenes that dally with time's distress (although one is fact, the other fancy); both unify nostalgically, but undermine themselves elegiacally in the same gesture.

Egoyan's reductions of plot not only collapse for the sake of abbreviation, they eliminate the very explanation that constitutes narrative. Egoyan resees the crash according to a lineage of poetic cinema. He highlights the accidental nature of events, not only in the plot, but also in the form, the telling. If, as David Bordwell has argued, classical narrative is fundamentally concerned with causality, then Egoyan's manner of proceeding reduces explanation by highlighting the contingent. The central pivot of the entire plot, Nicole's invented testimony, now has no clear causality; a reviewer in Sight and Sound calls it "inexplicable." The relationship of father and daughter is not so straightforwardly portrayed in the film—we have only a lyrical fantasy from her point of view to suggest her father's abuses. What is more, Egoyan expressly and self-consciously announces the generic alteration of the novel into poetry—the cutting back and forth in time, the reduction of causality, the atmospheric, brooding music—by adding a poem, "The Pied Piper of Hamelin" by Robert Browning. This is the bedtime story that Nicole reads to the two Ansel children, and it organizes thematically and rhythmically the totality of the film. Like Arseny Tarkovsky's lyrics in Mirror, the Browning poem names The Sweet Hereafter as part of the tradition of cinematic poetry.

Cocteau repeatedly emphasizes the accidental in true poetic creation. He insists that one cannot analytically interrogate poetic form, and this is why both Heurtebise and the Princess tell Orpheus to ask no questions. Poetry is related to dreams, to self-consciousness, thus accounting for the seemingly random structure and continuity of The Blood of the Poet (which is not surrealist, according to Cocteau). Cegeste's accidental death brings the underworld into the mundane, and Orpheus finally dies when a gun accidentally goes off in a scuffle. All of these accidents also have the force of inevitability, however, as we know that it is death's stormtroopers who are running down their victims. As elsewhere in the tradition of poetic film, these accidents are characterized both by chance and by a deeper truth.

Cocteau contrasts the mundane, everyday world with the poetic underworld, but slyly, self-consciously (one of the clubs where Orpheus may be found is called "The Two Worlds"). The normal, unpoetic world is found, first of all, in Orpheus's bourgeois household, decorated with tasteful furnishings and framed engravings. Cocteau's film savages the ideals of romantic love at the heart of both the Orpheus myth and middle-class marriage by showing Orpheus and Eurydice as a squabbling, quarreling couple. The myth's poignant prohibition—"Do not look at your wife!"—becomes an intolerable inconvenience to Cocteau's Orpheus. This Orpheus does not turn to look at Eurydice out of the yearning of infinite passion, but rather glances up at her in a car's rearview mirror, accidentally. In addition to this tedious and impatient middle-class household, the other characteristic space of the unpoetic world in the film is, tellingly, the café where all the poets hang out. Here, where fashion and convention reign, is premeditated poetry, and so not poetry at all, as Cocteau always maintains in his prose, along with Tarkovsky.

In contrast both to the walled, middle-class household, with its garage, garden, and classical statuary, and to the trendy club, with its intoxicated fights and trendy journals (such as "nudism," a book of empty pages), Cocteau imagines an underworld—the poetic otherworld—a space of ruins and night. Using relatively few special effects (eschewing mists and double exposures), Cocteau creates a space of slow, timeless absence. The worldly irritability and time-consciousness of middle-class home and literary club give way to slow motion, silence, and a deliberateness suggestive of eternity. Not that Cocteau creates a world that is entirely alien or remote from the human. For the helmeted motorcyclists still stand around like Nazi stormtroopers, and the tribunal of old men performs a readily identifiable interrogation and show trial. The Princess smokes, and asks for a pen. This Orphic underworld is a deserted, ruined contrast to the bourgeois and the trendy, but it is also connected through the logic of dreams to the real world, to the visual imaginary of our historical world.

The archetypal logic of the original Orpheus myth (as may be found in the narrative at the close of Virgil's Georgics) shows the relationship of the poetic to transcendental emotion and experience: Orpheus can charm the rocks and trees, is able even to charm Hell and win his wife back from the confines of Death. The power of poetry overwhelms itself and the poet himself as well when the frenzied bacchantes shred Orpheus limb from limb. Yet in many versions of the myth, poetry is so powerful that the decapitated head sings out even while it floats down the river.

Cocteau deidealizes this myth of transcendence and ecstasy in a number of significant ways. Orpheus himself is not particularly likable, and his quarrels with Eurydice seem trivial and fussy, rather than heroic resistances in the name of art. Once again, therefore, the rejection of what is not art or not poetry is not self-righteous. This Orpheus is inspired by a car radio in the garage. Cocteau certainly aims to demystify conventional, more obviously Parnassian sites of poetic inspiration, but the image of Orpheus obsessively bent over the radio, copying down numbers, removes from him all trace of the heroic or the transcendent. This Orphic creativity appears secondhand, and poetic madness just looks kooky in these scenes. This waffling Orpheus does not know what he wants, Eurydice or Death; he does not know where creativity resides. Yet this may be, Cocteau implies, a truer Orpheus after all; instead of one rapt with power and creative will, this irritable, indecisive Orpheus may be a truer conduit of the truly poetic—the poetry that occurs obliquely through accident and chance.

Cocteau's homosexuality means that his cinematic representations of poetry can be retrospectively related to representations of sexuality. Cocteau's films do not overtly represent homosexuality, nor, for the most part, do Pasolini's, but they can be related to traditions of homosexual art that directors like Jarman eventually drew from. The theme of the marginalized criminal, for example, can be seen to link directors such as Genet, Pasolini, Cocteau, and Kenneth Anger, all filmmakers with a formative influence on Jarman's work. A film like Cocteau's Orpheus can be read retrospectively as a gay text, though most overly particularized interpretations do not seem very convincing. There is no doubt that Cocteau radically disrupts the Orpheus-Eurydice couple on which the myth is found, but it seems that his target is as much middle-class respectability as it is heterosexuality. Yet Pasolini often makes the same strike: that is, he does not necessarily represent homosexual relationships with the positive, visual force that we find in Jarman's films, yet his attacks on middle-class norms and attitudes, as in Teorema, certainly attempt to remove a main obstacle to sexualized judgmentalism.

Among Jarman's various references to Cocteau's Orpheus, the following passage provides the most comprehensive explanation and context:

DEAD TO THE WORLD

***

Part of my interest in the magician John Dee was his preoccupation with secrets and ciphers.

Why this obsession with the language of closed structures, the ritual of the closet and the sanctuary? The prison cells of Genet's Un Chant d'amour, the desert encampment of Sebastiane; [Kenneth] Anger, insulating himself with magick, screening himself off; Cocteau's Orphée, an attempt to steal through the screen into the labyrinth and usurp the privileges only the cabal of the dead may confer; the wall of unreality that girds the house in Salo, and its victims, who are told: What is about to take place here will have never happened, you are already dead to the world outside.

Jarman does not read homosexuality into Cocteau's Orpheus through character, but rather through theme and image: the thematics of secrecy and the imagery of enclosure. It may be that Jarman's queer poetry is born here, at the junction of the accidental and the secret. How is one part joined to another? We don't know: it is a secret; we do not know: it is an accident. Does the act of reading (of viewing, of interpretation) lead to a kind of freedom—the secret expressed? the truth exposed? What is the nature of poetry and symbol when it has to do not with the formal relationships of aesthetics, but with the socially problematic expression of same-sex desire? We will begin to address these questions by looking, in the final section of this chapter, at several films that are, like Jarman's, both contributions to the tradition of lyric film and also clear expressions of homoerotic sensibility.

IV. Genet's Un Chant d' amour and the Invisible Underground

By setting Jarman next to Cocteau, Pasolini, and Tarkovsky, rather than Anger, Deren, and Jonas Mekas, I am following Jarman's sense of his own context and audience. Once he began to make feature-length films, he always insisted on the relatively traditional nature of his work. The films are low-budget, necessarily, but art-house, not alternative. Certainly, after Sebastiane one would not imagine that Jarman made underground films. Yet in both their antinarrative impulse and their representation of homosexuality, Jarman's films may be seen as rerouting underground themes into the feature-length light of day. Each of his feature-length films may be read as a parable of the visionary (and of visibility) in which poetry and homosexuality, which often make for short films or private showings, now appear to be jostling for feature-length attention, no longer satisfied with their relegation to the avant-garde. In this concluding section of the chapter, we need to fill out our sense of lyric cinema by looking at film that is more apparently to the scale of lyric poetry and by focusing especially on one of the most famous gay movies in history, Genet's Un Chant d'amour (1950).

By rebuilding the underground aboveground, as it were, Jarman continues the interest in poetic cinema embodied by a number of mid-century American filmmakers and definitively described by Sitney in Visionary Film: The American Avant-Garde. Filmmakers like Deren and Brakhage not only seem to be visionary, lyrical filmmakers in Sitney's terms ("the lyrical film postulates the film-maker behind the camera as the first-person protagonist of the film"), they use the language of poetry throughout their own extensive prose works. We have noted earlier Deren's important contribution to the symposium on "Poetry and Film"; her Collected Works includes many poems and also elaborates the interest in ritual and the occult that led to her film on Haiti. Brakhage's numerous prose works, including The Brakhage Lectures, cannot but remind readers of prose by Charles Olson or Robert Creeley.

Brakhage's film Thigh Line Lyre Triangular (1961), for example, might be read as a lyric poem that self-consciously renders the invisible visible. The "birth film" is, by genre, a species of underground film; one does not show, in any detail, the birth of a child to the public; on the contrary, the showing is necessarily private, restricted. Brakhage's birth films are home movies taken of his wife, Jane Brakhage, giving birth to their children. But these private images, especially those of Jane Brakhage's body, are taken into the realm of imaginative poetry and art. The delta of Jane's vagina becomes an Orphic lyre, inspiring the painterly filmmaker with color and emotion. Thus the film becomes more than a documentary revelation of a nude woman giving birth; it renders Stan Brakhage's wild responses to the event, as paint flies over and around the pictures. The act of projection in itself may thus be said to signal a liberation from oppressive distinctions between public and private. Whereas Window Water Baby Moving records a birth with conventionally beautiful images and without too much artificial mediation or postproduction interference, Thigh Line Lyre Triangular attains the complexity of a lyric poem by balancing subjective, painterly vision with emancipatory, documentary effect. Formal consciousness ("line," "triangular") is balanced against social consciousness, making for an uneasy viewing experience.

Genet's Un Chant d'amour (A Song of Love) also embodies, and may be said to perform, these paradoxes of underground visibility. Genet made the twenty-five-minute film with money from producer Nico Papatakis, nonprofessional actors, and professional film technicians. Genet intended the film for private viewings only, although it eventually found its way into various avant-garde and alternative theaters. The slow progress and controversial itinerary of the film (sometimes leading to arrests and always to polemics) make it a true underground film, one of the most famous in cinema history. For the book jacket of Jane Giles's The Cinema of Jean Genet: Un Chant d'amour, Jarman writes, "There's no smoke without fire; Un Chant d'amour is a communion in which Jean Genet takes us into the prison in order to liberate us from it."

Un Chant d'amour buries itself in the underground two times over, through its cinematic lyricism and through its explicit representation of homosexuality. The film makes itself more incomprehensible through its manner of signification at the same time that it makes itself less socially acceptable by showing men's desires for one another in graphic detail. Where the film's plot is expected to be clear, it is obscure; it unfolds through symbolism rather than narrative. And where the film is expected to be obscure—in its representation of sexuality—it is explicit. Like Brakhage's Thigh Line Lyre Triangular, Un Chant d'amour is more than a documentary liberation—here we are shown the repressed, physical body in all of its joy—since the expression of naked love is organized and framed in an obviously subjective, artificial way. Hence Jarman in the above quotation speaks not only of liberation, but also of "communion"—the ritual of art.

Why are there symbols in Un Chant d'amour? For these are not symbols of sexual implication; these are not the (simple) sexual symbols of, say, Alfred Hitchcock—the fireworks in To Catch a Thief or the train going through the tunnel at the end of North by Northwest, where the symbol in each case stands for a sexual act. Un Chant d'amour does not need phallic symbols since it already shows us, quite clearly, the phallus. At the beginning and end of Un Chant d'amour an arm swings a bouquet of flowers back and forth out of a window; at the end of the film, the bouquet is finally caught by another arm. The image of the flowers is in part used to frame the film, to provide it with structure in the absence of concrete narrative; so that finally, at the end, we feel a kind of consummation or success. The flowers also substantiate the title, assuring us that this is a poem, a song, and a song of love. One way of reading the flowers is that they help compose and dignify what may otherwise be jumbled images of erotic desire; they help us see art and civilization rather than chaos; they help us see love rather than lust.

Yet in all of his works, Genet's figuration readily deconstructs; no wonder he is one of the two heroes in Jacques Derrida's Glas (the other is Hegel). As Edmund White points out, "Genet declared that he liked only works of art that destroyed themselves, that were both player and target in an artistic shooting gallery." Lyric symbolism—the flower, the dance, the song—does not exist in order to raise the material into art or to add beautiful imagery to otherwise bleak regions. For the flower in Genet is inextricably, one might say dialectically, related to filth, as beauty is to ugliness, poetry to prose. The flower is born of the prison. Were there no prison, no obscure gesture would be needed; the men would need not to exchange cigarette smoke through a thin straw. Were all the walls down, there would be no need for figurative communication—communication that is both joyful and desperate, ecstatic and onanistic. The contrary energy of poetry and symbol does not perfect a good world; it comments desperately on a bad world, enclosed and oppressive. In poetry here we do not return to the primitive expressions of prelapsarian men. Instead, poetry itself signals the fall: communication has been reduced to gestures and images that may travel no further than cell to cell, window to window, private theatre to theatre.

Without much interpretive strain, Un Chant d'amour may be read as entirely self-reflexive, that is, as an allegory of underground film. Like many of Jarman's films, Un Chant d'amour models explicitly relationships between viewer and the viewed, thereby providing analogies for cinematic experience, for the relationship between film and audience. Un Chant d'amour unfolds as an exercise in voyeurism: we do not just objectively see the prisoners in their cells, at a distance; rather we see from a point of view that is aligned throughout the film with the spying guard.

It has been noted that the voyeuristic observer who looks in on one cell after another is reminiscent of the poet in Cocteau's Blood of the Poet; we know also that Cocteau watched the filming of Un Chant d'amour and that he provided the setting, although probably not actual assistance. The two voyeurs do look superficially similar, but turn out very differently in the end. Cocteau's poet is a passive, floating viewer in a hotel hallway; his exotic visions (opium smokers, a hermaphrodite) appear through the keyhole like strange, short films—magical, but rather dead, without affect. Cocteau toys with the idea of poetic inspiration and the marvelous, but demystifies the mythology as soon as it appears. Although not idealized, Cocteau's approach still aligns the viewer with the poet, and our creative participation with the images and hieroglyphics in the film is certainly encouraged.

In Genet, however, creative expression is squarely on the side of the prisoners in the cells, and the viewer is aligned with the police. We watch, fascinated and perhaps even intrusive, as the guard steps through the door and into the room. The film is intentionally underground, private, but it reenacts in its development the interruption of its private spectacle by an intrusive public, an audience that both judges and participates. The underground is already visible from the moment of filming; there is no hiding place. The underground is not a place of safety, of hiddenness, but always a space where public and private share tenuous borders, where visibility becomes a necessary tool for liberation, but one that occurs only through contention and fight.

The tradition of lyric cinema that leads toward Derek Jarman is a many-branched tree, multifarious and always changing. As we have seen, even the directors who most deliberately invoke the lyrical also often substantially undermine it. These lyric films are self-conscious to the degree that they include actual poems and poets. Yet this self-conscious artifice is often attended by irony, pathos, and subversion. Another film historian may well dispute the existence of such a tradition; with different purposes in mind, with a different goal, the films brought forward here would certainly be seen differently. As I confessed at the outset, these brief descriptions of lyric cinema form a rather tentative narrative that aims tendentiously toward Jarman. The names assembled here are partly those directors that Jarman mentions himself (Pasolini, Cocteau, Genet), on whom he has intentionally modeled his work, but also those that I feel are rewardingly placed nearby (such as Paradjanov). I will have occasion to invoke these filmmakers throughout the rest of this study, and the reader can expect other names as well, which they might very properly have expected in an overview such as this: Fellini, Buñuel, and Anger.

Jarman's feature-length films work themselves out from the underground, from short films and Sebastiane to Edward II and Blue, yet still bear many of the problems and themes associated with visibility and representation that we have touched on in Genet's Un Chant d'Amour. Genet's image repertoire in his lone film and throughout his novels repeatedly draws on the figure of the criminal—the prostitute, the prisoner, the thief. Jarman's imagery may be more idiomatically connected to the light and the light-writing that is cinema, but it is equally suited to the complex analysis of homosexuality that takes place in his art. The imagery that his films and prose return to, over and over again, has to do with the mirror and the sea, and it is to an extended discussion of those images that we turn to now.