Cinema and Painting

[ Film, Media, and Popular Culture ]

Cinema and Painting

How Art Is Used in Film

By Angela Dalle Vacche

In this discerning new approach to cinema studies, Angela Dalle Vacche discusses how the use of pictorial sources in film enables eight filmmakers to comment on the interplay between the arts, on the dialectic of word and image, on the relationship between artistic creativity and sexual difference, and on the tension between tradition and modernity.



33% website discount price

This is a print-on-demand title. Expedited shipping is not available.


6 x 9 | 320 pp.

ISBN: 978-0-292-71583-7

The visual image is the common denominator of cinema and painting, and indeed many filmmakers have used the imagery of paintings to shape or enrich the meaning of their films. In this discerning new approach to cinema studies, Angela Dalle Vacche discusses how the use of pictorial sources in film enables eight filmmakers to comment on the interplay between the arts, on the dialectic of word and image, on the relationship between artistic creativity and sexual difference, and on the tension between tradition and modernity.

Specifically, Dalle Vacche explores Jean-Luc Godard's iconophobia (Pierrot Le Fou) and Andrei Tarkovsky's iconophilia (Andrei Rubleov), Kenji Mizoguchi's split allegiances between East and West (Five Women around Utamaro), Michelangelo Antonioni's melodramatic sensibility (Red Desert), Eric Rohmer's project to convey interiority through images (The Marquise of O), F. W. Murnau's debt to Romantic landscape painting (Nosferatu), Vincente Minnelli's affinities with American Abstract Expressionism (An American in Paris), and Alain Cavalier's use of still life and the close-up to explore the realms of mysticism and femininity (Thérèse).

While addressing issues of influence and intentionality, Dalle Vacche concludes that intertextuality is central to an appreciation of the dialogical nature of the filmic medium, which, in appropriating or rejecting art history, defines itself in relation to national traditions and broadly shared visual cultures.

  • List of Illustrations
  • Acknowledgments
  • Introduction. A Thematic and Intertextual Approach
  • Chapter 1. Vincente Minnelli's An American in Paris: Painting as Psychic Upheaval
  • Chapter 2. Michelangelo Antomom's Red Desert: Painting as Ventriloquism and Color as Movement
  • Chapter 3. Eric Rohmer's The Marquise of O: Painting Thoughts, Listening to Images
  • Chapter 4. Jean-Luc Godard's Pierrot le Fou: Cinema as Collage against Painting
  • Chapter 5. Andrei Tarkovsky's Andrei Rublev: Cinema as the Restoration of Icon Painting
  • Chapter 6. F. W Murnau's Nosferatu: Romantic Painting as Horror and Desire in Expressionist Cinema
  • Chapter 7. Kenji Mizoguchi's Five Women around Utamaro: Film between Woodblock Printing and Tattooing
  • Chapter 8. Alain Cavalier's Thérèse: Still Life and the Close-Up as Feminine Space
  • Notes
  • Bibliography
  • Index

Browse the book with Google Preview »

While some may feel that film does not belong to the history of art, the fact is that filmmakers often use paintings to shape or enrich the meaning of their works. Thus the history of art is in film, even though, by evoking high art and creativity, rather than technology and mass culture, painting for the cinema constitutes a forbidden object of desire. This relation of love and hate between cinema and painting is further complicated by the tendency of art history to be evoked well beyond the boundaries of a text or the intentions of a filmmaker. It takes more than a study of the sources of a film to reveal the beauty of the encounter between cinema and painting; one must imagine all the possible elements of visual culture that a film, just by virtue of its circulation, has the power to attract into the textual orbit. It is that internal richness and outbound energy that I strive to convey in this book, by opening up film studies to the history of art and by trying to encourage other colleagues to join me in the new field of comparative arts. In short, my effort in this book stems from the belief that art history as a discipline cannot afford any longer to ignore film studies, for the advent of the cinema has forever changed the meaning of the word "art" and the meaning of the word "history."

By looking at eight films produced under different historical circumstances and in different cultural contexts, each directed by strong creative personalities, I ask: what happens to the paintings used or alluded to in these texts, and how do these films define painting as the realm of high art, creativity, and femininity, setting it against popular culture or industrial technology? Precisely because each film I examine offers a different definition of painting as art in film, my approach is thematic, the title of each chapter is meant to spell out the director's attitude toward painting. With Vincente Minnelli, for example, painting in An American in Paris (1951) is so compelling and subversive that it goes hand in hand with psychic upheaval, whether beneficial or harmful; for Michelangelo Antonioni, however, painting in Red Desert (1964) is both unsettling and irresistible, so that the director transfers the burden of creativity to the female protagonist.

Just as painting translates itself into different themes ranging from psychic upheaval to feminine sensitivity in the films I examine, the whole category of art emerges in a variety of manifestations across my case studies. In An American in Paris, Minnelli explores many views of art: as decoration, neurosis, temporary utopia; as a source of economic power; and as a mark of foreignness, elitism. In F. W Murnau's Nosferatu (1922) and in Kenji Mizoguchi's Five Women around Utamaro (1946), art is meaningful as long as it prevails over the laws of the marketplace. In Pierrot le Fou (1965) and Thérèse (1986), art is implicated in the establishment, whether this means aesthetic tradition or patriarchy; thus collage and still life, because of their historical reputation as subversive or marginalized genres, offer Jean-Luc Godard and Alain Cavalier a standpoint from which to challenge a wide range of dominant values, from fixed gender roles to the separation of art and life.

For Andrei Tarkovsky in the film Andrei Rublev (1966), art is a form of prayer and of communal energy. By contrast, in The Marquise of O (1975), art fosters a series of vices that Eric Rohmer denounces by developing a film style that opposes artifice, vanity, and self-indulgence, even as his pictorial images disclose the strength of temptation. Finally, for Antonioni, to practice art means to be caught in the struggle between experimentation and nostalgia. In short, the cinema's use of art in innumerable ways suggests that, even though my book focuses on painting as the most problematic but also the most alluring of art forms, all the films I examine must summon all sorts of sources in order to define art as a category.

This is the case because, by blurring the distinction between high art and popular culture, the cinema has always had a tendency to challenge not just painting in isolation but rather the whole system of the arts, thus disclosing the possibility of new configurations, hierarchies, alliances, and hostilities. In An American in Paris, for instance, while Minnelli associates painting with extreme expressions of creativity and emotion, he also turns to dance to stress communal values that reinsert the single artist within a national tradition. In Red Desert the contrast is not between solipsistic painting and communal dancing, rather, architecture, with its allegiances to mass culture, public space, and industrial technology, allows Antonioni to align painting with a feminine vision at odds with the values of rationality and efficiency but open to inner rebirth and new avenues of self-expression.

My approach is also intertextual. A broad category of analysis with overtones that range from narrative to stylistic detail, intertextuality for me includes the borrowing of images from art history to inflect the meaning of a text, the rejection of painting to stress the unique features of film, the insertion of cinema in broadly shared visual cultures and national traditions, and, finally, the power filmmaking has to redefine art history. Thus intertextuality enables thematic contrasts, iconographic similarities, and historiographic commentaries. In a sense, as broad as it may be, my definition of intertextuality shuttles between two extremes in the way film sees itself in relation to other arts: either film is plagued by a cultural inferiority complex and therefore obsessively cites other art forms, or it is self-confident enough to move beyond this state of dependency and arrive at the point where it can teach something new to art historians. To highlight this second extreme, I will present as my final case study the daring film Thérèse, by Alain Cavalier. Far from being overwhelmed by the history of painting, this film teaches us to look at the tradition of domestic genre painting with a still-life component in a new way.

My method also assumes that since the cinema is engaged in a dialogue with the other arts, the films under analysis can be read as allegorizations, self-conscious meditations on what is at stake in the encounter between painting and cinema, art and technology, tradition and modernity. This allegorical or metacinematic dimension is especially apparent in Murnau's Nosferatu and Mizoguchi's Five Women around Utamaro. Nosferatu thrives on Murnau's familiarity with the legacy of German Romantic landscape painting and also on the director's awareness that the invention of the cinema shortly preceded the development of art history as an academic discipline--a discipline to which he was exposed while a university student as well as through personal contact with members of the Expressionist avant-garde.

If the self-consciousness of Nosferatu is based on the generation to which Murnau belongs-the men and the women who lived through the transition from the nineteenth to the twentieth century, from the achievements of Romanticism to the displacements of Expressionism-likewise Utamaro owes its ability to reflect on the nature of the cinema to the timing of its creation during a traumatic historical moment: Japan's defeat in World War II and the subsequent occupation by American forces. Utamaro comments on the nature of the cinema in relation to the arts of woodblock printing and tattooing by overlaying definitions of artistic creativity with issues of national and religious identity.

Like the theme of painting as tradition and creativity, the battle of the sexes--or, if you will, the social construction of gender roles--is a preoccupation that informs all my case studies. Put another way, my analyses assume that artistic production in the cinema entertains a dialogue not only with mechanical but also with sexual reproduction. Most importantly my attention to the iconography of sexual difference is meant only to enhance our understanding of the power relations between cinema and painting, and is not an attempt to say something new about male and female identities on the screen.

Some readers may wonder if my intensive intertextual approach risks weakening my interpretation of individual films. My answer to this legitimate concern is that it is the film in all its aspects--genre, production history, reception, historical context, auteur--that channels, contains, and gives meaning to intertextual citations. While I acknowledge the art historical point of view, I am writing about what happens on the screen. The cinema is the protagonist of my book, and art history is considered only in relation to it. Film, in short, has the last word.

Besides relying on intertextuality, all together my chapters raise questions about influence and intentionality--two categories that have been traditionally hard to tackle because, like architecture, filmmaking is an industrial art characterized by collaborative authorship.

Influence and intentionality, for instance, are hard to prove in An American in Paris. Minnelli's primitivist notion that creativity involves psychological upheaval has affinities with Jackson Pollock's equivalence of painting with dancing. Both the brilliant director of Hollywood musicals and the painter with a masculine, aggressive public image sought a renewal of the self through a fusion of the arts. We have no record that either artist made statements about the other, but we can still account for the weakness of Minnelli's happy ending for An American in Paris by recalling the popular perception of Pollock's visceral approach to art-making. This is to say, in the fifties Pollock's laconic, intense persona must have pushed the boundaries of heterosexuality to the point where it threatened to slip into homoeroticism. Even if Hollywood appropriated Pollock's charisma to construct stars like Marlon Brando and Montgomery Clift, it certainly expected Minnelli to trade off the exhilaration of painting in Paris for a weak heterosexual romance with a rather insipid French girl. Thus, in my analysis of An American in Paris, I cannot speak of direct influence or conscious intentionality but only of meaningful cultural constraints or affinities, which the parallel of Pollock and Minnelli helps to illuminate, regardless of any recorded contact between the two.

Mizoguchi's Five Women around Utamaro is another extreme example, in which neither direct influence nor conscious intentionality seems applicable. Because Five Women around Utamaro is about the life of the artist, the connection between him and Mizoguchi is so obvious that influence and intentionality can exist only in their most excessive form: total identification. It is as if the Japanese filmmaker were telling a story about his own life and his own relationship to art. As a result of this overpowering congruence of the artist in the film and the artist making the film, we feel free to look outside the text, into the shared folklore of the national culture.

When I was at work on Murnau's use of art history in Nosferatu, I was able to reinforce my speculations on influence and intentionality with documents located in the archives of the University of Heidelberg; however, in my readings of Rohmer's Marquise of O, Antomoni's Red Desert, Godard's Pierrot le Fou, Tarkovsky's Andrei Rublev, and Cavalier's Thérèse, I have availed myself of the directors' published statements on painting and the arts in generalwhile always recognizing that the content of interviews is either fragmentary or implicated in the marketing process. Hence I read between the lines of my sources and often bring to the discussion art historical references that the films themselves, regardless of the directors' expressed intentions, seem to integrate into their textual space.

My book is structured like a necklace upon which eight different beads, so to speak, are strung on a thematic-intertextual red thread. Hence I must comment on the order of my chapters and on my choice of films. I start with Minnelli's American in Paris and follow it with Antonioni's Red Desert because by setting them next to each other I want to suggest how much common territory can be found between a European "art film" and an MGM Hollywood musical. This order makes it possible to see the European art film as a special genre that only deviates from or simply alters, but does not subvert, the Hollywood mode. Antonioni's reaching out to painting in film is as unsettling as Minnelli's. Whether creativity is set in the rigid context of the American industry or in the looser European milieu, it always and inevitably destabilizes male identity. An American in Paris and Red Desert both qualify as art films, but I have not constructed my case studies to emphasize the parallels between them. Each is conceived as a self-contained unit. Too many cross-references and comparisons would cloud the richness of the images, which are already multifaceted and have many levels of meaning.

Toward the end of this book, I discuss Mizoguchi's Five Women around Utamaro in a chapter that further develops issues of gendered identity and artistic creativity and thus returns me to where I started, but with this difference: the whole discussion has been based on the elision, not the exhibition, of art historical citations. Mizoguchi's disinclination to cite Utamaro's (or anyone else's) artwork, even as he develops a theoretical position on the nature of the filmic image, is both an effective ending point for the book and a way of reopening all the questions it raises.

After my discussion of An American in Paris come three chapters on Antonioni's Red Desert, Eric Rohmer's The Marquise of O, and Jean-Luc Godard's Pierrot le Fou, which I have grouped together because in these three directors' work, the dialectic of word and image characterizes the encounter of cinema and painting. In Red Desert, the words are by far less daring than the pictorial images, yet the female protagonist's visual sensitivity is limited by the alignment of her speech with a masculine perspective. While in Red Desert word and image are as difficult to reconcile as man and woman, for Rohmer, in The Marquise of O, literature and painting are categories that can be rearranged or broken down by attributing a plastic edge to words and by endowing images with an introspective aura. In Rohmer's film, the images that achieve the nuances of written speech and the use of language as an object of sensuous contemplation contribute to a steady, predictable oxymoronic logic; by contrast, with Godard the translation of the verbal into the visual and the visual into the verbal has a compulsive quality that well conveys the random proliferation of competing, unstable signs.

The dialectic of word and image and the playing out of artistic creativity within the scenario of sexual difference are two themes that are common to Antonioni's Red Desert and Jean-Luc Godard's Pierrot le Fou, a breathless philosophical essay on how collage works and what it can tell us about subjectivity in modern life. By adopting collage as his medium against painting, which he sees as obsolete, Godard assumes an iconoclastic, avant-garde stance that clearly contrasts with the conservative orientation and iconophilic project of Tarkovsky's Andrei Rublev. In this film, the director aims to use the cinema to restore the emotional power of religious icon painting in our modern world, which, he feels, has lost its spiritual values.

While in the films examined so far the encounter of cinema and painting redefines gender roles, expands the dialectic of word and image, and shows what creativity means to different filmmakers, in chapter 6 and chapter 7--on Murnau's Nosferatu and Mizoguchi's Five Women around Utamaro respectively--I dwell on the tension between art and technology. As a character, Nosferatu the vampire stands for Murnau's ambivalence about his historical position between cinema and painting. In a similar way, Mizoguchi situates his medium between the competing poles of mass appeal and forbidden artistry, mechanical reproduction and manual labor, that are respectively occupied by woodblock printing and tattooing--a tactic that also enables him to express his divided allegiance between East and West as well as his struggle with the cinema, which he views as being simultaneously a form of patriarchal oppression and a liberating force.

My book's final chapter is devoted to Thérèse because this film allows me to discuss what the cinema, instead of only borrowing or repressing paintings, can teach art history about itself. If I wanted to retitle the book so as to describe my sense of the encounter between cinema and painting, I would choose something like The Magic Mirror or Screen as Arabesque, for cinema does not merely reflect back to art history the image it received, but through intertextual activity it rearranges all its outlines, boundaries, and priorities. Thus I am offering my readers a brand-new, cinematic lens through which to view the older and more established discipline of art history.

Having explained the sequencing and the key issues of my chapters, I should now comment on my choice of films. Most readers will wonder why I have not included a chapter on Peter Greenaway, since this director's work is heavily concerned with the relation between film and the other visual arts, the dialectic of word and image, the tension between high and low culture, and the parallel between artistic and sexual activity. My answer is that precisely because a single Greenaway film, The Belly of an Architect (1987), explores all the issues I raise in my book, I have chosen to deal with other directors in order to spread my inquiry across many different personalities, instead of collapsing everything into one case study. To those who want to know what place Greenaway might have had in this book, I will say that he would have served as a summary of my first seven chapters, for The Belly of an Architect relies on a type of intertextuality in which film is more preoccupied with defining itself than with redefining art history.

My choice of films also reflects my attempt to introduce into film studies an awareness of different genres of painting and their relevance to the cinema. For example, the chapter on Murnau's Nosferatu is meant to make the reader think about the links between Romanticism and Expressionism, between landscape painting and subjectivity; Cavalier's Thérèse, a film I greatly admire, encourages us to pay attention to the connection of still life and the close-up, particularly in the way both art forms use domestic objects to explore gender roles. Furthermore, were we to compare these two films in a more direct fashion, we would quickly realize that they both draw upon art historical traditions, intent upon challenging what was formerly considered the highest and most central of genres: history painting.

My chapter on Godard and collage addresses the end of painting as art, and Pierrot le Fou is a film about the impossibility of portraiture in modern life, because this genre stands for a belief in a coherent and unified subjectivity. Besides Pierrot le Fou, two more films in my book shy away from portraiture: Tarkovsky's Andrei Rublev and Mizoguchi's Five Women around Utamaro. Yet these directors' rejection of portraiture is not based on a modernist crisis but rather on religious and social philosophies that do not equate individualism with mastery or physiognomy with identity.

Finally, my choice of films has also been determined by availability on video, and this is why I have not been able even to consider Raoul Ruiz's Hypothesis of a Stolen Canvas (1978) as a candidate for a case study.

Needless to say, my eight case studies are not the first attempts to discuss the dialogue between cinema and painting. In Moving Pictures, Anne Hollander searches through the history of art for the kinds of paintings that anticipated what cinema later on did with movement. The cinema, Hollander argues, is a mass medium, yet its pictures move each one of us in a secret and highly individual manner, so that the paintings that best pave the way for this psychological impact are from Northern Europe, where Protestantism favored an intensely private apprehension of images.

While Hollander feels that the tradition of Northern European art is responsible for the birth of the moving image, I would argue that throughout the history of art, starting with the flickering shadows on the wall of Plato's cave, we can find innumerable instances of a protocinematic imagination. As soon as we acknowledge cinema's tendency to borrow from all sorts of art forms and to employ many different visual sources within the same film, we have to conclude that Hollander's belief in one single, direct genealogy risks narrowing the richness of the encounter between cinema and painting. Hollander's work would rather teach us about which kinds of paintings anticipated the cinema than about the various ways in which cinema addresses painting. It is precisely this second direction of inquiry--from film to painting instead of painting to film--which I set out to explore.

Because Jacques Aumont in L'oeil interminable has no interest in citing paintings in film, it should come as no surprise that I reach conclusions different from the French scholar's on at least three topics we both discuss: Cavalier's Thérèse, color in film (my example is Antonioni's Red Desert), and Godard's relation to painting (which I examine by using Pierrot le Fou). To begin with, Aumont sees Thérèse as a failure, because it would like to be "painting" when it cannot or should not be. My intertextual reading of Thérèse, however, shows that the film teaches something valuable about the destabilizing power of still life and the multifaceted nature of the close-up.

Although Aumont offers a convincing explanation of how film technology dictates that filmmakers and painters have to work with color in completely different ways, in my analysis of Red Desert I want to expand on the current belief that Antonioni uses color as a way of making "abstract painting" in film. I contend that he uses color to depict and analyze a scenario of death and rebirth in Red Desert. As a nonverbal element of the mise-en-scène, an element invested with a poignancy that well fits the polarities and upheavals inherent in the death of an old order and the birth of a new one, color in Red Desert makes up "the text of muteness" that Peter Brooks argues is typical of melodrama. In other words, the historical trauma and psychological rebirth in Red Desert are so extreme that words cannot express them; only the silence of colors can convey the intensity of these changes. Thus color is movement in a film made like an abstract painting, for it charts the interface of public and private shifts.

Besides making a case for Red Desert as a melodrama concerned with historical upheavals and psychological transformations, my chapter on Antonioni intends to explain that his allegiance to both painting and architecture is predicated on the battle of the sexes, and to argue that Antonioni's visual ventriloquism through his female protagonist, Monica Vitti, is both experimental and exploitative in that while he uses his actress's eyes to work innovatively with color in film, her power resides more in the eloquent muteness of the images she shares with the director than in anything she says.

Thus, what I gain by reading Red Desert intertextually and in a melodramatic key is an expansion of the themes commonly associated with color: not only painting and abstraction but also the power of silence and femininity. It is precisely this broader thematic understanding of what color means in Red Desert that, in contrast to Aumont's technological preoccupations, enables me to assess how innovative Antonioni's film truly is.

Instead of concentrating on surface similarities and notions like influence, Aumont argues for an ongoing dialogue, a continuing relationship between cinema and painting that plays itself out in the norms of visual representation, in the historical changes of the "eye." Thus, for Aumont, the relationship between painting and cinema is not one of influence and borrowing but of joint participation in culture. Yet, since the cinema is the lens I use to look at art history, Aumont's metaphor of an eye is somewhat at odds with his intense historical consciousness.

Aumont's notion of the eye as a continuous but always incomplete system is both a problematic and provocative formulation; in addition, it matches the French theorist's definition of Godard's cinema. Yet precisely because Aumont refuses to deal with citation, we never quite grasp what is at the heart of Godard's work. My chapter on Godard will argue that his oeuvre is deeply logocentric and pictorial while it also entertains abstractions. In order to improve our understanding of Godard, I have decided to do exactly the opposite of Aumont and deal with intertextuality in Pierrot le Fou. First of all, my analysis of Godard's sources discloses that even if his cinema is visually stunning, Pierrot le Fou is built like a collage of words and images, graphics and colors, lines and volumes. All these oppositions, in turn, signal the director's preoccupation with sexual difference, a barrier he utilizes for his creative purposes but which he would also like to overcome for emotional reasons.

The fantasy at the heart of Pierrot le Fou, then, is one of overcoming the boundary between man and woman, the verbal and the visual pole, by depicting something between these poles that is neither but could become both. In particular, Godard's "endless eye" is ongoing to the extent that the dialectic of word and image in collage can reoccur between any two other sets of opposites, yet it seems to me that it is less incomplete than Aumont allows, for it is animated by a logocentric impulse, even if the images of Godard's cinema remain most compelling.

Unlike Aumont's work, which celebrates Godard as painter, my analysis of Pierrot le Fou leads me to conclude that Godard's cinema is iconophobic. I see Godard as first setting up an equation between the spectacle of the female body and the allure of the pictorial image, and then seeking to abandon the visual register for the sake of an empty canvas, a blank sky, a colorless sea, or a black screen--voids where the mind, through language, can project all its abstractions. In other words, Godard avails himself of collage rather than painting, because collage's ability to thrive on the boundary between word and image reflects Godard's use of semiotic permutations in the attempt to overcome the pain of sexual difference.

It is possible that Aumont might have missed the collagist-iconophobic thrust of Godard's work because he has been concerned with the late films of Godard such as Passion (19982). I have decided not to analyze Passion since so much powerful criticism has recently been published on this film. In addition, focusing on Pierrot le Fou enabled me to link Godard with collage and to demonstrate his intense antipictorial vocation, even though the appeal of painting constantly resurfaces in his images. Finally, I feel that the splendid canvases brought to life in Passion, although they seem to support Aumont's labeling of Godard as painter, do not at all refute my view of Godard as a collagist. In fact, the opening of Passion--the image of an airplane whose wake produces a sort of writing or trace across the empty sky--is one more reminder that even this film, a work built around classical painting, is animated by a voiding of visuality in favor of language, writing, abstraction, absence.

It must be clear by now that while I support Aumont's project of thinking in terms of representation and not of art, of considering cultural strategies and not aesthetics, my disagreements with him about specific directors or films are quite substantial. One reason for this is the completely different scale of our approaches: I like working with stylistic detail and the single image; Aumont tackles big issues at a much more abstract level. If Anne Hollander's argument is too narrow because it deals only with the influence of Northern European painting on film, Aumont's view is too broad. My project is to carve out a space of close reading that is centrally located between these two extremes; using the single filmic text as a unit of measure, I hope to use close reading to achieve broad theoretical ends.

My method is somewhat closer to Pascal Bonitzer's Décadrages and to Jean-Louis Leutrat's Kaleidoscope, in that these studies on cinema and painting capitalize on citation and intertextuality without shying away from such larger issues as realism, referentiality, temporality, framing, construction of space, and spectatorship. Yet I differ from them as well. Like Aumont, Bonitzer and Leutrat organize the discussion around a series of theoretical issues, whereas I have structured my book around individual films and have surveyed many fewer paintings and films than they. My analysis is clearly more intensive than extensive. No individual analyst, however, can produce a definitive interpretation of a work of art, so my readings neither claim nor wish to exhaust intertextuality.

My discussion of the use of paintings on the screen sets many competing definitions of the cinema against each other. In contrast, Brigitte Peucker contends that the cinema is a machine of dismemberment, an interesting argument but one that cannot account for Rohmer's and Tarkovsky's well-known opposition to the scissors of Eisensteinian montage, Mizoguchi's uneasiness with the cuts of editing, Murnau's fascination with movement in real time, his efforts to preserve the intergrity of real space, and his view of editing as empathy.

The collusion of cinema and painting works more like a subtraction than an addition, for it leads outside vision into feeling, thinking, and abstraction. In Nosferatu and in Andrei Rublev, Murnau and Tarkovsky respec tively tell of longing for something divine or transcendent; Godard and Mizoguchi, uneasy with the female body as an analogue for the image, adopt an antivisual stance, for Rohmer and Godard what is visual should always become something else so that cinema may convey thoughts or transform itself into poetry. Finally, from the encounter of cinema and painting in the films of Antonioni, Minelli, and Cavalier, we learn about psychic change, emotional upheaval, and feminine subtlety. Thus the meeting of screen and canvas makes visible the invisible and favours absence over presence, mind over body.

Angela Dalle Vacche is Associate Professor of Art History and teaches in the film studies program at Yale University. She is the author of The Body in the Mirror: Shapes of History in Italian Cinema.

"In this ground-breaking study, the author shows how eight films variously define painting as an art. . . . For all the wide range of art references, the readings of the films are rigorous and sensitive as film criticism. This art historian knows film."