In 1987 I was working towards my PhD at Princeton and was surprised to discover that the Department of Philosophy was offering a course on the philosophy of film. I had no idea that such a subject existed. I was interested in the philosophy of art and had a passing interest in film, so I decided to sit in on the lectures. The course was a revelation: one could actually do philosophy about film and moreover do it in a way that was both intellectually rigorous and also acutely sensitive to the aesthetic qualities of individual films. The visiting professor who taught that course was George Wilson, who has been a friend and something of a mentor ever since. My greatest intellectual debts in the philosophy of film are to him. Had he not taught that course, this book would probably never have been written.

Having been inspired by George's course, I attended several film courses run by P. Adams Sitney and Tony Pipolo at Princeton. I sat well back in a large lecture theatre, and I suspect that they never knew I was there. But their lectures showed me, along with George's wonderful interpretations of films, how powerful and interesting films could be and that films were capable of far greater depth than I had previously imagined. My debts to these two scholars are considerable.

This book addresses some central issues in the philosophy of cinema: the role of expression, realism, authorship, theories of interpretation, the nature of narration, character identification and audiences' emotional responses. In developing theories of these phenomena, two broad themes emerge. The first is a concern with cinema as an art. The second is an argument that the cinematic medium plays a role in explaining and evaluating central features of cinematic works. In both respects, the book reveals a strong debt to classical film theory, which was concerned with the question of what makes film an art, and argued that the nature of the film medium plays a central role in understanding and evaluating films. Contemporary film theory lost interest in the question of whether film is an art and in some of its modes was little concerned with the nature of the film medium, assimilating it instead to semiotic phenomena. And some contemporary philosophers of film, notably Noël Carroll, have argued at length both that there is no role for medium-specific explanations in film and also that, partly as a result of this, we should abandon the attempt to construct grand film theory, and instead adopt piecemeal theorising. If the argument of this book is successful, the classical film theorists were much closer to the truth in holding that medium-specific explanations and evaluations, as well as a good degree of systematising theory about cinema, are possible.

The scope of the discussion of cinema in this book is broad.

We saw in the previous chapter that it is in part the plasticity of the representational medium that gives film its expressive potential. This raises the question of how to characterise the representational capacities of film. There are two kinds of answers that have been given to this question. The first holds that film is a kind of language. This answer has been very influential in film studies, and will be criticised for traditional cinema in the first section below and for digital cinema in the second section. The second answer holds that film is a pictorial medium and that pictures are not language-like. This will be the answer returned in this chapter. It follows that we should pose the question of the nature of cinematic realism at least partly in pictorial terms. I consider seven kinds of realism in cinema, concentrating in particular on the view that pictorial images can be transparent.

FILM AS A LANGUAGE

Much in common thought about film supports the idea that film is a kind of language. We speak of the language of film, of film as text, of the development of new languages of film and of reading a film. The idea of film as a language has also received much (though not universal) support in film theory. The Soviet filmmakers and theorists Lev Kuleshov and Vsevelod Pudovkin held that the shot played the role of a word, and the edited sequence of shots the role of a sentence.

Having discussed cinematic authorship, the question arises of how one is to understand films, the products of authorial actions. I begin by discussing the most influential theory of interpretation, intentionalism, and criticise it particularly in respect of its account of collaborative arts, such as cinema. In Section 4.2 I discuss one of the best and most influential theories of interpretation developed by a film theorist, David Bordwell, and show that his global constructivist account should be rejected, but argue that there is a limited role for construction in some films. In Section 4.3 I defend my own account of interpretation, the patchwork theory, in the context of cinema and illustrate it with a discussion of Rashomon (1950). Finally, in Section 4.4 I argue that intentions are likely to be subject to some different defeaters in digital as compared to traditional cinema, and show how interactivity makes possible a new kind of constructivism in cinema.

INTENTIONALISM

I will argue that intentionalism as a theory of interpretation of collaborative art forms, such as cinema, is false. I will also briefly argue that intentionalism fails as a theory of interpretation of art in general. But I will focus mainly on the collaborative case and show that collaborative art forms, compared to solo (non-collaborative) forms, present extra hazards that undermine the artist's intentions and provide extra grounds for unintended but meaningful features of works.

A collaborative artwork I will understand as one in which two or more artists interact to produce the work.

The nature of cinematic narration is a central topic in the philosophy of cinema, and in particular for the project of the present book. Almost all films have a narrative (story) and therefore have a narration (they convey a story). This is true not only of fiction films; most documentaries are also narrative films; and narrative is central to determining viewers' responses to films. Moreover, a central question of this book concerns how the nature of the cinematic medium conditions cinema as an art. So we will examine the similarities and differences between film's narrational capacities and those of other arts, and in so doing will shed light on the nature of cinema and on how it differs from the other arts. Narration is a trans-medium capacity: many works in media besides cinema narrate – narrative works include some dances, musical works and paintings, and almost all comic strips and literature. There has to be some degree of commonalty between these different media by virtue of the fact that they can all narrate, but there may also be interesting differences between them in respect of how they narrate, differences which throw light on their different capacities as media. I will argue that there are some salient differences between cinema and literature in respect of their narrational capacities, particularly in respect of the greater role for implicit narrators in the case of literature than of cinema; and I will trace this to differences between the different media.

There is no doubt of the emotional impact that films can have on their audiences. The nature of that impact varies immensely, ranging from the gut-churning fear produced by an effective horror film to the nuanced emotional landscape of the best films within the European art-cinema tradition. The emotional power of cinema is central to its appeal and value as an art form, and the question arises of how it is achieved. I begin by examining how filmmakers employ aspects of the cinematic medium to foster emotional engagement. I then turn to defend the coherence and the role of identification, one of the most important devices of emotional engagement in cinema, and also show how it produces emotional learning. The discussion in the first four sections chiefly concerns emotion and identification in traditional films. In the last section I examine what difference digital cinema, in particular interactive digital cinema, makes to the ways that we can emotionally engage with films.

EMOTION AND CINEMA

In discussing the emotional power of cinema, I make two assumptions. The first is of the truth of the cognitive-evaluative theory of the emotions. This theory holds that emotions are not simply feelings, as are moods, but that they have an intentional object: I am afraid of something, I hope that something will relieve me from the danger, I pity that man.

At many points in this book I have invoked the notions of medium-specificity and of a medium; and the notion has structured our discussion of many of the artistic features of cinema. It is now time to examine these notions in greater detail and defend them from their critics. Medium-specificity is a concept that has figured in the discussion of many art forms besides cinema, and it helps to defend its application to film if we draw on examples from other media, to show that its use in cinema is not a mere product of the idiosyncrasies of the history of film theory. So this final chapter will range more widely than a discussion of cinema and deploy examples of how medium-specificity applies in other art forms too.

I begin by briefly examining some of the historical background to medium-specificity claims, particularly in the work of Gotthold Lessing. The second section distinguishes between three different versions of the medium-specificity claim. It also discusses and defends the notion of a medium and of differential properties, expanding on the brief discussion of these notions in earlier chapters of the book. The next three sections argue for the three versions of the medium-specificity claims, relating them back to the discussion in earlier chapters of the book. I conclude with a brief summary of the main claims of the book, highlighting how cinematic art is grounded on the features of the cinematic medium.

This book is about cinematic art. If there is an art form, there must, it seems, be artists working in that art form. If that is so, what can we learn about cinematic art by studying the cinematic artist, the so-called ‘cinematic author’? This chapter investigates that question, arguing that cinematic authors play a role in the understanding and evaluation of films and that all traditional films made by more than one person in the key production roles are multiply authored. I then argue that the multiple-authorship thesis also applies to digital cinema, and that the latter medium enhances possibilities for collaboration between different film artists. Finally, I show how multiple authorship is also true of interactive digital cinema, but argue that the audience are not among the authors of interactive works, though they do count as co-authors of interactive works' instances.

TWO PUZZLES

The notion that certain films are authored is one of the most powerful and pervasive views in current thinking about cinema. The enthusiast who looks forward to the film she thinks of as the new ‘Scorsese’, ‘Allen’, ‘Rohmer’, or ‘Tarantino’ is paying homage to the idea of the director-as-author. Rooted in the writings of Truffaut and other French critics in the 1950s, the view was transplanted to the United States by Andrew Sarris in the early 1960s, and dubbed by him ‘the auteur theory’.

Filmmakers during the first two decades of the new medium's existence thought of themselves sometimes as scientists, sometimes as explorers, sometimes as entertainers, but hardly ever as artists, as pioneers of a new artistic medium. D. W. Griffith in his earliest Biograph films from 1908 did not dare put his name on the credits, lest his ambitions in the legitimate theatre be undermined by his low-life escapades with celluloid. Only gradually did he come to think of his films as works of art; and if Birth of a Nation (1915) and Intolerance (1916) look now like deeply problematic achievements, they did at least represent the self-conscious striving to make films that are art. Though filmmakers took some time to think of their activity as a kind of art, theorists took longer. Hugo Munsterberg's pioneering The Photoplay: A Psychological Study (1916) represents the first sustained attempt to defend film as an art. But even in 1933 Rudolf Arnheim in his Film (revised as Film as Art in 1957) could think that the leading issue in film theory was whether film is an art. Indeed, Arnheim famously defended silent film as an art, but looked with some apprehension at the newly invented sound film, which he thought of as a threat to the artistic status of cinema. A great deal of classical film theory, as we noted in the Introduction, was concerned with arguing that cinema, despite its mechanical, photographic basis, is an art form.