Article body

The boundless growth of apparatuses in our time corresponds to the equally extreme proliferation in processes of subjectification. This may produce the impression that in our time, the category of subjectivity is wavering and losing its consistency; but what is at stake, to be precise, is not an erasure or an overcoming, but rather a dissemination that pushes to the extreme the masquerade that has always accompanied every personal identity.

Agamben 2009 : 15

The conference panel on which this article is based reflected upon the continuing usefulness of the concept of cinematic apparatus within film and media studies[1]. As Francesco Casetti suggested in an advance email to panellists, it is perhaps the case that “the ‘new life’ of cinema [...] outside the film theatre breaks with three basic elements tied with apparatus : we do not have anymore an individual in a crowd, an immobility of the spectator, and a concentration on the film”. The diagnosis of this “break”, of course, was in reference to the current viewing of film on mobile and multiple devices under conditions of human movement. Casetti further asked how we might “reconfigure an apparatus based on privatization, mobility and multitasking?”[2]

In this essay, I will explore the fate of apparatus theory by pursuing three themes. One of these has to do with the development of apparatus theory in the decades before an experience of cinema “outside the film theatre” had become a significant concern of film studies. I shall draw attention here to what I see as the slowly achieved independence of spectatorship studies from apparatus theory within film studies in the English-speaking world. As I will argue, this movement was bound up with the effort within English-language scholarship to “sociologize” the spectator, to ground it more firmly in individual biographies and social identities. A second theme is the manner in which theories of the apparatus have labored to recuperate, within the concept of apparatus itself, a variability (of historical conditions and viewer dispositions) it was long thought to lack. Finally, I will show how the fate of apparatus theory in recent years has been bound up with the opening of film studies onto two significant terrains of cultural analysis : one concerned with the place of cinema within cities (the so-called “cinematic city” turn) and another setting film studies within the larger project of a visually-oriented art history (an “art historical turn”).

Sociologizing the Apparatus

“Apparatus” is one of those concepts in media studies which, like “gaze” or “network”, was subject to ongoing theoretical revision even before new media and digital devices arrived to demand that one rethink it in even more radical ways. Indeed, the unravelling of apparatus theory has played itself out at multiple levels and followed several directions within English-language film studies over a period lasting at least forty years. One way of describing this unravelling, I suggest, is as the slow detachment of questions of spectatorship from those of apparatus. If, in an early stage of film theory, the “variability” of the spectator (as gendered, for example) was conceptualized in order merely to refine or to “correct” canonical accounts of the cinematic apparatus, the spectator quickly became a relatively autonomous focus of film scholarship. This autonomy has partly coincided with a second feature of apparatus theory’s unravelling within English-language film studies, one we might describe as the sociologization of spectatorship. This sociologization replaced a purely formal notion of the subject with a concern for the variety of identities that spectators brought to the viewing situation. While this identitarian turn is widely recognized in film studies’ recounting of its disciplinary histories, its effect on the notion of the apparatus still merits scrutiny.

The detachment of spectatorship from apparatus came with successive critiques of the model implicit in the classic works on le dispositif, like those of Comolli and Narboni or Baudry. These critiques, as is well known, focused on the alleged determinism and invariability of an apparatus seen to install, lodged within itself, versions of subjectivity. The move to complexify is a familiar one within the history of theoretical models, but in the case of film studies the complexification of apparatus theory proved exceptionally transformative. Since the late 1970s, at least, spectatorship has been pluralized (as in claims that “there are different gazes for different identities”), rendered conflictual (through the idea that the act of spectatorship may struggle against the constraining effects of the apparatus or text) or subject to empirical study (as a result of persistent claims that we need more information on how real people watch films in specific social and historical contexts). At the same time, notions of the viewing subject as a punctual effect of the spectatorial situation have given way to an account of viewer-subjects as produced in long processes of identity-formation which occur at least partly outside of the cinematic institution.

It was not inevitable that those seeking to revise the foundational model of the cinematic apparatus would find the variability they wanted outside of the viewing situation, in the heterogeneous and socially-grounded identities that spectators brought to the viewing situation. That variability might have been elaborated simply within more expansive models of the psychic processes through which spectatorship and subjectivation take shape in the viewing of films. Indeed, the history of apparatus theory has been marked by moves which, while seeking to displace Lacanian models of vision and subjectivity (of the sort to be found in classic statements on the cinematic apparatus), nevertheless restrict themselves to the circumscribed viewing situation and retain a notion of the subject as formal-psychoanalytic entity. These models include those which have advanced alternate versions of psychoanalytic theory, like that concerned with object-relations and their vicissitudes. For the most part, though, the building of models which limit themselves to the viewing situation (which concentrate, in other words, on the subject as formal-psychoanalytic category) have become peripheral to English-language film studies.

Instead, I suggest, it is through the breaking of spectatorship from the apparatus, and through the development of a more socially-grounded account of the former, that English-language film studies cast off the apparatus as a significant feature of film theory. In her book Spectatorship : The Power of Looking On, Michele Aaron usefully offers her own list of the principal critiques of apparatus theory to have emerged since the 1970s. In Aaron’s account, apparatus theory came to suffer from what she calls “the three D’s”:

The issue of difference : how the gender of the spectator, as one example, impacts upon spectatorship. The issue of done-to-ness : how the problem of submission for the spectator needs to be re-considered with regard to post-liberation culture (of women, of gays, of ‘sex’) and the pleasures of submission. The issue of disavowal : how spectatorship as self-affirming or salve requires a denial of, and distance from, various implications that it, at the same time, depends upon.

Aaron 2007 : 23

These critiques would generate new forms of analysis. They included well-known studies of resistant reading (as in Pam Cook’s analysis of her own responses to images of suffering masculinity) (Cook 1982) and sophisticated accounts of polysemic texts open to multiple reading positions (as in Rockler 2001). More recently, film studies has followed literary studies (e.g., Love 2010) in their embrace of what Aaron calls the “self-affirming” effect that the experience of texts may have upon social identities. In some cases – as with those critiques which called for empirical studies of audiences – the result has been the emergence of important new sub-fields within the discipline, like reception studies. In these, we see broader conceptualizations of a cinematic institution replacing the seemingly more limited notion of an apparatus (e.g., Stacey 1995). What is clear is that, while a concern with spectatorship may have emerged out of debates over the apparatus, spectatorship has generated its own sets of questions.

The Living Being, the Apparatus and the Subject

The classic vision of the cinematic apparatus, of course, posited an immobile spectator caught in a fixed relationship to a screen. In this vision, we see the elements which Agamben has described in his account of the apparatus : “we have then two great classes : living beings (or substances) and apparatuses. And between these two, as a third class, subjects” (Agamben 2009 : 14). As we shall see, other versions of apparatus theory will break with this model by including subject and “living being” in the apparatus itself. Agamben will quickly move to characterize the relationship of living being to apparatus as a “relentless fight”, in which subjectivity is endlessly transformed. In doing so, he is offering us a way out of the simple determinism of which the classic model of the cinematic apparatus was accused.

Arguably, though, Agamben’s way out is not the one which has always prevailed within film theory. English-language film studies came to be more interested in the different sorts of pre-existent subjectivity brought to the viewing situation than in the ways in which subjectivity itself was produced and transformed within the act of viewing. If, for example, we recognize that the sexual identities or inclinations of spectators modify the act of viewing, we are left with the question of whether these identities or inclinations come carried by the “living being” or are produced (as forms of “subjectiviation”) in the interaction of this living being and the cinematic apparatus itself. Much work in film studies would come to presume (if not insist upon) the former : that subjectivities are formed in long and complex processes, through the inhabiting of bodies and the performance of social identities across individual biographies. While these subjectivities are brought to the viewing situation, and while they may enter into conflict with the imperatives of the apparatus (the technology-text of the film), they are not “produced” in that conflict. The “relentless fight” posited by Agamben as unfolding between human being and apparatus on the separate terrain of subjectivity is more often imagined, in English-language film studies, as one between embodied subjectivities and the operations of the cinematic apparatus. In the 1980s and 1990s, it might be argued, film theory in its English language versions clawed subjectivity back into the space of what Agamben called “living beings”.

A longstanding critique of simplistic Althusserian-Lacanianism stressed that theorists should stop confusing the subject with the empirical/biological being. In fact, much of the work of Anglo-American film studies, at least since the 1980s, has been rooted in the implicit argument that some melding of the two is necessary. Studies of spectatorship assumed some of their independence from apparatus theory when the social-biological being and its continuities were offered as ballast against a vision of subjectivity that saw it as produced in the moment of encounter with the apparatus. If, for Agamben, the “living being” struggles against the apparatus, in a conflict in which subjectivity is produced, English-language apparatus theory, as it slowly drifted away from its psychoanalytic roots, sought to locate more of the variability of spectatorship in relatively stable characteristics of viewer identity (such as those formed biographically in experiences of difference or subjugation), rather than in the conflictual moment of the viewing act itself.

In this way, we might say, Anglo-American film theory armed (or re-armed) the spectator, by letting her biographical formation ground a set of coherences or positions with which to confront the cinematic institution and its texts. Arguably, this has set in place one of the key points of divergence between that version of film theory and those practiced elsewhere, most notably in France. In one of his questions to panel members, Francesco Casetti asked us to comment on the different trajectories of apparatus theory in English-speaking and other (primarily European) bodies of work. In the Montréal context in which I work (and in which the IMPACT conference was held), this difference of trajectories is usually experienced as a distance between the identitarian concerns of English-language film/cultural studies and the formal/aesthetic preoccupations of Francophone colleagues. A more fundamental divergence has to do with the criteria applied to theoretical propositions in order to judge their political effectivity. In English language film studies from the 1980s onwards, one often encounters the conviction that a theoretical model which shores up the viewing subject against the constraining features of the apparatus would offer a more hopeful resource for political and identitarian transformation than one which casts the subject as the product of a dynamic in which the apparatus played a central role. Implicit, here, is the sense that a theoretical model which permits one to imagine moments of liberation and resistance in acts of film spectatorship will itself be more liberating and resistant than one which does not.

The openness of theoretical models to variability and resistance has not always gone hand in hand with a resistant politics, of course. Models of resistant spectatorship may just as easily serve as alibis for cultural texts or institutions looking to disguise their ideological work behind the façade of spectator empowerment and an apparent openness of the text to oppositional readings. Nor have closed, determinist models necessarily served to block a resistant politics (though they may force us to relocate them.) In the current moment, for example, amidst the widely diagnosed corporatization and security state co-optation of digital communications, we find radically progressive models of media consumption that invoke the more tightly circumscribed and deterministic relationships proposed by Agamben in his model of the apparatus. Resistance, for those advancing such models, is directed towards the apparatus itself and the regimes that sustain it, rather than towards trying to build resistance in spaces of subjectivation opened up in practices of reception or spectatorship.

Outside of the English-speaking world, and in French film analysis in particular, one finds a more persistent enterprise of renewing and revising the models of spectatorship proposed in the founding texts on the cinematic apparatus. In the psychoanalytically-inflected work of Barbara Le Maître, for example, we find an innovative description of spectatorship which is limited to the relationship between subject-viewer and apparatus-text but offers a richly expanded model of the ways in which, in successive moments, the desiring subject and the text itself generate images in relation to mechanisms of desire (Le Maître 2004 : 33). In French writings on film as different from each other as those of Jacques Rancière (2011) and Marie-Josée Mondzain (2011), one finds a continued commitment to locating the politics of film within the operations of the film-text itself, in the configurations of knowledge, sensation and social discourse which it is able to produce and transform.

Dispositif and Apparatus

It has become a truism of film studies that every film is seen under conditions shaped by the historical context of its exhibition and the predispositions brought by spectators to the viewing act. One of the key questions posed to apparatus theory is the extent to which this context and its predispositions are part of the apparatus itself or function as external causes of its variable effectivity. Must we include, within a model of the apparatus, the historical sensibilities which inflect its functioning, or define the apparatus more narrowly, as a set of technological/institutional arrangements with which these historical sensibilities interact? This returns us to the question, raised earlier, of whether the viewer/subject and its predispositions are part of an apparatus or constitute the elements with which an apparatus is confronted. Agamben, following Foucault, will describe as apparatus the network established between various elements : “discourses, institutions, buildings, laws, police measures, philosophical propositions, and so on” (3). A little later, the apparatus is defined as “literally anything that has in some way the capacity to capture, orient, determine, intercept, model, control, or secure the gestures, behaviours, opinions or discourses of living beings” (Agamben 2009 : 3, 14). There is a slight uncertainty here as to whether those things which are captured, secured or oriented may still be seen as external, in a first instance, to the apparatus itself, or whether they are constituent parts of the apparatus-network described a few pages earlier. Is the apparatus-network that entity within which the “gestures, behaviors, opinions or discourses of living beings” are determined (one of Agamben’s options)? Or, rather, is the apparatus the logic of their arrangement and interconnection, of the manner in which they are disposed?

It is impossible to pursue these questions further without confronting the difficulties of translation which haunt apparatus theory in English. While, notoriously, the English word “apparatus” is used to translate both “appareil” and “dispositif” from the French, and while even French usage sometimes obscures differences between the two, it seems clear that each of the French terms operates on a different scale. (For a lengthy discussion of these problems of translation, see Buhler 2013 : 385). These differences become particularly acute when it is a question of the deterministic role played by technology in the production of subjectivities.

The distinction between appareil and dispositif is examined at length in a useful text on the apparatus, Bernard Vouilloux’ “Du dispositif” (2008). Vouilloux offers, as illustration, the technological and spatial systems established after 1870 by Alphonse Bertillon, the Parisian police criminologist, for taking photographs of prisoners. In order to ensure the conformity of angles, shading and other elements from one photograph to another, the heads of prisoner were constrained by technological means and the distance of the camera from the subject was stabilized through the interconnection of all components of the system. The aim of these arrangements were to ensure that, irrespective of who was taking the picture or being photographed, the repeatability of key elements of the photographic situation would ensure the direct comparability of all images.

If the apparatus here is the collection of technological-mechanical elements (the camera, the stand on which the prisoner’s head rests, etc.) which ensure standardization from one photo to the next, the dispositif is the set of protocols which govern the interconnected operation of these elements for a pre-determined end. (Bertillon himself, Vouilloux notes, uses the term dispositif to describe the larger set of arrangements operative in his picture-taking system.) The dispositif is not the physical arrangement of technologies, but the name we give to the regulatory operations that govern the relationship of all elements in a situation like that of Bertillon’s criminal photography laboratory. The disposition of the different parts of the appareil (camera, head-stand, etc.), their arrangement relative to each other, and the conditions and relationships they establish, give us a dispositif precisely as the effect of this disposition. Vouilloux goes on :

A dispositif , then, cannot be reduced to the internal arrangement [agencement] of elements – that is to say, it should be seen in its structure, and thus as a structure. Let us take an appareil, whether automatic or not; we see that it requires a subject who invents it, builds it and uses it; a material to which it is to be applied, a result to be produced, a recipient [destinataire] of the resulting product and, finally, acting upon both the sender and the receiver, a certain number of procedural rules, which themselves can give rise to singular, even deviant uses or to practices of variable normalcy and legitimacy. (...) A dispositif is an arrangement [agencement] that results from the investment and mobilization of means which are called upon to function towards a determined end.

Vouilloux 2008 : 17-18. My translation

We are close here to the broader arrangement of behaviours, dispositions, technologies and relationships that inhabit the most expanded definitions of the dispositif . There are affinities here with the Foucauldian sense of dispositif as Buhler describes it : that “formation” which is not simply discursive but, modelled on the concept of “discursive formation,” includes a variety of other elements (institutional, technological, spatial) (Buhler 2013 : 285). The dispositif, Vouilloux further specifies, is not an arrangement for simply imposing behaviours or other results; rather, it regulates the movement of forces which traverse it : “The dispositive captures and channels energy in order to redistribute it. The dispositive does not have the purity of structures; its form is never anything but the temporally and spatially determined disposition within which are stabilized the flows with which it deals” (Vouilloux 2008 : 20; my translation).

The distinction here, between an appareil (as a set of machinic elements) and a dispositif which regulates and names the disposition of the energies which flow between them, does not fully resolve the status of “living being” or “subjectivity” within Agamben’s account of the apparatus. Does the living being struggle with the appareil in the constitution of the dispositif or is the very status of this “living being” an effect of the process of disposition which Vouilloux describes? In other words, does the dispositif involve the production of subjects in relation to an appareil and its institutional, discursive and ideological contexts, or is it merely an endlessly regulated state of the conflict between these?

Of the many questions that might be pursued here, the most practical is perhaps that of the expansiveness of the notion of dispositif. To what extent might film studies recuperate, within the notion of the apparatus, the variable conditions and inclinations of spectatorship over time? If the dispositif is that regulating entity which binds together the different elements of any historically-situated practice, does the term prove useful in gathering up all the conjunctural elements of those practices, in a model which imposes a necessary unity upon them? Uricchio, describing the wave of scholarship on early cinema which followed the influential Brighton conference of 1978, suggests that dispositif was adopted precisely because it offered the promise of such unity :

Post-Brighton scholarship looked into production histories, stylistic trends, the period’s reception, and so on, effectively breaking with the teleological trends of the past by repositioning this body of films simultaneously as the culmination of various nineteenth-century representational efforts, and as a catalogue of unexpected possibilities for a yet-to-be disciplined medium. In this sense, it effectively embraced the notion of a media dispositif (a concept which links apparatus, the cultural imagination, and constructions of public).

Uricchio 2004 : 28-29

The strengths of this conception of dispositif are many, even if they are no longer controversial. Few would deny that the making and viewing of films occurs within complex structures and regulatory fields within which meanings, affect and ideological effects are produced in their historical contingency and specificity. However, the designation of these structures or fields as dispositifs does not resolve the problem of determinism, or where one locates processes of subjectivation, which haunted the earliest uses of apparatus theory within film studies. If any given moment in film history produces a variety of enunciative and spectatorial positions, are all these positions recuperable within the model of the dispositif, as outcomes of its complex and variable productivity? Or must we still map an outside to the dispositif – a space, for example, for the elaboration of spectatorial subjectivities that escape the regulatory operations of the dispositif or which anticipate emergent subjectivities that will outlive it? Post-Brighton film historiography broke clearly with the formal notion of viewing subject which was at the core of the foundational writings on the cinematic apparatus. The choice ever since has been between (a) an ever expanding model of the dispositif, one able to accommodate all the variability of identities which had been invoked as the basis of early critiques of apparatus theory; and (b) acknowledgement of the impossibility of this endless expansion, and recognition of the necessary fragmented character of film studies, divided now between studies of spectatorship, exhibition, form and style which transpire outside of any project of eventual unity.

Apparatus, Device and the Ethnography of Forms

If there has been a resurgent interest in recent film studies in the concept of the apparatus, it has perhaps come with the reinvigoration of media theory and its take-up within film studies. The pluralization of spectatorship in early critiques of apparatus theory has been partially displaced, at least, by a new interest in the diversity of spectatorial locations and technologies. This shift is signalled in the suggestion that the ontological question “What is cinema?” should give way to the question “Where is cinema?” (Dercon 2002). Typing “dispositif” and “appareil” into Google Translate, one gets “device” as the English equivalent for both, and while these translations may have long histories, it is hard to escape the sense that they are shaped by the current ubiquity of the mobile device. While “device” might have once seemed too small in scale and modest in ideological intent to capture the functioning of the cinematic appareil, that reduced scale now seems to express something of what has happened to the cinema. In his questions to the panel. Francesco Casetti asked whether the

“‘new life’ of cinema (not necessarily a “virtual life”) outside the film theatre breaks with three basic elements tied with apparatus : we do not have, anymore, an individual in a crowd, an immobility of the spectator, and a concentration on the film. How may we – if we may – reconfigure an apparatus based on privatization, mobility and multitasking?”.

cf. endnote 2

It is obvious, I think, that most of the features of the classical cinematic apparatus, as described above, are withering and fragmenting. What is not so clear is the appropriate analytic vantage point from which to register and respond to this withering and fragmentation. In another question, inseparable from that just cited, Casetti asked whether “the dismissal of the notion of apparatus implies a dismissal of the notion of ‘mode of vision’? Are problems tied with visuality and the gaze out-of-fashion today? My personal conviction is that visuality and vision will shrink as mobilizing questions within the study of film and media. If this occurs, it will have little to do with an ascendant interest in other senses or modes of subjectivation. Rather, it will occur because the encounter of a perceiving subject and a technologically-instantiated text will come to seem trivial and impoverished as a key feature in the life of cinema.

I have argued elsewhere that the most fruitful approaches within contemporary media analysis involve the articulation of two modes of analysis (Straw 2010). One of these is concerned with what Gaonkar and Povinelli have called the “ethnography of forms”, with the materiality of media as they move through social space. These authors ask “[w]hy is it that some forms move or are moved along? What limits are imposed on cultural forms as the condition of their circulation across various types of social space? What are the materialities of form that emerge from, and trace, these movements...?” (2003 : 385-97). Like much post-Kittlerian media theory, from which they nevertheless diverge in important ways, Gaonkar and Povinelli build on Foucault’s imperative to “think the outside”, to theorize exteriorities, to examine the edges of form, the surfaces of interconnection between cultural artefacts as the move through social space.

A second productive mode of analysis, I suggest, is rooted in an attention to circulation, to the movement of media forms through the social world and their attachment to distinctive surfaces and sites of consumption. If this sounds like a concern more appropriate to media than to cinema studies, I nevertheless see the most interesting renewals of film studies as those that follow film along its circulatory paths. Those pursuing film spectatorship into the crevices of public life may seek to track down ever more fleeting and ephemeral instances of that spectatorship, but it will be more interesting, I think, if we retreat from such instances in order to capture the broader patterns by which audio-visual texts insinuate themselves within our everyday environments and practices. Apparatus, appareil, and dispositif are everywhere in the contemporary world. However, the encounter with a textuality centred on that apparatus is no longer the most significant carrier of ideological effects or the most vital terrain for a struggle of meanings and interpretations. Film and media studies will be renewed if they elaborate their own practices of distant reading, in which audiovisual messages will mean, more than anything else, positions and linkages within social space.

With new media platforms, I suggest, the most interesting things happen at two levels. At the highest level of generality, we find infrastructure and system, studied from perspectives which have conventionally been those of political economy and “institutional analysis” but which now open onto analyses of public space, networked sociability and visual environments. At ground level, we find the busy movement of impulses and signals – movement which reveals, in diagrammatic form, the rhythms and trajectories of human or social action. Mid-level phenomena, like the semantic substance of communicative textuality, may easily wither as significant concerns in media analysis.

This unsettling of questions of visuality had already begun two decades ago, with the wave of research on early cinema in relation to urban life, and with debates over the so-called “modernity thesis” (e.g., Singer 2001). As is often acknowledged, the emphasis of this work on the fleeting, mobile visuality of the late 19th century has found at least tentative echoes in recent writing on the status of the image in the contemporary city. While he says little about device-based media, Alain Mons suggests that cinema strains more and more to capture the elusive, transversal and fragile forms of looking typical of present-day urban life (2004 : 114). In fact, it might be argued, the practices with which we carry, employ and attach ourselves to our everyday devices may be a stronger indicator of our contemporaneity than the forms of visuality these devices allow us to experience. With respect to new media-based arts, but with a broader applicability to audio-visual forms in general, Couchot and Hillaire argue that we experience technologies as pulling us toward a future, while works of media culture, labouring to inscribe themselves in real time, act as mechanisms of cultural delay or slowness. If true, this suggests a fractured temporality of the new audiovisual apparatus, one split between the anticipatory, technologically-mediated ways in which this apparatus occupies space and the residual demands of the audiovisual text itself for attention and contemplation through time (Couchot & Hilaire 2003 : 219.) Unlike the cinematic apparatus described by Comolli/Narboni or Baudry, then, the present-day audiovisual device might be an appareil which produces an ongoing decentering of the spectator/subject.


Finally, and very briefly, I want to take up the last of the questions which Professor Casetti’s posed to those participating in the panel on the apparatus. “[A]pparatus theory”, Professor Casetti suggested, “has been the last attempt to define a ‘specificity’ – even if in early film theories we have a number of scholars [who] define specificity from other points of view : historical, linguistic, anthropological. In a epoch of media convergence, [must we] get rid of any specificity, or must we rethink it in another way?” (cf. endnote 2).

What seems clear is that the specificity of cinema has been pulled apart at what we might call both “ends” of cinematic history. At the beginnings of this history, through writings dating at least as far back as Jonathan Crary’s Techniques of the Observer (1992), the origins of cinema have been dispersed across 19th century technologies of amusement and modernity. At the end that we inhabit, of course, the specificity of cinema has unravelled through the dispersion of cinematic or paracinematic spectacle across multiple media platforms and screen types. One interesting aspect of this dispersion, as noted, is how the wave of writings on cinema and the city has been central to this opening up of cinema at both “ends” of its history.

These separate unravellings have each opened film studies to new multidisciplinary configurations of knowledge and focus. One of these is what some have called the art-historical turn in film studies : the wave of writing on such phenomena as colour, ornament, landscape, paratexts and other features of film whose study has gone on for much longer in the study of the non-cinematic visual arts. This move is dispersive, setting film within broader regimes of taste, judgement and figuration. However, it is also reterritorializing in a sense, pulling back from technological mutation and recentering film studies within histories of aesthetic form. If the apparatus gave cinema a specificity in the founding moments of film theory, it has lost its weight amidst the new interest in styles and histories of figuration. The idea of any specificity to the cinematic apparatus sits uneasily in Rosalind Galt’s book on ornament or in much of the recent work on colour, but in both there are clear attempts to recast the cinematic in terms of a relationship to visual opulence or to fields of colour (Galt 2011; Street 2010). This work retains much of the longstanding interest in cinema as a machinery of vision. However, the political concerns implicit within this art-historical turn have to do with the functioning of collective regimes of taste and judgement rather than with the acts of spectatorship on which classical apparatus theory came to rest.

We may point to even more fundamental shifts in theoretical attitude in recent cultural theory. While the unfolding of French-derived theory since the 1960s organized itself around the psychoanalytic idea of split subjectivities, an idea that enshrined the absolute distinctiveness of the human, a variety of vitalisms and materialisms within recent cultural theory have dissolved the human within models of fluctuating matter. If these have transpired, mostly, outside of film theory, they are nevertheless eating at its edges and inspiring novel new ways of thinking about media and the audiovisual. They include the undulating, machinic assemblages described by Gaonkar and Povinelli (2003 : 391), the “nonhuman, thingly power” theorized by Jane Bennett (2010 : xiii) and, with specific reference to cinema, the “lived ecologies” described by Ivakhiv (2013). These theoretical moves may be resisted or embraced, but they have each moved far from the analysis of media spectatorship and the apparatuses that were seen to sustain it.