An Elegy for Theory*

D. N. RODOWICK

OCTOBER 122, Fall 2007, pp. 91–109. © 2007 October Magazine, Ltd. and Massachusetts Institute of Technology.

From the late 1960s and throughout the 1970s, the institutionalization ofcinema studies in universities in North America and Europe became identifiedwith a certain idea of theory. This was less a “theory” in the abstract or natural sci-entific sense than an interdisciplinary commitment to concepts and methodsderived from literary semiology, Lacanian psychoanalysis, and AlthusserianMarxism, echoed in the broader influence of structuralism and post-structuralismon the humanities.

However, the evolution of cinema studies since the early 1980s has beenmarked both by a decentering of film with respect to media and visual studies andby a retreat from theory. No doubt this retreat had a number of salutary effects: areinvigoration of historical research, more sociologically rigorous reconceptualiza-tions of spectatorship and the film audience, and the placement of film in thebroader context of visual culture and electronic media. But not all of these inno-vations were equally welcome. In 1996, the Post-Theory debate was launched by

* This essay was originally prepared as a keynote lecture for the Framework conference on “TheFuture of Theory,” Oklahoma State University, Stillwater, November 3–4, 2006. I would like to thankBrian Price for his invitation and perceptive comments. I would also like to thank the participants atthe Radcliffe Exploratory Seminar on “Contesting Theory,” co-organized by Stanley Cavell, TomConley, and myself at the Radcliffe Institute for Advanced Study, May 4–5, 2007—including RichardAllen, Sally Banes, Dominique Bluher, Edward Branigan, Noël Carroll, Francesco Casetti, Joan Copjec,Meraj Dhir, Allyson Field, Philip Rosen, Vivian Sobchack, Malcolm Turvey, and Thomas Wartenberg—for their challenging discussions of these and other matters.

Éloge. n. m. (1580: lat. elogium, pris au sens gr.eulogia). 1. Discours pour célébrer qqn. ou qqch.Éloge funèbre, académique. Éloge d’un saint.

—Le Petit Robert

He sent thither his Theôry, or solemn legation forsacrifice, decked in the richest garments.

—George Grote, A History of Greece (1862)

92 OCTOBER

David Bordwell and Noël Carroll, who argued for the rejection of 1970s GrandTheory as incoherent. Equally suspicious of cultural and media studies, Bordwelland Carroll insisted on anchoring the discipline in film as an empirical objectsubject to investigations grounded in natural scientific methods. Almost simultane-ously, other philosophical challenges to theory came from film scholars influencedby analytic philosophy and the later philosophy of Ludwig Wittgenstein. Thesedebates emerged against the vexed backgrounds both of the culture wars of the1990s and the rise of identity politics and cultural studies.

Confusing “theory” with Theory, often lost in these debates is the acknowl-edgment that judgments advanced—in history, criticism, or philosophy—in theabsence of qualitative assessments of our epistemological commitments are ill-advised. To want to relinquish theory is more than a debate over epistemologicalstandards; it is a retreat from reflection on the ethical stances behind our styles ofknowing. In this respect, I want to argue not for a return to the 1970s concept oftheory, but rather for a vigorous debate on what should constitute a philosophyof the humanities critically and reflexively attentive in equal measure to its episte-mological and ethical commitments.

A brief look at the history of theory is no doubt useful for this project.Retrospectively, it is curious that early in the twentieth century film would becomeassociated with theory, rather than with aesthetics or the philosophy of art.Already in 1924, Béla Balázs argues in Der sichtbare Mensch for a film theory as thecompass of artistic development guided by the construction of concepts.1 The evo-cation of theory here is already representative of a nineteenth-century tendency inGerman philosophies of art to portray aesthetics as a Wissenschaft, comparable inmethod and epistemology to the natural sciences. From this moment forward, onewould rarely speak of film aesthetics or a philosophy of film, but rather, always, offilm theory.

“Theory,” however, has in the course of centuries been a highly variable con-cept. One finds the noble origins of theory in the Greek sense of theoria as viewing,speculation, or the contemplative life. For Plato it is the highest form of humanactivity; in Aristotle, the chief activity of the Prime Mover. For the Greeks, theorywas not only an activity, but also an ethos that associated love of wisdom with a styleof life or mode of existence.2

1. Béla Balázs, Der sichtbare Mensch (Frankfurt am Main: Suhrkamp, 2001). The original citation is:“Die Theorie ist, wenn auch nicht das Steuerruder, doch zumindest der Kompass einer Kunstentwicklung.Und erst wenn ihr euch einen Begriff von der guten Richtung gemacht habt, dürft ihr von Verirrungenreden. Diesen Begriff: die Theorie des Films, müsst ihr euch eben machen” (p. 12). Balázs does, however,associate this theory with a “film philosophy of art” (p. 1). 2. On the question of ethics as the will for a new mode of existence, see Pierre Hadot, What IsAncient Philosopy?, trans. Michael Chase (Cambridge, Mass.: Harvard University Press, 2002). An influ-ence on Michel Foucault’s later works on the “care of the self,” Hadot argues that the desire for a philo-sophical life is driven first by an ethical commitment or a series of existential choices involving theselection of a style of life where philosophical discourse is inseparable from a vision of the world andthe desire to belong to a community.

Bringing together thea [sight] and theoros [spectator], theory has often beenlinked to vision and spectacle. (Perhaps this is what Hegel meant in the Aestheticswhen he names sight as the most theoretical of the senses.) In Keywords, RaymondWilliams identifies four primary senses of the term emerging by the seventeenthcentury: spectacle, a contemplated sight, a scheme of ideas, and an explanatoryscheme. With its etymological link to theater, no doubt it was inevitable that theyoung medium of film should call for theory. However, although the persistenceof associating thought about film with theory might be attributed to the deriva-tions of the term from spectating and spectacle, a contemporary commonsensicalnotion follows from the last two meanings. Theories seek to explain, usually byproposing concepts, but in this they are often distinguished from doing or prac-tice. In this manner, Williams synthesizes “a scheme of ideas which explainspractice.”3 This is certainly the way in which someone like Balázs or SergeiEisenstein invoked the notion of theory.

In The Virtual Life of Film, I argue that one powerful consequence of the rapidemergence of electronic and digital media is that we can no longer take forgranted what “film” is—its ontological anchors have come ungrounded—and thuswe are compelled to revisit continually the question, What is cinema? Thisungroundedness is echoed in the conceptual history of contemporary film studiesby what I call the “metacritical attitude” recapitulated in cinema studies’ currentinterest both in excavating its own history and in reflexively examining what filmtheory is or has been. The reflexive attitude toward Theory began, perhaps, withmy own Crisis of Political Modernism and throughout the 1980s and ’90s manifesteditself in a variety of conflicting approaches: Carroll’s Philosophical Problems ofClassical Film Theory and Mystifying Movies, Bordwell’s Making Meaning, JudithMayne’s Cinema and Spectatorship, Richard Allen’s Projecting Illusions, Bordwell andCarroll’s Post-Theory: Reconstructing Film Studies, Allen and Murray Smith’s FilmTheory and Philosophy, Francesco Casetti’s Theories of Cinema, 1945–1995, Allen andMalcolm Turvey’s Wittgenstein, Theory and the Arts, and so on.4

In detaching “theory” as an object available for historical and theoreticalexamination, these books take three different approaches. Natural scientific mod-els inspire one approach, both philosophical and analytic, which posit that the

An Elegy for Theory 93

3. Raymond Williams, Keywords: A Vocabulary of Culture and Society (New York: Oxford UniversityPress, 1976), p. 267.4. See D. N. Rodowick, Crisis of Political Modernism: Criticism and Ideology in Contemporary Film Theory(1988; Berkeley and Los Angeles: University of California Press, 1994); Noël Carroll, PhilosophicalProblems of Classical Film Theory (Princeton, N.J.: Princeton University Press, 1988); Noël Carroll,Mystifying Movies (New York: Columbia University Press, 1998); David Bordwell, Making Meaning(Cambridge, Mass.: Harvard University Press, 1989); Judith Mayne, Cinema and Spectatorship (London:Routledge, 1993); Richard Allen, Projecting Illusions (Cambridge: Cambridge University Press, 1995);David Bordwell and Noël Carroll, eds., Post-Theory: Reconstructing Film Studies (Madison: University ofWisconsin Press, 1996); Richard Allen and Murray Smith, eds. Film Theory and Philosophy (Oxford:Clarendon Press, 1997); Francesco Casetti, Theories of Cinema, 1945–1995 (Austin: University of TexasPress, 1999); and Richard Allen and Malcolm Turvey, eds., Wittgenstein, Theory and the Arts (London:Routledge, 2001).

epistemological value of a well-constructed theory derives from a precise concep-tual framework defined in a limited range of postulates. This approach assumesthere is an ideal model from which all theories derive their epistemological value.Alternatively, Casetti’s approach is both historical and sociological. Agnostic withrespect to debates on epistemological value, it groups together statements madeby self-described practitioners of theory, describing both the internal features ofthose statements and their external contexts. In The Crisis of Political Modernism,my own approach, inspired by Michel Foucault’s Archaeology of Knowledge, assumesthat the conditioning of knowledge itself is historically variable. Discourse producesknowledge. Every theory is subtended by enunciative modalities that regulate theorder and dispersion of statements by engendering or making visible groups ofobjects, inventing concepts, defining positions of address, and organizing rhetori-cal strategies. This approach analyzes how knowledge is produced in delimitedand variable discursive contexts.

As a first move, it might indeed seem strange to associate theory with history.Introducing a series of lectures at the Institute for Historical Research at theUniversity of Vienna in 1998, I astonished a group of students by asserting that filmtheory has a history, indeed multiple histories. Here the analytic approach to the-ory, on one hand, and sociological and archaeological approaches on the other,part ways. The fact of having a history already distinguishes film theory, and indeedall aesthetic theory, from natural scientific inquiry, for natural and cultural phe-nomena do not have the same temporality. Aesthetic inquiry must be sensitive tothe variability and volatility of human culture and innovation; their epistemologiesderive from (uneven) consensus and self-examination of what we already know anddo in the execution of daily life. Examination of the natural world may presume ateleology where new data are accumulated and new hypotheses refined in model-ing processes for which, unlike human culture, we have no prior knowledge.

*

I believe we need a more precise conceptual picture of how film becameassociated with theory in the early twentieth century, and how ideas of theory varyin different historical periods and national contexts. But let us return to the morerecent, metacritical attitude toward theory.

By the mid-1990s, film theory and indeed the concept of “theory” itself werechallenged from a number of perspectives. This contestation occurs in three over-lapping phases. The first phase is marked by Bordwell’s call throughout the 1980sfor a “historical poetics” of film and culminates in the debates engendered by thepublication of Making Meaning: Inference and Rhetoric in the Interpretation of Cinemaand by the special issue of iris on “Cinema and Cognitive Psychology,” both pub-lished in 1989. The capstone of the second phase is the 1996 publication ofPost-Theory. Subtitled Reconstructing Film Studies, the book represents an attempt toestablish film studies as a discipline modeled on cognitivist science and historical

OCTOBER94

poetics, and to recenter “theory” according to the epistemological ideals of nat-ural scientific reasoning. If the second phase may be characterized by the attemptto return theory to a model of “scientific” investigation and explanation, the thirdphase subjects the association of theory with science to philosophical critique. Asfound in the recent work of Allen and Turvey, and deeply influenced by Wittgen-stein’s critique of theory in the Philosophical Investigations, this perspective calls fora new orientation in the examination of culture and the arts through a philoso-phy of the humanities. In this manner, throughout the 1980s and ’90s there is atriple displacement of theory—by history, science, and finally, philosophy.

It is important to appreciate Bordwell’s contribution to what I have charac-terized as the metacritical or metatheoretical attitude in cinema studies. Amonghis generation, Bordwell was among the first to exhibit fascination with the his-tory of film study itself, and to focus attention on problems of methodology withrespect to questions of historical research and the critical analysis of film formand style. Throughout the 1980s, Bordwell produced a number of path-breakingmethodological essays promoting a “historical poetics” of cinema. From Narrationand the Fiction Film (1985) to Making Meaning, the broad outlines of his approachare made apparent. Bordwell cannot be accused of a retreat from theory—noone’s commitment to good theory building is greater or more admirable.5 Instead,he wants to recast theory as history, or rather, to ground theory in the context ofempirical historical research. In this way, Bordwell responds to what he perceivesas the twin threats of cultural and media studies. On the one hand, there is a riskof methodological incoherence for a field whose interdisciplinary commitmentshad become too broad; on the other, the risk of diffusing, in the context of mediastudies, cinema studies’ fundamental ground—film as a formal object delimitingspecifiable effects. The aim of historical poetics, then, is to project a vision ofmethodological coherence onto a field of study perceived to be losing its center,and to restore an idea of film as a specifiable form to that center. In this respect,poetics concerns questions of form and style. It deals with concrete problems ofaesthetic practice and describes the specificity of film’s aesthetic function whilerecognizing the importance of social convention in what a culture may define as awork of art. In Narration and the Fiction Film, the historical side of poetics addressesthe proliferation of distinct modes of narration (classical Hollywood, Soviet ordialectical materialist, postwar European art cinema, etc.) as delimitable in timeand sensitive to national and/or cultural contexts. Here Bordwell makes his bestcase for basing the analysis of individual works upon sound historical investigationand explicit theoretical principles in a way that avoids arbitrary boundariesbetween history, analysis, and theory.

An Elegy for Theory 95

5. See especially Bordwell’s introduct ion to Cinema and Cognit ive Psychology, “A Case forCognitivism,” iris 5, no. 2 (1989), pp. 11–40. Here I am especially interested in Bordwell’s characteriza-tion of theory as “good naturalization.”

By 1989, however, Bordwell’s attack on interpretation and his promotion ofcognitivism as a model of “middle-level research” recast theory with respect tothree particular propositions. First, his appeal to middle-level research calls forpulling back from broader concerns of ideology and culture to refocus attentionon film’s intrinsic structure and functions. Second, he promotes a comparableturn from psychoanalytic theories of the subject to the study of filmic comprehen-sion as grounded in empirically delimitable mental and perceptual structures.Finally, his renewed emphasis on history also signals a withdrawal from high-levelconceptual concerns to refocus research on the fundamental data of films them-selves and the primary documentation generated from their production contexts.Thus, Bordwell accuses interpretation of reaching too high in grasping forabstract concepts to map semantically onto its object. Here the film-object itselfdisappears in its particularity, becoming little more than the example of a concept.Moreover, the interpreters are reflexively insensitive to the cognitive operationsthey execute. They produce no new knowledge, but rather only repetitively invokethe same heuristics to model different films.

The sometimes unruly responses to Making Meaning and Cinema andCognitive Psychology demonstrate that Bordwell’s criticisms touched a nerve, andthere is little doubt that these works are a genuine and important response to theimpasse in theory that cinema studies began to confront by the end of the 1980s.In the critique of so-called Grand Theory, what is most interesting here is theimplicit alliance between historical poetics and analytical philosophy. In the twointroductions to Post-Theory, Bordwell and Carroll promote strong views of whatcomprises good theory building in stark contrast to the then current state of con-temporary film and cultural theory. Here I am less concerned with assessing theircritique of contemporary film theory than in evaluating the epistemological idealsembodied in their common appeal to natural scientific models.6 Looking at thereverse side of Bordwell and Carroll’s criticisms, I think it is important to examinetheir ideal projection of “good theory” as the ethical appeal for a new mode of

OCTOBER96

6. Ironically, one consequence of this appeal, strongly implicit in Carroll’s contribution, is thatfilm theory does not yet exist. Carroll, for example, criticizes both classical and contemporary filmtheory according to three basic arguments: they are essentialist or foundationalist, taking films asexamples of a priori conditions; they are doctrine driven rather than data driven, meaning not sus-ceptible to empirical examination and verification; and finally, they deviate to widely from film -basedproblems, that is, the concrete particularity of filmic problems disappears when they are taken up toillustrate broader concepts of ideology, subjectivity, or culture. Characterized by “ordinary standards oftruth” as a regulative ideal, good theory seeks causal reasoning, deduces generalities by tracking regu-larities and the norm, is dialectical and requires maximally free and open debate, and, finally, is char-acterized by fallibilism. In this sense, good theory is “historical” in the sense of being open to revisionthrough the successive elimination of error. In this respect, middle-level research presents the provi-sional ground for a theory or theories of film projected forward in a teleology of debate, falsification,and revision. The “post” in Post-Theory is a curious misnomer, then. For what has been characterizedas Theory is epistemologically invalid, and, ironically, what comes after may only appear after a periodof long debate and revisionism. A legitimate film theory remains to be constructed, the product of anindefinite future.

existence where, in their view, politics or ideology has not supplanted reason.Here “dialectics,” as Carroll presents it, become the basis of an ideal research com-munity of rational agents working on common problems and data sets with resultsthat are falsifiable according to “ordinary standards” of truth and error.7 But theseideals, I would argue, rest on no firmer philosophical grounds than the ideologicaltheories they critique. For example, while Grand Theory is criticized for itsobsession with an irrational and unconscious subject that cannot account for itsactions, Bordwell promotes a “rational agent” theory of mental functioning,which is in fact the subject of good theory recognizing itself in the object it wantsto examine.8 The concept of the rational agent functions tautologically here as aprojection where the ideal scientific subject seeks the contours of its own image inthe model of mind it wishes to construct or to discover. In a perspective thatstrives to be free of ideological positioning and to assert an epistemology that isvalue-neutral, the introductions to Post-Theory nonetheless express the longing fora different world modeled on an idealized vision of scientific research: a commu-nity of researchers united by common epistemological standards who are strivingfor a universalizable and truthful picture of their object.

Richard Allen and Murray Smith’s critique of contemporary film theory inFilm Theory and Philosophy echoes Bordwell and Carroll’s perspective. AccusingTheory of an “epistemological atheism” powered by an exaggerated ethical con-cern with the critique of a capitalist modernity, Allen and Smith’s criticisms makeclear a number of philosophical assumptions absent from the Post-Theory cri-tique. From the analytic point of view, arguments for and against “theory” takeplace against the background of a philosophy of science. One engages in theorybuilding or not according to an epistemological ideal based on natural scientificmodels. In employing the methods and forms of scientific explanation, however,philosophy becomes indistinguishable from science, at least with respect to theoryconstruction. Philosophy disappears into science as “theory” becomes indistin-guishable from scientific methodology.

In this manner, I want to argue that from the beginning of the twentiethcentury analytic philosophy has been responsible for projecting an epistemologi-cal ideal of theory derived from natural scientific methods. This ideal produced adisjunction between philosophy’s ancient concern for balancing epistemologicalinquiry with ethical evaluation.9 Here, theory, at least as it is generally conceived

An Elegy for Theory 97

7. In a so-far-unpublished essay, “Film Theory and the Philosophy of Science,” Meraj Dhir has pre-sented an excellent defense of Carroll’s position.8. For related arguments, see Richard Allen’s essay, “Cognitive Film Theory,” in Wittgenstein, Theoryand the Arts, pp. 174–209.9. Bertrand Russell’s 1914 essay “On Scientific Method in Philosophy” presents a succinct defini-tion of this ideal: “A scientific philosophy such as I wish to recommend will be piecemeal and tentativelike other sciences; above all, it will be able to invent hypotheses which, even if they are not wholly true,will yet remain fruitful after the necessary corrections have been made. This possibility of successiveapproximations of the truth is, more than anything else, the source of the triumphs of science, and totransfer this possibility to philosophy is to ensure a progress in method whose importance it would be

in the humanties, disappears in two ways. Not only is the activity of theory givenover to science, but philosophy itself begins to lose its autonomy and self-identity—it would seem to have no epistemological function save in the light reflected fromscientific ideals. Analytic philosophy attacks theory on more than one front.There is the implicit tendency to delegitimate extant film theory to the extent thatit draws on concepts and methodologies influential in the humanities that falloutside of the reigning norm of what W. V. Quine would call a “naturalized philos-ophy.” Consequently, because so little aesthetic thought on film conforms toscientific models, Carroll concludes that, for the most part, a theory of film doesnot yet exist, though it might at some future date. The conflict over theory in filmstudies thus reproduces in microcosm a more consequential debate, one thatconcerns both the role of epistemology and epistemological critique in thehumanities and the place of philosophy with respect to science. Analytic philoso-phy wants to redeem “theory” for film by placing it in the context of a philosophyof science. At the same time, this implies that the epistemologies that were charac-teristic of the humanities for a number of decades are neither philosophically norscientifically legitimate. And so the contestation of theory becomes a de factoepistemological dismissal of the humanities.

Throughout the 1990s, then, in cinema studies philosophy allies itself withscience as a challenge to theory. In this phase of the debate, “theory” is the con-tested term. Very quickly, however, “science” becomes the contested term, as aphilosophy of the humanities gives over theory to science and opposes itself toboth. Important keys to this transition are the late works of Wittgenstein, especiallyhis Philosophical Investigations, as well as G. H. von Wright’s calls for a philosophy ofthe humanities in works like The Tree of Knowledge, and Other Essays (1993).

The interest of the later Wittgenstein for my project, and for the humanities ingeneral, concerns his attack on the identification of philosophy with science. Inasserting that “Philosophy is not one of the natural sciences” (Tractatus Logico-Philosophicus 4.111), he presents a formidable challenge to Bertrand Russell’sconception of philosophy as allied with epistemological models drawn from thenatural sciences. In contrast to Russell, Wittgenstein argues that science should notbe the only model of explanation and knowledge, and so he insists on the specificityof philosophy as a practice. It is important to examine carefully Wittgenstein’s attackon “theory” as an inappropriate form of explanation for the arts and humanities.However, my central concern here will be to explore arguments favoring a philoso-phy of the humanities as distinguishable from both science and theory.

OCTOBER98

almost impossible to exaggerate.” In Mysticism and Logic: and Other Essays (New York: Longmans, Greenand Co., 1918), p. 113 (my emphases). This is an admirably succinct summary of the epistemology towhich Carroll subscribes. Theories are built piecemeal out of preliminary and falsifiable hypotheses,and one must establish the factual character of the parts before the whole can be understood. Thetheory then advances teleologically as successively closer approximations to the truth as hypotheses arefurther tested, refined, or rejected in light of new evidence.

If philosophy involves another mode of explaining and knowing, why doesthe alternative not amount to a theory? As Allen and Turvey summarize in theirintroduction to Wittgenstein, Theory and the Arts, philosophy differs from science inthat its subject matter is not empirical in nature—only nature is subject to investi-gation by empirical methods. “Empirical” has a precise definition here as that ofwhich we can have no prior knowledge. Alternatively, philosophy is concernedwith problems of sense and meaning, and these problems are not empirical in thesense that language use and creative expression are already part of a commonlyaccessible stock of human knowledge.

This involves a second criterion: statements about empirical phenomenonare, and must be, necessarily falsifiable. Philosophical investigation, however, onlyconcerns testing the limits of sense and meaning of given propositions. In thisway, Wittgenstein’s case for philosophy as the best alternative to theory for study-ing human behavior and creativity is based on what he calls the “autonomy oflinguistic meaning.” This concept is exemplified in the distinction between rea-sons and causes. In a causal explanation, each effect is presumed to have a causeidentified by a hypothesis, which may and must be rejected or revised in light offurther evidence. Causal explanations are legitimate in scientific contexts becauseactions have origins that derive from states of affairs of which we have no priorknowledge. Most human action and behavior, however, is ill served by causalexplanation, for agents have the capacity to justify their behaviors with reasons.“Autonomy” now indicates that agents have the capacity for authoritative self-examination and self-justification. Therefore, a key difference between scientificand philosophical inquiry is that science tests its hypotheses against external phe-nomena, that is, the natural world. But philosophy admits only to internal orself-investigation. This is less a question of truth and error than judgments concern-ing the “rightness” of a proposition tested against prior experience and knowledge.

This is one way to begin to unravel the conceptual confusions surroundingthe idea of theory in cinema studies; for example, why Bordwell and Carroll havebeen so wedded to a certain idea of science, but also why theory, even from a cul-tural or psychoanalytic perspective, remains so compelling for a great many fairlyintelligent people. As Turvey puts the question, “Why is there a lack of basicempirical research in film theory if the nature and functions of cinema are likethe laws governing natural phenomena? Why does such research, somehow, seemunnecessary to film theorists? And how is it that film theories ever convince any-one that they are plausible in the absence of such sustained research?”10 Becausethese criteria are irrelevant for cultural investigation. Film theories, like all

An Elegy for Theory 99

10. Malcolm Turvey, “Can Science Help Film Theory?,” Journal of Moving Image Studies 1, no. 1(2001), http://www.uca.edu/org/ccsmi/journal/issue1_table_contents.htm. The passage reads differ-ently in the latest published version of the essay. See Turvey, “Can Scientific Models of TheorizingHelp Film Theory?,” in The Philosophy of Film: Introductory Texts and Readings, ed. Angela Curren andThomas E. Wartenberg (London: Blackwell, 2005), p. 25.

humanistic investigation, concern human activities and thus presume a highdegree of prior, even self-, knowledge and examination. Like any cultural activity,cinema is a human creation and thus is embedded in practices and institutionsthat form the basis of our quotidian existence. We may not have conscious knowl-edge of these practices and institutions, nor any desire to construct theories aboutthem in the form of propositions or concepts, yet we act on and through them incoherent and consistent ways. This is why cultural theories are able to solicit agree-ment in the absence of empirical research and experimentation. Their power andplausibility is based on the extent to which they seem to clarify for us what wealready know and do on a daily basis. Here we need no external examinationbeyond the critical investigation of our own practices as they evolve historically.However, what film studies has called theory, in its multiple and variegate guises,might more appropriately be called aesthetics or philosophy. And indeed, perhapswe could achieve much methodological and conceptual clarification by settingaside “theory” provisionally in order to examine what a philosophy of the humani-ties, and, indeed, what a film philosophy might look like.

*

I would prefer to title this essay Éloge de la théorie, for in composing an elegyfor theory I have kept in mind the subtle variations present in French. Combiningthe English sense of both eulogy and elegy, and something more besides, an élogecan be both praise song and funereal chant, panegyric and chanson d’adieu. (Inaddition, it conveys the second meaning of a legal judgment expressed in some-one’s favor.) Certainly I think the enterprise of theory is still a worthy one. Yetwhy, in contemporary critical discourse, are there so few left to praise and none tolove it?

We must first examine the debate on theory from the point of view of com-peting epistemological stakes. Accused of “epistemological atheism,” theory as aconcept has been wrested from the Continent to be returned semantically to theshores of science and the terrain of British and American analytical philosophy.Initially, this debate was posed as a conflict between theory and philosophy. Butthe late Wittgenstein took this argument in another direction, one that alsoquestioned theory but as a way of turning philosophy from science to restore it tothe humanities. In so doing, Wittgenstein was less concerned with the epistemo-logical perfectibility of philosophical language than with reclaiming philosophy’sancient task of theoria. If the politics and epistemology of theory have been subjectto much soul searching and epistemological critique, it is important nonethelessto find and retain in theory the distant echo of its connection to philosophy, or totheoria, as restoring an ethical dimension to epistemological self-examination. AsWittgenstein tried to teach us, what we need after theory is not science, but arenewed dialogue between philosophy and the humanities wherein both refash-ion themselves in original ways.

OCTOBER100

Ultimately, I want to argue that Wittgenstein’s attack on theory is both toobroad and too restrictive, but here it is more important to foreground what thelater Wittgenstein brings to a philosophy of the humanities. In liberating human-istic inquiry from the bonds of empirical and causal explanation, a philosophy ofhumanities may make propositional claims, but these claims need not be fallible—they only require suasion and clear, authoritative self-justification. This is becausehumanistic theories are culture-centered. Unlike the investigation of natural phe-nomena, philosophical investigations examine what human beings already knowand do, and this knowledge is in principle public and accessible to all. InBordwell’s sense of the term, “naturalization,” whether good or bad, has little rele-vance here as humanistic (self-) inquiry does not require finding new information,but rather only clarifying and evaluating what we already know and do, or knowhow to do, and understanding why it is of value to us. In its descriptive emphasis,Wittgenstein’s philosophical investigations do support strongly one importantaspect of historical poetics—the analysis of the internal norms of culturalobjects and of our everyday sense-making activities in relation to those objects.Nonetheless, a “nonempirical” notion of history is wanted here, and for specificphilosophical reasons. Natural laws are time-independent, at least in a humancontext, and thus are appropriately explored through falsifiable causal explana-tions. Alternatively, cultural knowledge is historical in a particular sense. Itemerges and evolves in the context of multiple, diverse, and conflicting socialinteractions that require constant reevaluation on a human time scale. Humanhistory and natural history may not be investigated by the same means, even if,with respect to certain problems, their domains may overlap. Unlike the scientist,the humanist must examine phenomena that may be shifting before her very eyes.She must account for change in the course of its becoming, while she herselfmight be in a process of self-transformation.

To what extent, then, is the enterprise of theory still possible? And howmight we return to philosophy the specificity of its activity? The two questions aredifferent yet related, and both are linked to the fate of humanities in the twenty-first century and the place of film in the future of the humanities. Possibleanswers begin in recognizing that epistemological atheism does not follow froman ethical critique of modernity. And indeed what links philosophy today to itsmost ancient origins are the intertwining projects of evaluating our styles of know-ing with the examination of our modes of existence and their possibilities oftransformation. I want to conclude by briefly exploring these questions in dis-cussing two contemporary philosophers as exemplars of the twinned projects ofethical and epistemological evaluation: Gilles Deleuze and Stanley Cavell. Deleuzeand Cavell are the two contemporary philosophers with the strongest commit-ment to cinema, yet with distinctly original conceptions of the specificity ofphilosophy and of philosophical expression in relation to film. Though anunlikely pairing, reading these two philosophers together can deepen and clarifytheir original contributions to our understanding of film and of contemporary

An Elegy for Theory 101

philosophy. Here I want to make the case that a (film) philosophy may and shouldbe distinguished from theory. At the same time, I want to distinguish for thehumanities a fluid metacritical space of epistemological and ethical self-examinationthat we may continue to call “theory” should we wish to do so.

Deleuze’s cinema books present two pairs of elements that show what a filmphilosophy might look like. These elements recur throughout Deleuze’s philo-sophical work. On one hand, there is the relation of Concept to Image. Here thecreation of Concepts defines the autonomy of philosophical activity, while theImage becomes the key to understanding subjectivity and our relation to theworld. The second set involves Deleuze’s original reconsideration of Nietzsche’spresentation of ethical activity as philosophical interpretation and evaluation.

Deleuze ends Cinema 2: The Time-Image with a curious plaint for theory.Already in 1985, he argues, theory had lost its pride of place in thought about cin-ema, seeming abstract and unrelated to practical creation. But theory is notseparate from the practice of cinema, for it is itself a practice or a constructivismof concepts.

For theory too is something which is made, no less than its object. . . .A theory of cinema is not “about” cinema, but about the concepts thatcinema gives rise to and what are themselves related to other conceptscorresponding to other practices. . . . The theory of cinema does notbear on the cinema, but on the concepts of cinema, which are no lesspractical, effective or existent than cinema itself. . . . Cinema’s con-cepts are not given in cinema. And yet they are cinema’s concepts, nottheories about cinema. So that there is always a time, midday-midnight,when we must no longer ask ourselves, “What is cinema?” but “What isphilosophy?” Cinema itself is a new practice of images and signs, whosetheory philosophy must produce as a conceptual practice.11

A slippage is obvious here with theory standing in for philosophy. But that beingsaid, what does Deleuze wish to imply in complaining that the contemporarymoment is weak with respect to creation and concepts? The most replete responsecomes from the most obvious successor to the problems raised in the cinemabooks—Deleuze and Félix Guattari’s What Is Philosophy?

For Deleuze and Guattari, the three great domains of human creation areart, philosophy, and science. These are relatively autonomous domains, each ofwhich involves acts of creation based on different modes of expression—percep-tual, conceptual, or functional. The problem confronted in What Is Philosophy? isknowing how philosophical expression differs from artistic or scientific expres-sion, yet remains in dialogue with them. Percepts, concepts, and functions aredifferent expressive modalities, and each may influence the other, but not in a way

OCTOBER102

11. Gilles Deleuze, Cinema 2: The Time-Image, trans. Hugh Tomlinson and Robert Galeta(Minneapolis: University of Minnesota Press, 1989), p. 280.

that affects the autonomy of their productive activity. An artist or scientist nodoubt profoundly engages in conceptual activity, and so is influenced by philoso-phy. Yet the outputs of that activity—percepts, functions—retain their autonomyand specificity.

From one perspective, the distinctiveness of these outputs is easy to explain.The aim of science is to create functions, of art to create sensuous aggregates, andof philosophy to create concepts, but the devil is in the details. In art, perceptsrefer to the creation of affective experience through constructions of sensuous mate-rials. In painting, these expressive materials may be blocks of lines/colors; in cinema,blocks of movements/durations/sounds. Alternatively, the role of functions helpsclarify the relation of philosophy to theory in the scientific sense. There is a function,Deleuze explains, as soon as two wholes are put into a fixed correspondence.Newton’s inverse square law provides an apposite example. A function is a mathe-matical expression orienting thought (first whole) to a natural phenomenon (thepropagation of energy). As expression, the function is not the specific phenomenon,of course, nor is it analogous to thinking. The function is a descriptor or algorithm.Its descriptiveness of behaviors in the natural world is important, but this is notthe key to its specificity. It is abstract and general, and its generality derives fromits time-independence. It produces descriptions, and these descriptions are validfor all times and all places—thus, the proposal of a second whole. In its predictive-ness of future behaviors, then, the function is exemplary of what science calls“theory,” and when this predictiveness becomes regular, functions become “laws.”

Contrariwise, the concept is abstract yet singular—it relates to thought in itsown temporality and human specificity. For these reasons, philosophy is much closerto art than it is to science. The expressiveness of art finds its instantiation in thesensuous products of art and its human affects, and the expressiveness of sciencefinds its confirmation in the predicted behaviors of natural phenomena. But con-cepts express only thought and acts of thinking. Does this mean that thinking ispurely an interior activity cut off from the sensuous and material world? Art providesimportant answers to this question in relating concepts to ideas, signs, and images.

In 1991, Deleuze gave an important lecture at FEMIS [École nationale supérieuredes métiers de l’image et du son], the French national film and television school, anexcerpt of which was published as “Having an Idea in Cinema.” What does it meanto have an Idea in art and how do Ideas differ from Concepts? Ideas are specific toa domain, a milieu, or a material. And so Deleuze writes, “Ideas must be treated aspotentials that are already engaged in this or that mode of expression and insepara-ble from it, so much so that I cannot say I have an idea in general. According tothe techniques that I know, I can have an idea in a given domain, an idea in cin-ema or rather an idea in philosophy.”12 Now, ideas in philosophy are already

An Elegy for Theory 103

12. Gilles Deleuze, “Having an Idea in Cinema,” trans. Eleanor Kaufman, in Deleuze and Guattari:New Mappings in Polit ics, Philosophy, and Culture, ed. Eleanor Kaufman and Kevin Jon Heller(Minneapolis: University of Minnesota Press, 1998), p. 14.

oriented by a certain kind of image, what Deleuze calls the “image of thought,”and so a connection or relation must link them. In What Is Philosophy? the imageof thought is defined as the specific terrain or plane of immanence from whichideas emerge as preconceptual expression, or as “the image thought gives itself ofwhat it means to think, to orient one’s self in thought.”13 To have an idea, then, isto express thought through particular constructions, combinations, or linkages—what Deleuze calls signs. As Spinoza insisted, signs are not an expression ofthought, but rather of our powers of thinking. Ideas are not separable from anautonomous sequence or sequencing of ideas in thought, what Spinoza calls con-catenatio. This concatenation of signs unites form and material, constitutingthought as a spiritual automaton whose potentia expresses our powers of thinking,action, or creation.

The importance of Deleuze’s cinema books is that they present his mostcomplete account of a philosophical semiotic modeled on movement and timeand show how images and signs in movement or time are conceptually innovative;that is, how they renew our powers of thinking. In this manner, art relates to phi-losophy in that images and signs involve preconceptual expression in the same waythat the image of thought involves a protoconceptual expression—they preparethe terrain for new concepts to emerge. The cinema may be best able to picturethought and to call for thinking because like thought its ideas are comprised ofmovements, both spatial and temporal, characterized by connections and con-junctions of particular kinds. Every instance of art is expressive of an idea whichimplies a concept, and what philosophy does with respect to art is to produce newconstructions or assemblages that express or give form to the concepts implied inart’s ideas. It renders perspicuous and in conceptual form the automatisms thatmake a necessity of art’s generative ideas.

There is also an ethical dimension to the various ways Deleuze characterizesimage and concept in relation to the image of thought. For Deleuze, this implies aNietzschean ethics encompassing two inseparable activities: interpretation andevaluation. “To interpret,” Deleuze writes, “is to determine the force which givessense to a thing. To evaluate is to determine the will to power which gives value toa thing.”14 What bridges Deleuze and Cavell here are not only their interest inNietzsche, but also their original concept of ontology. Though Cavell uses theword and Deleuze does not, both are evaluating a particular way of Being. This isnot the being or identity of film or what identifies film as art, but rather the waysof being that art provokes in us—or more deeply, how film and other forms of artexpress for us or return to us our past, current, and future states of being. In bothphilosophers, the ethical relation is inseparable from our relation to thought. For

OCTOBER104

13. Gilles Deleuze, What Is Philosophy?, trans. Hugh Tomlinson and Graham Burchell (New York:Columbia University Press, 1994), p. 37.14. Gilles Deleuze, Nietzsche and Philosophy, trans. Hugh Tomlinson (New York: Columbia UniversityPress, 1983), p. 54.

how we think, and whether we sustain a relation to thought or not, is bound upwith our modes of existence and our relations with others and to the world.

The key to grasping this relation in Deleuze is to understand the originalityof his characterization of the image as both an ontological and ethical concept.Especially in the cinema books, the image is not the product of cinematic creationbut rather its raw material, the worldly substance that it forms and to which itgives expression. Hence the key place of Henri Bergson’s assertion from Matterand Memory that there is already photography in things. Like energy, images canneither be created nor destroyed—they are a state of the universe, an asubjectiveuniversal perception or luminosity that evolves and varies continuously. Humanperception is therefore largely a process of subtraction. Because we must orientourselves in this vast regime of universal change according to our limited percep-tual context, we extract and form special images or perceptions according to ourphysiological limits and human needs. This image is the very form of our subjec-tivity and persists in the crossroads between our internal states and our externalrelations with the world.

The image is thus in relation with ourselves (interiority) and in relation withthe world (exteriority) in an intimately interactive way. It is absurd to refer to sub-jectivity as pure interiority as it is ceaselessly engaged with matter and with theworld. By the same token, thought is not interiority but our way of engaging withthe world, orienting ourselves there and creating from the materials it offers us.Thus, another way of considering the autonomy of art, philosophy, and science isto evaluate the different though related images of thought they offer us. The per-cept is visually and acoustically sensuous, provoking affects or emotions in us.Concepts and functions are more abstract. What the function is to scientificexpression, the sign is to aesthetic expression. Art’s relation to thought, then, liesnot in the substance of images, but in the logic of their combination and enchain-ment. No doubt every artistic image is an image of thought, a physical tracing andexpression of thought given sensual form, no matter how incoherent or inelegant.However, while the aesthetic sign may imply a precise concept, it is nonethelessentirely affective and preconceptual. Yet there is a philosophical power in images.The artist’s iIdea is not necessarily the philosopher’s. But images not only tracethoughts and produce affects; they may also provoke thinking or create new pow-ers of thinking. In so doing, we are thrown from sensuous to abstract thought,from an image of thought to a thought without image—this is the domain ofphilosophy. And in moving from one to the other, art may inspire philosophy togive form to a concept.

What does philosophy value in art? To ask this question is to demand whatforces expressed in art, in images and signs, call for thinking? Philosophy partsways with science to the extent that time is taken as an independent variable—infact, the simplest way of describing Deleuze’s (or Bergson’s) philosophical projectis as the will to reintroduce time and change to philosophy’s image of thought.Philosophy finds inspiration in art because there the will to create is brought to its

An Elegy for Theory 105

highest powers. Here, as in many other ways, Deleuze goes against the grain ofcontemporary philosophy. While happily science has never renounced its powersof creation, it has become less and less conceptual. And of course, it does notneed concepts as philosophy does. Contrariwise, philosophy has moved closer andcloser to art, and vice versa. This is the great untold story of twentieth-centuryphilosophy that the twenty-first century must recount: that philosophy’s greatestinnovations were not made with respect to science, but in dialogue with art. Andfurther, that the modern arts came closer and closer to philosophical expressionwhile nonetheless amplifying their aesthetic powers.

That art may be considered philosophical expression is an important linkbetween Deleuze and Cavell’s interest in film. Like Deleuze, Cavell’s cinema booksare not studies of film but rather philosophical studies—they are works of philoso-phy first and foremost. Nonetheless, it may also be reasonable to read them asstudies of film culture in their deep awareness of how cinema has penetrated thedaily life of the mind and of being in the twentieth century. Though in very differ-ent ways, both Deleuze and Cavell comprehend cinema as expressing ways ofbeing in the world and of relating to the world. In this respect, cinema is alreadyphilosophy, and a philosophy intimately connected to our everyday life. Deleuzeexemplifies this idea in pairing Bergson’s Matter and Memory with the early historyof cinema. At the moment when philosophy returns to problems of movementand time in relation to thought and the image, the cinematic apparatus emergesneither as an effect of these problems nor in analogy with them. In its own way, itis the aesthetic expression of current and persistent philosophical problems. Norshould one say that Deleuze’s thought is simply influenced by cinema. Rather, it isthe direct philosophical expression, in the form of concepts and typologies ofsigns, of problems presented preconceptually in aesthetic form.

Cavell presents a similar perspective, though one more clearly framed byproblems of ontology and ethics. In my view, Cavell’s work is exemplary of a phi-losophy of and for the humanities, particularly in his original attempt to balancethe concerns of epistemology and ethics. In this respect, two principal ideas uniteCavell’s philosophical and film work. Moreover, these are less separate ideas thaniterations of the same problem that succeed one another more or less chronologi-cally, namely, the philosophical confrontation with skepticism and the concept ofmoral perfectionism. The question here is why film is so important as the com-panion or exemplification of this confrontation. One clue resides in the title of animportant Cavell essay, “What Photography Calls Thinking.”15 What does it meanto say that art or images think, or that they respond to philosophical problems as away of thinking or a style of thought? In the first phase of Cavell’s film philosophy,represented by the period surrounding the publication of The World Viewed, theresponses to this question are ontological and epistemological. But this ontology

OCTOBER106

15. Stanley Cavell, “What Photography Calls Thinking,” in Cavell on Film, ed. William Rothman(Albany: State University of New York Press, 2005), pp. 115–34.

refers neither to the medium of art nor the identity of art works, but rather tohow art expresses our modes of existence or ways of being in the world as the fallinto and return from skepticism.

Here an ontology of film is less concerned with identifying the medium offilm than with understanding how our current ways of being in the world andrelating to it are “cinematic.” In its very conditions of presentation and percep-tion, cinema expresses a particular philosophical problem, that of skepticism andits overcoming. If, as Cavell argues, cinema presents “a moving image of skepti-cism,” it neither exemplifies nor is analogous to the skeptical attitude.16 Rather,cinema expresses both the problem and its possible overcomings. The quality of“movement” in this philosophical image is temporal or historical in a specificsense. In its very dispositif for viewing and encountering the world, cinema pre-sents philosophy’s historical dilemma (skepticism’s perceptual disjunction fromthe world) as past, while orienting the modern subject toward a possible future.That skepticism should reproduce itself in a technology for seeing might meanthat it is no longer the ontological air we breathe, but a passing phase of ourphilosophical culture. If, as Cavell argues, the reality that film holds before us isthat of our own perceptual condition, then it opens the possibility of once againbeing present to self or acknowledging how we may again become present to our-selves. (Indeed Cavell’s examination of cinema’s relation to the fate of skepticismhelps clarify a Deleuzian cinematic ethics as faith in this world and its possibilitiesfor change.17) For these reasons, film may already be the emblem of skepticism indecline. Cinema takes up where philosophy leaves off, as the preconceptualexpression of the passage to another way of being. This is why cinema is both apresentation of and withdrawal from skepticism—the almost perfect realization ofthe form of skeptical perception as a way, paradoxically, of reconnecting us to theworld and asserting its existential force as past presence in time. The irony of thisrecognition now is that modernity may no longer characterize our modes of beingor of looking, and we must then anticipate something else.

In the major books that follow, culminating in Cities of Words, the temporalityof this epistemological condition is reconsidered as a question of art and ethicalevaluation. The key concept of ethical evaluation is what Cavell calls moral perfec-tionism. Moral perfectionism is the nonteleological expression of a desire forchange or becoming. Here our cinematic culture responds not to a dilemma ofperception and thought, but rather a moral imperative. This trajectory from onto-logical to ethical questions is exemplary of how Cavell uses cinema to deepen hisdescription of the subjective condition of modernity as itself suspended between aworldly or epistemological domain and a moral domain. In both cases, cinema

An Elegy for Theory 107

16. Stanley Cavell, The World Viewed: Reflections on the Ontology of Film, enlarged edition (1971;Cambridge, Mass.: Harvard University Press, 1979), p. 188.17. See my essay, “A World, Time,” in The Afterimage of Gilles Deleuze’s Film Philosophy, ed. D. N.Rodowick (Minneapolis: University of Minnesota Press, forthcoming).

confronts the problem of skepticism. In the first instance, this is an epistemologicaldisappointment, in that we are disconnected from the world by our own subjectiv-ity—all we can know of the world is from behind the screen of our consciousness.The second responds to a moral disappointment in the state of the world or withmy current mode of existence. This division is not only formal; it is also, and per-haps primarily, temporal. As Kant posed the problem, the province of under-standing, of knowledge of objects and their causal laws, defines the modern scien-tific attitude whose formidable power derives from making time an independentvariable. What is unknown in the natural world could not become known throughthe powers of causal reasoning if the rules could change in the course of time. Butthe problem that so provoked Kant was that atemporal reason was in conflict withmoral freedom. To be human is to experience change. So how might philosophycharacterize humanity as at once subject of understanding and of reason, as sub-ject to causal relations and expressive of moral freedom? Given that as materialcreatures we are in bondage to the empirical world and its causal laws, philoso-phy’s task is to explain how we are also free to experience and to anticipatechange in the projection of future existences.

Therefore, in Cavell’s account moral perfectionism takes us from the formof skepticism to the possibilities of human change, and to the deeper moral prob-lem of evaluating our contemporary mode of existence and transcending it inanticipation of a better, future existence. In the first stage, the problem is to over-come my moral despair of ever knowing the world; in the second, my despair ofchanging it and myself. Thus, Cavell’s interest in Emerson (or in Wittgenstein,Nietzsche, or Freud) is to heal this rift in philosophy exemplified by Wittgenstein’sdisappointment with knowledge as failing to make us better than we are or to giveus peace. Alternatively, moral perfectionism begins with this sense of ethical disap-pointment and ontological restlessness, catching up the modern subject in adesire for self-transformation whose temporality is that of a becoming withoutfinality. “In Emerson and Thoreau’s sense of human existence,” Cavell writes,“there is no question of reaching a final state of the soul, but only and endlesslytaking the next step to what Emerson calls ‘an unattained but attainable self’—aself that is always and never ours—a step that turns us not from bad to good, orwrong to right, but from confusion and constriction toward self-knowledge andsociability.”18

This idea forms the basis of Cavell’s later books on comedies of remarriageand melodramas of the unknown woman. The interest of film here is to show it asthe ordinary or quotidian expression of the deepest concerns of moral philoso-phy. And just as Wittgenstein sought to displace metaphysical expression intoordinary language and daily concerns, film brings moral philosophy into the con-text of quotidian dramatic expression:

OCTOBER108

18. Stanley Cavell, Cities of Words: Pedagogical Letters on a Register of the Moral Life (Cambridge, Mass.:Belknap Press of Harvard University Press, 2004), p. 13.

These films are rather to be thought of as differently configuringintellectual and emotional avenues that philosophy is already in explo-ration of, but which, perhaps, it has cause sometimes to turn from pre-maturely, particularly in its forms since its professionalization, or acade-mization. . . . The implied claim is that film, the latest of the great arts,shows philosophy to be the often invisible accompaniment of the ordi-nary lives that film is so apt to capture.19

Where contemporary philosophy has reneged on its promise of moral perfection-ism, film has responded, though in the preconceptual manner of all art andsensuous expression. Thus the great project of film philosophy today is not only tohelp reinvigorate this moral reflection, but to heal by example the rift in philoso-phy’s relation to everyday life.

In the prologue to Cities of Words, Cavell reprises Thoreau’s lament that“There are nowadays professors of philosophy, but not philosophers. Yet it isadmirable to profess because it was once admirable to live.” How well Thoreauforesaw the difficult life of philosophy in the twentieth and twenty-first centuries.If one must compose an elegy for theory, let us hope it awakens a new life for phi-losophy in the current millennium.

An Elegy for Theory 109

19. Ibid., p. 6.