Wiley
What a Documentary Is, After All Author(s): Carl Plantinga
Source: The Journal of Aesthetics and Art Criticism, Vol. 63, No. 2 (Spring, 2005), pp. 105-117 Published by: Wiley on behalf of American Society for Aesthetics Stable URL: http://www.,jstor.org/stable/3700465 Accessed: 29-11-2015 13:07 UTC
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/ info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
	(Si
m	4
mi	
American Society for Aesthetics and Wiley are collaborating with JSTOR to digitize, preserve and extend access to The ■Tournal of Aesthetics and Art Criticism.
STOR
http://www.jstor.org
This content downloaded from 153.106.124.134 on Sun, 29 Nov 2015 13:07:33 UTC All use subject to JSTOR Terms and Conditions
CARL PLANTINGA
What a Documentary Is, After All
The question of how best to define the documentary film and video and to distinguish it from the fiction film continues to fascinate and baffle philosophers and film theorists. It is clear that the special nature of the film medium—and in particular its use of photographic images and sound recordings—has proven particularly difficult to conceptualize in relation to the fiction/ nonfiction film distinction.
In this paper I offer a characterization of the documentary that can account for the visual and aural nature of the medium and that furthers our understanding of what we mean when we use the word 'documentary.' I call my theory a characterization rather than a definition, because rather than posit necessary and sufficient conditions, I will be content to identify and describe the central tendencies of the typical, or usual, documentary film.
Terminological confusion often results from various uses of the word 'documentary' and the phrase 'nonfiction film.' In its most expansive sense, a nonfiction film is any film not fictional, for example, instructional films, advertisements, corporate films, or historical or biographical documentaries. The Scottish filmmaker and theorist John Grierson called the documentary the "creative treatment of actuality," a characterization that simultaneously distinguishes the documentary from the fiction film (not thought to be primarily a treatment of actuality) and the non-fiction film (not thought to be creative or dramatic).1 Although the distinction between nonfiction film and documentary cannot bear much theoretical weight, it might be useful to think of the documentary as a subset of nonfiction films, characterized by more aesthetic, social, rhetorical, and/or political ambition than, say, a corporate or instructional film.
Even so circumscribed, the category "documentary" embodies a wide range of films in the various moving-image media. Documentary types can be variously categorized; the most influential conceptual mapping comes from Bill Nichols, a film scholar who proposes six subgenres or modes of the documentary: expository, observational, poetic, participatory, reflexive, and performative.2 The modes each emerged at a particular time, some have come into and fallen out of favor, and all are subject to the vagaries of fashion and critical practice. Yet films continue to be made in each of the modes and so they remain a viable way to chart the documentary terrain. As I will demonstrate, any attempt to characterize the documentary must take into account the differing natures of these various subgenres.
It would be useful to begin by identifying and briefly examining the two best candidates for traditional definitions of the documentary. These are what I call the Documentary as Indexical Record (DIR) and the Documentary as Assertion (DA) accounts. In the next two sections of this essay I give descriptions of the basic claims of these accounts, noting internal problems and proposing a plausible statement of each. In the third section, I show how both accounts fail as traditional definitions of the documentary. In Sections IV and V, I develop an alternative account, in which I argue that the typical or usual documentary is what I call an "asserted veridical representation."
I. DOCUMENTARY AS INDEXICAL RECORD
Documentary as Indexical Record (DIR) accounts, in their most plausible form, characterize
The Journal of Aesthetics and Art Criticism 63:2 Spring 2005
This content downloaded from 153.106.124.134 on Sun, 29 Nov 2015 13:07:33 UTC All use subject to JSTOR Terms and Conditions
106
The Journal of Aesthetics and Art Critici
documentaries as films comprised predominantly of moving photographic images that are indexical records or traces of the pro-filmic scene(s).3 Charles Sanders Peirce defines an indexical sign as one that bears a relationship of causality or proximity to that which it represents. He distinguishes the index from the icon, which resembles its referent, and the symbol, which bears an arbitrary or purely conventional relationship with its referent.4
Photographs and sound recordings can (and often do) function as icons, indices, and symbols. It is their indexicality, however, that has been most intriguing for filmmakers and theorists. It is a well-known claim that the photograph is in part the product of a series of mechanical cause-and-effect operations performed in and through a machine—the camera. Insofar as the photograph is produced by causal processes governed by physical laws (and not by human intentionality), this allows us to impart a veracity to photographs that we do not allow for a painting. The filmmaker often must choose what to shoot and how to shoot it; photography certainly involves intentions and plans on the part of the photographer. Nonetheless, the mechanical nature of the photograph's provenance allows us to attribute to the photograph an evidentiary status that we would not grant to a painting.
DIR theories have often made much of the ability of the documentary photograph to record the world and tend to underestimate the creative, interpretive nature of documentary filmmaking. Some early practitioners of direct cinema or cinema verite talked as though their purpose were merely to record reality and leave all interpretation to the spectator. This led some to think of documentaries as mere "re-presentations" of reality, or simple records, rather than creative interpretations, of their subjects.6 Poststructur-alist theorists were quick to note that no documentary can perfectly re-present or reproduce anything, and they declared the very idea of documentary to be suspect.7 The problem, however, is not with the documentary, but with confused theories of documentary; a solution would be to provide a better conception of what a documentary actually is, as I attempt to do in this essay.
Though the practitioners of direct cinema and various theorists have overstated the degree to
which a documentary is a mere recording of subject (and not an interpretation of it), it nonetheless undeniable that the document; has relied on the power of the moving pho graph to "show us the world," and to do so w an authenticity that depends not only on the v ual wealth and detail of the photograph, but a on the indexical, causal bond between pho graph and pro-filmic scene.
Gregory Currie has recently taken up the E banner.8 To begin to describe Currie's theo we must first explore his notion of photograp] representation. Currie distinguishes betwe what he calls "traces" and "testimonies." A t timony, for Curry, is a representation that i; record of "what someone thought the facts the matter were" (p. 287). Testimonies, unl traces, are thoroughly mediated by the p ducer'-s intentions. Moreover, persons are ca] ble of giving testimony about all kinds of thii that might never have existed, while only r things can leave traces of themselves. Examp of testimonies include paintings, drawings, h tories, and journalism.
Photographs, like footprints and death mas are traces of the world left by the subjects the selves. Photographs are traces in part becai they are independent of belief in a way ti paintings are not.9 The painter may hallucin while painting and paint an empty room though it were full of apparitions. The pho grapher, similarly hallucinating while pho graphing the room, will be surprised to fine photograph of an empty room. To some degr then, the making of the photograph is indepei ent of belief, and the photograph is a trace.
Moving photographs in fiction films are a traces, however, so the use of photographs traces cannot by itself define the documenta Currie argues that the "ideal" documentary is filmically sustained narrative the constituti film images of which represent only pho graphically: they represent only what they : of (p. 291). A fiction film may use an image Gregory Peck to represent the fictional char; ter Atticus Finch. An ideal documentary, contrast, "may not represent things and eve other than the things and events they are tra< of."
Let us leave aside for now Currie's proble atic claim that documentaries must be nar lives.10 Currie's account contains a fundamer]
This content downloaded from 153.106.124.134 on Sun, 29 Nov 2015 13:07:33 UTC All use subject to JSTOR Terms and Conditions
Plantinga   What a Documentary Is, After All
107
confusion that is more germane to the present discussion. He sometimes (as above) implies that a documentary is a film that uses photographs to represent what the photographs are traces of, such images being employed to support an "asserted" narrative.11 At other times, however, Currie writes that a documentary film itself is the trace of that which it represents. He writes, for example, "to be a documentary the thing in question must be a trace" (p. 289).
These seem to be two quite different notions of documentary. The former defines the documentary as a filmic narrative supported by visible traces used to represent what they are of, while the latter defines a documentary itself as a visible trace. But what would it mean to claim that a documentary film is a trace? Were Currie to suggest that documentaries themselves are traces in the same sense that individual photographs are, then Currie should want to attribute the same kind of belief-independence to documentaries that he does to discrete photographic images. This cannot be done, however.
Let us grant that individual documentary shots, in addition to their status as interpretations or expressions (through all the creative choices involved in cinematography), are also traces in the sense that Currie claims. Documentary films are also edited, and editing almost invariably further interprets the event and involves intentionality in a way that indexical signs such as traces do not. When one adds music or titles or voice-over narration, additional mediation between documentary and subject is added. A documentary itself might be considered a trace only under conditions that very few, if any, documentaries ever meet. The surveillance film would seem to be the best example of such a documentary.12
Currie recognizes this problem and attempts to resolve it by claiming that most documentary films have parts that are not documentary (such as bits of voice-over narration, nondiegetic music, animated maps, and so forth.). Yet this attempted resolution, rather than clearing up the issue, foregrounds the basic mistake in Currie's formulation. Like many before him, Currie confuses a document with a documentary. A photographic document can be a physical trace, and documentaries often make use of such traces. There are very few documentaries, however, that can legitimately be said to function as traces.
For the purposes of this paper, then, we will formulate the DIR account to be claiming the following: a documentary is a sustained discourse of narrative, categorical, rhetorical, or other form that makes use of moving or still photographic images predominantly as traces to represent what the photographic images are of.
II. DOCUMENTARY AS ASSERTION
Documentary as Assertion (DA) accounts have been formulated in various ways, but their similarities legitimate taking them as a single category of definition. In a 1987 essay, I made use of Nicholas Wolterstorff s theory of projected worlds to argue that a nonfiction film is a film in which a filmmaker takes an assertive stance (as opposed to a fictive stance) toward the world projected by the film.13 Wolterstorff argues that through every representational work of art an agent projects a world, or state of affairs. When a writer or filmmaker takes an assertive stance toward the world projected through the work, he or she asserts that the state of affairs making up that projected world holds or occurs in the actual world. When filmmakers take such a stance, I write, they "make assertions about the actual world" through the work. When a text is recognized as a documentary, this "mobilizes relevant expectations on the part of the audience."14 Thus the fiction/nonfiction film distinction was placed within the realm of speech act theory.
This proved to be the direction taken in subsequent work by Noel Carroll and Trevor Ponech, work that clarified the sense in which nonfiction film could be said to fundamentally involve something like "assertion," an "assertive stance," or "assertorial intentions."15 This is typically contrasted with fiction, in which the filmmaker or writer takes a fictive stance toward the world of the work, presenting the state of affairs for our delectation, edification, education, amusement, or what have you, but not to have us believe that the state of affairs that constitute the world of the work holds in the actual world.
In a conceptual analysis of the word 'documentary,' Carroll introduced the idea of the "film of presumptive assertion," which he also terms the film of "putative fact" and "presumptive fact."16 This essay is characteristically clear and
This content downloaded from 153.106.124.134 on Sun, 29 Nov 2015 13:07:33 UTC All use subject to JSTOR Terms and Conditions
108
The Journal of Aesthetics and Art Criticism
insightful, even if his idea of presumptive assertion is faulty in one regard. Carroll invokes what he calls an intention-response model of communication, which presupposes that the artist or maker communicates with an audience in part by indicating that the audience is meant to respond in a certain way.
In the film of presumptive assertion, "the filmmaker intends that the audience entertain the propositional content of his film in thought as asserted."17 Carroll calls documentaries films of presumptive assertion (rather than simply "films of assertion") in part because the audience presumes that it is to entertain the propositions as asserted; this is the response part of the intention-response model of communication.
In this case, one wants to know whether it is the filmmaker's assertorial intentions or the audience's response (or both) that defines a work as a film of presumptive assertion. Carroll's definition explicitly makes audience response a necessary condition when he argues that one requirement for a film to be a film of presumptive assertion is that "a entertains p as an asserted thought," where a is the (or an?) audience, and p is the propositional content of the film (p. 188). If the (or an) audience fails to adopt such a stance, presumably, this would disqualify a text as a film of presumptive assertion. It strikes me that in such cases, the producer's intention, together with the textual cues and markers that signal such intention, makes the work one that can be said to make assertions, and not the actual presumptions of any particular audience. An intention-response model of a type of film need not rely on the actual response of spectators. To do so would imply a thorough-going subjectivism, such that, depending on its audience, a film could be a documentary for some and not for others.18
It makes more sense to leave the actual spectator response out of the definition since what is most important about such a relational definition is that a filmmaker intends that the text be received in a certain way, and that he or she design the text according to that expected reception.19 It is quite plausible, then, for Carroll to say that documentaries are films for which the relevant propositional content therein is meant to be taken as asserted, but the qualifier 'presumptive' in 'presumptive assertion' ought to be dropped. Why not call it, simply, "the film of assertion"?
Trevor Ponech's work on these issues also merits serious attention.20 Ponech writes that documentaries are "cinematic assertions," at their core consisting of "the action of indication." Ponech writes that in "producing non-fiction, a communicator uses some unit of motion picture footage in an effort to assert that something is (or was, or will be, or could be) the case." "To perform a cinematic assertion," he writes, "is to employ a motion picture medium... with the expressed intention that the viewer form or continue to hold the attitude of belief toward certain states of affairs, objects, situations, events, propositions, and so forth, where the relevant states of affairs etc. need not actually exist."21 Ponech goes on to write about how various types of documentaries embody cinematic assertions.
Trevor Ponech, then, defines the nonfiction film as one in which its makers "openly signal their intention that viewers take the attitude of belief toward" the states of affairs presented in it. His is an intentionalist theory, one that locates the essence of nonfiction film in the intentions of the filmmaker(s). Ponech writes that those intentions are discoverable in the plans the filmmaker develops in making the work, plans that become manifest in the finished film.22
These DA accounts, then, share much in common. They go beyond the formal elements of films to distinguish between fiction and nonfiction on the basis of the illocutionary act performed through or with the work. Moreover, they all implicitly appeal to the intentions of filmmakers. Roughly speaking, DA accounts hold that documentaries are moving picture texts in or through which filmmakers assert that the states of affairs represented in the work hold in the actual world. In other words, filmmakers take an assertive stance toward the world of the work. These definitions also take into account the response of the spectator as a factor that enters into the filmmaker's plans in making the film. At the receiving end, the spectator of a documentary is meant to form or continue to hold an attitude of belief toward the state of affairs so represented.
III. THE FAILURE OF DIR AND DA DEFINITIONS
The question I ask here is whether the DIR and DA definitions, when plausibly stated, capture
This content downloaded from 153.106.124.134 on Sun, 29 Nov 2015 13:07:33 UTC All use subject to JSTOR Terms and Conditions
Plantinga   What a Documentary Is, After All
109
what we mean when we use the word 'documentary.' To begin to answer this question, we need to explore the usage of the word 'documentary' a bit further. What kind of moving-image nonfictions do we have in mind when we use the term 'documentary'?
Films that are considered to be documentaries come in many varieties. If we survey the territory, we see journalistic documentaries such as those found on the public television series Frontline, associational and poetic documentaries such as Anima Mundi (1992) and Koyanisqaatsi (1983), propaganda films such as Why We Fight (1942-1945) and Triumph of the Will (1935), the films of the direct cinema or cinema verite movements, films that make heavy use of reenactments such as the documentaries of John Grierson, Robert Flaherty, and Humphrey Jennings, and documentaries in the making of which the filmmaker becomes a kind of provocateur {Chronicle of a Summer [1960], Sherman's March [1985]).
There exist many ways to carve out this diverse body of films, but perhaps the most influential has been Bill Nichols's description of six documentary modes, which I mentioned in the introduction to this essay: the poetic, expository, observational, participatory, reflexive, and performative.23 For my purposes, it will be sufficient to describe just two of the six modes—the expository and the observational— to show inadequacies in the DIR and DA accounts.
• Expository. Typically, a voice-over narrator provides an explanatory conceptual framework, and images and sounds are used to illustrate or provide (loose) evidence for what is stated by the voice-over narrator. Expository documentaries tend to be heavily scripted and many make an overt argument for a position or for a particular interpretation of history. Examples include the Why We Fight series, The Sky Above, the Earth Below (1962), and most journalistic television documentaries, such as CBS Reports' Harvest of Shame (1962) and most of the current films of the PBS Frontline series.
• Observational. Eschews voice-over narration and many other traditional techniques in favor of the observation of the pro-filmic event and a more open-ended and ambiguous
treatment of its subject. Often thought to allow greater freedom of interpretation on the part of the viewer than the expository mode. Associated with American direct cinema and, to a lesser extent, with cinema verite. Examples include any of the documentaries of Frederick Wiseman such as High School (1968), Hospital (1980), and Racetrack (1985) and the Maysles brothers' Salesman (1969) and Grey Gardens (1975). More recent examples are The War Room (1993) and Startup.com (2001).
Equipped with this new terminology, let us return once again to the DIR and DA accounts, with a view toward assessing whether either, taken as a traditional definition, seems to fit both of these rather central modes of documentary. The DIR account, which I will consider first, has little trouble with observational films. The observational documentary, of course, is directly rooted in the ability of the moving image and sound recording to provide a kind of indexical record, or trace, of the pro-filmic event.
It is the expository documentary, in many of its historical manifestations, that DIR accounts describe poorly. This is obviously true in the case of historical documentaries about subjects that existed before the invention of photography. In these cases there can be no photographic trace of such subjects. Currie admits that his version of DIR would preclude documentaries about Napoleon, for example.
DIR accounts would also have trouble with the first sixty-five or so years of documentary history. The films produced under the aegis of John Grierson, the man who did so much to fix the meaning of the word 'documentary,' commonly used recreations and stagings of events, as did other pioneers of the documentary form such as Robert Flaherty and Humphrey Jennings. Shots of reenacted events clearly do not represent what they are photographs of and are thus problematic for DIR accounts.25 The kind of cinematography favored by DIR accounts, in fact, did not become strongly associated with the documentary until after the direct cinema and cinema verite movements of the 1960s.
This issue aside, DIR accounts, it seems to me, do not capture the most important features of expository documentaries. Currie writes that
This content downloaded from 153.106.124.134 on Sun, 29 Nov 2015 13:07:33 UTC All use subject to JSTOR Terms and Conditions
110
The Journal of Aesthetics and Art Criticism
under his definition, documentary films are those in which "meaning passes from image to narrative, while in nondocumentary meaning goes the other way."26 Although Currie admits that this passage is put "loosely" (and thus I may be misinterpreting it), I take Currie to be saying that whatever meaning documentaries might have originates in or stems from the photographic traces that make up the documentaries, and not from some prior argument, previously researched historical account, political analysis, scientific explanation, and so forth. This claim, however, is implausible for a wide range of documentaries. Well-known documentaries such as The Life and Times ofRosie the Riveter (1980), The Thin Blue Line (1987), and Roger and Me (1989) are carefully crafted films organized around an argument, broadly conceived. It is quite obvious that the images support a scripted argument or narrative, the meaning of which does not necessarily arise from the images used.
Neither would it be right to find the essence of these films, qua documentary, to lie in the particular use of motion picture photographs as traces. In these cases, it seems to me, the use of cinematography is harnessed to the broader argumentative strategy of the filmmakers, which, I believe, DA accounts can account for. DIR accounts, then, fail as traditional definitions in part because they are too narrow. They would not only rule out many paradigm examples of the documentary, but they do not fit one central mode of the documentary—the expository documentary.
DA accounts, in my view, are far more plausible, but nonetheless must contend with conceptual problems. With their emphasis on truth claims, the assertion of propositional content, and/or cueing spectators to take a stance of belief toward what is presented, DA accounts are well able to distinguish prose fiction from prose nonfiction, since the assertion of propositions and/or the assertive stance are well suited to linguistic discourse. DA accounts do less well in characterizing the documentary, however, in part due to the peculiar nature of the photographic and sonic, as opposed to linguistic, discourse.
Consider an observational film such as Frederick Wiseman's Hospital (1970). In the tradition of American direct cinema, Hospital
has no voice-over narration and exclusively uses images and sounds recorded on location to present a portrait of New York's Metropolitan Hospital. Wiseman clearly implies much about the hospital through the selection of footage and through editing, so one might find in the film propositional content that is implied or asserted. Yet the film's epistemic voice, as we might call it, is open or hesitant.27 For much of its running time, the film seems just as content to observe and to display, as to make assertions about propositional content.
One might claim that in observing or displaying such images, Hospital should be taken to assert that the state of affairs represented, stated in propositional terms, occurred in that actual hospital. Trevor Ponech takes this position. In relation to the anthropological film Trance and Dance in Bali (1952), with a nonsynchronous music track as its only sound, Ponech first describes, in linguistic terms, the features of the ritual dance performances shown, then concludes that the filmmakers' objective "is to assert that this ritual performance has the aforementioned features." In other words, such a series of images, without voice-over narration, should be taken to assert a series of propositions about its subject, stated in linguistic terms.
This claim is problematic, however. If photographs are traces, as Currie claims, then we should say that they have a communicative life that in part escapes the intentions of the filmmakers). The filmmakers cannot have in mind, when making the film, all the propositions that might plausibly be gleaned from the film's images. Wiseman and the makers of Trance and Dance in Bali need not be committed to any particular propositional account of what occurs in each moving image. Why is this? It is because the moving photograph and the sound recording are to some degree belief-independent. Their communicative richness extends beyond the intentions of the filmmakers and leaves something for interpretation and discovery by audiences.
In addition, it may be that certain images and sounds, or sequences thereof, are meant to approximate some element of the phenomeno-logical experience of the event, such as how it looked or sounded from a particular vantage point, or how it was full of energetic good cheer or a strong sense of foreboding. Thus the film
This content downloaded from 153.106.124.134 on Sun, 29 Nov 2015 13:07:33 UTC All use subject to JSTOR Terms and Conditions
Plantinga   What a Documentary Is, After All
111
may be taken to assert that the relevant scenes give a sense of how the filmmakers were "appeared to" aurally and/or visually. This is still a case of assertion in some sense because the filmmakers might be taken to be asserting that a scene shows what the pro-filmic event looked like, or approximates how the filmmakers "were appeared to." The apprehension conditions of such scenes, however, cannot be linguistic in nature. That is, we can grasp those phenomenological qualities the scene embodies only by viewing the scene.
We might get at this by drawing a distinction between saying and showing. Saying, in the context of a documentary, characteristically involves the assertion of specific propositional content. It is something like making an assertion or assertions about the representee, saying that it is thus and so. Showing, on the other hand, is something like standing in for the representee and may not involve the assertion of specific propositional content. For example, showing a person a series of snapshots taken of an event need not commit the shower to an assertion of the propositional content of the photographs. The shower is simply presenting the photographs as veridical representations of the event and allowing the viewer to learn and perhaps form beliefs about the event on the basis of those photographs.
Most documentaries, it seems to me, are representations that combine saying and showing and do so in different proportions depending on the type of documentary. The observational documentary often leaves more to the viewer's share than does the expository documentary and is further toward the side of showing. The viewer of Hospital might come to have certain beliefs about the subject based on the documentary cinematography. Yet what is being asserted about the propositional content of the photographs is underdetermined; in some cases, the propositional content is unspecified and in many cases it is unclear. The propositional content of many documentaries, or of many parts of many documentaries, escapes the control and intentions of the filmmaker.
In some cases, moreover, what is communicated is the look, the sound, the feel of an event, as much as propositional information. It is as though the director is content to show us the "look" of the pro-filmic event, or what the
event sounded like or felt like from a particular perspective. The spectator may infer propositional knowledge from these ways of seeing, hearing, and feeling, but the filmmaker is not necessarily asserting, or intending the spectator to take as asserted, all the propositional content that can reasonably be inferred from the shot or series of shots.
In a documentary, what the filmmaker asserts, in the first instance, is that the images, sounds, and other materials presented are what I will call veridical representations of whatever the documentary takes as its subject. As I describe below, documentary representation commits the filmmaker to assert the reliability or functionality of whatever materials are used to show the spectator how something is, was, or might be in the actual world.
DA accounts, then, miss or minimize the extent to which the moving photographic image and recorded sound, as rich, sensual, indexical records of the subject, cannot be reduced to the intentional assertion of propositional content. Photographic and sonic communication in the documentary require that we consider the notion of asserted veridical representation (AVR) as embedded in our concept of the typical or usual documentary.
IV. ASSERTED VERIDICAL REPRESENTATION
My argument is that central to our idea of the typical or usual documentary, and prior to any notion of the photograph as a trace, is the implicit directorial assertion of veridical representation, representation that is, in the case of implicitly or directly asserted propositions, truthful; and in the case of images, sounds, or combinations thereof, a reliable guide to relevant elements of the pro-filmic scene or scenes. When a filmmaker presents a film as a documentary, he or she not only intends that the audience come to form certain beliefs, but also implicitly asserts something about the use of the medium itself—that the use of motion pictures and recorded sounds offer an audiovisual array that communicates some phenomenological aspect of the subject, from which the spectator might reasonably be expected to form a sense of that phenomenological aspect and/or form true beliefs about that subject.
This content downloaded from 153.106.124.134 on Sun, 29 Nov 2015 13:07:33 UTC All use subject to JSTOR Terms and Conditions
112
The Journal of Aesthetics and Art Criticism
I have introduced the notion of AVR in an attempt to account for what people often mean when they use the word 'documentary.' In claiming that AVR is expected of documentary films, I am not claiming that audiences, critics, and filmmakers share a well-defined conception of what constitutes AVR. Far from it. Audiences need not have a philosophically precise idea of what constitutes AVR for the concept, vague though it is, to play a central role in thinking about the typical or usual documentary. People do expect of the documentary that it is intended to offer a reliable record, account of, argument about, or analysis of some element of the actual world, that is, they expect an assertedly veridical representation.
What counts as AVR, however, differs in various contexts. For example, what is accepted as a veridical representation depends in part on the mode of documentary in question. In expo-sitional documentaries, the assertion of propositions or truth claims becomes central. The implicit rules for veridical representation through images are relaxed somewhat, allowing for animated maps, occasional reenactments, the relatively loose use of archival footage, and so forth, as long as such images and sounds are not fundamentally misleading.
Typical observational documentaries have stricter conventions for the use of motion picture photography. Within the context of the observational film, AVR requires that the filmmaker refrain from overt manipulation and staging in the making of recorded images and sounds. In any documentary, however, when photographic images and sound recordings are used as documents, that is, as evidence that the pro-filmic event occurred in a certain way, the requirements of veridical photographic representation are quite strict.
Conventions of veridical representation also change with history. A quick look at the history of documentary shows that the staging and reenactment of scenes was routine and commonly accepted as legitimate documentary practice for the first sixty-five years of documentary history. The films of Robert Flaherty, John Grierson, and Humphrey Jennings, arguably the most important documentary filmmakers of the first half of the twentieth century, commonly make use of staged and reenacted scenes.
The development of lightweight cameras and sound-recording equipment in the late 1950s contributed to the rise of a new ethos of authenticity, fully developed in the direct cinema and cinema verité movements of the 1960s. The project of the documentary film, some cinema verité filmmakers claimed, was to record and represent reality, and not to make interpretations. The documentary filmmaker became, then, not an artist or teacher so much as a facilitator, one who self-effacingly records the pro-filmic event in order to re-present it, as is, to the spectator. The sense that the filmmaker's duty was to record and not interpret led to conventional practices of documentary film production. Voice-over narration was rejected as manipulative and patronizing; the spectator should be allowed to interpret the film himself or herself. The filmmaker refrained, as much as possible, from manipulating or influencing the pro-filmic event, and attempted to become a proverbial fly on the wall. Cinema verité filmmakers used images and recorded sounds predominantly as traces, in Currie's sense. Some rejected the use of program music because it did not originate from the pro-filmic scene.29
With the influence of cinema verité, conventions of asserted veridical representation for documentaries became much stricter. Staging and reenactment were looked on with suspicion, and cinematography was largely limited to the trace functions Currie describes. At the height of the cinema verité movement, the conventions of asserted veridical representation were far different than they are for most filmmakers today.
Contemporary documentarians and critics, for the most part, reject the more extreme claims of the cinema verité movement, and freely use scripts, voice-over narration, animated simulations, program music, and so forth. Errol Morris, among whose films is The Thin Blue Line (1988), which makes use of dramatic recreations, argues that documentary filmmaking must be as personal and creative as fiction filmmaking, and that cinema verité set documentary filmmaking back twenty years.30
The point of this brief historical interlude is this: my argument is that inherent in our use of 'documentary' is that the typical or usual documentary is an asserted veridical representation. I do not argue that filmmakers, critics, and audiences will agree about what qualifies as asserted
This content downloaded from 153.106.124.134 on Sun, 29 Nov 2015 13:07:33 UTC All use subject to JSTOR Terms and Conditions
Plantinga   What a Documentary Is, After All
113
veridical representation. In fact, disagreements about such issues are common. This is in part due to historical variability and variability across movements, modes, and contexts.
My notion of AVR encompasses the insights of DA and DIR definitions by affirming, but not claiming too much for, the assertion of propositions or the use of trace images in the documentary. Thus I do not deny that in many documentaries, propositional assertion plays a central role. Asserted veridical representation requires that the relevant portions of the propositional content of a documentary be asserted as holding or occurring in the actual world.31 Yet even in the expositional film (the mode most associated with the explicit assertion of propositional information), images and sounds are used not solely to assert, but to provide a sense of the look, sound, or overall perceptual experience of a scene or scenes. In the observational film, on the other hand, AVR more fundamentally takes the latter form, as the filmmaker is taken to assert that a shot embodies the look or sound or "experience" of a scene from some perspective, and thus can serve as a reliable record or model of its subject. Secondarily, the film might also be taken to implicitly or explicitly make assertions about its subject, beyond the implicit claim that the scene captures some aspect of the look or sound of the pro-filmic event.
As discussed above, Currie characterizes the documentary as a film that uses photographs to represent what the photographs are traces of. But the notion of trace neglects an essential element of the use of images and sounds for the purposes of AVR. As every cinematographer knows, a shot that represents what it is of may nonetheless imply much that is misleading or just plain false about its subject. The cinematographer can use angle, focal length, shot composition, lighting, filters, camera movement, film stock, background, and context to represent the subject in diverse ways, all while using said moving photograph to represent what it is of. Moreover, the director might ask a subject to strike an uncharacteristic pose or perform some highly unusual action.
If we expect that a documentary provide us with reliable visual information, we expect more assurance from a director than the mere fact that she or he used only traces. Traces are a central element of AVR in the observational
documentary, but always in relation to various requirements for their production and use. There are requirements, for example, about the treatment of the film's subjects, the filmmakers' influence on the pro-filmic event, the use of cinematography, and how such traces should be edited.
Photographic traces are often important in the expositional film as well, though expositional documentaries loosen requirements for cinematography and allow for occasional recreations, reenactments, charts, maps, animations, and computer simulations. Conventional protocols of AVR, then, might also allow for computer simulations, animated maps and charts, various sorts of directorial manipulation and staging, and a host of other creative manipulations. It could well be the case that a computer simulation might provide more accurate and detailed information about a subject than a trace photograph, as in the case of James Cameron's use of computer simulations in Ghosts of the Abyss (2003), of the interior of the Titanic before the ship sank.
Note that while this conception of the documentary finds the assertion of veridical representation to be central to the documentary film, it in no way precludes the possibility that documentaries can be used to lie, or that they may sometimes be highly misleading, inaccurate, or mired in any of a number of epistem-ically troubled waters. The mere assertion of having followed conventional protocols for veridical representation, of course, guarantees nothing. Moreover, the director may intend that audiences presume that relevant protocols were followed, but not in fact have followed them.
For example, it is commonplace in nature documentaries that filmmakers eliminate any trace of human existence in films about wildlife and wild areas.32 The cinematographer may reframe to leave the telephone pole out of the shot and the sound recorder might wait for the right moment to record the sound of young birds in a nest without the sound of an airplane flying overhead. Although we may not want to call such practices deceptive in themselves, the convention overall may lead to a romantic and inaccurate conception of nature as something that is wholly free from human influence and presence.
It is certainly possible that some conventions of documentary representation are mired in naive
This content downloaded from 153.106.124.134 on Sun, 29 Nov 2015 13:07:33 UTC All use subject to JSTOR Terms and Conditions
114
The Journal of Aesthetics and Art Criticism
mythmaking, romanticization, self-conscious rhetorical and persuasive strategies, and various forms of self-delusion. The question of which documentary practices and techniques are in fact reliable extends far beyond the boundaries of this paper. My claim is not that documentaries are in fact veridical (although I believe that in some cases they are), but that the documentary filmmaker typically intends the spectator to take what is presented as asserted veridical representation.
Clearly, much of the most interesting discourse about the documentary has to do with the nature of veridical representation since, in my view, accepted notions of asserted veridical representation are clearly subject to change. When, for his 1991 documentary about physicist Stephen Hawking, A Brief History of Time, Errol Morris constructs a set to look like the office of the famous scientist, does that qualify as asserted veridical representation? When James Cameron uses animated computer simulations of the interior of the Titanic in his film about exploring the wreck of that ship, Ghosts of the Abyss, can we expect audiences to accept this as asserted veridical representation? When, in order to get spectacular footage of flying birds, the makers of Winged Migration (2003) imprint young birds to follow an Ultralight in flight, does this constitute asserted veridical representation? I am inclined to answer yes in each case, because in each the filmmaker intends that the audience take the relevant film scenes as reliable representations of some element of the actual world from which true beliefs can be formed about the film's subject.
Any characterization of the documentary had better work to distinguish the fiction from the nonfiction film, and the notion of asserted veridical representation does so. If the practices of asserted veridical representation are conventional, historical, and contextual, then we should not expect asserted veridical representation to distinguish fiction from nonfiction film in the same way in all contexts. Behind all the historical change and contextual variations, however, lies a more or less constant functional difference between fiction and documentary that the notion of AVR embodies. Typical documentaries are first and foremost meant to be taken, in both their particularity and their broad thematic outlines,  as reliable  accounts of,
records of, and/or arguments about the actual world. Fictions may also muse about the actual world, but do so indirectly through fictional characters and events.
In some cases, differences between documentary practices of asserted veridical representation and fictional practices might be subtle and complex. In almost no case, for example, would we accept actors playing purely fictional characters as asserted veridical representation, yet we might accept actors playing historical figures if we were convinced that quality research had figured into the historical accuracy of what the actors wore, said, and did. Some fiction films intend the audience to take a stance of belief toward portions of their propositional content, but we rarely accept as asserted veridical representation the offering of fictional characters, imaginary worlds, and made-up stories.
Although the distinction between fiction films and documentaries is most often clear, various hybrid films stand at the fuzzy boundaries of fiction and nonfiction, never settling comfortably into either category. Although Lost in Translation (2003) is clearly fiction and The War Room (1992) clearly documentary, certain kinds of "historical fictions," such as Oliver Stone's JFK (1991) and the "nonfiction movies" we sometimes see on PBS, such as Woodrow Wilson (1999) and The Saga of the Israelites (2003), do not fit easily into either category. Of course, fuzzy boundaries do not by themselves call into question the legitimacy of the categories.33
V. WHAT A DOCUMENTARY IS
Now I am prepared to say what a documentary is, after all.
I propose that the typical or usual documentary film be conceived of as an asserted veridical representation, that is, as an extended treatment of a subject in one of the moving-image media, most often in narrative, rhetorical, categorical, or associative form, in which the film's makers openly signal their intention that the audience (1) take an attitude of belief toward relevant propositional content (the "saying" part), (2) take the images, sounds, and combinations thereof as reliable sources for the formation of beliefs about the film's subject
This content downloaded from 153.106.124.134 on Sun, 29 Nov 2015 13:07:33 UTC All use subject to JSTOR Terms and Conditions
Plantinga   What a Documentary Is, After All
115
and, in some cases, (3) take relevant shots, recorded sounds, and/or scenes as phenomeno-logical approximations of the look, sound, and/ or some other sense or feel of the pro-filmic event (the "showing" part).
I will not assert necessary and sufficient conditions, and thus I refrain from offering a traditional definition. If one insists on the usefulness of traditional definitions, it may be more fruitful to attempt to define the various modes of the documentary, the modes being sufficiently circumscribed to admit of a better chance at such definition. In the case of the documentary film broadly considered, it is enough to describe central features of the usual or prototypical documentary to enable a fuller understanding of this kind of film text.34
The characterization I propose fits prototypical documentaries rather than those at the periphery, and describes some documentary modes better than others. For example, this characterization will not fit well the poetic mode of documentary, which uses trace images to an aesthetic more than informational intent. But one could argue that the poetic mode is itself far less central to the documentary genre than the expository or observational modes.35 In any case, I would argue that a traditional definition of the documentary cannot be given, since the variety of films we call "documentary" are too variegated to allow for the attribution of necessary and sufficient conditions.
Two possible objections to my account of AVR are (1) that it is either too vague to be assessed by comparison with DA and DIR accounts, or (2) that it is equivalent to a plausible version of DA. The first objection, that of vagueness, might be that to describe the documentary as asserted veridical representation, that is, as a set of ever-changing practices dependent on documentary mode, history, context, and so forth, is equivalent to no more than claiming that a documentary film is one that a filmmaker produces using conventional documentary techniques. I have two responses to this objection. First, my purpose in this paper is to claim that the notion of asserted veridical representation captures something central about our sense of the functions and purposes of the documentary. Insofar as the concept is applied ambiguously, and differs according to historical context, mode, and so forth, then any precise,
historically unchanging definition of AVR risks all the sorts of problems I found with the DA and DIR accounts. Thus, any attempt to further circumscribe the concept of asserted veridical representation, or the general concept of documentary, must take into account historical and contextual, and not merely conceptual, factors.
Second, I do give AVR content. In making a distinction between saying and showing, I suggest that a documentary combines both, but that in either case, the documentary is intended as a reliable account of, argument about, record of, or approximation of some aspect of the actual world. The practices of AVR that distinguish the documentary are oriented toward one or more of these functions.
The second objection to my characterization might be that my account of AVR is merely an improved version of the DA account. It is true that in AVR, both saying and showing involve assertion: saying does for obvious reasons, and showing does because when a documentary filmmaker openly signals an intention that the audience take a shot, for example, to approximate what the pro-filmic event looked like from a certain perspective, the filmmaker is implicitly asserting a relation of resemblance (appearance) between the shot and the scene. The AVR account, then, denies only that the assertion of propositional content be taken as the characteristic or defining feature of the documentary; it does not deny that such assertion plays an important role in most typical, or usual, documentaries. For as I have argued, exaggerating the role of specific assertions in the documentary risks overlooking showing in the documentary, that is, the presentation of images and sounds to allow for the apprehension of phe-nomenological qualities on the part of the spectator, and to allow more to the "viewer's share." At the end of the day, it matters little whether AVR is a new and improved version of DA, or whether it is a sufficiently novel characterization of the documentary to warrant a new title. Of most interest to us is whether it is useful in characterizing the documentary and improves on prior accounts of what a documentary is.
The interesting task now would be to explore the conventions of asserted veridical representation in various documentary modes or exemplars, in the docudrama or what some call "nonfiction movies," and in various documentary
This content downloaded from 153.106.124.134 on Sun, 29 Nov 2015 13:07:33 UTC All use subject to JSTOR Terms and Conditions
116
The Journal of Aesthetics and Art Criticism
techniques and practices. Veridical representation is widely assumed, but poorly understood, and much work remains to be done. Yet the notion of asserted veridical representation is clearly needed to account for what people typically mean when they use the word 'documentary.'37
CARL PLANTINGA
Department of Communication Arts and Sciences Calvin College
Grand Rapids, Michigan 49546
internet: cplantin@calvin.edu
1. Grierson's phrase is quoted in Forsyth Hardy's introduction to Grierson's collected writings, Grierson on Documentary, ed. Forsyth Hardy (University of California Press, 1966), p. 13. Grierson's most important writing on defining the documentary is found in the chapter, "First Principles of Documentary," pp. 145-156.
2. Bill Nichols, Introduction to Documentary (Indiana University Press, 2001), pp. 99-138.
3. The pro-filmic event or scene for a moving photograph is whatever was in front of the camera while the camera recorded the scene.
4. C. S. Peirce, "The Icon, Index, and Symbol" in Collected Papers, 8 vols., ed. Charles Hartshorne and Paul Weiss (Harvard University Press, 1931-1958), vol. II.
5. See my Rhetoric and Representation in Nonfiction Film (Cambridge University Press, 1997), p. 59. The implications of the indexical bond between photograph and referent can easily be overestimated and misunderstood. See pp. 40-82 for an extended discussion of the nature and uses of moving-image photography in nonfiction films. It is also worth noting that the moving photograph can function simultaneously as an index, icon, and symbol.
6. For an account of various attempts to define the documentary outside of philosophy, see Plantinga, Rhetoric and Representation in Nonfiction Film, pp. 7-39.
7. For a critique of postmodernist and poststructuralist theories of the documentary, see Noel Carroll, "Nonfiction Film and Postmodernist Skepticism" in Post-Theory: Reconstructing Film Studies, ed. David Bordwell and Noel Carroll (University of Wisconsin Press, 1996), pp. 283-306; see also my essay, "Moving Pictures and the Rhetoric of Nonfiction Film: Two Approaches" in Post-Theory, pp. 307-324.
8. Gregory Currie, "Visible Traces: Documentary and the Contents of Photographs," The Journal of Aesthetics and Art Criticism 57 (1999): 285-297. Further references to the Currie essay will be made parenthetically within the text. The two published responses to this essay and Currie's replies are very interesting. See Noel Carroll, "Photographic Traces and Documentary Films: Comments for Greg Currie," The Journal of Aesthetics and Art Criticism 58 (2000): 303-306; Gregory Currie, "Preserving the Traces: An Answer to Noel Carroll," The Journal of Aesthetics and Art
Criticism 58 (2000): 306-308; Jinhee Choi, "A Reply to Gregory Currie on Documentaries," The Journal of Aesthetics and Art Criticism 59 (2001): 317-319; Gregory Currie, "Response to Jinhee Choi," The Journal of Aesthetics and Art Criticism 59 (2001): 319-320.
9. Here Currie makes reference to Kendall Walton's essay, "Transparent Pictures: On the Nature of Photographic Realism," Critical Inquiry 11 (1984): 246-277. Currie notes that Walton acknowledges a debt to the work of Paul Grice.
10. Narrative is one sort of organizational structure for documentaries. Documentaries can also be organized categorically or topically, as arguments or cases, and in loose associative form.
11. Currie does not say what he means by 'asserted narrative,' but this does suggest that he recognizes the need for some type of DA account, at least as a supplement to his theory.
12. In his response to Currie, Noel Carroll suggests that Currie's theory would seemingly make the surveillance video a paradigmatic case of documentary (p. 304).
13. Carl Plantinga, "Defining Documentary: Fiction, Non-Fiction, and Projected Worlds," Persistence of Vision 5 (1987): 44-54; see also Plantinga, Rhetoric and Representation in Nonfiction Film, and Nicholas Wolterstorff, Works and Worlds of Art (Oxford: Clarendon Press, 1980).
14. Plantinga, "Defining Documentary," pp. 52-53.
15. One might also mention here the work of Roger Odin, whose "semio-pragmatic" approach to the documentary is similar in some regards to action-oriented definitions. Odin argues that while the fiction film posits a fictional enunciator (or narrator), the documentary posits an actual narrator. I find this approach to be highly problematic, however. First, not all fiction films have an enunciator or narrator; they have a narration, but not necessarily an anthropomorphized fictional narrator. When we view a film, we assume that the story is being presented, or narrated, but not necessarily, I believe, by a narrator or enunciator. Second, even if all films did have a narrator or enunciator, the distinction between fiction and nonfiction does not seem to reside in whether the enunciator is actual or fictional, but more centrally in the stance taken by the actual filmmaker toward the state of affairs projected. See Roger Odin, "A Semio-Pragmatic Approach to the Documentary Film" in The Film Spectator: From Sign to Mind, ed. Warren Buckland (Amsterdam University Press, 1995). On the idea of narration, see David Bordwell, Narration in the Fiction Film (University of Wisconsin Press, 1985).
16. Carroll, "Fiction, Non-Fiction, and the Film of Presumptive Assertion: A Conceptual Analysis" in Film Theory and Philosophy, ed. Richard Allen and Murray Smith (Oxford: Clarendon Press, 1997), pp. 173-202. The essay also appears in Carroll's Engaging the Moving Image (Yale University Press, 2003).
17. Carroll, "Fiction, Non-Fiction, and the Film of Presumptive Assertion," p. 186.
18. Another reason Carroll gives for calling this sort of film a film of "presumptive assertion" is that documentaries may lie. Carroll writes that documentaries "are presumed to involve assertion even in cases where the film-maker is intentionally dissimulating at the same time that he is signaling an assertoric intention" (p. 187). In this case, however, one wants to say that 'presumptive' is misleading because it implies that an assertion must be true to qualify as an
This content downloaded from 153.106.124.134 on Sun, 29 Nov 2015 13:07:33 UTC All use subject to JSTOR Terms and Conditions
Plantinga   What a Documentary Is, After All
117
assertion. But if a statement must be true to qualify as an assertion, this would rule out thinking of a lie as the assertion of a falsehood. I would argue that when we see a film as a documentary, we may exercise a certain skepticism about the truth of its assertions, but not typically about whether its assertions are in fact assertions.
Carroll seems to recognize this difficulty, and on a few occasions writes of the "film of presumptive fact" or "films of putative fact" as synonymous with the "film of presumptive assertion" (for example, p. 187). But this confuses two very different concepts—fact and assertion. If Carroll were to consistently call the documentary "the film of presumptive fact," then the 'presumptive' would neither be unnecessary nor misleading. This definition, however, would have the unfortunate consequence of throwing its weight on the response of the spectator (who presumes that the prepositional content of the film is factual) rather than on the intentions and actions of the filmmakers as embodied in the text through cues or markers of various sorts.
19. For a critique of subjectivist definitions of the documentary, see my essay, "The Limits of Appropriation: Subjectivist Accounts of the Fiction/Nonfiction Film Distinction" in Moving Images, Culture and the Mind, ed. Ib Bondebjerg (University of Luton Press, 2000).
20. See Trevor Ponech, "What is Non-Fiction Cinema?" in Film Theory and Philosophy, pp. 203-220; see also Trevor Ponech, What is Non-Fiction Cinema?: On the Very Idea of Motion Picture Communication (Boulder: Westview Press, 1999).
21. Ponech, "What is Non-Fiction Cinema?" pp. 204-205.
22. Ponech, What is Non-Fiction Cinema?: On the Very Idea of Motion Picture Communication, pp. 8-39.
23. Nichols introduced this taxonomy in a different form in his Representing Reality: Issues and Concepts in Documentary (Indiana University Press, 1991). His significantly revised taxonomy can be found in Introduction to Documentary (Indiana University Press, 2001), pp. 99-138.
24. Clearly, these films are not strictly or merely observational, but films that to some degree abdicate the formative work of the filmmaker in favor of the capacity of the documentary to record and observe.
25. Shots of reenacted events are photographs of a reenacted, staged scene, but are meant to represent the actual historical event. Thus they are not photographs of what they represent.
26. Currie, "Visible Traces: Documentary and the Contents of Photographs," p. 296.
27. See Plantinga, Rhetoric and Representation in Nonfiction Film, pp. 101-119, for a discussion of "voice and authority" in nonfiction films.
28. Ponech, "What is Non-Fiction Cinema?" p. 205.
29. For a discussion of the philosophical implications of cinema veríte, see Carroll's "From Real to Reel," Philosophic Exchange 14 (1983): 5-46.
30. Errol Morris and Peter Bates, "Truth Not Guaranteed: An Interview with Errol Morris," Cinéaste 14 (1989): 17.
31. I say "relevant portions" because I do not want to imply that the only illocutionary act performed in or through a documentary is assertion. There is clearly much else going on. My claim here would be that assertion is a central element of what is characteristic of the documentary, not that assertion is all there is to documentary communication.
32. This example was suggested by Ronald Tobias.
33. For more on the nature of definitions and categories in relation to the nonfiction film, see Plantinga, Rhetoric and Representation in Nonfiction Film, pp. 7-25.
34. Many philosophers have given up attempts to find necessary and sufficient conditions in favor of characterizing paradigm cases of the concept in question, as I do here. See, for example, Robert C. Roberts, Emotions: An Essay in Aid of Moral Psychology (Cambridge University Press, 2003). Ronald de Sousa, for another example, suggests that a greater understanding of the emotions will occur once we avoid the partiality of, but nonetheless learn from, each of the various theories of emotion. See Ronald de Sousa, The Rationality of Emotion (MIT Press, 1987), p. 22.
35. See Plantinga, Rhetoric and Representation in Non-fiction Film, pp. 171-190, for a discussion of what I call the "poetic voice" in documentary.
36. This is the term used by filmmaker Carl Byker for his historical films, for example, Woodrow Wilson and The Saga of the Israelites, which make heavy use of historical reenactments.
37. I owe thanks to several people. Members of the Calvin College Department of Philosophy—Terrence Cuneo, Ruth Groenhout, Lee Hardy, and James K. A. Smith— read and discussed this paper with me. I also benefited from the response of audiences at Montana State University and the 2003 American Society for Aesthetics Conference. Jinhee Choi, Arild Fetveit, Dan Flory, Alvin Plantinga, two anonymous reviewers, and the editor of The Journal of Aesthetics and Art Criticism, Susan Feagin, all provided useful comments on earlier drafts of this essay.
This content downloaded from 153.106.124.134 on Sun, 29 Nov 2015 13:07:33 UTC All use subject to JSTOR Terms and Conditions