“Photographers, especially, can easily be unaware
Kendall L. Walton
The questions to be discussed here are easily formulated: can a photograph show something fictional or not? And if so, what are the conditions for this? It is difficult to give a conclusive answer to these questions, however, for at least two reasons.
Firstly, photography seems to be a medium that is bound more firmly to the concrete object or process being recorded than any other media technology (with the possible exception of analogue sound recording). Following Peirce’s concept of the index as a sign that signifies by means of causality, the photographic image is essentially understood, in a number of key positions in the theory of photography, as a trace of the real objects and processes which were in front of the camera at the moment of exposure. This connection to an actual ‘this’ which has existed in the past seems at least to severely limit the possibilities for fictionality in the photographic image, if not — as some authors have argued — to exclude this possibility altogether.
Secondly, the concept of ‘fiction’ is extraordinarily difficult to define. A number of very diverse attempts have been made, mainly in literary studies and philosophy, to define the term and the relevant factors that must be present for a ‘fiction’ to exist. These approaches and their mutual criticisms are so wide-ranging that it would be difficult to cover them in a book, let alone an essay. It seems that one must inevitably make a decision in favour of one approach and then apply it. It makes sense to avoid those approaches that are tailored to a medium quite distant from photography — usually ‘literature’, insofar as it can be understood as a medium.
Photography, the index, and the ‘this’ which has existed in the past
Let us begin by defining photography in a flexible and fairly broad manner: photography includes any technical process in which particular effects of electromagnetic radiation are recorded by a sensor (as a rule, the amplitude and the colour values; the phase of light is only recorded in holography).
This definition a) does not specify whether lens optics are used or whether this is a kind of photogram; b) does not specify whether the radiation corresponds to that section of the electromagnetic spectrum which is visible for humans, thus the recording of ultraviolet or infrared light (or other forms of radiation) is also included; c) does not by any means apply only to chemical recording — the general concept of the sensor refers here to any material which can be changed through electromagnetic radiation and which allows this change to be preserved (referred to as ‘fixing’ in classical photography) — so it also includes, for example, electronic sensors as in older video cameras, or quantum electronic sensors such as the CCDs normally used in digital cameras today; d) does not specify whether this is an analogue recording, i.e. one created through continual changes in the materials concerned (disregarding quantum effects), or a digital recording, i.e. one based on discrete sampling and quantification.2
The definition of photography as a record of a this, by means of the radiation emitted by this this, has led to the deduction that the recording is indexical. The term comes from the sophisticated semiotics of Charles Sanders Peirce, and refers to signs that signify by means of causality.3 The photographic sign, according to this, signifies indexically (as well as in other ways) because it is causally connected with the recorded scene by means of the radiation.4 In this sense, it is a trace. And photography means ‘writing by light’ — nothing here stipulates how this recording has taken place, whether chemically, in an analogue-electronic process (as in video), or digitally. The nature of the recording thus has no logical effect on the indexical potential. For the question of whether ‘the fictional’ is possible in photography, the analogue/digital distinction is basically irrelevant. In this discussion I will deliberately exclude the manipulation of the image, which is of course possible in both digital and analogue recordings. Even when recording the radiation of a scene digitally it is not necessary to edit it, and in this context the fundamental question is especially pertinent: to what extent can the recording of particular effects of the radiation of a scene be fictional?
As is typical of discourses that seek the ‘specific features’ of a medium, the indexical aspects of the photographic recording have encouraged theorists to formulate a number of consequences that are of direct relevance to the question of fictionality in photography.5 Two of these consequences are: a) Photographs only ever show a singular this; it is impossible to photograph ‘horses’ in general (though it is perhaps possible to draw them).6 One can only ever photograph this horse that was standing in front of the camera at that moment. According to Roland Barthes, a photograph says: “that, there it is, lo! but says nothing else.”7 And this specificity naturally implies that the this is a real this, one which can reflect radiation. It thus follows that: b) photographs only ever show things that have been — because they show a this which has, in the past, reflected radiation.8
The ‘fictional incompetence’ of photography
The reference to a past this seems on the one hand to make photography ideally suited to what can broadly be termed documentary representations in science, police work, the military, medicine, and, not least, journalism, while on the other hand rendering any fictional use impossible. This was the argument of Roger Scruton in his 1983 essay Photography and Representation. Working from the definition which we have just discussed, “that the relation between a photograph and its subject is a causal relation,” he concludes:
Of course I may take a photograph of a draped nude and call it Venus, but insofar as this can be understood as an exercise in fiction, it should not be thought of as a photographic representation of Venus but rather as the photograph of a representation of Venus. In other words, the process of fictional representation occurs not in the photograph but in the subject: it is the subject which represents Venus; the photograph does no more than disseminate its visual character to other eyes. This is not to say that the model is (unknown to herself) acting Venus. It is not she who is representing Venus, but the photographer, who uses her in his representation. But the representational act, the act which embodies the representational thought, is completed before the photograph is ever taken.9
The situation is clear, then: photography is, in Scruton’s view, characterized by a fundamental “fictional incompetence.”10The photograph of a girl dressed up as Venus, embodying Venus, adds nothing to this scene. Scruton asserts that the photographic image is equivalent to the scene photographed; in a commentary on Scruton which will be discussed shortly, McIver Lopes will call this the ‘equivalence thesis.’ The photograph here takes on no other function than to transmit the scene to the eyes of a temporally distant observer — similar to a mirror — and Scruton refers here to Sir Oliver Wendell Holmes’s often-quoted description of photography as a “mirror with a memory,” transmitting a scene to a spatially distant observer.11
Obviously it is problematic to equate the scene photographed with the photograph of the scene, although Scruton’s argument is not easy to rebut. Scruton argues that understanding a photograph as a photograph means understanding that the image is causally linked with the scene it depicts. This corresponds to the widespread tendency, in various contemporary theories, to describe the photographic sign as an index. It is therefore necessary to find an argument that will show that understanding a photograph as a photograph means more and/or something different than understanding that it points causally, indexically to a scene.
Dominic McIver Lopes’s Aesthetics of Transparency – Photography, Fiction.
The response to Scruton’s position has been a complex discussion, of which I can only give an abbreviated account here. The most important counter-argument, in my view, was expounded in a 2003 essay by Dominic McIver Lopes. Again, I can only give a brief outline of his complex argument here. He refers to Kendall Walton’s concept of ‘seeing through.’12One looks at the object through photographs; in this sense they are transparent. He writes: “To say that photographs are transparent is to say that we see through them. A person seeing the photograph of a lily, literally sees a lily.”13 This actually sounds just like Scruton’s argument, but after a few steps of argumentation the author stresses:
Seeing an object through a photograph is not identical to seeing it face-to-face. The transparency claim shows only that the interest one may properly take in seeing a photograph as a photograph is necessarily identical to the interest one may properly take in seeing the photographed object through the photograph. [...] Seeing through the surface does not block seeing the surface itself: photographic transparency is not photographic invisibility.14
McIver Lopes explicitly stresses — and shares with Scruton — the view that considering a photograph as a photograph means considering the object. Considering a photograph only in abstract terms, i.e. only with regard to its formal structure, means not considering it as a photograph. But considering a photograph as a photograph also means taking account of the difference between the photographed object and the object itself, a difference which can reveal to us aspects of the object that we would not be able to perceive in a direct encounter with the object — this would be one way of understanding Benjamin’s often-quoted phrase, the ‘optical unconscious.’ McIver Lopes describes five aspects of this difference — and these aspects’ seeming banality merely shows that his description is about the everyday perception of a photograph as a photograph. These points are: a) photographs fix a moment (if the exposure is short enough, in any case); b) photographs show us the object in its absence; c) photographs isolate the object more or less from its context and place it in a new context when the photo is exhibited; d) photographs generally show the object in the presence of the camera, which is important in photographs of people in particular, and finally, e) photographs show the object and themselves at the same time (for example in the absence of colour or in the changed size of the object).
Working from this list of differences, McIver Lopes attempts to establish an aesthetics of photography — recalling Arnheim’s efforts to base the aesthetics of film on this very difference from everyday perception.15
Contextualization and Fiction?
But another question is of interest here: can one, based on this argument, make any statement about the possibility of fictionality in photography? McIver Lopes, unfortunately, does not do so; in fact he only talks about “documentary aesthetics.”16 It seems to me that if this is to be done at all, it can only be by way of Lopes’s point c), i.e. decontextualization and recontextualization. Let us take as an example the photo of horses (see fig. 1): I can look through the photo at these animals, which have existed in the past, and can admire them for their beauty.
But I can also set up a Wikipedia entry on the species ‘horse,’ and take the same photo as an illustration for this article. Through this decontextualization and recontextualization, something strange happens: the photo of the horses (see fig. 2), since it is now illustrating an article about ‘horses’ in general, no longer signifies just these horses, but also all white horses etc.
The picture goes from being singular (or at least partly singular) to general. But this also means that it now refers to the future as well as the past. Because, of course, an article about horses in general, including its illustrations, also claims to describe all those horses that are yet to be born.
Edward Branigan explicitly links this potential for desingularization with the possibility of using photographs fictionally. In the chapter ‘Fiction’ from his study Narrative Comprehension and Film he writes: “A person in a photograph can simultaneously [be] both specific and (fictionally) nonspecific in the same way a photograph of a tiger in a dictionary can be both a specific tiger and many tigers [...].” And:
The object photographed seems to testify its own existence. Nevertheless, when the photograph is construed as a fictional entity, it becomes a picture of a nonspecific object. Our interpretation is not constrained by the particularity of detail in the photograph but acts to hold reference open while building complex predicates about what the photograph pictures.17
This hints at a possible answer to Scruton’s damning judgement of ‘fictional incompetence’: even a photo of an object or process which was not intended to embody, represent or perform a fiction at all (as opposed to Scruton’s example of the photo of the girl embodying Venus, in which the scene being photographed is supposed to represent a fiction) can nonetheless, under certain conditions, be understood to be the representation of a fictional entity. Thesis: if it can be successfully proven that a photograph as a photograph, by virtue of its potential for de- and recontextualization, can transform the photo of an object intended to be non-fictional into the photo of an object which can be read as fictional, then this refutes Scruton’s argument and shows that the photograph has genuine fictional competence.
And yet Branigan’s phrasing poses new questions. What does “a photograph is construed as a fictional entity” mean? Obviously a cognitive process of interpretation plays a part here, a process apparently concerned with “building complex predicates about what the photograph pictures.” What does this mean?
Let us first return again to the clearly central issue of the de- and recontextualization of the photographed object. If the question of the possible fictional competence of the photograph hangs on the fact that a photograph decontextualizes the photographed object and then recontextualizes it in a different way, then fictional competence clearly does not depend solely on one photograph, but on one (or several) photograph(s) in a specific context. An isolated photograph, whatever it depicts, is in itself neither fictional nor non-fictional.18 This can also be turned against Scruton’s argument on the fictional incompetence of photography: his argument that the fictionality of a photo-of-a-girl(-as-Venus) depends on the pre-photographic scene of the girl-as-Venus already implicitly presupposes a context in which the photo is not understood as the record and documentation of a specific girl-as-Venus style of costume (which is by no means a curious special case, but rather the bread and butter of art and media historians). This means that the thesis offered above can be negatively confirmed: no photograph of a fictional scene can guarantee that the scene will be understood as fictional. Which suggests, conversely, that no photograph can guarantee that it will be understood as a trace or document — this is evident both in the fact that photographs used in journalism or science require interpretation,19 and in the phenomenon of ‘docufiction’, i.e. fictional pseudo-documentaries.20 So the question of the fictional competence of photography must be reformulated into the question of what kinds of de- and recontextualizations invite what we might call a ‘fictional way of reading’ for a given photograph. And this question, of course, implies another, namely: what does a ‘fictional way of reading’ actually mean?
We have already obtained one clue — and this takes us back to Edward Branigan. He wrote that if “a photograph is construed as a fictional entity, it becomes a picture of a nonspecific object.” Branigan bases this argument on a specific interpretation of the concept of fiction proposed by Hartley Slater in 1987, a concept which is of interest here because — in contrast to the definition given at the outset of photography as an immediate form of reference — it stresses that reference understood as fiction is an only partially determined form of reference.21 Branigan:
Interpreting a symbol fictionally requires that one qualifies the immediacy of the symbol itself: its material presence must not imply an immediate reference, nor a simple reference to something atomic, nor indeed any reference at all, much less one that is true or false in our familiar world. Further information and calculation is required. [...] A fiction does not determine exactly which object or objects it represents, and this openness is what distinguishes fictional reference from other sorts of reference.22
In other words: a photo of a horse is, on the one hand, the photo of this specific horse, but read fictionally — e.g. in the context of an illustrated story about a nobleman — it becomes, on the other hand, the picture of a horse. Yet, unlike the general horse picture in a lexicon, the context is not “overwhelmingly denotative” and does not act “to limit and confine the range of possible referents to a specific class.”23 Instead, we as observers and readers have the task of finding out in the course of the story what attributes the horse has. Is it just a horse on which the nobleman rides? Is it a sort of companion, or a symbol for friendship? Perhaps it will turn out to be an enchanted prince? Thus ‘fictional’ does not mean anything that is interpreted as fiction — that would be a tautology — but anything that is processually interpreted as a partially determined form of reference — and that interpretation depends on the context, not on whether this is a record of a scene which has already been staged.
Branigan relates fiction to a specific cognitive process in which, step by step, an initially un-determined reference is augmented.24 This process has to be learned: Branigan stresses that, according to some studies, children can only understand what fiction is from the age of about seven. The process is generally triggered by conventionalized markers (in literature, for example, the phrase ‘Once upon a time’)25 which do not, however, guarantee that one will be reading fiction, just as the markers of the documentary can be simulated (the example of docufiction again).
An interesting example of this can be found in the work of Douglas Crimp, who has been described by one commentator as the “first theorist of staged photography.”26 In 1979 Crimp wrote of Cindy Sherman’s Untitled Film Still #21 (see fig. 3):
Here is a picture. It shows a young woman with close-cropped hair, wearing a suit and hat whose style is that of the 1950s. She looks the part of what was called, in that decade, a career girl, an impression that is perhaps cued, perhaps merely confirmed by the fact that she is surrounded by the office towers of the big city. But those skyscrapers play another role in this picture. They envelop and isolate the woman, reinforcing with their dark-shadowed, looming facades her obvious anxiety, as her eyes dart over her shoulder ... at something perhaps lurking outside the frame of the picture. Is she, we wonder, being pursued? But what is it, in fact, that makes this a picture of presentiment, of that which is impending? Is it the suspicious glance? Or can we locate the solicitation to read the picture as if it were fiction in a certain spatial dislocation — the jarring juxtaposition of close-up face with distant buildings — suggesting the cinematic artifice of rear-screen projection? [emphasis added]27
Crimp’s reading of Cindy Sherman’s photo shows the process of interpretation towards a partially determined reference (“to read the picture as if it were fiction”). He asks himself what exactly can be seen. And: “We do not know what is happening in these pictures, but we know for sure that something is happening, and that something is a fictional narrative. We would never take these photographs for being anything but staged” [emphasis added]28. We do not know exactly what is happening, but we try to find out – and this constitutes a “fictional narrative.” Sherman’s title, Untitled Film Still, does not specify what sort of film this still comes from or, to be more exact, what film it might be alluding to. Nor does it specify whether it is actually based on a film at all. This non-specificity marks an entry into the fictional reading of the photo as only partially determined. We try, in the process of interpreting, to devise around the picture a diegesis that gives us a sufficiently consistent explanation. Of course the example of Untitled Film Still will not convince a sceptic like Scruton, since the pre-photographic scene itself, is, once again, staged — though hardly to the same extent as the girl-dressed-as-Venus. It is therefore easy to imagine a ‘documentary’ photo of a street scene that looks exactly the same and which nonetheless, through its title and context, does not serve (or is not meant to serve) as a document, but instead initiates a process in which the reference is interrogated.29
The context here, in Crimp’s work, is film, the field from which Branigan also takes his main examples. This points to a fundamental problem in the fictional reading of photographs — and also of other static images. Fludernik notes: “Fictionality typically surfaces in narrative form (including narrative poetry, drama and film); it is not generally employed to define poetry, sculpture or music, and in painting is restricted to specifically narrative representations.”30 As a rule fiction is connected to the process of narration, because this process allows us, step by step, to form, confirm, or reject assumptions about the referential status of the objects presented.31 For an image that is not extended temporally, such as a painting or a photo, this process is more difficult to create. Individual photos can less easily tell stories, and it is thus more difficult to describe them as fictional. But as the example given above suggests, it is possible to think of strategies in which this lack of genuine temporal extension is compensated for: the series, the picture story, installation-based arrangements containing photographs.32 Such processes facilitate fictional readings of photographs.
Medium and sign
I wish, lastly, to discuss a final consequence that confronts the question of fictionality in photography. Clearly this question is an excellent point for showing the difference between medium and sign.33 In the above discussion I had moved unconsciously from the question of the photographic medium — though it was defined with a semiotic term as index — to the question of processes of signifying, i.e. semiosis. In terms of medium and media technology, photography only seems suited to the specific, singular and past; nonetheless it can signify things that are unspecific, general, and future. One could say, with Wolfgang Iser, that in the “act of feigning […] the recurring reality [becomes] a sign.”34 If this is so, however, it has ruinous consequences for the concept of medium specificity, assuming firstly that one accepts this concept at all, and secondly that one understands medium specificity to mean that, on the basis of the clearly describable structure of a given medium, positive conclusions can be drawn about what can be done with it, or even normative demands for desirable aesthetic strategies (as in the case of High Modernism in painting, which took inspiration from Clement Greenberg’s texts in the field of art criticism). Instead, media specific analyses seem to draw negative boundaries, defining what definitely cannot be done with a given medium (e.g. a photograph is not suitable for receiving radio programmes), while the kinds of semiotic processes that can be produced with a given medium are not determined. The example of photography shows that the possible semioses can contradict the limitation on such semioses that is thought to be deducible from the medium’s specific qualities. Branigan uses an expression which seems particularly apt for photographs which are read fictionally in certain contexts: “to interpret a symbol fictionally is to operate in a precarious, intermediate zone between sets of possible references [...] and a specific reference.”35
From this one could — tongue in cheek — derive a kind of normative aesthetics after all. If a given photograph always exists in the tension between the technological limitations of the medium and the more open semioses, then a photograph that concentrated reflexively on displaying the indexical relationship to the world would by no means be a particularly successful example for an aesthetic of photography. On the contrary, staged photographs, photo series, photographic picture stories, and installations with photographs (i.e. photographs which encourage fictional readings) would be suitable strategies for displaying this tension between medium and sign, working on it, and continually finding new forms for it. This is precisely what Cindy Sherman shows: she photographs herself (i.e. in the right context the person depicted can be identified quite specifically as Cindy Sherman), but at the same time she stages herself in various film stills with fairly non-specific contexts, hence ‘untitled’. Crimp writes, after all, about “read[ing] the picture as if it were fiction.” He does not write that the Untitled Film Still is fictional, but that we (can) read it as if it were fictional — thereby referring pointedly to this oscillation. Ultimately, this oscillation can come as no surprise:
In order to be intelligible, fictional or imaginary universes have to be related to the world. An embodied sign, for instance a work of fiction or a painting, that is totally disconnected from — or better yet, ‘unconnectable’ to — our world is not only an impossibility but also would be beyond intelligibility. In this sense the ultimate object of our representations, including fiction, can only be reality (the one and only).36
Looking at it this way, one can draw an even more radical conclusion: precisely because photography possesses a particular indexical potential, one can use it for the construction of elaborate fictions. Fictional films — as long as they are based on photographic processes — operate in the same way. It is because they are based (partly) on indexical photographic images that the fictions brought forth with them are able to be so convincing, and so oppressive. Scruton did not understand this; Cindy Sherman, on the other hand, shows it with great clarity.
Photography and Fictionality von Jens Schröter steht unter einer Creative Commons Namensnennung-NichtKommerziell-KeineBearbeitung 3.0 Unported Lizenz.