About CA
Contact CA

Search Journal

Editorial Board

Permission to Reprint


Site Map


Experiencing Photographs Qua Photographs:
What's So Special about Them?

  Jiri Benovsky

Merely rhetorically and answering in the negative, Kendall Walton has asked:  "Isn't photography just another method people have of making pictures, one that merely uses different tools and materials; cameras, photosensitive paper, and darkroom equipment, rather than canvas, paint, and brushes? And don't the results differ only contingently and in degree, not fundamentally, from pictures of other kinds?"

Contrary to Walton and others, I  answer with a resounding "Yes" to Walton’s questions in this article.  It is a widely shared view that photographs are somehow special and that they fundamentally differ from hand-made pictures such as paintings, both from a phenomenological point of view (in the way we experience them) and an epistemic point of view (since they are supposed to have a different that is, greater, epistemic value from paintings that gives us a privileged access to the world).  I almost reject the totality of these claims and, as a consequence, there remains little difference between photographs and paintings.  As we shall see, “photographs are always partly paintings,” a claim that is true not only of retouched digital photographs but of all photographs, including traditional ones made using photosensitive film and development techniques.

digital photography, metaphysics, painting, perception, phenomenology, photography

1. Introduction

Merely rhetorically and answering in the negative, Kendall Walton[1] has asked:  "Isn't photography just another method people have of making pictures, one that merely uses different tools and materials; cameras, photosensitive paper, darkroom equipment, rather than canvas, paint, and brushes?  And don't the results differ only contingently and in degree, not fundamentally, from pictures of other kinds?"

Contrary to Walton and others, I answer with a resounding "Yes" to Walton’s questions.  It is a widely shared view that photographs are somehow special and that they fundamentally differ from hand-made pictures such as paintings, both from a phenomenological point of view (in the way we experience them) and an epistemic point of view (since they are supposed to have a different, that is, greater, epistemic value than paintings that gives us a privileged access to the world).  I almost reject the totality of these claims, and as a consequence there remains little difference between photographs and paintings.  As we shall see, “photographs are always partly paintings,” a claim that is true not only of retouched digital photographs but of all photographs, including traditional ones made using photosensitive film and development techniques.

2. Perception of pictures

Let me start with something that has nothing to do with photography but that concerns ordinary perception.  Suppose you see a bottle of beer on a table in front of you.  Setting aside sceptical scenarios (hallucinations, Descartes' evil demons, and the like), what you see, according to many standard ontologies, is the bottle of beer because there is a bottle of beer in front of you, and your perception is somehow caused by the bottle (along with other factors involving light, your eyes, your optic nerve, and so on). Eliminativism is a metaphysical theory that comes in many different varieties but all of them have in common the claim that entities such as bottles of beer do not exist.[2] 

According to eliminativism, there are only fundamental components arranged bottle-of-beer-wise.  The nature of these fundamental components is subject to controversy and varies from one version of eliminativism to another (particles, properties, or other); for our current purposes let us simply call them 'atoms.' The central claim of eliminativism is, then, that atoms arranged bottle-of-beer-wise can do all the metaphysical work bottles of beer can do, and consequently bottles of beer can be eliminated from our ontology without any loss of explanatory power.  For instance, bottles of beer can be bought and sold, they can be used as weights on a paperback book on a windy day, or they can occupy a rather well-defined spatio-temporal region in your fridge – but atoms arranged bottle-of-beer-wise can do all of that too. No need then to postulate extra entities, namely bottles of beer, in one's ontology.

Furthermore, eliminativists typically claim that their view is not contrary to common sense and that it actually is a rather intuitive one.  This is where an objection concerning ordinary perception comes into the picture.  Indeed, on the one hand eliminativists say that there are no bottles of beer but on the other hand they want to say that we see them even when we are not under an evil demon's influence or hallucinating.  This is a seeming contradiction.  The correct reply to this worry, nicely put by Trenton Merricks, is the simple but significant claim that our experience is the same whether there is a bottle of beer in front of us or whether there are atoms arranged bottle-of-beer-wise.[3] 

Thus, the phenomenal character of our experience is neutral in relation to the eliminativist's metaphysical claim.  Our experience is caused, in short, by light reflected by a bottle of beer, and since atoms arranged bottle-of-beer-wise reflect light in the same way bottles of beer do, our experience is qualitatively the same in both cases.  The fact that we have non-hallucinatory perceptions  of bottles of beer thus cannot be used as an argument against eliminativism.  The general idea here is that our sensory experiences can be accounted for by more basic and genuinely fundamental (and existing) entities, such as atoms arranged x-wise.  This means  there is no need to postulate a further entity such as x.

In this paper, I am not interested in eliminativism but I am interested in what the situation described above teaches us about phenomenology.  What it teaches us is that phenomenology parts ways from epistemology or metaphysics. Whether we know that there are (or aren't) bottles of beer or whether there are any (or not) simply does not matter for what our experience is like. Beliefs we have about what there is and how things are are irrelevant to what we see (perceive, in general) in a purely qualitative and phenomenal sense.  The eliminativist's response to the objection above illustrates this point nicely, I think.  Beliefs do not intervene in what we see.

Now we can talk about photographs.  A first, simple, and perhaps even trivial claim I want to put on the table is the following: photographs and paintings are both pictures and are both experienced in the particular way in which pictures are, but there is no significant difference in our visual experience when we look at a photograph or at a painting.  What we see is simply a picture. Whether a picture is a sharp photograph or a hyper-realistic painting, or whether it is a digitally manipulated, heavily retouched  photograph or an impressionist painting, our visual experience may be the same; indeed, for the viewer these cases may sometimes be  visually (that is, phenomenally) indistinguishable.[4]  

The point here is not to say that we can make mistakes (although we can) and take a photograph to be a painting or a painting to be a photograph.  Rather, what I want to highlight here is the fact that our visual experiences qua phenomenal visual experiences are of the same kind:  they are visual experiences of pictures.  This simple fact shows us that, here again, our phenomenology exists apart from what we know about the picture (especially about the way it was produced) or from the way the picture is (its metaphysical nature).

What I want to do here is to clearly distinguish between phenomenological issues on the one hand and epistemic and metaphysical ones on the other.  This is not always the case, as for instance Robert Hopkins and Mikael Pettersson both recently put it, independently of each other:

[Traditional] photographs have an epistemic status that “handmade” pictures such as drawings, paintings, and etchings do not.  (Hopkins [5])  Both photographs and handmade pictures can be sources of knowledge, but photographs offer us a way of finding out about the world that is more secure than that offered by handmade pictures....[T]his epistemological difference is accompanied by a difference in phenomenology:  we experience photographs differently from the way in which we experience other pictures. Photographs seem to put us in a relationship to the objects they depict in a way that is somehow more intimate and direct than the relationship we bear to the objects that handmade pictures depict....What we see in traditional photographs is, of necessity, true to how things were when the photograph was taken....It is this that explains traditional photography's special epistemic status and the special experience it instills.  (Hopkins [6])

[M]ore than whether photographs actually provide epistemic access to what they depict, it is viewers' beliefs that they do so that matter for the phenomenology of photography.  (Pettersson [7])

The link between phenomenology and epistemology is obvious in both citations.  Both Hopkins and Pettersson mention the influence one's beliefs allegedly have on one's phenomenal experience when perceiving a photograph.  But as I tried to show above, it is a mistake to mix the two issues in this way. What we see (that is, what the phenomenal character of our visual experience is like) is one thing, and what we believe to be the case about what we see is another.  Perhaps I am insisting too much on the trivial, and perhaps I am not interpreting the citations above in a charitable way.  But perhaps once we do make the conceptual distinction between phenomenology, epistemology, and metaphysics more precisely, we will have a better starting point for the discussion of the alleged differences between photographs and paintings; namely, we learn that it is not a phenomenological affair, but an epistemic and metaphysical one.  These are the claims to which I will turn my attention.

3. Photographs and reality

The difference between paintings and photographs is that, typically in the case of photographs, when we know that we are looking at a photograph, we have a piece of knowledge about a metaphysical truth that we don't have in the case of paintings. More precisely, the relevant epistemic situation is that we know how the picture was produced, and this gives us access to a simple but important metaphysical truth:  there was something before the camera.  This is a claim that is widely shared by virtually everyone, including Walton, Hopkins, and Pettersson. Indeed, given the way photographs are made, it is necessary that, at the beginning of the causal process that leads to the existence of a photograph the is something that has been photographed.  In short, that something  reflected light that was then recorded by a camera.

Now, what I want to insist on is how weak this claim is.  Let us start by having a look at these three photographs I took of a bottle of beer.

Photo 1:  a photograph of a bottle of beer, f/29, 1sec, 28mm

Photo 2:  a photograph of a bottle of beer, f/3, 1/160, 16mm

Photo 3:  a photograph of a bottle of beer, f/8, 1/80, 36mm

These three photographs are of the same subject, under the same light conditions, taken at (almost) the same time.  They are photographs of “the same metaphysical reality.”  All three are such that we have the piece of knowledge about the metaphysical truth on which everyone agrees: something was there.  The weakness of this claim is most obviously apparent in Photo 1 where the “something” is unrecognizable (because of a long exposure and a shaking hand), but Photo 2 and Photo 3 illustrate the claim I want to make, as well, namely,  that in the case of a photograph, when we know that we are looking at a photograph, we know that there was something that has been photographed but we do not know how this something was. 

Sometimes we do not even know what this something was, as in the case of Photo 1, but this is only a matter of degree.  It is because we know so little about how it was that we are not even able to see what it was.  Always, we do not know how the something was, for the simple reason illustrated by Photo 2 and Photo 3, that the entities that have been photographed are never pictorially represented (depicted, shown, visually given to us) as they are “in the world.”  Indeed, as a matter of necessity, in any normal process of creation of a photograph, there are steps where some features of the entities represented are altered or even “erased” and replaced by other apparent features. 

All three photographs above, for instance, “misrepresent” the colors of what they are photographs of since they are black and white.  Photo 2 “misrepresents” the entities located in the background by representing them as blurred because of a shallow depth of field.  Both Photo 2 and Photo 3 “misrepresent” the shape of the bottle (most apparent in the case of Photo 2, but Photo 3 is actually deformed as well) because of the choice of a particular focal length; also, all photographs always represent what they are photographs of only from a certain angle, and so on.

An important thing to note is that all of these “misrepresentations” are due only to a normal use of traditional and standard photographic techniques:  aperture, shutter speed, angle of view, focus, and focal length.  Photo 2, for instance, is thus no less normal than Photo 3, while Photo 1, relevantly, is no less normal than the other two.  It would simply be entirely arbitrary to claim the contrary.  No special effects have been used here, only standard settings on a standard camera.[8]

Now what we see here is that even normal photographs, using standard settings and photographic techniques, tell us in principle very little about the “true properties” of what they depict.  The shape of the bottle, for instance, is “misrepresented” in all three photographs above (and so are colors, sharpness, etc.).  Thus, again, since we know that we are looking at photographs and not at paintings, we know that there was something that was photographed but we do not know how it was.  For instance, by looking at the photograph of the bottle, we do not see what  its true shape was.  We can perhaps guess at it, or even calculate it if we know all of the settings and the distance from which the photograph was taken, and if we know the equations that allow such a calculation.  But even if something like this were at least partly possible, this would not be a normal way to interact with photographs (and it would definitely not work very well in the case of colors or of a blurred background).

A photograph does not give us the world.  It gives us a pictorial representation that in normal and standard cases misrepresents the world in a more-or-less interesting way.  A photograph tells us that there was a world, and in some cases (but not always, as in Photo 1) tells us approximately how the world was.  The latter is often true of paintings, as well, and only the former metaphysical claim constitutes a principaSl difference between paintings and photographs since it can (but does not have to) be false in the case of paintings.

4. Photographs and photographers

Perhaps then, as many have claimed, the difference between photographs and paintings comes not from the resulting picture but from the way it was produced, in the sense that photographs are made mechanically without human intervention, while paintings are necessarily subject to human intentions, beliefs, and interventions (see these quotations: [9]).  In a previous article ("Three kinds of realism about photographs," 2011), I argued at length that this is incorrect, so let me only focus on the main point here:  it is not possible not to make decisions when one takes a photograph.  Any time a photograph is made, a decision has to be made at the very least about aperture, shutter speed, focal length, exposure, and usually many other settings. These decisions can be either purposefully, consciously, and manually made by the photographer herself, or they can be made by the engineers who programmed the automatic mode of the camera that a Sunday snapshooter can use to avoid making these decisions by herself.  But in any case, human decisions and human interventions are unavoidable.

These decisions make a big difference to the resulting picture, as Photo 1, Photo 2, and Photo 3 illustrate.  Indeed, the differences among these three pictures are entirely because of my decisions. Big aperture can be chosen to create a shallow depth of field, resulting in a blurred background. Long exposure time can be used to produce photographs like Photo 1.  A wide angle lens (short focal length) can be chosen to produce deformations like those in Photo 2.  And so on.  These tools, as well as many others, are the standard tools the photographer is meant to use to produce a picture according to how she wants to represent the metaphysical reality in front of the camera and not to how the reality is.  In the same way painters can (and often do) give us a pictorial representation of the world according to how they want us to see it.  Photographers use the various settings and techniques at their disposal to make us see the world the way they want to show it.

Keeping this in mind, we see here again how weak the epistemic and metaphysical claim is.  Indeed, in the case of photographs the claim that "there was something" is necessarily true, while it is only contingent in the case of paintings.  But that's about the only principal difference between these two types of pictures, and as we have seen above, it is not a big one.  In both cases, the entities that are pictorially represented are only given to us after some human decisions have been made to represent them in such-and-such a way.

5. Photographs and (post-)production

Furthermore, both digital and traditional photographs require a certain amount of “post-production steps” where either a RAW file is converted into a final image file or a negative is developed to produce a final picture on photographic paper.  These manipulations, digital or chemical, are necessary to  produce a photograph; without them no photograph would even exist. Indeed, after the shutter has been pressed, there is only a negative or a RAW file, but these are not yet photographs.  Additional steps need to be taken in order to bring a photograph into existence.  These steps can be done quickly inside the body of a camera (as in a Polaroid camera or in most compact automatic digital cameras), or later manually (in a darkroom, or on a computer), but this practical difference does not constitute a principal   difference.  However, if these steps are taken, and they have to be taken, here again they involve human decisions (either the photographer's own or somebody else's).

Minimally, these are decisions about contrast, colors, and brightness, which are decisions that have to be taken to produce any photograph at all.  But these decisions can also concern more sophisticated techniques in order to produce a particular effect (such as a sepia effect, for instance) or to chemically or digitally manipulate the negative or the RAW file to produce a retouched photograph.  Such retouches can be small and light or they can be heavy, involving instruments used by the photographer to finish her work; that is, to better achieve her goal of showing us the world the way she wants us to see it. 

In part, then, these manipulations are necessary.  They are an essential part of any normal process of creating  a photograph. Partly they are contingent and the photographer can choose to take such additional steps or she can choose not to do so. How many such steps must be taken before the resulting picture ceases to be a photograph and becomes a painting is a vague matter.[10]  But long before we reach that limit, we are in a position to see that the mere existence of any normal photograph requires some amount of post-production technique and human decision, and that in standard cases the amount of chemical or digital manipulation goes well beyond these minimal necessary steps.  As before, we see here again how human intervention plays a crucial role in photographs coming into existence and in the picture that results; a way that tells us not how the world is but rather how the photographer wants us to see it.

6. Photographs are always partly paintings

What stems from the preceding sections, I hope, is a clear picture of the nature of photographs and of the way we produce them, experience them, and interact with them.  I hope it is clear enough  to render the following claim obvious: photographs are always partly paintings.

The photographer deals with a metaphysical reality in front of her camera (and the difference between her and a painter concerns the fact that this is necessary for her and only contingent for the painter), and uses the various photographic tools and techniques at her disposal to create a pictorial representation of that reality. These tools are such that they require her to make important decisions.  Thus, even if she wanted to, she could never simply represent reality.  Rather, she necessarily has to misrepresent it, and by making such-and-such a decision rather than another, she shows us the world, again, not as it is but as she decided to show it. Photo 1, Photo 2, and Photo 3 are examples of such decisions. Thus, not only photographs always partly paintings, but photographers are always partly painters.  This is true even of those who limit themselves to the strict (and necessary) minimum when it comes to post-production.

What all of this shows us is how small the principal difference is between photographs and paintings.[11]   Of course, they are pictures produced using different tools (in a narrow sense, a painting is made using paint and is, in this sense, trivially different from a photograph), and the epistemic and metaphysical claim "We know that there was something" is only contingent in the case of paintings.  Furthermore, when it comes to paintings, a change in the reality will only make a difference for the painting if it also makes a difference in how the painter sees the reality.  In the case of photographs, a change in the reality will make a difference in the photograph even if, say, it goes unnoticed by the photographer.  This is true provided that the change is big enough to be visually noticeable in the resulting picture.

Despite these differences, we have seen that, first, there is no phenomenological difference between these two types of pictures.  That is, there is no difference in the qualitative experiences we have of them and, second, the metaphysical claim is a weak one.  The weakness of this claim, I suppose, will become more obvious and significant with the evolution of digital photography, because the ease with which digital manipulation can be done during production and post-production will make photographers become even more painters than they already are.[12]


Jiri Benovsky

Jiri Benovsky is a “Scientific Collaborator” at the University of Fribourg, Switzerland.  His interest in the philosophy of photography comes from the fact that he is both a philosopher (see www.jiribenovsky.org) and a photographer (see www.benovsky.com).

Published on August 19, 2013.


[1] Kendall Walton, "On Pictures and Photographs: Objections Answered" in Film Theory and Philosophy, eds. Allen and Smith (Oxford University Press, 1997), p. 67-68.

[2] See for instance Peter Van Inwagen, Material Beings (Ithaca: Cornell University Press, 1990), Mark Heller, "The Donkey Problem," Philosophical Studies, 140(1), (2008), 83-101, or Trenton Merricks, Objects and Persons (Oxford: Clarendon Press, 2001), pp.8-9.

[3] Trenton Merricks, Objects and Persons.

[4] I will have more to say about digital manipulation later.

[5] In Hopkins' view, digital photographs are different in this respect:  they are closer to paintings than traditional photographs made by using photosensitive film.  In my view, which I develop below, I will treat both types of photographs in the same way.  I insist on equal treatment also in my article, "Three kinds of realism about photographs," The Journal of Speculative Philosophy, 25:4 (2011), and in my manuscript, "The Limits of Photography."

[6] Robert Hopkins, "Factive pictorial experience:  What's special about photographs?" Noûs, 46 (4), (2012), 709-731.

[7] Mikael Pettersson, "Depictive Traces:  On the Phenomenology of Photography," The Journal of Aesthetics and Art Criticism, 69:2, (2011), 185-196.  Quote on p. 191.

[8] Namely, the Canon EOS 7D, the lens used here being the Canon EF-S 10-22.  The fact that this is a digital camera is irrelevant here, since the very same settings (and results) are standard on traditional film cameras, as well.

[9] "[...] the relation between a photograph and its subject is a causal relation.  If a is the cause of b, then the existence of b is sufficient for the existence of a."  Roger Scruton, "Photography and Representation," Critical Inquiry, 7:3 (1981), 588, my italics.

"[...] photographs are things, whose state – i.e. the configuration of marks on their surfaces – depends, as Walton points out, belief-independently and counterfactually on visible features of what they are photographs of."  Mikael Pettersson, p. 190, my italics.

"[...] by a mechanical reproduction in the making of which man plays no part.  The solution is not to be found in the result achieved but in the way of achieving it.…For the first time, between the originating object and its reproduction there intervenes only the instrumentality of a nonliving agent.  For the first time an image of the world is formed automatically, without the creative intervention of man."  André Bazin, "The Ontology of the Photographic Image," transl. Hugh Gray, Film Quarterly, 13:4 (1960), 7.

"Traditional photography, in contrast [with hand-made pictures], involves a causal chain free from the influence of people's beliefs and experiences […]."  Robert Hopkins, Noûs (2012).

"Photography overcame subjectivity in a way undreamed of by painting, one which does not so much defeat the act of painting as escape it altogether:  by automatism, by removing the human agent from the act of reproduction."  Stanley Cavell, The World Viewed (New York: The Viking Press, 1971), p. 23, my italics.

[10]  This is an issue I discuss in detail in a manuscript in preparation.  

[11] To make it clearer, the claim here is not that photographs and paintings are the same thing and that they can be identified.  The claim is that the difference between them is smaller than what we might have thought.

[12] I would like to thank Rob Hopkins, Lynda Gaudemard, Mikael Pettersson, Gianfranco Soldati, and an anonymous referee of this journal for interesting discussions and comments that helped me think about the issues I raise in this paper.