welcome covers

Your complimentary articles

You’ve read one of your four complimentary articles for this month.

You can read four articles free per month. To have complete access to the thousands of philosophy articles on this site, please

The Arts

How Do Pictures Represent?

Marek Soszynski considers whether it couldn’t look like resemblance after all.

How do pictures represent in the way they do? How come a picture of a cat, for example, is indeed of a cat – actually has a cat as its object? Or to put it the other way around, how is a cat pictorially represented or depicted by its picture? If your first thought was to reply that it is simply because the picture looks like or resembles what it represents, then think again. I shall call this idea Resemblance, with a capital R. The Resemblance theory says that resemblance straightforwardly explains pictorial representation.

This idea is highly problematic. To give a couple of brief examples; we do not know what certain historical characters such as John the Baptist looked like, yet we can still successfully represent them in pictures; likewise with imaginary or impossible objects, which by their very nature are unavailable to be resembled. I could provide further examples, but instead I want to lend some support to Resemblance; not by defending it comprehensively – perhaps that cannot be done – but by pointing out a couple of major misconceptions by its detractors.

The two most important modern writers on pictorial representation have been Nelson Goodman (d.1998) and Richard Wollheim (d.2003). Both have written at length on the subject, and both linger over Resemblance only to dismiss it, so it is with them that we must engage. We shall begin with Goodman. Right from the start of his Languages of Art, he goes on the offensive against Resemblance. A particular target of his is the different logical relations of representation and resemblance. While resemblance is symmetric and reflexive, representation is usually neither. Thus, a man resembles himself, but ordinarily does not represent himself. Most inconveniently for supporters of representation being explained by resemblance, the picture of the cat will resemble the cat as much as the cat resembles its picture; yet it is the cat that is represented by the picture, not the other way round.

One weak point in Goodman’s argument is in his characterisation of ‘resemblance’. Like most writers in this field, he leaves resemblance largely unelaborated beyond the (implied) textbook definition of ‘a symmetric relation between items sharing properties’. But is that really what we as laymen are talking about when we say of a picture of a cat, that the picture looks like the animal? Rob van Gerwen in his Art and Experience gives us an alternative account. He considers cases of attributed resemblance, and points to a single-directedness from an object that is standard-setting and primary, so that instead of a symmetry between things there is sequence – as when a boy is said to take after his grandfather, or indeed, a portrait to resemble the sitter.

It is easy to come up with our own examples in support of van Gerwen. Consider the Queen and a coin that shows her head. Compare the ordinariness of saying that the coin resembles the Queen with the oddness of saying that the Queen resembles the coin. This appears to support van Gerwen’s notion of single-directedness from what is primary.

Not convinced? If instead you think it is the Queen who resembles the coin, then all you have done – quite legitimately, though perhaps unusually – is to make the coin primary. What was the resemblee (thing being resembled) is now resembler (thing doing the resembling), and vice versa: resemblance’s sequentiality is maintained. We should remember Mr Pooter’s clever remark in the comic novel The Diary of a Nobody by George and Weedon Grossmith: “Without an original there can be no imitation.” A similar thing applies to the resemblance attributed between pictures and their objects: without a resemblee there can be no resemblance. In this way the inconvenient apparent symmetry of the resemblance relation is broken.

Before we finish with Goodman, I must just mention a notable passage from early on in Languages of Art (p5): “A Constable painting of Marlborough Castle is more like any other picture than it is like the Castle, yet it represents the Castle and not another picture – not even the closest copy.” Elsewhere he claims that “a postcard of Paris is about as unlike the city as can be.” This is all good knockabout stuff: but actually, unless a context is provided, the degree of resemblance between items cannot be determined. Here Goodman has provided his own context: what matters most here is the physical size and suchlike details of painting and castle, postcard and city, rather than any ‘image similarity’. He can adopt this context if he wishes, but it seems perverse, as if he is wilfully misunderstanding the nature of pictures for only so long as he is arguing against Resemblance.

Wollheim’s principal work on aesthetics is Art and its Objects, but he has written about pictorial representation in many other places besides. Here I will deal with just a part of his criticisms of Resemblance. Remember, Resemblance has it that a picture represents what it does by being like the thing it represents. But Wollheim says that this is to get the matter reversed, since it is not the picture that is being compared to something absent, but the object in the picture. But that would mean that the object in the picture (the painting’s subject or sitter) has already been identified. There’s no need for any further search for its resemblance to something else:

“Sometimes, it is true, we exclaim of a drawing, ‘But how exactly like A!’ But this is not the counter-example to my argument that it might at first seem to be. For if we try to expand the ‘this’, of which in such cases we predicate the resemblance, we are likely to find ourselves much closer to ‘This person is exactly like A’ than to ‘This configuration is exactly like A’. In other words, the attribution of resemblance occurs inside, and therefore cannot be used to explain, the language of representation.” Art and Its Objects

We can argue with Wollheim over the real significance of this. Were we to encounter someone who disagreed with us about a portrait – over whom it represents, or whether it represents at all – how could we win that person over? “Look at the profile” we might say, or “look at the shape of the head.”

Now while ‘head’ and probably ‘profile’ are still within the obvious language of representation, we should discriminate between the necessity and the convenience of such terms. If the word ‘head’ escapes us, as it may do if we were struggling to speak in a foreign language, we could say ‘blob’ or ‘squashed circle’ instead. Are those terms within the language of representation? And if we can use them, then isn’t Wollheim simply wrong? We are not directly identifying a head, but a squashed circle.

Firstly, if ‘squashed circle’ is disallowed, then it would hardly be possible to talk about the surface of the canvas in non-representational terms at all, and the anti-representation speaker would have nothing to say anyway. If however ‘squashed circle’ is allowed – as surely it must be – then there is indeed something we can say to someone who does not see the resemblance we see, ie “Doesn’t that squashed circle look like A?” The discussion need not be prejudiced by being within the everyday language of representation. ‘Head’ and ‘profile’ etc may be used for convenience or for economy of expression, but they are not unavoidable terms for the pro-Resemblance discussant.

Like many other writers on depiction, Wollheim also goes astray in another respect. All kinds of things can be represented, he tells us, from abstractions such as cuboids, to complex events such as the aftermath of a storm.

Here we may wonder whether the concept of ‘representation’ isn’t being overworked. Even when we narrow our concern to aesthetics, still it covers a range of concepts. Thus sometimes for the sake of clarity and precision we ought to use the idea of symbolisation instead. A picture may represent a heart which in turn symbolises love. It would be confusing to claim that the picture directly depicts love. But hasn’t Wollheim made the same sort of error with the breadth of his idea of what ‘pictorial representation’ is supposed to cover? We may wonder whether a picture of the aftermath of a storm doesn’t involve some sort of symbolisation or second-order representation. The picture will show objects – a seascape, damaged rigging and so on. How those objects are themselves depicted is problematic enough, without claiming that the events in which they have taken part are directly depicted too.

A similar mistake is committed with respect to caricatures. A caricature may not look much like its target at all. Yet some theorists still like to claim that the victim is directly represented by its caricature. I would say that this is another example of increasing the strain on ‘representation’. In the first analysis a caricature represents, say, a man with bulging eyes. How that representation is then a caricature of its target is another question, and unfortunately beyond my scope here.

The notion of representation has become so general that accounting for it has become convoluted and actually impossible. That’s the problem with discussions of pictorial representation which dismiss Resemblance, like Goodman’s and Wollheim’s. On the one hand they take ‘representation’ as broadly as possible, on the other hand they take ‘resemblance’ as narrowly as possible. No wonder, then, that resemblance is not up to the task of explaining pictorial representation.

© Marek Soszynski 2006

Marek Soszynski did an MPhil. in Philosophy at Birmingham University in the UK. He is the co-author with Jan Przewoznik of How to Think in Chess (Milford: Russell Enterprises 2001).

This site uses cookies to recognize users and allow us to analyse site usage. By continuing to browse the site with cookies enabled in your browser, you consent to the use of cookies in accordance with our privacy policy. X