Interview by Richard Marshall.

'A child babbling for no other purpose than the sheer joy of it—that’s the prototype of art production. Aesthetic enjoyment isn’t tied to the value of particular things, as standard evolutionary accounts maintain. It is rooted, rather, in the joy that infants get from activities that have no immediate purpose, but which have a long-term benefit in terms of learning.'

'Here’s the take away: perceptual images are not things we are aware of; we are aware of objects in the external world. A perceptual image is a brain state that gives us awareness of objects and their qualities arrayed in three-dimensional space; it is a structure carries the spatial content of perceptual awareness. This content is produced by the brain, which works on material provided by sensory receptors such as those in the eyes, ears, and skin.'

Mohan Matthenwas born in Bangalore India and studied Physics and later Philosophy at Mayo College, Ajmer and St. Stephen’s College, Delhi. Then, he went to Stanford University in California, where he got a PhD. He has worked in Canada since 1976: at Calgary, McGill, Alberta, and UBC. Since 2006, he has been Professor of Philosophy and senior Canada Research Chair at the University of Toronto. Here he discusses his work on perception and aesthetics. He talks about subjectivism and its history, his alternative theory of perception, his theory of image content, why he disagrees with Fodor, Marr and Gregory even though they take issue with subjectivism, the multi-modality of active perception, whether we need different theories of perception depending on which sense we're talking about and how tricky it is to count how many senses we actually have. Then he talks about art, what aesthetic pleasure is, why humans do art, and why it makes evolutionary sense. This is one hell of a clear view...

3:AM:What made you become a philosopher?

Mohan Matthen:I backed into it, really. I studied Physics as an undergraduate. I loved the mathematics but hated the lab work. Unfortunately, my talents reversed that preference: I wasn’t all that good at the math, though I wasn’t bad in the lab. (I got a First in Lab, but barely passed mathematical physics.) I had dabbled in Philosophy, and thought it might be a good way to combine my love of abstract structure with my dislike of using my hands. So, much to the consternation of my father, who warned repeatedly that he didn’t have much money to leave me, I took philosophy. It turned out that I loved it.

Strangely, I didn’t want to do philosophy of physics. I think that I was still too close to a physicist’s territorial view that it was inappropriate for a philosopher to meddle. Philosophical speculation about the nature of physical reality seemed just too madly metaphysical. Certain other paths were closed off too. I didn’t feel that the “big” questions were for me. I wasn’t much moved by scepticism, for example, in the way that Moore and Wittgenstein were, as something that I passionately wanted to escape. And, true to my background in science, I was reluctant to engage in a topic where progress is almost impossible.

My real interest has always been in tinkering around with the philosophical problems that scientists (other than physicists!) generate. I started with Aristotle, but grew frustrated by constantly having to explain brilliant falsehoods. I gravitated to evolutionary biology as a replacement for Aristotle—Does it solve the problem of biological function? What is the evolutionary take on biological classification? Why is natural selection probabilistic? Questions like these.

My main and most abiding interest has been the nature of perception. There are lots of really neat questions here. Just as examples: Why do we perceive colour? How does colour vision help us when so many other animals manage quite well without it? How do we perceive space? Is this just a matter of perceiving spatial relations among other perceptibles? Questions like these demand quite a bit of empirical knowledge, which has to be massaged and reconceptualised in order for it to yield answers to philosophical questions. These questions have kept me busy and happy for decades.

3:AM:Well let’s start with your ideas about perceptionthen. You argued against the ‘subjectivist’ approach to perception back in 2005. This was the dominant approach, and had big names of the past setting it up didn’t it – Descartes, Hume, Berkeley, Ayer, Quine all had a part to play in it. So could you sketch it for us so we can see what you were up against?

MM:Subjectivism (in the relevant sense) claims that all we know in perception is states of our own subjective consciousness, or the sense-data that arise from it. To understand this idea, think of what Descartessaid in his Second Meditation: “If I look out of the window and see men crossing the square, as I have just done, I say that I see the men themselves . . . Yet do I see any more than hats and coats that could conceal robots? I judge that they are men. Something that I thought I saw with my eyes, therefore, was really grasped solely by my mind’s faculty of judgment” (Jonathan Bennett’s translation). It’s a cold day; the men are all bundled up and you can’t see a square inch of skin when you look down at the street from your window. Descartes’ idea is that it is only the hats and coats that make an impression on the eyes; they obscure the men themselves, much as the wrapping on a birthday present prevents one from seeing the gift itself. (Descartes is centred on vision here, and I’ll stick with him for the sake of the example, though there is much to be said about how the other senses operate in similar contexts of occlusion and partial concealment.)

Descartesfamously extended this reasoning to the wax of a candle; we say we see the wax, he argues, but what we really see is “its colour, its figure, its size.” A thought experiment proves this. When we melt the wax, we destroy the colour, figure, and size of the wax. We say we see the same wax, but nothing we see is the same. This shows that contrary to what we say, it isn’t a matter of vision. When we identify the wax as the same, we are making a “judgment.”

Berkeley and his successorstook this idea to an extreme. Are we really aware even of the colour of the wax? No, all that we know is the colour it appears, which is an idea in subjective consciousness. This kind of subjectivism gradually became the foundation of a certain kind of empiricism. Ultimately, it lies behind the idea that Quine describes this way: “Physical things generally, however remote, become known to us only through the effects they help induce at our sensory surfaces.” These superficial sensory effects do not mandate the conclusion that we see men; it does not mandate even that we see external objects.

You could say that this kind of attitude focuses on sensory information. When you look at the men from your window, perception reveals mounds of cloth moving up and down the street. Those mounds of cloth could, for all that you actually see, be “robots;” they could even be self-moving. You don’t see anything that decides among these possibilities. So why do you think they are men? Your belief goes beyond the information contained in what you see. The information available to you is just a certain state of consciousness, or looking at the matter slightly differently (as Quine does) the stimulation of your retina.

Let’s take this argument at face value. Superficial sensory effects don’t account for the knowledge we take from perception. My response is that to account for perceptual knowledge, we need to change the focus. Here is another way of accounting for perceptual knowledge—you judge that you see men below you because it looks as if there are men walking down the street. There is a characteristic way men move; this distinguishes them from self-animated mounds of cloth, from automatic machines, and even from women. The moving objects below you move like that. You believe, or judge, them to be men, because they appear to be men, not because you possess sensory information that they are men.

In this way of thinking, perceptual knowledge is based on appearance, not just sensory information. It is not forced by appearance, because you may come to realize in a particular case that appearance is misleading. When you reason from appearances, you have to accept the possibility of error. To understand perceptual belief, we have to shift our gaze from the certainty of sensory information to the sometimes misleading world of perceptual appearance. This shift is what philosophers as different as J. L. Austin and John Pollock recommend. The question that I wanted to address in my book was this: Where does perceptual appearance come from? Turning Quine’s reasoning on its head, one can say, “OK, it is not just a matter of sensory effects. What is it then?”

One kind of answer is the one that Humeoffers. Appearance is an “association of ideas.” You have seen men move that way; when a mound of cloth moves that way you “infer” by association that it covers a man. There is a pile of scientific evidence that Hume’s theory is incomplete. Undoubtedly, some perceptual appearance results from learned association. But consider the richly structured scenes our perceptual systems provide us from infancy on—three-dimensional objects, their features and locations, their persistence through time, and various causal relations among them. It’s pretty clear that association can’t explain all of this. Other processes have a hand in it.

My aim is a more complete understanding of the process by which perception does all of this. I have been concerned primarily with three large themes. The first is what I called the “classificatory” activity of perception, whereby it attributes similarity-ordered qualities to objects. We see various things as matching in colour; we also see them as more or less similar in colour. What’s the topography of such groupings? The second is “pluralism.” Different kinds of animals construct different quality orderings. The colours birds see are different from those we see. Why and how? The third theme is particularity: how does perception engage with particular objects in the environment? How is it different from memory and imagination in this respect?

Sorry to be so telegraphic about these issues, but I just want to give you an overview of the topics that I have engaged with in order to respond constructively to subjectivism.

3:AM:No need to apologise - trust me, I need all the help I can get! For you sensation isn’t anything like an image is it? To think it is is to make basic mistakes about the visual processes in the brain and sense-organs. The Gestalt phenomenon of occlusion helps show this doesn’t it?

MM:I now use the term ‘image’ slightly differently, as I’ll explain in a moment.

Holding this in abeyance, I certainly think that what we get from our perceptual systems is nothing like a picture or a two-dimensional image. In particular, perceptual awareness is not a simple transfer from the retinal image. As you say, our awareness of occlusion helps show why. In the scene below, there appears to be a red lozenge in front, casting a shadow on a partially occluded shape behind. That’s not what you find on the retina—there’s no occlusion in a two-dimensional display—but it’s how it appears.

There are lots of reasons why one should reject a correspondence between perceptual awareness and “the” retinal image. First, there are two retinas; we are conscious of a three dimensional scene that the brain constructs in part by extracting and combining the information available in each. Second, auditory awareness is of a spatial layout of sounds; yet there is no spatially organized display like a retina in the ear. (The cochleas are spectral frequency analysers; they are not spatially organized.) Finally, even the visual qualities we see are not, in any simple way, transferred from the retina. Colour is an instructive example. Traditional philosophers thought of the retinal image as a mosaic of colour pixels. This is completely wrong-headed. The brain gets its information about colour by combining the outputs of four different kinds of cells. So even at this supposedly basic level, a lot of brain processing is needed—retinal activity and colour awareness are not isomorphic. This all illustrates the magnitude of work the brain has to do in order to get from sensory receptors to perception.

This brings me back to the term ‘image.’ We are perceptually aware of objects and their qualities located in the space around us. Cognitive scientists like Roger Shepard and Steven Kosslyn try to capture this by saying that we have perceptual (and other mental) images. Their point in using this term is to emphasize the spatial form of perceptual awareness. Some are tempted to reify perceptual images; Shepard and Kosslyn do not do this. Images are not, for them, spatially organized structures (or pictures) in the brain; they are, rather, informational structures that encode things and their features as spatial arrays.

Here’s the take away: perceptual images are not things we are aware of; we are aware of objects in the external world. A perceptual image is a brain state that gives us awareness of objects and their qualities arrayed in three-dimensional space; it is a structure carries the spatial content of perceptual awareness. This content is produced by the brain, which works on material provided by sensory receptors such as those in the eyes, ears, and skin.

3:AM:How should we understand image content?

MM:Image content is how an image represents whatever it represents. Consider a picture like this one (by Constable, courtesy of Sailko and Wikimedia):

The picture represents Salisbury Cathedral and you can ask what spatial information it conveys. For example, you can ask: Where are the cows relative to the Cathedral? The answer is that they are closer to the viewpoint of the picture than the Cathedral. This is part of the content of the image; this is part of how it represents the cows and the Cathedral.
Now think of mental and perceptual images. You are aware of various things in your vicinity; you are aware of them laid out in space. For example, imagine that you are at a dining table, and that you are aware of a dinner plate closer to you than your water glass, a fork to your left and a knife to your right, and so on. The mind/brain has provided you with this representation of the scene in front of you. The person sitting across from you is aware of the same objects in the same spatial layout, but there are differences in the way they are represented to her. They are all reversed in her perceptual awareness; your dinner plate is further away from her than your glass, your fork is to her right, and so on. What I want to say here is that she has a different image of the same things in the same spatial relations.

Notice how this externalizes the space we perceive. One of my current preoccupations—I can’t go into detail here—is how best to capture the objectivity of spatial perception. Too many philosophers are distracted by the apparent subjectivity of our points of view on external scenes—how things appear different from where each of us stands—but I want to argue that when one moves and looks at things from different perspectives, one’s representation of spatial relations stays the same in important ways.
It is important here not to confuse the properties of mental images with those of pictures. Most importantly, there is this difference. There is no actual Cathedral you see when you look at the picture. What you see is an expanse of paint that visually represents the Cathedral. In the case of mental images, there is no such intermediary, no equivalent of the expanse of paint on canvas. If you were standing where Constable was positioned, you would see the Cathedral; you would not see a mental image of the Cathedral. Your brain/mind would make you spatially aware of the Cathedral in a certain way. The perceptual image is the brain structure in virtue of which you would be aware of it in this way. You do not see this brain structure.

3:AM:Why did you take issue with theories that rejected the subjectivist view – a range of theories put forward by David Marr and Jerry Fodor and Richard Gregory?

MM:Fodor thinks we are perceptually aware of objects and their features, not of internal mental entities such as sense-data and the like. This is where the Representational Theory of Mind takes him. This makes him an opponent of subjectivism, as you say, and I follow him in this. However, he thinks that when people talk about perceptual images, they are committed to pictures in the brain. And he thinks that pictures don’t (without a viewer’s interpretation) tell us about objects. So he rejects the view that perception is imagistic. You can see how my view of mental images keeps me out of Fodor’s quandary, if I may call it that.

As for Marr and Gregory, their question is how the brain extracts content from sensory information. For example: How does the brain work out, from the data available to it, the particular arrangement of objects in three-dimensional space? This is a dauntingly difficult question, and these psychologists (following in the footsteps of a great mid-century generation: Barlow, Lettvin, Hubel, Wiesel) made giant strides toward solving them. My approach is a little different not because I disagree with them, but because I have increasingly come to think that we perceive, in part, by deliberately exploring the world. Perception does not begin with the state of the sensory receptors; rather it begins with what we do to create and change the state of those receptors.

In the simple occlusion diagram I provided earlier, we can theorize about the hints or “cues” that enable the perceptual brain to make it look like a lozenge in front of partially occluded rectangles. But in the real world, we also move our heads and our feet so as to get a view of what lies behind what from other angles. This is what I call sensory exploration or active perception. We act on the world and see how it responds. For example, I might run my thumb across (not along!) a kitchen knife to test whether it is sharp. Or I might look at somebody’s face to improve my comprehension of what they are saying. J. J. and Eleanor Gibson talked a lot about such methods of exploration. Our instinctive ways of perceptually investigating the world are as much a part of how animals gather knowledge through perception as the brain’s data processing pathways. Marr and Gregory have to be supplemented by the Gibsons to get a fuller philosophical appreciation of how perception enables us to gain knowledge and to get on in the world.

3:AM:How does your idea of perceptionas doing solve the problem with these approaches? Is this theory of active perception Kantian in that it takes space and time to be premodal representations?

MM:Active perception is multimodal. When I test the sharpness of a knife by running my thumb across the blade, I integrate awareness of my own movement with the tactile sensations that arise from it. When I look at somebody’s facial gestures in order to better comprehend their speech, I have to put visual and auditory perceptions together. In a 2014 paper on the representation of space (which appeared in a volume edited by Dustin Stokes, myself, and Stephen Biggs), I argued that cross-modal coordination requires a common spatial framework that serves all the perceptual modalities. Such a framework would be “premodal” in the sense that it exists prior to information received through the senses. It would not be shaped by modal experience, by definition, and so it would circumvent Molyneux’s problem (which has to do with the alleged incommensurability of different modal representations of space). And it is a priori, since it is not shaped by experience, but is rather a pre-condition for sensory experience.

In all of these respects, my account of the perceptual representation of space follows Kant(though I dissent from his idealism in these matters). But the argument is different. Kantthought his conception followed from the “necessity” of Euclidean space. I don’t believe that Euclidean space is necessary in this way. I derive the premodality of space from the multimodality of active perception.

All of the above holds for time, though it is less controversial to argue that time is not experienced through the senses. (Nobody has ever propounded a Molyneux problem for time, have they? That is, nobody thinks that our visual awareness of time is somehow incommensurable with our tactile or auditory awareness of it.)

I’ll mention two more significant things about active perception. The first is that (as I argued in another 2014 paper, “How to be Sure”) sensory exploration is a crucial part of how we externalize things. One way we discover that something is an objectively existing external object is by moving around and changing our perspective on it. If it thereby appears different in perspectivally dependent ways, it appears to be objective. (I owe a lot to Susanna Siegel’sinsights into perspectival variation.) Since sensory exploration is bound up with how we represent space, this links up with the Kantianidea that things appear to be external and objective by appearing to be located in space.

Another central point that I made in “How to be Sure” was that sensory exploration is how we surmount uncertainty. I can’t get into the detail here, but the idea is that when we doubt perceptual appearance, we try to test out alternative possibilities by sensory exploration. For example, if I am in doubt about the colour of my socks, I take them to the window to expose them to good sunlight, and I turn them over in my hands to make sure the light falls on them from different angles. When I do this, there is little room left to doubt what colour my socks are. My suggestion was that in this kind of way, it is possible to approach “empirical certainty” about the perceptible world.

3:AM: Is your approach an alternative to Andy Clark’s ‘Predictive coding’ and Nicholas Shea’s emphasis on past experience or can these approaches be fused?

MM:As far as I can see, my theory is about what perception is—I say that it’s an interactive exploratory process. Andy Clark’s “predictive coding” is about how it works. His claim is that “higher” level perceptual processing generates predictions about what is to come based on an estimate of the nature of the thing being perceived. These predictions are then checked against lower level inputs. When discrepancies are encountered the higher-level processes are corrected. As I see it, this theory can tell us something about how active perception is processed. As I run my thumb across my knife, for example, my perceptual system forms an estimate of how sharp the knife is, and based on this estimate, it generates predictions about future inputs to my tactile receptors as I continue to test the knife. These predictions are checked against the inputs when they arrive, and the estimate of the knife’s sharpness is either reinforced or revised. There are connections here with the notion I talked about a minute ago—that sensory exploration yields empirical certainty. All in all, I think Andy’s ideas work very well with mine.
Nick Shea is interested in identifying top-down effects on perception. This too is an idea that resonates with my general approach, though I won’t go into the detail here. I do want to mention, though, that my former students, Kevin Connolly and Dustin Stokes, are doing some amazing work on a particular kind of top-down effect, namely perceptual learning and perceptual expertise. We are all trying to reorient our notion of perception into this framework of active interactions with the world.

3:AM:Do we need to have different approaches to perception depending on which sense we’re looking at – so listening isn’t the same as hearing for example? Or should we approach the senses as a team event with all the senses working together?

MM:There are two important ideas in play in your question. The first is a distinction between listening, which is active and something we do, and hearing, which is passive and something that happens to us. We listen to things; we listen for them. When I am in conversation, I listen to my companion. Sometimes when I drive a car, I listen for signs that something is wrong with the engine. In both cases, I am doing something that’s outwardly directed. Hearing is a part of this, but it’s a part that I don’t entirely control.

As a kind of active perception, listening is normally multimodal. When I listen for funny sounds in my car, I direct my attention here and there, and this is usually accompanied by looking at various things, testing the accelerator, feeling the vibrations, and so on. How I perceive the situation depends on what I hear, but also see and feel, when I test the car in these ways. This kind of activity involves the senses working as a team as you put it.
The second idea is that, all of the above notwithstanding, we think of listening as an auditory activity. This is because listening focuses on what one can hear. What I do in order to listen is aimed at getting a better auditory fix on the engine. And this is where the peculiarities of listening come into view. Think of what you do when you want to look at something closely. It’s different from what you do to listen. This illustrates the fact that though it’s a team event involving the same players in both cases, the focus of the event is different and the players occupy different roles. This is what makes us think that listening is an auditory event and watching a visual event, even though in reality things are more complicated.

To sum things up, visual perception isn’t different from auditory perception because the first involves just the eyes and the second just the ears. They both involve multiple sensory organs and multiple perceptual activities. They are different because in visual perception, the ultimate focus is to produce varying inputs to the eyes, whereas in auditory perception, the focus is the ears.

3:AM:And just how many senses do we have? Are Aristotle, Grice and MacPherson right to think about whether there are more than the traditional five? How should we approach this consideration?

MM:This is a really tough question and really worth thinking about because of certain claims widely made in the scientific literature—namely that two of the traditional senses, touch and flavour perception, are multisensory.

Let me quickly say why the philosophers you mention don’t give us much help with these questions. It’s difficult to reconstruct Aristotle’sthinking in a way that is relevant to modern science, and so I won’t try, at least not here. Grice thought that there were a number of different ways of differentiating the senses; he tried to show that most of them end up either in confusion or exposed as not fundamental. Fiona Macpherson thinks that each of Grice’s ways marks a different sense of the term “sense,” and she tends to treat them on a par. I don’t think this is right because, as Grice showed, there is little value in some of the ways of distinguishing the senses that he enumerates. We shouldn’t be promiscuous in our taxonomic habits.

To be frank, I don’t think that Grice’s own favourite distinguishing mark, phenomenology, is of much use either. It’s all very well to say “It’s obvious that vision and audition are different because I can phenomenologically distinguish between sights and sounds.” But is it really obvious that we make this distinction by subjectively examining visual and auditory sensations? This seems to be what Grice thinks, and I have a hard time coming to grips with this seemingly introspectionist claim.

Moreover, as I said a moment ago, we want to be able to adjudicate whether it makes sense to say (for example) that flavour perception is multisensory. Phenomenology is of no use here at all. Scientists say that the tongue and the nose give us the flavour of a cherry, acting together. But flavours are phenomenological unities. There’s nothing about the phenomenology of tasting a cherry that helps us decide whether the nose is involved. There is nothing here that enables us to think about whether that flavour is unimodal or multimodal.

In my view, there are two ways to think about “the senses.” You can think of them as receptor classes. According to this point of view, the light sensitive cells of the eyes define vision; the cochleas define audition; the receptors in the nose and tongue define olfaction and taste. This is the way most scientists think about the senses. However, the classification they arrive at is quite at odds with common sense. For example, as I mentioned, they say that flavour perception is “multimodal” because it involves the smell receptors in the nose as well as the taste receptors in the tongue. Yet, flavour perception is what we ordinarily think of as “taste,” which is one of Aristotle’s traditional five. We’d like to know what lies behind this. Is the tradition simply a factual error? Louise Richardson argues that traditions of this kind have a special status in our thinking, and cannot be errors. I don’t agree, but it’s still interesting to figure out where they are coming from.

A second way of thinking about the senses is that each of them corresponds to a kind of perceptual activity. A little while ago, I said that listening was an activity focused on hearing though multiple senses participate in it, and that watching or looking is a different activity in that it focuses on seeing. My view is that this points to a way of grouping these kinds of activities so as to yield something very close to the traditional five senses. It’s a little bit technical, but my suggestion (in my 2015 Oxford Handbookentry on “Individuating the Senses”) was that a perceptual modality is a group of mutually certifying perceptual activities. I won’t go into the details, but it seems as if this yields something like the traditional list. So I think I can save something close to the tradition.

3:AM:If aesthetic pleasure is universal then does it need an evolutionary account to explain it? Doesn’t your approach mean that evolutionary accounts of aesthetic appreciation as a matter of appetites and ‘telic’ pleasure fail because they depend on the benefits that aesthetic pleasures bring? Where does that leave an evolutionary explanation?

MM:Thanks for asking me about aestheticpleasure, which is a new area of investigation for me.

Pleasure is a felt evaluation. It is a conscious state that is closely connected to reasoned evaluations. Interestingly, it is bi-directionally yoked to these evaluations in humans. What I mean is this. A cool drink of water gives you physical pleasure if you are thirsty. This pleasure gives you a reason to evaluate what you are doing positively. (Of course, this reason is not conclusive; your drink might be bad because the water is contaminated.) This is a pleasure-to-evaluation connection. But the connection goes the other way too. Suppose you calculate your finances, and find that you are doing better than you thought. This complicated piece of rational evaluation can make you literally giddy with pleasure.

Aestheticpleasure is bi-directional too; you can spontaneously enjoy a piece of music and thus have a reason for thinking it meritorious; you can also intellectually work out its brilliant use of counterpoint and enjoy it as a result.

Now, aestheticpleasure is universal, as you rightly say. It is not culturally specific to experience aesthetic pleasure. However, culture influences what things we find aesthetically pleasing; there are relatively few cross-cultural aesthetic universals aside from the fact that all humans are capable of aesthetic pleasure. This means that the universality of aesthetic enjoyment cannot be grounded in the value of the specific things we enjoy. This invalidates standard evolutionary accounts of aesthetic pleasure. These accounts appeal to the beneficial characteristics of certain specific objects. For example, they say that we enjoy certain landscapes because these landscapes are safe to live in. The problem is that first of all we don’t all enjoy the same landscapes. And secondly, even if we did, this tells us very little about the extensive range of objects each of us finds aesthetically pleasing.

Some say that aesthetic enjoyment is an evolutionary “by-product.” I don’t think this is tenable. On the face of it, aesthetic enjoyment detracts from evolutionary fitness because it takes resources away from more pressing needs such as foraging and vigilance and sleep. This means that even if aesthetic pleasure were an evolutionary by-product, it would be selected against, and thus unstable. We need to explain why it is preserved.

3:AM:Why do humans contemplate art? Is it to help us develop our learning skills?

MM:Something like that. We have certain skills that, though they are innate, depend on repeated practice in infancy. Language is an example. Little children learn to talk by first babbling and then chatting away incessantly. Why do they do this? Because it is fun for them. This fun has a long-term purpose; it enables the development of language skills. Evolution reinforces otherwise pointless word play because of this. The utility of babbling doesn’t arise from the specific syllables uttered.

I argue that a lot of activities are like this. Our motor skills are developed through physical play; our perceptual skills by playing at looking and listening “pointlessly”—the infant stares, and by doing so she learns how to look. (Notice how my emphasis on active perception plays into this notion of perception as a skill.) It is this fun, I claim, that is the basis of aesthetic pleasure. A child babbling for no other purpose than the sheer joy of it—that’s the prototype of art production. Aesthetic enjoyment isn’t tied to the value of particular things, as standard evolutionary accounts maintain. It is rooted, rather, in the joy that infants get from activities that have no immediate purpose, but which have a long-term benefit in terms of learning.

3:AM:How should we understand the pleasure component of art?

MM:I said just now that aesthetic pleasure is rooted in the fun we get from doing certain “pointless” things. But the pleasure of making and appreciating art involves a lot more than this. There are two reasons for thinking this. First, though it is rooted in a kind of play that we all indulge in, art involves a very high level of skill, in both the producer and the consumer. I may enjoy whistling a tune; that’s the kind of play that is at the bottom of musical appreciation. But my whistling is not yet art; art has to be self consciously “artful” (Alva Noëis interesting on this point), and it must also be skilful. Secondly, art is, as I said earlier, culture specific. You can go to Chinese opera and enjoy it, but if you are a beginner, you will not appreciate it as much as somebody who understands the musical conventions and the stories. In general, art does not reward a novice’s attention as much as it does an expert’s.

So, one aspect of appreciating art is a virtuous spiral of skill, in which producers make things with an increasing level of skill and consumers become increasingly skilled in discerning the skill it takes to make these things. As the art historian, Michael Baxandallwrote, “Much of what we call ‘taste’ lies in . . . the conformity between discriminations demanded by a painting and skills of discrimination possessed by the beholder.” In a way, this spiral is economic. Various artists produce items with varying levels of skill and imagination; consumers attach the greatest value to those that incorporate the greatest skill and imagination. The laggards among the producers fail to be consumed; the less skilled consumers exit the market because it holds no interest for them.

When a skilled consumer engages with a work of art, she finds much that holds her interest. Thus she finds it both rewarding and meaningful to continue to engage with the work of art. There is a self-reinforcing cycle here. You start to read a great novel. The more skilled and discerning you are as a consumer of this novel, the more you find to hold your interest, and the easier you find it to continue reading it. Culture plays an essential role in this, because the more attuned you are to the relevant culture, the more you will find to enjoy.

I have been thinking a lot about this reinforcing role of pleasure recently, and my paper about “the pleasure of art” will appear in the new journal, Australasian Philosophical Review, in 2017, along with commentary by a dozen or so philosophers. I really look forward to the discussion.

3:AM:Is art dependent on the object or the response and is the communal and the communicative always important? Can there be an art that no one ever responds to or is purely personal?

MM:As I have been saying, the more you find to engage with in a work of art, the more you are motivated to continue to engage with it. The artist uses culture to salt her work with attractors; the consumer needs to share the artist’s culture to be drawn by and engage these attractors. This much is obvious. But the implication is somewhat less obvious: art is not a matter of “beauty” as such, but rather of qualities that psychologically absorb an audience. Great works of art are often beautiful, but what makes them great is that they reward and compel continued and repeated attentive engagement. This could be idiosyncratic, I guess—I am not committed to objectivity with regard to aesthetic response. Maybe for purely personal reasons, I am mesmerized by “Good Night Moon,” and return to it again and again. But this would be unusual, to say the least.

3:AM:Your approach aims to illustrate the psychological foundations of aesthetic norms doesn’t it? But if these norms are not stable and change then surely there’s a problem with this approach?

MM:I don’t think my approach runs into this problem. I distinguish between the nature of aesthetic appreciation, which is psychologically similar wherever it occurs, and the nature of the things particular people appreciate, which is different from person to person, culture to culture, and time to time. I’ve tried to say something about aesthetic pleasure, which I think is at least roughly the same whether it occurs in New Guinea or Japan. I haven’t tried to come up with universal principles of art criticism, which are very different in different places.

3:AM:And finally for the readers here at 3:AM are there five books you can recommend to get us into your philosophical worlds?

Gareth Evans, Varieties of Reference.

Fred Dretske, Knowledge and the Flow of Information.

The above are indispensible classics. To get a good idea of what philosophy of perception is about nowadays, see:

Casey O’Callaghan, Sounds: A Philosophical Theory

Matthew Fulkerson, The First Sense: A Philosophical Study of Human Touch

And finally one of my favourite books on aesthetics:

Cynthia Freeland, Art Theory: A Very Short Introduction.

Richard Marshallis still biding his time.

Buy his book hereto keep him biding!