Friday, February 04, 2005

The Cognitive Science of Art: Beauty and the Brain

If you've read my two previous posts on Ramachandran's principles of art (here and here), you've probably got a good idea of what Ramachandran's concept of beauty is. While his 10 principles of art are concerned more with the production of art and the visual principles that apply to the viewing of art, and he therefore does not discuss beauty at length, there does seem to be a fairly clear conception of beauty contained within his explanation of the principles and why they work in art. Specifically, beauty is seen as the product of perceiving rewarding features in a stimulus (be it art or something else). Rewarding, in this case, means that the visual system (or perhaps some higher-order conceptual system), when it processes certain types of features or combinations of features, sends a signal to the limbic system, which then releases rewarding chemicals (e.g., dopamine). Ramachandran concentrates on largely innate properties of the visual system, because he is concerned with universals in art, but it could also be the case that the perception of beauty can be caused by features in art that are rewarding for experiential reasons. In fact, a straightforward implication of Ramachandran's theory is that works of art that produce activate positive associations (e.g., an approach motivation) are seen as beautiful, while those that activate negative associations (e.g., avoidance motivations) are not. Since approach and avoidance motivations are largely the products of associations derived through experience, then some of the subjective aspects of beauty in art might be explained this way. This explanation would fit nicely with theories of evaluation in non-aesthetic domains, in which all objects have valences for us.

Still, there seems to be something missing from Ramachandran's concept of beauty (if you can call it that). Certainly we can have aesthetic experiences when encountering objects that are not "art," such as a sunset or a member of the opposite (or same) sex, there does seem to be something that it unique about the aesthetic experience, or beauty in general, that distinguishes it from other rewarding experiences. All manner of objects and thoughts can be rewarding, in the neurological sense, without being experienced as aesthetically pleasing or beautiful, and outside of art, many of the phenomena Ramachandran describes in his ten principles fail to produce aesthetic pleasure. The problem, then, which Ramachandran does not solve, is to explain the qualitative aspects of the experience of beauty or aesthetic pleasure which differ from the qualitative aspects of encounters with other types of rewarding stimuli. Jennifer McMahon, in a discussion1 of the problem of beauty (the problem of reconciling the pre-reflective and reflective, or conceptual aspects of beauty), argues that an explanation of beauty must do two things:
  1. Account for the experience of beauty in such a way that both its subjectivity (I know something is beautiful by how it makes me feel rather than by first identifying the presence of necessary or sufficient conditions of beauty in the object) and objectivity (a judgment of beauty is grounded in objective properties in the object) can be understood as complementary rather than contradictory. In other words we need to provide a rational basis for beauty which does not translate into principles (logically necessary or sufficient conditions for beauty).
  2. Provide grounds for differentiating between the pleasures of the agreeably sensuous,
    the good and the beautiful.
In her view, Ramachandran's principles provide a good start for an explanation of (1), but as I've just argued, fail to address (2). She writes:
Rather than making a contribution to our understanding of beauty, the perceptual principles discussed by Ramachandran and Hirstein could be drawn upon to explain what kind of perceptual principles are exploited through certain art styles. In particular, their principles could be drawn upon to explain and identify the kind of relationship between an artistic representation and the world out there.
Interestingly, McMahon's own explanation of (2) is actually reflected in one of Ramachandran's principles, "perceptual problem solving." McMahon argues that the experience of beauty arises out of the processing of "within-object relations," which, as McMahon puts it, would "cause us to experience perception (consciously the object) as a solution to the problem of constructing a cohesive form, which itself is pleasurable." I have to admit that I'm not sure how McMahon's own conception, which she believes accounts for (2) in a way that Ramachandran's principles do not, is different from Ramachandran's principle of "perceptual problem solving." I'm also not sure how it a.) explains the experience of beauty, or aesthetic pleasure, as separate from other forms of pleasure, and b.) is supposed accounts for the beauty all art. Some objects, artistic or otherwise, that we experience as beautiful may in fact be quite simple (e.g., vivid colors, in a sunset or a work of modern art), and it's not clear where the "problem solving" takes place in these cases. I suspect that while currently deficient, Ramachandran's principles, because they provide for many different types of visually pleasurable experiences, are more likely to point in the direction of an explanation of the particular qualitative character of aesthetic experiences than McMahon's overly simplistic single principle is.

Regardless of whether you agree that Ramachandran's or McMahon's principles might someday provide a theory of the qualitative differences between aesthetic and other pleasurable experiences based on neuroscience, it's safe to say that we don't currently have one. In fact, despite being published almost six years ago, Ramachandran and Hirstein's original "principles of art" paper has yet to inspire much research testing their principles or predictions. In fact, neuroaesthetics in general is still largely a speculative research program. We know very little about the effects of art on the brain, and we know even less about the neuroscience of experience of aesthetic beauty.

There is, however, one study which sheds some light on the localization of "beauty" in the brain, and its contrast with the experience of "ugly." Hideaki Kawabata and Semir Zeki2 recently published an imaging study in which they presented participants with several paintings and asked them to rate them as "beautiful," "neutral," or "ugly," and then had them view the paintings again while being scanned with a functional MRI machine. They found that all three types of paintings produced activation in areas of the visual cortex specific to the types of features in the paintings. This finding is consistent with Ramachandran's principles, which require that visual art activate certain specialized areas in the visual cortex to produce an aesthetic experience. Kawabata and Zeki also found that all three types of art produced activation in the orbitofrontal cortex, which is an important part of the brains emotion system, as well as the reward system. In this area, the amount of activation was dependent on how beautiful the painting was perceived to be, with the most beautiful paintings producing the most activation, and the ugliest producing the least.

A third general area of activation, the motor cortex, would likely not have been predicted by Ramachandran. As in the case of the orbitofrontal cortex, all three types of paintings produced activation in the motor cortex, but in this case, the "ugly" stimuli produced the most activation. Increased activation in the motor cortex in response to ugly paintings is consistent with findings in other areas in which participants observed aversive stimuli, such as the perception of faces rated as ugly, and the "violation of social norms," but the reason for its activation in these contexts and that of ugly art is unknown. Kawabata and Zeki speculate that this may be part of a "preparation for action" in response to aversive stimuli, though there it's difficult to know what sort of tests of this hypothesis we could produce. The activation and their explanation are also consistent with my speculation that approach and avoidance motivations may be involved in the perception of "beautiful" and "ugly" stimuli.

Finally, Kawabata and Zeki found that, when contrasted with neutral stimuli, beautiful stimuli produced a significant amount of activation in the anterior cingulate and parietal cortex. As with the orbitofrontal cortex, these areas are associated with reward and emotion, and the parietal cortex is also associated with spatial attention. It's not clear what the role of activation in these areas is, and since it only appeared in the contrast of beautiful and neutral stimuli, Kawabata and Zeki don't really offer any speculations.

Figure 3 from Kawabata & Zeki (2004), p 1701.


Kawabata and Zeki argue that their results may show how the brain makes judgments of beauty, perhaps solving the problems McMahon raised for Ramachandran's principles. Beautiful objects produce greater activation in emotional centers of the brain, while ugly objects produce greater activation in the motor cortex. I'm inclined to think that, while this is an improvement over Ramachandran's concept of beauty, in that we can see quantitative differences in different types of judgments (beautiful and ugly), their findings still don't provide us with a suggestion for how aesthetic experiences come to be different from other experiences of pleasure. The orbitofrontal lobe is also active when we perceive types of pleasurable objects that aren't usually associated with aesthetic pleasure (e.g., sexual objects), and the motor cortex is active when we perceive a wide range of aversive stimuli (from the visual to the social, or conceptual). Perhaps better imaging data might demonstrate more subtle differences in the localization of beautiful vs. other types of pleasurable experiences and ugly vs. other types of unpleasant experiences, but for now, we'll still have to consider the specific character of aesthetic experience to be a mystery.

1 McMahon, J.A. (2000). Perceptual principles as the basis for genuine judgments of beauty. Journal of Consciousness Studies, 7(8-9), 29-35.
2 Kawabata, H., & Zeki, S. (2004). Neural correlates of beauty. Neurophysiology, 91, 1699-1705.

3 comments:

Bhisma Chakrabarti said...

very nice critique. i feel that the chance that there would be a separate functionally specialized module for 'aesthetic pleasure' is remote : the very search for one is a somewhat scary reminder of the bad old days of phrenology, where people went upto the extent of localising almost everything, including e.g.'paternal instinct'.

Chris said...

Hah! there's actually a very good book on cognitive neuroscience titled The New Phrenology.

Anonymous said...

Twice in my life I have experienced beauty in that something has come into my view that could be felt (rather than seen) as powerful emotion, beginning as sort of shock in the heart and then spreading through my whole being as a sense of great awe and wonderment. I think true beauty can only be felt not thought.