Perspectives: Neuroscience. Science.285:673-4 (1999)


Is Art Lawful?

Christopher W. Tyler


It must be an eventual hope of brain science to address the complexities of the aesthetic experience. Perception may seem to some to be a phenomenological experience inaccessible to scientific rigor, but the efforts of generations of perceptual psychologists have shown that many aspects of perception are governed by a body of lawful relationships no less tractable than those of quantum physics, for example. The extension of such relationships to the subtleties of aesthetics is another kettle of slippery fish taken up by Ramachandran and Hirstein (1999) in a thought-provoking article in the Journal of Consciousness Studies. They bring a deeper viewpoint to bear on this challenge, combining principles not only from perceptual psychology, but also from evolutionary biology, neurological deficits and functional brain anatomy to provide a new overview of the nature of art. Their treatment is novel not so much in the idea of finding perceptual rules in art, but in addressing the essence of art, the evolutionary meaning of beauty, by means of such principles.

Fig. 1 a. ‘Madonna of the Meadows’ by Raphael (1505) is a composition of classic balance and composure within a naturalistic style. b. This 11th century Indian depiction of the Goddess Parvathi exemplifies the principle of caricature in exaggeration of the gender-specific features of the female form.


Taking their inspiration from the sinuous poses of Indian sculpture (see Fig. 1b), Ramachandran and Hirstein propose that one route to the perceptual essence is the principle of caricature. By exaggerating the characteristics of the subject, the artist draws attention to its essential features, in an analogy to the "peak shift " or enhancement of trigger features in animal behavior that has evolved to maximize the behavioral response to critical stimuli (Hansen 1959). These authors go beyond the conventional analysis of caricature in cartoons alone to argue that a similar process of perceptual caricature is one of the key principles of art as a whole. By enhancing the essential in the chosen subject matter, the artist can create an image (graphic or sculptural) that resonates in our memory and stands as a cultural symbol (see also Martindale, 1990).

The trick is to define which features are essential and which are incidental. Another way to say this is to ask which direction in N-dimensional feature space should the features be moved in order to make the desired enhancement? The usual assumption for facial caricatures is to assume that the feature shift for the caricature of a particular face should be to enhance its difference from the form of the average face (Brennan, 1985). Generalizing to the core artistic subject of the human figure, Ramachandran and Hirstein propose that the relevant anchor point should be the form of the average body. The artistic representation of female form, therefore, should be enhanced in a way to exaggerate the features that distinguish the two sexes, accounting for the enhancement of sexual expression in Indian statuary. For each kind of enhancement, this explanation requires that one identify both the form of the anchor point and the appropriate direction away from it, which may not be so easy for landscapes or scenes of particular human activities. Indeed, can one talk of a definitive essence at all when art may be viewed not as an object but as a developing relationship between the artist and the art work, and subsequently with the viewer?

An ultimate problem with this analysis, as the authors make clear, is that art is a vastly diverse enterprise that may not be so readily amenable to a simple treatment. Given a principle like the peak shift away from average, one is naturally anxious to try it out on a variety of classic examples. For example, enhancement of sexual features does not seem to have been a hallmark of Renaissance art in the West. That had to wait for the Baroque and Rococo eras, when the riot of depictions of Gods and Goddesses conform well to the predicted peak shift away from each other, with bold muscular Gods bending lithe delicate Goddesses to their will. It seems that Renaissance art is rather characterized a deeper principle whereby the artist tries to capture the average, or typical, image of the subject matter itself rather than shifting away from it, as in Raphael’s ‘Madonna of the Meadows’ (Fig 1a). This is perhaps exemplified by the earlier history of art, from the Greek sculptures of the era of the Venus di Milo or the discobulos, whose appeal seems to be the elegant modesty of their sexual features, to the Renaissance ideal of the Mona Lisa, which is almost androgynous. Capturing the typical pose of the subject matter in this way is such a basic principle of art that Ramachandran and Hirstein seem to have missed it, but no comprehensive analysis of the issue can afford to exclude it. Much Renaissance art consists of depicting the same pose repeatedly until a canonical form is achieved, as in the 'madonna and child' or 'Christ on the cross'.

Following the shift principle, these authors outline seven other interesting principles, drawn from Gestalt psychology and neurological sources that they believe constitute the deep structure of the art experience. The case for each principle is generally made with ingenuity and flair. In brief, these are the isolation of key aspects of the composition, grouping of related features, contrasting of segregated features, perceptual problem solving to extract the relevant information, a preference for generic views, the use of visual metaphors and, finally, symmetry. These two neurologists reveal much more of the brain processes that are likely involved in implementing these principles than would most art critics. But it is unlikely that many art critics would regard these principles as exhausting the repertoire of art principles (even accepting the caveat that there is much that is individual in art and not amenable to a principled analysis). Nothing is said about the widely recognized principle of balance in composition, of the power of the center emphasized by Rudolph Arnheim, or of the dynamic interplay of visual forces emphasized by Wassily Kandinsky, for example.

Of the eight principles delineated, the one that may be of most interest to artists is that of perceptual problem solving. Curiously, the section addressing this principle seems to have been lost in the main text, but it is clearly identified in the summary. Had they included the section, perhaps it would have included an extensive analysis of how the principle accounts for many of the diverse manifestations of 20th century art. One of the characteristics of 20th century art is that it included (among other things) a succession of styles of abstraction: fauvism, cubism, dadaism, futurism, abstract impressionism, op art, minimalism, and so on. Even in earlier representational trends, extremes of representational style such as impressionism, pointillism, expressionism, photo-realism and so on are themselves forms of abstraction. The point is that any form of abstraction requires the viewer to make a perceptual effort to extract the theme of the painting (compared with the essentially effortless perception of representational art). This effort itself forms an essential component of the artistic experience; by slowing down the perceptual processes of decoding the art work the viewer becomes aware of their evolution and interplay over time, and then experiences a sense of achievement when the full composition falls into place (or of continued mystery if it does not). Thus, the problem solving principle, evoked to account for the appeal of hiding the female form under diaphanous garb, could account for much of the development of 20th century abstract movements (which are hard to shoe-horn into the other seven principles!).

Ramachandran and Hirstein state that their goal is to foster debate on the principles underlying art. This they will undoubtedly do (as exemplified by the Commentaries in the same journal), but it is clear that the debate will extend far beyond the eight principles that they have enunciated.



Brennan S. E. (1985) Caricature generator: The dynamic exaggeration of faces by computer. Leonardo 18, 170-178.

Hansen H. M. (1959) Effects of discrimination training on stimulus generalization. Journal of Experimental Psychology 58, 321-334.

Martindale C. (1990) The Clockwork Muse: The Predictability of Artistic Change. Basic Books: New York.