THE BLOG

# A Mathematical Formula for Beauty?

01/24/2013 07:05 pm ET | Updated Mar 26, 2013

Almost everyone would agree that the image of the radio galaxy Hercules A (Figure 1), taken by the Karl G. Jansky Very Large Array radio telescope and the Hubble Space Telescope, is beautiful. Similarly, few would object to the statement that Vermeer's Girl With a Pearl Earring (Figure 2) is a breathtakingly beautiful masterpiece. But what is it that induces these responses in people? Is there a mathematical formula that can evaluate, even approximately, the beauty of images, objects, literary works or pieces of music? Believe it or not, but in a book published in 1933, the famous mathematician George David Birkhoff (1884-1944) attempted to do precisely that -- to develop a mathematical theory of aesthetic value. Birkhoff realized that there are many components to the concept of "beauty," and he attempted to discuss what he regarded as that intuitive feeling which is "clearly separable from sensuous, emotional, moral or intellectual feeling." To that goal, he divided the aesthetic experience into three phases: (1) the effort of attention needed for perception; (2) the recognition of certain order; and (3) the appreciation of value -- the reward for the mental effort. Birkhoff then took a stab at quantifying these three stages. He argued that the initial effort is proportional to the complexity of the object being observed, which he denoted by C. He called the order exhibited by the object O, and the assigned mental value, or aesthetic measure, M. He then suggested that within each class of aesthetic objects, such as flowers, vases or pieces of music, one can actually define an order O (which may depend e.g., on symmetry) and a complexity C. Birkhoff's formula for the feeling of aesthetic value was then simply: M = O/C. In other words, for a given level of complexity, the more order the object possesses, the higher its aesthetic measure. Alternatively, for a specified amount of order, objects are considered more aesthetic if they are less complex.