Science and the search of beauty (2): the halflings’ view

In my past entry I described the tribe of the ‘Halflings’ as those authors who try to find a middle road between the ‘Platonist’ that identify beauty as one essential goal of scientific research (even conflating it, in the end, with truth itself), and the ‘Sceptics’ that assume that aesthetic criteria are essentially subjective and hence can play absolutely no epistemic role in science (or, at most, a secondary role as an additional personal incentive to researchers’ effort). The philosopher James McAllister has been the most influential of the Halflings, thanks to his 1996 book Beauty and Revolution in Science¹, a work in its turn deeply influenced by the long lasting debate around Thomas Kuhn’s epochal The Structure of Scientific Revolutions and the conflict between post-positivist and historicist accounts of science development this work helped to raise.

According to McAllister, amongst the criteria for theory choice that the scientists of a particular historical period employ (criteria that, according to Kuhn, cannot be reduced to any kind of logical algorithm, but always contain in some essential way subjective, or at least not totally objective value judgments, and constitute a central part of what since the sixties is known as a Kuhnian ‘paradigm’), something that can be called the ‘aesthetic canon’ is a non-negligible part. This canon varies in history, as well as from discipline to discipline, and consists of those (in some loose sense) ‘aesthetic’ properties of those scientific theories, that, according to the evolution of a particular discipline or ‘paradigm’, have been associated with a high level of empirical success. That is, McAllister substitutes the Platonist idea of an eternal, immutable and objective ‘essence of beauty’, for a collection of aesthetic criteria that are subject to change, and that, instead of showing us the road to truth by some kind of a priori intuition, leads us to appropriate theories just thanks to the fact that those criteria have been empirically successful. McAllister’s argument (or, as he himself calls it, his ‘aesthetic induction’) turns in this sense the traditional Platonist argument upside down: it is not that scientists’ look first for an ‘eternal’ beauty and end finding out the truth, but they produce first empirically successful theories, and then they learn to appreciate some of the virtues of those successful theories as aesthetic ones. And, since the successfulness of those theories will probably reflect some important deep properties of the structure of the world, it usually happens that the pursuit of further theories within the same field taking into account those aesthetic values and judgments ends producing new empirical successes. For example, the discovery of the importance of the geometrical shape of chemical molecules in the second part of the nineteenth century favoured the application of an aesthetic appreciation of those geometrical relations as part of the ‘know-how’ a specialist in chemistry had to develop to become a successful scientist within his or her domain.

These criteria are historically determined, and hence they are subject to change, and usually will be substituted by others once the pursuit of good theories using them as ‘hints’ becomes impossible, i.e., when the ‘paradigm’ to which they belong stagnates and is superseded by some new paradigm associated to a more or less different system of criteria (aesthetic, methodological, or whatever). For example, determinism and mechanism were once important aesthetic criteria in the development of physics, during the eighteen and particularly the nineteenth centuries, but they had to be abandoned with the advent of quantum mechanics once physicists admitted that at a deeper level they didn’t capture the objective properties of physical systems. In cases like this, a scientific discipline experiments what Kuhn called a ‘revolution’: not only a change in the theories that are accepted, but also in the value judgments that support some theories and methods above other alternatives, and McAllister explores in his work the changes in aesthetic criteria that have also accompanied those revolutions.

One possible criticism to McAllister ² is that it is not clear that what he considers ‘aesthetic judgments’ (and particularly taking into account his explanation of why those judgments are important and how they evolve) are ‘really’ aesthetic, or are just another way of talking about methodological or cognitive criteria (e.g., ‘elegance’, in the mouth or the pen of a scientist, can mean something like ‘easy to understand’ or ‘easy to grasp’). The truth is that, actually, some authors have even identified ‘beauty’ (as it plays some important role in science) with ‘intelligibility’ ³. A similar argument has been advanced by Catherine Elgin⁴, who asserts that the pervasive and non-eliminable use of metaphors in science is in itself an aesthetic phenomenon that needs to be assessed, partially at least, by means of aesthetic judgments. However, if beauty reduces to intelligibility (be it of theories, or of the metaphors allowing us to build and understand those theories), then nothing deep is asserted when Platonists claim that truth is a privileged route to our comprehension of the (physical or mathematical) cosmos, for in this case this would be equivalent to the truism that, when we are trying to understand the world, what we are trying to do is… to understand it.

To end with Todd’s words:

“It seems that when, for instance, a scientist exclaims that the theory is ‘beautiful’, she may be doing no more than expressing pleasure in it being a ‘good’ theory, which is to say empirically successful, and appealing to epistemic properties to support her judgement. In other words, there are strong grounds for suspecting that what appear to be aesthetic claims may often be, if perhaps not always are, really masked ‘epistemic’ functional ones and the aesthetic terms at issue – ‘beautiful’ and ‘elegant’ and so on – must be being used in a ‘quasi’, or metaphorical way”. (Todd, 2008, 72).