Ciencia vs. Arte. Donald Knuth

En clase dejamos de lado la discusión epistemológica acerca de la diferencia entre las ciencias y el arte. Aquí podemos leer un poco sobre el tema en un fragmento del ensayo "Computer Programming as an Art" (La Programación Computacional Como Arte) de Donald Knuth. Está en ingles como podrán ver. Si algun compañero acomedido y anglohablante se avienta la chamba de traducir sería muy generoso de su parte y muy agradecido por todos.


Science vs. Art

The word "science" seems to have been used for many years in about the same sense as "art"; for example, people spoke also of the seven liberal sciences, which were the same as the seven liberal arts [1]. Duns Scotus in the thirteenth century called logic "the Science of Sciences, and the Art of Arts" (cf. [12, p. 34f]). As civilization and learning developed, the words took on more and more independent meanings, "science" being used to stand for knowledge, and "art" for the application of knowledge. Thus, the science of astronomy was the basis for the art of navigation. The situation was almost exactly like the way in which we now distinguish between "science" and "engineering."

Many authors wrote about the relationship between art and science in the nineteenth century, and I believe the best discussion was given by John Stuart Mill. He said the following things, among others, in 1843 [28]:

Several sciences are often necessary to form the groundwork of a single art. Such is the complication of human affairs, that to enable one thing to bedone, it is often requisite to know the nature and properties of many things... Art in general consists of the truths of Science, arranged in the most convenient order for practice, instead of the order which is the most convenient for thought. Science groups and arranges its truths so as to enable us to take in at one view as much as possible of the general order of the universe. Art... brings together from parts of the field of science most remote from one another, the truths relating to the production of the different and heterogeneous conditions necessary to each effect which the exigencies of practical life require.

As I was looking up these things about the meanings of "art," I found that authors have been calling for a transition from art to science for at least two centuries. For example, the preface to a textbook on mineralogy, written in 1784, said the following [17]: "Previous to the year 1780, mineralogy, though tolerably understood by many as an Art, could scarce be deemed a Science."

According to most dictionaries "science" means knowledge that has been logically arranged and systematized in the form of general "laws." The advantage of science is that it saves us from the need to think things through in each individual case; we can turn our thoughts to higher-level concepts. As John Ruskin wrote in 1853 [32]: "The work of science is to substitute facts for appearances, and demonstrations for impressions."

It seems to me that if the authors I studied were writing today, they would agree with the following characterization: Science is knowledge which we understand so well that we can teach it to a computer; and if we don't fully understand something, it is an art to deal with it. Since the notion of an algorithm or a computer program provides us with an extremely useful test for the depth of our knowledge about any given subject, the process of going from an art to a science means that we learn how to automate something.

Artificial intelligence has been making significant progress, yet there is a huge gap between what computers can do in the foreseeable future and what ordinary people can do. The mysterious insights that people have when speaking, listening, creating, and even when they are programming, are still beyond the reach of science; nearly everything we do is still an art.

From this standpoint it is certainly desirable to make computer programming a science, and we have indeed come a long way in the 15 years since the publication ot the remarks I quoted at the beginning of this talk. Fifteen years ago computer programming was so badly understood that hardly anyone eventhought about proving programs correct; we just fiddled with a program until we "knew" it worked. At that time we didn't even know how to express the concept that a program was correct, in any rigorous way. It is only in recent years that we have been learning about the processes of abstraction by which programs are written and understood; and this new knowledge about programming is currently producing great payoffs in practice, even though few programs are actually proved correct with complete rigor, since we are beginning to understand the principles of program structure. The point is that when we write programs today, we know that we could in principle construct formal proofs of their correctness if we really wanted to, now that we understand how such proofs are formulated. This scientific basis is resulting in programs that are significantly more reliable than those we wrote in former days when intuition was the only basis of correctness.

The field of "automatic programming" is one of the major areas of artificial intelligence research today. Its proponents would love to be able to give a lecture entitled "Computer Programming as an Artifact" (meaning that programming has become merely a relic of bygone days), because their aim is to create machines that write programs better than we can, given only the problem specification. Personally I don't think such a goal will ever be completely attained, but I do think that their research is extremely important, because everything we learn about programming helps us to improve our own artistry. In this sense we should continually be striving to transform every art into a science: in the process, we advance the art.