Patrick Suppes


Lucie Stern Professor of Philosophy, Emeritus
Stanford University, CA, USA

5. What are the most important open problems in philosophy and what are the prospects for progress?

I think there are many open problems that will be of great interest in this century, but I cite three that are especially important and in terms of which great progress will be made. The first concerns our conception of the physical universe, the second is about our understanding of how the brain represents the world around it, and the third is the topic of free will, prominent for centuries in philosophical discussion.

Nature of the physical universe. Because of recent research in many different directions in physics, from experiments in quantum entanglement, which challenge the correctness of the special theory of relativity because of the ability to affect, faster than the velocity of light, the behavior of entangled particles separated by indefinitely large distances, to a variety of findings in astrophysics, we will undoubtedly develop a new set of ideas for thinking about the structure of the universe. New concepts concerning space-time and the nature of matter and energy will cause, in many ways, changes in our conceptual thinking as large as any that took place in the twentieth century. If this prediction is correct, our philosophical conceptions of the physical world in which we live will be changed in the kind of fundamental way they changed with the displacement of the Aristotelian and Ptolemaic conception of the heavens, which dominated our world view for fifteen hundred years, to that of Newtonian mechanics, which itself lasted until special relativity and quantum mechanics were developed in the twentieth century. We are now ready for another such conceptual set of changes, and we will find them at first equally mysterious.

Brain representations of the world. There is a long history of speculative ideas about mental representation, but in mental representations we have the kind of freedom we do not have in dealing with the actual processes of the brain that must embody such mental representations. The deep problems we have only begun to touch are the identification of the kinds of structural isomorphisms that exist between our brain representations of our perceptions or our cognitive patterns of dealing with the world. These structural isomorphisms, as we discover them, should show us in detail how the structure, including the processing structure, of the universe as we perceive it, is represented in our brains. As yet we have barely begun the research required to have a detailed understanding of such structural isomorphisms, but as we do, it seems very likely our conception of our own mental lives will change as well. Let me give just two examples. A widely accepted thesis about the structure of spoken language is that sentences are not understood holistically, but are understood by recognizing the words which are their constituents. No one, as far as I know, defends the thesis that we have a purely holistic understanding of sentences. Our conception of how we compute the meaning or truth of even the simplest sentences depends on some kind of process of recognizing the individual constituents that make up sentences. The corresponding question for the brain is how are these constituents isomorphically mapped into the representation of sentences in the brain. The word or phrase constituents of sentences must be mapped into individual representations in the brain, in order for us to compute, in a feasible way, the meaning and truth of sentences. We surely do not do this computation in some completely holistic way

A second and related example of structural isomorphism is understanding how we recognize or represent in the brain the various natural properties of visual perceptions. Such structural analysis is also needed for perception using the other senses as well. But to take a simple example, consider the perception of a red circle. This simple abstract figure has two properties, one of color and one of shape. How is each of these properties represented in the brain? How do we disentangle the structural isomorphic representation of color from that of shape in the electromagnetic brain waves or signals that reach the cortex and permit us cognitively to recognize and to express that we recognize what we see? The problem when formulated this way in some sense seems simple, but scientifically it is apparently far from being simple. The important point that I want to make is that it is the relation between the constituents, or the parts, and the whole that must be reflected in the structural isomorphism. Finding these isomorphic mappings is a task that is only barely begun, and until greater progress is made our detailed understanding of what the mental, embodied in the brain, really consists of will be poorly understood.

Free will. For me the most important open scandal in philosophy is the problem of free will. I can understand why Hume gave the solution he did to the problem, because he felt he must have an answer that was consistent with the necessary views that held sway, at the time, of the laws of nature. When he said “necessary” he meant in most respects what we mean by the laws being deterministic. We now have a much deeper and more sophisticated understanding of the physics of matter. Determinism is not an idea that has the sway it once did, but we have not properly absorbed what I think is the conversion of the problem of free will from being a traditional philosophical problem to a scientific one. The scientific problem is to give a detailed account of how it is that matter can be intentional in character. In some sense, we all recognize that matter can be animate. This is the great lesson of modern biology. We also need wide dissemination of a thorough analysis of the intentional character of matter. Even though many details are missing, just as they are in case of the evolution of life from inanimate matter, the outlines of the story are pretty clear. The scientific and philosophical mistake was to believe, as Hume, Kant and other great philosophers of the past did, that in some sense the laws of the universe are necessary and deterministic in character. Kant’s antinomies were the apex of this tradition. The great problem for philosophy in this area is to disentangle itself from all the arguments of the past that were mistaken and to develop a view of free will that is consistent with modern conceptions of matter, how matter can be intentional, and how it is realized in some very simple intentional processes like elementary conditioning in biologically primitive animals with a relatively small number of neurons.

There is another side to this story, also important from a different scientific angle. This is the realization that there in not really a strong empirical distinction between deterministic and stochastic behavior. This is a new kind of invariance not well enough recognized in scientific or philosophical circles. The main results are beautiful and surprising ones that come out of ergodic theory, which I shall not try to describe here, but just mention some of the consequences that follow from them. An excellent account is to be found in Ornstein and Weiss (1991). I will end with this example, because of its beauty and profundity at the same time. Consider that hallmark of determinism of the past, an ideal billiard table with an ideal ball moving about it periodically and endlessly in exactly the same endless pattern. Now put on that billiard table a convex object in the middle, off of which the ball must be reflected in its course, just as it is from the sides of the table. There must be errors of measurement, —one of the great truths of modern experimental physics. And these errors of measurement must be bounded away from zero. Then no matter how many observations we may take of the behavior of this idealized ball, which could be a photon “on a billiard table of mirrors,” we cannot decide between the correctness of a stochastic theory of the motion, and that of the classical deterministic theory. The two mathematically formulated theories, one stochastic and the other deterministic, are mutually inconsistent, but due to the existence of errors in the measurement, it is impossible empirically to distinguish one from being more correct than the other. This is a new invariance undoubtedly present in many complex physical systems and of great conceptual importance in the proper view of intentionality and freedom. (A view I defended in a 1993 article on the transcendental character of determinism.)

Read the remaining part of Patrick Suppes' interview in the book Formal Philosophy

ISBN-10    87-991013-1-9    hardcopy
ISBN-10    87-991013-0-0    paperback
Published by Automatic Press ● VIP, 2005

Order now from Amazon!    US    UK    CA     DE