Metaphysical Functionalism

The standard approach to philosophy has been to begin with a philosopher -- a rational adult of considerable education, uncommon openness to new ideas, exceptional aptitude for logic, and other outstanding traits. Next, put the philosopher to work on problems or puzzles about the universe and our place in it. With sufficient effort applied to set aside the philosopher's preconceptions and biases, it is hoped, the resulting beliefs might be the result of reliable deliberations -- of sound or cogent arguments.

For some time, now, many philosophers have been concerned about the dramatic dependency our professed beliefs have on other, deeper, beliefs (such as a rather complex folk psychology that almost transparently guides our daily lives) -- often "self-evident" notions that are rarely noticed, let alone examined or justified. For this reason, I think it's important to consider that a philosopher begins forming beliefs as an infant -- and that beliefs accumulate and interact, eventually to reach in adulthood what might have seemed to be the starting point for philosophy.

Any individual's philosophy begins with a metaphysics (defined, for my purposes, as the study of what exists [ontology] and how it all fits together [cosmology])-- developed through childhood and adolescence as a natural part of learning to deal with our environment. We have robust beliefs about the world and how it works long before we develop the capacity for philosophical reflection. We cheerfully accept the view of the world given to us by our perceptions until, realizing that our senses are fallible, we come to accept the modern view that our senses give us representations of material objects (rather than the objects themselves). Informed by this more sophisticated epistemology, we theorize that the representations we are given genuinely (under normal conditions) provide us information about properties of material objects we observe. We refine our theories based on this information -- or, in philosophical terms, we refine our metaphysics according to the deliverances of our epistemology, which in turn must be supported by our refined metaphysics or face refinement itself. This is a constructive feedback loop whereby our beliefs about ourselves and our world can be regularly improved.

Let me put all this in somewhat different terms. Among the functions we perform as human beings are (1) generating perceptual representations of our environment and its fixtures; (2) making mental models of how the physical world in which we exist works -- of rules the world apparently obeys; (3) adjusting our mental models to fit our perceptions, and our interpretation of our perceptions to fit our mental models; and, additionally, (4) communicating, as best we can, these models and rules to our fellows (while, conversely, learning some of them from others).

Together with standard evolutionary theory, the foregoing account of human perceptual and theoretical functioning supports the "trialist" theory of Sir Karl Popper that Sir John Eccles (1985) discusses. According to it, three "worlds" contain all of experience and existence: World 1 is the material world of physical objects and processes; World 2 is the experiential world of perceptions, consciousness, and thought; World 3 is the world of culture, of communicable ideas -- all of history, literature, science, art, philosophy, etc.

On this view, we experience a purely mental world -- Popper's World 2 -- whose sensory and theoretical representations we trust correspond (in most essential aspects) to the material World 1 we believe we are a part of. If we are correct, then these representations are generated entirely by physical processes -- i.e., World 2 emerged from World 1. We may be attracted to the view that our representations are those physical processes rather than some emergent, non-physical entities supervening on the physical, and that World 2 is therefore contained in World 1. Like Paul Churchland and others, we might go so far as to say that World 2 has no actual existence, but is merely the way we experience the workings of our World 1 brains. Those who hold mental experience to exist in a non-physical world do so without the support of the hugely successful tradition in science of explaining our theories in terms of physics and its derivative theories (called "Galilean reduction" by Allan Gibbard), for an explanation of the non-physical in physical terms is ruled out by definition.

I have noticed that Galileans, though inclined to seek theoretical reduction in physical terms, nonetheless typically view both physically recorded information (e.g., books, paintings, electronic data) and human behaviors that are emulated and shared without recourse to a tangible, physical medium of exchange (e.g., natural language, folk traditions, cultural norms) as belonging to World 3. The words, musical compositions, and social mores of human culture, as examples of this category of World 3 entities, may appear to be non-physical. We easily distinguish their material representations (in print, in performance, and in legislation, say) as instantiations of these abstract entities -- not the entities themselves. To put a finer point on it, humanity (born of material processes) has created a realm of non-physical, abstract entities.

Strictly speaking, I think this view is mistaken -- but how?

The elements of World 3 might have some manner of existence in World 2 (via individual experience) and World 1 (as, say, neurochemical processes, or arrangements of ink on paper) -- but what manner of existence would that be? I think it is the instantiation of non-physical, abstract entities, which are the actual denizens of World 3: things such as sets, numbers, relations, functions, logical operations, natural laws, possibilities, concepts, and propositions. Most especially, I need to include in World 3 what I call valences -- dispositions of a situation, either actual or counterfactual, which are simply ways the components of a situation (objects with interactive properties) can interact.

Such abstract entities might be instantiated in World 2, World 1, or both -- or neither. Four sticks of equal length, properly arranged, can instantiate a square; a pair of words can instantiate the number two, though a written numeral or a spoken word can only mention it.

World 3, I think, is a purely abstract world. It doesn't contain our meanings; those are World 2 entities -- dwelling in individual minds -- that merely instantiate World 3 entities. Instead, World 3 contains all conceivable propositions related to all possible worlds. Combined, the actual arrangements (past and present) of the stuff of the universe, plus the theoretical furniture of every individual's mental world of experience, lifelong -- all instantiate a negligible subset of these infinitely numerous abstract entities.

The amount of overlapping content between any two individuals' mental representations of abstract entities will depend on an array of cultural and idiosyncratic influences. Even with a tightly circumscribed culture in common, I think it's possible (even likely) that most concepts we think we share with our fellows are actually just very similar concepts representing similar abstract entities. (Worries like this inspire talk of "radical interpretation"; Gibbard (101, 156-160) discusses Donald Davidson's views on this matter.) On a functionalist, Galilean, view, it matters little whether a genuine "meeting of the minds" (concept for concept) is ever achieved; what matters is that each individual's concepts are useful in promoting survival and reproduction.

Certain propositions in the abstract World 3 will be true or false according to their relation to some semantic system that Carnap might have called a 'language'. Indeed, a single proposition can be true or false according to the language chosen. We can imagine distinct domains carved out of this abstract world for individual metalinguistic systems and semantic systems, with overlap wherever a concept or proposition is meaningful in multiple systems. There are semantic facts pertaining to individual semantic systems. "And -- true propositions in World 3 about properties of the physical World 1. We have no direct access to this physical domain, and can only approach it by way of the second domain with which we are specially concerned: metaphysics. This domain of World 3 contains all of the propositions represented by our World 2 theories about what exists (in the present, past, or future, or about what could possibly exist) and about what relations can or do obtain among entities. Once again, we can imagine distinct territories carved out for individual metaphysical systems.

Galilean metaphysics deals with the physical, mental, and abstract worlds. It includes the belief that there is a realm of semantic facts in some possible language that has the same content as the realm of substantive facts about the world. The functionalist variety of metaphysics to which I subscribe also includes the belief that it is possible for human ideas to represent many such semantic/substantive facts, and harbors the hope that there is available to us some collection of these whose web of entailments will explain all aspects of the physical world (including ourselves) that are important to us.

The Galilean task is to refine metaphysics to conform to substantive facts. Its goal (perhaps unattainable) is for metaphysics to include the sort of core semantic/substantive facts just mentioned (and no propositions contradicting those facts). Its method is straightforwardly empirical, and Popperian in the sense that it seeks to falsify particular theories by using them to predict how certain behaviors we might undertake will result in certain sensory experiences, and taking the failure of those experiences to occur as proof that one's theory is false. Its greatest obstacle is theoretical holism, which acknowledges the impossibility of determining for certain where within our metaphysics the flaw revealed by experiment lies -- in the particular theory in question, or in some theoretical presupposition(s).

The best pragmatic response to the problem of theoretical holism, I think, is a faith in probability. The Galilean core has a constructive circularity: our physical theories explain how creatures like us would evolve to have the sorts of sensory experience and cognitive mechanisms we do, and this in turn explains how we developed our physical theories. While confirmation of sensory expectations in any given case is very weak support for one's theory, continued accumulation of such confirmation carries increasing weight. It would be more than strange if an adult's lifetime experience only accidentally corroborated the Galilean core of beliefs. It would be bizarre indeed if hundreds of years of apparent sensory confirmation of this core, reported in historical accounts of both daily life and methodical investigation, were purely coincidental.

Much more impressively, however, it would be beyond bizarre if the last billion years of life for our ancestors -- relying on the veracity of sensory impressions, perceptual appearances, and cognitive expectations -- flourished due, not to an adequate, rudimentary metaphysics (or its behavioral, functional equivalent), but to chance alone! Thus justified, the Galilean core is much less suspect when predictions fail than are its more recent theoretical extensions. This is clearly a very Quinean view, bolstered by evolutionary psychology.

There may be Galileans who would disagree with aspects of this brief account; indeed, I have never heard one quite like it from any philosopher or scientist. Its components, however, have each had ample support from within the Galilean tradition, so I have some hope that their synthesis will meet few serious objections. For the purposes of this project, however, the indispensable part of it is the distinction between semantic and substantive facts, and the claim that evolutionary (and, recently, cultural) selection has brought about (and can refine) a remarkably close correspondence between these for propositions that are of great import for organismic survival and reproduction. If that is granted, the rest can be harmlessly disputed.