This would be part 2 of my review of The Last Superstition. Last time, we took a quick look at the general logical structure of the book, without actually evaluating any of the arguments within it. In the remaining posts in this series, I will attempt to evaluate these arguments and determine if Feser's case is successful.
To recap: Feser starts by arguing for realism about universals. He then moves into Aristotle's metaphysics, arguing that the existence of change entails the distinction between actuality and potentiality, that the existence of universals entails formal causes, and that the existence of causation entails final causes. He then argues that this combination - actuality and potentiality, formal and final causes - entails the existence of God, the immortality of the soul (specifically, hylomorphic dualism), and natural law morality. He then argues that the mechanical philosophy that denies formal and final causes is incoherent and reduces itself to absurdity, thereby giving both a positive and a negative argument for Aristotelianism.
Now, this seems all fine and dandy on the face of it. Feser is certainly a very competent philosopher, and I agree with his thesis that Aristotelianism is just as defensible today as it ever was. His case certainly needs to be taken seriously, especially by anyone who wants to defend modern secular morality from the conservative, religious morality of natural law. But I do not think that Feser's conclusions are rationally unavoidable. In fact, I will argue that Feser needs a very specific formulation of Aristotelianism to be true in order for many of his arguments to go through, and that this formulation (which is presented in the book) faces some problems.
As stated before, Feser begins by arguing for realism about universals. He gives 9 arguments for realism - 5 direct arguments, 2 arguments against nominalism, and 2 arguments against conceptualism. Now, for some universals - like mathematical entities or propositions - I think these arguments work well in establishing realism. But Feser needs more than just those kinds of universals to be real - what he needs, in order for his case for hylomorphic dualism and natural law to go through, is for complex biological forms - like "dog", "squirrel", "frog", and "human" - to be real, mind-independent universals. Otherwise, an individual dog (ignoring consciousness for the moment) is very plausibly nothing more than a collection of elementary particles, with "dog" being an organization that we project onto the particles, similar to us seeing a meaningless ink splotch as the letter "A". I will argue that conceptualism, at least about universals for composite biological entities, is perfectly defensible, and that Feser's arguments against it don't quite hold up.
Surprisingly, only a few of Feser's arguments for realism apply to biological universals. The first of these goes like this:
1. The “one over many” argument: “Triangularity”“redness,”“humanness,” etc., are not reducible to any particular triangle, red thing, or human being, nor even to any collection of triangles, red things, or human beings. For any particular triangle, red thing, or human being, or even the whole collection of these things, could go out of existence, and yet triangularity, redness, and humanness could come to be exemplified once again. They also could be, and often are, exemplified even when no human mind is aware of this fact. Hence triangularity, redness, humanness, and other universals are neither material things nor collections of material things, nor dependent on human minds for their existence (Kindle Locations 925-930).
On first glance, this argument seems sound. At least the first part, which argues against nominalism, seems correct. But a conceptualist would not grant that universals (at least biological ones) are "instantiated" when no human mind is aware of this fact. For whether or not something even counts as instantiating the universal "dog" in the first place, the conceptualist would argue, depends on whether we decide that it counts. What is there objectively is a collection of particles; the concept "dog" is something we apply to it in order to make sense of it. Additionally, our concept "dog" is actually a set of arbitrarily-chosen properties; had we chosen different properties as relevant for classification, objects that we now count as dogs may or may not have still counted as such, or we may not have even had the category "dog" at all. So while this argument may hurt the nominalist, it does nothing to hurt the conceptualist.
The second, third, and fourth arguments are about mathematics and propositions, which as I said before are not relevant to the kinds of formal causes Feser needs to make his case. The fifth argument goes like this:
5. The argument from science: Scientific laws and classifications, being general or universal in their application, necessarily make reference to universals; and science is in the business of discovering objective, mind-independent facts. Hence to accept the results of science is to accept that there are mind-independent universals. Science also makes use of mathematical formulations, and since (as noted above) mathematics concerns a realm of abstract objects, to accept the results of science thus commits one to accept that there are such abstract objects (Kindle Locations 946-950).
Again, this argument, on the face of it, seems logically airtight. But I would challenge the idea that science, above the level of basic physics, is really in the business of discovering facts that are entirely mind-independent. The fact is that the higher sciences, like biology, exist because of human interests; in principle, science could describe the entire universe in terms of particle physics, but because humans have special interest in (what we see as) biological systems, we have the science of biology, which operates within the framework of our biological concepts. And whatever facts hold about the physical world given these biological concepts, hold because of the facts about the physical world describable by physics. Again: in principle, science could describe the entire universe in terms of particle physics, but because of human interests, we find it convenient to describe certain parts of the universe in terms of "cells", "reptiles", "mammals", "plants", and other concepts that we project onto the physical world to make sense of it. This, at least, is what the conceptualist would say.
Feser gives 2 arguments against nominalism, which, as far as I can tell, are sound (at least against "traditional" versions of nominalism; trope theory may or may not be able to answer them). But then he gives 2 arguments against conceptualism, which are very closely related. The first of them goes like this:
8. The argument from the objectivity of concepts and knowledge: When you and I entertain any concept – the concept of a dog, say, or of redness, or of conceptualism itself for that matter – we are each entertaining one and the same concept; it is not that you are entertaining your private concept of red and I am entertaining mine, with nothing in common between them. Similarly, when we each consider various propositions and truths, we are entertaining the same propositions and truths. So, for example, when you think about the Pythagorean theorem and I think about the Pythagorean theorem, we are each thinking about one and the same truth; it is not that you are thinking about your own personal Pythagorean theorem and I am thinking about mine (whatever that would mean). So, concepts (and thus universals) and propositions do not exist only in the mind, subjectively, but independently of the mind, objectively (Kindle Locations 985-992).
The ninth argument says roughly the same thing, but extends it a bit:
9. The argument from the possibility of communication: Suppose that, as conceptualism implies, universals and propositions were not objective, but existed only in our minds. Then it would be impossible for us ever to communicate. For whenever you said something – “Snow is white,” say – then the concepts and propositions that you expressed would be things that existed only in your own mind, and would thus be inaccessible to anybody else. Your idea of “snow” would be entirely different from my idea of “snow,” and since your idea is the only one you’d have any access to, and my idea is the only one I’d have access to, we would never mean the same thing whenever we talked about snow, or about anything else for that matter. But this is absurd: we are able to communicate and grasp the same concepts and propositions. Hence these things are not subjective or mind-dependent, but objective, as realism claims (Kindle Locations 992-999).
I do think that these arguments have something to them. But they don't quite prove what they may, at first glance, seem to. What these arguments prove, at best, is that our concepts - like "dog", "cat", "squirrel", "elephant", etc. - exist in some sense independently of any individual mind. What they do not prove is that these concepts existed as mind-independent universals before any mind thought of them. It is entirely possible that they were indeed invented by us in order to make sense of the material world, and that they occupy a "third realm" beyond the material world and the world of the mind. In fact, philosopher of science Karl Popper believed something much like this. But if universals like "dog", "cat", "squirrel", and "elephant" only exist because we decided to invent them, then they are not "objective" in the way that is needed to establish that individual dogs, cats, squirrels, and elephants have objective "essences" or "forms" over and above the parts that compose them. And without that, the cases for hylomorphic dualism and natural law fall apart.
Besides Feser's arguments not quite holding up, there are positive reasons to think that biological universals like "dog" are arbitrary concepts that we invented rather than objective "essences" that are instantiated in nature. As mentioned before, properties like "dogness" actually refer to a collection of properties: there are certain properties that, if exemplified by a particular, make that particular a dog. As Aristotelians would say, there are certain "essential properties" that make something a dog, along with some "accidental properties" that individual dogs will have that are not essential to them being a dog.
But why choose this particular collection of properties as essential, rather than some other properties? Why classify based on these properties, specifically? It seems that we have simply chosen certain properties as relevant for classification, and created a taxonomy based on them. Had we chosen different properties as relevant for classification, we may have ended up with radically different categories. And in that case, the categories we have - like "dog", "cat", "squirrel", and "elephant" - cannot be objective universals. The properties that make up these categories may in fact be objective universals (although that could be contended as well), but the categories themselves are arbitrary and dependent on our whims. And this is true regardless of whether or not, as Karl Popper would likely hold, these categories turn out to exist in a "third realm" of abstract concepts that we have invented.
So much for formal causes at the level of biology, then. In my next post, I'll tackle final causes, and show why I think they are problematic (at the macro level) as well.
No comments:
Post a Comment