Friday, July 27, 2012

Universals and an Argument for the Existence of God: More on Edward Feser's The Last Superstition an Unpublishable Review, part 2

This post is the second installment in a review of Edward Feser’s The Last Superstition, a polemic against the New Atheists, their movement and their beliefs. In this installment I discuss Feser's presentation of an argument for the existence of God based on the existence of universals or abstract objects.

Universals are properties or features common to many individual things or concrete particulars, as philosophers call them. An example makes this a whole lot clearer. Bob is a tall human being with a limp. Bob is the particular thing that instantiates (are instances of), exemplifies (are examples of), partakes of, or participates in certain attributes or features which are not strictly parts of himself, and in his case these are the properties humanity, tallness, and gimpiness (the limp being not a thing that Bob has like a coin in his pocket but a feature or aspect of Bob’s gait which itself a property of Bob). Now, there are philosophers who think these features are real aspects or features of the world, and others who deny this. There are so many subtle distinctions among the different beliefs one can have about these supposed features of the world, and the reasons for and against each such position, that even the most basic introduction to the question takes about 65 pages in Michael Loux’s excellent introductory volume Metaphysics: A Contemporary Introduction (which I highly recommend). So it is difficult for Feser, and it is more difficult for me, to provide any fair, balanced, and complete explanation of this issue. So, I will settle for a link to this excellent article by one of my former teachers on the Stanford Encyclopedia of Philosophy on Properties and the following brief categorization of the views. Some of my terminology is standard, but I'm not sure about the term "conceptualism", and I make no attempt to comprehensively describe the alternative positions.

Realism
Features of concrete particulars are real properties that are shared among these particulars.

Two subcategories of Realism are Platonism and Aristotelianism.

Platonism
These features are real properties that exist independently of the individuals that have them.

Aristotelianism
These features are real properties shared by individuals but they depend on those individuals and cannot exist without any instantiations.

Nominalism
There are no such features; there are no shared properties that particulars instantiate.

Two (non-exhaustive) subcategories of Nominalism are trope theory and conceptualism. This is not all the ways one could be a nominalist, but they are worthy of mention given Feser's argument.

Trope theory
There are features that particulars have, but these features are themselves individual or particular (so there is no such thing as a redness that is shared among all red things, but each individual red thing has its own particular redness).

Conceptualism
There are no universal features of particulars; categorizations of particulars under the same type (red things) is only an idea in people’s minds. There is no external or real feature of the world that justifies the categorization into types of things; the categorization is taken as basic.

My point here is not really to attack or defend these positions, but to note the distinction between different positions that Feser does not distinguish and which lead to incoherence in his conclusions.

Feser argues for the existence of abstract objects, universals, that exist independently of their instantiations. Feser presents a set of arguments for Platonism, ignores the arguments for the other side, and fails to distinguish different attitudes towards universals which, if noted, would undermine his conclusions. Feser talks about the reasons for the reality of universals but almost none of the problems with them. Presenting only the arguments for your own side but not addressing the arguments for the other side is like claiming to have won a basketball game while only mentioning your team’s score. Yeah, a hundred points is a lot, but. . .

More problematic is Feser’s incoherent view of universals. His arguments for the existence of universals focus on the Platonic notion of abstract objects (the famous Platonic Forms) that exist independently of any instantiation of them. Feser recognizes there are problems with this view and endorses an Aristotelian account of universals without noting that his arguments for them are inconsistent with the Aristotelian concept of them. He can either have the arguments he provides or he can have the Aristotelian he tries to draw from those arguments, not both.

I will attempt to support the above claims in the following post. First, I will outline the Augustinian argument Feser backhandedly endorses. Then I will explain why the premises are actually inconsistent with his Aristotelian account of universals. Then I will show why this argument does not prove the existence of God because treating universals as ideas in God’s mind does not resolve the difficulties in the concept of universals.

Feser introduces the following Augustinian argument for the existence of God as one Alvin Plantinga says must be taken seriously. I assume this means that Feser also thinks it is sound, but there are two levels of plausible deniability here. First, saying that Plantinga endorses it does not mean Feser endorses it. Second, saying that we should take the argument seriously is not the same as saying it is sound. Still, if we are not to think the argument is sound, then Feser is just jerking us around. So, I'm going to assume Feser is not jerking us around. Here's the Augustinian argument following my own standardization.

1. If universals exist, they must exist either as abstract objects, universals that depend for their existence on their instances, concepts in imperfect or in perfect minds.
2. Universals exist.
3. Universals cannot exist as abstract objects.
4. Universals cannot exist as universals that depend for their existence on their instances.
5. Universals cannot exist as concepts in imperfect minds.
6. Therefore universals must exist as concepts in perfect minds.

An abstract object is an object that exists in some Platonic third realm, as a kind of individual thing that exists in the abstract. Potential examples of abstract objects include the number 1, the meaning of sentences, propositions, perfect geometrical figures. The first premise obviously looks a bit like a false dilemma. Why think these are the only ways that universals might exist? I include these because they are the views that Feser implicitly or explicitly argues about. Feser runs together the Platonist and Aristotelian accounts of universals, but I will treat them as distinct since Feser’s failure to distinguish them undermines his argument.

I should take a moment to define the terms I introduced: “perfect minds” and “imperfect minds”. An imperfect mind is either a finite mind, such as a human one, that cannot cover the entirety of the universe in both time and space, or that cannot conceive all possible propositions or all possible mathematical truths. I also wanted to convey the problem that a finite mind might make mistakes in its categorization of objects into groups, and that this could cause problems if one tries to define the existence of properties as being only concepts in minds (thus, an infinite but still fallible mind will not do the trick). If a mind failed to adopt a grouping that we intuitively take to be legitimate, then the property at issue would not exist, and that’s clearly not an acceptable result. So we need to distinguish the kind of mind we have and the kind of mind God is purported to have. There may well be only the one perfect mind, or there may be an infinite number, but I’m not concerned with that claim because I don’t think ultimately solving the problem of universals by appeal to any perfect mind is going to work.

Feser argued for premise 2, the existence of universals, at some length in an earlier chapter of the book. In particular, Feser argues for Platonism. Platonic objects are abstract entities that exist independently of any instances, that can exist without instances. For example, if all physical triangles are imperfect in some way (the angles total something very slightly different from 180 degrees), but triangularity is still real, then triangularity exist independently of any instance. Feser argues for the this view of universals as things that exist independently of any instances. Hence, Feser, in his argument for the existence of universals, argues for a particular view of them, and that I indicate in premise 4.

What are Feser’s arguments for the existence of Platonic universals? There are several, but they all focus on the truth value of claims that Feser thinks can only make sense if universals exist independently of their instances. For example, mathematics, geometry and logic are thought to involve true statements yet none of the terms need refer to any physical existents. If so, then these entities exist independently of their instances (since they exist but possibly have no instances). There are difficult and complex arguments in many areas of philosophy (philosophy of mathematics, philosophy of language) that dispute Feser’s claims, so it is not easy to be as sanguine as Feser that we should accept the conclusions. Nonetheless, I will not worry about these arguments since I am not writing a metaphysics textbook.

Feser does acknowledge that there are problems with Platonism, problems that may or may not be debilitating, and then claims that Aristotelianism provides a better account of universals. Now, these forms of realism about universals are inconsistent with respect to the existence of uninstantiated universals. Platonism says that these abstract objects exist, independently of their instances, in a third realm distinct from our minds and from the physical world. This view causes all sorts of ontological and epistemic problems. By contrast, Aristotelianism says that there are no uninstantiated universals. So, if you argue for the existence of uninstantiated universals, you cannot then take those arguments to support Aristotelianism. If Aristotelianism is the correct view of universals, then it cannot be because the arguments Feser presents are convincing. Still, in chapter 2 of his book, Feser endorses an Aristotelian account of universals despite having just made arguments against it (by arguing for the existence of uninstantiated universals). Perhaps it is because of this inconsistency that Feser then almost-endorses the Augustinian argument for the existence of God. He may be unhappy with the earlier realist solutions to the problem of universals and adopts a different one, especially since he thinks it can be used to prove the existence of God. And if he endorsed either of the other two views of the universals, he would not be able to use them to attempt to prove the existence of God.

In short, Feser thinks we have to accept that there are universals and he thinks that they must exist independently of their instances, so he needs an account of universals that does not endorse either Platonism or Aristotelianism. Existing as ideas in God’s mind is, then, Feser’s way out of his dilemma. If universals are concepts that exist in God’s mind, then they can be objective (God’s mind being perfect after all) and real without entailing all that problematic stuff about a Two Worlds ontology that Plato had. Plus, they get to exist even if there are no instances since God can think about things that do not exist. The problem is that this “solution” contains the worst aspects of Platonism and conceptualism.

Here are some problems with universals that Feser's solution does not actually solve.

Knowledge:
One problem for Platonism is how finite, physical, limited minds can gain information about the world of abstract universals. If these items exist in a Platonic heaven (i.e. without any instances), how could we come in contact with them, how could we know they exist, how could we know facts about reality that depend on them? Plato offers only metaphors (the myth of the cave, the divided line, etc.) in trying to explain how we know them. Other Platonists, such as Kurt Godel and Gottlob Frege, have not been much more help. However, as problematic as Platonism is, placing the ideas in God’s mind does not solve these problems. God’s mind is at least as inaccessible as Plato’s heaven. We have no more access to God’s mind than we do to Plato’s world of the Forms, so Feser’s solution does not solve problems of knowledge for universals.

Communication:
Feser thinks that communication would be impossible if there were not Platonic universals that constitute the meanings of at least some of our terms. There certainly is a good case to make that nominalism can provide no adequate substitute for cases of apparent abstract reference (e.g. “Blue is a friendlier color than red.”) If universals are ideas in God’s mind, however, communication becomes again mysterious. Communication would only be possible if we could refer to ideas in God’s mind. Yet, since we lack any access to God’s mind, these ideas cannot show how communication is possible. It’s mysterious enough how I could refer to an idea in someone else’s mind; it is completely mind-boggling to consider how I could refer to an idea in God’s mind. The degree of difficulty involved in referring to ideas in God’s mind is so great that it would seem to be impossible. Hence, on Feser’s Augustinian proposal, communication is apparently impossible.

Objectivity:
Since universals must exist objectively in order to allow for communication, knowledge and the rest, we have to ask: Are God’s ideas objective? One might think so, but answering this question raises a parallel to the classic Euthyphro dilemma (are actions good because the gods love them or do the gods love them because they are good?). Do objects instantiate the universals they do because God conceives of them as doing so? Does God impose universals on the world by his choice? If so, then God can have no reason for ordering the universals as he/she does. (If Fido was a dog this morning could God make Fido a cat this afternoon if God only changed the way he conceived of Fido?) Or does God conceive things as he/she does because they are objectively like that? It’s obvious that if we are to take God’s conceptions as objective, they must be based on an independently existing reality, on independently existing universals. If not, then the distribution of universals to objects must be completely arbitrary and God must have no reason for conceiving of objects as he/she does.

Similarity/one over many:
The problem that universals are primarily introduced to solve, that objects appear to be similar or resemble each other in certain respects, is not really solved by appeal to ideas in God’s mind. Do objects resemble each other because God conceives them as similar, or does God conceive them as similar because they bear an objective resemblance to each other? If we take the former answer, God’s categorizations must be brute unexplainable facts (which is the same answer the nominalist gives, so there is no advantage for the Augustinian here). The answer must again be, for reasons given above, that God must recognize objective resemblances rather than create resemblances. Hence, we must have some other view of the similarity objects appear to have to each other.

This is not to say that the other solutions to the problem of universals are good, but trying to solve the problem by appeal to God does not help matters. It’s possible that one could solve these problems by fiat, and simply claim that the nature of God’s mind is such that these problems are solved. But if you allow that then you might as well allow the Platonist or the conceptualist to solve the problems in the same way. Ultimately, you get nowhere by appeal to God to solve this problem.

In sum, Feser argues that there must be universals understood as Platonic, abstract objects. Then he claims that the best view is Aristotelianism about universals which entails that universals are not Platonic, abstract objects. Then he presents an argument for the existence of God as ideas (and hence not universals or abstract objects) that exist in God’s mind. Thus, according to Feser, universals are apparently all of the following: abstract objects that exist independently of any instance (Platonism), universal resemblances that depend for their existence on their instances (Aristotelianism), and ideas in the mind of God (and hence not independent entities). Feser seems to settle on the last response, but that response does not seem to resolve any problem that universals might be introduced to solve.

One of the oddest facts about Feser’s discussion of nominalism and realism about universals was the weight he placed on it. Somehow nominalism is to blame for the excesses of liberalism, atheism, the sexual revolution, the murdering of millions of helpless babies, and, no doubt the Holocaust and the evil that is Def Leppard too. This is a peculiar stance since some of the most cogent nominalist positions were put forward by theists (most famously, William of Ockham) and many contemporary atheist and materialist philosophers are realists. Bertrand Russell was perhaps the 20th century’s most famous atheist and also a realist (at least in his Problems of Philosophy). Most importantly, there seem no links of entailment or likelihood between any of the supposed social ills and the philosophy of nominalism; nor do there appear to be particular historical connections among them. What could the belief that there are no universals have to do with immorality? Perhaps Feser believes that nominalism entails moral nihilism or relativism, but you can still believe there are individual good or bad actions (tropes), just that there are no universals that explain a resemblances among them. Even if one particular argument that God exists would be rejected by nominalists, it does not immediately follow that they reject the existence of God. So, even if nominalists reject Feser's arguments, there is no necessary connection between nominalism and any moral ills at all.

Since this post has been technical and mostly no fun at all, I wanted to give one more example of the conservadroid mind at work for its sheer humor value. In his second chapter, on the topic of the ancient Greeks, before talking about the universals which are important to the arguments I just discussed, Feser mentions Socrates and his execution. His thoughts on Socrates and the analogy to contemporary politics and society once again indicate the peculiar sense of persecution and victimization so often found on the political right in America. Feser writes,

Socrates (469-399 B.C.) vigorously opposed the Sophists and the moral corruption they had fostered within the Athens of his day. . . When put on trial by a jury of 500 of his fellow citizens for purportedly denying the gods of the city and replacing them with new ones, and in general corrupting the youth . . . he defended himself, Plato tells us, by claiming that he was divinely called to lead others to the improvement of their souls. Naturally, this democratic assembly had him executed. (Today, they’d probably just denounce him as a “neo-con” or part of the “religious right” and haul him off for multicultural sensitivity training.) (p. 31)

Just to get this straight, Feser is claiming that the great iconoclastic thinker Socrates, who was executed for his heretical views on the polis and the nature of the gods, is the equivalent of the religious right, who ran all of the Bush domestic policy and are one of the most powerful domestic interest groups in American politics, and the neo-cons, who ran American foreign policy for the entire Bush era (and who not only have not been reeducated but have, despite demonstrable falsehoods that led the U.S. into a war in Iraq, remained mysteriously unindicted) Let’s just have a little comparison of Socrates, the religious right, the neo-cons, and the new atheist movement.

Socrates
Accused of atheism and impiety

Neo-cons
No

Religious right
No; they’re the ones making these accusations

New atheists
Yes

Socrates
Had no political power

Neo-cons
No—they had a prominent role American foreign policy for years


Religious right
No—one of the most powerful interests in American politics


New atheists
Yes; just for example, there is only one openly non-theist in Congress

Socrates
Represented a minority opinion


Neo-cons
Maybe, but not really in political circles


Religious right
Yes, the religious right is not a majority, but Christians are a majority and the religious right is disproportionately influential because they are seen as representing Christians


New atheists
Yes

Socrates
Committed to critical thought, doubt, and questioning established beliefs


Neo-cons
No


Religious right
It is to laugh.

New atheists
Yes

Socrates
Argued for inconvenient truths that those in power did not want to hear

Neo-cons
Hah!


Religious right
Is that milk coming out of your nose?


New atheists
Yes

Socrates
Said that the gods spoke to him to encourage him to tell the truth


Neo-cons
No; pretty clear they have no divine mandate to tell the truth


Religious right
Yes; they claim a direct line to the divine, but not in any way Socrates would recognize*


New atheists
No but they do see themselves as bringing the truth to people

Socrates
Was executed for his ‘crimes’

Neo-cons
No; strangely unindicted for potential war-crimes

Religious right
No; historically these are the executors not the executes

New atheists
Not lately

So, Feser’s sense of persecution is so strong that he perceives the most politically powerful elements of America, with which he is aligned, as being powerless, besieged by political correctness and crushed under the boot-heel of postmodernist literary theorists and Michael Moore (also on p. 31). Compared to this, the dissociation from reality of Feser’s religious views is fairly mundane stuff.

*In contrast to the contemporary religious right, Socrates never relies on divine inspiration for any of his claims. In the Ion Socrates ridiculed the poets for their inability to defend their beliefs, ironically suggested they must be divinely inspired. Have you ever seen a cogent defense of the religious right’s views? Must be divine inspiration. Socrates doesn’t claim divine knowledge; in fact he denies that anyone has divine knowledge.

1 comment:

  1. Good discussion! I think you may have the problem of knowledge backwards; it is not so much a mystery for the theory of forms that we have knowledge of forms, as it is part of the theory that forms play a dual role in structuring our minds and the world (well, it remains a mystery how anything can play that dual role, but that doesn't seem like specifically a problem about knowledge). Rather, the knowledge puzzle for theories of forms is how are we able to make mistakes, if the knowledge of forms is built into our minds? Hence Plato's lengthy (and inconclusive) discussions of how error is possible, which often seem odd to modern readers.

    I have a series of posts about Feser's book on my own blog (protagoras.typepad.com), btw.

    ReplyDelete