In chapter 2, Feser discusses Plato and Aristotle. One feature of his discussion is that he accepts a fairly standard view on which Plato is the crazy metaphysician, and Aristotle takes the good parts of Plato's metaphysics and grounds them with a healthy mix of common sense. As I understand it, this is roughly Aristotle's interpretation, and I think it has misleading aspects, but it's at least partly true. It's also why I like Plato better than Aristotle, which is of course the reverse of Feser's judgment. The problem with tempering your philosophy with common sense is that it's actually pretty common for common sense to be wrong, and if you make a mistake as a result of faulty common sense, people may fail to notice the mistake for centuries, or even millenia. On the other hand, if you make a mistake in your wild metaphysical flights of fancy, people are sure to call you on it, as they apparently did with Plato; he wrestles with very serious objections to his various theories in the course of the dialogues.
But that's all very general; on to specifics. First, on the progression leading up to Plato, it is perhaps not surprising that Feser bashes the sophists as early examples of liberalism. Their presentation in Plato's writings is actually much more nuanced than this, and it is not clear that Socrates is well described as "vigorously opposing the sophists" (Aristophanes didn't think so, and Aristophanes was a fairly perceptive guy). Gorgias comes off pretty badly in the dialogue named for him, but most of the other sophists who actually appear in the dialogues are much more complicated characters.
On to Plato. Most of what Feser has to say about Plato is fairly orthodox, which is not to say I agree with all of it, but most of my suspicions don't seem very relevant. Feser classifies Plato as a mathematical Platonist, which is reasonable enough though probably shouldn't just be taken as given. But Feser gives an odd argument for mathematical Platonism, saying, for example, that "it is not up to us to decide that the angles of a triangle should add up to 38 degrees instead of 180." I assume that he has heard of non-Euclidean geometries, so I am curious as to exactly what he means when he says this is not up to us. Why isn't it up to us whether we adopt a Euclidean geometry or a hyperbolic geometry? What forces one choice rather than another? Or does Feser have some view on the relationship between the two which explains this?
Einstein thought that we should make a distinction between "pure" geometry and "applied" geometry. Pure geometry is an entirely abstract science in which both Euclidian and hyperbolic geometries, as well as of course elliptic geometries, are in their own ways true as all they say is that certain conclusions follow from given axioms and in each case those conclusions do follow from the particular given axioms, but none of them tell us anything about how the world actually is. Applied geometry, on the other hand, describes the shape of space, but is for that reason dependent on empirical results (and Einstein of course argued that an elliptic geometry best modelled the shape of space). But Feser wishes throughout to argue that claims about the world can be established by purely rational means, so the sort of distinction Einstein draws here (and which is accepted by many others, including myself) would seem to undermine Feser's attempt to use mathematics as an example of the legitimacy of using purely rational means to discover facts about the world.
It seems to me that one of the major reasons for the decline of rationalism, not mentioned by Feser, is that in the early modern period it became quite apparent that traditional logic was entirely inadequate, and that traditional mathematics also had unacceptable limits; Leibniz and Newton introduced calculus, and of course mathematical innovation continued and provided many valuable tools to the advancing sciences. However, the new mathematics was also a theoretical mess, now known to have contained not a few outright contradictions which the mathematicians were simply fortunate in not accidentally drawing consequences from. The criticisms of calculus by people like Berkeley were far from completely unwarranted, even if the needs of science meant they were largely ignored.
The situation was largely cleaned up in the 19th century, of course, with the work of a wide variety of mathematicians and an especially large contribution by Frege. Still, the cleanup was not complete, and some of the remaining puzzles (like the Euclidean vs. Non-Euclidean geometry issue already mentioned, plus in the early 20th century Russell's paradox, Gödel's theorem, and questions surrounding controversial mathematical principles like the axiom of choice), meant that confidence in mathematics as a source of eternal truth has not returned; to many of us, mathematics just seems more like something we make than a matter of eternal truths we discover. Carnap's "in logic there are no morals!" seems the principle best suited to the variety and flexibility of modern logic and mathematics.
Of course, Plato does not only discusses mathematical forms. At least he appears to discuss non-mathematical forms; there are at least two important caveats. First, Plato expresses doubt on occasion as to whether some of his example forms are genuine forms, and indeed some of his forms seem to be intended as jokes rather than serious examples (e.g. the form of the bed in Republic). Second, Plato may have believed that even forms which appeared not to be were ultimately mathematical, in line with the Pythagorean influences on his thought. Still, setting aside those caveats, there is certainly discussion of apparently non-mathematical forms. Plato has the form of the Good, and while I don't recall him ever using Feser's example of the form of the squirrel, he does seem to have thought there were forms for biological species. This is one of the points on which the ancients and Feser seem to be most clearly wrong; evolutionary biology has given us a very good understanding of biological species, and it is in conflict with this account based on forms, but more on that later since Feser does address evolution directly in later chapters.
It is not clear how much damage it does to Plato's system to eliminate the form of the squirrel. He may still be able to maintain the form of the Good, and insist that morality is based on reason, by taking something like the Kantian view of ethics (though I expect there are details on which Plato would not have agreed with Kant). How viable that is remains a hotly contested question; I reject Kantian ethics myself, but I think Korsgaard et. al. deserve considerably more respect than Feser gives them when he directly mentions Kantian ethics later in the book.
When he turns to Aristotle, Feser presents some quick arguments that forms are necessary. He gives the example of mathematical forms, which he thinks are necessary for us to make sense of mathematics. Maybe, or maybe not, but he continues to fail to note how problematic it is to equate mathematics with his own idiosyncratic examples of rationally acquired knowledge. For example, some of the things which Feser says can be established by conclusive rational argument are quite obviously contingent. Whether something is contingent or not is itself a necessary truth (at least in S5; perhaps Feser is employing some other view of modality, but that would raise other problems of its own). So it's necessary that those claims can't be established by conclusive rational arguments, since no contingent claim can be established by conclusive rational arguments.
Now, of course, it could be that the obvious contingency of the claims Feser thinks can be rationally proven is on a par with the obvious truth of Frege's Basic Law V. But this applies equally to Feser's principles; if there's any way to settle who has the right of the matter here, Feser never supplies it, as he doesn't even consider this problem.
Feser defends a realism about categories against nominalism and conceptualism by claiming that our practicies implicitly assume realism, and that nominalism and conceptualism couldn't do the job. He waves his hand at the (entirely adequate) reply available to modern nominalists and conceptualists, namely an evolutionary story of concepts (Millikan has done especially impressive work in this area). His criticism is that we need concepts to engage in evolutionary explanation in the first place, and that on this story the concepts we get from evolution have no "objective validity." Unfortunately, this is only true if "objective validity" is defined in such a way that nothing has it. The evolutionary story provides a perfectly adequate account of the origin of concepts, including providing us with reasonable justification for relying on the concepts produced by that process.
Though Feser isn't willing to give enough credit to the evolutionary story for him to consider this question, it occurs that somebody could ask about pre-Darwin nominalists and conceptualists. They didn't have any adequate account of the formation of concepts, and yet they rejected realism anyway. One might argue that this shows that they must have rejected realism for reasons involving the kind of biases Feser constantly attributes to his opponents, and so that this at least provides support for his position by showing how strong the biases are. However, I would draw a different lesson. Realism doesn't actually explain concepts; it just says comforting and familiar-sounding things about them until we stop worrying and think everything is all right with concepts. The early nominalists and conceptualists decided that no explanation at all was better than the comforting pseudo-explanation offered by realism (in the tradition of Socrates' view that wisdom is knowing when you don't know something), and so they rejected realism even before there were rival explanations. This did in fact turn out to be beneficial, in just the way Socrates says awareness of one's own ignorance is beneficial, as it spurred efforts to find genuinely good accounts of concepts, which we now have. It seems to me that the early modern rejection of final causes also fits the same pattern as the nominalist/conceptualist rejection of realism; the early moderns rejected final causes as unhelpful despite lacking any replacements, and science proceeded in new directions, producing vastly more adequate explanations of the world as a result of rejecting unhelpful pseudo-explanations. Again, things have now reached the point where the old pseudo-explanations aren't even superficially appealing in comparison to the genuine explanations modern science can provide.
Feser's criticism of Kant appears to be that while Kant avoids the relativism Feser sees in the nominalists and conceptualists, the deeper problem, the one already mentioned, remains. I'm not sure what his basis is for attributing relativism to the nominalists and conceptualists (I suppose that as usual much depends on what is meant by the very slippery term "relativism"), but of course I've pointed out that the deeper problem isn't one, so Kant seems to be off the hook.
There's some interesting projection when Feser says liberals like Plato because they like his idea of philosopher-kings. I've never thought Plato was very serious about that; after all, he made the rather impressively unpolitical Socrates the great hero of his dialogues. I also find it somewhat odd that a person with a Ph.D. is so knee-jerk skeptical of intellectual elites, while being equally knee-jerk submissive to political and economic elites, but of course Feser is a standard conservative in that respect. Further, accoring to Feser, the rejection of Aristotle is responsible for Star Wars Episode I (well, probably. Feser blames everything else he doesn't like about the world on it. Of course, perhaps I shouldn't assume that Feser didn't like "The Phantom Menace;" he doesn't show that much evidence of good taste elsewhere in the book).
In any event, the idea of potentiality is obviously very important to Feser, so we should be careful with it. Potentiality could be defined purely logically in terms of the actual future of things like the actual thing in question, or the range of futures of things identical to the thing in question in specified respects. So defined, it would be unproblematic. I would not quibble with someone wishing to introduce such a concept. "In logic there are no morals!" But Feser means more, and the question is whether his arguments establish that there is an interesting or even meaningful concept stronger than these bare logical potentialities which can do the work he requires.
The first feature that Aristotelian potentiality has is that it must be activated by something outside of the actual thing with the potentiality. Feser says that absent this, it would be inexplicable why a potentiality would be actualized at one time rather than another, and so suggests that he thinks even bare logical potentiality would have to have this feature, but if that's what he thinks, he's obviously wrong. It is not apparent why there would need to be an explanation, and considerably less apparent why the explanation would need to be something external to the actual thing. Maybe some things just have it in their nature to actualize some potentiality that they have randomly, or some specific amount of time after they come into existence. Or maybe something is going on that we haven't thought of. Feser's usual pattern is to jump to a plausible-sounding but uninformative explanation to escape the horror of not being able to explain something, but there are lots of things that we don't know how to explain. Another thing I like about Plato is that he seems to have realized this.
The second feature that Aristotelian potentiality apparently has that bare logical potentiality lacks is that it is, we are told, possible for something to be merely actual, with no potentiality at all. At least, logical potentiality rules that out for anything except perhaps the final stage of something which exists at the end of time. But there's something which is not the final stage of something which exists at the end of time which Feser says has no potentiality, namely God.
The other features of potentiality that Feser comments on are that potentiality depends somehow on a thing's nature and that one can make distinctions between different kinds of potentiality; both of those appear to be true of bare logical potentiality.
Feser's discussion of form and matter leaves a number of issues unclear. In particular, can one say anything true and informative about Aristotelian matter at all? It seems that all of what one would ordinarly consider to be the properties of things are, on the Aristotelian picture, aspects of the form of the thing. It is thus quite misleading when Feser attributes to the materialist the view that anything, such as his example rubber ball, is "just a piece of matter." No materialist thinks it's just a piece of something like prime matter; that wouldn't make any sense. And, whatever Feser may want his readers to think, materialists generally avoid saying things that are that obviously nonsense.
Feser jokes about getting a martini before writing about Aristotle's four causes. I kind of wish he was drunk while writing this section, and really the whole book; it wouldn't really be an excuse, but it would at least make the sloppiness somewhat more explicable. In any event, his arguments do seem to become more sketchy and gappy whenever his writing gets more cutesy. And, yes, his writing is fairly cutesy in most of the book. The obvious inference is sound.
He continues to fail to clarify puzzling aspects of the matter/form distinction (perhaps because it's so obvious and common sense it doesn't occur to him to wonder if it remains coherent under closer examination)? He identifies the rubber as the material cause of the rubber ball. But surely it is some kind of form that makes something rubber. And in fact it seems to be forms all the way down to the most basic substances that make up reality. So what makes the cause "material?" What does it have to do with matter? And, at the risk of getting ahead of ourselves, if the forms are doing all the work, how can Feser even tell when matter is present and when it isn't (as he claims to be able to do with allegedly immaterial souls)? Also, Feser will eventually explain that the material cause determines a thing's potentiality. But surely what a thing is potentially depends on what it is actually like, and so depends on the properties (aspects of the form) of the thing. I don't understand what role matter plays in this.
The formal cause is what a thing is like, and the efficient cause is whatever actualized its potentiality. If potentiality is understood in the bare logical sense mentioned earlier, this would seem to make the efficient causes of Aristotle very similar to the view of later philosophers on all causes.
The final cause is the goal, end, or purpose of a thing. Aristotle attributes final causes to inanimate things, making it less than clear how appropriate this "purpose" talk is. In any event, it is obvious that one could construct a bare logical version of final causes along the lines of the bare logical versions of potentiality, and indeed the two would be closely related. Again, the question is whether anything stronger than the bare logical version is coherent, and whether there is any reason to postulate or invoke it.
Feser casually dismisses Hume's discussion of causation, which has been tremendously influential for very good reasons. Just because we encrust correlations with reassuring names doesn't mean that we understand them any better, and in fact once we remove the encrusting Aristotelian nonsense, what remains may look strange, but functions just as well or better. Certainly Feser's rejection of Humean causation on the basis that efficient causation is at work when an event is identical with itself strikes me as bizarre.
Feser says "an attentive reader may have noticed that Aristotle's account seems to entail a series of simultaneous causes and effects, and might also wonder where such a series terminates and how it can be explained." Actually, I noticed instead that it's not obvious how to put simultaneous events into a series; if it really is simultaneous, it is not clear why it needs to terminate, so that worry didn't occur to me.
Feser also asserts an Aristotelian principle that whatever is in the effect must in some sense be contained in the cause as well. It can't be said that he argues for it, though "in some sense" is so incredibly vague that it's also not clear what counter-examples there would be, or what the principle even means. Certainly there's nothing about his rather narrow examples to suggest how this would generalize to all causation ever.
He says that naturalists reject this principle, and I suppose I reject some forms of it (Feser would have to say what he means by it before I could decide if I reject his version). According to Feser, it is "sometimes suggested" that evolution contradicts this principle. I have no idea who would suggest that; Feser's note on this sentence does not concern where this suggestion comes from. I certainly don't see why evolution would be more in conflict with the principle than anything else in science; if the principle is stated in an unacceptable form, it probably contradicts evolution (and lots of other facts), and otherwise not.
There are a number of points Feser indicates he will explain in more detail later. It's rather distressing how infrequently his later discussion is actually any more detailed or helpful, but more on that in future installments.
