The Last Superstition – A Refutation of the New Atheism: The Scholastic Dead End.
In Part 7 I discussed how mathematics broke from Platonism and from a pre-dominant metaphysics. As I noted there, the various foundational crises in mathematics did not disprove or even really discredit Platonism, instead they made it viable to be agnostic on those issues. In addition the metaphysics no longer directly informed the epistemology. Indeed a philosophical rival to Platonism, mathematical formalism (in which axioms are like the arbitrary rules of a game) while flawed in a number of ways could still be a productive stance. The old question of whether mathematics was discovered or invented is not one that can be settled but in an age of invention and in an age of computer model ‘invented’ made a lot of sense as a way of working. While computer programming had old roots in the age of the Industrial Revolution, in the second half of the Twentieth Century it became a major part of humanities culture both intellectually and economically. Inventing formal languages, developing data structures and discovering surprising features of invented logical structures became something familiar and which we had a shared cultural experience of.
The break with Platonism within mathematics is notable because it was relatively recent, it was part of deep intellectual changes (relativity, the birth of the electronic digital computer), and it occurred in relatively secular times. The break with scholasticism occurred much earlier but confident with that break was a sea-change in the intellectual life of Western Civilisation. What was cause and effect is hard to say but Feser’s position was that the baby was thrown out with the bath water. This is true in so far as it is still possible to rationalise modern science with a Thomistic/Aristotelian framework. The problem for Feser is what does that framework bring to the party? Feser’s answer is that it brings a coherent metaphysics but the permanent question for science when regarding a philosophical framework is not the metaphysics but the epistemology and the methodology.
Put another way, is reasoning within the kind of Thomistic framework that Feser proposes a good idea? Is it productive intellectually or is it prone to error? Back in the day, we know the answer. Scholasticism was the source of errors that became entrenched and difficult to displace. Teleological arguments were used to justify the Earth being the centre of the universe – it provided a simple explanation for gravity, that masses belonged to the centre of the universe and that is where they would tend to go. In the light of understanding of gravity was that conclusion a necessary outcome of Thomsitic thinking? No, but it was an outcome of Thomistic thinking and it was an obstacle that our understanding of gravity and the heliocentric view of the solar system had to overcome. Likewise the view that planetary orbits must be based on perfect circles was also Thomistic in nature, appealing to a Platonic/Aristotelian notion of perfection – that planetary orbits are very nearly circular ellipses doesn’t count as a near miss. Yes, again it is possible to reframe the work of Kepler (who was himself seeped in Platonic thinking) into Feser’s framework.
The point here is not that Thomistic thinking could not be repeatedly tweaked to deal with intellectual advances but rather that it acted like an intellectual handbrake. It provided very little that was good in terms of intellectual advance in the science and rationalised a lot of the bad. People dumped it by virtue of it being the bulwark of people who were repeatedly demonstrably wrong. To some extent that was simply because it was the established philosophical position of the intellectual elites during a time of an intellectual revolution. For reason more socio-economic than philosophical, control over the sciences was shifting from the churches and church institutions to more secular sections of society.
However, I would also contend that Thomistic scholasticism made it very easy to be confidently wrong about abstract matters and to confuse contingent or physical issues with metaphysical ones. By tying Aristotle to Church doctrine, Aquinas (probably unwittingly) also tied this kind of rationalisation to any topic on which the Catholic Church might wish to pronounce – it amounted to the worst possible combination of making the sciences (and everything else) subject to theology. To dispute geocentrism was to dispute the rationalisations for geocentrism and those rationalisations were based on the SAME metaphysics as the rationalisations for all of Thomas Aquinas’s theology (irrespective of whatever Aquinas would have though if he had been around to see the evidence for heliocentrism).
Feser’s version of Aquinas’s thinking has multiple ways of being wrong on an issue. That is not a flaw. A scheme that is incapable of being in error is a scheme that is essentially tautological and a tautological scheme is necessarily a scheme that says very little. However, it is worth considering the specific kinds of error that arise from it. Fewer contrast Platonic realism regarding universals with two alternative: nominalism (universals are essentially just semantic categories) and conceptualism (universals are cognitive categories we use for making sense of the world). I’m not going to go into great depth on either of those because it would take too long but it is worth noting that even if we accept Platonic realism the universals we discuss are 1. described using words and 2. do also refer to concepts that we have i.e. a triangle may exist as some kind of abstract entity in a Platonic realm but we still have an English word ‘triangle’ with a definition and we have all formed our concepts of a triangle. Putting Platonism aside for a moment, in a discussion if there is a mismatch between the words we use and the concepts we are attempting to refer to then misunderstanding will occur. With Platonic realism we have a three-way error – potential mismatches between Platonic reality, words and concepts. In theory what Platonic realism offers as a consequence is a way of grounding both words and concepts to the Platonic reality. It makes it possible that there really is an objectively right definition of a word/concept, except that, aside from mathematics there isn’t any direct way of checking our words against this abstract realm.
To illustrate consider this Proposition X: a rectangle is a trapezium. Is it true or false? First of all there is a superficial but significant problem with words. British English and North American English had an exact swap in usage for the words ‘trapezium’ and ‘trapezoid’. A British trapezoid is a four-sided shape with no parallel sides and in the USA that is a trapezium. There isn’t a conceptual difference, it is just that the words have swapped for reasons that are obscure. So an American will say that Proposition X is false but if the American knows that I’m using British English they will know that I’m not using the same concept of a trapezium – I’m thinking of a different geometrical category. As it happens I am NOT using the US term – so having sorted that out can we now decide whether Proposition X is true or false? Unfortunately if we turn to normal non-technical dictionaries there is an ambiguity. A trapezium is often defined as a four sided shape with one pair of parallel sides. Now a rectangle does have one pair of parallel sides because it has two pairs of parallel sides and if you have two of something then you must have one of something…That is a somewhat odd way to read the definition but it is logically consistent. What we have is two different readings: 1. a trapezium has at LEAST one pair of parallel sides (Prop X is true) versus 2. a trapezium has EXACTLY one pair of parallel sides (Prop X is false). Technical mathematics dictionaries will vary on this but in general 2 wins out because people don’t tend to think of rectangles as trapeziums but definition 1 has advantages because it makes for a neat hierarchy of shapes. In usage, examples like ‘the trapezium rule’ (for approximating integrals) imply that rectangles sort of do count. Either way we have two different concepts and in Platonic realism two different abstract categories (one a subset of the other) but only one word and a fuzzy distinction.
Surprisingly mathematics copes with this confusion. In theory it is problematic whatever kind of reasoning we use but in reality we all cope with loose definitions and loose concepts. Scholastic philosophy though, was very much an inquiry into the nature of these (assumed real) universals and as a consequence they had to pay very close attention to fine distinctions between words and needed a highly technical vocabulary. Fewer describes the issue like this:
But suppose that we interpreted this vocabulary in terms of a nominalist or conceptualist metaphysics, rather than a realist one. Then all those complicated technicalities would reflect, not objective reality, but only our subjective ideas or the way we decided to use words. the Scholastic philosophy that inherited this terminology would come to seem an exercise in mere wordplay and irrelevant hair-splitting, rather than a serious investigation of the real world.
This tendency in Scholasticism is, I believe, fundamental to how it became such a stifling dead end. Fewer sees this as primarily only a problem when the pressure for terminological exactitude was presented in a nominalist framework but the problem was deeper than that. The Platonic view of universal also encouraged the notion of the universe ordered into distinct categories that could, with a fine enough definition, be separated one from another. In addition the primary logical tool used was the syllogism which was essentially categorical in nature. Messy categories with awkward boundaries and exceptions could not only not be well handled by the logic but they also ran somewhat counter to the spirit of the approach. Platonic realism could cope with such categories in so far as it was accepted that we live in an imperfect world but attach this concept to Catholic theology you end up with an implication that the exceptions, the messy boundaries and the hard to classify were either analogous to sin/evil or actually were an expression of sin.
This categorical approach was a cognitive straight jacket. Science still needed technical language and clear definitions but the shifts in methodology allowed for more flexibility both because of innovations such as numerical continuity (measurement) but also because rather than mounting single long chains of reason, observed data was used as a means of testing reasoning against reality. As Scholastic metaphysics lost currency so did the syllogism as the primary form of logical reasoning and the emphasis into reasoning about the world shifted from reasoning about what kind of thing a given subject of inquiry was to how that thing worked.