On Leibniz's Mill

There was a recent post on Dr. Feser's blog reviewing a book on the challenge to materialism presented by Leibniz's Mill, which refers to the illustration Leibniz offers in section 17 on his The Monadology. I haven't read the book, and so offer no opinions there. However, The Monadology and Dr. Feser's post are both items I will to comment upon below the fold.

Starting with The Monadology, the first four parts discuss what these Monads are, and what is means for them to be simple. In particular, Monads are simple, without parts, are not collections of things, can’t be extended, divided, created, nor destroyed through natural means. So, from the start, the grounding of this notion is outdated physics, as monads are created/destroyed by phenomena such as pair production or pair annihilation. Since monads are proposed to have a property that is counter to reality, any discussion of them based on that property will not be descriptive of reality.

The example of Leibniz’s Mill comes in section 17 of The Monadology; the claim is that perception can’t be explained mechanically (if we enlarge the brain to a size where we could walk around in it, we can’t point to any given activity as a perception), and is therefore simple. Leibniz then uses this notion of simplicity to talk about the notion of a soul, how it starts, etc. Since the notion of simplicity itself is not descriptive of reality, there is little point in going into detail about the results from this notion. That’s one of the advantages of arguing in a formal system, a mistake at the beginning invalidates the entire argument.

However, according to Dr. Feser, the book under review uses the other aspect of Leibniz’s Mill (the inability to point to a perception in a physical model) to show that mechanical descriptions of nature can’t account for perceptions, a position Dr. Feser endorses. The basic style is a careful presentation of positions that Dr. Feser feels can be refuted. I don’t find some of the refutations to be particularly convincing.

The first is the notion that since we would not recognize collections of nerve activity as being perceptions on sight, asserting they are is basically doing an end-run around Leibniz’s position. However, this is a play to an argument from our ignorance. The real issue is not a lick of connection between brain activity and mind activity, but the lack of the ability to form sight interpretation. Indeed, given a precise translation table between brain states and thoughts, you would expect even under Dr. Feser’s Scholasticism that observing particular brain states exactly corresponds to particular mind states, and the lack is the ability to make the translation.

I do agree with Dr. Feser that taking the position there are no thoughts (eliminative materialism) concedes that Leibniz is correct you can never see a thought. However, I don’t see where that is a problem for eliminative materialism, any more than conceding you can never see a unicorn would be a problem.

Dr. Feser discusses the idea that this could be compared to a computer, where you could see computation going on but not understand the output of the program (again, unless you had a precise translation table). His response is that there can be no such thing as computation without something that assigns meaning to the processing, just as there is no meaning to a written word except as assigned by the reader. However, there is a fundamental difference between an active collection of objects and a passive collection of markings. The water cycle has many of the properties we assign to algorithms, and we can discuss its effects and consequences without making reference to whether the cycle has been designed. There is far less discussion to be had regarding a small collection of rocks lying on the ground. Saying that mental activity which is not previously programmed does not qualify as computation is a matter of terminology, not effect.

Finally, while I’m agnostic on whether the activity cycles of a computer/brain can be pointed to as a separate level of existence or not, such existence would be a far cry from embracing the entire Scholastic metaphysical system or some sort of mystical existence. Noting that activity patterns have cumulative effects does not require that they point to anything outside themselves (so no final causality) and certainly does not entail an exterior mind to direct them. Thus, there is nothing in the computationalist response that need disturb the materialist in the slightest.