To: Mark S.
From: Geoffrey Klempner
Subject: Locke's distinction between primary and secondary qualities
Date: 20th October 2009 11:51

Dear Mark,

Thank you for your email of 12 October, with your essay for the University of London Modern Philosophy: Descartes et. al. module, in response to the question, 'Why did Locke think it important to distinguish between primary and secondary qualities? Is his way of explaining the distinction a satisfactory one?

This is in many ways an excellent essay, but I disagree with your conclusion; which is just as well, otherwise we wouldn't have much to talk about!

At one point, your discussion of counterfactual or dispositional properties reminded me of an infuriating argument I once had as an undergraduate at Birkbeck, with a student from my year, whose daytime job was a Tax Inspector, and was also a very proficient lutenist (not that this is exactly relevant, but he was a bit of an oddball). The argument was over the very idea of a 'dispositional' property. All properties, he claimed, were dispositional. 'What about being round?' I asked. Surely being round is definable independently of an object's dispositional properties. 'No,' he said, 'An object is round if you can roll it!' and to prove the point, he took a glass (we were sitting in the refectory at the time) and rolled it on the table. 'See!'

Who was right? Obviously, the question is more general than (and indeed prior to) the question of the validity of the primary/ secondary quality distinction. Intelligent beings whose only perceptual sense was proprioceptive feedback (as in Helen Keller) could distinguish between the capacity of an object, e.g. to bear a given weight without being crushed, and it's size, shape, weight (mass) etc. There are infinitely many dispositional properties (with sufficient ingenuity, one could keep adding properties to the list) but, intuitively, the non-dispositional properties depend upon -- and are therefore limited to -- what would be a complete description of that object (e.g. sufficient to enable us to make an exact copy).

We're both agreed that Locke was wrong about what he took to be the 'primary qualities' of an object. As you remark, science has moved on. What our senses report as being solid, physics tells us is 'mostly empty space', and that very description is itself fatally contaminated by our pre-scientific understanding of the world. String theory takes us even further way from that pre-scientific view.

But so what? Wouldn't that just mean that the primary qualities of objects are different from what Locke thought them to be, and not that there are no primary qualities, only secondary ones?

We need to take a step back: why is the distinction important?

You list a number of reasons why Locke thought it important to distinguish primary and secondary qualities. However, there is an unclarity here about how relative this distinction, and Locke's drawing of it, is to the context in which he was writing. Obviously, Boyle's corpuscularian theory (derived originally from the metaphysical speculations of Leucippus and Democritus) was an important influence. However, as the example of the Greek philosophers shows, the idea wasn't new.

Locke's antipathy to the scholastics is relevant, because Scholastic philosophy developed from Aristotelian doctrine: Aristotle was implacably opposed to the atomist view and the very idea of microstructural explanation, on the grounds that human reason and our powers of perception ought to be sufficient to discover all that there is to discover about the natural world.

If you want to do science in a non-scholastic way, discover the hidden structure of things, how can you avoid a primary/ secondary quality distinction?

There are serious questions about the tenability of Locke's theory of perception and knowledge, in particular the way it generates a 'veil of perception' problem. However, that is just the context in which Locke argued for the distinction, which survives the rejection of Lockean empiricism.

I take on board the point that the 'positive resemblance thesis' cannot be salvaged, at least in the form that Locke intended it. As you argue, 'resemblance' is far too flexible a notion anyway (Nelson Goodman in 'Languages of Art' offers a powerful argument against the view that a representational painting 'resembles' its object more or less closely). However, something survives: namely, the fact that an explanation of why we perceive the world in the way we do will be couched in terms which ultimately refer to whatever physics posits as the 'ultimate' physical structure of things.

Or maybe not: we are assuming that in principle there is a route from physical theory to the explanation of the phenomenal quality of experience; there are many philosophers who would disagree. From the assumed truth that facts about experience supervene on physical facts, it does not follow that there is, in principle, a reductivist explanation available of experience in terms of those facts. However, I take it that the basic (negative) point of the primary/ secondary quality distinction would still apply.

All the best,

Geoffrey