blog.kennypearce.net

Newton's Rationalism

One of the problems for the traditional 'Rationalists and Empiricists' story of early modern philosophy is that it is surprisingly difficult to define 'rationalism' and 'empiricism' appropriately (see here for a previous discussion). One traditional way of drawing the distinction, derived from Locke, is over the existence of innate ideas. This distinction, however, does not capture what is of importance to many other early modern philosophers, and oddly excludes Malebranche and his followers from the rationalist camp. (Since Malebranche holds that no ideas are ever in the human mind—they are all in God—he holds that no ideas are innate to the human mind.) Another traditional way of drawing the distinction, derived from Kant, is over the existence of a priori knowledge. This is perhaps somewhat more promising, for Locke's critique of innate ideas is presented as a component of a broader critique of innate knowledge. However, most of the philosophers usually classified as empiricists accept at least some a priori knowledge: for instance, in mathematics. Kant would say that the rationalists accept synthetic knowledge a priori, but the analytic/synthetic distinction is a Kantian innovation with no precise parallel in earlier modern philosophers.

An approach which is perhaps more promising, in terms of its ability to connect to explicit subjects of debate in the period is the definition of rationalism in terms of ratio, i.e., reason. The rationalist, on this telling, distinguishes between the faculty of sense/imagination and the faculty of pure reason/intellect. The empiricist collapses them. On this way of drawing the distinction, Cartesianism turns out to be a paradigmatic form of rationalism (good), and Malebrancheans get to be rationalists for the same reason other Cartesians do (also good). Further, Hobbes and Gassendi offer explicit arguments in favor of empiricism in this sense, and Berkeley and Hume appear to presuppose such an empiricism. Still, there are some odd consequences. The question whether Locke is an empiricist turns out, on this approach, to be a difficult interpretive question rather than a straightforward textual one, though Locke does strongly suggest empiricism (in this sense) by his intentional collapse of the distinction between 'species' and 'notions' (EHU §1.1.8). A stranger consequence (which perhaps suggests that this account should not be pushed back before the mid-17th century) is that the traditional Aristotelian/Thomistic picture turns out to be a form of rationalism, despite holding that "there is nothing in the intellect which is not first in the senses," since it does affirm a distinction between sensory and intellectual representations.

One more curious feature of this approach (which is the reason I am thinking about it today) is that it turns out that Newton offers an explicit argument for this kind of rationalism in De Gravitatione:*

If anyone now objects that we cannot imagine extension to be infinite, I agree. But at the same time I contend that we can understand it. We can imagine a greater extension, and then a greater one, but we can understand that there exists an extension greater than we can imagine. And here, incidentally, the faculty of understanding is clearly distinguished from imagination (Janiak 38).

Now, in a way this is not surprising. In Descartes (and Plato), as in Newton here, there is considerable evidence that the affirmation of rationalism (in this sense) arises from reflection on the phenomenology of mathematics: many people who have a great deal of experience in mathematics report the experience of encountering an object not revealed by the senses, hence one supposes that there is a faculty of understanding that has objects of its own, distinct from the objects of the senses. Perhaps these objects may be somehow derived from the senses, in a manner consistent with the Aristotelian dictum ("nothing in the intellect that was not first in the senses") as the Aristotelians interpreted it, or perhaps not. Nevertheless, the idea/notion of extension contemplated by the intellect is unlike anything known by the senses, for the senses know only particular images of extension, all of which are finite.

Hobbes, Berkeley (at least on my reading), and Hume all hold, on the contrary, that this mathematical activity, which may be somehow and in some sense about infinite extension, nevertheless employs, as the mind's immediate object, only finite determinate sense images. These images, which according to Descartes and his followers are the objects of the faculty of sense/imagination and are not even properly called 'ideas', are in fact all the ideas there are. It can be seen now why Locke's empiricism is somewhat ambiguous: although he rejects the species/notion distinction, whether his abstract ideas are imagistic in this way is highly controversial. One can also see here that Newton's rationalism (in this sense) is part of a broader tension in the development of physics, which to some degree continues to this day. Galileo, Leibniz, and Newton all insist that a proper approach to physics must be both mathematical and experimental, but math itself is, of course, precisely not experimental. For Newton (at least in De Gravitatione), just as much as for Descartes, many of the fundamental concepts of physics (most notably, in both cases, extension) are mathematical concepts attained by the pure intellect and differing radically from anything perceived by the senses. Yet (against Descartes, in agreement with Galileo and Leibniz) Newton holds that the laws of physics, employing those concepts, must be derived from sensory experience. And of course even Descartes holds that the laws ought to be applicable to what we experience by means of the senses. So there is no obvious contradiction between Newton's rationalism and his empirical/experimental methodology, but there is an apparent tension, or at least a collection of difficult philosophical questions (which are, again, still very much alive) concerning the very concept of (what we now call) applied math. Though these sorts of questions are by no means absent from (e.g.) Plato, the mathematization of physics through the 17th century suddenly places them among the most important questions in natural philosophy, a role they had not previously occupied.

Cross-posted at The Mod Squad.

* I should note at the outset of this discussion that I am not a Newton specialist and, other than looking at the introduction on the Google preview this morning, have not read this important book on the subject I am about to discuss. I am aware that the relationship of De Gravitatione to Newton's published works is a vexed question. Blog posts are not journal papers.