On Isaac Newton as Philosopher, and the split between Philosophy and Physics

One project that was delayed by my long covid is The Oxford Handbook of Newton (here) that I co-edit with Chris Smeenk. OUP allows online pre-publication, so nearly all of the projected (ca 35!) chapters are already online and have been impacting scholarship and (hopefully) education. I am especially pleased that we include three impressive chapters from the late George E. Smith, “Newton’s Laws of Motion” (view chapter); “Experiments in the Principia” (view chapter); “The Principia, from Conception to Publication” (view chapter).

Newton’s status in the history of mathematics, the history of sciences (especially astronomy, optics, mechanics, and topics like experimentation, laws of motion, telescope design, and cosmogeny) is secure barring global catastrophes that destroy all civilization. (See even Neil de Grasse Tyson on Newton (although ignore the implied chronology).) And indeed, several chapters of the Handbook are written by distinguished historians of science and mathematics. However, the vast majority of the chapters are written by historians of philosophy and philosophers of science. In fact, The Oxford Handbook of Newton appears under the subject, ‘philosophy.’ (In addition, Smeenk and myself are originally trained philosophers.) Since few philosophers tend to treat Newton as a canonical philosopher, this philosophical bias in the Handbook is worth elaborating on.

Before I get to that it is worth noting that historians of science have been retreating from Newton scholarship for over a generation. This is worth remarking on because in many ways Newton scholarship was foundational to the development of the history of science as a distinct profession in the twentieth century: some names immediately spring to mind, IB. Cohen, Alexandre Koyré, Guerlac, the Halls, Herivel, Whiteside, Alan Shapiro, Thomas Kuhn, etc.* Of course, some of their generational colleagues (Stillman Drake, Sarton, Dijksterhuis, Hoykaas, Struijk, etc.) focused on other seventeenth century scientists. The names are not exhaustive because I have not focused on the great historians of astronomy of that era (Gingerich, Curtis Wilson, etc.)

Now, for all kinds of reasons, historians of science have broadened their study away from Newton. This makes sense, of course, in light of the rapid developments in the sciences during the twentieth century, especially. There is a lot more exciting science to cover than there was (say) seventy-five years ago. In addition, the historians of science have gotten a lot more interested in cultural and social issues, and so if they write about Newton at all, it is increasingly less focused on the scientific and mathematical details of Newton.

Thanks to Howard Stein, Ted McGuire, Mary Hesse, Bill Harper, Michael Friedman, and George E. Smith there is a robust tradition of scholarship on the scientific and mathematical details of Newton’s corpus within philosophy. (Of these only Harper is still active, alas.) They and their students (and those they inspired) have been filling the gap that historians of science have opened. My friend, Andrew Janiak, deserves special mention for his edition of Newton: Philosophical Writings, which eased work on Newton by philosophers (and, of course, his monograph Newton as philosopher). The table of contents to the Handbook actually understates the number of philosophers working on Newton in some fashion or another today.

Of course, trained philosophers, even historians of philosophy — however sensitive they are to historical sources and detail — may well ask slightly different questions than historians would. Philosophers have been especially interested in a limited number of themes: Newton’s method(s) of discovery and evidential arguments; Newton’s account of space (and, to a lesser extent, time); Newton’s metaphysics/ontology; Newton’s epistemology (and his rules of reasoning); Newton’s controversies with Descartes and Leibniz (and, more recently, with Spinoza); Newton’s cosmology and cosmogeny; Newton’s theology. So it stands to reason that with the retreat from Newton by historians of science so-called Kuhn losses will pile up.

If you check out the table of contents of The Handbook you will notice that there is a lot more attention to Newton’s own views on the building blocks of reality and that he is put into conversation with many historically salient philosophers. These are two major effects of scholarship during the last half century. Now we have a much better sense how philosophers reacted to Newton and how particular philosophers were building blocks or targets of Newton’s projects.

One may well wonder why it matters to include Newton in the history of philosophy, especially because Newton did not write a major philosophical work. After all, not every person who was a significant interlocuter in his own day should be subject of study by a community of historians of philosophy today. (We don’t do this for Digby, Mersenne, Riccioli, William Harvey, Kepler, Hooke, Halley, or De Volder, etc.) That Newton was seminal to history of science and mathematics is insufficiently explanatory (because we barely write on Euler, the Bernoullis, etc.)

For a long time my own preferred explanation was that the reception of Newton decisively changed something about the way philosophy was practiced in two closely related ways (see here for a scholarly account of what follows): first, a certain kind of argument from authority which could block or silence certain moves became popular in philosophy—the authority was Newton’s works or how they were taken by others. This move was diagnosed early by Berkeley (critically) and Toland (ironically). And detestation of this move animates much of that is great in Hume’s work. Second, as this indicates, Newton, perhaps unintentionally, sets up a split between what came to be known and institutionalized as ‘science’ (especially mathematical physics) and what came to be known and (partially) institutionalized as ‘philosophy.’

In a recent majestic work, Philosophical Mechanics in the Age of Reason (OUP 2023), Katherine Brading and Marius Stan, suggest a different story about this split. On their view, post Newton, the eighteenth-century was unable to unify the physics (a sub-discipline of philosophy) and mechanics (studied by so-called mixed mathematicians) of bodies. In particular, it turned out to be impossible to offer a “single, well-defined concept of body that is simultaneously (i) consistent with intelligible theory of matter, (ii) adequate for a causal-explanatory account of the behavior of bodies, and (iii) sufficient for the purposes of mechanics.” (p. 2) If you read their book, they show that even leaving aside (i), it turned out that combining (ii) [that is physics as practiced by philosophers] and (iii) [that is, mechanics as practiced by the mathematically trained] was impossible for body.

So, on the view of Brading and Stan, those who aimed at this unification (Boscovich and Kant) ended up being treated as philosophers, whereas those who developed mechanics (Lagrange, Laplace, Navier, Cauchy, etc.) ended up being treated as physicists (p. 397).** For them physics — which combined mathematical and empirical methods — took over a century to emerge as a self-standing subject after Newton (p. 397). One nice feature of the story that Brading and Stan tell is that it fits with the rise of physics and chairs in physics as an autonomous discipline.

Before I comment on my disagreement with Brading and Stan it is worth noting a feature of their achievement. I used to think that after consensus was developed on the mathematics of the laws of collision (for perfectly elastic bodies) by Wallis, Wren, and Huygens in the late 1660s, it became perfectly legitimate to screen off and hold at arm’s length certain metaphysical and causal questions from consideration. I still think that’s true; but what Brading and Stan show in their book is that, of course, these laws are perfectly useless for other kinds of bodies (that much was clear) and that the search for how to understand and classify the collision and constitution of other kinds of bodies continued without interruption for the next century. This was eye-opening to me (and it’s interesting to see how important Malebranche was). So, what they show is that there was a widespread felt philosophical need for understanding that no mathematical consensus could block. So, historically my story was simply impoverished.

That rational mechanics (as practiced by mixed mathematicians) and physics (a causal-explanatory account within philosophy) are two distinct and relatively autonomous practices is common ground between Brading & Stan and people like myself. Our real difference is to what degree the aim to offer a “single, well-defined concept of body” should be thought of as the key to understanding eighteenth century philosophy.

Let me close with two observations that show my hand. One person who barely registers in their very rich narrative is John Locke. Now, I am myself not especially fond of Locke, and it’s easy to underestimate him. But the reason why Locke does not fit their narrative is that Locke basically thought that this unificatory project whose history Brading and Stan pursue throughout their book was basically hopeless from the start. For starters he thought collision itself was unintelligible. And this is common ground among all so-called subsequent folk known as ‘empiricists.’

So, the empiricists get off the boat from the start and deny the presuppositions of the whole project of the unification of body. (Some of these empiricists even deny the extra-phenomenal reality of bodies!) So, even before the failure of offering a unified account of body became clear, within philosophy there were (ahh) philosophers who disowned the whole project.

But Locke, Berkeley, and Hume (and the folks they inspired not the least Condillac) do also recognize and diagnose that people appeal to the authority of Newton and mathematical-empirical knowledge as such to generate burden shifting arguments within philosophy not just the science of body. (These arguments are also noticed by folk who do believe in the project of unifying accounts of body.) So, at least some seventeenth and early eighteenth-century philosophers quite clearly recognized that there was already a well-established, fairly robust division of epistemic labor between philosophy and mathematical-empirical knowledge.

---

* Cf. Bernard I. Cohen, "Newton's scholarship in historical perspective." The foundations of Newtonian scholarship (2000): 11-28. Cohen emphasizes the work with manuscript evidence.

**Laplace himself is an interesting case because there may be a split between how he conceived his own project and how he was treated subsequently. I thank Jeroen Van Dongen for this observation.