Metaphysics is dead. Long live metaphysics! Assorted musings of a philosopher-scientist

(image from

I am a scientist and a philosopher. As such, I have a problem with metaphysics. A serious problem. Apparently, the word “metaphysics” was coined by an anonymous first century editor of Aristotle’s works. The person in question assembled a small number of the philosopher’s writings and called them “ta meta ta phusika,” literally meaning, the stuff that comes after the Physics, the latter being one of Aristotle’s most famous books.

Ever since, metaphysics has been described as the study of what exists, in the most general manner possible. Aristotle did not invent this kind of study, as it was carried out by the Pre-Socratics of the 6th and 5th century BCE. It continued with Plato, Aristotle himself, and a number of other Greco-Roman and Christian philosophers (in the west, then of course we need to mention the Arabic, Judaic, and more far eastern traditions as well).

It ended, as far as I’m concerned, with Descartes. To be more precise: a particular, traditional style of metaphysics ended with the famous French thinker. That style, sometimes referred to as “first philosophy,” and nowadays as analytic metaphysics, hinges fundamentally on the notion that we can discover things about the way the world is by thinking about it, as people do in mathematics or logic. Call it the rationalist approach.

That approach, as I said, died with Descartes and his bold attempt at radical doubt. Descartes understood that he could reasonably doubt pretty much everything he thought he knew. Not just about the external world, but also his notions of mathematics. The only thing he couldn’t doubt was the very fact that he was a thinking being: cogito ergo sum. I think he was right, by the way. But the trouble started when he attempted to use that simple notion as the foundation to rebuild the entire edifice of knowledge: start with something you absolutely cannot doubt, and go from there.

He got stuck immediately, and made the predictable bad move: he invoked God as a guarantor of his “clear and distinct” perceptions about things, assuming that God wouldn’t trick him. This is known in philosophy as the Cartesian circle, as in circular logic. And it’s the shoal against which the purely rationalist program in philosophy crashed and sank.

Not at all coincidentally, that was also the time that marked the beginning of modern science. Descartes thought of himself as a physicist, and he engaged in metaphysics in order to pave the way to his physics. Galileo, his contemporary, thought of himself as a (natural) philosopher, but his inquiries in physics helped develop the empiricist program that eventually fueled the empiricist philosophy of David Hume, and finally resulted in the separation of science from philosophy.

Ever since Galileo, we turn to science, not metaphysics, to understand how the world works. Before you accuse me of scientism, let me provide you a spoiler: I will argue below that this does not mean that metaphysics is out of business. It just means that it has to reinvent itself. (Another spoiler: it already has, in certain quarters!) Indeed, I have argued before that what I am describing here is typical of the evolution of philosophy as a field: whenever a piece of it spins off into its own area of inquiry (physics, biology, psychology, etc.) the new field itself is now subject to philosophical inquiry (philosophy of physics, of biology, of the social sciences, and so on). But no serious philosopher would today consider doing the job of a biologist or a psychologist. So why do analytic metaphysicians insist in doing the job of scientists?

Let me give you an example, to ground this discussion a bit. Take the perennial problem of personal identity: what constitutes a “person”? What criteria shall we deploy to identify persons? What makes me ten years ago, or ten years in the future, “me”? As it happens, I’ve been teaching precisely this material to my students at the City College of New York over the past couple of weeks, as part of a course in science fiction and philosophy.

The literature on personal identity is huge, and the number of philosophical accounts (I don’t really like to call them theories) is vast. Should we say, with Locke, that memory is what makes a person that particular person? But then what happens if a neurological disorder wipes out most of your memories? And since you presumably don’t remember what you did when you were a 4-yr old child, does that mean that it wasn’t you? Well, perhaps the right criterion is physicalist one: what makes you “you” is that you are made of the same stuff at different points in time. Except that you aren’t: the actual stuff of which our bodies are made changes continuously (a problem known as Theseus’ ship). And moreover, what would happen if the two hemispheres of your brain where transplanted in another body? Or if you stepped into a Star Trek transporter?

This is lots of fun, but even my students — with relatively little background in philosophy — quickly zeroed in on the problem: all these accounts assume that there is a fact of the matter about what constitutes a person. If that were the case, the question would become empirical, and would be settled by the proper special science (presumably, cognitive science, or social psychology). But there is no such fact of the matter. “Person” is a human cultural construct, not a thing out there in the world. So yes, both memory and physical continuity (as well as personality, among other things) are the kind of criteria we normally deploy to figure out if a person is a person. The apparently problematic counterexamples mentioned above are simply anomalous, pathological, or fantastic (Star Trek) cases, so it is no surprise that our everyday notion of personal identity runs into trouble. But this is because personal identity is a Wittgensteinian family resemblance concept, the result of a “language game” we have agreed to play. Not a description that corresponds with an objective external reality to be discovered.

Now, I’m certainly not the first one to point out all of the above. David Hume did it in his An Enquiry Concerning Human Understanding, published in 1739:

“If we take in our hand any volume; of divinity or school metaphysics, for instance; let us ask, Does it contain any abstract reasoning concerning quantity or number? No. Does it contain any experimental reasoning concerning matter of fact and existence? No. Commit it then to the flames: for it can contain nothing but sophistry and illusion.”

He was reacting to the increasing futility of Scholastic metaphysics, but he may just as well be responding to modern analytic metaphysicians. More recently, W.V.O. Quine argued that philosophy is continuous with (which doesn’t mean subsumed by!) science and the methods of science. One way to cash out that notion is to say that one’s metaphysics ought to never wander too far from one’s epistemology: if we are making statements about what does and does not exist, then we better be prepared to back up those statements with evidence. Such evidence — after Descartes — cannot be simply in the form of a logically coherent argument, because there are literally infinitely logically coherent accounts of how the world might work, and yet, we think, there is only one actual world. That means that we have to use empirical evidence to bring down the number of accounts to a manageable set. Ideally, a set of one, just like scientists try to do when they do science in a philosophically sound way. (Which is not the case for those scientists that indulge in metaphysical speculation about multiverses and such.)

Incidentally, whenever I have tried to make this point on social media of late I have been accused of simply adopting a particular metaphysical position — usually referred to as physicalism, sometimes as reductionism — without providing empirical evidence for such position. Gotcha, Prof. Pigliucci! But this is an elementary case of confusing metaphysics with epistemology. Regardless of whether you subscribe to physicalism, dualism, idealism, or any other kind of metaphysical “ism,” it is still perfectly reasonable to ask you what sort of evidence you bring to bear on your specific claims, such as “persons really are X,” or “the multiverse exists.” My epistemological stance, again, is that if your claim is about how the world works (as distinct, for instance, from prescriptive statements in ethics, or from mathematical and logical theorems), then you better have some kind of empirical evidence to back it up. I’m open to better ways of supporting factual claims, but my critics are curiously silent on that count.

Despite Hume and Quine (not exactly minor names in the history of philosophy), some people haven’t learned Descartes’ lesson. One of them is Philip Goff, with whom I recently engaged in an 8-letter correspondence on the notion of panpsychism, the idea that consciousness is an elemental property of matter. Physicist Sabine Hossenfelder has pointed out that panpsychism is empirically incompatible with modern physics. You would think that would be the end of it. When I pressed Goff on this, he gingerly admitted that panpsychism is not empirically testable. Indeed, it does not make any contact with the empirical at all. Then what on earth are we talking about? This is, I think, precisely the sort of stuff that gives philosophy a bad name, especially among scientists.

What’s the alternative? Something called scientific, or naturalized, metaphysics, a project articulated by philosophers like James Ladyman and Don Ross, for instance in their book, Every Thing Must Go, as well as in a collection of essays aptly entitled Scientific Metaphysics. The basic notion takes its inspiration from the above mentioned general trend in philosophy: when field X spins off (usually because empirical approaches have become available, so it turns into a science), then we see the onset of a philosophy of X. The general project of a naturalized metaphysics is to make sense of the disparate pictures of the world emerging from the various sciences, a job that no individual science can do, and for which most scientists are not adequately prepared — both because they tend not to think philosophically, and because they are too busy getting funded to run experiments. I highly recommend Every Thing Must Go, a tour de force of modern fundamental physics and a splendid example of how to do naturalized metaphysics.

More generally, Wilfrid Sellars has argued already decades ago that something along the lines of what I’m proposing here could really be construed as the overall task of (modern) philosophy itself. As he put it, philosophy is in the business of reconciling what he called the scientific and the manifest images of the world. The scientific image is, of course, what comes out of the best science available. The manifest image is how we normally perceive and talk about things. 

It is a no starter to try to reduce the scientific to the manifest image: science has discovered countless aspects of the world that are simply beyond our normal senses, and when we try to use everyday concepts to understand what science tells us we run into trouble. For instance, I think much common misconceptions about quantum mechanics are due to the fact that people keep trying to wrap their human minds around concepts that do not admit of descriptions in everyday terms. There is, for example, not contradiction or mystery in saying that light behaves both as a wave and as a particle. Photons are quantum level entities, which behave in no way that is directly reducible to our everyday words “wave” and “particle.”

One might be tempted, as many people are these days (especially scientists), to go the other way and reduce the manifest to the scientific image. Even some philosophers got onto this path, for instance Patricia and Paul Churchland (who, ironically, were students of Sellars). But this won’t do either. Not only it is cumbersome to use, say, descriptions from neuroscience or fundamental physics for everyday objects and concepts, there are some important notions that play a fundamental role in human life and that do not have equivalents in science: specifically anything to do with prescriptive language, values, moral oughts, and the like. In short, anything to do with rules and normativity.

Sellars’ vision, then, was one of philosophy working to develop what he called a “stereoscopic” image of the world, bringing together the scientific and the manifest images. After all, science itself is a human cultural construction, it serves human needs, and it is constrained by the limits of human understanding. Philosophy — when done properly — is uniquely suited to further that understanding. Now that’s a program of inquiry I can fully get behind as both a scientist and a philosopher!

By becoming a patron, you'll instantly unlock access to 17 exclusive posts
By becoming a patron, you'll instantly unlock access to 17 exclusive posts