I've been studying physicalism for a presentation I'll be doing on Shelly Kagan's book Death. One of the slides is on its problems, and one of those problems is that we don't have a clear definition of what the physical is.

In the Stanford Online Encyclopedia of Philosophy Hempel's Dilema is summarized as:

  • If physicalism is defined via reference to contemporary physics, then it is false.

  • If physicalism is defined via reference to a future or ideal physics, then it is trivial.

After some thought it came to me that a good definition of the physical, fitting into the latter category, would be that they are entities which are necessary for a maximally complete physics. Maximally complete meaning as predictive as possible.

At first sight it seems that it does no work for the reason that Hempel gives, that we can't say anything about a complete physics. In this case meaning that we can't know that a future physics would posit entities. On the contrary, not only has is it the case that physics has historically relied on entities for it's theories, but I think I might have an argument for the statement that it necessarily relies on them.

I'm not sure how this will hold up, but here's my attempt. Physics, as science, theorizes and judges the validity of its theories with methods. Then it is necessary that at the very least the method itself would be one of the entities in a theory of physics.

  • 1
    Unfortunately, this "definition" is trivial because it is vacuous. We want to know what it is possible to predict, not to rename we-know-not-what into "physical".
    – Conifold
    Feb 23, 2020 at 10:59
  • This definition does not let us predict anything, it's not a scientific theory. The kind of philosophy I am concerned with here is the investigation of the conditions of thinking anything as being physical. Conjoined with a suitable definition of physics it may be possible to deduce truths having to do with these objects, but at no point do I expect to be able to point to something and tell you "Hey, that's a physical object!" Feb 24, 2020 at 18:14
  • How would you know that a theory was "as predictive as possible?" What would it mean for one theory to be "more predictive" than another?
    – puppetsock
    Feb 24, 2020 at 18:18
  • @puppetsock One way to answer that, though currently impractical, would be to simultaneously measure and simulate the observable universe for a period of time. Then take the root-mean-square of the difference between the numbers found in the simulation, and those measured. The larger the number, the less predictive. Feb 24, 2020 at 18:28
  • @BrendonEspinoza Um... That leaves a situation where you are requiring something that is impossible. Certainly impossible now, and I would suggest impossible in principle. So, at least right now, there is no way to decide between potential theories. That's quite a serious problem. Maybe you need to reformulate?
    – puppetsock
    Feb 25, 2020 at 14:28

1 Answer 1


Your question presumes a variety of framing assumptions which are now generally held not to be true by most philosophers.

First, you are presuming that "completeness of physics" would mean that "physics is everything", IE a reduction of all thought to physics, including the philosophy of science. But reductionism is no longer considered viable in most science, and certainly not OUTSIDE science, in philosophy of science! https://plato.stanford.edu/entries/scientific-reduction/

This point was very effectively demonstrated when the Logical Positivists tried to subsume philosophy into science using the Verification Principle. The VP itself is meaningless per its own terms, https://www.newworldencyclopedia.org/entry/Verifiability_principle The point of this failure is that science is part of a suite of thought disciples -- IE pluralism of valid modes of thinking applies not just within the sciences, but across all fields of thought. And notably, philosophy in the form of epistemology and logic, are both preconditions for science to even operate. Therefore physics cannot ever be logically closed.

Additionally, a dominant movement in philosophy is analyticity -- the premier role of logic in philosophic thinking. Analytics depend on precise definitions, hence the problem you face with the definition of physics, and its fuzzy boundary where science is still being done. Quine, in Two Dogmas of Empiricism, showed that NO language will ever be precise enough to support analyticity https://www.theologie.uzh.ch/dam/jcr:ffffffff-fbd6-1538-0000-000070cf64bc/Quine51.pdf by inference, the very fuzzy nature of the boundaries of physics (or of science) is even less precise than language, so your effort to precisely define the terms of physicalism in terms of physics is doomed to failure. For reference, this book explored the definitional approach to physicalism, and found it to be fruitless: https://www.amazon.com/Physicalism-Problems-Philosophy-Daniel-Stoljar/dp/0415452635/ref=cm_cr_srp_d_product_top?ie=UTF8 (Notably, Stoljar concluded that physicalism was actually refuted by the peculiarities of modern physics).

Hempel's Dilemma only applies to this definitional/analytic/rational approach to physicalism. A potentially more fruitful approach is to treat physicalism as an empirical "hypothesis". In empiricism, definitions are not as critical -- they can be adjusted on the fly as one does further investigations. An excellent generalization of how to do empiricism is found in the Research Programme language of Imre Lakatos. https://antimatter.ie/2011/02/11/kuhn-vs-popper-the-philosophy-of-lakatos/ And one defender of physicalism has cast it explicitly in Lakatosian terms: https://core.ac.uk/download/pdf/74162.pdf

Papineau's paper explains very clearly why physicalism became a dominant view in philosophy in the 20th century. It also implies why physicalism is struggling today. Hempel's Dilemma is but one minor problem. For physicalism to be a global worldview, everything has to reduce to physics, and reductionism has failed as a project, both in science, and across knowledge generally. When science relies upon philosophy (epistemology and logic) and on math, and neither of these fields reduce to physics -- the claim that physics is everything there is -- is pretty explicitly self-refuted! A plurality of philosophers hold by both abstract object realism, and moral realism, both of which are effectively incompatible with physicalism. Plus there is widespread consensus that the "hard problem of consciousness" has not been solved, and there are no current promising routes to solve it -- which leaves consciousness too as a major unaddressed subject outside physics. https://www.amazon.com/Physicalism-Something-Princeton-Monographs-Philosophy/dp/0691133859 Worse, most theoretical physicists actually appear to be IDEALISTS -- IE they consider matter to be a reflection of MATH. So even if reductionism were somehow salvageable -- reduction itself is now problematic for physicalism.

So, as an answer to your question, no, you cannot solve Hempel's Dilemma definitionally, nor can you include epistemology in this redefinition of physics. And the problems physicalism faces are much worse than just Hempel's Dilemma.

  • Thank you for your answer. Much of what you said will be helpful for the presentation, but I want to clarify some things before I can accept your answer. I was referring to a maximally complete physics, one which is as predictive as possible. I'll edit the question to reflect that. Feb 22, 2020 at 18:36
  • Also, I didn't try to define physics, I tried to use what I think in it as science to argue that all science must posit entities in it's theories. Feb 22, 2020 at 18:44
  • 1
    Well, this parsed to me like you tried to subsume the philosophy of science within physics by defining philosophy of science as an entity, that defines a methodology, and if physics is the science of the behavior and existence of entities, then Phil of Sci would then be part of physics. What the failures of reductionism shows is that what is or isn't an entity, AND the appropirate methodologies of empiricism, AND the appropriate logics to use -- are all outside physics. Therefore physics is not the study of all entities and the methodologies of their study, and cannot be.
    – Dcleve
    Feb 22, 2020 at 19:03
  • First, the method would be the entity, not that which produced it. So philosophy of science would not have to be an entity. Second, I didn't try to define physics, but I see you came up with a definition for me. I believe your definition of physics doesn't hold up very well because philosophically educated physicists tend to be wary of believing in the entities which their theories posit. My favorite definition is that physics is the study of matter, matter being the sensory content of experience. Feb 22, 2020 at 19:36

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .