Skip to content

The Relationship between Physics and Philosophy

April 30, 2012

The comments in the An Explanation from Nothing? post by Sean Carroll (linked here to his blog) and Lee Smolin are very helpful.  They move the discussion  in a good direction. So let’s put an end to the kerfuffle (or brouhaha)  over whether David’s review of Krauss’ book was fair and whether Krauss response was inappropriate.

The issue was raised by Carroll and Smolin and Krauss himself  of exactly what the relationship between physics and philosophy of physics can be and should be. Obviously, there are some questions and problems that properly belong to one field but not the other. For example, physicists make proposals about what laws and chances there are while philosophers of physics are interested in what laws and chances are.  Physicists produce explanations and argue that one theory is better supported by evidence  than another without having explicit accounts of explanation or support. These jobs properly belong to philosophy.

However,  there are some issues and problems where collaboration seems the way to go. I (and I hope others) would be interested in some discussion of issues in the history of philosophy/physics where collaboration did pay off (even if is just Einstein collaborating with himself!) and some current issues where it looks like collaboration may lead to progress. My own two cents is that the discussion of non- locality (exactly what are the consequences of nature failing to satisfy Bell’s inequality) has been advanced by work that involved both physicists and philosophers and interactions between them though there is still a lot of confusion on this topic. (One can still find people saying that what Bell showed is that hidden variable theories are impossible because they are non-local).

Here is another question involving quantum mechanics where some collaboration may be useful. Sean in the post in Cosmic Variance  mentions that in quantum mechanics the states of the universe are “wave functions.”   I am not sure whether he means  by “wave functions”  mathematical entities or whatever concrete things or stuff satisfy a certain mathematical description. My question concerns the second construal; In other words,  what is the ontology of quantum mechanics?

It would be really good to see more collaboration between physicists and philosophers on this issue.   Anyway, the blog is open for successful past and current collaborations (and also unsuccessful ones).

8 Comments leave one →
  1. cormac permalink
    May 1, 2012 6:48 am

    Excellent post. If I may add a suggestion, it’s also worth exploring whether there have been incidences in the history of physics where an adherence to a particular philosophical viewpoint has hindered progress (or not). For example, Einstein’s great interest in the philosophy of Mach was hugely important, yet Mach’s trenchant opposition to atomism on philosophical grounds cannot be denied, nor indeed his opposition to relativity (which became clear after his death, much to Einstein’s dismay).

  2. tmaudlin permalink
    May 2, 2012 1:20 am

    It is true that there are some particular questions that fall more into the domain of philosophy (e.g. accounts of the nature of scientific practice) and others that are firmly in the domain of physics (e.g. methods of calculating scattering cross-sections), but for the particular sorts of questions we are largely interested in here, I can see no way to assign the topic to “physics” or “philosophy”. Take the case of the “nature of the wavefunction”, for example. In one sense a “wavefunction” (as its name implies) is a mathematical object—e.g. a complex function on another mathematical object called “configuration space”—that is employed in physics as a representation of a physical system. That immediately raises many questions. One, which Einstein, Podolosky and Rosen famously raised, is whether that particular mathematical representation is complete. That is, does it explicitly or implicitly represent all of the physical features, the values of all of the physical degrees of freedom, of the system. Can two systems represented by the same wavefunction nonetheless be physically different in some respect? Can the same system be properly represented by two different wavefunctions? Are there mathematical degrees of freedom in the representation that do not correspond to physical degrees of freedom in the system itself? (As an example of the latter, should the physical state of the system correspond to vector or a ray in a Hilbert space? If a ray, then it is misleading to say that the physical state is represented by a vector in the space: the vector has mathematical properties that do not correspond to physical properties of the system.) If one asks whether this sort of question is one of philosophy or of physics, I (and Einstein) would say it is a matter of physics. Indeed, it seems to be an absolutely essential question if one is to understand the physical account of the world being provided by the mathematics. But it is a sociological fact that while it is perfectly acceptable and even expected to discuss questions like this in a philosophy department, or a philosophy course, it can be rare, or even frowned upon, to discuss them in a physics department or physics course. We all know the phrase “shut up and calculate”. This phrase was not invented by philosophers, and in my experience physics students immediately recognize what it describes in their physics courses. Steven Weinberg tells the cautionary tale of the promising physics student whose career was ruined because “He tried to understand quantum mechanics”. The point of the story is that the physicist, as physicist, should not try to have a clear, exact understanding of the physical meaning of the mathematical formalism. But certainly this ought to be a question in the domain of physics! It is just a weird sociological fact that, since the advent of quantum theory and the objections to that theory brought most forcefully by Einstein, Schrödinger, and later Bell, a standard physics education does not address these fundamental questions and many physics students are actively dissuaded from asking them. But anyone with a philosophical temperament cannot resist asking them. So, for better or worse, discussing these questions is more universally recognized as an important and legitimate task in philosophy departments than in physics departments (even Bell characterized his seminal work in foundations as secondary to his “real” physics work at CERN). And the habits of mind—a certain sort of precision about concepts and arguments—that are needed to pursue these questions happen to be exactly those habits instilled by a good education in philosophy. So while Krauss and Hawking lament that many philosophers don’t know enough physics (which is true), it is equally the case that many physicists are sloppy thinkers when it comes to foundational matters. I’m not sure that collaboration is the proper model here—as the “example” of Einstein collaborating with himself suggests!—but rather an appreciation of both which details of the physics are important and where the physics is simply not clear and precise as physics. Learning sophisticated mathematics, which a a large part of a physics education, does nothing to instill appreciation of the sort of conceptual and argumentative clarity needed to tackle these foundational issues.

    Let me give a quick example. Take the “vacuum state” in quantum field theory. (No, I’m not raising the question of whether it is “nothing”!) It is commonly said that the vacuum state is positively a buzzing hive of activity: pairs of “virtual particles” being created an annihilated all the time. But it is also commonly said that the quantum state of system is complete: to deny this is to posit “hidden variables”, and those are not regarded by most physicists with favor. But it is clear that these two claims contradict each other. Consider a system in the vacuum state over some period of time. Over that period of time, the quantum state is static: it is always the same. So if the quantum state is complete, nothing physical in the system can be changing. But the “buzzing hive” of virtual particles is presented as constantly changing: particle pairs are being created and destroyed all the time. So which is it? This strikes me as a straightforward question of physics. But it is more likely to be asked, I think, by a philosopher. And insofar as the observable predictions of the theory do not depend directly on the answer, it is even likely to be dismissed by the physicist as “merely philosophical”. But without an answer, we really have no understanding of the vacuum state, or the status of “virtual particles”.

    • May 12, 2012 3:12 pm


      The “virtual particle” status issue raised in your last paragraph could be a poster child for the need for greater collaboration between the physical and philosophical modes of analysis. Virtual particles currently occupy an important, but ontologically self-contradictory position in physics. Important because they enter in an essential way into all perturbative solutions in quantum field theory, as well as in the analysis of static force fields, near-field phenomena, and others. The confusion regarding the reality of virtual particles is multifold. A virtual particle exists within limited time and space intervals. But, if a virtual particle is detected (i.e., interacts with an “observer”), then its existence is thereby prolonged and it becomes a real, rather than virtual particle. The propagators in Feynman diagrams are considered to be virtual particles because they are never themselves separately detected. But when the same types of propagators (such as photons) are detected, they are considered to be real. Similarly, virtual particles in the vacuum are not considered real due to their non-detection, even though real phenomena such as the Casimir effect can result from them.

      Relativity has already been proven to be a very powerful philosophical concept, and perhaps deserves much more attention from the philosophical community interested in the ontology of QM objects and concepts. The relativistic interpretation of QM (RQM) discussed in my other comment here can perhaps resolve the question of the ontology of virtual particles, including in the “vacuum state”. In RQM, even a single particle is considered to be a QM system that potentially can play the same observer role as a more complex system in establishing the states of other QM objects. Thus, in QFT, where a virtual particle carries the interaction between two real particles, the virtual particle is very much real relative to the absorbing particle, which in this case acts as an observer of the state of the virtual particle. Similarly, in vacuum pair production, when a virtual particle is annihilated, it is observed by the annihilating particle. Where virtual particles account for static fields or near-field effects, the particles are observed by the other systems affected by the fields or effects. In RQM, any QM object serves as a potential observer, but the state of all systems exist only relative to other systems, and there are no absolute system states. Thus, in this view, the particles involved in all of these phenomena are considered to be real, not virtual, but the notion of reality has been loosened. For example, there is no unique vacuum state, since the state of the vacuum is only defined relative to its potential observing systems, including any particles that might be created and annihilated. Similarly, the RQM interpretation clearly explains the EPR paradox by viewing all interactions as strictly local, but more loosely defining reality (in terms of the state of an object) as a relative property. This all seems like very fertile ground for philosophical thought.

  3. May 3, 2012 12:45 pm

    This blog subject pushes us into more interesting territories. Both physics and philosophy must enter into any worthwhile discussion of the interpretation of QM, particularly where issues related to cosmology are concerned. Without an improved understanding of the ontology of QM, it is difficult to imagine that some of the more poorly understood aspects of cosmology will be satisfactorily addressed. The status of the quantum state of the universe as mathematics or “stuff” questioned above by Barry Loewer and the issue of the nature of the vacuum state raised by Tim Maudlin are two such examples. Another is the apparent non-conservation of energy for the photons in the cosmic microwave background (CMB), which seems to have no widely accepted explanation at this point.

    But new suggestions regarding QM interpretation continue to be advanced, providing different and perhaps more compelling insights into these issues and others. One worth mentioning is the “relational” interpretation (RQM) originally put forth by Carlo Rovelli, a physicist that has also worked and published in philosophy. In this approach, QM strictly represents the description of physical systems relative to other systems, where all systems are quantum mechanical, much in the philosophical spirit of relativity theory. RQM has already been used to convincingly address many of the more general interpretive issues regarding QM. For example, it provides a good explanation of the results of experiments related to Bell’s inequality.

    Regarding the above mentioned cosmology issues, in RQM the wave equation for an observed system changes for each observer, so any wave must be purely mathematical, not physical (i.e., “stuff”). Moreover, RQM is actually inconsistent with any concept of a state of the universe. Every state in RQM defines the relationship between two physical objects, the system and the observer. Defining a state for the entire universe would thus require a second physical object outside the universe, which is a contradiction. As for the apparent non-conservation of energy for CMB photons, in RQM the photon is not considered to have a unique QM state unless it is observed. (More striking, one can perhaps ask in what sense a CMB photon even exists before it is observed.) Once the photon has been detected, its red shift can be viewed as a purely relativistic effect produced by the difference in motion between the rest frames of the emitting particle and the detector, and is thus energy conserving. Such an approach has been pursued by Alasdair Macleod with some success. Thus, while interesting in its own right, RQM also serves as a good example of the synergy between philosophical and physical reasoning in advancing our understanding of the universe.

  4. Dmitri Tymoczko permalink
    May 4, 2012 10:10 am

    A few thoughts from a nonphilosopher who went to philosophy grad. school for a few years, and who has worked through a few field theory textbooks.

    1) My sense is that physicists and philosophers have a hard time collaborating because they begin in fundamentally different places: physicists with standard-issue relativistic quantum field theory, philosophers with elementary quantum mechanics (or maybe in some cases algebraic QFT). And that’s not to mention all the many metaphysicians who seem to assume a roughly Newtonian world (“imagine all the electrons in my body were replaced with the electrons in your body”). In my experience, this sort of thing makes working physicists distrust philosophers in general.
    For me it’s easy to see what philosophers often lack, and it’s something like: a working knowledge of the content of field theory textbooks (along with the associated math). When I think about what physicists lack, it’s something more like intellectual patience; in many cases they’re content to say something like, “that’s just not an interesting question” or ” we just don’t know the answer to this question right now.”
    Come to think of it, I remember once talking to a famous string theorist about John Norton’s “Dome.” He genuinely thought it was “sad” (his word) that someone would “waste their time” thinking about this issue when there were so many more interesting things to do (e.g. calculate the entropy of a black hole from first principles). I’m sure he would’ve felt the same about the “Hole argument.” It wasn’t that he lacked any knowledge of the underlying physics, but he did have a very different sense of what was an interesting question.
    (I say this not to take sides, but just to point to a difference in intellectual values that I think could be important.)

    2) I think the development of QFT in the 1960s provides some interesting examples of the sort of interplay you’re looking for, though the “philosophy” comes mostly from the physicists. For instance, the proposal that QFT was an effective, low energy theory (rather than a complete theory valid to all energies) is as much a philosophical as physical proposal; it doesn’t change any equations, but allows you to interpret the process of renormalization (and the appearance of infinities) very differently. The philosophical idea here is something like fallibilism, or that our current best theories are merely provisional.
    Another possible example is the failure of S-matrix theory and the development of nonperturbative methods in field theory, which arguably involve different attitudes toward the reality of the field itself (as opposed to the perturbative Feynman rules).

    3) My own sense is that physicists are very open to speculative ideas that I would consider philosophical: for instance, the idea (not provable at present) that local gauge invariance explains why we have the photon field, or the laws of electromagnetism; or the idea that we have the particles we do because of some SO(10) GUT group, or the shape of some extra dimensions. These are grand speculative ideas about how the world might be.
    But somehow it seems like this is not what you’re looking for — perhaps the sense of “philosophy” that’s relevant to your question is not “grand speculative ideas” but something more like “the careful drawing of distinctions taught in philosophy grad. school.”

    • tmaudlin permalink
      May 4, 2012 7:49 pm

      I understand the appeal of this sort of analysis of the situation, but it is just factually inaccurate. Off the top of my head, here are some philosophers of physics that know relativistic quantum field theory perfectly well: David Albert, David Wallace, Paul Teller, David Malament, John Earman, Laura Ruetsche, Gordon Belot, Hans Halvorsen, Wayne Myrvold, Frank Arntzenius, Michael Dickson, Richard Healey, Jeff Bub, Doreen Fraser. There are many more. Relativistic quantum field theory raises some new conceptual questions, such as renormalization, but it does not present any new resources if one is interested in the measurement problem, the status of the wavefunction, the implications of Bell’s theorem, etc. The measurement problem, for example, is derives from the claim that the wavefunction always obeys a linear equation. Whether that equation is the Dirac equation or Schrödinger’s equation is neither here nor there. These problems can be more simply can directly presented using the resources of non-relativistic quantum mechanics, and the additional complications of using field theory would just get in the way of understanding things.

      Nor are the questions that interest philosophers the product of philosophy grad school. The best, most searching discussions of the problems I just mentioned are due to Einstein, Schrödinger and Bell. When Bell wrote “Against ‘Measurement'” he did not talk about field theory, but not because he was unfamiliar with it or the mathematics it uses. He did not talk about because the way it differs from non-relativistic quantum mechanics is simply not at all relevant to the problem. Plain old non-relativistic quantum mechanics also suffices for the EPR argument, and Schrödinger’s cat argument, and to show the violation of Bell’s inequality. The questions raised by these arguments are the ones that interest philosophers, and they are not questions that are produced by arcane philosophical distinctions. They are, as I have said, directly questions of understanding physics as physics.

      In his famous 1963 Scientific American article, Dirac divides the problems confronting the understanding of quantum theory into the Class One and Class Two problems. The Class One problems include the ones I mention above (except, of course, Bell’s theorem), and the Class Two problems include technical issues about renormalization and so on. Dirac appreciated the Class One problems, and saw that they are problems. He just did not see how to make progress on them. On the other hand, it was more obvious how to work on the Class Two problems. One can even do so in the hope that the solution to the Class Two problems might shed light on the Class One problems, but there is really not much in the way of good argument to expect this fortuitous result. No doubt philosophers tend to focus more on the Class One problems than the Class Two problems, but not because they are “philosophical” rather than “physical” problems. Maybe it is because as philosophers we are more used to working on problems—e.g. the Liar paradox— that have resisted solutions for millenia, and we are not so frustrated not to get quick results. But the main thing is not to lose sight of the Class One problems, even if they resist solution. For there is physics—indeed, the most important physics— in the solution.

      • Dmitri Tymoczko permalink
        May 5, 2012 9:37 am

        Hi Tim, thanks for writing!

        My sense is that at least some of the figures you mentioned are more involved with Algebraic QFT than with the standard QFT that physicists use. (Wallace, who does seem to write and think like a physicist, emphasizes the importance of the distinction.) Others, while surely familiar with QFT, manage to convey a fairly idiosyncratic understanding of the subject. For instance, in his recent NYT review, Albert seems to have a realistic interpretation according to which a quantum field is a kind of “stuff” over and above its particles; I found this odd, and struggle to understand how someone who knew, say, Weinberg’s QFT textbook, could just assume this. (He also seems not to realize that many QFT texts propose local gauge invariance as a reason *why* we have the photon field.) So I’d probably distinguish “familiar with QFT” from something like “having deeply internalized the working physicist’s perspective on the subject.”

        I do like your point that physicists want to get tangible results and to make progress, while philosophers are accustomed to dealing with questions that take a long time to get resolved. I take that to be part of this important difference in intellectual values that sometimes prevents conversations across disciplines. A physicist who says “don’t try to understand QM” may simply be saying “you’re unlikely to make progress, or to get tangible results, by doing this” and that’s can be a pragmatic statement, rather than a judgment about whether the question is interesting.

        I didn’t quite understand whether you brought up the class one/class two distinction as a response to my point about effective field theory — are you saying that the effective field theory paradigm isn’t philosophical in the sense that you are interested in? Because if only class one problems are truly philosophical then that does limit the possibility of interaction even further …

        • tmaudlin permalink
          May 6, 2012 6:39 pm

          Quick points. Of course David was discussing an understanding of QFT in which particles (or better, particle-like behavior) is only emergent and the basic ontology is a field ontology…because that’s just how almost every physicist would describe the situation. If you think there is an alternative ontology accepted by the “average physicist”, please try to state clearly what it is. David is (of course) perfectly aware of issues of gauge invariance: you really ought to expect someone with a Ph.D. in theoretical physics to be familiar with all of these topics. If you think they are somehow relevant, then explain exactly how, but suggesting that David “seems not to realize” what is in standard QFT texts is again just factually false. Standard texts happen not to be very good sources for clear accounts of the fundamental issues we are discussing, in any case.

          It is true that some philosophers of physics show more interest in axiomatic QFT than is shown by practicing philosophers. They also show more interest in “hidden variables” theories (e.g. Bohmian mechanics) and in explicit and clear collapse theories (e.g. GRW) and more interest in getting really clear about just what Many Worlds theory claims. It is notable that Bell was a major expositor and advocate of both Bohm and GRW, and I think it is obvious that philosophers are interested in them because they are clear and exact and solve the measurement problem. The preference for axiomatic QFT feels to me different, as arising for a demand for a certain sort of mathematical rigor. That may arise from studying logic, and indeed be traced to a philosophy background.

          I brought up Dirac for a completely different reason: to show that he appreciated the importance of the Class One problems, even though he preferred to work of the Class Two problems because he thought he could make more progress on them. I want to contrast that with anyone who wants to dismiss the Class One problems as “just philosophy”.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: