Perhaps it is a stereotype, but I assumed that most physicists are empirical realists (external reality affects our senses, and science infers a representation of it from sensory data). At the same time, the Copenhagen interpretation is the majority interpretation of quantum mechanics. These two positions seem to be in conflict with each other.

Bohr's original explanation of the uncertainty principle, consistent with empirical realism, was that the measurement apparatus disrupts the measurement precluding simultaneous determination of positions and momenta with precision. Unfortunately, the same disruption occurs in classical mechanics without precluding assignment of positions and momenta, or even modeling the disruption itself and correcting for it. Einstein designed an experiment (EPR) which demonstrates that similar assignment of "local elements of reality" does not work in quantum mechanics and declared it "incomplete". Bohr eventually abandoned his early view.

More generally, the famous collapse of the wave function occurs simultaneously all over the universe (what Einstein called "spooky action at a distance"), so treating it as a physical process contradicts relativity. Bohr's mature position, which became Copenhagen, was that the wave function describes maximally possible "state of knowledge" about quantum objects, and the collapse amounts to re-conditioning of probabilities, a purely mathematical adjustment with no physical content. But this position seems to be outright Kantian, we have a wave of knowledge that describes phenomena only, and of noumena thou shalt not speak. Moreover, according to von Neumann's no-go theorem quantum mechanics is not just incomplete but incompletable, it is inconsistent with any "local elements of reality" whatsoever (not just positions and momenta). So the noumena, if they exist at all, must exist outside of space and time, making this even more Kantian.

David Bohm came up with a non-local completion that gives an idea of what these noumena might look like, "quantum potential" and "implicate order". But Bohmian mechanics is a small minority position, and the implicate order in its own right would be hard for an empirical realist to swallow. In Kantian terms, it is an illegitimate "metaphysical speculation", a misapplication of the categories.

Are most physicists really Kantians, or is there a confusion as to what the Copenhagen interpretation actually says (e.g. taking it for Bohr's early position)? Or perhaps there is some other philosophy that reconciles external reality with Copenhagen epistemology?

  • 2
    The collapse of the wave function link in this question provides an indication of more modern interpretations of QM. It's my impression that physicists who actually work in this area have dropped the CI in favour of working on the entanglement of quantum systems with macroscopic ones (and physicists who don't work in the foundations of QM, for the most part, don't worry about these problems). At the time of its development, QM generally, and the CI specifically, was strongly influenced by logical positivism (Bohr in particular if I remember correctly).
    – Dave
    Commented Nov 19, 2014 at 20:43
  • @Dave But there is a position of physical community as a whole, not just narrow experts, and Wikipedia claims "there is widespread but not unanimous acceptance of the interpretation". Second, it's true that many experts moved on to decoherence and many worlds, but that only takes care of indeterminism, I don't see how it resolves the realism issue in a particular world. Objective collapse theories would do that, but they were broadly rejected for multiple reasons. Also, there is a direct "line of succession" from Kant to neo-Kantians to log positivists, which supports that CI is Kantian.
    – Conifold
    Commented Nov 20, 2014 at 0:38
  • 6
    My observation is that there is widespread acceptance that the CI is widely accepted, if you see the distinction I'm making. I suspect that in fact most physicists personally prefer one variant or another of the Everett interpretation, to the extent that they worry about it at all. Commented Nov 20, 2014 at 1:06
  • Maybe my education experience is atypical, but it had a decidedly "just calculate and don't worry about the meaning" flavor to it, e.g. my QM text (R. Shankar, 1980) doesn't mention anything about interpretation at all. It's not clear that the CI is "widespread", c.f. arxiv.org/abs/quant-ph/9709032 where if you read the paper you see that "undecided" has the plurality (the sample is arguably from "narrow experts" though).
    – Dave
    Commented Nov 20, 2014 at 13:11
  • 2
    @Harry Johnston "Widespread acceptance of widespread acceptance", that's a neat way to put it. It seems that Bohr succeeded at creating a default position cryptic enough for people to either shrug at or read their favorite meaning into and say "close enough".
    – Conifold
    Commented Nov 20, 2014 at 23:30

8 Answers 8


The best expression of how quantum mechanics conflicts with realism is the so-called measurement problem. Assuming scientific realism (that the theory describes an independent reality), one faces the following trilemma:

  • either the wave-function is an incomplete description (Bohm's theory)
  • either its evolution is non-linear (collapse theories)
  • or measurement outcomes are not definite, but superposed (many-world interpretation).

It is not very clear where to situate copenhagen's interpretation, because of its vagueness. One option is to view it as a collapse theory, but it does not specify when, where and how the collapse occurs (only that it is related to measurement).

Another option is to view it as a realist theory with regards to classical states and an instrumentalism with regards to quantum theory ( which indeed has kantian flavours). The idea is this: quantum theory is a tool for predicting classically described outcomes, given classicaly described initial state, but it is not in any sense a description of reality. Bohr emphasised the fact that you need a "classical" vocabulary to interpret a wave-function and the observables. It has been claimed that it assumes an arbitrary microscopic-macroscopic dichotomy (and a collpase) but this is probably a realist misreading of Bohr's view.

Note that some of the physicists who elaborate quantum theory participate to the Vienna circle (logical positivist philosophy group) which was more or less influenced by neo-kantian philosophy. Logical empiricist's attempts to reduce theoretical knowledge to an observation vocabulary (phenomenism, operationalism) is clearly anti-realist.

As for the philosphy of contemporary physicists: I think there are various views, some being instrumentalists, others assuming many-worlds or a somehow vague collapse interpretation, and many being pragmatists without an explicit philosophy.

  • 3
    Nice breakdown. I almost forgot how much Bohr emphasizes the uninterpretability of quantum reality to classical observers. He is probably right, the appeal of "reality building" is based on coming up with something that can be understood by direct analogy with what is directly perceived. As we move away from familiar scales such analogies increasingly break down. So theories either have to embrace phenomenology like CI, or unobservable and non-unique speculations like Bohm's or Everett's. Neither can meet the classical "external reality" ideal.
    – Conifold
    Commented Nov 20, 2014 at 23:23
  • " One option is to view it as a collapse theory, but it does not specify when, where and how the collapse occurs." The intrinsic nature of the uncertainly makes it impossible to predict such things Commented Apr 11, 2019 at 22:53

The philosophy here surely doesn't obligate one to be a Kantian, because we are talking about a model here. Whatever lies behind the model, we realize that we have reduced it to terms we can handle. There is no presumption of a deeper reality, just of a lack of perfect fit between the model and things we have not managed to observe.

I do believe that it stretches the notion of materialism past its breaking point, but not realism, or even 'physicalism' to the degree that we make no presumptions upon what is and what is not physical without experimental evidence. The spirit of realistic reductivist monism remains, and we are not forced into accepting idealism by the fact that we treat mathematics as ideal.

Everyone in all the sciences has always done so, assuming the issues of reducing idealist mathematics to something consistent and realistic will be resolved inside mathematics. Acting as though mathematics is an ideal domain does not mean you believe it is one, only that its issues are outside your purview. For instance using continuous space does not imply real acceptance of the actual infinity of points, only acceptance that it is a good model, not worthy of breaking down farther.

The observation about fitting with relativity is unfair. We have no idea whether or not relativity and quantum mechanics are actually compatible, but the idealization that the collapse of the wave function occurs simultaneously is just that, and idealization. We have not yet come up with any way of testing whether this is information that travels at the speed of light, or slower, because we have not manage to find an unobserved frame large enough. There is a good reason to think we can't, in that the odds of making or sustaining one become too small quite quickly.

Edit in response to @Confold:

If physics is truly required to explain and not just to model, it does seem true that the Copenhagen Interpretation either requires a belief in the "ultimately un-analyzable cause" of the wave function, or a belief in some magical quality of intelligence. You can have Kant or magic. In practice, I think more people choose magic: they accept some version of 'observer' indirectly tied to parsing of the information by some intelligence. That intelligence still does not have to be ideal, just macroscopic and hugely more complex than individual quantum phenomena, like us.

The easiest form of magic is to say "magic is simply technology we have not yet explained" and stop worrying about it. Yes, we can just believe, with Einstein that the wave equations are incomplete, rather than inexplicable. "The magic of intelligence will eventually give us a more complete framing. Just wait. in the meantime, do some physics."

Everett's "magical quality of intelligence" is "selection of world".

Mine would be as final arbiter of the arrow of time. I would dodge the objectivity of collapse completely with reversible time and the weak anthropic principle. Due to CPT symmetry, time is free to reverse itself during periods when its observations will never affect humans, who are, in effect the cause of time. Yes, that sounds even more Kantian. But if you assume our own shared reality is only an aspect of the fact brain chemistry is exothermic, there is still a way out of the collapse dilemma. It can collapse or not collapse, and only along timelines we eventually care about, can we even find out.

I see the former as a perfectly acceptable answer, and the latter two as further specification of, rather than as alternatives to, the Copenhagen position.

  • I agree that QM as a mathematical model does not compel one to be a Kantian, but it seems that CI does. In "realistic" models we have what "is", how it behaves during measurements, and what will be "measured" all represented. In phenomenological models we only have the last part without as you say a "deeper reality". But QM is a very peculiar phenomenological model, not only does it not have "deeper reality", it provably can not have it (on acceptable terms). And CI openly embraces that (Bohr characterized questions beyond the wave function as meaningless).
    – Conifold
    Commented Nov 20, 2014 at 0:23
  • The idea of "de-idealizing" the collapse was thoroughly explored in "objective collapse" theories, and they are frought with problems even aside from relativity. The "best" of them, GRW, had to re-adjust its parameters continuously and ad hoc to stay consistent with experiments. I think GRW has even fewer supporters than Bohm. en.wikipedia.org/wiki/…
    – Conifold
    Commented Nov 20, 2014 at 0:23
  • @Conifold (and see prior) Basically, Kant plays a huge role in giving us tools to even think about such things, but materialism really can invent ways to reassert itself, and then ways to test if those reassertions hold water, whereas real idealisms always seem to find themselves suspended in intellectual space without purchase, and incapable of moving forward, at some point. Saying Kant is the inspiration is different from saying he is the basis.
    – user9166
    Commented Nov 20, 2014 at 15:33
  • (The 'prior' went into the answer as an edit.)
    – user9166
    Commented Nov 20, 2014 at 16:02

But this position seems to be outright Kantian, we have a wave of knowledge that describes phenomena only, and of noumena thou shalt not speak. Moreover, according to von Neumann's no-go theorem quantum mechanics is not just incomplete but incompletable, it is inconsistent with any "local elements of reality" whatsoever (not just positions and momenta). So the noumena, if they exist at all, must exist outside of space and time, making this even more Kantian.

Unless nonlocal state is incoherent [at least: to our comprehension], this is false—I think all of it. We must remember how deeply the mechanical philosophy, atomism, and reductionism have infiltrated our Western consciousnesses. Quantum mechanics blew that synthesis to pieces, as is described by David Bohm:

    Indeed, when this interpretation is extended to field theories,[7] not only the inter-relationships of the parts, but also their very existence is seen to flow out of the law of the whole. There is therefore nothing left of the classical scheme, in which the whole is derived from pre-existent parts related in pre-determined ways. Rather, what we have is reminiscent of the relationship of whole and parts in an organism, in which each organ grows and sustains itself in a way that depends crucially on the whole. (Causality and Chance in Modern Physics, xi)

When one hears talk of a universal wavefunction, the idea is that any wavefunction other than the universal one, while it can be used in computation, is only ever an approximation. One cannot merely start with basic building blocks and construct a universe. Hegel had something right in his focus on needing to understand the whole to understand the parts. The more technically correct form is that the less of the whole you understand, the less complete your understanding will necessarily be. Samuel Taylor Coleridge may have something helpful to say as well:

Our minds will be forced, as Coleridge so often lamented, to "divide in order to distinguish."[27] There will be no real unity, no mutual penetration, among the items of the world, but only (literally) super-ficial associations. Such is the necessary legacy of solids, and such the enemy of transformation. (The Form of Transformed Vision, 56)

Coleridge was reacting against the mechanical philosophy; he saw it was insufficient for fully comprehending reality. Indeed, he saw it as sufficiently only for understanding appearances, to the extent that one can understand appearances without understanding what is underneath/inside them.

The Copenhagen interpretation refuses to say that the appearances are reality. This is wonderfully illustrated by René Magritte's famous painting, The Treachery of Images. Contrast this to Sean Carroll's Fluctuations in de Sitter Space, in which he argues that "the quantum state is the physical thing; there's no sort of hidden variable underneath" (18:13). One result of this assumption is that there is no true time-dependence in [mind-independent] reality; it only appears that way because we're in it. Coleridge certainly didn't know quantum mechanics, but his "enemy of transformation" could be seen as a dislike of that which would make time fictional, and according to Michael Tooley, make causation fictional (quotations here).

As you note, collapse of the wavefunction under the Copenhagen interpretation violates relativity, and we like relativity. So, we're highly tempted to either reject it as a physical event, or get rid of it, e.g. through decoherence. We desperately want to be talking about something real; instrumentalism is not a position many like to take. When it comes to violations of relativity this isn't surprising: superluminal causation really messes with one's ability to model reality, as it threatens the principle of locality. Nevertheless, rejecting the Copenhagen interpretation on these grounds is an iffy procedure; while many physicists are inclined towards the Many-Worlds interpretation, it has potentially serious problems.

Alasdair MacIntyre sheds some light on the situation in his 1977 The Monist article Epistemological Crises, Dramatic Narrative and the Philosophy Of Science:

Instrumentalism, like attempts to refute scepticism, is characteristically a sign of a tradition in crisis. (459)

I argue that the Copenhagen interpretation acknowledges the truth of this and refuses to hide the fact of the crisis behind a façade of certainty. If one takes this instrumentalist approach, one is open to the possibility of a deeper dimension to reality which can be accessed with further understanding. In contrast, the MWI argues that those other worlds are 100% causally disconnected from us. David Bohm makes it clear that the issue here is philosophical, not scientific:

The assumption that any particular kind of fluctuations are arbitrary and lawless relative to all possible contexts, like the similar assumption that there exists an absolute and final determinate law, is therefore evidently not capable of being based on any experimental or theoretical developments arising out of specific scientific problems, but it is instead a purely philosophical assumption. (Causality and Chance in Modern Physics, 44)

  • I am not sure I understand, there seems to be a disagreement between the beginning and the end. Do you agree that CI is Kantian (or instrumentalist, which is essentially the same) or not? Also, collapse doesn't violate relativity in CI, it's similar to saying that "P happened at x" becomes true all over the universe instantly once P happened at x, even though actual information about P can not travel from x faster than light. One can not use collapse to transmit any information or energy in QM, so unless collapse is hypostatized as in Bohmian mechanics, there is no issue with relativity.
    – Conifold
    Commented Nov 21, 2014 at 23:52
  • @Conifold: that CI cannot tell us about a deeper reality doesn't mean that nothing can. Indeed, CI resists saying that 'nothing can', in contradistinction to MWI. As to collapse under CI violating relativity, I read that in a few places that I now closed; I would have to look into it when I have more time—probably after Thanksgiving week.
    – labreuer
    Commented Nov 22, 2014 at 1:11
  • Oh, I see. Instrumentalism is like Kantianism but without the finality.
    – Conifold
    Commented Nov 25, 2014 at 0:43

The Copenhagen model and Bohm's model certainly get a lot of attention, but it seems to me the many-worlds model developed by Everett deserves more consideration.

You also might like to check out the paper "Quantum Phenomena Modeled by Interactions between Many Classical Worlds".

  • CPT reversibility and imaginary time, to me, make even more sense, as it seems bizarrely anti-Occam to create a bunch of realities we never visit. If time simply reverses itself often, but is then pushed back into a forward direction by the pressure of low entropy, you don't need infinitely many alternate universes, just a bunch of retrograde corrections. If it takes seven reversals to accommodate some circumstance, that is better than taking a continuum of alternate forward moves to explain the same thing.
    – user9166
    Commented Nov 20, 2014 at 15:26
  • @jobermark I'm unaware of a formulation of QM that uses this time reversal & entropy approach; where is it from?
    – Dave
    Commented Nov 20, 2014 at 15:36
  • 2
    I don't know, and I cannot even trace it down. I encountered it in discussions of 'Boltzmann Brains', and then lost track of it. In spirit it goes back to Loschmidt and Boltzmann, but that was before quantum dynamics. Feynman also talked about it as a way of removing the specialness of antimatter, but I have not seen that explanation laid out in a way that addresses collapse. I don't think he even really believed in 'collapse', just in all paths and elimination of the unobserved, so he would not have written such a thing.
    – user9166
    Commented Nov 20, 2014 at 15:38
  • @Dave, sorry, forgot the annotation, the comment above is to you.
    – user9166
    Commented Nov 20, 2014 at 16:22
  • Re: the two upticks on my last comment. I am glad the fact I don't know the source of my favorite theory is a useful comment in some way. But if anyone has more information, please pass it back to me.
    – user9166
    Commented Nov 21, 2014 at 3:46

You seem to imagine that there is some advanced philosophy behind the Copenhagen interpretation (CI), but this is not true. The CI is a strategy to avoid admitting the actual implications of quantum mechanics, and nothing else.

The substance of the CI is that (1) Quantum mechanics (QM) is so strange that it defies the laws of logic and is impossible to understand. (2) QM makes accurate predictions about some things, like atomic spectra. (3) So QM should be used to make predictions about some stuff but doesn't apply to other stuff we are familiar with.

Statement (1) contradicts statement (2) because predictions require explanations of how the experiment works, otherwise there is no account of which parts of the theory correspond to what you observe, or how to tell when you've set up the experiment correctly. If QM was actually incomprehensible, the right strategy would be to replace the theory, not fiddle about with it an ad hoc way as practiced by advocates of the CI. Advocates of the CI just prefer to ignore the implications of QM while using its predictions.

You talk about wave function collapse and non-locality. Any theory that features collapse violates the Schrodinger equation and so should be counted as a distinct physical theory that may in principle have different predictions, like GRW. The Bohm theory is also a distinct that makes different predictions, not an interpretation of QM and should be treated as such. Such theories may be right or wrong, but they are not QM.

QM itself does not feature wave function collapse and is entirely local, as explained here



The option of taking QM seriously as a description of how the world works is commonly called the many worlds interpretion (MWI). But as noted above, the other interpretations are either unclear or distinct physical theories.

The link provided above to objections to QM say things like the following. "I can't see how QM reproduces the everyday world." No clear explanation of what is wrong is offered. An objection that can be offered with no explanation can be rejected with no explanation. The objections to the explanation of probability in the MWI similarly miss the mark, as explained here:


Now, as far as epistemology is concerned, I don't think most physicists have any epistemological ideas. Most of them have some vaguely inductivist ideas. The CI fits with such ideas fairly well since they both have an astonishingly crude view of how experiments and science work. In practice these ideas recommend the following: if at first you look at something and you don't understand it, give up immediately and say it is incomprehensible. Scientists don't actually do this, which is why science is not a complete failure. Scientific knowledge is created by noticing problems, proposing solutions to them and then criticising them until only one is left and is has no known criticisms, as explained by Karl Popper. See "The Fabric of Reality" by David Deutsch chapters 2,3,7 and "The Beginning of Infinity" by Deutsch, Chapters 1,2,11,12 for explanations of quantum mechanics and epistemology that lack the problems pointed out above.


If you want to read what different members of the Copenhagen group thought the 'meaning' is, you can read Bohr's, Schroedinger's, and Heisenberg's thoughts in Quantum Physics and Ultimate Reality: Mystical Writings of Great Physicists by Michael Green. Available as a Kindle ebook. They were non-dualists.


Quantum language ( i.e., the linguistic Copenhagen interpretation of quantum mechanics) was proposed as the metaphysical and linguistic turn of the Copenhagen interpretation. This turn from physics to language does not only realize the remarkable extension of quantum mechanics but also yield the linguistic turn of Descartes=Kant epistemology. Also, this has a merit that the wave function collapse can be completely understood.

  • Welcome to Philosophy.SE. Do you think you might be able to elaborate on the linguistic turn as it relates to the present topic? The relevance is a little obscure and it might help to clarify.
    – commando
    Commented Mar 27, 2017 at 2:17

As far as I can tell there is no philosophy officially associated with the Copenhagen interpretation. You might like to check out Ulrich Mohrhoff's text book on the maths of QM, for he shows how the link can be made to the philosophy of Kant and back to the Upanishads, but he is not mainstream.

One thing. If you examine the definition of Kant's 'noumenon' it becomes clear that there cannot be more than one of them. This would be an important point when trying to reconcile Kant with the nondual idealism of the Perennial philosophy for it makes it much easier.

It seems to me that QM pioneers were fascinated by the philosophical implications of QM but somewhere around 1965 this fascination evaporated and now exploring these implications is frowned upon. This leaves us having to read Schrodinger, Eddington, Jeans and so forth, all long dead but still well ahead of the now moribund field.

You must log in to answer this question.

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