In Science and the Unobservable Nature (1937) it says

An outstanding characteristic of modern physics is that only that which is observable is significant.

... the followers of Einstein maintained that if the physical world were regarded as including entities or conceptions which were unobservable either directly or indirectly, there was no criterion for distinguishing the real from the unreal.

So unobservable things are considered unreal, i.e. non-existent.

Three classes of unobservables are described

  1. logically unobservable
  2. physically unobservable
  3. practically unobservable

Something practically unobservable may be a theoretical particle for which testing would require a more advanced particle accelerator than is currently available.

For some it is a point of dissonance that if the particle is subsequently found to exist this is not problematic with respect to its earlier status of not existing.

This is fine with Kantian phenomenology, where "Logically, [existence] is merely the copula of a judgement." A598/B626 i.e. the judgement of a cogito in joining subject with predicate: e.g. physicist declaring "the particle is found". ('is' denoting the copula of existence.)

The article concludes

It follows that modern physics is justifiable only in terms of an idealistic philosophy.

Is there a flavour of philosophy where the particle would be considered to exist before it is observed?

  • 1
    "unobservable either directly or indirectly" Most of contemporary physics' theories are based on indirectly observable elements: fields, subatomic particles, space-time. Commented Jan 29 at 14:01
  • @MauroALLEGRANZA Indeed. I'm really most interested in how much physics is in agreement with Kant on the meaning of objective existence. For example, Schrödinger's Cat is perfectly Kantian. Commented Jan 29 at 14:12
  • 1
    Kant's quote point at well-know kantian assertion that existence is not a predicate. Re: "Is there a flavour of philosophy where the particle would be considered to exist before it is observed?" see Scientific realism. Commented Jan 29 at 14:28
  • 2
    Is there a flavour of philosophy where the particle would be considered to exist before it is observed? - pretty sure that's the norm, not the exception. If we discover a particle on Tuesday, most physicists and philosophers would happily agree that the particle existed on Monday, and they were just incorrect on Monday.
    – TKoL
    Commented Jan 29 at 16:43
  • 1
    @ChrisDegnen sounds like you already have a name for the flavor of philosophy you're asking about then. And of course even more branches of philosophy believe that unobserved things exist than just that.
    – TKoL
    Commented Jan 31 at 13:13

10 Answers 10


I perhaps wouldn't call Kant's transcendental enterprise "a phenomenology", but that's besides the point.

It's also untrue that Kant was an idealist in the sense in which the term is used in the quote. Kant wasn't a phenomenalist, he didn't cast doubts on entities available only through inference from observation. In his philosophy there is no clear ground for making them a separate ontological category even. That's characteristic of phenomenalism, but Kant is, in fact, critical of phenomenalism (of Berkeley etc.).

Also, yes, unobservables exist. If we are phenomenalists we would consider them a part of a separate ontological category (what corresponds to "logical unobservables" on your account). If we aren't, then we would restrict ourselves to entities that are unobservable due to some practical circumstances or due to, for example, insufficient data - "physical unobservables" on your account (a theory might postulate entities which can never be directly observed, at least given what we currently know, but based on numerous observations we can claim some knowledge of them).

It's untrue what the author of the article says. Nowadays most philosophers aren't even phenomenalists - for various good reasons. And a phenomenalist wouldn't say that theorethical entities don't exist but rather that they're "useful fictions".

A general category under which various methods of determining what exists according to a theory fall is called "criterions of ontological commitment".

  • I think the decisive quote is in B287 : "while possibility was merely a positing of a thing in relation to the understanding (to its empirical use), actuality is at the same time its connection with perception." Actuality (existence) depends on perception, so an unobservable cannot exist in this scheme. Commented Jan 31 at 14:44
  • @ChrisDegnen This quote doesn't say that. It says: "What is not merely possible, but also actual, is not only in agreement with the formal conditions of the possibility of experience but also with the material conditions, i.e. with matter supplied by the sensation". Yes, a perception is alone insufficient to establish the truth of anything but perceptual judgements (I do not mean Kantian judgements of perception), that's why Kant speaks of, if I remember correctly, "agreement" - today we would speak of "coherence". Cf. Postulates of Empirical Thought in General.
    – user71009
    Commented Jan 31 at 16:05

Generally when used, scientific observation refers to the empirical apprehension of things in the world classically through sense data, though there are objections to objective notions of sense data today by conceptualists and less extreme views in the post-positivist philosophical ecology. It is a classic tenet of empiricism that phenomena are said to exist scientifically if and only if there is strong empirical evidence for them. In contemporary strong forms of ontological commitment (SEP), one can posit the existence of truthmakers. From WP:

Truthmaker theory is "the branch of metaphysics that explores the relationships between what is true and what exists".1 The basic intuition behind truthmaker theory is that truth depends on being. For example, a perceptual experience of a green tree may be said to be true because there actually is a green tree. But if there were no tree there, it would be false. So the experience by itself does not ensure its truth or falsehood, it depends on something else.

In physics, where the ontological status of subatomic particles stretches conventional notions of empiricism given the heavy application of mathematics and the exotic laboratory procedures that govern the metaphysics, it is of course natural that operationalization was first developed and applied. Hempel in his monograph on definitions, for instance, considers Bridgman's articulation of the theory in his survey of definitions. Thus, the correspondence theory of truth remains a central pillar of modern physics despite the challenges that the uncertainty principle and the wave-particle duality present. Russell's teapot is still a good way to understand existence in the philosophy of physics.

Yet, there seems to be to me an extension of philosophical principles to make claims about the existence of things in modern physics that goes beyond this classic and conservative view about objects, let us call it admitting the Meinogian jungle to the classically conservative physical ontology. It is simply admitting that abstract objects should be counted as physical objects if the abstract objects (SEP) exist in physical theory. Let us call this the "liberal interpretation of physical existence".

Physics, like all sciences, requires metaphysical speculation. For instance, phlogiston was speculated to exist prior to it being ruled out through empirical activity. Luminiferous aether is another example. Both were posited to exist, but subsequently discarded. That is, sometimes we have to posit a teapot in space before we can design tests and use them to determine whether or not the teapots exist. There are such teapots that are well-funded and well-received in today's mathematical physics: dark matter, strings, branes, and modal realist universes distinct from ours. None of these things have a lick of empirical support, and yet many accomplished physicists commit to them ontologically. What's going on?

In this case, while immediate apprehension and operationalization are still considered valid truthmakers for things that are directly and indirectly observable, physicists in question seem to be implicitly blurring the distinction between real objects and abstract objects. If, for instance, we accept the wavefunction as a real object, and rely on its predictive capability as a truthmaker by ignoring the distinction between model of reality, and physical reality itself, then we are now admitting abstract objects as real objects. Thus, the wavefunction IS the particle. This is consistent with Wheeler's It From Bit and his Participatory Anthropic Principle.

Under Wheeler and other physicists who are pushing against a radical physicalist interpretation such eliminative materialism, idealism, which 150 years ago was the default metaphysical presupposition, is attempting to reassert itself. Thus John Archibald Wheeler, Seth Loyd, Max Tegmark and other mathematical physicists are pushing against the strict dichotomy between real and unreal that the logical positivists attempted to prove was beyond question. Now, we can answer your question

You asked:

Is there a flavour of philosophy where the particle would be considered to exist before it is observed?

Yes. It's philosophical realism in the broadest sense: from vanilla, naive realism, Searle's direct realism, and scientific realism. In realism, we may not commit ontologically to an object until it's verified physically, be we do consider things to exist physically before they are observed.

Now consider Wheeler's anti-realist position, something that goes beyond using mathematics to merely predict existence. If one admits as a truthmaker mathematical physical theory and the existence of abstract objects and then ignores the Korzybski's maxim that the map is not the territory, we find ourselves at Wheeler's position where we show what is real through math first, and then bring about their physical existence by observing them. Note that this is not phenomenalism because the criterion of existence establish the Ding-an-sich-ness through the proof of the abstract object. In this way of thinking, objects are therefore mathematically real and then are created physically after the fact.

So, hidden variables, model-independent existence, and using mathematical physics to predict physical observation is actually nothing other than plain old scientific realism pushing back against the ontological ambiguity of instrumentalism and radical forms of phenomenalism. Kant introduced ontological ambiguity with his phenomenon/noumenon dichotomy, and how a thinker responds to his argument is key in understanding their respective position. Today, there are flavors of philosophy of science where the particle would be considered proven to exist before observation, and these are the anti-realist positions. (And somewhat unscientific in my estimation.)

  • I wonder if anyone has proposed Eliminative Idealism - nothing that we think exists does? (and I would add "as such" to that) Well, the Buddha did, but that might be cheating.
    – Scott Rowe
    Commented Feb 7 at 11:35
  • 1
    @ScottRowe On a serious note, elimnativism is to collapse one into another, so eliminative idealism would be collapsing all physical into the mental. I'm rather sure that's what Berkeley had in mind with his subjective idealism (though he didn't call it that himself IIRC). On a more observant note, the denial of the existence of anything would leave a philosopher with nothing to engage in endless metaphysical speculation about; where's the fun in that?!? : \
    – J D
    Commented Feb 7 at 16:48
  • 1
    "I philosophize, therefore it is." :-)
    – Scott Rowe
    Commented Feb 8 at 0:10

It depends on what you mean by "observable".

Distant stars and bacteria are unobservable by the human eye. Then microscopes and telescopes are invented, and suddenly both those things become observable. But such observation did not create distant stars and bacteria, which existed before those inventions. They were simply rendered visible, allowing observations.

Infrared rays and X-rays were similarly unobservable by the human eye until photographic film was invented, which allowed us to "observe" things our eyes could not on their own.

Note also that modern physics is under no obligation whatsoever to be congruent with Kant.


Broadly speaking, I understand your three classes of unobservable things as follows, using slightly different names and characterizaions:

  1. Logically unobservable: These are concepts which cannot have a referent because they are logically impossible: A round square.

  2. Physically unobservable: All ideas, e.g.,mathematical concepts like Hilbert space, curved manifold, integral schema, …

    These ideas may be useful as human constructions to formulate scientific theories. But they cannot be detected by observation, because ideas have a completely different ontological status than physical objects.

  3. Practically unobservable: Physical things, which are observable, but have not yet been observed.

    There are different causes why such things have not been detected until now. But in any case, the existence of the thing does not depend on whether it is discovered or not. There is no ontological change when the thing is discovered. In particular, the thing is considered to exist before it is observed.

    Examples: The neutrino, the Higgs' boson, the genes - all before their discovery.


"Does any philosophy define 'existence' such that unobservable things exist?"

You also ask about the distinction between "logically unobservable and physically unobservable".

Leibniz's philosophy does just that. He distinguishes between entities that are accidentally impossible and those that are absolutely impossible. The latter is a logical impossibility, a.k.a. contradiction. For instance, Leibniz considered infinite wholes as being impossible in this strong sense. On the other hand, accidental impossibility has to do with the unavailability of such entities in the world we happen to live in. Leibniz applied this distinction to infinitesimals, which in his mature period (starting about 1676) he viewed as only accidentally impossible. These and related issues are dealt with in our publication

Katz, M.; Kuhlemann, K.; Sherry, D.; Ugaglia, M. "Leibniz on bodies and infinities: rerum natura and mathematical fictions." Review of Symbolic Logic 17 (2024), no. 1. https://doi.org/10.1017/S1755020321000575, https://arxiv.org/abs/2112.08155


Cosmologists have no problem with the idea that the universe extends beyond the observable universe (but unobservable because the universe has not been in existence for long enough for light from there to have reached us). Depending on how the universe is expanding, the unobservable universe may never be observable. That doesn't mean it doesn't exist.

The singularity within a black hole would be another example. We can't observe it in any way (I think we can only determine the black holes mass, spin and electric charge, but nothing else?). Yet we don't seem to find it that difficult to consider the existence of something we can never see (even if we "visualise" it, that almost certainly wouldn't be what it actually "looked like").

  • If one can make confident experimental calculations about something, that can be considered as aided perception. The notion of 'unobserved stars' together with the aided perception of them does grant them existence in the Kantian form, e.g. B287 as "the conjunction of the thing with perception". I guess the unobservable universe is not really completely unobservable to the extent that we can confidently model the creation of early stars. Commented Feb 5 at 15:58
  • 1
    @ChrisDegnen there is no way to observe stars beyond the observable univers using any instrument - it is prevented by the finite speed of light. We can't make calculations about them, we can't observe them. There can be no "conjunction of the thing with perception" as there is no perception. The reason we think they exist is simply that there is (as far as we know) nothing special about the boundary of the observable universe - it is just a consequence of our location. We can confidently model the creation of the early stars we see, it is an assumption that the others are the same. Commented Feb 5 at 16:23
  • 1
    "it is an assumption that the others are the same" Indeed it is an assumption that there are unobservable stars, but it is supported by the science that models the Big Bang. Like donning a pair of super-science spectacles to aid perception. Commented Feb 5 at 16:30
  • 1
    @ChrisDegnen that is not the same as observing it though. As lot of philosophy, a much of this seems to be strained use of natural language. For me not having a basis to distinguish something that is real from something that is unreal doesn't mean that it is unreal. But that is perhaps because I am a Bayesian and am not unduly worried by uncertainty. The multiverse is an example of something unobservable that is supported by science (in the same sense, but perhaps not as strongly) where opinion is divided on its existence. Commented Feb 5 at 16:36
  • @ChrisDegnen I don't think we can know whether the the model of the Big Bang necessarily means there are stars outside the observable universe. It is possible that we are at the center of a seed from which the observable universe inflated and there is nothing else. How could we tell? More assumptions. It sounds to me like the authors mentioned in the OP are trying to put physics on too sound a foundation, perhaps as Hilbert tried for maths. Commented Feb 5 at 16:41

As is very often the case in seemingly deep and grandiose statements about physics, this one is utterly boring. The truth is usually boring. You get to deep, exciting and counterintuitive truths on the far side of a mountain of achingly dull truisms.

"Observable event" in the cited passage means "a hypothetical experiment of arbitrary sophistication could, in principle, distinguish between a universe in which this event happened and this event didn't happen."

Let's rephrase:

Modern predictors of the state of the universe believe that if there's no difference between the state of the universe in which something happened and the state of the universe in which it didn't happen, there's no difference between those states.

If events were considered to be such that no possible criterion could distinguish events from non-events, there would be no possible criterion for distinguishing events from non-events.

There are, in fact, many exciting and counterintuitive truths - indeed, the whole of modern science - on the far side of this achingly dull truism. The reason you can trust any of it is because all that follows is built upon the solid rock of the utterly boring.

No philosophical idealism, or any other metaphysics, is indicated. It's just Identity, Noncontradiction, and the Excluded Middle.

  • I love boring things! Where would we be without them?
    – Scott Rowe
    Commented Feb 5 at 18:44

Is there a flavour of philosophy where the particle would be considered to exist before it is observed?

when you say about a particle that existed before it is observed, you indirectly imply the existence of some non-particle also because these non-particles must be necessary to differentiate each other.

You categorised unobserved as 3, viz., logical, physical and practical. But just imagine 'self' or 'consciousness'. Does it come under these 3 categories? Won't it escape all these 3 you mentioned because it transcends the mind?

If the observer, the observed and the act of observing are ultimately one, the 3 categories using the term observe or observable becomes meaningless when the ultimate is reached ... or when it transcends the mind.

Since I know little about other philosophies, I didn't answer your main question. But I can say one thing -- if any philosophy that contradicts what I have mentioned here about non-duality, if they are 'particle based' or accept particles as fundamentals, they will have to agree with you, I think ... I mean, the particle is assumed to exist before it is observed.

  • So I could speculate on tachyons for example, but none has been detected, so not reified. (Might be tricky if it were though 'cos it could change the past, lol.) This is all about reifying objective things though. In contrast, the observer's Dasein is already a given, like Descartes' cogito. The Dasein-cogito reifies objective things by correspondence of sense data with mental modelling and rationalisation. The thinking of "cogito ergo sum" comes from somewhere before thought, e.g. Kant's gedankenlose Anschauung (thoughtless intuition) or any other indeterminate originary mystery noumenon. Commented Feb 1 at 9:57
  • So the self (Dasein) or consciousness (cogito) is about experience itself. Where thought comes from is before thought so unobservable except perhaps in a non-mentational zen-type blank intuition, (which comes with purported clarity benefits). Dasein itself seems to be partly subjectively existing and partly unobservable/mysterious. Of course it can always consider itself as objectively existing but I mention that only to differentiate objectivity from subjectively existing and experiencing and being in the zone. Commented Feb 1 at 10:21
  • Re. "Won't it escape all these 3 [categories]"? It may be logically unobservable. Heidegger calls the noumenal origin of experience 'Being', and by his logic states "Being and the structure of Being lie beyond every entity and every possible character which an entity may possess." (Being & Time, H.38) Due to the subject matter (being) the logic gets quite elusive, but basically thought is tangible and determinate and what it originates from is intangible and indeterminate, which makes it logically unobservable. Commented Feb 1 at 17:14

Does any philosophy define 'existence' such that unobservable things exist?

If something existed which was unobservable in principle, then it would have to have no effect whatsoever on ourselves, either directly or indirectly, and then, though it would exist, it would be as if it didn't, and it would therefore have not the slightest importance to anyone because it could have no effect on us.

It should be noticed that we have the ability to conceive of such things, the proof being this discussion, but we don't have the ability to imagine them, as if our imagination couldn't be bothered.

Is there a flavour of philosophy where the particle would be considered to exist before it is observed?

This is the default position of every particular science, otherwise, scientists would have to say that the world didn't exist before humans started to look at it. This is also the default position of every human being on this planet. Nobody thinks that other people only start to exist when we meet them the first time. This would make for a world without causation as we think of it. We would have to say that our own progenitors don't exist before we their children see them for the first time, and then how exactly were we born to this world?


Yes. According to David Lewis' modal realism, things only have to be possible in order to be "real."

Lewis held that this world was just one among many like it. A proposition, p is possibly true if and only if p is true in one of these worlds.

I believe that things could have been different in countless ways; I believe permissible paraphrases of what I believe; taking the paraphrase at its face value, I therefore believe in the existence of entities that might be called ‘ways things could have been.’ I prefer to call them ‘possible worlds.’ (1973a: 84)


It's actually a good counterpart to modern physics, since the "many-worlds" theory is one interpretation of quantum indeterminacy.

You must log in to answer this question.

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