Bohr famously said in relation to quantum systems:

Nothing exists until we measure it

This can't be right, for how can we measure Nothing, something that doesn't exist. It seems it must come into existence just prior to our measurement. But this means the quantum system must anticipate our measurement - which also sounds bizarre.

Could we say, instead that it exists but its value is indeterminate?

  • 5
    I think he just meant that it doesn't have certain characteristics with probability 1 until it's measured.
    – Xodarap
    Dec 30, 2013 at 23:48
  • The Participatory Anthropic Principle (by John Archibald Wheeler), even states that the universe requires observers, because without observers the universe could not actually exist.
    – draks ...
    Dec 31, 2013 at 0:24
  • The universe does not need us. There is plenty of evidence that it existed without us. So I would say Bohr is simply wrong.
    – Dave L.
    Jan 2, 2014 at 3:08
  • This seems to be a variant of the "If a tree falls in the forest, does it make a sound?" dilemma.
    – Geremia
    Feb 20, 2014 at 21:02

5 Answers 5


This is a particularly sticky wicket because (a) we are delving into the philosophy of quantum mechanics (which is beset on all sides by those who wish to pervert it to their own ends) and (b) because we are confounding our understanding of the problem by confusing what we mean by "exists."

To make this question more meaningfully answerable I will address precisely what is meant by the idea that "nothing exists until we measure it" and then go on to provide an answer to this newly reformulated question.

Who can really know what Bohr was thinking when he made the statement that you quote in your question, but what I assume that he meant is not that the "particle" (as one example) does not exist before you "measure" it but rather that information about the particle's "observable" attributes (such as position, momentum, etc.) does not exist before that observable is measured. In other words, quantum theory claims that all the information that we have about a particle is contained in its wave function, which generally represents a linear superposition of possible states in which the particle might be found if a measurement is performed.

Therefore, one might ask, "What is going on before we make a measurement?"

Historically, there were three answers to this question:
(1) Something totally predictable is happening, but quantum mechanics doesn't know what it is because quantum mechanics is incomplete. (The "Hidden Variable" Answer)
(2) Maybe something totally predictable is happening, but we'll never be able to know what, so don't even bother asking.
(3) Whatever is happening looks just like quantum mechanics and therefore doesn't affect the outcomes of our experiments, so who cares?

As it turns out, the first answer is basically dead because of Bell's Theorem, which essentially says, "If the predictions of quantum mechanics are correct, then local hidden variable theories do not describe reality." In other words, no theory which allows us to know exactly what the particle is doing at every moment in time (and also preserves causality) can ever reproduce all of the predictions of quantum mechanics.

The second answer is demonstrably false because a theory can be constructed which tells us exactly what the particle is doing at every moment in time by violating causality. It is known as the "pilot-wave theory." The theory exactly reproduces the predictions of quantum mechanics but is widely disregarded as a "toy theory" because it doesn't make any new predictions of its own.

As a result, answer #3 in conjunction with the Copenhagen Interpretation tends to be the default position for most physicists.

The take-home message is this: if you insist on things being "classical" (i.e. having definite observable values) to exist, then you must sacrifice local causality. On the other hand, if you are willing to accept the existence of things that don't have definite values for their observables (that is, accept the idea that an electron is its wave-function - whatever that means), then things certainly exist before they are measured. However, just as easily, you could say that "nothing exists" before a measurement and the act of measurement forces the system to "take a stand."

Any way you cut it, there is definitely one thing that you can't say: namely, that the system has some definite value but we just don't know what it is. Either it has a definite value and we know it, or it doesn't and we don't. You can have one or the other but not half-and-half.

  • +1:How 'local' is local causality? Can any bounds be put on it? I think I'd go for indefinite reality. Do many worlds return definiteness? Though I find this idea distinctly unappealing. Dec 31, 2013 at 20:00
  • I find the idea of 'nothing' ontologically suspect. But I can accept Bohr meant it as shorthand for something else. Dec 31, 2013 at 20:04
  • Local causality is the idea that a "cause" must exist within the past light cone of its "effect." If something violates that (i.e. a cause is outside of the past light cone), then it is said to violate local causality. The many-worlds interpretation is a whole other can of worms which does not cast any useful light on the nature of our universe. I suspect that "nothing" is shorthand for "no observable" (i.e. if we take a particle to be merely the label we give to a particular collection of observables, then it doesn't exist until we measure it).
    – Geoffrey
    Dec 31, 2013 at 20:12
  • Presumably, a violation of local causality means that a cause must come from the future light-cone? Unless of course the future & past light-cone together doesn't exhaust all the possibilities for causes. Isn't violation of local causality a prediction? I find it interesting that such a theory can be constructed with the same physical content as the usual theory. Dec 31, 2013 at 20:31
  • An effect could conceivably come from the causally unconnected region outside of both light cones. Such events are called "space-like separated" - to be distinguished from "time-like separation" for events whose light cones overlap.
    – Geoffrey
    Dec 31, 2013 at 22:31

I've been daydreaming about this, and have come up with an interesting thought. Equivocation. The way you observe this sentence defines it. Here on this page we have already three ways of observing it. Your own which is

"There is a thing called nothing, which ceases to be when it is observed"

The second one, by Chris that says:

he's saying a thing doesn't exist until it's accurately perceived

Then on the comments by draks:

the universe requires observers, because without observers the universe could not actually exist

My own interpretation is that if I cannot observe something (even through some convoluted means like neutrinos with Heavy Water), I cannot say it exists (similar to Chris), i.e. something that interacts with nothing may as well be nothing itself.

It would appear that the question answers itself. The perception of anything depends so much on the observer state, that the "thing" depends on the observer. That is not to say that things would not "exist" without the observer.

The very sentence is a play on words since its meaning depends on the observer, and thus without the observer it has no meaning. It is the observer of the sentence that gives it meaning.

I am observing this as a clever self-explaining premise. Thanks to you guys.

  • You are close, but modern (continental) philosophy specifically defines existence (extantness) as being codependent on the observed and the observer. Things that you describe as existing without (ever) being observed can be admitted to exist under the category of "unobserved things", which we can be justifiably confident about. Jan 7, 2014 at 9:06
  • For instance: "... what is going on when we assert "the coffee mug exists"? If we are not talking about an essential aspect of the coffee mug, then what are we doing? According to Heidegger, Kant's answer to this question is that actuality or extantness is equal to absolute position, which is equal to perception. Heidegger's charitable interpretation of this cryptic equation is that actuality is equivalent to being-perceived, or perhaps, the possibility of being perceived." - (Link) Jan 7, 2014 at 12:25
  • @Degnan: Does Heidegger discuss the possibility of matter perceiving matter? Or is perception a solely human perspective. Apr 25, 2014 at 21:09

He isn't saying nothing exists; he's saying a thing doesn't exist until it's accurately perceived. (The possibility of the existence of nothing is another matter: ref. mystical speculation.)

The idea that existence depends on perception comes from Descartes and Kant et al, who changed the orientation of existentialism so that the existence of phenomena was understood to be codependent on the perception of the perceiver. I.e. I think therefore I am ... For a thing to exist (to someone perceiving) the idea of it has to be comprehended, and then the actual existence of the thing has to be confirmed, by sense or measurement. Formally this type of existence is called being extant, and it applies to perceived objects, e.g. tables and chairs, atoms. The existence (being) of the perceiving being is a transcendentally different class of existence which Bohr is not alluding to.

If you find the idea that the existence of objects being depended on the observer doesn't gel with your sensibility that's normal, but it is an accomplishment of modern philosophy that the existence (as extantness) was formulated to be relative to cognition.


There is a description of what a quantum system is like before, during and after a measurement. You get that description by doing experiments to test theories and by looking at the equations of motion that have survived testing and working out what they imply about how the world works. When it comes to quantum mechanics, physicists have often been remarkably unclear about what explanations are ruled out by experiment and what the equations of motion describe. The correct description of what is going on is that every object around us exists in multiple versions that are approximately sorted into layers each of which approximately resembles the universe as described by classical physics. See


and "The Fabric of Reality" and "The Beginning of Infinity" by David Deutsch.

  • 1
    I think the answer expresses one of the theories, but I think you're overstepping a bit by saying "the correct description." Part of the point is that this cannot be resolved scientifically.
    – virmaior
    Apr 25, 2014 at 11:13
  • It can be resolved in the same way as any other scientific controversy. You propose solutions to problems, criticise the solutions by doing experiments and by arguing about issues like whether the theory concerned is consistent with other ideas that solve problems, whether it solves the problem it was intended to solve, whether it solves problems other than the problem it was invented to solve and so on. If the theory in question is wrong, then it is wrong because it fails to solve some specific problem, not because of "overstepping" whatever that means.
    – alanf
    Apr 25, 2014 at 12:32
  • Not, it really cannot. Most scientific controversies can be resolved due to an explanatory difference (i.e. an experiment that matches one solution but not another). Due to the very nature of this disagreement, that's not possible. The many minds hypothesis, the many world hypothesis, and others are about the observer problem -- which occurs as a question of explaining the state of things in both observed and unobserved state. There's no experiment that can resolve that.
    – virmaior
    Apr 25, 2014 at 14:11
  • I didn't say an experiment can resolve it. The alternatives to the MWI fit into two categories. (1) Alternatives to quantum theory that can in principle be experimentally tested such as the Bohm theory. Those often clash with other theories like relativity and don't solve any problems, so they are wrong. (2) Ideas such as the Copenhagen interpretation that say quantum theory is wrong but provide no alternative account of how the world works. These are eliminated from the controversy by their vagueness.
    – alanf
    Apr 25, 2014 at 14:44
  • You claimed that "it can be resolved in the same way as an scientific controversy"... which would be through experiments. You then say that other interpretations are disproven through ... experiments. You then dismiss another theory due to "vagueness." This only serves to validity my claim that the locution "the correction description" is errant.
    – virmaior
    Apr 25, 2014 at 16:03

Bohr famously said in relation to quantum systems: "Nothing exists until we measure it"

Bohr is not debating whether 'nothing' does or does not exist, as that would be nonsensical. Nothing is nothing. It does not refer to anything. It is just a grammatical stand-in for what is lacking.

However, this is a correct, philosophically sophisticated explanation of existing things in the physical world, the empirical side of the world of physics.

In the subatomic world of quantum physics, there are literally no existent physical objects prior to an actual quantum interaction. It is the interaction that not only measures measurables, thereby making them real, but also creates the self-same measurables as permitted by the interaction's potentialities.

Existence isn't there prior to quantum interaction. Existence arises from the interaction of two potentialities.

Our apparatus only manipulates what is possible so measurement becomes possible. Measurement can only measure properties, as at the quantum level, that is all there is to see. It is the theorists who classify clusters of properties into particles or waves.

There really aren't any tiny particles or waves prior to quantum measurement.

  • 1
    Bohmian mechanics shows you can get a realist ontology if you're prepared to sacrifice non-locality; this position at least makes us think critically about Bohrs position. Apr 29, 2014 at 0:22
  • Non-locality is well established by now. Quantum tunneling is strange enough, and it has been the basis of all modern electronics for 50 years.
    – user6323
    Apr 29, 2014 at 11:51
  • I should have said sacrifice locality. Quantum tunnelling isn't non-locality; non-locality refers to information signalling that is faster than the speed of light; of course this is managed such that causalty isn't broken; Apr 29, 2014 at 15:48
  • Under [link="physics.stackexchange.com/questions/81190/whats-inside-a-proton/… in a proton?[/link] @David Z said: [quote]A proton is really made of quantum fields. Remember that. ... One of the more confusing things about quantum fields is that they react differently depending on how they are observed.[/quote] illustrates that even from solely a particle perspective, instrumentation (energy level) creates what will be seen, i.e. what is 'real'.
    – user6323
    Apr 30, 2014 at 12:15

You must log in to answer this question.

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