I'm not sure what it's called, but Russell famously resolved problematic statements like "The present King of France is bald," simply by saying that it contains a hidden assumption of existence. The statement itself says in full "There exists a present King of France, such that..."

Isn't there a similar hidden assumption of an observation in all scientific statements? The scientific statement would properly begin "It is observed that..." But in reading about the various paradoxes surrounding "observations" in science, I've never seen this rather simple qualification introduced. In other words, it simply makes no sense to say "It is observed that when the cat is not observed..."

Perhaps this is so obvious and accepted as to be trite, but I'm not sufficiently familiar with the literature. Isn't there always the assumption of an observer in physics, so that any statements about unobserved states are strictly metaphysical? Or is this just another way of stating the Copenhagen interpretation?

  • 1
    I think you need to be a bit more precise about what you’re asking! After all, hypotheses are perfectly reasonable scientific statements Nov 27, 2020 at 21:47
  • 1
    "so that any statements about unobserved states are strictly metaphysical?" - No, because Bell's theorem can be physically tested with real experiments, and it makes a statement about unobserved states. Yes, because Bell's theorem says that hidden variable theories are wrong or incomplete (under the assumption of classical locality), which greatly limits our ability to invoke unobserved states as an explanation for many phenomena in practice. Maybe, because I'm not sure what you mean by "metaphysical" in that sentence.
    – Kevin
    Nov 27, 2020 at 21:57
  • 1
    By "observation" I think you are referring to measurement in the sense that measurement provoke the collapse of the wave function in quantum physics (I.e. force the particule to have a fixed position, speed, spin...instead of just a probability). Lots of woo theories have been made about this formulation of an "observer" that makes stuff happen just by observing, but the current consensus is that it is not the observation but the interaction with other matter that provoke the collapse. Simply put, wave function collapse happen wether someone looks at it or not.
    – armand
    Nov 27, 2020 at 22:11
  • 1
    In science one can have models that agree well with observations and which also model the behavior of things that haven't actually been observed, like what general relativity says the curvature of spacetime is like beyond the event horizon of a black hole, or how quantum physics would model an isolated system between observations (your example of Schrodinger's cat). Sometimes these modelled-but-unobserved states are used to check the consistency of different theories (GR and QM being incompatible at the Planck scale for example), or with other types of assumptions (local realism in QM).
    – Hypnosifl
    Nov 27, 2020 at 23:58
  • 1
    Quine once suggested using Russell's paraphrase as a device for teasing out ontological commitments. This idea is largely rejected now. One can not extract "hidden assumptions" from local grammar, even grammar of "cleaned up" language. Using existential quantifier does not commit one to an ontology. The theory/observation dichotomy is also largely abandoned, see SEP. Analysis of commitments is semantic and global, that theories need to be empirically testable as wholes does not mean that individual statements share the property.
    – Conifold
    Nov 28, 2020 at 1:30

2 Answers 2


Short answer: no, statements in physics don't always assume an observer.

There are many perfectly legitimate statements in physics that do not (and often cannot) assume an observer. I'll give two examples.

  1. The Schrodinger equation tells us how the quantum state evolves in time. But there is no way to observe the full quantum state. So the Schrodinger equation is a statement in physics that talks about an entity many of whose aspects (coefficients in a superposition) cannot in principle be observed.

  2. Physicists believe that the whole universe is much bigger than the observable universe, and this often plays a role in cosmological models. In other words, physicists find it perfectly legitimate and even necessary to make statements about parts of the universe that cannot in principle be observed.

  • I know that in science, as with everything else, we discuss things that are not observable, but in science these must always be linked to observations and observers to be meaningful, so that in "science" per se existence is never apart from an observer. I believe Wigner said something like this, while Einstein felt otherwise. Nov 28, 2020 at 0:14

The problem is not just scientific, but moreover philosophical (philosophy being the mother of all sciences).

Classical science dismissed the subject as if it would not exist. The perspective of the subject was considered always as an absolute truth, which implies that no other truths are valid, and that there's almost no subject, assuming that any scientific statement is totally objective. Perhaps we can describe such approach as this: objectivity is the absolute negation of the subject. Modern science, on the contrary, has been required to consider the subject to keep advancing. That means that even if we want to exclude any subjective bias, any observation has always a biased (subjective) content, so a philosophical approach is always necessary for any interpretation (considering that most interpretations of reality are biased by our senses).

The most precise example of this issue is Kant's Copernican Revolution. It is called so, because Kant switched the problem of knowledge from the object to the subject. Probably since then, science had a solid foundation for the inclusion of the subject in the production of knowledge.

Examples in science, there are thousands. Possibly, the difference between the interpretation of quantum mechanics between Einstein and Bohr was the attempt of Einstein to have an objectivist perspective. Although that Einstein already had introduced the subject in Newton's laws of motion. For me, Boltzmann's equation has been always a proof of absolutism in science: his mathematical interpretation of entropy is perfect, for a nature made of particles, of small balls. But nature is not made of balls. Pre-darwinian approaches of survival were focused on the object (nature). Darwin introduced in science the role of the subject on the behavior of the ecosystem and on its own survival. Pre-Bertalanffy approaches of general entities follow the same lines, the theory of systems becoming a description of how the whole is impacted by its parts, each one related to a particular subjective behavior (which is not necessarily human). etc.

Remark that science deals with empirical truths, that is, the truths that result from experience, which might be incompatible with absolute or philosophical truths. Empirical truths are usually accepted, as long as they are useful (e.g. Boltzmann equation). That's why Newton's laws, creationism (against evolutionism), reductionism (against the systems theory), or the classical atomic theory (against quantum mechanics) were considered valid in a certain point of history.

There's a similar issue in philosophy, regarding ontology: most ontological approaches (study of the object) become epistemological problems after some analysis. That is just a case of a Copernican Revolution. An ontological approach of a rainbow will show that the colors, the position, and any possible knowledge of the rainbow is usually determined by subjective elements. So, the problem becomes, from ontologic (focused on the object) to epistemologic (focused on the interaction object - subject). But there's more! In multiple situations, the problem comes to be generic, so, a purely epistemic approach (focused completely on the subject) could be the last resort. Precisely what Kant did. But even if Kant set the agenda, the task of understanding the observer, us, is far from being completed.

You must log in to answer this question.

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