Watching quantum mechanics lectures and it was mentioned that it is pointless/meaningless to try to talk/question things that can not be tested/measured.

Is this a principle? And if so what is it's name?

Also does this apply to questions other than Quantum mechanics? E.g. does it make sense to ask if earth was the only object in universe if gravity would still exist? Although it seems intuitive to answer yes, yet it is something that can never be tested/verified. It seems almost as meaningless as to ask if the universe was only made of a unicorn would it have gravity?

As an analogy consider glottogony, at some point it was banned as it seemed to be an unanswerable problem. http://en.wikipedia.org/wiki/Origin_of_language


8 Answers 8


It's not really a single principle - it's a philosophy and in the context of philosophical discussions about science, it is usually known as positivism.


As any philosophy, it cripples the penetrating power of science if it is extended too far - and every philosophy ultimately fails. The thought experiment about the Earth in the Universe is just one among millions of examples. Positivism could have a problem with the whole concept of thought experiments.

While it was very useful and important for the development of quantum mechanics to realize that science doesn't have to talk about things that can't be measured, i.e. that theories that deny the existence of things that can't be observed are just fine, it still remains true that science also can talk about concepts that can't be observed, such as quarks.

It's up to the scientific method to decide whether an auxiliary concept or a theory that isn't accessible to observations has an explanatory power that justifies its validity - and the answer may be different in each individual situation.

  • 3
    $\begingroup$ I would up-vote this if it didn't disparage philosophy so much. While I agree philosophy can only tell us so much, it is complementary to physics (and other fields), and is in general a very worthwhile pursuit. $\endgroup$
    – Noldorin
    Jan 24, 2011 at 22:00
  • 2
    $\begingroup$ Actually, physics is "natural philosophy"... :-) $\endgroup$
    – Sklivvz
    Jan 24, 2011 at 22:12
  • 3
    $\begingroup$ That's right, Sklivvz. Physics is a natural philosophy while the other philosophy is either unnatural philosophy or man-made philosophy. ;-) Jerry: yes, the statement that "every philosophy ultimately fails" cannot be proved within the system, as dictated by Gödel's theorems. However, the proof that it's unprovable within the system also shows that it's true. $\endgroup$ Jan 25, 2011 at 6:40
  • 6
    $\begingroup$ Quarks can be observed as well as anything else. Positivism does not forbid inference. $\endgroup$
    – Ron Maimon
    Aug 30, 2011 at 5:34
  • 2
    $\begingroup$ The key word for positivism is «meaning». Positivism does forbid inferences to thing which cannot be measured, since it forbids us to even talk about them. It says all discussion of them, including inferences using them, are meaningless. Positivism was not actually very helpful in the development of QM: it inspired Heisenberg, but if he had never done it, Schroedinger would have, and Schroedinger was inspired quite differently. $\endgroup$ Dec 19, 2011 at 8:06

This principle is called "positivism". But I prefer the term "logical positivism".

Positivism is a basic principle of thought--- it distinguishes questions which are meaningful and meaningless. It is not meaningful to ask "How does Argentinian property law taste?", it is not meaningful to ask "How many angels can dance on the head of a pin?", and it is not meaningful to ask "Wny is there something rather than nothing". These questions have the property that, whatever the answer, there is no consequence to our perception of the world, or to anything that we can measure. Such questions do not need an answer, because they are just words put together without meaning.

Philosophers do not accept positivism, because it moots many of their favorite questions. For example "how can we have free will in a deterministic universe?". This question, although superficially so clear, is positivistically meaningless, because there is no measurement which can be made upon a creature's decision making power which can in any way determine whether the underlying laws are deterministic or not. This question is moot, because if you define "free will" properly, it cannot possibly have anything to do with determinism. You can't figure out that quantum mechanics is not deterministic by introspection, because your introspection has nothing to do with the determinism of the underlying laws.

Similar jibberish:

  • If the universe exists, doesn't it require a prime mover?
  • Is ethics/mathematics defined independently of human beings, or do we produce it?
  • If God is good, how come there is so much suffering in the world?

Basically, half the philosophy curriculum is brain-dead nonsense, and the rest requires massive restructuring to eliminate redundancies. Religion, when formulated positivistically, is not nonsense, and it is possible to formulate religion, including monotheistic religion, in positivist terms, only throwing away the metaphysics and keeping the meat of it.

Logical positivism is an early twentieth century tradition that combined the notion of scientific positivism with the new concept of a formal language, then being developed in Logic by Boole, Frege, Russell, Hilbert, and others. The logical positivists wished to get rid of the other annoyance of philosophy, the fact that we have to deal with imprecise natural language. The goal was to replace natural language with a formal language, in which all the terms of discourse are precisely defined. In this way, you would remove another huge branch of philosphy, namely that philosophy whose goal is to make sense of the various ambiguities in the writings of other philosphers.

The formal languages at the time were barely adequate for formalizing mathematics. It was a bold leap to assume that formal languages could encompass a large enough domain of discourse to approximate natural language. This leap was associated with a bunch of people who I am not going to name, because I haven't read any of them, because philosophy is so trivial compared to any real intellectual work, that you can reproduce any of the results of the non-mathematical philosophers for yourself by thinking for ten minutes.

Logical positivism moots a whole bunch of other philosophy, because it suggests that the right way to think of terms in a language is in terms of a reduction to a formal language, much like in mathematics. If you take the positivists seriously, most of the field of philosophy is pointless and stupid.

At first, philosophers celebrated the revolutionary ideas, and positivism was the ascendent philosophy until the 1950s. But with the emergence of horrific totalitarian states with scientific materialism as their religion, scientific materialism lost ground, and by the 1970s, it was killed off in the west.

The positive influence of positivism

People think positivism equals quantum mechanics, because quantum mechanics was formulated strictly in positivist terms, using "observables". But it wasn't only in quantum mechanics that positivism was important:

  • Fields: the notion of fields, which is so abstract, acquires meaning through positivism. How can you tell if a field is there? Put a charge there and see it move! This positivist formulation was important in making clear that a concept that seems so immaterial to many people at first glance is in fact real.
  • Luminiferous ether: the ether lost its material characteristics one by one, and it was an act of positivism by Einstein to reject the ether completely, because it had become unobservable.
  • Equivalence principle: In order to get from the fact that you can't observe gravity in a free-falling frame to the principle that gravity is a geometric force, Einstein made an act of positivism. If the effects of acceleration are indistinguishable from gravity, then gravity and acceleration must be the same thing in essence. This is a very predictive statement.
  • String theory: String theory emerged from a positivistic question--- how can you make a measurement in a space-time that is not well defined at short distances? The answer was to speak about s-matrix states, and their scattering. In physics, the resulting S-matrix theory led to string theory, which is the only candidate for a theory of everything.
  • Black hole interiors/holography: Within string theory, the rejection of the simultaneous existence of the interior and exterior (since each are observed by different observers) led to the fruitful principles of black-hole complementarity and holography as developed by Susskind.

There are also several cases where positivism was misapplied overzealously, and I think this clarified things:

  • Quantum field theory is meaningless--- because it is difficult to imagine measuring a quantum field. In fact, the original paper of Bohr-Rosenfeld analyzed measurements of the quantum electric and magnetic field, and decided that it made sense to quantize them. the positivist's complaint was about ultra-short distances, and for those distances you need to deal with gravity anyway.
  • Quarks don't exist--- it is impossible to isolate quarks. This is no better an argument than Mach's argument that atoms don't exist because he can't see them. If there are observable phenomena which are best explained using quarks, then quarks exist to the extent that they are included in the explanation. The real battle here was over quantum field theory, and the postivists were the folks doing the s-matrix theory.

Outside of physics, positivism is identified with this ridiculous claim:

  • Interior experience doesn't exist--- since all we can measure are inputs and behavior, we should model organisms as a black box without internal experience. This idea, due to Skinner, is pretty idiotic.

In my view, positivism is an engrained part of science, and in my view, it has never steered us wrong, even in cases like S-matrix-theory/string-theory when almost everybody thought it did. With the presence of computers, and computer languages, formal languages are no longer abstract and remote. People program computers all the time. Scientists know when what they say has meaning when they can program a computer to do their model. So I think it is fair to say logical positivism is the only correct functioning philosophy.

All the logical positivists are now dead, and their work is basically ignored within their own field. If you ask philosophers why they ignore logical positivism, they often say: "Positivism contradicts itself. It says all truths need to be experimentally verifiable, but positivism is not experimentally verifiable!" To call this stupid, self-serving and intellectually vapid is too kind.

  • $\begingroup$ "If the effects of acceleration are indistinguishable from gravity, then gravity and acceleration must be the same thing in essence..." By no means. If that is true then, because it is a very predictive statement, it should also be true this sentence: "If the effects of acceleration are indistinguishable from "motion in an electrostatic field"", then "electrostatic field" and acceleration must be the same thing in essence. They are distinct notions. $\endgroup$ May 31, 2012 at 22:07
  • $\begingroup$ @HelderVelez: The effects of accelerations for electrostatic fields are not indistinguishable unless all particles have the same charge as mass. Otherwise, you use a neutral particle to see that other particles are accelerating. But there is such a limit, in Kaluza Klein theory, in cases where all particles have the same momentum in the fifth direction, and it is precisely in this case that gravity and electromagnetism are welded together to make electrostatics a geometric force. $\endgroup$
    – Ron Maimon
    Jun 1, 2012 at 1:22
  • $\begingroup$ Skinner never said inner experience doesn't exist. What he denied was the scientific usefulness of introspection and traditional informal language (words such as love, fear, consciousness, etc) as a method of construction for a theory with well-defined elements that can be measured and can explain (to some reasonable approximation) "inner experience". $\endgroup$
    – zyx
    Aug 29, 2012 at 23:12
  • $\begingroup$ @zyx: Then why would Skinnerian linguists treat language as a regular automaton? This is suggesting that they view the inner experience as a simple computation between input and output. $\endgroup$
    – Ron Maimon
    Aug 30, 2012 at 2:13
  • $\begingroup$ Modern computational linguistics is Skinnerian, behaviorist, data driven positivism. Crunch the data, train the machines, and extract useful lessons. None of this requires the automata to be regular or stackless, but even so, as summarized in kornai.com/Papers/efsed.pdf, use of finite state models actually increased over time as the Chomsky program failed to handle real data sets. $\endgroup$
    – zyx
    Aug 30, 2012 at 3:11

This is a very good question and we have very good discussions.

I feel that a meaningful scientific question should satisfies the following condition: The different answers to the question should have different measurable consequences.

Also a meaningful scientific statement should satisfies the following condition: The statement being true or false should have different measurable consequences.

For example, "a sign shows 'stop'" is a meaningful scientific statement since it can be tested by looking at the sign. But "What does a sign show before anyone looks at it?" may not be a meaningful scientific question. (Or is it?) This kind of issue appear in the measurement theory of quantum mechanics.

How to ask a meaningful scientific question? One may first think about what the potential answer might be. Then try to design an experiment to test the validity of the answer. A scientific question is meaningful or meaningless depend on the ability to design an experiment that can test the validity of its potential answer.

Here is an example: "Is existence timeless or in time?" The potential answers are: Existence is timeless and Existence is in time. The next task is to design an experiment to test the validity of the statement "Existence is timeless" or "Existence is in time". Here I have trouble. I do not know what experiment can test the validity of the above statement. Being able to design such an experiment will make a great scientific progress. One big area of scientific research is to design such kind of experiments and to make those used-to-be meaningless questions meaningful.

Not having such an experiment, the question is not a meaningful scientific question. The burden is on the person who ask the question to provide the experiment that, at least in principle, can test the validity of the potential answer.

  • $\begingroup$ I get it if you want to say that 'the answer should only be about something measurabe'. I find "The different answers to the question should have different measurable consequences" confusing. Not all answers are right and isn't the answer already a prediction of what will be mesured? How can an answer have consequences? The answers "the sign is red" and "the sign is blue" don't have any consequences for the so called actual color of the sign. The only consequence is one regarding the truth of the answer. "answer" shouldn't be equated with "true statement" in a discussion about meaning. $\endgroup$
    – Nikolaj-K
    May 28, 2012 at 14:03
  • 1
    $\begingroup$ There are meaningful answers and there are meaningless answers. There are right answers and there are wrong answers. A wrong answer can be a meaningful answer if we can show the answer is wrong through a experiment. A wrong answer is better than a meaningless answer. A meaningful question, by definition, allows meaningful answers. In physics.stackexchange, we should try to ask meaningful questions, by thinking about weather the potential answers can be tested by experiments. This is the fundamental division between scientific questions and non-scientific questions. $\endgroup$ May 28, 2012 at 14:28
  • $\begingroup$ By "have measurable consequences", I mean "can be tested by experiments". "the sign is red" and "the sign is blue" do have measurable consequences. They can be tested experimentally by looking at the sign. $\endgroup$ May 28, 2012 at 14:35
  • $\begingroup$ Okay, I see what you're saying. Although your use of the term "consequences" is a little unconventional. By this undestanding, the mere statements "This desk is a red.", "This desk is blue." and "this desk is not red." all have consequences. Baiscally, I can map colors to the reals and say arbitrary many sentences, which all have "consequences". $\endgroup$
    – Nikolaj-K
    May 28, 2012 at 15:12
  • $\begingroup$ @NickKidman: this is not always the case as in the example of certain widely accepted postulates for instance the rest mass of a photon as observing a photon at rest is not possible or measurable but somehow it is still generally accepted $\endgroup$
    – Argus
    May 28, 2012 at 20:36

I'm not an expert on philosophical terminology, but I'd go for operationalism.


Old answer: the "verification principle". The context was positivism, operationalism, or instrumentalism. The idea was that if a statement cannot be empirically verified it cannot be said to have meaning. This "principle" fell by the wayside when philosophers such as Bertrand Russell realized that the principle itself could not be empirically verified.

Recent: Please see the excellent discussion of this question in the context of string theory by Gordon Kane in the Nov. 2010 issue of Physics Today.

  • $\begingroup$ The argument "verificationalism cannot be verified" is ridiculous. Verificationalism does not need to be verified, it is a definition of what it means to have knowledge. Everything other than this needs to be verified. It's like saying "you proved this from Euclid's axioms, but how do you prove Euclid's axioms?" You don't prove an axiom, you accept an axiom. Russell was not hostile to verificationalism anyway. You are probably thinking of the elder Wittgenstein, who broke with the positivists (some would say, and "some" includes me, that he sold out positivism). $\endgroup$
    – Ron Maimon
    May 29, 2012 at 5:38
  • $\begingroup$ +1 because this is the only answer that is correct - the term is "verification principle". "Positivism" means something much more specific. Positivism had the verification principle at its core, but also included all sorts of other stuff about null hypotheses and the propensity interpretation of probability theory, all of which sounded good at the time but has now been pretty much superseded by Bayesianism in most practical applications. However, I agree with Ron that "the verification principle cannot be verified" isn't a good argument, since the principle isn't a statement about the world. $\endgroup$
    – N. Virgo
    May 31, 2012 at 9:47

Positivism, in this context, hinges on the principle that

« the meaning of a statement is the way you would verify it. »

(Or falsify it... the variations between verificationism and falsificationism are too small to warrant attention in this answer, which is trying to focus on the big picture.)

So the definition of a concept has to be laying out how you would measure it or experimentally verify it. It inspired Einstein (through Mach) and Heisenberg, but not Dirac or Schroedinger. Einstein later abandoned it, saying « one must not repeat a good joke too often ». Nowadays, most scientists would say that laying out how you would verify a concept is its operationalisation, not its meaning or definition.

To understand this principle, contrast it with Wittgenstein, who was never a positivist. Wittgenstein, as a good engineer and student of Hertz, said that

« the meaning of a proposition is the state of affairs which would hold good if the proposition were true.»

According to Weinberg in Dreams of a Final Theory, most physicists have by now abandoned positivism. According to Feynman in his Physics Lectures, positivism is false: he gives the same reasons as prof. Motl but states them even more strongly.

But where Dirac and Heisenberg agree, and some posters here, too, is that what we have learned is that if an objection to a scientific theory cannot possibly be re-cast into experimentally verifiable (and replicable) form, then the scientific theory should not be rejected simply because of that objection.

(Wienberg, Dreams of a Final Theory, p. 181: «The positivist concentration on observables like particle positions and momenta has stood in the way of a "realist" interpretation of quantum mechanics, in which the wave function is the representation of physical reality.» See also pp. 167, 304. Feynman, Lectures on Physics vol. III, p. 2-8: «the idea that we should not speak about those things which we cannot measure. ... The idea that this is what was the matter with classical theory is a false position

  • $\begingroup$ I agree with you on the numbers, but the real reason Feynman abandons positivism is because it led to Chew, Frautshci, Mandelstam physics, and he hated that stuff. The thing is, that stuff is now called string theory, and without positivism, you can't make sense of string theory. So while you are right that physicists don't say they are positivists, they are deep down inside anyway. All physicists are, and will always be verificationalist positivists period. $\endgroup$
    – Ron Maimon
    Dec 18, 2011 at 6:41
  • $\begingroup$ Also, Dirac was certainly fine with positivism, it went without saying in those days. Positivism was the normal standard philosophy until about 1975 or so. It is depressing that we've actually gone backwards a little--- philosophers sitting around a haze of marijuana smoke, saying "Dig, How can I know that the feeling of seeing red that you see is the same seeing red feeling that I see, like, in positivism, man?" What a waste. $\endgroup$
    – Ron Maimon
    Dec 18, 2011 at 6:42
  • $\begingroup$ I said Dirac was not inspired by positivism. $\endgroup$ Dec 18, 2011 at 6:48
  • $\begingroup$ I am not sure about this. Dirac followed heisenberg in the observable/matrix quantum mechanics, which was directly inspired by positivism. I agree it wasn't the major driving force behind his research, but he doesn't postulate unmeasurable quantities, as others who reject positivism have done. In particular, Feynman's rejection of positivism is based on the feeling that off-shell operators are meaningful, a feeling shared by Mandelstam and others, which is true in field theory, but false in string theory. $\endgroup$
    – Ron Maimon
    Dec 18, 2011 at 7:02
  • $\begingroup$ I am. (first note inspiration is not transitive). He was inspired by the beauty of Heisenberg's equations and the parallel with Poisson Brackets. did he not postulate unmeasurable constructs? he, famously, thought that the state vectors were real, that the Hilbert space of the Universe was genuinely, um, actual (complex), that a piece of chalk was always in a pure quantum state and could even be in a superposition of states, one localised here, and the other localised there. which is precisely what led him to postulate the collapse of the wave function as a real physical process. (I disagree.) $\endgroup$ Dec 18, 2011 at 7:21

There is a difference between to measure and to test.

If something can't be measured (by devices or the senses) it is not a scientific concept according to old positivism. Pure positivism in use was/is for instance operationalism which thus meant that a concept is meaningful only if we have some way to measure it.

The way a theory can or cannot be tested has more to do with the scientific method. According to Popper a theory is not scientific if it can't be falsified in contrast to positivists considering a theory scientific if it (among other things) can be proved true by confirmation.


A positivist will tell you only questions which can be answered in terms of my empirical sense perceptions right now can have meaning. Other questions are meaningless.

For example, is there an objective world out there going more or less as my sense perceptions seem to tell me they are, or is reality really completely different and I am on an acid trip hallucinating my current sense perceptions, or am I a computer program having a cyberdelic electric dream in a virtual reality temporarily taking on the persona of a human? Positivism will command you to reject this as a meaningless question. A mere bunch of words evoking some thought patterns in the brain! Aah, I might not be able to tell now, but in the future, when I wake up, I will be able to tell? Nonsense! Meaningless gibberish, says the positivist. There is no empirical experience of the future right now. Another meaningless question. The future is meaningless. I might as well go read a Phillip K. Dick story... except maybe I am a character in a Phillip K. Dick story I have not read (because that would spoil the script), and some advanced technology in the 22nd century allows people to have immersive first hand experiences of fictitious characters brought to life... Can I perform a test to measure the difference? No! Another meaningless question to be tossed into the trash can.

  • $\begingroup$ You are making a caricature of positivism, but it is an axiom of thought. It does not require you to say that reality is real, it does not require you to say that you are hallucinating it, it allows you to say either, and to believe both, and to translate between the two positions and see that there is no real difference between them. This is like a gauge transformation for philosophy, and the true philosophy has to be gauge invariant. The philosophy people debate is "I want realist gauge", "I want instrumentalist gauge", and this gets tired quickly. $\endgroup$
    – Ron Maimon
    May 31, 2012 at 11:09
  • $\begingroup$ "There is no empirical experience of the future right now." ??? ALL the PAST is an "empirical experience of the future" because any past moment is in the future of an infinite previous moments. By no means I can say that this moment of now, and the ones in the future are different from the ones in the past. There is no basis to argue that 'talk/project' the future is meaningless. $\endgroup$ May 31, 2012 at 23:31

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.