Popper’s Indeterminism

I read Popper’s The Open Universe for the first time yesterday. What do you think of Popper’s arguments against determinism, especially the metaphysical variety?

Although I appreciate Popper’s critique of “scientific” determinism, I confess to assuming that some kind of metaphysical determinism is true, and I found little in Popper’s words to dissuade me of that position. Unfortunately, I was unable to read the entire book. However, I plan to return to it soon, and so perhaps Popper will convince me yet.

Given Popper’s indeterminism, one issue that confuses me is the truth status of propositions about the future. Normally, Popper would say that a proposition is true when it corresponds to the facts, but what can a proposition about the future correspond to if the future facts are indetermined? Perhaps this is just a minor quibble, but it was a nagging question in the back of my mind while reading.

In any case, what do you think of Popper’s indeterminism? Have you been convinced?

About Lee Kelly

Amateur philosopher
This entry was posted in CR scholars, science. Bookmark the permalink.

17 Responses to Popper’s Indeterminism

  1. Elliot says:

    I haven’t read the book. What’s he say?

    In any case, physics is deterministic across the multiverse as far as we know. But that has nothing to do with free or morality, which have never made claims about physics.

  2. Lee Kelly says:

    Hi Elliot,

    I hadn’t expected any contibutor here but me to have not read Popper’s critique of determinism!

    In The Open universe, Popper comes out firmly against all forms of determinism, even in physics. He is a thoroughgoing indeterminist. Primarily, his critique is directed at “scientific” determinism, i.e. the assumption that, with complete theoretical and factual knowledge of any particular moment, the future can be predicted with absolute precision. For Popper, although the past and present may limit possible future events, the future is not implicit in the past.

    “Scientific” determinism is stronger than metaphysical determinism. Metaphysical determinism does not require that the future be derivable from the past (with complete theoretical and factual knowledge or otherwise), but merely claims that future events, in a sense, already exist. Since “Scientific” determinism is a stronger claim (i.e. can more easily be criticised), Popper spends the lionshare of his time, at least in the first half of the book, critiquing that.

  3. Matt says:

    Let me put down some thoughts that might easily be dismissed by someone who knows more about this than me.

    First, methodologically, Popper argues we should search for regularities. To the extent we find these, and they are true; these represent something that’s been determined. I guess you could awkwardly call this “methodological determinism”. But just because there is stuff out there that seems to be determined, does that mean that that is all there is?

    So methodological determinism is important, metaphysical determinism isn’t. (In other words, search for regularities, but don’t presume that that’s all there is.)

    The problem traditionally was thought of how do we go from a particular to a universal. That is how do we go from taking an observed instance and justify making a generalization of this. This appeared to be a really pressing and intractable problem — and still is to many people.

    So, I’m guessing materialism or determinism and so on were formulated as possible ways to help explain how we go from particular to general. I don’t think they ever managed to adequately accomplish this but generally only pushed the problem back farther and probably only complicated it further.

    Of course, Popper argues we don’t ever go from particular to general. That in a sense, it’s universals as far down as we can *see*. There are regularities out there to be seen, surly, but is that all there is? Why would anyone want to presume that?

    What is the problem that metaphysical determinism answers?

    It’s been a long while since I’ve read _The Open Universe_ but I guess one way to look at it might be like this; *metaphysical* determinism doesn’t readily solve any problem we are really concerned with, and not only that, here are a list of trivial (or not so trivial) problems we *create* by presuming it.

    I mean, it’s fine to have problems, but do we want to create them needlessly?

  4. Matt says:

    >>Normally, Popper would say that a proposition is true when it corresponds to the facts, but what can a proposition about the future correspond to if the future facts are indetermined?<< Is Popper saying *all* the facts are indetermined or only that some of them might be?

  5. Elliot says:


    Determinism is the theory that the laws of physics, given some initial conditions, always take it to the same state/result (say, 1 second later). You’ll have to imagine you could set up the same initial conditions again.

    Another way to state this is you could perfectly predict the future, given perfect knowledge of the present, perfect knowledge of the laws of physics, and unlimited computational resources.

    Or in other words, determinism says the laws of physics contain no randomness (or magic, or as-yet-unimagined other possibility).

    Determinism should not be put forward as a theory in its own right. It’s just as aspect of the laws of physics. We can examine the laws of physics and note if they do anything random or not.

    The current best understanding of physics is that it is deterministic (across the whole multiverse — events can look random if you don’t see the whole picture). That is an issue for physicists, not philosophers.

    There is a second theory, also called determinism, which states, roughly, that people don’t make choices or have free will (or that morality doesn’t exist, which is about the same). This is sometimes confused with the first type of determinism, or thought to follow from it. This second theory is an issue for philosophers, as is the claim that it follows from the first determinism.

    Sometimes people say we have fate or destiny and are ambiguous about whether they mean the first or second type of determinism, or both.

    I don’t find the second type of determinism very interesting b/c I don’t believe any decent argument for it has ever been given.

    I don’t see how going from particulars to universals, which is related to induction, comes into it.

  6. Matt says:


    >>I don’t see how going from particulars to universals, which is related to induction, comes into it.< < I agree what I stated could easily be in error, at some point I'll try to explain this more fully so that the idea can be better criticized. >>Determinism is the theory that the laws of physics, given some initial conditions, always take it to the same state/result (say, 1 second later). You’ll have to imagine you could set up the same initial conditions again.<< To the extent we are speculating about the empirical world, isn't what you are saying true about our expectations for *any* theory, not just physics. In what cases would we not expect to be able to take our theory plus initial conditions and arrive at the same results? (And then still regard the theory as true ...) Aren't you just describing Popper's necessary conditions for a theory to be falsifiable?

  7. Elliot says:

    Consider the theory that when two photons collide there are two ways they can scatter, and which of the two happens is random with a 50% chance for each (the theory specifies both ways based on the trajectories the instant before the collision). That is falsifiable. And it’s indeterministic.

    I don’t see anything wrong with that theory, except that it contradicts what we know about physics. (Photons can’t collide with each other at all).

  8. Alan Forrester says:

    I think that most scientists and philosophers would admit that the future is impossible to predict for most practical purposes. Some might think it would be possible to predict it in principle. I think Popper was right to pour cold water all over that idea and the way he did so was basically correct.

    As for his metaphysical arguments. I don’t see how metaphysical indeterminism helps us to understand free will or morality. Also, there are lots of reasons to be uneasy about metaphysical indeterminism in physics. Quantum mechanics is commonly said to be indeterministic. I think this idea is badly mistaken and a lot more costly than many people seem to realise. If what happens after a measurement is that all of the possible outcomes are weeded out except one, then there are correlations among quantum systems that cannot be explained by any local influence from one system to another (EPR experiments) and on top of that whatever influence tells another system what to do also cannot be Lorentz invariant (Hardy experiments). So any theory that states that one of the possible outcomes happens is incompatible with basic principles of both special and general relativity.



    The Everett interpretation, according to which the wave function describes how the world really behaves can be made both local and Lorentz invariant:


    Finally, it seems to me that most indeterministic interpretations of quantum mechanics, including the statistical interpretation, are a lot vaguer than the Everett interpretation. In the Everett interpretation, you have one equation of motion for a system and you can say when one system becomes correlated with another, when the wave function acts in a manner that could give rise to probabilities and so on. In an indeterministic interpretation you have the Schrodinger equation or something and the collapse postulate and a very vague set of rules about when to use one or the other. A noble exception to this vague flummery is the Ghirardi-Rimini-Weber theory, which, although it suffers from all of the problems with non-locality and so on, at least gives a specific set of predictions.

  9. Lee Kelly says:


    In my opinion, metaphysical determinism makes sense of propositions about the future (including universal propositions). For example, if the proposition “every swan is white” is true, then what is entailed about swans that have not been born yet? Does the universality of that proposition extend to all places at all times, or merely our particular frame of reference. If “every swan is white” does entail that unborn swans will be white, then how could it possibly be true in an indeterministic world? Since that portion of its logical content about the future cannot possibly correspond to the facts, in what sense can it be true? However, if metaphysical determinism is true, then talking about events that haven’t happened yet (from a particular point of reference) doesn’t run into any difficulty.

    In any case, metaphysical determinism does not preclude the appearence of indeterminism, i.e. there could still be events which are scientifically unpredictable. I also like and agree with your comment regarding methodological determinism.

  10. Peter D Jones says:

    Does Popper really have an argument for indeterminism? Or just against determinism?

  11. Pingback: Elliott Wall — | There is no good reason whatsoever for the belief in free-will

  12. Elliott Wall says:

    Hi— I saw that my site was linked to, then I stop by to see what’s going on and I was alarmed to see that (one of) my doppelgänger(s) was on here arguing about one of my favorite topics.

    Though the subject may be long settled by now as I’m coming quite late into the discussion, I wanted to say that I think it probably isn’t necessary to prove determinism in order to disprove the existence of free-will. I fully accept that the universe may be entirely random! In fact it’s pretty likely.

    Yes, Popper gives arguments for indeterminsim, but his arguments are epistemological in nature, in that he basically only proves (to my satisfaction) that there is no rigorous way to prove determinism. He doesn’t manage to rescue free-will though. Somewhere in a foot-note in tOU he says that he isn’t making an attempt to contend with logical determinism.

    Excellent website. It’s very flattering to me to be linked to; I’m self-educated and everyone here is so bright, thank you.

  13. Elliott Wall says:

    Ah, it’s me again. I sound like a big twat in my previous comment and I was wondering if you could delete that last sentence! PLEASE! I’m in total earnest but it never quite sounds right on the internet unfortunately. Cheers

  14. Frank Lovell says:

    I’m about 3 years late, my just now having discovered this thread of dialogue (sorry ’bout that). I want to add a few thoughts from one who has come to embrace Popper’s philosophy of science and objective knowledge and critical rationalism from the educational and professional pathway and traditions of empirical science (rather than from the educational and professional pathway and traditions of philosophy).

    Elliot said, “…physics is deterministic across the multiverse as far as we know…,” and later, “Determinism is the theory that the laws of physics, given some initial conditions, always take it to the same state/result (say, 1 second later). You’ll have to imagine you could set up the same initial conditions again.” And Matt said, “…Of course, Popper argues we don’t ever go from particular to general.” Well, we certainly SHOULD not ever go CONCLUSIVELY (as distinct from conjecturally or tentatively) from the particular to the general , for we lack justification for doing so, yet an awful lot of people (particularly theists and even some philosophers) still seem powerfully want to do so (errantly, of course).

    But think for a moment what physics (sort of) tells us in a case of (sort of) going from a (sort of) general to a (sort of) specific, namely:

    Physics at the MACRO level (large molecular ensembles and “upwards” through specks of dust and grains of sand and marbles and billiard balls and boulders and moons and planets and stars and galaxies and galactic clusters) appears to be deterministic — indeed it appears to be rigorously (inexorably) mechanically deterministic, such that if one could repeat any macro-physical event by resetting ALL of the initial physical conditions identically it would be perverse to expect DIFFERENT outcomes.

    Physics at the MICRO level (individual or very teensy ensembles of molecules, atoms, neutrons, protons, electrons, and “downward”) appears to be increasingly indeterminate, and there are events at the macro-level for which, if repeated by resetting ALL of the initial physical conditions identically, it would be perverse to expect the precise SAME outcomes.

    For one example (there are others), if we start with a macro-quantity (say, one mole or gram-atomic weight, which is indeed a large ensemble of about 6×10^23 atoms) of an unstable (radioactively decaying) element, say, a 226 gram chunk of solid radium-226, we know that half of that mass will have decayed to radon gas (leaving 113 grams of radium) after 1,601 years. And we reasonably expect that THAT will be the SAME outcome we obtain if we repeat that event time and time again starting with the precise same (or, as it happens, even surprisingly variable) initial conditions.

    BUT, if we start with one single atom of radium-226, we have NO IDEA when it will decay into radon – it might decay to radon by noon tomorrow or still be radium when in the far future our sun passes though the red giant stage on it’s way to white dwarfism (it MIGHT be a GOOD bet that it will not still be radium in 16,010 years, but it will not be a SURE bet that it will not still be radium in 5 billion years). And if we could repeat that event with ALL the same initial conditions, we have thus far discovered NO empirical reason to expect the same time-to-decay outcome ever again. At the MICRO level, physics appears to NOT be rigorously mechanically determinate — quantum uncertainty appears to run clean-to-the-bone (or said another way, it appears that not even Mother Nature knows EVERYthing).

    My point with this: it is not strictly correct that physics (the dynamics of mass/energy in space-time) is rigorously determinate throughout; physics appears to be solidly determinate at the MACRO-level of physical events, but significantly INdeterminate at the MICRO-level of physical events. And there does not appear any clear-cut or sharp line between where strict/inexorable mechanically determined physics prevails and where it does not prevail

    For this reason I (for one) presently have no good reason to think that the past rigorously determines all of the future (though it might limit or rule-out the possibility of outcomes, as Popper conjectured); I recognize that in science and philosophy, legions disagree with me — there is almost always found somewhere on the Internet a raging debate about whether or not any degree of genuine “free will” volition is possible in the (predominantly) determinate physical world.

    Also, even in a NON-deterministic world there could still be statements which have epistemic value – probabilities are not AS usefully valuable as certainties but probabilities ARE usefully valuable. And while many EVENTS in the micro-physical and macro-physical realms are chance/random events, there appears to be VERY few PROCESSES (especially in the macro-physical realm) that are chance/random processes (even chemical bonding, which occurs at the atomic level, is an example of a NON-chance/NON-random PROCESS; gene mutations appear to be random/chance events but natural selection is a NON-chance/NON-random PROCESS).

    Still, I do find very useful concept of methodological determinism (which I have always named “physical determinism”) that prevails at the macro-level of physics but which gives way to a probabilistic(/”statistical”) determinism at the micro-level of physics. The concept of metaphysical determinism I have not myself yet found any use for (or problems which are solved by).

    I could be wrong (and if I am wrong I welcome critical arguments for concluding that I am wrong), but presently I think “we don’t presently know” and even “we cannot ever know” are perfectly legitimate epistemic states which do not imply or impute or insinuate or entail or necessitate trivalent logic concerning propositions about future events – and so, for the present, I do not think much about or of metaphysical determinism. (Maybe I should.)

    There is MUCH more that COULD be said – and volumes and volumes more have indeed already been said and published pro and con, to and fro, on the presently unsettled issues in both physics and philosophy of determinism/indeterminism and on the issue of whether “free will” cognitive volition is genuine at least to some degree or is only realistic illusion to any and every degree – but I can’t think of anything else that I could BRIEFLY add that might put a dent in that raging controversy which frequently surfaces on the Internet, so here I’ll put a cork in it. (I probably shoulda corked it up 7 or 8 paragraphs ago.)

  15. Frank,

    FYI here is what’s going on:

    You’re making physics claims which are incompatible with MWI. You’re disagreeing with me. But you do not post any refutation of MWI.

    Criticisms of everything other than MWI are readily available (e.g. see material by David Deutsch), but no quality/correct criticisms of MWI exist as far as I know.

  16. Lee Kelly says:

    I wish to clear one thing up: this was never a post about free will. In my view, it’s mistaken to assume that determinism contradicts free will. What people care about is not whether (or to what extent) the universe is determined, but rather what is doing the determining. Are our choices determined by physical events, or are physical events determined by our choices? Are our choices the explicandum or the explicans, the cause or the effect, the input or the output, the culprit or the victim? Materialists and reductionists see responsibility flowing from physical events to choices, while dualists and holists see it flowing from choices to physical events. Each explains the same phenomena from a different level of emergence, but also tries to deny the causal potency or reality of the other. Each implicitly agrees that one or the other is the “real” cause, and that moral responsibility hinges on which. Ugh.

    Anyway, from what I understand, the multiversal interpretation of quantum mechanics, which David Deutsch has convinced me is the best explanation we currently have, is actually deterministic. It doesn’t, however, mean we can predict the decay of a single atom of radium-226, no matter how precise we measure the initial conditions, because every possible universe will be realised and “we” will have copies in each.

  17. Bruce Caithness says:

    “Emergence” as the preceding posts highlight is critical. Methodological behaviourism tries to demand that researchers confine themselves to the study of behaviour as measured by a disinterested observer. This is analogous to the mathematisation of physics. There are situations when a behaviourist model is applicable but not when imaginative and critical aspects of human cognition are to be considered. When a person is in a coma a different level of freedom or lack of it exists compared with when awake.

    Life is an emergent property of physical bodies and knowledge is an emergent property of life, world two emerges from world one and world three from world two.

    One can conceive of punctuated equilibria, or umwelt boundaries between “objects” existing at the speed of light, sub-atomic umwelts, physical matter, biological organisms, human intentionality.

Leave a Reply

Your email address will not be published. Required fields are marked *

please answer (required): * Time limit is exhausted. Please reload the CAPTCHA.