Radium atoms decay by emitting alpha particles at random. Are these events without cause?

Of course one may take a closer look at radium nuclei to determine a possible reason why they decay; for example a hidden-variables model of QM may provide an explanation.

But the essential question I'm driving at is this: if random events occur in nature are they then by definition without cause? (The above example being simply a concrete illustration).

example: If we find a particular atom decays, and we later find out that its a Radium atom - we could say it decays because it is a Radium atom. In fact this is a tautological statement. We still haven't found the cause.


If an event is determined by its circumstances, only that event occurs and no other. If its undetermined by circumstances, then given something must happen, a range of possible events present themselves. Either all occurs, or one of these possible events occur and this choice must be taken at random, ( for if not, it is determined, and we're not interested in this case here).

An epistemologicaly random event is one which has only the appearance of randomness, given more precise knowledge of the events circumstances, one can determine why this event occurs as opposed to another. For example, a dice throw given Newtonian Mechanics is such an event. If one could precisely know how the dice is thrown and thus model its trajectory one can explain why a five occurs rather than a six.

An ontologically random event is one which knowledge of circumstances is not a barrier; it is random by fiat. In principle completely precise knowledge of circumstances will not enable prediction in a precise sense (a statistical prediction is still possible).

  • Lightnings fall randomly through the most conductive paths, that depend on the random placement of the particles in the air. What is random? What is a cause?
    – Trylks
    Commented Sep 16, 2013 at 21:37
  • 1
    According to Kant, applying reason beyond the limits of sense experience is an error (transcendental dialectic: the logic of illusion). Causality and randomness are macroscopic empirical intuitions, which should not be used to interpret the microscopic universe, which, as far as we know, does not hold for causality or randomness. Macroscopically, Newton's first law says it all: there are no causeless or random. The appearance of randomness would just be complexity. Microscopically, (QM, alpha particles, etc.): nothing to discuss, that's beyond sense experience.
    – RodolfoAP
    Commented Dec 10, 2021 at 20:14
  • @RodolfoAP: Simply throwing around philosophical jargon around doesn't mean very much. Given that our senses habe bern extended by instruments you imvocation of Kant is pointless. The notion that chance is a cause is already in Aristotle's Metaphysics and so was discussed then. Moreover, your understanding of Newtons laws leaves something to be desired. His laws allows deterministic chaos. But none of this is at all relevant to the question, it's merely answering your vacuous charges. Commented Dec 12, 2021 at 9:42
  • @MoziburUllah Sorry for the jargon, was not the intention. The comment do has a meaning. Otherwise, you can apply the intuition of causality to the atomic behavior. Regarding Newton, the first law essentially means that no effect can be produced if there is no related cause. That is the direct answer to your question: macroscopically, there is no "causeless", and "randomness" is just systemic complexity, or as you call it, "deterministic chaos".
    – RodolfoAP
    Commented Dec 13, 2021 at 15:58
  • 9 answers so far, but not a single correct one. (1) Every random process must have a source for its randomness. There cannot be randomness without a source. (2) The popular notion of "wave-function collapse" is just a pretty fiction. I have met many physicists who refuse to face the incoherence of this notion and instead just say "shut up and calculate". (3) Why do radioactive atoms have lifespan that appears to be drawn from a probability distribution? It is simply because the wave-functions describing different atoms are all slightly different, and the environment amplifies such differences.
    – user21820
    Commented Dec 15, 2021 at 21:40

9 Answers 9


Randomness and causation are in different categories. Something can be both random and caused, or random and uncaused (if you believe in such things). Randomness is not a property of origin (cause) but of comprehension (understanding the origin).

Random can mean simply "unpredictable", or "of or characterizing a process of selection in which each item of a set has an equal probability of being chosen." The roll of a die is sufficiently complex such that the outcome is effectively random, but that doesn't mean the outcome was uncaused. It just means the cause was not reasonably predictable with human faculties alone.

EDIT: Response to the updated question
I see how you are defining randomness now, but I can't conceive of any examples in practice because I can't conceive of a universe in which such randomness exists. Radioactive decay is, in my opinion, only stochastic because our science hasn't figured the causal chain yet. I hold that this is more likely than the opposite case because we already have innumerable examples of causation and non-randomness yet we have no examples of the latter. Occam's razor compels me to chose the one that rests upon the least new assumptions.

That said, I can vaguely conceive of a world like you describe. In this world, for something to be truly random, it cannot be part of any causal chain. This suggests to me the answer to your question is yes, ontologically random events are causeless. But this is all very counter-intuitive; we tend to not like the idea that something can come from nothing...

  • I'm not asking about epistemological randomness but ontological. "but, that doesn't mean the outcome was uncaused". You're using an implicit assumption that nature is determinstic; this is something that I specifically said I'm not assuming in the question. Commented Sep 14, 2013 at 13:58
  • Could you explain the distinction more clearly in your post? I'm not sure I understand the difference.
    – stoicfury
    Commented Sep 14, 2013 at 20:30
  • I've edited the question to try and bring out the differences. Commented Sep 16, 2013 at 22:24
  • @stoicfury - Right now there is something and if it did not come from nothing then something has always existed (something is eternal). When something has always existed, then there is no first cause and no finite description how or why everything came to be. Our modern views have an internal disagreement - on one hand, we assert that something cannot come from nothing, on the other hand we assert everything must have a beginning but that contradicts with the logical consequences of the first assertion.
    – Saul
    Commented Sep 17, 2013 at 8:59
  • @Saul - Yes, that is the one area we still don't really understand yet. But everything else, turtles, stars, corndogs, pencils.... these things all have causes. I can grant we don't know what the deal is with the very first cause but still rationally hold that everything else does indeed seem to be caused. To me, it's still more rational to believe this than to believe in uncaused events, purely from the number of examples we have in reality alone.
    – stoicfury
    Commented Sep 17, 2013 at 21:54

Here is an answer from István Zachar on the physics.SE site which seems to directly answer your question:

It was John Stuart Bell in 1964 who proved by simple arithmetics that there are no hidden local variables behind the statistical nature of quantum processes, and behind the spooky non-locality displayed by entangled particles. Consequently, the paradox presented in the 1935 Einstein-Podolsky-Rosen paper upon which they claimed that quantum physics cannot be complete ("since it relies on statistical laws, it cannot give the ultimate full description of nature") is inherently wrong.

We understand causality as a relation that links post-events (effect) to prior-events (cause) (note that this does not necessarily mean similar chronological sequence, see here). In this sense, observable phenomena are dependent on deeper, possibly hidden variables, that nevertheless can be usually uncovered, at least at the macroscopic level. However, as Bell has proven, there are no hidden variables responsible for lowest-level quantum processes e.g. the random decay of radioactive elements. Therefore I would say that there exist no lower-level, ultimate cause for these processes.

  • 2
    If this is solely a quote from another site (and I see that it is), this is not an answer. You could have just posted the link in a comment.
    – iphigenie
    Commented Sep 16, 2013 at 10:02
  • 2
    @iphigenie - I quoted it simply for the sake of convenience.
    – Saul
    Commented Sep 16, 2013 at 11:09
  • 1
    For the record, to my knowledge we have no specific policies on this (reposting someone else's answer from another site). In this case, it's clear that Saul is not trying to make any claim that the answer is his and is correctly referencing it as an answer from elsewhere, so I think it's fine. That is to say, this answer is useful and does not violate any known policies (again, to my knowledge). :)
    – stoicfury
    Commented Sep 17, 2013 at 20:27
  • 1
    @Saul: the point is that if we insist on locality, we may sacrifice precision of description. Some systems appear noisy until you expand your view of them.
    – labreuer
    Commented Oct 15, 2013 at 0:11
  • 1
    @labreuer - Your present definition of local is different and narrower than it can be in quantum physics. However, in other interpretations, the experiments that demonstrate the apparent non-locality can also be described in local terms: If each distant observer regards the other as a quantum system, communication between the two must then be treated as a measurement process, and this communication is strictly local. Source
    – Saul
    Commented Oct 16, 2013 at 7:18

The answer depends on how you define causation. The question 'are random events causeless?' is the result of conflating causation with necessitation. If 'to cause' simply means 'to bring about', then the problem is dissolved. Whether or not the events are random has no bearing on whether or not they have a cause. If the random events are brought about by something, then they are not causeless.

"Causality consists in the derivativeness of the effect from the cause. This is the core, the common feature, of causality in all its kinds” and "Causation, then, is not to be identified with necessitation". These are two quotes (p. 136) from G.E.M. Anscombe's paper ‘Causality and Determination’, The Collected papers of G.E.M. Anscombe Vol. II Metaphysics and the Philosophy of Mind (1981), pp. 133-147, Blackwell). This excellent paper is a great critique of theories of causation which identify causation with necessitation.



The intuition (as given in one response) that this question involves a distinction between an objective 'origin' and a subjective 'comprehension' is obviously correct, but making such a distinction does not answer the question so much as identify what makes it difficult.

It is true that when we speak about randomness in observed results, we are speaking about limits to our own understanding. However, we dismiss what this question is really asking about if we simply say that everything is caused in reality, and that therefore what we experience as random could and would be experienced as non-random if our faculties were sufficiently powerful.

The question was posed about "actually random events."

This question is related to the old question: is it the case that everything is caused, and do we mean by that that someone or some thing with sufficient faculties could accurately predict the location and qualities of every particle in the universe at some arbitrary date in the future? The current problem begins by assuming up front that the answer to this question is 'no.'


They key to this whole set of related questions is making them even more difficult: just like randomness, the concept of a 'cause' is also an element of our understanding, and therefore a 'cause' is not something straightforwardly real or 'actual.' (The use of the word 'origin' as a substitute for 'cause' tries and fails to avoid this problem by connoting the object which presents itself to us.) The important idea is that "causes and effects are merely a semiotics of the results" of a real process which is more complex and which exists independently of schemas that try to represent it with concepts like 'cause' and 'chance.'

From this perspective, I take the question to be one about (a) the existence of randomness in nature, (b) how we might detect randomness in nature, and (c) whether observed randomness is an observation of the absence of causes.

Limited answer.

My own intuition about these things is that we necessarily experience the world in terms of causes and effects, and that observed randomness is a way of experiencing orders of reality which are not and cannot be experienced directly. They are causeless in the sense that they are not experienced.

There does appear to be a trade-off between how randomly observed variables behave and our ability to explain them in terms of causality. But this speaks to deeper epistemological problems, rather than simply identifying a way of dividing phenomena into two classes: caused and un-caused.


I'd say this is at heart a semantic question.

Events may happen due to two distinct preconditions: either that event is possible, or it is necessarily going to happen. In the latter case, whatever made it necessary is the cause. In the former case, one could say its possibility of occurence is its cause. Why did it happen? Because it could.

Others may add a conditional to causation however: something only has a cause if happened due to logical necessity.

Most dictionaries I've checked either define cause as the reason something happens, or the person or thing that makes something happen. So, the question one must thus ask is this:

Is a possibility a thing and/or a reason?

  • Some might argue your first claim applies to all of philosophy. :D
    – J D
    Commented Dec 9, 2021 at 18:12
  • 1
    @JD I would too!
    – A. Kvåle
    Commented Dec 9, 2021 at 19:26

To enlarge slightly upon Gnasher's answer:

We know why radium nuclei decay: the radium has too many protons in it for it to be stable. We also know why the exact time any given radium nucleus will spit out an alpha particle is unpredictable: the ejection process involves statistics at the quantum level. Those statistics yield only an average time for the ejection to occur based on an extremely large number of actual ejections, and yields no answer for a single ejection event.

In this sense, radium has a well-known cause for decaying and a well-known reason for being in principle unpredictable. This means it may not be a suitable example for you to ponder in a philosophical context.


Actually, there are no completely random events at all, all events are partially random. All events have a cause, but the cause never determines its effect with absolute accuracy. Ontological (=true) randomness is the observed inaccuracy in all effects.

Dice-rolling results are ontologically random, because even if we know the cause (a hand throws the dice), the person throwing cannot decide the result.

Ontological randomness is not about unpredictability, epistemological randomness can be equally unpredictable. The difference is that ontologically random outcomes are not deliberately decided.

Epistemologically random outcomes are also known as pseudo-random (=fake random) outcomes. They are actually deliberately decided by someone (or products of an algorithm chosen by someone) for the purpose of creating a false impression of randomness.

To summarize: Random is not causeless. Random is purposeless.


I see 2 main problems limiting an answer to your question:

  1. In that standard QM does not "clearly and forthrightly address two fundamental questions: what there is, and what it does", there is no ontology. Without any ontology, we can't say if there is ontological randomness. https://ndpr.nd.edu/reviews/philosophy-of-physics-quantum-theory/

  2. Events (your usage) are within the "ontology" of general relativity. QM and GR do not agree on an ontology, nor many other things. The closest corollary in QM to an event is measurement, so imprecise it "[has] no place in a formulation with any pretension to physical precision: system, apparatus, environment, microscopic, macroscopic, reversible, irreversible, observable, information, measurement." Bell, J. S. (2004). Speakable and Unspeakable in Quantum Mechanics: Collected Papers on Quantum Philosophy. United Kingdom: Cambridge University Press. pg.215

Without a direct answer, there can still be some insight from QM. In standard QM, there is no measurement/observation without disturbance. Even the radium atom is disturbed. But disturbances (measurements), while not "random", force us to "revise our ideas of causality..."

"...Causality applies only to a system which is left undisturbed. If a system is small, we cannot observe it without producing a serious disturbance and hence we cannot expect to find any causal connexion between the results of our observations. Causality will still be assumed to apply to undisturbed systems". Dirac, P.A.M. (1991). The Principles of Quantum Mechanics, Fourth Ed. pg. 4

In summary, what you are asking, "if random events occur in nature are they then by definition without cause" is not within the domain of standard quantum mechanics because standard QM does not have an ontology of events, nor any other ontology. Given that standard QM (and by extension QFT) are our most leading-edge scientific theories, science has no answer to your question. The question you are asking is thus entirely philosophical, and there is no agreed upon answer. See Robert Nozick as someone who was comfortable with true ontological randomness, but most aren't.


"Radium atoms decay by emitting alpha particles at random. Are these events without cause?"

Radium atoms decay at an unpredictable time. Each and every radium atom will decay within the next second with the same fixed probability. If it doesn't decay in the next second, then in the second after, with the same probability. There's no effect that "old" radium atoms decay quickly or slower.

There is no reason for the atom decaying at the time when it decays. There is a cause for it decaying eventually: It is in the nature of a radium atom to decay at some point in time. It's in the nature of a radium-226 atom to decay with a probability of 0.5 in the next 1600 years, and with a probability of 0.25 in the 1600 years after and so on. You can call this "the cause" of the statistical behaviour. But the exact time of decay - there is no cause for that.

Regarding comments: An atom must be decay at some point in time in the same sense as throwing a fair dice repeatedly must land on a six eventually. The chance that it doesn’t decay in the next two million years is the same as winning the lottery every week with a single ticket for a whole year. But if it survives for two million years, it’s chances to last longer are exactly the same as for a “brand new” atom.

  • "It is in the nature of a radium atom to decay at some point in time." Not a physicists, so I can't rule out the possibility that perhaps element have a "after x amount of time, the element MUST decay". However, if the aforementioned possibility is not true, then your statement is not true, unless your "in time" refers to an infinite amount of time. If we're using a logical framework, any event with a non-zero probability will happen certainly happen within an infinite amount of time, proven by $\lim_{n\rightarrow \infty} \frac ab^n = 0$, a/b being the probability the event doesn't happen.
    – A. Kvåle
    Commented Dec 9, 2021 at 19:32

You must log in to answer this question.

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