In philosophy of mind, epiphenomenalism is a form of dualism (the view that the body and the soul are made of two different substances), where it it said that the physical fully dictates the constitution and behavior of the mental, but not vice versa (the other way around).

One rough analogy would be that between fire and smoke. The configuration of the smoke is dictated by the configuration of the fire, but the configuration of fire is not dictated by the configuration of smoke. However, this doesn't hold up since the fire will be affected by the smoke (as long as the fire in question is within the smokes light cone), although to a very (very, very) small degree. So, this can't be a case of genuine epiphenomenalism.

My question is: Are there, or could there theoretically be, any confirmed cases of epiphenomenalism (not necessarily having anything to do with philosophy of mind)?

Here, a tentative (more or less), formalized definition of epiphenomenalism would be:

Epiphenomenalism := when the state and/or postion of an object a, at any point in time at or after time t (at which b gets affected by a), is different from the state and/or position it would have if b wasn't affected by a's action

To me, the statement that there could be confirmed cases of epiphenomenalism sounds like a contradiction. If x affects y (but not the other way around), an observer confirming the changes of y would have to be isolated from x (as in not part of the universe since everything affects everything to a very small degree), otherwise there would be a causal chain from y to x.

One example I can think of is that of quantum entanglement. Let's say that particle a and b are entangled and that we keep particle a on earth and transport particle b to the sun. If I were to measure particle a, particle b would instantaneously be affected, right? (Sorry if my knowledge in physics lets me down here. I don't know if it is correct to say that particle a affects particle b when a measurement is made on particle a, this question would suggest it isn't but in that case we can redefine quantum entanglement, so that it works in the way we want it to, and think of it as a made up concept that still is relevant to the question of the theoretical possibility of ever detecting epiphenomenalism). Also, particle b wouldn't affect particle a back. This would then seem like a case of epiphenomenalism. However, the epiphenomenalism would only be temporally true since after 8 minutes (the time it takes to travel between the sun and earth with the speed of light), particle a would lie within the light cone of particle b in its new, measured state, right?

  • 1
    Past and future? Aug 2 '11 at 21:34
  • 2
    Sorry, could you expand your comment? I'm not sure I understand what you're asking.
    – Speldosa
    Aug 2 '11 at 22:17
  • 1
    Apologies, was wondering if, by the definition you were using, the future is an epiphenomenon of the past. Aug 2 '11 at 22:45
  • Oh, I see. Yes, that would qualify. Now, I just have to think about what consequences that has for my question :) Give me some time to think.
    – Speldosa
    Aug 2 '11 at 23:08
  • 1
    This seems peculiar to me too. I am familiar with Stanford's definision: "Epiphenomenalism is the view that mental events are caused by physical events in the brain, but have no effects upon any physical events." I suspect the reason there is no such definition for physical objects is that cause and effect themselves are difficult to distinguish, Newton's 3rd being what it is. Aug 13 '11 at 1:24

Physical objects produce shadows, but shadows (as the absence of something, i.e. light) do not affect the producing object.

  • I would say that they do. If the ground would have received light (that is, the shadow isn't cast), then this would have affected the object in another way then when the shadow actually is cast.
    – Speldosa
    Aug 3 '11 at 18:18
  • Yes, you are correct, but this says nothing about my answer. :P If the ground would have received the light, then it would have affected it in a different way. But my example is that ground didn't receive the light, and through this absence of light the producer was not affected in any way. So your statement is mutually exclusive to mine. In other words, the absence of a physical thing cannot be said to be the cause of something else. That said, I can think of ways in which my answer would not work. Say, if the object was a person, and they were afraid of/spooked by their own shadow.
    – stoicfury
    Aug 3 '11 at 19:26
  • This seems to be a question of how you should look upon causation. In my view, we have to look upon it as if a certain concept is difference-making compared to if something else would have been the case. Also, to claim that a shadow is something is problematic, since its, by definition, is the absence of something. Also, the fact that the shadow can create a concept in our head (which we then for example get spooked by) shows that there is a causal connection from the shadow back to the object (how else could the object detect this fact?).
    – Speldosa
    Aug 4 '11 at 11:01

If we assume that:

-the brain is the substrate of the mind.
-the brain is an object in the physical world and is only affected by causal chains that are constrained to that physical world.
-something epiphenomenal, by definition, is something outside the physical world and or something that does not affect the physical world.

Then it seems that you could never obtain evidence of the existence of anything epiphenomenal.

Also, you could always doubt epiphenomenalism on purely epistemological grounds (how do I know what I think I know?)

On the other hand, you could imagine a circumstance where we started with the epiphenomenal, and then the physical world was somehow contrivedly created to coherently host physical mirrors of minds (brains). Then, at some point, by examining the physical world, we might find clues that suggest it was contrived for that purpose. Far fetched yes, but, perhaps, never say never.

  • Fine, but this is really an answer to why epiphenomenalism in philosophy of mind doesn't work. What I'm after is an general refutation (or a positive answer) of something detectable epiphenomenal in any domain.
    – Speldosa
    Aug 5 '11 at 13:21

Your claim as to the impossibility of your 'physical affecting physical' definition of epiphenomalism is a direct consequence of the Copenhagen interpretation of quantum mechanics. But the same cannot, I think, be said for certain when the epiphenomenon is allowed to be non-physical (See my comments on stoicfury's answer).

Edit: The copenhagen interpretation of quantum mechanics states that the effective uncertainty of observables is entirely a consequence of our inability to measure such observables without altering the state of the system.

If we assume that particles obey the laws of quantum mechanics (as your question must do, if it is to get an answer) then this uncertainty is endemic- and thus so (if we also accept the copenhagen interpretation) must be the problem of measurement without alteration.

If a quantum mechanical system is affected by another, then the former measures the latter in some way and so alters it.

  • Sorry, I don't really see how it is a consequence of the Copenhagen interpretation. Could I ask you (once again, sorry :P) to expand a little?
    – Speldosa
    Aug 5 '11 at 10:40
  • Done!........:) Aug 5 '11 at 11:07
  • Hmm...But even if one could measure without alterate, some information would escape from the particle, right? And that would be enough to causate back.
    – Speldosa
    Aug 5 '11 at 14:14

Despite the quantum physics and the philosophy of the mind part of it, I find really hard to believe that A can affect B while B does not affect A at all.

If we take time out of causality and we speak in a pure theoretical metaphysical way, then A affecting B could we thought as A => B. This is equivalent to ¬B => ¬A, which, by the way that the implication can be read, that B is, in a way, affecting A.

If we abstract it even a little more, we could say that the direction of causality is subject to the arrow of time direction and to the perception of what is actually the cause and what is actually the effect of that cause. At such a point, I would say that a one-direction-isolation of causality is impossible, which would actually discard epiphenomenalism as even theoretically possible.


You can think about the answer to this question on a level far more rudimentary than quantum physics. To say that the indeterminate (seemingly random) state of a particle at the subatomic level grants human beings a method by which they can span the gap between the physical and the mental is nonsense. You couldn't even begin to make such a claim. Dualist theories of mind are flimsy for this very basic reason: there is an explanatory gap between how the physical interacts with the non-physical. Epiphenomenalism takes this to a whole new level: somehow not only does the interaction occur, but it can only go one way. So now not only are you denying the laws of planar interaction, but also causality too.

Also, even if you find an analogy that works with your definition, I'm not sure what you think that will get you. Epiphenomenalism as a theory of mind is still bounded by the same basic constraints, and an analogy alone isn't going to liberate it.

  • 1
    I disagree with the premise that epiphenomenalism (as I understand it, not as it is defined here) is 'a form of dualism' in the sense that it suffers from the same problems as dualism proper. The non-physical can easily be affected by the physical. If I build a machine that makes a mark at one place on a tape if, say, it is a sunny day, and at another if it isn't, the meaning of the mark on my tape at the end of the day is a non-physical thing depending entirely on the physical. Aug 3 '11 at 7:51
  • First, this answer doesn't really answer my question but rather talks about what I could do with my example. That's pretty much another topic altogether. Second, the reason I ask this question is that I find epiphenomenalism (in dualism) an unlogical position. As soon as anyone would have knowledge of a part of a epiphenomenal process and state this in the world where the origin of the epiphenomen took place (in the case of dualism; the world where the brain is, that is, the physical world) it would be a contradiction. So I'm on your side regarding epiphenomenalism in theory of mind.
    – Speldosa
    Aug 3 '11 at 11:07
  • Tom: How do you define the "mental" here? Without knowing, it sounds like you have some kind of eliminitavistic view on it.
    – Speldosa
    Aug 3 '11 at 11:11
  • Also, stoicfury, I never said that this is any theory of mine of how dualism works (quantum entanglement between physical particles). This is purely a search for anything epiphenomenalistic in the world (which I don't think can be find).
    – Speldosa
    Aug 3 '11 at 11:13
  • @Tom I respectfully disagree. In fact, I think your analogy only works with epiphenomenalism as defined here, not with epiphenomenalism proper ("EP"). EP is clearly a dualist-type theory which posits the existence of two distinct substances (mind and body). One is located in and bound by the physical universe and laws (e.g. our brain). The other exists in some other "realm"/universe, which is typically described as non-physical (e.g. our mind). Your example suggests that mark on the tape is directly proportional to the physical as the meaning of the mark on the tape is to the nonphysical.
    – stoicfury
    Aug 3 '11 at 17:04

This example answers the "or could there theoretically be"-part of the question I believe:

Person A looks at object x which is being manipulated with parameter p (which is an extremely complicated pattern). A then travels through a one way wormhole which takes him to another universe, U2, distinct and isolated from our universe U1. He then observe an object y and find imprints of parameter p (something which cannot be a coincidence due to the highly complex nature of p).

He can then confirm that parameter p effected object y from object x without y affecting x back (remember that U2 is isolated from U1). The only problem is that person A never can communicate this fact to anybody living in U1 without also making it untrue.


Lets imagine a form of epiphenomenalism called pseudo-epiphenomenalism (pEP). In pEP there are two causal chains A, and B. each chain has its own type of particle a for A, and b for B, which are the substrate of the causal chains. Also, interactions between a 's somehow affect b 's, but not the other way around. Lets say A is our physical world. One day, scientists conduct an experiment that modifies a 's in such a way that they can be temporarily affected by b 's. Thus, b 's can be detected based on their effects on modified a 's (m-a 's), and pEP can then be detected by measuring a -> b -> m-a. Perhaps this does not meet your definition of epiphenomenalism, but it does show an example of how an epiphenomenalism-like situation could exist and one day be detected.

  • Yes, I find it to be a great example! This is kind of the revered answer from the one I suggested above where the scientist observed the manipulation of b (he lives in B), travels to A and then observes m-a's. In your example, the b's are inferred and not directly seen but still plausible to postulate. I guess the only problem is to make sure that the a's doesn't affect back, sort of speak, to some other entity in B, for example on c's.
    – Speldosa
    Aug 9 '11 at 0:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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