Friday, August 06, 2021

Commonsense Epiphenomenalism

I recently came across a popular article claiming that epiphenomenalism was "one of the most disturbing ideas in all of philosophy." The author seemed to confuse causal inertness with "irrelevance" in some broader sense, so I won't waste time addressing that further.  But it may be helpful to offer a more commonsensical conception of epiphenomenalism.  Consider:

We may think of mental states as having both physical and experiential components: their physical effects are due entirely to the physical aspects of our thoughts. The non-physical (experiential) component, on the other hand, constitutes what it feels like to be in that state. There's then an obvious sense in which our mental states have causal effects, insofar as their physical aspects do. That doesn't require that the causal 'oomph' come from the experiential aspect -- indeed, how could it? Experiential feels aren't the kinds of things that push atoms around. You need other particles to accomplish that!

On this view, we may (speaking loosely) say that I pulled my hand away from the hot stove "because it hurt", and this can be perfectly informative, without implying that the hurty feel itself provided the causal force that moved my hand...

This sort of property dualism strikes me as perfectly commonsensical, so I find it weird that so many other philosophers seem to regard it as a non-starter.  I mean, I get that many are ideologically committed to physicalism.  But there often seems an assumption that there's something distinctively unacceptable about epiphenomenalist dualism in particular.  Whereas I really don't see any advantage to interactionism -- the latter view introduces conflicts with our scientific understanding of the world (epiphenomenalism, by contrast, respects the causal closure of the physical), and once we bear in mind that our mental states have causally-oomphy physical components, what motivation is there to attribute causal oomph to their phenomenal correlates in addition?  It seems to me that the intuitive force of this objection to epiphenomenalism depends upon momentarily forgetting that we have brains.


  1. Hi Richard,

    It looks difficult to understand, and I can imagine very weird interpretations, and I'm not sure I can imagine one that isn't weird (there are aspects of your interpretation I'm not familiar with, though).

    One question about causal closure: Is it be one-way closure, or is it bidirectional? In other words, do particle movements cause pain, but pain does not move particles around?

  2. > "do particle movements cause pain, but pain does not move particles around?"

    Yes, that's the idea.

    1. Hi Richard,

      It looks difficult to understand, and I can imagine very weird interpretations, and I'm not sure I can imagine one that isn't weird (there are aspects of your interpretation I'm not familiar with, though).

      One question about causal closure: Is it be one-way closure, or is it bidirectional? In other words, do particle movements cause pain, but pain does not move particles around?

      Alright, so I've been thinking a bit more on this, and it's still very weird. I do not know why the philosophers you have in mind regard it as a non-starter (though there probably is more than one reason), but it does not strike me as commonsensical. The main reasons for me are that:

      1. It seems to go against our ordinary practices and indeed common sense. For example, suppose Mary has a headache, consciously ponders whether to take an aspirin or an ibuprofen, and consciously chooses the latter; on the basis of her choice, she actually takes an ibuprofen (and then calls the doctor in case she needs to have a COVID test, but that aside).

      It seems to me that her having a headache played a causal role in her consciously thinking about whether to take an aspirin or an ibuprofen. It also seems to me that her conscious deliberation on the subject played a causal role in her eventually choosing (for some reasons she consciously thought of) ibuprofen, and then her conscious choice is one of the causes of her action, namely the ibuprofen.

      In other words, it seems natural - and, at least intuitively - obvious to me that our sensations like pain, our conscious deliberation, and our causal choices, are generally causally effective.

      2. There is no good reason I can think of to believe it's true. Granted, sometimes some of our intuitions and practices can be misguided. But I would need some reason to believe an error theory is in order. And I do not see it. The main argument I see in support of this sort of epiphenomenalism is that you need particles to push particles. But I do not see that as a good argument, because we have the particles - are a certain bunch of particles acting in some way.

      On that note, I do not believe in souls or some kind of substance dualism. But are conscious. We have minds. We suffer pain, experience pleasure, consciously deliberate, etc. Yet, we are not made of bunches of particles. The point is that, in addition to the properties physics deals with, bunches of particles - when combined in the right way - also have properties (or states, or however one calls them) - like being in pain, thinking about stuff, etc.
      So, more precisely, it's a bunch of particles in pain that brings about a bunch of particles deliberating about whether to take an aspirin or an ibuprofen, etc. But that's all done by bunches of particles - by us! -, so I do not see any strength in the objection (after reading Aron's post, if I understanding it correctly it seems to me he had something like this in mind too).

    2. Sorry, the first two sentences were repeated. Copy/paste error.

    3. Hi Angra, could you say more about what you find unsatisfying about my account of the sense in which "her having a headache played a causal role"...? That is, we can think of headaches as physical-phenomenal composites, and the causal impact comes entirely from the physical properties of the headache (i.e. the neurons that generated the pain sensation, rather than the pain sensation itself).

      Likewise for "conscious deliberation", and all the rest. It all has neurological underpinnings, which are undeniably causally efficacious. The physicalist claims that those neural states *constitute* pain -- that the latter is nothing more than the former -- whereas the property dualist claims that these are distinct states and that the neural one gives rise to the phenomenal one. Epiphenomenalists further clarify that the causal oomph all comes from the physical side of things. But I'm suggesting that this doesn't require denying commonsense claims about headaches and conscious deliberation playing a causal role in our subsequent behaviour.

      You wrote: "in addition to the properties physics deals with, bunches of particles - when combined in the right way - also have properties (or states, or however one calls them) - like being in pain, thinking about stuff, etc."

      Now you, too, sound like a property dualist! And since I hope you'll agree that the causal powers of things stem from the properties physics deals with, it would seem to follow that these additional properties must be epiphenomenal.

      I'd also argue that physicalists aren't really in any better position here. They claim that by identifying pain with some neural state N, they thereby allow for pain itself to do causal work. Now, it's true that N does causal work, and they identify pain with N. But the fact (if it is one) that N constitutes pain, or feels painful, does no explanatory work here (as we can see from the conceivability of zombies, for whom N has the same causal effects without truly constituting pain). The causal influence of N is entirely explained by its familiar physical properties.

      Why think epiphenomenalism is true? Because (1) it's the best alternative to physicalism, and (2) physicalism is false.

      Why think physicalism is false? Because it's incompatible with taking boundary disputes about consciousness to be substantive. (Also: zombies.)

    4. Hi Richard,

      I will try to elaborate, but I'm not sure I can be very precise, since I'm not sure I understand the account well! Please let me know what I got wrong. But in a nutshell, it comes down to this: I do not see a reason to remove some properties from the causal equation so to speak.

      Longer version: (I'm not sure how to make a distinction between physical/phenomenal, or more precisely the 'physical' part of it. I'll take it that the physical properties are those physics studies (which is also controversial, but it will have to do for now)).

      So, you say your theory doesn't deny common sense claims that our conscious deliberation plays a causal role in, say, our conscious choices. But I do not see why it does not deny it. If the conscious deliberation is not causally efficacious, it seems to me it does deny it - though it says there is something that causes both the conscious deliberation and the latter conscious choice, perhaps some neurons...or is the particles? (and what if it's turtles all the way down? What would be causally efficacious then? )

      In re: the neurons that generated the pain sensation, rather than the pain sensation, I would say one is looking at the matter from different perspectives: it's a bunch of particles experiencing pain what plays a causal role in bringing about the next state of affairs of a bunch of stuff deliberating about what to do, and the separation seems to be a question of focusing on some properties. For example, if asked about the causes of my consciously typing this, I could say it's my previous conscious reasoning and decision to try to explain what I think. I might, however, not mention the conscious reasoning and say it's a bunch of neurons doing some stuff, which is also correct though far less informative. Or I might not mention any neurons and say it's a bunch of subatomic particles doing some stuff, which is also correct but far less informative.

      In re: physicalism: I'm not sure in which sense neural states might constitute pain. Surely, there is no semantic identity. So, one might thing it is like 'water is H2O', but I think this would be false too. 'Water is not H2O' is metaphysically impossible (impossible as counterfactual, in 2D semantics). But then, I think zombies are possible (so, we agree! I don't agree about Moral Twin Earth, but the zombies and similar things suffice in my view). Also, which physical correlates? I mean, one can think of pain existing with very different correlates from those in the real world, or the current physical correlates associated with different sensations. And I do not see a good reason to think it is impossible as counterfactual.

      So, leaving some room for the possibility I haven't understood physicalism well enough to tell or I made a mistake about the different correlates and the zombies, I'd say physicalism is very probably false.

      As for your question about the causal powers of things, I do not know that they come from some properties to the exclusion of others. For example, let us say P is one physical property, whereas Q is not one of those. P does not necesitate the power to bring about conscious states (e.g., zombies). In the actual world, having physical properties correlate with some causal powers involving phenomenal properties (though we do not know what those are!). But then, having phenomenal properties also so correlates, and we have a much better idea of the correlations. As before, I do not see a good reason to exclude properties that are not physical from causal efficacy...and here I should point out that they are "non-physical" in the sense they're not studied by physics, but not in the sense of some spooky stuff going on!

  3. 1. Property dualism is the idea that a person consists of 1 substance with (at least) 2 clusters of properties, physical and mental. It seems to me that property dualists can avoid epiphenomenalism if we insist that causal actions are, most properly speaking, performed by substances, rather than by properties. The properties are simply ways of speaking about ways that the substance is, or things that it can do. (This is an attractive solution since, if the unity of the substance plays no role in our metaphysics, then we are merely substance dualists pretending to be property dualists!)

    One can, of course, say in a loose manner that something was caused by property X, but I would regard this as an imprecise shorthand way of talking about things that the substance itself can do, by virtue of being the sort of thing that it is. On this view, your mental properties don't really cause anything; but then again strictly speaking, neither do your physical properties! Rather, your physical and mental properties are simply ways of talking about the various acts (thinking, feeling, moving, pushing, falling) that you are able to perform, many of which have both physical and mental aspects.

    2. One can generalize this problem away from philosophy of mind, by distinguishing between the "causal" and "noncausal" properties of a given object. This leads to a "Generalized Epiphenomenalism" puzzle. Since logic appears at first sight to imply that noncausal properties are, by definition, causally inert, it seems that they can have no effect on the causal properties! Assuming we are realists about causality, I only see 3 possible conclusions here: A) try to argue that all valid properties reduce entirely to causal relations (which may be attractive to some, but I don't think it is plausible); B) accept that some properties (all the purely noncausal ones) are epiphenomal; or C) acknowledge that while we can indeed---formally or abstractly---sort properties into "causal" and "noncausal", that this division is more in the mind of the person doing the sorting, and that we certainly shouldn't think of any substance as if it were a causal "part" attached a noncausal "part" (as if they were two independent entities, adhering to each other only because of some metaphysical version of Scotch tape!)

    In ordinary language we say "I tripped over a rock", not "I tripped over the causal properties of a rock". I am proposing that this ordinary way of speaking is actually more metaphysically correct. That does not mean that the phrase "the causal properties of a rock" is just meaningless gibberish, but we have to treat it more cautiously since it refers to a more abstract and convoluted semantic construction: "the fact that rocks are able to trip people, etc." Such abstract phrases may be helpful when trying to describe causality in specific detail, but not if we excessively reify them by acting as if such complex phrases are the most natural subject of causal relations.

    1. This seems like a nice diagnosis, though I'm inclined to think there's a perfectly clear sense in which physical properties are causally efficacious while phenomenal ones aren't, namely that when we cause things, it's in virtue of our physical properties that we had those causal effects.

      Could you give an example of what other "noncausal properties" you have in mind, for the generalized puzzle?

    2. This probably depends on one's other metaphysical beliefs, but personally I think that leaving aside cases like psychological/normative facts, there are also probably structural or intrinsic properties of physical systems that could qualify as being noncausal.

      It's easiest to give an example in a more artificial context, so let's pretend that the laws of physics were actually given by Conway's "Game of Life" cellular automaton. In this case, a given cell is defined to have 2 basic types of properties, its spatial location (/spacetime location on a 3D view), and a binary state (usually called "alive" vs. "dead"). Conditional on the evolution rules of the cellular automaton, the causal effects of a cell on other cells (/future cells) are determined by its location and state. I won't describe the evolution rule here because it is well-known, but it is important for the thought experiment that there is a specific rule.

      However, in Chalmers style p-zombie arguments, one is supposed to ignore the actual laws of Nature and just think about conceivablity. From a logical point of view, Conway's evolution rule is not logically necessary, and thus you can (at least in a purely conceptual way) separate in your mind the state of the cell considered in itself, from its causal effect on other cells. For instance, you could imagine without contradiction a "p-alive" cell which is dead, and yet it affects future cell-states just as if it were alive! Once you have separated in this way the intrinsic state, from its causal effects, my generalized argument would suggest that the intrinsic state is now "purely epiphenomenal".

      But, I would suggest that this conclusion arises from an excessively particuarized analysis of the properties of the cell, and does not really tell us too much about the actual metaphysics of cells (assuming Conway's "Game of Life" were the real world). In reality the state of the cell is always accompanied by certain predictable causal effects on the world. But of course, the same can be said of our conscious experiences.

  4. I think the problem some people have is that goal and function doesn't easily fit in to our usual ideas about "physical aspects". Consider the "terminal value" of blue being your favourite colour, and so predicting your clothing purchases. The sensual nature of a colour is one reason for such a preference. Now, I think such basic aesthetic experiences are completely determined by the structure of your visual processing pathways - which holds true over time despite a complete turnover of all the proteins involved - but it seems dumb to say the sensual pleasure qua experience is not somehow important and a thing that persists despite the evanescent nature of our physical being.


Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)

Note: only a member of this blog may post a comment.