Blog Coordinator

« Free Will Skepticism and Criminal Behavior: A Public Health-Quarantine Model | Main | Life after Death in Slate Magazine »



Feed You can follow this conversation by subscribing to the comment feed for this post.

Hi Angra, thanks for the clarification. We disagree at a quite fundamental level, as I reject the functionalist account that you are sketching of Bot. I reject functionalism for the reasons that are frequently cited in the philosophy of mind literature, but adequately discussing this would require an entirely different blog post, so I’ll set it aside for now.

About the issue of the complexity of the events that might be involved here, the point that I’m trying to make is not empirical in that way. There is nothing that we could empirically discover that would alone determine the correct ontological category to which you belong as an agent. Empirical data is insufficient for that task.

And about your revised argument, I agree that there is no obvious contradiction, but your (3’’’) does not follow from your (2’’’), given (1) and (2) as stated. I think much more work would be required to generate that conclusion.

Hi Chris,

Thanks for pushing back against the push back. My apologies if I’ve been unclear or kept you befuddled.

Before I attempt to clarify my argument, let’s not forget about my (2*) above, made in response to your previous comment. As you envision it, does agency-reductionism introduce another kind of causal relation, one that secures a causal role for the agent in addition to the causal role of the agent’s mental events? If so, does that not amount to an abandonment of the Event-Causal Assumption (ECA)?

One way to see the validity of my argument would be to make explicit the fact that (ECA) is intended to cover not only the presumed causal interactions between mental events and corresponding bodily movements, but causation in general. Thus, if event causation is the only kind of causation that there is, and if (1) and (2) obtain, then (3) would follow.

Importantly, this requires the assumption that event causation is the only kind of causation that there is, but it does not require a commitment to the denial of agency-reductionism.

Hope that clarifies rather than mystifies!


Thanks for the reply. I think I need to clarify a couple of points:

Regarding (3'''), I agree it doesn't follow from (1), (2), and (2''). It wasn't meant to follow from them (side note: I'm not trying to defend (3'''), either).
The point I was trying to make in that part of my post is that (3''') is apparently not incompatible with the conjunction of (1) and (2). But (3''') is incompatible with (3), so if there is no contradiction in (1), (2), and (3'''), your argument is invalid because (3) does not follow from (1) and (2). I added (2'') only to highlight one alternative to the ones you gave for the defender of ECA (namely, identifying a person with either a single event or an abstract object).

Regarding functionalism, I don't endorse it, either. For example, I don't hold that zombies are metaphysically impossible (they might be possible for all I know), even though they wouldn't have minds.

That said, I do think that behavior such as the one I described provides evidence of minds (even if not conclusive), so I think they would probably conscious as well, and I think they would qualify as agents even if the causes of their movements are also events in their processors. (perhaps, there is some disagreement here on the semantics of the word "agent"). But regardless of whether they would be conscious, my intuition is that they would be causing their movements, here also even if the causes of those movements are also events in their processors. But at this point, we seem to have quite different intuitions on these scenarios.

Hi Andrei, Michael --

I see. Thanks. I agree that if one thinks that it is the properties (all powers, for the pandispositionalist) that do the causing, and not the propertied-thing ("thing" as a count noun only), then the attachment problem is reproduced for the powers theorist. I just don't think that anyone should think that. The antecedent, I mean. :-)

Michael, I guess that it depends upon what you'd say re: my moving my fingers, i.e., when that is the "outcome" in question. I don't think that I cause my fingers to move; I think that I just do move them. I exercise my power of manual dexterity. Jonathan Lowe would say (did say) that I do so BY willing my fingers to move. I had understood - misunderstood, it seems - you to be saying that I cause my fingers to move; that doesn't seem right to me.


ps. I think that I am ahead of the game re: dropping the ECA in that, never having adopted it, I don't have to drop it. (I'm teasing -- though not about never having adopted it.)

As I see it, there is a real danger of reifying the powers of things. I think that ironically enough (because Locke was good on this one point, at least) the tendency to do so is helped along by a tacit Lockean-style ambivalence and ambiguity about the concept of a substance and the nature of that to which the concept of a substance refers, an ambivalence that is ubiquitous in the Humean-inflected analytic environment. (Of course, it's even stronger in Hume, but I think that insofar as they are thinking about substances, Humeans - especially otherwise-Humeans who like powers - often fall back on Lockean substances.) My sense is that Aristotelians are less apt to loose track of the actual entity that has the powers.

I seriously need to be able to edit these posts. :-) Sorry for them being harder to read than they should be. Apparently I never met a nested clause that I didn't nest even further.

Hi, Angra, thanks for the further clarification. As I meant (2) to be read, it is the claim that you are not identical with any particular mental event, or with the set of all your mental events, or with any collection of mental events, or “connected network” of mental events, as you put it. Thus, I took (2) to include the kind of modification that you mentioned in your (2’’).

I believe that the denial of the conclusion of my argument is indeed incompatible with (1) and (2), and I think this is more readily seen by making explicit the fact that the Event-Causal Assumption (ECA) is an assumption that applies to causation in general. When formulating the argument in the original post, I assumed that any philosopher who tacitly or explicitly endorsed (ECA) in the context of explaining human action *also* tacitly or explicitly endorsed (ECA) as a general thesis about causation.

So, if you assume that all causation is event-causation, and if you assume that you are not identical with any mental event (or with the set of all your mental events, or with any collection of your mental events), then, I claim, there is no way for *you* to play any causal role in any causal relation whatsoever. For if events are the only kind of entity that can play a role in causation, and if you are not that kind of entity, it follows that you cannot play any role in causation. Hence, I believe that my original argument, with the background assumption that all causation is event-causation, remains valid.

Thanks for pressing me to clarify. I hope that helps.

Hi Ruth, thanks for the reply. Sorry that I misspoke: you were never an adopter of (ECA) in the first place. Even better, I suppose!

I agree that there is a real danger of reifying the causal powers that objects possess. I worry that this is precisely what happens to those who assume that only manifesting causal powers play a role in causation, and not the objects that possess those powers. As a faithful (neo-) Aristotelian, I try to keep my sights clearly focused on the particular persisting objects that have causal powers. It sounds like you are doing something similar.

In cases where moving your fingers is a basic action of yours, i.e., where you “just do it”, rather than where you are doing so because you are typing words on a keyboard, for instance, even here I would say that you are performing this basic action *by doing something*. After all, the fingers are not moving by themselves, like in a spasm or twitch, out of the blue. Rather, you are making your fingers move, and you are doing so *by exercising* the relevant bodily capacity. Thus, even at the basic or fundamental level, what takes place is intentional action. Or so I would suggest!


Thanks for your clarification of premise (2).
That blocks avoids the validity objection I was raising, but without the implicit premise that all causation is event causation, it seems to me the argument would still be invalid.
Once the implicit premise is included, I'm not sure. For example, what about the following claim:

(3'''') A person is partially but not entirely constituted by (or identical to) a causally connected chain of events, and a person is causally effective because some of the events that constitute her are. In particular, a person plays a causal role in generating her own bodily movements.

(3'''') is compatible with (1) and (2), but not with (3), so a question is whether the implicit premise that all causation is event causation is worded in such a way that, together with (1) and (2), it entails (3), and thus in particular is incompatible with (3'''').

If you word it in such a way that, as you say, it follows that events are the only kind of entity that can play a role in causation (excluding even complex entities partially but not entirely constituted by events, if there are any, or anything else), then your argument is valid as far as I can tell (I'm assuming the "only kind of entity" statement is equivalent to something like "For all X, if X plays a role in causation, then X is an event, or a chain or network of events, etc.")
But I wonder whether [most] philosophers who tacitly or explicitly endorse (ECA) in the context of explaining human action but do not endorse the view that humans are identical to an event (or a causally connected network of events, etc.), also [explicitly or implicitly] endorse the view that the only kind of entities that can play a role in causation are events, excluding even entities constituted or in one way or another identified with (but not identical to) events.

Hi Angra,

Thanks for the reply. I intend the background assumption about the generality of event causation to be the sort of claim that you mention (e.g., “For all x, if x plays a role in causation, then x is an event, or a chain of events, or network of events, etc."), and with that claim in mind, I think the original argument remains valid.

My hunch is that, at least in the context of philosophy of action, a great many philosophers tacitly or explicitly assume that event causation is the only kind of causation that exists. Those who assume otherwise would likely accept the kind of argument that I present above. Or so I imagine!

One thing I'd press for here is that event language is merely a precise way of expressing causal relations that (i) is capable of capturing the uniqueness of cause and effect (by e.g. an intraworld rigid designation schema) and (ii) is compatible with a non-event ontology of both physical and mental entities that crucially relies on time measurement as an empirical foundation for relating event-talk with any larger-scale ontology. One huge motivation for event-language is that it is compatible with any and all solutions to the mind-body problem in terms of expressing causal and non-causal relations in fine-grained event terms. But the relationship of the events of the many ontologies are thus expressible, and the fact that event language is useful for talking about them entails nothing about the several ontologies themselves.

I'm glad that you clarified that the ECA you're targeting is exclusionary, e.g. "For all x, if x plays a role in causation, then x is an event, or a chain of events,... etc". I'm against any such monopolizing tendencies.

Physics provides examples, such as the AdS/CFT correspondence , of theories which are "dual" to each other. Sean Carroll writes in _From Eternity to Here_ (282) that "dual" is "a fancy way of saying that [the theories] look very different but really have the same content." I suspect that event-theories and object-theories are also, at their best, duals in that sense. As long as no exclusionary claims are made, no offense should be taken.

Or, in short: what Alan said.


I'll note that I don't actually think that the sort of dispositionalist/powers theory of causation I prefer has the consequences you insist it does. This is because it is the powers of the agent that are doing the work (most likely in concert with varying manifestation partners that are outside of the agent--think here of Mumford and Anjum's vectors only with both mental properties and various non-mental properties). I take powers to be ways objects are that are interacting with one another in causal processes.


I follow Martin, Heil, Ingthorsson, and take all properties to be powerful qualities. Is that pan-dispositionalism? I suppose. I don't follow Holton and others and take there to be pure powers that are not also qualities. Anyhow, I'm not sure this makes a difference. I do think there is a debate among folks who endorse an ontology of powers over whether objects cause outcomes holus bolus or not. If you say they don't, then it seems it would be some modes or ways of that object that are doing the work (that is, causal relations are relations between manifesting dispositional properties and their outcomes). I take this to be the view of the folks I mentioned in my earlier comment. To quote C.B. Martin, “The object is causally operative in some event for particular effects only in virtue of some of its properties rather than others. It is not operative holus bolus for each and every effect” (“Properties and dispositions” In *Dispositions: A Debate*, edited by Tim Crane, (New York: Routledge, 1996), 71).

Hi, Alan, and Hi Paul (since you’re in agreement), if I understand you correctly, I would agree that nothing about the myriad ways in which we speak about events and objects entails anything about which ontological commitments we ought to make. Deciding among potentially competing ontological commitments requires more than merely relying on our linguistic practices, I’d say.

Hello Andrei, thanks for the follow-up. If it’s only the powers that you possess that are doing the work of causation and not *you* as an agent, then I think the Problem of the Missing Agent remains in full force. We will need to ensure that the agent is playing a causal role in making her body move during an action, otherwise there is nothing but powers manifesting together with other powers, even if those powers belong to the agent as properties. Even if those powers belong to the agent as properties (or, ways in which she is at any given moment in time) what takes place would be nothing but mere events, viz., manifestings of reciprocal causal powers. And mere events, as they say, do not an action make.

Regarding the connection to C.B. Martin: if an object is going to play a causal role in producing an effect, it’s not the case that it *must* do so “holus bolus” or all at once (assuming that notion is clear and uncontroversial). It’s not the case that either (a) the object plays a causal role all at once or (b) only its causal powers are producing the effect in question.

That is, we can agree with Martin that, on any given occasion, an object is causally operative in producing an effect only in virtue of some of its causal powers rather than others. This is a contrastive claim about *which* causal powers are relevant on a particular occasion. But, contrary to what Martin seems to say in the passage you quoted, it does not follow that it is *only* the object’s causal powers that are relevant.

In the Woodwardian interventionist model of causation, if I intervene in your action by talking to *you*, doesn't that make you a cause? (Or to a dog, say). The intervention only works if there is a reasonably complex intact intentional system. Which is full of events (speaking as a humble reductionist), but best understood causally at a high level (the same verbal command and motor output will be differently represented in different nervous systems).

Hi David,

Thanks for your comment. It’s nice to meet you here. If I intervene in your action by talking to you and thereby make you a cause of some result, we do not (yet) know *how it is* that you are thereby producing the relevant effect. My concern here has been that, if your production of the relevant effect is accomplished via nothing but the causal powers of your mental events, then we yet to address the Problem of the Missing Agent, given that you are not identical with any of your mental events. By itself, the interventionist model of causation does not address this worry, it seems to me.

Hi Michael. I guess this comes down to the slipperiness of what a cause is. If I am a complex series of pipes that you shout into, leading to a flap moving at the other side, in the interventionist way of thinking I am a cause - by virtue of my (static) structure. I suspect you won't think me particularly agential.

Hi David, thanks for the response. There are other senses of "agent" that might apply to the kind of example that you mention, but as you correctly suspect, in the sense of "agent" that I'm interested in here (i.e., intentionally performed bodily actions) I think that your example is not agential.

The comments to this entry are closed.

Books about Agency

3QD Prize 2014: Marcus Arvan