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.

Could the argument be defused if we rejected premise 2 by arguing that there is a sense in which you are identical to (supervene on?) a large set of mental processes (or activities), including occurrent conscious mental activity involved in deliberation and planning but also the 'background processes' (character traits, goals, values) that causally contribute to those occurrent processes? I'm not sure why the causal contributions mentioned in premise 1 need to be narrowed down to occurrent mental events, much less one (!) mental event. And I don't think it's counterintuitive to think that we are constituted by a large set of mental processes, both standing and occurrent, both conscious and non-conscious. But maybe I'm betraying some cherished principles of action theory here.

Interesting stuff Brent.

There is a third option: what I call identification reductionism. This option rejects premise 2 and argues that while an agent is not identical to any state or event or any set of states and events, the agent is identified with some range of states and events and given that the agent is identified with these states and events, the causal role of these states and events counts as the agent's causal role.

Here we can think of identification as playing a structurally similar role that the notion of realization plays for Shoemaker or constitution for Pereboom in there nonreductive accounts of physicalism.

As I see it, the argument that you raise (I present a very similar argument in my 'If Anyone Should Be an Agent-Causalist, Everyone Should Be an Agent-Causalist' forthcoming in *Mind*), is structurally similar to Kim's Exclusion Argument in the philosophy of mind. Thus, those who think that only states and events play a fundamental causal role are faced with the choice of either reducing the agent to states or events or finding some relation of non-identity between the agent and states and events that confers (in a derivative sense) a causal role on the agent.

Thanks for the comment, Eddy. I don’t think you’re betraying any cherished principles of action theory. In fact, my hunch is that many would share the sorts of assumptions that you stated, e.g., that you are (identical with) a set of mental processes, both occurrent and non-occurrent. It’s just those sorts of assumptions that the argument I shared is meant to challenge.

There are at least two worries here:

One is that no clear sense can be made of the claim that you are (identical with) a set of mental processes, however large in number or diffuse in type its members might be. After all, sets are abstract entities, whereas creatures like us are not.

Another is that if you are (identical with) a set of mental processes, the Problem of the Missing Agent is not resolved, only worsened. It is worsened because sets are not in the business of causing anything, let alone an overt bodily action. Thus, even if we assume that you are (identical with) a set of mental processes, the question would remain: exactly how are you playing a causal role in generating your own bodily movements when acting?

Thanks for chiming in, Chris. I look forward to reading that paper.

In place of my premise (2) above, your suggestion is that you are (identical with) some range of presumably mental states and events. Perhaps you address these questions in the forthcoming paper, but I wonder: What, exactly, is a “range”? Is it an abstract entity? How does it relate to its members? Does it play any role in causation? If so, how? And most importantly for present purposes, what is it for you, presumably a human being, to be (identical with) such a thing?

The worry is that claiming that you are (identical with) some range of states and events is as nasty a metaphysical bullet to bite as my premise (2).

Although I haven't thought about these things enough, why is there not a third option: accept ECA, but recognize that there are events other than mental events like desires, belies, or intentions. That is surely uncontroversial. Of course, we'd still need to do something to identify the agent with the relevant event, but since we've opened up what counts as an event, this may be easier to do.

If you take it to be axiomatic that there is _no_ sense in which an agent is an event, then you cannot both accept (an exclusive version of the) ECA and also claim that agents can cause the things they do. But unless this is axiomatic (and I don't see why it would be), this leaves an open question to investigate.

Of course, you may not like this option. But then I take it you've got concerns about the other options. In this poor company, maybe this third option is the less distasteful road to explore!

(It could be that this road is plainly unavailing. Like I said, I need to think more about these things.)


Not identical with, but identified with. Identification is a non-identity relation. When an agent is identified with an attitude, say a desire, he is not identical to it. Rather that desire has authority to speak for him--it plays a privileged role in his agency. Agents are not identified with a range of anything, rather there is at least one and member more distinct attitudes each of which the agent is identified with, at least for a time.

Obviously this is sketchy and there are worries about it. You just asked if there is a third option and it seems clear to me that the answer is yes!

Also, there may be a forth option. Why not think the 'being the subject of' relation is sufficient? I am an agent in virtue of being the subject of attitudes that play certain causal roles.

Finally, with respect to your point about sets. Suppose I think I am identical to a set, the members of which are certain states, events, process and the like. Suppose you're right that sets don't cause things. Still the members of the set might cause things and so why not think I cause things in virtue of the members of the set I am identical to causing things?

I suppose all these responses I am running point to the fact that your argument assumes agency reductionism is false. According to agency reductionism human agency is not a fundamental phenomenon. On this account, my causal role as an agent will be played by things other than me (states, events, process, etc.). Does it follow that I don't cause anything? Well, not according to reductionism. On this account I do cause things, but wholly in virtue of the causal role of things none of which are identical to me.

So why think we need a non-reductive role for the agent? Unless we have a strong case that we do need such an account, then it seems easy for proponents of ECA to avoid the conclusion of your argument.

Hi Craig, nice to meet you here, and thanks for commenting.

I strongly believe that this is an open question, and that it must be investigated with more care and attention than it typically seems to receive.

As you suggest, there are plenty of third options. The good thing is that some options do not require us to make the implausible claim that you are (identical with) an event. For instance, the ontological category to which you belong could be that of a particular persisting object, or a simple mental substance, or a property, etc., etc. Part of the point of my above argument is to challenge us to think about such alternative options, given that committing to the claim that you are (identical with) an event seems to be a gross category mistake of the kind that would make Gilbert Ryle roll over in his grave.


We've talked about this stuff many times before. I thought I'd chime in quickly with some half-baked thoughts based on stuff I've been working on with Jesús Aguilar.

While, like you, I reject the standard events oriented metaphysics of causation assumed by many (in favor of a richer metaphysics of causation involving powers interacting and manifesting in causal processes a la Mumford and Anjum, Chakravartty, Heil, Molnar, etc), I think you'd insist, Michael, that folks like me have the same problem as proponents of the traditional Humean event-causal story.

I think proponents of an ontologically (but not necessarily conceptually) reductive picture of agency must emphasize that the complete story of our exercising intentional agency (whether free or not) that makes reference to mental items producing behavior is a complex one. Both the events and powers-in-processes folks can take agents to be complex systems of interacting element (these may be events or, in my case, simple substances and their powers). The elements of such systems are causally interacting to bring about intentional behavior. And not all of the elements at a time are causally relevant to the production of a given act-token. But the sorts of processes that occur in such a system are sufficient to provide us with the truth-makers for our agency talk.

I think this is not too distant from what Chris and Eddy are suggesting. The responses at least bear a family of resemblance of sorts to one another.

In the interest of putting pressure on you, Michael, I think that the natural response to philosophers of action who raise the sort of objection you raise here is that they mistakenly reify agents (so they take agents to be more than systems of interacting items). We need to know why we need to drink the ontological Kool-Aid you are asking us to swallow.

Hi Chris, thanks for the response. You mention many interesting points.

As I mentioned to Craig, there are plenty of third options out there that should be explored. It sounds like you offer another option in your forthcoming paper, which I look forward to reading.

About the “being the subject of” relation, here’s an analogy that shows why this relation does not serve the purpose you might think it does:

I’m the subject of the Queen of England, in virtue of the fact that I’m a Member of the Commonwealth, in virtue of the fact that I’m subject to the glorious Charter of Rights and Freedoms as a proud Citizen of Canada. Now, the Queen, the Commonwealth, and the Country of Canada each play particular causal roles when bringing about particular effects. For instance, suppose the Queen makes you a Knight. If it were the case that the “being the subject of” relation is sufficient to make me an agent, given that I’m the subject of the Queen and given that she played a particular causal role in making you a Knight, it would follow that I, too, played a causal role in your becoming a Knight. But that is absurd!

About my point regarding the fact that sets do not possess causal powers: if you want to say that it’s the members of a set that are causing the relevant results, we need a principled way of individuating just those members that are playing the sought-for causal role on any given occasion in which you are making your body move while acting. How do we individuate the specific and correct members of this set? And why not drop reference to sets now, if it is (some of) the members that are doing the work of causation?

About my argument: it in no way assumes that agency reduction is false. Rather, by arguing that agency reductionism is committed to an unpalatable metaphysical claim, my argument purports to *demonstrate* that agency reductionism is false.

Lastly, nothing I’ve (yet) said makes a positive case for an irreducible causal role for the agent, so I appreciate the prompt. One reason to prefer such a role is this: if you think that human beings are particular persisting objects (rather than events), your account of the causal role that such an object plays in producing its own movements during an action will likely be non-reductive.

Thanks for adding your voice to the chorus, Andrei, and for the pressure to say something more about my own view.

You raise a lot of interesting points, and I'll reply to three, as I interpret them:

First, I deny that the kinds of processes that occur in a system of interacting items (as you put it above) could provide the truth-makers for talk about you, qua agent, or the overt bodily actions that you are intentionally performing. When we are talking about your actions, we are not talking about mere happenings. Rather, we are talking about bodily movements that *you are making happen* during an action. The sorts of processes that take place within the kind of complex system that you describe are mere happenings, not actions. As a result, such processes cannot, by themselves, be truth-makers for talk of agents and the actions that such agents are performing.

Second, I suggest that the ontological category to which you belong is a particular persisting object, rather than an event or any other kind of entity. As an object, you persist through a finite duration of time, you possess particular properties, you occupy some relations, you undergo various processes and changes, you partake in numerous events, etc., etc., but you are not any of those kinds of entity. Assuming that the ontological category to which you belong is an event (or system of interacting items) strikes me as a category mistake.

Third, for those numerous philosophers who tacitly or explicitly adopt the Event-Causal Assumption (ECA) that I mentioned above, there is serious pressure to claim that, if you are going to have any causal role when making your body move during an action, your role must be identical with some event. This, I suggested, leads to a nasty dilemma: either you are identical with the causally efficacious event in question, or you play zero causal role with respect to your own bodily movements when acting. As a way out of this dilemma, I would reject the Event-Causal Assumption (ECA) in favour of a Causal Powers-Based (CPB) conception of causation, and reject the claim that you belong to the ontological category of an event (or system of interacting items) in favour of the claim that you are a particular persisting object. That would be some sweet, sweet summery Kool-Aid, indeed.

What say you?

Hi Michael--

I'm interested in this because somewhat like Davidson and Whitehead I think that event language is the best metaphysically descriptive language of reality. But it is not the only language available to us. I can use event language to describe a particular wave of Lake Michigan that breaks on the shore of the beach behind my campus (perhaps I photographed it and thus have its particular spatiotemporal dimensions recorded). But, using another more "thingy" kind of language referring to the Lake as a whole and across time and space, it is a wave of the Lake. So I take it that event language just isolates and particularizes aspects of the universe that might well be described otherwise as an event part of a greater thing. Agency language typically is focused on a transspatiotemporal referent--the agent. Event language picks out parts of that agent, mentally and physically, for more detailed work. Now none of this constitutes a justification for the "thing" we call an agent, but given that we can justify that non-event reference, then mental events are just particular waves breaking on the universe that are part of our lake of agency.


Premise 1 holds that the movements of your body are caused by the occurrence "your mental events". In that context, the causal attribution is not limited to a single event. Nor is there a causal attribution to a set. But premise 2 only leaves those two options. Why can't the person who endorses ECA hold that you are identical (for example) to all of those events in question, but not to any set in the sense of an abstract object?

I'm not sure I'm explaining the worry clearly, but perhaps the following parallel might be of help:

Let's consider a rock-collecting autonomous robot. (e.g., ). Let's call it "Bot".

(1') When Bot collects a rock, the movements of its robotic arm are non-deviantly caused by some events in its microprocessor(s).
(2') There is no sense in which Bot is identical to any of the specific events in its microprocessors that caused any of the movements of its robotic arm.
(2'') There is no sense in which But is identical to any set or collection of events.
(3') Thus, when collecting a rock, Bot is not playing any causal role in generating its own bodily movements during its performance of that action.

We can establish (2'') by means of one of the arguments you gave: sets are abstract objects, but Bot is not. Additionally, sets are not causally effective, so if Bot were a set (of anything), then (3') would be true.
As for (2'), we can tell it's true because - for example - sometimes Bot moves forward, but none of the specific events in Bot's microprocessor(s) moves forward. Alternatively, we may point out that for any specific event in one of its microprocessors that caused the movement of its robotic arm, Bot existed before the event happened, and/or after the event happened.

Would you conclude that either (1') is false, or (3') is true?

In the case of (1'), granted, there are other causes (other events), but for that matter, those who endorse ECA would agree that if, say, you move your hand, the mental events are not the only causes of the movement (for instance, some events in the nerve connections going from your brain to your hand also play a causal role).

Hi Al, thanks for the comment. I spent two years living in Milwaukee, Wisconsin, not too far from the water, so I like that your image involves Lake Michigan.

It’s not clear to me that this is the sort of thing that you had in mind so forgive me if not, but following a line of argument found in the work of folks like E. J. Lowe, I think that talk of causal relations among discrete events is shorthand for talk about the activity of particular persisting objects and the effects brought about by that activity.

For instance, consider a stick of dynamite that explodes, where its explosion causes the collapse of a nearby small building. We could say that the event of the dynamite’s exploding caused the event of the building’s collapsing to the ground, but that would not be correct. Here, our talk of one event causing another event is merely derivative. It is derivative because what is doing the work of causation is the dynamite, a particular persisting object. The dynamite, by exploding, is causing a particular effect, the collapse of the building. We have an object, behaving in a particular way, thereby causing an effect. I would claim that this is the correct model for thinking about causation in the context of agents and the actions that they perform.

Is that the sort of thing that you had in mind?

Hi Angra, thanks for the interesting post.

Your analogous argument involving Bot is great. It shows that there is no sense in which Bot is playing any causal role in generating its own its movements and actions. In fact, Bot is just the sort of complex system of interacting items to which Andrei alluded above. When such robots and systems go about their business, we are talking about mere happenings, not actions. Since every movement of Bot’s body is non-deviantly caused by some events within Bot’s microprocessors, there is no sense in which Bot is *making* such movement happen, and so there is no *action* in view here at all. There is nothing but mere movement. As a result, such movements cannot, by themselves, amount to an agent performing an action.

But perhaps you don’t share my intuition here?

Hi Michael,

Thanks for your interesting reply as well. As you suggest, I don't share your intuition on the matter. Whether we call Bot an agent on not, my intuition is that it's causally effective. But given our different intuitions, that example is not going to help much I think.

Here's a variant:

Let's say the capabilities of the robots increase gradually. Eventually, some robots can interact with humans, pass for humans in on-line conversations, etc., and actually beat humans at any intellectual task we can come up with, from math to philosophy. Do you think the movements of the robots in question would no longer be caused by some events in their processors? Or perhaps that the robots wouldn't be causally effective, in spite of their great intelligence? Or maybe some other option?

Leaving the scenarios aside, I still think that the person who endorses ECA may reply that a person is not identical to a single mental event but to - for instance - a complex causally connected chain or rather network of them. I don't think that single mental event vs. abstract entity (i.e., "set") are the only options. After all, events caused by multiple other events seem to be common, regardless of whether some people are also involved.

Alan and Michael,

Alan really gets this one, I think. Let me take it further. I don't think there is any competition at all between event-speak and object-speak.

Imagine a raging debate between economists, where one faction insists that world GDP is explained by adding Eastern Hemisphere and Western Hemisphere GDPs. The other says no, that has no explanatory power, the real explanation is gotten by adding Northern and Southern Hemisphere GDPs. I see no reason to view the philosophical event-causation/object-causation debate any more seriously than my imaginary Longitudinal-Economics/Latitudinal-Economics debate. It's the same world. Cut it up however you like.

Craig writes, "you cannot both accept (an exclusive version of the) ECA and also claim that agents can cause the things they do." Indeed - but the parenthetical part is the entire problem.

Thanks for the responses Michael.

Here's a couple push backs. Your argument as stated is invalid. You need to either fix up premises (1) or (2) or add some further premises. I suspect that the only way to do this will be to add some explicitly anti-reductionist clause. This is because you try to move from the fact that you are not identical to the events causing your actions to the conclusion that you do not cause your actions. But this is *exactly* what a reductionist will deny. She will argue that agents are not identical to any of the things playing a fundamental causal role in agency.

Whether or not this is a problem for your argument depends on what you are trying to do. What I like about your argument is that it seems to me to isolate an interesting challenge for reductionists. But it does not seem to me that your argument by itself gives us *any* reason to think reductionists cannot answer it. That will take a different kind of argument. Would you agree?

That wasn't the subject of relation I was talking about. I am wondering why a human's being the subject of some mental states that bring about some human behavior is not sufficient for saying that the human brought about that behavior?

Hi Angra, thanks for responding again. I would say that, even with the supposed increase in capabilities of Bot, the *kind* of thing that Bot is remains the same. That is, an increase in complexity and sophistication alone does not change the fact that Bot does not exert any causal influence over its own movements. Every such movement is initiated, sustained, and controlled by its microprocessors. If all agents and their movements operated like this, then there would be no sense in which any agent performs an action. Or so it seems to me.

I agree that a defender of ECA could claim that you are identical with an entire chain or network or system (or pick your favourite metaphor) of events, but notice that does not address the Problem of the Missing Agent. In fact, it makes is worse, because now we have to isolate just those specific events with which you are supposed to be identified on any given occasion on which you are performing a bodily action, and then provide a principled reason why this identification should amount to your having a causal role in the production of your own action. That would seem to make the task even more difficult!

Hi Paul, nice to meet you here, and thanks for the comment. There might not be a competition between our *speaking* of events and objects, but from this it does not follow that events and objects are members of one and the same *ontological category*. I believe that they are members of distinct ontological categories, and so we cannot, and should not, divide up the world any way we like. Like it or not, the world itself places constraints on how we can go about dividing it up. Or so it seems to me.

Hi Chris, many thanks for pushing back, I appreciate it. There’s no invalidity in my original argument, though of course agency-reductionists can attempt to respond!

Here’s the argument again:

(1) When you are performing an overt intentional bodily action, the movements of your body are non-deviantly caused by the occurrence of your mental events (ECA)

(2) There is no sense in which *you* are identical to any of your mental events, or to the entire set or collection of your mental events


(3) When you are performing an intentional bodily action, *you* are not playing any causal role in generating your own bodily movements during your performance of that action

The kind of agency-reductionism that you seem to have in mind accepts (1) and (2), but denies (3). In order to avoid the conclusion of my argument, the agency-reductionist must add another premise, something to the effect that:

(2*) When you are performing an overt intentional bodily action, the movements of your body are non-deviantly caused by the occurrence of your mental events, *and* you are playing some other (e.g., derivative) causal role in generating your own bodily movements during your performance of that action

My claim is that, in order to find a suitable occupant for that other causal role while still maintaining that you are not identical with any mental event, the defender of agency-reductionism must abandon the Event-Causal Assumption (ECA) in favour of another account of the metaphysics of causation.

To put the point slightly differently, the difficult task for the agency-reductionist is to establish that, although you are not identical with any of the mental events that are non-deviantly causing your bodily movements during an action, nevertheless you are playing some (other, non-event?) causal role in the production of your bodily movements while acting.

I think my analogy with the actions of the Queen stands as it is, as the “being the subject of” relation that I happen to presently occupy with respect to the actions of the Queen is isomorphic to the “being the subject of” relation that I happen to presently occupy with respect to whatever behaviour is produced by my current mental state of being hungry. For in both cases, I am subject to the behaviour of someone or something other than myself, and in neither case can I be said to be doing anything, or bringing about any of that behaviour. Thus, the “being the subject of” relation is not sufficient here. Or so it seems to me!

Hi Michael, all --

I'm for option (b), though (what a shock) it seems to me that there are a couple of different issues in play. The event-causal picture in relation to which the contemporary free will problematic is configured doesn't JUST involve events rather than substances. It also - and even more fundamentally, in my view - involves the idea that causation does not involve real activity, irreducibly oomphy do-ing. (Plus, there's the post-Cartesian substance and/or property dualism(s) as the only presumptive alternative(s) to reductive physicalism. And there's a commitment to reduction or atomism generally (bolstered by various kinds of allowable fudging in order to make it palatable). So I'm for (b), but for reasons independent of the attachment or intelligibility problem to which you point, i.e., that there is a problem with getting the causation to connect to the agent regardless of whether one is defending event-causal compatibilism or event-causal libertarianism -- though I agree with you that it is a problem.

Andrei, I'm not seeing why a powers-theorist (with respect to what causation is) need have the same attachment problem. I agree that s/he will if s/he thinks that it is properties (here dispositional ones, or powers) that do the causing, rather than propertied-things (either because s/he thinks that the bearers of properties are themselves inert or because s/he thinks that what apparent "things" are is in reality nothing other than properties), but one needn't, and oughtn't, think any of those things in order to think that causation is about the expression of powers-to-do, that it is productive (i.e., straight-forwardly, sans stipulation that what "productive" really means is "constantly conjoined, and so counterfactually dependent"). Can you say more about why you think any powers person will have the attachment/intelligibility problem?

Also, Michael I'm not sure that I agree with you that I am causing myself to move my fingers so as to type this. I'm pretty sure that I don't think that I am causing myself to think this thought, either. I think that these are powers that I have, and that I am exercising them. Heuristically speaking (and, you know, for what it's worth), on this point I understand myself to be closer to the late Jonathan Lowe than to, say, Tim O'Connor. Tim's got that same gap in there that you have.

What a great conversation! Thanks for encouraging me to chime in!


Thanks again for the reply.
I disagree about the complex bots. We're talking about entities one can talk to, entities that can do math, science, etc.; the fact that they're silicon-based other than carbon-based seems irrelevant as long as they have those capabilities, so I don't see any reason not to call them "agents". I guess we'll just disagree on this issue.

As for the second point, I don't think there would be a problem of a missing agent, for the following reason: Just as we don't know which specific events we would identify a car or a bot with - it's really complicated -, the person who holds that ECA is true or probably true may say that it's a network of mostly or entirely mental events, but it's not known at this point which ones, and it might take centuries of research or more to figure it out - these are very complicated matters.

That aside, I have to agree with Chris point on the validity of the argument, as it's part of what the alternative I'm suggesting implies.
For example, one may consistently assert the premises and deny the conclusion as follows:

(1) When you are performing an overt intentional bodily action, the movements of your body are non-deviantly caused by the occurrence of your mental events (ECA)

(2) There is no sense in which *you* are identical to any of your mental events, or to the entire set or collection of your mental events.

(2''') You are probably either identical to a complex causally connected network of many mental events, or to a complex causally connected network involving both many mental and non-mental events.

(3''') When you are performing an intentional bodily action, *you* are playing a causal role in generating your own bodily movements during your performance of that action.

As far as I can tell, there is no contradiction, even if I'm asserting the premises and denying the conclusion. So, I think your argument would need more premises, in order to deal with alternatives such as 2'''.


That's an interesting response about the being the subject of relation--I hadn't thought of it like that. Maybe that response works.

I am still confused on why you think your argument is valid. I cannot think of any way to translate it into propositional or first-order logic in a way that renders it valid. What's the inference rule that lets you move from (1) and (2) to (3)? My guess: you won't find any.

Again I imagine there is an easy fix, but I think the fix will *have* to have an anti-reductionist clause.


I think your questions for me should really be directed at Michael. Why? Because Michael has indicated in conversation with me that he regards a powers approach that takes interacting powers to be the causal relata to face a version of the attachment problem. This is because intentional agency on my preferred account takes some interacting powers of an agent versus the agent holus bolus to be the cause of the outcome that is an intentional action. I think he would argue that the former is not sufficient to provide the truthmakers for claiming that *I* moved my arm.

For what it is worth, Ruth, I don't think powers-causation people have the same problem as the event-causation people (but there is a question about whether substances or interacting powers are the relata that I think serves to divide proponents of an ontology of powers). I've assumed the problem is similar for the sake of argument because of some background conversations I have had with Michael about (1) the role of powers in causation and (2) the disappearing agent problem.

Hi Ruth, thanks for joining the conversation, it’s great to hear from you! I’m glad that you’re willing to drop the Event-Causal Assumption (ECA) in favour of an alternative account of the metaphysics of causation.

As mentioned by Andrei, I think that as an alternative account of the metaphysics of causation the causal-powers framework is on the right track. But, for folks like Andrei, given this account of causation what is doing the work of causation is the manifesting causal powers, not the object that possesses those powers. That is, the causal powers alone, by interacting as they do, are producing the effect in question. If this were correct, then the object that possesses such causal powers would be playing no role in producing the relevant effects, and so we would face another version of the Problem of the Missing Agent.

About moving your fingers as you are typing on your keyboard: I think we are in agreement, for I would say that as an agent, you cause the completed *effect* of your typing, which in this case is the words that comprise your comment. But, and crucially, you cause this result *by doing something*, namely, by moving your fingers in that way. Hence, as an agent, you are the cause of an effect—the words that comprise your comment—and you are causing this effect by acting in some particular manner—by moving your fingers as you do. On my view, we find intentional actions all the way down to the basic or fundamental level. But perhaps you would agree with this?

The comments to this entry are closed.

Books about Agency

3QD Prize 2014: Marcus Arvan