Blog Coordinator

« Still A Compatibility-Monist, After All These Years? | Main | Dennett Reviews Waller »



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

What I think, Saul, is that you have just discovered the perfect antidote to HD. Why in the world would I lend any credence at all to my post hoc reflections? In fact, why would I Monday morning quarterback myself on this score in the first place? Everything was settled in the temporal vicinity of my choice. I was then certain of what was happening inside of me- if anything the passage of time would distort my understanding of that situation. It‘s not like I was acting then in haste or hotheadedly, where I might have a good reason for questioning my judgment. Nor do I have any reason to accuse myself of wishful thinking. I wasn’t walking around then hoping to prove HD wrong, but sensing that it was true. The fact is I couldn’t even bring myself to entertain that notion as I was trying to figure out what to do. And here’s something else. At the very same time that my post hoc doubts begin to creep in, I will also know full well that the next time I am faced with a decision they will be disregarded as before. That feeling of freedom is natural; it accompanies me every time I am making a “normal” choice. My Monday morning quarterbacking seems contrived and otiose in comparison. Nothing requires it of us, explaining HD’s minority status, even amongst philosophers. I have always thought that Reid’s point was extremely telling in this regard: if I could be wrong when it comes to something as simple, natural, and obvious as my own freedom of choice, then Katie bar the door regarding all other forms of skepticism.

Saul--very interesting question. Metaphysically you're right; mere passage of time should be irrelevant as to whether you are/were really free in a given instance. But time certainly does have an effect on our perspectives about decisions. Think of any present time as that point between forward and rearward light-cones. What might be a locally future choice you must make in that narrow present with what you know drawn from the past that is now relevant to your decision, but contemporary/simultaneous facts are not available to you even if they might (after the fact) have been relevant to your decision if you'd known them. But later when that decision is in your rearward light-cone, such surrounding circumstances are available to you as past knowledge and you can better appreciate the limitations imposed on you at that previous time. I can see how that later broader perspective might ally with a judgement that you were limited/determined to choose as you did. Light-cones are also epistemic "enlightenment" cones in such cases.

No, Professor White, lamenting unavailable information is certainly not the same thing as seeing one's decision as unavoidable/inevitable. You are talking about an epistemic limitation, not a metaphysical one. The sort of facts that would have been "relevant to (a) decision" would have nothing to do with the question of determinism. Looking through my rearward light cone, I might say something like 'Had a known Jones was injured, I would not have bet so heavily on his team'. But this lament would not make me sense that I had to bet as I did; my ignorance will still seem consistent with an alternative course of action. After all, I still have my memory of how things looked from my forward light-cone, from which I saw nothing constraining my choice, the matter depending solely upon me, and the world awaiting my fiat.

Robert - the argument from introspection cannot prove determinism wrong. There is no reason to think that we are always able to know all the causes of our behavior. But I did not claim otherwise. On the contrary, I said "I can say that I believe that, whatever I will end up doing, it will have been determined that I will do it, and if determinism is the case, that will be true". I am not trying to prove the existence of libertarian free will (in the relevant forms, I even think that it is incoherent). My post says that the phenomenology of choice seems to exclude a HD stance when choosing and acting, but not after the fact. This seems to me interesting. For example, it arguably means that HDs cannot fully function as HDs at the time of action. If you think otherwise, tell me how to reply to Zelda. Does this matter? Well, it seems to, let's all talk about it.

Alan - perhaps that's all there is to it, we just know more after the fact. But then, once I take back with me into real-time choice the new gained insight of the possibility of HD, but still cannot apply it, then what should I do?

"how to reply to Zelda" :

My earlier action was unthinking.

I was not myself at the time.

I have previously observed such a tendency in myself, and am taking steps to avoid it in future.

I am consistently rude to everyone like that - it isn't anything personal. Maybe I have Asperger's.

Saul - I still believe that, being an HD, I am functioning as an HD at any given moment. Even as I type these words I know full well that I am not "choosing" them in the libertarian sense. I could not have chosen to type an alternate phrase. This is where the concept of time really kicks into gear. I just took a sip of coffee after writing the last sentence. At the precise moment that I lifted the cup I thought to myself could I have done otherwise? The answer of course is "no". It felt as if I chose to pick the cup up because my "awareness" of that action occurred post facto of its occurrence. I use the term "awareness" for lack of a better word in this scenario.


I wonder if we as philosophers are sometimes subject to something like an akrasia about beliefs. Clearly there are lots of instances here in the States of outright denial of reasonable beliefs based on unmovable deep motives (global-warming/evolution denial). Isn't it plausible that in close instances of epistemic warrant (such as we find among at least many wildly different FW positions) that we sometimes exhibit a "weakness of will" with respect to plausible, maybe even evidentially-favored beliefs? One aspect about FW disputes that hasn't been explored nearly enough is what I've called the foundational values or "world-view" problem (yeah, it was in a chapter about FW in a pop culture work "The X-Files and Philsosophy" but hey, if it applies to Mulder and Scully, it could apply to us): even as self-described seekers of truth can we really suspend beliefs about our basic values/world-views in order to be receptive to evidence, especially in close cases of epistemic warrant? I have my doubts about that.

Thanks for making me think about all this Saul.

David - "My earlier action was unthinking"; "I was not myself at the time" - that's exactly my point, a hard determinist can say such things, use such excuses, about the PAST, even right after acting, but not (under normal conditions) AT THE TIME of acting. He does NOT have to be nasty to Zelda, he CAN just control himself and stop. If he says to Zelda otherwise then he is deceiving her. The phenomenology doesn't lend itself to a HD interpretation of the absence of freedom and responsibility. This in no way establishes libertarian FW (how could it?), and for the sake of discussion we can just assume determinism. What I am arguing is that a hard determinist perspective seems impossible, at the time of acting.

R.D. - I don't see it. You start lifting the cup, you hold it, you think: "I can set it down"; you do, you think "I can lift it, but also not". That's how it seems to me (I really tried, just now, with a lovely brew of Master Liu’s Autumn Moon tea). After a moment, I can understand why I raised it or not (e.g. I was trying to make a point), but I don't see what is absent AT THE TIME of cup handling. What stops you from being responsible for your action? Does this support compatibilism? Perhaps. Compatibilism by default, coupled by an illusory libertarian phenomenology? Maybe. In any case, HD seems out.

Alan – your doubts here are reasonable, and in the FW case, with so much at stake (personally, morally, socially) we probably deceive ourselves more than usual. But which way do you think it goes, in this case? Should we dismiss the phenomenology at the time of choosing and acting as a quirk of human nature which means nothing? Or does it say something deep about HD? Assume beings with no back-view-mirror, i.e. all the time just choosing and acting, with no reflection. HD for them cannot enter it seems, does that teach us anything? Or is the after-the-fact perspective, where HD is certainly a possible interpretation, more dominant? Is philosophical understanding here dependent on avoiding the at-the-time-of-action OR the after-the-fact perspectives, OR does it come from combining both? And what does that dual perspective show?

Saul, I think that the Zelda example shows why a dualist position is necessary, but I don't think it is time that is the key to the question why we experience a different situation before and after the decision. Rather, I think it is what Thomas Nagel discusses in his book The view from nowhere, "how to combine the perspective of a particular person inside the world with an objective view of that same world, the person and his viewpoint included", the inside view and the outside view. Before a decision we are fully occupied with figuring out what to do, we don't have any use to know that the world is determined. After the decision we can look back and try to analyse why we did as we did and then we notice that we could not have done otherwise, we handled the information that was available, our desires, values, knowledge, memories etc.

It seems impossible to take the outside view before the decision because the outside view will then become an inside view. After I raised my arm to drink a cup of tea, I can evaluate the reasons why I did it, but if I try to do that before I make the decision I get into trouble. I may reason Even if it seems to me that I can or can't raise my arm, the world is determined so I can only do one thing. But this, the outside view, is of no help so after having thought about that a while I just decides to raise my arm because I want the tea. Afterwards when I try to explain why I raised my arm I can describe how I tried to incorporate the outside view but that only added confusion.

This seems to be kind of a system issue, is it possible for a process to incorporate an outside view of the process in the process itself?

Hi Saul, great posts! Thanks for stimulating the discussion. I don't want to sidetrack that discussion, so we can ignore the following point. But I want to point out that the way you present things in this post is great for an "explaining-away-incompatibilist-intuitions" theorist like me.

You write: "I cannot (in a normal case) view myself as being carried away by the forces determining me" and "The typical phenomenology of choice is not of being swept away by forces beyond one’s control." (You also talk as if determinism suggests we are passive not active and our choices are inevitable.)

Here, you use the "bypassing" language common in incompatibilist's informal presentations of the problem of determinism, language and imagery that I think explain why people have allegedly (but not really) incompatibilist-looking intuitions. The image is dualistic. We are one thing and the forces governing the universe are another thing, and if they are "in control," they push us around so that we are not in control. (I can offer similar imagery from van Inwagen and others--Ginet's image of the Disney ride where we think we control the steering wheel but don't is a classic.)

But this imagery is misleading. Determinism without dualism means there's just the one system and we are (barring evidence otherwise) important parts of it--we are difference-makers. We are not swept along by forces beyond our control. Our consciously envisioning various futures, each possible depending on what we choose, and our choice to realize one of them make a difference in what happens. Yes, yes, all that deliberation and decision-making may be part of a deterministic system, but that does not make them sea foam pushed along by the waves of causation.

Time is indeed essential. As we begin deliberating we envision many conditionals and we do not know which antecedent is true: If I decide A, then (likely) X; If I decide B, then (likely) Y; etc. And we do not yet know what we will decide. After we've made a decision, we know which antecedent is true and we also know more about what made it true (i.e., the course of our deliberation). Deliberation itself does not feel forced upon us (we typically do not feel swept along). But neither does deliberation (typically) feel like a process of creation (ex nihilo!). (Of course, it's possible my phenomenology is just a barren wasteland compared to some libertarians...)

'There is no reason to think that we are always able to know all the causes of our behavior.'


Isn’t the whole question here whether introspection affords us knowledge of the workings of our minds as we make choices? Why should we be unwilling to consider what it reveals as a solid piece of evidence against HD and in favor of LFW? Why should I simply dismiss my remembered inability to adopt HD and lack of passivity when presented with the deliverances of neuroscience supporting Determinism? If I automatically favor the latter, am I not guilty of Scientism, not to mention bad faith? As Reid points out, isn’t such a philosophy untenable? If we can be wrong about the workings of our own minds, if FW is an illusion, then how can we trust our senses to provide accurate impressions of the external world?

Thank you for your yeoman's work on our behalf.

Thanks to everyone for their posts.

Gunnar - I also think that the dual results here help a compatibility-dualist perspective, and since that is a view I hold on other grounds, I find this welcome. The idea of a system issue seems promising and worthy of further development in this context.

Nagel's discussion is of course relevant here, but I am inclined to think that the important cut is not between the inner/outer but between the while-acting/in-retrospect. If you see me being nasty to Zelda, and you ask me to stop, then, if I say to you "I cannot" then (in a normal case) that will be false - I CAN, and YOU can recognize this. Partly it is because you reasonably assume that I am a normal human being, and you know what it's like to be one. But also, if you take a gun and point it at me, I will at once stop. This will show you what you already very much suspected, namely, that I have a motivation problem - I don't want to stop being nasty to Zelda - rather than an inability. But in any case, all this was not done in the first person, you can reasonably assume that I CAN stop being nasty to Zelda, even if you have no access to my internal awareness.

Eddy - one of the advantages of being a compatibility-dualist is that you are happy when compatibilists score points against HDs, and when you can help them out (and vice versa) :) Does our when-choosing sense of freedom feel like compatibilist free will or like (an illusion of) libertarian free will? Here I am not sure. But in any case I agree that it doesn't lend support to the idea of helplessness. However, your problem as a compatibilist still resisting the truth of compatibility-dualism is to account for the implications of the retrospective view. Here I will often view myself as not having been free, once I recognize why I did what I did, and it can make sense to challenge one's moral responsibility, in retrospect. So even if the phenomenology of choice supports compatibilism, surely you would not want to say that compatibilism can only be sustained if we discount what we can come to believe in retrospect.

Robert - I think that a robust libertarian free will (not the Kane or Ekstrom variety which is just compatibilism with some indeterminism thrown in, without real control over it) is incoherent, it simply doesn't make sense. So no introspection could grasp it; to the extent that it feels like libertarian free will when we choose, that must be an illusion. In any case, as libertarians like van Inwagen have acknowledged, there is no reason to think that if our choices are determined, we would necessarily know it. So I simply don't see the attempt to prove the existence of LFW through introspection as promising.


But there is a time at which LFW does not seem incoherent to you and that is when you choose. There is otherwise no time problem to be considered. You generate the problem by saying that while you are harassing poor Zelda your conduct does seem avoidable, within your control, and determined by nothing but your own fiat (ruling out a Compatibilist interpretation). When you say that LFW is incoherent, then, you must be speaking from the post hoc perspective. The time problem only becomes pressing if we take seriously both perspectives. If you are going to straight away dismiss the "while acting" perspectice, there is no time problem. But the fact is you don't and you can't, at least until some time has passed. And then my question becomes, why should we put more stock in reflection than immediate awareness?

I also believe that you are failing to appreciate Reid's warning: once you dismiss as illusory LFW you are inviting a form of skepticism inimimical to science itself.

AS for VI, I would ask why wouldn't I know something as important as the source of my choices? How could I be so radically deceived as to think that I can control myself when in fact I am powerless? I might just as soon deny my own existence. Better: to deny self-control is to prove its existence. Only a being in control of himself could form the the thought that he lacks self-control in the face of the overwhelming evidence (you yourself cite) in favor of self-control extant while choices are being made. This Scotistic thought experiment is the gist of a disproof of Illusionism: unless we exercise self-control we cannot pretend to believe in FW per the Illusionist's advice.

Hi Saul,

Another interesting post!

I am not sure that the fundamental point here concerns a temporal distinction. In fact, it seems to me that there is a different distinction that you have drawn that might be more fundamental. The distinction is between (a) an agent’s actively controlling an action while she is in the midst of performing it, and (b) an agent’s retrospectively thinking about an action after its completion.

It seems to me that this distinction concerns different relations in which an agent can stand with her own actions. On the one hand, there is a relation that we might call “active control”, and this relation obtains while the agent is performing the action in question. In the case of typical and rather simplistic bodily actions where we have set aside the question of freedom, the sense in which an agent enjoys “active control” over her own action can be understood in terms of the way in which she initiates and sustains the activation of the relevant bodily capacities as she performs that action.

In contrast, the way in which the agent relates to her own action when it is no longer occurring and she is retrospectively thinking about it is very different: in such cases, the performance of that action has ended, she is no longer initiating or sustaining the activation of the relevant bodily capacities therein, and so she is no longer enjoying that form of "active control".

Crucially, these distinct relations do not concern the mere passage of time itself, but something more fundamental. Would you agree?

Robert - I suspect we will still disagree on this one in the end. I don't share the Reidian fear of overwhelming philosophical skepticism if LFW is denied (as you know, I think that it is probably a useful public illusion, but not for philosophers). I find the philosophical arguments against the argument from introspection for LFW too strong: first, I am convinced by the Galen Strawson type argument that a robust LFW is simply incoherent. Second, the idea that we could be sure we know the causes of our own actions just seems implausible, as many libertarians concede. Hypnosis is one example - apparently people can be made to do things under hypnosis, and it will be undeniable that it is the H which lies behind their actions, but they have no awareness of this, and make up excuses for their odd behavior. So I do think that we need to pay attention to the phenomenology of choice, hence my post, but also that taking it to prove LFW is the wrong way to go.

Michael - I agree, I didn't really think that the mere passage of time makes the difference, and what you say sounds plausible.

GENERAL POINTS - let's review a bit what we have come up with. I would also like to connect this to some of my general interests on the free will problem, in Free Will and Illusion (compatibility-dualism and Illusionism) and more recently:

(1) The phenomenology of choosing "feels" free. Assuming LFW is indeed out, then it is either compatibilism or an illusion of libertarian free will. ILLUSION SEEMS DOMINANT IN HUMAN EXPERIENCE HERE. Compatibilism seems much more natural in retrospect, as the fruit of retrospective experience - or as a normative stipulation. It would be good to have compatibilist attempts to model the phenomenology at the time of choice.

(2) As you know from the discussion of my "funishing" piece, I have grave doubts whether hard determinism can be seriously applied at all. As far as I can see, the phenomenology of choice (e.g. my interaction with Zelda story) shows that HARD DETERMINISM IS IN TROUBLE HERE AS WELL. It seems very difficult to deny free will and moral responsibility while choosing and acting; perhaps impossible. What does this mean for hard determinism?

(3) The striking contrast between the results in the when-choosing and in-retrospect perspectives (in normal cases, and at least frequently), SEEMS TO LEND FURTHER TENTATIVE SUPPORT TO THE IDEA THAT WE NEED TO COMBINE COMPATIBILISM AND HARD DETERMINISM. Time might be one way in which the cut is made, i.e. it might be more plausible to be a compatibilist when acting (and judging people at the time of action, and the like), while HD could recover in retrospect.

(4) Beyond the seeming role of illusion in the phenomenology of choice, ILLUSION SEEMS TO PLAY A CENTRAL ROLE IN OTHER WAYS AS WELL. For example, living with a "time-based" dualism seems very problematic, and we can see how lack of awareness of this could, in practice, be helpful. Likewise, if HDs concede that the phenomenology of choice is inconsistent with HD, then HDs would seem to depend on a view that the phenomenology of human choice is illusory. So the way to discount the importance of this result for HD lies through enlarging the role of illusion in our lives.

All these points need further development, and a blog post is not the place for it. But I hope you find some of them suggestive, and that (whether here or in the future) they will generate further discussion.

Turn your scenerio in "Is it about time" on its head. Suppose you are "overCOME" by rage (COME from where?) because Zelda did something that infuriated you and you got in her face big time. You're not even sure why what she did hit such a raw nerve. It was hurtful on any account, but with you, it was a blow below the belt. So you lash out. You stew for several days trying to reconscile what she did with how you responsed. You acknowledge you were out of control...but it had to be done. She had to be taught a lesson.
A week later, with the help of your psychiatrist, you determine that a childhood trauma deep in your unconscious stirred up unknown emotions that drove your behavior.
You recognize that a better decision would have been to sit down with Zelda and tell her how hurtful what she did was to you and perhaps (after reflection) even why.
This better decision was not even apparent to you at the time. However, you hold out the possibility that the next time your psyche is attacked you will pause long enough to consider an explanation for your deeply hurt feelings that does not result in over the top retribution. I'm not sure if this demonstrates free will or determinism. I call it growth towards maturity.

Saul, thanks for the comments.
I don't remember Nagel's definition of an inner/outer view by as I use it the inner view will include all agents' actions in the actual case. So if I take a gun my action is also included in the inner view and neither you nor I can take the outside view and examine the case using an external perspective including determinism.

If you try to include the outer view, the thoughts of you and an outsider may be like this
Your (Saul's) inside view: I act (or decide or deliberates etc)
The outsider's view: S acts.
Your view of the outsider: The outsider thinks that S acts.
The outsider's view: S thinks The outsider thinks that S acts.
Your view of the outsider: The outsider thinks S thinks The outsider thinks that S acts.

This informally indicates that it impossible to you to incorporate the outside view. After the action is completed you can take the outside view because now you are not interfering in the decision. During the decision process only agents that are not included in the process can examine it. But time is not the critical factor. For instance, if there is an omniscient observer that knows everything about you and your local deterministic universe he/she/it may be able to forecast the result of the decision process even before you made the decision. This forces me to conclude that the important cut is between while-acting and while-not-acting. This can be seen with an inner/outer view or in a before/after view.
However, I am not certain that this cut is the same cut as you discuss in your last post, between C and HD.

Lindsey - I can accept this, there are cases like that and the retrospective perspective can induce growth and overcoming. My aim in the Zelda story was however different.

Gunnar - you make a number of interesting points. But note that in my gun test it is the external person who uses the gun, to see whether (in the Zelda example) I indeed cannot avoid doing what I say I cannot, namely, being nasty to Zelda. So I still claim that while we do not have access to another person's inner processes, we can reasonably believe, and even test, whether that other person is helpless during action. The gun example is too crude, for even addicted people (whom compatibilists would not classify as by and large free) can be deterred if one points a gun at them, although in other situations they would find it very hard to resist temptation. But if we remain with normal people then I think that the gun example does show that one can find out whether another person is helpless, even if that person claims to be.

Perhaps, as Eddy jokes, my phenomenology is (also) just a barren wasteland compared to some libertarians. But I don't see my freedom disappearing in the rear view mirror. I can see the reasons why I did what I did, but that's not a *problem*. Those are usually just the very reasons I was searching for, or weighing up, while I was still deliberating.

Even if I see that my decision was wrong - say there was an excellent reason to do C, while I foolishly deliberated between A and B - that doesn't usually make me feel my choice was unfree. Only if I was literally obsessed with A and B, or something like that, will this feel in retrospect like unfreedom.

I hope this is all exactly what you'd expect a compatibilist to say. But I thought it worth a mention, anyway.

"...the phenomenology of choice seems to exclude a HD stance when choosing and acting, but not after the fact."

The simpler explanation is that you can't introspect and make decisions at the same time because it is physically impossible. There are a few papers, I now find, on neural bases of "metacognition" ie introspective knowledge;jsessionid=4igpTzh5BSCEDiHK8VAB.8

The latter asked “How much personal control did you feel when thinking the name of the color?”

So, you can only know yourself predictively or retrodictively.

Paul - that's fine, this post didn't really present much of a case against compatibilism (the next post will focus on compatibilism). What seems to me striking is that hard determinism does not seem to make sense when choosing and acting (in typical cases), but it can after the fact, even if it is but a short time later. Whether we would be attracted by a hard determinist interpretation of the past event or not will depend on many factors. But all I need here is that the HD mode seems available, in retrospect, but not when choosing and acting.

David - that seems to me too easy. We are not speaking about trying to balance two balls on your head while standing on a third, a state where indeed too much reflection is precluded, but about a fairly lengthy interaction with another person, where normal awareness of what one is doing, and of how it relates to what one believes, does seem possible. If we go back to my Zelda example, we can imagine that my discussion with her goes on for half an hour. Let's say that despite my being nasty to her, she does not leave, because she does not want to lose her job. So I have plenty of time to become aware of what I am doing, reflect whether it is justified, decide to continue or back off, and what not. There is a question whether the phenomenology here is compatibilist-like in its quality, or a libertarian illusion, but let's set this aside. In any case, as far as I can see there is no problem for me to be a compatibilist, in such a situation. I can say to myself, "you are free, after all, to let her be, are you sure you want to do this?". Or, when I perceive that Zelda is being hurt by what I am saying, I can say to myself, AT THE TIME ALL THIS IS HAPPENING, "you know that you are responsible for the fact that she is about to cry, do you feel you wish to continue", and the like. On the other hand, I don't see how a hard determinist can engage in choice and action WHILE HIS HD IS "ON". As I said above, when choosing and acting (in normal cases), the sense of free control and responsibility seem unavoidable. So there seems to be a difficulty here that is specific to HD.

The comments to this entry are closed.

Books about Agency

3QD Prize 2014: Marcus Arvan