Why Two Source-Incompatibilist Arguments are Question-Begging

This is going to be my last post. I want to thank everyone for this terrific opportunity and for the helpful comments I got. I really appreciate your having taken the time to think through these (in some cases, quite underdeveloped) ideas with me. Thanks to Thomas, especially, for the invitation, and for doing such a great job at running this blog. This has been my first serious blogging experience, and I feel like it has been a very positive one. I hope you enjoy my last post!

Many of us are interested in “the problem of determinism and free will.” But what is the problem of determinism and free will? Last year Joe Campbell came to Arizona (as an Arizona alum) and we discussed some of his work on free will in a workshop. As most of you probably know, Joe wrote a very interesting paper in Analysis (“Free Will and the Necessity of the Past”) where he argued that the consequence argument is flawed or at best proves a weaker conclusion than incompatibilism. My reaction to Joe’s objection to the consequence argument (a reaction that I share with others, although not with everyone, by all means) was, and still is, that we shouldn’t conclude that this is a problem for incompatibilism (or for incompatibilists that rely on the consequence argument) but, rather, that incompatibilism is a weaker thesis than we have typically taken it to be. And this is because the threat to our freedom is not determinism per se, or not just determinism, but determinism plus something else. Let’s call this claim “the demotion claim” about the problem of determinism and free will.

I know that it’s a debated issue whether the demotion claim is really true, and I’m not going to try to persuade everyone of its truth (although we can of course discuss this, if you want). Instead, I’ll briefly explain my reasons for believing in the demotion claim and what I take the “missing ingredient” in the problem of determinism and free will to be, and then I’ll argue that, on the assumption that my thoughts about the demotion claim are true, it follows that two important source-incompatibilist arguments fail. (“Source-incompatibilist” arguments are arguments that attempt to show that determinism is a threat to our free will, not on the basis of alternative-possibilities considerations, but on the basis of considerations having to do with actual sources or actual causal histories.) The moral will be that, whereas the demotion claim may be bad news for leeway-compatibilists, it’s good news (in fact, great news!) for source-compatibilists, in particular, for actual-sequence theorists like me.

According to at least some plausible definitions of determinism, if determinism is true, what we do now is determined by past events, but it’s also determined by future events. Now, whereas we think that the existence of past-determination is a threat to our freedom, we don’t think that the existence of future-determination is a similar threat. What’s the difference, if in both cases there’s determination (by other events and the laws)? Presumably, the difference is that, whereas we can causally influence the future by acting in the present, we cannot similarly influence the past. Now, this suggests that the threat to our freedom is not determination per se, but determination by events that are beyond our causal reach. That is, the problem arises for us because there exist events that are such that, if determinism is true, they make our acts inevitable and they fall outside of the sphere of our causal influence (and thus are out of our control).

So this is basically why I think that the demotion claim is true and what I think the missing ingredient in the problem of determinism and free will is. (I warned you my explanation here would be brief!) Now consider two main “source-incompatibilist” arguments:

1. Ultimacy Arguments

 Roughly, these go as follows:

(a) Determinism entails that we are not the ultimate sources of our acts.

(b) We can’t be free unless we are the ultimate sources of our acts.

 Therefore,

(c) Determinism rules out freedom.

Different understandings of the key “ultimate sourcehood” condition result in different versions of the argument. However, all plausible understandings of the ultimacy condition seem to entail the following “minimal” condition:

Ultimacy (Causal Access): Being the ultimate sources of our acts requires having some kind of causal access to all the actual sufficient sources of our acts.

I say this is a minimal condition because, according to some formulations of the ultimacy condition (for example, Kane’s), ultimacy requires more than that (it requires being responsible for the actual sufficient sources; but, then again, he formulates the whole ultimacy requirement in terms of responsibility, not freedom).

Now, assuming my thoughts on the demotion claim are true, if ultimacy arguments are to establish incompatibilism, they should look more like this:

(d) If our acts are determined by factors beyond our causal reach, we are never the ultimate sources of our acts.

(e) We can’t be free unless we are the ultimate sources of our acts.

Therefore,

(f) If our acts are determined by factors beyond our causal reach, we cannot be free.

But, if, as we have seen, ultimacy is understood in terms of Ultimacy (Causal Access) or something stronger, the second premise of this argument is blatantly question-begging (and the first premise is true, but trivially so). The second premise says that freedom requires (at least) causal access to the actual sufficient sources of our acts, where the absence of causal access of that kind is something that’s needed for the problem of determinism and free will to even arise. 

2. Direct arguments (or “transfer of non-responsibility” arguments)

 Roughly, these go as follows:

(g) We are not responsible for the remote causes of our acts.

(h) If we are not responsible for X, and we are not responsible for the fact that, if X then Y, then we are not responsible for Y. (Transfer of non-responsibility principle)

Therefore,

(i) If determinism is true, we are not responsible for our acts.

As Ravizza, and then Fischer and Ravizza, noted, as it stands the argument doesn’t work against a source-compatibilist because Frankfurt cases and other overdetermination cases seem to undermine the transfer of non-responsibility principle (the agent in a Frankfurt case is not responsible for the presence of the neuroscientist, and he’s not responsible for the fact that, if the neuroscientist is present, he’ll make a certain choice, but he still seems responsible for his choice). As a result, “one path” versions of the transfer principle have been suggested (McKenna, Stump): 

(One-Path Transfer) If we are not responsible for X, where X is the only path that is (at some time) actually causally sufficient for Y, then we are also not responsible for Y.

The thought is that this version of the principle could still do the required work in the direct argument (and Frankfurt cases are not counterexamples to it). But note that, contraposing, this principle is equivalent to:

If we are responsible for Y, and if X is the only path that is (at some time) actually causally sufficient for Y, then we are responsible for X.

And, assuming that we could only be responsible for that single path X to the extent that we had some kind of causal access to it, this principle entails Ultimacy (Causal Access). Now, as we have seen, assuming the demotion claim is true, the appeal to Ultimacy (Causal Access) is question-begging. As a result, direct arguments that rest on one-path principles (which are the only direct arguments that could work in this context) are also question-begging.

In sum: these arguments presuppose, at some level, that freedom requires causal access to the actual sufficient sources of our acts. However, if the demotion claim is true, the problem of determinism and free will only arises to the extent that we assume that we lack causal access to some of the actual sufficient sources of our acts. As a result, the arguments are question-begging. 

Sorry if this is a bit sketchy, but hopefully you get the idea!

Comments

[John Fischer]

Thanks for this post, and all your interesting posts this month. It is cool to have at least one ally in this hard, cruel philodophical world!

Just a quick historical note: I have also argued that the One-Path Principle, developed in reply to the Ravizza/Fischer critique of the Direct Argument, is question-begging: "The Transfer of Non-Responsibility" (reprinted in MY WAY).

I essentially agree with you that it is hard to formulate any Direct Argument or Ultimate Sourcehood Argument in a way that seems clearly non-question-begging (or, given the slipperiness of the notion of begging the question), dialectically helpful (or likely to advance the debate, perhaps appealing to thoughtful, uncommitted agnostics). On the other hand, some have also of course argued that the Consequence Argument begs the question, or is dialectically unhelpful; this group includes great philosophers such as David Lewis and Gary Watson, so go figure.

In my view, and sorry to simply pontificate (as a Jewish atheist, may one permissibly pontificate??), I think that dialectical issues are incredibly important--and delicate--in just about every are of the free will/moral responsibiliity debates. Taking a klunky methodological or dialectical view will spoil one's arguments. Adopting a subtler, more nuanced approach to these issues often reveals additional complexity and makes me at least a bit less confident of my own views (but, fortunately, also less persuaded by at least some others' views).

[V. Alan White]

Hi Carolina--another thoroughly enjoyable, challenging, and insightful post. Thank you so much for your terrific work this month.

When you say that “what we do now is. . .determined by future events” I take it you are referring to the possibility of deducing present acts retrodictively from future events and laws much as we can predict them likewise from past events and laws. As Salmon pointed out, this logical relationship of event-subsumption-under-law is conceptually separate from questions of causation, which must invoke explanatory temporal asymmetry. So any account of determinism as specifically involving causation must align along cause-before and effect-(simultaneous or)-after relations, and that in part shows why causal explanations are temporally reversible, and why retrodiction isn’t actually a deterministic relationship even if it is a logical one associated with determinism.

So, to say that “the threat to our freedom is not determination per se, but determination by events that are beyond our causal reach” uses “determination” in place of backwards-or-forwards deducibility and subtly asserts the asymmetry by referring to past events (only) as “beyond our causal reach”. But that then says that unidirectional causality threatens freedom, and that seems (to me) much of what the usual-suspect CA is about. Is this not some version of Beta otherwise stated?

Okay, besides that. Now to your criticism. It seems to me that your ultimacy arguments rely on two senses of “factors beyond our causal reach”. One is past-oriented, noting that we can’t affect the past. The other is atemporal, referring indifferently to anytime that we can’t affect things causally. That explains the first sense used in (d) and the second in (e). So, given the equivocation, this supports your petitio claim. That’s how I see it.

I agree with your direct argument approach as well (even preserving the equivocation point) but wonder if someone would object that your contraposition as logically posed is against a proposition that may be parsed as temporally asymmetric. It’s not clear that such propositions are susceptible to contraposition preserving truth, aside causal/explanatory assumptions.

Thanks again for your great work.

[Carolina Sartorio]

Hi John,
Yes, it seems like we agree about this too. In fact, if you and I are both right about the direct argument in particular (or its one-path version) then there may be more than one reason why the argument is question-begging. You claim it’s question-begging because of considerations having to do with the move from two-path overdetermination cases to one-path cases being unmotivated or wrong-headed; I, on the other hand, claim it’s question-begging because, once one of the conditions or circumstances that need to obtain for the problem of determinism and free will to arise is revealed, we see that the principle simply assumes that we can’t be free in those circumstances. So perhaps there is more than one reason why the argument is question-begging, I’m not sure. In principle, I’m inclined to say: the more the merrier!

[John Fischer]

Hi Carolina,

Yes, not only do we have "the more the merrier", but a felicitous kind of over-determination, of which we are both so fond!!!

[Chandra]

Hi Carolina,

Thanks for another thought provoking post, and for a really terrific month of blogging!

I think I agree with most everything you say in your post, but I am wary about the conclusion you want to draw--“on the assumption that my thoughts about the demotion claim are true, it follows that two important source-incompatibilist arguments fail.” Here is a softer way of stating your conclusion, and perhaps you can let me know if you agree with this way of putting it.

Let me start by retracing your arguments a bit. You agree with Joe that the CA can’t deliver incompatibilism full stop, but rather only the weaker claim that free will is incompatible with determinism in those worlds with a remote past. And then you say that this in turn suggests that “the threat to our freedom is not determination per se, but determination by events that are beyond our causal reach.” Call this latter claim Causal Reach. You then say that certain key source incompatibilist arguments (Ultimacy Arguments and the versions of the Direct Argument) seem to require Causal Reach or something very close to it as the key premise. You conclude that these arguments are question-begging, and thus they fail. Correct me if I am wrong or if I have misunderstood the flow of your argument.

If the above is right, I think I feel uneasy about saying that these source incompatibilist arguments “fail”. Often when people say an argument fails, they mean that the premises are false, or that the premises do not support the conclusion. But you are not saying this. Rather, I think a softer lesson to draw is something more like this: We previously thought the CA, Ultimacy Arguments, and versions of the Direct Argument each provided *independent* grounds to support incompatibilism. But if the CA yields only a demoted conclusion, i.e. Causal Reach, then it follows that other arguments such as Ultimacy Arguments and versions of the Direct Argument aren’t after all so independent, since these latter arguments use Causal Reach as the main premise. So if incompatibilists succeed in showing the truth of the demoted CA, then they don’t add any *new evidence* for their view with Ultimacy Arguments and certain versions of the Direct Argument.

What do you think of this slightly more mellow version of your own conclusion? Does it seem right or have I lost something important?

[Carolina Sartorio]

Hi Alan,
Thanks, yes, I was thinking that there are at least some standard definitions of determinism that are “neutral” in that respect—they entail that the future determines the past as much as the past determines the future (van Inwagen has a definition like that; Lewis has another). Actually, one thing I was wondering about is, how common is it to use one definition instead of the other? Have people thought that there could be different problems involving determinism and free will depending on which definition we pick (and, perhaps, that the different problems have different answers)?

At any rate, I used the case of the future because, assuming one of those definitions of determinism, it’s very easy to see that determination itself cannot be the culprit. But this is not really essential for my main argument to work. Imagine you define determinism in a way that the past determines the future but the future doesn’t determine the past. Still, there are good reasons to think that determination by itself doesn’t create a threat to our freedom. For imagine that we had causal access to the past (imagine, for example, that we could travel back in time and causally influence the events in the past). Then we could have (at least in principle) control over the past (in the same way we now have some control over the future), and thus the fact that the past determines the present wouldn’t be enough to threaten our freedom.

On your comments on the two arguments. First, on the direct argument: Are there any good reasons to think that the transfer principle doesn’t contrapose? It’s typically presented as something like an inference rule.

On the ultimacy argument: I’m not sure I follow what you say about the “equivocation” in the argument, and that equivocation supporting the petitio claim. My thought was this: If we assume the demotion claim, and we understand the ultimacy requirement along the lines of Ultimacy (Causal Access) (or something stronger), then the first premise becomes trivial. It basically says:

If our acts are determined by (some) factors beyond our causal reach, then we don’t have causal access to all of the actual sufficient sources of our acts.

This is trivial, because the determining factors beyond our causal reach can only be certain causes of our acts, and their being “beyond our causal reach” means that we don’t have causal access to them.

And then the second premise carries all the weight, which is why the argument ends up being question-begging. The second premise now says:

We cannot be free unless we have causal access to all the actual sufficient sources of our acts.

But this is just begging the question, for it simply states incompatibilism (as I think we should understand the thesis of incompatibilism)…

Last-minute thoughts (that I had after I wrote all this):
Perhaps I see what you mean by the equivocation now. Assuming a neutral formulation of determinism (one that doesn’t presuppose that determiners are causes), there are two possible ways of understanding incompatibilism, in accordance with the demotion claim:

1. Determination by factors (ANY factors) beyond our causal reach is incompatible with freedom.
2. Determination by CAUSES beyond our causal reach is incompatible with freedom.

The second premise in the argument is just incompatibilism only if we understand incompatibilism as claim 2. Does the argument still beg the question if incompatibilism is understood in the stronger sense, as claim 1? Arguably, yes, because it is hard to see how anyone who doesn’t already believe in 1’s truth would be tempted to believe in premise 2. But I'm not totally sure, it's an interesting question.

[John Fischer]

I should have been a bit more careful.
I should have written that I have argued that *certain* restricted, or one-path principle, cannot be used in Direct Argumrnts without begging the question or at least being dialectically unhelpful.

Again: the dialectical issues are often extremely important, and also extremely delicate.

[Carolina Sartorio]

Chandra,

Thanks for your comment. Yes, I think that’s *pretty much* what I want to say. The arguments fail *in the context where they are typically offered*, a context where the incompatibilist is trying to offer reasons in favor of incompatibilism (reasons that could potentially move a reflective agnostic toward incompatibilism, say). I’m certainly not arguing that the premises are false or anything like that.

Let me add just a few clarifications, because I don’t think we agree about all the details. When I said “the threat to our freedom is not determination per se, but determination plus something else” [this is the claim you called “Causal Reach”] I of course didn’t intend to be suggesting that this was a real threat (since I’m a compatibilist), but only a potential threat (the kind of thing that should make someone think that there is a problem to be discussed). So in that sense Causal Reach wouldn’t be the conclusion of the demoted CA. (Also, the CA is an argument about freedom understood as the ability to do otherwise, but I had in mind a broader conception of freedom: freedom as the metaphysical component of responsibility, whatever that is.)
Finally, I don’t just think that the lesson we should draw from Joe’s objection to the CA is that we should take the conclusion of that argument to be weaker but, more generally, that the problem of determinism and free will is slightly different from what we took it to be (and thus compatibilism and incompatibilism themselves are also different theses from what we have taken them to be). So I guess I wouldn’t say that the problem with the ultimacy and direct arguments is that they’re “borrowing on” the demoted conclusion of the CA and thus are not independent arguments. The demoted conclusion is just incompatibilism, and the problem with those arguments is just that they’re assuming incompatibilism or something close to that.

But, in general terms, I think that we agree about what conclusions I should be drawing from my arguments. (Do we?)

[Carolina Sartorio]

Chandra: Sorry, I just realized I said something misleading. The demoted conclusion of the CA is not really incompatibilism, as I’m understanding incompatibilism (in terms of a broader conception of freedom, possibly different from the ability to do otherwise). But hopefully you can still see what I meant.

[Chandra]

Aha! I think I get it now. So you are saying that Joe’s argument doesn’t just reveal something about the proper conclusion of the CA, but on a more fundamental level changes our understanding of what incompatibilism is. So in your view Causal Reach is in a way baked into the very notion of incompatibilism. What Joe’s argument reveals is that incompatibilism is better understood as the thesis that free will is not compatible with determination *by events that are beyond our causal reach* (Is this right? Have I understood your view correctly now?). If this is correct, then I do see why you prefer to put things more starkly in terms of overt question-beggingness. If Causal Reach is indeed baked into the very notion of incompatibilism as you suggest, then Ultimacy Arguments and certain versions of the Direct Argument do seem to have a worrisome whiff of assuming their conclusions.

[Carolina Sartorio]

Yes, Chandra, that's *exactly* what I want to say :)

[V. Alan White]

Thanks so much for your response Carolina. And yes, your last reflections on how I took the equivocation to work is on target--I should have been more careful to frame it in terms of the "neutrality" of determination with respect to incompatibilism.

Thanks again for taking my thinking in new directions on all this.

[Jiajun Hu]

Hi Carolina,

Thank you so much for such an interesting post. And I really like and appreciate all your thoughtful posts and careful responses in the month.

I think I agree that source incompatibilism or direct argument is likely to involve something similar to begging the question. And I am not trying to defend source incompatibilism here. What interests me is your demotion claim, which makes me a little bit uncomfortable. I haven't been able to think it through. So I just report my thread of thoughts and hope you can help me sort it out.

First of all, I am not sure what is the best way to understand the term "determinism per se". If my present decision (more accurately, the decision I am going to make right now) is a free one, however, it will causally determine everything follows it (similar to God's free creation of a deterministic world). In such case, maybe we should say there is determinism per se. And this determinism per se surely doesn't threaten my freedom of making the current decision. But is it a relevant determinism for my present decision? For incompatibilist, not any remote piece of determinism would pose threat to our freedom. Considering I could make different decisions which produces different futures, and when I make the present decision the causal necessity hasn't obtained, such a determinism seems to be an irrelevant determinism even though my present decision is going to determine the future and the future determines the present decision. Strictly speaking, such kind of determinism per se may not be determinism at all with respect to my present decision.

So, can we find a case of relevant determinism per se in which I do have causal reach? This might be one (or not, I am not very sure). Suppose that the world is created by God and God allows it to be indeterministic until now. So my present decision is not determined by its causal history. However, God makes me believe that from now on the world will change to be a deterministic one which leads to an inevitable future event (e.g., justice prevail or eternal world peace). And the future event is inevitable not merely due to the causal determination, but also because of my present decision (if I make different decision, the inevitable future would fail to obtain). In this way, God makes me believe that my present decision (or the decision I am going to make) has already been determined by the fixed future.

In such a case, how will incompatibilist react? Do they feel their freedom is being threatened by the fixed future? As an incompatibilist, I want to say yes. But admittedly, it is a delicate situation. I am afraid that my intuition is no longer reliable here, not only because I don't believe in God but also because the modern worldview has already been framed by a physical and open-future idea about time and history. Maybe Christians or Christians in history have a better grasp of a fixed-future and teleological idea about time and history.

However, if I just honestly report my untrustworthy intuition, I will say this: when I view the necessity of my present decision from an external perspective (in which God and the fixed future plays the essential roles), I feel my freedom threatened. But I still have the feeling that I am in control and free when I start to think what I am going to decide anyway from the first person perspective. So it is a mixed feeling. But doesn't the antecedent causal determination also produce a similar mixed feeling for us? I guess my question is: is there any difference between determinism with causal reach and determinism without causal reach (perhaps, other than we are more susceptible to threat from fixed history than threat from fixed future due to a modernized worldview)?

Do you think this could be a legitimate worry? How do you think about it?

[Carolina Sartorio]

Hi Jiajun,

Thanks for your comments and questions.

By the claim that “determinism per se” is what threatens our freedom I meant the thought that what threatens our freedom is *just* the assumption that we live in a deterministic world; no other assumption is needed. But note that there’s some room for embracing this view while rejecting the idea that future-determination is a threat to our freedom. (It could be that the assumption of determinism always results in both past-determination and future-determination, but, even then, only past-determination is threatening. In my view, this is what a reasonable incompatibilist should say.)

Imagine that I make a certain important decision at T, say, I decide to kill a person in cold blood. As a result, the rest of my life goes a certain way (I’m sent to jail, I’m unhappy, etc.). If determinism (in its neutral formulation) is true, my decision at T is determined by the state of the world before I was born and the laws, but it’s also determined by the state of the world at that later time when I’m sent to jail (T+). Is my freedom at T threatened by the fact that my decision is determined by the state of the world and the laws at T+? Arguably not. The state of the world at that time includes certain important events (such as my being thrown in jail) that I brought about *by* making that decision at T. In most cases, those events wouldn’t even have taken place if I hadn’t made that decision. So the fact that they (together with other events and the laws) determine my decision doesn’t seem to be a good reason to think that I’m not free at T.

So now imagine, in line with your example, that our world was indeterministic up until time T and that at that time God turns it into a deterministic world. In that case my decision at T is not determined by any prior events, but only by future events. This is a good case to motivate the demotion claim, I think (assuming we can make sense of all the required assumptions of the case, of course). For, in this world too, there seems to be no good reason to think that the existence of future-determination is a threat to my freedom. Given that there is only future-determination (no past-determination) at time T, determinism doesn’t seem to be a threat to my freedom at T. This suggests that the threat to our freedom, when it exists, is not just determinism.

One of Campbell’s cases was interestingly similar to this. He imagined an agent who is born, as an adult, fully rational being, at the very beginning of time, and who makes a certain decision at that time. So his decision isn’t determined by prior events, only by future events. Again, I’d argue that the existence of future-determination is not a threat to his free will. But then, how can determinism be a threat to his free will, if the only form determinism takes at that time is as future-determination? (Another case in the literature involves an agent who has existed forever, as an adult and fully rational being, in an eternal world. In that case there is past-determination at every time but, since he has existed forever and thus has exerted his causal influence at every time, determinism doesn’t seem to be a threat to his free will either.)

Hope these comments help motivate the demotion claim a bit more!

[Jiajun Hu]

Carolina,

Thank you for your comments. They are helpful. Now I think I know what is the common ground for us. But I still have some doubts. I apologize if I am not seeing your ultimate point.

Let's begin with what I agree with you: mere future-determination does not threaten my present freedom. A case of mere determination could be your example above: You are making a free decision to murder a person or not, and such a decision will determine your future (whether you are in jail or not). And your future also determines your present decision. Such a case of mere determination obviously pose no threat to the freedom of your present decision. So, to threaten freedom, there should be a missing condition.

However, I tend to disagree with your suggestion that causal reach is the missing condition which distinguishes between past-determination and mere future-determination. I think the relevant difference between past-determination and mere future-determination is that past has been fixed and future hasn't (if your present decision is free). If the future has been fixed, and your decision right now is only a necessary step leading to the fixed future, it won't matter that the future is in your causal reach. Suppose you think you are making a free decision, but there is a magical force "forcing" you to make a particular decision (e.g., to murder this person, which could be a situation similar to the movie "Minority Report"). Moreover, to use the Rollback Argument here, if you see thousands of rollback and find out you make the same decision every time which will put your into jail eventually. I suppose, an incompatiblist would say that you are not free about the decision even if the fixed future is caused by you.

But by "magical force", I don't necessarily appeal to a real magical force. The point is: in a deterministic world, the future is fixed. And to pose a threat to my present freedom, I needn't to discuss how the future is fixed since we don't want to say that the future is fixed by the past. Noted, when we appeal to a fixed past, we usually don't discuss how it is fixed as well. Therefore, all facts incompatibilist need to appeal are: my present decision is determined by the future, and the future is a fixed future. If so, my freedom is under threat. It remains so even if the future is in my causal reach.

Another case you mentioned above could also be helpful here. Imagine the whole universe is just a play in which I play a role. And I play from the beginning through the end. However, the whole plot of the play is fixed. There is only one way to play it out. After I realize this, should I still believe I am free? Of course, this is a open question. But it shouldn't be a surprise that incompatibilists would say NO.

This is my concern about the demotion claim.

[Robert Allen]

Carolina,

Your post is a straw man. No philosopher contends that determinism by itself rules out free will. The skeptic is always committed to some other principle he fears could not be satisfied by agents living in a deterministic world. To take the most obvious example, holding PAP in conjunction with determinism motivates some thinkers to become Hard Determinists. I myself believe that determinism is also inconsistent with self-control: being the sole cause of one's choices. Were I to to lower my standards and treat it simply as a matter of choosing in accord with one's beliefs and desires, a la Hobart, determinism would no longer threaten my commitment to FW. Thus, it is my understanding of self-control along with that adhesion that requires me to eschew determinism.

[Carolina Sartorio]

Jiajun,

I’m not sure I follow everything you said, but at least some of your comments suggested to me that only incompatibilists who are also fatalists would want to argue in the way you argue. Imagine that we embrace a view of time according to which it is true now that certain events will take place tomorrow (in that sense the future is “fixed” now). Does that threaten our freedom? Only fatalists will think so; non-fatalists would say that the mere fact that it’s true now that certain events will take place tomorrow doesn’t threaten our freedom with respect to those events, precisely because we can still help to bring those events about, or make them happen.

[Carolina Sartorio]

Hi Robert,

I didn’t mean to suggest that, according to some incompatibilists, “only determinism rules out free will” in the sense that there is no explanation for why determinism rules out free will. Presumably, there’ll be some explanation of this, and that explanation will take the form of certain principles about the conditions for freedom. But those principles will be necessary metaphysical principles. The claim is only that, besides the assumption of determinism, other *contingent* assumptions (about us, or about our world) are needed for those necessary principles to “kick in” and thus for determinism to be a threat to our free will.

In other words, if incompatibilism were the claim that there is no possible world where determinism and free will coexist, then, assuming it is a contingent truth about us that we have a past beyond our causal reach, determinism and free will could still coexist in worlds where those contingent assumptions are not true (worlds in which agents don’t have a past beyond their causal reach). The thought is that this itself is not a reason to embrace compatibilism. Rather, it’s a reason to understand incompatibilism slightly differently, as the claim that there is no possible world where free will and determination *by events beyond the causal reach of agents* coexist.

[Jiajun Hu]

Carolina,

Thanks for the warning of fatalism. I think you are right about the danger. I should have been more careful when I use the term "fixed". It sounds like even if I make a free decision, after making it, the decision has been fixed. I surely don't want to mean that. What I should mean is something similar to necessity, although not necessity per se. So allow me to give it a final try, see if I can grasp what concerns me intuitively.

Your mentioning of fatalism reminds me of a case discussed by van Inwagen in his Essay. Suppose I get myself caught in a fire, and I try to escape. There are two exits in front of me. But unbeknownst to me, the nearer one is blocked and the remote one is good to go. In short, to choose the nearer exit is a necessary and sufficient condition for my death. Moreover, I have a strong rational character which always recommends a pure and simple rational choice to myself (I don't second guess an obvious rational decision). And to choose the nearer exit appears to be a rational choice to me at that moment. Further, let's assume it is a contingent fact that I am caught in such a circumstance and I have this rational character (and it is ok to assume that I partly contributed to them).

Trying to avoid begging the question, let's NOT stipulate already that I necessarily choose the nearer exit. Instead, a conditional necessity could be true: given the circumstance and my character, I necessarily end up dead. This conditional necessity, together with the deterministic relation between a wrong choice and my death, implies that it is conditionally necessary for me to choose the nearer exit given the circumstance and my character. So, if the future is "fixed" in the conditional necessity sense, it is good enough to threaten my present freedom.

Now, my point is: if determinism starts to obtain since my present decision through the future, no matter what my character is, the future is a conditional necessity. So my present choice is a conditional necessity. In such a case, we come back to the old debate: incompatibilist take this choice to be unfree, compatibilist disagree. (Actually, this sounds weird, why do I need to prove something obvious by deliberating the future first and tracing back? Why not just say my choice is necessitated by the present (thus the relevant) determinism? Anyway, the idea of future-determination sound weird from the beginning, let's just live with it.)

After considering this, I would like to take the force of the demotion claim to be this: the consequence argument cannot be applied to future-determination, because it needs the premise of "remote past is not up to me" (to use your term, it is beyond my causal reach). However, CA doesn't equate to incompatibilism. CA just makes it easier to prove incompatibilism (if PAP) with the background setting-up (no one lives forever and the universe has been deterministic for a long time).

So, on my view, CA helps build a strong case against compatibilism by lively revealing an obvious threat to free will from a certain deterministic setting-up. Many incompatibilists certainly feel happen about CA. But it doesn't mean that incompatibilists cannot live without CA. At worse, CA needs to be revised if compatibilists don't want to assume such a background setting any more (maybe they find it too friendly to incompatibilism). The merit of the demotion claim is to suggest such a possibility. But it is simply too quick to start revising incompatibilism.

[Carolina Sartorio]

Jiajun,

Indeed, incompatibilism (understood in the stronger sense, as the claim that determinism itself rules out free will) could, in principle, still be true even if the CA failed to establish it (but only established the “demoted” version of incompatibilism). But what I think is that it’s actually very natural to regard the demoted version of incompatibilism as the only one that’s really plausible, on reflection.

Could the inevitability of my acts (given the past and the laws) *by itself* threaten my freedom? On reflection, it’s hard to see how that could be. If I could be in control of those events that make my acts inevitable, then, arguably, the inevitability of my acts given those events would not itself be an obstacle to my freedom. Intuitively, it’s only when I *lack* control of what makes my acts inevitable that the inevitability of my acts starts to look like a problem. This just strikes me as a very natural thing to say.

[Jiajun Hu]

Carolina,

I think we are discussing two questions. One is how to understand CA. Another is how to understand incompatibilism.

For the first question, I think a common understanding is that CA essentially depends on PAP (according to Fischer$Ravizza). If so, any discussion which takes the force of CA seriously should take PAP seriously. If we have already taken PAP seriously, how could the inevitability of my acts by itself not threaten my freedom? Of course, I am open to the possibility of reinterpret CA. Maybe, as you suggest, the inevitability only becomes a threat when it is an out-of-reach or out-of-control inevitability. This could be an interesting idea (I can feel the pull), however, it would also be a controversial one for many of us (including some compatibilists I think).

As for the second question, I agree to a certain degree that it is natural to think inevitability doesn't contradict with freedom. Especially for philosophers who approaches freedom in a metaphysical way (including myself), freedom is a factual issue. If so, actual sequence view already seems a little bit plausible from the very beginning. But reflecting as an incompatibilist, I also have a strong intuition that PAP is true. The idea of living a life in which every detail is inevitable is horrible. It is horrible in a metaphysical way: are I still an agent who has control or just a small part of a big machine? It remains horrible even if I have lived forever. I think this is a natural thought as well, at least for incompatibilists. Acknowledging these different things would just bring us back to the old debate.

[Justin Capes]

Carolina,

I wonder if your "determinism by factors beyond our control" and "plain old determinism" don't, in the end, come to the same thing (at least for purposes of the free will debate). In a fully deterministic world, aren't all our actions going to ultimately be rendered inevitable by things beyond our control. Even in cases of instant agents and actions that occur at the first moment of time (if such a thing is really metaphysically possible), the agent is acting on the basis of mental states and laws that she finds herself with and over which she has no control. Off the top of my head, I can't think of how a fully deterministic world would go in which human actions aren't determined by factors beyond their causal reach.

[Carolina Sartorio]

Jiajun,

You write: “The idea of living a life in which every detail is inevitable is horrible… It remains horrible even if I have lived forever.” But recall that the type of inevitability in question is inevitability at a certain time *given the state of the world at other times.” And if I can (in principle) have some control over the state of the world at those other times by existing and exerting my causal influence at those times, then the inevitability in question seems much less horrible, or not at all horrible. My life develops inevitably in a certain way, but I get to contribute to what that way is.

On PAP: PAP says you’re responsible only if you could have done otherwise. But van Inwagen is clearly thinking that the incompatibilist can’t just assume that I couldn’t have done otherwise with respect to times that are not like the remote past (in that I exist at those times and can exert my causal influence), that’s why he picks the remote past to run his argument. So I think that even in van Inwagen we can find something that implicitly suggests that taking PAP seriously doesn’t involve committing ourselves to the idea that determinism itself threatens our freedom (or, in this case, the ability to do otherwise). Determinism by itself doesn’t rule out the ability to do otherwise; only determination *by factors with respect to which we couldn’t have done otherwise* does. At least, that's how I'm inclined to see this...

[Carolina Sartorio]

Hi Justin!

It’s an interesting question how exactly the two assumptions (of “plain old determinism,” as you called it, and determination by factors beyond our control) come apart, if at all. I’m inclined to say that the cases that are most likely to suggest that they can come apart in ways that matter are (1) the eternal agent scenario, and (2) a time-travel scenario:

(1) In the eternal agent scenario, at every time the agent finds himself in certain states, including mental states, that play a role in the choices he makes; however, each of those states is the result of earlier events that he could (at least in principle) have some control over, given that he existed at those times and could have exerted his causal influence then.

(2) In the time-travel scenario, we may imagine a very powerful time machine that could take you to any time and place and give you the power to causally influence a wide range of past events. This could give you (at least in principle) the relevant kind of control over those events to have the relevant kind of control over the causes of your current choices.

(The case of the agent who comes into existence at the very first moment of time is interestingly different, I think. In that case, the choice he makes at T1—the first moment of time—is also not determined by antecedent factors beyond his causal reach, because there are no antecedent factors because there is no antecedent moment of time. However, it’s also not the case that his choice *is* determined by antecedent factors *within* his causal reach. And it could be that this threatens his freedom too. But, arguably, then it’s not determinism itself that threatens his freedom, but the fact that he lacks the right kind of causal history.)

On the other hand, imagine that the scenarios described in (1) and (2) are not genuine metaphysical possibilities, or that for some reason they still can’t give you the relevant kind of control (not even in principle). Isn’t it plausible to say, even then, that the real threat to our free will is not determinism per se, but determination by events beyond our causal reach? After all, it seems like we’re still assuming that it’s only because the assumption of determinism *also* results in the assumption of determination by events beyond our causal reach that the problem arises.

[Jiajun Hu]

Carolina,

I think your claim that mere determinism itself doesn't threaten our freedom could be more acceptable in a loose sense. For example, if I was free and responsible for a bad character. Then I might be responsible for an act determined by the bad character (given some relevant conditions satisfied). In other words, I am free or have control about the determined act in a loose sense.

However, I wonder if this is the picture which interests you (many incompatibilists would welcome such a picture, such as van Inwagen, Kane). If not, how should we understand "if I can (in principle) have some control over the state of the world at those other times by existing and exerting my causal influence at those times"? If at those times, your contribution was also determined, what is the difference between your earlier causal contribution and your current causal contribution? It seems unnecessary to appeal to an earlier causal deterministic contribution since no one is denying your current causal contribution.

As for your understanding of CA and PAP, I think it is interesting, and open to discussion. All I can say right now is that I always take the particular form of CA to be a good argumental strategy (which is the reason why he argues like this). But the essential point is just that I couldn't have done otherwise if determinism obtains. The only qualification is what if I was free in the causal history of my current act, which makes me free now in a loose sense.

But this is something I haven't thought about carefully, and I could be wrong about all these. But honestly, I would be happy to see CA being challenged in a new way. I tend to believe that the particular form of CA, although pretty powerful (if PAP is correct), hides certain important issue related to moral responsibility. Let's say, it assume a too friendly background which makes it even harder to reach the truth about compatibilism/incompatibilism. It feels like trying to find the moral truth about capital punishment by merely deliberating on an extremely evil criminal who raped and murdered young kids in cold blood. It helps one side win the debate, but miss the whole truth.

(Of course, I am not claiming incompatiblists are winning the debate. On the contrary, if someone is winning, it is compatibilists thanks to Frankfurt.)