Is the will free or determined? This is a crude way of posing the traditional problem of free will and determinism. But the traditional problem presupposes that free will and determinism are incompatible. Since this cannot be legitimately presupposed, the fundamental problem is the compatibility problem: Are free will and determinism compatible or incompatible?
I view them as incompatible, and, influenced by Kant, I see compatibilism as a 'shabby evasion' of the underlying difficulty. But since one cannot shame a philosophical position out of existence, pace Daniel Dennett, I had better present an argument. An argument one finds in the literature is the Consequence Argument. (See for example Peter van Inwagen's An Essay on Free Will.) Here is a version of it that draws upon van Inwagen and also this discussion by Tomis Kapitan.
0. We first of all need to understand what determinism is. It is the doctrine that the past determines a unique present and future. That is, past states of the universe in conjunction with the laws of nature render only one present state of the universe physically or nomologically possible. The laws of nature might have been otherwise: they are not metaphysically (broadly logically) necessary. And the past might have been otherwise. But given the actual past, and the actual laws, the actual present state of the universe could not have been otherwise. Things have to be the way they are given that things were the way they were and the laws are what they are. As the word 'given' indicates, the necessity here is conditional rather than absolute. There is no logical or metaphysical necessity that I be blogging now -- please don't confuse determinism with fatalism -- but if determinism is true, then my blogging now is physically necessary. Therefore,
1. If determinism is true, then all our actions and thoughts are consequences of events and laws of nature in the remote past before we were born.
2. We have no control over circumstances that existed in the remote past before we were born, nor do we have any control over the laws of nature.
3. If A causes B, and we have no control over A, and A is sufficient for B, then we have no control over B.
Therefore
4. If determinism is true, then we have no control over our own actions and thoughts.
Therefore, assuming that responsibility requires control,
5. If determinism is true, then we are not responsible for anything we do or think.
Therefore, assuming that freedom entails responsibility,
6. If determinism is true, then we are not free, which is to say that every form of compatibilism is false.
If you don't accept this argument, which premise will you reject?
Bill,
Perhaps the word "control" needs further unpacking here, given that the causal chain that results in our future actions does indeed pass *through* us and our deliberations. My deliberation genuinely affects my future, even if the steps of that deliberation are determined (or perhaps random) at the microlevel (and how would I feel the difference?).
But determined or not, it is still true that if I *want* to do X, and *decide* to do X, I will do X. So would you agree that in some important sense, it is still I who am "controlling" my future, in that my dispositions, principles, experiences, etc. - i.e., all those things we would *want* to be the important factors influencing our decision - still critically affect the outcome?
Where we draw the boundary around 'I' is, I think, an important parameter here.
Posted by: Malcolm Pollack | Wednesday, May 06, 2009 at 04:08 PM
Malcolm,
The causal chain that eventuates in your drinking a beer does indeed pass through you with your desires, beliefs, abilities, and whatnot. But each of these causal factors is determined by earlier and earlier events over which you have no control. So yes there is a sense in which you control whether you drink a beer in that it is events internal to you that are the proximate causes of your beer drinking. But unless you control these events, then by (3) you have no control over the beer drinking if determinism is true. So (3) looks to be the crucial premise.
Dennett has a response to this which I will examine tomorrow.
Posted by: Bill Vallicella | Wednesday, May 06, 2009 at 04:40 PM
Hello Bill,
Longstanding readers may remember this argument from a post in September 2006. I recall it well because I submitted one of my first comments to you on it. I've been reading you daily, with great pleasure, ever since. Hearty congrats and many thanks for five years of thought-provoking postings. I've learned much.
My comment though is as before, namely, doesn't the argument here prove too much? Consider this parody argument obtained by the purely syntactic substitutions
we ---> thermostat
all our actions and thoughts ---> temperature
and fixing the verb number:
1. If determinism is true, then the temperature in my house is a consequence of events and laws of nature in the remote past before my central heating thermostat was made.
2. My central heating thermostat has no control over circumstances that existed in the remote past before it was made, nor does it have any control over the laws of nature.
3. If A causes B, and X has no control over A, and A is sufficient for B, then X has no control over B.
Therefore
4. If determinism is true, then my central heating thermostat has no control over the temperature in my house.
Posted by: David Brightly | Wednesday, May 06, 2009 at 05:15 PM
Right, David: this is just the sort of thing I meant when I suggested we unpack "control" a little more.
Posted by: Malcolm Pollack | Wednesday, May 06, 2009 at 07:17 PM
David,
Thanks for the kind words. I am glad to have you as a reader and critic. On 13 September 2006 I responded to your 'probative overkill' objection in a separate post. Let me review and possibly repost that post tomorrow.
Posted by: Bill Vallicella | Wednesday, May 06, 2009 at 07:26 PM
Bill,
I accept the first three premises but I do not think (4) follows from them. It seems to me that it assumes that ‘we’ must be apart from the causal chain exerting influence on our actions. Suppose I claim that I simply am my thoughts/beliefs/mental activities. To quote CS Peirce: “My words are the sum total of myself; for the man is the thought.” I do not control the distant past, I do not control the laws of nature and these things together (according to determinism) determine the future. But if control means “to exert a causal influence over” then I nonetheless do control (i.e. exert influence over) some of the future at the sequence in time when I enter the causal chain as my thoughts/beliefs/ mental activities.
Posted by: Spencer | Wednesday, May 06, 2009 at 09:47 PM
Spencer,
There is no problem with the logic of the argument. (4) indeed follows from (1)-(3). So if you reject the conclusion you must reject one or more of the premises.
Posted by: Bill Vallicella | Thursday, May 07, 2009 at 01:51 PM
Bill,
As I pointed out in your last post on free will, Fischer helpfully distinguishes between guidance and regulative control. He argues, successfully IMO, that responsibility requires guidance control (and other things not directly relevant to the topic at hand).
Anyway, your argument isn't stated as the traditional consequence argument is (i.e., "can do otherwise"), which I accept. So, regarding your argument, I reject (3) and (4) as ambiguous. Indeed, this ambiguity afects everything below (3). So I deny it all from (3) down except for:
* Therefore, assuming that responsibility requires control,
* Therefore, assuming that freedom entails responsibility,
until the post takes into account some distinctions compatibilists have drawn.
Posted by: Paul | Thursday, May 07, 2009 at 07:42 PM
"See for example Peter van Inwagen's An Essay on Free Will."
See also PvI's "Mind Argument" against libertarianism. That's why he's a mysterian.
Posted by: Paul | Thursday, May 07, 2009 at 07:48 PM
Where is PvI's 'Mind Argument'? And what exactly is the distinction between guidance control and regulative control?
Posted by: Bill Vallicella | Friday, May 08, 2009 at 06:51 PM
Hi Bill,
Mind Argument: PvI offers it in The Oxford Handbook of Free Will (see, Free Will Remains a Mystery 158-177, esp. 167-175). He also discusses it in An Essay on Free Will (1983, 142). However, a very rigorous defense of the Mind Argument, MA, that takes into account many of the contemporary rebuttals to PVI-type versions (e.g., Timothy O'Connor's specific kind of agent causation rebuttal, which, BTW, he thinks is the only successful response to MA-type arguments) can be found in Sean Choi's, The Libertarian Dilemma: A Study in the Logic and Metaphysics of Indeterminist Free Will (VDM Verlag, 2008; basically his UCSB '07 diss., and ultimately what I'd recommend over PvI)
Control: Fischer discusses the various kinds of "control" in his The Metaphysics of Free Will (Blackwell 1994); Responsibility and Control (Cambridge 1998); My Way: Essays on Moral Responsibility (Oxford 2006); and Four Views on Free Will (Blackwell 2007). The account is spelled out most fully in (1998).
Basically, guidance and regulative control are assumed to go together. Say you were driving to a chess tournament and you decide to stop at Fivebucks for some coffee before the tourney starts. Your car is functioning properly, your choice to go to the coffee house is based on your own reasons, etc. Your turning right into Fivebucks is to exhibit "guidance control" over the car's going right (this is distinctive because various other factors could have caused the car to go to the right (sneezing, indeterminist twitch in your arms, etc).
Now, if we assume that there's nothing else in play (e.g., no determinism, etc), then you had it in your power to turn the car left or continue to the tourney or stop or whatever. You exhibited guidance control but had regulative control, then. Regulative control over, say, the car's movements, stems from the ability to do otherwise (now, I'd agree with Choi here in denying that libertarianism affords the kind of control needed for freedom, but Fischer doesn't make those moves and allows libertarians to have regulative control).
So, libertarians would normally place both of these types of control together. Fischer thinks they can be prized apart. For example, suppose that the reasons are the same, you're properly functioning, etc., but the steering mechanism on your car breaks such that if you were to try to go straight, left, whatever, the car would go to the right anyway, along the same path it would have went, etc. However, everything else remains the same and you turn the wheel into Fivebucks coffee house. The defective steering wheel (or apparatus) plays no role in the "actual sequence of events" (Fischer and Ravizza's actual sequence account comes into play here). You exhibit guidance control but not regulative.
Of course, you could have tried to go right and thus exhibited some regulative control, but Fischer makes moves to account for these "flickers" of freedom (mostly by appeal to Frankfurt Counter Examples--which I am convinced of, same with some libertarians, BTW, e.g., Hunt, Craig, etc). Yet I trust the basics of the distinction have been laid out. It's obviously a much more detailed case (cf. the above literature). I think it is extremely successful (when you build in everything, e.g., actual sequence, reasons responsiveness, etc) especially in showing how an agent can be morally responsible for actions. And since I, like you, take it that if an agent is morally responsible for an action, then he did it freely (and vice versa). So, I take it that Fischer's account shows how we can be free even given determinism.
Posted by: Paul | Saturday, May 09, 2009 at 06:14 AM