Flickers of Freedom

The Basic Problem

On almost all formulations of the Consequence Argument (CA), the argument uses both a “grounding principle” and a “transfer principle”. Much attention has been given to the latter, taken as a logical principle. But I want to continue talking about the natural-language CA, which I take to be an argument about the metaphysical relationship between (uninterrupted) causal determination and free action, and not some distinct argument about the validity of certain rules of inference. Looking at these two principles in metaphysical terms, determination provides the “transfer mechanism” (or, better) “preservation mechanism” and a timeslice of the universe at some time in the remote past provides the “ground”. Recent arguments (see Joe Campbell) seem to suggest that neither the ground nor the preservation mechanism is sufficient to undermine free will on its own. In reaction, the emerging standard line on CA seems to be that the remote past (the ground) and the deterministic laws (preservation mechanism) work together to undermine our free will (e.g. Shabo, Bailey). On this line, if CA is sound, the deterministic laws are menacing because they do not allow us to escape from some freedom-undermining ground in our pre-personal past.

Notably, CA’s ground-plus-preservation structure is also visible in the diagnostic version of Pereboom’s Four-Case Argument. Pereboom claims that the upshot of his manipulation stories is that “an agent cannot be responsible for decisions that are produced by a deterministic process that traces back to causal factors beyond her control—decisions that are alien-deterministic events” (2001: 126). Here, “alien” points to the ground; “deterministic process” points to the preservation mechanism. Pereboom claims that the causal source of the manipulation victim’s action is “alien” and “beyond his control” because it is the result of progamming originally implanted by neuroscientists--and our own "pre-programming" is no different. Now, isn't this just another way of saying that the manipulation victim suffers from freedom-undermining “constitutive luck”? If so, it seems that Pereboom’s argument concludes (like CA) that uninterrupted causal determination is menacing because it does not allow a person to overcome the freedom-undermining features of his pre-personal past, i.e. his problem with constitutive luck.

------~-------

But the Basic Argument, too, rests upon a transfer principle...doesn’t it?

------~-------

There are multiple ways to summarize the Basic Argument. On some, it seems that we have inherited our current lack of freedom from our past: We are not free now because we didn’t bring ourselves into the world ex nihilo at our first moment of existence. On this reading, it seems that (just as in CA and Pereboom’s manipulation argument) the ground of non-freedom doesn’t get the whole metaphysical job done. It seems that there must also be some diachronic preservation mechanism that prevents us from overcoming our problematic origin over time. 

But let’s look at the metaphysical picture painted by the Basic Argument. The upshot of the argument appears to be that there is a necessary “starting-point condition” for free action:

 “In order for one to be truly or ultimately responsible for how one is in such a way that one can be truly responsible for what one does, something impossible has to be true: there has to be, and cannot be, a starting point in the series of acts of bringing it about that one has a certain nature; a starting point that constitutes an act of ultimate self-origination.” (Galen Strawson 1998)

Necessarily, the starting-point condition cannot be satisfied, as it can only be satisfied by someone who self-creates ex nihilo. From the first moment I existed, I have been unable to self-create from nothing, since I’ve been something at every moment of time since then. Prior to my creation, there was nothing of me. But, as there was no self at any point prior to my creation, there was no self to do any creating. In sum, there’s no point in time in one's past, present, or future at which she could satisfy the starting-point condition, and the nature of the causal links in between one's states has nothing at all to do with this. We might say that the ground of non-freedom picked out by the Basic Argument is so sweeping that it does all the preserving of non-freedom that needs doing; an independent preservation mechanism is not required.

As has come up in earlier threads, the counterexample strategy usually used by incompossibilists to undercut the Basic Argument depends upon the intuition that indetermination "helps" certain agents act freely. The idea, it seems, is that indetermination introduces a freedom-relevant starting-point or ground of free action. But if this intuition is irrational (and there are arguments other than the Basic Argument to suggest that it is!), then the counterexample strategy fails. If the counterexample strategy fails, all substantive differences between the Consequence Argument, the Manipulation Argument, and the Basic Argument disapear. None is an argument for incompatibilism, but each pinpoints the same basic problem: the problem of constitutive luck. The purported "problem of determination" was born from a proposed solution to the problem of constitutive luck, and the problem disappears if the proposed solution fails.

As I see things, Determination has served philosophers as a neutral guide. She patiently holds our hands and leads us back to our beginnings and reveals to us the basic problem that we face: our shockinginly difficult problem with constitutive luck. We don’t like what Determination helps us to see, of course--but doesn't it seem that incompatibilists are just trying to shoot the messenger?  

The Consequence Argument at its Best

I hope all the fathers out there had a nice Father’s Day weekend!  In this post, I would like to extend some of the reasoning from earlier posts and suggest that the Consequence Argument is like the Zygote Argument: in its extant forms, it is just a defense of a qualified incompossibilism, and only by adding a best-explanation argument would it pinpoint deterministic laws as a threat to free will:

THE CONSEQUENCE ARGUMENT

If determinism is true, then our acts are the consequences of the laws of nature and events in the remote past. But it is not up to us what went on before we were born, and neither is it up to us what the laws of nature are. Therefore, the consequences of these things (including our present acts) are not up to us.   [van Inwagen, p. 16]

Just over 30 years ago, Peter van Inwagen provided three formal statements of this argument. He referred to them as “three arguments for incompatibilism” or “one argument for incompatibilism done three ways”. The Consequence Argument has been popularly known as an argument for incompatibilism ever since.

van Inwagen, who seems to have coined the term ‘incompatibilism' in the early 1970s, uses the term to denote the view that the following two theses (or very similar variants) cannot both be true:

• PvI’s Thesis of Determinism (P-DET): There is at any instant exactly one physically possible future.
• PvI’s Qualified Free-will Thesis (P-FWT): We have free will.

In modal terms, van Inwagen’s ‘incompatibilism’ denotes ~◊(P-DET & P-FWT).  Technically, then, the “incompatibilist” conclusion of the Consequence Argument may be true even if deterministic laws pose no threat whatsoever to anyone’s free will.

But how could it be that the Consequence Argument is sound and yet deterministic laws make no freedom-relevant difference? After all, the Consequence Arguments identifies the freedom-undermining role played by the laws, namely: deterministic laws undermine free will by perfectly transferring or preserving non-freedom and non-responsibility from a point in the past up to the present. So the modest, non-explanatory conclusion must be an artefract of some technical glitch in the formal summaries of the argument, right?

It certainly seems so, at least when we assume that free will is possible—as van Inwagen seems to do in the first lines of his 1983 chapter on the Consequence Argument. If free will is possible but exists only in universes with indeterministic laws, then it seems that the best explanation for the lack of free will in universe with deterministic laws is that deterministic laws do something to undermine free will. And, the Consequence Argument appears to tell us just what this “something” is. 

For the sake of argument, though, let us reject the working assumption that possibilism is true. Yet, let us grant that deterministic laws transfer non-freedom and non-responsibility in just the way suggested by the Consequence Argument. Let us also grant that there is a valid “transfer principle” (e.g., some variant of Rule Beta) and that some version of the Consequence Argument using this principle is sound. Now, against this set of background assumptions, let us ask:

Does the Consequence Argument pinpoint deterministic laws as a threat to free will—at least for beings who, like us, cannot perform miracles, are not “agent causes”, are not causa sui, and have a remote past?

It seems not: Source impossibilists argue that self-creation ex nihilo would be the only way to overcome freedom-undermining “constitutive luck”. However, the natural laws of a universe seem to have no bearing on whether someone can self-create ex nihilo. A nothing-self is a contradiction in terms: for x to be a self is for x to be something, and if x is something then x is not nothing. Natural laws (whether deterministic or indeterministic) play no role in making it the case that a nothing-self is a contradiction in terms, or that there can be no x such that x is simultaneously both something and nothing. In short, it is impossible for any x to answer to the concept CAUSA SUI, and this has nothing at all to do with natural laws.

Still, it’s true that deterministic laws preserve our non-free status insofar as they in no way help us to “escape” from the freedom-undermining features of our past. But indeterministic laws cannot facilitate a successful escape either—not even indeterministic laws can help a person to be/become a nothing-self, i.e. a self that exists even though it does not. Surely we can't count everything that fails to help us to be/become a causa sui as a positive threat to free will--following  that practice, it seems that we would have to include the kitchen sink as a threat to free will, too. 

Might it be, then, that there is no technical glitch? Might it be that the lesson of the most famous “argument for incompatibilism”  is that we all have a freedom-undermining feature in our past, but deterministic laws pose no threat whatsoever to anyone’s free will? 

The Dull Edge of Manipulation Arguments

In my first post on manipulation arguments, I argued that the Zygote Argument may be an argument for incompossibilism, but it’s not an argument for incompatibilism. From that conclusion, a new question arises: Could any manipulation argument provide a free-standing defense of incompatibilism?

First off, a manipulation argument for incompatibilism would need a logical form something like:

The Diagnostic Manipulation Argument Template

D1. Victim Premise: The (apparent) manipulation victim S is not free or responsible for performing an action A. 

D2. Diagnostic Premise: The menacing feature of the (apparent) manipulation scenario--the feature that accounts for S’s lack of freedom and responsibility—is F.

D3. Same-feature Premise: F is found in the (apparent) manipulation scenario and also in any normal determination scenario.

D4. Conclusion: No one (like us) is free or responsible in the determination scenario (or any scenario in which F obtains) owing to F.

While the Zygote Argument offers no support for an instance of the Diagnostic Premise, such support is provided in Derk Pereboom’s Four-case Argument. Specifically, Pereboom claims to have identified the best-explanation of the freedom-undermining feature in his manipulation cases:  F = causal determination by factors beyond one’s control (2001: 114). Plugging the details of Pereboom’s manipulation argument into the above template, we get incompatibilism as the conclusion. We have just shown that the Four-case Argument is an argument for incompatibilism—right?

Not so fast. Pereboom’s proposed explanation points to causal determination and to factors beyond our control. The thought, crudely, is that we didn’t have control over anything before we came into existence, and causal determination hurts us (after we come into existence) by keeping it that way. The incompatibilist view that deterministic laws undermine free will by preventing us from becoming the “right sort” of sources of our actions is now aptly called“source incompatibilism”. It seems widely accepted that the Four-case Argument should persuade us of the truth of source incompatibilism, if we find it persuasive at all (e.g., 1, 2).

But now consider a nearby view, what I’ll call “source impossibilism”. The source impossibilist holds that necessarily, no one acts freely just because no one ever has control over the source of their actions. Notably, there are many controversial claims that (taken literally) both source incompatibilists and source impossibilists accept as true, for example:

1. Necessarily, if determinism is true, then the thesis that someone performs a free action is false; incompossibilism is true.
2. An actor does not freely perform actions that are produced by a deterministic process that traces back to causal factors beyond her control, a.k.a. “alien-deterministic” events.

But source impossibilists disagree with source incompatibilists about the metaphysical relationship between deterministic laws and free will. If source impossibilism is true, then source incompatibilism is false (and vice versa). Pereboom's best-explanation argument does not give us reason to favor one of these views over the other.

Could a free-standing manipulation argument settle who is right?  

My thought is “no”, and I’m wondering what others think. My reasoning goes roughly as follows:  The incompatibilist might argue that when the deterministic laws are removed and replaced by indeterministic laws, then freedom can be restored. Since indeterministic laws “help”, it must be that deterministic laws “hurt”. But, of course, that rests the case for incompatibilsm on a highly controversial view that the impossibilst rejects--and one that deterministic manipulation stories themseles provide no reason to accept. Coming from the other side, the impossibilist might try to show that deterministic laws are irrelevant by pointing to a (counterpossible) deterministic scenario in which someone self-creates ex nihilo and, nonetheless, acts freely—after all, even many libertarians accept that causal determination after a suitable “self-forming” action is not a threat to free will. It seems like we’re heading for a dialectical stalemate if we try to settle the dispute between incompatibilists and impossibilists with a best-explanation argument—and I see no reason to think this problem is going to be limited to the Four-case Argument.

But beyond a best-explanation argument, what could be used to defend an incompatibilist instance of the Diagnostic Premise? After all, a manipulation argument is just a façade of another argument if it requires another argument for incompatibilism to get its incompatibilist conclusion--and we should just scrap manipulation arguments and go look at that other argument. So, say we grant that manipulation arguments are suitably designed to persuade philosophers of the truth of incompossibilism and of the need to reconsider our conception of freedom-relevant sourcehood (as I’ve been persuaded). It still seems like they can’t, on their own, deliver incompatibilism.

To carve things up properly at this fine-grained level of explanation, are manipulation arguments too blunt a tool?

            

Manipulation Arguments and the Phantom Menace

Here is a formal summary of what is, I think, the best “global” manipulation argument around, Alfred Mele’s Zygote Argument:

Z1. Ernie is not morally responsible for anything he does.

Z2. Concerning moral responsibility of the beings into whom the zygotes develop, there is no significant difference between the way Ernie’s zygote comes to exist and the way any normal human zygote comes to exist in a deterministic universe.

Z3. So in no possible deterministic world in which a human being develops from a normal human zygote is that human being morally responsible for anything he or she does. (Mele 2013) 

Clearly, the Zygote Argument is not a defense of incompatibilism, understood as the view that if deterministic laws obtain, then someone subject to those laws lacks free will because those laws obtain. Z3 mentions natural laws and zygotic development, but does not pinpoint these things as threats to free will—which is good, because this argument would be invalid if it had a positive “diagnostic” conclusion. The purpose of these phrases is actually to make the conclusion more modest, to restrict Z3’s negative claim to a qualified incompossibilism: there is no possible world at which beings relevantly like us (i.e. developed from a zygote and subject to the laws of nature) live in a deterministic universe and act freely/responsibly. 

This qualified conclusion leaves open the question of whether someone might be a free and responsible agent so long as they were not subject to deterministic laws and/or did not develop from a zygote. As such, the above Zygote Argument does a good job of maximizing unity among incompossibilists in their effort to refute “compossibilism”.

But incompossibilists are a motley crew. Incompossibilists may believe that free will is possible or impossible. They may believe that natural laws are irrelevant to free will, that some sorts of laws would undermine free will and others would not, or that each type of law is freedom-undermining its own way. In other words, incompossibilists must agree to certain claims regarding what exists and where, but they may wildly disagree about why.

I care about the why. 

As I see things, the qualifications that appear in Z3 may be distorting the dialectic because they concede that there is good reason to worry that deterministic laws pose some threat to free will (and, remember, the Zygote Argument itself doesn't give us reason to believe think this.) For proponents of the Basic Argument, these qualifications are a distraction (zygotes and natural laws are irrelevant to the truth of incompossibilism) that obscure the real lesson of the Zygote Argument: that failure to self-create alone precludes free will. With deterministic laws in the spotlight, it can be hard to keep in mind that natural laws may be totally irrelevant to Ernie's lack of freedom and responsibility. 

Since I am devoting my posts to the search for an argument which, if sound, would show that incompatibilism is superior to its incompossibilist rivals, I ask:

1. Should the Zygote Argument be understood as an independent defense of incompatibilism--its formal summary just needs to be retooled a bit?  That is, can Ernie's story be turned into a defense of incompatibilism without appealing to some other (logically independent) argument for incompatibilism along the way?

2. What should we make of the Zygote Argument if (even retooled) it cannot deliver a specific, contrastive diagnosis of the freedom- and responsibility-undermining feature of Ernie’s story? Should compossibilists be expected to forsake compossibilism because of a phantom menace?

The Terms of the Debate

To begin, I would like to thank Thomas for offering me a spot in this Featured Author series. His lineup has been amazing so far, which makes it a *bit* daunting to step up to the plate--but here goes!

In order to set the stage for my upcoming posts, I want to generate some concern about the current use of the terms ‘compatibilism’ and ‘incompatibilism.’  I’ll forgo the history of the terminology here (but if you’re curious, just ask), because it doesn’t much matter how the terms were defined a half-century ago. Whether due to semantic shift, confusion, or what have you, the term ‘incompatibilism’ is now used to refer to two distinct theses, roughly:

1. Deterministic laws and free will are incompossible, i.e. these phenomena cannot possibly co-exist (e.g., van Inwagen 1983, and more explicitly 2004).
2. Necessarily, if deterministic laws obtain, then no one (subject to the laws of nature) has free will because deterministic laws obtain, from which it follows that deterministic laws and free will are incompossible (e.g., McKenna 2010; Vihvelin 2008).

Because the distinction between these two theses will come up frequently in my posts, I’m going to name them. I will use ‘incompossibilism’ to refer to Thesis 1 (because the term is so apt). I will reserve ‘incompatibilism’ for Thesis 2 (because the term is already connotes this view, and I cannot think of a better term). Notably, incompossibilism may be true even if incompatibilism is false, but not vice versa.

Here are the two related questions that I would like to explore this month:

Is there even *one* argument in the extant free-will literature that has incompatibilism (Thesis 2) as its conclusion? If the answer is “no”—as I will suggest—what should we make of that?

In an effort to answer the first question, I will discuss these famous arguments for “incompatibilism” in turn:

1. The Manipulation Argument (The Zygote Argument, The Four-Case Argument)
2. The Consequence Argument
3. The Basic (Hard-Luck) Argument

In my final post, I will offer my thoughts on how to answer to the second question.

                                                                                                     And, here we go……

The Featured Author of the Month: Kristin Mickelson

For starters, I would like to thank Randy Clarke for doing a great job of stimulating discussion here on Flickers of Freedom last month as the Featured Author.  Since I launched the series on the blog more than a year and a half ago, I have repeatedly been impressed with how much interesting work is being done at the crossroads of action theory and ethics. With an allstar cast of guest bloggers including Fischer, Sartorio, Mele, Pereboom, Levy, Campbell, McKenna, Nelkin, Nahmias, Sommers, Vargas, Smilansky, Waller, and others, the series thus far has been a pleasure to organize and the posts and threads have been fun to follow. Clarke's stint as a Featured Author was no exception. That said, I have nevertheless wanted the series to be more inclusive when it comes to the younger generation of folks working on free will. So, while I have some invitations to send out to some additional senior philosophers down the road, for the next few months I have lined up some of the new guard of our exciting field of philosophy.

First up is this month's Featured Author: Kristin Mickelson.  Mickelson earned her PhD from the University of Colorado-Boulder in 2012, and is currently a Visiting Assistant Professor at the University of Minnesota at Morris. Her research interests are primarily in metaphysics (free will, natural laws, philosophy of mind, personal identity) and philosophy of education. She published an essay which brings philosophy of mind literature to bear on Derk Pereboom’s famous Four-case Argument, and she will have a chapter on manipulation arguments in the Routledge Companion to Free Will (forthcoming). Mickelson has several draft manuscripts (available upon request) on metaphilosophical issues related to the free-will debate, including the relative merits of persuasive and non-persuasive manipulation arguments, and the proper management of technical terminology.

Mickelson also recently starting working on a book about autonomy. The book conceives of autonomy in terms of self-knowledge, self-control, and the pursuit of excellence. While primarily a work of metaphysics, the theory will also draw support from recent empirical work on phenomena such as willpower, boredom, and psychological reactance. A central goal of this project will be to respond to the “hard luck” problem (as defended by G. Strawson and Neil Levy), which she takes to be the most significant challenge to the metaphysical possibility of free will.

Because it's important for the next generation of free will scholars to have their voices heard when it comes to agency and responsibility, I hope the readers of this blog will be especially active in the comment threads during the next few months. So, please check out the posts and join the ensuing discussions. Given that Mickelson and the next few Featured Authors are at the early stages of their careers, helpful feedback is especially important. So, help spread the word and take the time to participate.

Here is the upcoming line up:

July:                    Joshua Shephard

August:               Marcus Arvan

September:         Matt King

October:             Justin Coates

Hopefully, everyone joins the fray and enjoys the show!