John Martin Fischer on Free Will and Moral Responsibility: Where the Action Is

This is the first in a series of postings I've commissioned by leading philosophers on "where the action is" in various subfields of the discipline.  (I wrote about Nietzsche studies a couple of weeks ago.)  Hopefully these short essays will be illuminating for both graduate students interested in the particular subfield and philosophers working in other areas who want to keep abreast of developments outside their specialty.

Today, John Martin Fischer at the University of California at Riverside writes about "where the action is" on free will and moral responsibility.

=====================

The topics of free will and moral responsibility continue to be the subjects of lively and active debates in contemporary analytic philosophy.  Not surprisingly, no consensus about results has been reached (cf. the fallen angels in Milton's Paradise Lost), and, like much contemporary philosophy more broadly, there is a fairly wide range of methodologies.  Additionally, some philosophers approach these topics with an emphasis on ethical issues (such as moral responsibility, blameworthiness/praiseworthiness, judgments such as "ought/ought not", "right and wrong," and so forth), whereas others approach the topics from the perspective of metaphysics and/or philosophy of mind. 

There was a tremendous amount of work on the relationship between God's omnisicience and human freedom (and responsibility) about twenty-five years or so ago,  but this discussion has been relatively less lively recently (with certain exceptions).  The doctrine of Molinism has been a hot topic, especially in certain parts of Indiana.  According to Molinism, God can know prior to any providential activity on his part a set of "counterfactuals of freedom"--subjunctive conditionals about what human beings would freely do in various circumstances.  He combines this "middle knowledge" with his "natural knowledge" of the various possible initial situations to derive truths about what human beings would freely do, and he uses this information to guide his creation and providence.   There has been much debate about whether there can be true "counterfactuals of freedom," with Robert Adams leading the charge against this possibility.  The problem is that such conditionals would seem to need "metaphysical grounding," but if causal indeterminism is true, all that would seem to be grounded are "might-conditionals".  (Molina and his followers believe that causal determinism is incompatible with the relevant sort of freedom.)  Alfred Freddoso and Thomas Flint are contemporary exegetes and defenders of Molina.  Whereas there is much of interest in these debates, I believe they are orthogonal to the problem about the relationship between God's foreknowledge and human freedom.  In my view, Molinism provides a picture of how God could know about future actions of humans, and how he could use this knowledge in his providential activity.  But it does not provide an answer to the problem about the relationship between God's foreknowledge and human free action; rather, it simply presupposes some answer to this problem.  Indeed, Molina's favored solution to the latter problem was that of Aquinas, according to which God is outside of time (and thus has omniscience but not foreknowledge), and Flint's solution is Ockhamism.  Some philosophers (for example, Robert Kane in his otherwise excellent introductory book on free will in Oxford University Press's Fundamentals of Philosophy series) seem to think that Molinism is an answer to the problem about omniscience (or foreknowledge) and freedom, but this is a mistake. 

I think it might be interesting to explore whether any contemporary work on the nature of belief (in philosophy or perhaps the cognitive sciences, but probably not neuroscience!) might be employed to seek to make some progress on the traditional problem about the relationship between God's omniscience (foreknowledge) and human freedom.  This is not where the action is, but it is one place where I'd like to see some (more) action.  (For discussion, see the Prosblogion blog.)

There is some intriguing work on the nature of causal determinism.  J. Howard Sobel's book, Puzzles of the Will, has a subtle discussion of the various forms of determinism and how they might be thought to threaten human freedom.  He (in my humble and contentious opinion!) successfully shows that the Consequence Argument (or Basic Argument for Incompatibilism) can be formulated rigorously in various ways, not all of which require a modal principle such as the Principle of Transfer of Powerlessness (the principle that acts like a slingshot that can transmit powerlessness from one thing [the past] to another [the present] via the natural laws).  Some philosophers, notably Peter Van Inwagen and Ted Warfield, maintain that all valid versions of the Consequence Argument rely explicitly or implicitly on such a modal principle; I have been a persistent (they might say pesky!) critic of their view.  Carl Hoefer has a fascinating paper in which he points out that contemporary physics may not sit well with the basic intuitions that go into the Consequence Argument; in Hoefer's view, the intuitions that play a role in this argument (and buttress the Fixity of the Past and Natural Laws) come from common sense (and the A-Series picture of time), whereas physics presupposes the B-Series (without a moving now, as it were, and thus without past, present, and future). 

Recently philosophers have distinguished at least two importantly different ways in which causal determinism might threaten moral responsibility.  One way stems from the Consequence Argument.  If causal determinism rules out the sort of freedom that involves alternative possibilities, and this freedom is required for moral responsibility, then causal determinism is incompatible with moral responsibility.  Some compatibilists deny that the Consequence Argument successfully rules out the sort of freedom that involves genuine alternative possibilities.  Others contend that moral responsibility does not require this sort of freedom.  There are at least two different versions of this latter strategy.  Harry Frankfurt and I (and others) have contended that "Frankfurt-style" examples (first invented by John Locke) show that moral responsibility does not require alternative possibilities.  The debate about the Frankfurt-style examples still rages; a good recent collection is edited by David Widerker and Michael McKenna: Moral Responsibility and Alternative Possibilities: Essays on the Importance of Alternative Possibilities (Ashgate).  David Widerker has been an important and influential critique of the strategy based on the Frankfurt-examples; there are rumors that he has a new paper in the works on Frankfurt-examples that might surprise many.  (Is it too much to hope that Widerker will defect to the side of the good guys?) Another strategy builds on the seminal work by Peter Strawson; there are important recent developments of Strawsonian approaches in Paul Russell's Freedom and Moral Sentiment and R. Jay Wallace's Responsibility and the Moral Sentiments. 

Another way in which causal determinism might threaten moral responsibility is directly and not via threatening to expunge the sort of freedom that involves alternative possibilities.  A whole range of different options are available here.  "Source incompatibilism" is the term typically used to refer to this family of worries; this view has it that if causal determinism is true, then an individual would not be the "source" of his choices and behavior in the way required by moral responsibility.  Important developments of this view are found in Robert Kane's The Significance of Choice and Derk Pereboom's Living Without Free Will.  Although there is still much interest in the Frankfurt-style examples and Strawsonian approaches, the "action" seems to have shifted somewhat in the direction of Source Incompatibilism recently.  (I cannot help noting--please excuse this--that in my 1982 Journal of Philosophy article, "Responsibility and Control," I identified Source Incompatibilism as a significant view with which compatibilists must come to terms; but of course in subsequent years I have sought to do just this, whereas such philosophers as Kane and Pereboom have developed and defended the position with considerable detail and skill.)

Timothy O'Connor's book, Persons and Causes, seeks to develop an agent-causal view.  It is the most sophisticated and comprehensive development of such a view of which I know.  In an interesting and important recent book, Libertarian Accounts of Free Action, Randolph Clarke surprisingly backs off from his former defense of agent-causal libertarianism; there is still hope that he will come all the way over to the compatibilist's side, although he remains unconvinced.

A major problem for any version of libertarianism--event-causal (Van Inwagen, Kane, Ekstrom) or agent-causal (O'Connor) is the problem of arbitrariness or luck.  Much recent work has addressed this set of problems, and Alfred Mele has written a book on this subject forthcoming in early 2006 with Oxford University Press.  I hope he solves the problem, because I believe that moral responsibility should not "hang on a thread"--should not depend on whether the conditionals that link past to present in physics have associated with them 100% probabilities or merely 99.9% probabilities!  Since I believe our most basic views about ourselves as free and responsible should not hang on a thread (at least a thread of this sort), I believe that these views should be resilient with respect to a discovery that causal determinism were true; also, they should be equally resilient with respect to a discovery that causal indeterminism (of certain sorts) were true.  So good luck, Al--no pun intended!

Various important challenges have been raised to our views of ourselves as robustly free and responsible, including Galen Strawson's Freedom and Belief, Saul Smilansky's Freedom and Illusion, and Derk Pereboom's Living Without Free Will.  (One can imagine a whole series of Living Without X books: Living Without Ethics, Living Without Knowledge, Living Without Aesthetics, and so forth…)

There is much intriguing and suggestive work employing somewhat nontraditional methodologies.  Neurophilosophers such as the Churchlands have written recently about freedom and responsibility.  Also, there is a booming industry in "experimental philosophy" which seeks (among other things) to clarify and establish what people actually believe about the range of issues pertaining to freedom and responsibility.  I think that progress in the future on free will and moral responsibility will most likely be made through a broad-minded and eclectic approach whereby one takes insights from both traditional methodologies and also the developing informational sciences, psychology, and neuroscience.  I think it is misguided and not fruitful to brush aside centuries of thought on these ideas as my friend Daniel Dennett does in his intelligent, lively but in my view disappointing book, Freedom Evolves.  Similarly, we need to be open to the insights potentially available through the work of neuroscientists, psychologists, and philosophers employing these intellectual tools.

I shall end this (too long but also obviously too short) post with my puzzlement over a "splinter" in the work on free will/moral responsibility.  Some theories are called "normative" accounts (Watson, Wolf, Wallace, Benson, et. al.), whereas others are called "metaphysical" (Van Inwagen, Ginet, Fischer and Ravizza, O'Connor, Clarke, et. al.)  It is striking to me that the "normative" group is highly miscellaneous; the specific ways in which they are "normative" are very different, and it would be useful to get a clearer understanding of the distinction.  I doubt that it is as sharp or important as some suppose.  The difference may be as much a matter of "conversational networks" (as Manuel Vargas has pointed out to me) than content.

I should note that much of interest had to be omitted from this discussion--I'm sorry.  I welcome you all to the discussions at "The Garden of Forking Paths."

(Please note that comments may take awhile to appear.)

Comments

[C.J.Colucci]

This is probably something the professionals have addressed, but my 4 year old grand-niece knows what I will do in certain situations (I'll spare you the charming story) and my wife has even more knowledge of what I will do. How does their knowing what I will do affect whether I was free to do something else? I can understand the argument that the kind of causality presupposed by such knowledge is inconsistent with freedom, but what does my grand-niece's or wife's accurate knowledge add to the mix?

[John Fischer]

A quick reply to C.J. Colucci: thanks for the question. God is supposed to be essentially omniscient (on certain tradtional views), and this is what really generates the problem. If He is essentially omnsicient and in time, then I can't do otherwise (it would seem) without giving him a false belief (something incoherent). On the other hand, I assume that your wife and daughter are not essentially omniscient; so even if they were to genuinely know what you are going to do (in advance), you could still so act that they would have been wrong. No human being is essentially omniscient, so even if we assume humans can have foreknowledge of future contingents, this does not threaten the freedom to choose or do otherwise (at least not in the same way as the foreknowledged of an essentially omnscient being).

[Michael Drake]

What would you say are the main objections to, say, Pereboom's project of developing an ethics independent of the twin notions of free will and moral responsibility. (It may help you in outlining an answer to know I am a free will skeptic and continue to be puzzled by the resistance of contemporary naturalist compatibilists to the skeptical conclusion--a conclusion that to me seems quite natural.)

[Robert Allen]

Prof. Fischer,

What do you think is to be gained philosophically by "establish(ing) what people actually believe about the range of issues pertaining to freedom and responsibility"? How can the work of the experimental philosophers help us determine what free will is and whether or not it exists?

Also, I'm going to look into Prof. Kane's position, but right now I've gotta believe that you are right about Molinism being orthogonal to the problem of freedom and divine foreknowledge (to the extent that a belief in divine foreknowledge should raise questions about freedom; cf. my previous post). Indeed, it seems to beg the question of how could we be freely doing things if God knew in advance that we would be doing them.

[Sean Aas]

Professor Fischer:

Very interesting post. A brief quibble with your response to C.J. Colucci: if his wife and daughter genuinely know that he is going to A rather than not-A, then isn't it thereby *true* that he is going to A? Does it matter whether the knower is essentially omniscient, given the factivity of 'knows'? Does this create a problem for freedom (and, if so, what kind?)?

[Eddy Nahmias]

Like Robert, I'd like to hear if John has any thoughts about what an empirically informed understanding of ordinary people's beliefs about free will and moral responsibility could offer to the philosophical debate.

As someone engaged in this endeavor, I will offer just three possibilities (I know they won't convince Robert but perhaps others will find something to chew on):

1) If a philosophical theory begins with the view that the relevant concept of free will is one that, in large part, best describes the conditions an agent must have to be morally responsible, then assuming our actual practices and beliefs about moral responsibility provide some indication of those conditions, it would be helpful to understand those practices and beliefs. Empirical investigation, as well as armchair consideration, seems necessary to achieve such an understanding of ordinary people's practices, intuitions, pre-philosophical theories and concepts. Notice that if one think Peter Strawson is on the right track, this approach seems particularly appropriate.

2) Understanding pre-philosophical intuitions may shed some light on the way certain philosophical arguments work. Here's a controversial example: I think many incompatibilist arguments derive their intuitive appeal by conflating determinism with reductionism--specifically, by suggesting that determinism implies a sort of reductionism that makes conscious deliberations and decisions causally epiphenomenal. This claim obviously requires philosophical argument to back it up. But I think it derives *some* support from this result I've obtained from surveying the folk: when a group of people read a scenario describing complete causation of choices and action in terms of neurobiological states (which are completely caused by genes and environment), the vast majority say that this rules out free will and moral responsibility; when another group reads a very similarly worded scenario describing complete causation of choices and action in terms of psychological states (which are completely caused by genes and upbringing), the vast majority say that this does *not* rule out free will and moral responsibility. I interpret this result, as well as others in the literature, to suggest that determinism per se is not intuitively threatening to freedom and responsibility, but reductionism is (I suspect indeterministic reductionism would be perceived as nearly as threatening).

3) As long as philosophers in the free will (or any other) debate claim that their own theory, premises, or conclusions best accord with (or account for) ordinary intuitions, commonsense, or folk concepts or theories, then empirically informed investigation of folk intuitions, etc. seems entirely appropriate either to (a) support or undermine the philosophers' claims or ... (b) get them to stop making such claims! Many philosophers make such claims about ordinary intuitions, etc. because it seems to offer prima facie support for their views and shift the burden of proof to the supposedly counter-intuitive theory (incompatibilists have used this tactic).

[John Fischer]

"Q" the Enchanter: What kind of name is that? Anyway, I like Derk Pereboom's book a lot, and he goes a long way to convincing me that we could do with a leaner, but not meaner (!), notion of moral responsibility. But. I do still have the intuition, and it is hard to get rid of, that under certain circumstances certain individuals deserve, in a full and robust sense, harsh treatment and condemnation for what they have done. That is, I still have some sort of residual view that people deserve (sometimes) to be punished, or morally blamed (or rewarded or morally praised); I would suggest that this is not entirely straightforward to expunge. And I'm not inclined to seek to expunge it on the basis of the possibility that causal determinism obtains.

Robert: I will leave this for the practioners of experimental philosophy to fill in. I think they have various ways of proceeding, and that some are (to me) more promising than others. I don't think it is the best strategy to rule such strategies out from the beginning; let's see what develops and evaluate it then.

An anecdote about a different but related subject matter: I recently attended a fascinating "mini-conference" at Princeton's Center for Human Values on the Doctrine of Double Effect (organized by Walter Sinnott-Armstrong). It was a good chance to meet some psychologists who work on what they call "moral psychology"--and Gilbert Harman reminded me that some philosophers also would use that name for the same subject matter! (We trogodytes think of "moral psychology" as practical reasoning, deliberation, reasons for action, free will, and so forth.) Among the interesting results presented by such psychologists as Marc Hauser and Josh Greene (and the law professor John Mikhail), was the following. Fully 30% of the people surveyed said that it is morally permissible to push the "large individual" (originally, the "Fat Man", now perhaps the "Inidividual with a [large]) backpack) off the footbridge to stop the trolley from killing five innocent persons. This struck me as amazing, especially given Judith Thomson's statement in her original presentation of the example to the effect that EVERYONE she had asked about this case said that it was NOT permissible. So we went from zero to thirty percent by actually doing the survey.

One thing that comes out of this research for me is that demonstrably people's intuitions or responses change dramatically based on small changes in the presentation (the wording, the order, and so forth). These are identical to or related to what Kahneman and Tversky have called "framing effects". It seems to me that philosophers should be attuned to such effects.

[AspiringPhil]

I first have to grant that I’m not completely clear on the definition of “essentially omniscient” so I may be off target here. Putting that aside, if we place God in time, it seems to me that a good way to explain what it is for God to be all knowing is to say that He knows perfectly the laws of nature (however they are to be defined) and knows every fact of the world at (at least) one time. From here, God looks at the time when he knows every fact and then He applies the laws of nature. From this He is able to derive the facts at the next moment and the next and so on. If this were the case, I would contend that God is doing more than making predictions – he really does know what will happen next.

Now, I’ll bet that in some way this conflicts with a traditional view of God’s omniscience (and I’d be interested to know how and why), but I think it does cover omniscience in some sense or other. Furthermore, this will allow for Frankfurt style examples to be applied to our problem. God may know what will happen one moment from now based on the laws of nature and the previous moment, but this says nothing about how events came about. So, for example, let us suppose that I am sitting in a chair. God’s information may accurately tell him that I am about to stand up in the next moment (this amounts to my being determined to stand up), but I may still be the source of this action in some relevant sense. To expand, God’s knowledge that I am about to stand up does not necessarily presuppose that I was not in control of this action in some sense. It may be that when I am exposed to certain stimuli I am disposed to act in certain ways. However, a compatibilist can argue that there is a relevant sense in which I’m free and a relevant sense in which I can be held responsible for my action. As it happens, I think I would not have stood up freely in this case, but this certainly a debatable conclusion. To fill in the enormous blanks here I suppose one would have to take a look at some compatibilist arguments surrounding the Frankfurt-examples.

[Michael Drake]

"What kind of name is that?" Why, a made-up one. (If I were the type to use emoticons, I would have inserted one here.)

I share your retributive intuitions, and I'm not even sure there's any "can" to the question of whether to expunge them (I mean in my case, of course; some persons do seem able to vanquish their retributive instincts).

On the other hand, I harbor equally robust intuitions that (1) retribution is justifiable iff there is free will and (2) there is no free will.

I guess I'm just destined to live in antinomy, then...

[Manuel]

Hi John-

I'm inclined to think that you've identified the biggest areas of growth: source incompatibilism, skeptical views, and the rise of work that has a strong connection to empirical data. If I had to make bets about where things are going to go, I'd guess that we'll see more discussion of agent causation, increased attempts to respond to skeptical views, more empirical stuff, and (I suspect) more discussion of how the metaphysics stuff connects to "normative" issues and vice-versa. In particular, I think we'll see more attempts to sort out how various proposals for metaphysical requirements on free will do normative work (i.e., what does indeterminism buy you, normatively speaking; do reasons views have some normative considerations that favor them over identificationist views (and vice-versa)?). At any rate, these are some of the kinds of things that I tend to end up talking about with other young-ish philosophers working on free will, which seems like a reasonable index of what at least SOME new developments will look like in 5-10 years.

[Chris Franklin]

Hey John,

Great post! It was interesting reading Manuel's post. Being a "young-ish" philosopher, I guess my interests serve as empirical data for his points (if interests are empirical!). Particularly, I am interested in providing a safe-guard for moral responisbility againts certain skeptical views. By safe-guard I mean an argument(s) that is compatible with a wide array of views about free will and (in)determinism (e.g. compatibilist and libertarian). I would like to put some (more) force behind van Inwagen's claim that he would give up the claim that free weill, and hence moral responsibility, is incompatible with causal determinism should causal determinism turn out to be true. But I am not sure what the argument might be, or exactly how much of the "content" of moral responsibility such a gerenal argument could defend.

I also think you nailed it on the freedom and foreknowledge issue. Hopefully both sides will take your cue and make some head way in the debate.

[Tom Clark]

John,

You can’t find a good reason to support your “residual view” that we should treat people harshly if it serves no good consequence, but acknowledge that it’s hard to drop. It’s difficult, perhaps impossible, to expunge the retributive impulse since we’re hard wired to retaliate. But of course that doesn’t make it right, all things considered. Moral praise and blame survive, but when naturalized don’t any longer need to include imposing suffering without consequential benefit, e.g., see http://www.naturalism.org/morse.htm and http://www.naturalism.org/fiction.htm#sidebar . Absent some positive support for the retributive intuition, I think it’s somewhat intransigent to retain it in your view, even if you’re “not inclined to seek to expunge it on the basis of the possibility that causal determinism obtains.”

Q the Enchanter,

Yes, to some extent we’re condemned to live in antinomy between our instincts and our considered judgments. But intuitions – our first blush untutored pretheoretical notions about things – can change radically under pressure from rational considerations. So do you really any longer have the retributive intuition, given your free will skepticism? Or is it simply that you know that in certain situations you’d very much want to behave retributively, despite your best judgment that it’s unjustified, all things considered?