Evil and Libertarianism

I want to follow up on one of Eddy’s older posts.  On July 10, 2006, in a post titled “Free Will in the World Cup”, Eddy asked:

“For that matter, why is it that in the case of other foolish and seemingly irrational acts, like suicide terrorism, people are so much less interested in applying their theory of mind modules to figure out what drives the behavior and so much more likely to just say it's evil (is there a boundary past which we give up trying to explain or simply cannot explain certain actions)?”

    Those words really resonated with me.  Surely, I thought, Eddy was picking up on a real phenomenon: people were less able or willing to understand the behavior of wrongdoers and more willing to attribute their behavior to evil simpliciter.  Indeed, Eddy was making a point I tried to make on September 11th 2004, in a post titled “Who Was Morally Responsible For September 11?”: “The world's response to the events of September 11, 2001 suggests that our intuitions about freedom and responsibility are, in some ways, mistaken…”

    When I wrote my post on September 11th 2004, I had not yet read about cognitive biases.  But last summer I did research on many such biases that might be relevant to the free will problem.  Fortunately, I discovered research documenting exactly the phenomenon that Eddy and I were talking about.  I quote my findings below:

“Asymmetrical attributions of blame may reach their zenith in the process of demonization (Ellard et al. 2002). Demonization is an unwillingness to empathize with another such that the person regards this other as evil.  Despite correcting for any potential differences in personality, by having the same person write both a story as a perpetrator and a story as a victim, Baumeister, Stillwell, and Wotman found significant differences between the two kinds of stories: perpetrators recalled much shorter time spans, victims regarded the actions of perpetrators as more inexplicable, and victims regarded the perpetrator’s actions as more harmful (1990).  One might worry that these findings are only relevant to attributions of evil but Baumeister and Vohs are quick to suggest that demonization admits of degrees and so may be a common feature of our responsibility practices:

‘To be sure, our research sample consisted of many everyday conflicts and misdeeds, few of which were sufficiently important to qualify for the grandiose term evil.  Our assumption, however, is that similar processes operate in everyday transgressions as in large-scale misdeeds, and that if anything, the gap between victim and perpetrator would probably be even larger in horrendously evil events than in petty, everyday conflicts.’
(2005: 87)

    In the context of the free will debate, the most important feature of demonization may be the failure of victims to understand the actions and motivations of perpetrators.  Baumeister et al. reported that although “[b]oth victims and perpetrators distorted their stories—and to almost identical degrees,” “the weight of the evidence tends to be closer to the perpetrators’ accounts” (2005: 89-90).  This is so because “[p]eople rarely attack for no reason” even though the “perpetrator’s motives are often opaque to the victim” and “victims cannot or will not see this perspective” (2005: 88-89).”   

    This research fits well with Nichols & Knobe’s data suggesting that affect tends to make people more free willist.  But I don’t think it is quite right to suggest that these persons become more compatibilist because such actions do not “come out of nowhere.”  The observer of wrongdoing in a deterministic world can always explain, in principle, where the behavior came from: just trace the states of the universe back in time, according to the laws of physics.

    Instead, I take the data from Baumeister and his colleages to suggest that angry or affected persons simply lose their grip on determinism.  Bully soccer players, or suicide terrorists, tend to make us more libertarian, not more compatibilist.  Indeed, it is most interesting that, according to the Luck Objection, the actions of libertarian agents “come out of nowhere” and cannot be attributed to the agent, just as victims see transgressors as “attacking for no reason.”

    I’m optimistic about the prospects of cognitive science to help us solve the free will problem.  Baumeister et al.’s research helps explain third person attributions of evil and may be relevant to understanding third person attributions of freedom and responsibility.  It suggests that such attributions may vary in degree, according to context.  But further research would need to explore first person attributions of freedom and responsibility and the ways in which different contexts inspire different attributions.  Other biases (such as the illusion of control), and other contexts, may also be important.  Such research has not yet solved the free will problem.  But it suggests to me how we might go about trying to solve it.

Comments

[Cihan Baran]

Kip,

Thanks for breaking the looooooooooooong silence. I want to comment immediately, just because I miss commenting at GFP so sorely. So I am sorry if comments are just nonsense - not that they are all that good anyway.

"people were less able or willing to understand the behavior of wrongdoers and more willing to attribute their behavior to evil simpliciter... This research fits well with Nichols & Knobe’s data suggesting that affect tends to make people more free willist."

I think I have read that paper by Nichols & Knobe you are referring to. I think there might even be an evolutionary/biological basis of such behavior. That the retributive instinct is especially great at the face of great evil/tragedy might be explained by evolutionary/survival advantages such attitudes offer.

So, indeed, people love to blame.

"Such research has not yet solved the free will problem. But it suggests to me how we might go about trying to solve it."

What free will problem are you talking about? According to you, isn't it already solved - doesn't non-realism win the day?

And also re:Eddy's comment:

I don't want to change the subject (I apologize if I do so as in the Vihvelin discussion) but there is a topic that I am really curious about. I don't know how much we can allow our philosophical convictions about free will affect our daily, practical lives - especially for non-realists.

Is the hard incompatibilist entitled to any normative claims about revising our attitudes? Even if it is rational that I withhold from having any morally reactive attitude, the fact that I don't desire to and therefore won't just a determined fact, just like any other fact?

There is a better way to make my point but I'll let it slide. Perhaps, I should read Pereboom's "Living Without Free Will".

[Kip Werking]

Cihan,

Thanks. You're right; it has been a little quite around these parts and I'm happy to try to stimulate some discussion.

You wrote:

"I think I have read that paper by Nichols & Knobe you are referring to. I think there might even be an evolutionary/biological basis of such behavior. That the retributive instinct is especially great at the face of great evil/tragedy might be explained by evolutionary/survival advantages such attitudes offer.

So, indeed, people love to blame."

I agree entirely. Tamler Sommers and I are two people who have tried to explore, not just how the illusion(s) of free will work, but also how they evolved. It is interesting to contrast our approaches with Dennett's, who cites evolutionary forces to explain why free will does exist, not why it doesn't.

"What free will problem are you talking about? According to you, isn't it already solved - doesn't non-realism win the day?"

When I first got into this stuff, I would have agreed with you: I would say that free will doesn't exist, without exception or qualification. The more I've learned about the problem, the more I've realized that it is too subtle and complex for that kind of answer.

I'd still probably say that free will doesn't exist. But it might not exist in either of two ways: (i) the term is sufficiently well defined that it refers to something logically impossible or (ii) the term is so vague or poorly defined that it cannot refer to anything real (just as muhabutasille, a word I just made up, doesn't refer to anything real). You can call the former the G. Strawson approach and the latter the Double approach. I probably wouldn't say that the problem is solved until I know which of these approaches is right and can demonstrate its truth to others.

My primary skepticism towards the G. Strawson approach is that I think all philosophers---including non-realists like my (former) self---tend to be overconfident when they define terms. There is a very natural inclination to think:

"'X'? Oh, I've always thought 'X' means X. And, because I think 'X' means X, and I like to be right, I tend to think that everyone probably agrees with me. What? You think 'X' means Y? Well, you must be mistaken. You must be using some idiosyncratic definition of 'X'. You need to get with the program. I've used 'X' to refer to X my entire life, and nobody has ever corrected me."

And this presents a sort of bias that can be seductive and dangerous. Just because a definition of a term feels so right and natural to you doesn't mean it's right. Perhaps you grew up using the term wrong. Or perhaps the term is vague or poorly defined.

Of course, that person might still be right. Maybe 'X' does mean X. Maybe the people who say it means Y have been tricked by the fundamental attribution error, or the illusion of control, or wishful thinking, or demonization processes (as I describe in my post). But maybe they haven't.

So, I hesitate to say "free will doesn't exist" because I've lost some confidence that I know what what the term "free will" means, and I want to reserve judgment until I see data showing:

1. how "free will" is commonly used by ordinary people
2. whether there are individuals differences between different camps in the free will debate, such as differences in vulnerability to cognitive biases, which might explain why one groups so adamantly believes they are right despite being (according to the other camps) wrong

[Anibal]

The dualistic property of mind or as here is called cognitve biases, reason Vs emotional cognitve style, and how we use them in terms of the free will debate and realted fields such as the allocation of punishment (retibutive justice); is now more or less a well accepted fact thanks to advances of cognitve science in general (thanks to specefic people such as Kahnneman, Greene, Knobe, Hauser, Haidt, Nichols...)

But if our evolutionary trajectories shpaed the things in that way, why is not more economical in terms of survival (preventing fighting or conflict with conspecifics)to blame ourselves instead of others and never developing a reason based explanation to construct the concept of evil or others metaphisical instances that are pointless.

Why utilitarians, aristotelians, or kantians use more reason based explanations and humeans more emotive ones by default, when probably it is mixture of bothdepending on context.

We are not "cognitve monothematics" we never use all the time reason or all the time emotion to deal with moral problems we face daily
.
It is the free will debate sink or suspended because before to make any progress in this intellectual activity or field of theorizing we have to scan the brains of our leading philosophers to see their biases and not let them to spread theirs?

[Jim]

Jerry Fodor's review of Galen Strawson's essay in the recent London Review of Books states near the end that "...anyhow, Strawson is right that the hard problem is really very hard(explaining how subjective experiences arise from neural computation); and I share his intuition that it isn't going to be solved for free. Views that we cherish will be damaged in the process; the serious question is which ones and how badly."

Some sense of the set of damages that might be inflicted on the determinist camp can be inferred from the arguments of Henry Stapp. He says in the precis of his new book "The Mindful Universe: Quantum Mechanics and the Participating Observer," that "Of course, this "hard" problem is--and will remain--a mystery insofar as one's thinking is imprisoned within the fundamentally invalid conception of classical physics, which has no rational place for consciousness. Within that framework the problem is seen to be explaining how subjective experience arises from neural computation," since all that is given is mindless matter. But the mystery immediately dissolves when one passes over to quantum theory which was formulated from the outset as a theory of the interplay between physical descriptions and conscious thoughts, and which comes with an elaborate and highly tested manchinery for relating these two kinds of elements."

[Cihan Baran]

Hey Kip,

Now, you write that:

When I first got into this stuff, I would have agreed with you: I would say that free will doesn't exist, without exception or qualification. The more I've learned about the problem, the more I've realized that it is too subtle and complex for that kind of answer.

I'd still probably say that free will doesn't exist. But it might not exist in either of two ways: (i) the term is sufficiently well defined that it refers to something logically impossible or (ii) the term is so vague or poorly defined that it cannot refer to anything real (just as muhabutasille, a word I just made up, doesn't refer to anything real). You can call the former the G. Strawson approach and the latter the Double approach. I probably wouldn't say that the problem is solved until I know which of these approaches is right and can demonstrate its truth to others.

I have been reading your posts and comments on the blog with great interest on this blog but I have never quite understood your fascination with the vagueness of what free will means. I don't see how the vagueness of a term can have an effect on our philosophical conclusions.

Note that (i) and (ii) don't differ all that much for most ordinary folks - or so I'd imagine. When people study the problem of free will, they usually are interested in learning whether humans can be free or morally responsible in any deep or meaningful way. Since (i) and (ii) converge on this point, the metaphysical issues seem to a large extent settled.

However, let me now argue why the vagueness or the use of terms should have no bearing on our philosophical conclusions. Imagine a planet where people use the word "two" to refer to 1 and "one" to refer to 2. So when these people say "two plus two equals one", what they are really expressing is "1+1 = 2". However, the common usage of the terms has no effect on the conclusion that "1+1=2". "A rose would smell sweet by any other name," as Shakespeare wrote.

Similarly, if a term or expression is vague, you could disambiguate it in various ways and tell us your thoughts on these various disambiguations. In fact, a paradigmatic example of this is what Russell does with the sentence "King of France is not bald." in On Denoting.

Thus `the present King of France is not bald' is false if it means
`There is an entity which is now King of France and is not bald',
but is true if it means
`It is false that there is an entity which is now King of France and is bald'.

So, to an agnostic observer who agrees that free will is a vague term, I would just go through each disambiguation and argue that it doesn't hold up to scrutiny - and I don't really have to be interested in which disambiguation is the correct one. I'd say "Look, if free will is the freedom to do whatever you want, then B.F. Skinner's Walden II's behaviorally engineered people are also free. If to have free will means to be reasons-responsive, then Ernie in the Zygote argument is also free. If free will means to have second-order desires, then Richard Double's example of the youth in a cult also has second-order desires. If to have free will means to have indeterminate decisions, then random decisions would also count as..." and so on.

The interesting part of the free will debate is that the debate revolves not only around the various disambiguations of the term but also how exactly it should be disambiguated.

You also write,

I want to reserve judgment until I see data showing:

1. how "free will" is commonly used by ordinary people
2. whether there are individuals differences between different camps in the free will debate, such as differences in vulnerability to cognitive biases, which might explain why one groups so adamantly believes they are right despite being (according to the other camps) wrong.

For (1), I think I have already explained why ordinary people's common usage wouldn't be relevant - at least for me. People could be using the term free will to denote Free Willy - the friendly killer-whale or some such thing.

As for (2), well, you can only find biases that help the non-realists. I don't think there can be a bias (maybe depression?) that favors non-realists.

[Cihan Baran]

Let me rewrite the last part:
For (1), I think I have already explained why ordinary people's common usage wouldn't be relevant - at least for me. People could be using the term free will to denote Free Willy - the friendly killer-whale or some such thing - I couldn't care less.

As for (2), well, you can only find biases that help the non-realists. I don't think there can be a bias (maybe depression?) that leads people into non-realism.

[Kip Werking]

Cihan,

Thanks for the thoughtful post. I'm flattered to hear that you find my ideas interesting. It would take me quite a while to answer your questions in detail, but let me try my best:

1. I don't think (i) and (ii) converge on the same point. I don't think (ii) converges on "whether humans can be free or morally responsible in any deep or meaningful way." I tend to think "deep" and "meaningful" are too vague to decide questions in this debate.

2. Regarding the Twin Earth hypothesis: imagine that, on this world, half of the people call one "two" and the other half call one "one". Imagine that, instead of being separated from each other on different parts of the planet, they are intermingled and try to talk to each other. What will happen? Chaos.

But I think the situation is even worse than this. Imagine that people are disagreeing, not about important things like one and two, but about a term that people rarely if ever use (e.g. "free will") and, furthermore, do not have absolute commitments to either definition---even proponents of one definition can feel a little pulled in the direction of the other. This is more like the situation with "free will" I think.

3. You're absolutely right that we can "divide and conquer" by talking about free will', free will'', and so on. But that strategy is silent on the question of whether "free will" exists. Perhaps you are not interested in that question, but I am, even if I think it is "arguing semantics".

4. Regarding your argument about the relevance of common usage ("Free Willy"), if "free will" meant "Free Willy" (which it clearly does not), then I would want to know that. It would help settle questions like "does free will exist?" and "is compatibilism true?" I would also be interested in exploring the ways in which people have "deep" or "meaningful" freedom/responsibility. But this latter interest does not negate my interest in the more semantic debate.

5. I'm glad that you mentioned "depression" because one of my favorite pet theories (however much of a long shot it may be) is that there is some truth to the psychological ideas of "depressive realism" and the "positive illusions", and this phenomena inclines more depressed or melancholy persons to accurately denying the existence of free will. Psychologists have argued that depressed people are less vulnerable to wishful thinking, the fundamental attribution error, and the illusion of control. And I would hypothesize that they are less vulnerable to the just world phenomenon and the tendency to demonize perpetrators.

There is one bias, called anchoring, which I suspect may favor realists over non-realists. Fischer, in one of his articles, suggests that non-realists suffer from "metaphysical depression" because they focus too much on responsibility undermining features of agents, and too little on responsibility enhancing features. They "anchor" on the wrong aspects of agency. Despite all of the caveats above, Fischer's idea strikes me as a fascinating possibility.

But I doubt any of us can decide these questions from our arm chair. You say that "you can only find biases that help the non-realists", but perhaps tomorrow Eddy will describe, at length, how ten different biases actually favor compatibilists. I don't think many philosophers, realist or non-realist, have much investigated the heuristics and biases literature and its relevance to the free will problem, so I think it would be premature to declare any sort of victory.

[Paul Torek]

On September 12 2001, a friend of mine (and this is an otherwise very intelligent person) remarked something to the effect that it would be wrong to try to understand the motivations of the terrorists. You can't get much more blatant than that. In his mind, apparently, evil is inherently inexplicable and to try to explain its genesis is somehow belittling its evil.

I should have said something, but I was too flabbergasted.

[Kip Werking]

Thanks for sharing that, Paul. It is precisely the sort of phenomenon I am talking about. And, even if it is anecdotal, it reinforces my suspicion that something like the phenomenon I describe in my post is a very real aspect of human moral psychology.

[Eddy Nahmias]

I suspect people have a hard time distinguishing an explanation from a justification. So, if you explain terrorist's actions, you may thereby justify it (so don't explain it, for pete's sake!). This may be fodder for incompatibilists, since it may look as though people think that you can't be responsible for an action that can be (fully) explained. But we should not forget that people's beliefs and desires and reasons may also be part of the explanation for what they do.

Notice that what people were really upset about after 911 were not explanations in terms of the terrorists' evil beliefs and desires (their lousy reasons for doing what they did) but justifications for the terrorists' beliefs and desires (e.g., they were reacting to Western hegemony or the plight of the Palestinians or whatever). And perhaps people also get upset about explanations in terms of responsibility-mitigating causes (they were insane or they deprived or something).

[Cihan Baran]

This may be fodder for incompatibilists, since it may look as though people think that you can't be responsible for an action that can be (fully) explained. But we should not forget that people's beliefs and desires and reasons may also be part of the explanation for what they do.

This motivates a regress argument. I don't know how good. I offer it for others to criticize.

(1)An action A by S is causally, fully explained either in terms of (a)beliefs and desires and reasons or (b)other impersonal facts, i.e. random, uncontrolled twitching. [I am really doubtful about the way I set up (b).]

(2) If explained by (b), then S can't be held morally responsible for A.

(3) If explained by (a), then those beliefs, desires and reasons must also be in turn causally, fully explained by (a*) other reasons or (b*) other, impersonal facts.

(4) If explained by (b*), then S can't be held morally responsible.

(5) But we can't indefinitely fully, causally explain someone's actions in terms of their reasons, beliefs, desires. The regress must terminate at some point with some form of (b).

(6) Hence, S can't be held morally responsible for A.

Okay, so I don't think it's "formally" valid but I hope at least writing it this way makes it easier to understand.

As I mentioned, I think th chief weakness of the argument is (b)? Just what are "other, impersonal facts"? Perhaps a better way to define (b) would be as non-(a).

Somewhere here, Kim's "Mechanism and Explanation" plays a large role but I still haven't quite figured it out.

[Mark]

Cihan,

As stated, (2) is patently false. It would need to say, "If explained only by (b), and not (a), then S can't be held morally responsible for A." Call this revised version (2*). Moreover, as stated, (5) has nothing to do with your conclusion (6). (5) is only speaking about how (a) can happen by means of (a*) or (b*), and both (2) and (2*) are compatible with (b*) being the explanation of (a) regardless.

You need some additional premise that says something to the effect of, "personal facts cannot proceed from impersonal facts," which could be derived by combining your premise (3) with (2*). There are two main lines to challenge this premise, both of which involve rejecting premise (3). First, naturalists who want to salvage moral responsibility will say personal facts can be produced by impersonal facts (e.g. by natural evolutionary processes) and thus avoid the infinite regress. Second, super-naturalists can accept the premise and avoid the infinite regress by maintaining that the source of all facts in the world is intrinsically personal in nature.

So, to repair the argument you would need to supplement it with a killer premise like, "there are no personal facts", but good luck selling that premise! Most of us would be prone to think that premise is outright self-defeating: for if there are no personal facts, who is making the argument in the first place and to whom is it being made?

What you are going for here really is no different than Galen Strawson's Basic Argument since (2*) is exactly what Strawson has in mind. And I think in the end your version turns out to be just as weak as Strawson's...

[Cihan Baran]

Mark,

I don't see how (2) is patently false. (Maybe I am missing something.) A simple manipulation case is enough to see that both (a) and (b) can explain someone's actions without moral responsibility. Suppose Manipulator installs a physical process P into Victim's brain that will make Victim have reasons R so that Victim will do A. Now you can explain Victim's actions both by A and P but he is not morally responsible.

You need some additional premise that says something to the effect of, "personal facts cannot proceed from impersonal facts," which could be derived by combining your premise (3) with (2*).

I certainly don't need such a premise and such a premise is so obviously false. I think here you are confusing "personal facts" with "moral responsibility granting facts". A person's taste buds may have such a structure and make-up such that she is more sensitive to sweetness and thus very much likes a certain exotic food X more than other adults. That this person likes X is a "personal fact" about her but it can be explained by natural facts.

Now consider substituting "liking an exotic food" for some morally charged trait like "being against abortion".

If I give up the formalisms, I think the point will emerge more naturally.

Suppose there can be two sorts of explanations for someone's actions: one in terms moral responsibility granting vocabulary such as (a)desires, beliefs and reasons, (b)the other in terms of vocabulary that's not granting moral responsibility, one that makes no reference to agent's reasons, desires.

Suppose you explain someone's action by (a), i.e. reasons, beliefs and desires. And now you need to explain (a). You can either explain (a) in terms of some other (a*) or (b*). If you explain in terms of (b), then by hypothesis, (b) is not moral responsibility granting and agent is not morally. You could explain in terms of (a*) but eventually, keep explaining and at the you would eventually run into some form of b^(x*) - b to the power of x many *. Since (b) is not granting moral responsibility by hypothesis.

What you are going for here really is no different than Galen Strawson's Basic Argument since (2*) is exactly what Strawson has in mind. And I think in the end your version turns out to be just as weak as Strawson's...

I don't think so. Of course there is a similarity but my argument is presented in terms of "explanations" whereas Strawson's argument is in terms of "self-making". I think the point is more summarily made by Alfred Mele, as quoted by John Fischer's book "My Way" (emphasis mine):

Alfred Mele has offered a useful suggestion here. That the idea: an agent is independent, in the relevant sense, according to the incompatibilist, insofar as he makes an explanatory contribution to his behavior, the making of which cannot be fully explained by the laws of nature and the state of the world at some time prior to his having any sense of the apparent options.

[Mark]

Cihan,

Okay, so I think I read too much into your language. It seemed to be that you were referring to (a) in a way that bestowed or conferred responsibility. However, as you noted, of course there are cases of an agent's action being attributed to an (a)-type explanation in a way that is responsibility undermining (such as manipulation, etc.). So, we should distinguish two kinds of (a)-type explanations: those that are responsibility conferring, (a+); and those that are responsibility undermining, (a-).

Your (2) said that an agent cannot be responsible for an act if the act has a (b)-type explanation. However, (2) is patently false because a single act could have both a (b)-type explanation and an (a+)-type explanation. Fischer calls this phenomenon over-determination. (For relevant examples, see any paper he's ever written...)

So, I attempted to repair this defect in (2*), but now that we are distinguishing between (a-) and (a+) explanations, I see the need to specify a (2**) as follows:

(2**) If explained only by (b)-type or (a-)-type explanations, and not (a+)-type explanations, then S can't be held morally responsible for A.

Since you'll say that (a-)-types must be explained by either (a+)-types or (b)-types, your argument still rests on the idea that (a+)-type explanations cannot come from (b)-type explanations, correct?

If so, then the rest of my original criticism is still applicable. Just replace references to (a) with (a+). Naturalists desiring to defend responsibility will say that (a+)-type explanations can have natural causes (i.e. (b)-type explanations), and super-naturalists who want to defend responsibility can also say that (a+)-type explanations are a basic feature of our world given that reality is fundamentally personal in nature. In other words, both groups have a way of explaining the grounds of (a+)-type explanations that are required to ground responsibility.

So, your argument seems to fall flat on its face: the only people who will accept it are those who already deny responsibility.

[Mark]

Cihan,

I thought I would add that I have a very hard time swallowing this statement you made in your last post:

A person's taste buds may have such a structure and make-up such that she is more sensitive to sweetness and thus very much likes a certain exotic food X more than other adults. That this person likes X is a "personal fact" about her but it can be explained by natural facts.

I can see how you can account for facts about taste buds, facts about neurons, facts about the human nervous system, facts about humans in general, but I still lack a framework to understand what you mean by "her", "likes", and "personal" (surely liking sweets isn't reducible to the sensitivity of one's palate). Moreover, without a responsibility endowing notion of personhood at play even the verb "to explain" becomes quite suspicious...

[Mark]

David,

Granting that you reject the compatibility of determinism and responsibility, how do you suppose that indeterminism helps? I understand the significance and role of conscious experience in your argument, but I don't see how granting the truth of indeterminism helps your case. Your assertion requires at the least a minimal account of consciousness that demonstrates its incompatibility with determinism.