More on science and free will

Some empirical evidence that the belief in free will affects behavior:

    Belief in free will can be manipulated with methods developed by Vohs and Schooler (2007). These researchers had some participants read an essay by a well known scientist (Francis Crick, a Nobel laureate) rejecting and indeed mocking the notion of free will. Others read a neutral essay. Another manipulation involved a procedure in which participants read aloud a series of statements emphasizing either freedom of action or lack of freedom and determinism. In those studies, participants who had been induced to disbelieve in free will were later more willing than controls to cheat on a test.
    Likewise, manipulations of belief in free will have been shown by Baumeister, Masicampo, and DeWall (2006) to affect other social behaviors. Participants who had been led to disbelieve in free will were significantly more aggressive and less helpful toward others.
    We are not suggesting that these studies be taken as proof of the existence of free will. Nevertheless, taken together, these findings indicate that not only is the belief in free will normative, but it is also socially beneficial. Undermining that belief leads to an increase in antisocial actions (cheating and aggression) and a reduction in socially desirable behavior (helping).
    Also, in these studies, manipulations aimed at promoting belief in free will typically yielded results identical to neutral controls, which suggests that encouraging people to believe in free will simply reaffirms their ordinary state. That is, people normally believe in free will, and getting them to disbelieve in it is the departure from normal. This pattern indicates that belief in free will is woven into the fabric of everyday social life and the assumptions according to which people perceive and interact with each other.

[from Baumeister, Sparks, Stillman and Vohs (forthcoming) Free Will in Consumer Behavior: Rational Choice and Self-Control. Journal of Consumer Psychology]

Some necessary caveats: the notion of free will that was manipulated was almost certainly some kind of contracausal notion (it’s hard to be sure; both the studies cited are unpublished). The authors of this paper note that many philosophers are compatibilists, but having noted it they seem immediately to forget it. Second, we need to distinguish the following two situations: shaking belief in free will affects behavior (for the worse) and shaking belief in a cherished notion affects behavior (for the worse). Still, some very interesting results for optimistic disbelievers in free will to mull over.

Comments

[Q the Enchanter]

The results don't surprise me, but I'd also distinguish shaking a belief in a society that holds the belief to be a sine qua non of morality and meaningful action, on the one hand, and shaking that same belief in a society that by and large has gotten over or beyond the belief, on the other. (I.e., it's still an open question on these data whether the apparent practical dependence on belief in free will is innate or cultural.)

[Jasper Yate]

I find it very hard to believe that any reasearchers could contrive a plan in which they could manipulate any given person's belief in free will and have this data be at all relevent to actual social conditions. Furthermore I think it extremely irresponsible of these researchers to be toying with these peoples minds when they could be actually effecting their lives and influencing them in positive ways instead of treating them as, well, test dummies. As philosophers, these people must see the errors in treating people as objects for ones study and how these kinds of problems are a plague to humanity in the modern hyper-real commercially controlled world.

[Neil]

Jasper, the manipulation consists in getting them to read Francis Crick! If that's impermissible, then we're all in trouble (some of us perhaps a bit more than others). BTW, they're not philosophers.

[Eddy Nahmias]

More ammunition for neurotic compatibilism: see this ad
http://www.allstate.com/content/refresh-attachments/Brain-Ad.pdf

Here's a relevant quotation:
"Even bright, mature teenagers sometimes do
things that are “stupid.” But when that happens, it’s not really their fault. It’s because their brain hasn’t fi nished developing. The underdeveloped area is called the dorsal lateral prefrontal cortex. It plays a critical role in decision making, problem solving and understanding future consequences of today’s actions. Problem is, it won’t be fully mature until they’re into their 20s."

No wonder I was so stupid when I was 16. I only wish I had this research back then to show my parents.

[Kip Werking]

I remember when Josh Knobe mentioned this research (or research like it), in response to Feltz and Nadelhoffer's critique of Smilansky's Illusionism at Inland 2006.

http://experimentalphilosophy.typepad.com/experimental_philosophy/2005/09/trying_to_test_.html

(It's been a while, so please correct me if I am getting the details confused.)

Does this post suggest that Smilansky's ideas may rest upon a foundation that is not so shaky after all?

[Neil]

Kip - yes, but the evidence is no more than suggestive at the moment. I drew attention to some caveats, and Q added another: we need to be sure that the effect is specific to beliefs in free will and that the effect will persist. The second is big problem, since it is experimentally intractable (though maybe there is cross-cultural data that could be used).

[Cihan Baran]

Fascinating paper, even though I must admit I liked the parts unrelated to free will the best.

As far the excerpt, I don't think its central claims are true. Knobe and Leiter wrote a paper on Nietzsche's moral psychology, in which they defend the idea that it's mostly the psycho-physical facts about us that determine our moral character. Perhaps, Leiter and Knobe's more important tenet was that for most people, it is not actually the conscious deliberations that form our moral characters. Genes and heredity actually play a more causal role.

Now if anything like that is true, I don't think people will be swayed all that much by losing their belief in free will. I don't think *any* philosophical idea could move people that much.

Also consider this: Ask a religious person what he or she would do if he or she found out one day that there was no God. It might happen that he or she might say, "I'd kill and steal" in "If there is no God, everything is permitted" style. However, statistics shows that a great majority of prisoners believe in a God.

I think something also similar would also hold true for people's opinions about free will.

It seems to me that there is a cognitive bias at work here, when people imagine themselves as believing in a view that they don't currently believe in.

Here is a way to explain it: When a religious person imagines herself as an atheist, her imagination would exclusively focus on the "loneliness" or dread having such a view would indicate. Nevertheless, when one becomes an atheist, one doesn't stop eating, walking, breathing, having fun and so on. By exclusively focusing on one aspect in their vicarious exercise, people ignore other aspects. This might explain why they get it wrong.

Now similarly, when people imagine themselves without free will, their imagination will exclusively and selectively (as a sign of "metaphysical depression") focus on the fact they don't have the contra-causal freedom they thought they had had. However, wine doesn't cease to be wine once you lose free will - life is still life without free will. (Or is it?)

So, I don't believe that despite this experiment, people's behavior will change substantially if they lose the belief in free will.

[Neil]

Cihan,

I believe I read the Knobe/Leiter paper, though I don't recall it right now. But I find it *very* hard to believe that they argued that "genes and heredity actually play a more causal role" than conscious deliberation. I'm not sure what Josh's views on genes are, but Brian is well-informed on this topic and wouldn't make this claim, which is either false or meaningless. In any case, the experimenters are not particularly interested in manipulating people's conscious beliefs. There is mountains of evidence that system 1 processes - which are non-conscious - play a more important role in guiding most behavior than system 2 (which isn't to say that system 2 isn't important, just that is rare). The experimenters did not try to manipulate system 2; they tried to get at system 1. Note they did not try to manipulate beliefs! It's attitudes that matter here. The manipulation *did* alter behavior, whatever the causal route. That's not to say that the modification will persist, though.

[Cihan Baran]

Neil,

I apologize for attributing some "false or meaningless" views to Leiter and Knobe.

What I meant was people's values are not consciously formed.

Or in their words:
"The Nietzschean account of moral psychology differs from the Aristotelian and
Kantian accounts along almost every dimension. What is decisive is not upbringing,
particular habits, or conscious choice; what matters most are heritable psychological and
physiological traits." (emphasis mine)

This is what I meant. When I said, "conscious deliberations" play a less causal role, I didn't mean that people don't think when they act. (Thank you for the uncharitable interpretation.)

Here is the abstract of their paper:
Contemporary moral psychology has been dominated by two broad traditions, one usually associated with Aristotle, the other with Kant. The broadly Aristotelian approach emphasizes the role of childhood upbringing in the development of good moral character, and the role of such character in ethical behavior. The broadly Kantian approach emphasizes the role of freely chosen conscious moral principles in ethical behavior. We review a growing body of experimental evidence that suggests that both of these approaches are predicated on an implausible view of human psychology. This evidence suggests that both childhood upbringing and conscious moral principles have extraordinarily little impact on people's moral behavior.

This paper argues that moral psychology needs to take seriously a third approach, derived from Nietzsche. This approach emphasizes the role of heritable psychological and physiological traits in explaining behavior. In particular, it claims that differences in the degree to which different individuals behave morally can often be traced back to heritable differences between those individuals. We show that this third approach enjoys considerable empirical support - indeed that it is far better supported by the empirical data than are either the Aristotelian or Kantian traditions in moral psychology.

Here is the link:

[Cihan Baran]

Okay more clarification:
The only question is about whether
these factors actually play any important role in the etiology of people’s moral behavior.
So, for example, Nietzsche would say that people do have conscious moral principles but
that these principles have only a limited impact on the behaviors they actually end up
performing. More typically, people first perform certain behaviors and then develop
principles that serve to justify the behaviors they have already performed. The most
important factors in the origin of moral behavior are people’s basic psychological and
physiological traits; the conscious moral principles largely serve as post hoc justifications
for behaviors that would have been performed either way.

So if something like that is true, I don't think a lack of belief in free will cause anything. That's what I am trying to say.

[Neil]

Cihan, It was unnecessarily bad tempered in what I said. The reason was you hit one my buttons, by referring to genes and heredity. And you seem right in your claim about
Leiter and Knobe - from the abstract, I would say they are making the very mistake I attributed to you. My mistake, and theirs - not yours.

[Tom Clark]

If the methodology and results are valid and can be replicated, these findings suggest that we have to be careful in how we dismantle the widespread intuition of contra-causal free will (CCFW), one of Dennett's perennial and well-taken concerns. As a promulgator of naturalism, I spend a good deal of time reassuring people that without CCFW we can still be held responsible, that we make real choices, that we don't lose our dignity and individuality, that fatalism is false, etc., etc. To simply deny one of people's central assumptions without providing supports and reassurances is irresponsible. Of course this was an experimental set-up so I'm not getting on the authors' case, only saying that as the word gets out it's important to take steps to avoid the demoralization that Saul thinks is inevitable. My guess is that if the experimental subjects were told of the positive aspects of abandoning CCFW, plus were reminded that they can (and will) still be held responsible under compatibilism, then we wouldn't see an increase in anti-social behavior.

[Tamler Sommers]

To repeat more tendentiously what Neil and Tom said...

The authors write:
"Nevertheless, taken together, these findings indicate that not only is the belief in free will normative, but it is also socially beneficial. Undermining that belief leads to an increase in antisocial actions (cheating and aggression) and a reduction in socially desirable behavior (helping)."

Now I'm a big fan of cool social science experiments, but I can't see how anyone could draw that conclusion from these results. First of all, we don't know if the belief in free will was truly undermined--we just know it was jarred somehow. Second, at least in the Schooler study, the subjects were tested right after reading the essay. No time to reflect, no time to talk to anyone, or read or get a more nuanced view about what it would mean to give up the notion of free will. Anyone who is capable of genuinely denying free will is capable of engaging in this reflective process. So these results tell us absolutely nothing about how people would behave after coming to settled reflective beliefs about free will.

A more accurate claim I think would be this: taken together these results show that in the after having beliefs about free will shaken up, before having any time to think about it, people are more likely to cheat and be anti-social in the moments immediately following the shake-up.

I don't think this is nitpicking--this distinction is at the heart of the issue. As an analogy, imagine the following experiment. Assume for the sake of argument that today's atheists and agnostics are not more anti-social or inclined to cheat than contemporary theists. (A plausible assumption, I think, there may even be data supporting it somewhere). Now imagine an experiment where they took committed theists and gave them some passage out of Dawkins where he's mocking the idea that the world has any purpose behind it at all. And then right after they test for cheating. It would not surprise me if the people reading the Dawkins essay did a little more cheating. But I would never draw the conclusion that (a) the people who read it genuinely gave up their belief in God, and certainly not (b) that the belief in God is socially beneficial. At best, I might think that when a cheristed belief is challenged or threatened, people are more likely to have 'what the hell' kind of attitude immediately following the threat.

So to answer Kip: I think illusionsism needs considerably more than this to serve as foundation.

[Neil]

Good points, Tamler. It's worth adding that Crick - unlike Saul - doesn't know whereof he's mocking; it's not like the subjects are being convinced on good grounds that there is no FW. Tom, it's all very well for you to go around reassuring people that responsibility survives free will. It's no help to me, given that I think responsibility is by far the shakier of the concepts.

[Saul Smilansky]

I have an intuition that Kip is right:) ... But opponents of Illusionism shouldn't be completely (as the authors call it) "ego deflated". Seriously, the studies are interesting, but like previous ones aren't decisive. And of course we should have more confidence in studies run by Garden-variety philosophers who know the terminology and are aware of the complications. But I do think that these last studies should make the "don't worry" guys pause. Two related thoughts:

1. It seems interesting that the "don't worry" camp is composed of such very different sorts of views. You have (a) the P.F. Strawson Humean position according to which nothing can really change whatever happens; (b) various compatibilist views that admit that people have incompatibilist views on PAP, FW and moral responsibility, see a possibility for big changes, but feel that a suitable compatibilist replacement (still speaking about free will and moral responsibility) could provide, in practice, a very good alternative; and (c) "happy hard determinists" who think that we can ditch the whole thing but that that would make things much better. Clearly these are starkly different views holding many contradictory stances on both theory and practice. Which leads me to skepticism about all of them; and hence to a skeptical conservatism backed up by Illusionism.

2. Secondly, think what a demanding idea is involved in a significant revisionism (of either a compatibilist or a hard determnist type). You need to be somehow confident that the Folk won't fall too far, that you could somehow control the direction of their fall, and that you have the resources to catch them. And so many things could go wrong! Illusionism begins to sound pretty prudent to me.

[Jasper Yate]

I consider any self-thinking person a philosopher and I consider it a fallacy to say that only people who write books and essays about concepts that they never care to act on in the actual thing that they so obsessivey study, which happens to be human life in the physical universe that we percieve, are philosophers. With that said, yes, these people may only be providing information for certain individuals to do with what their minds will, but the are using these people as objects, not as human beings, and are distancing themselves from any concern with other humans' lives as intellectual beings; they do not care about these peoples minds or their intellectual state - they don't actually want to help or change the world or involve themselves in the human community directly, they want to tell others about a study they did on a few people. They also didn't give all of these people this paper to read, and if you deem it an intelligent piece of writing that an intelligent person should read, than you cannot say that they have any concern for these people because they had a control which did not read this paper. In conclusion to this, the research that treats humans who have apparent free will as objects, as controls, as this and that, is not worth doing if so chooses to disengage itself from some of the people (and therefore assert that they do not care for any actual people) it seeks to enlighten with its results. Furthermore what sense is there in doing research on free will in which the researchers in essence take the free will of these people by telling them to do things and then testing how their minds process this information and how they act on it - how can a study on something that may or may not consist of something be at all useful or possible to get results from if it does not have the thing which it is testing about; i.e. testing water for it's reaction to an addition of electrons, but first removing the hydrogens, effectively making it not water anymore, then adding the electrons, and then again adding back the hydrogens to see how the whole package reacts, it is not observing water, it's observing somethin else; it is the same with this reasearch, they are not researching human beings, they are conducting tests of the mind and how it reacts to something being shown to it without respect to any other contingents of human life and the human experience, and last time I checked this reality isn't going anywhere so respect for this human experience shouldn't be an issue here...

[Thomas Nadelhoffer]

Saul,

Rather than illusionism--which involves intentionally hiding the truth from people--why wouldn't you simply lobby instead for a better primary and second education system that would prepare children to survive disillusionment about LFW and UMR with their emotional and social well-being intact? It is not so much the prudence of illusionism that I object to but rather the paternalism.

Kip,

I agree with Saul, Neil, Tamler, and others that we should view the results of these studies with caution. First, I suspect the Crick excerpt suggested not just lack of free will but rather some kind of by-passing or epiphenomenalism. But Saul cannot appeal to this kind of data in motivating his view since his point is that if awareness of the fact that from the ultimate perspective "luck swallows everything" were to gain traction with the folk, they would lose sight of the fact that they really do have the kind of local, reflective, compatibilist control that is necessary for grounding the community of responsibility. The non-existence of LFW and UMR simply does not entail that our conscious states are epiphenomena. Of course, perhaps Saul would like to be an illusionist about epiphenomenalism in the event that Wegner and others are right--but that's not the view he's been pushing so far.

For what it's worth, I have been working on a paper about this very issue with an undergrad of mine. I hope to have a draft posted in a few days. We rely on the literature from social and developmental psychology on perceived control and experiential control to suggest that mere disillusionment about LFW and UMR would not produce the negative consequences Saul predicts unless people also (mistakenly) conclude that the ultimate perspective shows that their conscious wills are bypassed. I am also working on a follow-up paper about the objective attitude with Tamler, Adam, and my RA Tatyana that relies on data about impact bias and theism/atheism. But more on that later...

For now, I wanted to follow Tamler in pointing out another reason to be cautious when looking at the results of the aforementioned studies (it would be nice if I had actually read the papers in question!). Consider, for instance, the following quote from Shelley Taylor and Jonathon Brown (1988: 198):

"Research indicates that when a positive (as opposed to a negative or neutral) mood has been induced, people are generally more likely to help others...to initiate conversations with others...to express liking for others and positive evaluations of people in general...and to reduce the use of contentious strategies and increase joint benefit in bargaining situations. Summarizing the research evidence, Isen (1984) concluded, 'Positive affect is associated with increased sociability and benevolence'."

If, on the other hand, negative moods are induced, we should expect the contrary to be true. In the Schooler studies, for instance, perhaps people who believe they have free will were put in a bad mood by the except that suggested otherwise and this bad mood (i.e., negative affect) is what caused people to behave less morally not the fact that they were somehow convinced by a paragraph from Crick that they lacked free will and moral responsibility (which seems highly implausible). I am not suggesting that this is the right explanation of the results--as I have not even seen the studies--but I suspect this has just as much explanatory power as the model that was used to explain the data. By my lights, the suggestion that a paragraph from Crick induced people to disbelieve in free will is less believable that the claim that people who believed in free will were pissed off by the suggestion to contrary--which in turn affected their behavior. But I am no psychologist, of course...

Minimally, I think illusionism is not off the hook just yet.

[Tom Clark]

Speaking of disillusionment, Shaun Nichols will be presenting a folk-version of his paper "After incompatibilism: a naturalistic defense of the reactive attitudes" at a meeting of the Center for Inquiry of Southern Arizona, a secular humanist group (he's retitled the talk "After Free Will" for the occasion). The paper is at http://dingo.sbs.arizona.edu/~snichols/Papers/AfterIncompatibilism3r.pdf . I gather he'll reassure everyone that even without LFW and UMR we're still justified in moral anger. I'll be addressing the same group a month or so later, saying that our reactive attitudes might be moderated in the light of giving up LFW and UMR, and that this is a good thing.

Whoever's right, I'm pleased that this discussion is starting to get outside the academy so that people can, as Thomas suggests, learn how to cope with disillusionment. Better yet, don't even install the illusion!

[Cihan Baran]

Saul,

Assuming you are not a bad person, what exactly is the difference between you and the common man that causes this supposed difference in behavior?

What's keeping you from being all evil and horrible?

[Thomas Nadelhoffer]

Cihan,

If I understand Saul's position correctly--and I very well may not--the claim is not that people will be evil and horrible, the claim is that not only will they be disturbed emotionally if they are disillusioned (in an existential angst sort of way), but they will also lose their grip on the local reflective control and compatibilistically grounded responsibility they do in fact have. It wouldn't follow from these two consequences that people would become evil--just less happy and less pro-social. Relatedly, Adam and I tried to make the same point in our paper a few years back, and Saul's response (in person at the INPC--which is where we presented the paper) is that he, too, often falls back into illusion as he goes about his daily affairs. Of course, if it is that easy for him to fall back thusly, then it is unclear why the unphilosophical masses wouldn't simply do the same. So, it is unclear how helpful this response is for Saul's overall view. In essence, it's the Humean idea that after realizing the truth of skepticism, I go back to playing backgammon with my friends as if all were right in the world. And, of course, in this context, I know precisely which moves are allowed in the game and which ones are not. But the lesson from Hume seems to be that disillusioning people about their mistaken beliefs about absolute capital-K knowledge wouldn't do much harm. Life would suck them right back into their former patterns of behavior. I would argue the same thing would happen with disillusionment about LFW and UMR--but that is another story for another day.

On a related note, having now looked at the Baumeister paper, I am even more convinced that the explanation I offered in my earlier comment of their results is correct. Why would any of these psychologists think that they induced a disbelief in free will in the participants just because they primed them with anti-free will excerpts and the like? If nothing else, I would require follow up questions that showed that some of the participants' beliefs in FW and MR really were disturbed, upset, or undermined by the vignettes before I would accept their own preferred explanation of the results. It seems that the negative-affect-leads-to-less-social-behavior model I suggested earlier is much more parsimonious. If I am correct about this, then the results of these studies offer no help to the illusionist about FW.

[Micky]

Our moral freedom, like other mental powers, is strengthened by exercise. The practice of yielding to impulse results in enfeebling self-control. The faculty of inhibiting pressing desires, of concentrating attention on more remote goods, of reinforcing the higher but less urgent motives, undergoes a kind of atrophy by disuse.
PEACE BE WITH YOU
MICKY

[Cihan Baran]

Thomas,

Thanks for the clarificatory note. You say,
"the claim is not that people will be evil and horrible, the claim is that not only will they be disturbed emotionally if they are disillusioned (in an existential angst sort of way), but they will also lose their grip on the local reflective control and compatibilistically grounded responsibility they do in fact have."

Well, I haven't studied Saul Smilansky's views in great detail, so I am probably wrong in my original assertion that disillusionment (in his sense) will cause people to be "all evil and horrible." However, note that you say that "they will also lose their grip" - what exactly does this consist in?

Sure, people will feel bad - in fact lack of free will may be the most depressing philosophical truth - but surely, this won't change people's morally relevant behavior in any significant way.

Some theist on reflection might come to renounce her faith and she might feel depressed about her atheism, too. I don't think this emotional disturbance would change her moral atittudes.

[Thomas Nadelhoffer]

Cihan,

I happen not to think that people will really feel all that bad post disillusionment about LFW and UMR, to be honest (I know I didn't)--especially if all we're talking about is being disillusioned about LFW and UMR. So long as people still feel that their mental states are causally efficacious, that they have local reflective control, and the like, I am not even sure they'll be all that depressed. Of course, if epiphenomenalism or some other bypassing story ends up being told, then of course there would be pause for concern.

And as far as what I meant by "lose their grip"--I simply meant that people will lose sight of the fact that they are still locally accountable for x-ing (even if they're not ultimately responsible for x-ing). As a result, people may feel less guilt as well as less vengefulness. Of course, this may lead people to misbehave more than they did when they had the purportedly mistaken beliefs LFW and UMR, but it does not follow that the ordinary person will start murdering, raping, pillaging, etc. just because they realize that from the ultimate perspective, "luck swallows everything." I take that to be Saul's position--but he would obviously know better than I! I was just pointing out that Saul doesn't seem to move from disillusionment to evil--rather, he focuses instead on the overall emotional and social disutility that would purportedly follow.

[Tamler Sommers]

Just to follow up on what Thomas said, Saul talks about the danger of 'retrospective disassociation,' meaning that hard determinists or skeptics would come to see past behavior as actions that cannot be their fault, and over time this might corrupt their behavior in the future. (I think that's nmore or less right, I don't have the book with me now.) I think that's a worry that needs to be taken seriously. Not that people would become like the characters in Natural Born Killers. My sense is that most people have more that's preventing them from becoming serial killers than the belief that they'll deserve blame for it. But it might have long term effects in the smaller decisions, those close calls between immediate self-interest and doing the right thing. Cheating on taxes maybe, or acts of infidelity.

All that said, nothing in the Schooler study supports or undermines this worry. This is a long-term danger as I see it, people would have to really incorporate hard determinism into their lives before it would take effect. Coming up with an experiment that would shed light on this would be a challenge to say the least.

[Neil]

Thomas,

A smallish point: it doesn't much matter whether the experiment changes explicit beliefs or only implicit attitudes, as you suggest: if the hypothesis is that belief in FW promotes pro-sociality, then altering implicit attitudes without altering explicit beliefs is a valid test (compare: racist beliefs support racist behavior; test hypothesis by priming racist stereotypes).