Free Will and Cognitive Biases Part 4: The Most Important Bias of All?

When I researched free will and cognitive biases last summer, I found myself in an awkward position. The plausibility of some biases having influence on the free will debate was obvious enough: the illusion of control, the fundamental attribution error, the just world phenomenon, etc. But the plausibility of other biases being relevant was often less obvious. Yet, sometimes, it was these “hard sells” that I, personally, thought would be most important to the debate.

In this shorter post, I want to describe one of the harder sells. To convey the notion, I first want to focus upon the idea of a happy coincidence: people might find certain constraints on their alleged freedom to be less disturbing if they are more confident that things turned out, nevertheless, the “right” way.

Consider this interesting data point: those who tend to be non-realists about free will also to also be non-realists about moral truths. This is true of at least myself, Tamler Sommers, Richard Double, and Joshua Greene (correct me if I have gotten this wrong). Why might this be? It may just be that these persons find themselves attracted to deflationary views in general (e.g. perhaps they enjoy shocking people with their outrageous claims). I want to suggest something deeper is going on: such non-realists feel it is less of a happy coincidence that they are who are they are and do what they do. For example, a person who both (i) believes that killing innocent people is wrong and (ii) finds himself as the sort of person who neither kills innocent people nor wants to do so probably considers this a happy coincidence. Even if he lacks the freedom to be, or have become, such a killer, this freedom does not undermine his sense of free will much, because he has nevertheless found himself on the one, true path (so to speak). A non-realist (or anti-realist) about moral truths, however, does not have this luxury.

Given this consideration about happy coincidences, I wonder: is there a cognitive bias that might be relevant here? I had not thought so until I remembered Smilansky’s lovely phrase “we are merely the unfolding of the given.” Smilanksy is right, I think, to characterize the free will problem as one about the self, and one’s relationship with the self, and this worry about given-ness. Just how disturbing is this given-ness and how disturbing should this given-ness be? There is a prominent cognitive bias that might distort our judgments about that which we are given: it is called the endowment effect. According to the endowment effect, people develop undue liking, or irrational fondness, for that which is given to them. There is similar bias called the mere exposure effect: people develop undue liking, or irrational fondness, for that with which they are familiar.

Suppose that the endowment effect and mere exposure effect are real. Then one tantalizing possibility is that compatibilists and libertarians finds Smilansky’s given-ness to be less disturbing because of a happy accident—an arguably irrational one. The idea here is that, because we are given the original traits and characters we have (the “slant” of our constitution, to use Fischer’s phrase), that alone gives us a preference for being ourselves. And as we grow more familiar with ourselves, at the expense of familiarity with others, this irrational preference grows and continues. Because we prefer being ourselves, we feel less disturbed at the given-ness which Smilansky describes. By analogy, a boy at Christmas might say “wow, I thought it was so awful that I couldn’t choose my own Christmas presents, but then I discovered that I got just the presents I wanted! With a Santa Claus like that, who needs to make their own decisions about presents?” We regard ourselves as happy accidents but, to the extent that these biases are irrational, this may be a mistake. I actually suspect that this bias may be the most relevant of all to the free will problem (which is not to say that I suspect its absolute relevance is great; further experiments may help settle that different question).

Comments

[Kip Werking]

Let me add some questions for other Gardeners, to see if I can stir up some conversation:

1. Do you think the endowment effect is real?
2. If so, do you think it might apply to the original charcters and constitutions with which we are endowed?
3. If so, do you think we therefore tend to regard worries about luck as being less relevant (e.g. "it's ok that I didn't choose or determine my own character, because it turned out the right way anyway")?

[RICHARD DOUBLE]

Kip,

Are you saying:

1. Comps and libs who are satisfied with themselves morally gain consolation for their happy coincidences from the belief that they have objective morality on their side,

2. Whereas we moral non-realists tend not to gain consolation for our happy coincidences. So, we are less sanguine about our lack of moral freedom.

I find this suggestive [worth investigation] and ingenious. [Especially because it is fully in accord with my attitudinal view of free will.]

One caveat. I suspect that attitudes toward happy coincidences will be very idiosyncratic. I believe psychology gives insight into general processes [e.g., perception, memory, language, general cognition, but is weak at explanations for specific phenomena [‘individual differences’], especially something as rarified as attitudes about moral freedom.

[Kip Werking]

Richard,

Thanks for the kind words.

I would rephrase your 1 to say:

1. Compatibilists, and to a lesser extent libertarians, feel less disturbed by constitutive luck, because they feel lucky: they feel capable of still walking the straight and narrow. In the same way, a lottery ticket winner probably doesn't worry that the selection of winning lottery numbers wasn't under hir control. It all worked out in the end. Our lives, in this weak sense, are happy coincidences.

In contrast, non-realists like you and me do not recognize any straight and narrow, in a rigorous sense. Moral talk, and the human condition, are infected with a certain arbitrariness.

Consider this sort of scenario: we could have been born into a world where everyone feels it is objectively right and good to rape and torture innocent people, just as, in this world, everyone (largely) feels it is right and good to not rape and torture innocent people. Yet no person, in either world, seems to have a good argument, against the other, for why their beliefs are the right, true beliefs about morality. This consideration helps show how objective moral truths don't exist.

The moral realist looks at this sort of situation and says "oh no, the people in that other world just got it all wrong; it actually is objectively wrong to rape and torture innocent people, even if nobody thinks so." Then the realist might add: "I did not have any ultimate control over my character, but fortunately, I was endowed with just the sort of character I need: I don't want to rape and torture innocent people."

The non-realist cannot really say this. S/he feels "thrown into the world", with an arbitrary character. And this character cannot, happily or unhappily, track or fail to track what a good character would be, because there is no such thing as the right, good character. There are just arbitrary points in character space.

Second, you raise an interesting point about psychology and I would want to hear more about it. I certainly agree that psychology has a miserable record of explaining certain phenomenon and treating certain problems.

[Cihan Baran]

Hey Kip,

Today, I received an email that said I was rejected for an internship that I really wanted.

Suddenly, the non-reality of free will seemed so crystal clear. :P

What do you think of that? The reverse-endowment effect?

*In fact, one aspect of Fundamental Attribution Error is that people tend to come up with situational explanations, assign blame to other people when things go badly. That is, people have a self-serving bias.

**Also, after I read your blog entries on cognitive biases, I checked out an introductory social psychology textbook from the library. There are so many more similar biases - biases that would lead to a belief in free will. In fact, it wouldn't be wrong to say that the whole field is constantly producing evidence for the non-reality of free will. However, I think you take these biases far more seriously than their worth.

[Kip Werking]

Cihan,

I wouldn't call that the reverse endowment effect, because it says nothing about how much the denial/rejection affected your attitude towards the internship.

"In fact, one aspect of Fundamental Attribution Error is that people tend to come up with situational explanations, assign blame to other people when things go badly. That is, people have a self-serving bias."

I'm not sure if this is an aspect of the same bias, the FAE, or a different bias altogether (either way, this is just a semantic point).

But this is not the first time at the GFP that someone has mentioned this problem. Only future research can show which biases are relevant, and to which extent, in explaining belief (or non-belief) about free will. (Note that, according to the positive illusions, people feel they are better than they actually are, and expect their futures to better than they it turns out to be.)

Fortunately, the FAE is not the only bias I think is relevant. The FAE may account for mistaken moral judgments about others, if not the self. But humans also seem vulnerable to wishful thinking, the illusion of control, the fundamental attribution error, even the endowment effect (etc.).

"Also, after I read your blog entries on cognitive biases, I checked out an introductory social psychology textbook from the library. There are so many more similar biases - biases that would lead to a belief in free will. In fact, it wouldn't be wrong to say that the whole field is constantly producing evidence for the non-reality of free will. However, I think you take these biases far more seriously than their worth."

Well, I'm very glad to hear that you think there are plenty of biases favoring free will, and few favoring disbelief in free will. That is my own (unpopular)hypothesis and the impression I have of the relevant literature. I talk about the relevance of many of these biases (as many as 15, although some of them cluster together in groups of similar errors) in the article I wrote last summer.

In fact, psychology has a history of being unfriendly towards free will (e.g. Freud, Menninger, Skinner); a recent informal poll published in Nature showed that evolutionary psychologists are very skeptical of the existence of free will. Some might say that psychology itself presupposes that we lack (libertarian) free will and, and it must do so, in order to both explain behavior and treat patients. I suspect that one reason the belief in free will remains so popular is that philosophy has isolated itself, too much, from psychology. Instead, these two should work together.

Do I take these cognitive biases too seriously? This is very possible. It is important to remember Nichols' comments on automaticity and cognitive biases: because these results are sexy and counter-intuitive, the institution of psychology itself is biased (meta-biased) towards publishing these results. Few people document the mundane extent to which people are actually rational and self-controlled; those experiments are boring. People want to learn about how shockingly stupid and vulnerable human beings are (e.g. the Milgrim experiments). This is an important insight to remember: given this meta-bias, we should take the emphasis in the social sciences upon the irrational of human beings with a grain of salt.

[Cihan Baran]

Kip,

Well, to get the semantics right, FAE is the tendency to come up with dispositional as opposed to situational explanations. And I believe the self-serving bias is another kind of bias that says people like explanations that serve them. So if you combine both, people will come up with dispositional explanations and ascribe praise to themselves for doing something good. People will come up with situational explanations and/or ascribe blame to others when something bad happens. Or at least tend to...

There is another bias called the confirmation bias. It says that people are likely to find confirming evidence for their beliefs and disregard evidence that might refute their beliefs. So that might explain why a free will non-realist might find all these biases so relevant but, say, a compatibilist may not.

However, I feel like I have given that book a fair reading and I am not necessarily a free will non-realist (I think I am a true skeptic -i.e. can't make my mind up or change from time to time). It seems as if that all these biases would also favor a belief in free will.

Also, bear in mind, all these social psychology bits operate against the backdrop of intuitive views about free will, justice and so on. There is the control illusion because genuine control is a possibility. Only some situations give rise to the illusion of control. There is the just world phenomenon because there can be legitimate justice.

I would also add another institution bias to the one you give. Social psychologists are interested in doing experiments. I may be wrong but it seems to me that it's much easier to measure the effects of situations as opposed to dispositions. Because of this, if you wanted a social psychology experiment to publish stuff, you would just look for situations that affect people's behavior. That itself is a meta-bias that favors research that would undermine free will.

[Rob]

Kip,

You may appreciate Daniel Wegner's response ("The Godfather of Soul," p. 30) to this article in BEHAVIORAL AND BRAIN SCIENCE article in which "the experience of conscious will" is correlated with belief in the afterlife:

http://www.wjh.harvard.edu/%7Ewegner/pubs.htm

[Kip Werking]

Rob,

Thanks for the pointer. That sounds like exactly the sort of thing that would interest me.

Kip