A Species of Unconscious Autobiography?

Nietzsche famously wrote that every great philosophy was a 'species of involuntary and unconscious autobiography.'  The suggestion here is that our philosophical theories, no matter how complex and technical they may appear on the surface, ultimately have their origin in certain facts about the kinds of people we are. 

In a pair of recent papers, Adam Feltz, Edward Cokely and Thomas Nadelhoffer have been applying this basic insight to the study of free will.  They have conducted series of intriguing studies exploring the ways in which different people can end up with quite different intuitions about the free will problem. 

In one of these papers, Feltz, Cokely and Nadelhoffer show that different people truly do have different intuitions about free will and that these different people each hold remarkably consistent views.  (This article also includes some very intriguing philosophical reflection on the significance of the findings.) 

Then, in a second paper, Feltz and Cokely show that the differences between people's intuitions can to some degree be predicted by their differing levels of the personality trait extraversion.

These studies open up the tantalizing possibility that the free will problem has been so persistent, not because either side is making an error in reasoning, but rather because the two sides truly do start out with quite different intuitions.  I would love to hear any thoughts you might have about whether this possibility is actual. 

Comments

[Manuel Vargas]

Joshua, thanks for calling attention to these really interesting papers.

A couple of quick thoughts:
1. If Nietzsche said it, it must be true

2. FCN do a nice job of identifying and probing some interesting issues in the Nahmias et al/Nichols and Knobe pieces

3. Some explicit autobiography: It strikes me as plausible that some adults give consistently compatibilist and others consistently incompatibilist intuitions. Indeed, it was some of this experimental work that pushed me in that direction. I've confessed to this a bit in print, in Ch. 4 of Four Views on FW, after discussing the Nahmias et. al./Nichols&Knobe work. And, I've even briefly floated a half-baked explanation of how this might happen in "Philosophy and the Folk" (2006). The basic idea: many of us might start out with mixed intuitions but come to lean one way or another in light of various pressures to systematize our convictions. In the (2006) piece I mainly have professional philosophers in mind, and I mention some mostly non-rational sources (there's another Nietzsche-consistent view!) but I also think there are some rational pressures towards consistency, conceptual unity, and the like that may arise.

4. On the extraversion idea, interested readers might also want to compare the chapter by Dweck and Molden in the recent collection _Are We Free?_ ed. by Baer, Kaufman, and Baumeister.

[Mark Smeltzer]

On the topic of relating people's intuitions to their takes on the free will question, I tend to side with Keirsey (and Plato) in identifying the following distinct kinds of persons: Artisan (roughly 40% of the population), Guardian (40%), Idealist (10%), and Rational (10%).

Those personality types are in turn related to propensities toward the following meta-ethical positions (respectively): egoism, deontology, virtue ethics, and utilitarianism.

I tend to relate those metaethical propensities to propensities toward free will positions: for the egoist, optimism yields (a propensity towards) libertarianism and pessimism yields fatalism; for the deontologist, optimism yields compatibilism and pessimism yields fatalism; for the virtue ethicist, optimism and pessimism yield compatibilism; and for the utilitarian, optimism yields compatibilism and pessimism yields fatalism.

Much empirical work would need to be done to test these associations (I've certainly got some of them wrong, because I can only look at things through the lens of an Idealist), and optimism versus pessimism may strongly correlate with extraversion. Moreover, I tend to break down the propensities further than the high level positions and into their subtypes, but when we get down to that left of empirical prediction, there are bound to be a lot of errors in the absence of actual empirical data, even if the propensities exist.

Maybe someone will get inspired enough to look into the matter...

[Michael Drake]

These results of course mean that the free will skeptics were right!

[Neil]

It was encountering neurotic compatibilsim that convinced me of the truth of the autobiography idea.

[Neil]

It was encountering neurotic compatibilsim that convinced me of the truth of the autobiography idea.

[Fritz Warfield]

Josh -- you write:

"These studies open up the tantalizing possibility that the free will problem has been so persistent, not because either side is making an error in reasoning, but rather because the two sides truly do start out with quite different intuitions. I would love to hear any thoughts you might have about whether this possibility is actual."

But hasn't the record of the philosophical work on both (major) sides of the incompatibilism debate shown that most participants on both sides of the debate do make repeated mistakes in reasoning? This wouldn't rule out the possibility that there are different starting points that take people to different conclusions but it surely suggests at least one alternative explanation to the one you're pointing towards.

[john dell]

Hi Josh,

Reading your post reminds me of the Social Intuitionist Model developed by Haidt and Borklund (see e.g. Moral Psychology vol. 2; ed. Sinnott-Armstrong). Let's say something like this model is correct and much of the moral reasoning people engage in is a post hoc effort at justifying intuitions.

It seems possible then, that under similar cognitive circumstances people could have various personality dispositions that give rise to varying intuitions about free will (or lack there of). And so, much of the debate is structured around attempts at justifying these intuitions. Thus, the driving force behind the debate is a difference in starting point (i.e. personality).

[please take extra note to the claim that *it seems possible*; that is all I am endorsing here]

[Mark Smeltzer]

Although I tend to endorse the idea of a range of starting points, I fail to see why that would have any bearing on whether there is a truth to be sought.

The thing we ought to look for is the commonality amongst the starting points, if they even exist.

[doris]

Personality trait? Extraversion?

[john]

Mark: Given your post above (the one after Vargas's) I am a little confused as to why you would not consider "starting points" relevant. [maybe I was not clear?] By starting point I mean a personality disposition (or propensity if you prefer) that gives rise to certain intuitions about various subject matter (or situations). For example, as you point out above "optimism yields (a propensity towards) liberalism". My suggestion was *it seems possible* that the same sort of cognitive dynamic could exist with respect to positions taken on the free will debate.

Now, why does this have any bearing on whether there is a truth to be sought?

By 'truth', it is not clear to me whether you are speaking in the fashion of cognitivism or non-cognitivism (which I think is going to matter here). But regardless, I think it is possible that the above phenomenon has bearing on the "truth" of the matter in at least one important way.

If the type of theory we endorse is ultimately derivative of the type of person we are, knowledge of this will be extremely important in understanding how to proceed. It might even provide better control over the reasoning process and its outcomes. Let's say for example you know you are an optimistic person and that having this propensity leads you to endorse more liberal views. If there is a truth about the best political position (this is probably a big IF), then doesn't it seem like knowing how you are going to be generally disposed gives you the opportunity to take a more objective stance on the issue in question? This definitely seems possible to me.

Returning back to your question, understanding how starting points effect reasoning may not bear *directly* on the truth about free will. But, it does seem like understanding this aspect of ourselves can remove *a barrier* to finding the truth (whatever that is). So, at the very least, it appears what we are talking about has *indirect* bearing on the truth.

But again, I think relevance here is going to depend on what you mean by truth.

[Situations?] :P

[Tamler Sommers]

Fritz and Josh,

Here's what seems to me a plausible way of looking at this issue.

The persistent disagreement in the free will debate is a result of BOTH errors in reasoning and differences in starting intuitions due to factors like temperament (and, I would emphasize, one's cultural environment). Studies like these don't make it likely that errors in reasoning have played little or no role in the long standing disagreement over this issue. Rather, they suggest that even one day when the errors in reasoning are corrected, we'll STILL have persistent disagreement about issues relating to free will and moral responsibility. Becoming more rational may very well change the way we look at free will and MR. But it won't make the problem go away.

This may be another way of stating Josh's original point, but I think there's a distinction in there somewhere...

[john]

Tamler: I don't want to suggest that what you have posted sounds implausible, but it is not clear to me how you parse 'errors in reasoning' and 'differences in starting intuitions'. These things are not necessarily distinct entities.

As I understand it, part of what is at issue here is that there are certain automatic (or unconscious) processes that effect the way we think about various problems (e.g. free will).

Trying to justify an unconscious intuition is something that can be done in a very rational way (of course it can also lead to errors in reasoning). But, if these automatic intuitions are real, then they are not separate from reasoning, they are the driving force in how our reasoning proceeds.

Quite simply, intuitions and reasoning often (maybe not always) go hand in hand.

As I see it, the problem is what to do about these intuitions because of how they impact our reasoning and our judgment. Certainly, becoming aware of them offers the opportunity for a greater level of objective rationality. It is true that this sort of awareness, in and of itself, will not make the problem go away (as you suggest). But I think it is a step in the right direction.

[Tamler Sommers]

John,

I agree with you that intuition and reason often go hand in hand and "are the driving force in how our reasoning proceeds." My point is only that two people with radically different starting intuitions can reach different conclusions without either one committing an error in reasoning. Their reasoning may have been sound, given their starting intuitions. It just led them in different directions.

[Kip Werking]

Tamler,

I pretty much agree with you 100%.

I think the errors+starting-points idea maps nicely onto my cognitive-biases+semantic-ambiguity view, which I've mentioned at the Garden several times.

Semantic ambiguity can result in people just having different starting points: some people think "free will" means X, others think Y. This is not surprising, considering that just about everyone agrees that free will's definition is not very clear (and yet we expend so much time and energy arguing about whether it exists!). There is no clear error, or right or wrong, here.

But I agree that both forces are at work. There are also errors in reasoning involved. That's what initially attracted to me to the subject, and that's what keeps it interesting.

Of course, errors in reasoning are things that we expect Ph.d-types to fix, if they have a few centuries to work on it. If there are errors in reasoning, we wonder why they haven't been positively refuted, the way that Copernicus refuted the idea that the earth revolves around the sun.

My ongoing theory is that there are two reasons for this. The first is just the semantic ambiguity (or different starting points). The errors in reasoning are mixed in with the different starting points, making it difficult to talk about either.

But, more importantly, the topic of free will raises issues that invoke many cognitive biases. These are moral illusions, like the visual illusions. Their important feature, for our purposes here, is that the errors are hard to see, even if they're pointed out to you.

You can try it yourself. Look at Adelson's check-shadow illusion:

http://www.michaelbach.de/ot/lum_adelsonCheckShadow/index.html

Even after you click "Question 1," and the answer is pointed out to you, it *still* looks as if the two shades of gray are different, even though we know they are the same.

The relevant cognitive illusions seem to have the same feature. The illusion of control, the fundamental attribution error, the mind projection fallacy, victim-blaming or the just world phenomenon, the empathy gap, the illusion of motiveless evil, reactance, etc. All of these may help explain why these errors don't go away.

[Kip Werking]

I'll add: in response to the original post, I agree with Josh's suggestion, and I find the new work by Feltz, Cokely and Nadelhoffer to be fascinating.

The extroversion result seems intuitively right to me. I also conducted a preliminary study with 21 random subjects outside the public library. I found that how people answered "I have always had difficulty making friends," on a 1-10 scale, was significantly correlated with incompatibilism. I also found that a composite score, averaging 10 primary features of Asperger Syndrome, was significantly correlated with incompatibilism (my understanding is that introversion is a common feature of AS).

[john dell]

Tamler: Thanks for the response. I see what you mean now. Like Kip, I agree.

Kip: The analogy you draw with the visual illusions is really fascinating. As you point out, no matter how aware we are that the 2 squares are the same shade or say that the devil's triangle is an impossible object, we still see them a certain way. Donald Hoffman does a nice job describing why we are susceptible to these illusions in his book *Visual Intelligence*(I imagine you know this book).

Prima facie, it certainly seems like this sort of cognitive process could be taking place with respect to our intuitions and moral reasoning. So, by analogy, no matter how much awareness we have that our disposition for optimism is going to manifest itself in liberal tendencies, we will not be able to help but see things in a liberal light. Again, this does seem very plausible.

But, I am not certain that the relationship between intuitions and reasoning about free will is going to be immune to awareness in the same way visual illusions are.

What I have in mind here is based on what Hoffman says about visual illusions. As he sees it, the illusions arise because certain features of the object observed exploits the probabilistic and rule based functioning of our visual system. More simply put, our visual system follows certain rules that we can't help but follow, and this is why we experience visual illusions.

For the analogy to hold up, it seems to me our intuitions would have to work this way as well; i.e. we can't help but follow them. However, I am not so sure this is always true. It does seem like we can change what intuitions we have, and even adjust some features of our personalities (Haidt and Bjorklund give some examples of ways we can change our intuitions).

Nevertheless, you might be right. For me this is a very difficult and interesting question.

[Tamler Sommers]

Kip, yes, really interesting, I'd add one thing to what John said. For the analogy to hold up, you'd have to have a principled way of deciding what counts as an illusion and what doesn't. With the visual illusions, we have an uncontroversial method of doing this. I don't think the same holds for many competing intuitions about moral responsibility. Yes, we have scientific ways of confirming, say, the FAE at times, but what about the intuition (shared by some) that you don't have to have intended to perform an act in order to be blameworthy for it? Is that an illusion? Why? Why couldn't the competing intuition positing a robust control condition for moral responsibility be the illusion? How do we tell?

[Kip Werking]

Tamler,

I am not saying that every alleged moral illusion is provably illusory. I grant that we might not be able to prove that other alleged illusions are actually illusions. The example you mention may be such an alleged illusion.

I would stick by the examples I mentioned:

the illusion of control
wishful thinking (positive outcome bias)
the fundamental attribution error
the empathy gap (projection)
victim blaming (the just world phenomenon)
the inability to understand a hurter's motives (the illusion of spontaneous evil)
reactance

I don't there is much controversy that most of the above are real, actual illusions (granting that one can nitpick any given one). This may be true, even if the magnitude of any given illusion is not that great, and even if they overlap somewhat in their origin.

In response to John Dell, we can talk about the strength of an illusion, on a 1-10 scale. The strength just reflects how resilient the illusion is in the face of contrary information. As we agree, the visual illusions are pretty strong, perhaps 9s or 10s in strength. To answer your question, I agree that the moral illusions might not be as strong as the visual illusions. Perhaps they vary from 3s to 7s. That's an empirical question and I don't really know the answer.

[Edward Cokely]

Hi All. Thanks so much for the thought provoking comments. I'll weigh in more specifically soon (along with Adam). However, first I wanted to ask a question (or two):

1. Is one of the major implications of these data the demonstration that fundamental philosophical intuitions (measured by vignettes) actually DO (or can) tell us something about people's actual experiences and lives?

To simplify, there is real world data showing that measurement of heritable personality traits predicts (or at least is modestly related) to actual life outcomes. For example, personality has been shown to be related to many aspects of people’s lives, experiences, judgments, health, relationships, happiness, etc… Thus, the fact that philosophical intuitions can be related to personality provides at least some evidence that some parts of the arguments about the fundamental importance of basic philosophical intuitions may be valid (i.e. convergent validity), even if others are not.

Is there other (better) evidence that such intuitions (e.g. judgments on X-phil scales) can be related to real life outcomes?

[john dell]

Hello Edward, you are doing some really interesting research.

One thing that I think would be intriguing to look at is how individuals respond in dialogue (or just in group settings) to various thought experiments.

It looks like some of the correlations are getting strong enough where you should be able to predict--based on heritable personality traits--how a person will respond to specific types of thought experiments. From here I think there are some interesting ways to proceed. One possibility is to look at how hearing the responses of others impacts a person's original intuition and response given. For example, A person high in extroversion might consistently assert his interpretation of a given thought experiment (or vignettes), even when he is surrounded by others who dissent (what I have in mind here is a variation of the Asch paradigm). But, a person low in extroversion might initially respond to the vignette one way, however, when he is put in a situation with others high in extroversion responding oppositely, he might not embrace his original intuition and endorse the view of others.

What I am trying to get at is the idea that it is one thing to be of a certain personality type and respond to a question about a vignette in a more private setting. But, it is another thing entirely to maintain that response in a social situation. In other words, there is also an interpersonal aspect to personality that may impact intuitive responses to vignettes. Ultimately, I think an extremely valuable personality based predication would be one that explains why people react the way they do when interpreting philosophy in a social context.

I think understanding the relationship between personality, intuition, and the interpersonal would give a more ecologically valid picture of why people respond differently to various philosophical problems, and might even provide some insight into why (in some cases) philosophical dialogue proceeds as it does.

[Neuroskeptic]

Kip's visual illusions analogy is excellent, and it's one of I've often thought of myself in reference to free will, also moral intuitions, and others.

The "persistance" of visual illusions, even once one becomes aware of them, was pointed out by Jerry Fodor in The Modularity of Mind (1983), although I can't remember off-hand whether he was referring to someone else at that point.

Fodor used visual illusions to make the point that our minds are "modular" in the sense that large swathes of our mental processes are "sealed off" from each other - our faculty of concious reasoning does not have access to the "internal workings" of our visual systems, all we have access to is the "final output" of our visual pathway.

Another good example is language - just try to cast your eyes over these words without actually reading them - you can't do it. As soon as you see (or hear) certain patterns of shape (or sound), your mind interprets them linguistically. You can't just decide to switch that off, and you can't decide that you want to read "Cat" as meaning "dog" for the next five minutes.

My view is that our intuitions about minds and morals (including free will) are just like that. Speaking as someone who would, in an argument, defend determinism to the hilt, I *have* free will intuitions, and the fact that I'm an avowed determinist doesn't change that in the slightest. I think anyone who really didn't have the usual intuitions about the minds of others would be very odd indeed (possibly this occurs in severe autism? - that's something I've wondered for a while. A lot of determinists seem to be somewhere on the austistic spectrum. I'm not aware of any research on that, though.)

My suspicion is that determinists are not so much people who don't have certain intuitions, so much as people who see them *as* intuitions. We're all wearing free-will-tinted spectacles, but some of us know it and some don't. The implication is that people could change their minds about free will, which does seem to happen (it happened to me).

[Kip]

Neuroskeptic, I agree with you 100% about autism-spectrum disorders. I think it's the research of the future.

There are surprisingly few people, in history, who have defended uncompromising no-free-will views (for secular reasons). The biggest names I know are Einstein, Spinoza, Russell and Clarence Darrow.

Einstein and Russell were friends, and both were gushing admirers of Spinoza.

Both Einstein and Spinoza have been diagnosed, posthumously, as having Asperger Syndrome. And anyone who's read Russell's biography knows of the problems he had in his personal life (e.g. the extreme coldness he showed his first wife).

And, while I don't dream that my mind compares to these guys, I have Asperger tendencies myself.

I also ran a 30 person survey, and found that Asperger Syndrome traits were correlated with incompatibilism. I didn't find a link between no-free-will-ism and AS, but this may because of the small sample size---hardly anyone doubts the existence of free will.

A link between AS and beliefs about free will raises fascinating possibilities. Here are just a few (all 100% speculation):

1. As Nichols and Knobe suggest, compatibilism is based on affect-triggered irrationality. Something about AS or autism-spectrum disorders prevents such persons from exhibiting this irrationality. This fits with the generally cold, unemotional, robotic, etc., impression that autistic persons have.

2. Compatibilists regard free will as an important part of a Community of Responsibility. They are, like most neuro-typical people, concerned about upholding the Community of Responsibility and making sure that it continues to function properly. Out of that concern, they find a way to reconcile free will and determinism. AS people, on the other hand, are famously oblivious to social norms and customs, and therefore see no problem with throwing the Community of Responsibility out the window.

3. Humans have specialized, fast, modular, imperfect systems (system 1) for thinking about agents, and generalized, slow, more accurate systems for thinking about abstract ideas (system 2).

Neurotypical people use system 1 to think about other agents. System 1 is fine for coordinating social relations in hunter-gatherer tribes. But it is horrible at producing an accurate picture of how the human mind actually works. System 1 tends to produce the impression that brains are "black boxes" that magically produce indeterministic decisions, immune from the influences of genetics and heredity. System 1, in short, produces the illusion of libertarian free will.

AS types, being impaired in social processing in general, do not use system 1 to think about other people (or use it less). Instead they use system 2. So they think of human brains as predictable, mechanical systems, they way everyone thinks of computers and gears and pulleys. Thus they don't see the illusion of libertarian free will as much.

4. Neurotypical people have such a strong emotional response of empathy for a victim, and anger towards a harmer, that they are incapable, or less capable of feeling empathy for the harmer. They can easily put themselves in a victim's shoes. But they find it very difficult to put themselves in a harmer's shoes.

AS types, because they feel a muted empathy response to victims, are less blinded by their anger towards the harmer, that they can actually empathize with the harmer. That is, they can empathize with both the victim, and the harmer, almost equally. To neurotypical people, this seems perverse, because they have the emotionally healthy/normal response of anger and rage towards the harmer.

[Neuroskeptic]

My thoughts exactly... I tend to favor Explanation 3, I think because I'm generally a fan of dual-process theories, but we don't really know at this stage.

Another possibly ASD philosopher, in my opinion, was Jeremy Bentham - posthumous diagnosis is a fool's errand, but: He never married, he dedicated his entire life to writing (and wrote very prolifically) and little else, he was very eccentric (he had himself embalmed and put on display after his death), he was a child prodigy - it seems likely.

I'm not aware of whether Bentham was a proclaimed determinist - does anyone know? - but it wouldn't surprise me, and it also seems to me that consequentialism could be correlated with ASD. For example, I strongly suspect (just on intuition) that "utilitarian" responding on Trolley Dilemma questions is correlated with ASD traits. Again, does anyone know of any evidence?