The "Deep Self" Model

Hi all,

Like many others,  I've contracted a bad case of "Knobe Effect" fever!  I've written a paper on a model, which I call the Deep Self Model, that attempts to explain a variety of asymmetries in intentionality and responsibility judgments using a unified framework.

Quoting from the abstract, "The model proposes that people make an intuitive distinction between two parts of an agent’s psychology, a Deep Self that contains the agent’s stable and central psychological attitudes and a Superficial Self that contains attitudes that are less central and more fleeting".  An intriguing aspect of the model is that the distinction between an agent's Deep Self and Superficial Self is one already made by, or is implicit in the work of, a number of philosophers including Harry Frankfurt, Gary Watson and others.  Thus I hope the Deep Self Model will help anchor the recent asymmetry findings from the experimental literature in an already familiar and well accepted philosophical framework.

I would much welcome any comments and feedback.

-Chandra

Download sripada_the_deep_self_model_and_asymmetries_in_folk_judgments_about_intentionality_and_responsibility.doc

Comments

[Doug]

I think if you change the Bad Shot example to where John is a skilled rifleman (shooting accurately is part of his deep self), but whose hands slips but still accidentally makes the shot, people would still not think he intentionally shot it. That would contradict the Deep Self prediction.

I think there are some issues with the examples. The chairman helping the environment story would never happen in real life. With the 'bad shot' example, it's not really good or bad.

With the other examples, the reducing or increasing of something bad is crystal clear. Stopping Germans from killing innocent civilians, killing your aunt for money, and harming the environment.

Also in the chairman helping story, there is a second person, the one who came up with the idea that would both make money and help the environment. He might be viewed as more responsible for helping the environment than the chairman.

What if the chairman said no to the second person when he suggested a plan that hurt the environment? The chairman would be seen as good in that example.

Doug,

Good point about the Bad Shot example. I think the vignette you have in mind might go something like this:

******
Good Shot

Jake has become good at shooting his rifle through years of practice and carefully honing his skills. He desperately wants to win the rifle contest, and knows that he will only win the contest if he hits the bulls-eye. He raises the rifle, gets the bull’s eye in the sights, and presses the trigger. His hand slips on the barrel of the gun, and the shot goes wild . . . Nonetheless, the bullet lands directly on the bull’s-eye. Jake wins the contest.

Did Jake hit the target intentionally?
*****

I am not sure what people might say about a case like this. I asked a couple of folks who happen to be around the office and they both say he did hit the target intentionally, which is what is predicted by the Deep Self Model. I guess it's an empirical question worth pursuing.

But there is another issue that is lurking in the background. One thing I notice about the Good Shot Case versus the Bad Shot Case is that one’s interpretation of the word “slip” is quite different in the two cases. In the case where Jake isn’t skilled at using a rifle, then “slip” is the kind of movement one would expect Jake to make(the slip is construed as compatible with Jake’s antecedent means-end beliefs about how to move his hands to cause the firing). But in the case where Jake is highly skilled at using a rifle, then his “slip” seems highly unexpected. In fact, it seems so discrepant from his antecedent means-end beliefs about how to move his hands to cause the firing, it is unclear whether what Jake does is even an *action*. If it not an action, then of course it cannot qualify as an intentional action.

-Chandra

[tnadelhoffer]

Chandra,

Thanks for posting your paper. I enjoyed reading it. I think the first thing worth pointing out is that it would be easy enough to test your view. As it stands, however, I think your view fails to explain all of the data about ascriptions of intentional action even if it happens to explain some of Knobe's results. Take, for instance, my Butler Problem examples, Non-Moral Sniper examples, and Nuclear Reactor examples. In most of these examples, the attitude of the agent is kept constant while the morality is manipulated (bad, good, or neutral). It appears that in these cases, the participants normative judgments are doing part of the work. Similar conclusions can be drawn, I think, from the work done by Phelon and Sarkissian as well as Mele and Cushman. And while I have admittedly not gone back and looked at all of the examples that have been used thus far by all of the people working on this issue--which might be a helpful thing for me to work on over the winter break!--I suspect that your deep self view won't be able to explain all (or perhaps even most) of the results thus far. Of course, as far as I can tell, none of our models explain all of the data. So, your view's no worse off in this respect than others.

But I leave looking at all of these other studies up to you. Perhaps I am not remembering them correctly (it is the end of the semester, after all--so, I am a bit drained). For now, I wanted to suggest the easiest way for you to test your view. But first, I wanted to point out that I have always worried about Knobe's use of the "don't care at all" language in his vignettes since it fits much more naturally in the bad/blameworthy conditions than it does the good/praiseworthy conditions. In some sense, your deep self view picks up on this problem as well. The easiest way to test it would be to rerun the CEO studies along the following lines:

"The vice-president of a company went to the chairman of the board and said, ‘We are thinking of starting a new program. It will help us increase profits, but it will also harm the environment.’ Normally the chairman is a good person who cares about the environment. But because today has been uncommonly hectic and stressful, he is in a really foul mood. The chairman just found out that his main competitor recently came up with an amazing new invention that will enable him to corner the market. Consequently, the chairman of the board answered, ‘I don’t care at all about harming the environment. I just want to make as much profit as I can. Let’s start the new program’. They started the new program. Sure enough, the environment was harmed. (2003a: 191)"

You would then use a second vignette that made it clear that the chairman is miserable and bitter S.O.B. who doesn't recycle and thinks global warming is a hoax, etc. I predict that you would get roughly the same responses. But your view predicts that the intentionality rating in the second vignette ought to be higher than the first. For now, I am curious what your intuitions tell you.

My suspicion is that you are distinguishing deep self from superficial self in a way that will allow you to suggest that both of these cases are of a piece with respect to the CEO's deep self. After all, you think that in order for an agent P's x-ing to count as merely reflecting his superficial rather than deep self, P's x-ing needs to be done as the result of not just temporary, unstable and non-central attitudes, but irrational or impulsive emotions/reactions. I take it that's the upshot of the Pizarro cases. But this condition seems too strong. If nothing else, your claim that "most actions performed by a non-impaired agent in ordinary circumstances arise from motivations that belong to the agent's Deep Self." This seems to fly in the face of the gathering literature on situationism. But that's another story for another day. For now, I am curious to see what you think about the modified CEO vignettes I mentioned above.

Thanks Thomas for your very incisive and helpful comments. In this reply, I want to focus on your claim that the Deep Self Model can’t handle many kinds of cases, such as “Butler Problem examples, Non-Moral Sniper examples, and Nuclear Reactor examples.” In the Deep Self paper, I do show that the Deep Self Model explains Knobe’s moral and non-moral sniper cases (i.e. “Bad Shot”, “Rich Aunt” and “Klaus”). So I disagree with you; the Deep Self Model *does* have quite a bit to say about these general sorts of cases.

But most importantly, I want to focus on your claim that “In most of these examples, the attitude of the agent is kept constant while the morality is manipulated (bad, good, or neutral).” Let us call this the “Only a moral difference” claim. I think the key point I am trying to make in the Deep Self paper is that the “Only a moral difference” claim is simply false. Here’s why.

In virtually all of the cases used in the literature, there is a systematic confound that arises because the protagonists in the moral versions of the cases act from psychological attitudes that have greater “depth” than the attitudes in the non-moral versions of the cases (where “depth” refers to the centrality and enduringness of a psychological attitude, as I explain in greater detail in the paper). Let me explain this point by considering two cases drawn from your “The Butler Problem Revisited” paper (Analysis, 2004).

C1
Brown wants to kill Smith now. So, he takes out his six-shooter, places a single bullet in the chamber and spins the chamber. After spinning the chamber, Brown takes careful aim at Smith from a distance of ten feet, pulls the trigger, and shoots Smith directly in the heart in just the way Brown hoped he would. As a result, Smith dies.

C4
Brown is playing a game with a six-shooter and a paper target. The game requires that he place a single bullet in the chamber and spin the chamber before firing. After spinning the chamber, Brown takes careful aim at his target’s bull’s-eye from a distance of ten feet, pulls the trigger, and hits the bull’s-eye dead centre in just the way Brown hoped he would. As a result, Brown wins the game.

In this paper, you conclude:
“Overall, these results strongly suggest that moral considerations do fuel the asymmetry of our intuitions about Butler’s problem… [F]or example, subjects in C1 were more likely to say that Brown shot Smith intentionally than subjects in C4 were to say that Brown shot the bull’s-eye intentionally, *even though all of the other factors in the two cases were identical*." (the stars are mine)

It is this final clause between the stars that troubles me. The Deep Self Model suggests that a key difference between C1 and C4 is that the attitudes that are the source of the action in C1 (the desire to kill Smith, the devaluation of human life) are much more central and enduring (i.e. they have greater “depth”, to use the language of the Deep Self Model) than the attitudes that are the source of the action in C4 (the desire to win at a game). I claim that, other things being equal, subjects judge that an agent’s bringing about an outcome is more intentional in so far as the action(s) that give rise to the outcome arise from psychological attitudes that possess greater depth.

To illustrate that depth of psychological attitudes, rather than moral considerations, plays a role in explaining the difference between C1 and C4, consider a non-moral analog of C1, which I’ll call C*.

C*
Brown desperately wants to be a policeman and he has been training in the academy for years. In order to pass his academy final test, he must shoot a paper target. Specifically, the exam requires that he place a single bullet in the chamber and spin the chamber before firing. After spinning the chamber, Brown takes careful aim at his target’s bull’s-eye from a distance of ten feet, pulls the trigger, and hits the bull’s-eye dead centre in just the way Brown hoped he would. As a result, Brown passes his exam.

This vignette is designed to make sure that the attitudes that give rise to the agent's action have significant "depth" (since the agent has a long-standing desire to be a policeman and it's plausible that this desire is central and enduring). I make two claims: The first is that in response to the question, “Did Brown bring about the outcome intentionally”, more subjects will say “yes” in C* than C4. The second claim is that “yes” responses will be roughly similar for C* and C1, suggesting that it is subjects’ assessment of the depth of underlying psychological attitudes, rather than moral considerations, that explains the difference between C1 and C4. Even if I am wrong about the second claim, if the first claim is true, then it suggests that the widely accepted “Only a moral difference” claim is false.

Now I do realize that in the preceding discussion, I treated the “depth” associated with psychological attitudes as a continuous variable. In the paper I posted, I in effect treated depth as a dichotomous variable since I simply distinguished between attitudes that are part of an agent’s Deep versus Superficial Self. So I am indeed revising my view so as to treat “depth” as a continuous variable (and will be revising the paper shortly), and I thank your very incisive comments for helping me see that this modification was needed.