More on My Way

I’m reviewing My Way for the Philosophical Quarterly. It will probably be a short review, so I won’t have time to do much more than express admiration. Nevertheless, I’m reading the book carefully. I wanted to try out some (rather inchoate) thoughts on fellow gardeners.

How Many Mechanisms?

The question of mechanism individuation is one that has always worried me about the theory. Here’s my latest way of attempting to get at the problem.

Let’s begin with Alison McIntyre’s Insects. The case goes like this:

Bill is offered a plate of fried crickets to eat. He doesn’t find the idea of fried crickets particularly appealing (though he isn’t disgusted by the idea), so he declines. However, had he decided to accept the offer, he would have had to look more closely at the crickets, which would have prompted an overwhelming revulsion to the idea of eating them, rendering him incapable of doing so.

Fischer (and Ravizza) think that Bill is responsible. They hold that the mechanism that Bill acts on, in the actual sequence, is reasons-responsive. Were Bill to accept the offer and look more closely at the crickets, a different mechanism would have been triggered. That mechanism would not be reasons-responsive, but that fact is irrelevant on the actual sequence. Bill’s revulsion is like a Frankfurt intervener, and like such an intervener its presence ought to be bracketed in assessing the reasons-responsiveness of the mechanism upon which he acts.

But now consider Insects*. Suppose that upon being offered the fried crickets, Bill experienced mild revulsion, a revulsion insufficient to render him incapable of accepting the offer. Nevertheless, he declines. Everything else is as before. In Insects* revulsion is part of the actual sequence. Will Fischer still say that he is responsible for turning down the offer, because the mechanism upon which he actually acts is different to the mechanism that would be triggered were he to accept the crickets? In that case, he is committed to saying that there is a mild revulsion mechanism, as well as an overwhelming revulsion mechanism – and the way is open for us to introduce a variety of further revulsion mechanisms (mild, moderate, and so on).

Perhaps Fischer will say there are just two revulsion mechanisms: non-overwhelming and overwhelming revulsion, and so long as Bill acts on non-overwhelming, he is responsible.  There’s an empirical problem with taking this route. Real compulsions are experienced as mounting states of discomfort, which is finally relieved by giving in. If we go the two mechanisms route, we are committed to saying that when people give in too soon to their compulsions, they are responsible for doing so. But when does a compulsion become overwhelming? There’s a strong case for thinking that there is no moment at which this occurs. At very least we can say this: for any actual degree of compulsion, the person could hold out a second longer were she given sufficient incentive to do so. So the compulsion is not yet literally overwhelming. But then it is never overwhelming.

Fischer might say that a compulsion counts as overwhelming just in case it is sufficient, in the actual circumstances, to cause the agent to act upon it. But that comes uncomfortably close to saying that a compulsion is overwhelming when it causes the agent to act – and then we want to know how this kind of causation differs from ordinary determinism.

Suppose, on the other hand, Fischer says that in Insects* Bill is not responsible for refusing the offer, because his revulsion, which would become overwhelming were he to accept the offer, operates in the actual sequence (that is, because he acts upon a mechanism that is not moderately reasons-responsive).  Here’s a challenge to this line. I introduce a variant on the counterfactual intervener here. Counterfactual interveners are generally poised to remove agents’ abilities. My counterfactual intervener is poised to restore them. So suppose Bill experiences mild revulsion at the thought of eating crickets, and refuses the offer. But the counterfactual intervener stands by to ensure that were Bill to accept the offer and look at the crickets, his revulsion would not intensify.

Now, Fischer is committed to holding counterfactual interventions fixed. So he must say that since Bill is not responsible for accepting the offer in the absence of the intervener, he continues to be not responsible once the intervener is added. But this now seems implausible. Since Bill’s revulsion is not overwhelming, and nothing would prevent him from eating the crickets were he to accept the offer, he seems responsible for refusing. Thus this response fails.

Finally, it is worth thinking a bit more about positive versions of the counterfactual intervener. Recall Sharks. In this case, an agent who decides not to rescue a drowning child is not responsible for failing to do so, because (unbeknownst to him) a ring of sharks would have eaten him had he dived into the water. Now suppose that there is a counterfactual intervener, poised to intervene to stun the sharks were the agent to dive in. Fischer seems committed to holding this counterfactual intervention fixed. Intuitively, this seems to me the wrong result.

It proves more difficult to transfer this kind of case into the head of the agent. But the very difficulty provokes further questions. The difficulty is this: we need an answer to the question, what makes it the case that the agent acts from a compulsion in the actual sequence, given that the counterfactual intervener remains poised to make the agent moderately reasons-responsive (ie, to remove the compulsiveness, if the agent makes more of an effort to resist)? The dilemma here is that Fischer will not want to say that what makes it the case is that the desire simply causes the action, because that comes to close to saying that causal determinism rules out moral responsibility.

Comments

[Kip Werking]

Here are some scattered thoughts:

"They hold that the mechanism that Bill acts on, in the actual sequence, is reasons-responsive. Were Bill to accept the offer and look more closely at the crickets, a different mechanism would have been triggered. That mechanism would not be reasons-responsive, but that fact is irrelevant on the actual sequence."

If Bill hate crickets sufficiently, wouldn't he be strongly reasons-responsive (with respect to not eating crickets) not reasons-unresponsive? I suppose we might consider the case in which terrorists tell him to eat crickets or else they will start a thermonuclear war.

This raises an interesting possibility. Certain strong deontological views would hold reasons as sufficient in every world, and defy consequentialist "lesser evil" thought experiments (like the terrorist example above). Would Fischer consider these deontological duties or rights to be sufficient reasons in every world or does his view build into it a certain hostility to such deontological views? And if he can accomodate these views, would we distinguish between less controversial rights/duties like those prohibiting lying or the killing of innocents from, for example, not eating disgusting food?

I agree with your skepticism about sharks, although I'm not sure that your (very clever) thought experiment helps motivate that skepticism. All of this talk of reasons and counter-factual intervenors can get horribly thorny/technical, at the risk of obscuring the underlying point. I think sharks is just wrong to the extent it defends an externalist account of moral responsibility. How can something outside of an agent's awareness factor into whether the agent was morally responsible for the decision s/he made?

As someone who sympathizes with TNR, I find Fischer's critique of TNR, his response to Pereboom (and Strawson/Smilansky in The Cards That Are Dealt You), and also his more recent discussion of the value of acting freely to be more fascinating than, for example, his rejection of PAP and discussion of guidance control. In My Way, Fischer says that acting freely is valuable to the extent that it allows us to creatively express ourselves. It seems to me that talk of creativity in this sense recalls the stronger notion of free action, which Fischer rejects as necessary for moral responsibility, just as much as it recalls his weaker sense---if "metaphysical megalomania" infects the one, it infects the other.

I think we can see this if we can imagine a meta-author programming an author to creatively express himself just as the meta-author wishes. Fischer's response to Pereboom seems to show that Fischer is committed to holding such an agent morally responsible for this creative self-expression (so long as the agent satisfies Fischer's other requirements for mr). But this sort of programmed self-expression does not seem to be valuable in the sense we ordinarily value creative expression. We tend to think of such creative self-expression being un-tracable to another---that the self-expression originates within us and is, in a strong sense, buck-stopping. And this raises the fascinating question, which Manuel is always pressing: to what extent is Fischer's view (and any other compatibilist view) tacitly revisionist?

[Neal]

About the crickets case and mechanism individuation -- I'm not sure that we would have to posit two different "revulsion" mechanisms. After all, even if he experiences mild revulsion in the actual sequence, that doesn't mean the revulsion is what characterizes the actually-operative mechanism. Presumably the mechanism here is just ordinary practical reason. His mild revulsion might be a factor about which he deliberates, but it need not be central to the sort of mechanism at play. On the other hand, in the alternative-sequence when he has an irresisitible desire, the entire mechanism *is* characterized by the revulsion.

So it seems to me that Fischer and Ravizza can plausibly claim (though of course without a fully spelled out account of mechanism individuation, as they admit) that even in your Insects* case, he is responsible because in the actual sequence, he acts on ordinary practical reasoning, which is his own moderately reasons-responsive mechanism.

[Mark Smeltzer]

Neil,

I am not sure that I agree with your assessment of "real world compulsions".

First, when I see other people bleeding, my knees go out, I get light headed, and my stomach turns into knots. I usually have to sit down and catch my breath, and due to this physical impairment (I've tried to get over it, but the beast won't budge) I am absolutely no help in a situation where someone is bleeding and needs help.

Second, I have a strong aversion to carrots, but it is not a revultion. I dislike them, and try to avoid them, but I can will myself to eat them if I feel like it is appropriate to do so. For instance, if I were over to your house for dinner and your wife had prepared salad with carrots, I would smile and chow down, keeping my discomfort to myself.

Third, on occasation when I do eat carrots I have tred onto the dangerous path of contemplating my discomfort. If I start on that process, my stomach will usually start churning and lead to a downward spiral that results in me having to excuse myself. However, as long as I keep my mind off of the fact that I am eating carrots, I am fine.

My intuition is that only the first example is a case of non-reasons-responsive mechanism. In the third case, if I puked all over your dinner table, I should be held responsible because that outcome stems from a reasons-responsive mechanism.

Could you rephrase your worry in context of these examples?

[Neil]

Thanks for the replies, guys. A few thoughts.

Kip,

I too think that we need to think more about the internalism/external distinction, in Fischer's work and in MR more generally. Here's one way of getting at the problem: in Frankfurt-style cases, we typically ask whether the agent possesses the right set of dispositions to be morally responsible, which is an internalist question, but cases like 'Sharks' turn, in part, on actual external sensitivity of the causal path. Of course, the problem with going internalist is that we get cases like Rain Dance, in which someone deliberately refrains from performing said dance in order to make a drought continue. She expresses a bad will, but do we actually want to hold her responsible? Its cases like that that make me hesitate over internalism.

Neal,

That's a good response, I think. On your view, there is 'ordinary practial reasoning', up till the time that revulsion becomes so severe that it overwhelms it and another mechanism replaces it. As I said (and here Mark's objection becomes pertinent) I doubt that we can always identify a moment at which ordinary deliberation is overwhelmed. Moreover, it seems to me that introducing a sudden lurch disrupts what is intuitively a continuum: the greater the revulsion the less the responisibility. I need to think more about this.

Mark,

It seems to me that your cases (I take it they're really autobiographical) are realistic. My description of genuine compulsions is based on a lot of empirical reading, in psychopathology (esp. OCD and Tourette's), in the cognitive psychology of self-control and in neuroscience (if you're interested, see my chapter with Tim Bayne in Sebanz and Prinz (2006) Disorders of Volition). Your second example is textbook: self-control is often achieved through indirect means, like redirecting attention. But these strategies have their limitations, especially in the pathological cases. OCD sufferers eventually give in to their urges if they can't remove themselves from the triggers. The evidence that it can't be resisted forever is overwhelming, and comes in two forms: the price they will pay to avoid the triggers, and the costs to themselves of some of the actions they perform. But - if I'm right about these cases - they're on a continuum with many ordinary losses of self-control. In such cases there are genuine limits on how long self-control can be exerted (how long can you eat carrots for?) On the other hand, you might find that were the situation to demand it, your ability to control your actions in the blood cases would increase.

Hope that helps!

[Mark Smeltzer]

Neil,

I see what you're getting at now. For quite some time now I have held the belief that responsibility comes in degrees, and your discussion of the continuum of compulsion fits well with my view. Fischer's view seems accustomed to the idea that only one mechanism operates in a sequence, but I have little reason to believe that is true.

Although Frankfurt-style cases usually pit mutually exclusive mechanisms against each other (such that only one or the other can operate in the actual sequence), it seems that other mechanisms play a role in the unfolding of normal scenarios in addition to the agent's own MRR mechanism.

I think Fischer would respond that if we could hold the situation fixed, subtract out the interference mechanisms, and leave the agent's MRR mechanism, would the agent still identify with the action it took with the interference mechanisms in place? I do not believe that Fischer has explicitly addressed this kind of case, but I believe that his account does provide him sufficient resources to develop a response.

I am suspicious of the single-acting-mechanism concept, but I have not determined whether Fischer's account actually makes that stipulation.