Enchanting Causes

In Gail Carson Levine's fairy tale Ella Enchanted, the  imprudent fairy Lucinda casts a spell on Ella which causes her to ‘always be obedient’, her behaviour triggered by ‘commands’ which she literally cannot resist ‘obeying’.  Here is how Ella describes her own predicament in the novel:

 Anyone could control me with an order. It had to be a direct command, such as “Put on a shawl,” or “You must go to bed now.”…against an order I was powerless…I could never hold out for long.

 Levine’s idiosyncratic use of the word ‘obey’ aside, there is no contradiction in the scenario put forward. The tale describes a possible world in which one’s behaviour can be manipulated in just such a way. The nature of the curse entails that Ella’s behaviour - which is without question beyond her immediate control – is non-intentional (i.e. neither intentional nor unintentional). Ella’s behaviour is in no way motivated by the curse; she does not reason (consciously or otherwise) that she us under a spell and therefore has reason to obey:  the spell is not a consideration she acts upon and her belief that she is under it plays no part in the behaviour it causes. Rather, the curse causes her to do whatever she is told by paralyzing her will to do otherwise.

   So characterized, the causation in question might be thought to be ‘brute’, perhaps akin to that involved in reflex action or nervous ticks. But that would be quite wrong, for the meaning of any ‘command’ uttered to Ella (typically by one of her wicked half-sisters) plays a clear structural role in the causation of her behaviour. The expressions which cause Ella’s behaviour are effective because of their semantic content, because Ella (who speaks American English) understands what the words used mean. Commands in languages that Ella does not understand are useless: they cannot make her do anything.

   It seems to me that Ella’s actions seem to fit the most popular accounts of intentional action, including those of Donald Davidson and Fred Dretske. According to Davidson’s account of mental causation, intentional actions are caused by a combination of one or more beliefs and pro-attitudes, together identified as primary reasons for the action; on Dretske’s view actions are triggered by physical events (e.g. of perception) but structured by the representational content of what was seen or heard i.e. it is the meaning assigned to things by the agent which explains why certain events triggered the behaviour they did. It is plausible that on either proposal, Ella’s behaviour must count as a case of intentional action, and yet it seems to me that it clearly isn't.

But perhaps this is too quick.  Any thoughts?

Comments

[Eddy Nahmias]

I don't know the story. But it sounds like one reason Ella's commanded actions should not be considered intentional is that she is not acting on *her* desires (or pro-attitudes). The causal structure of these actions may parallel that of her intentional actions (and hence they do not seem to be *un*intentional). But it seems Davidson and Dretske could (if they don't already) stipulate that intentional actions are only those that are caused by one's own desires. Am I missing something?

[V. Alan White]

I'd say that this is a human-platform instance of computer programming, except that the human-platform includes an element of self-consciousness awareness that one is performing the programmed action. Any imperative that is given to Ella (my Mom's name!) is analogous to a programming input that must be well-formed syntactically to activate the cursing program and produce a result congruent with it. The curse of course supervenes on the antecedent syntactic structure of Ella's mind and the semantics involved in that structure as fixed by both prior structuring (learning, maybe a priori endowments) as well as the semantics of self-consciousness. But Ella is simply activated by proper syntactical inputs, whether they come from people or machines ("swipe your card now"). She may understand that she is obeying, but the underlying syntax of input-output is governing her behavior. Her responses are programmed given the proper form of command, though she can understand what's she's doing. A real-life example might be certain forms of OCD, where one can grasp (for example) that one is stealing, and even evaluate it as wrong, but be helpless to prevent the outcome given the stimulus.

[Neil]

I don't see why you are so sure that this isn't intentional action. Do you think it is part of the meaning of the phrase that.someone can't be forced to act intentionally? Suppose the causal path is to make Ella have a very strong urge to phi; the urge rapidly grow in power until until it is her strongest desire. Well that just seems to be an unusual case of a usual experience. (I don't know how similiat the cases are, but there are examples of TMS-induced behaviors that the person reports as chosen).

[Constantine Sandis]

Thanks for these comments, all very helpful. I think there's a notion of desire that means something like 'voluntary', and its clear that Ella doesn't desire to do what she does in this sense. But there's also a notion of desire that just signifies that the action was intentional.
Whether or not Ella desires to act in this sense depends on what one takes the criteria for having this sort of desire to be. If they are behavioural, then the answer seems to be yes,and it also seems that she could also function just as one with such a desire would. On the other hand, it seems that she doesn't intend, set out, or otherwise will to do what she does in any sense. Perhaps the example merely tests our intuitions about what sort of desire makes an action intentional, without settling things either way (?)

[neil]

It seems to me that desires are *never* directly voluntary. Desires are just things we find ourselves with, at a time (of course, we can intentionally alter our desires over time). So Ella just seems like a normal agent to me. She does intend to do what she does. That's the nature of the curse: to bring it about that she intends to phi and phis.

Eddy's point seems well taken as a reason not to think that Ella is responsible for her action (for standard historical reasons). But it doesn't seem to entail that she doesn't act, or doesn't act intentionally. I find myself with a new desire together with a strong urge to act on it. I act. Seems pretty ordinary to me. OCD (by the way Alan) seems to be a real-life phenomenon of what is described here, where was is described is indeed intentional action ("I do the tic", as Leckman and Cohen quote on Tourette's patient).

[Constantine Sandis]

I agree that we either find ourselves with desires (or bring them about indirecly over time). But can we not intend to act (or refrain from acting) upon our desires? Perhaps this is just a matter of further desires, but I should have added that in the book Ella is constantly trying to stop herself from doing what she's told to do; the attempts always fail, which frustrates her.

[neil]

I guess the case is underdescribed. The situation in which I have a first-order desire to phi and a second-order desire that that first order desire not be my will is ordinary enough. It is not implausible that we always act on our strongest desire.

[Chandra Sripada]

Very interesting case! I do not think the fact that the commands arise from outside Ella is what drives our intuitions. If the commands were reflectively endorsed by Ella, so each time a command was issued she endorsed doing what it says, then we *do* say she acts intentionally. So the fact that she resists the commands seems critical to our intuitions coming out the way they do. But why? My take is that since she resists with all her might, the desire that eventually moves Ella to action must not be reflective or expressive of her practical identity. Our intuitions about intentional action appear to be sensitive to this fact.

If this hypothesis is correct, then this is important. Traditional models of intentional action (Davidson, E.J. Lowe) require the agent have certain mental states such as means-end beliefs, desire, foresight, etc… Constantine’s Ella case highlights that having these mental states might not be sufficient. There appears to be an additional ‘aretaic factor’ that our intuitions about intentional action are sensitive to. I call it an ‘aretaic factor’ since the condition that one’s action reflects or expresses one’s underlying evaluative stance closely resembles the key condition for moral responsibility defended by aretaic/attributionist theorists such T.M. Scanlon, A. Smith, and others. I have defended a model of intentional action, called the Deep Self model, that highlights the role of aretaic factors in intentional action. I think that in addition to having a ‘volitionist face’, intentional action also has an ‘aretaic face’. I think this Ella case might be seen to support this kind of view.

[R. Clarke]

Suppose that when desires generate actions in us, they do it this way: the desire generates an executive state that triggers and guides the behavior. We might call this kind of executive state an intention. Now suppose that the spell works this way: it causes Ella to be disposed to come to desire to A when she's commanded to A, and (more specifically) to come to desire so strongly to A that (provided Ella is capable of A-ing) the desire will generate an intention (with appropriate content) that triggers and guides her A-ing.

We could so strengthen the requirements on intentional action, and so weaken the requirements on action for which one is morally responsible, that they come to the same thing. But why would we do that?

[Chandra Sripada]

Randy, I think your account of how the spell works is quite plausible. But that only shows that, where A is the content of some command, an intention to A precedes Ella’s A-ing. But the question of whether Ella A’s *intentionally* (or her A-ing is an intentional action) is different than the question of whether she has the intention to A. At least, if the two are tightly connected, then this has to be established by some argument.

My own view is that an intention to A is neither necessary nor sufficient for A-ing intentionally. Cases like Harman’s Sniper (who alerts his enemy intentionally but does not intend to alert his enemy) and Butler’s dice roller (who intends to roll a six but does not intentionally roll a six) support this view, though I admit there is ample room for debate here.

By saying that our notions of acting intentionally and acting responsibly are similar in that both have an aretaic face, I do not mean to suggest that the two come to the same thing. A drunk driver is morally responsible for killing the pedestrian but does not kill the pedestrian intentionally. The difference is clear. So I readily admit that even if both intentional action and moral responsibility have an aretaic face, the correct account of these concepts will still have to capture the difference in cases like the drunk driver.

[Josh Shepherd]

The case does seem underdescribed. It seems possible, as many have mentioned, for the process to involve an intention. In such scenarios Ella's action seems intentional. But I wonder how Ella experiences these actions? Constantine's 6:12 remarks suggest the following scenario.

Ella is like one of Frankfurt's bystanders to the action. Perhaps Ella is intending all along to do anything but A - she's sitting there, intending B, then C, then D, with increasing desperation - all the while she's A-ing.

The problem with such scenarios is that Ella's action-system seems so compromised that it's no longer clear that the case generalizes to human agents (our action-systems are plausibly never so hijacked).

In order for the case to generalize, it seems the case would need to make the following claim of Constantine's plausible: 'it seems that she doesn't intend, set out, or otherwise will to do what she does in any sense.' If Ella's action-system has been bypassed, then this claim is true, but the case doesn't generalize. If Ella's action-system is not bypassed, but merely coerced, then there is a sense in which she does intend.

[R. Clarke]

Suppose I can't say no to dark chocolate. I can avoid seeking it out, but when it's offered, "I can never hold out for long." It's not as though I behave convulsively. I'm quite adept at breaking it into pieces, eating each one slowly so as to savor them, etc. Am I not acting intentionally when I do so?

[Constantine Sandis]

I agree that the case is underdescribed, but perhaps all that matters is whether there is a possible description of it that would be problematic for certain views to even make sense of. The version I find most interesting is the one in which Ella is trying to resist obeying but is physically caused to do what she is told. I think it is plausible to say that she is not following the rules given to her by her sisters but merely behaving in a way which accords with them. So my intuitions are closest to Chandra and Josh's though I realise that others posting above do not share them. Must read more of this aretaic stuff!

[Chandra Sripada]

Randy, your Dark Chocolate case is well taken. But since it involves a pretty harmless indulgence, it may not bring out the sense in which the person genuinely stands against the desire (and intention) that moves him. Consider a pair of matched cases:

An Unwilling Addict has a desire to use narcotic but hates this desire and constantly struggles against it. The desire to use is too strong and overwhelms him. He takes out the syringe, loads it with the drug, rolls up his sleeve, and injects himself. Even as he injects, he deeply despises what he is doing.

A Willing Addict has a desire to use narcotic just like the Unwilling Addict, but loves his desire. Were it extinguished somehow, he would reinstate it. He takes out the syringe, loads it with the drug, rolls up his sleeve, and injects himself. All the while as he injects, he strongly endorses what he is doing.

I think that the Unwilling Addict does not use the drug intentionally, while the Willing Addict does use the drug intentionally. If we allow that a person can do something intentionally to a greater or lesser degree (which I think is reasonable), then at the very least the degree of intentionality exhibited by the Unwilling Addict’s action is less than the Willing Addict’s action.

What explains the difference? The cases are matched with respect to the desire that moves each to action (i.e., the narcotic-directed desire). And the means/end beliefs, foresight, etc.. are the same. Also I am fine with saying that an intention to use the drug precedes and guides the action – but this too is matched between the two cases. I think the difference between the two cases in our intuitions about intentional action has an aretaic/attributionist source. The fact that the Unwilling Addict’s action strongly conflicts with his underlying evaluative stance, while the Willing Addict’s action clearly expresses his underlying evaluative stance seems to be influencing our judgments of whether he uses the drug intentionally.

[Grant Rozeboom]

Specifying sufficient conditions for an action's being intentional, on the basis of intuitions drawn from various cases, seems to me to be an unreliable way to get at what's going on here, because our intuitions about whether an action is intentional or not are notorious for getting pushed around by factors (seemingly) extraneous to human agency (the "Knobe effect"). I, for one, don't share Chandra's intuition that the Willing Addict acts more intentionally than the Unwilling Addict.

But, if you want to ask whether or not the Ella character exemplifies morally responsible agency, in the sense with which Davidson and Frankfurt and others were concerned, it seems that whether or not the "effect" of the curse leads Ella's behavior to be an output of a process that bypasses, or proceeds through, Ella's "control system" (whatever you think that to be ... desires and beliefs, intentions, deliberation, decisions, etc.; simply grasping the meaning of a command and then behaving in accordance with this command is not sufficient to employ such a control system, on any plausible account -- consider Davidson's worries about "deviant causal chains") is the relevant matter to settle. If it always involves the employment of a control system, then it seems that all the curse has done is make Ella deferential. And deferential people, no less than rebellious ones, are morally responsible for what they do. And if it always "bypasses" the control system, then it is hard to understand how Ella could display any measure of agency in behaving as she does in response to commands. Note that both the bypassing scenario and the control-system-employing scenario are compatible with Ella resisting her obedience to the commands (after all, sometimes we resist what we ourselves have made up our minds to do!).

[Chandra Sripada]

Grant, your point about not getting the intuitions in Unwilling/Willing Addict is well taken. I concede this case does not yield the clearest intuitions, and it seems to depend on what one think happens when a person eventually yields to powerful urges (does it make sense to say that, at the very time they act, they still genuinely stand against the desire that moves them?). I have my own take on this that may contribute to my thinking about whether the Unwilling versus Willing addict acts intentionally. So I agree this is all controversial.

But your point about avoiding relying on intuitions about hypothetical cases altogether, because intuitions may be unreliable, seems too extreme. Whether an intuition about a case represents our competence with the relevant concept an ‘extraneous’ performance error can itself only be established by looking carefully at our intuitions about a whole range of cases, among other sources of evidence, and generating a systematic theory. Only then could one come to know that intuitions about certain cases are driven by factors external to the concept being analyzed. For example, I think the asymmetry that Knobe identifies is part of our competence with the concept of intentional action, while others disagree (for example, some think pragmatics explains the pattern of intuitions). It will take a lot of close examination of our intuitions about a range of cases, among other kinds of evidence, to sort this out.