This is Chandra Sripada's Typepad Profile.
Join Typepad and start following Chandra Sripada's activity
Join Now!
Already a member? Sign In
Chandra Sripada
Recent Activity
Hi Josh, I wonder if some readers of this post are scratching their heads and thinking, "Why would Josh, a leading advocate of Xphi post *this*? Isn't he undermining the very rationale for doing XPhi?" To those who might be thinking this, I guess it would be useful to underscore that the project of identifying instability in intuitions is really only a small part of Xphi. I saw one place where you (Josh) estimated it was about 1% of studies in XPhi. So the questions you are raising about the instability of philosophical intuitions (or lack thereof) are important and fascinating. But whatever the answers turn out to be, the day-to-day business of doing XPhi should go on just fine.
This is a really interesting paper. Thanks to Kim and Yuan for their terrific work and to you, Josh, for sharing it here. Let us assume these intuitions are universal. What could explain this? Kim and Yuan distinguish nativist and social explanations (i.e., culture). I lean towards nativism. Here’s why: There is another aspect of these intuitions that wasn’t highlighted in the paper that is relevant: their remarkable sophistication. These intuitions track very subtle and non-obvious features: the difference between a Gettiered and non-Gettiered agent, probabilistic versus perceptual evidence, means versus foreseen but unintended effects (i.e., side-effects). The processes and principles that produce these intuitions thus reflect a very sophisticated psychological competence. The socio-cultural explanation thus has to propose that there are these highly abstract and sophisticated competences across cultural groups that arose more or less independently but, nonetheless, these competences ended up all being highly similar. To the extent this seems unlikely, this pushes us in the direction of nativism. There are also Chomsky/John Mikhail-style arguments about learnability, poverty of the stimulus, etc. in the vicinity here. But I will stop here.
Thanks Paul for reminding us that cog sci might very well vindicate the folk view (as Josh has characterized it). This was option 2 on my list and I kind of did assume that this option was deeply implausible. I tend to take a dim view view of the idea that stochasticity during decision might be harnessed to deliver genuine libertarion freedom (for reasons that many on this blog are familiar with). But then again I am not aware of the details of all the views out here (such as Peter Tse's). Would be delicious if the pointy heads all got it wrong and the ordinary shmoe was all along right on!
Thanks Thomas for these elaborations. I especially agree with your point 1 that lays out some important considerations when considering a full on error theory. You asked what would it look like for cognitive science to show that I do not stand above my mental states or have an immaterial soul. I agree it would be hard to definitively show *that*. But I had in mind something weaker. We have a pretty good handle of what happens when someone resists an urge in terms of specific computational/psychological processes: reasoning processes, evaluative inputs, inhibitory control circuits, etc. And these seem quite sufficient to fully explain why a person resists when they do. Given we have the makings of a complete and detailed materialist explanation that fully captures the data to be explained, the postulation of something immaterial that stands above and intervenes seems otiose. So that is the sense in which our best view from cog sci points to falsity of the folk contracausal view.
Hi Josh, I've been following all these fascinating posts, but because of travel, haven't had a good chance to chime in. But just wanted to let you know I am really enjoying seeing your research distilled into these juicy pieces so we can see the connections. Let's suppose that you are right and people think the self stands outside the causal order -- it gets to ratify or resist motives without itself being subject to determination. But surely you also agree that this idea of a contra-causal "willing" self is likely just plain wrong, or at least the folk impression that it exists can be readily explained away. Our best cog sci tells us that when a person resists wayward motives, there are in fact a slew of motivational/decisional factors that cause and fully explain why they resisted. These factors, however, operate largely outside conscious awareness. So it is not at all surprising that ordinary people "mythologize" a bit -- they think there is some spooky contra-causal source of their willings when there is none. I am curious to know what you think about the tension between the folk view you are uncovering through your experiments and the actual findings from cognitive science. Do you think, for example,: 1) the folk view should be abandoned; 2) cog sci might yet vindicate the folk view; 3) it doesn't matter because probing the folk view is interesting in its own right; 4) something else??
The difference between the Leiter 11 and the remaining programs appears statistically significant for women (p=.03) though not for non-whites (p=.13). I used permutation methods as the data appear unlikely to conform to distributional assumptions for parametric tests (sample 11 of the 97 programs randomly, calculate proportion of women/minorities in the 11, repeat 10,000 times to get null distributions for the proportions). Despite the statistically significant difference, however, the actual magnitude of the difference is quite small. And there is striking heterogeneity amongst programs as David points out. Given the heterogeneity, a good use of this data is to motivate folks to investigate what a few programs (Az State, Binghamton, U. Memphis, Georgetown) are doing right.
Thanks Mark for taking the time to read and respond to “fan mail” (I am a fan!) In Slaves of the Passions you defend a Humean analysis of reasons that goes (very, very roughly) like this: (1) a state of affairs R is a reason for Aing iff R helps to explain why Aing promotes one’s desires And in chapter 8, you say something like the following at least initially seems plausible as part of the analysis of desire: (2) a mental state D is a desire iff, among many other things, D disposes S to take certain states of affairs (namely, those that help to explain why Aing promotes D) as a reason for Aing It appears that (1) and (2) are circular, and indeed that’s what you claim on page 155. But the issue of circularity isn’t probed in any detail, so we don’t hear detailed arguments about the nature of the circularity and whether it is vicious or benign. I want to press you on this issue. I don’t think there is really any vicious circularity going on. Things only appear a little circular because DESIRE and REASON are *interdependent* notions. But interdependence and vicious circularity are two different things. Consider the notion of the LAW and an OUTLAW. The law has many functions, one of which is to specify punishments for outlaws. An outlaw is a person whose actions stand in the violate relation to the rules specified in the law. Each notion uses the other in its analysis, but there is no vicious circularity going on here. Rather, law and outlaw exhibit healthy interdependence. (I could offer a detailed argument for this, but I am hoping it is intuitively clear that LAW and OUTLAW are not viciously circular notions.) DESIRE and REASON stand in a broadly analogous relation. From a design perspective, it makes sense that among the sequelae of desires (dispositions to act, attend, and feel certain ways) there would be one further disposition: the disposition to recognize those states of affairs relevant to promoting the satisfaction of the desire as being in some way special. These special states of affairs are of course what are picked out by the term “reason” (so says the Humean analysis). So reason appears in the analysis desire and desire appears in the analysis of reason. But I don’t see any vicious circularity here. Rather like LAW and OUTLAW, DESIRE and REASON appear to be only healthily interdependent. I am curious to hear your take on the plausibility of this line of argument.
Toggle Commented Jun 11, 2013 on Open for Questions...? at PEA Soup
Neil, this is a thought-provoking post. My question to you is what role is evolution playing in your argument. Might the argument go through without invoking evolution at all? To clarify, let me state a bit more formally the argument I see you articulating. (1) The reactive attitudes evolved to perform some fitness-promoting function F, namely, supporting social norms and cooperation. (2) Due to differences between current and ancestral environments, the reactive attitudes no longer perform F. Indeed the reactive attitudes are counterproductive to F and we could achieve F more readily by abandoning the reactive attitudes. (3) Therefore, we should abandon the reactive attitudes. I think the transition from (2) to (3) is plausible only if you insert a premise along the following lines: (2.5) We care (or should care) about achieving F (i.e., supporting social norms and cooperation). But once (2.5) is made explicit, can’t you just drop premise 1? In other words, evolutionary observations about what functions out attitudes evolved to perform don't seem to be doing the work in your argument.
Hi Tim, I’d like to chime on a topic related to the very illuminating exchange between you and David. I like the new view you are putting forward: Blame and gratitude are appropriate as responses to conduct that *indicates* something problematic about one’s attitudes that carries interpersonal significance. If one’s conduct fails to indicate anything (such as my forgetting a birthday due *exclusively* to failures of short-term memory), then blame responses would not be appropriate. Indication is, you suggest, an *epistemic relation*. You talk about “evidential” links between attitudes, and what information they “reveal”. I have a worry, however, that epistemic/evidential links can’t do the work that this view needs it to do. Here is an example that brings out my worry. Consider Frankfurt’s Unwilling Addict who rails against but is ultimately defeated by an irresistible desire for the narcotic. To make the example go better, let’s us assume there is really is something morally problematic with using narcotics (or substitute something genuinely morally wrong for “using narcotics”). In assessing blame in this case, one might reason this way: What one does due to an irresistible desire can’t provide evidence about there being something problematic about one’s underlying attitudes. Since the addict’s using the narcotic fails to indicate anything about his relevant underlying attitudes, blame responses are not appropriate. So far, so good. But now consider Frankfurt’s Willing Addict who has the exact same irresistible desire, but he loves his addiction and is thoroughly identified with it. I, along with Frankfurt and others, think a blame response would be appropriate for such an individual when he uses the narcotic (you appear to agree in WWOE, ch.6). But we just said acting on an irresistible desire can’t indicate anything about one’s underlying attitudes. Since the indication relation fails, we can’t make sense of the appropriateness of the blame response in this case. The problem, as I see it, is that the indication relation is only an imperfect guide to something else, and it is this something else that really matters for justifying the blame response. Here is my quick take on what this something else is: With regard to blame, the key issue with the Unwilling Addict is not that his actions fail to indicate anything about his relevant attitudes, but rather the other way around—his relevant attitudes fail to be *expressed* in his actions (where expression is construed as a species of semantic, i.e., content-based connection). He stands against using the narcotic, and what he does contradicts the content of where he stands. In contrast, the content of the Willing Addict’s relevant attitudes *are* expressed in his actions. It is the obtaining of this semantic expression relation, rather than the epistemic indication relation, that determines whether blame responses are appropriate. My question to you, Tim, is what do you make of these addict cases, and my broader argument. Are you committed to indication as the key relation? Or are you open to seeing indication as an indirect, and perhaps inexact, way of talking about certain semantic/content-based connections between attitudes? (Sorry for the length of this, and let me echo David’s apology—sorry to Pea Souper’s if all this is too “inside baseball”.)
Toggle Commented May 30, 2013 on Featured Philosopher: T.M. Scanlon at PEA Soup
Michael, another really awesome post!! I am really intrigued by the problem you highlight regarding RR views--that they do not pay enough attention to our internal psychic structures. I love the way you put it here: “My relation to my own internal states and the problems of agency posed by my complex nature as a person is not like my relation to the reasons afforded to me by the world as I find it.” I think an RR theorist can bridge some of this gap by offering a more detailed theory of reasons. I favor so-called Humean “attitude-dependent” accounts of reasons that say you have a reason to phi if phi-ing advances some element of your motivational set, e.g. your desires. Externalists about reasons respond with cases in which a person desires something crazy (e.g., the case from Allan Gibbard of Polly who desires to have a trimmer figure even if it risks death), and point out we don’t think the person as a reason to perform these crazy actions. A standard Humean response is to refine their theory to say it is not just any old element of one’s motivational set that grounds a reason, but rather some privileged subset. Sharon Street for example calls this subset one’s “values”, and I have an account that I am writing up says this subset consists of one’s cares. Obviously, this dialectic has interesting parallels with debates about deep self theories of moral responsibility. Susan Wolf famously makes a similar a demarcation move saying you are morally responsible not when your action flows from any old desire (think of OCD and kleptomania) but only when your action flows from a privileged subset of attitudes, which she calls one’s Self. Suppose an RR theorist adopts a Street-style Humean theory of reasons. Wouldn’t this collapse some of the differences between RR theories and mesh theories? In virtue of responding to Streetian reasons, a person would in effect be responding to the attitudes of her Self (since these reasons are grounded in expressions of her self). This is one way for the RR theorist to bring in some of the complex relations between one’s internal psychological states into the RR framework. To be clear, I am a partisan of the deep self view of MR. But I do think the RR theorist is getting a bum rap here so I wanted to offer this as a way to help.
Randy, I think C is true in Mele’s Diana story. But as Patrick’s paper makes clear, there is perfectly good *compatibilist* explanation for this: Given that “wicked and conniving” Diana desired, intended, and brought about that Ernie does A, we are entitled to assume that Diana possesses attitudes towards Ernie’s doing A (for example, the attitude of endorsement) that makes her blaming Ernie *hypocritical*. As Patrick points out in the paper, this is a perfectly good explanation for why C is true that offers no help to the incompatibilist. In order for Patrick’s incompatibilist argument to work, even though 1) the Designer desires, intends, foresees, and brings about that S does A; it must nonetheless be the case that 2) the Designer lacks attitudes that make a charge of hypocrisy appropriate if she blames S for doing A (since the charge of hypocrisy is a compatibilist-friendly explanation for the truth of C). This is where the tension comes in; 1 seems very much at odds with 2. Patrick does an excellent job arguing that 1 & 2 are co-possible, but I have reservations and various complaints. I don’t want to turn this comment into a discussion of Patrick’s Moral Standing argument (though that would be a great topic for another post). The point is that it is plausible that the Moral Standing argument does involve certain significant internal tensions. And, again, the charge being pressed by us manipulation skeptics is that when cases involve stipulations that are in significant tension, reliability of intuitions is degraded. Looks like this thread is winding down. Thanks to Michael for insightful comments and remarkably deep engagement with a large and unruly bunch! Much much love from all of us!
Randy, I am curious to know why Patrick’s Moral Standing argument (which I agree is an absolutely superb argument) stays at the “object level”, where we evaluate the truth or falsity of p, and avoids the need to go to the “meta-level”, where we evaluate the reliability of judgments that p. The charge being pressed against manipulation cases is that they make stipulations that are importantly in tension (conceptual, nomological, or some other form of tension), and thus make for unreliable intuiting. Patrick’s argument relies on an intuition we have about IDA cases--that the manipulator cannot justifiably blame the agents they manipulate. But IDA cases are cut from similar cloth as manipulation cases, so it seems that the very same “stipulations are in tension” charge would appropriately apply to Patrick’s IDA case, and so we are off now at the meta-level. Actually, the “stipulations are in tension” charge has, if anything, *more* bite against Patrick’s specific IDA case. We normally take someone who desired, foresaw, intended, and brought about that S does A to have an attitudinal structure that makes blaming S for A an act of bad faith (in the sense of hypocrisy). But for Patrick’s argument to work, we need to read the IDA case as taking place in the fairly exotic circumstances in which this link does not hold (that anyhow is my understanding and if I am wrong, I should be corrected). So if anything, Patrick’s case exhibits still more stipulations that are in tension than standard manipulation or IDA cases, thus warranting even more so the detour into the meta-level.
Derk writes: “A philosophical example may feature a stipulation that is at odds with other features of the example, and this may make it hard or impossible to evaluate in a way that is illuminating.” I think this is an important point. Many, many compatibilists (including me) have suggested that it is hard not to see a manipulated agent as volitionally damaged in some way. My version of this kind of argument says that we can’t help but see manipulation as producing discordance between the agent’s action and his deepest self. If so, stipulations that the agent is manipulated and at the same time his action is expressive of his deepest self are at odds, creating poor conditions for reliable or “illuminating” intuiting. But Derk sees another, different internal tension. He suggests that we (or at least many ordinary people) naturally understand genuine expression of one’s self as involving a kind of indeterministic origination. For example, Frankfurtian higher-order endorsement may covertly appeal to (or be seen to appeal to) a self that stands above one’s various motives and chooses among them, where this self is not itself subject to causal/motivational determination. If this is right, then once again stipulations that the agent is manipulated and at the same time his action is expressive of his deepest self are at odds, though this time for very different reasons. My and other compatibilists’ version of the “stipulations are in tension” argument is well known and compatibilist-friendly. Derk’s is new (at least it is new to me) and incompatibilist-friendly. One response to this development is to say the two kinds of arguments—one compatibilist-friendly and the other incompatibilist-friendly—cancel each other out. No harm, no foul. Another response is to say that Derk’s argument is just more grist for the mill for the view that manipulation arguments are problematic—now there are two completely different ways in which they exhibit significant internal tensions and thus make for unreliable intuiting. I favor the latter response. I am not sure about Derk’s or others’ take.
Michael: You write, “I think it only fair in this debate to give Pereboom and others a little here. Isn't there *some* pull toward a judgment of non-responsibility in the "extreme" cases? And isn't Pereboom correct that they make vivid hidden causes?” I think it may depend on what the audience is. To a lay audience, I do think manipulation cases help to bring out certain potentially problematic implications of determinism, which is why I do use them when teaching undergraduates. I don’t know how else to get them to “see” certain points in a vivid way. To a philosophically sophisticated audience (folks who have thought carefully and reflectively about free will, determinism, etc.), I have to say I don’t really see how manipulation cases clarify or reveal hidden causes/issues over and above what we can expect these philosophers to already have thought about and carefully considered. If you or others want to take this on, it would be helpful to hear exactly what manipulation cases help to reveal or make clear that was not already made clear in standard formulations of determinism, and standard source incompatibilist or skeptical arguments (such as the Basic Argument) that don’t involve the idea of manipulation.
Robert: It seems I am more skeptical than you that we are *authoritative* about what we are responding to when we have an intuition. Here is what I mean. A paradigmatic feature of an intuition is that while the target intuited proposition seems true, the feature(s) of the case that justify the intuitive judgment are not themselves apparent or even introspectable. For example, presented with a Gettier case, I spontaneously judge the agent lacks knowledge. The feature of the case that justifies this judgment is not itself consciously available. Indeed this must be true to make sense of the fact that epistemologists cleverly construct hypothetical cases to try to figure out whether the relevant intuition is a response to tracking, safety, etc… Call this attribute of intuitions “Feature Opacity”. If intuitions are indeed feature opaque, then we should have more humility about saying that we know exactly what we are responding to when we have an intuition. It may very well be that when presented with an emotionally provocative scenario such as a manipulation case or my silly stick-in-the-eye case, we automatically (i.e., unconsciously/subpersonally/mandatorily/etc.) impute features into the case that are contrary to the case stipulation, and because of Feature Opacity, we are not aware that it is these contrary-to-stipulation features we are responding to. I just want to emphasize that none of this implies the method of cases should be abandoned (I love the armchair!). It is just that we have to have more humility and recognize we are not always authoritative when we believe we have mentally constructed the hypothetical case in exactly the way stipulated and are responding to certain features of the case and not others.
Consider the following philosophical exchange. X proposes a hedonic theory of well-being. Y proposes the following counterexample: A man experiences a calm and happy feeling when he violently shoves a long jagged stick through his eyes. Y triumphantly exclaims, “In response to this case, clearly we have the intuition that shoving a jagged stick through his eyes does NOT contribute to this man’s well-being. Therefore X’s theory of well-being is false.” I think the case proposed by Y is a bad one, i.e., it has little probative value. It is not because his case is bizarre (it is bizarre, but that is not why it is a bad case). Rather, the case works only because it leverages our ordinary and familiar abhorrence of ramming a jagged stick through one’s eye to generate Y’s favored intuitive response. More specifically, we have a very hard time envisioning Y’s case as stipulated because of deeply ingrained tendencies. We automatically mentally “fill in” that shoving the stick through his eye will cause the man excruciating pain. Of course Y can always say that he has *stipulated* that the man will have a calm happy feeling when he rams a jagged stick through his eye. And in deference to Y, we can work REALLY HARD to try to envision the case in just the way that Y intends. But keep in mind that intuitions are fast, automatic, *spontaneous* intellectual seemings. If you have to carefully and effortfully construct a mental scenario and work really hard to keep in mind layers of stipulations, it is doubtful that this is the kind of thing that yields *spontaneous* seemings of any sort. The stick-in-the-eye case helps us see what goes wrong with manipulation cases. It is not bizarreness that is the problem. Rather, it is that these cases require us to reverse deeply ingrained thought patterns in order to envision them properly. Sure we can stipulate that a victim of comprehensive manipulation acts from a reasons-responsive mechanism, his action is expressive of his deepest self, he is normatively competent, and so on… But it takes great diligence to fight layers of ingrained thought and spontaneous inference patterns in order to envision manipulation cases in just this way. At the end, I don’t see that there is anything remotely like an intuition—a spontaneous intellectual seeming—to be had after all that effort. There is only a puzzling, complex, and hard to imagine scenario that incompatibilists and compatibilists will make different considered judgments about due to antecedently held theory.
Michael, you suggest that we should put less credence in intuitions when they are responses to “bizarre cases”. But this can’t be right. Galileo imagined planes with no friction, were of infinite length, and so on. Pretty bizarre stuff, or at least far removed from ordinary experience. But his thought experiments, though bizarre, are immensely clarifying. The intuitions they elicit are deep and important, and established important principles in physics. The problem with Manipulation Cases is that not that they are too bizarre. Kind of just the opposite. The phenomenon of manipulation is all too ordinary and ubiquitous and it is something we genuinely hate, and for good reason. That’s why when presented with the text of a standard manipulation case, we have a REALLY difficult time imaging the case as intended. We import our ordinary reaction to everyday varieties of manipulation into the case. The result is the world famous “manipulation intuition”—manna for the incompatibilist. I agree with you that the way to deal with manipulation cases is work really hard to make manipulation cases clearer, and make possession of the relevant agential factors really salient (i.e., your PPR 2008 paper). But the fact that we have to do this—and more specifically the fact that we have to work REALLY HARD to do this— illustrates that Manipulation Cases are unreliable not because they too bizarre, but rather because they are too close to ordinary, real life cases in a way that distorts our intuitions.
@Alan: I see why you think that Bonnie has better quality of will. After all, she is unconflicted in wanting to give. But I was trying to set the case up as one in which the degree to which they care for the good outcome is exactly equal (I stipulated that this is the case). I suppose the case may be slightly internally inconsistent, for the reasons you raise. @Michael: I am indeed trying to offer a ‘single’ factor quality of will account of praise/blameworthiness. And the case you raised in which the quality of wills of two agents are stipulated to be identical is just the kind that will cause a single factor theory like mine to stumble. Thanks a lot :) But here is my suggestion to deal with your case: We need to be very careful to count the number of praiseworthy actions an agent performs when making comparisons between agents, as such comparisons are problematic if the number is not matched. The non-depressed person does one action: she helps the person in need. The depressed person does two actions: 1) she exerts great mental effort in fighting off the fatigue, ennui, and pessimistic thinking that arise from her depression; 2) she helps the person in need. Consider the following principle: (N) A person who does two very praiseworthy things (both of which express excellent quality of will) merits more praise than a person who does one praiseworthy thing. Here is a case that illustrates that (N) is very plausible: Suppose you and I walk to the department on different paths. I encounter a young girl and help her across the street. You also encounter a young girl and help her across the street. In addition, you encounter a kitten in a tree and help it get down. I didn’t help a kitten because there was none on the path I was on. Were I to have encountered a kitten, I would have helped it. But this counterfactual doesn’t matter. You merit more praise because you have done more good deeds. So (N) seems pretty plausible. If (N) is true, then the depressed person merits more praise because she has done two praiseworthy actions (resistance & helping) both of which express excellent quality of will, while the non-depressed person has done just one. This is one way for a quality of will theory to ‘save the appearances’ in the face of your counter-example. I wonder if folks find this ‘number of actions’ approach plausible?
Great thread! Sorry to join it so late. Here are a couple of thoughts: I favor a quality of will approach to praiseworthiness and blameworthiness: An action is praiseworthy if it expresses positive moral concern (my caring about what morality requires or encourages); it is blameworthy if it expresses negative moral concern (my caring about what morality forbids or discourages). The degree of praiseworthiness/blameworthiness depends on the degree of my moral concern (how strong is the care that is expressed in the action). This is basically Nomy Arpaly’s view, though it’s tweaked a bit. A view like this can make sense of the fact that difficulty can enhance the praiseworthiness of an action, but the link is largely *epistemic*. Doing something good in the face of tremendous difficulty provides strong, albeit defeasible, evidence that the person cares very deeply about the good outcome. Since a will of higher quality is expressed in her action, she is more praiseworthy. Here is a problem for this this view: Bonnie and Ronnie both care equally about some morally good outcome, helping the poor. And both on this basis save 20% of their weekly paycheck for a year to give to Oxfam on Christmas Day. This is a piece of cake for Bonnie. But it is hard as hell for Ronnie because he is a shopaholic and has powerful desires to spend his paycheck on clothes. He has to fight these desires each and every week. It seems that Ronnie is more praiseworthy than Bonnie when he sends that check to Oxfam on Christmas Day. Yet the degree to which they care about bringing about the morally good outcome is stipulated to be equal. The quality of will view seems to suggest that they should be equally praiseworthy, as their underlying cares are equally strong. I suspect the reason we praise Ronnie more is because *he has done more*. They are actually both equally praiseworthy for *sending the check to Oxfam on Christmas Day*. But in addition, Ronnie has done a large number of other praiseworthy actions that Bonnie hasn’t done—each and every week he has fought off temptation-directed desires to spend the money on clothes (I view resisting temptation as thoroughly actional; I guess others might disagree). The cumulative tally of praise on Ronnie’s ledger is thus higher than on Bonnie’s largely because the cumulative number of actions he has performed is higher.
Good points Mark. I don’t think I’m alone in moving back and forth between de dicto and de re readings for closely related judgments. Here is a case from Michael Zimmerman (his 1997 Ethics paper Moral Responsibility and Ignorance) where he is clearly taking the de re reading for ‘freely’, but de dicto for ‘knowingly’, ‘intentionally’ and ‘trying’ (or this is at least implied). Here is the gist of the case: Perry comes onto the scene and sees Doris crashed into a tree and lying unconscious in her car. Perry thinks the car will explode, and resolves to save Doris by pulling her from the car. The car doesn't explode and there was really no danger of this. Instead, Perry's action paralyzes Doris. Zimmerman says the following about the case -Perry “acted freely in paralyzing Doris” -Perry was ignorant of the risk of paralysis -Perry did not have the intention of paralyzing Doris -Perry was trying to save Doris rather than paralyze her In other words, according to Zimmerman, Perry freely As but does not knowingly or intentionally A, and is not trying to A. This is just the quartet of judgments that Al said was odd. I agree with Al that this is odd. But the odd quartet seems to be shared by others.
Hi Al, It does sound odd that a person can freely do something she does not knowingly, intentionally, or ‘tryingly’ do. I am not so worried about knowingly and freely coming apart because that is just the question under consideration (whether one can fail an epistemic condition for MR and still fulfill a freedom condition). As for the other two dissociations, maybe they aren’t so problematic when we reflect on other cases where they dissociate. Freely/intentionally: Knobe’s Chairman freely helps the environment, but does not do so intentionally. A man playing a dice game freely rolls a double six but does not do so intentionally. Freely/tryingly: A woman knows a procedure will save her life but put her baby at risk. She chooses the procedure. She freely puts the baby at risk, but does not try to put the baby at risk. In the end, I think you might be right that in Immunization the Mom does not freely give her child a harmful vaccine. But intuitions are not clear-cut here, and the costs for letting freely dissociate from intentionally/tryingly may not after all be that high. I guess I am on the fence.
@Al: My answers are also true for all the first members and false for all the second members except this one (probably the one that matters most): Mom freely gave the child a vaccine. Mom freely gave the child a harmful vaccine. I say true to the first, but I am having trouble with the second. This is one where subtle changes in formulation matters. Consider: Mom was free in giving the child a vaccine Mom was free in giving the child a harmful vaccine. I say true to the first and lean towards true for the second. I think it would be helpful to check the negations as well: Mom was unfree in giving the child a vaccine Mom was unfree in giving the child a harmful vaccine. I say false to the first, and I lean towards false to the second.
I was operating on the Fischer and Ravizza account of control as a starting position. I just read Al's paper from Phil Explorations where he is questioning the F&R account, and where he addresses just the question I ended with. So I withdraw that question. That paper really nicely sets out the background for Al's question if folks want to check it out. @Neil, your surgery case involves knowing how. I propose we stick with knowing that cases, before we go to the more challenging knowing how cases. I know from chapter 5 of Hard Luck (an absolutely superb book by the way) that you put epistemic conditions within the control condition. That may be right, but I have my doubts. While it doesn't sound quite right to say that I do have control over the fact that the kitten is run over, it doesn't sound quite right either to say that I lack control over the fact that the kitten is run over. Control, by my lights, seems orthogonal to the epistemic condition for MR. If you doubt this, reflect on Philip's cases (especially Immunization) that are structurally similar to kitten but even better intuition-elicitors. Does Mom lack control over whether she gives her child harmful vaccines?
Angel, that's a low blow quoting my own words. :) I don't want to digress from Herman's book so I will keep this brief. I am not a skeptic about folk competence and I do think it would be bad for a philosophical theory if it made appeals to common sense that turned out, on careful study, to diverge from folk opinion. But I think that almost never happens, because from the armchair we have a pretty reliable grasp of folk opinion. That is, I think Billy, Anna, and David are pretty much spot on here: www-personal.umich.edu/~dunaway/TheFolkProbablyDo.pdf XPhi can have a lot of very useful missions. But by my lights, testing philosophers' claims about common sense is not one of them.
Why do we need to look at such exotic cases? Consider standard cases such as the gruesome classic: A tiny kitten has been sleeping behind my car tire, and I am non-culpably ignorant of this. I back my car out of the driveway and run over the kitten. It is clear I am not responsible for running over the kitten. But it appears to be true that I am in control of running over the kitten. If this sounds odd, keep in mind standard accounts of control. For example, my running over the kitten issues from a reasons responsive mechanism (such that, for example, in at least one world in which there is reason to do otherwise, I recognize this and do otherwise). Aren't standard cases like this all we need to in order to separate satisfaction of epistemic versus control conditions for moral responsibility? Those that disagree need to supply an alternative, precise notion of control.