Blog Coordinator

Knobe's X-Phi Page

X-Phi Grad Programs

« Registration Open: Second Workshop of Experimental Philosophy Group UK | Main | CFA: 2011 MPA "Experimental Philosophy and Its Critics" »

07/19/2011

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

Tamler Sommers

Very cool. We just read the JJT paper in my intro class last week. One student said the cases were disanalogous because the violinist was not related to the woman in any way. She suggested that people might react differently if the violinist were a relative of the woman. And apparently they do!

Marie

I can see some concern with using stats to back up a philosophical question in that it leans towards an ad populum fallacy. If a statistically significant number of people agree on the answer to the question, it's meaningless.

However, I think your explanation about using the results to clarify audience addresses these concerns.

Kitagawa

I am still confused about exactly what this kind of experiment is purported to show or establish. I mean, teach any intro level contemporary moral issue class and you will always have some students who make such "insightful" comments that our intuition will be affected by what kind of the relationship is there between the violinist and the woman and whether the woman is "you" or someone else (well, at least, in my experience. But I am more or less certain than anyone who taught an intro course has had the same kind of responses from a good number of students) Yes, indeed, our intuition about the case does seem affected by all the factors mentioned in the JJT paper. But, so what? The paper only reconfirms that piece of empirical fact, which itself by the way is neither surprisingly new nor contributing to the extant discussion about Thomson's violinist case negatively or positively. (All this might be due to my ignorance, so I reserve some room that my critical comments are off the target, though.)

Alexandra Bradner

Just a note to all of the (very welcome) non-xphi philosophers who are reading and commenting (especially the care ethicists and feminist philosophers, whose feedback we hope to incorporate into our work): We were not polling people to find out how they would respond to Thomson’s case. We were running a controlled experiment, which means that (ideally) we changed one word in the thought experiment, and held everything else constant, in order to see what effect would result from changing that one word. By simply changing “violinist” to “half-sibling” we got a significant number of people to change their answers. It’s not important how they responded to the thought experiment. Rather, it’s important that one little change in the vignette got them to alter their original answer. So, our question is: What do you think this shift demonstrates?

Mark Phelan

I wonder if you might report the means for the different comparisons you find significant in your results section. I have a few questions about the findings that may or may not make sense, depending what those means are.

Also, I wonder if Kitagawa and Marie might explain what the significance of Thomson's arguments are supposed to be, assuming they don't elicit the intuitions they purport to elicit.

Clayton

Hi Alexandra,

I think this is a very interesting research project, but I had some concerns about the way that the results are described in the post above. My concerns largely stem from the fact that I have a hard time identifying Thomson's position in that paper and separating out the points that she's actually arguing for and the points that she accepts but doesn't intend to defend. So, I suppose that my general worry has to do with the extent to which Thomson is actually committed to denying the observations that you've made.

First, it's always seemed to me that Thomson's primary concern is to show that a specific argument for the extreme view fails and not to argue in favor of a fairly permissive view of abortion. (Yes, she says on pp. 66 that she is arguing for the permissibility of abortion in "some cases", but this strikes me as an incautious remark since mostly she's trying to show that the claim that the fetus has certain rights is consistent with the further claim that it is permissible to terminate a pregnancy. And _that_ consistency claim is certainly consistent with the claim that there are additional considerations that don't have to do with rights that can show that abortion is impermissible. So, it's not clear to me that she's committing herself to rejecting arguments from ethics of care as much as she's arguing that the issue cannot be as simple as the rights-based arguments show. (She does say that she's not arguing that abortion is always permissible, after all on pp. 65.)

Still, it might be sufficiently obvious that she likes a fairly permissive view that this is a bit nit picky. I'm interested in the objection on pp. 64. Thomson's remark:
"But of course there are arguments and arguments, and it may be said that I have simply fastened on the wrong one. It may be said that what is important is not merely the fact that the fetus is a person, but that it is a person for whom the woman has a special kind of responsibility issuing from the fact that she is its mother ... I have in effect dealt (briefly) with this argument in section 4 above; but a (still briefer) recapitulation now may be in order. Surely we do not have any such "special responsibility" for a person unless we have assumed it, explicitly or implicitly. If a set of parents do not try to prevent pregnancy, do not obtain an abortion, but rather take it home with them, then they have assumed responsibility for it, they have given it rights, and they cannot now withdraw support from it at the cost of its life because they now find it difficult to go on providing for it. But if they have taken all reasonable precautions against having a child, they do not simply by virtue of their biological relationship to the child who comes into existence have a special responsibility for it."

It seems two issues here are separable--
(i) whether responsibilities have to be voluntarily undertaken and
(ii) whether mere _biological_ relations between a pregnant woman and a fetus are enough to generate special responsibilities.

The data collected causes trouble for the claim she makes about what is undertaken voluntarily. (And, I agree that this claim isn't terribly plausible. Duties of gratitude seem like a perfectly good example of special responsibilities that don't arise because of some voluntary choice.) The data doesn't seem to cause trouble for her claim that a mere biological connection between the fetus and pregnant woman isn't enough to generate special responsibilities (esp. if the woman took reasonable measures to prevent those relations from being established but became pregnant because of failed contraception or because of rape)?

Also, could you maybe say more about what you take the objection on pp. 64 to be an objection to? As I read her, Thomson takes it to be an objection to this claim:
(*) The relation between a pregnant woman who became pregnant unwillingly is analogous to the relation between the subject and the violinist using her kidneys.

The objection that Thomson is trying to deal with, I take it, is this:
(**) There is an important difference between cases because of special responsibilities grounded in the relation between the pregnant woman and the fetus.

The data offered suggests that people will find the following claim intuitive:
(***) There can be special responsibilities that are not voluntarily undertaken between you and members of your family.

Is that the dialectic? I don't yet see how (***) bears on (**) or (*). (In fact, it's not clear that Thomson is all that opposed to the intuitions that seem to support (***). She describes situations where she thinks there is an obligation to keep the violinist alive even if the violinist has no rights to your kidneys ["So my own view is that even though you ought to let the violinist use your kidneys for the one hour he needs, we should not conclude that he has a right to do so--we should say that if you refuse, you are, like the boy who owns all the chocolates and will give none away, self-centered and callous, indecent in fact, but not unjust. And similarly, that even supposing a case in which a woman pregnant due to rape ought to allow the unborn person to use her body for the hour he needs, we should not conclude that he has a right to do so; we should say that she is self-centered, callous, indecent, but not unjust, if she refuses. The complaints are no less grave; they are just different."]. Even if she agreed that relations between family members can generate stronger reasons to assist, wouldn't she stress that no reason has been given for thinking that such relations exist between a pregnant woman and the fetus? )

Alexandra Bradner

Mark asks above: What is "the significance of Thomson's arguments ... assuming they don't elicit the intuitions they purport to elicit?"

Unfortunately for anyone wanting to debunk Thomson's armchair TEs with x-phi data, because she is making a *plausibility* argument with her TEs, her conclusion stands, as long as our data locates *some* group of people willing to say that remaining attached would be "outrageous," and our data does find such a group. This is what Clayton and Kitigawa are getting at, I believe.

However, I would add/argue, this plausibility argument is in the service of her ultimate conclusion, that assessments about the morality of abortion are going to depend on the details of the case. Given this, her hostility to x-phi is mysterious to me.

First, our data can make her plausibility argument much better than a hypothetical TE can. Second, our data identifies an additional factor that influences our judgments about the morality of detachment, a factor which is not personhood status, namely the relationship you/Lee bear to the attachee. This helps her case against those with tunnel vision about personhood. Third, if we're going to avoid reductive (in this case, pro-life) ideology and turn to a closer examination of cases, that is, to the actual multitude of factors partners consider when making abortion decisions, what could possibly be a better way to start than x-phi?

All of this has become a lot sharper for me in the course of this conversation, so Mark's question is perfectly reasonable, given the way I started the post above, with the p.64 quote. But thanks to this discussion, this is how I would respond to his worry now.

Alexandra Bradner

Hi Clayton,

I really appreciate your careful comments. You and Kitagawa have helped me to better articulate what I'm thinking re my "additional question 1" above.

(1) I agree completely with your reading of Thomson. As you say: "[M]ostly she's trying to show that the claim that the fetus has certain rights is consistent with the further claim that it is permissible to terminate a pregnancy. And _that_ consistency claim is certainly consistent with the claim that there are additional considerations that don't have to do with rights that can show that abortion is impermissible.... [I]t's not clear to me that she's committing herself to rejecting arguments from ethics of care as much as she's arguing that the issue cannot be as simple as the rights-based arguments show."

(2) I also agree with the following (and the way you sorted this out is *really* helpful): "It seems two issues here are separable--(i) whether responsibilities have to be voluntarily undertaken and (ii) whether mere _biological_ relations between a pregnant woman and a fetus are enough to generate special responsibilities. The data collected causes trouble for the claim she makes about what is undertaken voluntarily."

I agree, our data suggests that we recognize responsibilities that are not undertaken voluntarily, which I take to be a fact better dealt with by care views than justice views. But why isn't the data relevant to (ii) as well? Certainly one of the differences between the violinist and the half-sibling is the presence of a biological connection. So, given that the switch to half-sibling shifts responses, wouldn't it be reasonable to look further to see whether the reason for the shift was the biological connection? Are you saying that the half-sibling is not analogous to an uninvited fetus, because the half sibling is somehow more welcome? If so, in the case, "you" and "Lee" wake up to find the half-sibling attached, so the half-sibling is arguably uninvited. Yet, participants *still* say they will and are obligated to remain attached.

(3) You write: "The objection that Thomson is trying to deal with, I take it, is this: (**) There is an important difference between cases because of special responsibilities grounded in the relation between the pregnant woman and the fetus. The data offered suggests that people will find the following claim intuitive: (***) There can be special responsibilities that are not voluntarily undertaken between you and members of your family. Is that the dialectic? I don't yet see how (***) bears on (**) or (*). (In fact, it's not clear that Thomson is all that opposed to the intuitions that seem to support (***).)"

First, how you put (***) is terrifically clear, so thanks. And I agree with the last sentence there, that is: Thomson's sensitivity to a multitude of conditions and factors (her sensitivity to cases) recognizes that, as you say, "there can be special responsibilities that are not voluntarily undertaken between you and members of your family." What follows is going to be kind of long, so I apologize to everyone, but here is a better statement (than my post above) of the thoughts that motored this project for me. (I can't speak for Jeanine and Seth.)

Thomson and, now, Kitigawa seem to say re xphi: It’s an empirical fact that our intuitions about remaining attached *are* affected by a number of factors, including facts like: we value bodily integrity, we value life, we value relationships, etc. These tell us nothing about whether there is an *obligation* to remain attached, one we can’t ignore, on pain of wrongdoing.

I would want to respond to this in the following (sorry-so-long) way:

The start of Thomson’s article tells us that her interlocutor is anyone who argues the following: The mere fact that the fetus is a person, and that fact alone, is enough to obligate us to remain attached.

Thomson’s paper simply responds that the fact of personhood *alone* is not enough to obligate us to remain attached. *Other factors matter.* Given some of those other factors, we must remain attached, on pain of wrongdoing; and, given other of those factors, we can detach without violation. The decision to detach is to be made on a case-by-case basis.

She uses analogies to argue that personhood alone is not enough to obligate us to save a life. That is, there are plenty of “persons” the law does *not* obligate us to save. Under the present Good Samaritan laws, for instance, we don’t have a legal obligation to rescue strange people we never promised to rescue. Thus, we don’t have an obligation to remain attached to a strange and uninvited fetus we have taken certain reasonable steps to avoid producing either. Why should pregnant women bear a disproportionate burden? It’s seems unfair, because the fetus, violinist, and Good Samaritan cases are all analogous in relevant ways.

The argument is not made, as xphi-ers might think, by using readers’ intuitions as *evidence* for her position or even on the exactness of the analogies. Rather, the argument succeeds on the weakness of her claim. Remember, all she is saying is that personhood status *alone* is not enough to obligate us to remain attached to a fetus. So she simply has to persuade us that, in some relevant cases, some factor other than personhood status, like whether or not there is an invitation, intervenes and makes it impossible to apply the narrow/reductive pro-life ideology, which is not fine-grained enough to fit real world circumstances. In short, she is simply motivating something like: it’s unreasonably prescient and omniscient to argue that personhood status will *always* be the make-it-or-break-it property.

Her aim is to manufacture a reasonable doubt about the personhood-alone-matters ideology. That doubt about *ideology,* then, is supposed to shift our focus to casework.

Our thesis is that experimental philosophy and care ethics have *a lot* to say in regard to this casework, much more than Thomson (or anyone else) could ever say a priori. Casework is, after all, “casework” because we are looking to the details of real cases to understand (not justify) our moral judgments. Some might be tempted to worry about is-ought concerns here, but we muted that doubt when we realized how vacant and unsatisfactory ideology is/was. The dissatisfaction with ideology led to casework, so ideology no longer has the baseline reasonableness it would need to undergird a doubt about casework.

My initial concerns in this study were more meta-philosophical, stemming from an interest in the way philosophers of science use historical cases as philosophical constraints, in response to worries about a priori theorizing, and the (underdetermination) arguments against this work. I think there is an interesting argument in the Thomson paper for casework. And I think both care ethics and x-phi offer many more resources for this project than she seems to think. So it's not that we're arguing *for* care in response to her argument *against* care. We're arguing for *x-phi,* in response to her statement against x-phi, by showing that care concerns do, in fact, come in to play in a paper of hers and explaining just *how* they come in to play, which is something she could never do from an armchair. (We have more work to do re the "how.")

Anyway, *thanks* for your comments(!). They helped me to articulate what I was thinking, and your (i) and (***) gave me a very compelling way of thinking about our data. Moreover, your comments focused my attention more on the details of the Thomson, rather than on the pro-con x-phi issues.

Alexandra

philfemgal

I found two aspects of this jumped out at me. First, I am very surprised that only about 1/3 of the participants had the initial intuition that they were "definitely not" obligated to remain attached. I have always found things to go the other way, with only 1 or 2 students (if any) believing that maybe they are required to stay attached.

I wonder if one thing that might explain the low number of participants who thought they were definitely not morally obliged to remain attached might have to do with what the common sense understanding of "morally obliged" is. The last time I taught "A Defense of Abortion" we spent a lot of time trying to visually map out the categories of actions which are morally obligatory vs. those which are not obligatory (which seem to break down into at least three categories--actions which you would be indecent/callous if you failed to do, actions which are neutral, and actions which are praiseworthy.) We put this paper in conversation with other papers about the duty to rescue and tried to determine what different authors meant when they talked about "duty" or "moral obligation"--e.g. when Thomson says one is morally obligated to do x, must it be the case that someone else has a right that one do x? It often sounds like she means this. But perhaps this is a stricter way of using the term "moral obligation" than our common sense meaning which might be more like "you really, really ought" to do this and you'd be a really shitty person if you didn't. That is something different than meaning someone else has a right to your doing x. So I wonder if we can assume that the participants are using the right conception of obligation if Thomson's understanding of it hasn't been explained to them.

A second thought I had was regarding what the "half-sibling" case was intended to get out--whether that case supposed to be about mere genetic relations or about an actual relationship. From the discussion in the comments it seems like you meant to leave room for both.

My immediate thought upon reading the description of the cases, though, is that I don't know what "half-sibling" means and my answers would be very different depending on my understanding. As a participant I would probably have assumed you had in mind a sibling relationship in which, say, a divorce occurred after a marriage which produced a child and then when of the divorced parents had a child with a new spouse. And I would assume that these people have lived as children as siblings, either primarily in one household or perhaps in two separate households depending on the custody arrangement. (In my own family where "half-siblings" exist this is the situation. Though I can't say I have ever heard anyone in my family use the term "half-sibling" since this term might be taken to indicate a lesser emotional relationship, which is absolutely not the case amongst the sisters I have in mind.)

So I would think of my cousins who are technically half-siblings (same mother, different fathers), in which case the "half" does no work for my intuitions at all. Alternatively I might wonder if the survey writer had intended me to think about a non-sibling like relationship. Say Ann finds out at age 40 that her father had an extramarital affair and Bob is was the result. But now I need to fill in details about how Ann and Bob feel about each other and how Ann feels to her father and how her father feels toward Bob, etc, etc. Finally, the "half" could mean mere genetics--i.e. Sarah was conceived by donor sperm. There are another 150 people out in the world who were conceived by the same donor's sperm. She's never met them and has no interest in ever doing so and then she wakes up hooked up to one of them. Finally, I might also imagine a family formed through adoption in which the parents split and then a parent adopts another child with the new spouse. This would also be a "half-sibling" relationship--one in which my intuitions are no different than the first case of my cousins even though in my cousins case there are biological relations and in the adoption case there are not.

That is just to say, I'm not sure it is accurate to say that the "half-sibling" relationship necessarily involves biology or to assume that all participants would have had biology in mind when imagining such a relationship. Similarly, I wonder with so many ways to interpret what a "half-sibling" is, whether--as you suggest to Clayton--the results really can be taken to mean anything about biological relations.

Wesley Buckwalter

Wait, just so I have this straight. You’re claiming that JJT’s plausibility argument (that there exists at least one case in which personhood status itself does not alone morally obligate us to remain attached) does not bear on people’s intuitions about a particular case as evidence for what counts as moral obligation? I found this curious because you also say that “her conclusion stands, as long as our data locates *some* group of people willing to say that remaining attached would be "outrageous," and our data does find such a group.” What I took this to mean here was that some group of participants have the intuition in response to the proposed thought experiment that there is no moral obligation to stay connected, so therefore one such plausible case exists in which personhood alone does not obligate. I was thinking though, it seems like intuitions are pretty relevant to this “plausibility” argument in sort of typical xphi ways, since if nobody or very few people had them it wouldn’t be all that plausible!


Now about the results. Certainly there were *some* people in the experiment who answered all different kind of ways, but I thought you were taking the data to show 2 main things. First, that in the original-ish violinist case, despite perceptive, a majority of people (just barely unfortunately!) had the intuition that there is no moral obligation to stay connected. Second, that when people receive the relationship manipulation, they are significantly more likely to say there was a moral obligation to stay attached. Given these main things, I was just wondering what you would say about two different ways people might want to be interpreting these data. One way of course, is to use it to support JJT’s original thesis, as evidence that it’s not personhood itself that buttresses judgments in these sorts of situations, but rather personhood+other stuff, in this case your relationships to the individual shaping obligation intuitions. But also, another way people might want to take the data as mentioned above is to argue that if people think there is an obligation to stay connected in personhood+relationship cases, and if those exact cases are closer in analogy to situations of abortion, then independent of what JJT was originally doing perhaps, you’ve ended up with evidence that people have intuitions that support the moral obligation not to terminate a pregnancy. So, I was just wondering what people think, 1) if this is a good way t summarize what is going on, and or 2) what they would have to say about these two different conclusions.

Alexandra Bradner

Hi Wesley,

Thanks for your comment. You write: "[Alexandra is] claiming that JJT’s plausibility argument ... does *not* bear on people’s intuitions about a particular case as evidence for what counts as moral obligation?"

No, I'm claiming the *opposite,* that is, I agree with exactly what you say next. This might be more clearly stated in my response to Mark, where I wrote: "First, our data can make her plausibility argument much better than a hypothetical TE can." You and I agree: What could possibly make plausibility arguments better than *data*? Sorry if that was not clear.

On your second paragraph, I think what you have written is an excellent way to summarize what has been going on. Re the data, lately (that is, in all of these comments), I have been thinking that our data can be used, as you so nicely write: "to support JJT’s original thesis, as evidence that it’s not personhood itself that buttresses judgments in these sorts of situations, but rather personhood+other stuff, in this case your relationships to the individual shaping obligation intuitions."

Re your second interpretation: "[Alexandra has] ended up with evidence that people have intuitions that support the moral obligation not to terminate a pregnancy." I think this is right as well, and I do not think Thomson would be hostile to this conclusion, because she is open to the thought that there are cases in which it would be *extremely* indecent to abort. As Clayton writes above, "She describes situations where she thinks there is an obligation to keep the violinist alive even if the violinist has no rights to your kidneys."

In general, this is what I'm thinking: As I read Thomson, she ends up with a very care-ethics-ish conclusion (sensitive to details, no one-size-fits-all claims, open to the relationship factor, etc.), but persists with more global rights talk. I think she should jump ship, and x-phi could help her do it.

Alexandra Bradner

Hi to philfemgal,

You write: "[T]he low number of participants who thought they were definitely not morally obliged to remain attached might have to do with what the common sense understanding of 'morally obliged' is." I think you could be right, but of course we can't know for sure. (Only more experimental data could support anyone's particular intuitions over anyone else's here!) Participants may have read: "Are you/is Lee morally obliged?" as something like: "Do you/does Lee absolutely *have* to do it," especially when you remember that they're answering this question just after or before answering the "Will you...?" question in contrast. However, we used this phrase because *Thomson* does. So any ambiguity that might have befallen us, will affect her as well.

Your in-class activity sounds terrific, by the way.

I really appreciate your comments re the term "half-sibling." I think you're right that this term is dated--very 70s. Selecting the particular person to oppose Thomson's violinist was one of Jeanine and my longest conversations. We *really* agonized over it. (A little pre-testing might have helped us here.) We were looking for something that, as you wrote, had both a genetic and a potential relationship component, i.e. something relevantly *like* a fetus, but not *obviously* a fetus (because we didn't want to trip the *compulsion* to rescue).

You focus mostly on the biology remark I made to Clayton. I guess I think the biological connection is important and not important. On the one hand, it seemed important to locate something like a fetus, to satisfy the curiosity of those who complain that a strange violinist is no potential child. So, in this regard, we did need something with a biological connection. But, in the end, whether or not there is a biological connection to the thing is not important to our overall argument, which is, simply, that our data show, better than Thomson can, that it's plausible to say there are other factors at play in abortion decisions than personhood alone. And, given this, care ethics and x-phi offer more theoretical and methodological resources than more "traditional" (essentialist/reductive, universalist/one-size-fits-all, a priori, etc., etc.) philosophy.

We are presently preparing to continue our study by testing several other options. And I think this future data will help us understand more about what was in play in the half-sibling responses. Thanks so much for the time you took to offer this feedback.

The comments to this entry are closed.

3QD Prize 2012: Wesley Buckwalter