Counterintuitive intuitions

by John Quiggin on May 5, 2006

I’m reacting a bit late to Brian’s post on thought experiments in ethics. Like some commenters, I’m unimpressed by such exercises. In too many cases the approach seems to be to postulate a totally counterintuitive situation (for example one in which pushing people onto railway tracks has good consequences) then claiming that people’s intuitions about such a situation tell us something useful.

Here’s another one quoted by Walter Sinnott-Armstrong in the Stanford Encyclopedia of Philosophy. I’m reposting my response from a few years ago.

One common illustration is called Transplant. Imagine that each of five patients in a hospital will die without an organ transplant. The patient in Room 1 needs a heart, the patient in Room 2 needs a liver, the patient in Room 3 needs a kidney, and so on. The person in Room 6 is in the hospital for routine tests. Luckily (!), his tissue is compatible with the other five patients, and a specialist is available to transplant his organs into the other five. This operation would save their lives, while killing the “donor”. There is no other way to save any of the other five patients (Foot 1966, Thomson 1976; compare related cases in Carritt 1947 and McCloskey 1965).

We need to add that the organ recipients will emerge healthy, the source of the organs will remain secret, the doctor won’t be caught or punished for cutting up the “donor”, and the doctor knows all of this to a high degree of probability (despite the fact that many others will help in the operation). Still, with the right details filled in, it looks as if cutting up the “donor” will maximize utility, since five lives have more utility than one life. If so, then classical utilitarianism implies that it would not be morally wrong for the doctor to perform the transplant and even that it would be morally wrong for the doctor not to perform the transplant. Most people find this result abominable. They take this example to show how bad it can be when utilitarians overlook individual rights, such as the unwilling donor’s right to life.

I don’t know if it’s been pointed out before, but this example doesn’t work as claimed. The proposal of killing the test patient is dominated by the following alternative: With the agreement of the five needy recipients, draw lots. The unlucky one is cut up (but of course, they would have died anyway) and their healthy organs are transplanted into the others. The number of lives saved is the same as in the proposed case, no rights are violated, it’s a Pareto-improvement on the status quo ante and so on. We even save one transplant operation relative to the proposal.

Of course, you can impose some sort of ad hoc assumption to rule this out, but this just points up the other flaws of this example.

First, it’s an appeal to intuition, but it’s based on so many counterintuitive assumptions that it’s hard to believe that intuition is going to be a reliable guide. As Sinnott-Armstrong notes, utilitarian critics have already made this point.

An equally serious problem is that the example is like a bad cryptic crossword clue – you know what answer you’re supposed to get, but not exactly how you got there. Is the problem that an innocent person is killed in order to preserve the lives of others? If so, why not look at conscription or war in general? Or is it some sort of claim about rights to bodily integrity? In that case, why not ask about a nonlethal violation of this right e.g. a compulsory blood donation to save a life? Or is the problem that explicit and active agency of the doctor? After all, implicit decisions to sacrifice one life in order to save others are made in hospitals every day.

FWIW, I’d suggest that the core problem is that we know that it would be a very bad idea [from a consequentialist or any other viewpoint] to let doctors kill people for their organs , and that no amount of counterfactual assumptions is going to shake this belief.

PS Coincidentally, Maureen Dowd has an excellent piece on this topic in today’s NYT.

{ 1 trackback }

Crooked Timber » » Meatscaping
05.05.06 at 9:33 am

{ 59 comments }

1

Daniel 05.05.06 at 5:20 am

I really hate the “you just know” stipulations by which the entire problem of calculating an expected value is assumed away.

2

abb1 05.05.06 at 5:44 am

The test patient should shoot the doctor as soon as the tests are done, just in case; better safe than sorry.

3

Harald Korneliussen 05.05.06 at 5:55 am

When did we all agree that killing one person is better than letting five die? If killing is the same as letting people die, wow, I’ve effectively killed every person who has died since 1981, including many of my relatives.

When did we agree that killing one person is better than killing five?

When did we even agree that it’s an obligation to let one die if the alternative is letting five die?

HK, unrepentant deontologist.

Live and let die, but do not live and kill, please.

4

Harry B 05.05.06 at 7:18 am

I got lost in Brian’s thread, and couldn’t struggle through it. But I think you are reading the purpose of such thought experiments the wrong way. It is not to tell us what we should do in those circumstances. It is to help us isolate our intuitions so that we can find out what general principles we really believe. Hence the “you just know” conditions, which help you not to have stray thoughts. So, think about Thomson’s violinist. The experiment is designed to get you to say that you have a right not to stay hooked up to the violinist. But the violinist has a right to life (no-one will deny that). You are then asked to think whether, in unhooking yourself, you would be violating his right to life. Intuitively, not. But (she claims) the situation is structurally identical to the case of the relationship between the pregnant woman and the fetus. Does it follow that she has a right to abort? Well, no, at least not without lots of further argument. But it does follow that she would not be violating the right to life of the fetus if she aborted. So, if abortion is impermissible, that is not because the fetus has a right to life (which Thomson grants it does, for the sake of argument) but for some other as-yet-to-be-specified reason. So, the thought experiment successfully undermines one argument for the impermissibility of abortion by undermining the principle “Abortion violates the right to life of the fetus”. This seems like a result, to me. I think the paper itself, by use of several more interesting thought experiments, establishes an even more interesting conclusion; that denying the right to abortion commits you to a very strong principle of a duty to aid others in great need. No surprise to good Catholics, but a hell of a surprise to some American conservative anti-abortionists.

I haven’t looked at the Sinnot-Armstrong, and not all thought experiments are as successful and well-wrought as Thomson’s (she’s brilliant) but once you see the rather limited ambition of the thoought experiment it is a useful tool. The BBC site obscured this, but that’s only to be expected; they were trying to have fun.

5

Dan Kervick 05.05.06 at 7:23 am

An equally serious problem is that the example is like a bad cryptic crossword clue – you know what answer you’re supposed to get, but not exactly how you got there.

Well the usual procedure, when addressing these examples in a classroom, is to propose various principles that, along with the contingent circumstances of the example, do indeed provide an account for how we get there, and then to test those principle by considering other hypothetical circumstances.

I also seem to be missing something in your proposed alternative action – wouldn’t that action lead to saving only four lives, rather than five? Any way, you say that that alternative dominates the others. Fair enough … the next step would be to propose an alternative moral theory, presumably some version of ideal utilitarianism augmented with justice as an intrisic value, or with side constraints of some kind, and then to test that theory against other hypothetical circumstances.

When considering hypothetical circumstances such as these, people often find that principles they are committed to seem to produce results in certain hypothetical circumstances that run counter to their own intuitive judgments about what one should do in such circumstances. Faced with this situation, the next rational step would appear to be to (a) reject the intuitive judgments, (b) reject the principles, or (c) show that the principles do not really recommend the counterinuitive actions, and that there is an error of reasoning in the argument that purports to show they do.

However, the exercise of considering hypothetical examples often generates a certain amount of discomfort. People often resent being made to evaluate the consequences of their beliefs. They prefer to leave their motives and principles shrouded in ambiguity as they are in real world, where several potentially conflicting moral principles may all agree on the same course of action in the well-behaved options the real world typically presents. Why sow discord, they feel, by showing us we actually disagree – just when we were all getting along so well?

The particular kind of discomfort felt by the committed moral theorist, though, may generate deeper methodological criticisms of the procedure. They may argue that the cognitive and emotional dispositions that generate intuitive judgments are useful but very imperfect guides to action, and that their reliability breaks down rapidly as the hypothetical circumstances become more remote from the concrete circumstances we typically encounter in the actual world. That is a plausible conjecture, and many moral theorists appeal to this conjecture in response to certain hypothetical examples, as a way of defending their theory in the face of what seem to be contrary considerations. But I have a conjecture of my own based on experience: the hypothesized boundary of applicability always seems suspiciously to lie in just about the same region where the theorist’s favored moral theory starts generating counterintuitive results.

Another approach to the hypothetical examples routine is to stick one’s fingers in one’s ears and say: “Satan be gone! I know there is something wrong with this procedure, which tempts me to give up some of my cherished beliefs – even though I can’t say what exactly is wrong with it.”

6

greensmile 05.05.06 at 7:44 am

If the postulated, “… classical utilitarianism implies that it would not be morally wrong for the doctor to perform the transplant and even that it would be morally wrong for the doctor not to perform the transplant“, is a correct application of utilitarianism then that is a rediculous and narrow principle. Among the flawed assumptions here are that conditions as natural and universal as illness and aging, though they be dire, somehow endow their victims with higher priority entitlement to treatments and services.

This straw-patient is not worth operating on!

7

soru 05.05.06 at 7:47 am

wouldn’t that action lead to saving only four lives, rather than five?

Test-guy gets to live to.

8

Dan Kervick 05.05.06 at 7:55 am

Test-guy gets to live to.

Ooops… duh. Thanks soru. Now I get it. But so this seems to me a good example of the usefulness of considering hypothetical circumstances.

9

Matt 05.05.06 at 7:56 am

Harry Said, about these examples,
“It is not to tell us what we should do in those circumstances. It is to help us isolate our intuitions so that we can find out what general principles we really believe.”
But I think this is exactly what John and I and others think that they cannot do, or at least cannot do very well. For this to work we would have to have some reason to think that the “intuition” that is isolated from such cases are reliable, but given the highly unusual nature of the cases I don’t really think there’s a good reason for thinking that. Now, the examples we’ve discussed here are, as anyone who’s spend some time reading analytic ethics from the last third or so of the 20th century knows, pretty mild in their weirdness. Even Thompson’s violinist example, which I do rather like, is a lot more comprehensible than ones she comes up with later in the paper (Plant people blowing on thing wind and growing in your carpet, etc.) The same with the more and more unusual permutations of the trolly problem. I guess I don’t want to say that such thought experiments are worthless (in ethics or in any other branch of philosophy) but I’m skeptical that they are really worth all the energy that’s been put in to them. It’s hard enough to make a good moral theory that works for the core cases, after all, and I still see no reason to think that our intuitions in these very strange cases should be thought to be good guides to much of anything at all. They may make good pedagogical tools (though even that has to be done more carefully than it usually is) but I don’t think they tell us anything very deep.

10

abb1 05.05.06 at 7:58 am

So, the thought experiment successfully undermines one argument for the impermissibility of abortion by undermining the principle “Abortion violates the right to life of the fetus”.

Strictly speaking, I don’t think it does that, necessarily. Violinist and fetus are different things; try “your mother” instead of “violinist” and the result might be the opposite – what are you going to do then?

11

emmanuel goldstein 05.05.06 at 8:31 am

For this to work we would have to have some reason to think that the “intuition” that is isolated from such cases are reliable, but given the highly unusual nature of the cases I don’t really think there’s a good reason for thinking that.

Timothy Williamson defends an interesting principle of charity on which intuitions are generally reliable in (sections 5 & 6 of) ‘Philosophical ‘Intuitions’ and Scepticism about Judgement’ (pdf).

12

Harry B 05.05.06 at 8:34 am

Abb1’s example in fact bolsters the success of the case. If you put in “your mother” I get a different intuition myself, but its clear that the right to life has remained constant. So, the right to life is irrelevant to the change in intuition. So if I’m obliged to stay hooked up to my mother it is not because she has a right to life (the violinist has that, too) but because of some special obligation that I don’t have to the violinist. Look for that special obligation, and arguments for it. Forget about the right to life as a grounding of this sort of obligation. Similarly the obligation to carry a fetus to term.

I don’t think they tell us anything very deep; its precisely the expectation that they do or should that leads you to be disappointed with/hostile to them. They are indeed useful pedogagical tools; not just in the classroom, but in our own heads when we are trying to map out the conceptual space. Sure, I agree that some people use them badly, and also that some people who use them well are interested in some of the corners of conceptual space that don’t interest me (and don’t interest most people). That’s ok, no? Would we be better off without them? I doubt it.

13

emmanuel goldstein 05.05.06 at 8:34 am

14

abb1 05.05.06 at 9:01 am

I don’t believe I find myself defending ‘pro-life’ people, but I think you may be arguing against a strawman here. Do they actually put forward this generic catch-all “right to life” argument? I doubt it, because (I assume) most of them are also pro-death-penalty and that’s already a contradiction right there. No, I am sure vast majority of them are arguing this one very specific case – aborting a fetus; and they couldn’t care less about your violinist.

Inconsistent? Maybe, but maybe not, because the fetus is different: it’s a ‘child’, it’s ‘innocent’, it was ‘never given a chance’, or maybe pregnancy is special altogether, a ‘miracle’ or whatever.

15

Tim 05.05.06 at 9:28 am

Thanks, John, for pointing out the the fundamental problem with these things: counterintuitive conditions 1,2,3,4, and 5, and then you’re supposed to see what your intuition thinks?

As I see it, these ‘experiments’ are rhetorical devices to get a discussant into committing himself to an untenable position, or at least an uncomfortable one, so that the arguing philosopher can step in and say “ah, this is the solution” and it looks very clear.

Does this “think, think, think, then let go of your thoughts and go with your gut” method have a place in moral philosophy? The only example I can think of is Leon Kass’s defense of “revulsion” (is that his term? I can’t remember this time of morning) as an ethical principle — are there other, better ones out there?

16

Maria 05.05.06 at 9:35 am

I think abb1 has a point. I was wondering what would happen if you substituted “your mother”, “your child”, or “your brother/sister” for any of the “fat people” in the examples.

17

marcel 05.05.06 at 9:46 am

JQ wrote:

I don’t know if it’s been pointed out before, but this example doesn’t work as claimed. The proposal of killing the test patient is dominated by the following alternative: With the agreement of the five needy recipients, draw lots. The unlucky one is cut up (but of course, they would have died anyway) and their healthy organs are transplanted into the others. The number of lives saved is the same as in the proposed case, no rights are violated, it’s a Pareto-improvement on the status quo ante and so on. We even save one transplant operation relative to the proposal.

This situation does not successfully counter the original one: a key assumption is missing, to wit:

Luckily (!), his tissue is compatible with the other five patients, and a specialist is available to transplant his organs into the other five.

I suspect that compatibility is not transitive, so that tissue of the unlucky, unsuspecting, sixth man [sic] is compatible with that of each of the other five does not mean that any one of them is necessarily compatibile with each of the others’. Unless JQ is willing to make an unrealistic assumption (he is an economist, after all, so perhaps this not a big objection), another counter is needed.

18

Bill Gardner 05.05.06 at 9:54 am

I do not think that John’s example is ‘totally counterintuitive’. It wouldn’t happen in a hospital in a liberal democracy, but the example doesn’t say that. However, there is great concern that China has been selling organs for transplantation that have been harvested from executed prisoners. This is a more interesting case, first because it is probably real. Second, because the expected value argument is even stronger here, if you grant the premise that the prisoner is going to die anyway.

19

Bill Gardner 05.05.06 at 9:56 am

Oh, and in case anyone is worried, I think the selling of organs harvested from executed prisoners is wrong / appalling / etc…

20

Richard Bellamy 05.05.06 at 10:12 am

Abb1 has a good point, but complementary to that is the intuititions of the “pro-choice” position. If there is a one-hour operation that can hook the violinist (or your mother) onto someone else instead, do you have to go through with it, or do you have the right to kill the violinist right now?

Imagine a not-to-distant technology in which 12 week old or younger fetuses can be extracted unharmed from the woman seeking an abortion, and implanted unharmed in the volunteer pro-life activists picketing outside.

Now the pro-choice advocates have to decide whether they are really “pro-choice”, or if they are “pro-death” or “pro-abortion” as the pro-lifers claim. Do the fetus’s rights (whatever they are) trump the woman’s right to have the fetus not exist?

21

Sebastian Holsclaw 05.05.06 at 10:33 am

Richard Bellamy makes a point that I was thinking about–the time factor.

Does the intuition change if you only have to be in bed and in the hospital for one more minute? One more hour? One more month? It is my feel that people are more and more uncomfortable with abortion as the time before birth approachs. Also the level of discomfort might be useful to look at. What if your hookup to the violinist wasn’t a big inconvenience until the end of the nine months?

And personally I think if I woke up in the violinist hypothetical, I would feel obligated to continue the treatments and would feel that I had participated in something akin to manslaughter if I didn’t. So the ‘obvious’ intuition that Thomson is going for isn’t obvious to me.

22

Moriarty 05.05.06 at 10:54 am

I always thought these experiments had value in “bracketing” the issue. You pick obvious cases on both sides – of course you wouldn’t be allowed to kill the guy and harvest his organs – and then work inward to less obvious ones. They aren’t meant to be the final word on the subject.

23

pedro 05.05.06 at 11:31 am

As much as I sympathize with John Q.’s distrust for these experiments (I distrust them myself, for similar reasons), I wonder if he’s stacking the deck by electing to use the adjective “counterintuitive” to describe the situations. Doing so makes it easier to suggest that there is little we can learn about “intuition” by looking at people’s reactions to them. I’d prefer the word “implausible” to describe the situations.

I do not think that choosing unfamiliar situations as the setting in which to study moral intuition–whatever that may be–is a bad thing. (Psycholinguists learn about language production from studying speech errors.) But I confess that, at a visceral level, I have almost the same reaction to these sorts of thought experiments that John Q. does.

24

Matt Austern 05.05.06 at 11:47 am

The point in #21 is specifically addressed by Thomson’s article. Actually, that article addresses almost all of the points people have been making in this thread. The violinist scenerio is just the beginning of the discussion, and Thomson understands perfectly well that you need further argument to get from there to a pro-choice position.

What the violinist argument does show is that you also need further argument to get to an anti-abortion position. Some people in this thread have responsed to Thomson’s argument by saying that the violinist scenerio is different from pregnancy because one’s share of responsibility for getting into the situation is different. But that misses the point. Thomson’s point is precisely that you have to look at issues like responsibility and special duties of care that wouldn’t apply more generally and things like that. You can’t go from “fetuses are people” to “abortion is always wrong” without some additional argument in the middle. Most of Thomson’s article deals with what that additional argument might look like.

And there is also a sense in which Thomson’s article rescues the anti-abortion movement from the charge of incoherence. People on the pro-choice side often ask: if anti-abortion people sincerely believe that abortion is murder, then why are they willing to pass laws with any exceptions? Why aren’t all US anti-abortionists pushing for harsh enforcement regimes like what they have in El Salvador? Answer: an anti-abortion position with exceptions isn’t necessarily incoherent precisely because of that implicit argument in the middle that Thomson’s article calls attention to.

(In real life, by the way, Thomson does not concede the “fetuses are people” premise. Her article explores what consequences you might get to if you do accept that premise.)

I recommend that anyone who is interested in this article go and read it. It’s short, clearly written, and very easily available. It’s a good example of how hypothetical situations can help to clarify a situation instead of muddying it. You’ll probably find this article in every ethics anthology for undergrad philosophy classes written in the last 25 years.

25

cd 05.05.06 at 11:48 am

On the subject of John Quiggin’s draw-lots- alternative, which he uses to object to the anti-utilitarian thought experiment above:

Yes, the same # of lives is saved in the draw-lots-alternative. But utilitarianism cares about aggregate happiness, not # of lives. So the anti-utilitarian can stipulate that the person in room 6 for routine tests is a homeless guy leading an unhappy life, with no loved ones who will grieve in his absence, etc. Then cutting him up looks back to being the clear winner as regards aggregate utility (provided we stipulate that the folks needing organs won’t go on to lead horribly miserable lives or cause others great misery).

At least this is so if short run effects are all that one takes into consideration. Utilitarianism can argue that the long run effects of the public learning about this killing will be bad–people will stay away from hospitals in droves, and so die prematurely of all sorts of preventable ailments. So utilitarianism can in the best analysis plausibly be said to condemn the killing. But in response, anti-utilitarians have an effective reply: They can say that even if utilitarianism gets the right answer, it does so for the wrong reason. For it essentially condemns the killing as bad public health policy. But surely it is wrong because it wrongs the homeless guy, in demoting him to the status of a sack of spare parts, rather than respecting him as a person with his own life to lead. (Yes, I’m a Kantian.)

26

abb1 05.05.06 at 12:10 pm

You can’t go from “fetuses are people” to “abortion is always wrong” without some additional argument in the middle.

I suspect that most anti-abortion people justify their position not by any argument, but by a very similar thought experiment: imagine that you discovered a small innocent child growing inside you, totally dependent on you… do you you think it’s permissible to hire a strange man to kill this baby inside you, cut it into pieces and pull it out by metal hooks… suck the brains out with a vacuum cleaner… collapse the skull with a set of plyers…

Phrase it this way and you intuition will leave you no choice…

Ah, but now imagine that a strange parasitic creature crawls into your stomach uninvited in the middle of the night… it sucks your blood and your precious body fluids… it makes you vomit… it destroys your teeths… it makes you ugly, it tortures you, ruins your life… Are you permitted to throw it out? No?? WTF?!!

27

blah 05.05.06 at 1:22 pm

This sounds like a great idea!

Does that make me a bad person?

28

theogon 05.05.06 at 1:23 pm

It’s also worth noting, since it so often isn’t, that an additional argument is needed to get from “abortion is always wrong” to “abortion should be illegal.” There are, after all, difficulties with enforcement of concern to both utilitarians and deontologists.

29

nik 05.05.06 at 1:25 pm

Sorry, I’m having trouble following John’s counter-example.

Doesn’t the fact that his dominating alternative needs the “agreement of the five needy recipients” means is isn’t a dominating alternative? The reason for the example is that people find it hard to commit to classical utilitarianism because it overlooks individual rights. But if you need the “with the agreement of the five needy recipients” line then you aren’t overlooking individual rights and you aren’t a classical utilitarian. So the rational for the thought experiment is lost.

Or am I missing something?

30

jt 05.05.06 at 1:49 pm

I think that the example of the patients does show something, but as many commenters have pointed out, the absurdity of the example reduces its credibiliy. I actually think that the example carries a lot of extra baggage which is unnecessary to make the point. Why not just ask the question directly:

1. Is it, or can it be morally acceptable/obligatory to kill one person in order to save several people?

2. If one can save someone or several people by either (a) letting someone else die, or (b) killing someone else, is (a) morally preferred?

These questions can also be generalized so that they don’t involve life and death, but rather other consequences.

The examples alert us to these questions, but anything can alert you to a question. We can have inutions directly about the answers to these questions rather than the specific examples, and I cannot see why the intuitions about the questions in general would be less reliable. If examples are a way of moving our thoughts along about the general principles, then fine, but it is also true that embedding strange examples in the analysis reduces the credibility of the analysis in many people’s eyes, and it doesn’t really add much, and can often be gratuitous. However, understanding the principles that the examples point to is worthwhile.

31

Jake 05.05.06 at 2:24 pm

It should also be pointed out that transplantees virtually never live a normal lifespan. The more serious the transplant, the shorter the expected lifespan. If patient #6 has an expected remaining lifespan of more than say 40 years, letting him live instead of the five others would almost certainly give a result of more total life-years. Dumb hypothetical.

Not to mention that lung transplants for adults are usually because of smoking-related problems . . . .

32

Mike Otsuka 05.05.06 at 2:27 pm

John Q wrote:

An equally serious problem is that the example is like a bad cryptic crossword clue – you know what answer you’re supposed to get, but not exactly how you got there.

Dan K rightly responds:

Well the usual procedure, when addressing these examples in a classroom, is to propose various principles that, along with the contingent circumstances of the example, do indeed provide an account for how we get there, and then to test those principle by considering other hypothetical circumstances.

Dan K description of what goes on in the classroom is also to be found in print. Nobody who was familiar with, and had an at least rudimentary grasp of, the classic papers by Foot and Thomson to which Sinnott-Armstrong refers would have raised John’s objection.

33

Mike Otsuka 05.05.06 at 3:05 pm


This excellent comment
by Tom Hurka in Brian’s thread should also have pre-empted John’s recycling of his cryptic crossword objection.

34

John Quiggin 05.05.06 at 3:06 pm

Mike (and others), the example isn’t being put forward for pedagogical purposes. It has been put forward in a series of papers, and in Stanford Encyclopedia of Philosophy as the basis of a moral argument against consequentialism.

I agree that this kind of example might be useful pedagogically. Students might be encouraged for example to find defects like the one I noted. But this use is almost directly opposed to the idea of taking intuitions on cases like this as being data with which an ethical position should comply.

35

John Quiggin 05.05.06 at 3:38 pm

Nik, the point you’re missing is that the recipients should rationally agree (they gain a 1/5 chance of living) and so I don’t need an extra counter-intuitive assumption to generate this feature. But the dominance claim works without this feature, if it worries you.

36

Mike Otsuka 05.05.06 at 4:10 pm

John,

In the Stanford Encyclopedia of Philosophy piece, Sinnott-Armstrong makes reference to two of the earliest and most influential discussions in the philosophical literature of the transplant case – those by Foot and by Thomson. (He also refers to cases of Carritt and McCloskey having to do with the framing of the innocent.) Though Foot and Thomson are anti-consequentialists, and surely they and many others think the transplant case causes problems for consequentialism, they don’t deploy this case in these articles in order to refute consequentialism. Rather, they pose this example, alongside several others, in order to gain a better understanding of the moral significance of various distinctions, among them the distinction between killing and letting die, between intended and merely foreseen killings, between negative duties to avoid harm and positive duties to provide aid, and between the initiation of a threat versus the redistribution of an already-existing threat. Their discussion also has the related aim of trying the unearth the legitimate grounds we might have to affirm the duties to harm and aid that we think we have. Their discussion proceeds in roughly the manner that Dan Kerwick and Tom Hurka describe. Just about everything I’ve just said about the Foot and Thomson pieces can be said about subsequent discussions of the transplant case in the philosophical literature – with which I suspect I’m somewhat more familiar than you are, since I specialize in moral philosophy, and you don’t. Your cryptic crossword remark is simply a travesty of this literature.

37

Seth Edenbaum 05.05.06 at 4:25 pm

The military is modeled on very specific principles, utilitarian, I guess, according to very specific goals.

What are he strengths and weaknesses of a system of justice predicated on such terms?

38

abb1 05.05.06 at 5:00 pm

Isn’t it a serious flaw in this exercise that the result is so susceptible to even subtle rhetorical tricks, the ways these scenarious are presented, phrased; just like the well-known problem with opinion polls, ‘estate tax’ vs. ‘death tax’ being a classic example? This itself could be the subject of a study of course…

39

bi 05.05.06 at 11:49 pm

Mike Otsuka: re Tom Hurka: I fail to see the need to create implausible, artificial test cases just to obtain disagreement, when there’s no shortage of perfectly good real-life cases where people actually disagree (remember Terri Schiavo?).

I recall someone else pointing out the problem with “what if” scenarios. What if dead people can be resurrected? What if zygotes have fully-functional brains? What if we can go back in time? What if pigs can fly? Well, if these are true, then won’t we’ll need to rethink a lot of things? Sure… but how relevant are the results in a world where pigs can’t fly?

40

Mike Otsuka 05.06.06 at 4:19 am

Bi: In real-world cases such as abortion or euthanasia where people disagree, we still need to assess the different justifications that people offer for their views. And sometimes it’s illuminating to appeal to hypothetical rather than actual cases in order to test the plausibility of these different justifications. See, for example, Harry B’s comment #4 above on Thomson’s violinist case, which serves as a test of the claim that abortion is wrong because the fetus has the same right to life as an adult human person.

Abb1 #39: Yes, I agree that’s a genuine problem. Among other things, the order in which cases are presented affects people’s responses, as they try to render their later responses consistent with their earlier ones. And different descriptions of what is essentially the same case will also make a difference to people’s reactions. So it’s important to randomly vary the order of presentation of cases and also to track the ways in which different descriptions elicit different responses.

41

MIchael Kremer 05.06.06 at 8:17 am

Re: the Thompson violinist example. It’s always seemed to me to be an unreliable intuition pump because it abstracts not just from the relationships involved in pregnancy, but from the simple fact that pregnancy is a normal feature of the human form of life. We may now soon be able to choose to reform our form of life in fundamental ways, so that it’s continuation does not any longer rely on pregnancy. If we do that, it will be a very momentous step indeed. But as it stands, becoming pregnant just isn’t like either of the scenarios described in #27. Because pregnancy is *normal*, not weird and artificial. And if there are “intuitions” here they’re surely being distorted by the equation of a normal state with weird and artifical states.

I haven’t read the Thompson article in a long time, so I can well imagine this point is met in it. Perhaps I will reread it now. Probably not a bad effect of this thread, if it happens.

42

abb1 05.06.06 at 9:03 am

In fact we have already reformed our form of life in fundamental ways, creating such fundamentally new concepts as ‘planned pregnancy’, ‘accidental pregnancy’, ‘unwanted pregnancy’, ‘pregnancy by rape’, etc., and thus your statement “pregnancy is a normal feature of the human form of life” seems about as informative as, say “violence is a normal feature of the human form of life”.

43

Hob 05.06.06 at 10:22 am

This serious discussion is all well and good, but if we want real answers to the basic questions, it’s time to roll out the brain in a vat at the wheel of a runaway trolley.

44

bi 05.06.06 at 11:22 am

Mike Otsuka: but why bring up a “what if”, when we can just look to cases like e.g. Abu Ghraib and Hurricane Katrina, where we can clearly see how highly the anti-abortionists value adults’ right to live? Bringing in flying pigs, square circles, and sentient zygotes gains you nothing, at the expense of introducing nonsensical stuff into your arguments.

Hob: is the runaway trolley situated in Hypotheticopolis?

45

John Quiggin 05.06.06 at 3:36 pm

Mike, as I said before, I’ve got no problems with the use of examples like this in pedagogy, though real-world examples would seem to do better in many cases. Certainly, if you’re trying to understand your moral intuitions, the “cryptic crossword” aspect of the example is a benefit not a cost.

But I was first presented with the organ bank example as a knockdown argument against consequentialism. And Sinnott-Armstrong introduces the example as follows

Another problem for utilitarianism is that it seems to overlook justice and rights. One common illustration is called Transplant …

In this context, the illustration purports to show
(i) consequentialism violates reliable/widely held moral intuitions
(ii) it does so because it overlooks justice and rights
I don’t think this example works on either score, and I think the same is true in general for ethical examples based on implausible problems.

46

MIchael Kremer 05.06.06 at 3:44 pm

abb1: I don’t believe “pregnancy by rape,” “unwanted pregnancy,” and the rest are some sort of new -fangled concepts. People have been trying to control pregnancy for a long time and have wanted or not wanted to be pregnant for a long time. And, most terribly, rape has been around, and people have known that it could lead to pregnancy, for perhaps an even longer time.

That said, I didn’t intend to be saying something especially “informative.” More like a reminder of something we all know, that’s too obvious to be stated, except when philosophical imaginings lead us to temporarily forget it. The point is, finding yourself pregnant isn’t just like finding yourself hooked up to a violinist in a hospital. People don’t normally find themselves hooked up to violinists. Women do normally find themselves pregnant. And have been for the entire history of the human species.

Let me make my point another way. Suppose that as part of our biological life, we periodically grew tentacles that hooked us up to other humans, from whom we obtained some sort of regenerating substance allowing us to continue our lives. Suppose that we’d evolved that way, that we all depended on this process, that we all had to be on the “receiving” end of the process at various points in our normal biological lives, etc. What would we then say about the case of someone who wakes up to discover himself to be “host” to another person in this way? I don’t have clear intuitions about this scenario as so described.

That’s all.

47

Baal_Shem_Ra 05.06.06 at 3:46 pm

Let’s play.

You have to choose between World A and World B. In World A, people only have good intentions and only bad consequences occur. In World B, people only have bad intentions and only good consequences occur.

Which one do you choose?

48

abb1 05.06.06 at 5:07 pm

Well, Michael, as people smarter than me noted above, the point of this exercise is to demonstrate that claiming some kind of a universal, unconditional “right to life” will lead to at least one hypothetical counter-intuitive situation. And this exercise does exactly that, no more and no less, so that strictly speaking it’s irrelevant whether being kidnapped and hooked to a violinist is a normal every-day situation or a highly unlikely one.

Now, I personally don’t see much value in refuting the “unconditional right to life” claim, because I don’t think anyone is seriously making this claim (well, some buddhist monks, maybe?), but if someone want to deal with it, I don’t see any problem.

49

Mike Otsuka 05.06.06 at 6:02 pm

John,

You say that the cryptic crossword aspect of the transplant example is a benefit rather than a cost if you’re trying to understand your moral intuitions. I can’t figure out why you think this. Would you please elaborate a bit? And could you also explain how this cryptic crossword aspect renders the example problematic as an objection to consequentialism? As far as the question of whether this example embarrasses consequentialism is concerned, all that would seem to matter is whether it’s true both that we ought not carve up the one and that consequentialism says we ought to. The alleged fact that one knows that it’s wrong to carve up the one, but not which of several competing explanations regarding its wrongness is the right one, would seem neither here nor there insofar as the truth or falsity of these two claims is concerned.

(Incidentally, you said, in #35, that these sort of thought experiments might have pedogogical value: for example, it might be useful to present these examples to students to teach them how worthless they are. I then pointed to the much more positive, and not merely pedogogically valuable, insights which Foot’s and Thomson’s discussion have yielded. So it’s disingenuous for you to reply by saying in #46: “as I said before, I’ve got no problems with the use of examples like this in pedagogy”.)

50

Matt 05.06.06 at 6:19 pm

Mike,
I suppose much of the discussion here is over whether Foot and Thompson’s examples have really yielded much valuable insight. You think so, obviously, but it’s possible to be familiar with the literature and think that this has rather been a huge waste of time without very significant results. It’s not as if its uncontroversial, after all, that such examples are really useful, even among philosophers. And saying so doesn’t make one an enemy of analytic ethics, obviously enough. (Do Hurka and Brighouse use such examples in their own work? Not all the time, at least, and not in the papers I’ve read. They seem to do fine without them, and I’ve profited more from their work that I’ve read than I have from papers using such examples.)

51

Mike Otsuka 05.07.06 at 4:39 am

Matt,

I grant that there are good and serious objections to the use of artificial, hypothetical examples, though, like Tom Hurka, I don’t know of any better way to assess competing explanations of why it is that we think that certain acts are right or wrong.

But there are bad as well as good objections to the use of such examples, and I was simply objecting to John’s claim that an “equally serious problem [with the transplant case] is that the example is like a bad cryptic crossword clue – you know what answer you’re supposed to get, but not exactly how you got there.”

As I understand him, John’s cryptic crossword point is that the transplant example is a bad objection to consequentialism because the example doesn’t settle the question of whether it’s wrong to carve the one up because (i) that would be killing an innocent, (ii) that would be violating the person’s right of bodily integrity, or (iii) that would involve the explicit and active agency of the doctor. (Puzzlingly, John also now maintains that, insofar as the purpose of this example is to shed light on our moral intuitions, it is a virtue that this example doesn’t settle this question.)

If you or anyone else can show how this makes the transplant example a bad objection to consequentialism, I’d be grateful.

52

John Quiggin 05.07.06 at 6:29 am

Mike, a good counterexample to a principle should not only show that the principle gives the wrong result, but should also give a clear indication what is wrong with the principle. The organ bank example fails this test, in my view, even if you overlook my more important objections to it.

By contrast, if the object is to get students thinking about their moral intuitions, then “cryptic” problems require more thought than clear ones, which is presumably a good thing in pedagogical terms.

53

bi 05.07.06 at 7:06 am

Artificial scenarios can constrict thought just as much as they can encourage it. Once more, supporters of torture like to bring up nonsensical scenarios involving Abu Muhammad al-Hitler the epitome of evil. It’s a borked, unreliable method, and the argument that There Is No Better Way just doesn’t wash (and it’s false anyway).

54

Mike Otsuka 05.07.06 at 7:50 am

a good counterexample to a principle should not only show that the principle gives the wrong result, but should also give a clear indication what is wrong with the principle

Surely it would be unwise to insist that a good counterexample to a principle must satisfy this criterion.

Example A might be well-suited as a counterexample to principle p because it clearly shows that this principle yields the wrong result. But it might be ill-suited as a test of which of competing principles q or r explains the right result. Example B, though ill-suited as a means of showing that principle p yields the wrong result, might be well-suited to adjudicate between principles q and r. There might, moreover, be no other example that is well-suited to perform both tasks simultaneously. Tailoring an example so that it tests well for one thing might ensure that it tests less well for another thing. That’s in the nature of controlled experments, including thought experiments. Example A would not reveal exactly why principle p is false, because it would not settle the question of whether principle p is false because it conflicts with true principle q or because it conflicts with true principle r. It would be bad to offer Example A and then to shut down one’s inquiry. One should go on to consider example B. But that fact wouldn’t make Example A a bad counterexample to principle p. That one needs two examples rather than a single one to get at the full truth is no objection to either example in the set of two.

55

abb1 05.07.06 at 8:48 am

Actually, I think it would be interesting to analyze the ‘ticking bomb’ scenario together with this ‘organ bank’ scenario and try to figure out why the intuition gives two exactly opposite answers to two seemingly similar questions.

It’ll probably be something along the lines of what Mike suggested in the other thread in regards to “the fat guy on the bridge” vs. “Big Jack”.

56

W. Kiernan 05.07.06 at 11:21 am

I, me, mine! I have way more than five times more right to my personal internal organs than all you bystanders. Even though you’re all successful, productive, well-beloved academics with two, three or four college degrees apiece, while I am an illiterate unwashed vagrant with fleas.

Anyway, how about let’s not talk about the simple binary abstractions “right” and “wrong” here. Let’s talk about this: how motivated are you fuckers to fight me in order to get hold of my liver? You and your gang may outnumber me, but I’m sure gonna break some bones and gouge out some eyeballs before I go down, in an effort to push your sigma(utility) down into the negatives. And you bitches best watch your backs too, ’cause if I see you moving on my neighbor with your steely grins and your shiny scalpels out you’re apt to meet up with my friend, Mr. Lou Slugger.

57

Tom Hurka 05.07.06 at 5:19 pm

Coming in late: excellent points by Mike Otsuka, especially in #55.

58

bi 05.08.06 at 3:35 am

I agree it’s not a counterexample’s job to explain why a principle does not hold. But if one’s to give a counterexample, let it be a real-life counterexample, instead of something with flying pigs, square circles, and ticking bombs from Abu Muhammad al-Hitler.

59

abb1 05.08.06 at 4:07 am

I think that while excessive abstraction and simplicity is a weakness of these hypothetical scenarious, what you’re likely to have with your real-life examples is excessive complexity.

Think of the McDonalds hot coffee incident, for example. If you use an one-liner version of it, the jury’s verdict seems counter-intuitive; but then, as you process more and more information about the case, you’ll begin to understand why they reached this verdict.

In the end you may find out that you haven’t really learned much about anything other than this very specific case. You just have to go abstract, at least to some extent.

Comments on this entry are closed.