I was browsing some of the blogs over at Experimental Philosophy the other day, and it got me thinking about something weird that has happened in every ethics course I have taught to date. The class will be discussing hedonism (about welfare), and I present Nozick’s Experience Machine thought experiment in an argument against this view.
For the uninitiated, hedonism about welfare claims (roughly) that the only intrinsically valuable thing is pleasure, and the only intrinsically disvaluable thing is pain; a person’s life, therefore, goes better for her the greater the balance of pleaure over pain there is in that life. Robert Nozick argued against such a view in his Anarchy, State, and Utopia by asking us to imagine The Experience Machine (EM). This is a machine, created by super-duper scientists, that can give a person any experiences she desires. These experiences are indistinguishable from veridical experiences (i.e. experiences that are obtained by interacting in the real world in the normal way rather than triggered in our brains by a machine). Imagine that you were looking for a way to get the best possible life for yourself (the life with the most welfare value for you). Would you hook into the machine?
If hedonism were correct, the answer should be yes. However, almost all of my students say they would not hook into the machine. I am sure that anyone who has taught a course on ethics has had a similar experience. But this is where things start to get weird.
In every class I had some students who thought that the experience machine does promise the best possible life for them in terms of welfare value. So I tried to develop another case that I thought would be even more convincing than the Experience Machine. I presented them with The Duplicitous Significant Other (DSO):
Compare the following two possible worlds:
In world1, you are in love with a person, A, and A loves you back. You have a variety of experiences with A, and these experiences make you extremely pleased.
In world2, you are in love with A, but A only pretends to love you back. In this world, however, A hates you. A puts up with you only because you buy A things. A cheats on you on a regular basis, but you never catch on to this. In fact, the experiences you have in this world are identical to the experiences you have in world1.
To my shock and dismay, every time I present this case to the class, I find that the majority of students believe that world1 and world2 are identical with respect to the welfare value for you contained in them.
Admittedly, these results are anecdotal and not very scientific. However, the trend has always been the same: most students would not hook into the EM, and yet most students think that world1 and world2 are equally good for me.
When presented with the DSO, students always respond, “But what you don’t know cannot hurt you”. I then ask them why the EM will not provide them with the best possible life: certainly, the fact that they are in the EM is something they will not know and so should not affect their welfare. It is tempting to explain these results by pointing out that our students are not always consistent, and may not be able to see that there is a tension in their responses. But I am not sure that this explanation is correct anymore. I have come to the conclusion that the EM is not a good test of what people actually believe about value, and that the DSO is.
The problem with the EM is, I think, its strangeness. For those of us who entertain thought experiements that may be even weirder than this on a day-to-day basis, it is easy to forget just how unusual is the question we are asking our students. When confronted with the choice of hooking into the machine or not, they may not be able to keep certain thoughts out of their minds: for example, what if the machine breaks? What if there is a power failure? What if the program I choose is really bad? We are able to bracket these concerns and focus on the point of the question: if the machine were perfect, and power failures never happened, and the scientisits could match the program with your current desires, etc., then would the EM be best for you? I believe that non-philosophers cannot do this bracketing. They may try, but they fail. So when confronted with the choice, students opt not to hook into the machine, but this is not because they think something other than pleasure is valuable. It is because they are allowing distorting influences to affect their answer. [It is not just students who claim that the question is too strange to be helpful. L. W. Sumner argues that it is in “Welfare, Happiness, and Pleasure” Utilitas 4 (1992).]
But in the DSO, the question being asked is not so strange. In effect, it is a question any person who has ever been in a relationship asks themselves: am I harmed if my SO cheats on me and I never find out about the infidelity (or, most likely for some of my students, the question they ask is: am I harming my SO when I cheat on him/her)? The DSO just takes this question to the extreme: what if your SO always cheated on you, did not love you, etc? There is some bracketing required here (e.g. can someone who hates me really act towards me in the exact same way as someone who loves me?) but it is less severe, less strange.
I still present students with both scenarios, but I am contemplating dropping the EM from my classes. The EM is distracting for students and has them focus on things other than their views about value. Furthermore, the DSO has an added advantage: those students who claim that “what you do not know cannot hurt you” believe that if you were to find out about the cheating, you would be harmed. I can ask them to explain why it is a harm to learn of this fact. Many students begin to see that the best explanation is that the cheating was harmful even before they knew about it, which is, of course, the point I wanted to make.
I am curious to see what you think: is the EM a worthwhile thought experiement for our students, or is it too weired to be of help? Which example allows people better to reveal their true beliefs about value? Is it worth keeping in our classes?
Scott,
As I read your post, an alternative explanation came to mind.
About the Experience Machine, your question is one of *choice*: “Would you enter the EM?”
But in Duplicitous Significant Other, your question is one of *value*: “Are your worse off in the DSO-world?”
I think this difference might account for the difference in your students’ answers. If you asked your students not, “Would you enter the EM?”, but rather, “Are you worse off in the EM?”, you might get a different answer. Likewise, if you asked your students not “Are your worse off in the DSO-world?”, but rather, “Would you choose the DSO-world?”, you might also get a different answer.
For my part, I think I agree with those who would say “No” to entering the EM but who agree that one is no worse off in the DSO-world. But this is not because the EM thought experiment is too far-fetched. Indeed, I accept that I would be no worse off in the EM either. What explains my refusal to enter the EM is that *I care about more than just welfare*. I prefer that I not be deceived, even though I am no worse off as a result of it. I prefer really to accomplish things, even though I wouldn’t know the difference in the EM.
P.S. I do notice that you pose the EM question with the constraint that “you were looking for a way to get the best possible life for yourself (the life with the most welfare value for you).” I think this does effectively change the question from one of choice to one of value. But perhaps it is easy for students (as it was for me) to forget about this constraint when answering the EM question about choice.
I agree with Chris and had the same idea about choice and value. It’s quite possible that the students don’t like the EM choice for two reasons. Firstly, they are being asked to consciously decieve themselves which they might be reluctant to do. If they don’t know they are in the EM, they are being asked to choose a lie. If they know they are in the EM, it is of course impossible to enjoy it as much since you know your pleasures are inauthentic (which is why Prozac is not the same as being genuinely happy).
With the DSO world students will simply feel they are being asked to distinguish the two experiences, which will be identical.
BTW I’m not sure your point about the DSO is correct. I agree that the harm that is done by the cheating spouse originates in the acts of lying and infidelity but nevertheless, for the harm to be actualised requires the cheated upon to be aware of it. The misdemeanors are necessary but not sufficient conditions of the harm.
1. In support of Chris’s point about choice and value, I often have students offering other-regarding reasons for not hooking up to the EM. They say “sure, I would be having a wonderful time hooked up to the EM, but I’d also be neglecting my family and friends, failing to contribute to society, etc.” In order to block such considerations I usually resort to that good old philosopher’s prop: the desert island. “Imagine you’re isolated form the rest of the world on a desert island, and you come across an EM … ”
2. Regarding your comments about finding out that your duplicitous partner has been cheating on you, we can distinguish two questions:
(1) Are you harmed by your partners cheating?
(2) If you are harmed, when are you harmed?
Suppose we say, in answer to (1), that you’re harmed only if you find out about the cheating. So far as I can tell, it’s consistent with this to say, in answer to (2), that you were harmed all along (i.e., for the full duration of the cheating, including times prior to your finding out). That is, the following seems a consistent position: if you find out (at some time), then you were harmed all along; whereas, if you never find out, then you were never harmed.
Scott,
I think that the EM is pedagogically useful, if only to isolate the issue between hedonists (and other subjectivists) about value from their objectivist opponents. I do think it helps students begin to ask questions about hedonism which do not tend to strike them naturally. Students have trouble critically scrutinizing hedonism largely because they have trouble distinguishing two possible roles pleasure, or other psychological states, might play in a theory of value, namely, pleasure as evidence for something’s being valuable and pleasure as a thing’s value. (These have reverse directions of explanation.) In my experience, the EM gets students to ask ‘what, if anything, does the EM leave out of what we value?’ Once students appreciate that question, they can at least begin to entertain doubts about a purely subjective theory of value.
But as for its utility in getting students to probe their own understandings of value, I’m not so optimistic. I agree with previous posts that the difference between the EM situation being chosen and the DSO worlds not being chosen might be important to many people, our students included. Perhaps the thought is this: Suppose that from a purely experiential standpoint, the EM is indistinguishable from actually having the corresponding EM experiences. Likewise, DSO w1 is experientially indistinguishable from DSO w2. Perhaps someone might think that (a) the EM is worse than the real world EM-equivalent because of whatever (non-experiential) factor is lacking in the EM, but (b) DSO w2 is no worse than DSO w1, though DSO w2 presumably lacks that same non-experiential factor that made EM worse than its real world equivalent. Maybe (a) and (b) can be made consistent if we say that EM is worse because the agent chose to divorce herself from this non-experiential factor, whereas DSO w2 is no worse than w1 because the non-experiential factor (that the DSO is actualy cheating on you) was not chosen by the agent. I’m not sure that this amounts to much more than ‘ignorance is bliss’, and of course, the EM could be described so that the agent who has entered it is no longer aware of having chosen to enter it.
I also wonder if, incidentally, our students might not be intuitively turning the logical relationship between welfare and morality around. That is, perhaps they think that one is worse off, all other things being equal, if one is a victim of a moral wrong. And if we add that one must know that one is such a victim in order for it to be a wrong, then we might be able to account for the view the consistency of (a) and (b). Just a thought.
Campbell,
I’m not sure that I agree with your assertion that if you are harmed by the cheating at any time, you are harmed from the moment the cheating began. I think the inuition comes from an analogy with physical harm. In the case of physical harm, you are harmed the moment you are inflicted with some physical ailment whether you are aware of it or not. I’m not sure that emotional harm works in the same way. With emotional harm you really do need to find out about whatever it is to be harmed. With physical harm you can be unaware of it and it still makes sense to say you were harmed.
What I’m getting at is that finding out about the cheating is every bit as much a component of the harm as the cheating itself. Without the knowledge there is no harm, therefore, the harm must be done at the moment of revelation even though the possibility of that revelation occurred earlier.
If students are in fact more capable of “bracketing” in the case of the DSO than they are in the case of the EM, I’m suprised. It seems to me that the closer to one’s actual experience an example is, the more difficult it is to distinguish from one’s actual experience. Since students have never encountered anything like the EM, I would think they would be willing to believe just about anything you tell them about it. After all, nothing you say will be any more plausible or natural than anything else. On the other hand, since students have probably had significant others, and might even have been cheated on, I would think it would be difficult for them to set aside the details of their own experience and deal strictly with the rules of your hypothetical case.
Off the cuff, it would appear that in the EM case one knows that experience is not veridical. (At least, students could get that impression.) Whereas in the DSO case you don’t know this. And on the principle that “what you don’t know can’t hurt you” the cases are asymmetrical.
I teach an introductory political philosophy case and ask the EM differently. Instead, the question is whether Cypher is wrong to want back into the Matrix (the movie) — the ignorance is bliss Prisoner in the Cave view. You might try this approach and see if your results are any different.