The virtue ethicist endeavors to be the sort of person who doesn’t go around pushing fat men from bridges, and so recognizes it as a terrible, tragic situation.
It’s important when thinking about that thought experiment to picture yourself running up to the stranger, shoulder-checking him, wrapping your arms around him, feeling the fabric of his shirt press against your face and smelling his sweat. And then listen to him scream and feel his blood and brains get splattered all over your clothing.
The virtue ethicist, like most people, probably freezes and watches the whole thing unfold, or panics, or futilely tries to get the folks off the tracks before the trolley hits them. Do you expect an actual consequentialist human to do better?
As for the right thing to do, it’s probably to have better procedures for stopping people from being in the way of trolleys.
Do you expect an actual consequentialist human to do better?
Another interesting question is how all these consequentialitsts who insist that pushing the fat man is the right thing to do would react if they met someone who has actually followed their injunctions in practice. It seems to me that as soon as they’re out of the armchair, people’s inner virtue ethicist takes over, no matter how much their philosophy attempts to deny the relevance of his voice!
A real-world example would be a mountain climber who cut the rope that his partner was attached to, because if he didn’t, both people would have fallen and died. If I met a mountain climber who did that, I wouldn’t react negatively, any more than I would to someone who killed in self-defense.
That’s not a very good analogy. One could argue that by engaging in a mountain-climbing expedition, you voluntarily accept certain extraordinary risks, and the partner merely got unlucky with his own share of that risk. Whereas one of the essential premises in the fat man/trolley problem is that the fat man is a neutral passerby, completely innocent of the whole mess.
So, the real question is if you’d be so favorably inclined towards a mountain climber who, in order to save multiple lives, killed a completely unrelated random individual who was not at all entangled with their trouble.
That’s a good point. What about the following scenario: some crazy philosopher holds A and B at gunpoint and forces them to go mountain climbing. They do, and A starts to slip. B realizes he has to cut the rope or he’ll fall also. In this case, A didn’t voluntarily accept any risk. I’d still be favorably inclined to B.
Hm… according to my intuitions, this example features another important premise that is lacking in the original fat man/trolley problem—namely, a culprit who willingly and maliciously brought about the problematic situation. Going by my intuitive feeling, it turns out that in such scenarios, I’m much more inclined to look favorably at hard-headed consequentialist decisions by people caught in the mess against their will, apparently because I tend to place all the blame on the main culprit.
Note that this is just an impromptu report of my introspection, not an attempt at a coherent discussion of the issue. I’ll definitely need to think about this a bit more.
This is reminding me of some long discussions of “The Cold Equations”, a short story which is an effort to set up a situation where an ideally sympathetic person (pretty young woman with pleasant personality) has to be killed for utilitarian reasons.
The consensus (after decades of poking at the story) is that it may not be possible to rig the story to get the emotional effect and have it make rational sense.
I’m not absolutely certain about this—what if the girl had been the first stowaway rather than the nth, so that there wasn’t as good a reason to know that it shouldn’t be so easy for stowaways to get on ships?
If I remember correctly, she still would have died even if she hadn’t been jettisoned—the ship would have crashed and she would hardly walk away from that. That makes her unsalvageable. In standard trolley problems I don’t switch tracks, but if there were a way to switch the track so the train killed only one of the same five people it would already have killed, that person is unsalvageable and can be singled out to save the salvageable.
The SciFi Channel usually does a pretty poor job at making original movies, but their adaptation of “The Cold Equations” was pretty good, covering most of the problems with the original story. The pilot and the girl frantically look around for excess mass to jettison, and find some, but it’s not enough. The issue of what measures were taken to stop people from stowing away simply weren’t discussed; she’s there, and they have to deal with it. And at the last minute, the pilot does offer to sacrifice himself to save the girl, but she refuses to let him.
Ideal consequentialist would push the fat man in the standard trolley scenario. I was asking whether an ideal virtue ethicist does. It doesn’t matter (for me, now) that actual (if that means average) people, moral philosophers included, don’t always follow their principles. Neither it matters whether they recognise the situation as tragic and feel uneasy with all the blood and screams. I ask, what is the right thing to do under virtue ethics, when there are no available procedures better than pushing the fat man. And I find your answer a bit ambiguous.
(Disclaimer: My interest is purely theoretical. I don’t hold any definite position on what’s right in trolley scenario, and I would almost certainly not push the fat man, although I can imagine killing him in some less personal way.)
Ideal consequentialist would push the fat man in the standard trolley scenario. I was asking whether an ideal virtue ethicist does
You are confusing ethics and metaethics. Consequentialists, deontologists, and virtue ethicists all might or might not push the fat man, but they would all analyze the problem differently.
It’s not true that all possible consequentialists would push the fat man. A consequentialist might decide that one pushed death would be a worse consequence than X train deaths. Consequentialists don’t necessarily count the number of deaths and choose the smaller number; they just choose the option that leads to the best consequence.
This criticism is exactly right except that both the form question (rules, consequences or character traits) and the content question (pleasure, preference, the Categorical Imperative, Aristotle’s list, etc.) are part of normative ethics (what I assume you mean by ‘ethics’). Metaethical questions are things like “What are we doing when we use normative language?” and “Are there moral truths?”
OK, I should have said “typical consequentialist”. Of course a consequentialist may value the life of the fat man more than the sum of lifes of the people on the track, or find other consequences of pushing him down enough bad to refrain from it, or completely ignore humans and care about paperclips. I am not confusing ethics and metaetics, but rather assuming we are speaking about consequentialists with typical human values, for whom death is wrong and more deaths are more wrong, ceteris paribus. For such a consequentialist there may always be some critical number of people on the track whose common death will be worse than all consequences of pushing the fat man. On the other hand, deontologists typically hold that killing an innocent person is bad, and should, at least in theory, not push the man even if survival of the whole mankind was at stake. At least this is how I understand the difference between consequentialism and deontology.
Speaking about all possible consequentialists is tricky. Any moral decision algorithm can be classified as consequentialist when we try hard enough. I want to get an idea about what is the main difference between consequentialism and virtue ethics, given typical human values. The OP has said that they are the same except in bizarre situations like the trolley problem. So what is the difference in the trolley problem?
(If there is a consequentialist who disagrees with me and would not push the man even if it could save five billion lifes, let me know, ideally with some justification.)
assuming we are speaking about consequentialists with typical human values, for whom death is wrong and more deaths are more wrong, ceteris paribus.
I would question whether these are typical human values. People generally think the deaths of some people are more wrong than the deaths of other people. Most people do not value all human life equally. For typical humans ceteris almost never is paribus when it comes to choosing who lives and who dies.
At the risk of getting downvoted for nitpicking, I must point out that if you really insist on using Latin like this, the correct way to say it is: cetera almost never are pares.
Sorry, but the sight of butchered Latin really hurts my eyes.
That’s actually a matter where some interesting linguistic judgment might be in order.
The “ain’t” part is grammatical in some dialects of English, though, as far as I know, not in any form of standard English that is officially recognized anywhere. But the wrong cases for cetera and pares are not grammatical in any form of Latin that has ever been spoken or written anywhere.
On the whole, I’d say that “ain’t” is less bad, since in the dialects in which it is grammatical, it has the same form for both singular and plural. Therefore, at least it respects the number agreement with the Latin plural cetera, whereas “is” commits an additional offense by violating that agreement.
I sympathize with this logic, but I don’t completely agree. Languages frequently take words from other languages and regularize them, and when this occurs, they are no longer inflected the way they were in the original language. When we use Latin phrases in English often enough, they become part of the English language. ‘Ceteris’ and ‘paribus’ are in the ablative case because they were taken from a particular Latin expression, so it’s reasonable to keep them in that case when using the words in that context, even though they’re not being used in exactly the same way.
Yes, that’s a good point. Out of curiosity, I just searched for examples of similar usage in Google Books, and I’m struck by how often it can be found in what appear to be respectable printed materials. I guess I should accept that the phrase has been reanalyzed in English, just like it makes no sense to complain about, say, the use of caveat as a noun, or agenda as singular. (Though I still can’t help but cringe at singular data, despite being well aware that it’s a lost cause...)
Nitpick alert: You probably know this, but it’s an
important distinction that the non-plural usage of “data”
not only is grammatically singular, but is also a
mass noun.
(People say “I have some data, you have more data”, not *”I
have one data, you have two data[s]”.)
Datum is the neuter singular of the perfect passive participle of the Latin verb dare “to give.” This grammatical form is roughly analogous to the English participle “given.” However, in Latin, such participles are sometimes used as standalone nouns, so that the neuter form datum by itself can mean “[that which is/has been] given.” Analogously, the plural data can mean “[the things that are/have been] given.”
In English, this word has been borrowed with the meaning of “information given” and variations on that theme (besides a few additional obscure technical meanings).
I think of “ain’t” as either standard in some dialects, or as a tool for emphasis in standard English (usually spoken rather than written).
It seems reasonable that if you’re using informal English for emphasis, then it’s stylistically consistent to use the sort of colloquial mangled Latin that an English speaker who doesn’t know Latin would use.
The word ain’t can be used in both speech and writing to
catch attention and to give emphasis, as in “You ain’t
seen nothing yet,” or “If it ain’t broke, don’t fix it.”
Merriam-Webster’s Collegiate Dictionary gives an example
from film critic Richard Schickel: “the wackiness of
movies, once so deliciously amusing, ain’t funny anymore.”
Apologies, the only Latin I remember from school is Caecilius est in horto. I actually spent several minutes with Google trying to figure out what it should be but there appears to be a shortage of online Latin translation services. Gap in the market?
One problem is that such a service is in much less demand compared to the living languages currently supported by translation programs. However, another major difficulty is that Latin is a far more synthetic language than English, and its inflectional suffixes often carry as much information as multiple-word clauses in English. For example, the mentioned ceteris paribus packs the entire English phrase “with everything else being the same” into just two words. Similarly, the last word in quod erat demonstrandum (a.k.a. “QED”) packs the last four words of the English “that which was supposed to be demonstrated” into one. This makes it much harder to come up with satisfactory translation heuristics compared to more analytic languages, especially considering the extreme freedom of word order in Latin.
Similar difficulties, of course, exist in automatic translation of English to other highly synthetic languages, like e.g. the Slavic ones.
I am clearly unable to express myself clearly today.
I haven’t said that it’s typical to value all life equally. I tried to say that set X of x deaths is typically worse than set Y of y deaths, if x>y. Almost always it holds when Y is a subset of X (that was the intended meaning of ceteris paribus), but if x>>y, it often holds even if the sets are disjoint.
Also, the context of the trolley scenario is that the fat man isn’t your relative or friend; he’s a random stranger, fully comparable with those on the track.
The virtue ethicist endeavors to be the sort of person who doesn’t go around pushing fat men from bridges, and so recognizes it as a terrible, tragic situation.
It’s important when thinking about that thought experiment to picture yourself running up to the stranger, shoulder-checking him, wrapping your arms around him, feeling the fabric of his shirt press against your face and smelling his sweat. And then listen to him scream and feel his blood and brains get splattered all over your clothing.
The virtue ethicist, like most people, probably freezes and watches the whole thing unfold, or panics, or futilely tries to get the folks off the tracks before the trolley hits them. Do you expect an actual consequentialist human to do better?
As for the right thing to do, it’s probably to have better procedures for stopping people from being in the way of trolleys.
thomblake:
Another interesting question is how all these consequentialitsts who insist that pushing the fat man is the right thing to do would react if they met someone who has actually followed their injunctions in practice. It seems to me that as soon as they’re out of the armchair, people’s inner virtue ethicist takes over, no matter how much their philosophy attempts to deny the relevance of his voice!
A real-world example would be a mountain climber who cut the rope that his partner was attached to, because if he didn’t, both people would have fallen and died. If I met a mountain climber who did that, I wouldn’t react negatively, any more than I would to someone who killed in self-defense.
That’s not a very good analogy. One could argue that by engaging in a mountain-climbing expedition, you voluntarily accept certain extraordinary risks, and the partner merely got unlucky with his own share of that risk. Whereas one of the essential premises in the fat man/trolley problem is that the fat man is a neutral passerby, completely innocent of the whole mess.
So, the real question is if you’d be so favorably inclined towards a mountain climber who, in order to save multiple lives, killed a completely unrelated random individual who was not at all entangled with their trouble.
That’s a good point. What about the following scenario: some crazy philosopher holds A and B at gunpoint and forces them to go mountain climbing. They do, and A starts to slip. B realizes he has to cut the rope or he’ll fall also. In this case, A didn’t voluntarily accept any risk. I’d still be favorably inclined to B.
Hm… according to my intuitions, this example features another important premise that is lacking in the original fat man/trolley problem—namely, a culprit who willingly and maliciously brought about the problematic situation. Going by my intuitive feeling, it turns out that in such scenarios, I’m much more inclined to look favorably at hard-headed consequentialist decisions by people caught in the mess against their will, apparently because I tend to place all the blame on the main culprit.
Note that this is just an impromptu report of my introspection, not an attempt at a coherent discussion of the issue. I’ll definitely need to think about this a bit more.
This is reminding me of some long discussions of “The Cold Equations”, a short story which is an effort to set up a situation where an ideally sympathetic person (pretty young woman with pleasant personality) has to be killed for utilitarian reasons.
The consensus (after decades of poking at the story) is that it may not be possible to rig the story to get the emotional effect and have it make rational sense.
I’m not absolutely certain about this—what if the girl had been the first stowaway rather than the nth, so that there wasn’t as good a reason to know that it shouldn’t be so easy for stowaways to get on ships?
If I remember correctly, she still would have died even if she hadn’t been jettisoned—the ship would have crashed and she would hardly walk away from that. That makes her unsalvageable. In standard trolley problems I don’t switch tracks, but if there were a way to switch the track so the train killed only one of the same five people it would already have killed, that person is unsalvageable and can be singled out to save the salvageable.
You’re right.
The SciFi Channel usually does a pretty poor job at making original movies, but their adaptation of “The Cold Equations” was pretty good, covering most of the problems with the original story. The pilot and the girl frantically look around for excess mass to jettison, and find some, but it’s not enough. The issue of what measures were taken to stop people from stowing away simply weren’t discussed; she’s there, and they have to deal with it. And at the last minute, the pilot does offer to sacrifice himself to save the girl, but she refuses to let him.
Ideal consequentialist would push the fat man in the standard trolley scenario. I was asking whether an ideal virtue ethicist does. It doesn’t matter (for me, now) that actual (if that means average) people, moral philosophers included, don’t always follow their principles. Neither it matters whether they recognise the situation as tragic and feel uneasy with all the blood and screams. I ask, what is the right thing to do under virtue ethics, when there are no available procedures better than pushing the fat man. And I find your answer a bit ambiguous.
(Disclaimer: My interest is purely theoretical. I don’t hold any definite position on what’s right in trolley scenario, and I would almost certainly not push the fat man, although I can imagine killing him in some less personal way.)
You are confusing ethics and metaethics. Consequentialists, deontologists, and virtue ethicists all might or might not push the fat man, but they would all analyze the problem differently.
It’s not true that all possible consequentialists would push the fat man. A consequentialist might decide that one pushed death would be a worse consequence than X train deaths. Consequentialists don’t necessarily count the number of deaths and choose the smaller number; they just choose the option that leads to the best consequence.
This criticism is exactly right except that both the form question (rules, consequences or character traits) and the content question (pleasure, preference, the Categorical Imperative, Aristotle’s list, etc.) are part of normative ethics (what I assume you mean by ‘ethics’). Metaethical questions are things like “What are we doing when we use normative language?” and “Are there moral truths?”
Thanks for the correction: I didn’t realize that. Are there better terms for expressing the difference between form and content in ethics?
Not that I know of, I’m afraid. In fact, I may have invented the form and content language.
OK, I should have said “typical consequentialist”. Of course a consequentialist may value the life of the fat man more than the sum of lifes of the people on the track, or find other consequences of pushing him down enough bad to refrain from it, or completely ignore humans and care about paperclips. I am not confusing ethics and metaetics, but rather assuming we are speaking about consequentialists with typical human values, for whom death is wrong and more deaths are more wrong, ceteris paribus. For such a consequentialist there may always be some critical number of people on the track whose common death will be worse than all consequences of pushing the fat man. On the other hand, deontologists typically hold that killing an innocent person is bad, and should, at least in theory, not push the man even if survival of the whole mankind was at stake. At least this is how I understand the difference between consequentialism and deontology.
Speaking about all possible consequentialists is tricky. Any moral decision algorithm can be classified as consequentialist when we try hard enough. I want to get an idea about what is the main difference between consequentialism and virtue ethics, given typical human values. The OP has said that they are the same except in bizarre situations like the trolley problem. So what is the difference in the trolley problem?
(If there is a consequentialist who disagrees with me and would not push the man even if it could save five billion lifes, let me know, ideally with some justification.)
I would question whether these are typical human values. People generally think the deaths of some people are more wrong than the deaths of other people. Most people do not value all human life equally. For typical humans ceteris almost never is paribus when it comes to choosing who lives and who dies.
At the risk of getting downvoted for nitpicking, I must point out that if you really insist on using Latin like this, the correct way to say it is: cetera almost never are pares.
Sorry, but the sight of butchered Latin really hurts my eyes.
I had a teacher once who liked to say “ceteris ain’t paribus”. Is that better or worse?
That’s actually a matter where some interesting linguistic judgment might be in order.
The “ain’t” part is grammatical in some dialects of English, though, as far as I know, not in any form of standard English that is officially recognized anywhere. But the wrong cases for cetera and pares are not grammatical in any form of Latin that has ever been spoken or written anywhere.
On the whole, I’d say that “ain’t” is less bad, since in the dialects in which it is grammatical, it has the same form for both singular and plural. Therefore, at least it respects the number agreement with the Latin plural cetera, whereas “is” commits an additional offense by violating that agreement.
I sympathize with this logic, but I don’t completely agree. Languages frequently take words from other languages and regularize them, and when this occurs, they are no longer inflected the way they were in the original language. When we use Latin phrases in English often enough, they become part of the English language. ‘Ceteris’ and ‘paribus’ are in the ablative case because they were taken from a particular Latin expression, so it’s reasonable to keep them in that case when using the words in that context, even though they’re not being used in exactly the same way.
Yes, that’s a good point. Out of curiosity, I just searched for examples of similar usage in Google Books, and I’m struck by how often it can be found in what appear to be respectable printed materials. I guess I should accept that the phrase has been reanalyzed in English, just like it makes no sense to complain about, say, the use of caveat as a noun, or agenda as singular. (Though I still can’t help but cringe at singular data, despite being well aware that it’s a lost cause...)
Nitpick alert: You probably know this, but it’s an important distinction that the non-plural usage of “data” not only is grammatically singular, but is also a mass noun. (People say “I have some data, you have more data”, not *”I have one data, you have two data[s]”.)
Virtually everyone who makes “data” grammatically plural actually uses it as a mass noun, too.
...so what’s “datum”, then?
Datum is the neuter singular of the perfect passive participle of the Latin verb dare “to give.” This grammatical form is roughly analogous to the English participle “given.” However, in Latin, such participles are sometimes used as standalone nouns, so that the neuter form datum by itself can mean “[that which is/has been] given.” Analogously, the plural data can mean “[the things that are/have been] given.”
In English, this word has been borrowed with the meaning of “information given” and variations on that theme (besides a few additional obscure technical meanings).
It’s the singular that plural “data” is a plural of. Someone who strictly uses “data” as a mass noun would say “piece of data”.
I think of “ain’t” as either standard in some dialects, or as a tool for emphasis in standard English (usually spoken rather than written).
It seems reasonable that if you’re using informal English for emphasis, then it’s stylistically consistent to use the sort of colloquial mangled Latin that an English speaker who doesn’t know Latin would use.
Wikipedia:
(Which is exactly how it’s used in “ceteris ain’t paribus”. See also this post by Geoff Nunberg.)
Apologies, the only Latin I remember from school is Caecilius est in horto. I actually spent several minutes with Google trying to figure out what it should be but there appears to be a shortage of online Latin translation services. Gap in the market?
One problem is that such a service is in much less demand compared to the living languages currently supported by translation programs. However, another major difficulty is that Latin is a far more synthetic language than English, and its inflectional suffixes often carry as much information as multiple-word clauses in English. For example, the mentioned ceteris paribus packs the entire English phrase “with everything else being the same” into just two words. Similarly, the last word in quod erat demonstrandum (a.k.a. “QED”) packs the last four words of the English “that which was supposed to be demonstrated” into one. This makes it much harder to come up with satisfactory translation heuristics compared to more analytic languages, especially considering the extreme freedom of word order in Latin.
Similar difficulties, of course, exist in automatic translation of English to other highly synthetic languages, like e.g. the Slavic ones.
I am clearly unable to express myself clearly today.
I haven’t said that it’s typical to value all life equally. I tried to say that set X of x deaths is typically worse than set Y of y deaths, if x>y. Almost always it holds when Y is a subset of X (that was the intended meaning of ceteris paribus), but if x>>y, it often holds even if the sets are disjoint.
Also, the context of the trolley scenario is that the fat man isn’t your relative or friend; he’s a random stranger, fully comparable with those on the track.