xkcd

by **SpiritOfRock** » Fri Mar 16, 2012 3:43 pm UTC

Do the ends justify the means?

Imagine a commonly used thought experiment to express the concept. Say there was a train, five passengers on board, traveling along a track headed towards a pit of fire. You hold the switch to change the course of the train to a different track, which does not lead to direct harm for the passengers. Should you pull the switch?

Most people's answer to this is yes, it is moral to save lives. That's a good start, unless your answer is no, it is immoral to save lives, in which case you're a sociopath and should contact a psychiatrist.

Now, imagine on the alternative track is a person, tied up on the tracks. If you pull the switch, he'll be run over. Should you pull the switch?

Most people will still say yes, five lives are worth one. Some will say that it is immoral for an individual to pull the switch, because then it is his action that ends a life, rather than five lives ending due to circumstance. To illustrate the second point, if there were a person tied to the track and only one person was aboard the train, almost nobody would pull the switch (although some would inquire about the demographics of the people involved). They would simply let the train take its course, as it's better for a life to end due to circumstance than by their will.

Then next part is a bit tricky. Suppose there is no alternate track or switch, but instead a very fat man standing in between you and the track. The man is heavy enough to stop the path of the train in a head-on colission. Is it moral to push him into the train?

I recognize the question is absurd as no combination of a train and a human could realize this situation, but try and understand the principle behind it. While most people would be content merely making a decision that ends a life and saves five, when the decision involves personally murdering someone with your own two hands, almost nobody is willing to make the decision which is equivalent in terms of lives saved and lost.

Even ignoring the instinctive repulsion to shoving a person into a train, you might say that it would be immoral for an individual in general to push a person into a train to save five lives. If so, you believe that the ends do not justify the means. On the other hand, if you would have pushed a person into the train to save five lives, you believe that the ends do justify the means. The principle is, does committing an immoral action become moral if the consequence is favorable to the alternative? Favorable, of course, in this case, means more living people (unless you really like trains).

Now, let's imagine something even more absurd than everything I've just asked you to think about. Let's imagine that instead of a person able to push a fat man into the train, there is a robot. The robot can make decisions, but does not have feelings or emotions. Should it push the fat man? Think about it.

If your answer was no to the human shoving the fat man, but yes to the robot shoving the fat man, your opinion implies that the ends do not justify the means because of the consequence the means have on the mean-er. If a person shoves a fat man into a train, regardless of the lives he's saved, he will have to live with the memory of murdering a human being for the rest of his life. A robot will have simply done what it was programmed to do, which is presumably to seek the most morally favorable outcome for its surroundings.

The phrase "ends justify the means" is misleading. The only "end" you're referring to is the end of some abstract event you separated from the rest of spacetime in your imagination. In reality, time continues to go on, and will continue to be affected by your decision in more ways than you'll ever be able to see or imagine. Taking this into consideration, it would be morally shortsighted to think only about the near consequences of your choice. The longterm should also be considered, which involves mainly the lives of the people affected. Passengers aboard a train may not be very happy to know that a human life had to be sacrificed in order to save their lives, while a fat man probably isn't going to lose too much sleep over not being thrown into a train to save five lives.

If you are an individualist, then what matters more than these things is the effect your decision will have on yourself and your life thereafter. You are probably not going to have nightmares about your refusal to kill a fat man in order to save lives, but you are liable to have nightmares about the opposite. With the improvement of your character as the moral purpose of your life, you may decide not to push the fat man because such an action would corrupt your character.

Stepping away from the train track, think about the kinds of people who would have pushed the fat man. Are their characters compromised or damaged in any way? Do you think sane, civilized and rational human beings would push the fat man into a train? What about after the fact? Do you think their decision to push the fat man might compromise or damage their character?

Regardless of your perspective on this thought experiment, one thing is for certain: Robots are bound to kill us all, and it will be for the greater good.

by **Роберт** » Fri Mar 16, 2012 4:29 pm UTC

Some people like to consider the responsibility* of the people. If you're on the train tracks or in the train you're generally considered more responsible* than if you're just a random "fat man" (kinda dumb hypothetical) who is nearby.

*Not quite the right term, but I'm talking about involvement in the situation already. Most people would find it immoral to kill an uninvolved healthy person to harvest their organs and use as organ donations on lots of otherwise healthy people to save their lives, even though, in the end, more people live longer.

by **Bharrata** » Fri Mar 16, 2012 6:24 pm UTC

If a person shoves a fat man into a train, regardless of the lives he's saved, he will have to live with the memory of murdering a human being for the rest of his life. A robot will have simply done what it was programmed to do, which is presumably to seek the most morally favorable outcome for its surroundings.

Wouldn't the robot's programmers have to live with the fact that they've unleashed a Neutral Good monster upon the world?

Would their guilt be greater because of the effects of their extrapolating imagination, or lesser because they did not have to physically witness the consequences?

Taking this into consideration, it would be morally shortsighted to think only about the near consequences of your choice. The longterm should also be considered, which involves mainly the lives of the people affected.

I'm glad you brought this up, as most people forget about the time dimension of Utilitarian calculations when engaging in ethical debates.

If you are an individualist, then what matters more than these things is the effect your decision will have on yourself and your life thereafter.

Well, isn't this the heart of the issue? An individualist in this case is operating as if they are an actor/thing that can be separated from the rest of existence, which thereby causes them to give their own individual self more value than other selves (or "the Other") when they create the framework through which they perceive the world and decide to interact with it.

It doesn't/shouldn't matter what the subject of your focus decides to do, but rather what the fat man in question decides to do. The subject is not presented with a choice, but rather the fat man, and he, as a separate autonomous individual is the one with the responsibility to act or not to act.

The idea that individuals are entirely free to make decisions for other individuals (at least when it comes to ethical dilemmas) is inherently contradictory. To paraphrase or terribly misquote Emmanuel Levinas: "The desire for freedom is the spark of tyranny."

Further, I don't much see the difference in the severity of trauma from witnessing a fat man get hit by a train (or causing it) and save five lives or watching the fat man live while witnessing five people die in a horrible train accident.

Do you think sane, civilized and rational human beings would push the fat man into a train?

Loaded question is loaded. :cry:

by **UniqueScreenname** » Fri Mar 16, 2012 6:42 pm UTC

I think a lot of the decisions made depends on the person's ego. A lot of people will not feel bad for not doing anything in their power to save the lives of people that they don't know from an accident they were not responsible for. A person who feels the need to be a hero would be more willing to make these decisions than someone who would be able to sleep at night thinking there was nothing they could have done.

At first I thought that we were hypothetically on the train also, that the switch to change tracks was for some reason on the train. While absurd, it brings up another factor: self-preservation. Would you be willing to kill a person tied to the tracks to save five people and yourself? Many people may consider it a necessary evil. IMO very few would be so altruistic (don't know if that's the word I'm looking for) to sacrifice themselves over a blip in their conscience.

by **Bharrata** » Fri Mar 16, 2012 7:22 pm UTC

UniqueScreenname wrote:I think a lot of the decisions made depends on the person's ego.

Making a decision says nothing about whether the decision was ethical or unethical, and depending on your definition of ego - are we defining it psychoanalytically or colloquially? - if it's the latter I'm more agreeable with the assumption that it would cause the individual to make the wrong choice.

IMO very few would be so altruistic (don't know if that's the word I'm looking for) to sacrifice themselves over a blip in their conscience.

This type of criticism is why I've always found these ethical thought experiments to be poorly constructed and not quite capable of making clear the ethical questions they're attempting to raise. (Note: I'm talking broadly about the "ohemgee, strangers on a train" scenario that is thrown around in undergrad Intro to Ethics rather than something more interesting and useful like the Prisoner Dilemma)

edit: I will admit the OP does come at the thought experiment in a relatively more novel way.

by **Роберт** » Fri Mar 16, 2012 9:42 pm UTC

If we aren't already harvesting usable blood and organs from people who are dead and using them to save other people, I see no reason to kill living people to save other people.

by **PeterCai** » Sat Mar 17, 2012 7:37 am UTC

Роберт wrote:If we aren't already harvesting usable blood and organs from people who are dead and using them to save other people, I see no reason to kill living people to save other people.

Appeal to tradition is not a legitimate argument. Why is harvesting organs from dead people bad?

by **Technical Ben** » Sun Mar 18, 2012 7:11 pm UTC

If it was a real life instance (and thus one not limited by the rules of the thought experiment) I would ask the person what their choice was.
Who knows, perhaps they're a hero, and would save the train and it's passengers even at their own expense.

In which case, no one was pushed and I helped save the train. If I was the person able to stop the train myself, without involving anyone else, then I'd happily give up my own safety to save the passengers.

Plus as has been said, without consent of the persons involved, would it not still class as murder? At least in it's legal definition.

[Edit]
Oh, and if I caught the hint correctly, and this is about the ending of ME3? Then I always choose the unlisted options in stories I see as illogical. I've not played it, but when the "Killer Robots" asked "how should we destroy the universe" I'd reply "You decide!" and hope for a critical exception fault (they left no option for that in ME3, so I assume it's not in the Reapers programming!

) .

by **SunAvatar** » Mon Mar 19, 2012 5:44 am UTC

Re. the "push the fat guy in front of the train" thought experiment:

If you present the situation in a vacuum, I'll interpret it in a vacuum. That is, I'll assume you're postulating that
1. I know the train's five passengers will die if I do nothing;
2. I know this guy is heavy enough to stop the train;
3. I know I am capable of pushing him just so to get him onto the tracks in such a way as to stop the train;
4. I know that the passengers won't die if I do this;
5. I can't think of another plan to stop the train in time to actually carry it out.

If I accept all those postulates, then sure, that guy's getting pushed. The reason this seems morally counterintuitive is probably because our moral intuitions are developed around situations that aren't completely impossible.

How in the world could someone ever justify being certain of all these things? Particularly #2 since it isn't true, but #3 and #4 are quite iffy as well, and #5 is deeply suspicious. It's usually considered a cop-out to challenge the assumptions of a hypothetical question like this, because it seems like one is trying to avoid the choice. But in the absence of a source of perfect knowledge about the world, challenging assumptions and looking for third alternatives is a really good idea.

In practice I think, in a case where you are about to sacrifice the lives of others for the greater good, even a diehard act utilitarian would suggest there is nearly as strong a chance that you have miscalculated as there is that sacrificing lives is really the best plan, and demand very strong evidence that you are not crazy or otherwise impaired.

by **Bharrata** » Mon Mar 19, 2012 7:04 am UTC

SunAvatar wrote:How in the world could someone ever justify being certain of all these things? Particularly #2 since it isn't true, but #3 and #4 are quite iffy as well, and #5 is deeply suspicious. It's usually considered a cop-out to challenge the assumptions of a hypothetical question like this, because it seems like one is trying to avoid the choice. But in the absence of a source of perfect knowledge about the world, challenging assumptions and looking for third alternatives is a really good idea.

In practice I think, in a case where you are about to sacrifice the lives of others for the greater good, even a diehard act utilitarian would suggest there is nearly as strong a chance that you have miscalculated as there is that sacrificing lives is really the best plan, and demand very strong evidence that you are not crazy or otherwise impaired.

And it all comes back to epistemology, perhaps the more relevant question is, do we know the ends of our means?

Hume would have had a field day with chaos theory. :twisted:

by **TheGrammarBolshevik** » Mon Mar 19, 2012 1:30 pm UTC

I think it would be helpful, assuming we don't want to just turn this thread into trivial "Well like nothing is certain man," to just suppose that all the assumptions of the thought experiment are known and ask what would be the right thing to do if we were to have all the relevant information. After all, it's not like there aren't analogous situations that deal away with these worries (the unsuspecting patient, for example).

by **Роберт** » Mon Mar 19, 2012 2:03 pm UTC

PeterCai wrote:
Роберт wrote:If we aren't already harvesting usable blood and organs from people who are dead and using them to save other people, I see no reason to kill living people to save other people.

Appeal to tradition is not a legitimate argument. Why is harvesting organs from dead people bad?

It's not. But it's senseless to refuse to harvest organs without permission, thereby letting people die, and than actively kill someone to keep from letting people die.

I think if I thought about it I would be for mandatory organ donation of dead people, but would not agree to killing an uninvolved third party to save people.

by **jules.LT** » Mon Mar 19, 2012 2:22 pm UTC

The way our minds work accounts for uncertainty, and uncertainty supposedly doesn't exist in these scenarios.
That's a pretty major flaw, when discussing whether you'd commit murder in those situations.

Роберт wrote:I think if I thought about it I would be for mandatory organ donation of dead people, but would not agree to killing an uninvolved third party to save people.

You might want to check out this thread: Mandatory Organ Donation.

Making organ donation opt-out rather than opt-in is enough to make a huge difference in organ supply.

by **TheGrammarBolshevik** » Mon Mar 19, 2012 3:17 pm UTC

jules.LT wrote:The way our minds work accounts for uncertainty

Sure. Like, I'm "uncertain" whether the world is real, so I'm "uncertain" whether I'll find my hallway or the gaping mouth of Cthulhu when I open my door in a minute. But the kind of certainty that makes reasoned action impossible? We clearly don't have that.

by **jules.LT** » Mon Mar 19, 2012 3:37 pm UTC

TheGrammarBolshevik wrote:
jules.LT wrote:The way our minds work accounts for uncertainty

Sure. Like, I'm "uncertain" whether the world is real, so I'm "uncertain" whether I'll find my hallway or the gaping mouth of Cthulhu when I open my door in a minute. But the kind of certainty that makes reasoned action impossible? We clearly don't have that.

No.
More like "Am I really sufficiently certain of the outcome that I should kill a man based on this?"

by **TheGrammarBolshevik** » Mon Mar 19, 2012 3:43 pm UTC

OK. Suppose you live in a world where you can have such certainty. There's no conceptual difficulty in supposing that you're a very strong person pushing a very heavy man.

The question is still whether it's OK to do it.

by **jules.LT** » Mon Mar 19, 2012 3:47 pm UTC

And I'm saying that this kind of world is nothing like the world we live in, and we are therefore ill-equipped to answer that question.
Also, that makes the answer to the question irrelevant to situations that happen in the real world.

by **TheGrammarBolshevik** » Mon Mar 19, 2012 3:49 pm UTC

Seriously? A world with big, strong people is so unlike ours that we can't think about it?

by **jules.LT** » Mon Mar 19, 2012 3:51 pm UTC

A world where 100% certainty exists. Duh.

by **TheGrammarBolshevik** » Mon Mar 19, 2012 3:55 pm UTC

Who says you need 100% certainty in order to kill someone? If I kill someone in self-defense, there's a non-zero chance that they were just playing a harmless practical joke. Self-defense is still OK.

by **jules.LT** » Mon Mar 19, 2012 4:00 pm UTC

I'm not saying that you need to be 100% certain to kill someone. Only that 100% certainty scenarios (or even scenarios where you somehow know the exact odds) are not transposable to real world decision-making because our evaluation of risk is such a huge part of that decision-making.

by **Роберт** » Mon Mar 19, 2012 4:14 pm UTC

jules.LT wrote:
Роберт wrote:I think if I thought about it I would be for mandatory organ donation of dead people, but would not agree to killing an uninvolved third party to save people.

You might want to check out this thread: Mandatory Organ Donation.

Making organ donation opt-out rather than opt-in is enough to make a huge difference in organ supply.

Changing it from opt-in to opt-out is a great idea for a real-world situation.

I'm still wondering how I'm strong enough to get the train-stopping Fat Man onto the tracks...

by **Technical Ben** » Mon Mar 19, 2012 4:15 pm UTC

Yep. I see the problem as worse than that though. In assuming "perfect information" for the scenario, are we not assuming we already know the answer? In that case it's a statement (X = good/bad) not a question (is X a good/bad result). "Does the end justify the means" depends on what the "means" or the "end" is. If we say "no" then can any action be justified (such as walking the dog or drinking coffee)? I don't think many would say all actions are justified either. So we do need more information! We can never make a general answer to that question, because each end, means and answer will be different. For example in the train thought experiment we do not know what the "justified end" is. Unless I've missed something.

By asking for a moral choice or an opinion, are we opening the question up to unknowns? Given perfect information in the thought experiment, we can give a logical answer. But that also depends on what out goal is. We would need to define a "best case" for the thought experiment, none is given. Instead we are asked "is this favourable". We still first need to define what is favourable. That forces us to define which is the correct answer before asking the question (hence the 3 rules problem for robots!

). If we want to consider "morals" we need to know things that are not specified in the thought experiment, so we do not have the "perfect information" assumed in the question.

I agree with the TheGrammarBolshevik in asking "what would be the right thing to do if we were to have all the relevant information". However, we cannot make a decision on this ahead of time. Because it depends on what the relevant information is. That is not given in the thought experiment, only the fact we have a choice, hence why people ask further questions. The thought experiment does help in getting people to think about what questions or considerations to make in each instance though. That's a good thing in of itself.

So, back to the "your faced against a killer robot" problem. I still choose "option n!" and hope it reaches a P versus NP problem halting it or leaving it stuck in a loop. Else hopefully the robots are programmed to not give answers to questions they do not know how to compute.

by **Zamfir** » Mon Mar 19, 2012 4:24 pm UTC

I don't think the main problem is to imagine a situation with high enough certainty. The difficulty is whether the lessons you might learn are generalizable to situations that we actually care about. If you live in a world where people regularly run into runaway trains with convenient switches and fat men and lots of certainty, etc, it would make sense to carefully consider and debate the ethics of such situations.

But as it is, such intuition pump scenarios hve a high bait-and-switch content. First, you're asked to think away the uncertainty. But whatever conclusions you arrive at, you people will try to apply those conclusions to more realistic cases. And then the temptation is to underestimate the importance of uncertainty. After all, you already have worked hard on these perfect-certainty cases, it would be a shame to start frm scratch.

There are specific activities where people do expect to encounter life-and-death choices with enough regularity to prepare for them. Doctors, soldiers, mountaineers. Police officers, especially with respect to self-defense. Such fields tend to develop their own fairly elaborate ethical rules and guidelines, that don't easily generalize or translate to different situations.

If runaway trolleys with fat men on bridges really were a regular occurence, I'd expect something similar to happen. The development of somewhat acceptable principles for dealing with runaway trolleys. Perhaps a philosophical tradition of train ethics, just as there is for medicine.

by **morriswalters** » Mon Mar 19, 2012 5:24 pm UTC

Certain neurological deficits can cause someone to take the utilitarian approach and push the fat man. fMRI's on the other hand show a high degree of conflict when acting directly, in normal people. Apparently we are built this way. They call it empathy.

by **ahammel** » Mon Mar 19, 2012 7:02 pm UTC

I have a friend who's doing his MA in ethical philosophy. His supervisor used to drive a trolley in an amusement park on the weekends. According to her, any competent conductor could intentionally derail the train at the points, thereby saving everybody's life. So, problem solved, I guess :lol:

[/off topic]

Rule utilitarianism is a system that was devised to deal with the uncerrtainty problem. It goes something like this:

Act Utilitarian: The morally correct action is the one that results in the greatest good for the greatest number of people.
Rule Utilitarian: But how do you know what that action is a priori? When one has to make moral decisions "in the field", so to speak, it's very difficult to forsee the consequences of one's actions accurately enough to make a guess at which produce the greatest good.
Act Utilitarian: Well, what would you suggest?
Rule Utilitarian: The correct approach is clearly to devise a system of moral rules which will, when followed consistently, produce the greatest good. That way, we can rationally evaluate the consequences of actions from the armchair instead of having to go through a bunch of complicated and uncertain utilitarian calculus every time we do anything.
Kantian: But that suffers from the same problem, doesn't it? How do you know what the utilitarian consequences of a rule will be until it's let loose in the world? And besides, you're glossing over the fact that there's no general method to decide whether one action has more "utility" than another.
Rule Utilitarian: Ah...um...well...
Kantian: Now, if you'll take a look at my paper on the Categorical Imperative...
Rule Utilitarian and Act Utilitarian (in unison): Oh, shut up!

by **Bharrata** » Mon Mar 19, 2012 8:10 pm UTC

TheGrammarBolshevik wrote:I think it would be helpful, assuming we don't want to just turn this thread into trivial "Well like nothing is certain man," to just suppose that all the assumptions of the thought experiment are known and ask what would be the right thing to do if we were to have all the relevant information. After all, it's not like there aren't analogous situations that deal away with these worries (the unsuspecting patient, for example).

I appreciate your attempt to keep this thread going (just because I'm critical of the thought experiment doesn't mean I think it shouldn't be discussed, quite the opposite) but as others have pointed out there is a problem with acting as if we have full knowledge of the results of our actions, especially since we are human and fallible. However, my Hume quip about chaos theory wasn't just a joke for the sake of a joke or a "nothing is certain braaahhh" but more of the problem that every moment and every action is not a small slice that is separate from the past and future but is influenced by and influences both in turn. How do we know what the fat man will do in 20 years time if he lives, or for that matter, what the five people will do? Will one of the five people write a story about the experience that causes people to think our pushing of the fat man was justified, even though (and obviously this is still up for debate) it was the ethically wrong decision? Would this cause everyone who reads the story to act ethically wrong while believing it is ethically right? This is why I said Hume would have a field day with the idea/observation that small changes in the starting variables of a system produce huge changes over time, especially when coupled with human ignorance.

We cannot know all the ends of our actions in every moment but I don't think it's fair to say that we can't know all of them, neither is it fair to say that we cannot make value judgements on the means to get to those ends. If I work out and lift weights, I can be fairly sure it will improve my overall health, barring injury. But if I take steroids to get to my goals faster I may negatively impact other parts of my health. I think in this case that unless I'm taking PEDs under the strict supervision of an expert that the means I used to achieve my goals/end are very much a question of ethics and even moreso if the means I used end up affecting my future personality and attitude. Maybe in the case of the PED user they look down on people who don't use them as "wimps" or "chumps" who aren't willing to game the system so they can sit on their moral high horse, or maybe the PED user always feels a sense of guilt or insecurity over their gains which they know they didn't put the natural time in for....and both of those situations aren't even taking into account the health ramifications.

The structure of this argument applies to your health care example as well, though first I'd like to note that I do not know much about the healthcare field and its developed ethics though I did work in a pharmacy for a number of years when I was younger as a technician....I just really wanted to add that the question would be better posed to my younger sister who is not a philo geek but a nursing/med student and one day blew up just this sort of example-type I'm criticizing when for an essay on Utilitarian ethics which proposed that you either had to kill a child to save the world population and what was your choice...she said that she'd, based on utilitarianism, let the world end (they would be dead immediately, not slow suffering) because once it did there would no longer be any suffering and no humans to judge good or bad. <- I was both the proudest older brother at that moment and struck by the simplicity which with a lot of these ethical thought experiments are posed. The professor realized the question needed revising :lol:

Anyways, my treatment of the healthcare question is based solely on my understanding of the Hippocratic Oath, so take it for what it's worth. I can understand the problem of triage (using that term for the sake of brevity) and I would say in the case of one patient coming in who has the money to pay for care and one coming in who doesn't that it makes more sense to take the one who has the money for it and sending the other patient on their way without doing any harm...for no other reason than treating Patient A instead of B will allow the hospital to continue running and provide care to Patient C tomorrow. I don't know if that's where you're going with the health care question but that's my take on what I assume to be the toughest ethical decision in healthcare.

But let's go back to the question I posed in my last post: do we know the ends of our means? This is one of the things Heidegger is attempting to answer, or at the very least begin to question again, in his "Question Concerning Technology" essay. Later on (maybe even before that essay, I'm not that great with specific timelines on their work) I believe Levinas develops this question of means and ends into a more full ethical philosophy.

The relevant bit in the essay, at least at the beginning:

Martin Heidegger wrote:According to ancient doctrine, the essence of a thing is considered to be what the thing is. We ask the question concerning teehnology when we ask what it is. Everyone knows the two statements that answer our question. One says: Technology is a means to an end. The other says: Technology is a human activity. The two definitions of technology belong together. For to posit ends and procure and utilize the means to them is a human activity. The manufacture and utilization of equipment, tools, and machines, the manufactured and used things themselves, and the needs and ends that they serve, all belong to what technology is. The whole complex of these contrivances is technology. Technology itself is a contrivance—in Latin, an instrumentum.

The current conception of technology, according to which it is a means and a human activity, can therefore be called the instrumental and anthropological definition of technology.

Who would ever deny that it is correct? It is in obvious conformity with what we are envisaging when we talk about technology. The instrumental definition of technology is indeed so uncannily correct that it even holds for modern technology, of which, in other respects, we maintain with some justification that it is, in contrast to the older handicraft technology, something completely different and therefore new. Even the power plant with its turbines and generators is a man-made means to an end established by man. Even the jet aircraft and the high--frequency apparatus are means to ends. A radar station is of course less simple than a weather vane. To be sure, the construction of a high-frequency apparatus requires the interlocking of various processes of technical-industrial production. And certainly a sawmill in a secluded valley of the Black Forest is a primitive means compared with the hydroelectric plant on the Rhine River.

But this much remains correct: Modern technology too is a means to an end. This is why the instrumental conception of technology conditions every attempt to bring man into the right relation to technology. Everything depends on our manipulating technology m the proper manner as a means. We will, as we say, "get" technology "intelligently in hand." We will master it. The will to mastery becomes all the more urgent the more technology threatens to slip from human control.

But suppose now that technology were no mere means: how would it stand with the will to master it? Yet we said, did we not that the instrumental definition of technology is correct? To be sure. The correct always fixes upon something pertinent in whatever is under consideration. However, in order to be correct, this fixing by no means needs to uncover the thing in question in its essence. Only at the point where such an uncovering happens does the true propriate. For that reason the merely correct is not yet the true. Only the true brings us into a free relationship with that which concerns us from its essence. Accordingly, the correct instrumental definition of technology still does not show us technology's essence. In order that we may arrive at this, or at least come close to it, we must seek the true by way of the correct. We must ask: What is the instrumental itself? Within what do such things as means and end belong? A means is that whereby something is effected and thus attained. Whatever has an effect as its consequence is called a cause. But not only that by means of which something else is effected is a cause. The end that determines the kind of means to be used may also be considered a cause. Wherever ends are pursued and means are employed, wherever instrumentality reigns, there reigns causality.

For centuries philosophy has taught that there are four causes: (1) the causa materialis, the material, the matter out of which, for example, a silver chalice is made; (2) the causa formulis, the form, the shape into which the material enters; (3) the causa finalis, the end, for example, the sacrificial rite in relation to which the required chalice is determined as to its form and matter; (4) the causa efficiens, which brings about the effect that is the finished, actual chalice, in this instance, the silversmith. What technology is, when represented as a means, discloses itself when we trace instrumentality back to fourfold causality.

But suppose that causality, for its part, is veiled in darkness with respect to what it is? Certainly for centuries we have acted as though the doctrine of the four causes had fallen from heaven as a truth as clear as daylight. But it might be that the time has come to ask: Why are there only four causes? In relation to the aforementioned four, what does "cause" really mean? From whence does it come that the causal character of the four causes is so unifiedly determined that they belong together?

So long as we do not allow ourselves to go into these questions, causality, and with it instrumentality, and with this the accepted definition of technology, remain obscure and groundless.

For a long time we have been accustomed to representing cause as that which brings something about. In this connection, to bring about means to obtain results, effects. The causa efficiens, but one among the four causes, sets the standard for all causality. This goes so far that we no longer even count the causa finalis, telic finality, as causality. Causa, casus, belongs to the verb cadere, to fall, and means that which brings it about that something turns out as a result in such and such a way. The doctrine of the four causes goes back to Aristotle. But everything that later ages seek in Greek thought under the conception and rubric "causality" in the realm of Greek thought and for Greek thought per se has simply nothing at all to do with bringing about and effecting. What we call cause [Ursache] and the Romans call causa is called aition by the Greeks, that to which something else is indebted [das, was ein anderes verschuldet]. The four causes are the ways, all belonging at once to each other, of being responsible for something else.

http://www.wright.edu/cola/Dept/PHL/Class/P.Internet/PITexts/QCT.html (*note that this translation is one I came across while looking for an easy way to post what I wanted, it seems correct but I do not have the time at the moment to check it with the translation I normally use)

It may not be readily apparent how this relates to the question posed in the original thought experiment of this thread. However, when we think about the way in which we interact with the world, in other words our ethics, it has to be kept in mind that the ways in which we think about the things and people we interact with determines how we interact with them. Further, when we do not even fully understand our relationship with the means of mere objects (as Heidegger is pointing out) how is it that we are so sure we know we understand the relationship we have with other people and the means and ends therein? This is where Levinas goes beyond Heidegger to develop a different ethics, namely that our notions about the things external to us end up affecting how we view other people and how we interact with and value them; the individual judges the world based on their preconceived framework of reality that may have no correlation to reality as it is. So when you say that we should approach this question as if we can have absolute certainty about the outcomes of our actions you are asking us to participate in a framework that operates as if the conscious life is a uniform plane, this then reifies the individual as an independent holder of absolute knowledge (a god unto their own universe in which others are held to their judgements) rather than an interdependent participator in the commons of the undiscovered and discovered. (That may seem accusatory, it is not...and yes I got a bit poetic en route to making my point :oops:

) So then every moment becomes an ethical choice between treating the world and the people in it as a means to your individual end, or as an end which (by which I mean the physical world and the people other than yourself) treats you as the means.

By that I mean the external is the framework which alters the internal rather than the common notion at this moment in time that the internal is what alters the external. This is the ethical question posed in this thread, or I believe the fundamental ethical question, because your answer to that question will affect your decision. If you believe you couldn't live with yourself because you failed to save 5 people when you could've killed one and saved them you are not acting morally when you do save them and kill the fat man, you are acting selfishly because you do not want to live with the trauma and/or you wanted to be the hero with divine knowledge that saved the day - though yes you will have trauma if you choose to not save them. Like I stated earlier in the thread though, the ethical thing to do in this case is to tell the fat man what you know, then it is up to him to either act or not act, and your duty is fulfilled. But if you would rather just kill the fat man to save the five, you are essentially a tyrant who is acting as if your happiness supersedes that of other individuals' autonomy and self-determination, to say nothing of the fact that you believe it is worth setting the precedent that another's autonomy is null and void when an accident is taking place (i.e. what caused this situation to develop in the first place...if anything those five deaths would be on the hands of the conductor).

I realize that I'm setting up a slippery slope argument in that last, non-parenthetical, clause but that is the problem with ethical systems and frameworks, they end up reifying themselves and in the case of the current common system we end up with people that think it is morally correct to be entirely self-serving because the invisible hand will take care of it.

** it should also be said that there may be a physical, instinctual urge to push the fat man in front of the train, regardless of your ethical choice, but I think it goes without saying that the task of humanity, or at least the task we've attempted over the centuries is to rise above our basic instincts.

by **TheGrammarBolshevik** » Mon Mar 19, 2012 8:29 pm UTC

Zamfir wrote:I don't think the main problem is to imagine a situation with high enough certainty. The difficulty is whether the lessons you might learn are generalizable to situations that we actually care about.

I'm not sure that trolley problems are often used to make generalizations about moral theory. It seems like they usually come either as A) interesting casuistic questions (where the question is not "which moral theory is right" but "what does moral theory tell us about this situation") or as B) objects of research in moral psychology (where the question is not "what is right," but "how do people decide what is right").

by **Technical Ben** » Mon Mar 19, 2012 9:13 pm UTC

Thanks. I agree with that TheGrammarBolshevik. (I was even arguing that we could not find out a right/wrong answer from the thought experiment too.)

I'm not sure what you mean by a). I see the decision to make as a difficult situation, I'm not sure what comment moral theory can make on that.
But it does bring up interesting questions as regards to b). Most peoples mechanism in this thread was to ask more questions when posed such a decision. This is often one of the best things to do. It's how we learn. Sometimes we can decided ahead of time, like the police or doctors mentioned above, but even then more information = a better decision. Also, do most people agree that they'd respect any decision someone makes? The reason is, to the one involved, it seemed the correct decision.

I suppose that's why the question becomes difficult when AI or robotics is involved. We think of robots and computer programs as either/or decision makers. Where as humans are a lot more flexible. There are a lot of problems in the real world which would take inordinate amounts of time to make into a binary decision. Even our laws can reflect that. What would a court rule in the thought experiment? Would a Jury decide that as every other action was ruled out, the accused could be considered acting against their own consent (such as crimes committed when under life threatening circumstances. Like stealing someone's phone to call 911).

by **krogoth** » Tue Mar 20, 2012 1:22 am UTC

I would like to pose a similar question.

Is it MORE immoral to save yourself if you are tied to the tracks and the train is already coming at you, to deviate the trains path to crash.
Explicitly, You are in danger if you don't change the train to the other tracks, but they arn't endangered unless you cause them to be.
It seems like the same question posed to the fat man whom could save them, but makes it seem more selfish to not save them.

Maybe this is the opp-ed out rather than opped in of organ donations...

by **ahammel** » Tue Mar 20, 2012 2:33 am UTC

krogoth wrote:I would like to pose a similar question.

Is it MORE immoral to save yourself if you are tied to the tracks and the train is already coming at you, to deviate the trains path to crash.
Explicitly, You are in danger if you don't change the train to the other tracks, but they arn't endangered unless you cause them to be.
It seems like the same question posed to the fat man whom could save them, but makes it seem more selfish to not save them.

Maybe this is the opp-ed out rather than opped in of organ donations...

I don't think you'll find anybody this side of Ayn Rand who'd be willing to argue that allowing the train to kill you rather than save yourself and kill all of its ocupants is anything short of heroic.

by **SunAvatar** » Tue Mar 20, 2012 5:21 am UTC

TheGrammarBolshevik, I understand what you're saying. I really do. I've said it myself before, when I'm using an extreme case to put some concept or another into sharp relief and someone gets on my case because it's "unrealistic." You are saying it doesn't matter if the scenario is unrealistic, it's still possible to talk about what we would do in an unrealistic scenario. Trying to get around it with "but what if..." quibbles is missing the point.

I'm saying that so you don't think I'm missing the point myself when I say that these kinds of idealized moral questions are usually useless and even dangerous. It's not just that they're unrealistic and inapplicable, though they are: figuring out what I would do in a universe of perfect information if I were incapable of error might be a great party game, but it isn't really going to help me out once I leave the dorm. The danger is that they seem applicable. They are designed as edge case tests, but treated as applied problems.

As Zamfir pointed out, a question like this is almost always a setup for a bait-and-switch. First you are told explicitly that this is just an exercise, not meant to be an accurate model of reality, and you should just accept all the premises at face value and ignore questions like "How do I know what I think I know?" that would be relevant in real life. If you try to ask for more information you are assured it's irrelevant. So you silence the mental alarms that would normally sound if someone tried to tell you to push someone in front of a god-damned train, and give the only solution the questioner hasn't defined out of existence. Then suddenly you are presented with an actual, complicated moral dilemma, which the questioner insists is perfectly isomorphic to the toy problem you just solved, deftly brushing away any details. If you protest that this seems different, the questioner will sigh, "Didn't we just go over this? You aren't being very consistent." It's a classic foot-in-the-door technique: no one wants to go back on what they've just said. It's far easier to just do something morally repugnant based on a lazy analogy with some fantasy scenario.

Since I do think it is possible to solve fantasy scenarios like this, I came right out earlier and said that, sure, we should push the guy in front of the train. After all, we've already postulated that I know it will work, so I can't invoke uncertainty. If I asked about setting a precedent we'd probably postulate that no one will see me do it and no precedent will be set. If I asked about looking for another way to stop the train, I'd be told there is no other way. If I asked whether I might have missed some relevant detail, I'd be told no, I can't possibly be missing a detail, that would mean the questioner left out relevant information and we all know the questioner wouldn't make a mistake like that. So, sure, I'll push this man from nowhere into the path of this train going from nowhere to nowhere.

Are you proud of me for making the tough choice? But that isn't a tough choice at all. I was left a single moral option, and I took it. Everything about the scenario was constructed so that there was no meaningful trade-off to make. A tough moral choice is one where you have to lose something, not one where the solution is counter-intuitive because the scenario is impossible. Questions like this, where there is no trade-off to be made and the only question is who can stomach the conclusion, are boring. They imply that the choice between good and evil is a matter of saying "Good!" rather than a matter of determining which is which.

by **TheGrammarBolshevik** » Tue Mar 20, 2012 7:04 am UTC

I don't know why you're treating this like I'm part of some grand leading-question conspiracy to convince you that it's OK to shove people in front of trains. Somebody made a post mentioning trolley problems. The initial responses were pretty universally "But like that can't happen!" I came along and suggested that whether this can happen is a separate issue from what would be right if it were to happen. That's all I've said. So who do you think is trying to trap you? The OP, who hasn't returned to the thread at all? Me? Do you think I'm trying to make you commit to hard-line, short-term utilitarianism, just because I think informed, intelligent adults can be trusted with thought experiments?

If you're worried that this might become a bait-and-switch, I invite you to point it out if and when it actually becomes a bait-and-switch. Right now it seems like you're accusing me of hypothetical future intellectual dishonesty, which is a rather difficult charge to answer. Yeah, trolley problems are often parts of shifty arguments. So are most things on the Internet.

As it happens, I don't think it's OK to push the guy in front of the train. No, not even if you're certain that it's going to work and that your action won't have any unintended consequences. So why the mocking expectation that I'm going to be proud of you for taking the "only" choice in this "boring" scenario?

by **Zamfir** » Tue Mar 20, 2012 7:44 am UTC

If you're worried that this might become a bait-and-switch, I invite you to point it out if and when it actually becomes a bait-and-switch. Right now it seems like you're accusing me of hypothetical future intellectual dishonesty, which is a rather difficult charge to answer. Yeah, trolley problems are often parts of shifty arguments. So are most things on the Internet.

My apologies for that introducing that term. I did't want to imply that everyone using abstracted thought experiments was attempting a bait-and-switch, more that the potential is there. That's a big reason for me to dislike them, especially the (rather common) ones that involve whimsical deaths and pain.

As far as I can tell, the appropriate response to such scenarios if they occur in reality, is along the lines of 'oh god that's horrible. For those people, their families. And for the person having to make such a horrible decision. There's not really a good answer, I guess it will haunt you whatever you do'. And you should surely not make heavy moral judgements on whether the real-life persons acted 'correct'

So partially it's an aesthetic objection, but it's a serious one. Because the stories we tell do somewhat shape how we think. Trolley problems (and violin players and the like) have become a prominent public face of ethical questioning, which is a development I don't feel very good about. In a similar sense that I don't like movies where the hero casually executes the bad people with only token hints of necessity.

by **Technical Ben** » Tue Mar 20, 2012 11:53 am UTC

SunAvatar, I see the difficulty in deciding being more to do with the fact the scenario goes against everything else we know. The difficulty may not be in the logical conclusion, but in getting us to accept we have been transported to a universe where there is these "perfect examples". So isolated within the scenario, as you said, we seem to have an easy conclusion to make. Outside and in the real world, even with the perfect information, we still have a difficult time deciding. Same with other people thinking you are a hero for saving the other 5 lives. If we compare your decision to others we make day to day, it's still a decision to take note of. Plus the result is no easier to swallow just because it was unavoidable. :/

I'd guess that your means must not contradict your ends (or visa versa) either (Oh, I don't know, like murdering people to prevent them from being murdered etc).That could cause a whole lot of other problems.

by **SunAvatar** » Tue Mar 20, 2012 4:25 pm UTC

TheGrammarBolshevik wrote:I don't know why you're treating this like I'm part of some grand leading-question conspiracy to convince you that it's OK to shove people in front of trains.... etc.

Sorry, I don't actually think you're doing that, at least not deliberately. I just think it's all the question is really useful for outside of a social science lab. I know something about this particular question, which is that its original purpose was to demonstrate experimentally that people's moral sense is not really based on the principles they think it is based on, by constructing it so that (under certain moral frameworks) there is one obvious correct answer, and then watching people generate ad hoc explanations for why the obvious correct solution isn't correct after all.

Now, this doesn't make it impossible that some other value can be found in it. But, speaking as an aspiring mathematician, it really reminds me of a certain bad habit mathematicians have of breezing into some field like mechanical engineering, abstracting away the things they consider mere details, easily answering the oversimplified questions that remain, and declaring that mechanical engineering is trivial, all the while failing to notice that they haven't done anything engineers actually care about. Sometimes the details actually matter.

TheGrammarBolshevik wrote:As it happens, I don't think it's OK to push the guy in front of the train. No, not even if you're certain that it's going to work and that your action won't have any unintended consequences. So why the mocking expectation that I'm going to be proud of you for taking the "only" choice in this "boring" scenario?

Well, that's unexpected. Do you give the same answer for the version where you are only switching the train to another track where it will incidentally hit someone? The doctrine of double effect says that one may take an action with both good and bad effects, as long as the good effects outweigh the bad, but not if the bad effect plays a causal role in bringing about the good one. In this example, we have one case in which the victim's death is part of the plan for saving five others, and another where the victim's death is merely a side effect; someone employing the doctrine of double effect might reject the first, but accept the second. Honestly I've never met someone who actually buys into this doctrine, so if you give a different answer in the latter case that would be interesting indeed.

by **morriswalters** » Wed Mar 21, 2012 3:53 am UTC

There is no possibility of the trolly problem happening the way it is presented it in any real sense. You will do what your programmed to do. You never make decisions consciously when you are pressed by events, which is why the military and emergency responders train. What the scenario allows you to do is test the way that people think about decisions.

by **ahammel** » Wed Mar 21, 2012 4:07 am UTC

SunAvatar wrote:
TheGrammarBolshevik wrote:As it happens, I don't think it's OK to push the guy in front of the train. No, not even if you're certain that it's going to work and that your action won't have any unintended consequences. So why the mocking expectation that I'm going to be proud of you for taking the "only" choice in this "boring" scenario?

Well, that's unexpected. Do you give the same answer for the version where you are only switching the train to another track where it will incidentally hit someone?

It's not especially surprising to me. It was my impression that "flip the switch but don't kill the fat man" was the most common response (it's certainly my response). I can't dig up a reputable citation for this at the moment, however. Anybody have any data?

by **jules.LT** » Wed Mar 21, 2012 10:26 am UTC

I'll add a data point: I'd flip the switch and my gut won't let me push the fat man.
When we put intermediaries between our actions and their consequences, we feel less implicated. Like when economic experiments use tokens exchangeable for dollars instead of actual dollars.

by **TranquilFury** » Wed Mar 21, 2012 5:05 pm UTC

The ends always justify the means, it is always correct to take the course of action that satisfies your ultimate goals with the least amount of risk.

We have an intuition about the predicted utility of varying options, with the switch, it feels like there are a few possible outcomes if you flip the switch, in order of expected probability:
1 person dies because the switch works as expected
5 people die because switch doesn't work at all
5 people die instead of 1 because you were wrong about which track the train was actually on

alternatively with the fat man, pushing him has some very different predicted consequences, in order of expected probability:
6 people die because the fat man was unable to sufficiently slow the train, this feels more probable because trains are so much more massive than even very fat humans, such that if one fat man was able to absorb all that momentum, the train was going slow enough that it could be evacuated before falling into the pit.
1 person dies 5 people live that may or may not have actually been in any danger in the first place.

The other factor that contributes is personal risk, Pushing a man to certain death, who may or may not fight back, is dangerous immediately, and the consequences if you fail or are witnessed are severe.

Flipping a switch feels much less personally risky, switches normally don't fight back, and aren't going to report you to police, additionally, bystanders that see you flip the switch are either not going to know what the switch was responsible for killing someone, or are going to understand that the alternative was killing everyone on the train, so those people are less likely to report you.

In short, the problem with the question is that the consequences of pushing someone onto the track disagree with our expectation of reality, and as the consequences feel unreliable we are unable to make the call that it's worth the risk to push him.

Who is online