Should You Kill the Fat Man?

[Gorghandi]

What defines good and evil? What if one person's good is another persons evil?

the aeranu tribe, doesn't believe there is good, and evil. we believe these 2 things to be the same, only said differently.  we do believe in right and wrong. and progression. that regardless of what a person is doing or what their 'job' is, whether it is "evil" or "good" that they should do it right. to the best of their abilities. or progression will halt for everyone.

[Ginyu]

1. Tough one. Rationally, it is whether saving 1 or 5 innocent people. So I would probably turn it, but in the real scenario I am not sure (I said "yes").
2. Even harder. Rationally, it is the exact same scenario as before, but still actively pushing someone is entirely different on your personal view. This might seem inconsequent, but I said "no".
3. Yes, easily. The man had intentions in murdering 5 people, and saving 5 innocent people is clearly the better way than one criminal.
4. Yes, easiest. With no other ways to save the innocent people as described in the task, I would torture the murderer to hell and back until he opens his mouth. He wants to take millions their right of living, why should I be graceful to a murderer and let all those innocent people die?

Through all of this, I considered "not saving someone" on a similar level of "actively killing someone". This made 3. and 4. so easy for me.

[Espithel]

Yes, Yes, Yes, Yes.

Take the harm principle, swap harm with merit, and you basically get the same thing.

[Blacksmith]

I don't know why but for me the link in the op leads no where.

[Fippe94]

Probacon made a mistake with the link, it doesn't actually link to the stated url. Copypaste the link to the addess bar instead.

[Blacksmith]

Probacon made a mistake with the link, it doesn't actually link to the stated url. Copypaste the link to the addess bar instead.

K thank you.

I also answered yes on everything. The choices was not really hard tbh but quite straight forward. If you want to maximize happiness you only have 1 option.
What makes it a bit tricky is that in the real world you can never be 100% sure that killing 1 person instead of 5 people or even 1 million people leads to greater happiness. The 5 people might be miserable planing to take their own lives and do so the day after, and thus the fat man died for nothing. The fat man might have been super smart and about to come up with a medicine that would kill all diseases and save billions of lives. And that's why I think people find question 2 hard. In the given example yes is the natural answer but since you can't compare it to a real life situation some people will still say no.

[Higurashi]

Well, this test was too presumptuous to be correct, and in its assumptions thought it useful to chide me for my "inconsistency". Bad design.

While it's morally wrong to kill if avoidable, it's even more morally wrong to allow even more people to die because of your inaction (inaction is also making a choice). The fact that I believe the latter doesn't make my former belief inconsistent at all. A morally wrong action can still be less wrong (i.e. do less harm) than another action, and therefore you can be morally obliged to choose the lesser evil. It's odd that the test designer didn't consider this when you can clearly see that he considered killing one over many was morally acceptable.

The second problem was that the designer considered "total happiness" to be the only way to measure the morality of an action. I don't believe it's right to kill one over five because of total happiness, but rather because of total contribution to humanity's progress.

[OldTrees]

Well, this test was too presumptuous to be correct, and in its assumptions thought it useful to chide me for my "inconsistency". Bad design.

While it's morally wrong to kill if avoidable, it's even more morally wrong to allow even more people to die because of your inaction (inaction is also making a choice). The fact that I believe the latter doesn't make my former belief inconsistent at all. A morally wrong action can still be less wrong (i.e. do less harm) than another action, and therefore you can be morally obliged to choose the lesser evil. It's odd that the test designer didn't consider this when you can clearly see that he considered killing one over many was morally acceptable.

The second problem was that the designer considered "total happiness" to be the only way to measure the morality of an action. I don't believe it's right to kill one over five because of total happiness, but rather because of total contribution to humanity's progress.

The designer did not consider "total happiness" to be the only way to measure morality of an action. That was merely a "utilitarian ethics or other ethics?" question.

I would be curious about your 8 answers to see if your false positive was a result of differing definitions of "morally wrong". In the context of the questions "morally wrong" would mean "invalid answer" rather than "sad answer" and thus any action that can sometimes be the valid answer would not be label as always morally wrong in this context despite it potentially being always lamentable.

As for consistency both the following answer chains are scored as 100% consistent:
no,no,no,no,1 dies,1 dies,1 dies,1 dies && no,no,no,no,many die,many die,many die,many die


That said, yes this particular author is rather pretentious in their implementation of the trolley problem. I could try to present a less pretentious set of examples if that would be beneficial.

[Higurashi]

There is no reason, or excuse, as a utilitarian, not to consider other methods of measurement. That would be tantamount to blind faith. "Other ethics" aren't even hinted at, much less discussed.

[antiaverage]

There is no reason, or excuse, as a utilitarian, not to consider other methods of measurement. That would be tantamount to blind faith. "Other ethics" aren't even hinted at, much less discussed.

I completely agree with you. I was also bothered by their statement of "fact" that torture had a 75% chance of revealing the true information. That's arbitrary and baseless. Torture more often than not leads to the victim saying whatever you want to hear, not necessarily saying the truth. You could tell the "evil fat man" that you'll torture him until the bomb is defused, but there is no 75% chance of getting the truth. It's just as likely this "evil fat man" will stall for 24 hours. Torture doesn't lead to a rational and predictable outcomes, even with historical data from having previously tortured the "evil fat man." So between the illogical "inconsistencies" and the baseless claims, I take umbrage with the test creator. Maybe we should throw them off the metaphorical bridge.

[OldTrees]

There is no reason, or excuse, as a utilitarian, not to consider other methods of measurement. That would be tantamount to blind faith. "Other ethics" aren't even hinted at, much less discussed.

I did not say the author was following utilitarian ethics.

I said the question was asking if you followed utilitarian ethics or some other ethics. The reason this question is asked with the trolley problem is that utilitarian ethics gives a clear answer in every case and the initial sampling of moral intuitions differed from case to case. Now when people are replicating the experiment they also replicate the question about utilitarianism to compare the data sets.

Silly example:
Do you believe that more pasta is strictly better in every case than less pasta?
You are at home, you can either order 1 bowl of pasta delivered to you or go out and get 5 bowls of pasta. Which would you do?
The first question helps inform me if an answer to the second question is determined by 1<5 or if it was determined by a qualitative difference between order delivery vs going out.
If I were to ask a question with both a quantitative and a qualitative difference between the outcomes, it would benefit me to separate out those answers based upon which difference governed their choice. Doing so would give me more information about those answers.
Note: This analogy is not perfect since there are more people that would say "less pasta is strictly better in every case" than would say "less happiness is strictly better in every case". But both groups are exceptionally small so we can include them as known error rather than alter the question.

[Higurashi]

He doesn't need to be. The test being from a utilitarian perspective is fine. Using it as the only option for a line of thinking is not, especially when it's a test designed to test your consistency. You can't test consistency by only using one measuring method, especially if your only options are "utilitarian principle or evil douchebag governed only by chaos".

In other words, this test can test utilitarian consistency only, because if your beliefs aren't utilitarian they score as "inconsistent".