Seeing as I didn't really express my intention with the Black Box question effectively, here's a thought experiment created by an actual professional. It's called the Mere Addition Paradox.
My question is, what is the proper resolution of the paradox? It's rather clear that something is wrong with it, but which of Parfit's conclusions is wrong?
As much as I hate these darned "happiness plots", I'll bite.
What's going on is there's absolutely no basis for any of these claims. The goodness of a 'case' is entirely subjective and the 'happiness' of people in a population, even if we could give that a numerical value, is... entirely subjectively interpreted for value and desirability.
But, let's ignore my crazy ideas, and do this:
One cannot say that A is better than B, or that A is better than A+, if one thinks it is bad to cease the existence of certain people so as to increase the happiness of the remaining people. Assuming that coherence is a logical necessity (assuming).
In fact, A+ is not worse than A (the argument given on Wiki is quite... vacant, though), and B is not worse than A.
The argument that B can't be better than A seems reasonable. So it would seem that B is of equitable appreciableness to A. Which is curious. Very, very curious.
This of course allows one to conclude that A+ bears to A the same goodness-difference as it does to B. Which means A+ is not worse than B.
This is also curious.
If A+ were better than B, then it would be good to reduce the happines of each member of half the population by some amount, so as to increase the happiness of each member of the other half of the population by a smaller amount. In effect, some form of servitude, where half the population serves so as to increase the happiness of the other half, even at the expense of great happiness for the serving half, would be A Good Thing^TM.
I don't think people will agree to that one.
So it seems that all three cases are of equitable goodness. Assuming that people don't believe that ending existence to increase average happiness is good, and that people don't believe that a great loss of happiness is a price at a bargain for a small gain in happiness of an equal number of people.
Private Mod Note
():
Rollback Post to RevisionRollBack
Epic banner by Erasmus of æтђєг.
Awesome avatar provided by Krashbot @ [Epic Graphics].
It seems like the philosophical equivalent of a blivet. The first statement shows you one end. The second shows the other. The third then shows you the middle. But it's just an illusion. A+ is exposed to different rules than both A and B.
The statements are so divorced from any realistic scenario that it's really impossible to say.
However, here's my question: How does one determine happiness?
Quality of life is not linked to happiness. People can have a terrible quality of life and be happy. People can have an amazing quality of life and be unhappy.
Therefore, how do we measure happiness?
It occurs to me the fundamental problem is that happiness must be measured in reference to something else. One is happy or sad compared to people. Therefore, it seems that all statements have a flaw. If everyone is within the same level of happiness as all around them, how can their happiness be any different from another populace with similar circumstances?
The problem with the paradox is that you're applying "better" and "worse" concepts where they don't really belong. "Better" for whom? "Worse" for which purpose? Those aren't defined at all here. Until a rigorous definition of what is better is brought up, the paradox can't even be scientifically discussed.
It seems like the philosophical equivalent of a blivet. The first statement shows you one end. The second shows the other. The third then shows you the middle. But it's just an illusion. A+ is exposed to different rules than both A and B.
I agree that one of the problems is that rules are being applied inequitably.
The comparison between A+ and A deliberately ignores the average happiness, and implicitly assumes that the fact that the average decreases is not an indication that the society has become "worse". It tires to disguise this by calling the two groups "separate".
However, the statement B is better than A+ is entirely founded on the fact that the average happiness is higher in society B. You can't have it both ways - either average happiness is an indication of better, in which case the first conclusion is erroneous, or average happiness is not an indication of better, in which case the second conclusion is erroneous.
Another problem I see with this argument is that it assumes that happiness levels are both linear and quantifiable. I don't actually believe they are either.
Pretext: recently finished my college semester, have been staying up all night (it's now about 5:30am), can't sleep/not tired, and have a *lot* of new free time on my hands... That being said, I've never thought much about philosophical debate with others (I guess the people I surround myself with are simply of similar mindset) but lately some changes in my life have got me to see "things" differently, so here I am, finding myself in a philosophy forum on one of my favorite websites, deciding to try something new! Whew sorry for the long "introduction"... here goes:
I dislike statement 2, that Population B is happier than Population A+. B is only better than (lets say) 25% of A+ (remember, the two sections of A+ are completely independent of one another!), while B is not better than 75% of A+. It seems to me that including A+'s two populations into one to compare as one to Population B would be breaking out of the parameters that hold and define the entire argument together.
To extrapolate (ooh fun word!), if we are going to compare A+ as one object to B as one object, then we must also compare A+ as one object to A, which would then change statement 1 to "A+ is worse than A". Agreed, this is an incorrect statement, so the real problem simply lies in statement 2, which is incorrect. B is not better than A+, not if we want to maintain universal comparative guidelines... do you agree? If not, please be gentle, it's my first time
*edit*
oops, should have read everyones' posts before making my own... sorry for pretty much repeating what you said, Goryus!
I agree with many of the above ideas, but just to add my own voice to it:
To me, the problem is the seemingly arbatrary choice that A+ is as good as A. It does not matter to me that the two groups do not interact, the fact that there is a new group of less happy people, demonstrates to me that A is > A+
So the way I see it. A > B > A+
In A+, we find the lowest quality of life. This is why it scores lowest on my scale.
It seems that Parfit's arguments 1) and 2) are based on saying that the total volume of happiness is what matters, since:
a) by the "mere addition" line of reasoning, any additional people (with any level of happiness) makes the situation better, and
b) he mentions that B is better than A+ because it has the same population size but with a higher average happiness
However, this obviously is inconsistent with the claim that B is worse than A, since it does not follow the same reasoning that the total volume of happiness is the determining factor.
He attempts to base the paradox on argument 3) leading to a Repugnant Conclusion, but his argument 1) alone leads to a similarly repugnant conclusion. That is, let's take population A and now add a separate enormous population whose lives are barely worth living. By his "mere addition" claim, this population should be better than A, even if that additional population is so enormous that it makes the original population negligible in size (and the average happiness approaches that of the additional "barely happy" population).
I think it makes the most sense to reject argument 1.
Another point at which I would consider it reasonable to disagree: His whole A>B argument is a slippery slope argument. This doesn't make it wrong...but it doesn't make it right, either. Just because we are all happy to say Z<A. In fact it does nothing of the sort! It could easily be that B>A and Z
This argument hinges upon the two groups in A+ being completely separate from each other, which is the actual contradiction. If they're separate, they're not part of the same population; we might as well call every group in this example part of A. If, on the other hand, they aren't kept completely separate, then the unhappiness of one group impacts the other, and lowers the overall happiness, and there's no paradox at all.
This argument hinges upon the two groups in A+ being completely separate from each other, which is the actual contradiction. If they're separate, they're not part of the same population; we might as well call every group in this example part of A. If, on the other hand, they aren't kept completely separate, then the unhappiness of one group impacts the other, and lowers the overall happiness, and there's no paradox at all.
But remember that no group is unhappy - in fact, they are all happy.
If we say that a nonexistant person has happiness = 0, then it becomes clear that we should be maximizing total happiness (ie. area of the rectangle). Thus A+ > A, and B > A or A+ iff B has a greater area than that rectangle.
So we reject 2 of the arguments: 2 and 3.
For 2, B is not always greater than A+; it depends on their area. For 3, if B is worse than A, then it must have less area than A. So it must thus be less than A+, and so 2 does not hold.
Why are we trying to increase the size of the rectangle at all? It has yet to be shown that this is something inherently good at all. You're trying to confuse the issue by adding multiple contradicting standards of goodness, but there's no reason we shouldn't simply measure which society is better by the happiness of the average person.
I'd determine the type of average you're looking at. The mean average isn't the best way to go, imo, for a few deliriously happy individuals offset scores of depressed. Once you say the mode or median average, I'm with ya, IBA.
Private Mod Note
():
Rollback Post to RevisionRollBack
[The Crafters] | [Johnnies United]
My anecdotal evidence disagrees with yours! EXPLAIN THAT!
Why are we trying to increase the size of the rectangle at all? It has yet to be shown that this is something inherently good at all.
I would say that Parfit's "mere addition" argument (argument 1) is what implies that increasing the total area = good, which is why it needs to be rejected.
Why are we trying to increase the size of the rectangle at all? It has yet to be shown that this is something inherently good at all. You're trying to confuse the issue by adding multiple contradicting standards of goodness, but there's no reason we shouldn't simply measure which society is better by the happiness of the average person.
I'm getting rid of the use of average as a measure of goodness - it is useless.
If the average happiness is the measure of goodness, then we are morally obligated to kill anyone of less than average happiness in order to increase average happiness. And after we have killed those people, the average goes up, and there are more people we should kill; everyone but the happiest person should die.
Using total happiness (ie. area of the rectangle) gets around this problem.
Why is the average happiness a useless measure of the happiness of a society? And what problems does using the total area solve, since it invokes the aforementioned paradox?
Using the area only evokes the aforementioned paradox if you make the unreasonable assumption that the progression is linear. I think the best method is to use some combination of average H and area H. You'd need to (somehow) specify that Area ceases to influence V(H) when V(AH) goes below a certain threshold.
Average happiness is a better measure than total happiness in that it avoids Parfit's paradox, but since we're still talking about ethics, I think that average happiness is still ultimately a failure. Using that as the sole measure would indicate that a population where everyone has a "happiness rating" of 50 is equally as good as a same-sized population where half the people have a happiness of 100 and the other half has a happiness of 0.
Average happiness is a better measure than total happiness in that it avoids Parfit's paradox, but since we're still talking about ethics, I think that average happiness is still ultimately a failure. Using that as the sole measure would indicate that a population where everyone has a "happiness rating" of 50 is equally as good as a same-sized population where half the people have a happiness of 100 and the other half has a happiness of 0.
I agree. Trying to judge 'better' using only the average happiness of an individual has two problems:
1) It assumes that happiness levels are linear (meaning that each gain in happiness is exactly as important as any other). I don't believe this is true. Making a really happy person 'more happy' doesn't improve the society as much as making a really un-happy person 'more happy'. There is also probably some sort of cap on how much happiness any one person can feel.
2) It ignores other factors that may influence which society should be considered 'better'. While I'm willing to grant that happiness is a significant factor, I'm not yet convinced it is the only factor.
For example, imagine two societies (call them S1 and S2). Both societies have the same number of people, the same amount of available space, and the same amount of available resources. However, their behaviors are different, and this effects their happiness levels.
The people in S1 are deliriously happy, because they are gluttonously consuming their available resources without worrying about the future.
The people in S2 are less happy, because they ration out their resources and worry about things the future might bring.
Taking a snap-shot of these two socieities, can we really say that S1 is better than S2 because the people are all happier? Or does the fact that S2 is more stable and will not quickly self-destruct need to factor into the equation of which society is 'better'?
I agree. Trying to judge 'better' using only the average happiness of an individual has two problems:
1) It assumes that happiness levels are linear (meaning that each gain in happiness is exactly as important as any other). I don't believe this is true. Making a really happy person 'more happy' doesn't improve the society as much as making a really un-happy person 'more happy'. There is also probably some sort of cap on how much happiness any one person can feel.
2) It ignores other factors that may influence which society should be considered 'better'. While I'm willing to grant that happiness is a significant factor, I'm not yet convinced it is the only factor.
For example, imagine two societies (call them S1 and S2). Both societies have the same number of people, the same amount of available space, and the same amount of available resources. However, their behaviors are different, and this effects their happiness levels.
The people in S1 are deliriously happy, because they are gluttonously consuming their available resources without worrying about the future.
The people in S2 are less happy, because they ration out their resources and worry about things the future might bring.
Taking a snap-shot of these two socieities, can we really say that S1 is better than S2 because the people are all happier? Or does the fact that S2 is more stable and will not quickly self-destruct need to factor into the equation of which society is 'better'?
For 1), I would say that happiness levels can be linear, but due to decreasing marginal utility, it gets harder and harder to make a happy person happier. So, instead of making that happy person even happier, you could instead make many many unhappy people happy.
In answer to 2), I think you should look at total happiness over the entire future of the society. So glutony would reduce how good the society is, because there is no future.
And for that matter, how can you say one happy person is happier than another happy person?
This, I'm thinking, is an important point.
I remembered this thread and it seems to touch on a point related to an issue I'm thinking about. Essentially, the question is, can you compare the happiness (utility, rather) of one person with another? At most, you can figure out that one person prefers, say, chocolate ice cream to vanilla ice cream, while another prefers vanilla to chocolate, and it can only be found on the margin at that (this can be done by granting each one the same amount of money to spend on either chocolate or vanilla ice cream); preferences and values are a result of, it appears, a mix of innate tendencies and external circumstances. But even if you find that person A prefers chocolate to vanilla and person B prefers vanilla to chocolate, you can't say that person A derives less utility from vanilla ice cream than person B derives from chocolate ice cream.
Of course, holding this view consistent requires that I cannot solidly assert that, for instance, the loss of somebody else's arm causes them more unhappiness than a loss of two cents to me. On the other hand (har har), am I still able to make a sort of empathetic guesstimate?
But without your two cents, how can you offer your oh-so-valuable token comments, Einsteinmonkey?
You caught the problem; what is the principled difference between ice cream choices and and physical mutilation, if we can only use this "happiness" measure?
I'm just more and more skeptical whether the entire hypothetical is flawed, because it might be nonsensical to posit such a thing as a "happiness amount."
Private Mod Note
():
Rollback Post to RevisionRollBack
Epic banner by Erasmus of æтђєг.
Awesome avatar provided by Krashbot @ [Epic Graphics].
My question is, what is the proper resolution of the paradox? It's rather clear that something is wrong with it, but which of Parfit's conclusions is wrong?
Discuss.
Experiments Series: #5 (Courtly Intrigue Mafia) | #4 (Drunken Tracker) | #3 (Big Red Button) - coming soon | #2 (Pope Mafia) | #1 (Iso's Inflammable Mafia)
Mini Games: MTGS Mafia Redux II (Invitational, Evil Mirror Universe) | Unreal City
Old Games (bad): The Greenwood Affair | Blood Moon Mafia
What's going on is there's absolutely no basis for any of these claims. The goodness of a 'case' is entirely subjective and the 'happiness' of people in a population, even if we could give that a numerical value, is... entirely subjectively interpreted for value and desirability.
But, let's ignore my crazy ideas, and do this:
One cannot say that A is better than B, or that A is better than A+, if one thinks it is bad to cease the existence of certain people so as to increase the happiness of the remaining people. Assuming that coherence is a logical necessity (assuming).
In fact, A+ is not worse than A (the argument given on Wiki is quite... vacant, though), and B is not worse than A.
The argument that B can't be better than A seems reasonable. So it would seem that B is of equitable appreciableness to A. Which is curious. Very, very curious.
This of course allows one to conclude that A+ bears to A the same goodness-difference as it does to B. Which means A+ is not worse than B.
This is also curious.
If A+ were better than B, then it would be good to reduce the happines of each member of half the population by some amount, so as to increase the happiness of each member of the other half of the population by a smaller amount. In effect, some form of servitude, where half the population serves so as to increase the happiness of the other half, even at the expense of great happiness for the serving half, would be A Good Thing^TM.
I don't think people will agree to that one.
So it seems that all three cases are of equitable goodness. Assuming that people don't believe that ending existence to increase average happiness is good, and that people don't believe that a great loss of happiness is a price at a bargain for a small gain in happiness of an equal number of people.
Awesome avatar provided by Krashbot @ [Epic Graphics].
However, here's my question: How does one determine happiness?
Quality of life is not linked to happiness. People can have a terrible quality of life and be happy. People can have an amazing quality of life and be unhappy.
Therefore, how do we measure happiness?
It occurs to me the fundamental problem is that happiness must be measured in reference to something else. One is happy or sad compared to people. Therefore, it seems that all statements have a flaw. If everyone is within the same level of happiness as all around them, how can their happiness be any different from another populace with similar circumstances?
When you add a less happy population, it's not a better society. It's less happy on average (logic used for A > B). the order should go A+ < B < A
4th place at CCC&G Pro Tour
Chances of bad hands (<2 or >4 land):
21: 28.9%
22: 27.5%
23: 26.3%
24: 25.5%
25: 25.1%
26: 25.3%
I agree that one of the problems is that rules are being applied inequitably.
The comparison between A+ and A deliberately ignores the average happiness, and implicitly assumes that the fact that the average decreases is not an indication that the society has become "worse". It tires to disguise this by calling the two groups "separate".
However, the statement B is better than A+ is entirely founded on the fact that the average happiness is higher in society B. You can't have it both ways - either average happiness is an indication of better, in which case the first conclusion is erroneous, or average happiness is not an indication of better, in which case the second conclusion is erroneous.
Another problem I see with this argument is that it assumes that happiness levels are both linear and quantifiable. I don't actually believe they are either.
If your answer is "No," then your morality does not come from God's commandments.
If your answer is "Yes," then please, please reconsider.
I dislike statement 2, that Population B is happier than Population A+. B is only better than (lets say) 25% of A+ (remember, the two sections of A+ are completely independent of one another!), while B is not better than 75% of A+. It seems to me that including A+'s two populations into one to compare as one to Population B would be breaking out of the parameters that hold and define the entire argument together.
To extrapolate (ooh fun word!), if we are going to compare A+ as one object to B as one object, then we must also compare A+ as one object to A, which would then change statement 1 to "A+ is worse than A". Agreed, this is an incorrect statement, so the real problem simply lies in statement 2, which is incorrect. B is not better than A+, not if we want to maintain universal comparative guidelines... do you agree? If not, please be gentle, it's my first time
*edit*
oops, should have read everyones' posts before making my own... sorry for pretty much repeating what you said, Goryus!
To me, the problem is the seemingly arbatrary choice that A+ is as good as A. It does not matter to me that the two groups do not interact, the fact that there is a new group of less happy people, demonstrates to me that A is > A+
So the way I see it. A > B > A+
In A+, we find the lowest quality of life. This is why it scores lowest on my scale.
a) by the "mere addition" line of reasoning, any additional people (with any level of happiness) makes the situation better, and
b) he mentions that B is better than A+ because it has the same population size but with a higher average happiness
However, this obviously is inconsistent with the claim that B is worse than A, since it does not follow the same reasoning that the total volume of happiness is the determining factor.
He attempts to base the paradox on argument 3) leading to a Repugnant Conclusion, but his argument 1) alone leads to a similarly repugnant conclusion. That is, let's take population A and now add a separate enormous population whose lives are barely worth living. By his "mere addition" claim, this population should be better than A, even if that additional population is so enormous that it makes the original population negligible in size (and the average happiness approaches that of the additional "barely happy" population).
I think it makes the most sense to reject argument 1.
In addition, the combining of A into A+ is a false solution; if the populations are *totally* separate, then if you combine them to compare to B, you must eb able to do the same and compare them with A; and then if A > B, A> A+. It is the initial assumption that A=A+ which makes this a paradox, but EITHER A+ is both better and worse than B (not impossible, but unlikely)y) or A+ <A.
But remember that no group is unhappy - in fact, they are all happy.
If we say that a nonexistant person has happiness = 0, then it becomes clear that we should be maximizing total happiness (ie. area of the rectangle). Thus A+ > A, and B > A or A+ iff B has a greater area than that rectangle.
So we reject 2 of the arguments: 2 and 3.
For 2, B is not always greater than A+; it depends on their area. For 3, if B is worse than A, then it must have less area than A. So it must thus be less than A+, and so 2 does not hold.
I would say that Parfit's "mere addition" argument (argument 1) is what implies that increasing the total area = good, which is why it needs to be rejected.
I'm getting rid of the use of average as a measure of goodness - it is useless.
If the average happiness is the measure of goodness, then we are morally obligated to kill anyone of less than average happiness in order to increase average happiness. And after we have killed those people, the average goes up, and there are more people we should kill; everyone but the happiest person should die.
Using total happiness (ie. area of the rectangle) gets around this problem.
I agree. Trying to judge 'better' using only the average happiness of an individual has two problems:
1) It assumes that happiness levels are linear (meaning that each gain in happiness is exactly as important as any other). I don't believe this is true. Making a really happy person 'more happy' doesn't improve the society as much as making a really un-happy person 'more happy'. There is also probably some sort of cap on how much happiness any one person can feel.
2) It ignores other factors that may influence which society should be considered 'better'. While I'm willing to grant that happiness is a significant factor, I'm not yet convinced it is the only factor.
For example, imagine two societies (call them S1 and S2). Both societies have the same number of people, the same amount of available space, and the same amount of available resources. However, their behaviors are different, and this effects their happiness levels.
The people in S1 are deliriously happy, because they are gluttonously consuming their available resources without worrying about the future.
The people in S2 are less happy, because they ration out their resources and worry about things the future might bring.
Taking a snap-shot of these two socieities, can we really say that S1 is better than S2 because the people are all happier? Or does the fact that S2 is more stable and will not quickly self-destruct need to factor into the equation of which society is 'better'?
If your answer is "No," then your morality does not come from God's commandments.
If your answer is "Yes," then please, please reconsider.
For 1), I would say that happiness levels can be linear, but due to decreasing marginal utility, it gets harder and harder to make a happy person happier. So, instead of making that happy person even happier, you could instead make many many unhappy people happy.
In answer to 2), I think you should look at total happiness over the entire future of the society. So glutony would reduce how good the society is, because there is no future.
This, I'm thinking, is an important point.
I remembered this thread and it seems to touch on a point related to an issue I'm thinking about. Essentially, the question is, can you compare the happiness (utility, rather) of one person with another? At most, you can figure out that one person prefers, say, chocolate ice cream to vanilla ice cream, while another prefers vanilla to chocolate, and it can only be found on the margin at that (this can be done by granting each one the same amount of money to spend on either chocolate or vanilla ice cream); preferences and values are a result of, it appears, a mix of innate tendencies and external circumstances. But even if you find that person A prefers chocolate to vanilla and person B prefers vanilla to chocolate, you can't say that person A derives less utility from vanilla ice cream than person B derives from chocolate ice cream.
Of course, holding this view consistent requires that I cannot solidly assert that, for instance, the loss of somebody else's arm causes them more unhappiness than a loss of two cents to me. On the other hand (har har), am I still able to make a sort of empathetic guesstimate?
You caught the problem; what is the principled difference between ice cream choices and and physical mutilation, if we can only use this "happiness" measure?
I'm just more and more skeptical whether the entire hypothetical is flawed, because it might be nonsensical to posit such a thing as a "happiness amount."
Awesome avatar provided by Krashbot @ [Epic Graphics].