Monday, May 14, 2007

The 1000 Sadists Problem

There are two Big Questions that people trying to understand desire utilitarianism will eventually ask. Atheist Observer asked one of them two days ago.

This question asks me to imagine a world with 1000 people who seek do to harm of one form or another (child rapists, Nazis, slave owners) and 1 person whom they desire to harm (a child, a Jew, a slave). The question then is, “Doesn’t your theory say that the harmful acts in this type of situation are obligatory, since they fulfill the most desires?

[Note: For the sake of completeness, Big Question #1 is, “Why should I care whether the desires of others are fulfilled or not?” I have addressed that question in the post, “The Hateful Craig Problem.”]

My answer is that the theory still condemns the harmful act.

We must distinguish between two theories:

Desire fulfilling act utilitarianism: That act is right that fulfills the most desires.

Desire utilitarianism: Desires (like everything else in the universe) are good or bad according to whether they tend to fulfill or thwart (other) desires. Acts are right or wrong according to whether or not a person with good desires would perform that act.

Desire Fulfilling Act Utilitarianism would say to harm the child in these imaginary circumstances.

Desire Utilitarianism says that we need to look at the desire to see if it tends to fulfill or thwart other desires.

So, take the desire to rape children, for example. In any society, the more prevalent and the stronger this desire becomes, the more other desires are thwarted. As we turn the desire up, making it stronger and more common, either more children (and those who truly care for children) are having their desires thwarted, or those with this desire to rape children are having their desires thwarted.

Either way, ‘up’ in strength and prevalence means more desires being thwarted. ‘Down’ on the other hand means fewer desires being thwarted. If we can dial this desire all the way down to zero, then children would be safe at least from this type of harm, and nobody in society would be suffering the frustration of not acting on such a desire.

So, this desire counts as a bad desire.

Then, an act is right according to whether or not a person with good desires would perform that act. A person with good desires would have no desire to rape a child (and several reasons not to). So, a person with good desires would not perform that action. Which means that the action falls in the category of “morally prohibited” in desire utilitarian terms.

The key difference, again, is the difference between evaluating actions according to whether or not they fulfill the most desires, and evaluating desires according to whether they tend to fulfill or thwart other desires.

Intrinsic Value vs Relational Value

This distinction actually has at its roots another distinction – a distinction between the idea that value is an intrinsic property (found, in this case, in states of affairs where desires are being fulfilled), versus the idea that value is a relational property (a relationship between states of affairs and desires).

I can illustrate this difference by asking you to imagine two possible universes. In Universe 1there is a single person, Agent A, with a desire that P (for some proposition P) and a state of affairs in which P is true. In Universe 2 there is no such agent A. The question is: Which universe is better?

The theory that says that desire fulfillment has intrinsic value would argue that Universe 1 is better than Universe 2. This is because Universe 1 contains this intrinsically valuable state of desire fulfillment, while Universe 2 does not.

The theory that says that value is a relationship between states of affairs and desires says that both universes have equal value. Within Universe 1, states of affairs in which P is true has value to Agent A. That is to say, Agent A has ‘reasons for action’ to create or preserve states of affairs in which P is true. However, the fact that A exists and P is true has no value unless there is somebody with a desire that can be fulfilled by a state of affairs in which A exists and P is true.

Many readers would likely choose Universe 1 over Universe 2, saying that Universe 1 is better. However, this is because you, my reader, have desires. If you had to choose between Universe 1 existing and Universe 2 existing, you would likely choose Universe 1, because this would fulfill the more and stronger of your own desires.

However, imagine how a creature with no desires would choose in this case? By definition, this person would not care one way or the other. He has no reason to choose Universe 1 over Universe 2 if he has no desires that can be fulfilled by Universe 1 but not Universe 2.

Argument Against Intrinsic Value

The section above is not an argument that aims to show that desire utilitarianism is correct. It simply exists to illustrate the difference between theories that state that desire fulfillment has intrinsic value (which would support desire-fulfillment act utilitarianism), and that value exists as a relationship between states of affairs and desires (which supports desire utilitarianism).

The argument against the intrinsic value theory is simply an application of J.L. Mackie’s “Argument from Queerness”. We can quite adequately explain all relevant observations in the real world by postulating desires (as motivational propositional attitudes – a desire that P motivates the agent who has it to bring about or preserve states of affairs in which P is true). We do not need any other type of entity. By the powers vested in me by Occam’s Razor, intrinsic values do not exist. Only relationships between states of affairs and desires exist.

If somebody wants to claim that desire fulfillment has intrinsic value, I am going to ask, “What is this intrinsic value? Do states with intrinsic value somehow emit some form of ‘goodon radiation’ that we have evolved a faculty to detect? Why would evolution have given us such a faculty? Evolution is going to give us a tendency to like that which promotes our genetic replication, not that which radiates goodons, unless there is some reason to believe that these are the same thing. You have no way of making sense of this ‘value as an intrinsic property’ theory. So, I am going to stick with my ‘value as a relational property’ theory.

By the way, relational properties exist. They are as much a part of the real world as the objects or states they relate. So, the theory that value is a relational property is still a theory that says that values are real, they can be known, and that they are things about which people can be (objectively) mistaken. So, this is still, very much, a realist theory of value.

It is a theory that says that the value of desires themselves rests in the relationships that they have to other desires – that claims about the value of a desire are objectively true or false.

Acts and Desires

The next question is: Why is the evaluation of actions tied to desires? Why not evaluate actions directly by their consequences?

The answer to this question is because there ain’t no such thing as a free will.

A lot of moral philosophy is wasted on the idea that we can evaluate actions on their own intrinsic value. However, if this is the case, then we must somehow be able to perform an action based on its intrinsic value. That is to say, the property of the intrinsic value of an action must somehow be causally linked to the muscle contractions that cause the action.

The fact is, there is no such power. Intentional actions are the product of beliefs and desires. Whenever we say that an agent “ought to have done X” we must infer that the agent “ought to have had those desires that would have caused him to do X.” Since desires are persistent entities that will influence a number of actions over time, in saying that an agent “ought to have done X” we have to ask what those desires would then cause the agent to do under other, common, every-day circumstances.

If that desire would produce a good result in this highly unusual circumstance, but bad results in more common circumstance, then it is NOT the case that we have reason to promote that desire. That is to say, it is not the case that the agent ought to have done X.

For quite some time moral philosophers have been trying to argue that morality requires some form of free will. This conclusion came from the fact that moral philosophers were busy trying to evaluate actions independent of the desires that caused them. The fact that this practice requires such an absurd metaphysics is reason enough to discard it, and replace it with the idea that actions are evaluated according to whether or not they would fulfill good desires, where good desires are desires that people generally have reasons to promote or inhibit through social forces.

Do people generally have more and stronger reasons to turn the dial on a desire up – making the desire stronger through social forces? Or do they instead have more and stronger reasons to turn the dial on a desire down, making it weaker and less common? Which option will lead to more overall desire fulfillment, and less desire thwarting?

What will be the effect, in terms of desire fulfillment, of turning the dial for the desire to rape children up, or down? Down is the only option that reduces the thwarting of somebody’s desires. This tells us that there is more and stronger reason to bring the weapons of social influence to bear on inhibiting such a desire.

8 comments:

Anonymous said...

Alonzo,

I’m left with just a few problems with your arguments.
First In your alternate worlds, one has a desire and a state of affairs and a relationship, therefore as you say, value to an agent. Since one has something of value and the other does not, I fail to see how they are equivalent.
Second, you discuss desires fulfilling desires. Desires cannot fulfill anything, just as propositions cannot fulfill propositions. Only states of affairs, or acts can fulfill desires or propositions.
Finally your argument about wishing to extinguish a bad desire would also work with the desire to win a tournament. Only one person or team can win the tournament, leaving everyone else’s desire thwarted. To avoid this we should try to eliminate the desire to win?

Alonzo Fyfe said...

First: The two universes are equivalent to whom?. In order to have a preference for one universe over another, one must have a desire. The only agent with a desire in my example is Agent A, who has a desire that P.

Now, if P is true in Universe 1, and P is false in Universe 2, then A will have a reason to choose Universe 1 over Universe 2.

However, it is also the case that if P is false in Universe 1, and true in Universe 2, he has a reason to choose Universe 2 over Universe 1.

However, a neutral third party, with no desires, has no reason to choose one universe over the other. Only an entity with a desire has a reason to choose.

Second: A state of affairs in which a desire exists can fulfill other desires - either directly (there is a desire that the desire exist) or indirectly (the desire causes an agent to act in ways that create states that fulfill other desires).

Desires are entities that exist or not, so they are parts of states of affairs. They also have effects - specifically, through their influence on intentional actions - thus creating states of affairs.

Third: I believe that the desire to compete exceeds any desire to win. Only one person gets to fulfill the desire of winning a tournament, but the everybody gets something of value from a tournament. Many people choose to participate even when they they have no chance of winning. This would be inconsistent with the hypothesis that there is only a desire to win.

Indeed, the desire to win is expected to be tempered to a level that it enhances, not detracts from, the value of participation.

Anonymous said...

Another question from the studio audience:

These sorts of problems are facinating, but it almost goes without saying that the 1000 Sadists are always morally wrong. Seeing how a particular theory explains this (or fails to) can be very informative, but it's fairly irrelivant on a practical scale.

I'm interested in how a more contemporary issue would be tackled. For example, pornography (of the legal variety). Views on this seem to have flucuated in recent history, and vary widely right now from those who recoil in horror at the mere mention of it, to those who consider it a healthy indulgence. Would a desire to view pornography be considered good, bad, or neutral? Is it even a malleable desire to begin with?

Anonymous said...

Your assertion contra free will is wrong.

Example 1:
I can have two desires that both have supporting beliefs and societal praise/indifference, but I can one fulfill one and the other is lost. I have the capability to make a choice between those two desires.

i.e. saving the life of two people, or being with a dying relative versus spending one night with the love of your life before she marries someone else, etc. etc.

Example 2:
I can have a desire that is backed up by a belief and that has either the praise or indifference from societal/legal pressures. I can choose to not fulfill that desire. I can even choose to thwart that desire and/or do the opposite.

i.e. getting married, saving a life, giving to charity

Both of these simple examples show that free-will/compatibilism are possible. Even if our desires, beliefs, mental states, and culture limit our choices we still have the cognitive and metaphysical ability to choose between alternatives.

Emu Sam said...

Very few acts that we take are a result of only one desire, or only conscious desires. At the precise moment that we make a decision, that choice fulfills the more and stronger of all our desires.

I think you are right that the possibility of desires being perfectly balanced should be considered. However, I suspect it is difficult for this perfect balance to last long. After a moment of consideration, the strengths of many of our desires might change subtly, which makes it possible to make a decision.

Anonymous said...

Anon - yes, you can choose between two competing desires, or choose to not fulfill a certain desire. But why do you choose A over B? What reasons-for-action are there for you that compelled you to take action A over action B? I would guess that there are more/stronger reasons for A than for B, which explains why you chose A and not B. And if any reasons-for-action other than desires exist, they have yet to be demonstrated.

So sure, you have the freedom to choose, and you'll choose that which you feel fulfills the most/strongest of your desires. Freely. Or deterministically. Which word you wish to use is up to you.

Jayman said...

Why, according to desire utilitarianism, should we attempt to remove the desire to rape from the 1000 instead of creating the desire to be raped in the one? Either way we would have a society where nobody's desires were being thwarted, correct?

You may say that, by definition, one cannot desire to be raped for it would become consensual sex. However, I think similar 1000 sadist problems could be constructed where the same basic question applies. For example, a world where 1000 people want to kill one person. Why not create the desire to be killed in the one person?

Marc said...

@Jayman
The answer to your last example is very simple: after killing the one person with the desire to be killed, the remaining 999 people would have their desires thwarted.

Same goes for your other questions.