Sunday, August 03, 2008

Recap - Desire Utilitarianism

A couple of members of the studio audience made some comments recently that give me an opportunity to recap some of the central claims of desire utilitarianism. Since I put a lot of weight on that theory, I should take the time to explain it, from time to time, to new readers.

So, let me address those concerns.


That's an incoherent answer. A moral question is not one of fact. It's like asking "what color should we paint the car", instead.

First, the answer is obviously coherent, since you understood it. If it were truly inherent, you could not say that it was wrong. "2 + 2 = 5" is a perfectly coherent statement. It is because its meaning is so clear that we can be so certain that it is false.

Second, I disagree with your statement that a moral question is not one of fact.

Third, if moral statements are not matters of fact, then we should abandon them. There are two possible realms. There is the realm of fact, and there is the realm of fiction. If moral statements are not a part of the realm of facts, then they are a part of the realm of fiction and e should treat them as such. The idea that there is some 'third thing' - a realm other than fact and fiction - is false.

You claim to disbelieve in inherit value but your arguments rely on moral judgments being objective.

Moral statements are objective, but not intrinsic. They are relational properties.

Take the property of being '50 years old'. The statement that something is 50 years old is a scientifically objective statement. Yet, it does not describe an intrinsic property. It describes a relational property (a relationship between the age of the object and the number of times the Earth goes around the sun). Moral properties are objective relational properties.

Value statements relate states of affairs to desires. An object is good to the degree that it is such as to fulfill the desires in question. Moral properties are a species of value properties that relate desires to other desires. A malleable desire is good to the degree that it tends to fulfill other desires; bad to the degree that it tends to thwart other desires.

You asked whether I was a subjectivist, and I answered that this depended on what you meant by 'subjectivist'. I am a subjectivist in the sense that value depends on mental states (they are relationships between states of affairs and desires). However, I am an objectivist in the sense that beliefs and desires are real. And relationships between states of affairs and desires are real-world properties that we can discuss just like we can discuss the age of a rock or the distance between stars.

Moral properties exist only in regards to personal perspective, thus subjective.

You are merely making an assertion here. It is an assertion that has a great many problems. One of the worst problems is that is morality is something that we can make up as we go along, then that puts morality in the realm of fiction, and all statements in the realm of fiction are false.

Except that you consistently say desires that fulfill the desires of others are superior to other desires. So you are claiming that altruism is a higher moral value.

Not necessarily.

There is a difference between a desire that tends to fulfill other desires, and a desire that aims to fulfill other desires (altruism).

It may well be that the hard-core capitalists are right. It may be that greed or 'rational self-interest' in a system that protects property rights is a system that makes everybody better off (fulfills the most and strongest desires). In which case, this system would be good. It would be the system that people generally had the most and strongest reasons to promote (whether they knew it or not).

Now, when I call a state of affairs 'good' I simply mean that people have reason to bring that state of affairs about. The only reasons for action that exist are desires. (But desires do, in fact, exist.) A 'desire that P' is a reason to bring about a state of affairs in which 'P' is true. These relationships between states of affairs and desires exist in the real world. I do not make them up.

You cannot demonstrate moral principles or oughts the same way you can properties of nature. They are fundamentally dissimilar and none of your semantic tricks change this.

Again, this is a mere assertion. I have over 1000 blog postings where I do exactly what you say cannot be done. Now, some of my statements may be false. In fact, I often assert that these blog postings are guaranteed to contain at least one false statement. But the possibility of error does not prove that something is not objectively knowable. In fact, it assumes objectivity.

Fire is hot is an objective fact that can be determined regardless of beliefs or mental states.

Here, you are equivocating. I fully agree that moral values depend on mental states. But mental states are real. A person's mental states are as real as his blood pressure, his age, his height, his body temperature. All of these are scientifically knowable properties. His beliefs and desires are also scientifically knowable properties. Relationships between states of affairs and desires are scientifically knowable properties.

They are not a 'different kind of thing'.

Since we can study these relationships, claims that we make about them are objectively true or false. We should treat them as such.


You cannot claim that your ideas are as objective as scientific conclusions until you can show us the property of good and the property of bad under a microscope.

Well, I assume that 'under a microscope' is metaphorical. You can't get the property of being 50 years old under a microscope, but it is still an objective property.

But, yes, we can get moral properties 'under a microscope' as you say. Beliefs and desires exist as real-world. They are a part of the explanation of the movement of real-world objects through space (intentional actions, such as writing a post). We can study these entities like we study any entity we cannot see directly, by looking at their effects, and from the effects discovering facts about the things that bring about those effects (people's beliefs and desires).

Just like scientists do.

Just as we can study beliefs and desires, we can study relationships between other states of affairs to beliefs and desires. Those relationships, too, reflect scientifically objective claims about the real world. They are things about which we can be right, or we can be wrong.

You have found a system of moral appraisal but cannot use objective methods to compare it to other systems of moral appraisal.

Actually, yes I can.

I can compare them to religious systems because religious systems postulate entities that do not exist. Therefore, all religious systems can be thrown out.

I can compare them to systems that postulate some sort of intrinsic value, because intrinsic values do not exist. They can be thrown out.

I can compare them to dualist theories - theories that say that there are are two types of reality. These theories postulate separate and distinct realms of 'is' and 'ought' that are completely different from each other, but can somehow interact. I call this metaphysical nonsense. There is the realm of 'is', and whatever is not in the realm of 'is' belongs in ther realm of 'is not'.

'Should' and 'ought' have to do with reasons for action. Desires are the only reasons for action that exist (or, at least, the only ones we have discovered to date). So, a 'should' or 'ought' statement has to either relate the object of evaluation to real-world reasons for action (in which case they are objectively true), or they do not draw such a relationship (in which case they are objectively false).

This is especially true since you not only go beyond a non-cognitivist subjectivism (desires shaped by beliefs) to state a moral truth that desires which fulfill others are superior to other desires. Superior how? Show me how you come to this conclusion without using your own system's semantics as a basis.

Well, your demand that I cannot use my own system's semantics to explain the theory is something that no scientist could meet.

Try explaining evolution without speaking about genes, mutation, or natural selection?

Try explaining the heliocentric theory of the solar system without talking about stars and planets.

I will defend my theory using the same standards and procedures that scientists generally use to defend their theories. That does not include a prohibition on 'using your own system's semantics'.

Now, if you are going to recommend states of affairs, then you are going to have to talk about reasons for action that exist for realizing those states of affairs. It scarcely makes sense to recommend a state of affairs based on reasons for action that do not exist.

Desires are the only reasons for action that exist.

So, if you describe relationships between the objects of evaluation and desires you can make objectively true statements. Otherwise, your statements are objectively false.

Your idea of ethics begs too many questions and ignores too many meta-ethical problems. Too many holes to be taken seriously I'm afraid.

This is an amusing statement. I can tell how much time and effort you have put into studying my ideas, and yet you claim the expertise to claim to know what I have and have not ignored. You could not even put my ideas into your own words before criticizing them.

There's a book listed up there to the right, "A Better Place: Selected Essays in Desire Utilitarianism" that will tell you what I believe. And, if you do not want to buy the book, you can search for the phrase "desire utilitarianism" in this blog and find all sorts of postings that give the details to the theory.

The 'holes' that you have found are the holes you see in a straw man that is of your own construction. If you were to take your attention away from your straw man for a moment and look at the actual theory, I think you would not discover so many holes.

Though, I am more than happy to be proved wrong.


Anonymous said...

Very nicely handled. Say, could you point me in the direction of a link where you explain your criteria for judging desires. I would buy the book but I'm on a student's income and philosophy is not my major.


Anonymous said...

I'm a different anon, I swear. Anyway I was curious, is your idea of the best desires being those that tend to fulfill others based on game theory? Like how the prisoners dilemma shows that cooperation gets more optimal results?

Emu Sam said...

Adam R.,

I think these may help answer your question. The first has a section on the function of desires. The second gives an example of analysing a desire.

The short answer might be a quote from

A good desire is desire that tends to fulfill other desires.

The desire utilitarian does not measure utility exclusively in terms of pleasure, happiness, well being, or preference satisfaction. Desire utilitarians say that the good is found in all of the things that we desire. It is found in happiness to the degree that we desire happiness, and it is found in pleasure to the degree that we desire pleasure. It is found in the company of family to the degree that we desire the company of family. All value exists in the form of desire fulfillment.

Alonzo Fyfe said...


[I]s your idea of the best desires being those that tend to fulfill others based on game theory? Like how the prisoners dilemma shows that cooperation gets more optimal results?

No, it is not.

Ultimately, I hold that game theory does not have as much to do with morality as many game theorists claim. There are many flaws with game theory.

Game theory does not have provisions for people 'defecting' without getting caught.

It assumes that all games have the same payoffs - as opposed to the real-world fact that each game has a different payoff. So, a person can adopt the policy of cooperating on all of the trivial games only to defect when the payoff gets high.

Game theory does not make provisions for differences in power - areas where people can defect without the possibility of any sort of retribution.

Game Theory and Morals

Game Theory and Desire Utilitarianism

Free Ridership and Game Theory

The Ultimate Game

Alonzo Fyfe said...


I wanted to add a tad bit to my previous answer.

Desire utilitarianism proposes a different way of handling 'prisoner's dilemma' situations. That is to alter the payoffs so that prisoner's dilemmas do not happen (or they are extremely rare).

If people generally are given a desire to cooperate - if the mere act of choosing 'C' is taken to be an end in itself with its own distinct payoff - then prisoner's dilemmas become increasingly rare, and we do not need to worry about the complexities of game theory.

Alonzo Fyfe said...

adam r.

In addition to emu sam's suggestions, readers have reported that the following two articles have been particularly helpful in explaining desire utilitarianism:

The Hateful Craig Problem. "Why should I, who for purposes of this question hate everyone, wish to fulfill the desires of others?"

The 1000 Sadists Problem. "[I]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?"

Anonymous said...

Wait a minute, according to one of the posts you link Alonzo you are arguing for a moral system of judgement but you deny the existence of free-will?

Am I misunderstanding something? How is a moral system not useless in determinism? If I can't make a choice between alternatives then how can I be held responsible for my actions?

Even if we accept your ideas on controlling desires through praise, condemnation, and punishment isn't it thus useless to even discuss the subject since it will occur regardless in a determined world?

Anonymous said...

Let me see if I've gotten this straight. The reason that "desires that tend to fulfill other's desires" are better than other possible types of desires is because those desires are more likely to receive praise, not be thwarted, and be fulfilled?

I always try to sum stuff up in little sound bites, it makes it easier for me to understand. Am I anywhere near right?

i don't see what Anon is talking about with the free will issue. In this post you are discussing developing habits and changing desires.

Where is this post about free will, Anon?

Emu Sam said...

If there are real random elements in the universe, then determinism is not the case either. It is my understanding that we don't know if there are random elements.

If determinism IS the case, but we don't know what has been predetermined, then we still have reason to try to influence people to act in ways which fulfill our desires. Perhaps without a few words from us, our child would be more likely to become a thief and be sent to jail. We don't know if it's been determined that we will say those words or not. However, we desire our child to not go to jail. So we say the words. Perhaps it was going to happen, set in stone, and perhaps not.

If everything is predetermined, then our efforts to change the course of history are also predetermined. However, we can look back over history (our personal history, world history, or points in between) and see how decisions changed the way things happened. Those decisions may have been predetermined, but it doesn't change that someone changed that way thing looked to be going.

Martin Freedman said...


You have it upside down, one could not have moral system with free will. Now note that Alonzo is a compatibalist whereas I reject libertarian free will (LFW - which Is what I think you mean) - period - without needing to retain any notion of free will. In what follows, the difference between Alonzo and myself probably makes no difference practically.

It is a double fallacy to think that without LFW there is no responsibility. The first fallacy is an argument from consequences, that it may have been in fact the case that there is no such thing as responsibility and arguing for free will in order to have such a fiction is to no avail. However the second fallacy is that responsibility cannot work if there is free will.

Now responsibility is, in fact, possible but it only works if antecedent and proximate factors - specifically beliefs and desires - are the necessary and sufficient conditions for intentional actions. Free will breaks this connection. Whatever beliefs and desires one has - which are in turn the result of the individual's unique combination of nature and nurture or genes and environment - LFW can ignore all these antecedent factors, otherwise it is not LFW. However incorporated into one's belief and desires, when one deliberates over alternatives, is the prospect of credit and blame, reward and punishment, this prospect being due to being held accountable for one's action. These prospects are a constitutive part of the weighing of alternatives and so incorporated into finding the more and stronger of one's desires to act upon.

That is how responsibility works in properly socialized individuals, one of the challenges that DU shows how to resolve is how to increase the number of properly socialized individuals through social and material reinforcement. LFW argues that there is no necessary relation between reinforcement/ expected reinforcement, desires, beliefs, actions and consequences. What else is left but random behaviors independent of antecedent proximate factors and so no responsibility?

Alonzo Fyfe said...

adam r.

Let me see if I've gotten this straight. The reason that "desires that tend to fulfill other's desires" are better than other possible types of desires is because those desires are more likely to receive praise, not be thwarted, and be fulfilled?

Not quite.

It is not necessarily the case that people are more likely to praise desires that tend to fulfill other desires, because they might have the facts wrong.

For example, people might praise piety thinking that it will encourage God to protect them from natural disasters. They think that 'love of God' is a desire that tends to fulfill other deisres, so they praise it highly. Only, since there is no God, piety has no such effect. In this case, what they are likely to praise, and what they have real-world reason to praise, are not the same thing.

And desires that tend to fulfill other desires, perhaps, are easily thwarted or fulfilled.

Desires that tend to fulfill other desires are desires that people have the most and strongest reasons to promote.

That a particular desire tends to fulfill other desires is a fact about which the whole population could be ignorant. There is, in this way, moral facts that are independent of what people believe. Yet, they still have the most and strongest reasons to promote desires that tend to fulfill other desires.