Friday, January 09, 2009

Robert Nozick's Utility Monster

Mail from the studio audience is coming in at a fairly fast rate right now. Which is a good thing. I am trying to answer as many questions as I can.

One piece of correspondence came with a question that I really should have answered long ago.

I was curious if you would be willing to or already have (and would be willing to provide me a link) addressed the concept of the utility monster and how it relates (or doesn’t) to desire utilitarianism.

The “utility monster” was one of philosopher Robert Nozick’s objections to utilitarian theory.

Nozick postulated a creature who received 100 units of utility (pleasure, happiness) per unit of resource consumption, in a universe where everybody else received 1 unit of utility per unit of resource consumption. In this type of universe, Nozick argued, utilitarian would require that all of the people who got lesser utility be sacrificed (give up any and all resources) to the utility monster. This moral demand for sacrifice, however, is absurd. Therefore, basic utilitarianism is defeated by means of a reduction to absurdity.

How does desire utilitarianism handle the utility monster?

I will begin by asserting that morality is a tool used for the fulfillment of real-world desires. If I were to build a hammer, somebody might raise the objection, “How would that hammer work in a world where nails were all built of clay?” The answer is, it wouldn’t. However, this is still a perfectly good hammer for the real world, where nails are made of steel.

Having said that, as a desire utilitarian, I would also note that the utility monster, in this case, has desires that tend to thwart other desires. Desire utilitarianism focuses on the evaluation of desires - counting as good those desires that tend to fulfill other desires and bad (evil) those desires that tend to thwart other desires.

Let us assume that we had a choice between a utility monster who received 100 units of utility per unit of resource consumption, and one that received 100 units of utility for each act of kindness he performs. We certainly have more and stronger reason to prefer the second type of utility monster to the first. That is to say, we have more and stronger reason to use our social tools to promote the formation of the second type of utility monster and to inhibit the second type. That is to say, we have many and strong reason to call the latter 'good' and the former 'evil'.

It is better, in other words, for the utility monster to get its 100 units of utility from states of affairs in which the desires of others are fulfilled, than from states of affairs where the desires of others are thwarted through the loss of resources.

This, I would wager, is where the emotional reaction to Nozick’s hypothetical come from. We read the description and immediately note that we have an aversion to that type of situation – the desires we have that will be thwarted by the greedy consumption of all the resources by this monster.

Those desires are "reasons for action that exist" for each of us to act so as to prevent the realization of such a state – to avoid a state in which there is a utility monster commanding the consumption of all the resources. They give us reasons to promote a aversions to overconsumption, waste, and greed, so that more desires are fulfilled through the use of fewer resources.

This is, in fact, what we see in the real world – the use of social forces (praise and condemnation) to promote aversions to the over-use of resources. It is because of the desire-thwarting qualities of "utility monsters" that people act to inhibit the creation of a world in which people find utility in the over-consumption of resources.

Individuals will seek to act so as to fulfill the more and stronger of their own desires. As a result, if we assert that the utility monster is a creature that acts in ways that thwart the desires of others, it follows that others have little or no reason to feed the utility monster. Advocating that people do that which, by hypothesis, we are told that they have more and stronger reason to refrain from doing, is nonsense. It’s a contradiction built straight into the example.

Let us say, instead, that we are dealing with a non-malleable desire. The utility monster has extremely strong desires (obsessions, perhaps) that require the consumption of vast amounts of resources. Other people have non-malleable desires to use those resources to fulfill weaker desires.

In this case, desire utilitarianism says that we have stepped out of the realm of morality. We have a universe in which these two sets of beings are in an unavoidable conflict. As a matter of fact, each faction will continue to act so as to fulfill the most and strongest of its own desires, given its belief. The fighting will continue until one group or the other has been wiped out or conquered.

Moral concepts only apply when we have a faction that has malleable desires. Then (and only then) it makes sense to ask how social forces should be brought to bear to mold those desires. Which is the same as asking about the desires that the most and strongest people have reason to promote or inhibit - asking about what to call good and what to call evil.

The utility monster's resource-consuming desires are not desires that people generally have reason to promote.


Martin Freedman said...

Hi Alonzo

Let me first say I am enjoying the high quality of argument and clarity of your recent posts.

Now you say something intriguing here, that when two groups who have non-malleable desires clash over resources that DU does not apply, in the sense that this is no longer a moral issue. You also say that "The fighting will continue until one group or the other has been wiped out.". Well here are four questions or issues on this.

1. Is it the case that one side necessarily eventually wipes out the other? Indeed looking at predator/prey relationships,hawk/dove models or your favourite lion/antelope example, various equilibriums such evolutionary stable strategies can occur?

2. Surely there is some sense that a broad conception of DU (that is broader than just a moral framework) - certainly based on DF if nothing else - can be used as a framework to understand/model/predict such "equilibriums" as noted in (1)?

3. We can look at examples using, let us say, groups of "militant extremists" (religious or not) - they are "extremist" in the sense they are immune to social forces and "militant" in the sense they act upon these desires (obsessions) regardless of thwarting the desires of others - Now if the others are moral agents (amenable to social forces with malleable desires) one can certainly judge whether such a group is at fault or evil. And when two such groups clash one can still judge which, if any, if more or less evil - based on what the differential affect would be on moral agents - and so use DU to select which side to support (e.g. international sanctions) - if there is a significant difference.

4. Finally what about psychopaths/sociopaths who, by definition, are not influenced by social forces (or am I over-stating this)? Again there could be two such clashing groups. Is this any different to (3) or do they amount to the same thing? (Arguments as to whether sociopaths are born or made are I think beside the point, we could be talking about "philosophical sociopaths")

P.S. Have you tried switching on embedded comments? It makes for a much better commenting experience and only takes a couple of minutes. It now works fine AFAIK and various other bloggers I follow have all switched over.

Friar Zero said...

If I were to build a hammer, somebody might raise the objection, “How would that hammer work in a world where nails were all built of clay?” The answer is, it wouldn’t. However, this is still a perfectly good hammer for the real world, where nails are made of steel.

That's often my first intuition whenever I hear or read the utility monster argument.

Thanks for answering my question on the blog. I hope to start school (phi major) this fall and I find your blog a great asset in to think through topics philosophically.

JD said...

With regards to the hammer analogy, do I understand you correctly to be saying that there is no such thing as a utility monster, therefore the concern is unfounded?

It's true that there's probably no such thing as a utility monster in its most extreme interpretation - an entity that gains infinite utility from its consumption of resources - I don't think this invalidates the entire concept. Even if a utility monster only gets, say, twice the utility per unit of consumption that you do, what's to say that that doesn't give it a moral claim to more resources than you? After all, nails are not made of clay, but there's no shortage of entities in this world which seem to have a near-infinite hunger for resources.

Anonymous said...

You say that a utility monster that gets 100 utils from acts of kindness is preferable to one that gets 100 utils from a unit of a resource. This is no longer a utilitarian theory, because it places value on something besides utility. What constitutes an 'act of kindness?' Why does an act of kindness have moral value? Utilitarian theory cannot answer that without by definition appealing to utility. Your 'desire utilitarianism' must reach outside of utilitarianism to make moral judgments. Furthermore,by placing value on a kind action (emphasis on action) your theory becomes deontological and no longer consequentialist (utilitarian). You are not a utilitarian, and you do not adequately address the utility monster, through utilitarian / consequentialist means.

Anonymous said...

The Anonymous guy above me sums it up pretty well, but I'll throw my hat into the discussion as well.

There's nothing Good about "acts of kindness" Inherently under a system of utilitarianism. The only reason they may Seem good is they might bring more utility/happiness overall to the world. Kindness tends to do this. But that's missing the point entirely. Nozick's argument Assumes that More utility/happiness occurs in the world when everyone sacrifices themselves to the utility monster. Maybe the monster just gets absolutely no happiness out of doing kind things. In fact, maybe he even becomes Less happy doing kind things.

In other words, let's say the utility monster eating all the resources at the expense of everyone else equals 1000 units of overall happiness in the world.

And the utility monster doing "acts of kindness" for everyone equals 900 unites of overall happiness in the world.

In this scenario, for a utilitarian it would be More moral for everyone to sacrifice themselves to/be killed by the utility monster. Yet this seems wildly immoral. That is his point.

Anonymous said...

That's what people forget - we can choose what kinds of institutions we create.

For instance you can have a social institution like the British NHS, which is very similar to your idea of a modified utility monster that gains utility from acts of kindness. It could not function if it didn't heal the sick, free of charge.

On the other hand, you can have a typical business corporation, which essentially consumes human beings and natural resources and produces nothing except an ever-expanding balance sheet.