Hedonic Utilitarianism

Essay from the book Lucifer’s Question.

Hedonic utilitarianism is the tacit moral theory of the modern West. It is rarely stated explicitly, but it is implicit in most moral discourse.

Jeremy Bentham and John Stuart Mill proposed hedonic utilitarianism as a philosophical theory of value. See Classical Utilitarianism in the Stanford Encyclopedia of Philosophy.

Hedonic utilitarianism has two core assumptions:

• Hedonism: Pain is intrinsically bad for the experiencer, pleasure is intrinsically good for the experiencer, and nothing else is intrinsically bad or good.
• Altruism: We have a moral obligation to act for the good of others, not just for our own good.

Altruism depends on the definition of a moral circle: the others to whom we should be altruistic. The moral circle could be all human beings or all sentient beings. In this essay, I will assume the latter. Given the hedonism assumption, it is the least arbitrary choice.

The hedonic utilitarian believes that we have a moral duty to decrease the pain and increase the pleasure of all sentient beings. In other words, we should try to maximize the hedonic utility of the universe.

Hedonic utility is total pleasure minus total pain over time. This assumes that pain and pleasure are on the same dimension of experience, with one being negative and the other positive. For multiple minds, it also assumes that different minds have the same hedonic dimension, so we could in theory (if not in practice) sum pain and pleasure from different minds.

For an individual, hedonic utility is the integral of a single variable over time. Let’s call this variable “P”. When P is zero, the individual experiences no pain or pleasure. When P is positive, he experiences pleasure. When P is negative, he experiences pain. The intensity of the feeling is the magnitude of P.

For a collective, hedonic utility is the sum of individual hedonic utility for all members of the collective.

Pain and pleasure are qualia. They are qualities of subjective experience.

This has some important implications. We can’t measure, observe or compare qualia. They exist only to the subject, and only in the moment. We cannot directly study subjective experience, even our own. We can observe behavior that is correlated with pain and pleasure, such as facial expressions. We can also introspect about our mental states, but introspection isn’t scientific measurement or observation. Also, the data of introspection are memories about qualia, not the qualia themselves. There is no external scale by which we can measure pain and pleasure, and no way to compare qualia between different subjects.

We should keep those limitations in mind when we theorize about the nature of subjective experience. We have to rely on some beliefs about how qualia correlate with the data of observation and introspection.

The Case Against Hedonic Utilitarianism

Now, let’s consider some problems with hedonic utilitarianism:

• Hedonic utilitarianism violates our intuitive notions of fairness.
• Hedonic utilitarianism would produce a dystopia.
• Hedonic utility can’t be estimated in a practical way.
• Hedonic utility might be zero-sum for the individual.
• Hedonic utility might be zero-sum in the aggregate.
• Altruism is self-defeating.
• The hedonism axiom is unjustified.
• The altruism axiom is unjustified.

Two well-known thought experiments illustrate the conflict between ordinary moral intuitions and hedonic utilitarianism.

Ursula K. Le Guin gave a critique of utilitarianism in the science fiction story The Ones Who Walk Away from Omelas. Omelas is a society whose members are very happy, on average. But their happiness is sustained by the torture of a single innocent person. At a certain age, the citizens of Omelas are told about the basis of their happiness. Some accept it. Some reject it, and walk away from Omelas.

The story shows that ordinary moral intuitions are not just about average outcomes, but also about how outcomes are distributed.

Robert Nozick’s “utility monster” thought experiment also demonstrates the importance of how outcomes are distributed. The utility monster is a being who receives a disproportionate amount of benefit or harm per unit of resource consumed or withheld. Let’s assume that benefit and harm are defined hedonically. Compared to an ordinary person, the utility monster would experience much more pleasure from eating a cupcake, and much more pain from being kicked in the shin. If a utility monster existed, hedonic utilitarianism would prescribe outcomes that seem very unfair.

Suppose that the utility monster’s emotional experiences are a hundred times more intense than an ordinary person’s. According to hedonic utilitarianism, it would be worth depriving 99 people of a cupcake to give the utility monster a cupcake. Likewise, it would be worth kicking 99 people in the shin to prevent the utility monster from being kicked in the shin.

Now, suppose that the utility monster’s emotional experiences are a billion times more intense than an ordinary person’s. To maximize hedonic utility, we should organize society around the welfare of the utility monster. It might be worth throwing children into wood-chippers to make his lunch.

Thought experiments demonstrate that an intuitive moral theory can have counter-intuitive implications. There is no requirement that a moral theory must agree with moral intuitions. However, most moral theories are attempts to rationalize moral intuitions. If we reject moral intuitions, why posit a moral theory? Why not just reject morality entirely?

Ordinary moral intuitions are based on how society works, and societies are not aggregate-utility maximizers. Societies need to distribute benefit and harm in ways that make society function. They must reward socially productive behaviors, and punish socially destructive behaviors. So, any type of utilitarianism will violate ordinary moral intuitions. And no society could be based on utilitarianism.

However, most people don’t think about morality as instrumental to society. Instead, they think about morality as cosmic normativity and imperativity: cosmic good and bad, and cosmic rights and obligations. Morality involves the delusion that collective values and rules are cosmic. That’s why no moral theory fits moral intuitions.

See What is Morality? and Morality and Bad Faith.

The “experience machine” thought experiment (also proposed by Robert Nozick) shows the counter-intuitive implications of hedonism. Imagine a machine that allows you to experience anything, including pure ecstasy at all times. If you choose, you can have your brain removed from your body and plugged into the machine, allowing you to live out any fantasy, or simply wallow in ecstasy, for a hundred years. Then you would cease to exist. In reality, you would have done nothing, but you would have experienced paradise for a hundred years. Would you choose real life, with all of its struggle and suffering, or an artificial paradise?

Most people would choose real life. The thought experiment shows that feelings lose their meaning when they are detached from real-world consequences.

Let’s consider a utilitarian version of that thought experiment: a mouse brain utopia.

Suppose that we could biologically engineer mouse brains to be in a state of bliss for 10 years or so. Let’s imagine that each brain exists in a jar, with tubes and electrodes that keep it alive and generate certain types of mental activity.

Now, suppose that this is the most energetically efficient way to produce pleasure. According to hedonic utilitarianism, we should turn the Earth into a giant mouse-brain factory. We should use all the energy on the Earth for that purpose, eliminating all other life forms. Only a few humans would remain, to run the mouse-brain industrial complex as long as possible. Perhaps we could even create a mouse-brain Dyson sphere, to convert all the energy from the Sun into pleasure.

Most people would view this hedonic utopia as a horrific dystopia. The thought experiment demonstrates the counter-intuitive implications of an intuitive moral theory.

Now, let’s consider the problems with estimating net hedonic utility. To maximize it, we must estimate it, and predict the effects of action on it.

We cannot observe or measure qualia. We can theorize about the causes of pain and pleasure, and their effects on behavior. We can use that theory to estimate hedonic utility and/or predict the effects of action on hedonic utility. But there are many problems with that.

If we use observable behavior as a measure of hedonic effects, then emotionally expressive individuals would have greater moral significance than stoic individuals. That would incentivize squeaky wheels, making observable behavior a less accurate measure of emotional states. (See Goodhart’s Law.)

How could we estimate hedonic utility for other species? We don’t even know what types of life are sentient. Can we define sentience for organisms with different brains? Birds seem to be sentient, but their brains are different from mammalian brains. Fish are more different. What about an octopus? Could insects be sentient?

Even if we could predict the hedonic effect of an action on a single sentient being, we might not be able to predict the aggregate effect on all sentient beings. Actions have many consequences across space and over time. This is a problem for any type of utilitarianism.

What if we can’t increase hedonic utility?

I believe that pain and pleasure balance out. The gas pedal in a car goes up and down, up and down, but its net displacement over time is zero (within the car). I believe that the motivation mechanism in the human brain works in a similar way.

See Motivation.

If hedonic utility is net zero by its nature, then there is no way to increase it or decrease it. Actions might change how sentient beings feel at any given moment, but not the sum of their feelings over time.

Even if hedonic utility is not zero-sum for a single individual, it could be zero-sum (or close to it) in the aggregate. Life is competitive. Organisms compete for resources. Helping one organism is often harming another.

Trying to eliminate conflicts between sentient beings would lead to something like the mouse-brain utopia, in which nature has been replaced by a technological system that mass-produces hedonic utility.

Another criticism is that altruism is self-defeating, and so any hedonic utopia would destroy itself. Altruism is not adaptive, and so it would be eliminated by evolution. Those who try to increase hedonic utility would be replaced by those who try to reproduce.

Finally, and most importantly, hedonic utilitarianism is based on unjustified assumptions. It takes hedonism for granted as a personal value theory, and it takes altruism for granted as a moral theory. There is no reason to accept those assumptions.