Utilitarianism

Some Terminology

Consequentialist: According to consequentialist theories, the normative status of an action or practice is determined solely by the consequences of the action or practice in question.

Deontological: According to deontological theories the normative status of an action or practice may depend upon some features other than their consequences. On these accounts, the concept of duty is, in some sense, independent of the concept of good.

Hedonism: The doctrine that pleasure and pleasure alone is good.

Utility counted in hedons and negative hedons is pleasure - pain. Notice that the beauty part of talking about utility rather than some other good is that it is a common coin: it is precisely because different people have different preferences that we need to talk about utility.

The Principle of Utility (PU) : An act is right iff there is no other action that the agent could have done which has higher utility that it has. (i.e. iff it "maximizes utility.")

Act Utilitarianism (AU) as we shall understand it, or "classic" act utilitarianism, is the doctrine that the Principle of Utility (stated above) is the criterion for morality.

Some Objections to AU Classic

Doctrine of the Swine Objection: Pig Heaven isn't good enough: even if we could bring about at state in which everyone was happy as a pig in swill it would not be right to bring it about at the expense of higher pleasures.

No Time Objection: If we had to calculate utility every time we acted we wouldn't be doing anything but calculating utilities!

Trivial Actions Objection: Where trivial amounts of utility are at stake, it really doesn't matter whether you maximize. Failing to maximise in such situations could hardly be wrong.

Supererogatory Actions: On the utilitarians account, whatever is permitted is mandatory (except where more than one action produces maximal utility) so no action can be supererogatory, i.e. "over and above the call of duty." Clearly, however, some actions are and this poses a problem for the Utilitarian.

Self-Regarding Actions Objection: On the Utilitarian account, we have the obligation, all other things being equal, to maximize our own utility. My intuition is that we have no obligation to be good to ourselves.

Intentions Count Objection: On the Utilitarian account, only the consequences of an action count when assessing its normative status: the beliefs and intentions of the agent do not count at all. Surely this is counterintuitive.

Promise to the Dead Man Objection: There are some conceivable circumstances in which breaking a promise would produce higher overall utility than keeping it (if, e.g. one has made a promise to a person who subsequently dies, on a desert island where breaking one's promise will not set a bad example for anyone else, etc.). In some such cases it would seem that we are still morally obliged to keep our promise, contrary to what the utilitarian must hold.

Response: This may be a decisive objection. Notice, that for it to be decisive against consequentialism it is not necessary to show that promises ought always to be kept regardless of the consequences but only that there is at least one case where maximizing utility would not be right because it would be promise breaking.

Punish the Innocent Objection: In some cases, scapegoating an innocent victim may maximize utility and thus be, on the utilitarian account, the right thing to do. For most people this is counterintuitive.

Response: This may also be decisive.

Stepford Wives Objection: On the Utilitarian account it would seem that there is nothing particularly objectionable about a world of happy, lobotomized zombies and, if bringing about this state of affairs maximizes utility we should go for it--even at the cost of obliterating individual autonomy.

Many Vessels Objection: (1) Maximizing utility is not sufficient for rightness: consider big utility pie, inequitably divided. (1) Maximizing utility is not necessary for rightness: consider slightly smaller equitably divided utility pie.

The fundamental problem is that without some account of distributive justice, the utilitarian treats people (and other sentient beings) as mere vessels for holding utility. Intuitively however the distribution matters.

Rule Utilitarianism: An Alternative to AU

Act Utilitarianism, as we saw, makes it difficult to account for our intuitions concerning truth-telling, promise-keeping and, in general, basic honesty. Maybe we can live with this; maybe we can't. If we can't, then one "patch" for utilitarianism is to reject act utilitarianism, the doctrine that what we weigh in calculating utilities are the consequences of token actions, in favor of some version of rule utilitarianism, according to which the normative status of token actions is determined by their conformity to rules or codes of conduct which are in turn assessed according to the utility their adoption would produce.

Many different formulations of rule utilitarianism have been proposed. The common feature of such accounts however is the suggestion that the normative status of token actions is not determined directly by its consequences but rather by its conformity (or non-conformity) to some moral rule or code which is in turn assessed with respect to its consequences.[1] Thus, schematically, rule utilitarian theories look like this:

An act is right iff it is permitted by a correct moral rule or code.

A moral rule or code is correct iff _____

The difference between various versions of rule utilitarianism concerns conditions for the correctness of moral rules or codes, that is, the way in which we fill in the blank in the above schema. The version of rule utilitarianism we shall consider stipulates that a correct moral rule or code is one that has higher "currency utility" than any alternative rule or code that could be adopted.

The Problem of Extensional Equivalence

Before formulating this version of rule utilitarianism, one caveat for any proposed utilitarian theory: in formulating a rule utilitarian theory it is crucial that we avoid having it turn out to be extensionally equivalent to act utilitarianism. To say that two moral theories are extensionally equivalient is to say that they assign exactly the same normative status to any action: whatever is right (wrong, obligatory, etc.) according to one is likewise right (wrong, obligatory, etc.) according to the other. In otherwords the two theories are, in a sense, really just saying the same thing. Why is it so crucial to avoid formulating rule utilitarianism in such a way that it turns out to be extensionally equivalent to act utilitarianism? Because, silly, it would be pointless! The whole purpose of the move from act utilitarianism to rule utilitarianism is to avoid some of the problem cases where act utilitarianism seems to assign an intuitively incorrect normative status to various actions--punishing the innocent, certain cases of promise-breaking, etc. A rule utilitarian theory that is extensionally equivalent to act utilitarianism will assign exactly the same incorrect normative status to exactly the same actions, so it won't be any improvement at all!

It is precisely to avoid the problem of extensional equivalence that we have to formulate rule utilitarianism in what may seem on the face of it an overly complicated and fussy way. If we aren't fussy however Extensional Equivalence could sneak in the back door and we could be stuck a disguised version of act utilitarianism. As an example of how this could happen, consider the following version of rule utilitarianism that fails through being extensionally equivalent to act utilitarianism.

A Crumby Version of Rule Utilitarianism[2]

In formulating rule utilitarianism we may be tempted to fill in the blank in the account of what constitutes the correctness of a moral rule or code as follows:

(1) A moral rule or code is correct iff it maximizes utility.

This, however won't do because it's not clear what it is for a rule or code to maximize utility. Remember "maximize utility" means "causes the maximal amount of utility." Rules and codes as such don't cause anything: strictly speaking it is rather the actions which are done in accordance with them that cause things to happen. Consequently we need to reformulate (1). We may be tempted to reformulate it as follows:

(2) A moral rule or code is correct iff everyone's acting in accordance with it all the time would maximize utility.

The problem with (2) is that it renders the resulting rule utilitarian theory extensionally equivalent to act utilitarianism. Just think about it: the one rule which guarantees that everyone's acting in accordance with it all the time would maximize utility is precisely "Whatever happens, do whatever maximizes utility!" And this gets us right back to where we started: act utilitarianism. Consequently, if we want to avoid this, we need a more sophisticated version of rule utilitarianism. One such might be formulated as follows:

A More Sophisticated Version of Rule Utilitarianism

On this account, the correctness of a rule or code depends not upon everybody's following it all the time, but rather upon it's having maximal "currency utility." We examplain the notion of currency utility as follows:

A rule or code is current in a society iff most normal adults in the society believe that it is correct and sincerely attempt to act on it most of the time and believe that most other people believe that it is correct and sincerely attempt to act on it most of the time.

The currency utility of a rule or code for a society is the amount of utility produced by its being current in that society.

Given this notion of currency utility we can formulate an account of the correctness of moral rules and codes in terms of it as follows:

(3) An moral rule or code is correct iff it has maximal currency utility.

It can be shown that a formulation of rule utilitarianism in terms of (3) is not extensionally equivalent to act utilitarianism. We leave it an open question whether such a theory is satisfactory in other respects and whether it in fact adequately accounts for our moral intuitions.

---

[1] In talking about the consequences of moral rules or codes here I'm being deliberately vague--rules or codes, strictly speaking, do not have consequences.

[2] This version of rule utilitarianism is unacceptable because it is extensionally equivalent to act utilitarianism. The reason why we formulate an unacceptable theory like this is to show what we have to avoid in formulating an acceptable theory and why such a theory will have to be a little fussy.