a p p e n d i x

not presuppose this [theological] basis (if this is possible). I shall assume that all the views we discuss accept this background objective.

§2. The Statement of the Classical Principle of Utility (Sidgwick)

(1) Sidgwick gives a careful statement of the principle in Methods of Ethics, Bk. IV, Ch. I. I shall go over the main points, making a few comments for purposes of clarification: I define utilitarianism (“Universalistic Hedonism,” as he sometimes says) as the ethical conception that holds that the (objectively) right institution or set of institutions, or the objectively right conduct (of individuals), in any given circumstances is that which will produce the greatest amount of happiness on the whole. Or which will lead to the greatest net balance of happiness (agreeable feeling).

In this sum of happiness that is to be maximized, we are to include all individuals (persons), whoever they are, that are affected by the institution or conduct (that is, whose happiness is affected positively, or negatively). Actually, the classical utilitarians thought that in principle it was necessary to include all sentient beings and therefore all animals or living things that can experience pleasure and pain. Capacity and liability for these feelings requires them to be included. This is an important aspect of utilitarianism that we shall note later; for the moment let’s suppose that the consequences of institutions and actions are limited to human individuals and subsequent generations of individuals.

Formally, we can write the utilitarian principle the following way: Let u1, . . . , un be the utilities (numbers representing the degree of happiness) of the n individuals affected by the institution (or system of institutions) or the actions in question: say the n individuals in society, or whatever [group]. Let a1, . . . , an be the weight of these utilities. Then, the principle is:

To maximize: ∑ ai ui = a1 u1 + ... + an un

That is: the right alternative (institution or act (or whatever)) is that alternative (institution or act) which belongs to the set of feasible (possible) alternatives that maximizes this function. (Assume for the moment that there are no ties.)

It is immediately obvious that this principle is not that of ethical egoism: everyone’s happiness is taken into account and given some weight (assuming all the ai > 0).

[ 396 ]

Copyright © The President and Fellows of Harvard College

Four Lectures on Henry Sidgwick

(2)Let us at this point notice a very important feature of this principle: the ui’s are numerical measures of happiness, and for Sidgwick the ultimate good is agreeable feelings or agreeable experiences (or consciousness) (more on this later). These are states of mind or aspects thereof, and they are known, as it were directly, by introspection: they are, so to speak, complete in themselves (over a certain interval of time) and good in themselves (or in the case of pain, bad in themselves). The recognition of these feelings does not presuppose, or use, any principle that involves the concept of right, or justice, etc., or any concepts that fall under these. Thus, classical utilitarianism uses a notion of happiness and of ultimate good that is defined independently and, as it were, prior to all other moral notions, or at any rate, prior to those of right and justice and of moral virtue and moral worth. This is characteristic of teleological conceptions, and so utilitarianism is a teleological doctrine.11

Where it differs from other teleological conceptions is in its definition of good—of what it is that is to be maximized. Thus perfectionism says that we are to maximize certain forms of excellence (human and other perfections) or certain other values: beautiful things, or knowledge of the world (or the main structural parts, etc. thereof ), or some mix of these.12 (Sometimes the term “ideal utilitarianism” is used for this view, but this is a misnomer.) Examples of perfectionism are found in G. E. Moore and Hastings Rashdall and many other writers who give some weight to perfectionist values.

But classical utilitarianism defines the good to be maximized subjectively, that is, in terms of agreeable feelings or experiences (consciousness) of (human) individuals.

(3)This may seem like an overly narrow definition of good. I have used it first because it has a certain clarity and simplicity about it; and second, because it is Sidgwick’s (and Bentham’s and Edgeworth’s) view; and Sidgwick has several interesting arguments for it (which I shall mention later). His view is the sharpest statement of the strict classical doctrine and he resists all efforts to depart from it, especially those of Mill (whom we take up next); and Moore, etc.

If we wish, however, the utilitarian notion of goodness can be inter-

11.[See John Rawls, A Theory of Justice (Cambridge, Mass.: Harvard University Press, rev. ed., 1999), pp. 21–23, 35–36, 490–491, 495–496, for a discussion of teleological doctrines and their contrast with contract doctrines. —Ed.]

12.[See A Theory of Justice, section 50, on the Principle of Perfection. —Ed.]

[ 397 ]

Copyright © The President and Fellows of Harvard College

a p p e n d i x

preted much more widely; thus as the satisfaction or fulfillment of non-he- donistic human interests; or as the satisfaction or fulfillment of rational (human) interests, imposing certain tests of rationality, provided that these do not involve the other moral concepts (rights and moral worth, etc.): that is, we allow for a certain appropriately limited class of corrections to human interests or desires (by rational deliberation, etc.). Or more generally, we may think of the good as happiness defined as the successful execution of a rational plan of life (again with “rational” appropriately defined).13 Utilitarianism can be enlarged to include these variations, and many common objections to the doctrine are not as plausible against these forms of utilitarianism. Certainly Mill, for example, wants to categorize the good to be maximized in this sort of way (at least).14

The crucial feature of the definition of the good in utilitarianism is this:

(a) that it defines the good independently (from concepts of right and moral worth); and (b) subjectively: what is good is (i) agreeable feeling (or consciousness) (pleasure), or (ii) the satisfaction of rational individual inter- ests—defined relative to people’s actual interests—(with “rational” suitably limited), or (iii) execution of rational plans of life (happiness); and (c) it is in a certain sense individualistic: ultimate good is attributed solely to the conscious experience of individual persons and presupposes no objective relations. In any case, (d) it is the sum of this good (these goods of individuals) that is to be maximized. Perhaps the best way to be clear about what this sense of individualism amounts to is to contrast it with other views.

(4) Permissible Variations or Refinements: Let’s explain in more detail the notion of a permissible variation (or refinement) of classical utilitarianism: namely, what is the point of this notion and what do we mean by it? First its point: there is a tendency to use the term “utilitarianism” extremely loosely, so that there are many distinctly different kinds of moral conceptions said to be utilitarian. This looseness has an unfortunate effect: it obscures the structure of different moral doctrines, and we fail to keep in mind what is special about each. So we need a notion of permissible variation of utilitarianism to specify those variations all of which share the characteristic or special structure of the classical utilitarian view.

Now what is this characteristic structure? (a) First, it is the characteristic

13.[See here Rawls’s account of a person’s good in terms of the rational plan of life that a person would choose under conditions of deliberative rationality, taking into account the Aristotelian Principle. A Theory of Justice, sections 63–66. —Ed.]

14.See the Appendix to this lecture on permissible variations of utilitarianism.

[ 398 ]

Copyright © The President and Fellows of Harvard College

Four Lectures on Henry Sidgwick

structure that utilitarianism shares with teleological doctrines generally: namely, that the notion of the good is defined prior to and independent from the right (and all the concepts that fall under it); and then right is defined as maximizing the good. This way of introducing the right is one aspect of the natural intuitive idea that underlies utilitarianism. It is the idea that rational conduct and decision is maximizing the good: striving for the greatest good. (Contrast Social Contract Theory.)

(b) Second, the characteristic feature of utilitarianism as distinct from other teleological conceptions is that it defines the good subjectively, roughly speaking from the point of view of the subject: the individual human agent. What this means in this case is this:

(i)The good is defined as agreeable or desirable consciousness; or pleasure rather than pain; or as the satisfaction of desire, according to its intensity and duration.

(ii)The capacities for pleasure and pain, or the relevant desires and aversions, are those which, at any given time, people actually have. We start at each moment in our deliberations from those aspects of people as they are, or can be foreseen to be. Practical reason is based upon given propensities and desires.

Thus what is characteristic of classical utilitarianism is that it treats the person according to its capacities for pleasure and pain, satisfaction, etc. Its claims upon social resources depend upon these. And this is in contrast with some other views which regard the claims of the person differently, e.g. Social Contract Theory, Kant’s theory.

Now to define a permissible variation of utilitarianism: namely, one that preserves these features and does not introduce elements inconsistent with them. The idea is that in examining utilitarianism, we want to see if any view that has these features can be correct. It would be progress to show that all views with these features must be unsatisfactory.

For this reason, we might be willing (as I suggested earlier) to permit the view—when it strengthened its case and made it a better conception— to suppose that the good is defined as satisfaction of rational desires, where these are the desires individuals would have if they subjected their present actual desires to certain forms of rational assessment (by the principles of rational choice). This gives a different view (a variation), but we might want to count it as a permissible variation within the same structure.

The point is: we would not have found the basic fault of utilitarianism if we did not allow it this variation when it made the view better. (As dis-

[ 399 ]

Copyright © The President and Fellows of Harvard College