A theory of Justice: Revised Edition

How To Do Things with Words
(Oxford, The Clarendon Press, 1962), esp.
pp. 99–109, 113–116, 145f.
8. Here I borrow from P. T. Geach, “Good and Evil,” 
Analysis,
vol. 17 (1956), pp. 37f.
356
Goodness as Rationality

---

expressions does not change in those statements that are counted as
advisory. It is the context that converts what we say into advice even
though the sense of our words is the same. Climbers, for example, have a
duty of mutual aid to help one another, and hence they have a duty to
offer their considered opinion in urgent circumstances. In these situations
their words become advisory. And so as the situation warrants, what we
say may be, and in some cases must be, reckoned as advice and counsel.
Accepting the theory of right already sketched, the constant descriptive
sense together with the general reasons why persons seek out the views of
others explain these characteristic uses of “good.” At no point must we
appeal to a special kind of prescriptive or emotive meaning.
It may be objected to these remarks that the theory of illocutionary
forces allows all that has been claimed by those who have proposed a
prescriptive or an emotive theory of meaning. If so, there may be no
disagreement. I have not denied that the understanding of the illocution-
ary forces of the various uses of “good,” its being employed in statements
of praise and advice, and the like, is relevant to grasping the meaning of
the term. Nor do I oppose the view that a certain illocutionary force is
central to “good,” in the sense that one cannot accept as true the statement
that something is good and at the same time dissent from its illocutionary
force (assuming this force to obtain in the context).
9
The question is how
these facts are to be explained.
Thus the descriptive theory maintains that “good” is characteristically
used with the force of a recommendation or advice, and the like, precisely
because of its descriptive sense as given by the definition. The descriptive
meaning of “good” is not simply a family of lists of properties, a list for
each kind of thing according to convention or preference. Rather in the
way that the definition explains, these lists are formed in the light of what
it is rational to want in objects of various kinds. Therefore understanding
why the word “good” (and its relatives) is employed in these speech acts
is part of understanding this constant sense. Similarly, certain illocution-
ary forces are central to “good” as a result of its descriptive meaning, just
as the force of factual narration belongs to some utterances in virtue of
their descriptive meaning. For if we assent to the statement that some-
thing is best for us when it is offered as advice, say, we will indeed accept
this advice and act upon it if we are rational. The dispute, if there is one,
is not about these recognized facts but concerns the place of the descrip-
9. For these and other points, see J. O. Urmson, 
The Emotive Theory of Ethics
(London, Hutchin-
son University Library, 1968), pp. 136–145.
357
62. A Note on Meaning

---

tive meaning of “good” in explaining them. The descriptive theory holds
that conjoined to a general theory of speech acts the definition of “good”
yields an adequate account of these facts. There is no occasion to intro-
duce a distinct kind of meaning.
63. THE DEFINITION OF GOOD FOR PLANS OF LIFE
63. Definition of Good for Life Plans
To this point I have discussed only the first stages of the definition of
good in which no questions are raised about the rationality of the ends
taken as given. A thing’s being a good X for K is treated as equivalent to
its having the properties which it is rational for K to want in an X in view
of his interests and aims. Yet we often assess the rationality of a person’s
desires, and the definition must be extended to cover this fundamental
case if it is to serve the purposes of the theory of justice. Now the basic
idea at the third stage is to apply the definition of good to plans of life.
The rational plan for a person determines his good. Here I adapt Royce’s
thought that a person may be regarded as a human life lived according to
a plan. For Royce an individual says who he is by describing his purposes
and causes, what he intends to do in his life.
10
If this plan is a rational one,
then I shall say that the person’s conception of his good is likewise
rational. In his case the real and the apparent good coincide. Similarly his
interests and aims are rational, and it is appropriate to take them as
stopping points in making judgments that correspond to the first two
stages of the definition. These suggestions are quite straightforward but
unfortunately setting out the details is somewhat tedious. In order to
expedite matters I shall start off with a pair of definitions and then explain
and comment on them over the next several sections.
These definitions read as follows: first, a person’s plan of life is ra-
tional if, and only if, (1) it is one of the plans that is consistent with the
principles of rational choice when these are applied to all the relevant
10. See 
The Philosophy of Loyalty,
lect. IV, sec. IV. Royce uses the notion of a plan to characterize
the coherent, systematic purposes of the individual, what makes him a conscious, unified moral
person. In this Royce is typical of the philosophical usage found in many of the writers cited in §61,
note 2, Dewey and Perry, for example. And I shall do the same. The term is given no technical sense,
nor are the structures of plans invoked to get other than obvious common sense results. These are
matters I do not investigate. For a discussion of plans, see G. A. Miller, Eugene Galanter, and K. H.
Pribram, 
Plans and the Structure of Behavior
(New York, Henry Holt, 1960); and also Galanter’s
Textbook of Elementary Psychology
(San Francisco, Holden-Day, 1966), ch. IX. The notion of a plan
may prove useful in characterizing intentional action. See, for example, Alvin Goldman, 
A Theory of
Action
(Englewood Cliffs, N.J., Prentice-Hall, 1970), pp. 56–73, 76–80; but I do not consider this
question.
358
Goodness as Rationality

---

features of his situation, and (2) it is that plan among those meeting this
condition which would be chosen by him with full deliberative rational-
ity, that is, with full awareness of the relevant facts and after a careful
consideration of the consequences.
11
(The notion of deliberative rational-
ity is discussed in the next section.) Secondly, a person’s interests and
aims are rational if, and only if, they are to be encouraged and provided
for by the plan that is rational for him. Note that in the first of these
definitions I have implied that a rational plan is presumably but one of
many possible plans that are consistent with the principles of rational
choice. The reason for this complication is that these principles do not
single out one plan as the best. We have instead a maximal class of plans:
each member of this class is superior to all plans not included in it, but
given any two plans in the class, neither is superior or inferior to the
other. Thus to identify a person’s rational plan, I suppose that it is that
plan belonging to the maximal class which he would choose with full
deliberative rationality. We criticize someone’s plan, then, by showing
either that it violates the principles of rational choice, or that it is not the
plan that he would pursue were he to assess his prospects with care in the
light of a full knowledge of his situation.
Before illustrating the principles of rational choice, I should say a few
things about the rather complex notion of a rational plan. It is fundamen-
tal for the definition of good, since a rational plan of life establishes the
basic point of view from which all judgments of value relating to a