10 PHILOSOPHICAL STUDIES

implied by it. It may just be possible that this alternative is exactly the one
that would allow the "good-reasoners" in ethics to indulge in truth-talk
about nomative and valuative sentences without going over to objectivism.
I do not myself, however, think it would break the present "deadlock in
ethics" between the objectivists and their opponents.
Received September 2, 1958
NOTES
1 The Moral Point oi View: A Rat/anal Basis of Ethics, Contemporary Philosophy,
Max Black, ed. (Ithaca, N.Y.: Cornell University Press, 1958).
The Structure oi Metaphysics (London: Routledge and Kegan Paul, 1955).
8I seem to find an equivocation in Baier's thought which I cannot explore here
between a true reason for an action and truly a reason for an action or between stating
a true rule and truly stating a rule, which may have helped bridge the commonly recog-
nized boundary between truth (of statements) and goodness (of making them).

The Senses of "Ought"

by P E T E R GLASSEN
UNIVERSITY OF MANITOBA

IN A recent paper 1 Jason Xenakis has asserted that it is a mistake to hold,

as many ethical philosophers do, that the word "ought" as it occurs in such
pairs of sentences as "You ought to go there because you promised" and
"You ought to go there because he'll remember you in his will" is used in
two different senses, a "moral" sense in the first case, and a "nonmoral,"
"prudential," or "motivational" sense in the second case. Rather, he main-
tains, "ought" is used in the same sense in these two sentences and it is
only the ground for saying that the agent "ought to go there" that is differ-
ent. The ground in the first case is a moral ground, and in the second case
it is a prudential ground. In this paper I want to do two things: I want to
show that Xenakis himself is mistaken and that "ought" is used in two
different senses in the above sentences; and I want to explain what is
common to, as well as what distinguishes, the different senses of "ought"--
for that word does, indeed, have several senses.
What are Xenakis' reasons for saying that "ought" as it occurs in these
two sentences has the same meaning? I can discover only two, neither of
them a cogent one. His first reason seems to be that if one takes the posi-
tion that "ought" in these sentences has two different senses, "one might as
THE SENSES OF "OUGHT" 11
well say that in 'John is dejected because he had a bad night's sleep' and
'John is dejected because his wife came back to him,' the copula occurs in
two different senses." ~ Now this is surely a non sequitur. Even if "ought"
does have two senses, it would not follow that the copula in these examples
has two senses; in other words, there would be no inconsistency in main-
taining that "ought" in the examples given has two senses while the copula
in the examples given has only one sense. One might as well say that since
the copula in the examples given has only one sense, the word "may" as
it occurs in the following examples has only one sense: "He may vote
because he is a citizen" and "He may vote because he's become fed up
with the present government." But, of course, it is clear that in the first
sentence "He may vote" means "He is allowed to vote" and in the second
"It is possible that he will vote."
Xenakis' second reason is that the thesis that "ought" in the examples
given has two different senses "leads to absurdities, for example, the ab-
surdity that what is enjoined in our examples (if not indeed the actual
going there as well) is both 'moral' and 'nonmoral.'" 3 Now I am not
certain that I understand this, but insofar as I do, I fail to see any absurdity.
It is surely quite possible that the same act regarded from one point of
view should be moral and regarded from another point of view should
be, say, prudential. Thus my keeping of a contract is, from one point of
view, the morally right thing to do; but from another point of view, it is
the prudent or wise thing to do, since it will keep me out of trouble with
the law, and furthermore, will leave my credit unimpaired.
I conclude that Xenakis has not shown that "ought" as it occurs in his
two examples has only one sense. Still, it may have only one sense even
though he has not shown that it has; what reason, then, is there to believe
that it has different senses? First, however, let us note that Xenakis him-
self admits that "ought" is not unambiguous: he would not wish to main-
tain that "the weatherman's 'ought,' as when he says 'It ought to rain
tomorrow,' is not different from, say, the reformer's." 4 But if "ought" has
a sense here different from the first two examples, it may have senses in
those examples different from each other. And now, what is the difference?
To see what the difference is, let us ask ourselves whether there is any
other way in which we could say much the same thing as we want to say
by the sentence "You ought to go there because you promised." Clearly
there is. Instead of saying "You ought to go there," we could say "It is
your duty to go there," or "You have an obligation to go there," or "You
are morally bound to go there," or "It is morally obligatory for you to go
there," or "It is morally incumbent upon you to go there." All these ex-
pressions amount to much the same thing, there being only differences
12 PHILOSOPHICAL STUDIES
in degree of emphasis and formality of tone. But in the sentence "You
ought to go there because he'll remember you in his will," the "ought"
cannot be replaced by any of these expressions: we are not saying "It is
your duty to go there," "You are morally bound to go there," etc. In other
words, there is a sense of the word "ought" in which expressions in which
it occurs can be replaced by expressions in which the word "duty" or some
other equivalent moral term occurs, and this sense may, therefore, properly
be called the moral sense of "ought", and there is a sense (indeed, there
are several senses) of the word "ought" in which expressions in which it
occurs cannot be replaced by expressions in which a moral term occurs,
and these senses may, therefore, properly be called nonmoral senses of
"ought" (one of these nonmoral senses being the prudential sense).
Is not Xenakis himself forced to recognize the distinction, at least im-
plicitly? For, in discussing the reasons why the speaker says "You ought to
go there," he says: " . . . we may, if we wish, call the reason in the first
example 'moral' and the one in the second 'nonmoral.' . . . And we may
do this on the ground that the former involves (tacit) appeal to a duty
(or prima-facie duty), what some of us like to call a moral rule, namely
that promises ought to be kept . . . whereas the latter involves no such
appeal." 5 Now in what sense is the word "ought" used in the "moral rule,"
"promises ought to be kept"? Surely if this is a moral rule it is used in a
moral sense; and that it is so used becomes evident when we consider that
the same moral rule could be expressed in a different way, e.g., "It is one's
duty to keep one's promises," or "Keeping promises is morally obligatory."
(We may point out in passing that the former example does not make a
tacit appeal to a duty, but is an explicit assertion that the agent has a duty,
or that something is the agent's duty.)
Granting, then, that "ought" is used in different senses in Xenakis' two
examples, a moral sense and a prudential sense, is there anything in com-
mon between the two and between them and any other senses it may
have? And is there any systematic way of exhibiting the difference among
its senses? The answer to both questions is in the affirmative. Let us begin
by considering the moral sense of "ought" again. To say to X, speaking
morally, "You ought to go there" is to say much the same thing as "It is
your duty to go there" (except that the latter way of speaking is more
forceful than the former). Very well; now if it is true that X ought to go
there or that it is his duty to go there, what can we say about him if he
does not go there? W e can say that he violated his duty--and to violate
one's duty is to do wrong. If he had gone there, he would have performed
his duty--and to perform one's duty is to do right. Hence, to say "You
ought to go there," speaking morally, is to imply, "If you do go there you
THE SENSES OF "OUGHT" 13
will be doing what is right and if you don't go there you will be doing
wrong." Indeed, this is the way in which we can find out whether some-
one does intend to be using "ought" in a moral sense: if someone tells us,
"You ought to do x," we can ask him, "Do you mean that I would be
doing wrong if I didn't do x?" and if the answer is in the affirmative, we
know he is speaking morally. Thus, in this case, the sense in which "ought"
is being used can be specified by reference to the contrary moral terms
"right" and "wrong": the action that one ought to do, in this sense, is the
one that it would be right to do and wrong not to do.
What about the prudential sense of "ought," the sense in which it is
used in the sentence "You ought to go there because he'll remember you
in his will"? Here we can say that if the person addressed does go there
(and it is true that he ought to go there) then he will be doing the wise
or prudent thing, or acting wisely or prudently; and if he does not go there,
he will be acting unwisely or foolishly. Thus this sense of "ought" can be
specified in terms of the contrary prudential terms "wise" or "prudent" and
"unwise" or "foolish."
Now moral terms and prudential terms belong to a genus of terms vari-
ously called "normative terms," "value terms," "evaluational terms," and
(the expression I shall use) "terms of appraisal," this genus being usually
contrasted with another, that of descriptive terms. (Other terms of ap-
praisal would be such terms as "beautiful," "delightful," "horrible," "cor-
rect," "valid," "contemptible," "good," and their contraries or contra-
dictories.) On the basis of this distinction between terms we can state
what is common to and what distinguishes the different senses of "ought"
as follows: what is common to the different senses of "ought" is the fact
that each of its senses can be specified in terms of two contrary or contra-
dictory terms of appraisal; and what distinguishes its different senses is
the actual terms of appraisal in terms of which they may be specified. Let
us illustrate this by some additional uses of "ought." And first let us con-
sider the sentence in which, according to Xenakis himself, "ought" is used
in a different way from that in which it is used in his first two examples,
namely, "It ought to rain tomorrow." Xenakis himself does not tell us
what the difference is--or what the similarity is! But we can do both. For
if it ought to rain tomorrow, what can we say when tomorrow comes and
it does not rain? Why, that it is surprising or puzzling that it does not rain;
and if it does rain, it is not surprising (there being, as far as I know, no
other well-established contrary or contradictory to "surprising" in English).
Similarly, when we say "He ought to be here any minute now," we imply
that it would not be surprising if he were here any minute now and that
it would be surprising if he were not. Thus the sense of "ought" as that
14 PHILOSOPHICAL STUDIES
word is used in these sentences can be specified in terms of the appraisal
terms "surprising" and "not surprising." Further, in the sentence "You
ought to divide here rather than multiply" the sense of "ought" can be
specified in terms of "correct" and "incorrect," for what this sentence
says is that if you divide here rather than multiply you will be doing what
is correct, and if not, you will be doing what is incorrect. Again, in the
sentence "After a date the man ought to see the gift home," the sense of
"ought" is specifiable in terms of "proper" and "improper": he will be
doing what is proper if he takes her home, and he will be doing what is
improper if he lets her make her way home by herself.
There are, no doubt, still other senses of "ought," but I shall not at-
tempt to give an exhaustive list of them. Once we know the formula for
discriminating among these senses, as just explained, it is not difficult to
define any additional senses. To show this, let us consider a use of "ought"
which has presented a puzzle to many moral theorists for a long time, the
use we find exemplified in the following sentence: "'Of course it's his duty
to fight for his country; but if he sincerely believes that he ought not to
fight, then he ought not to fight." Two other examples of the same sort,
though less obviously so: "If they think it's wrong for them to get a
divorce, then, even if it isn't wrong, they ought not to get one" (here "it's
wrong for them to get a divorce" is equivalent to "they ought not to get
a divorce"); "She thinks it's her duty to give up her career for the sake of
her family. I don't happen to agree with her, but in view of her convic-
tions, I wouldn't respect her if she didn't" (here "I wouldn't respect her
if she didn't" indicates that the speaker would be prepared to say "I think
she ought to do it"). These examples are puzzling because they seem to
be either self-contradictory or paradoxical in some other way. The first
example seems clearly to say that he both ought and ought not to fight for
his country; the second, that even if it were not the case that they ought
not to get a divorce, they still ought not to get one; and the third that she
both ought to give up her career and that there is no reason why she ought.
Moreover, the implication seems to be that a man's duty, or what he ought
to do, depends on what he thinks he ought to do; but that what a man
ought to do should depend on what he thinks he ought to do seems very
odd to moral philosophers for a variety of reasons, not the least of which
is that the reason why a man would think he ought to do something could
not very well be that he thought he ought to do it.
Now this puzzle can be solved very simply if we realize that the sense
in which "ought" is used can always be specified in terms of two contrary
or contradictory terms of appraisal, and if we remember that there is a
difference between right conduct and morally good conduct on the one
THE SENSES OF " O U G H T " 15
hand, and wrong conduct and morally bad conduct on the other. 6 Right
conduct depends on the nature of the act itself, independently of any
motive of the agent; morally good conduct, however, must not only be
right, it must be done from a certain motive, namely, the sense of duty,
or the desire to do what is right simply because it is right. A man who
donates a million dollars to charity will be doing what is right, but his ac-
tion will not be morally good if his motive is merely to win prestige for
himself; and similarly, a couple who get a divorce may be doing wrong (ff
the views of certain religious groups are correct), but their conduct will
not be morally bad if they do not know that divorce is wrong, and if, had
they believed that it was, they would have refrained from getting one.
How does this help us to solve the puzzle presented by such a judgment
as "He ought1 to fight, but if he thinks he ought2 not, then he ought3
not"? Clearly, by suggesting that the various "oughts" in the above judg-
ment are not all used in the same sense. If "ought1" and "oughtz" were
used in one sense, and "ought3" in another, then there would be no con-
tradiction and no paradox, and the puzzle would disappearY Once we
realize this, the solution is obvious: "oughh" and "oughtd' are evidently
used in a sense which is specifiable in terms of "right" and "wrong," while
"ought3" is used in a sense which is specifiable in terms of "morally good"
and "morally bad." W h a t the above judgment is equivalent to is this:
"It would be right for him to fight and wrong not to, but if he thinks it
would be wrong, then it would be morally bad of him to do so." From
this we see that "ought," so far from not having any specifically moral
sense, as Xenakis claims, has at least two moral senses, s
Received February 24, 1959

NOTES
I " A Mistaken Distinction in Ethical Theory," Philosophical Studies, 8:69-71
(October 1957).
Ibid., p. 69.
Ibid., p. 70.
' Ibid., p. 71.
Ibid., p. 70.
6 Perhaps the best modem discussion of this distinction is to be found in Sir David
Ross's Foundations of Ethics (Oxford: Clarendon Press, 1939), Chapter 12.
This is, indeed, suggested by A. C. Ewing's discussion, in his The De~nition of
Good (New York: Macmillan, 1947), p. 121. But Ewing's failure to realize what is the
general formula for distinguishing the senses of "ought" makes his discussion unduly
prolix.
8 Ewing distinguishes (ibid., pp. 118-2 3 ) among "three different usages of ["ought"]
in ethics" as follows (I use his numbers but present them in a different order): (1) in
one usage, " 'The action we ought to do' may mean that action which is really preferable,
taking everything into account"; (3) in another usage, "we ought to perform an act if
in the light of the availab]e evidence it seems the preferable act to choose"; (2) in still
16 PHILOSOPHICAL STUDIES
another usage " 'ought' . . . is . . . used in a sense in which not to do what one
ought, or to do what one ought not to do, is always morally blameworthy."
Now Ewing's sense (2) is equivalent to my second sense, since to do an action that
is morally bad is to do something for which one is morally blameworthy, and vice versa.
Ewing's senses (1) and (3) are not clearly moral senses at all, or any other specific
senses, unless we state from what point of view an action is preferable--the moral point
of view, the prudential, or some other. However, even if we state that we are speaking
about an action that is preferable from the moral point of view, we are still being am-
biguous, for we might mean from the point of view of what is right, or from the point
of view of what is morally good. Hence it will not do to discriminate among the moral
senses of "ought" in terms of what is preferable on the whole or what is preferable on
the basis of the available evidence; we must discriminate among them by reference to
the pairs of contrary moral terms in terms of which the senses of "ought" can be
specified. And so far as my observation goes "ought" is always used in moral discourse
in a sense which can be specified in terms of "right" and "wrong" and their equivalents,
or in a sense which can be specified in terms of "morally good" and "morally bad" and
their equivalents.

Note on Contingent Properties

of Abstract Objects
by J A M E S W I L L A R D OLIVER

IN m s paper on definitions of 'Existence,' 1 Rescher asserts, w i t h o u t argu-

ment, "abstract objects . . . have no c o n t i n g e n t properties: all the prop-
erties of 3--e.g., being half of 6--are necessary (demonstrable)." This view
serves as the basis for criticism of a definition proposed by Leonard, b u t it
seems questionable, even apart from the usual questions of the justifiability
and desirability of making assertions about properties or contingencies at all.
A n example suffices: a " c o n t i n g e n t property" of 3 is surely the property
of having as a m e m b e r the class of books on m y desk now. If, in view of
the example, the notion of property is modified so that the having of a
particular entity as a m e m b e r is n o t a property, we m a y use another: the
property of being one of the classes which have as a m e m b e r the class of
books now on m y desk. Surely one making assertions a b o u t properties and
contingencies would have to regard this as a " c o n t i n g e n t property" of 3.
Received February 22, 1958
NOTE
1 "Definitions of 'Existence'," Philosophical Studies, 8:65-69 ( 1957).