TIMM TRIPLETT

CHISHOLM'S FOUNDATIONALISM

(Received 14 November, 1979)

In the second edition of his Theory o f Knowledge (1977), Roderick Chisholm

proposes several refinements on the foundationalism which has characterized
his approach to epistemology) I wish to show that these refinements are
unsatisfactory as presently formulated.
Central to any successful foundationalist theory is an account of how the
basic or 'directly evident' propositions which serve as the foundation for
knowledge can confer evidence upon nonbasic or 'indirectly evident' proposi-
tions. But Chisholm's present system provides no such account, for, as I shall
argue, his epistemic principles and definitions do not explain how a directly
evident proposition can confer evidence upon an indirectly evident one, in
spite of the fact that they appear designed to explain this. And in the absence
of such explanation, as I shall also argue, Chisholm's system does not allow
for knowledge of propositions about the external world.
My arguments are not advanced in order to undermine a foundational
approach to knowledge, but rather to criticize what is perhaps the most care-
fully worked out version of foundationalism, in the hope that such criticism
will lead to an improved formulation.
Section I provides a brief exposition of the basic concepts and principles
of Chisholm's system. My arguments then follow in Section II.

One of the central concepts in Chisholm's revised foundationalist system is

introduced in the context of his new definition of knowledge. As this defini-
tion, once unravelled, also introduces us to the other central epistemiq
concepts to which we shall be referring, let us begin with it: 2

D6.4 h is known by S =df h is accepted by S; h is true; and h is non-

defectively evident for S. (p. 110)

Philosophical Studies 38 (1980) 141-153. 0031-8116/80/0382-0141501.30

Copyright 9 1980 by D. Reidel Publishing Co., Dordrecht, Holland, and Boston, U.S.A.
142 TIMM T R I P L E T T

The acceptance condition can be equated with the belief condition in the
traditional definition of knowledge. The third condition, which is designed
to avoid Gettier-type counterexamples, Chisholm defines as follows:

D6.3 h is nondefectively evident for S =df Either h is certain for S,

or h is evident for S and is entailed by a conjunction of proposi-
tions each having for S a basis which is not a basis of any false
proposition for S. (p. 109)

The terms 'certain' and 'evident' refer to significant points on Chisholm's

carefully delineated scale of epistemic value. The certain is for Chisholm
that which it is most reasonable to believe. More precisely:

D1.4 h is certain for S=df h is beyond reasonable doubt for S, and

there is no i such that accepting i is more reasonable for S than
acceptingh. (p. 10)

The phrase 'beyond reasonable doubt' is also defined:

DI.1 h is beyond reasonable doubt for S=af Accepting h is more
reasonable for S than is withholding h. (p. 7)

To withhold a proposition is to accept neither it nor its negation (cf. p. 6);

but the term 'accept' itself and the phrase 'more reasonable than' are
primitives in Chisholm's system. 'Accept' may be equated with 'believe' (cf.
p. 6), but neither term is defined independently of the other.
Certainty is the highest epistemic status a proposition can attain. Falling
immediately below what is certain on the epistemic scale is the evident:

D1.5 h is evident for S =df (i) h is beyond reasonable doubt for S and
(ii) for every i, if accepting i is more reasonable for S than
accepting h, then i is certain for S. (p. 12)

A proposition may be directly or indirectly evident. This distinction

brings out the essentially foundationalist character of Chisholm's epistemology.
Intuitively, a directly evident proposition is one which is evident without any
other proposition serving to justify or confer evidence upon it. Chisholm
expresses this intuition as follows: for any directly evident proposition p,
he holds it acceptable to say 'What justifies me in thinking I know that p
is simply the fact that p' (cf. p. 19). Indirectly evident propositions then are
those which have their epistemic status as evident propositions conferred
upon them by some other proposition or propositions.
 CHISHOLM'S FOUNDATIONALISM 143

What kinds of propositions turn out to be directly evident on Chisholm's

account? Chisholm's approach is foundationalist in a Cartesian sense in that
those propositions which are directly evident for a person S have to do in
some way with S's mental states - his thoughts, beliefs, feelings, perceptual
experiences, etc. "Thinking and believing provide us with paradigm cases of
the directly evident" (p. 21). Chisholm calls such directly evident propositions
'self-presenting', after Meinong, and provides the following definition:
D2.1 h is self-presenting for S at t =dr h is true at t; and necessarily, if
h is true at t, then h is evident for S at t. (p. 22)

A paradigm case of a proposition which is, when true, self-presenting for me

is that expressed by 'I seem to have a headache'. Examples of other
propositions which can be self-presenting are those about what a person
believes, what he seems to perceive, and what he seems to remember. The fact
that S believes that Socrates is mortal, for example, is said to be enough to
make it evident to S that he does believe that Socrates is mortal (pp. 21-22).
Chisholm does not equate the self-presenting with the directly evident, but
rather sees the former as a subset of the latter. A directly evident proposition
is roughly one that is entailed by a proposition that is self-presenting (cf.
pp. 23-24, especially D2.2). 3
Even the broader class of directly evident propositions does not extend
far enough to embrace any of those propositions which are often characterized
as being about the 'external world', or what Chisholm calls propositions about
"the ordinary things we know" (p. 11). An example of the latter would be
the proposition expressed by 'There is a cat on the roof'. Such a proposition
is not self-presenting for any S - for it may be true without being evident
to S - nor is it directly evident, since it is not entailed by any self-presenting
proposition.
Propositions about the external world, then, are for Chisholm propositions
that can only be indirectly and not directly evident. Since Chisholm
emphasizes that such propositions can be and are known (pp. 11, 16, 120-
121), he must, by D6.4, conclude that they do achieve the status of being
evident. And they must be rendered evident, ultimately, by propositions
that are directly evident (cf. p. 85).
Although directly evident propositions are certain, sharing with
mathematical and other necessary truths the highest epistemic status possible,
indirectly evident propositions are not certain, according to Chisholm, since
144 TIMM T R I P L E T T

it is not as reasonable for us to believe in them as it is to believe, for example,

propositions about the elementary truths of arithmetic and propositions
about what seems to us to be ttie case (pp. 10-11).
The relation between the directly and indirectly evident is of course a mat-
ter of great importance for Chisholm. As a foundationalist, he must explain
just how the former confers evidence upon the latter. He sets forth a series
of epistemic principles in Chapter 4 which appear intended to give at least
a partial account of the relation between the directly and the indirectly
evident. One of the most important of these principles will figure prominently
in the discussion to follow:
(C) For any subject S, if S believes, without ground for doubt, that
he is perceiving something to be F, then it is evident for S that
he perceives something to be F. (p. 78)

The predicate letter F is to allow as substituends only those predicates which

connote sensible characteristics, e.g. being white. Chisholm uses 'perceives"
in such a way that if S perceives something to be white, then it follows that
there is some ordinary external object which is white. What Chisholm means
by belief 'without ground for doubt' will be explained below.
Directly evident propositions do not entail indirectly evident ones even
if we take Chisholm's epistemic principles to be necessarily true, as Chisholm
himself surely wishes to. Even principle C, which moves us from the
directly to the indirectly evident, does not express this as a matter of the
directly evident entailing the indirectly evident. Consider two states of affairs:
(p) S believes without ground for doubt that he perceives something
to be white.

and

(q) S does perceive something which is white.

q can be indirectly evident to S, but p does not entail q. Principle C tells us

that it is necessary that if p is true, then q is evident to S (in this case
indirectly evident). But it does not follow that q itself must be true, for this
may be one of those cases in which a proposition is both evident and false, a
possibility which Chisholm explicitly allows (p. 15).
 CHISHOLM'S FOUNDATIONALISM 145

II

Let us return now to Chisholm's definition of the ~nondefectively evident

(D6.3). Since Chisholm holds that indirectly evident propositions are among
those we know, he must allow that some indirectly evident propositions are
nondefectively evident. By D6.3 there will then be indirectly evident propo-
sitions which are entailed by a conjunction of propositions each of which has
a basis for S.
When an indirectly evident proposition is entailed by a conjunction of
propositions, at least some of the propositions conjoined must themselves
be indirectly evident. For as we have seen, no directly evident proposition -
and hence no conjunction of them - could entail an indirectly evident propo-
sition. And since by D6.3 each of these conjoined propositions has a basis,
it follows that at least some indirectly evident propositions are such that they
have bases.
What does it mean for one proposition to be a basis for another? Chisholm
provides the following definition:

D6.1 e is a basis o f h for S =df e is self-presenting for S; and necessarily,

i f e is self-presenting for S, then h is evident for S. (p. 106)

As we have seen, some h's are going to have to be indirectly evident propo-
sitions. This requires, by D6.1, that there be some self-presenting proposition
e such that, necessarily, if e is self-presenting for S, then some proposition h is
indirectly evident for S.
It is at this point that a careful articulation of Chisholm's foundationalism
becomes essential to his account of how ordinary propositions about the
external world can be known. It has been noted that such propositions can
only be indirectly, not directly, evident. If~.they are to be known, they must,
on Chisholm's definition of knowledge, b e nondefectively evident as well.
Hence there must, by our above argument, be some indirectly evident
propositions which have bases. And therefore, if Chisholm is to explain how
propositions about the external world can be known, he must show that we
can plausibly expect there to be propositions (self-presenting by definition)
which can serve as bases for indirectly evident propositions. However,
Chisholm's definitions and principles as they stand do not explain how there
can be such propositions. Those of his principles which explain how proposi-
146 TIMM T R I P L E T T

tions become indirectly evident turn out to exclude self-presenting proposi-

tions from serving as bases for indirectly evident ones.
This result, while posing a problem for Chisholm's account of our
knowledge of ordinary propositions, is of course more directly a problem for
his foundationalism. For the source of the difficulty is a problem which must
be resolved by any adequate foundationalist theory: that of explaining how
directly evident propositions can serve as bases for indirectly evident ones.
Let us consider precisely why Chisholm's system does not provide such an
explanation: We have already seen that every proposition which is self-pre-
senting for S will be directly evident for S, and that no directly evident and
hence no self-presenting proposition can entail an indirectly evident proposi-
tion. How then can there be some proposition e such that necessarily, if e is
self-presenting for S, then there is some proposition h which is evident for S,
but only indirectly so? (This is, of course, what is required by D6.1 if there
is to be any indirectly evident proposition h which has a basis.) Since
this relationship between the self-presenting and the indirectly evident cannot
be a matter of entailment, it must be a matter of one of Chisholm's epistemic
principles saying that the relationship holds. (Again, we may presume that
Chisholm intends his principles to be necessarily true.) This leads us to
principle C as the keystone which states as a necessary truth the conditions
under which some directly evident proposition makes it the case that some
logically distinct proposition is indirectly evident.
Initially, principle C looks promising. It states that necessarily, if S
believes without ground for doubt that he is perceiving something to be white
(for example), then it is (indirectly) evident to S that he is perceiving
something to be white. According to principle C, it will always be the case
that a directly evident proposition will, under specified conditions, determine
another proposition to be indirectly evident. And this seems close to what is
required for a directly evident proposition to be a basis for an indirectly
evident one.
But a closer look at D6.1 and principle C brings difficulties to light. If
principle C is to express the conditions under which some e can be the basis
for some indirectly evident h, what precisely is our e and what is our h?
Obviously, h is the proposition that S perceives something to be white. As
required, h is indirectly evident for S.
Note that e cannot be the proposition that S believes that he is perceiving
something to be white. For although this proposition may be self-presenting
 CHISHOLM'S FOUNDATIONALISM 147

for S, it is not the case that necessarily, if it is so then h is evident for S. It is

consistent with the necessity of principle C that there be cases in which (i)
the proposition that S believes that he is perceiving something to be white is
tree and is self-presenting for S, yet (ii) h is not evident to S. As Chisholm
recognizes (pp. 74-75), this situation will arise when S believes that he
is perceiving something to be white and yet there also exist grounds for S
to doubt this belief. Hence by D6.1 the proposition that S believes that he
is perceiving something to be white cannot be a basis o f h for S.
Clearly, what we need for e is the whole antecedent of principle C (sub-
stituting 'white' for the predicate letter F):

(e) S believes, without ground for doubt, that he is perceiving

something to be white.

But if e is to be a basis for another proposition, it must by D6.1 be self-

presenting, s
However, it does not appear plausible to maintain that e is self-presenting.
It seems, first, that an analysis of e in terms of Chisholm's stated definition
of one of its central concepts - that of a belief that is without ground for
doubt - indicates that e is not self-presenting, and second, that Chisholm
does not in any case intend the concept of the self-presenting to embrace
epistemic propositions such as e. Indeed, we shall see that, in response to
recent criticism of his second edition definitions and principles, Chisholm
has chosen to revise and narrow his definition of the self-presenting in a way
that more clearly excludes epistemic propositions such as e from falling under
that concept.
But let us first consider an analysis of e in terms of Chisholm's second
edition definition of the self-presenting (D2.1). Such an analysis requires that
we consider more precisely what it means for a belief to be without ground
for doubt. Chisholm provides the following definition:

D4.3 S believes, without ground for doubt, that p = af (i) S believes

that p and (ii) no conjunction of propositions that are acceptable
for S tends to confirm the negation of the proposition that p.
(p. 76)

When one's belief is without ground for doubt, then, a quite complex
epistemic condition obtains with respect to that belief. (The complexity is
here not spelled out in full. That would require replacing 'acceptable' and
148 TIMM TRIPLETT

'tends to confirm' with Chisholm's definitions of these concepts.) Since this

complex epistemic condition is equivalent by definition to the condition one
is in when one's belief is without ground for doubt, it would appear that e
comes to the following:

(f) S believes that he is perceiving something to be white, and no

conjunction of propositions that are acceptable for S tends to
confirm the negation of the proposition that he is perceiving
something to be white.

Given the definitional equivalence expressed by D4.3, it would seem that e

is self-presenting if and only if f is. But f is not plausibly construed as self-
presenting. Its second conjunct, for example, is a contingent general proposi-
tion, and it appears that such a proposition can be true without being evident
for S. 6 The view that general propositions require, in order to be evident, the
support of more specific and epistemically more basic propositions would
seem to be both independently plausible and a position that Chisholm would
accept. (Cf. Chisholm on contingent general propositions pp. 64-67, especial-
ly p. 66.)
Thus, upon analysis, it does not look as though e can be counted as self-
presenting under Chisholm's second edition definition of that concept. Before
we consider his more recent, revised definition, it is instructive to examine
several passages in Theory of Knowledge that bear on his concept of the self-
presenting, to see both what Chisholm appears to intend by this concept and
what considerations have led him to revise his definition of it.
No passages, to my knowledge, suggest or clearly imply that epistemic
states are self-presenting; there are at least two passages, however, which
suggest that Chisholm does not intend epistemic propositions in general to
count as self-presenting. One such passage is to be found in the first edition
of Theory of Knowledge, where there is a fuller discussion than in the second
edition of the intended range and limits of self-presenting states. Chisholm is
careful to l~ote that propositions about one's thoughts (which are shown by
example to include those about one's beliefs, memories, and emotions) may
not be the only sorts of propositions capable of being self-presenting. He sug-
gests that some propositions concerning what one is "trying or undertaking
to do" are also capable of being self-presenting (1966, p. 29). There is at
least some reason to think that if Chisholm held that epistemic states could
be self-presenting (either because they are themselves thoughts or because
 CHISHOLM'S FOUNDATIONALISM 149

they constitute another category of potentially self-presenting propositions),

then he would mention this possibility in the present passage. But no such
mention is made.
And when Chisholm does, in the second edition, directly discuss the
nature of epistemic propositions, he does so in terms very different from
those that would be suggested if he believed epistemic states could be self-
presenting: he discusses epistemic states as normative states. To assert that
one is in some epistemic state with respect to a proposition, Chisholm says, is
to assert that one has some responsibility or duty with respect to one's belief
concerning that proposition (p. 14). But on any traditional account of
normative states, a person can have a duty without it being evident to him
that he has that duty. 7
Chisholm's second edition also includes an extended discussion of what
might be called iterative epistemic principles - for example, the principle
that knowing entails knowing that one knows. This discussion bears
importantly on the question whether e is self-presenting. An iterative
principle which asserted that necessarily, if Ksp, then S knows that Ksp
(where 'Ksp' abbreviates 'S knows that p') would, given D2.1, entail of some
epistemic states that they are self-presenting. For, on this principle, when-
ever Ksp were true, it would be evident to (indeed, known by) S that Ksp.
Hence Ksp would be self-presenting to S by D2.1. While it would not follow
that e is self-presenting, one would be rather hard pressed to explain why
an epistemic state such as e is not self-presenting given that one as complex
as Ksp is. However, Chisholm explicitly rejects such a strong iterative
principle, and thus appears to avoid the consequence that Ksp is self-presenting.
He bases this rejection on the following considerations:
We have emphasized that a proposition cannot be evident to a person unless the person
understands the proposition. Now it is possible that there is a person who does not yet
have the concept of evidence or of knowledge, but for whom, all the same, a certain
proposition is known. Such a person, then, would be one for whom it could not be
evident that anything is known or evident. Therefore a proposition may be evident
without it being evident that it is evident, and a proposition may be known without it
being known that it is known. (p. 114)
Chisholm does affirm a series of qualified iterative principles. The first o f
these, and the one most relevant to our purposes, is as follows:

(K1) If S considers the proposition that he knows that p, and if it is

evident to S that p, then it is evident to S that he knows that p.
(lO. 114)
150 TIMM T R I P L E T T

K1 appears to allow that the epistemic propositions it pertains to may be

true without being evident. For an epistemic proposition, e.g. Ksp, may be
true of some person S without having been considered by S. In such a case,
K1 cannot be used to establish that Ksp is evident to S.
Unfortunately, K1 does not avoid the conclusion that Ksp is self-
presenting as neatly as it appears to. Notice that in the quoted passage above,
Chisholm affirms a psychological principle to the effect that if a proposition
is evident to a person, it is understood by that person. On the basis of this
principle, Chisholm rejects a strong iterative principle in favor of the weaker
K1. Chisholm also affirms as another psychological principle that if one
understands a proposition, one has contemplated and reflected upon it
(p. 41). In a recent article, Herbert Heidelberger has shown that these psycho-
logical principles, together with D2.1, entail that no propositions are self-
presenting (cf. Heidelberger, forthcoming). D2.1 can be revised in an obvious
way to avoid this consequence, but Heidelberger notes that the revised
definition will then entail that Ksp is after all self-presenting, given K1 and
the psychological principles. Implausible results are shown to obtain if one
rejects the second, but retains the first, of the psychological principles.
Heidelberger further points out that the consequence that Ksp is self-
presenting presents serious difficulties for Chisholm. Either he must deny
that Ksp is self-presenting or adopt the view that self-presenting states need
not be confined to what is epistemically basic or prior. For if Ksp is self-
presenting, then p is directly evident, since it is entailed by Ksp. Butp may,
of course, be a proposition about the external world - the sort of proposi-
tion Chisholm's foundationalism cannot accept as directly evident or basic.
According to Chisholm (p. 20), propositions about the external world do not
constitute a 'proper stopping place' in the process of justification, for one can
reasonably ask what justification there is for accepting such a proposition.
In responding to Heidelberger, Chisholm writes (forthcoming, Section 10):
I believe it is accurate to say that, if we ...reject the remarks about the psychological pre-
conditions, and if we restrict self-presentation to the certain (as distinguished from the
evident), then we escape the difficulties that Heidelberger points out.

It would seem enough, in order to defuse Heidelberger's criticisms, simply

to reject the psychological 'pre-conditions' or principles noted above. Why
then does Chisholm go further and offer a revised and restricted definition of
the self-presenting. Apparently because he does not wish Ksp to be self-
presenting. Recall that Chisholm rejected a strong and unqualified iterative
 CHISHOLM'S FOUNDATIONALISM 151

principle because he held the psychological principle that a proposition must

be understood if evident. Now that he rejects that psychological principle,
there does not appear to be any reason for Chisholm to prefer the qualified
principle K1 to a stronger iterative principle such as the following:

(EK) If S knows that p, then it is evident to S that he knows that p.

Apparently, Chisholm would now accept EK, or something very close to

it. Otherwise, he would not feel the need to revise his definition of the self-
presenting. EK together with his original definition would entail that Ksp is
self-presenting. But he can retain EK and avoid the api~arentty undesirable
consequence that Ksp is self-presenting by requiring that a self-presenting
proposition be certain, rather than merely evident, whenever true. This is,
in effect, just what his new definition does. s Chisholm's decision to thus
restrict the self-presenting also indicates that he does not wish to opt for the
alternative of allowing some self-presenting states to be nonbasic, and this
means that he cannot allow propositions about what one knows to be self-
presenting. Again, his restricted definition neatly avoids such a consequence.
It follows, of course, that if e is to be self-presenting on this revised defi-
nition, it must be certain. Recall that e states that S's belief that he is
perceiving something to be white is without ground for doubt. It seems
unlikely that such a proposition can be accorded the highest possible epistemic
status for S. Propositions such as those expressed by 'S seems to have a
headache' or 'S believes he is perceiving something to be white' seem, when
true, to be considerably more reasonable for S to believe. It also seems more
reasonable to affirm basic arithmetical truths and other simple necessary
truths than to affirm a proposition such as e. But if other truths are more
reasonable to believe than e, it follows that e is not certain (cf. D1.4). Even if
it were maintained that S could ultimately, through reflection, achieve
certainty about a propositiofl such as e, it would not follow that e is self-
presenting on the new definition, for prior to reflection e would be true yet
not certain.
If it is true that e is not self-presenting, then principle C does not provide
the required linkup between directly and indirectly evident propositions. It is
not this principle that shows us how directly evident propositions can be
bases for indirectly evident ones. Nor do any of Chisholm's remaining
epistemic principles appear capable of showing this. The only other principle
that is a candidate for explaining how indirectly evident propositions can
152 TIMM T R I P L E T T

have bases is principle I (p. 84). But like the a n t e c e d e n t o f principle C, that
o f principle I contains the r e q u i r e m e n t that one's belief be wit.hout ground
for doubt. Hence any a t t e m p t to show that the antecedent o f principle I is
self-presenting will e n c o u n t e r the same difficulties f o u n d in trying to show
this for the a n t e c e d e n t o f principle C.
It thus seems d o u b t f u l t h a t indirectly evident propositions do have bases,
as Chisholm understands this term. This presents a p r o b l e m , as we have seen,
b o t h for Chisholm's foundationalism and for his a t t e m p t to account for our
knowledge o f ordinary propositions about the external world. I f Chisholm
wishes to retain his foundationalist structure and his anti-skeptical approach
to knowledge o f the external world, then his epistemological system appears
to be in need o f revision and supplementation.

University of Massachusetts, Amherst

NOTES

For a good characterization of the main tenets of foundationalism, see Annis (1977,
p. 345). Chisholm's foundationalist approach is apparent in his conclusion that "every
proposition we are justified in believing is justified, in part, because of some relation that
it bears to the directly evident" (Chisholm, 1977, p. 85). See text below for a more
complete discussion of the directly evident.
2 I retain the numbers and letters Chisholm uses to identify his definitions and
principles. All page references in the text are to Chisholm (1977), unless otherwise noted.
3 Chisholm's actual account of the directly evident is somewhat different from that set
forth here, but it is convenient, and approximately correct, to think of the relation
between the self-presenting and the directly evident as one of entailment in the sense of
strict implication. Nothing in the present discussion turns on the difference between
my account and the relation Chisholm actually has in mind. Compare Chisholm's defi-
nition of the directly evident (p. 24) with his definition of entailment (p. 137).
4 It will not do to attempt to avoid this conclusion by omitting from principle C the
condition that S's belief be 'without ground for doubt'. This would result in a principle
entailing one formulated in the first edition of Theory of Knowledge (Chisholm, 1966,
p. 45) and shown by Herbert Heidelberger (1969) to be false. Chisholm acknowledges
and discusses this problem in his second edition (p. 75).
I shall henceforth assume that e is true. Thus the issue more conveniently becomes
whether e is self-presenting, rather than whether e is self-presenting if true.
6 Even a contingent generalization entailed by a self-presenting proposition will not be
self-presenting. For example, although the self-presenting proposition that S is thinking
entails the general proposition that someone is thinking, the latter is not self-presenting,
for it may be true without being evident to S, e.g., when S is asleep and nothing is
evident to him (ef. Chisholm, pp. 23-24).
Nevertheless, the generalization that someone is thinking is one that can be directly
evident, since it is entailed by a self-presenting proposition (cf. Note 3 above). This
 CHISHOLM'S FOUNDATIONALISM 153

might lead one to conjecture that the second conjunct of f is also directly evident. How-
ever, there is no reason to suppose that this is so. The proposition that someone is
thinking achieves its status as directly evident only because it is entailed by a self-
presenting proposition (in this case, that expressed by 'S is thinking'). But it is neither
obvious nor likely that there is some self-presenting proposition that entails the second
conjunct of f.
Even those who hold that we discover by intuition what our duties are would not, I
believe, hold that in every case in which a duty obtains for a person, it is immediately
evident to that person, without reflection, that it does obtain. Reflection would seem to
be required before one's duties become clear to one. If this is so, it will not be self-
presenting to a person that he has a particular duty, for he may have the duty without it
being evident to him that he has it. Cf. Sir David Ross's remarks (1930, pp. 30-32) that
our actual duties are not self-evident to us and that even careful reflection can at best
produce a 'greater likelihood' that we will discover what our actual duties are. Cf. also
R. B. Brandt's discussion (1959, pp. 188- 189) summarizing the views of nonnaturalists
who hold that our duties are discovered by intuition.
s Chishohn's revised definition (fully stated in section 10 of his forthcoming article
in Grazer Philosophische Studien) is complicated by an attempt, not relevant to our
present purposes, to avoid 'I-propositions'. It is quite acceptable for us to think of
Chisholm's new definition simply as D2.1 with 'certain' substituted for 'evident'. In
Person and Object (1976, p. 25), Chisholm had already proposed a definition of the self-
presenting which was just such a variant on D2.1.

BIBLIOGRAPHY

Annis, D. B.: 1977, 'Epistemic foundationalism', Philosophical Studies 31, pp. 345-352.
Brandt, R. B.: 1959, Ethical Theory (Prentice-Hall, Englewood Cliffs, New Jersey).
Chisholm, R.M.: 1966, Theory of Knowledge, first edition (Prentice-Hall, Englewood
Cliffs, New Jersey).
Chisholm, R. M.: 1976, Person and Object (Open Court, LaSalle, Illinois).
Chisholm, R.M.: 1977, Theory of Knowledge, second edition (Prentice-Hall, Englewood
Cliffs, New Jersey).
Chisholm, R. M.: forthcoming, 'Replies to critics', Grazer Philosophische Studien.
Heidelberger, H.: 1969, 'Chisholm's epistemic principles', Nous 3, pp. 73-82.
Heidelberger, H.: forthcoming, 'The self-presenting', Grazer Philosophische Studien.
Ross, W. D.: 1930, The Right and the Good (Oxford University Press, London).