Epistemic virtues in the first person

Paper with Manuel is to appear in the Croatian Journal of Philosophy
The idea was to show that epistemic virtues could also be not first-personal enough. I transcribe below a bit of the argument in the middle of the paper.


Surely, rejection of the
privileged status of first-person access admits of degrees: we can merely claim that avowals are sometimes insufficient to establish everything one believes (or desires, or fears); in circumstances where first-
person access is silent, one could appeal to third-person procedures.
This weaker version of the claim that the first-person access falls short
as the ultimate authority fails to have all the authority over my beliefs
(and other mental states) can be correlated with the following versions
of Moore paradox (which I shall call MI for Moore Ignorance):
(MI) I don’t believe p, but p.
In this version, I claim ignorance about p and, at the same time, I take
p to be the case—presumably because I have some reason (for example,
another source of information) to take p to be the case. A stronger ver-
sion of Moore paradox (Moore Error) could be correlated with a stron-
ger version of the claim that first-person authority is fallible:
(ME) I believe not-p, but p.
The two versions of fallibility of the first-person authority could then be
expressed through these two paraphrases of MI and ME:3
(PMI) I fail to avow the belief that p while attributing to myself the
belief that p.
(PME) I avow the belief that not-p while attributing to myself the
belief that p.
Both paraphrases express failures in the authority of a first-person ac-
cess to my beliefs and the second indeed holds that it could be corrected
by the third-person approach. In both cases, what I take through avow-
als to be my belief about something is not the last word about what I
actually believe. In both cases my reflective investigation of the world
is not enough to establish what I believe (or fail to believe); there may
be beliefs of mine that could only be accessible by means of a third-
person procedure. In the case of PME, we have diverging accounts of
what I believe: only further (perhaps specific) considerations could help
to establish (and help me to make up my own mind) whether I in fact
believe p or not-p.


Failures of transparency could follow from failures in first-person
authority concerning beliefs expressed by situations such as PME (and
to some extent such as PMI). If PME is possible, I have transparent
access to what I avow as my beliefs but transparency can be overruled
by other considerations related to other beliefs (or other mental states)
that I hold and that have to be taken into account in order to attribute
beliefs to me (in a third-personal manner). Transparency is a feature
of beliefs that are acquired through a process of examining how things
are that should last until something else manages to persuade me. If
(some of) my beliefs are not accessible by me in a first person manner,
they could fail to display this transparency. Such a failure could be ex-
pressed by saying that for some of my beliefs I cannot infer what I (fully
and on reflection) take to be the case and there could be cases where
what I take to be the case is not enough to determine what I believe (or,
perhaps, not even what I ought to believe). Transparency is important
here because it enables me to complete an inference from what I take
to be the case to what I take to be what ought to be believed and then
to what I believe. PMI-like cases also entails some failures of transpar-
ency: some of my beliefs can fail to display transparency because I can-
not acquire them (nor can I access them) by a first-person investigation
of the world—when this investigation proves to be insufficient, I can
appeal to other strategies to determine what I believe. With PMI, I can-
not infer from what I believe to what I take to be the case—by reflection
and by sufficient (by my own standards) investigation.
Some of the consequences of these failures could be quickly appreci-
ated if we consider our efforts to persuade. I can fail to persuade you of
something solely because there is something in your mental set-up that
is incompatible with the beliefs I’m trying to make you acquire (by giv-
ing as many good reasons as possible). In a scenario where PME is pos-
sible, the most I can hope for is to persuade you to avow some beliefs;
that is, you will have first-personal access to whatever I am giving you
reasons to believe while still stopping short of believing it. If we use an
example reminiscent of Evans (1982: 225), I can be persuaded to avow
that there will be a war but for some reason—to do with my further
beliefs, fears or desires—I cannot believe anything but that there won’t
be a war. In the PMI scenario, I can still say that I cannot establish
that there won’t be a war by investigating the world but I that I still
cannot help believing that there won’t be a war. I establish my belief by
looking at, say, my behavior or my other expressed beliefs, desires etc.,
and this evidence eventually completes what is accessible of my beliefs
through avowals.
If first-person authority has no ultimate authority over my beliefs,
do we have to give up the idea that beliefs are transparent? Moran
(2007 and personal communication) argues that transparency is rather
a norm than a description of our common practice concerning our be-
liefs. Our beliefs, he insists, ought to be transparent so that we can take
them to be responding to the world (to the best of our knowledge and
relevant abilities). Further, we can argue that not all my beliefs can fail
to be transparently accessible—at least some of my beliefs have to be
avowed, or there will be little sense in saying that those beliefs are in
any sense mine. Suppose all my beliefs—including beliefs like ‘I believe
this belief can explain my behavior’—are accessible solely in a third-
person manner; there will be no way for me to say anything further
than that those beliefs are somehow connected to my body (assuming
I have further reasons to assert that this body is mine). It seems that
the rules of ascribing beliefs to people presuppose that at least some of
my beliefs are, in principle, accessible in a first-person manner. Then,
in a PME scenario, we can take some of my beliefs not to be avowed (or
even not capable of being accessible through avowals) but only against
the background of avowed beliefs that determine the belief-set I hap-
pen to own. Beliefs are understood as items that are owned by people;
owning involves accessibility in a way that is somehow privileged. That
at least some of my beliefs are to be accessible in a first-person man-
ner (and transparently) seems to be a limit on how thoroughly we can
deflate the notion of first-person authority.4 That our beliefs ought to
be transparently accessible is a prerequisite of (rational) persuasion
and, at the same time, transparency is what makes our more accepted
ideas seem more worthy: they have been scrutinized through different
perspectives. In section 3 we will examine more carefully this last point
and submit that both avowed beliefs and beliefs self-attributed in a
third-personal manner play a role in our doxastic economy.

Our perspective on how things are seems to be inevitably shaped by
values and practices that we have acquired in our communities. An is-
sue akin to that of the requisite transparency of my beliefs is that we
can end up being excessively sensitive to the way others describe our
actions and thoughts. Bernard Williams has aired the suspicion that
“thinking about your possible states in terms of the virtues” is “to think
about the way in which others might describe or comment on the way
in which you think about your actions, and if that represents the es-
sential content of your deliberation, it really does seem a misdirection
of the ethical attention”. In other words, paying too much attention
to how one’s action would be perceived from the outside—from a third-
personal perspective—somewhat taints a moral act. One could, for ex-
ample, be motivated by an attempt to look modest or courageous in the
eyes of somebody else (or in one’s own eyes while observing one’s own
action) instead of looking at the world with courageous or modest eyes
and acting accordingly. It is one thing to act moved by a virtuous way
of seeing the world and another to act guided by models of virtuous ac-
tion. Of course one has to learn, at some point, how to see the world in
a virtuous way and that would most likely involve looking at virtuous
actions in order to acquire the capacity to spot in the world the salient
features that would guide a virtuous judgement. Still, according to Williams’ thesis, an action guided by an attempt to be virtuous—and that,
therefore, sees virtue from an external point of view instead of seeing
the world through virtuous eyes—is not fully morally adequate because
it is not first-personal enough. This can be understood as a consequence
of the asymmetries between how I evaluate my own action and how
I evaluate somebody else’s: it is morally fine to judge other people in
terms of how virtuous their actions are but it may not be so to judge
one’s own action likewise because evaluation can taint one’s motivation
and therefore action itself. Williams—in a vein that often resembles
Sartre’s notion of bad faith—believes that we can consider an action
from the moral point of view in terms of how much first-personal it is.
If Williams is right about moral judgements, we can apply the idea
that mental content is sometimes not as first-personal as it should be
to beliefs in general. We can start by considering epistemic (or doxas-
tic) virtues, instead of moral virtues. Consider a case of someone that,
in the process of acquiring and managing her beliefs, pays excessive
attention at how reliable (empirically adequate, coherent, or widely
accepted) her beliefs are when considered from a third-personal point
of view. The suspicion is that she can be misdirecting her capacity to
have a third-personal access to her beliefs. We judge anyone’s beliefs
in terms of their virtues—in terms of features that we use to evaluate
their justification or their likelihood to be true—and this is at least one
way to assess someone else’s beliefs. Since other people’s beliefs are
not transparent to me—I cannot establish their beliefs by just looking
at the world—it is open to me to consider the epistemic quality of their
beliefs from an external point of view. In our case, however, the person
who pays excessive attention to the epistemic qualities of her beliefs
is neglecting the transparency that is open to her as a resource for es-
tablishing her own beliefs. She will then be paying too much attention
to the standards of evaluation for beliefs (that maybe she is ready to
recommend and further to maintain) to an extent that would neglect
her capacity to examine the world from the perspective of her beliefs;
in this sense she can end up holding beliefs that are not first-personal
enough—she could be guilty of what we can describe as an epistemic
bad faith.
[...]