A Broad Perceptual Model of
Privileged Introspective Judgments
William
S. Larkin
__________________________________
On a broad perceptual model of introspection, there
is a merely contingent causal connection between our introspective judgments
and the occurrent thoughts that are their targets. Such a model is metaphysically unassuming and naturalistically
appealing, but it would seem to have difficulty accounting for the special
epistemic status of introspective judgments.
It is intuitively undeniable that the introspective judgments we make
about the contents of our own occurrent thoughts occupy a position of privilege
in our cognitive economy: Privileged
introspective judgments about the contents of our own thoughts seem to be on a
qualitatively better epistemic footing than our ordinary perceptual judgments
about the external world.
But now it is difficult to
see how a model of introspection that explicitly likens that faculty to
ordinary perception can possibly account for a qualitative difference in
epistemic status between introspective and perceptual judgments. What is called for here is a philosophical
explanation, in Nozick’s sense, of how it is even so much as possible for a
perceptual model of introspection to account for the privileged epistemic
status of introspective judgments. My
aim here is to provide such an explanation.
1.
Privileged introspective judgments (PIJs) seem, to
both common sense and philosophically sophisticated intuition, to be
qualitatively better off epistemically speaking than ordinary perceptual
judgments (OPJs). Thus the irresistible
tendency to think of PIJs as better suited to a foundational role than OPJs;
not merely because PIJs are contingently more likely to be true or
epistemically secure than OPJs, but because they are intrinsically more
reliable. PIJs are privileged over OPJs
not because PIJs are produced by mechanisms that happen to be more reliable in
the sense of more frequently producing true or secure judgments, but because
PIJs by their very nature are immune to some epistemic defect(s) to which OPJs
are susceptible.
PIJs
are thought to be qualitatively more reliable than OPJs in the sense
that PIJs are utterly immune to some epistemic defect to which OPJs are
not. Thus the traditional attempts to
account for the special epistemic status of PIJs in terms of infallibility,
indubitability, or incorrigibility—immunity to error, doubt, and correction
respectively. Judgments of a certain
type are infallible when it is not even possible for such judgments to be
false. The difference between fallible
and infallibile judgments is, in the sense intended here, a qualitative
difference in their likelihood of being true.
Fallible judgments may be extremely reliable and may in fact
(contingently) always turn out to be true, but it is still possible for them to
be mistaken; whereas infallible judgments are not even possibly mistaken. With infallibility we get a qualitative
difference in reliability from a modal difference in likelihood of truth.
The situation is likewise
with respect to indubitability and incorrigibility, except that instead of a
qualitative modal difference in truth, there is a qualitative modal difference
in epistemic security. Indubitable and
incorrigible judgments are qualitatively more secure, in the sense that they
are immune to compelling counter-evidence.
S’s judgment J is indubitable when it is not possible for S to have
evidence sufficiently cogent to rationally warrant S giving up J. S’s judgment J is incorrigible when it is not
possible for someone else to have evidence sufficiently cogent to outweigh S’s
testimony. In both cases we have an
immunity to a certain epistemic defect that springs from the impossibility of
there being sufficiently rationally compelling counter-evidence. The difference between a judgment that is
extremely secure but still possibly subject to evidential defeat and a judgment
that is utterly immune to defeat is, in the sense intended here, a qualitative
one. With indubitability and
incorrigibility we get a qualitative difference in reliability from a modal
difference in epistemic security.
Traditional attempts to
account for the privileged status of PIJs claim that PIJs are immune to error,
doubt, or correction; thereby accounting for a qualitative difference between
PIJs and OPJs in terms of a modal difference in likelihood of truth or
epistemic security. OPJs might be
extremely reliable, secure, or likely to be true; but it is still possible for
them to be mistaken or defeated by cogent counter-evidence. So if PIJs were either infallible or
indubitable or incorrigible, then there would be a modal difference in
epistemic status between PIJs and OPJs; and we would have an account of the
qualitative difference in reliability between PIJs and OPJs.
The problem is that on a
naturalistically appealing broad perceptual model of introspection, PIJs are
neither infallible nor indubitable nor incorrigible. On such a model, PIJs would seem to be subject to the very same epistemic
defects as OPJs. Perhaps introspection
would turn out to be contingently more likely to produce true or secure beliefs
than ordinary perception; but still there is no modal gap. On a perceptual model of introspection, it
is still possible for PIJs to be mistaken or defeated. Thus, though there might be a quantitative
difference in reliability between PIJs and OPJs on a broad perceptual model of
introspection, there can seemingly be no qualitative difference in
reliability. I will argue however that
even on a broad perceptual model of introspection, there will remain an
epistemic defect to which PIJs but not OPJs are immune. This I will contend is a modal difference
sufficient for a qualitative difference in reliability between PIJs and OPJs.
2.
I am attempting here to provide a philosophical
explanation for how a perceptual model of introspection can nevertheless
account for a qualitative epistemic difference between introspective and
perceptual judgments. A philosophical
explanation of this kind works by putting forth an explanatory hypothesis,
which is to be judged not so much by its initial plausibility as by its ability
to effect some desirable outcome. In
this case the desirable outcome would be reconciling a naturalistically
appealing perceptual model of introspection with the intuitively undeniable
idea that PIJs enjoy a privileged status is desirable.
My explanatory hypothesis is
this: If judgments of type J are immune to subjective irrationality whereas
judgments of type G are not, then J-judgments are qualitatively more reliable
than G-judgments. I will argue that
even on a broad perceptual model of introspection PIJs are immune to subjective
irrationality whereas OPJs are not.
It will follow then from the explanatory hypothesis that even on a broad
perceptual model of introspection PIJs are qualitatively more reliable than
OPJs. Thus if we accept the explanatory
hypothesis, we can reconcile a broad perceptual model of introspection with the
special epistemic status of PIJs.
Two beliefs conflict when one
provides sufficiently compelling evidence against the other. It is subjectively irrational to knowingly
maintain conflicting beliefs. That is,
it is irrational for one to maintain a belief B even though one is aware that
some of one’s other beliefs conflict with B.
Two conditions, then, are necessary for subjective irrationality—the
cogency condition and the awareness condition.
The cogency condition insures that the counter-evidence is sufficiently
compelling; and the awareness condition insures that one is aware that she has
sufficiently compelling counter-evidence.
Thus:
It is subjectively irrational for S to maintain a
belief B on the basis of some faculty F at time t only if:
(Cogency) S has certain beliefs C at time t, and C
is sufficient to undermine S’s trust in F at t.
(Awareness) S is aware at t that the cogency condition
is met.
Ordinary
perceptual judgments are not immune to subjective irrationality. It is possible for S to have a belief on the
basis of some perceptual faculty and yet to be aware that she has compelling
evidence against that belief. Imagine
that Mary forms a visual belief that there is a table in front of her. Though her vision is highly reliable, in the
sense of producing a high ratio of true to false beliefs, it is conceivable
that Mary could be told by an eminent and implicitly trustworthy source that
her vision is faulty at the present time.
This auditory input could produce in Mary beliefs that she recognizes to
conflict with her visual belief, as they are sufficient to undermine her trust
in her vision at the present time. Mary
would be subjectively irrational if she were to continue to maintain her belief
that there is a table in front of her on the basis of her vision.
3.
Privileged introspective judgments are utterly
immune to subjective irrationality; and this is so even on a broad perceptual
model of introspection where PIJs are not immune to either error or compelling
counter-evidence. In the case of PIJs,
the awareness and cogency conditions for subjective irrationality cannot
possibly be jointly satisfied.
On a broad perceptual model,
there can be compelling counter-evidence to any PIJ. Imagine that Sam judges, on the basis of her introspective
faculties at time t, that she is now (at t) thinking that the Orioles last won
the World Series in 1970. Now imagine
that at t S is hooked up to a highly reliable and carefully calibrated brain
scanner that prints out that S is now thinking that the Orioles last won the
World Series in 1983. The evidence of
the brain scanner conflicts with the evidence of Sam’s introspection. If introspection is merely a highly reliable
contingent causal process, then there might well be a highly reliable brain
scanner device and the evidence from that contingent causal mechanism could
compete on a level par with introspective evidence. Thus, given a broad perceptual model of introspection, there can
be compelling counter-evidence to PIJs (and this is why PIJs are neither
indubitable nor incorrigible on a broad perceptual model).
However,
no subject S can be aware that there is sufficiently compelling counter
evidence to one of her introspective judgments at time t. Let S form a belief B on the basis of her
introspective faculty F at time t, and let there be some compelling counter-evidence
C. C is sufficiently compelling only if
there is a cogent argument from C to the conclusion that F is not trustworthy
at t. S is aware that she has
sufficiently compelling counter-evidence to B only if she is aware of C and
reasons from that to the conclusion that F is not trustworthy. Thus it is subjectively irrational for S to
maintain B on the basis of F at t only if there is a cogent argument from S’s
awareness of C at t to the conclusion that F is not trustworthy at t. But there can be no cogent argument from S’s
awareness of C to the conclusion that F is not trustworthy. For S will have to rely on F (i.e., her
introspective faculties at time t relevant to the content of B) in order to be
aware of C and reason to the conclusion that the source of B (i.e., F) is
untrustworthy. Thus, any argument from
S’s awareness of C to the conclusion that F is not trustworthy is
self-defeating.
To
put the argument in a slightly different way:
It is subjectively irrational for S to maintain B at t on the basis of F
given counter-evidence C only if it is subjectively rational for S to give up B
at t on the basis of C. It is
subjectively rational for S to give up B at t on the basis of C only if there
is a rationally compelling argument from S’s awareness of C to the conclusion
that F is untrustworthy. But there is a
rationally compelling argument from S’s awareness of C to the conclusion that F
is untrustworthy only if it is reasonable to think that F is both trustworthy
and untrustworthy at t. Since it is not
reasonable to think that F is both trustworthy and untrustworthy at a single
time, there can be no rationally compelling argument from S’s awareness of C,
and thus S’s relying on F at t, to the conclusion that F is unreliable at t. So it can never be rational for a subject to
give up an introspective judgment on the basis of her awareness of some
conflicting beliefs. Thus it can never
be irrational for S to maintain an introspective judgment despite her awareness
of some conflicting beliefs.
4.
Introspective judgments are immune to subjective
irrationality, even if they are not immune to error or compelling
counter-evidence. Thus even on a broad
perceptual model of introspection PIJs are immune to a certain epistemic defect
to which OPJs are not. My explanatory
hypothesis is that immunity to subjective irrationality is sufficient for
epistemic privilege—that is, since PIJs are immune to subjective irrationality
but OPJs are not, PIJs are qualitatively more reliable than OPJs. If we accept this hypothesis, then we can
reconcile a perceptual model of introspection with a qualitative difference in
epistemic status between PIJs and OPJs.
If immunity to subjective irrationality is sufficient for epistemic
privilege, then a broad perceptual model of privileged introspective judgments
is viable.
William S. Larkin
Department of
Philosophy
Southern Illinois
University Edwardsville