I. INTRODUCTION
The analysis presented in this essay is primarily concerned with two doctrines
in their most general form: realism and skepticism. More specifically, it is
the language that is utilized in the theses of these two preceding doctrines
that is the major focus of the essay.
Modern skeptics have rigorously promoted the philosophical attitude that we
cannot acquire knowledge of the way the world really is. Such knowledge may
possibly include scientific, religious, metaphysical, ethical, and, depending
on how extreme the skepticism being advocated is, commonsensical knowledge.
In simple terminology, part of which is formal, the skeptic's attitude may
solidify into a thesis having the appearance of the following statement:
Given any x, if x is a person, and given any y, if y is a modicum of knowledge
of the way the world is, then x cannot have knowledge of y.
Realism, on the other hand, can be presented in two forms that are germane
to this investigation. The first, which is known as scientific realism, is the
metaphysical doctrine that maintains that there is a mind-independent world,
independent of perception and cognition, which is comprised of physical objects
having mathematical coordinates in objective space and time. The second is
naive realism, according to which we can be directly aware of the external
world and physical objects through sense-perception. As with scientific realism, naive realism
also maintains that the external world exists independent of perception.
What is pertinent here is that both scientific realism and naive realism
involve the notion of an external world. With the naive realist, this world
is readily available by means of sensory perception. For the scientific
realist, it transcends sense-perception.
It is not always the case that the realist simply
contends that there is a external world. Occasionally the realist
will construct, or at least propose, a theory of knowledge by which
one can acquire knowledge of the external world. For example,
in addition to the scientific realist positing the existence
of a mind-independent world of physical objects, he will also make
knowledge claims such as, 'The mass of an electron is 9.10939 X 10-31 kg'
or, 'The sensation of color is caused by light particles impinging upon
our retinae". Similarly the naive realist
may make knowledge
claims such as, 'Seattle is south of Vancouver' or "There are green
vines in my back yard even though I can't see them". The foregoing
knowledge claims would seem to be rather trivial and indisputable
to most people. In fact, questioning the truth of these claims would be
considered futile.
However, is it possible that the modern skeptic's central thesis
can be challenged and an argument constructed, according to
which skepticism is rendered unintelligible and thus cognitively
meaningless? Is it also possible that the realist's central thesis
is overstepping its own limitations and in actuality all knowledge
is restricted to the data given in direct awareness? This essay provides a
definite "yes" to the preceding questions. It will
be argued that in light of the unintelligibility of realism and
skepticism, a more restricted form of epistemological theory should be
considered, a theory that could be regarded a unique form of skepticism,
not because it carries with it implications concerning what we do or
cannot know (for this aspect of skepticism will be shown to be
inadequate), but because of its strong emphasis on the paucity of
what we do know.
II. THE PRINCIPLE OF DEDUCTIVE CLOSURE
Central to many arguments for modern skepticism is the employment
of the Principle of Deductive Closure (PDC). The principle
can be expressed symbolically as follows:
(1) PDC: [Ksp & Ks (p implies q)] implies Ksq
In ordinary language (1) can be translated as,
(2) If one knows that p, and one knows that p implies q, then one knows that q.
In recent arguments for skepticism, 'p' in the PDC stands for any proposition
expressing a modicum of empirical knowledge, whereas 'q' stands for any
proposition expressing a logical possibility, the denial of which will
undermine one's knowledge of p. For example, 'q' may express the possibility
that one is not hallucinating or dreaming, that one is not being deceived by
an evil genius, or, what some might take more seriously, that one is not
imperfectly constituted in such a way that deception and error in perception
and cognition are inevitable.
Thus, with reference to the PDC, we might say that if one knows that he has
five fingers on his hand, and that this knowledge implies that he is not
hallucinating, then he also knows that he is not hallucinating.
The strategy of the skeptic is directed toward denying the first conjunct
of the bracketed antecedent in the PDC. This is accomplished, for the
most part, by demonstrating that one cannot know that q, followed by
the inference that one cannot know that p by means of the modus tollens principle.
P.A. Schreck has constructed the following argument in which the structure of
the skeptic's strategy in connection with the PDC is employed: [1]
[2] 1. [Ksp & Ks (p implies q)] implies Ksq
2. ~Ksq
3. ~[Ksp & Ks (p implies q)] 1,2 MT
4. Ks (p implies q) assertion
5. ~Ksp 3,4 DS
Schreck points out that there are many ways one could go about
attempting to refute the skeptic's central thesis (i.e., premise 5).
One way of doing this is by denying the validity of the PDC itself. In his
paper, "Epistemic Operators", Dretske attempts this sort of refutation
by introducing the notion of penetrating, non-penetrating, and semi-penetrating
operators. An epistemic operator, such as 'it is true that', 'it is a
fact that', or 'one knows that', is penetrating when one statement can
be deduced from another by means of an intermediate statement, all of
which contain that same epistemic operator. The deduction is made
possible when the intermediate statement says something about the
original statement. For example, if the epistemic operator, 'knows that',
in the statement 'S knows that p' is penetrating, then from that
statement, conjoined with the statement 'p is r', one should be
able to deduce the statement 'S knows that r'. Similarly, if the
operator used in the PDC is penetrating, then one must be able
to deduce 'Ksq' from 'Ksp' by means of Ks (p implies q)'. According to
Dretske, the epistemic operator 'Ks' (i.e., 'one knows that') is
only semi-penetrating, with the result that one's knowledge of p,
and one's knowledge that p implies q, does not imply that one
knows that q [3].
The PDC can be constructed as a truncated version of the following argument:
1. Ksp
2. Ks(p implies q)
3. Ksq 1,2 ?
Dretske's argument appears to demonstrate that there is not
legitimate principle that one can put in place of '?' in
the argument above in order to establish the truth of its
conclusion and validate the PDC. Such a principle would allow
the transference of the epistemic operator 'Ks' from premise 1
to the conclusion by virtue of premise 2, and would entail the
penetrability of 'Ks'.
Two other strategies may be employed in order to refute the skeptic's
central thesis. First, an argument could be offered against
premise 2 in argument #3 above, where it is demonstrated that
one can know, for instance, whether he is hallucinating or dreaming.
Consequently, 'Ksp' could not be denied by steps 3 and 4 of
Screck's formal argument.
Second, one could argue that the truth of any statement in place
of 'Ksq' in the PDC is not a necessary condition for the truth
of any statement in place of 'Ksp', and that there is an independent
criterion for one to know that p, which does not require one to know
that q. In other words, he may be able to know that he has five fingers
on his hand without having to know that he is not hallucinating.
Possibly, a criterion for knowledge may be constructed according
to which hallucinations and possibilities with similar ramifications
are precluded, or where knowledge is based upon something other
than empirical data.
But there is still another strategy, one that is detrimental not
only to the skeptic, but to the realist as well. The strategy
requires a simple, straight-forward analysis of the language
that the skeptic uses when employing the PDC, language such as,
'one knows that' and 'one does not know that'. The strategy also
involves an inquiry into what the variables 'p' and 'q' in the PDC
could possibly express. The result will be an invalidation of the
skeptic's Principle of Deductive Closure, not because of the logic
of the principle, but because of the meaning (or lack of meaning)
of its terms once they are replaced with ordinary knowledge claims.
The strictures of the strategy will also extend to the thesis of
realism, since the realist uses language such as 'one knows that' and
'one does not know that' as well. Ultimately, the skeptic will
be barred from utilizing the PDC once the central thesis of
skepticism is shown to be cognitively meaningless. The realist
will be barred from attacking skepticism by invalidating the PDC
with knowledge claims about the external world, once the central
thesis of realism is shown to be cognitively meaningless.
III. THE ANALYSIS
Both the argument of the skeptic and that of one who contends
that we can have knowledge of an external world can be construed
in terms of some sort of formal argument that utilizes the
epistemic operator, 'Ks'. In other words, both arguments will
involve the claims, 'one does not know that p' and 'one can know
that p' respectively.
Before any argument concerning the possibility of obtaining knowledge
can be constructed, a complete analysis of the phrases, 'one knows
that p' and 'one does not know that p' is required. It is reasonable
to assume that the preliminary task, before one sets out to ascertain
whether the premises of an argument such as that of skepticism or
realism are true, or whether its conclusion follows from its premises,
is to determine the exact meaning of its premises and conclusion. This
includes an analysis of the terms and phrases by which the premises
are constituted. Hence, ascertaining meaning is methodologically prior
to determining truth and validity. To do otherwise
would simply be putting the horse before the cart.
First consider the statement 'S knows that p', where S is a person
and p is an empirical state of affairs. Two questions are relevant
to ascertaining what the preceding statement means: What does
the phrase 'know that' mean? And what is the nature of p for S,
at the time S knows that p?
Certain components of Locke's theory of knowledge may be helpful
in providing adequate answers the above questions. However,
for reasons explained below, these components will be
translated into slightly different terminology. In his Essay
Concerning Human Understanding, he contends that, "since the mind,
in all its thoughts and reasoning, hath no other immediate
object but its own ideas. . .it is evident that our knowledge
is only conversant about them" (book IV). This contention of
Locke's is quite attractive. An idea for Locke is basically an
object of the understanding or something with which one can be
acquainted when thinking. His terminology carries with it the
assumption that ideas are the limits of our understanding and thus
the limits of our knowledge. In other words, knowing is a mode
of the understanding, and, for that reason, it is not surprising
that knowledge is directly relevant to the contents of the mind
(i.e., ideas). However, instead of speaking of
ideas it may prove more beneficial to speak of items of direct
awareness. Speaking only of ideas seems to restrict knowledge
to the traces left by direct awareness (whether perceptual or
introspective), whereas speaking in terms of items of direct
awareness would not only encompass ideas and the objects from
which they originate, but give these two things equal precedence
in respect to them being directly relevant to
the knowledge. When the phrase 'items of awareness' is used,
it will serve to denote either sensory or introspective data.
In either case an item of awareness may be either simple or
complex. A patch of color, for instance is a simple sensory
datum. The mental image of that patch of color is an
introspective datum. Two patches of color in juxtaposition
may be a complex sensory datum. One's entire field of
vision at any given moment may a complex sensory datum,
analyzable into simple sensory data. There are no rigid
rules dictating how to identify a complex sensory or
entrospective datum. A simple datum cannot be desribed in
any other way, except by ostensive definition. In the
case of a patch of color, one points to the color. You
cannot point to a complex datum and expect someone to
immediatly understand that it is the complex datum
that you are pointing to, versus one of the many sensory data.
That requires the introduction of something more than
the ostensive definition.
One might object to the use of the
notion of items of direct awareness in place of Locke's
ideas by suggesting that, since knowledge involves
the process of knowing, which is essentially mental, it
is only concerned with ideas. And since the concept of
items of direct awareness encompasses items of sense
perception, which are supposedly ontologically distinct from ideas,
it is too general and thus partially incongruent with the
actual process of knowing. This objection is, of course,
grounded in the assumption that perception is not mental,
which renders its occurrence unnecessary during the actual
act of knowing as a mental process.
It is likely true that the process of knowing concerns
ideas (or what may be called items of awareness given in
introspection). But it is not entirely clear that the process
of knowing does not concern items of sense perception in the
same sense that it concerns ideas. In other words, the
restriction of knowledge to the domain of ideas may be unwarranted.
If one were to consider Hume's distinction between ideas and items
of sense perception (i.e., impressions) or even carefully examine
these two things, it will be discovered that ideas and impressions
differ only in respect to the latter being more lively and vivid
than the former. Hume does not make any rash ontological distinctions
between ideas and impressions,
which renders it feasible to regard
items of sense perception equally as pertinent to knowledge as ideas.
Hume's overall program seems to imply that there is a spectrum of awareness in
which we can acquire knowledge, but is restricted to impressions
and ideas, as opposed to things which are mind-independent.
With impressions and ideas being the only constituents of this
spectrum, they themselves cannot be used to demonstrate that impressions
and ideas are ontologically distinct without generating a circular
argument. All that is warranted by the spectrum in which knowledge
may be acquired is a psychological description of impressions and
ideas, which yields nothing more than the bare assertion that the
former are more vivid and lively than the latter. Thus both ideas
and items of sense perception are elements of the process of
knowing. The fact that knowing predominantly involves ideas as
objects of knowledge, and other operations that are mind-restricted,
does not preclude the possibility of perception being an element
of the process of knowing in the same way that ideas are elements
of the process of knowing. With regard to knowledge, and perhaps
the complete conception of reality, the distinction between the
mental and the perceptual can be placed on the same spectrum of
awareness--both are merely modes of consciousness.
Thus knowledge concerns items of direct awareness, and if we are
to follow through with Locke's contention, then knowledge
concerns only items of direct awareness. There are a number
of ways to establish the foregoing claim [it does not, of
course, follow from an account of the work of Locke and Hume],
one of which will be clarified later, and the other as follows:
The act of knowing is essentially cognitive in nature.
In other words, certain operations or processes that are
inextricably tied to mental phenomena are involved in the
act of knowing (here, the terms 'cognitive' and 'mental'
are taken to include sensory perception or anything that
occurs within one's sphere of awareness, which obviously
renders all forms of awareness as cognitive in nature).
Furthermore, that which is known will be an element of
the cognitive process of knowing. Anything that cannot
be or is not an item or complex of items of direct awareness will not be an
element of the process of knowing. Consequently, it
will be impossible for knowledge to pertain to things
that are not items of direct awareness--for the act of
knowing cannot extend beyond its own limitations to
things that are not a function of its own operation.
Locke goes on to suggest that "knowledge is the perception of
the agreement or disagreement of two ideas", which, with
regard to what has been said thus far, can be translated as
"knowledge is the direct awareness of the agreement or
disagreement of two or more items of awareness." For the sake of
simplicity, an agreement or disagreement may be regarded as a
relation. Accordingly, knowledge is the direct awareness of
a relation between two or more items, where these items are given in
the act of direct awareness as well. The phrase 'know that'
can now be translated as 'directly aware of', which merely connotes
a cognitive process involving either perception or introspection.
Furthermore, p will be a relation between two items of direct
awareness. Thus knowledge essentially involves two things:
1) two or more items of direct awareness, and 2) the relation
in which these things partake. For example, I can be aware
of two billiard balls and the relation of these items being
juxtaposed. I can also be aware of my awareness of each of
these balls individually, in which case the two items of
awareness are one of the billiard balls and myself. At any rate,
items of direct awareness, within the above context, need not
be restricted to relations alone, but should be construed as
encompassing the constituents of relations as well. [4]
Unfortunately, the traditional understanding of what the
variable 'p' may stand for generates a rather awkward situation
when the phrase 'knows that' is translated as 'directly aware of'.
The variable 'p' is most often used with the intention of it
standing for a proposition which expresses a certain state of
affairs (or fact). So the statement 'S knows that p' can
stand for a statement such as 'Hans knows that the eight ball
is next to the side pocket'. But it would be ridiculous to
translate the foregoing statement as 'Hans is directly aware
of the eight ball is next to the side pocket'.
It is clear that the awkwardness generated by the above
translation is merely grammatical and can be eliminated
by construing p in some other manner. The statement
'Hans knows that the eight ball is next to the side pocket'
can be translated as "Hans is directly aware of the eight
ball next to the side pocket' or 'Hans is directly aware of
the eight ball being next to the side pocket'. It is,
accordingly, tacitly assumed in the statement 'S is directly
aware of p' that if p is a relation of some sort, then S
is also directly aware of he constituents of p.
What, then, are the implications of the statement
'S knows that p' being translated as 'S is directly aware of p',
where p is a relation between two or more items of direct
awareness? The thesis of realism becomes problematic, unless
the external world were to be regarded as that of
which one is directly aware through sense perception. Even
if this were the case, knowledge of the external world would
be limited to that of which we are directly aware through sense
perception, not that of which we were or could be aware.
Accordingly, historical knowledge, the predictive knowledge
sought by science, and even what would be considered
commonsensical knowledge that there is a world independent
of the realm of direct awareness, are all impossible and
the notion that such knowledge could possibly be
acquired is rendered unintelligible.
The claim that the idea of commonsensical knowledge of the
external world is unintelligible may seem blatantly false;
however, it is not lacking a justification. To know what
it means (and thus find it intelligible) to know that p, where p
is something outside the realm of direct awareness, is impossible
because the act of knowing requires what is known to be a
relation between two or more items of direct awareness.
Suppose that S knows that q, and q is a relation between the
proposition 'S knows that p' and its meaning. So if S knows
that q, then S knows what it means for him to know that p. Part
of the meaning of the statement 'S knows that p' will be p
(p may be an idea or an object of sense perception). If p
is something outside the realm of direct awareness, then S
cannot know that p. And, since one linguistic component of
the relation q (viz. the statement 'S knows that p') has a
constituent that does not connote anything within the realm of
direct awareness, S cannot know what it means for him to know
that p. Hence, S can only know that q if and only if every
component of the relation denoted by 'q' contains only items
of direct awareness; and q in the above case is certainly
not this sort of relation.
As for the skeptic, it is indicative of him to
express the claim that one does not know a certain
modicum of knowledge, whatever that modicum may be.
In other words, the skeptic obviously feels that the
statement 'S does not know that p' is perfectly
intelligible, not to mention supported by his skeptical
arguments. However, such may not be the case.
Consider the statement 'S does not know that p'. Since the
phrase 'know that' can be translated as 'directly aware of',
the foregoing statement can be translated as 'S is not directly
aware of p'. So what exactly does it mean to say that S is
not directly aware of p? Suppose that S's not being directly
aware of p is a cognitive act C, in which case C involves p.
If C involves p and p is directly related to C for S, then it
is reasonable to assume that p is an element of C. Thus S's
not being directly aware of p is a cognitive act in which p is
an element. However, does it actually make sense to suggest
that p is an element of the cognitive act of S not being directly
aware of p? This appears to be contradictory in that if p is
an element of a cognitive act of S, then no matter what sort of
cognitive act it is, S must be directly aware of p. For a
cognitive act is basically a thought process, in which case
if p is an element of a cognitive act, then it must also be
an object of thought and thus something of which one is
directly aware. But is S's not knowing that p implies that
S is not directly aware of p, and S's not being directly
aware of p (if 'not being aware' expresses a cognitive act)
implies that S is directly aware of p, then S both knows that
p and does not know that p, which is contradictory.
Hence, S's not knowing that p cannot be a cognitive act
containing p as an element. But if any cognitive act
involving p, including the act of knowing, must contain p
as an element, and if S's not knowing that p does not contain
p as an element, then S does not, not know that p. Nor does
S know that p: p is simply not an object of thought that
can be known or not known [so much for the Exclude the
Middle principle!].
Following from the above demonstration, the only alternative
is to deny that S's not knowing that p is a cognitive act,
the implications of which are that S cannot comprehend not
knowing that p, and the assertion that S does not know that p
in unintelligible for S. For if S's not knowing that p is
not a cognitive act, then S's not knowing that p is not a
process of thought, in which case p is not an object of thought.
If p is not an object of thought for S, then S cannot comprehend
not knowing that p, and his assertion that he does not know that p
will, to reiterate, be unintelligible for him.
In summary, it is apparent that there are good reasons to
reject the realist's thesis. For it is impossible to have
knowledge of things outside the realm of direct awareness.
To say that S knows that p is to say nothing more than S
is directly aware of p, where p is a relation between two
or more items of direct awareness. The foregoing is
grounded in the assumption that the act of knowing is a
cognitive act in which the object of knowledge is, and
must always be, an element of that act.
One might object to the conclusion drawn above and suggest
that we can have knowledge of things outside the realm of
direct awareness in so far as we can have an idea of a
certain state of affairs and at the same time know that
there is an actual state of affairs, independent of the
act of direct awareness, that corresponds with that idea.
In other words, one might argue from the idea of something
to the reality of that something that is supposedly
represented by the idea.
The foregoing objection is inadequate for the following
reasons: First, to say that one has an idea of a certain
state of affair is to presuppose the realist's central
thesis and thus tacitly beg the question of whether one can
have knowledge of an external world. One does not have an
idea of a certain state of affair (unless, of course, there
is some actual state of affair that corresponds with that
idea and one has knowledge of this correspondence), but
instead one has direct awareness of a relation given in
introspection, which may resemble a relation of which one
has had direct awareness through perception. Thus merely
having an idea does not provide knowledge of things outside
the realm of direct awareness.
Second, there is nothing in an idea itself by which one
can deduce or infer the existence of anything beyond that idea,
namely something that has produced or corresponds with that
idea. To suggest that one has knowledge of the external world
through a mere idea is to naively confound a relation given
in direct awareness with something outside the realm of
direct awareness or, for that matter, something totally
outside the realm of thought. There is absolutely no
cognitive act whatsoever that corresponds or accompanies a
putative knowledge claim about the external world, except
one that is constituted by direct awareness and thus
items of direct awareness.
In addition to the problematic nature of the realist's
central thesis, there are also good reasons to suggest
that the modern skeptic's central thesis is also unintelligible
and cannot be comprehended as either true or false.
For to say that one does or cannot know that p is to say either,
1) that there is a cognitive act (viz. of not knowing) in which
p is an element, or 2) that there is no cognitive act of not
knowing (in which p is an element). The former alternative
is inadequate because an element of a cognitive act is an
element of an act of thought, in which case p is an
object of thought and thus an item of direct awareness.
Hence, to not know that p would be to know that p, which
is contradictory. Moreover, the latter alternative leaves
the skeptic's central thesis unintelligible; for if p is
not an object of thought, and if there is no thought process
whatsoever in which p is an element (that is, if p is supposed
to be something that is not known), then the claim that S
does not know that p will be a mere flatus vocis and thus
cognitively meaningless for S.
IV. CONCLUSION
The end result of what has been demonstrated in this paper is
the groundwork of epistemology that is incongruent with both the
modern skeptic's thesis and the realist's thesis,
according to which knowledge claims are limited to those
making reference to items of direct awareness. With
regard to Schreck's formal argument, premise 2 can be
rejected on account of its unintelligibility; and as far
as the PDC is concerned, the scope of p in the epistemic
operator 'Ksp' ('one knows that p') is restricted to relations
and their constituents of which one is directly aware.
Furthermore, because knowledge is limited to the direct
awareness of relations, the constituents of these relations
and the relations themselves should be regarded as
metaphysically neutral. In other words, coupled with the
direct awareness of a relation and its constituents there
can be no presuppositions or knowledge claims that cannot
themselves be grounded in the act of direct awareness.
Consequently, such claims regarding whether certain items of
direct awareness are physical or non-physical, real or ideal,
detached from or essentially connected to the perceiver,
and so forth, must all be disconnected
from the act of direct awareness
and all the knowledge claims it may ground.
__________________________________________________________________________________
NOTES
(1) Schreck, P.A. "Cartesian Scepticism and Relevant Alternatives". Eidos 8, (2), p. 127. (1989).
(2) Schreck's formal argument is actually incomplete and requires two more steps to infer his conclusion. A complete version can be constructed in the following manner:
(1) [Ksp & Ks(p implies q)] implies Ksq
(2) ~Ksq
(3) ~[Ksp & Ks(p implies q)] 1,2 MT
(4) Ks(p implies q) assertion
(5) ~Ksp v ~[Ks(p implies q)] 3 DeM
(6) ~~[Ks(p implies q)] 4 DN
(7) ~Ksp 5,6 DS
(3) Dretske, Fred I. "Epistemic Operators". The Journal of Philosophy 67, (24), p. 1007-1023. (1970).
(4) One could object here by suggesting that relations are metaphysical entities and are not accessible through direct awareness, and that only their constituents are legitimate items of direct awareness. I would circumvent this problem by defining a relation as the entire perceptual field or context within which the constituents of the relation are placed. For example, one can focus on one billiard ball, individually, as a separate item of direct awareness; or one can focus on both billiard balls, including the extension between them. The two billiard balls and the extension beween them could thus be defined as the relation.