University of Southern California Dissertations and Theses

Justificationalism Versus Nonjustificationalism In Philosophy: A Critique of The Theory Of Rationality In Karl Popper And W.W. Bartley, Iii

(USC Thesis Other)

Justificationalism Versus Nonjustificationalism In Philosophy: A Critique of The Theory Of Rationality In Karl Popper And W.W. Bartley, Iii

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content This dissertation has been
microfilmed exactly as received
69-13,061
HANNA, Maurice, 1934-
JUSTIFICATIONALISM VERSUS NONJUSTIFICATIONAL-
ISM IN PHILOSOPHY: A CRITIQUE OF THE THEORY
OF RATIONALITY IN KARL POPPER AND W.W.
BARTLEY, m .
University of Southern California, Ph.D., 1969
Philosophy
University Microfilms, Inc., Ann Arbor, Michigan
^ Copyright by
MAURICE HANNA
1969
JUSTIFICATIONALISM VERSUS NONJUSTIFICATIONALISM
PHILOSOPHY: A CRITIQUE OF THE THEORY OF RATIONALITY
IN KARL POPPER AND W.W. BARTLEY, III
by
Maurice Hanna-
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Philosophy)
January 1969
UNIVERSITY O F SOUTHERN CALIFORNIA
THE GRADUATE SCHOOL
UNIVERSITY PARK
LOS ANGELES. CALIFORNIA 9 0 0 0 7
This dissertation, written by
MAURICE HANNA
under the direction of Ais... Dissertation Com
mittee, and approved by all its members, has
been presented to and accepted by The Gradu
ate School, in partial fulfillment of require
ments for the degree of
D O C T O R O F P H I L O S O P H Y
^ 'Tna J j . °
Dean
Date Janu arY , 1969.
DISSERTATION COM M ITTEE
_ m / Chairman
..
TABLE OF CONTENTS
INTRODUCTION .....................................
Chapter
I. IS A NONJUSTIFICATIONAL PHILOSOPHY POSSIBLE?
The Problems of Knowledge and Rationality
Rationality and Ultimate Relativism
Rationalism's Two Crises
Comprehensive Rationalism
Critical Rationalism
Rationality and Authoritarianism
Comprehensively Critical Rationalism
Criticism of Nonjustificationalism
The irrefutability of logic
Intelligibility as a goal
Conclusion
II. IS TRUTH DISCOVERABLE? ....................
A Basic Problem of Epistemology
Truth and Error
Knowledge and Conjecture
The Scope of Popper's Position
Three Views of Human Knowledge
Essences and Laws of Nature
Essences and Natural Necessity
Popper's Principal Assumptions
The Consequences of Counter "Assumptions"
An Alternative View of Rationality
Conclusion
III. WHY ARE LOGIC AND ARITHMETIC APPLICABLE TO
REALITY? ................................
Rules of Inference
Limitations of Popper's Exposition
Three Main Views of Logic
Criticism of Popper's Analysis
The normative view
The semantic-conventionalist view
The descriptive view
The Truths of Arithmetic
Conclusion
IV. HOW ARE THE TRUTHS OF LOGIC KNOWN? ........
Bartley's Skeptical View of Logic
Popper's Fideistic View of Logic
Justifying Rules of Logic
Intuitional versus Nonintuitional Claims
Logic, Language, and Reality
Conclusion
V. ARE PRESUPPOSITIONS UNAVOIDABLE? ..........
Presuppositions and Rationality
Presuppositions and Evidenz
Presuppositions and Starting Points
Presuppositions and Justification
Conclusion
CONCLUSION......................................
BIBLIOGRAPHY
INTRODUCTION
Some of the most urgent problems facing philoso
phers today are concerned with understanding the nature of
philosophy itself. Is there one fundamental feature that
characterizes all philosophical methods and perspectives?
If there is* what is it, and why does it enjoy such uni
versality?
It is my contention that every philosophy does pos
sess a universal characteristic, so that whatever differ
ences distinguish it from other philosophical positions, it
is one with them in its justificational procedure (i.e.,
the procedure of attempting to establish propositions as
true or credible by showing their basis in positive evi
dence). Perhaps in the past this assertion would have been
dismissed as a trivial truism. But it is not so obvious
today. For, besides the advocates of various "irrational
isms" of recent times, there are those who, in_ the name of
rationality, have challenged the justificational procedure
and have claimed that nothing could be more inimical to the
enterprise of philosophy than, justificationalism. Philoso
phers cannot afford to ignore this criticism, especially
when it comes from thinkers whose philosophical influence
is undeniable. For* if these critics are correct, then
philosophy--if it is to continue at all--must don a new
garment of identity.
In the vanguard of this opposition to justifica-
tionalism are Karl Popper and W.W. Bartley, III. Although
Popper's published writings stop short of an explicit non-
justificationalism in philosophy, some of his statements
evidence a tendency in that direction. Moreover, there are
two reasons why the views of Popper and Bartley can, with
appropriate qualifications, be treated together. First,
Bartley's position demonstrably rests on the foundation
laid by Popper. Secondly, since Bartley has been a student
of Popper's, it is unlikely that he is wide of the mark
when he says, "I believe Popper would largely agree with
the view I call 'comprehensively critical rationalism'
. . ., a view that I should never have been able to formu
late without having had his views to guide me" (11:132).
It is my purpose in this work to show that non-
justificationalism is untenable because of the absurdity
of its consequences. This is established in the first
chapter. Then, in the second, third, and fourth chapters,
I proceed to attack some of the main prejudices on which it
is based. In the last chapter I examine and dispute the
assumption that the only form of justificationalism is one
that asserts the necessity of "criteria" for all knowledge,
thereby resulting in the "ultimate arbitrariness" of all
our views. Failing to see an alternative to such a self-
defeating justificationalism, Bartley attempts to construct
a program that can dispense with justificationalism alto
gether. That sense perception and logical principles do
not require "criteria" is a possibility which Bartley
either refuses to take seriously, or of which he is simply
unaware. Against Bartley, I argue that there is a defensi
ble form of justificationalism that can dispense with the
comprehensive demand for "criteria," thereby precluding
"irrationalism," whether in the guise of fideism, skepti
cism, or nonjustificationalism.
Initially my aim appears to be negative, but it has
an intended positive result. For the refutation of nonjus-
tificationalism shows the necessity of justificationalism.
These are the only two alternatives for philosophy. Never
theless, justificationalism does not exclude the subsidiary
role of refutation ("nonjustification") within its frame
work, but nonjustificationalism, by definition, excludes
any valid role for justification. This will be made clear
in the course of my exposition and criticism. Furthermore,
the meaning and grounds of a viable justificationalism will
be indicated in the critical application of insights stem
ming from Edmund Husserl's phenomenology. While it is be
yond the scope of this work to attempt a comprehensive dis
cussion of phenomenology, some of its important findings
can be brought to bear on the crucial problems under con
sideration. Moreover, since nonjustificationalism claims
to be a presuppositionless philosophy, it is illuminating
to contrast it with a procedure (viz., phenomenological
description) which provides a way to realize this ideal.
Expressed in a positive way, then, my thesis is not
only that a justificational procedure is inescapable, but
also that it is possible both to criticize and at least
partially to justify philosophical positions. And while I
do not agree with Husserl in every particular, I am con
vinced that phenomenological description is a propaedeutic
for all other branches of philosophical inquiry. The rec
ognition of its methodological validity is the first step
toward overcoming the charge leveled at justificationalism,
namely, that it is. wedded to arbitrary presuppositions, and
that, as such, it is the denial of philosophy's ideal of
maximum rationality. Undoubtedly, there are numerous phi
losophical positions which are subject to this animadver
sion. But phenomenology has been forged on the anvil of
such problems, and it has proven its mettle by doing more
justice to the multifarious factors in human knowledge than
any alternative. Among these factors are "givenness,"
"self-evidence," "insight," "universals," and "intention-
ality."
I think it can be rightly said that in recent phi
losophy the two most radical attempts to avoid presupposi-
tionalism are to be found in the phenomenology of Husserl
and the nonjustificationalism of Popper and Bartley. If
the arguments I present are essentially correct, then the
reasons for the indispensability of "phenomenological jus
tificationalism" will have two important consequences.
First, they will contribute to an elucidation of the nature
of philosophy and of philosophizing. Secondly, they will
serve to raise important questions about the implications
of justificationalism for understanding the nature of man
and the world in which he lives.
CHAPTER I
IS A NONJUSTIFICATIONAL PHILOSOPHY POSSIBLE?
In his book. The Retreat to Commitment, W. W.
Bartley, III argues that every philosophy, theology, and
religion that involves unremitting commitment to some be
lief, value, idea, or way of life is a form of irrational
ism. "The fundamental problem of modern philosophy," he
says, "is the problem of showing that it is possible to
choose in a nonarbitrary way among competing, mutually ex
clusive theories, and--more broadly speaking--among compet
ing 'ways of life'" (11:105). He believes he has found a
philosophical position that resolves this problem and that
escapes the various forms of irrationalism which result
from "ultimate arbitrariness." I shall argue that his pro
posed solution is a failure and that there is another way
of resolving the problem.
How, according to Bartley, does irrationalism
arise? And what is the precise nature of the problem it
poses for philosophy?
The Problems of Knowledge and Rationality
Bartley contends that the three main problems of
6
philosophy are "the problem of knowledge, the problem of
rationality, and the problem of reconciling knowledge and
rationality" (10:3). The second of these three problems is
the problem of resolving the conflict that arises when
one's theory of rationality makes rationality impossible.
This problem has long been recognized in philosophy, and it
has usually been considered in connection with the question
of the limits of rationality. These limits have been vari
ously conceived by philosophers. For example, there are
the limits imposed on reason by (l) the psychological and
biological conditions of human thought, (2) the restric
tions of sense experience, (3) the circumscriptions of
science, (4) the subjectivities and uncertainties of his
torical existence, and (5) the necessarily selective char
acter of description.
None of these limitations is Bartley's immediate
concern, however. The conflict between rationality and the
theory of rationality that is perpetuated by the so-called
logical limitation of rationality is what he is concerned
to resolve. The logical problem of rationality must be
seen as the attempt to overcome skepticism and fideism,
two positions which have emerged from the claim that ra
tionality is logically limited. As examples of these two
positions, Pyrrhonism would qualify as the most arrant
skepticism, and Pascal's perspective would be a clear case
of "fideistic irrationalism."
The task before Bartley, then, is twofold: (l) to
examine the philosophical claim made by skepticism and
fideism, and (2) to refute the argument which is used to
support this claim. Both views claim that only an arbi
trary choice can be made between fundamental beliefs, posi
tions, and ways of life,when such a choice is viewed from
the standpoint of rationality. The skeptic and the fideist
insist that it is logically impossible to make such choices
and to act on rational grounds when it comes to adopting
such positions, whether those beliefs or positions are re
ligious, metaphysical, moral, mathematical, scientific,
political, or whatever. This is of great importance for
philosophy; for the main task of rationality can be con
ceived as the making of such choices in a nonarbitrary way.
Rationality and Ultimate Relativism
Advocates of skepticism and fideism claim to ra
tionally demonstrate that these choices are arbitrary.
Their argument is based on the dilemma of ultimate presup
positions, i.e., when one is questioned about the reasons
for any claim to knowledge, either there is no end to the
giving of justificatory reasons or a halt is called at some
point in order to avoid an infinite regress. The stopping
point is one or more presuppositions, i.e., criteria or
goals. This poses the problem. If defending a position
rationally requires the adducing of good reasons, then it
appears that rationality is relinquished at the point where,
presuppositions are adopted without any justifying reasons.
Since starting points are diverse, "ultimate relativism"
seems inevitable, "for some way of choosing rationally
among competing ultimate stopping points by appeal to a
common standard is excluded in principle by the way the
problem is set" (10:6). Ultimate relativism results be
cause there is no way of resolving the differences between
advocates of different ultimate presuppositions. Such
opponents can only reply to the charge of irrationality or
arbitrariness by saying "tu quoque." No one is exempt from
this presuppositional predicament. At least, this is what
is very widely supposed, even among philosophers. Bartley
believes, however, that he has the one viable way out of
the predicament.
If the skeptic and the fideist agree in making the
same claim and presenting the same argument to support it,
what then is the difference between them? Bartley contends
that it is their opposed practical attitudes to the claim
that distinguishes them. The skeptic claims that he sus
pends his judgment about competing positions, whereas "the
fideist makes an irrational commitment to one or another of
them, or to some authority or tradition claiming to possess
the competence or the right to make such decisions for him"
(10:6). Nevertheless, Bartley points out that their agree-
10
ment is philosophically more fundamental, for "a position
that cannot escape skepticism cannot refute fideism" (10:6).
If their claim is correct, then it is philosophi
cally futile for men to rationally debate divergent ulti
mate presuppositions. These presuppositions set the limits
of rational argument, and everyone, whether an "irrational-
ist" or a "rationalist," is logically bound to make an ir
rational (i.e., without justifying reasons) commitment.
"This means that any irrationalist has a rational excuse
for irrationalism and a secure refuge from any criticism of
this irrational commitment" (10:7). The irrationalist in
sists that whatever one's position, and that whether a
position asserts or denies the "ultimacy" of logical prin
ciples, it is based on an act of "unjustified" faith. The
irrationalist maintains that on this basis the "rational
ist" and "irrationalist" are subject to the same limita
tion. By resorting to a tu quoque argument, then, the ir
rationalist believes he can retain his intellectual integ
rity.
Nevertheless, a tu quoque argument exacts a great
price from its user. One's belief that rationality is lim
ited, and that "irrational" commitment is universally un
avoidable, means that he loses the right to criticize. If
all commitments to ultimate presuppositions are arbitrary
from a rational point of view, then they cannot be criti-
11
cized rationally. Yet it seems that rational criticism of
basic presuppositions has been effective in contributing to
the growth of knowledge and the elimination of error. Ad
vocacy of a tu quoque argument, however, "makes nonsense of
the idea of the historical development and change of ideas
in the face of criticism" (11:97).
Rationalism's Two Crises
In order to bring the skeptico-fideistic argument
into sharp focus and to make his task of refuting the re
treat to commitment manageable, Bartley focuses his dis
cussion on a single example of a philosophical viewpoint
and way of life, namely, the rationalist tradition itself.
Does this position really require an irrational commitment
to reason? Or, in Bartley's own words, "what are the lim
its of rationality when it comes to choosing between the
rationalist and some other way of life?" (10:8) Is genuine
rationalism even theoretically possible if everything one
accepts must be rationally defensible? Can any "rational
ist" position be structured in such a way that its identity
does not lay the basis for its loss of integrity?
In view of these questions about the rationalist
tradition, it is important to consider the relevance of the
two concepts of "crisis of identity" and "crisis of integ
rity." Concerning their meaning, Bartley says:
In their technical psychological sense, these terms
are meant to refer to two turning points in the lives
12
of most people. The problem of identity, usually en
countered in late adolescence and early adulthood, is
to mold out of one's heritage, one's conception of one
self, and one's conjectures about the image one pre
sents to others, some satisfactory self-image, iden
tity, or personal identification. The problem of in
tegrity, which typically occurs in later life, although
it is often fused with the problem of identity in the
life of the homo religiosus or philosophicus, is this:
Given one's identity, purpose, and claims about oneself,
how can one live up to them, and cope with the fact
that one can rarely if ever be entirely successful in
such aims? (10:9)
Rationalists have fallen into a perpetual crisis of
integrity because they have neglected their crisis of iden
tity. Their theory of rationality gives rise to insuper
able problems. Since they wed rationality to the justifi
cation of their assertions, they are confronted with a loss
of integrity by being unable to fulfill the ideal of justi
fying all their claims. In the final analysis, therefore,
their own identity is not preserved in contradistinction to
irrationalists, for both base their ultimate points of ref
erence on an arbitrary choice.
The failure of rationalists to resolve these prob
lems is not due to the incapacity of rationality or reason
ing, but to an inadequate theory of rationality. Only when
the crisis of identity is overcome can the crisis of integ
rity be resolved. To explain this thesis, Bartley con
trasts three main stages in the rationalist quest for iden
tity and integrity, namely, (l) comprehensive rationalism,
(2) critical rationalism, and (3) comprehensively critical
rationalism.
Comprehensive Rationalism
Bartley observes that while comprehensive rational
ism is especially characteristic of modern philosophy, it
can be found as early as Epictetus, who said that every
thing reasonable may be supported. Its identity, there
fore, is made to rest on the acceptance of only those posi
tions that can be justified by an appeal to rational crite
ria. Both intellectualists (i.e., "Rationalists," in the
seventeenth-century sense) and empiricists were comprehen
sive rationalists, for the former appealed to intellectual
intuition and the latter appealed to sense observation as
the ultimate justifying criteria. That both were proven
inadequate is an integral part of the history of modern
philosophy. Hume effectively exposed the failure of empir
icism, and Kant revealed the defectiveness of intellectual-
ism.
Bartley avers that the resuscitation of comprehen
sive rationalism is impossible. He contends that its com
mitment to accepting only those positions that can be jus
tified by appealing to rational criteria is necessarily
self-referential, and thus it requires its own justifica
tion. But neither intellectual intuition, sense experi
ence, or any other rational authority can justify it. Fur
thermore, any attempted justification of the practice of
justification would be circular, and therefore futile.
Bartley contends that "the argument would be acceptable
14
and convincing only to those who had already adopted the
principle that arguments should count" (10:13).
Critical Rationalism
The apparent insurmountability of this dilemma has
led to attempts to formulate an alternative theory of ra
tionalist identity, viz., "critical rationalism." A well-
known exponent of this influential contemporary theory is
A.J. Ayer, who once adhered to the comprehensive rational
ism of sense experience (cf. his Language, Truth and Logic,
1936). But he sought to replace this inadequate theory
with the more sophisticated position of critical rational
ism, which he defended in The Problem of Knowledge (1956).
There are three main points on which Ayer and other
adherents of critical rationalism are in general agreement:
(l) They concede that rationality is limited in the
sense that some matters, like the rationalist position
itself, cannot be justified by appeal to rational
standards. ... (2) They claim that this concession is
unimportant, or at least not important enough to give
any consolation to irrationalism. (3) If challenged,
they tend to ground or justify their rationalist posi
tion in personal or social commitment to standards
which are beyond challenge. (11:125-26)
Thus, Ayer acknowledges the impossibility of providing a
rational justification of one's basic philosophical princi
ples and procedures. By not claiming to do more than he
logically can, he appears to circumvent the problem of in
tegrity. His theory of rationality does not demand that
ultimate principles be rationally justifiable, and there-
15
fore he can admit that the skeptic and fideist are correct
in asserting such justification to be impossible. But it
is a critical rationalism because he proceeds beyond this
concession to an attempt to demonstrate why his theory does
not need the justification of ultimate standards. If this
can be done, then the criticism of the skeptic and fideist
would lose its force.
What is the nature of Ayer's case against the de
mand for the justification of ultimate standards? This
question can be answered by noting what he maintains about
the method of induction. He says that "it could be irra
tional only if there were a standard of rationality which
it failed to meet; whereas in fact it goes to set the
standard: arguments are judged to be rational or irrational
by reference to it" (8:75). And if there can be no proof
that what we take to be good evidence really is so, says
Ayer, "it is not sensible to demand one. The skeptic's
problems are insoluble because they are fictitious" (8:8l).
Nevertheless, Bartley deems Ayer's approach to be
unsuccessful, because, as a theory of rationality, it fails
to defeat the crucial argument of the skeptic and fideist,
namely, that "since comprehensive justification is impossi
ble, the choice between competing ultimate positions is
arbitrary" (10:15). And the reason Ayer's theory cannot
defeat this argument is that he begs the question by assum
ing that his own particular standards and procedures of
rationality are the correct ones. But how does he know
these are the right ones? It seems possible to conceive of
different criteria of rationality which may invalidate all
other putative standards of rationality. "Whether regarded
historically or theoretically*" says Bartley* "the most im
portant criticisms of putative standards of rationality
have questioned whether they were correct" (10:16). This
is the crucial issue in the problem of rationality* and
Ayer has provided no way out of the impasse. Bartley cites
Popper's position as an alternative to Ayer's putative
standards of rationality. Popper denies the existence of
an inductive method* and he argues instead for a different
conception of scientific procedure. Hence* if mutually ex
clusive standards of rationality can be claimed* it is in
sufficient for one merely to assert that his standards are
ultimate or incontrovertible.
It follows* according to Bartley* that Ayer's posi
tion is fideistic* as is any position that is based on a
petitio principii. Ayer makes an irrational commitment to
certain standards like the principle of induction* for he
provides no rational means whatever for defending his posi
tion against critics or for deciding between alternative
views of ultimate standards and procedures of rationality.
Moreover, he has no defense against the irrationalist who
denies the rationality and primacy of standards like the
principle of induction. Hence* Ayer automatically guaran-
17
tees the rationality of his standards by creating a "per
suasive definition" of "rational." Ultimately appealing to
some form of conventionalism, Ayer can only retreat to his
commitment when his view is challenged by opponents outside
of the group of which he is a member. Critical rational
ism, therefore, is unable to solve the problem of rational
ity, for it cannot escape the skeptico-fideistic charge of
ultimate arbitrariness.
Rationality and Authoritarianism
According to Bartley, the inadequacy of the theo
ries of comprehensive and critical rationalism lies in
their inherently defective structure. He argues that "cer
tain features of these theories of rationality . . . pre
determine the kinds of questions asked in philosophical
discussion and limit the range of answers deemed appropri
ate" (10:19). This structure Bartley calls "authoritarian,"
and he maintains that it has dominated even the most lib
eral forms of our philosophical tradition. Its authoritar
ian character can be seen in two of its philosophical
dogmas which have generally gone unnoticed:
(l) The assumption that criticism is necessarily fused
with justification; and (2) the assumption that the
quality and degree of rationality pass through the
relationship of logical deducibility from justifying
premises to justified conclusions. (10:20)
Modern philosophy has made the mistake of pitting
rational authorities against irrational and traditional
18
authorities. In doing so it has merely rejected particular
authorities, but it has not rejected authoritarianism. It
has focused on questions that beg authoritarian answers.
It has asked for the justification of beliefs and the guar
antee of opinions, and it has answered with the appeal to
allegedly unquestionable authorities that purportedly guar
antee the correctness of a conclusion. Without the recog
nition of the authoritarian character of traditional philo
sophy, the following crucial question could not have been
raised: Is a nonauthoritarian theory of rationality possi
ble? Bartley not only believes that it is possible, but he
also believes that it is necessary if the problem of ra
tionality is to be solved.
Karl Popper is credited by Bartley as being the
first to call for a radical revolution in our way of view
ing the nature of philosophy. Popper goes beyond Ayer's
repudiation of the demand for rational proofs of our ra
tional criteria. He also calls for the rejection of "the
demand that everything else except the standards be proved
by appealing to infallible intellectual authorities (which
do not exist anyway). . ." (10:21). Indeed, the tradi
tional authoritarian questions should be replaced by the
criticism of standards as the main task of philosophy.
Philosophers have been radically misdirected in their
search for infallible intellectual authorities. Instead,
they should be primarily concerned with constructing a
19
philosophical program for the elimination of error.
Within such a program., the traditional "How do you
know?" question would not legitimately arise. And if
it arose in fact, the philosopher would have to reply:
"l_ do not know; I have no guarantees." To be more pre
cise, such a philosopher might elaborate: "Some of the
theories I hold may in fact be true; but since there
are no criteria of truth, I can never know for sure
whether what I believe to be true is in fact so."
(10:21)
Following this program, the philosopher would seek for
every possible means to keep ideas and practices subjected
to the maximum criticism, for only in this way could intel
lectual error be effectively counteracted. Bartley heralds
this shift from the demand for authoritative justification
to the demand for criticism as a "genuine innovation" in
philosophy.
The novelty of this concept of criticism, Bartley
contends, lies in the fact that it diametrically contrasts
with the traditional ideas of criticism, for the latter
always fused the idea of criticism with the idea of justi
fication. The failure to see that criticism and justifica
tion could be separated made it inevitable for professedly
anti-authoritarian philosophies to be authoritarian. Au
thoritarianism and justificationalism are inseparable, but
not criticism and justificationalism. It is impossible for
justificational philosophies of criticism to escape commit
ment to various ultimate authorities, for justification can
only be accomplished by the appeal to authority. Indeed,
20
criticism within this structure could only be made from
these justifying points of reference. In contradistinction
to this approach is Popper's position,, whose "main origi
nality . . . lies in the fact that it is the first nonjusti
ficational philosophy of criticism in the history of phi
losophy" (10:23).
Nonjustificational criticism can be more readily
grasped if it is seen against the backdrop of specific
cases of justificational philosophies. It was a common
assumption of intellectualists and empiricists alike that
the criticism of an idea was to be carried out by deter
mining whether or not it could be rationally justified.
For example, Descartes rejected as rationally unjustified
whatever ideas that could not be reduced to clear and dis
tinct ideas. Hume's rational authority was sense experi
ence, and every idea or theory that could not be shown to
be derived from this sole source of human knowledge was
rejected, i.e., it was held to be without rational support.
And other modern philosophers have likewise carried out
their evaluation of views by reference to logical deriva-
bility or underivability from their particular rational
criteria. Moreover, justificationists have generally held
that if a certain belief can be shown to conflict with
their allegedly ultimate standards, that belief is irra
tional. In any case, whether a belief is derivable or
21
underivable from., or in logical conflict with, the partic
ular rational authority, the latter is held to be immune
from criticism.
Comprehensively Critical Rationalism
At this point Bartley raises a question that leads
to a full explication of nonjustificational criticism. He
asks why justification and criticism have been repeatedly
fused in the ways described. This question he answers in
terms of the second assumption of traditional philosophies:
"Most "philosophical views have tacitly taken for granted
that rational character and degree of rationality are pro
perties that pass from premises to conclusion in the same
manner as the property of truth through the relationship of
logical deducibility" (10:24). The standard of rationality
was identified with intellectual respectability, and crite
ria of respectability were regarded as criteria of truth.
But criteria of truth gave way to a weaker standard, viz.,
the probability calculus. Not only is truth transmissible
from premises to conclusion through the deducibility rela
tionship, probability is also transmissible--i.e., "a con
sequence will be at_ least as probable as the premise from
which it is derived. It may of course be more probable"
(10:26). Nevertheless, Bartley argues that criteria of
truth are unavailable, and measures of probability are in
adequate for deciding rationally among competing scientific
22
hypotheses and less precise ideas. Since traditional phi
losophies take the transmissibility principle for granted,
they retain probability measures which they consider to be
logically transmissible.
Bartley argues, however, that it is a mistake to
think that empirical character can be transmitted through
the deducibility relationship. The "empirical character"
of statements is the property that makes it possible to
understand them in terms of reports of sense experience.
The statements in which sense experience is actually re
ported are termed "basic statements." It is an error to
think that induction from particular empirical statements
can establish the truth or probability of universal scien
tific hypotheses, for the latter "cannot be reduced to
truth functions of a finite class of basic empirical obser
vation statements" (10:27). According to Bartley, Popper's
theory of testability for evaluating scientific theories
avoids such difficulties.
The nonjustificationalism of Popper's view lies in
the attempt to falsify theories, in contradistinction to
the attempt of justificationalism to "probability" theories.
While truth and probability are supposedly logically trans
missible, falsity and degree of testability are logically
retransmissible (by modus tollens). Popper also seeks to
avoid what he considers the chief error of probabilistic
theories., namely, their assumption that any logical conse
quence of a hypothesis is as highly testable as the origi
nal hypothesis. He argues that it is a mistake to think
that testability is transmissible from premises to conclu
sion. For example, the proposition "All orbits of planets
are circles" is deducible from "All orbits of heavenly bod
ies are circles." Let us refer to the latter as p_ and to
the former as q. It is clear that q is not as testable as
p_, for with c i the degree of universality decreases, and
this means that the number of potential falsifiers of q_ is
less than those of p. The proposition with a higher degree
of falsiflability (p_) is logically less probable than the
one which is less well testable (q). The less falsifiable
a proposition is, the more probable it is, simply by vir
tue of its logical form (53:122). Hence, according to
Bartley, there is a fundamental difference between various
probabilistic theories of confirmation and Popper's theory
of testability: the former are inexorably bound to logical
transmissibility from premises to conclusion, while the
latter depends solely on logical retransmissibility from
conclusion to premises.
By adopting Popper's theory of falsiflability and
making it integral to his position, Bartley believes he has
found a basis for nonjustificationalism in philosophy.
Since it hinges upon logical retransmissibility alone, his
approach can be considered nothing more than a philosophi
cal program for counteracting intellectual error. At
least, this is what Bartley contends. Moreover, he says
that his view holds all beliefs open to criticism and con
tinued testing, including itself. Since rationality is un
limited in nonjustificationalism, it requires no commit
ment. Thus, a tu quoque argument cannot appropriately be
leveled at the comprehensively critical rationalist when
ever he accuses his opponent of claiming that some belief
of his is immune from criticism because of irrational com
mitment .
Since no theory can be decisively and finally re
futed, according to Bartley, the means for eliminating
error by criticizing our conjectures are not accepted as
uncriticizable. They are open to testing and possible re
jection as well. There are four basic checks that should
be applied in criticizing views: (l) logic (i.e., is the
theory in question consistent?), (2) sense observation
(i.e., is the theory empirically refutable by some obser
vation? And if so, do we know of any refutation of it?),
(3) scientific theory (i.e., is the theory, whether or not
it seems to be incompatible with sense observation, incom
patible with any scientific hypothesis?), and (4) the na
ture of the problem in question (i.e., what problem is the
theory intended to solve? Does it do so successfully?).
25
Since these tests are used to criticize a theory and not to
justify it, the position can be held rationally. And since
this position itself can be criticized, it satisfies its
own requirements and overcomes the crisis of integrity in
which other forms of rationalism have been caught (11:158-
59).
Criticism of Nonjustificationalism
From this summary statement of Bartley's position
I now turn to critical considerations. First of all, it
must be stressed that by the term "irrationalism" Bartley
means any position that holds some criteria or presupposi
tions to be beyond criticism. These presuppositions he
calls "commitments." Bartley insists, however, that cri
ticism is genuine only when the possibility of refutation
is not precluded. Hence, for rationalism to be truly com
prehensive and critical, every view and position must be
open to criticism, including the practice of criticism
itself.
There are two basic issues that reveal his position
to be untenable. These concern (l) the nature of criticism
and (2) the goal of intelligibility. It will be seen that
in both cases Bartley's view necessitates an "ultimate com
mitment." And it should be noted that the allowance of
only one ultimate commitment (i.e., a presupposition that
26
is irrefutable in principle or by stipulation) will serve
to destroy the integrity of comprehensively critical ra
tionalism.
The Irrefutability of Logic
To some extent, Bartley is aware of the problem
of avoiding commitment in the practice of criticism and in
the use of the principles of logic entailed by it. But
while he thinks that this problem only involves a serious
difficulty that can be hurdled, in reality it involves an
impossibility that cannot be overcome. What, then, is
this impossibility that destroys the integrity of his posi
tion?
Bartley recognizes that certain changes have taken
place in logic, but he points out that all of the alterna
tive systems of logical rules of inference have a funda
mental characteristic in common: "whenever we observe these
rules and, starting with true premises, argue in accordance
with them, we arrive at true conclusions" (11:169). W.V.O.
Quine, Morton White, and Reinhold Niebuhr, among others,
seem to contend that we can revise logic in the sense of
denying that true premises need always lead, in any valid
inference, to true conclusions. Bartley seems to argue
that logic is not revisable in this way, for he says that
"we cannot regard logic as part of the set of beliefs that
are put to the test in critical discussion, for the notion
27
of testing and revising in accordance with the results of
the test presupposes logic" (11:170). Criticism presup
poses the principle of deducibility. He concludes that
there can be no critical argument without it: "Thus I have
stated an absolute presupposition of argument to which we
are committed not as_ human beings, because of our biology,
psychology, or sociology, but a_s arguers about the world"
(11:172) .
In combating Quine, White, and Niebuhr, Bartley
says that logic cannot be part of the totality that is
brought under test (11:171). Yet, on the other hand, he
contends that all means for eliminating error are open to
refutation: "we must be willing to reopen to examination
and further criticism and possible rejection all the criti
cal arguments and critical institutions we have accepted"
(italics mine, 11:158). And then he says that the most
important critical institution is logic. Moreover, he ex
plicitly states that "everything, including the practice of
arguing and revising (and using logic), is open to criti
cism and rejection" (11:172).
Bartley is partially aware of the problem he has
created for himself. He tries to solve it by saying that
to reject logic (i.e., either to deny that the universally
acknowledged rules of valid inference can be known to al
ways lead to true conclusions from true premises when
28
applied in an argument, or to refuse the practice of eval
uating arguments according to logical rules) is to reject
the practice of criticism. And conversely, "if we reject
the practice of argument and revision, we may reject logic"
(11:172) .
The precise issue at stake, however, is the skep-
tico-fideistic contention that to accept the practice of
critical argument (and hence logic) is to make an "irra
tional" (i.e., unjustified and unjustifiable) commitment,
for there is no way to "justify" it or to refute it without
presupposing it.
This is the point at which Bartley relapses into
rationalism's crisis of integrity. The structure of his
theory necessitates an ultimate commitment to critical ar
gument and to the logic that it unavoidably presupposes.
The skeptic and the fideist will ask why he accepts these.
Any attempt to give an answer will either assume critical
argument and logic in the very act of giving the answer or
it will be a mere "confession" of what one believes without
rational justification. But even a "confession" must be
made in well-formed expressions if it is to be intelligible
or coherent. Moreover, if it involves a cognitive claim,
it thereby presupposes the law of noncontradiction, for not
only must its words have a determinable meaning but some of
its statements must be either true or false and not both.
The skeptic and fideist hold the choice between
critical argument and confession or silence to be arbi
trary. And they maintain that it is arbitrary because no
justification can be given for either one. Now Bartley
must either agree with them and say that the choice is
arbitrary or he must provide a justificatory answer. He is
compelled to do the latter, thereby destroying the integ
rity of his "comprehensively critical rationalism." His
recourse to justification, is seen in his contention that
"if we want to learn about, or even to describe the world,
we need to be able to derive true conclusions from true
premises" (11:171). And if one asks for a justification
for learning about the world or why it is that logic is
necessary for learning about it, a justificatory procedure
in providing an answer is ineluctable. Eventually one must
stop at some point or concur with the skeptical alternative
of an infinite regress. Thus, Bartley is caught in the
very skeptico-fideistic dilemma that he had hoped to over
come .
Can Bartley "nonjustificationally" assume, then,
that critical argument is better than the mere confession
of one's views or total silence? How can he refute the
latter positions? If the advocate of confession or si
lence maintains such a view because he believes that at
crucial points critical argument and logic either lead to
a distorted view of reality or fail to mediate reality al
30
together* what can convince him otherwise? Critical argu
ment will he powerless* for how can he he convinced by that
which he does not accept as the sole or ultimate method of
apprehending what is real? The principle of retransmission
of falsity cannot he appealed to as the ultimate touch
stone* for the advocate of confession or silence refuses to
accept it as such from the outset. He charges that to hold
to the validity of the principle of retransmissibility is
another form of fideism. And on Bartley1s terms* it must
he conceded that the fideist is correct in this allegation.
For* if Bartley's options are the only ones* then one must
be either a skeptic* a fideist* or a comprehensively criti
cal rationalist. But since nothing could conceivably re
fute the practice of critical argument (whose differentiae
include fundamental principles of logic and rules of in
ference)* Bartley is "committed" to critical argument* and
hence he is a "fideist."
Bartley would undoubtedly counter these remarks by
insisting that the principle of retransmissibility is crit-
icizable. But one might ask how it can be criticized.
Certainly not by other means* for* according to Bartley*
that would be to acquiesce to justificationalism and all of
its self-defeating problems. There is* for him* no prin
ciple of criticism above retransmissibility. And it is the
sine qua non of nonjustificationalism.
31
It is important to see., in this connection, that
Bartley is evidently misled by a confusion between the
psychological and the logical. Rejectability and refutabil-
ity, however, are not identical. While it is possible for
someone to reject the practice of criticism due to a vari
ety of psychological causes, it is impossible for one to
refute it by logical means. The repudiation of critical
argument, therefore, can never be on the basis of logical
refutation. Of course this holds only if critical argument
is understood to be characterized necessarily by the prin
ciples of logical refutation. And, indeed, this is not
only how Bartley understands it, but there seems to be no
way to construe it apart from its use of logical prin
ciples. Thus, the principle of retransmissibility could
never be used to refute itself, for in the attempt it
would be presupposed by being the means of criticism. And
if there is at_ least one criterion or presupposition that
is irrefutable in principle, one has an ultimate commit
ment. Consequently, Bartley's charge of fideism ineluc
tably boomerangs upon his own view.
Intelligibility as a Goal
When Bartley argues that the quest for intelligi
bility (i.e., linguistic and propositional clarity in des
cribing and speaking about objects) or knowledge (i.e.,
awareness of the truth or of true propositions) is impos
32
sible without critical argument, he merely takes the ques
tion of justification to another level. Why should anyone
be "committed" to this quest? If Bartley says that this is
not a legitimate question to ask, he contradicts himself--
for then there is some issue that is protected from criti
cism, and hence his view is no longer comprehensively crit
ical rationalism. He intended to make it comprehensive by
saying that everything must be open to criticism, whether
"standards, criteria, ends, or goals" (10:6) or "beliefs,
and positions and ways of life, whether scientific, mathe
matical, moral, religious, metaphysical, political or
other" (10:5).
The value judgment that says that intelligibility
or knowledge ought to be achieved is implicit in every act
of criticism. Without the assumption that they are good,
it is difficult to see any basis at all for rational dis
cussion. And, as has been noted, it is mandatory for
Bartley's form of "rationalism" that the object of criti
cism be open to the possibility of refutation. But how
can one criticize this value judgment without begging the
question? As in the case of criticizing retransmissibil
ity, so here one cannot avoid the paradox of attempting to
refute what he, at the same time, is presupposing.
There is no way out of this dilemma for Bartley
except to deny that nonjustificationalism should be applied
33
to goals and value judgments, or at least to the one in
question. But this he does not do. Neither is it some
thing that he can do without allowing his comprehensively
critical rationalism to lose its integrity.
Conclusion
I have sought to show that there are unwitting
"ultimate commitments" in Bartley's position, and that they
serve to destroy the integrity of his so-called nonjustifi-
cational philosophy. His attempt to construct a thorough
going nonjustificationalism is a bold and interesting con
jecture, but it is undermined by its own inherent contra
dictions .
Briefly summarized, my argument has been as follows.
Bartley maintains that every practice and proposition is
refutable--but, of course, not that each one will necessar
ily be refuted. Nevertheless, his claim of universal ref-
utability means that if it can be shown that there is at
least one practice or proposition that is irrefutable, then
nonjustificationalism is unrealizable. According to
Bartley, the practice of criticism is itself open to criti
cism. And criticism, he says, is not genuine if the possi
bility of refutation is antecedently precluded, either by
stipulation or by logical necessity. Now refutation can
only be accomplished by critical argument. Therefore,
there must be allowed the genuine possibility that critical
argument can refute critical argument. But this is mani
festly absurd. Apparently, Bartley failed to see this ab
surdity because he neglected to distinguish rejection from
refutation. While the rejection of critical argument does
not require critical argument, the refutation of critical
argument can be attempted only by critical argument. If
the latter were possible--which it is not--the absurdities
it would involve can be seen in the following antilogies:
retransmissibility can refute retransmissibility, logic can
refute logic, criticism can refute criticism, and reason
can refute reason. Clearly, then, Bartley's theory of ra
tionality makes rationality impossible.
Stipulating that everything is open to the possi
bility of refutation does not make it so. The practice of
criticism, the law of noncontradiction, and the goal of in
telligibility are irrefutable in principle, for without
them there can be no refutation. And since there is at
least one practice or proposition that is irrefutable in
principle, nonjustificationalism is untenable.
Bartley's unsuccessful attempt to retreat from com
mitment is a tour de force that leaves unscathed the re
treat to commitment. If skepticism and the fideistic re
treat to commitment are to be overcome--and I believe they
can be--it must be on the basis of a "rationalism" signifi
cantly different from comprehensive rationalism, critical
rationalism, and Bartley's putatively comprehensive, criti-
35
cal rationalism.
To find, out why Bartley believed that he could con
struct a new theory of rationalism on the epistemological
foundation laid by Popper, it is necessary to examine some
of the latter's main theses. As I proceed to do so in the
next chapter, it will be found that Bartley's position is
based on the assumption that truth is undiscoverable. And
this raises the question whether such an assumption is
warranted.*
*1 have thoroughly searched the leading philosophi
cal journals and periodicals, but I have been unable to
find any articles or books in which Bartley's position is
discussed. A number of the reviews of his book, The Re
treat to Commitment, are listed in my bibliography. How
ever, I have found that none of these reviews indicates
the precise nature of the paradoxes which make nonjustifi
cationalism untenable.
CHAPTER II
IS TRUTH DISCOVERABLE?
Karl Popper is an adamant foe of classical ration
alism and classical empiricism. He maintains that they
both err by turning their attention to the question of the
source of knowledge, arguing that it is either reason (in
tellectual intuition) or observation (sense perception), as
the case may be. He is also opposed to epistemological
relativism, or "the idea that there is no such thing as ob
jective truth" (52:4), and to epistemological pragmatism,
or "the idea that truth is the same as usefulness" (52:5).
Crucial for my discussion, however, is his criticism of
epistemological optimism, which he describes as the view
that man has the power to acquire knowledge because truth
is manifest. By this latter phrase is meant the idea that
truth either reveals itself or that it may be discovered by
us if we take proper steps to unveil it. Popper is con
cerned to dispute this claim that "once the naked truth
stands revealed before our eyes, we have the power to see
it, to distinguish it from falsehood, and to know that it
is truth" (52:5).
A Basic Problem of Epistemology
In his Conjectures and Refutations, Popper says
that he has "only one interest--to find out the truth about
the problems of eplstemology" (52:6). He argues that the
chief epistemological error is the view that "truth, if put
before us naked, is always recognizable as truth" (52:7).
The "natural light" of reason enables us to see the truth,
according to this view, so that further argument is un
necessary. Either a person "sees" it or he does not.
Hence, knowledge is the possession of truth, and if anyone
is ignorant, it is because of some adverse influence on his
mind or through his own "sinful refusal" to see the mani
fest truth. Popper calls this "the conspiracy theory of
ignorance," and terms both it and the idea that truth is
manifest "myths." While having had a number of incidental
salutary effects, these myths have also led, according to
Popper, to the disastrous consequences of fanaticism and
authoritarianism.
I doubt that Popper would want to argue that these
adverse effects serve to refute the view that truth is man
ifest. They do not disprove it anymore than its benign
historical consequences prove it. Popper contends that "it
is a case of a bad idea inspiring many good ones" (52:8).
But he uses the adjective "bad" before he has shown why it
is a "bad" idea. In response to his allusion to the "dis-
astrous consequences" of the idea, one could just as well
say, "it is a case of a good idea inspiring many bad ones."
Since examples of both kinds can be cited from history,
they do little to resolve the question of the truth or fal
sity of the claim that truth is manifest.
What kind of an argument, then, does Popper present
to rebut what he calls a "false epistemology"? First of
all, it must be noted that he argues from an ambiguous po-
sition--with the usual attendant advantages and disadvan
tages of equivocation. On the one hand, failure to define
his terms adequately seems to give him a stronger case than
he actually has. On the other hand, important issues are
clouded and crucial questions are glossed over because his
own epistemology is truncated and unclear. For example, he
appears to reject categorically the view that truth is man
ifest. But the issue is not so sharply drawn, for in one
place he says "truth is not manifest, as £ rule" (italics
mine, 52:9). He never explains what he means by this
qualifying phrase, however. If truth is usually not mani
fest, then there is a strong suggestion that sometimes it
is. But he does not afford even one example of a manifest
truth. In fact, he repeatedly denies that there is such a
thing. But if he does mean to concede that there are some
instances of manifest truth, it is incumbent upon him
(whose chief interest is the discovery of the truth about
the problems of epistemology) to tell us by what means he
39
distinguishes manifest truths from those which are not, and
to give us at least one example of a manifest truth.
Otherwise, he ought to drop the modification "as a rule,"
for it serves no purpose except to confuse.
It may be, however, that by the phrase "as a rule"
Popper only intended to express the Pyrrhonistic mitigation
of his claim that truth is not manifest. Perhaps he was
seeking to avoid the contradiction that results from
asserting the "truth" that "truth is not manifest." For so
stated, the latter is surely claimed to be a manifest
truth. But if he wished to den;y that this statement is
unqualifiedly true (i.e., if he intended to assert "It is
never the case that truth is not manifest"), then that
again is an implicit claim to a _ truth that is manifest, for
he thereby suggests that this proposition can be known to
be true. It is unlikely that he is unacquainted with the
elementary blunder of the Academic Skeptics who dogmati
cally asserted their skepticism to be the only "truth."
But then again even the most eminent philosophers have
sometimes made elementary mistakes. As we shall see later,
some combination of Pyrrhonism and fideism seems to be his
final position, but for the present it is important to con
tinue our analysis of this fundamental step in the con
struction of his view.
40
Truth and Error
Popper appears to use the terms "discoverable" and
"manifest" as synonyms. If truth is discoverable, then it
is manifest, and vice versa. This is clear from his state
ment, describing the doctrine in question, that "truth, if
it does not reveal itself, has only to be unveiled, or dis
covered" (52:7). Moreover, his references to Socrates's
maieutic, Plato's anamnesis, and Aristotle's induction, as
exemplifying the theory that truth is manifest, show un
equivocally that he does not see any important distinction
between the claim that truth is manifest and the claim that
truth is discoverable. This confusion is fatal. For exam
ple, Popper says that "our ignorance must necessarily be
infinite" (52:28), and that all we can hope to do is grad
ually approximate truth by eliminating error. In other
words, the refutation of theories is supposedly a method of
reducing the range of error. But if error, or ignorance,
is infinite, the method can never be effective. For an
infinity cannot be subtracted from or diminished, to say
nothing of the provisional character Popper attributes to
every "refutation" of an error. If_ the method is useful at
all, it is because the elimination of error takes place in
the context of the recognition of truth. The removal of
error is possible and meaningful because error is limited
by truth, apart from which it could not be known. Ilf truth
were undiscoverable, there would be no way to determine
4l
whether a refutation is the elimination of error and the
approximation of truth, or whether it is the commission of
error and the retreat from truth.
The only explanation of the limitedness of error,
however, is to be found in the fact that things have a de
terminate nature (i.e., certain qualities, forms, and rela
tions) which can be known. But this means that there are
knowable essences--a conclusion Popper resists. Often in
the history of philosophy an impressive theory, such as
Popper's, is shown to be specious once the flaws in its
fundamental premises have been detected.
Knowledge and Conjecture
Popper's theory of truth and error has been deeply
affected by the fact of flux and inconclusiveness in the
history of philosophy and science. Obviously it is impera
tive for an epistemologist to reckon with this fact, but it
is equally vital that he not be overwhelmed by it to the
extent that permanence and finality, however meager, are
obscured. It seems to me that Popper is the victim of this
myopic reading of history and experience, for the crux of
his position is the replacement of knowledge with conjec
ture .
Popper contends that Socrates, Plato, Aristotle,
Francis Bacon, Descartes, and a host of other philosophers
erred in the belief that certain knowledge could be had
and in the assumption that there was a particular method
for acquiring it. They were all convinced that certain
knowledge (episteme) was available to all who took the pro
per steps to attain it. Popper is just as convinced that
all of our vaunted knowledge is mere guesswork (doxa) based
on a method of hypothesis. But on what basis does he make
such a sweeping claim? Primarily on the grounds of pur
portedly psychological findings. While Bacon believed it
was possible to purge our minds of "all anticipations or
conjectures or guesses or prejudices" so we can "read the
book of Nature" (52:14), Popper is convinced that it is
impossible to purify our intellects of all distorting, sub
jective influences. Without the cleansing of our minds
from the latter, it is impossible, he says, to recognize
the manifest truth or truly "read the book of Nature."
Popper is most emphatic about this view. There is
no "pure observation," for all observations "are interpre
tations in the light of theories" (52:38). This creates
serious problems for him, however. First, he offers this
as a piece of certain knowledge, while at the same time he
denies the possibility of episteme. Of course he..can sim
ply declare all of his statements to be "conjectures,"
their dogmatic form notwithstanding. This Pyrrhonistic
ploy will be examined later, but an interesting problem to
note at this point is that there is an obvious difference
in the way certain claims are put forward in his writings.
If every statement is inconclusive and tentative, why is it
that some are expressed in dogmatic terminology while
others are not? Degrees of certitude certainly seem to he
reflected in the forms of expression. Moreover, specific
examples from his writings can be cited to show that he ex
presses doubt, liklihood, and assurance. If these mean
nothing in the final analysis, then what is their purpose?
And if they mean something, then what are the criteria by
which they are distinguished? Why is it that some state
ments are qualified by expressions like "perhaps," "maybe,"
"it seems," "undoubtedly," "probably," "certainly," "ob
viously," etc., and other statements are not? If there is
a significant difference between "perhaps" and "obviously,"
then there must be some way to decide which term should be
attached to a particular statement. But if this is the
case, then Popper is unwittingly making a claim to some
certain knowledge.
Or would Popper indeed adopt the position of the
radical skeptic who says that all propositions are equally
uncertain (including this one), and that the employment of
such diverse qualifying terms is either a mere expression
of the speaker's subjective certitude (in contrast to the
question of the objective certainty of the propositions)
or a mere rhetorical device to persuade others to adopt the
point of view of the author? But in the first instance,
the skeptic makes the implicit claim that he knows the dif
ference between subjective certitude and objective certain
ty, and that it is the former alone and not the latter to
which the qualifiers refer. In the second instance, he
makes the implicit claim that he knows the difference be
tween a rhetorical and non-rhetorical device, that other
individuals exist, that they can be persuaded, and that he
himself has a particular point of view. This is just one
of the serious problems created by Popper's epistemological
assumptions.
In the second place. Popper's theory of observation
ensnares him in a vicious circle. He avers that "even the
simplest observation statement" involves interpretation in
the light of theories, "and that it is therefore uncertain"
(52:4l). What kind of a statement is this, then, about
observation statements? Is it not an observation statement
itself, or more precisely and equally as fatal to his theo
ry, is it not a generalization based on observation state
ments? If not, how does he arrive at it? Is he privy to
esoteric knowledge, or is there some a_ priori source of
knowledge to which he has access? If he knows this a_ pri
ori, then there must be some reason why he is aware of it
while many other philosophers are not. Is it because they
have not adopted the proper method for recognizing this
"manifest truth"? Or is it perhaps due to their own "sin-
ful refusal to see the manifest truth"? If, as it seems
likely, it is an empirical generalization, then it is based
upon observations that are "theory-laden." Yet if no pure
observations are possible, "there can never be anything
like a completely safe observation" (52:4l). It follows,
then, that this itself is not a completely safe observa
tion. Since he has assumed that it is impossible to escape
(or know if we escape) the distorting effects of our sub
jective expectations and prejudices, then surely he has no
grounds for categorically asserting the unattainability of
a "completely safe observation." His own presupposition
has precluded such a generalization.
This question must not be ignored by Popperian
epistemology: How does one know that "every statement in
volves interpretation in the light of theories"? Does he
know it deductively? Then let him show the proof. This
Popper has not done, and I do not see any way it can be
done without begging the question. Does he know it induc
tively? Has he inferred that since so many observations
have been faulty and so many observation statements have
been erroneous in the past that therefore all of them are
"uncertain," "unsafe," "theory-determined" interpretations?
If anything is certain and safe, it is the fact that no
such generalization is warranted. Of course Popper would
be the first to repudiate an inductive basis for his gener
46
alization, for he is convinced that a method of induction
is itself impossible. On what, then, does he base his
claim if there is no inductive method, if deductive proof
is unavailable or inconceivable, and if observation itself
is undermined by the intrusion of subjective distortion?
Furthermore, does Popper mean that he cannot be
sure that there is a chair in front of him or that he has
two hands? How, then, can he know what a "completely safe
observation" is, if there never is any such thing? None
theless, he makes numerous observation statements that are
ostensibly certain and purportedly based upon "completely
safe observations." He says that "observations are always
inexact," and are always made under very special conditions
and are always concrete (52:186). How does he know that
observations have these characteristics unless he has ob
served them being made? What justifies his use of the term
"always," which he repeatedly emphasizes in this connec
tion? This term makes it clear that he is expressing an
ostensibly universal and necessary proposition. Is this
warranted in view of his intransigent opposition to both
Aristotelian and Baconian forms of induction, or for that
matter, any other view of induction?
Another important problem that emerges here is
Popper's failure to come to terms with the complexities of
observation. For example, there seem to be two different
levels of observations: (l) our recognition of the fact
that observations are made under very special conditions,
and (2) our recognition of the specific kinds of conditions
under which a given observation is made. The epistemologi-
cal significance of this distinction is that I can know
some "givens" without distortion, because what is referred
to by some observation statements is sufficiently clear and
simple to be apprehended as it is. What, then, is the na
ture of observations? Are they discrete, or are the con
tents of observations given in pre-structured relations?
How can we know--or, to adopt Popper's terminology, on what
basis do we even "conjecture"--that what we observe is dis
torted by our subjective equipment? While one may hypothe
size the possibility of observational distortion in every
instance, it is important to ask if there is any basis for
this assumption in view of the fact that it makes impossi
ble any comparison of our observations with things as they
really are. As the Ding-an-sich was an Achilles' heel in
Kant's system, so for Popper here is a similarly indefensi
ble assumption that makes his view untenable.
The Scope of Popper's Position
Since Popper is well known for his contributions to
the philosophy of science, it may be thought by some that
what he has to say about theory and observation should be
strictly limited to the context of science. There are a
48
number of things to be said in reply to such a claim. It
should be clear from the quotations already presented that
Popper makes no such restriction. While it is true that he
sometimes limits himself to what he calls "science," it is
also the case that his discussion oscillates between it and
a comprehensive theory of epistemology. That he intends to
establish fundamental principles for epistemology is evi
dent from the very titles of some of the chapters in Con
jectures and Refutations: e.g., "On the Sources of Know
ledge and of Ignorance," "Three Views Concerning Human
Knowledge," etc.
All doubt, should be removed about the scope of his
inquiry by his statement that his only interest is to find
out the truth about the problems of epistemology (52:6).
In pursuit of this interest, Popper avers that "nothing can
be justified or proved (outside of mathematics and logic)"
(52:51). But this statement itself is obviously not math
ematical or deducible from mathematics. And neither is it
purely logical. Of course, if he merely stipulates that
all "justification" or "proof" must be understood in terms
of purely formal mathematical and deductive logical princi
ples, then his statement would appear to be an arbitrary
definition. But the statement purports to give us some
information, and not to be a mere stipulation or tautology.
This obviously raises an important question: How does he
49
justify or prove this assertion? If it fits neither the
category of mathematics or the category of logic, then,
according to Popper's own analysis, it is itself an unjus
tified and unproved proposition. Why, then, should anyone
accept this arbitrary dogmatism rather than some other way
of drawing epistemological distinctions?
Three Views of Human Knowledge
To understand further why Popper maintains that
truth is not manifest, one must turn to his exposition of
the "Three Views Concerning Human Knowledge." These three
views are essentialism, instrumentalism, and conjectural-
ism. Popper is opposed to the former two and believes that
his own view (conjecturalism) steers a safe course between
them. I shall argue, however, that he fails to maintain
his course between Scylla and Charybdis and that a_ form of
essentialism is the most adequate view of knowledge. Then
I shall discuss the relationship between this issue and the
problem of discovering truth.
Popper organizes his discussion of these rival
views around three doctrines:
(1) The scientist aims at finding a _ true theory
or description of the world^and especially of its
regularities or 1 laws1), which shall also be an ex
planation of the observable facts. (This means that
a description of these facts must be deducible from
the theory in conjunction with certain statements,
the so-called 'initial conditions.')
(2) The scientist can succeed in finally estab
lishing the truth of such theories beyond all reason-
50
able doubt. . . .
(3) The best, the truly scientific theories, des
cribe the 1 essences1 or the 'essential natures' of
things--the realities which lie behind the appearances.
1527103 -TT
Popper defends the first doctrine, for it is integral to
conjecturalism. He argues that the second and third doc
trines are mistaken. These last two doctrines constitute
the crux of "essentialism."
According to Popper, essentialism is the view that
there are "essences" (i.e., the actual, but hidden, natures
of things) and that man is able to discover them (i.e., at
least some of them). Once discovered, they furnish the
basis for ultimate explanations. By this is meant that
further explanations are unnecessary and even impossible.
For Popper, however, it is obvious that such a view is
obscurantist, for it precludes the raising of further
questions and the posing of new and fruitful problems. It
fosters the notion that a long-standing or seemingly well-
established scientific theory is the conclusive description
of the actual nature of things included in its scope. Why,
then, should one look beyond these explanations for differ
ent interpretations?
This criticism of essentialism seems cogent on
first consideration. But its plausibility stems from an
insufficient analysis of the concept of "essence." Popper
contends that science is destroyed if it ceases to grow.
But he also maintains that the search for essences retards
its growth and tends toward its complete stultification.
Whether or not it does, however, largely depends upon the
controlling conception of "essence" and of the nature of
the "search." In view of the diverse theories of essence
in the history of philosophy, it is curious that Popper
fails to discuss rival and Incompatible views in an effort
to clarify the precise meaning of the term as he uses it.
As it is, one can only piece together two or three oblique
statements of his in an attempt to understand what he means
by "essence." In referring to Descartes and Newton he
speaks of an essence as a "true or absolute property of the
thing (i.e., a property which does not depend on the exist
ence of other things) such as extension" and as "a rela
tional property, i.e., a property which, like gravity, de
termines the relations (interactions in space) between one
body and other bodies" (52:106). But these remarks are
qualified by an important phrase, viz., "the hidden es
sences of things" (52:104). What he means by this locu
tion is partially elucidated by his interpretation of
essentialism as distinguishing between (l) the universe of
essential reality, (2) the universe of observable phenom
ena, and (3) the universe of descriptive language or sym
bolic representation (52:108).
In view of this one-sided characterization of
essentialism as assuming that a real world of substantial
essences lies behind the ordinary world of mere appear-
ancesj it is not difficult for Popper to discredit it. He
argues that essentialism is groundless because it has no
way of knowing if its alleged discovery of the real world
is any more than another layer of the world of appearances:
"the world of each of our theories may be explained, in its
turn, by further worlds which are described by further
theories--theories of a higher level of abstraction, of
universality, and of testability" (52:115). If this is the
case, then there are no ultimate explanations--at least we
cannot know if we are in possession of one. It is due to
the claim of finality and absoluteness that Popper finds
essentialism especially objectionable. He says:
The empirical basis of objective science has thus
nothing 'absolute' about it. Science does not rest
upon rock-bottom. The bold structure of its theories
rises, as it were, above a swamp. It is like a build
ing erected on piles. The piles are driven down from
above into the swamp, but not down to any natural or
'given' base; and when we cease our attempts to drive
our piles into a deeper layer, it is not because we
have reached firm ground. We simply stop when we are
satisfied that they are firm enough to carry the struc
ture, at least for the time being. (53:111)
This is in keeping with Popper's insistence that we can
only achieve doxa. For if there are no "essences" (at this
point I am using the word to refer to determinate struc
tures and determinate processes in things), then there is
no knowledge. If we can never apprehend anything but
appearances, then "reality" forever eludes us, and without
53
a cognitive grasp of reality, knowledge is impossible. How
else is knowledge to be understood if it is not the cogni
tion of particulars and universals, or essences (i.e.,
qualities, forms, and relations)? Even Popper concedes as
much when he denies the possibility of knowledge due to our
inability to grasp "essences" or to know when we have. If
we cannot have true information about objects, then know
ledge is impossible and skepticism is inevitable. Perhaps
Popper believes he escapes the pitfalls of dogmatic skepti
cism by granting that there may be essences while denying
that we can know them. But one wonders what value this
abstract concession has for epistemology or ontology.
Clearly, if we can never know essences, the actual result--
in terms of a useful understanding of the world and of our
acting in it--is hardly different from denying them al
together.
Essences and Laws of Nature
The important question, however, relates to the
meaning of "essence." It is crucial to distinguish "es
sence" as an objective feature of things and experiences
from "essence" as a "metaphysical substance." Why should
it be thought that the mere assertion of the reality of
"essences" presupposes a so-called noumenal world? Does
"essentialism" invariably commit one to the dualism of
"phenomenon" and "noumenon"? Are there any viable views
54
of essence that do not depend upon this bifurcation?
Popper’s quasi-Kantian point of view is by no means philo
sophically secure. Since Kant's time it has been shown
that the noumenal postulate creates more problems than it
solves. If the resolution of philosophical difficulties is
to have any criteria, and if simplicity and coherence are
among them, then the sooner the Kantian dualism is jetti
soned the better. There seems to be no good reason for
making the paradoxical assumption of an unknowable noumenal
world. Yet without it Popper's system, no less than
Kant's, resists formulation. Apparently, Popper, like
Kant, could find no alternative way of accounting for the
nature of scientific progress. And yet there are important
differences between them, both in their understanding of
epistemology in general and of scientific knowledge in
particular. Popper indicates a basic difference when he
refers to Kant with qualified approval of the latter's in
sistence that we are barred from the "real world":
As Kant puts it: "Our intellect does not draw its laws
from nature . . . but imposes them upon nature." While
I regard this formulation of Kant's as essentially
correct, I feel that it is a little too radical, and
I should therefore like to put it in the following
modified form: "Our intellect does not draw its laws
from nature, but tries--with varying degrees of success
--to impose upon nature laws which it freely invents."
(52:191)
The last sentence of this quotation is an apposite summary
of Pepper's view. And it is crucial to our understanding
of his perspective on essences.
55
Obviously, if we cannot know essences (qualities,
relations, determinate structures), then we cannot discover
the laws of nature. We will necessarily be restricted to
theorizing about the structures and processes of nature.
We can never arrive at the point where any "law" is immune
from critical testing and possible refutation. In fact, we
can never really know any law of nature. Every such "law"
is our own "imposition" of a theory upon the entities we
encounter. Yet, it is not clear how certain entities can
be related under a single theory unless they have an "es
sence." Nor is it clear how we can "impose" a theory on
"entities" which present no objective features whatever.
Moreover, Popper's generalizations about "laws of
nature" suffer from a disturbing ambiguity. In particular,
his use of the term "law" leaves its meaning vague. For
example, at one point he seems to distinguish between ter
restrial laws (e.g., meteorological laws that make possible
the use of thermometers, barometers, and hygrometers) and
laws that purportedly hold throughout the universe (e.g.,
the second law of thermodynamics). He states that an im
portant characteristic of scientific laws is that "the more
they prohibit the more they say" (53:^1). Hence, when
scientific laws are said to be expressed by universal
statements, "universal" must be understood within the con
text of "universes of discourse," i.e., a theory or law
may be said, for example, to describe (l) "All the planets
in our solar system," or (2) "All the stars in our galaxy,"
or (3) "All the stars in the universe." Natural laws,
then, may he said to apply only to the earth and things
within it, or only to the solar system, etc. And this
would appear to make it possible, contrary to Popper's
skepticism, to establish some laws of "limited" universal
ity "beyond all reasonable doubt." Popper seems to allow
for degrees of generality or universality in scientific
laws, which he says are like proscriptions which deny that
certain things or states of affairs exist. It is because
of this that they are said to be falsifiable by an observa
tion statement which expresses a negative instance.
However, at another point, strangely enough, he
seems to deny degrees of generality, for he maintains that
"we can never know, of course, whether a supposed law is a
genuine law or whether it only looks like a law but de
pends, in fact, upon certain special initial conditions
prevailing in our region of the universe" (53:^33). From
this statement it appears that a scientific law is not
"genuine" unless it is universal in the widest sense. If
this is Popper's meaning, then he is stipulating a defini
tion of scientific law that seems unfounded, for its con
sequence is that there are no "genuine" scientific laws
that are knowable (i.e., every use of the term "scientific
law" would be unwarranted by the mere fact that we do not
57
know the "initial conditions" prevailing in every region of
the universe). Nevertheless, scientists do speak of "sci
entific laws" when they are referring to certain regulari
ties on the earth or in the solar system, and this usage
appears to be justified by the fact that there are some
regularities whose relative simplicity makes them ascer
tainable .
Moreover, there are at least three kinds of scien
tific laws: (l) laws which refer to "constitutive" proper
ties, such as shape, (2) laws which refer to dispositional
properties, such as solubility, and (3) laws which are
idealizations rather than abridged descriptions of obser
vable facts, e.g., Boyle's law (that the volume of gas at
constant temperature is inversely proportional to its pres-
sure--a law no actual gas obeys exactly). Now generally,
good laws represent actual relations, and that is why they
have predictive power (3:421-24). With Popper's denial of
our ability to grasp actual relations, it is difficult to
see how he can account for the distinction between laws
which have predictive power and those that do not.
It is also problematic whether Popper's hypothet-
ico-deductive system can accept "idealization laws" as sci
entific laws, for that they are approximations and useful
ones, too, does not mitigate the fact that the gap between
the ideal and the actual appears to be a negative instance,
thereby serving to "falsify" the "law" or theory. However,
if its approximative character is not allowed to falsify a
theory like Boyle's law, then Popper faces the problem of
determining when approximation is to be allowed and how
much of a gap is to be tolerated. Furthermore, if para
digms of scientific laws are taken to be high-level theo
ries of categories (2) and (3) above, or if they must meet
the standard of unrestricted universality, then it is easy
to see the difficulty in establishing such laws "beyond all
reasonable doubt." But at least some of the laws of cate
gory (l), when they refer to observed shapes or relatively
simple structural relations, need not be subject to reason
able doubt. The same may hold true for some low-level the
ories or laws of categories (2) and (3).
What, it may be asked, is*the meaning of "reason
able," if one denies that it can be known "beyond all rea
sonable doubt" that the earth is (roughly) round, or that
the earth is larger than the moon, or that our planetary
system is heliocentric, or that all men need oxygen in or
der to live? Certainly there are good grounds for con
trasting these with theories that have a less evidential
basis, e.g., theories in quantum physics which are problem
atic, and theories of unlimited universal import which,
therefore, may well be open to reasonable doubt.
59
Essences and Natural Necessity
Another problem which Popper's theory encounters is
that of "natural necessity" in the world. Is it explicable
in terms of our minds "imposing upon nature laws which it
freely invents"? For example, the process of physical con
ception and maturation in human beings appears to be a
clear case of natural necessity. Of course this process
can be stultified under certain conditions, but it cannot
be reversed, i.e., once one is an adult, there cannot be a
precise or literal reversal of every temporal step in his
maturation that will return him to his physical conception.
The linguistic or symbolic formulation of this law, there
fore, expresses an intrinsic regularity in nature. What
ever this law may be termed, it certainly can. be described
as being "beyond all reasonable doubt."
It is unfortunate, however, that Popper nowhere
appears to state clearly what he means by a "reasonable
doubt." He seems to assume that as long as a theory or
"law" is conceivably falsifiable, it should not be con
sidered "beyond all reasonable doubt." Yet, he does speak
of degrees of "logical probability" and of degrees of test
ing. Hence, if a given theory has survived more rigorous
tests than has a rival theory, it would appear to be more
reasonable to provisionally hold the former. But since no
number of tests can guarantee the future impossibility of
60
a test being found to falsify the theory, no theory is "be
yond reasonable doubt." Hence, Popper appears to mean, by
this locution, the provisional character of theories.
It seems better, however, to understand "reasonable
doubt" as that incertitude (psychological state) which is
based upon relevant reasons. Thus, a clear case of reason
able doubt would be the incertitude of one regarding a
prior claim that a transplanted heart will not be rejected
by its recipient, for in many such instances of heart
transplantation rejection has occurred. Here is a reason
upon which the doubt is based. On the other hand, clear
cases of propositions "beyond all reasonable doubt", are the
assertions that "all men need nutrition in order to live,"
or that "the sun radiates light and heat," or that "the
tone 1G1 is not green." If there is no reason for doubting
these propositions, then it is unreasonable to say they
cannot be known "beyond all reasonable doubt." But this
kind of "certainty" is precisely what Popper rejects. It
must be stressed that all scientific laws, among other
knowledge-claims about the world, are, for Popper, nothing
more than conjectures or imaginative hypotheses which have
survived severe testing but which are also permanently open
to refutation.
Not only "natural necessity" but all cases of simi
larity and repetition are given this interpretation by
6l
Popper:
Generally, similarity, and with It repetition, always
presuppose the adoption of a_ point of view: ....
But if similarity and repetition presuppose the adop
tion of a point of view, or an interest, or an expecta
tion, it is logically necessary that points of view,
or interests, or expectations, are logically prior, as
well as temporally (or causally or psychologically)
prior, to repetition. But this result destroys both
the doctrines of the logical and of the temporal pri
macy of repetitions. (53:421-22)
Nevertheless, this interpretation itself presuppos
es a point of view which, to my knowledge, Popper nowhere
"justifies." That his argument presupposes a certain
"point of view" seems to follow from Popper's statement
that a "point of view" is "a system of expectations, antic
ipations, assumptions, or interests” (52:44-45) and that
a point of view is always prior to any repetition, and not
the result of repetition. His contention that similarity
and repetition are always "known" from a point of view is
itself a claim to know similarity and repetition.--which,
as he says, presupposes a "point of view." Therefore, if
having a "point of view" prevents us from knowing that any
case of similarity and repetition is objective, then we
cannot know that this case (i.e., what Popper claims) is
objective. This seems to be the consequence of Popper's
assertion that it is a "naive idea" that any events are
similar, and that we react to events "by interpreting them
as being similar" (52:45). Contrary to Hume who argued
that anticipations and expectations result from repeti
62
tions, Popper contends that repetitions result from antic
ipations and expectations.
Now this claim of Popper's itself presupposes doth
that there is "natural necessity" and that it can be known.
He seems certain that we necessarily arrive at notions of
similarity and repetition according to an invariable bio
logical and psychological pattern: "Instead of explaining
our propensity to expect regularities as the result of rep
etition, I proposed to explain repetition-for-us as the re
sult of our propensity to expect regularities and to search
for them" (52:46). Popper also asserts that it is naive
"to look upon repetition as something ultimate, or given,"
since "anything can be said to be a 'repetition' of any
thing, if only we adopt the appropriate point of view"
(53:422). This seems to undermine Popper's own theory of
falsifiability, however, for if "repetition" is not "given"
in any way, then there can be no intersubjective testing
of a hypothesis, and no particular test of a hypothesis
can be checked by subsequent repetitions (which are gener
ally the principal ways of determining the correctness of
a single test and of counteracting the possibility that a
test produced only an exception and not a counter example).
And yet Popper does recognize that "non-reproduci-
ble single occurrences are of no significance to science,"
and a theory can be "falsified only if we discover a re-
63
producible effect which refutes the theory" (53:86). Nev
ertheless, It Is difficult to see how this is explicable
apart from the givenness of some similarity and repetition.
Popper's epistemology will not allow him to admit that
there is knowable similarity, or repetition, or natural
necessity in the world, however, for that would be to con
cede that there are knowable essences. Yet it seems that
some natural laws are descriptive formulae for the recur
rent behavior of a class of phenomena, and as such they
are important indices to the reality of essences.
Although Galileo challenged Aristotelian science in
many respects, yet both he and Aristotle agreed that the
aim of science was to furnish a true description of nature.
Neither one of them stopped short of "essentialism" by re
sorting to views like instrumentalism and conjecturalism.
But the spectacular revolutions in the various sciences
have called their "optimistic view" into question in modern
times. And, in the twentieth century, the world of physics
--and especially subatomic physics--has represented certain
features of its domain as being so paradoxical that the
postulate of a real world behind this "contradictory phe
nomenal world" has become extremely attractive.
While not denying the incompatibility of various
theories, I believe that Popper, among others, has allowed
the perplexities of quantum physics to become virtually
paradigmatic of the nature of science. Along with revolu
tions in the history of science, however, it is necessary
to reckon with stability. To accentuate the former and
slight the latter is an egregious instance of special
pleading. In general, no doubt, the diachronic stability
in science is less interesting than its flux. But this
makes it no less important or real--particularly if one is
constructing an epistemology. What is needed is a coherent
account that avoids the shortcomings of those views that
excessively emphasize one end of the spectrum--the exagger
ation of objectivity on the basis of stability, on. the one
hand, and the exaggeration of subjectivity on the basis of
flux, on the other hand. Just as Popper seems to commit
the latter error, so he falls into the correlative fallacy
of enthroning negative criticism and rejecting positive
evidence. Even if it is true that flux and falsification
dominate the history and methodology of science, the mere
recognition of an integral (even if minor) role of their
contraries, stability and substantiation, will provide
grounds for a different view of scientific knowledge.
Popper's interpretation of science is undermined by its
failure to account for stability and permanence in the
midst of flux and change. Kant's interpretation of the
nature and grounds of science leaned too heavily on the
supposed finality of Newtonian physics. Conversely,
65
Popper's epistemology hinges on the skewing of science in
terms of its tentativeness.
Ancient Greek philosophy (Aristotle, in particular)
found "essences" in the objective realm of the real world.
They were not conceived as mind-created or mind-dependent.
Rather they constituted the objective structure of a tran
scendent (i.e., extra-subjective) reality. In modern times,
however, Kant argued that "essences" (i.e., "categories,"
not Aristotelian essences) are subjective, "imposed" on the
world according to the necessary formal structure of the
human mind. This immanent structure gives man the condi
tions for the possibility of universal and necessary "know
ledge." Following Hume, Kant had assumed that what is pri-
mordially given in sense experience was without structure
and meaning. Although experience is necessary for the
realization of knowledge, all order comes from the a priori
equipment of the human knower.
Popper rejects both of these views of "essence" and
the "laws of nature." The Greek view and Kant's view, in
their own ways, made certain knowledge possible, but
Popper's assumptions about the nature of essences and the
nature of the mind's functioning limit man to doxa (i.e.,
"opinion"). He disagrees with the Greeks by denying our
capacity to know objective structures of transcendent real
ity, and he disagrees with Kant by denying a formal a pri
ori structure in human thought which necessitates the rigid
66
"imposition" of laws., structures, or essences upon nature.
While agreeing with Kant that the human mind is "interpre-
tively active"--or, in Popper's words, "we are not passive
receptors of sense data, hut their active digestors" (52:
95)--in the acquisition of "knowledge," Popper differs with
him by contending that our intellect is free in its capaci
ty to contrive theories and "impose" them upon the world:
We do not stumble upon our experiences, nor do we let
them flow over us like a stream. Rather, we have to be
active: we have to 1 make1 our experiences. It is we
who always formulate the questions to be put to nature;
it is we who try again and again to put these questions
so as to elicit a clear-cut 'yes' or 'no'. . . . And in
the end, it is again we who give the answer; it is we
ourselves who, after severe scrutiny, decide upon the
answer to the question which we put to nature--after
protracted and earnest attempts to elicit from her an
unequivocal 'no'. (53:280)
Popper does not deny the possibility of objective proper
ties and relations in the world, but he does deny that we
can have any knowledge of them or be certain that we know
them. On first consideration, it may seem that Popper
"must" admit that we know some objective properties and
relations, because such knowledge would appear to be re
quired by the very possibility of an observation statement
being used to refute a hypothesis. But that this is not
the case is evident for two reasons. First, Popper says
that we become aware of "facts" only by observation, but
"this awareness, this knowledge of ours, does not justify
or establish the truth of any statement" (53:98). Hence,
observation does not provide us with any basis for know-
67
ledge, or knowably true statements, about the world or ob
jective properties and relations.
Secondly, Popper emphatically says that "all . . .
observations are interpretations in the light of theories"
(52:38) and that "theories are our own inventions, our own
ideas; they are not forced upon us, but are our self-made
instruments of thought" (52:117). Nevertheless, he says
that some of our theories can "clash with reality" (52:
117). This is puzzling, for if we can never know "reality"
and if our observations are "interpretations" in the light
of our own invented ideas, then how is it that there can be
any clash between our theories and reality? From what he
has said, it would seem that there can only be a clash be
tween different kinds or levels of theories. Since, ac
cording to Popper, we evidently have no way to get "behind"
observations, or to dispense with them altogether in our
contact with the world, "reality" is not a term that can
designate any knowable thing. Our "knowledge," then, is
restricted to our own theories, whose construction and
corroboration are "affected" in some vague and unspecifi-
able way by "reality"--vague and unspecifiable because
"reality" cannot be known at all. In a way reminiscent of
Kant, Popper seems to say that all synthetic wholes and all
cases of similarity are products of the mind's activity.
In fact, Popper states his agreement with Hume, who argued
that "observation could be only of singular (or particular)
instances, so that all theoretical knowledge was uncertain"
(52:94). While Popper does not appear to explicitly de
velop a theory of sense experience, his agreements with
Hume and Kant suggest that he believes we can never per
ceive any real connections between distinct sensations, and
that there is no similarity or repetition to be found among
them, but that "we actively try to impose regularities upon
the world" (52:46).
Popper's Principal Assumptions
Here, then, are the key assumptions which underlie
Popper's denial of the discoverability of truth and the
attainability of knowledge. (l) What is primordially given
in sense experience is devoid of structure, i.e., it pre
sents no "essences" (objective properties and relations).
When Popper says, for example, that "our attempts to force
interpretations upon the world" are "logically prior to the
observation of similarities" (52:46), he seems to tacitly
assume that what is originally given in sense experience
must be disparate sensations. For if "similarity" results
from the mind's interpretation of sensations, then what is
primordially given in sense experience can hardly be char
acterized by similarity or repetition. Moreover, if, as
Popper says, "all observation involves the recognition of
similarities (or dissimilarities)" (52:48), and if simi-
69
larity and. repetition always result from our interpretation
and "imposition of regularities on the world," then it is
difficult to see how any observation can yield awareness of
any objective regularities in the world. (2) Hence, from
an unstructured agglomeration one cannot derive any laws
(52:91)j and in it one cannot discover any truth, for
Popper says that we can only search for truth, but we can
never be sure we have found it (52:56).
(3) Nevertheless, there are numerous instances of
discovered error in our interpretations of nature. It is
possible to refute truth-claims (52:28). (4) Such refuta
tions would be impossible if the human mind did not have
the capacity to theorize about human experience. Since
there are only two poles in cognitive consciousness of na
ture, and since one pole has no structures of its own, it
must be the mind that "imposes" structures: "Our intellect
does not draw its laws from nature, but tries--with varying
degrees of success--to impose upon nature laws which it
freely invents" (52:191). (5) But since these structures
are "imposed" by the subject, they may be erroneous, i.e.,
they may be inaccurate interpretations of a given state of
affairs. Thus, every theory must be subjected to severe
tests to determine if it is false (52:51-53* 192; 53:78-92,
251-81). (6) But every refutation is only provisional,
since it is possible for any refutation to be based upon a
70
misinterpretation. Hence, every refutation is open to re
futation ad infinitum (52:51* 114-15). (7) Therefore,
truth is undiscoverable and knowledge is unattainable.
Popper writes:
The old scientific ideal of episteme--of absolutely
certain, demonstrable knowledge--has proved to be an
idol. The demand for scientific objectivity makes it
inevitable that every scientific statement must remain
tentative for ever. It may indeed be corroborated, but
every corroboration is relative to other statements
which, again, are tentative. Only in our subjective
experiences of conviction, in our subjective faith, can
we be 'absolutely certain'. (53:280)
The most we can express, then, is not certainty but opin
ion, for we can never be sure whether any given view will
escape refutation throughout the indefinite future. Truth
can only be approximated by the continuous elimination of
error. But truth itself is eternally elusive.
To these steps in Popper's argument Bartley's fur
ther contentions might be added (although it is doubtful
that Popper would subscribe to them without qualification).
(8) Popper's method of falsifiability can be extended to
the entire domain of alleged knowledge, and on this basis
a radically new epistemology can be constructed. (9) All
other epistemologies have been unable to overcome fideism
and skepticism. Their own starting points were unjusti
fied, for there was no way of making their rational cri
teria self-reflexive without begging the question.
(10) Since an alternative to "comprehensive ration
alism" and "critical rationalism" has been provided by
71
Popper's nonjustificationalism (which avoids both arbitrary
"ultimate presuppositions" and a vicious infinite regress),
it follows that the only rational position is "comprehen
sively critical rationalism." (ll) Now with the possibil
ity of a choice between inconsistent and arbitrary philo
sophies on the one hand,, and a consistent and self-reflex
ive rationalism on the other, those who opt for the former
are irrationalists, and their diverse positions are irra
tional "retreats to commitment." Every view other than
"comprehensively critical rationalism" is a form of irra-
tionalism--this includes all religious, theological, ideo
logical, and philosophical positions. Nonjustificational-
ism alone is thoroughly consistent (rational), and that is
because it is the only position that admits the possibility
of its own refutation. In this way one can avoid commit
ment to any view.
It has already been shown that Bartley's view is
untenable because of its fundamental self-contradictions.
Bartley assumed both that the method of nonjustificational-
ism in science was correct and that it could be fashioned
into a comprehensive epistemology. Yet even in science the
method is not without its problems. In fact, they are so
severe that they serve to undermine the claim that it is
the sole scientific procedure, because human knowing,
whether "scientific" or "nonscientific," cannot be made to
fit the model of a pure nonjustificationalism. This is
seen in Popper's unsuccessful attempts to exclude every
iota of alien methodologies from his position. That in it
there are surreptitious, if not avowed, elements of con
ventionalism and verificationalism can be seen in his own
admissions. For example, he says that "basic statements
are accepted as the result of a decision or agreement; and
to that extent they are conventions" (italics mine, 53:
106). Again, he says that a theory should not be refuted
too soon, and "it should be successful in some of its new
predictions. ... I admit that there may be a whiff of
verification here; but this seems to me a case where we
have to put up with it" (italics mine, 52:248).
The. Consequences of Counter "Assumptions"
Not only is nonjustificationalism impure, however,
it is also based upon dubious assumptions. As has been
observed, the most conspicuous of these is Popper's adop
tion of a quasi-Humean notion of sense experience (52:94)
and a quasi-Kantian distinction between phenomenal and
noumenal realms--!.e., in Popper's terminology, "reality"
(noumena) is unknowable because of interference by our
interpretations (phenomena). The importance of these as
sumptions can easily be seen by considering the conse
quences of their negation in contradistinction to the
aforementioned seven-point outline of Popper's argument.
73
(l) Sense experience primordially presents some
differentiated and organized structures, i.e., it presents
"essences." (2) From a structured manifold one can derive
some laws, and by examining it one can discover some
truths. (3) There are instances of both truth and error in
our interpretations of nature. It is possible to justify
some truths and eliminate some errors. (4) Such justifica
tion and refutation would be impossible if nature did not
contain some knowable structure and if the human mind were
unable to recognize it and theorize about human experience.
Since there are two poles in cognitive consciousness of
nature, and since both poles have structures of their own,
any given interpretation may or may not be a faithful ex
pression of an objective structure. (5) Since these inter
pretations may be unfaithful to the actual state of affairs,
they should be subjected to careful examination and severe
testing to determine the adequacy or inadequacy of evidence
for or against them. (6) In some cases the evidence may be
adequate; in others it may not be. In the former instances
justification is conclusive; in the latter it is not.
Moreover, the latter cases are open to revision and refu
tation on the basis of continued testing. (7) Therefore,
some truths are discoverable and some knowledge is attain
able. We can express both certainty and opinion, for if a
proposition is true at all, it is timelessly true. Some
truths can be apprehended, while other propositions, theo
ries, and "laws" may be conjectural, with little or no cor
roboration.
To these contentions contrary to Popper, it will be
helpful to add further assertions contrary to Bartley.
(8) Since Popper’s method of falsifiability is defective in
the field of science, it cannot be extended to the entire
domain of alleged knowledge. Hence, no new epistemology
can be constructed on the basis of it. (9) Whether all
other epistemologies have been unable to overcome fideism
and skepticism is open to question. And whether it is
possible to escape unjustified starting points depends on
the meaning of "justification." (10) Since "comprehensive
ly critical rationalism" is demonstrably inconsistent, an
alternative to it must be sought. (ll) Since Bartley's
position is self-contradictory, it is irrational. And
since it is logically impossible for a view to refute it
self, the meaning of rationality must be stated in such a
way as to avoid the demand of possible self-refutation.
An Alternative View of Rationality
With a different understanding of rationality,
then, it may be that some philosophical or theological po
sition can be established as rational, or it may be found
that one is more rational than another. The important
question, then, is : What is rationality, and what canons
75
does it provide for the assessment of various philosophical
and theological positions? In order to lay a partial
foundation for answering this question (to which I shall
return later in this work)., I shall briefly indicate some
important differences between Popper's position and mine.
First., as I have already indicated, there is no
necessity to understand "essences" as something occult or
hidden, i.e., as ghostly entities residing in or behind
things as they appear. I believe it can be cogently argued
that we are in direct contact with the world, due to the
intentionality of consciousness. If this is true, then it
would seem to provide an answer to the following important,
but apparently unasked, question in Popper's position: How
is it that, of the incalculable number of possible hypoth
eses, we can make any conjectures at all which are relevant
to a problem under consideration, testable by observations
or experience, and fruitful in understanding and coping
with the world in which we live?
By "intentionality" I mean what Husserl calls the
property of being a "consciousness of something" (30:109).
Certain acts have the characteristic of being intentional,
i.e., they are directed toward something. Such acts have
five defining elements: (l) the one who is performing the
act, (2) the fact of what he is doing (e.g., he is think
ing about something, or desiring something, etc.), (3) the
intentional object of the act, i.e., what the act is about,
(4) the way in which the intentional object is the object
of the act, and (5) the "coherence" of the act, i.e., the
characteristics which make it possible for an assertion
about its content to be true or false (by way of contrast,
"incoherence" can be illustrated by the assertion that "The
number five is blue and fragrant," or by saying today, "I
predict that World War I will begin in 1914"). Thus, if we
take the intentional act referred to by the statement,
"John wants to put the chair in his room," (l) John is the
one performing the act, (2) his wanting is what he is
doing, (3) the chair is the intentional object, (4) to put
the chair in his room is the manner in which the intention
al object is the object of the act, and (5) the coherence
of the act is seen in the fact that there is a way of find
ing out whether the statement is true or false.
By his insight into the nature of intentionality,
Husserl was able to expose the error characteristic of
those philosophies (such as Neo-Kantianism and "Psycho
logism") which interpret knowing in terms analogous to the
process of eating. Sartre aptly designates the latter
epistemological view as "an alimentary philosophy," and he
paints the following picture of its chief contention:
. . .the Mental Spider draws things into its web,
covers them with a pale spittle, and slowly swallows
them, ^turning them into its own substance. What is a
table, a stone, or a house? It is a certain assemblage
77
of "contents of consciousness," an arrangement of
those contents. . . . The world skeleton is eaten
away by these diligent diastases: assimilation,
unification, identification. (6l:l)
In contrast to this "digestive theory," consciousness is
necessarily characterized by self-transcendence, i.e.,
openness toward the world and toward things other than it
self. Consciousness is always consciousness of_ something.
The object of which one is conscious is not reducible to
consciousness any more than consciousness is reducible to
the thing.
Intentionality not only characterizes mental acts
such as judgments, meanings, and beliefs, but also valua
tions, emotions, volitions, and sentiments--for they, too,
intend objects. For example, I may not only have knowledge
of a mountain, I may also admire it, fear it, or detest it.
Hence, intentionality is a relation between a "meaning"
act and its object, i.e., that which is "meant." I may
think of or about any number or variety of things. I may
also love or hate things. And it is not necessary that
these things "exist." This means that one may have an in
tentional experience whether its apprehended object exists
(e.g., an actual mountain), is imaginary (e.g., a unicorn),
or even absurd (e.g., a round square). Husserl writes:
Cognitive mental processes (and this belongs to their
essence) have an intentio, they refer to something,
they are related in this or that way to an object.
This activity of relating itself to an object belongs
to them even if the object itself does not. And what
78
is objective can appear, can have a certain kind of
givenness in appearance, even though it is at the same
time neither genuinely (reell) within the cognitive
phenomenon, nor does it exist in any other way as a
cogitatio. (29: ^-3)
Intentionality, then, is the very essence of consciousness
itself, and one can direct his reflective attention to his
conscious life to explicitate and describe its content,
i.e., its perceiving, valuing, asserting, imagining, and
willing. While there are other characteristics of inten
tionality which Husserl explicated, I believe that what is
now before us is sufficient for my purpose.
The awareness and description of what is "seen"
makes possible the radical grounding of philosophy without
the distorting effects of prejudices. Husserl says:
Phenomenology is directed to the "sources of cognition,"
to general origins which can be "seen," to general ab
solute data which present the universal basic criteria
in terms of which all meaning, and also the correct
ness, of confused thinking is to be evaluated, and by
which all the riddles which have to do with the objec
tivity of cognition are to be solved. (29:^)
Not only can the modes of intending be described (noetic
description) but also the intentional object as such can be
described (noematic description). In noetic description
one describes the perceiving of a given mountain, for ex
ample, or the valuing of it, etc. In noematic description
one describes the mountain as perceived, or the mountain as
valued, etc. The noema is composed entirely of essences
which are timelessly what they are. These essences stand
79
in necessary or a priori relations with each other. Hence,,
"the question of essence" asks what the "perceived as such"
is, and what essential phases it harbors in itself in its
capacity as noema (30:240). The perceiving of the mountain
(i.e., a noesis) is related to the mountain as perceived
(i.e., a noema). The latter's structure and properties--
e.g., its specific "shape," "size," and "color"--correspond
to "perception." These are essences or Eideia, and if the
question concerning essences is to be answered, we must
"wait, in pure surrender, on what is essentially given. We
can then describe 'that which appears as such' faithfully
and in the light of perfect self-evidence" (30:240).
While there is a distinction to be drawn between
"acts" and "objects," acts of consciousness can also be
made the objects of reflection (i.e., of further reflective
acts). The act of perceiving or believing may be examined
just as the things perceived or believed may be examined.
In discovering their universal aspect, one becomes aware
of their essential nature, i.e., the properties that are
the necessary differentiae of a given "object"--e.g., one
can. remember without regretting, but he cannot regret some
thing without remembering it.
Nevertheless, phenomenology does not lose the world
by describing phenomena. It is retained both with respect
to particular realities and the world itself as the neces
sary background of our whole natural life. Now the percep-
tual noema (e.g., the perceived mountain appearing from a
certain side or aspect, i.e., in a one-sided manner of ad-
umbrational presentation) is presented before our mind and
we are in contact with it. If this is an accurate account,
then "representational epistemology" (e.g., the view advo
cated by Descartes and Locke, among others, which says that
we do not know the actual, external world directly, but
only by means of ideas or Images which represent objects in
it) is a faulty analysis of how we know perceptual objects.
Popper and Bartley seem to fall Into an error that
is similar to representationalism. They theorize with a
confused notion of how we "know, " because they fail to dis
tinguish between perception and those acts of consciousness
in which there is "the presentation of a_ symbol in. the form
of an image or meaning" (30:123). Husserl observes that
while with the latter types of presentation we are con
scious that what we are directly aware of copies something
else or indicates its meaning, in perception we are not
directed toward something else but are consciously aware of
the spatial thing as it is given in its bodily form. "We
are not given an image or a sign in its place. We must not
substitute the consciousness of a sign, or an image for a
perception" (30:123). There is an unbridgeable and essen
tial difference between these modes of presentation.
Popper's assumption that perception is never veridical be
cause it is theory-impregnated evidences a confusion be
tween our awareness of perceptual objects and of illusion
or imagination. Just as the image-theory and sign-theory
have proved misleading, so does Popper's theory. To assume
that we can never know sensory objects without a supposed
distortion that stems from "interpretation" is to ignore
the unbridgeable and essential difference between "spon
taneously theorizing" or "freely conjecturing" on the one
hand., and perception of things in their given "bodily form"
on the other hand.
In empirical consciousness, which gives us sensory
objects, "the individual object can 'appear,' one may be
aware of it as apprehended, but without .being spontaneously
'busied with' it at all" (30:83). If this is the case,
then there is a nucleus of irreformable truth in empirical
knowledge, even though it has its reformable aspects.
Moreover, if there were no veridical perception, it would
be impossible to argue that all perception involves a dis
tortion of what is perceived. It must be admitted that at
least some of the time error can be detected and corrected
by reflection. For example, we say that a stick that is
partly submerged in water "looks bent." But by examining
it carefully (e.g., taking it out of the water, touching
it, etc.) we find that it is straight (or that it actually
is bent). H. A. Prichard points out a basic distinction:
82
However much I see an object distorted through bad
glass or in reflections, it is the thing which I see.
There is strictly speaking no "look" or "image" of the
thing; there is only the thing with its characteristics
distorted. The word "look" or "appearance" properly
refers to a judgment, and that not a judgment I make
now, but one which I should make under certain condi
tions. (57:229)
Since Popper assumes that all observation is theory-laden
and that we cannot therefore distinguish between veridical
and non-veridical perceptions, it follows that observation
statements are always open to refutation. But with the
recognition that this is an unwarranted and self-defeating
assumption, and that there is a viable alternative in the
phenomenological view of perception, arrant nonjustifica
tionalism (even in science) is unacceptable.
Another assumption of Popper's which serves to
vitiate his nonjustificationalism is his interpretation of
biological and psychological "expectations" as species of
presuppositions (52:45ff.). It is unconvincing, however,
to maintain that the expectations of an infant fix and dis
tort its "awareness" of the world. There is no evidence
that infants are born into the world with a set of catego
ries or concepts (or "theories") which they "impose" on the
world. It seems, rather, that infants are initially con
fronted with innumerable sensations and objects which they
are unable to "interpret" within the framework of one or
more "theories."
Moreover, expectations can be reflected upon and
modified. There is no reason to think that they intrude to
make all perception defective. If this is the case, then
induction has a basis both in initial awareness of the
world and in phenomenological reflection upon the objects
intended by consciousness. Hence, Popper's dictum that
"theory precedes observation" does not always hold. Obser
vation, to some extent, precedes theory chronologically,
and it can be made to do so logically in some cases by
phenomenological description and reflection.
Furthermore, Popper never makes clear exactly what
it could mean for one to "impose" theories, or structures,
or regularities on the world. Can it be maintained seri
ously that I make one color darker than another, or one
tone higher than another, or that I put my heart in my
chest rather than in my leg, or that I make a flower red or
fragrant, or that I cause the succession of day and night,
etc.? Can it be meaningfully argued that there is no such
thing, for example, as a flock of birds, but that we "men
tally create" every flock by attributing to isolated birds
a group-character that they do not already have?
With the phenomenological examination of cognition,
one can remove Popper's assumptions about sense experience,
and establish the fact that "group characters" are often
given and may be observed directly. We often perceive or
ganized patterns (Gestalten) and not merely separate parts
84
that are simply added to one another. Moreover, there are
certain relationships that are lost if what is experienced
or intended is broken down into its parts. The perceived
relationships are one kind of meaning or essence, and we
can be aware of them as objectively given. And if we can
be aware of them, then at least some observation statements
and scientific laws can be established "beyond all reason
able doubt." Where qualities and perspicuous or simple
relations are brought to self-givenness, there is the basis
for some knowledge. Among even the simplest sense-fields
there are structures, e.g., a white spot against a homoge
neous, black background. There can be some true proposi
tions by the very fact that it can be known that that which
is intended or thought of is as it is intended or thought
to be.
Popper seems to have an. implicit theory of experi
ence, then, which prevents him from granting that truth is
discoverable. Apparently, certain reductionistic preju
dices have led Popper to reject the view that "essences"
can be known. He assumes that no distinction between
theory and data can ever be drawn, with the result that he
forever removes the latter from our knowledge. But this
rejection of "the original right of all data" (30:88) has
a boomerang effect on Popper's own epistemology. For with
out the recognition of essences as objective properties and
85
relations, there could be no theorizing and no testing of
theories, and consequently, no science. Accordingly,
Husserl says:
Our critics in truth see, and so to speak continuously
see, "ideas," "essences"--make use of them in thought,
formulate judgments concerning essences--only from
their epistemological "standpoints" they explain the
same away. Self-evident data are patient, they let
theories chatter about them, but remain what they are.
It is the business of theories to conform to the data,
and the business of theories of knowledge to discrimi
nate the fundamental types, and to describe them in
accordance with their distinctive nature. (30:80-81)
Perhaps if Popper had sought to bring his view of theoriz
ing into conformity with the data of theorizing, he would
have seen that it is impossible to construct theories with
out the intuition or perception of essences. And without
the phenomenological description of theorizing, one will
only theorize about theorizing ad infinitum.
Conclusion
Popper's thesis is misconceived and self-defeating;
for without some discoverable truths one cannot warrantably
assert that truth is undiscoverable. Most truths may not
be manifest (i.e., known by "seeing" alone), but many
truths are discoverable (i.e., known by reflection and
proof). And epistemology, in which Popper has an avowed
and intense interest, aims at discovering the primary
truths on which our knowledge is based. By "primary
truths" I mean those truths which are irreducible and
immediate, standing on their own as the basis of derived
truths. As Albert Dondeyne has said: "Every philosophy,
insofar as it is regarded as a work of criticism, is an
incessant effort to discover primary truths beyond the de
rived truths with which it is presented" (19:20).
By way of summary, it can be said that when Popper
asserts that theory always precedes observation, he is
making a claim that is arbitrary and indefensible. Its
arbitrariness is seen in the fact that he cannot justify it
--certainly not on nonjustificational grounds. Moreover,
he seems to suggest that "the given" is amenable to every
kind of theoretical determination because it has no know-
able determination of its own. But, as has been indicated,
descriptive faithfulness to what is primordially given does
not support the assumption that the given is formless or
completely unstructured. An indeterminate "reality" that
is amenable to every kind of form can hardly clash with
theories which we allegedly "impose" upon it. Error in the
empirical domain is only understandable in terms of the
recalcitrance of the given. The mind is not allowed, in
the interest of truth (or even of error), to excogitate and
"impose" theories with complete license, for the given must
already have a determinate structure that is independent of
the mind if "truth" and "error" are to have a sense. Of
course the very notion of the given is unacceptable to a
theory that views knowledge as nothing more than a set of
conjectures spontaneously or freely created by the mind.
87
Finally, Popper's view is indefensible because it
leads to a vicious infinite regress as soon as one asks
how it is known. The answer to this question is precluded
by Popper's conjectural theoreticism. In fact, his falli-
bilistic "hypotheticism” leads ineluctably to a fideistic
view of logic and rationality. And to this latter subject
I shall turn in the next chapter.*
*After a careful search of the leading philosophi
cal journals, I have found that while much of the litera
ture on Popper discusses his views on the philosophy of
science, there seems to be little that explicitly deals
with his epistemological views or that criticizes his vir
tual identification of epistemology with the philosophy of
science. Paul Bernays's article (which I have listed in
my bibliography) analyzes Popper's epistemology to some ex
tent, but he does so from the standpoint of basic concur-
rence--for he admits that he is "in full agreement with
Popper regarding the main tendencies" (l4:44). Several of
the critical reviews of Popper's books are also listed in
the bibliography.
CHAPTER III
WHY ARE LOGIC AND ARITHMETIC APPLICABLE TO REALITY?
In the twentieth century the conventionalist view
of logic has become widespread. Naturally, it has given
rise to a correlative view of the nature of rationality and
reason. And this in turn has led to important consequences
for epistemology and ontology. All of these features are
reflected in Popper's perspective. To understand his skep
tical conclusions concerning epistemology and ontology, it
is necessary to trace their roots in his logical conven
tionalism. This is made clear in his essay, "Why are the
Calculi of Logic and Arithmetic Applicable to Reality?"
(52:201-14).
Rules of Inference
After distinguishing rules of inference from formu
lae of logical calculi, Popper asks: "Why are rules of in
ference applicable to reality?" (52:204). He first con
siders Gilbert Ryle's thesis that the rules of Inference
are only rules of procedure. They are well-formulated in
structions for the procedure of drawing inferences. Apply
ing them means acting in accordance with them. It makes no
sense, according to Ryle, to ask if the rules of logic fit
the things and facts of the world. The question itself is
as meaningless as asking if the highway code or the rules
of chess fit the facts of the world. Such rules of per
formance "apply to reality" only in the sense that people
abide by them, and they do not "apply" when people fail to
observe them (60:20-29).
Although Popper describes Ryle’s analysis as "true
and important," he is dissatisfied with its reduction of
the problem to a "pseudo-problem." He argues that the
question might be interpreted to mean, "Why are the rules
of logic good, or useful, or helpful rules of procedure?"
(52:205). He is convinced that the man who acts or thinks
according to them does so because he finds them useful in
dealing with reality. In drawing inferences from true
premises according to the rules of logic, one finds that he
is able to obtain reliable and possibly valuable informa
tion.
This leads,- then, to another important question:
"What is the explanation of the fact that the logical rules
of inference always lead to true conclusions, provided the
premises are true?" (52:206). Popper contends that the
answer to this question is readily discernible. Logical
rules of inference are formalizations of intuitive proce
dures we use in describing facts. That is, beginning with
90
some original information we are able to obtain some kind
of secondary information merely by rational reflection.
But the formulation of this procedure into a rule of infer
ence must be subjected to testing to determine if it is
"valid." It must be tested by a diligent search for a
counter example, i.e., an instance in which the observance
of a proposed rule will allow us to obtain a false conclu
sion from true premises. If no counter example exists to
the observed rule, then the inference is called valid. And
the reason a valid rule of inference is useful is that we
can depend upon it as a technique that "leads us from true
descriptions of facts to true descriptions of facts" ( 52 : .
206). And since a true description can be said to fit the
facts, it is appropriate to interpret the application of
rules of inference to facts in this sense.
Limitations of Popper's Exposition
While this part of Popper's exposition is immedi
ately before us, let us consider some of its limitations.
(l) It is important, first of all, to recognize that he has
been dealing only with rules of inference, and not with
principles of logic, such as the law of noncontradiction.
Neither has he said anything about the relationship between
rules of inference and the basic principles of logic. Ap
parently he fails to see in what sense insight into the
principles of logic and the conditions specified by a rule
for its application are relevant to the process of rule
justification. In fact., it appears that Popper has as his
test of rules of inference an appeal to "factual" counter
examples. This seems to make rules of inference analogous
to empirical generalizations or scientific hypotheses. As
Popper argues elsewhere, the principle that demarcates sci
entific from non-scientific statements is that only the
former are empirically testable. Scientific statements are
always open to the possibility of refutation. And refuta
tion is accomplished by the discovery of factual counter
examples which serve to falsify the hypothesis in question.
In other words, the hypothesis is accepted when it survives
severe attempts to refute it by the adduction of supposed
counter examples. A rule of inference, he seems to con
tend, is accepted on the same basis: "we call a rule of
inference 'valid' if, and only if, no counter example to
this rule exists" (52:206). Now in effect, Popper says
that we can never know that a rule of inference is justi
fied. This is implied by his claim that rules of inference
are permanently open to refutation by the possible emer
gence of a counter example. This is the so-called non-
justificational method applied to logic.
Attempting to establish a rule of inference on the
basis of empirical considerations, however, implies that
every rule of inference is either an empirical generaliza
tion or a conjecture. Now a rule is not an empirical gener
alization, for the latter is usually understood to be ex
pressible only in a declarative statement. Nevertheless,
Popper repeatedly calls a rule of inference a "statement."
For example, he says: "Rules of inference are always state
ments about statements" (52:203). Now Popper does distin
guish a formula of a calculus (e.g., "If p_ and if p then q,
then q") from a corresponding rule of inference (e.g.,
modus ponendo ponens: p, if p then q, q). But his example
of a rule is actually a statement that expresses a proposi
tion about other propositions. It says: "For any two pro
positions whatever, if one is true and if that one is true,
then the other one is true, then the other one is true."
This proposition is true, but it is not a rule. The cor
responding rule should read as follows: "Given that one
proposition is true, and that if that one is true then a
certain other one is also true, assert or write that other
one." Thus, we are told, in general terms, what we ought
to do under certain conditions if we wish to be sure of not
going from truth to falsity.
Popper argues that every rule of inference can be
eliminated by using a corresponding formula of a logical
calculus as a premise. He says that the one exception to
this is the "principle of inference," or modus ponendo
ponens. But, in general, there are no corresponding rules
of inference for non-hypothetical formulae (e.g., "p or
93
non-p"). On this basis, Popper says, it is possible to
infer all hypothetical formulae from non-hypothetical for
mulae by using the "principle of inference" (52:203-9-).
Apparently, Popper either ignores or does not recognize the
possibility of using other rules of inference (e.g., modus
tollens or modus ponendo tollens) rather than modus ponendo
ponens for this purpose. Moreover, he contends that it is
possible to avoid making use of the principle of substitu
tion (52:203). Now to whatever extent it is possible to
construct a scheme of reduction and derivation approximat
ing the simplicity at which Popper aims, the problem posed
by the applicability of even one rule of inference is un-
avoidable--as Popper admits. That some rules must be re
tained as directives indicates that all rules cannot be
treated as statements.
How is it possible for a rule of inference to be
invalidated by a counter example? If we take the simple
analogy of a traffic rule--e.g., "Always signal with your
arm raised upward before making a right turn"--a counter
example is not a case of someone failing to signal in this
way before making a right turn. That is a violation of the
rule. Only a counter directive of overriding authority can
"invalidate" a rule, e.g., "Always signal with your arm
pointed downward before making a right turn." But traffic
rules are conventional to a limited extent (i.e., they are
conditioned or limited by the general character of the ac
tivity which they govern), and to that extent they can be
changed. Rules of inference are not conventions, however,
and thus they are not subject to alteration or replacement.
Now even if a "rule" of inference were "interpret
ed," per impossible, as an empirical generalization, then
Popper's strictures against induction would have to be al
tered radically or scrapped completely. To call "rules" of
inference "empirical generalizations" means that they are
based upon empirical experience, which is contingent and
incomplete. But how does one then account for the univer
sality and necessity of such rules, and the fact, in
Popper's own words, "that the logical rules of inference
always lead to true conclusions, provided the premises are
true" (52:206)?
If, on the other hand, rules of inference are con
jectures, then they are "freely imposed by our minds."
However, if this were the case, how would it be possible to
explain the fact that we cannot freely substitute other
conjectural rules that work just as well? Why is it that
certain rules of inference are universally valid and others
are not? I doubt, however, that Popper intends for them to
be viewed as conjectures "imposed" on reality by the human
mind. Yet, as Popper has characterized rules of inference,
there seems to be no other alternative. If they are not
empirical generalizations (and this seems totally unaccept-
able to Popper), then they must be conjectures. If this Is
the case, we must look for a different understanding of
them that will enable us to transcend the dilemma in which
Popper is ensnared.
Furthermore, since Popper admits that deductive in
ference can provide information in the conclusion that does
not appear in the premises, he cannot interpret the argu
ment as a complex "analytic" statement (i.e., in the sense
that the conclusion adds no new knowledge but simply makes
explicit what is "contained" in the premises). This means
that a deductive argument must be viewed as "synthetic,"
for when it is valid and proceeds from true premises,
Popper contends that it yields "reliable (and possibly val
uable) indirect information" in the conclusion. But he in
terprets all synthetic propositions as a posteriori, i.e.,
although not necessarily derivable from experience, yet de
pendent upon experience as the arena in which they must, if
they are not to be scuttled, continue to prove their mettle
by escaping counter instances. It is clear that there is
no logical inconsistency in asserting the negation of a
"synthetic a_ posteriori proposition." But--and this is
crucial--inconsistency is inescapable when the premises of
a valid deductive argument are asserted and the conclusion
is denied. On Popper's view, there is no adequate explana
tion of the fact that inconsistency is unavoidable in. every
such case, and of the fact that this is known to be so.
96
(2) That Popper does not intend to characterize a
valid rule of inference as a mere conjecture with the same
status as a scientific hypothesis seems clear from his
claim that we may he able to "establish" that no counter
example to a given rule exists. Apparently he recognizes
a certain necessity that attaches to valid rules of infer
ence that eludes even the most highly "corroborated" sci
entific theories. And indeed, the question of the neces
sity of logical principles and rules is crucial to any
attempt to understand their nature. However, Popper seems
muddled at this point. He has argued that every scientific
theory and every refutation of a scientific theory is pro
visional because one can never know when an empirical coun
ter instance will be found. Since we can never claim to
know all empirical states of affairs, the possibility of
refutation is never precluded. An experiment or observa
tion may be made tomorrow that will serve to falsify any
of our theories of today.
Popper seems to make two incompatible claims with
reference to valid rules of inference. First, he says that
a suggested rule is valid only if we cannot find a counter
example, i.e., an instance of obtaining a false conclusion
from true premises by inferring according to the rule. And
then he says that we may be able to establish that none ex
ists (52:206). "Counter example" here is a term that
seems to refer to "content" rather than, to "form." There
are several considerations which appear to support this in
terpretation of what Popper means by a "counter example."
Firsts he says that "a rule of inference, after substitu
tion of constants for the variables, asserts something
about a certain argument--an 'observance1 of the rule--
namely, that this argument is valid" (52:203). Then he
says that "an observance of a suggested rule which allows
us to obtain a false conclusion from true premises" is a
"counter example" (52:206). He defines an "observance of
a rule" as the substitution of "constants" (by which he
apparently means fixed terms which have a definite meaning)
in an argument that proceeds according to that rule. That
he evidently means this can be seen from the illustration
he provides: "If all men are mortal, and all Greeks are
men, then all Greeks are mortal" is his own example of the
substitution of constants for the variables of the formula,
"If M a P and S_ a M, then S £ P" (52:203). In other
words, Popper appears to say that by substituting various
terms in the different kinds of propositions (or "proposi-
tional functions") used in an argument, one can test the
validity of the rule of inference according to which the
argument should proceed. If, when making such substitu
tions, an instance of deriving a false conclusion from true
premises results from the application of the rule of infer
ence, then that rule is Invalid and should not be used for
logical deduction. But it Is obvious that If this proce
dure Is one of substituting concrete cases, then the pro
cess Is Interminable. Since our terms are referential, and
since referents--whether actual or imaginary--are numeri
cally indefinite, if not infinite, there seems to be no way
this method can establish that no counter example to a
given rule exists.
If deductive validity is construed as hinging upon
empirical tests, then no necessary, certain, or final rule
of inference can be formulated. On that basis no one can
know whether a counter example will someday be found.
Moreover, one may establish that a "counter example" to a
suggested "rule" exists, but it is hardly possible that
one, employing Popper’s a posteriori method, can establish
that none exists. However, if I have been misled by
Popper's terminology and illustrations, and if what he
actually means by "counter example" is an instance of ob
taining a false conclusion from true premises by the obser
vance of a suggested rule as_ demonstrated by truth tables,
then his method is not dependent upon empirical content,
and I would concur that it can establish that a rule of in
ference is valid. Hence, either valid rules of deductive
inference are a priori or their universality and necessity
cannot be established.
(3) There seems to be another serious defect in
Popper's analysis of rules of inference. While he makes
the attempted adduction of counter examples the test of the
validity of proposed rules of inference, he does not deal
with the problem of circularity in such an assumption. The
circularity is evident in the very attempt to test all
rules of inference by using a rule of inference. The
scheme is simply that of modus tollens. Let us suppose
that p is a proposed "rule" of inference (or rather a cor
responding formula of the calculus of propositions), and
that _q is an instance of inferring according to that rule.
Now if, in a particular instance of inferring (i.e., apply
ing p in a specific argument), it is found that a false
conclusion is derived from true premises, then the rule of
inference is invalidated: i.e., "If p, then q; not q,
therefore, not p." But in order to obtain the refutation
of a rule of inference by appealing to a relevant counter
example, one is employing a rule of inference, namely,
modus tollens. Since this is the case, there is no way of
testing modus tollens by the same method, for that would
involve the error of petitio principii. And since it must
be presupposed in any attempted test, it is irrefutable.
In view of the fact that there is at least one instance of
a rule of inference that cannot be refuted by adducing a
counter example, it follows that it is indefensible to ar
gue that a rule of inference is valid only if it survives
100
an assiduous search for counter examples. Popper's gener
alization of such a test, therefore, cannot he sustained.
In at least one instance the adduction of counter examples
is impossible.
How, then, is the necessity of modus tollens to be
understood? If it is not due to the fact that no counter
examples to it can be obtained, what, then, is its basis?
It seems that the necessity must not be material but for
mal, i.e., it must be rooted in the structure of logical
principles. This is not to imply, however, that the prin
ciples and rules are only "subjective” and have no descrip
tive applicability to the world. But that is another prob
lem. The important question at this point is whether a
rule of inference is valid because no counter example can
be found or whether no counter example can be found because
the rule of inference is valid.
Perhaps with the awareness that rules of inference
cannot be established with certainty by the method of es
caping counter examples, Popper resorts to faith. He says
that "the fundamental rationalist attitude results from an
(at least tentative) act of faith--from faith in reason.
Accordingly, our choice is open" (56:231). It is one
thing to say that the choice is open, however, but it is
another thing to show that it is and in what sense it is.
Popper says that rules of inference are formulated on the
101
basis of an analysis of our "intuitive procedure/' i.e.,
the procedure called "reasoning" or "arguing," which we
become conversant with in the process of learning to speak
and to use our language for the purpose of describing facts
(52:206). Perhaps his error is not so much that he fails
to see in this a clue to the nature of logic as it is his
prior refusal to admit a distinction between legitimate and
illegitimate "sources" of knowledge--or to concede certain
knowledge at all. He claims that one either has faith in
his "intuitive procedures" and their critical formulations,
or he does not. To believe or not to believe in them is,
according to Popper, in the power of the individual.
(4) Not only is there no grounding for logic apart
from fideism, in Popper's view, but there are no true (or
knowably true) premises. Every proposition is refutable:
"every bit of our traditional knowledge (and even our in
born knowledge) is open to critical examination and may be
overthrown" (52:28). In view of this statement, it is
difficult to know what he means by saying that a valid rule
of inference can be relied on as "a rule of procedure that
leads from true descriptions of facts to true descriptions
of facts" (italics mine, 52:206). He has categorically
denied the possibility of attaining a true description of
facts, for all we can hope for are "approximations" to
truth.
102
If one can never know whether he has true premises,
then even if he draws inferences according to valid rules
he can never know if he has true conclusions. According to
Popper's blanket statement about all propositions, formal
and material, there is no genuine certainty. It follows,
then, that it can hardly be a true description of the facts
for Popper to say that following a valid rule of procedure
will lead us from true descriptions of facts to true des
criptions of facts. In the final analysis, according to
Popper, there are no valid rules of inference or true des
criptions of facts, so far as can be known. The subjective
theorizing that we are compelled to "impose" on everything
we encounter throws up a smoke screen that inexorably hides
reality. For Popper there is no way of ascertaining any
truth. "Truth" is a term that he uses to refer to an eter
nally elusive postulation, the omega point toward which we
strive in our "approximations."
A serious question remains, however. How can we
know if our theories are approximations of the truth if we
cannot know the truth at all? It is inadequate and mis
leading for Popper to say that "coherence cannot establish
truth, but incoherence and inconsistency do establish
falsehood" (52:28). For there can be no recognition of
incoherence and inconsistency without coherence and consis
tency as the standards by which to judge that they are
such. One cannot criticize and refute theories without
- <**
103
some logical and epistemologlcal base on which to stand.
Popper does not explain how we can have only negative but
not positive knowledge, how we can know inconsistency but
not consistency, how we can know error but not truth. If
one does not imply the possibility of the other, then it is
impossible to know either. This means that some very ob
vious errors and some very obvious truths--even the most
elementary logical ones--must be viewed as inconclusive and
uncertain. This absurdity is enough to reveal the untena-
bility of such a position.
Three Main Views of Logic
We must first see how Popper understands the nature
of logic if we are to evaluate his answer to the question
of logic's applicability to reality. He distinguishes
three main views of the nature of logic:
(A) The rules of logic are laws of thought.
(Al) They are natural laws of thought--they describe
how we actually do think; and we cannot think
otherwise„
(A2) They are normative laws--they tell us how we
ought to think.
(B) The rules of logic are the most general laws of
nature--they are descriptive laws holding for any
object whatsoever.
(C) The rules of logic are laws of certain descriptive
languages--of the use of words and especially of
sentences. (52:207)
Popper rejects (Al) because it rests on the assump
tion that "a state of affairs for which they do not hold
good is inconceivable" (52:207). But inconceivability is
not a sufficient guarantee that they should be true. Only
104
if we knew that all objects are subject to the laws of
logic would we have an adequate basis for asserting their
inescapability and truth, i.e., "if (B) is correct, then
their compelling character would be clear and reasonable;
otherwise we may perhaps feel compelled to think in this
way merely because we have a neurotic compulsion" ( 52 :
207).
He finds that (A2) is equally unacceptable. It
asserts that we ought to avoid violations of the laws of
logic, i.e., "fallacies." But it does not tell us why, and
Popper is convinced that such breaches of logic are not
immoral or stupid. Justification for abiding by the rules
of logic is provided if, and only if, we are interested in
deriving true statements, i.e., "true descriptions of
facts." But to argue that only inferences made according
to the rules of logic will lead to true conclusions about
states of affairs leads to (B).
However, Popper finds three reasons why (B) appears
to be untenable:
First, because the rules of inference. . .are rules of
procedure rather than descriptive statements; secondly,
because an important class of logically true formulae
(viz., precisely those which Professor Ryle would call
the logician's hypotheticals) can be interpreted as, or
correspond to, rules of inference, and because these
. . .do not apply to facts in the sense in which a fit
ting description does. Thirdly, because any theory
which does not allow for the radical difference between
the status of a physical truism (such as "All rocks are
heavy") and a logical truism (such as "All rocks are
rocks" or perhaps "Either all rocks are heavy or some
105
rocks are not heavy") must he unsatisfactory. (52:
207)
Popper contends that the only reason such logical truisms
are true Is that there Is no way for them to be falsified
by any fact, and not because they describe the behavior of
all possible facts. Since logical truisms do not exclude
any possible fact, they do not assert anything whatever
about any fact at all. In view of these considerations,
(B) is found unacceptable.
Popper maintains that there are two chief interpre
tations of position (C). One version of it, which reduces
language to mere symbolism, is untenable because it makes
valid inference impossible. For interpreting language as
a symbolism without meaning precludes the possibility of
true or false statements. Before a statement can be said
to be either true or false, it must have a clear and deter
minable meaning. And if valid inference can only be de
fined in terms of truth, this version of (C) makes it im
possible both to construct rules of inference and to justi
fy the rules as valid, good, or useful.
Another interpretation of (C), however, avoids this
problem. Taking language to mean a symbolism that allows
us to make true statements (as explained by Tarski), Popper
argues that "a valid rule of inference with regard to such
a semantic language system would be a rule to which, in the
language in question, no counter example can be found,
106
because no counter example exists" (52:208). Since there
Is a diversity of semantic language systems, there Is a
sense In which the rules of Inference are always relative
to a particular language system. Nevertheless, this rela
tivism is limited by a common characteristic of all of
them: inferring in accordance with the rules leads from
true premises to true conclusions. This one absolute can
not be expunged from any language system "simply because we
have defined the term 'rule of inference' in such a way
that this is impossible" (52:208).
As one example of what he means, Popper considers
the inadequacy of "the language of categorical proposi
tions." Someone may argue, for instance, "Rachel is the
mother of Richard. Richard is the father of Robert. The
mother of the father is the paternal grandmother. Thus,
Rachel is the paternal grandmother of Robert." Schema
tized, this argument is as follows:
x R y
y S z
R S = T
x T z
Now if this same argument is formulated in the language of
categorical propositions, it is impossible to defend the
validity of the argument, because a valid rule of inference
which exhibits its general form cannot be formulated. The
language of categorical propositions presupposes that the
subject and predicate are single terms, i.e., that the
107
components of the term are inseparable. For example, the
predicate of the first premise, "mother of Richard" is
fused into one term. Thus, "the logical structure of this
language is too poor to exhibit the fact that this predi
cate contains, in some way or other, the subject of the
second premise, and part of the subject of the third pre
mise. Similar remarks hold for the other two premises and
for the conclusion" (52:209). The language of categorical
propositions allows only the following formulation of the
argument:
A is b
C is d
All e are f
A is g
That this rule is invalid can easily be shown by adducing
counter examples. In the former formulation it is possible
to substitute any proper name of individuals for "x," "y,"
and "z," and any names of relations between individuals for
"R," "S," and "T," etc., and yet find no counter examples
that would invalidate the rule. But such substitution in
the formulation according to the language of categorical
propositions is not invariably possible without the occur
rence of cases in which false conclusions are derived from
true premises. Hence, a language may be rich enough for
describing all the facts we wish to describe, but it may
not allow the formulation of the rules of inference needed
for all the cases in which we can safely pass from true
108
premises to true conclusions (52:209).
These considerations lend support to the view that
it is impossible "to construct one language permitting the
formalization of all valid intuitive inferences" (52:210).
This observation brings us to the heart of the question
posed in the title of this chapter. This question must be
viewed in the light of the fact that the applicability of
every calculus breaks down at some stage or other, i.e.,
there is no one language that serves as an adequate vehicle
for formalizing all inferences which we intuitively know
how to draw. Popper's answer to the question of the appli
cability of logical and arithmetical calculi to reality is
given in three statements:
(a) These calculi as a rule are semantical systems,
that is to say, languages designed with the intention
of being used for the description of certain facts.
If it turns out that they serve this purpose then we
need not be surprised.
(b) They may be so designed that they do not serve the
purpose; this can be seen from the fact that certain
calculi--for example, the arithmetic of natural num
bers, or that of real numbers--are helpful in describ
ing certain kinds of fact, but not other kinds.
(c) In so far as a calculus is applied to reality, it
loses the character of a logical calculus and becomes
a descriptive theory which may be empirically refu
table; and in so far as it is treated as irrefutable,
i.e., as a system of logically true formulae, rather
than a descriptive scientific theory, it is not applied
to reality. (52:210)
Deferring his discussion of (a), Popper remarks that (b) is
so evident in the restricted use of certain calculi that
"the belief that any one of the calculi of arithmetic is
109
applicable to any reality ... Is therefore hardly tena
ble " (52:211).
As to (c), Popper assumes that a logical calculus
as such has no empirical reference. In this sense It Is
only a system of logically true formulae and It has no
application to reality. He illustrates this by the fact
that the statement "2 apples + 2 apples = 4 apples" makes
no statement about apples or any other empirical fact, and
hence it is irrefutable. It is a logical truism that mere
ly asserts that a certain way of describing reality is
equivalent to another way. But if we interpret and use a
calculus to describe certain physical facts, then it is a
descriptive theory that may be empirically refutable.
Thus, if someone puts two apples in an empty basket in one
act, and then two more in another act, there will be four
in it. In this case, the symbol "+" is not logical but
descriptive, i.e., it stands for the act of physically
adding certain things to other things. Since this inter
pretation of the statement "2 + 2 = 4" is a physical theo
ry, it is empirically testable. We find that it does not
remain universally true. For example, two drops of water
added to two drops of water in a dry flask will not equal
four drops of water. To stipulate that the equation "2 +
2 = 4" only applies to a world of distinct and static ob
jects is to preclude its application to "reality," for
110
"reality" is characterized by innumerable kinds of changes.
Obviously, then, arithmetic is applicable to the extent to
which physical conditions resemble pure logical or arith
metical operations. Since we can imagine worlds in which
calculi of addition of measurements, for example, are dif
ferent from ours, it is important that we distinguish be
tween the logical and the descriptive uses of symbols. The
criterion of demarcation is that the latter incur the pos
sibility of empirical refutation. Accordingly, Popper
says, "Only if we are ready to accept refutations do we
speak about reality" (52:212). But why is it that we can
speak successfully about reality at all? Does this imply
that reality must have a definite structure? Popper re
plies that while our world is describable in our language,
we can imagine worlds that would not be. And even for a
finer description of our own physical world we find that
new linguistic means make it possible to talk about new
kinds of facts.
What, then, is the relationship between the means
of description and the described facts? It is of primary
importance, in Popper's view, that facts should never be
confused with things or processes which may be said to be
parts of the world. He says that "facts are something
like a common product of language and reality; they are
reality pinned down by descriptive statements" (52:214).
This means that they are the result of selection and
ill
abstraction, from the continuum of events, often made pos
sible by new linguistic means, e.g., a theory Invented
today describing facts that occurred a hundred years ago.
Criticism of Popper's Analysis
With this statement of Popper's position before us,
I now turn to critical evaluation. The first general ob
servation is that Popper has restricted his explicit asser
tions to the calculi of logic and arithmetic. "Calculi" is
a term that refers to methods, or in his words, "the rules
of procedure." At times, however, he seems to include ref
erence to the principles of logic, i.e., the lav/s of iden
tity, noncontradiction, excluded middle, and various state
ments that express judgments about rules of inference.
Popper seems to think that the question of the ap
plicability of logic to reality can be answered without
analyzing the status of the principles of logic. That this
cannot be done, however, can be shown by demonstrating the
inseparability of the rules of logic from the principles of
logic. Of course, one can be prescinded from the other,
just as thinking and the objects of thought can be distin
guished; but in actuality the two cannot be severed. If
this is true, then the status of logical principles is
crucial to the question of the applicability of logic to
reality. In fact, it is misleading to merely define logic
as "the doctrine of valid inference" if one thereby implies
112
that the principles of logic are dispensable or trivial.
It is difficult to see, for example, how a meaning
ful assertion can be made without presupposing the law of
noncontradiction. For instance, a given term must mean
something definite or refer to something definite in the
assertion in which it is used, for if a term can mean
everything, then it means nothing. Thus, if I say "I live
in a house,” the term "house" must mean house, and it must
not mean (at the same time) bird, door, man, cloud, etc.
If a term could mean everything, then the statement, "house
house house house house" could mean "The tree fell toward
the north," or "Five men walked into the office," or any
thing else--even the contradictories of these statements.
If we are to use terms to refer to things, then terms must
have a specifiable meaning. Moreover, one cannot say that
a certain term (t_) means x and that it does not mean non-x
without presupposing that a proposition (i.e., "t_ means x ")
cannot both be true and not true at the same time and in
the same way. For from two contradictory statements, any
statement whatever can be inferred--including the negations
of every derived statement--as Popper himself shows ( 52:
317-19). Furthermore, the principles and rules of logic
themselves cannot be formulated without some "meaningful
content," for even "If p then q, p, therefore q" is "mean
ingful" in the sense that "if," "then," "therefore," or
any symbols used in their place have specifiable signifi-
113
cance* and "p_" and "q" are symbols for propositions that
are taken to be unequivocal in the argument. Hence* to
state the rules of logic requires both the law of noncon
tradiction and the possibility of meaningful content.
The Normative View
Let us now examine Popper's assessment of the three
main views of the nature of logic. The first view main
tains that the rules of logic "either describe how we think
or how we ought to think. That they do not describe how we
actually think is easily established. From the context*
"think" evidently means "infer." But it is obvious that
human beings do not always infer correctly* i.e.* according
to valid rules of logic. We sometimes wrongly think that
an assertion is guaranteed by certain propositions that we
take to be premises in an argument. We know what it is to
have our arguments challenged and sometimes refuted. Not
only do people infer incorrectly, they sometimes accept
contradictory propositions. Psychologically this is pos
sible* and therefore it is indefensible to maintain that
the rules or laws of logic describe how we actually think.
Moreover* that contradictory beliefs cannot both
be true is hardly explicable in terms of the "limitations"
of our thought. For if it is said that we cannot conceive
of (i.e.* believe contradictory propositions to be true at
the same time) the moon existing and not existing at the
same time and in the same way, but that it is nevertheless
possible that it might in fact both exist and not exist
simultaneously, this double claim itself is contradictory.
To assert the empirical possibility of both the moon's
simultaneous existence and nonexistence is to assume that
both claims are conceivable (i.e., can be believed to ob
tain simultaneously). But if it is thus assumed that the
simultaneous existence and nonexistence of the moon is con
ceivable, then the advocate of (Al) may say that both of
these claims can be believed simultaneously, but he cannot
give any consistent sense to this "assertion." To say that
although we cannot understand how both sides of the "asser
tion" are true, but that in fact they may be so, means that
we are conceiving of p_ and non-p obtaining simultaneously,
which is exactly what is declared to be inconceivable.
This contradiction makes it difficult, if not impossible,
to defend the position that the laws of logic are nothing
more than laws of thought. Popper's criticism of the posi
tion seems to stop short of the fundamental issue. He ar
gues that since it is based on the notion of "inconceiv
ability," it is suspect. He observes that the laws of
logic must hold for all objects if their compelling char
acter is to be acceptable. But if this is meant to imply
that only an empirical knowledge of objects can justify the
laws of logic, then the laws of logic would always be ten
tative and open to testing as long as our knowledge of
empirical objects is incomplete.
At this point it is important to note two ambigui
ties that mar his dismissal of inconceivability and self
evidence as "always suspect." First* it is not obvious
that "inconceivability" and "self-evidence" are always
open to question. Both of these terms need explication
before they can be so characterized. For example* it seems
reasonable to distinguish empirical inconceivability from
logical inconceivability. While the splitting of the atom
may have been "inconceivable" to people in the nineteenth
century* its conception and actualization in the twentieth
century show that a misjudgment had been made. But the
laws of logic are categorially different from theories and
events like the splitting of the atom. What is it that
makes the difference? Why is the denial of the laws of
logic "inconceivable"? Is it possible to dispense with
self-evidence on every level of human knowledge? It has
already been observed that Popper interprets his acceptance
of the canons of rationality as a form of fideism. But is
there no distinction between faith and "self-evidence"?
If there is* what is it and on what is it based? I shall
return to these questions later.
Secondly* on what grounds does Popper base his
claim that "an argument that proceeds from inconceivability
is* like other self-evidence arguments* always suspect"
116
(52:207)? Perhaps one possibility is that it is inductive
ly grounded, i.e., since alleged "self-evidence arguments"
have sometimes proven incorrect in the past, all of them
are unreliable in the present. But this clearly does not
follow--and, besides, Popper has unequivocally repudiated
induction. Another possibility is that it is deduced from
other propositions. But it is not clear what these might
be. Moreover, those propositions themselves must be estab
lished inductively, deductively, or intuitively (i.e.,
"self-evidentially"). Induction is excluded, as we have
seen. Deduction leaves the problem unresolved, unless its
primitive propositions are mere assumptions or conjectures
--which would mean that the conclusion (viz., the assertion
in question) is itself a mere conjecture, and there is
therefore no obvious reason why we ought to take it seri
ously. The only other possibility is that the assertion
itself, or its basic premises, must be self-evident. But
that is impossible, for such a contention would be self-
defeating. How could it be self-evident that self-evidence
is always suspect? And if it is true that an argument that
proceeds from "inconceivability" is like other arguments
based on self-evidence, then, contrary to Popper, there
seem to be good reasons for viewing it as epistemological-
ly, and perhaps ontologically, important.
The other variant of (A) suffers from ambiguity.
To say that "the rules of logic tell us how we ought to
117
think" is a statement that needs qualification. For there
are plainly many details in any given area of inquiry of
which one should be cognizant if he is to think properly
about the matter at hand. Moreover, the rules do not tell
us that we ought to reason or that we ought to aim at true
conclusions. But there is a sense in which the laws of
logic are "normative" as the most general laws of reason
ing. All of our reasoning from premises to conclusions
"ought" to be in accordance with them, but the important
question is: Why should it? As Popper rightly observes, if
we do not abide by the laws of logic, we cannot derive true
statements about facts. If we are to think correctly about
facts, then, we must think logically. And this implies
that the objects of our thought are amenable to the laws
of logic.
The Semantic-Conventionalist View
Since I am interested in defending a version of
(B), I shall deal with (c) first. Popper adopts this view
when it is construed as asserting that the rules of logic
are laws of certain descriptive languages, i.e., semantic
systems which allow us to make true statements. He says
that so long as no counter example to a rule can be found
in a given language, that rule is considered valid. This
is a variant of conventionalism. It contends that the laws
and rules of logic are conventions, varying according to
118
particular semantic language systems as rules for games
differ according to the particular game that Is being
played. In order to play properly and successfully*
whether it be in the game of chess* or baseball* or foot
ball* or whatever* the rules must be followed. But there
is no necessity in any such set of rules. They are arbi
trary formulations that make possible the achievement of
some end. They could be replaced by alternative rules* so
long as the system in which they operate is conducive to
their application. Just as a language is a system of arbi
trary conventions (e.g.* there is no necessary connection
between a word and what it signifies or refers to)* so the
conventionalist holds that the laws of logic are statements
about particular linguistic usages* and hence that they
share the same character of arbitrariness.
This kind of conventionalism is based on a misun
derstanding of the nature of logic. Conventionalists fail
to see that logical laws and rules hold primarily for mean
ings, not symbols. Symbols* in the sense of mere marks on
paper or sounds in the air* have spatial and temporal rela
tions but no intrinsic meanings or meaning-relations.
Popper is aware of the problem involved in the positivist
notion that logical laws are propositions about the rela
tionship of symbols. He wants to avoid the absurdities in
volved in that position by defining language in terms of
meaning so statements can be true or false. This Is essen
tial to his definition of valid Inference. Just as he
sought to steer a middle course between essentialism and
Instrumentalism In his theory of knowledge, so he seeks a
mediating position in his theory of logic. On the one
hand, there Is the position that the laws of logic are
rules governing the use of symbols in a descriptive lan
guage. On the other hand, there is the position that the
laws of logic are propositions about meanings which refer
to the properties and relations of things. For Popper,
however, both views are untenable. He argues that since
there are numerous semantic language systems (whether actu
al or possible, natural or artificial), there is both an
arbitrariness and necessity about the rules of logic. The
arbitrariness lies in the fact that any given language is
conventional. The necessity lies in the fact that in every
language validity means that observance of the rules of
procedure lead from true premises to true conclusions. But
even here Popper's "necessity" is-arbitrarily based, for
his definition of valid inference is not adequately ground
ed. Rather it is "justified" fideistically and convention
ally, i.e., we decide to define it in such a way that our
interest in ascertaining the "truth" about facts will be
facilitated. But precisely why it is that one definition
of valid inference provides the basis for gaining a true
120
description of facts while other definitions do not Is left
without explanation.
Certain conventionalists* then., have tried to as
similate the principles and rules of logic to the rules of
language,, but the rules of language are arbitrary* while
the rules of logic are not. Whether one is reasoning in
English* French* Chinese* or Swedish* the law of noncontra
diction and the logical relations between propositions hold
without alteration by the linguistic medium in which they
are expressed. The laws and rules of logic are unrestrict
edly general and formal* whereas the rules of languages are
not. And when we take Popper's improved* but inadequate*
view of the nature of logic* we find that he becomes entan
gled in some serious problems. For instance* if we argue
that "Moby Dick is a whale* all whales are mammals* there
fore Moby Dick is a mammal*" is the conclusion merely a
consequence of observing the rules of the English lan
guage? Now let us suppose that Popper would agree that
each of these propositions has meaning and is true* and
that the premises imply the conclusion* then clearly he
must admit that the argument is sound. But why is it val-
is* according to Popper?" It is important to recognize ex
actly what he says in answer to this question. He main
tains* in effect* that it is sound because no counter ex
ample can be found in English to refute the validity of the
121
rule of inference by which the conclusion was derived from
the premises. This is not only an odd way of describing
validity, it is also an unnecessary way--once it is seen
that a proposition (i.e., a complex universal which con
sists of concepts, or simple referential qualities, and
which is about something being this or that) should be dis
tinguished from the linguistic symbols by which it is ex
pressed. The same proposition, being a universal, can be
expressed in many different languages. The language does
not determine the laws or rules of logic which govern pro
positions. Rather, it is the reverse situation that holds,
i.e., linguistic usage, if it is coherent, universally
exemplifies logical principles.
The radical conventionalist alleges that alterna
tive laws and rules of logic are equally acceptable. But
he cannot explain why it is that we have "adopted" certain
"conventions" and not others. In fact, it is difficult to
see how he could offer any reasons, since good reasons for
one choice and the absence of good reasons for another re
move the arbitrariness that is the sine qua non of conven
tionalism. Popper disavows this extreme version of conven
tionalism, however, and he believes that there is a good
reason for adopting certain rules of procedure, viz., that
their observance will enable us to infer true, descriptive
statements of fact. In other words, he says that it is not
because we grasp certain necessities, but solely because we
have an interest in the truth.
Nevertheless, it remains a perplexity how we can
know that certain rules of inference lead us to true des
criptions of fact when Popper has argued that truth is un-
discoverable. If the latter contention were "true," there
would be no way of knowing which set of conventions was
preferable as a means to the end of attaining truth, nor
would there be any way of knowing whether there were coun
ter examples to any proposed rule of inference, for in both
cases one must be able to discover the truth if it is to
serve as the final criterion. In other words, if the
choice of a "convention" is made by reference to an end to
which it is the means, the end must be knowable. Other
wise, it could never be ascertained whether certain means
did indeed subserve a certain end. Moreover, is all In
terest in truth a mere convention itself? Is it due to an
arbitrary decision that the attainment of truth is better
than its obscuration?
A valid rule of inference is "adopted" not because
no counter example can be found in a particular language,
but because of a perceived necessity that precludes the
possibility of any counter example being found at all.
There are certain relations between the meanings and truth-
values of propositions that hold solely on the basis of
necessary logical principles. We cannot treat these prin
ciples as arbitrary conventions of linguistic usage., for in
that case we could "conceive" of a language in which the
negation of the law of noncontradiction is possible. But
we cannot "conceive" of it, not merely because our minds
are so constructed but because we see that a proposition
cannot both be true and false in the same way at the same
time. For example, we perceive that a particular man can
not both have two hands and not have two hands. The two
contradictory propositions stating these mutually exclusive
states of affairs cannot both be true because the man him
self cannot have incompatible properties. The necessity
here is not so much in our understanding as it is in the
nature of things. Our minds (i.e., we, as persons) are
qualified by the propositions we hold, and psychologically
we may hold two contradictory propositions, but this does
not mean that these two propositions are compatible. The
fact is that they are logically incompatible even when one
may not recognize them as such. Hence, logical incompati
bility holds necessarily in spite of psychological compati
bility. Two contradictory propositions do not become logi
cally incompatible at the time that one recognizes them to
be mutually exclusive. Rather, one brings his thought into
conformity with the objective state of affairs, whether it
is ideal or empirical. Thus, necessity is not "imposed"
on the objects of our thought by our minds, but is discov-
124
ered as a transcendent quality of certain states of af
fairs .
Popper speaks of defining valid inference, however,
as if it could be defined in a radically different way.
That is, he seems to suggest that.the definition of valid
inference is arbitrary in some respect. His argument is
that if we have an interest in the truth, we shall want
rules that will secure the derivation of true conclusions
from true premises. And since the only "truth" we can
know is the product of the interaction of our free conjec
tures and reality, "truth" is never stable or final. More
over, an individual will meet reality with his own variable
conjectures, so that "truth” will not only be different for
others but "truth" may not even be the same for the same
person at different times. Now if truth is supposedly de
termined by the "spontaneous, free imposition of our theo
ries on reality," then obviously what we "know" can never
be known to be truth. It may be the result of "invariable
distortion" by our theorizing. Popper's assumption is
fatal to any attempt to arrive at a coherent epistemology.
And here the inseparability of logic and ontology can
plainly be seen. If reality has no structures or essences
(or even if it did, but we could not know them), then truth
(i.e., that things are the way they are thought to be) can
have no meaningful role in epistemology or logic. For if
125
we cannot know truth, then we cannot define valid Inference
In terms of truth and falsity.
Nevertheless, there Is a difference between valid
and Invalid Inference, and we can know when we have an In
stance of one or the other. Entailment Is not a conven
tional rule but a logical necessity. It holds whether one
agrees to it or not. Even arbitrary principles may have
logical implications, and it is impossible to interpret the
latter themselves as arbitrary. Conventional axioms may
logically entail certain theorems, but the entailment it
self is not a convention. For example, one may receive
instructions to write each of the letters of the alphabet
with a different color crayon, but doing so entails that in
every instance the colors will be extended. That conse
quence one cannot alter, for it is a necessary entailment
of writing at all with crayons. One can change the in
structions as many times as he likes (e.g., to write with
only one color crayon, or two, or fifty, or to use one
color for every two letters, or every three letters, etc.),
but he will encounter the necessity of extension In every
case. And he can know this before he goes through the act
of writing with any particular crayon. And try as he will,
he cannot make any proposals or rules that will nullify
this entailment of his arbitrary instructions about writing
with color crayons.
126
Furthermore, since even words and symbols are ob
jects, there can be no absolute separation between proposi
tions about symbols and propositions about objects. Logic
applies to the most general relations between all denoted
objects, whether ideal or empirical. This is the only
satisfactory explanation for the fact that a valid proof
(whether in mathematics, biology, or physics, etc.) is not
determined by linguistic syntax but is valid in every lan
guage, whether English, Russian, Japanese, etc. Thus,
there are definitive limitations put upon any attempt to
reduce all of logic to conventions. Whatever we refer to
has a certain character, and this character presents us
with a necessity that we are not free to treat as a conven
tion. For example, we cannot "meaningfully" assert that
x is y and that x is not y--or concretely illustrated, that
green is a color and that green is not a color, or that a
given table is completely composed of wood and that it is
not. Neither can we "meaningfully" assert that x is y, and
y is z, and then deny that x is z_.
Indeed, it is this kind of necessity that makes it
impossible to state meaningfully that "all logic is conven
tional, " for without the absoluteness of the law of noncon
tradiction, this statement is meaningless. It could mean
anything, and therefore it means nothing. If the law of
noncontradiction is arbitrary, then contradictories may
both be true. But if this is so, then the contradictory
127
of the conventionalist position has an equal claim to being
true. Hence* the conventionalist holds a self-defeating
position. He has no reason for denying that some logic is
not conventional* for by his own admission it may be true.
To interpret the law of noncontradiction as a con
vention is to erase the distinction between true and false
propositions. And without that distinction the basis of
knowledge is destroyed. Moreover* even to think the denial
of the law of noncontradiction is made impossible* for to
think that it is false and not true presupposes the dis
tinction between true and false* a distinction that cannot
obtain without the law of noncontradiction.
The Descriptive View
As I turn now to Popper’s criticisms of position
(B)* it is possible to evaluate them from the standpoint of
the foregoing arguments for the unconventionality of logic.
First of all, Popper rejects the view that the "rules" of
logic say anything about "reality*" for he says that rules
of inference are rules of procedure rather than descriptive
statements. And rules of procedure tell us how to do some
thing; they do not tell us anything about the actual char
acter or relations of things. But his failure to distin
guish the rules from propositions about the rules prevents
him from seeing that the corresponding propositions do say
something about reality* because they are descriptive
128
statements about ideal objects which are given to con
sciousness .
Moreover, rules of inference cannot be understood
properly apart from the distinction between the validity of
what is thought and the thinking itself. When thinking
aims at truth, it aims at something that transcends the
psychological activities in which objects are apprehended.
If we are to achieve this aim, our thinking must adjust to
the way things are, not to the way in which we desire them
to be. If our judgments are to be true and if our reason
ing is to be valid, then they must conform to the necessary
relations of concepts and propositions and the states of
affairs to which they refer. My judgment that a particular
table has a round top is false if it actually has a square
top. The falsity of my judgment is not determined by lin
guistic usage or arbitrarily defined rules, but by the
shape of the table and by the obvious fact that the top of
the table cannot both be square and not square (or round,
etc.) at the same time and in the same way. If my thinking
is to be sound, then, it is not required that it comply
with certain conventional rules, rather it is mandatory
that it accord with the character of its object.
Of course the rules of logic should function as
rules of procedure when we want to arrive at the truth, but
Popper has apparently failed to see what this implies. For
one thing, "alternative rules" will not serve to ascertain
129
the truth about actual states of affairs. If it is averred
that they can, then why are they not devised? Why is it
that only certain so-called conventions are adopted as the
rules of logic? Why is it that the law of noncontradiction
and the rule of modus ponens, for example, cannot be ig
nored or violated without the consequence that inferences
will then lead to falsehood rather than to truth?
A second implication is correlative to the first:
the reason that only some rules of procedure are successful
in leading us to truth is that propositions and their ref
erents have a determinate structure, i.e., reality has a
character that will be missed or obscured if approached by
unsound principles and invalid rules of logic.
Thirdly, Popper seems to think that the rules of
logic are only relatively compelling, for they are alleg
edly necessary only in the sense that if we are "conven
tionally" interested in truth, we are compelled to adopt
certain rules of procedure. Thus, there are no principles
or rules of logic that are absolute or intrinsically valid.
To assert that they are absolute would make them more than
instrumental, and to assert that they are intrinsically
valid would make them more than conventional. In other
words, logical necessity and ontological reference are both
implied, but Popper resists this conclusion. Perhaps this
is due to the fact that it would require the radical alter-
130
atlon of his entire perspective on knowledge and science--
in fact,, his view would face a crisis of integrity both as
a statement of nonjustificationalism and as a form of fide-
ism. Once it is recognized that not all interest in truth
is conventional (i.e., that we cannot arbitrarily adopt an
equal interest in error or let this latter interest become
regulative for our lives), then it will be seen that the
rules of inference which lead to truth are no more arbi
trary than is truth itself.
Popper's second criticism of position (B) is that
the "logician's hypotheticals" can be interpreted as rules
of inference, which, he contends, do not apply to facts in
the sense in which a fitting description does. What Popper
seems to mean here is that neither truths of logic or logi
cal truths are descriptive or informative of reality. They
are empty of content and depend upon other propositions
which are true descriptions of fact, for it is the latter
that may or may not be "fitting."
There are two considerations, however, which serve
to eviscerate Popper's argument. First, the adequacy of
his criticism is called in question by the fact that only
a subset of logically true formulae can be interpreted as
rules of inference. He himself recognizes that generally
there are no corresponding rules of inference for non-
hypothetical formulae. Hence, (B) can still be maintained
in the sense that non-hypothetical formulae are descriptive
laws holding for any object whatsoever. Secondly, If, as
has been shown, all rules of Inference cannot be Interpre
ted as conventions or mere Instruments relative to an arbi
trarily adopted end, then even if certain logically true
formulae are interpretable as rules of inference, this does
not preclude their descriptive reference to actual states
of affairs. Only if a "fitting description" is defined in
terms of concrete or specific empirical content can it be
said that no logical formulae are descriptive of the world.
It seems that if the fundamental principles of logic are
true and necessary at all, it is because they are true and
necessary ontologically. And if rules of inference are
useful at all, it is because they are useful ontologically.
This means that in some sense they are informative about
the world, and consequently there are true a priori propo
sitions that refer to actual states of affairs.
Popper's third objection to (B) is that it does not
allow for the radical difference between the status of a
physical and a logical truism. He assumes that a proposi
tion must exclude some possible "fact" in order to assert
something factually informative. With this antecedent
assumption, it is easy for Popper to dismiss so-called log
ical truisms as empty tautologies. For example, he offers
the statement, "All rocks are rocks," as an illustration of
a logical truism. Of course this is not empirically refu
table, for it applies to all rocks necessarily. But is the
proposition itself true or false? Obviously it is true.
But is it true of rocks or only of words? Since in this
case the word "rocks" is understood to refer to actual
rocks, the proposition is true of rocks. Hence, it "says
something" about empirical objects, and yet it is not em
pirically refutable. Popper has complained that "logical
positivists" confused meaning and verifiability. In like
fashion, however, he seems to confuse "empirically true"
and "empirically refutable," defining the former in terms
of the latter, i.e., he says that a proposition can be em
pirically true only if it is amenable to possible empirical
refutation. But just as there is no good reason for the
dogma of the logical positivists, so there is none for this
dogma of Popper's. No justification can be provided for
the claim that logically true propositions do "not assert
anything whatsoever of any fact at all" (italics mine,
52:208) .
Furthermore, is his claim about logical truisms
itself a logical truism? Obviously it is not, for it pur
portedly makes an informative assertion--not only about one
kind of "fact" (viz., logically true propositions), but
also about the relationship that this "fact" has to sensory
or empirical facts. It is a statement about all facts of
a particular class (i.e., the class of logically true pro
positions) just as the proposition about rocks referred to
133
all members of the class of rocks. Nevertheless, to be in
formative, in Popper's own words, it would be required to
"take the risk of being falsified by (some) fact." Does
it take this risk? The answer can only be negative, for
what fact could possibly falsify it? Popper has defini-
tionally precluded the possibility of its refutation, but
in doing so he has inadvertently robbed it of its inform
ativeness. Since it is a universal statement about logical
truisms, it does not exclude any possible "fact" about the
members of that class.
Apparently, then, there is for Popper another class
of propositions that may be "true" besides physical truisms
and logical truisms. For whatever his proposition about
logical truisms may be called, it is proffered as a true
statement even though it is not empirically refutable
(which is his test for physical truisms) 0£ uninformative
about facts (which is his characterization of logical tru
isms). With his rejection of the knowability of essences,
he has no way of accounting for his alleged discovery of
this "truth.” He cannot base it on "self-evidence" for two
reasons. One is that he rejects self-evidence as totally
unreliable. The other reason is that self-evidence leads
to the contrary conclusion. If one reflectively examines
the relationship between logically true propositions and
the objects to which they refer, he will find that there is
both an oblique and a direct claim about facts. The
13^
oblique claim lies in the fact that every logical truism
presupposes some content, whether symbolic or nonsymbolic.
But equally as important as this consideration is the sim
ple insight that logically true propositions say something
directly about facts in that what they assert about things
cannot be meaningfully denied. What would it mean, for
example, to deny the assertion that "I am either typing now
or I am not typing now"? Even if it does not exclude some
facts, that does not mean that it cannot apply to any fact.
But it does exclude the possibility that "I am both typing
now and I am not typing now." If this latter exclusion is
necessary, it is because the law of noncontradiction is
neither an empirical generalization or a convention. That
mutually incompatible properties cannot belong to the same
thing is intellectually apprehended as a necessity, and the
formulation of this necessity in terms of the ontological
law of noncontradiction is the expression of this "fact"
and the ontological counterpart of the logical law of non
contradiction.
In one sense it is true to say, with Popper, that
logically true propositions are "trivial," but in. another
sense they are extremely important, for they exemplify a
generality that is crucial for a proper understanding of
both epistemology and ontology. What is trivial for men
in the common affairs of everyday life may be of surpassing
135
importance for the philosopher in his attempt to provide a
coherent account of knowledge and reality. As D. S. Mackay
says :
It is a truism but not an empty tautology to say. . .
that the very nature of existence is such that a thing
cannot be otherwise than it is; or . . . that whatever
is, is what it is and not another thing. Hence, . . .
logic is inseparable from metaphysics, since its prin
ciples and laws have ontological reference; and fur
thermore . . . logical implication is an element in the
real meaning of things, although not, of course, the
whole of it. (46:200)
The Truths of Arithmetic
What has been said about logic also applies in
principle to arithmetic. Popper made a sharp distinction
between two different interpretations of an arithmetical
proposition. On the one hand, he says that it can be in
terpreted as a logical truism, i.e., it can be taken to
have no reference whatever to any factual reality. On the
other hand, it may be taken to be factually descriptive,
and in that case, it is a physical theory. The former is
irrefutable, and the latter is not. Popper concludes,
then, that the calculi of arithmetic are not applicable to
reality except in the form of conjectures which may be em
pirically refuted.
Popper's account of the matter is hardly adequate.
While there is no question that empirical objects are sub
ject to alteration and decay, nevertheless, to the extent
that they are objects, they are distinguishable. If they
were not, then we could not say that they change, merge,
multiply, etc. These operations are understandable only In.
terms of numerical concepts. There must be two or more ob
jects (such as drops of water) if they can be said to
merge. While arithmetic does not result from induction, it
does seem to depend upon the presentation of distinguish
able things and aggregates in our consciousness. That is,
without distinguishable objects and classes of units,
"number" would not enable us to speak truly of reality in
the way it now does. For while logic and arithmetic are
not dependent upon any particular content, they must have
reference to some possible content. Thus, it can be said
that whatever is distinguishable is also numerable. Every
group of distinguishable members has a certain numerable
quantity. Let us consider a flock of sheep, for example.
If there are fifty sheep in the flock, it is a correct des
cription to say that there are fifty sheep, and it is in
correct to say that there are forty-nine sheep in the
flock. Arithmetic is also applicable in the sense that
its various calculi (such as addition, subtraction, multi
plication) are descriptive of some of the abstract rela
tions between the fifty sheep. For instance, we can be
sure that the flock is divisible into only five groups of
ten, two groups of twenty-five, etc. And this is just as
certain when applied to any group of fifty distinguishable
things.
Of course, one must distinguish pure number theo
rems from their application. The theorems tell us about
numbers, not actual objects: for example, that the number
"5" is greater than the number "4". As Husserl points out,
"they can be applied, not merely to individuals, but rather
to 'general' objects as well, e.g., to species of color and
tone, to types of geometrical figures, and so on" (32:4/8).
It cannot be defensibly maintained, then, that arithmetic
says nothing about the actual world, or that when it does
apply to the actual world, that it is then reduced to
nothing more than an indefinite number of physical theo
ries which are all open to possible empirical refutation.
Once again we see how Popper's theory of observa
tion and his rejection of knowable essences vitiate his
analysis of the applicability of arithmetic to the world.
He contends that since "in 'reality' something happens all
the time" (52:212), an equation like "2 + 2 = 4" cannot
hold for reality. While "2 + 2 = 4" is a necessary state
ment, "2 apples and 2 apples are four apples" is not,
according to Popper. The latter must be contingent, he
says, since it is "open to empirical refutation." But
surely if I put two apples on a table, and in the next
moment I put two more on the same table with them, it
follows necessarily that there are four apples on the
table (although admittedly not all empirical entities are
138
as free of problems related to countability). Popper errs
by thinking that "continual flux" in "reality" precludes
the applicability of necessary arithmetical propositions to
it. The important question here relates to the nature of
the flux. The only kind of empirical flux that would pre
vent the applicability of arithmetic would be a world in
which there was no differentiation. But it is difficult to
see how such a world could be knowable, for however else it
is conceived, it seems there must be at least the distinc
tion between knowing and what is known.
Furthermore, to say that "something is happening
in reality all the time" is to imply that there are things
(acts, objects, qualities, relations) that can be differ
entiated. Something happening or being in flux requires
change which must always be relative to something, and it
also demands that the flux and the things undergoing flux
be distinguishable. How could flux be known without some
stability? Here, then, are two distinguishable states of
affairs, and this means that the descriptive applicability
of arithmetic to the world is entirely justified. Hence,
it is grossly misleading to base the repudiation of arith
metic's applicability upon the muddled assumption that be
cause "reality" is always in flux nothing in it can be cal
culated or numbered with certainty. It is necessary to ask
if there are any states of affairs that are stable enough
139
to be numbered. And obviously there are, for the kinds of
things that are characterized by relatively persisting
self-identity are legion. On the basis of these distinc
tions., number and arithmetical calculi can be applied to
the actual world. And when they so apply, they yield em
pirically true descriptions that are necessary propositions.
It is not necessary that particular apples remain
apples eternally for them to be describable in arithmetical
terms. Although the process of decay will dissolve apples
into their constituent elements, this does not alter the
fact that during a certain period of time they are suscep
tible to arithmetical operations. Hence, if I can judge
truly that "2 apples and 2 apples are 4 apples," the sub
sequent digestion or decay of those apples does not change
the truth or necessity of that judgment. But if I say
truly that "There are four apples on a certain table," such
a statement at another time when something has happened to
change the number of the apples would obviously not be
true. Nevertheless, this does not mean that a true propo
sition once expressed is falsifiable by subsequent empiri
cal processes. Such processes do not--indeed, cannot--
change a true proposition into a false one. What they can
do is change empirical states of affairs so that they are
no longer what they once were.
Not every kind of cnange, however, is pertinent to
the truth or falsity of an arithmetical proposition or cal
culation. For example, if one is counting apples irrespec
tive of their colors, and truly judges that there are four
apples on the table before him, the arithmetical truth is
not altered by the fact that a slow and imperceptible
change is taking place from green to red. Consequently,
one can agree that in "reality" something is happening all
the time, and yet he may also recognize that not all that
is happening is relevant to an arithmetical proposition
expressed at a given time. What is required for our judg
ing truly the outcome of arithmetical operations with em
pirical objects is that we recognize which things retain,
and which things lose, their distinctiveness or self-iden
tity when brought into varying relationships. But this
recognition means that states of affairs have determinate
structures, and this is another way of saying that they
have "essences."
Popper contends, then, that when we add one empiri
cal entity to another, we can never be sure what the re
sult will be. Since we cannot base certainty on induction,
and since essences are unknowable, the way in which any
given entities will behave is unpredictable. He uses the
illustration of adding two drops of water to two drops of
water, with the result that there is only one drop of
water (52:211). Does this show that the calculi of arith-
metic are inapplicable to the world?
It should be observed, first of all, that since he
is certain that this will happen, he is conceding the fact
that there is at least this determinate structure in things.
Secondly, while it is true that only one drop of water re
sults from the act of adding four drops, it should be noted
that the one drop is not one of the same class as each of
the four is--for it is four times larger than any one drop
of the four (assuming, for the sake of argument, that they
are equal in size). And this very fact is itself describ-
able by a true and necessary arithmetical description,
viz., "two drops of water and two drops of water (all of
equal size) added to one another, equal one drop of water
four times as large as any one drop of water." Thus, we
can see that it does not support the notion that the appli
cability of arithmetic may result in the proposition that
"2 + 2 = 1," for the units on the different sides of the
equation are different, i.e., it is not "1 + 1 + 1 + 1
= 1" but rather "l +14-1 + 1 = 1+1 + 1 + 1," a proposi
tion in which all the units are equal. And we know that
this is necessarily true of drops of water in our world.
Thirdly, Popper's interpretation of the example of
drops of water does not support the alleged inapplicability
of arithmetic to the actual world, because it assumes a
limited view of distinguishability. He seems misled by
142
leaning too heavily on optical perception for his conten
tion that the four drops of water added to one another are
consequently indistinguishable. Let us take a different
approach. If we weigh each drop of water, employing an
apparatus to insure that each one is equal in weight, then
we can add together as many as we like and by weighing the
resultant "drop" of water, we can distinguish the constitu
tive drops of water. On this basis we can say truly that
"two drops of water and two drops of water equal four drops
of water." The important question, then, is how a "drop"
is to be defined.
Popper says that arithmetic is applicable to the
world to the extent that physical conditions resemble the
pure logical or arithmetical operation of addition. But he
has not presented one example in which physical conditions
do not resemble the arithmetical operation of adding. Of
course, the term "resemble" here is ambiguous. How can
physical conditions add or perform an operation that can
actually be like adding? Human beings add things by rec
ognition of distinctions and by intentional grouping of
things that are distinguished. Now if we interpret the
term in the most favorable way for Popper's position, we
find that it seems to mean that characteristic about physi
cal conditions that may be called "stability." Popper
seems to be saying, then, that to the extent physical con
143
ditions are stable., to the same extent does arithmetic re
main applicable. But even this is to say too much. It is
conceivable that I and objects within my purview are con
stantly and noticeably moving ("instability") but that
arithmetic is still applicable to this state of affairs.
So it is better to say that arithmetic is applicable to the
extent things in the world are distinguishable. Hence,
since arithmetic is applicable to the world, distinctions
of some kind are discoverable in the world. Moreover,
since arithmetical propositions are necessary, then some
necessary statements can be made about the world.
If this is true, then contrary to Popper, it is not
a trivial statement that says that arithmetic is applicable
to the extent the world "resembles" (i.e., contains dis
tinctions like) arithmetical truths. If the implications
of this latter admission are carefully analyzed, it will
be seen that Popper's own neat classification of proposi
tions is indefensible. He seems to maintain a sharp divi
sion between statements that are (l) descriptive and there
fore refutable, and (2) logically true and therefore irref
utable (52:210). Against his view I have argued that
there are general descriptive propositions (logical and
arithmetical) that are irrefutable. Refutability, then,
should not be interpreted to mean that since "something is
always happening" to a given state of affairs, the possi-
144
billty of knowably true descriptive propositions is pre
cluded. If this were the case, then no empirical proposi
tion could ever be made that could justifiably be called
true.
Conclusion
It appears that Popper's unclear notion of refuta
bility seriously mars his interpretation of the applicabil
ity of logic and arithmetic. For instance, he does not
hold that a descriptive proposition is refutable because it
has incorrectly described a given state of affairs and can
be replaced by one that correctly describes it. For he
uses the term "refutable" to refer to every statement that
is empirical or allegedly descriptive of the world (52:
212). That is, one can never know if a statement correctly
describes any state of affairs--as Popper clearly claims in
his discussion of "basic statements." But if, as I have
argued, a statement can be a true description of a given
state of affairs and can be known to be true, then it is
confusing to say that that statement is "refutable" simply
because it has an empirical reference and empirical condi
tions are always "changing." Contrary to Popper (52:211-
12), the truth-value of a descriptive proposition is not
changed by changes in empirical conditions. A proposition
cannot be called false because it no longer applies as a
true description. Clearly, a proposition which correctly
describes a given state of affairs is not refutable or cor
rectable if that state of affairs is replaced by another
state of affairs. The truth of a given descriptive propo
sition is relative to the particular state of affairs to
which it refers. If I correctly describe a state of af
fairs by the statement, "There are ten acorns in a basket
in my backyard," and at some later time squirrels eat five
of them, it does not follow that the proposition is falsi
fied. The proposition can be said to be irrefutable be
cause it has been found to be a true description of a given
state of affairs. A proposition that is held to be true
but is actually false is refutable, e.g., I may say, "There
are nine acorns in the basket in my backyard," but on
closer inspection I find that there are ten. The important
question, therefore, is not whether a proposition refers to
static or changing states of affairs, but whether truth can
be known at all (i.e., whether a proposition can correctly
describe any state of affairs), and if so, how it can be
known.
The distinction between changing empirical condi
tions and unchanging truth-value must be drawn if we are
to avoid a muddled and self-defeating epistemological posi
tion. And once this is done, it will no longer be possible
to justify the assertions that "if a proposition is des
criptive of the world, then it is refutable," and "if a
146
proposition is irrefutable, then it is not descriptive of
the world,"--or, in Popper's own words, "Only if we are
ready to accept refutations do we speak about reality"
(52:212) .
That there are irrefutable descriptive propositions
means that necessity is not restricted to a purely formal
or logical sphere. Moreover, while logical principles and
arithmetical propositions tell us something about meanings
and classes, they also provide us with some knowledge about
the world, or "reality." As J. Loewenberg says:
Knowledge is nothing if not ontological. Its reference
to the real is twofold: the real is its prior object,
and the real is its ultimate objective. But the status
of the real, whether as prior object or as ultimate ob
jective, is a question with which reason alone can
grapple. . . . For better or worse, reason remains our
court of appeal even against reason's extravagances.
We cannot go behind or beyond reason without reducing
cognition to sentience. And sentience, whether stupid
or ecstatic is indeed acquiescence, acquiescence pure
and absolute and silent. (44:72)
CHAPTER IV
HOW ARE THE TRUTHS OF LOGIC KNOWN?
The thesis of the general, descriptive applicabili
ty of logic to the world has an important corollary, viz.,
that logic cannot be understood properly without consider
ing how the truths of logic are known. The "truths of
logic" should be distinguished from "logical truths." Ex
amples of logical truths are the tautologies, "All trees
are trees," "If all clouds move, then all clouds move," and
"Either it is snowing or it is not snowing." These sen
tences express propositions that are true by virtue of
their "form," and their denial would involve obvious self-
contradictions. Examples of truths of logic are "A propo
sition is either true or false, but not both,” "Modus tol-
lens is a valid rule of inference," "All Barbara syllogisms
are valid," and "'Either it is snowing or it is not snow
ing' is true by virtue of its form." The truths of logic,
then, are informative of certain relationships that neces
sarily hold between general meanings, or universals.
While it is possible to distinguish epistemological
and ontological factors in analyzing the nature of logic,
147
148
It seems Impossible to separate them altogether. The epis-
temological aspects of the problem are indicated in the
question, "How do we come to know the basic principles and
rules of logic?" The ontological aspects of the problem
find expression in the question, "Do logical principles
have an ontological reference or counterpart, and if so,
what is the nature of the relationship between logic and
ontology?" The answer given to one of these questions will
undoubtedly have bearing on the answer to the other one.
The inseparability of logic and ontology can readi
ly be seen in four main answers given to the question,
"How are the truths of logic known?"
(1) They are known by examining the psychological
processes of human thinking, i.e.-, they are psychological
generalizations of the way in which we think. (a) They are
natural laws of thought, i.e., descriptions of the way in
which we must think. (B) They are formulations of conjec
tures which we freely "impose" upon the world and upon the
structure of our arguments.
(2) They are known by investigating the most gener
al laws of nature, i.e., they are empirical generalizations
based upon sense experience. (A) They are "imposed" upon
our thinking as normative laws, i.e., as laws of empirical
necessity, and they may or may not be taken as defining the
conditions for the possibility of any empirical world. (B)
149
They are "abstracted" from empirical states of affairs, and
they may be changed by a fundamental change in the struc
ture of reality.
(3) They are known by analyzing the structure of
our language, i.e., they are generalizations about the way
in which we use language. (A) They are mere conventions,
i.e., they can be altered by the construction of different
languages or symbolic systems. (B) They are "imposed" upon,
us by our language, i.e., they cannot be altered or aban
doned any more than we can dispense with our language.
(4) They are known intuitively and phenomenologi-
cally, i.e., they are objects of direct intellectual appre
hension and reflection. (A) They are only "ideal," i.e.,
they are limited to meanings and relations between mean
ings, and they do not have ontological reference apart from
our "imposition" of them upon all experience. (B) They are
both "ideal" and "real," i.e., they are not only character
istic of meanings and relations between meanings, but they
also have an ontological reference or counterpart in the
world. The former (i.e., the "ideal" sense) can be illus
trated by the logical principle of noncontradiction, which
states that a proposition cannot both be true and not true
in the same respect. The ontological principle is ex
pressed by saying that a thing cannot both have and not
have a certain attribute in the same respect and at the
150
same time.
Bartley's Skeptical View of Logic
Bartley does not provide any explicit answer to the
question under consideration, because he apparently does
not believe the question is important. Following Popper,
he minimizes the value of asking for the source or evidence
for any alleged piece of knowledge. For instance, Popper
says, "I propose to replace . . . the question of the
sources of our knowledge by the entirely different ques
tion: 'How can we hope to detect and eliminate error?'"
(52:25). Similarly, Bartley is more interested in criti
cizing claims to truth--in fact, his entire philosophical
program is constructed around the single goal of eliminat
ing error.
The demand of Bartley's "comprehensively critical
rationalism," that every proposition be open to possible
refutation, means that the truths of logic have no perma
nent status. They might be refuted at any time. It has
been shown that this is impossible, however, for the law of
noncontradiction cannot be employed to refute itself, and
neither can the principle of deducibility or retransmissi-
bility be used to refute itself. Nevertheless, Bartley's
view involves the assumption that as long as there is the
mere possibility that the ontological structure of things
could be radically changed, there is the possibility that
151
the logical and ontological laws of noncontradiction will
no longer hold. If he maintains that both of these laws
are apprehended only In conjunction with our cognitive
grasp of things, this would involve him In self-contradic
tion. For he would thereby imply that there is some truth
(viz., that the laws of noncontradiction are ontologically
"determined," whether positively or negatively) which is
not open to refutation, and this would destroy the integ
rity of his "comprehensively critical rationalism."
Furthermore, to say that one state of affairs is
different from another is to presuppose that the ontologi
cal law of noncontradiction holds, at least as an overarch
ing principle. This can be explained by drawing an analogy
between individual objects which cannot have incompatible
properties and the totality of objects at any given time,
of which it cannot be said that the ontological law of non
contradiction both does and does not hold. In other words,
let us say that "reality" has ontological structure A_, but
that it is displaced by ontological structure B. In A_ the
ontological law of noncontradiction holds, while in B it
does not.
First of all, it is difficult to see how B can be
called a structure without the ontological law of noncon
tradiction being exemplified in it. But let us grant, for
the sake of argument, that an "unstructured" B is postulat
ed. Now if the law of noncontradiction holds for A and not
152
for B, yet insofar as A_, though superseded, obtained at all,
it was "part" of reality rn its most comprehensive sense,
i.e., including all "successive" ontological states of af
fairs, such as A_ and B. Now insofar as B is contrasted
with A, B, as a totality of "unstructured" objects, is
thereby subject to the ontological law of noncontradiction,
because A_ and B are incompatible states of affairs. That
things can be distinguished only if the ontological law of
noncontradiction holds seems to follow from the fact that
just as a logical contradiction entails every proposition
(and therefore none), so if everything can both have and
not have a certain attribute, that would mean that some
thing can have every attribute (e.g., an object could be
both red and not red in the same way at the same time), and
therefore, that it cannot be characterized or distinguished
from other things at all.
The ontological law of noncontradiction cannot both
obtain and not obtain i_n the same state of affairs, then,
and this means that A_ and B are distinguishable within
reality (in the most comprehensive sense) because the on
tological law of noncontradiction holds for them as subsid
iary totalities. If it did not, B could not even be postu
lated as a different ontological system in contradistinc
tion to A_. The only way B can be B and not A_ is on the
basis that there are actual incompatibilities in reality.
153
It seems, then, that we grasp the necessity of the
ontological law of noncontradiction hy an Intuitive aware
ness that incompatible properties cannot belong to the same
thing. And with respect to the issue under consideration,
I am not merely speaking about incompatibilities that may
or may not obtain within A_ and B, but also about the incom
patibility of A_ and B with each other. That they are dis
tinguishable means that the ontological law of noncontra
diction is indispensable, for without it there would be no
difference between A_ and B. And that would mean that A_
could be B and B could be A_. Moreover, there would then be
no reason for Bartley or anyone else to think that the on
tological system that now obtains is A_ and not B.
If this analysis is correct, it supports the con
tention that the necessity of the logical and ontological
laws of noncontradiction is not due to psychological com
pulsion but to its own givenness which we .apprehend in the
nature of things. Neither epistemological or ontological
considerations support attempts to interpret the laws as
refutable. There are only two alternatives: either the
laws are refutable or they are not. Since the case for
their refutability is found to be unsustainable, the only
alternative is that the laws are irrefutable. Their uni
versality and necessity seem to be supported by the fact
that they are not reducible to psychological characteris
tics but have a distinctively logical and ontological
154
status.
In addition to Bartley's relatively covert skepti
cism with respect to the principles of logic and the onto
logical law of noncontradiction, there is the more avowed
skepticism of Pyrrhonism. This position is expressed by
saying, "I cannot be sure that the principles of logic are
true; I can only guess. And I can only guess that I guess
that I guess ad infinitum."
Nevertheless, to say this, one must be able to con
ceive of the principles of logic and to distinguish them
from other propositions. Furthermore, he must distinguish
an infinite regress from a finite regress. All of this
presupposes the logical law of noncontradiction. So does
his interest in formulating a "perfectly consistent" theory
of skepticism. He mistakenly insists on an infinite re
gress as the only way of avoiding inconsistency. But why
should one care to avoid inconsistency when the status of
consistency as a necessary criterion is alleged to be with
out foundation? His profession of tentativeness, there
fore, is hollow. He dogmatically proclaims a skeptical
infinite regress and he tenaciously adheres to it. Thus,
he lives and dies protesting that his attachment to it is
only provisional, while in reality he could not have
thought or spoken rationally without presupposing the logi
cal law of noncontradiction.
Is it possible for anyone who can grasp the meaning
of the concepts in two mutually contradictory propositions
to say sincerely that he does not know that both of them
cannot be true? Similarly, can he seriously say that he
does not know that the conclusion of a valid Barbara syllo
gism cannot be false if the premises are true? Yet even
here there are professed skeptics who say that they doubt.
But such a skeptic "knows" that he doubts, and he is able
to demarcate doubt from certainty, his view from other
views, and this itself is the undoing of his professed
skepticism. Surely a difference that makes no difference
is no difference. But then it does seem to have at least
this consequence: it is unlikely that such resolute skepti
cism will allow one to take seriously the question, "What
significance does the irrefutability of the logical and
ontological laws of noncontradiction have for understanding
the nature of the world and man's place in it?" For the
Pyrrhonist professes not only that he cannot be sure of the
truth of elementary logical or ontological principles but
also that he cannot be certain that there is a universe and
whether man exists at all. Such a denial of all possibili
ty of discovering the truth destroys justification and mo
tivation for all human inquiry, whether philosophical or
scientific.
The essential point of my argument against skepti
cism is that it is self-contradictory, for it denies itself
156
in the very assertions by which it attempts to affirm it
self. That is, since it denies what belongs to the meaning
of every proposition, it denies the very propositions in
which it makes its denial. Hence, radical skepticism can
not have consistent meaning. The skeptic may refuse to be
persuaded, but that does not alter the validity of the ref
utation of his skepticism. Subjectively, or psychological
ly, the skeptic may remain unconvinced, but objectively, or
logically, his position may be refuted. Husserl expresses
the argument against skepticism in the following way:
The strongest objection which can be leveled at a
theory, and especially if it be a theory about logic,
is that it rejects the obvious conditions of the possi
bility of theory in general. To formulate a theory and
in its content deny, explicitly or implicitly, the
principles which provide the basis of the sense and
justification of all theory whatever--that is not mere
ly false, but rather is fundamentally absurd. (32:7/1)
Popper's Fideistic View of Logic
Popper's answer to the question, "How are the
truths of logic known?" is not too dissimilar from the
skeptic's reply. Popper says: "My rationalism is not self-
contained, but rests on an irrational faith in the attitude
of reasonableness" (52:357). He avers that men have a
choice between rationality and irrationality, and that his
choice of rationality is a matter of faith alone. More
over, the man who decides for irrationality cannot be ar
gued with or convinced that he is wrong, for critical argu-
157
ment presupposes rationality, and that is precisely what
the "irrationalist" will not accept.
There are two considerations, however, that under
mine Popper's view. First, can anyone choose rationality
or irrationality? A choice requires conceptual discrimina
tion between alternatives, and this discrimination is im
possible without the logical law of noncontradiction. That
rationality is not the same as irrationality, and, in fact,
that they are mutually exclusive, is due to the principle
of noncontradiction. Therefore, no genuine choice can be
made between the two, for if one is already presupposing
the principle of noncontradiction, it makes no intelligible
sense to say he is choosing it or rejecting it. And if a
man repeatedly and unwittingly violates the law of noncon
tradiction in his thinking and speaking, and cannot under
stand his logical error when it is brought to his atten
tion, then again it can make no intelligible sense to say
he is choosing irrationality. Hence, it is a fiction that
anyone can choose irrationality in the sense that he can
deliberately adopt a position that defies the principle of
noncontradiction. If this is the case, then it is also a
fiction to say that a man chooses rationality in the sense
that he can deliberately adopt the principle of noncontra
diction. Consequently, fideism, with the principle of non
contradiction as its object, is an impossibility.
158
The second vulnerability in Popper's position is
that it says nothing about how the principle of noncontra
diction can be an object of reflection in the first place*
so that it can be an alleged alternative for our "voluntary
faith." Enough arguments have been traversed for us to see
that it is not a "conjecture*" spontaneously excogitated by
the human mind in the course of history. A precondition
for the formulation of the principle is the principle it
self* i.e.* it is necessary that the meanings of concepts
be distinguishable and that propositions (e.g.* as ex
pressed by the formulation in question) cannot be both true
and false. The principle of noncontradiction did not begin
to obtain when it was explicitly recognized or formulated
in the course of human history. And neither is it a con
vention* for we cannot change it at will* nor can we con
ceive of it being "changed" without presupposing it. Since
Popper has rejected "intuition" and "self-evidence*" there
is nothing left for him to say but that one is simply con
fronted with it as part of "human tradition*" and thereupon
he decides to have faith in it or to repudiate it.
When Popper calls his "faith" in rationality an
"irrational" decision* he is saying that in_ the final anal
ysis we are all irrationalists and that the only difference
between "rationalism" and "irrationalism" is an uncertain
and relative one. This confusion is the inevitable result
159
of his failure to see that the principle of noncontradic
tion. is not chosen but intuitively grasped. And, thus,
since it is not a matter of faith or choice at all, it can
not properly be called an "irrational decision."
Justifying Rules of Logic
How, then, can we know the truths of logic? Now
the term "truths of logic" includes both the elementary
principles of logic (e.g., the law of noncontradiction) and
the propositions that serve to justify rules of inference.
To clarify the distinction between propositions and rules
of inference, let us take a specific example. The rule
called modus ponens is often formulated in the following
manner: "p, and if p then q, imply q. " However, this is a
proposition or statement, and not a rule. A rule tells us
to do something. It is differentiated from a command, in
that the latter usually lacks the generality of rules.
Rules are general directives to be followed in all cases
characterized by a specific set of conditions. They cannot
be evaluated in terms of truth and falsity, but they can be'
assessed in the way methods and plans are evaluated, viz.,
in terms of their effectiveness in realizing the end of a
given activity to which they are relevant. In logical cal
culi, then., rules of inference are directives for trans
forming sentences so that by following them, it is impossi
ble to derive false propositions from true ones. There
160
fore, modus ponens should be expressed as follows: "Given
that p is true, and if p is true, then q is true, assert
q. " An analogy from arithmetic will further elucidate the
distinction. The proposition "2 + 5 = 7" is not a rule,
for it does not prescribe some course of action or of
thought. The corresponding rule of calculation would be as
follows: "Given '2 + 5', assert '7'." Here we are told to
do something. Now to return to modus ponens, is its direc
tive justified? It is if it guarantees that its observance
will never allow us to derive false conclusions from true
premises.
How, then, is this guarantee established? It cer
tainly cannot be grounded in (l) psychological processes,
or (2) empirical generalizations from sensory experience,
or (3) conventions. The first two alternatives do not
yield the necessity and universality required, and the
third alternative yields arbitrariness by definition.
There seems to be no other answer to the question available
except the following: the universal and necessary truths of
logic are known by insight into essential relations which
hold between meanings. The rule of modus ponens is based
on the self-evidence (Evidenz) of the corresponding prin
ciple or judgment about the relations that invariably ob
tain between the meanings in such an argument. The term
"Evidenz" here means the perception of the agreement be
l6l
tween a proposition and what it refers to or intends* or*
in Husserl's words* "the pure viewing and grasping of some
thing objective directly and in itself" (29:6).
Now in considering the rule called modus ponens in
order to clarify the process of justifying logical rules*
it will be helpful to distinguish four aspects of the pro
cedure. (1) The conditions specified by the rule in ques
tion must be recognized. The conclusion* q* is to be as
serted when a certain proposition* p* is true* and when p
and q are so related that if p_ is true* q is true. (2)
Each of the propositions* p_ and q* cannot be both true and
false. (3) Since there are only two possible values for
each proposition* there are only four possible combinations
of the two values of the two propositions: (a) p is true
and q is true* (b) p is false and q is true* (c) p is true
and q is false* (d) p is false and q is false. (4) Accord
ing to (l) above* the required condition that p be true ex
cludes the possibility of (b) and (d)* and the specified
condition that "if p is true* then q is true" excludes the
possibility of (c). Since (a) is the only remaining possi
bility* and since it asserts that q is true* the rule of
modus ponens is a good rule* i.e.* it guarantees that its
application under the specified conditions will achieve its
purpose* namely, it will lead only to truth and never to
error. And by employing the rule* we can increase our
162
knowledge. Thus, inferences drawn according to modus
ponens are formally valid because the rule is justified.
Its justification is established, as D. A. Willard says,
"by reference to a _ true judgment about the judgments occur-
ring in the type of inference which the rule is proposed to
govern" (65:8).
In the final analysis, then, we must come to in
sight into the self-givenness of the truth of judgments
which justify corresponding rules. This makes the formula
tion of valid rules of inference possible. And with refer
ence to the logic of propositions, rigorous demonstrations
of validity can be provided by truth-value tables. But of
course the truth or falsity of specific propositions does
not determine validity. The premises and conclusion could
all be false, and yet the inference may be valid. In gen
eral, the logician is not interested in the question wheth
er the premises are in fact true. Rather, he is concerned
with the validity of inferring a conclusion if_ the premises
are true. Hence, an argument is valid if, and only if, the
conclusion is a logical consequence of the premises. Of
course, entailment is only one of a number of logical rela
tions. In general, then, logic is concerned with the rela
tions between propositions in virtue of which the possibil
ities of truth-values in the related propositions are re
stricted in certain ways.
163
Intuitional versus Nonlntuitional Claims
Of the four main positions on the question of know
ing the truths of logic, it seems that only one adequately
accounts for the necessity that characterizes them and the
concomitant "constraint" that we encounter when using them
and reflecting upon them. If they were psychological or
empirical generalizations, arrived at inductively, then
they would be only probable. But no generalization about
our psychology or about our empirical experience has the
same necessity as have the basic truths of logic. We could
be psychologically different, and the world could be empir
ically different, but no matter how great the difference,
it seems that the logical and ontological laws of noncon
tradiction would still obtain, just as "2 + 2 = 4" would
seem to hold in every possible world. As arithmetic does
not deal with the psychological operations of counting but
with numbers, so logic is not concerned with psychological
processes of judging but with ideal entities, such as con
cept, proposition, conclusion, etc. Moreover, while par
ticular instances of empirical generalizations are more
certain than the generalizations themselves, this is not
the case with the truths of logic and arithmetic. But of
course if one begins with the assumption that the only
realities are particular existents, and that all we can
know are particular "facts," then he has to attempt to
164
account for the truths of logic and arithmetic on that ba
sis, i.e., inductively. This would appear to be a case of
an erroneous ontology and epistemology determining a view
of logic.
Furthermore, it is to misunderstand the law of non
contradiction to say that it is simply the psychological
fact that a person cannot believe a proposition and its
denial at the same time. It is also to misconstrue the na
ture of logical validity to say that it is the psychologi
cal inability of a person to believe that the premises of
an argument are true and the conclusion is false. For it
is not only possible for persons to believe contradictory
propositions at the same time, there is abundant evidence
that they do. One of the great legacies of Husserl's early
phenomenological writings is his refutation of reduction-
istic psychologism and his demonstration that logic is not
concerned with what we can or cannot believe, but with
meaning, validity, and truth and falsity. Whether or not
a proposition is true does not depend upon whether or not
I believe it. And whether or not an argument is valid does
not depend upon whether or not I accept its inference. My
believing or accepting is a psychological event or experi
ence through which I live. But the intentional character
of judgments (which are acts of consciousness) is the pro
position, which is universal and timeless, and therefore it
can be the "one in the many," i.e., the one meaning or in
tention of many judgments (32:11/3/3). And, as such, pro
positions should be distinguished from experiences, for
propositions, not judgments, are true or false, and they
are formed out of concepts, not out of psychical acts (32:
l/3/lO). Minimally, then, we must distinguish concepts and
propositions from the acts of judging if we are to be able
to turn our attention to the intentions and objects of con
sciousness. When we do, we find that the principles of
logic and arithmetic are not empirical generalizations but
a priori truths. As Husserl remarks, "validity is some
thing which has no essential relationship to the person
empirically judging"(32:1/3/10). In this way the autonomy
and objectivity of the ideal objects of knowledge can be
acknowledged along with the subjectivity of knowing.
The inevitable result of psychologism, however, is
the reduction of all knowledge to the vagueness and incon
clusiveness of empirical processes. But the vagueness and
inconclusiveness of the laws of psychology cannot account
for the exactness and absoluteness found in the basic laws
of logic and arithmetic. This difference is unequivocally
revealed by a phenomenological analysis of the two kinds of
laws in question (viz., psychological and logical laws).
Moreover, they are seen to be disparate in terms of the
ways in which they are established. Psychological laws,
irrespective of their approximation of exactness, are
166
Inductively based on particular experiential facts, whereas
logical laws are known by insight as a priori truths (32:
4/lff.). This is why Husserl is uncompromisingly opposed
to the psychologistic confusions of "Evidenz and blind per
suasion; exact and empirical generality; logical incompati
bility of propositions and psychological incompatibility of
acts of belief; and hence, the impossibility of being true
together and the impossibility of being simultaneously be
lieved ..." (32:6/3). We know by apodictic Evidenz that
the law of noncontradiction is absolutely exact, and that
it is not a matter of probability but of necessity that of
two contradictory propositions one is true and the other is
false (32:5/9-10).
If the law of noncontradiction were a psychological
or empirical generalization, then no propositions could be
known with certainty to be true, for all meaningful propo
sitions presuppose the absolute correctness of the law.
If all propositions are only probable, then they can never
be known with certainty to be true or false, for probabili
ty never yields certainty. And even the assertion that all
propositions are only probable must be only probable. But
this, too, is a proposition, so it is also only probable,
and this implies an infinite regress. We have already seen
that, as in the case of radical skepticism, this is a self-
defeating position. Hence, one's view of the nature of the
167
principles of logic Is crucial to his entire perspective on
epistemology and ontology. If the laws which constitute
the necessary and ultimate foundation of all knowledge are
not exacts absolute; and objective--in short, if they are
not apodictically evident-then skepticism is unavoidable.
Logic, Language, and Reality
In Husserl's analysis of the problem under consi
deration, we have a clear example of the importance of ex
amining the sources (i.e., the rational evidence or
grounds) of our knowledge. And in view of his findings,
the indifference of Popper and Bartley to the question of
the sources of human knowledge is shown to be a serious
misjudgment. That their assumption is disastrous for both
epistemology and ontology is seen in their inability to
formulate a consistent and practicable nonjustificational-
ism in science or philosophy. Once it is recognized that
the fundamental principles of logic do not have their
source in psychological processes, or in empirical gener
alizations about the world, or in conventions based upon
language or conjecture, but rather are given as ideal, ob
jective truths which are known by immediate insight, then
these principles can serve as criteria for the rest of our
knowledge (i.e., the justification of all mediate knowledge
depends upon these fundamental principles). And by defini
tion, this is justificationalism.
168
Moreover, this form of presuppositlonless justifi-
cationalism overcomes what Bartley's allegedly presupposi
tlonless nonjustificationalism unsuccessfully tried to
avoid, viz., a vicious infinite regress (i.e., skepticism)
and the arbitrary adoption of presuppositions (i.e., fide-
ism). For apodictic Evidenz not only removes the need for
these alternatives, but it also precludes their very possi
bility. It must be made clear, of course, that the basic
principles of logic are necessary, but not always suffi
cient, criteria by which to judge claims to truth and know
ledge. Nevertheless, demonstrating their necessity is
enough to confute the possibility of an arrant nonjustifi-
cationalism in philosophy, and to establish the grounds for
justificationalism.
While the necessity of logical principles is all
that is needed to establish justificationalism, ontological
considerations also seem to support it. Aristotle pointed
out long ago that the meaningful assertion of anything in
language must presuppose that the logical law of noncontra
diction is true, and therefore the denial of the law is
self-contradictory. If it is supposed, however, that the
principles and rules of logic have no status apart from
their recognition and employment in human consciousness,
then all justification in philosophy would seem to be rela
tive to human consciousness. This would mean that if human
consciousness had not occurred at all, or if it had struc
169
tures other than those It has, then it may be that "the
ultimately real" is not characterized by laws which are the
counterparts of the principles of logic which we find to be
"normative" in our thinking. It appears that Aristotle was
aware of this problem, however, for he begins with things
and not with psychological processes or linguistic struc
tures. That is, he says that "the most certain of all
principles" is that "the same attribute cannot at the same
time belong to the same subject and in the same respect"
(5:524). For Aristotle, then, the ontological law of non
contradiction is basic to the possibility of knowledge. He
argues that the denial of the law would entail the possi
bility that everything is equally everything else.
Moreover, if the logical and ontological laws of
noncontradiction were dependent upon our subjective nature,
then it would seem that a change in our nature could result
in their negation. But such an interpretation would make
them merely probable, and the certainty that propositions
and things conform to them cannot be accounted for on that
basis. Our recognition of the laws is not merely a recog
nition that our consciousness is so structured that it
"must" or should think in accordance with them. This lat
ter recognition presupposes the truth of the laws. While
failing to clearly distinguish between the logical and on
tological laws of noncontradiction, Bertrand Russell once
expressed an important insight, which can be more fully
170
appreciated in the context of their distinction:
The belief in the law of contradiction is a belief
about things, not only about thoughts. It is not,
e.g., the belief that if we think a certain tree is a
beech, we cannot at the same time think that it is not
a beech; it is the belief that if the tree is_ a beech,
it cannot at the same time be_ not a beech .... If
this, which we believe when we believe the law of con
tradiction, were not true of the things in the world,
the fact that we were compelled to think it true would
not save the law of contradiction from being false; and
this shows that the law is not a law of thought.
(59:89)
It seems, then, that there is no reason that sup
ports the view that our minds make either the logical or
the ontological law of noncontradiction true. Even if one
assumes solipsism, the objects of his knowledge may be
"subjective," but they are also "real." He can distinguish
between consciousness and its "immanent objects," and the
distinctions introduce both arithmetic and the law of non
contradiction into "reality." Even if one takes the view
that reality is a monolithic unity and that all distinc
tions are illusory, he still confronts the problem of try
ing to dissolve the distinction implied in the assertion of
his view, viz., the distinction between the "illusion" that
he has of a multiplicity of distinguishable objects and the
undifferentiated reality that he postulates. That this
problem cannot be overcome shows that his ontological theo
ry cannot be stated with consistent meaning. The assertion
that all knowledge is illusory is self-contradictory, for
it presupposes that the assertion itself is not illusory.
171
It seems that there are good reasons for maintain
ing that logic has implications for ontology and that onto
logy has implications for logic. One view of the nature of
logic, however,, interprets the truths of logic as rules for
the proper use of a given language,, whether natural or ar
tificial. Nonetheless, the truths of logic are not about
words and sentences, but about meanings, i.e., concepts
(referential qualities with a single unitary meaning), pro
positions, and their relations. Now of the various charac
teristics of language, its most important feature is its
use to mean something beyond itself. Apart from intention
al qualities (i.e., "aboutness" or "consciousness of"),
language, which is only a system of empirical entities
called words or sentences, could not be used to refer to
something. Such manipulation of mere words would be an
exercise in acoustics or optics, but it would not involve
the communication of meaning at all.
Meanings are intentional qualities which are in
stanced in thoughts, and as such they are universals.
Meaning, or "to mean something," is an act of intention,
and as such it is characteristic only of minds or persons.
Mere words or sentences cannot mean anymore than raindrops
or rocks can mean. Therefore, meaning cannot be reduced to
the mere use of words (which are individual sounds or
marks), for meaning is a necessary condition of using words
to direct thought to something beyond themselves.
172
While it is indisputable that the symbols and gram
mar of particular languages are conventional, languages
fulfill their function as instruments of meaning and com
munication by meeting certain conditions that are universal
and necessary. The basic principles of logic are an indis
pensable part of these conditions. Rather than language
being the foundation for the soundness of the truths of
logic, it is the reverse that actually obtains. The con
ventionalism of language does not determine whether propo
sitions are true or false, or that a Barbara syllogism is
valid or invalid. And certainly the use of a language or
semantic system cannot dispense with the law of noncontra
diction. That law is a necessary condition both of propo-
sitional knowledge and systematic coherence.
Moreover, that there are certain necessary rela
tions among meanings is shown, for example, by the fact
that the Barbara syllogism exemplifies a necessary rela
tionship between the universals (concepts and propositions)
of which it is comprised. Accordingly, Husserl says:
The mode "Barbara" does not say that, when "All A1s are
B1 s" and "All B1s are C1s" are true, it is to be con
jectured that the corresponding proposition, "All A1s
are C's," is true .... Otherwise, we leave open the
possibility that the conjecture in question might be
disconfirmed by the broadening of the range of experi
ence, which always is only a limited one. (32:4/2)
The validity of the Barbara syllogism is objective, and one
has insight into the truths of logic and direct awareness
173
of certain given qualities and relations between meanings,
without which language and reasoning could not function as
instruments of communication and argumentation.
An example of the awareness of qualities and rela
tions is the differentiation of red from green, or the fact
that something is rectangular or not rectangular but not
both. Such insights are necessary conditions of reasoning
from premises to conclusions. Since there are relations
that are necessary, they cannot be accounted for by stipu
lation or linguistic convention.
Moreover, certain logical principles (e.g., the law
of noncontradiction) seem to have ontological counterparts
which constitute some of the general features of the world.
It seems that if a proposition cannot be both true and
false, then what it refers to cannot both have and not have
the property attributed to it by the proposition. To argue
that something can both have and not have an attribute is
to imply that it is possible for a proposition to be both
true and false. And this we have seen to be impossible.
For example, if a tree cannot both bear fruit and not bear
fruit, then the proposition that asserts that it cannot
both bear fruit and not bear fruit must be true. If the
ontological law of noncontradiction is true, then the logi
cal law of noncontradiction is true also; and if the logi
cal law of noncontradiction is true, then the ontological
law of noncontradiction must be true, too. If this were
174
not the case, It would be possible for any given proposi
tion to be both true and false despite the fact that the
thing it refers to is found to have a given attribute,
thereby excluding the possibility that it does not have
that attribute. Similarly, it would be possible for some
thing both to have and not to have an attribute, despite
the fact that the proposition which asserts that it cannot
both have and not have a given attribute is necessarily
true.
Conclusion
If the foregoing arguments are essentially correct,
then of the various attempts to answer the question, "How
are the truths of logic known?" all prove untenable except
the last one, namely, that they are known intuitively and
phenomenologically. That logical principles are ascer
tained phenomenologically means that reflection upon the
"meaning-contents" of our knowledge can bring the princi
ples to light and thus allow for their formulation as logi
cal laws. Thus, the primitive principles are intuited and
formulated before other truths of logic are derived from
them.
Logic cannot be reduced to psychology, or sensa
tions, or conventions. It is grounded in the very nature
of the ideal entities which it mentions. But this does not
mean that logic is separable from ontology, for our con
175
cepts and propositions are referential, and as such they
are about objects of various kinds. Generally, then, the
objects of our belief are not propositions; rather, they
are the objects (whether ideal or empirical) which are re
ferred to by propositions. This seems to indicate one kind
of connection between logic and ontology.
By way of summary, then, the justifying ground of
the laws of pure logic is their insightful givenness as
exact truths which exclude other possibilities. And since
science is a system of knowledge connected by logical rea
soning, logic is the foundation for all other knowledge.
Logic itself, being based on insight and reasoning (reflec
tion and proof), is a science, viz., the science of the
principles of valid inference, and hence it is concerned
with describing the conditions under which Evidenz may
arise.
What Husserl says about "extreme Empiricism" can
be applied to "logical conventionalism" as well: "Whoever
logically thinks it out to the end will have already aban
doned it when he gets there. If not, then . . . [it] would
provide a noteworthy example of how much stronger rooted
bias can be than the clearest testimony of insight" (32:
5/1). The unrestricted generality and absolute exactness
of the basic principles of logic are characteristics that
remain inexplicable on any basis other than immediate in
sight into given truths. And in view of this, Bartley's
assumption that all justificationalism is caught in the
dilemma of fideism and skepticism is misconceived. There
is a viable justificationalism that transcends the so-
called "presuppositional predicament." In the next chapter
I shall consider the question of presuppositions and the
resolution of it by insightful justification.
CHAPTER V
ARE PRESUPPOSITIONS UNAVOIDABLE?
Is philosophy possible without presuppositions?
This question has been crucial for philosophy since antiq
uity. Shadworth Hodgson has aptly stated the issue: "The
philosophical problem, is to find the means of philosophiz
ing without making assumptions" (quoted in 23:37). But it
is only recently that answers to this question have become
the most fundamental dividing line between philosophical
positions, depending upon whether it is answered affirma
tively or negatively.
The similarity between this question and the an
cient problem of skepticism should not be allowed to ob
scure the uniqueness and importance of demarcation by means
of "the presuppositional principle." It is true that rad
ical skeptics saw their doubt as an unbridgeable hiatus
that separated them from philosophers who advocated the
possibility of some genuine knowledge. But, in effect, the
radical skeptic abandoned philosophy as a consequence of
rejecting the availability of any means of determining
truth at all. If the apprehension of truth is considered
177
178
to be out of human reach* philosophy becomes no more than a
fruitless exercise with words. This has been manifested in
the tendency of sophists and skeptics to substitute rheto
ric for philosophy. By way of contrast* "the presupposi-
tional principle" is a line of demarcation that distin
guishes one class of philosophers from another class of
philosophers--that is* all philosophers who repudiate radi
cal skepticism are themselves divided by the position they
take on the question of presuppositions.
The philosophical dispute over presuppositions is
clearly evident in our own time. Arguments for and against
the possibility of achieving a presuppositionless philo
sophy are far from effecting unanimity or consensus among
philosophers. Part of the problem has been terminological.
Some have based their advocacy of presuppositionalism upon
the contention that the principle of presuppositionlessness
is itself the greatest presupposition (23:37). But this
charge is fatuous if it is* as it often seems to be* noth
ing more than the result of a failure to define the term in
question. It may merely evidence a confusion of different
meanings. For after we specify the different meanings of
the term* we can say* without contradicting ourselves* that
a presuppositionless philosophy is possible. Its presup
positionlessness in one sense need not preclude its being
"presuppositional" in a different and innocuous sense.
179
It can be argued, then, that beginning with the
"assumption" that a presuppositionless philosophy is possi
ble does not necessarily involve one in a self-defeating
position. The assumption may be purely methodological, and
if it can be shown to be realizable, then the assumption is
transformed into a valid principle. Only if the assumption
were insisted upon without probative argument would it be
"the greatest presupposition." Indeed, this has been the
character of my argument against Bartley's version of pre-
suppositionlessness. He adopts the principle as an osten
sible working hypothesis, but his argumentation fails to
support his thesis. His typo of "presuppositionlessness"
is self-refuting. But one case of failure to achieve a
presuppositionless philosophy does not preclude the success
of other attempts. Each must be analyzed on its own mer
its. Considerably more sophisticated and rigorous is
Husserl's attempt to forge a philosophy without presupposi
tions .
Presuppositions and Rationality
At the beginning of this work I discussed Bartley's
view of presuppositions and commitment, which he considers
to be inseparable. The problem at hand, however, is the
possibility of overcoming the dilemma that Bartley posed
with respect to them. What he said, in effect, was that
either one is an irrationalist because he is "committed" to
180
certain presuppositions or one is a rationalist because he
is not "committed" to any presuppositions. I have already
shown that his view of the latter alternative is untenable
because it is unattainable in principle. He implied., how-
ever, that if his presuppositionless philosophy is invalid
then there are no alternatives besides skepticism and fide-
ism. Now it is my contention that these alternatives are
avoidable and that it is possible to combine "presupposi
tions" and "rationalism." In order to achieve my purpose,
however, I must first examine the meanings of the terms
"presupposition," "commitment," and "rationalism."
Bartley's use of the terms "presupposition" and
"commitment" can be described in the following way. Any
view, value, principle, or perspective adhered to tena-
ciously--that is, to which one is "committed" in the sense
of refusing to concede any possibility of refutation--he
considered to be a presupposition. Since he refused to
allow the possibility of any position being irrefutable,
Bartley made no fundamental distinction between stipulated
absoluteness and logical absoluteness. His most basic dis
tinction was drawn between the attitudes in which princi
ples are held. If one holds to a principle with an intran
sigent unwillingness to consider it amenable to critical
refutation, then he is an irrationalist. However, if one
holds to the same principle with a genuine willingness to
relinquish it under the assault of negative criticism, then
181
he is a rationalist. This attitudinal distinction osten
sibly allowed Bartley to utilize certain principles (e.g.,
the laws of logic) as criteria for criticizing theories.
But since he believed that his position made even these
principles provisional and refutable, he denied that he was
"committed" to them--or to anything else. Hence, he
thought that he had escaped presuppositionalism. But we
have seen that he presupposes absolutely that views can on
ly be "falsified" and never justified. Moreover, while he
says that criticism and the principles of logic may be re
futed someday, it remains a mystery how "refuted" can have
any consistent meaning in. this alleged future possibility.
By defining rationality in terms of criticism, and
criticism in terms of refutability, and refutability in
terms of attitude (11:147), Bartley blurs the distinction
between a rational view and a rational person. He defines
refutability in terms of attitude because he leaves no
other means by which it can be defined. He says that one
either has or does not have the attitude that a principle
or theory is refutable. Since observation and the princi
ples of logic are held to be amenable to refutation, they
cannot be used to define refutability. So, in the final
analysis for Bartley, one's attitude is the decisive factor
in rationality. The decisive question is: "What is one's
attitude toward a principle or proposition?" Nothing can
182
be irrefutable in principle, according to Bartley, because
the locution "in principle" would mean that the laws of
logic are absolutized.
It is clear, nevertheless, that a view may be ra
tional (e.g., an argument may be valid), while a person may
reject it. Hence, contrary to Bartley's position, a ra
tional view and a "rational person" must be distinguished.
Even "rational persons" sometimes accept irrational views.
That a person may be persuaded that a view is rational when
in fact it is irrational can be illustrated by reference to
radical skepticism. Husserl writes:
Every genuine skepticism, whatever its type and
orientation may be, can be recognized by this fundamen
tal absurdity, that in the arguments it uses it presup
poses implicitly, as the conditions of the possibility
of its validity, precisely that which it denies in its
own theses .... He who merely says, I doubt the sig
nificance of reflexion for knowledge, maintains an ab
surdity. For as he asserts his doubt, he reflects, and
to set this assertion forth as valid presupposes that
reflexion has really and without a doubt (for the case
in hand) the very cognitive value upon which doubt has
been cast, that it does not alter the objective rela
tion, that the unreflective experience does not forfeit
its essence through the transition into reflexion.
(30:208-9)
Contrary to Bartley's assumption that the rational person
is one who holds all views to be provisional, the "rational
person" would seem to be one who accords a degree of assent
or conviction to a view according to its evidential and
logical merits. This means'not only that certain theories
will be doubted and held to be provisional, but also that
some principles will be recognized as irrefutable (e.g.,
183
the cognitive value of reflection, the law of noncontradic
tion, the principle of deducibility, etc.).
Presuppositions and Evidenz
On this point Husserl provides some useful distinc
tions. He contends that Evidenz (i.e., the perception of
the agreement between a proposition and what it refers to
or intends) is the ultimate criterion of truth, and that
"primordial dator intuition" (i.e., original object-giving
insight) is the ultimate source which justifies all our
knowledge. It is important to emphasize Husserl's distinc
tive understanding of Evidenz to prevent it from being con
fused with a variety of notions of "self-evidence." By
Evidenz he does not mean a strong feeling, or intense con
viction, or a disposition to take something for granted, or
the mere recognition of tautologies (29:48). Rather, it
consists of self-givenness grasped in mental seeing, i.e.,
it is an immediate "apprehending of the intended object it
self as it is" (29:28). Thus, I can "see" a phenomenon of
redness, and I can direct my "seeing" to the "seeing" of
redness, and so grasp the "seeing" of redness itself in
"seeing" (29:49).
Husserl says that "immediate 'seeing1 (Sehen), not
merely the sensory seeing of experience, but seeing in gen
eral as primordial dator consciousness of any kind whatso
ever, is the ultimate source of justification for all ra-
184
tlonal statements" (30:75-76). And again he points out,
"If we ask why the statement is justified, and ascribe no
value to the reply 'I see that it is so,' we fall into ab
surdity ..." (30:76). In contradistinction to Popper and
Bartley, then, he argues that primordial dator intuition is
a legitimate source of authority for knowledge, and that
"whatever presents itself in 1 intuition1 in primordial form
(as it were in its bodily reality), is_ simply to be accept -
ed as it gives itself out to be, though only within the
limits in which it then presents itself" (30:83).
Nevertheless, Husserl recognizes that since his
final criterion of truth is the direct evidence of intui
tively grasped data, Evidenz must not only be distinguished
from pseudo-evidence, but its own inner distinctions must
also be clarified. The former is achieved by utilizing the
phenomenological method in "the careful observation, anal
ysis and description of immediately given phenomena in
their irreducible nature and in their full variety" (39:
120) .
As to its inner distinctions, Evidenz may be either
adequate or inadequate (30:354-57). The latter admits of
degrees, but the former does not (30:357). Adequate Evi
denz is illustrated, for instance, by a simple arithmetical
insight, while inadequate Evidenz is exemplified in the
perception of a three-dimensional object, for such a per
ception is always limited to only one aspect of it.
Husserl draws a second distinction between assertoric Evi
denz and apodictic Evidenz (30:353-54). The former is
awareness of empirical individual objects or conditions,
whereas the latter is insight into general essences and
essential relationships which demonstrably excludes any
alternatives. Apodictic Evidenz alone is generally consid
ered to be adequate. Thirdly, Evidenz may or may not in
clude an element of fact. If it does it is impure, and if
it does not it is pure. This is found in the difference
between purely formal or analytic and material or synthetic
a priori Evidenz (30:354); between theoretical, axiological
and practical Evidenz (30:357-59)j and between immediate
and mediate Evidenz (30:357-64). Fourthly, in his Formal
and Transcendental Logic, Husserl differentiates Evidenz
of distinctness from Evidenz of clearness. The former is
found in "a judgment that is explicitly made and not merely
contained vaguely and confusedly in a causal idea or in a
sentence which is 'merely' read and repeated" (39:121).
Evidenz of clearness is not only distinct but it also pre
sents the object to which the judgment refers. And lastly,
in Husserl's analysis of Evidenz, there is prepredicative
Evidenz of individual objects, which is the most primary
and basic of all. Hence, Evidenz is not to be understood
as a "peculiar feeling" which supposedly guarantees the
truth of a judgment. Rather, Evidenz is the complete
awareness of the complete concordance between a proposition
and a fact--which concordance is_ truch.
In view of these epistemological distinctions and
the coherent account of our knowing that they appear to
provide, it is both unnecessary and inaccurate to consider
all propositions and principles to be provisional. Immedi
ate experiences can furnish Evidenz as the basis of indubi
table and irrefutable knowledge. Among the things we know
are certain universal and necessary truths, and an examina
tion of the way we know them reveals that we could not pos
sibly derive them from sense experience. For example, we
know "essential laws," such as the impossibility of regret
ting something without remembering it, or the impossibility
of two things being both left and right of something at the
same time in the same respect. We know, too, that remem
bering differs from imagining, and imagining differs from
inferring, etc. We also know that certain logical rela
tionships and principles are valid and true. If we know
the premises of a valid argument to be true, then we know
the conclusion of that argument to be true. But we can
know nothing if we do not know some things by intuition.
And since we do know some things, the only explanation is
that some things are given to us in direct awareness.
Some philosophers have tried to explain "necessary
truth" as the product of convention, apparently because
they believed this was the only way to escape the "dangers"
of admitting "intellectual intuition." But William Kneale
187
observes:
It may . . . seem hard-headed to say that all necessity
is Imposed by our own legislation. But if this is the
conventionalist's motive for developing his theory, he
is doomed to disappointment. Whenever he hopes to get
clarity and determinateness by setting out his rules
and considering their implications, he will find that
he is forced to rely once more on intellectual intui
tion, although this time it may be only insight into
logical connexions. The effort to dispense with intel
lectual intuition is, indeed, an attempt to dispense
with all thinking; for anyone who notices that some
thing is of a certain kind, e.g., red, must be intui
tively aware of the mutual exclusion of some kinds,
e.g., red and green. But the absurdity of the enter
prise is especially clear in the limiting case just
mentioned, where it is supposed that the conventional
ist has only to apply rules of his own making. (38:132)
Without intuition, therefore, there could be no logical ar
gumentation, for the latter presupposes awareness of valid
ity, cognition of some content or propositions that can be
true or false, and recognition of relevance. In the final
analysis, these cannot be argued to without circularity,
but must be argued from, with intuition as the means of our
knowing them.
Presuppositions and Starting Points
In view of the indispensability of intuition for
human knowledge, the problem of presuppositions can be put
in a new light. First of all, it is vital to distinguish
different kinds of presuppositions. Marvin Farber refers
to (l) material presuppositions (e.g., the continuity of
existence in time, causal uniformity, infinite extension),
(2) cognitive presuppositions (e.g., the reliability of
188
perception and memory), (3) formal assumptions In special
systems (e.g., the axioms or postulates In a system of ge
ometry), (4) the principles of logic (e.g., the law of non
contradiction), (5) motivation, (6) influences, (7) signi
ficance for society and history and (8) beliefs (23:39-
40). He observes that not all of these are of the same
type, however, for some of them are not to be treated as
assumptions but as results to be achieved. Furthermore,
cognitive presuppositions (such as the reliability of per
ception and memory, and the validity of knowledge) do not
have to be assumed if all known evidence overwhelmingly
verifies them. We have already seen that the law of non
contradiction is irrefutable and intuitively verified.
Thus, there are very important distinctions that need to be
drawn between "presuppositions."
It is unnecessary for my purpose to explore the
many facets of the problem of presuppositions. To estab
lish my thesis that justification in philosophy is possible
--and, indeed, unavoidable--it is sufficient to show that
there are certain indispensable starting points which serve
as the basis for grounding valid claims to knowledge. It
is misleading to call these starting points "presupposi
tions," for this term is often used to refer to both arbi
trary and necessary preconditions, whether tacit or explic
it. The important desideratum is not so much Bartley's
insistence that we examine our attitude with respect to
philosophical starting points* but that we determine the
nature of the starting points in question and adjust our
attitude according to their grounding in intuition. For
example* there are certain starting points* or "presupposi
tions* " that are necessary conditions for the establishment
of particular natural sciences. If one seeks to achieve
knowledge in one of these sciences* it is imperative that
he make certain "assumptions." With the relinquishment of
one or more of these sciences* some of the presuppositions
relative to them will also be abandoned* or at least sus
pended. Hence* they are not strictly unavoidable. But
there are "presuppositions" that are indispensable to human
life and knowledge as such. For example* if distinctively
human life and knowledge are not realized apart from con
ceptual thinking and the use of language* then the logical
law of noncontradiction is an unavoidable "presupposition"
for them. And as such it must be presupposed for all the
sciences* which obviously could not even be initiated apart
from conceptual thinking and language.
Another example may serve to highlight the impor
tance of the foregoing distinction between "presupposi
tions." In any of the branches of philosophy one must be
gin with the awareness of concepts and propositions (even
if they are only about one's own subjective states) and of
the value of knowledge (even if it is only considered as a
190
result to be achieved). One may refuse to engage in philo
sophizing, however. Yet If he continues to think concep
tually at all, or if he continues to prefer life over death,
he makes certain "assumptions"--namely, that life is dis
tinguishable from death, and therefore that the logical and
ontological laws of noncontradiction are true, and that he
knows that life is different from death, and that the know
ledge on which he bases his preference for life is a value
and not a disvalue, at least to this extent.
Now if, after careful scrutiny, certain truths are
recognized to be indispensable for conceptual thinking and
the use of language (which partially characterize human
life as human), then we have no choice but to openly ac
knowledge the ineluctability of justificationalism in phi
losophy. The kind of presuppositionless philosophy that
Bartley sought to construct is illusory, but the kind that
Husserl explicated seems to be a genuine possibility. In
contrast to Bartley’s attempted removal of all foundations,
Husserl pursued the ideal of a presuppositionless philoso
phy in the sense of discovering the most radical foundation
of all philosophy and science. Since it is "the things
themselves" that are the final court of appeal in determin
ing the truth of all our claims, he forged a method by
which one could let "the things themselves" present them
selves and be described as they actually are. In pursuit
191
of this goal,, he aimed at removing all prejudices and un
critical assumptions in philosophy.
Moreover, Husserl differentiated presuppositions of
an arbitrary and positive kind from those which are implic
itly presupposed in all questioning and answering. He did
not view the latter as assumptions, for they are justifi
able as well as indispensable. Such starting points are
found in all conceivable instances of Evidenz, and thus
precede them and are contained (implicitly) in them. The
totality of these necessary starting points constitutes the
absolute basis which philosophy secures for itself. Hence,
Husserl's phenomenological method begins with the "self-
validating" cognitive experience itself. And since assump
tions can be examined, and objects of experience are acces
sible to a truly descriptive method, a genuine presupposi
tionless philosophy is possible. Thus, the self-givenness
of the objects of experience is the basis for the critical
justification of all knowledge.
Whether justification or refutation is accomplished
by critical argument, reference to universals is unavoid
able. Although there are many ramifications to the problem
of universals, perhaps it will not be too misleading to
make some remarks on their significance for the question of
presuppositions. First it must be made clear, for example,
that both the color green and the concept "green" are uni
versals, i.e., cases of the one in the many. In the former
192
case* many things can have the property of greenness* and
in the latter case* the concept "green" can refer to many
instantiations of greenness. But not only qualities and
concepts are universals* for propositions are universals
too--and there are universal objects and states of affairs
(29:41).
Philosophers who have advocated sense experience as
the only source of knowledge have generally taken the posi
tion of conceptualists or nominalists with respect to uni
versals. As a result* they are compelled to adopt a con
ventionalist view of logic and inference. Yet* without an
objective or valid status for universals (i.e.* in contra
distinction to their being mere constructs of the mind* or
words arbitrarily adopted to refer to more than one partic
ular)* logical inference is unable to furnish us with con
clusions that are necessarily true. If universals are only
arbitrary conventions* then logical inference partakes of
that same arbitrariness. The conclusion of a logically
valid argument* on such a supposition* can only be known to
be true if it is empirically verified. Though this view is
not implemented in practice, it is inevitable in theory for
anyone who begins with the assumption of conceptualism or
nominalism.
Nevertheless* the clear cases of logical inference
leading to universally and necessarily true conclusions
support the contention that reason is operating with
193
genuine universals. If it were not for eidetic intuition
(i.e.* the direct awareness of universals), not only logi
cal inference but cognition (on the basis of which true
propositions can be made) would itself be impossible. And
it is because of eidetic intuition that a conclusion of a
valid argument can be known to be necessarily true prior to
empirical verification. When both premises are known to be
true, the conclusion can provide new knowledge by connec
ting the concepts of the major and minor terms in a dis
tinctly different proposition. This is made possible by
the concept of the middle term which is a universal. If
this is denied, then logic is made impossible, and without
logic scientific knowledge is impossible.
Presuppositions and Justification
It should now be evident that the kind of presup-
positionlessness claimed by nonjustificationalism is an
impossibility. That view, therefore, cannot even figure as
a genuine philosophical alternative. In the issue of justi
ficationalism versus nonjustificationalism, we are present
ed with a legitimate bifurcation. If one alternative is
not true to the phenomenon of knowledge (and in the case of
nonjustificationalism, it is not merely false but self-con
tradictory), then the only other alternative (viz., justi
ficationalism) is true to the nature of knowledge. And
since certain starting points are unavoidable, they should
194
be distinguished from those that are avoidable. The ideal
of presuppositionlessness means, then, that there should be
no unexamined presuppositions and that no prejudgment
should be allowed to obscure or distort the objects under
investigation.
In view of these considerations, it would appear to
be less confusing to use the term "commitment” to refer to
the personal acceptance of something that is not universal
ly and necessarily binding. Thus, the logical principle of
noncontradiction would not be viewed as a "commitment," nor
would one be "committed" to it. One either recognizes it
or he does not--and, in the latter case, the failure to
recognize it is due to physiological and psychological
causes and not to choice. For choice requires the discrim
ination of alternatives, which is impossible without the
law of noncontradiction.
The fact of error may be thought to vitiate the
distinction here. However, when one commits the logical
error of assenting to propositions that are mutually con
tradictory, it is because he does not recognize that he is
doing so, or he fails to grasp what he is doing, i.e., he
is not aware that he is judging both propositions to be
true when in fact they cannot be. This kind of error is
not a counter example to the non-optional character of
recognizing the principle of noncontradiction. On the con
trary, the fact that it is an error, and that apart from
195
mental Impairment by organic or functional causes It can be
shown to be an error, substantiates the universality and
necessity of the principle.
It is important, then, to distinguish at least four
main types of "presuppositions." (l) First, there are nec
essary "presuppositions" (i.e., starting points) which are
unavoidable in human knowledge and which are given in intu-
ition--and therefore preferably not referred to as "presup
positions." (2) Distinct from these are relative presup
positions which are indispensable to the pursuit of certain
goals, such as the various sciences. That is, just as it
is not necessary that anyone should engage in the study of
the science of physics, for example, so the special presup
positions which make physics possible as a science are not
necessary to that individual. Hence, there are certain
presuppositions that are relatively indispensable to each
of the various sciences. (3) Another class of assumptions,
which may be termed systemic presuppositions, can be dis
tinguished by their determinative role as basic premises in
philosophical, geometrical, and other systems where alter
native systems are possible. These are seen, for example,
in the geometrical systems of Euclid, Riemann, and
Lobachevsky, or in the metaphysical systems of Aristotle,
Spinoza, Hegel, Bergson, Whitehead, etc. (4) Lastly, ordi
nary presuppositions can be distinguished from the other
three types. These are the numerous assumptions that are
196
made In rather unsystematic fashion In everyday life, rang
ing from superstitions to highly sophisticated guesses.
The first two classes of "presuppositions" are
epistemologically essential, i.e., intrinsically and in
strument ally respectively, while the last two classes are
more properly considered to he conjectural, although they
are not without some degree of contextual determination.
It is not necessary for my purpose that these four classes
of "presuppositions" he viewed as exhaustive. No doubt,
there are other ways of classifying them--and, of course,
even as I have presented them, they are amenable to divi
sion into subclasses. Nevertheless, whatever alternative
terminology may be employed, the distinction between non-
conjectural, necessary starting points and speculative,
hypothetical assumptions must be maintained if the nature
of knowledge is to be understood at all. Moreover, this
partial analysis should be sufficient to indicate the need
for defining terms such as "presupposition" and "commit
ment" before they can be put to effective service in artic
ulating a philosophical position.
The problem that has been under consideration is
that of escaping both skepticism and fideism as a conse
quence of refuting and rejecting nonjustificationalism.
It has been argued that Bartley shunned justificationalism
because of his faulty analysis of the grounds and meaning
of rationality. He had assumed that all starting points
and presuppositions in philosophy must be either fideisti-
cally chosen or left problematic. But if one says that he
accepts logic on a fideistic basis, he tacitly concedes
that the skeptic and the misologistic irrationalist may be
right, for the fideist, by definition, offers no reason for
his choice. Thus, since he considers his choice to be ar
bitrary, he must admit that it may be wrong. And if "faith
in logic" may be wrong, then "irrationalism" may be right.
The reasons for this seeming impasse are clear by now. The
fundamental error stems from the assumption that "criteria"
must be given for "criteria" ad_ infinitum. However, the
claim that there are, for example, such things as colors
(or that black is darker than white) cannot be proved to a
sightless person by an appeal to "criteria." As for the
man who is not blind, the justification is in the seeing.
And, of course, Evidenz not only pertains to perceptual in
tuition but also to the awareness of ideal and mental ob
jects. If this is true, then there is a form of justifica
tionalism that overcomes both skepticism and fideism.
Conclusion
There comes a point in philosophy where construc
tion must yield to givenness, and explanation to descrip
tion. There are some things we know even when we cannot
offer an explanation of all the elements involved in our
knowing It appears that it is simply the nature of the
"mind" to know things other than itself. There is no co
gent reason why it is any less possible for a person to
know "extra-subjective" objects than it is for him to know
his own states. The givenness or certainty that there are
self-consciousness and self-transcendence is in no way de
pendent upon our ability to explain how they take place.
That there is something and not nothing is true despite our
inability to explain how it has come about. And it seems
that intuition* or direct awareness* is similarly undeni
able despite its defiance of complete explanation. As
Husserl observes* "to deny self-givenness in general is to
deny every ultimate norm* every basic criterion which gives
significance to cognition" (29:^9)•
In the final analysis* then* our knowing is based
upon direct awareness* i.e.* intuiting "the things them
selves" and examining them by reflection. This means that
if something is given it does not have to be presupposed
(i.e.* conjectured or postulated). This insight provides
the answer to the question posed at the beginning of this
chapter.
CONCLUSION
Since recourse to direct awareness of "the things
themselves" has been shown to be unavoidable, I can now
summarize my arguments against nonjustificationalism and
the skeptico-fideistic thesis that all choices are ulti
mately arbitrary. The problem has been to indicate the
means by which our choices can be made nonarbitrary. It is
plain that a nonjustificational philosophy cannot resolve
this problem, because its own logical paradoxes make it
unrealizable. Moreover, fideism and conventionalism cannot
account for the necessity of certain principles which they
mistakenly allege to be a matter of arbitrary choice. And
it has been established that a skeptic cannot even speak or
act on the basis of conceptual thinking without contradic
ting his skepticism. In view of these findings, it would
seem either imperceptive or disingenuous of the nonjustifi-
cationist, the fideist, and the skeptic to claim that all
positions and propositions are tentative or arbitrary.
The nonjustificationist cannot avoid holding his
view because he believes it is justified. When he pro
pounds and recommends it, he is seeking to convince others
that it is the only position that one is justified in es-
199
200
pousing. He may say that his attachment to his own view is
only an "engagement, " but that he is wedded to it follows
from the fact that it cannot be used to refute itself.
The fideist avers that everyone must choose some
position on nonrational grounds, because he is convinced
that there are no rational criteria by which to assess al
ternatives. But he cannot argue for his view without de
feating himself, for he must make use of rationality to
state his position and argue for it. He fails to see that
if rationality makes it possible to articulate such a
stance, then it can also make it possible to choose non-
arbitrarily between different philosophical and theological
positions.
The radical skeptic defines his position as the
total suspension of judgment. But he does so only on the
basis of awareness of his doubt and reflection upon it.
Consequently, he cannot suspend his judgment on rationality,
for without it he could not claim that he is a skeptic.
For him to conclude, as some ancient skeptics apparently
did, that he should not speak or act at all, still fails to
overcome the problem. For he infers that this is the only
consistent thing for a radical skeptic to do. He will live
and die without being able to voluntarily dispense with ra
tionality. Hence, rationality is not arbitrary.
What, then, needs to be said here about rationality
201
In order to add clarity to the case for justIficatlonalism?
In the way I have used the term "rationality" thus far, I
have been referring primarily to what it minimally entails,
viz., logical consistency (i.e., the principles and rules
of logic) and faithfulness to reality (i.e., the descrip
tion of "the things themselves"). Accordingly, Husserl
says that "faithfulness to reality" requires that we
"grasp what is momentarily perceived in faithful conceptual
expressions, of which the meaning is prescribed purely by
the object perceived or in some way transparently under
stood" (30:17^).
My contention has been that rationality in this
minimal sense is unavoidable for human knowledge, and even
for human action which presupposes conceptual thinking.
More explicitly, I have maintained that the phenomenologi
cal principle that every primordial intuition is a legiti
mate source of knowledge provides us with means for testing
knowledge-claims. It must be emphasized that intuition of
essences Is a primordial dator act analogous to sensory
perception. While empirical intuition is consciousness of
an individual object, "essential intuition" is conscious
ness of a pure essence. This makes it possible for funda
mental principles to be grounded on the basis of what is
given immediately in intuition. Hence, Evidenz based upon
intuition is the ultimate criterion of all truth, and in
202
tuition is the ultimate source which justifies all of our
knowledge.
In the discovery of truth, however, intuition must
be supplemented by reflection and inference. Inferring, or
reasoning, which is a certain kind of mental process, is
impossible without the standards of rationality. For with
out basic logical principles and some content, there can be
no process of drawing conclusions from premises and there
can be no way of distinguishing between correct and incor
rect reasoning. To be "reasonable," then, is to abide by
the standards of rationality.
Nevertheless, Bartley has made the claim that "ra
tionality is unlimited" due to the alleged open-endedness
of criticism (11:147). Yet logic (i.e., the basic princi
ples and rules of inference) constitutes the most defini
tive and general limiting conditions for all rational dis
course. In fact, the power of reason is manifested, as
Paul Marhenke argues, in one's capacity to (l) entertain
beliefs and express them by means of propositions, (2) draw
inferences, both deductive and inductive, from these propo
sitions, (3) apprehend or intuit necessary connections,
i.e., recognize the a priori truth of a proposition, and
(4) form a priori concepts (47:156). And since pure logic
alone cannot convey empirical information, it seems that
rationality should also be understood as the capacity to
grasp what is present in the world. For example, an
203
observation statement is "rational" if it is confirmed by
perception. Hence, these considerations warrant the char
acterization of "rationality in a person" as more than an
attitude of criticism and tentativeness, for it also em
braces a concern for consistency, objectivity, intersubjec
tivity, clarity, and universality.
Contrary to these claims, however, a number of phi
losophers have assumed that all facts are subjective or
biased "interpretations," usually made from some "point of
view," or within some metaphysical or quasi-metaphysical
framework, with the consequence that no one can be sure
that any fact is objective and undistorted by personal and
historical influences. In the "knowing" of anything, it
has been supposed, the thing "known" is distorted. But, as
it has already been noted, this fact itself could not be
known if such were the case. Any philosophical position
which argues that objectivity is impossible destroys it
self, for in withholding objectivity from everyone it
undermines its own claims. Its arguments cannot be taken
seriously, therefore, for on its own assumption neither it
nor any other view can claim truth. "Ultimate arbitrari
ness" --whether in the form of skepticism, fideism, or non
justif icationalism--destroys the very possibility of know
ledge.
If philosophy has known any progress in its history,
it is especially discernible in the clarification of its
problems and in the refutation of certain unfounded theses.
The removal of ambiguity and the elimination of error are
primarily negative accomplishments. But they are not nec
essarily trivial. And the means for achieving these goals
has been, and continues to be, the two standards of logical
consistency and material truth. Hence, if part of philo
sophy's task is the making of choices nonarbitrary, it must
apply these standards to an assessment of every theory and
knowledge-claim. Without some universal and necessary
truths, this assessment cannot even begin. But with even
one, my thesis that justificationalism is unavoidable in
philosophy is established.
More than one truth can be known, however, for in
primordial dator intuition and the judgments that faithful
ly fit the intuitively given data, we are provided with the
ultimate basis for all rational justification. On this
fact rests the possibility of reasoning, criticism, and
knowledge.
BIBLIOGRAPHY
BIBLIOGRAPHY
1. Adams* G. P. "Reason and Purpose*" University of
California Publications in Philosophy, XXI (1939)j
207-28.
2. Agassi* Joseph. "Science in Flux: Footnotes to
Popper." Unpublished paper* Boston University*
1965.
3. Alexander* Peter. "The Philosophy of Science* 1850-
1910*" A Critical History of Western Philosophy,
edited by D.J. O'Connor. New York: The Free
Press* a subsidiary of The Macmillan Co.* 1964.
4. Archibald* G. C. "Refutation or Comparison," The
British Journal for the Philosophy of Science,
XVII (February* 1967)* 279-96"!
5. Aristotle. "Metaphysics*" The Works of Aristotle.
Vol. VIII of Great Books of the Western World.
Edited by R.M. Hutchins. Chicago: Encyclopaedia
Britannica* 1952.
6. Ayer* A. J. Language, Truth and Logic. London:
Victor Gollancz* Ltd.* 1936. Second edition* 1946.
7. _______ . "Phenomenology and Linguistic Analysis*"
The Aristotelian Society* Supplementary Volume
XXXIII (1959), 111-24.
8. . The Problem of Knowledge. Edinburgh:
R. & R. Clark* Ltd.* 1956.
9. Bambrough* Renford. "Review of W.W, Bartley* Ill's
The Retreat to Commitment*" New Statesman, LXIX
(January* 1965)j 45-46.
10. Bartley* W. W.* III. "Rationality versus the Theory
of Rationality*" The Critical Approach to Science
and Philosophy* edited by Mario Bunge. London:
Collier-Macmillan* Ltd.* 1964.
206
207
11. ________ . The Retreat to Commitment. New York:
Alfred A. Knopf, Inc.~ , 1962.
12. Bennett, Jonathan. Rationality. London: Routledge
& Kegan Paul, 1964.
13. Bergmann, Gustav. Logic and Reality. Madison,
Wisconsin: The University of Wisconsin Press,
1964.
14. Bernays, Paul. "Reflections on Karl Popper's
Epistemology," The Critical Approach to Science
and Philosophy, edited by Mario Bunge. London:
Collier-Macmillan, Ltd., 1964.
15. Blanshard, Brand. Reason and Analysis. La Salle,
Illinois: Open Court Publishing Co., 1962.
16. Bunge, Mario (ed.). The Critical Approach to
Science and Philosophy. London: Collier-
Macmillan, Ltd., 1964. (This Popper Festschrift
contains twenty-nine articles on almost the
entire range of his views.)
17. Cranston, M. "Review of W.W. Bartley, Ill's The
Retreat to Commitment," The Listener, LXXII
(August, 1964), 316-17.
18. Dennes, W. R. "The Appeal to Reason," University
of California Publications in Philosophy, XXI
(1939), 3-42.
19. Dondeyne, Albert. Contemporary European Thought and
Christian Faith. Translated by E. McMullin and
J. Burnheim, Pittsburgh: Duquesne University,
1958.
20. Ewing, A. C. Reason and Intuition. London:
Humphrey Milford Co., Ltd., 1941.
21. Fain, Haskell. "Review of Karl Popper's The Logic
of Scientific Discovery," The Philosophy of
Science, XXVIII (July,1961), 319-24.
22. Farber, Marvin. The Foundation of Phenomenology.
Cambridge, Mass.: Harvard University Press, 1943.
208
23. ________. "The Ideal of a Presuppositionless Philo
sophy, " Phenomenology, edited by J.J. Kockelmans.
Garden City, N.Y. : Doubleday & Co., Inc., 1967.
24. Hepburn, R. W. "Review of W.W. Bartley, Ill's The
Retreat to Commitment," Philosophical Books, IV
(May, 1963), 1-2.
25. Hesse, Mary. "Review of Karl Popper's Conjectures
and Refutations," Philosophical Quarterly, XV
(October, 1965), 372-74.
26. Holmer, P. L. "Review of W.W. Bartley, Ill's The
Retreat to Commitment," The Journal of Religion,
XLIV (April, 1964), l6l-'UT.
27. Husserl, Edmund. Cartesian Meditations. Translated
by Dorion. Cairns. The Hague: Martinus Nijhoff,
I960.
28. _______ . Formale und Transzendentale Logik.
Halle: M. Niemeyer, 1929-
29. _______ . The Idea of Phenomenology. Translated by
W.P. Alston and G. Nakhnikian. The Hague:
Martinus Nijhoff, 1964.
30. _. Ideas: General Introduction to Pure
Phenomenology. Translated by W.R.B. Gibson.
New York: Collier-Macmillan, Ltd., 1962.
31. . Ideen zu einer reinen Ph4'nomen.ologie und
phgnomenologischen Philosophie. 3 vols. Haag:
Martinus Nijhoff, 1950 (Vol. I) and 1952 (Vols.
II and III).
32. . Logical Investigations. Unpublished
translation by D.A. Willard. Los Angeles:
University of Southern California, 1967.
33. . Logische Untersuchungen. 3 vols.
Halle: M. Niemeyer, 1922.
34. . Phenomenology and the Crisis of _
Philosophy. Translated with an introduction by
Q. Lauer. New York: Harper & Row, Publishers,
1965.
209
35. Jones, Peter. "Review of W.W. Bartley, Ill's The
Retreat to Commitment," The British Journal for
the Philosophy of Science, XVII (August, 1955)*
153-55.
36. Jordan, R. "The Revolt against Philosophy: the
Spell of Popper," The Return to Reason, edited by
John Wild. Chicago: Henry Regnery Co., 1953.
37. Kaufmann, Walter. "Review of W.W. Bartley, Ill's
The Retreat to Commitment," Ratio, LXIII
(December, 1955), 223-25.
38. Kneale, William. "Are Necessary Truths True by
Convention?" The Aristotelian Society,
Supplementary Volume XXI (1947), 115-33•
39. Kockelmans, J. J. A First Introduction to Husserl's
Phenomenology. Pittsburgh: Duquesne University
Press, 1957.
40. (ed.). Phenomenology. Garden City, N.Y.:
Doubleday & Co.~, Inc., 1967.
41. Laudan, L. "Review of The Critical Approach to
Science and Philosophy," The British Journal for
the Philosophy of Science, XVIII (August, 1957) , ~
154-57.
42. Lenzen, V. F. "Reason in Science," University of
California Publications in Philosophy, XXI
(1939), 77-95.
43. Lewy, C. "Why are the Calculuses of Logic and
Arithmetic Applicable to Reality?" The
Aristotelian Society, Supplementary Volume XX
(1945), 30-39.
44. Loewenberg, J. "Artifacts of Reason," University
of California Publications in Philosophy, XXI
(1939), 45-73.
45. . Reason and the Nature of Things. La
Salle, 111.: The Open Court Publishing Co., 1959*
46. Mackay, D. S. "Reason as Custodian," University of
California Publications in Philosophy7 ) XXI (1939),
153-204.
210
47. Marhenke, Paul. "Rationalism and Irrationality,"
University of California Publications in
Philosophy, XXI (1939), 153-79.
48. Martin, W. B. "Review of W.W. Bartley, Ill's The
Retreat to Commitment," The Christian Century,
LXXX (January 23, 1963), 114-15.
49. Moore, G. E. Philosophical Papers. London: George
Allen & Unwin, Ltd., 1959. (Especially chapters
nine and ten.)
50. ________. Some Main Problems of Philosophy. New
York: Collier-Macmillan, Ltd.7 1962. (Especially
chapters fourteen through seventeen.)
51. Passmore, John. A Hundred Years of Philosophy.
London: Duckworth, 1957"i (Especially chapters
eight and seventeen.)
52. Popper, K. R. Conjectures and Refutations. New
York: Basic Books, Inc., Publishers, 19^2.
53. ________. The Logic of Scientific Discovery. New
York: Harper & Row, Publishers, 1965•
54. ________. "Logic Without Assumptions," Proceedings
of the Aristotelian Society, XLVII (194777 251-92.
55. . "New Foundations for Logic," Mind, LVI
(July, 1947), 193-235.
56. . The Open Society and its Enemies. Vol.
II. Princeton: Princeton University Press, 1963.
57. Prichard, H. A. "Appearances and Reality," Mind,
xv (1906), 223-29.
58. Reinach, Adolf. Was ist Ph^nomenologie? Midnchen,
1951. (A lecture originally given in 1914.)
59. Russell, Bertrand. The Problems of Philosophy.
London: Oxford University Press, 1912.
60. Ryle, Gilbert. "Why are the Calculuses of Logic
and Arithmetic Applicable to Reality?" The
Aristotelian Society, Supplementary Volume XX
(1946), 20-29.
211
61. Sartre, Jean-Paul. "Intentlonality: A Fundamental
Concept In Husserl's 'Phenomenology,'" translated
by D.A. Willard, from La Nouvelle Revue Francaise
(January, 1939)* 129-32.
62. Spiegelberg, H. The Phenomenological Movement.
2 vols. The Hague: Martinus Nijhoff, I960.
63. Warnock, G. J. "Review of Karl Popper's The Logic
of Scientific Discovery," Mind, LXIX (January,
1900), 99-101.
64. Willard, D. A. "Meaning and Universals in Husserl's
Logische Untersuchungen." Unpublished Ph.D.
dissertation, University of Wisconsin, 1964.
65. . "The Role of Rules of Inference in
Logical Method." Unpublished paper, University
of Southern California, 1966.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Kant'S Doctrine Of Existence As A Predicate

PDF

An Analysis Of The Philosophical Beliefs Implicit Within General Semantics And Their Relevance For Educational Theory

PDF

Theory And Practice In The Black Mountain Poets: Duncan, Olson, And Creeley

PDF

An Investigation Of The Forms And Defenses Of Teleological Ethical Theories, With Emphasis On The Ethical Theory Of Brand Blanshard

PDF

A Study Of Philosophy Curriculum Dissonance In Selected Colleges And Universities In California

PDF

Rhetoricians On Language And Meaning: An Ordinary Language Philosophy Critique

PDF

In Defense Of Beauty: A Phenomenological Analysis Of The Hedonics Of Perception

PDF

A Critical Examination Of Heidegger'S And Jasper'S Interpretations Of Nietzsche

PDF

The Phenomenon Of Literature: Prolegomena To A Literary History

PDF

The Conventionalist Philosophy Of Empirical And Deductive Science

PDF

The Expressionistic Theology Of Horace Bushnell: A Study Of The Approach, Grounding And Form Of His Theology

PDF

An Ontological Interpretation Of The Concept Of 'Expression' In The Philosophy Of Merleau-Ponty

PDF

A Critical Analysis Of The Philosophical Fragments Of Epicharmus

PDF

'Difficult Balance': The Poetry Of Richard Wilbur

PDF

The Problems Of Jr. College Freshmen And The Development Of The Jr. College Problem Check List

PDF

Introduction To The English Edition Of Meinong'S "Ueber Annahmen" With Selections From The Translation: "On Assumptions"

PDF

The Good Life: The Development Of A Concept In Smollett'S Novels

PDF

A William Temple Word-Book: A Comprehensive Philosophical And Theologicalindex To His Major Published Writings

PDF

Fiction, Truth, And Reference

PDF

Husserl And Merleau-Ponty And The Problem Of The Cultural Studies

Asset Metadata

Creator Hanna, Maurice (author)

Core Title Justificationalism Versus Nonjustificationalism In Philosophy: A Critique of The Theory Of Rationality In Karl Popper And W.W. Bartley, Iii

Contributor Digitized by ProQuest (provenance)

Degree Doctor of Philosophy

Degree Program Philosophy

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag OAI-PMH Harvest,Philosophy

Language English

Advisor Willard, Dallas (committee chair), Cantelon, John E. (committee member), Engel, S. Morris (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-669452

Unique identifier UC11360928

Identifier 6913061.pdf (filename),usctheses-c18-669452 (legacy record id)

Legacy Identifier 6913061.pdf

Dmrecord 669452

Document Type Dissertation

Rights Hanna, Maurice

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA