University of Virginia Library

CERTAINTY SINCE THE
SEVENTEENTH CENTURY

After the seventeenth century, questions about the
certainty of knowledge agitated those who wanted to
understand and improve the lot of man on earth more
than those anxious to safeguard religious truth. Science
rather than religion became recognized as the realm
of the most venerable truths, and questions about the
character of scientific knowledge dominated discussions
about certainty.

The new context for these discussions was set by the
Scots philosopher, David Hume, who, unlike earlier
skeptics, did not merely emphasize that the human
intellect was imperfect, but denied that it had any
demonstrable link with the order of the universe.
Whereas Hume was able to reconcile philosophical
uncertainty with accepting other certainties, his suc-
cessors looked for a more secure relationship between
man and a cosmos made of alien stuff, and they hoped
to discover such security by redefining rationality. This
led them, with the exception of Hegel and his follow-
ers, to narrow the meaning of rationality by circum-
scribing either the objects that could be known or what
could be said about them or the act of knowing, ending
in a radically revised understanding of knowledge that
attached certainty to acts of will rather than intellect.

Hume's denial that man's reason gave him insight
into the rational order of the universe was inspired by
his opposition to the doctrine that moral perfection
consisted in the triumph of reason over passion. Instead
of dividing human nature, as previous accounts had
done, between a higher part, reason, that connected
man with God, and a lower part, passion, which men
shared with animals, Hume proposed that man was
simply a bundle of perceptions, all of whose ideas were
merely copies of sense impressions. Although some of
Hume's successors were to fasten on sensation as the
source of certainty, for Hume perceptions were a flow
of unconnected particulars which could not in them-
selves constitute knowledge because that had to do
with relations and general ideas. Far from denying that
men possessed well articulated, valid structures of gen-
eral ideas, Hume insisted upon it. He not only admired
Newton's work and accepted the laws of physics estab-
lished by him, but argued against miracles on the
ground that they violated the laws of nature. The
credentials of the belief in necessary connections be-
tween perceptions had to be examined precisely be-
cause it was intrinsic to human life.

Hume divided the connections men made between
ideas into two kinds—relations of ideas and of matters
of fact. Of the former, he considered mathematical
reasoning the most obvious example. Mathematical
truths, Hume believed, depended on comparing a finite
set of ideas with one another, the criteria for making
these comparisons correctly being given by a system
which is complete in itself and has no reference to the
outside world. “That the square of the hypotenuse is
equal to the sum of the squares of the other two sides”
has to be accepted by anyone who understands what
the terms of the proposition mean. But there were no
such well-defined criteria for propositions about mat-
ters of fact. How then did men come to believe not
only that fire burns, but that it always burns?

If the traditional explanation that we have insight
into the nature of fire were correct, Hume argued, the
belief that fire burns should arise after observing only
one instance. Instead we require repeated experiences
of a connection between cause and effect before be-
lieving that one necessarily follows the other. As there
is nothing in several resembling instances absent in one
of them to explain why two events must be joined,
the addition must come from the mind. The necessary
connection we discover is therefore “an internal im-
pression of the mind, or a determination to carry our
thoughts from one object to another” (Treatise, Sec.
XIV, “Of the Idea of Necessary Connexion,” p. 165).
Belief in necessity is nothing more mysterious than a
feeling attached to certain ideas and is produced by
repeated experiences of a conjunction between two
ideas, or, in other words, belief arises from custom.

Hume gave a similar explanation of abstract ideas
such as those of space and time. Here he followed the
suggestion made by Newton's critics, such as Leibniz
and Berkeley, that space and time were not inde-
pendent realities but a way of understanding the co-
existence or succession of objects. After repeatedly
experiencing the coexistence of patches or points of
color, Hume argued, we separate the idea of this co-
existence from particular perceptions and think of it
as the abstract idea of space or of time. General ideas
arise in the same fashion, from the custom of associ-
ating a number of objects with the idea of one of them.

The mind then consists wholly of sensations or feel-
ings, reflected in ideas, which appear to the mind in
certain relations. From these relations a natural instinct
of the mind leads men to conclude that objects exist
and that events are related as cause and effect. Our
justifiable certainty about the attributes of objects and
causal relationships has not been reached by grasping
the simple, unitary nature of a thing or event. Every-
thing we know is a compound of perceptions that
become something more than a heap only thanks to
a propensity of the mind to recognize resemblances
and connections and to reflect on this propensity.
Hume confessed that he had not accounted for the
person who did the perceiving, recognized resem-

blances, and made connections. He had nevertheless
suggested a radically different character for human
knowledge.

The result was a picture of knowledge that resem-
bled a pointillist painting. There are no sharp outlines,
no definite shapes or solid blocks of color. Yet, the
circus performer on her horse can be seen perfectly
distinctly. Anyone who looks closely can discover that
she is made of points of paint. But it is impossible to
know whether these dots are mind or matter. Whether
they constitute appearance or reality of the painting
is a meaningless question. The painting consists of
nothing but the dots arranged as they are.

As there is in Hume's view of knowledge no realm
of eternally fixed essences or forms to be distinguished
from the transitory world of sensations, no line can
be drawn between “knowledge” (of reality) and “opin-
ion” (about appearances). None of the connections men
make among their perceptions of the outside world
is a necessary one, for its contrary is always conceivable
without contradiction. Man cannot then aspire to any-
thing but “opinion” in the ancient sense. (Hume re-
served the name “knowledge” for reasoning about the
relations of ideas but made it clear that this was not
the “knowledge” of the ancients. He also sometimes
spoke, in the ordinary fashion, of “knowledge” about
matters of fact, but without meaning to suggest that
such knowledge was anything more than probable, and
therefore “opinion” in the ancient sense.)

Yet some things could be known with more assurance
than others. We feel that our assertions are less proba-
ble when

(ibid., p. 154). There is no ground for denying that
further experience may remove this lack of certainty.
But where uncertainty persists—and Hume was in-
clined to believe it would in some matters—it arises
not from any contingency in the operation of cause
and effect but “from the secret operation of contrary
causes” (ibid., p. 132) too minute or remote to be
discovered. But there are also other kinds of assertions,
“proofs,” which are about cause and effect and yet
“entirely free from doubt and uncertainty” (ibid., p.
124). It is impossible to establish that the sun must
necessarily rise tomorrow; yet men are and should be
certain that it will rise. Such certainty is not merely
a practical aid to living, nor just an approximation to
truth good enough for ordinary purposes, nor merely
a conclusion that we have not the power to contradict.
It is a genuine certainty in the sense that there are
no grounds for doubting its truth, even though its
contrary is conceivable.

What justifies the certainty about tomorrow's sunrise
is the experience not merely of countless sunrises and
sunsets, but of a whole concatenation of events and
the relations men have discovered among them. From
this broad intricate pattern of experience, men have
come to believe in the uniformity of nature and this
belief has become linked in a complicated fashion with
a series of other beliefs which constitute what men
think of as natural laws. In other realms of experience
as well, men have over the years built up elaborate
compositions of observations and general propositions
issuing in certain beliefs. None of these is necessarily
true yet each is certain until its falsity is established.

As there is no ascent from appearance to reality,
there is no hierarchy of knowledge in the ancient sense,
nor is there any reason for assuming that men can build
a single logically coherent system encompassing all
knowledge. Although Hume did not explore the precise
relations of different kinds of knowledge, he indicated
how he thought about the question. He never doubted
that men can and should be as certain about the truth
of Newton's laws as about the evil of murder. On the
other hand, when a man thinks philosophically, he can
become certain that there are no demonstrably neces-
sary grounds for believing any proposition about mat-
ters of fact. But each of these “certainties” has a differ-
ent character. None has an exclusive claim to our
attention or respect, nor can it constitute, as Hume's
predecessors might have argued, the perfect conso-
nance of mind with the architecture of the universe.
Each certainty is arrived at differently, through differ-
ent kinds of procedures, rules, discriminations, and is
related differently to other propositions within its own
realm of knowledge. There is not a single system of
knowledge, but a number of differently organized ac-
counts of human experience, related to different points
of view.

By recognizing a variety of kinds of knowledge,
Hume saved men from being sentenced to irresolution
and perplexity. Uncertainty belonged only to the phi-
losopher's study. There the philosopher learned that
the grounds upon which human beings had to judge
and choose were made entirely by men in the course
of building civilization. But if God or Reason did not
guarantee any human truth, that was no reason to deny
that there was a human truth, upon which men could
base firm decisions and assertions, whether in science,
morals, or politics.

This human truth made it possible also to understand
that error arose from an inclination to confuse one kind
of resemblance with another, to pass from one idea
to a similar one without noticing the change (ibid., pp.
60, 61). Therefore the problem common to all knowl-
edge is that of discriminating correctly among resem-
blances, distinguishing true or relevant resemblances
from false and irrelevant ones. Long chains of reason-
ings or conjunctions or objects are more likely to in-
clude errors than simpler propositions. But making
such discriminations correctly can be helped by the
general rules that civilized men have come to accept.
These draw our attention to regularities or differences
that may be obscured by the peculiar circumstances
or disposition of the moment. Thus we come to realize
that although an object moved to a greater distance
from us appears to be smaller, it has not decreased
in size. Similarly, in morals and aesthetics general rules
help us to correct improper or crude internal sensations
of approval or disapproval by helping us to distinguish
what appears to be one sensation into a number of
associated ones, or to recognize similarities that are
easily overlooked because they are associated with
striking differences.

(Enquiry Concerning the Principles of Morals, par. 185).

Although he had made it impossible to anchor human
knowledge to the order of the universe, Hume believed
that language and knowledge corresponded to the
constitution of the universe. For there exists “a kind
of pre-established harmony between the course of na-
ture and the succession of our ideas” (ibid., par. 44).
How this correspondence comes into being, what
maintains it, or how exact it is, Hume put beyond
human power to know. And with this limitation Hume
bade men to be content. For such was “the whimsical
condition of mankind.”

The boldest answer to Hume was given in Germany
by Immanuel Kant. He was concerned, as Leibniz and
Descartes had been, with establishing that the truths
of both physics and religion were certain. But Newton's
development of experimental physics, combined with
the questions raised by Hume, had given the problem
a new turn. Hume had undoubtedly proved that neces-
sary truths could not be derived from experience. Yet
Newton's laws were, Kant believed, necessarily true
and unqualifiedly certain, and had been established
with the aid of experiments, not by pure reason. Unless
he could find a way of reconciling these two convic-
tions, Kant was obliged to deny either that Newton's
laws were certain or that their discovery in any way
depended on experience. He was troubled besides by
another conflict: Newton had revealed a nature entirely
subject to inescapable laws; yet men were conscious
of being free from such determination and had built
their moral life on such freedom. The certain truth
of experimental physics had then to be reconciled with
human freedom as well as with Hume's argument
against experience as a ground for certainty.

Instead of inquiring, as Hume had, into the validity
of our belief in the necessary connection between cause
and effect, Kant started from the certainty that the
principle of causality was necessarily true and set him-
self to accounting for it. He found the clue to his
answer in revising Hume's view of mathematics as
identical with logic and consisting only of relations
between ideas. Kant argued that mathematical propo-
sitions correctly describe the external world of experi-
ence and are accordingly synthetic, not analytic. But
as they are obtained through construction, not obser-
vation, they are a priori, that is, not derived from
experience. It followed that mathematical truths con-
stitute a kind of proposition that philosophers had not
previously distinguished from others, propositions that
were valid about experience but reached independently
of experience. Unlike empirical propositions that are
drawn from experience, synthetic a priori propositions
could be necessarily true. They therefore held the
solution to the problems raised by Hume's demon-
stration that experience could not yield necessary
truths. If Newton's laws were based on synthetic a
priori propositions, the certainty of scientific knowl-
edge could be saved without denying its connection
with experience. It remained to explain how men could
construct synthetic propositions without referring to
experience.

Kant found that when all content that might be
derived from experience had been removed from
mathematical judgments, there remained the notions
of space and time. He concluded that space and time
“are therefore pure intuitions that lie a priori at the
basis of the empirical” (Prolegomena, par. 10). Further
analysis produced other concepts that were the pre-
suppositions of experience but not drawn from it. And
finally Kant connected these concepts into a logical
system, enabling him to say “that these, and this exact
number only, constitute our whole cognition of things
from pure understanding” (ibid., par. 39). These cate-
gories formed the framework of knowledge common
to all human beings. They were the formal conditions
of experience, the sorting and ordering machinery for
sense intuitions whereby such intuitions become em-
pirical judgments. Without them, no experience would
be possible.

Experience had then to be understood as a con-

struction the mind imposed on sensation. This con-
struction had the shape of an inverted pyramid resting
on synthetic a priori concepts, upon which the mind
built a hierarchy of rules, rising from general and
necessary to particular and accidental. Perceptions are
channelled through these categories and so made to
constitute experience. In acquiring knowledge, the
mind moves downwards from empirical consciousness
of particular and accidental rules to pure consciousness
of its a priori concepts. Whereas common sense is
aware only of the more superficial rules without dis-
criminating carefully among them, the scientific mind
disentangles the necessary from accidental connections
among perceptions so as to make explicit this basic
structure. The answers science produces come from the
same source as its questions, for the mind discovers
what the mind had originally made. It operates in the
manner of the mathematician who first constructs a
figure and then learns what he has put into it. Nature
is then nothing but the order framed by the mind. Thus
Kant rejoined the knower to the known.

The role played by experience in Newtonian physics
could now be made clear: Newton had abstracted from
accidental appearances until he arrived at hypotheses
about the necessary conditions of those experiences.
But it was only when he found no exceptions to these
hypotheses in experience, that he could be sure he had
arrived at the basic laws. Without making experiments,
Kant insisted, we cannot know whether our concepts
are truly concepts of the understanding or merely
imaginary. Without constructing hypotheses before-
hand, we cannot know how to question experience or
how to make our accidental observations yield a nec-
essary law.

As there was no pure content of the mind distinct
from the mental processes which made experience
possible, consciousness became identical with an act
of judgment. The mind forced experience to take up
a position within its own framework and by its own
constructive action produced knowledge. Knowing was
then no longer an activity of reception but a creative
organizing activity.

The certainty of knowledge had been secured, but
at the expense of a severe restriction on its domain.
Because we know by imposing our own categories on
experience, we can know things only in relation to
ourselves. What we know are the sensible effects of
things on us or what Kant called their “appearances”
or phenomena. To know “things in themselves” or
noumena, that is, what ultimately causes them to affect
us as they do or what governs their existence or consti-
tutes their essence, we should have to possess some
direct insight into their nature. This power, Kant was
as reluctant as Hume to grant to men. Reason in Kant's
view provided a framework that rendered experience
intelligible without offering any way of penetrating the
objects of experience: “For our understanding is not
a faculty of intuition, but of the connexion of given
intuitions [i.e., sense impressions] in experience” (ibid.,
par. 34). Knowing things as part of nature, as phe-
nomena, means knowing that they exist and how they
are related to other things as parts of chains of causa-
tion, but not why these causal chains are so or that
they must be so. Nevertheless we can be certain that
the sensible world exists in the form it presents to us.
Kant even believed that he could demonstrate the
reality of the outside world, and he severely repri-
manded philosophers who argued, as Hume did, that
it had to be taken on faith.

But for all the certainty that Kant could promise
about our knowledge of the sensible world, the restric-
tion of human knowledge was crucial for him. It en-
abled Kant to reconcile the world of mechanical ne-
cessity revealed by science with freedom of the will
and moral life. As there were two distinct ways of
knowing things, he could both defend the legitimate
claims of science against its enemies and reject its
arrogant attempts to encompass the whole of reality.
He could even insist on the possibility of discovering
universal laws of history without denying human free-
dom. For just as men can know the appearances of
other things, they can know themselves as objects of
sense subject to the same laws as other objects.

But by separating certain knowledge of appearances
from insight into things-in-themselves and into men as
moral beings, Kant set up an opposition between cer-
tainty and the aspiration after ultimate or higher truths.
Certainty pertained only to what might be called
“technical” knowledge of how things work as opposed
to knowing why they work so.

As all appearances are in one mind, they are known
only insofar as they are organized into one connected
world by that mind. The failure to put things in their
place in the mind's organization, Kant explained, is
what we mean by idiocy or madness. The search for
knowledge is therefore identical with the search for
some way of connecting all phenomena into a logically
coherent system. But unlike the disciples of Leibniz
and C. von Wolff, Kant never attempted to deduce
all truth from the principle of unity. He regarded it
rather as the basic regulative principle, guiding reason
in its organization of the data of experience by indi-
cating what problems it must solve. It is the assumption
that stimulates every intelligence to find a way of
reconciling apparent inconsistencies by discovering the
underlying unity.

Reason cannot, however, think this systematic unity,
Kant argued, without giving it an object. This object
is the concept of God. As experience cannot directly
display the object of its unity, there is no way of

establishing the existence, as in physics, of an object
corresponding to the conception of God. Nevertheless
God must be postulated in order that men may under-
stand the connection of things in the world of sense
as if they had their ground in such a being. Reason
has then shown both that the concept of God must
be assumed by reason and that it cannot be understood
by reason. The belief in God is therefore justified but
belongs to the realm of faith. In this fashion, Kant
supported religious faith while denying the rationalists'
extravagant pretensions to knowing things beyond ex-
perience.

By acknowledging that men had ideas of “the higher
world of reality,” even though he denied any possibility
of knowing it, Kant suggested that reason was capable
of operations other than discovering the laws of nature.
But just how such operations affected the conception
of reason as a whole was not made clear. For generally
Kant equated rational knowledge with the construction
of a system of laws. He described the understanding
as a “faculty of rules” occupied with investigating
appearances “in order to detect some rule in them”
and “rules so far as they are objective and therefore
necessarily depend upon knowledge of the object, are
called laws” (Critique of Pure Reason, A127). The
whole had to be constructed from the parts, but when
completed, the form of the whole would be the same
as if the parts had been deduced from a single princi-
ple. Its parts would be related to one another in the
manner of a mathematical demonstration and this
would be the mark of its certain truth. Although Kant
had denied that mathematics was identical with logic,
he tied certainty to the logic of mathematical reason-
ing. Certain knowledge in his view is always and only
a system of laws in which the more particular ones
are “only special determinations of still higher laws,
and the highest of these, under which the others all
stand, issue a priori from the understanding itself”
(ibid.). There is no variety within the realm of knowl-
edge such as Hume's picture permitted. In history as
in physics, what was knowledge had to have the order
of a system.

Kant saw no reason why all of experience should
not eventually come completely under the reign of law.
Although natural science, even after Newton, still con-
tained an area of “endless conjecture” (ibid., A481,
B509), there could be no retrogression in knowledge.
Logic was already complete because “since Aristotle
it has not required to retrace a single step, unless,
indeed, we care to count as improvements the removal
of certain needless subtleties or the clearer exposition
of its recognised teaching, features which concern the
elegance rather than the certainty of science” (ibid.,
B viii). Whatever laws had been established were es-
tablished forever.

Moral knowledge must be just as certain as any
other—“we must not venture upon an action on the
mere opinion that it is allowed, but must know that
it is so” (ibid., A823, 851). Because the unity of the
self is the only ground for certainty, we strive to impose
on our own lives a rational order and unity such as
we discover in nature and this is what we mean by
morality. We achieve morality by submitting all our
thoughts and actions to universal precepts and that is
the only way human lives can become stable and
dignified. Any principle that cannot provide such uni-
versal precepts, as for instance the principle of happi-
ness, is deficient as a moral principle.

Instead of varieties of knowledge, there are in Kant's
view only degrees of conviction, ascending from opin-
ion, which is known to be objectively deficient but is
subjectively satisfying, through belief to knowledge.
Even opinion, however, is somewhat dependent upon
certain knowledge. For we must never presume to
“opine” without knowing with certainty what connects
that opinion to truth, as well as the law governing that
connection, which although not complete, is yet more
than arbitrary fiction. Otherwise, “I have nothing but
opinion; it is all merely a play of the imagination,
without the least relation to truth” (ibid., A823, B851).

To affirmations that are justified only by the peculiar
character of the individual subject, Kant denied any
validity whatsoever. An affirmation worthy of being
a conviction must have some degree of objectivity, and
objectivity was for Kant strictly synonymous with being
universally valid. He never vacillated, as Plato did,
between understanding objectivity as what those en-
gaged in a particular enterprise can agree upon and
regarding objectivity as utterly independent of cir-
cumstances or person and the same for all men at all
times. Only the latter meaning was recognized by Kant
for whom universal validity was inseparable from truth.
But it is sometimes easier, Kant pointed out, to get
universal agreement than to demonstrate universal
validity. Universal agreement might therefore be used
as “the touchstone whereby we decide whether our
holding a thing to be true is conviction or mere per-
suasion” (ibid., A820, B848, cf. B4). By connecting in
this fashion the idea of certainty with that of agree-
ment, Kant suggested the possibility of making agree-
ment among men the criterion of truth.

But within his own philosophy, certainty is the es-
sence of any knowledge. It is equated with rationality
and endowed with its most comprehensive meaning,
from which later philosophers were to abstract some
aspect and reconstruct upon that a more limited ver-
sion of certainty.

What was thought about the question of certainty
after Hume and Kant can be understood as variations
on the themes they constructed. But two new subordi-

nate motifs were added by opponents of Hume—the
common sense philosophers, and by an avowed disci-
ple, Jeremy Bentham.

The so-called common sense philosophers, Thomas
Reid and James Beattie, wanted above all to refute
what they took to be Hume's denial that anything
could be known, and Berkeley's denial that matter
existed. Reid and Beattie refused to indulge in the
inordinate metaphysical reasoning and sophistry of
their predecessors and proposed instead to discover a
new ground for certainty in the constitution of human
beings. This was in keeping with the fashion established
by the end of the eighteenth century of contrasting
the hard truths of “common sense” with the fantasies
of philosophy. But the suggestion for a theory of com-
mon sense had in fact appeared earlier in Shaftesbury's
philosophy and had been developed somewhat in
Hume's theory of moral sentiments.

The only was to escape an infinite regress of grounds
for holding any truth and the uncertainty it produced,
Reid and Beattie argued, was to recognize that knowl-
edge was based on self-evident truths which, though
impossible to prove, are also impossible to disbelieve.
Otherwise there could be no reasoning because “All
reasoning must be from first principles; and for first
principles no other reason can be given but this, that
by the constitution of our nature we are under a neces-
sity of assenting to them” (Reid, An Inquiry into the
Human Mind [1823], p. 77). What identified these
intuitive first principles was

(Beattie, An Essay on the Nature and Immutability of
Truth, 8th ed., p. 34). Why we must trust to our consti-
tution or why it may not deceive us or lead to contra-
diction could not be explained. Men had to be satisfied
“to take things as they find them,” and “believe Nature
upon her bare declaration, without suspecting her of
any design to impose upon them” (ibid., p. 38).

Only prejudice or “inveterate opinion” might be
rendered more evident by reasoning or argumentation;
the truth of common sense could only be displayed.
The crucial distinction then is not that between con-
tingent and necessary truths, but between reasoned and
self-evident truths. Some truths of common sense are
however necessary because their contraries are incon-
ceivable, and these include the first principles of morals
along with the axioms of geometry. But other self-
evident truths command assent rather than conviction,
because the contrary of “I do not feel this hard body”
is equally conceivable. Beattie further distinguished
degrees of moral assent descending from the highest
“moral certainty” through several stages of opinion,
to “that suspense of judgment which is called doubt.”
We may “without absurdity,” Beattie insisted, speak
of probable truth as well as of certain truth. For a
probable truth can be known to be probable with the
same certainty as a certain truth is known. It can be
as “universal” and “permanent” because it is also a
truth which “a rational being is determined by... his
nature to admit” (ibid., p. 25). For all their antipathy
to Hume, the common sense philosophers followed him
in separating certainty from necessity and equating
certainty with “belief beyond doubt,” so that they were
able to include within its magic circle synthetic propo-
sitions as well as purely formal propositions without
encountering Kant's problem of having to justify syn-
thetic a priori propositions.

Implicit in the common sense philosophies are the
main assumptions of later attempts to escape from the
problems raised by Hume without recourse to meta-
physics or to a priori knowledge. The starting point
of all speculation was made the whole of consciousness,
within which, by careful reflection, one could distin-
guish the foundations from the superstructure. The
direct object of thought was taken to be not, as Locke,
Hume, and Berkeley had maintained, the contents of
the mind which mediated between the knower and the
external world, but the external object itself. This direct
relation between thought and the external world sug-
gested a new ground for certainty without recourse
to the pure reason of Kant.

Quite another avenue to certainty was laid out by
the utilitarian followers of Hume, who emphasized the
shortcomings of common sense. Jeremy Bentham was
concerned with the practical problem of how to make
life more pleasurable for more human beings rather
than with any speculative questions about how men
come to formulate and judge knowledge or truth. His
thinking began from a translation of Hume's descrip-
tion of ideas as reflections of sense impressions into
the postulate that all human thoughts and actions could
be traced back to sensations of pleasure and pain.
Confusion, fostered by language that obscured the
pains and pleasures inspiring or produced by some
course of action, Bentham was convinced, caused all
the difficulties and injustices afflicting men. Any state-
ment about human beings that could not be reduced
to pleasures and pains felt by someone must be re-
jected, and a rigorous application of logic would make
the references of words perfectly clear. Human en-
deavors that could not be subjected to such analysis
were delusive, fraudulent, and dangerous.

Bentham accordingly devoted himself to con-
structing an utterly clear, technical vocabulary for

politics and law, as well as to making exact definitions
of crucial concepts and formulating rules for combining
words and sentences without obscurity. This was to
be the foundation for a rational system of law entirely
derived from a single principle and its corollaries, that
would need no interpretation and could, with but
minor adjustments, do for all nations. Although he
himself never completed his code of laws, Bentham's
equation of certainty with unambiguity, reached
through analysis of language and precise definition of
basic terms, set a pattern taken up later by analytical
philosophers.

In the nineteenth century, the new meanings given
to the idea of certainty were the work of social re-
formers who had either professional or amateur inter-
ests in the natural sciences and hoped to find more
reliable answers to social problems by applying the
assumptions and methods of the natural sciences to
social phenomena. There was a model for them in the
efforts of the French Encyclopedists to make a new
social science continuous with the natural sciences. A
number of different schemes were proposed: P. H. D.
d'Holbach emphasized the reign of necessary laws in
all realms which made it as easy to construct a system
of ethical and political values on material grounds as
a system of physical laws. É. B. de Condillac hoped
to reach exact knowledge by discovering the simple
primary facts upon which purely logical operations
could build generalizations about the human mind. The
most clear-cut justification for expecting certain truth
about human affairs was given by a mathematician,
M. J. de Condorcet, whose ideas became influential
in England through the work of Auguste Comte.

Condorcet rested all his hopes on the calculus of
probabilities. There was no essential difference be-
tween the natural and social sciences, he argued, be-
cause all laws about experience are no more than
probable. Where the observation of facts is more diffi-
cult, as in the social sciences, the truth will be less
probable, yet there can be certainty about probable
truth because probability theory makes it possible to
evaluate mathematically all statements and to reach
a certain estimate of their reliability by applying the
calculus of probability. It should even be possible to
discover, Condorcet believed, under what conditions
the decision of an assembly or a tribunal could be
guaranteed to be more truly representative. Thus al-
though certainty was to be had only about the proba-
bility of truth, certainty remained the criterion of
acceptable knowledge and mathematics, the paradigm
of certainty, and at the same time social as well as
natural sciences were rendered valid.

The English attempts to produce a unified science
were not quite so simple. Bentham's outstanding disci
ple, John Stuart Mill, wanted to build a system em-
bracing all aspects of human life and exhibiting the
unbroken chain of causality from natural to mental
effects on utilitarian principles. But he recognized that
there were special difficulties about moving correctly
to a general theory from simple facts for which the
ordinary logic had no remedy, and he became absorbed
in what is now known as the problem of induction.

This problem had already been carefully discussed
by a distinguished scientist, William Whewell, with a
better claim to be heard on the character of scientific
discovery. Whewell had argued that logic could not
bring certainty into science because discovery de-
pended ultimately on personal genius and not on im-
personal correctness of method, and that the progress
of science would draw attention to different aspects
of reality rather than ascend to some one set of perma-
nent truths. But Mill promised more. He was convinced
that the key to correct induction lay in logic—“the
laws of the investigation of truth by means of extrinsic
evidence whether rationative or inductive” (Letters,
Sept. 28, 1839). His “Four Methods of Experimental
Inquiry,” described in his Logic, were designed to make
induction safe, and to guarantee that scientific truth
was unquestionably valid even if not demonstrably
certain.

Many of the leading scientists in the nineteenth
century shared Mill's dream of accounting for all natu-
ral and mental phenomena in a single theory. They
regarded science as a growing collection of certain
truths that could now go further to provide what T. H.
Huxley recommended as “a complete theory of life,
based upon a clear knowledge alike of its possibilities
and of its limitations” (Essays, III, 143). But the pattern
of their theories was different from Mill's. It had more
in common with the positivist philosophy of Auguste
Comte, in whom Mill had found his inspiration al-
though he had rejected Comte's social philosophy.

Suggestions for a unifying theory, that reduced the
most complex things to combinations of the most
primitive ones, came from both biologists and physi-
cists. A number of different lines of biological and
geological research converged in Darwin's Origin of
Species, which reduced superficial resemblances to
blood relationships explained by a purely mechanical
process of chance variations and natural selection.
From this Huxley drew a conclusion that became
widely accepted—that the difference between the
lowest plant and the highest animal was one of degree
not kind, and depended solely on the division of labor.
The physicists suggested another synthesis in terms of
matter and energy that made it seem reasonable to
hope that physics would one day deduce “Hamlet”
from the molecular forces in a mutton chop. The

grandest attempt to generalize the results of science
was made by Herbert Spencer. His philosophy of evo-
lutionism explained all scientific laws, social as well
as biological and physical, as variations on a single
fundamental law: that the universe and everything in
it were moving towards a state of equilibrium in which
all forces in the world would be perfectly balanced.
Although he relegated questions about the origin of
things or the existence of God to the “Unknowable,”
he believed that the answers to all other questions,
whether political or mechanical, could be found in his
theory.

Those who were influenced by Spencer's faith in
scientific synthesis, even when they disagreed with his
version of it, were confident that as long as they dealt
in scientific terminology and renounced excursions into
the “Unknowable” they were dealing in certainties.
The diversity of prescriptions based on supposedly
scientific ideas, such as the struggle for survival, did
not prevent Sidney and Beatrice Webb, along with
other social reformers, from recommending the politi-
cal organization that they favored as an indisputable
scientific truth. But beyond agreeing on the impro-
priety of metaphysical speculation and the exclusive
validity of the experimental method of science, none
of these reformers endeavored to explain how scientific
truths were discovered or justified.

The idea that had inspired the scientific syntheses—
that the truth about the universe would be discovered
when men could construct a system that displayed the
connection of all phenomena with some ruling princi-
ple—had been suggested by Hegel. He had adapted
Kant's principle, that truth was guaranteed by logical
completeness, to a system that was designed not only
to include noumena as well as phenomena, but also
to reveal that the whole was but the unfolding of all
the potentialities of Spirit in the course of history.
Unlike most of Hume's other successors, however,
Hegel recognized a variety of truths and certainties.
Absolute Truth is to be found only at the stage of
Systematic Science (represented by Hegel's philosophy)
when it has become possible to survey the whole and
to see how and when each new phase of experience
emerged. Nevertheless each of these phases, Hegel
emphasized, remains a permanent and distinct part of
human life. Knowledge has a variety of moments and
the moments assume the form of modes of conscious-
ness, each of which contributes to the whole. The
certainty of sense-experience is different from the cer-
tainty of moral self-consciousness or of demonstrative
reasoning. To questions such as when Caesar was born,
how large a space is, or how the hypotenuse and sides
of a right-angled triangle are related, it is proper to
give exact answers. But philosophic truth has a totally
different character and must therefore take another
form. Each certainty is appropriate in its own sphere,
and makes a distinctive contribution to the whole. In
Hegel's view it became most important to recognize
the partial character of all knowledge short of Absolute
Truth, and the question, “Can human knowledge be
certain?” could have no single answer.

This variety of certainties was ignored by the philos-
ophies that took their stand on an opposition to Hegel.
For G. E. Moore and Bertrand Russell, Hegel's attempt
to show that the cosmos and man were made of the
same spiritual stuff seemed to deny the reality of
everything that common sense believed in, above all
that there is an external world independent of experi-
ence yet perceived as “facts.” Truth, according to
Moore and Russell, consisted in the correspondence of
thought to the external world. Since Hegel described
the relation between thought and its objects differently,
they concluded that Hegel had rejected the possibility
of discovering any truth.

Moore reaffirmed the certainties of common sense.
He claimed that he knew a great number of proposi-
tions without a possibility of doubt, such as that he
had two hands or was standing up and speaking. These
were, he admitted, not necessary but contingent prop-
ositions, which were neither known a priori nor tauto-
logical and could be denied without contradiction.
They were nevertheless “absolutely certain.” But the
self-evident truths of common sense were not enough
for Russell, who more than any other modern philoso-
pher shared Kant's passion for discovering a realm of
pure certainty. He required besides that the ground
for certainty be totally independent of human beings.
In Russell's world men were only puny inhabitants of
a remote corner of the universe, not the center or apex
of creation. If what men knew were wholly derived
from themselves, it could reveal nothing significant
about a universe that was vastly more nonhuman than
human. Yet the traditional sorts of impersonal knowl-
edge were ruled out. On the one hand Russell was
committed to the empiricist belief that sensation was
the ultimate ground of knowledge, and he could not
therefore look to a metaphysical reality. But on the
other hand, neither could he accept an empiricist solu-
tion such as Hume's because he wanted above all to
establish that men could discover certain truth about
reality.

Russell began his search for certainty where Kant
had, by considering mathematics, but in order to dis-
prove Kant's view that mathematics depended on
human construction. He hoped to derive all of mathe-
matics from a few basic logical principles, which were,
if not literally Platonic Forms, at least analogous in-
sofar as they constituted a “region of absolute necessity,

to which not only the actual world, but every possible
world, must conform...” (“Study of Mathematics,”
in Philosophical Essays, p. 82). This dream of a perfect
logical structure on foundations independent of human
thought was broken when Russell discovered a number
of contradictions in his system that could be removed
only by more or less clumsy makeshifts. It was ended
finally by Wittgenstein, who convinced him that math-
ematics was after all, as Hume had argued, entirely
tautological. From the philosophy of mathematics,
Russell turned to seek a more modest kind of certainty
in science which “is at no moment quite right but it
is seldom quite wrong.... It is, therefore, rational to
accept it hypothetically...” (My Philosophical Devel-
opment, p. 17).

Although he no longer believed in the real existence
of logical concepts, Russell kept his faith in logic as
an instrument for achieving certainty. By logic he
meant the kind of analysis used in the Principia Mathe-
matica to break up complex propositions into their
components. Its success in reducing the whole of
mathematical knowledge to a more systematic unity
confirmed the validity of logical analysis for Russell. But
he had already declared his allegiance to it when he
had rejected idealism, and with it the tenet that made
analysis useless—that the world of human experience
is an organic whole and therefore more than the sum
of its parts. From this it followed, Russell believed,
that nothing could be known about a part in isolation
from the rest because anything short of complete
knowledge about an organic whole is an abstraction
and necessarily falsifies the relation of the part to the
whole. As human beings could never become omni-
scient, idealism implied that men could know nothing
for certain. Anyone looking for certainty had to turn
to a totally different conception of the universe—as
an agglomeration of discrete packets that can be bun-
dled together or undone into smaller ones without
affecting the character of the parts. If, as Russell as-
sumed, the language of common sense and science
mirrored this agglomeration, then every complex
proposition or set of propositions could be resolved
into smaller, more basic elements which might provide
a perfectly secure foundation for the rest.

Russell began by analyzing common sense beliefs to
discover whether there was “any knowledge in the
world which is so certain that no reasonable man could
doubt it” (Problems of Philosophy, p. 9). He found that
although we can question the existence of material
objects such as a table, we cannot doubt “the existence
of the sense data which made us think there was a
table” and he concluded: “Whatever else may be
doubtful, some at least of our immediate experiences
seem absolutely certain” (ibid., p. 27). Two questions
had then to be answered: What constitutes data? How
do we move from such data to knowledge, i.e., to our
belief in material objects and finally to the laws of
physics?

At first Russell hoped to establish an absolute dis-
tinction between a pure, luminously certain, nonhuman
source of truths that he described variously as “data,”
“hard data,” “primitive knowledge,” “knowledge by
acquaintance,” all indicating the hard core of knowl-
edge as opposed to inferred or derivative knowledge
where error arose. Analysis would arrange knowledge
in “a hierarchy of dubitables” which would make clear
the certain, sensory core along with the succession of
inferences based on it so that their correctness could
be checked and rendered more certain or appropriately
graded for lack of certainty. But the notion of data
was gradually enlarged. Even at the outset, Russell
included within it what he called “universals” (“yel-
low” as opposed to “this yellow ball”; relations of
before and after, up and down, resemblance), as well
as general propositions of logic. Later he added “per-
ceptions,” “beliefs,” and “propositions,” as he found
that he could not eliminate but only “whittle away
the element of interpretation” (Inquiry into Meaning
and Truth, p. 124). He had in the end to conclude that
once anything was said, even if only “there is a chair,”
the speaker was necessarily making generalizations and
inferences because the pure datum could not be ar-
ticulated. The notion of “data” became a limiting
concept that at most “we can approach asymptoti-
cally” (ibid., p. 124).

In much the same fashion inference also lost its
purity. There appeared to be different kinds of infer-
ences, the most primitive being those of common sense
which men were obliged to make by biological neces-
sity. But inductive inferences yielded at best only
probabilities which “are not disproved when what they
show to be probable does not happen” (ibid., p. 317),
and thus seemed more likely to produce falsehood than
truth when not corrected by common sense. And finally
Russell felt obliged to declare that induction rested on
“a priori non-demonstrative forms of inference [such
as the principle of causality] which experience can
neither confirm nor confute, but which we regard, in
some circumstances, as more certain than the evidence
of the senses” (ibid., p. 317). There was no mistakeproof
method for building up knowledge, even if an inde-
structible foundation for it had been discovered.
Knowledge had no simple structure with sharply de-
fined parts but rested on many mutually supporting
piers.

When he could discover no “set of beliefs which are
never mistaken” nor any test which “will always enable
us to discriminate between true and false beliefs”

(Analysis of Mind, p. 268), Russell's faith came to rest
on reducing the risk of error. If doubtful inferences
and unproved assumptions could not be altogether
eliminated, some could be avoided or translated into
less doubtful propositions by the technique of logical
construction. This was suggested by the theory of de-
scriptions developed in the Principia which distin-
guished two kinds of subjects in propositions: proper
names representing something that can be directly
perceived, and incomplete symbols denoting something
that may never have existed and gets its meaning only
from the context in which it appears. Russell found
that incomplete symbols could be replaced by an enu-
meration of qualities and were thus shown to be merely
linguistic devices summing up a collection of particu-
lars. As Russell dissolved more and more “proper
names” into descriptive phrases, he came to believe
that although common sense sees the world as com-
posed of things having certain qualities, it would be more
precise to speak only of qualities existing in a particular
place and time (that there is “redness,” “squareness,”
“heaviness” here, and not “there is a red table here”).
For qualities alone are known to experience. The sub-
ject to which they are ordinarily attributed cannot be
experienced and is therefore unknown and unknowa-
ble, indeed nothing but a fanciful and pointless imposi-
tion on what is known.

In his early enthusiasm over the technique of logical
construction, Russell hoped that he might be able to
purify science of any entities that could not be ana-
lyzed into perceptions of qualities. But later he ad-
mitted that at most the number of such entities could
be reduced, that scientific concepts such as “light
waves” were built up not only of perceived events but
also of “unperceived aspects” of reality. (Cf. Schilpp,
pp. 707f.) Objects of perception were only part of the
subject matter of physics, and it was irrelevant to the
truth of physical laws whether all that they described
could be perceived.

He also modified his view of what scientific truth
was about. Science gave an account not of the whole
of reality, but only of its “structure.” By “structure”
Russell meant a recurring pattern of relations in time
and space among events whose causes or subjects are
unknown to us. He understood it as analogous to the
form of a proposition which the logician discovers by
replacing the actual words of a sentence with symbols
or variables related in the same way so as to show how
sentences saying different things have in common the
same logical form. At other times Russell likened the
“structure” of reality to a map that indicates the spatial
relations between points without telling us anything
about the nature of things indicated by those points.
Because of its formal character, i.e., because it indi
cated possible relations without specifying the char-
acter of particular things, the structure revealed by
science seemed to share somehow in the certainty of
purely formal or analytical truth. Yet it was, Russell
insisted, just as real as the real world, only it left out
“that essence of individuality which always eludes
words and baffles description, but which, for that very
reason, is irrelevant to science” (Introduction to Math-
ematical Philosophy, p. 61).

With what might be called his “empirical formal-
ism,” Russell granted science something less than the
certainty established by Kant. Nor could he refute the
skeptic, “whose position must remain logically un-
assailable,” as he could affirm only that the con-
gruence between reality and the causal laws discovered
by science “must be regarded as a fortunate accident
and how long it will continue we cannot tell” (Analysis
of Mind, p. 271). Still, he had made truth more nearly
independent of human creation and will, and had
shown that science could achieve “definite answers”
and “successive approximations to the truth in which
each new stage results from an improvement but not
a rejection of what had gone before” (A History of
Western Philosophy, p. 789).

But Russell had extracted this promise of progress
“towards the unattainable ideal of impeccable knowl-
edge” (Analysis of Mind, p. 271) by renouncing much
of what has been commonly taken to be knowledge.
Although he followed Hume in giving knowledge the
character of a painting made up of many points, he
did not conclude, as Hume had, that we must resign
ourselves to a mystery about how the dots of paint
produce the figure on the horse. Instead Russell insisted
that we concentrate wholly on describing in mathe-
matical terms the relations among the colored points
in space. About that, he believed, men could be almost
certain. The rest, which was bound to be seen differ-
ently by each observer and might even be an illusion,
he preferred to ignore.

A more tenacious effort to carry out Russell's original
program for certainty was made by the Vienna circle
of logical positivists, who took their stand on the doc-
trine suggested by Russell and developed in Ludwig
Wittgenstein's Tractatus (1922), that there were ele-
mentary statements corresponding to simple facts. In
The Unity of Science (1934), Rudolf Carnap declared
that the “protocols” of “given direct experiences...
constitute the basis of the entire scientific edifice”
(Unity of Science, pp. 76, 93). Far more reluctant than
Russell to give up the search for an infallible sensory
core to knowledge, the positivists become engrossed
in the difficulties of explaining why one man's cer-
tainty, if derived from his sense experience (as they
insisted it must be) should be valid for any other.

When they could find no way of avoiding the con-
clusion that knowledge was either unverifiable or sub-
jective, the logical positivists retreated from seeking
certainty into arguing that there was no way of estab-
lishing a correspondence between knowledge and fact.
The possibility of discovering any reasonable ground
for believing in a relationship between words and the
world seemed to have been destroyed by Wittgenstein's
Philosophical Investigations (1953). The only certainty
left was the possession of language and the linguistic
analysts devoted themselves to analyzing the use of
language without committing themselves to any beliefs
about the nature of reality.

A curious reflection of the positivist progress from
the certainty of protocol statements to linguistic anal-
ysis is to be found in the development of the arts in
the twentieth century. The early dadaists hoped to
find the objective stuff of art by ridding themselves
of all social and aesthetic restraints and giving a free
rein to the unconscious or chance. Their more recent
successors, such as the so-called “new novelists” of
France, have been more inclined to look for the un-
adulterated truth in impersonal and detailed accounts
of objects, gestures, movements stripped of comment
and interpretation. But all these efforts have come to
the conclusion that order and truth are merely conven-
tional, and therefore fraudulent, and that reality is
nowhere to be found. The artist is left with nothing
to do but to reflect on the nature of his creative acts
and to proclaim the only certainty remaining—that he
can destroy what he makes.

While the positivists were attempting to find cer-
tainty by making knowledge independent of human
will, others were returning to the Kantian emphasis
on the intervention of the knower. They did so in a
variety of ways, but all refused to regard physical
science as the model of final truth.

The phenomenology of Edmund Husserl reaffirmed
the power of reason to intuit the content as well as
the form of truth and restored respectability to the
hope for certainty about the nature of being. There
is in Husserl's work an echo of Goethe's scientific
passion and his search for the Urpflanze (the funda-
mental or ideal form of the flower) as well as of his
criticism of the natural scientists for their indifference
to Truth. But whereas Goethe had insisted on the unity
of man and nature, Husserl maintained that we know
only what is present to our consciousness which is a
continuous whole. It is meaningless therefore to speak
of “reality” or “noumena” outside consciousness. By
methodically analyzing consciousness, we can however
discover within it the “essences” of the objects of
consciousness, that is, what is necessarily true about
an object as distinct from the accidental and changing
qualities attached to it by custom or prejudice or
evident only in some of its individual manifestations.

Every object of consciousness, whether the color red
or a horse, real or fictitious, has an essence if it can
be clearly distinguished from what is essentially some-
thing else. Each kind of essence must however be
described in a different manner. The essence of mathe-
matical concepts may be described “exactly” because
it is unchanging; other objects of consciousness can only
be described “strictly,” that is, narrowly enough to
distinguish absolutely and necessarily between one
essence and another. The essence of science is to arrive
at complete certitude. This aim is implicit in all sci-
ence, and yet cannot be realized by any positive sci-
ence which concerns itself with how things act, and
not with what they are. To discover the essences of
scientific concepts is the task of philosophy and it is
therefore the only true science, the science of the
sciences.

A description of an essence is valid when it could
not be otherwise for anyone else studying the object.
But there is no proof of phenomenological truth, only
the hope that others will see things in the same way.
Neither is there a system growing out of one funda-
mental principle into which each essential truth must
fit, nor can the truth about one essence be derived from
knowledge of another. Each must be examined by itself
and verified in itself and the edifice can never be
complete. But with the help of correct phenomeno-
logical analysis, scholars can discover, and know with
certainty, the essences of those things about which men
have been speaking since the beginning of time.

Some of the scientists, however, as a result of the
new quantum theory developed in the twentieth cen-
tury, were denying that certainty was the essence of
science. Quantum theory united two conceptions that
had formerly been thought to be mutually incompati-
ble—the wave theory and the corpuscular theory of
energy. When quantum theory declared both of them
to be equivalent and equally fundamental for explain-
ing microphenomena, the reality of scientific truth
seemed to be put in doubt. Besides, it was found im-
possible to determine precisely both the energy and
the position of an electron. The state of atomic parti-
cles appeared to be altered by the measurement of
them, so that the precise state of a particle after meas-
urement could not be told from the measurement and
the repetition of the experiment under apparently
identical conditions might produce different results.
Instead of one exact description of microparticles,
quantum theory yielded only packets of possible values
for the variables in question, that is, probabilities. This
limitation on the degree to which the state of micro-
particles could be precisely determined was summed

up by Werner Heisenberg in the Uncertainty Principle;
moreover, the uncertainty of position and momentum
was only one of many uncertainties intrinsic to quan-
tum theory. Nevertheless quantum theory successfully
explained a great deal and made a major revision of
physics.

Some scientists saw in the Uncertainty Principle
nothing more than an expression of the limits of human
knowledge and the crudeness of measurement which
did not rule out the possibility of discovering an en-
tirely new explanation, free from the restrictions of
quantum theory. But Heisenberg concluded that quan-
tum theory had shown elementary particles to be al-
ways in only partially defined states, thus endowing
nature with a random or free quality. The assumption
of classical physics—that nature obeyed fixed laws
unaffected by human observation—was therefore false.
Physics, as Heisenberg came to see it, did not describe
an independent nature but only our relationship with
nature.

That science does not establish certain truth has been
the thesis also of Karl Popper, but for quite other
reasons. Nor is Popper's argument based on the other
common ground for denying the certainty of science,
that it yields only probable truth. Such an assertion,
he believes, confuses two senses of probability, “verisi-
militude” and “the calculus of probability.” If the latter
is meant, then there is nothing uncertain about the
results obtained from the calculus of probability. One
need recognize only that the answers obtained pertain
to a range of events rather than to a single event. If
verisimilitude is meant, then scientific theories should
not have a high degree of probability.

Indeed “it can even be shown that all theories,
including the best, have the same probability, namely
zero” (Conjectures and Refutations, p. 192). For the
truth of a theory depends not on its being irrefutably
established, but on its having been submitted to rigor-
ous attempts to refute it that have failed. The less
content a theory has, the more difficult it is to refute
and the more likely that it will not turn out to be false.
If probability of truthfulness were the only criterion,
the ideal scientific theory would be a tautology. Valua-
ble scientific theories, however, have a high content
because they attempt to explain a great deal, and are
therefore more likely to be refuted. The heart of sci-
ence, according to Popper, is refutation; what matters
is not the “probability” of a theory, but what it can
explain and the degree to which it has survived severe
tests or been “corroborated”: “For us therefore science
has nothing to do with the quest for certainty or prob-
ability or reliability” (Conjectures and Refutations, p.
229).

Popper bases this un-Kantian view of the uncertainty
of science on a theory of knowledge in some ways
resembling Kant's; both Kant and Popper escape the
need to justify induction by denying that induction
takes place. Instead of trying to eliminate inter-
pretation and theorizing from science, as Russell had
done, Popper considers the invention of theories the
object of science, making observation a test, not a
source of scientific laws. There is neither an a priori
basis for knowledge nor any certainties based on sensa-
tion such as atomic facts or propositions. As in Hume's
picture, man remains a stranger in the universe, but
a stranger who is continually making guesses about the
nature of his world. The scientist attempts system-
atically to test these guesses. Logic, mathematics, sci-
entific imagination combine to suggest connections
among guesses established as true until a coherent body
of scientific theory is built up. At any given moment,
some parts of this knowledge will be better corrobo-
rated than others; failures of corroboration may lead
to minor or to more basic revisions either in the theory
being tested or in the methods of observation. Al-
though, in any given test, some part of existing sci-
entific theory must be assumed to be true, no scientific
truth is permanently proof against refutation. It always
remains possible that a new test will overthrow it. Nor
is there any certainty about whether the facts taken
currently to be basic or essential will continue to be
so. For today's basic facts may be resolved into more
basic elements by tomorrow's theory.

Nevertheless, as in Russell's view, the reality of an
external world (and the relation to this reality) remains
the final arbiter of truth, and the scientist may believe
that an accepted scientific theory truly describes real-
ity. Science aims at truth without ever being certain.

What appears to be the most radical condemnation
of the quest for certainty was made by John Dewey.
The foundation of his pragmatic instrumentalism—that
every idea is a design for manipulating experi-
ence—had been used differently by C. S. Peirce and
William James. Dewey's interpretation made both
truth and certainty irrelevant to knowledge.

Dewey was disturbed by the cleavage between the-
ory and practice, knowledge and action that made
some lives and activities seem worthier than others,
and put science into conflict with moral values; such
a cleavage suggested that men had to choose between
scientific and moral progress. These were spurious
dichotomies, he argued, based on the belief that there
was a realm of pure rational activity which could hope
to achieve certain knowledge about immutable Being.
By contrast, practical activity that dealt with perish-
able, changing things and was therefore full of risk and
uncertainty was bound to be considered low and
irksome. Men were consequently always trying to

escape from practical activity into the safer realm of
theoretical certainty and they neglected the real possi-
bilities for putting themselves in stable possession of
the worthwhile things of life. But the success of the
experimental method of science, by installing doing as
the heart of knowing, has helped to break down these
misleading distinctions. They disappear altogether once
it is recognized that there is no higher, fixed truth or
good, indeed nothing but constantly changing events
and their interactions.

Living means interacting with one's environment.
In the course of their interaction with the outside
world, men encounter conflict, confusion, discomfort.
When a man becomes aware of something troublesome,
“he can make a change in himself either by running
away from trouble or by steeling himself to Stoic
endurance; or he can set to work to do something so
as to change the conditions of which unsatisfactoriness
is a quality” (The Quest for Certainty [1929], p. 222).
If he tries to solve the problem, he will imagine “hy-
pothetical solutions” and then test them. When his
“inquiry” is successful, the original “indeterminate
situation” is transformed into a “unified whole” (Logic
[1938], pp. 104f.). The final solution is a “warranted
assertion,” which Dewey proposed as a more precise
name for knowledge.

Human life is not then divided between thinking and
doing, but between needs, problems, hopes, fears,
aversions and the responses to these—ideas, inquiries,
hypotheses, solutions. There is nothing purely mental
or theoretical because “nothing merely mental can
actually resolve doubt or clarify confusion. At most
it can produce only a feeling of certainty—something
best obtained by withdrawing from the real world and
cultivating fantasies.... In principle, the corre-
spondence of what we do when a situation is practically
unsatisfactory with what happens in the case of intel-
lectual doubt is complete” (Quest for Certainty, pp.
221-22). Thought is not separate from action, but “...
a mode of directed overt action. Ideas are anticipatory
plans and designs which take effect in concrete recon-
structions of antecedent conditions of existence...
their worth has to be tested by the specified conse-
quences of their operation” (ibid., p. 160). Science
cannot conflict with the values men cherish, because
it is their most effective instrument for changing their
outer conditions so as to secure what they prefer. The
problem of knowledge is only the problem of discover-
ing how to redirect events and is “never ended, always
in process; one problematic situation is resolved and
another takes its place” (ibid., p. 281).

That this was a correct view of being and knowledge
was moreover corroborated by the latest scientific
achievement, Heisenberg's principle of uncertainty or
indeterminacy, as Dewey called it: “The principle of
indeterminacy thus presents itself as the final step in
the dislodgment of the old spectator theory of knowl-
edge. It marks the acknowledgement, within scientific
procedure itself, of the fact that knowing is one kind
of interaction which goes on within the world...”
(ibid., p. 195). Natural laws are shown to be not state-
ments about immutable properties but “a way of
transacting business effectively with concrete exist-
ences, a mode of regulation of our relations with them”
(ibid., p. 198). It is made obvious that nature is being
understood “not by a mind thinking about it from
without but by operations conducted from within,
operations which give it new relations summed up in
the production of a new individual object” (ibid., p.
205-06).

Thus Dewey translated Kant's prescription for
reaching certainty, through eliminating intellectual
contradictions and producing a logically perfect whole,
into a prescription for achieving security by removing
tensions in the practical world so as to organize it into
an organic unity. He thereby drew out the most ex-
treme implications of Kant's suggestion that certainty
depended on what the knower constructed, and made
explicit a new understanding of knowledge that has
become widely accepted in the twentieth century, even
by opponents of pragmatism such as existentialists. The
object of knowing is no longer to grasp that which
is, but to create and securely establish what is wanted.
Man overcomes his strangeness in the universe by
transforming it to fit the image he invents, and the
perfection of knowledge that once was thought of as
certainty becomes the unceasing reconstruction of the
world.

BIBLIOGRAPHY

James Beattie, An Essay on the Nature and Immutability
of Truth in Opposition to Sophistry and Scepticism (London,
1812). Jeremy Bentham, An Introduction to the Principles
of Morals and Legislation (1776; Oxford, 1948; various re-
prints). Rudolf Carnap, The Unity of Science (London, 1934).
John Dewey, Logic, The Theory of Inquiry (New York, 1938);
idem, The Quest for Certainty (New York, 1929; reprint
1960). David Hume, A Treatise of Human Nature (1739;
reprints New York, 1941, Oxford, 1951); idem, Enquiry
Concerning Human Understanding (1748; various reprints);
idem, Concerning the Principles of Morals (1751; various
reprints). T. H. Huxley, Essays, 9 vols. (London, 1893-94;
New York, 1894-1908). Immanuel Kant, Prolegomena (Chi-
cago, 1933); idem, Critique of Pure Reason (1781; 1787;
London, 1929; New York, 1934). John Stuart Mill, System
of Logic, 2 vols. (London, 1843; 8th ed. 1872; New York,
1929); idem, Letters, 2 vols. (London, 1910); idem, Auguste
Comte and Positivism (London, 1865). George Edward
Moore, Philosophical Papers (London and New York, 1959).

Karl Popper, Conjectures and Refutations (London, 1963;
New York, 1968). Thomas Reid, An Inquiry into the Human
Mind on the Principles of Common Sense (1764; London,
1923). Bertrand Russell, Analysis of Mind (London and New
York, 1921); idem, A History of Western Philosophy (London
and New York, 1946); idem, An Inquiry into Meaning and
Truth (London and New York, 1940; Baltimore, 1963); idem,
Introduction to Mathematical Philosophy (London, 1919;
New York, 1963); idem, My Philosophical Development
(London and New York, 1959); idem, Philosophical Essays
(London, 1910; various reprints); idem, The Problems of
Philosophy (London and New York, 1912). P. A. Schilpp, ed.,
The Philosophy of Bertrand Russell (Chicago, 1944). William
Whewell, History of the Inductive Sciences, 3rd ed. (1857;
reprint, 1968). Ludwig von Wittgenstein, Philosophical
Investigations (Oxford, 1953); idem, Tractatus Logico-
philosophicus (London, 1922; a new trans. by D. F. Pears
and B. F. McGuinness, London, 1961; New York, 1963).

SHIRLEY ROBIN LETWIN

[See also Causation; Certainty in Seventeenth-Century
Thought; Necessity; Positivism; Pragmatism; Probability.]