Truth Is Established by Coherence
Classic Philosophical Questions Part 5 - Knowledge
Francis Bradley, from Essays on Truth and Reality
Kem Stone - 8 November 2007
I think one common feature of most philosophical arguments is the unfair treatment usually given to the opponent’s
argument. The coherence theory of truth is far stronger, I believe, than Bertrand Russell depicts it, and it is well-
defended in this text by Francis Bradley. Bradley himself does not even address the correspondence theory in this
text, but merely argues against foundationalism as though this were a complete theory of truth in itself, and the only
serious opponent of coherence.
Bradley’s stated purpose in this argument is to defend three claims: 1- that the test of truth or falsehood in matters
of facts of perception or memory lies in coherence within a system, 2- that this test works satisfactorily and that no
other test will work, and 3- that there are no judgments that are in principle infallible. I believe that Bradley shows
that coherence within a system is a test for truth but not the only test. I would give my complete assent to the
third claim only.
That infallible truths exist, Bradley begins by showing, is virtually beside the point when we are trying to construct
a system of knowledge. According to Bradley, “in your search for independent facts and for infallible truths you
may go so low that when you have descended beyond the level of error, you find yourself below the level of any
fact or of any truth which you can use” (290). As a test of how strong an ultimate fact is, we must ask what we
are able to contradict on the basis of the fact. For example, Bradley takes the case of simple sensations a, b, and
c, from which you can deny only the denial that a, b, and c exist in any manner or sense. If we take a complex
feeling containing both a and b, we can only deny the denial that a and b can ever co-exist in feeling. But these
truths do not get us very far.
In order to raise these simple truths to the level of a particular fact of perception or memory that we can use to
construct a system of knowledge, we must state them in terms such as “I am here and now having a sensation or
complex of sensations of such or such a kind” (290). Yet we have now introduced several fallible elements into
our judgment. For one, it is difficult to pin-point what is meant by “I” and if it is stipulated that this is a self existing
in time, memory becomes involved and the judgment is subject to error. Second, if the “here” and “now”
elements are to mean a special place in a special order, they too are fallible. Not only that, but in the case of the
sensations themselves there is always a chance of sense-hallucination or a morbid fixed idea, special psychological
cases that could arguably render even the fact that one is experiencing such and such a sensation false. Yet even if
we grant that what is felt is felt and cannot be mistaken, the judgment always becomes fallible when the “I” or the
“here” and “now” are introduced.
Foundationalists might declare that in spite of the weakness of fallible truths, we have no choice but to contend
that they exist, otherwise we would have no certain knowledge upon which to build and everything would be
subject to scepticism. But Bradley rejects both the premises and the conclusion of this claim, asserting that “such
infallible and incorrigible judgments are really not required for our knowledge, and, since they cannot be shown,
we must not say that they exist” (292). Bradley agrees that we cannot suppose that it is possible that all
judgments of perception and memory are false, but he rejects the notion that that knowledge is a construction that
rests on a foundation of certainties. The foundation of truth, according to Bradley, is merely provisional: “In order
to begin my construction I take the foundation to be absolute—so much certainly is true. But that my construction
continues to rest on the beginnings of my knowledge is a conclusion which does not follow” (292).
For Bradley, experience is not a superstructure but a system, and the object of knowledge is to have as
comprehensive and coherent a system as possible. To attain this, we must both reflect and continually have
recourse to the materials of sense. Some facts must of course be relegated to the world of error, as it would be
impossible to maintain a coherent system if all facts presented to us were accepted. But we need not assume that
any facts are absolutely certain, but quite the contrary—that all facts are subject to error. We may regard a fact
as being true in so far as it coheres within our system, and as a result we will certainly find some facts that will
never need to be discarded. But even these strong facts without which the rest of our knowledge would be
incoherent are not in principle infallible, and they need not be. The criterion for truth is not certainty, but
something much simpler and less demanding. “Facts are valid so far as, when taken otherwise than as ‘real,’ they
bring disorder into my world” (293).
One objection to thinking of all facts as fallible is that truth can never be more than a matter of probability.
Bradley recognises this, but believes that it is something we must accept. Facts are justified when they make one’
s world-view wider and more harmonious, and the more harmonious this structure, the greater the probability of a
fact being true. “And if we could reach an all-embracing ordered whole, then our certainty would be absolute.
But, since we cannot do this, we have to remain content with relative probability” (293). Coherence theory would
not suffer from this problem if we were omniscient beings, because if we knew all facts there would be no room
for new facts to call into question the old, and our all-encompassing system of knowledge would be coherent,
complete and certain. However, because our minds are limited, we can never reach “The Truth” and must accept
that certainty is only a matter of degree.
One might ask Bradley how we are to trust facts of memory within such a system, as if all our memories are called
into question it may seem that our structure of knowledge collapses. But for Bradley there is really no difference
between facts of memory and facts of perception, as both are justified according to the same criterion. He is
justified in believing that Louis XVI was executed because if it were not the case that this well-known historical
event took place, such an error would bring too much disturbance into his world. “To take memory as in general
trustworthy, where I have no special reason for doubt, and to take the testimony of those persons, whom I
suppose to view the world as I view it, as being true, apart from special reason on the other side—these are
principles by which I construct my ordered world, such as it is” (294). We can assume the truth of facts of
memory just as we can assume the truth of widely accepted scientific theories. We can not verify either, but
unless there is a reason to doubt we need not do so.
Finally, it may be objected that according to this system one can invent an imaginary world even more orderly that
the known world, and that this world can be so well-constructed that the facts of the actual world might justifiably
be considered to be erroneous while the facts of the fancied world are considered true. Bradley does not take
this objection seriously, as it fails to properly understand his criterion for justification, which demands the inclusion
of all possible material. One cannot confine oneself within an imaginary world and determine what is true or false
from within that framework alone, but all of the materials of the sense must be given due consideration. Since the
world of perception stands firmly opposed to the fancied world, to consider the imaginary world true and the
actual world in error would not be justified according to Bradley’s model of coherence.
I believe that Bradley makes an excellent case for coherence, which stands up well to most criticism. Russell’s
critique of coherence in his argument for correspondence is valid, but it fails to do much justice to the theory,
which I believe has more merit than many are willing to admit. Yet Bradley is also guilty of presenting the
alternative to coherence as a far weaker position than it actually is. He certainly shows the flaw in believing that all
of our knowledge is built on a foundation of certain, infallible truths, but this is not a critique of correspondence but
foundationalism, which is not a complete epistemological theory in itself but merely a feature of many such
systems. Coherence is superior to foundationalism, but I do not believe it superior to correspondence for one
reason only: when a person speaks of something as being true, what is usually meant is that it corresponds to
some external reality, not that it coheres within our existing system of knowledge.
Coherence does have two major strengths which are much to its credit. The first is its compatibility with
fallibilism, which is a feature that I believe is essential to any valid epistemological theory. I believe that it is utterly
impossible for any fact of any kind to be considered absolutely certain. Even facts such as “I am experiencing
such and such a sensation” are subject to doubt for the reasons Bradley has given, such as the ambiguity over
what is referred to by “I” and so on. The second major strength of coherence is that it eliminates our need for
direct knowledge of the external world, which is the fatal flaw in correspondence theory. We can never know
anything beyond our own thoughts, so under the correspondence theory we can never know if anything is true.
With coherence we can designate a proposition is true if it makes sense within our existing framework of
knowledge without taking that leap of faith and stipulating that it corresponds to some external reality. The belief
that an external reality exists can be justified under coherence theory, but its truth is merely provisional and if it
turns out to be false it would not undermine coherence theory as it would correspondence.
The major weakness of coherence theory, however, is one that Russell touched upon, which is that we can have
conflicting coherent systems of knowledge and be unable to tell within the framework of coherence which is closer
to “The Truth”. To use the belief in an external world as an example, it can clearly be imagined how one could
have a complete and coherent system of knowledge that includes the belief that the real world is merely illusion
and that corporeal objects do not actually exist. Another person might have a nearly identical system of
knowledge but believe in the existence of the material world. From the principles of coherence alone we would
not be able to tell which system is closer to the way things really are. A more practical example would be to
judge between one person’s religious worldview and another’s atheism. If both systems are coherent they are
both equally valid, though they will undoubtedly contain many beliefs which conflict with one another. While the
belief that there is a God may be justified and true in one person’s system, the belief that there is no God will be
justified and true in the other.
Russell’s contention that coherence is invalidated by the fact that it must presuppose the truth of the laws of logic
such as the law of contradiction is an unfair objection in my opinion. Of course the coherence theory would
collapse if we assume the law of contradiction to be false, but so would every other piece of knowledge we
thought we knew. If a proposition and its contradiction could both be true, we might as well give up on all of
philosophy and science because knowledge of any kind would be impossible. We must assume the truth of the
laws of logic in any system, and to say that the system is flawed because it depends on the truth of these laws is
completely unfair. In any case, Bradley would surely respond just as he has in the case of memory: we may
assume the truth of a memory or the testimony of an authority unless we have special reason to doubt. Since we
have no reason to doubt the truth of the law of contradiction, we may accept it and continue from there.
So while coherence does have its weaknesses, and it is not what most people have in mind when they think of the
concept of truth, I believe it is a far stronger position than it is usually given credit for.