Académique Documents
Professionnel Documents
Culture Documents
Tenability” Catherine Elgin presents a theory of the justification of beliefs which argues
self-supporting web of beliefs in which all of the various, individual beliefs receive
support from one another via mutual corroboration and confirmation. According to this
something other than other beliefs, nor does a belief have to be justified via a chain of
beliefs which extend to infinity and which justify one another in a linear fashion. Rather,
a set of beliefs can be justified simply by the manner in which they come together or
'cohere' to form a coherent picture of the world which, through its very coherency, then
lends support to the various, individual beliefs which make it up. In this paper, I will first
provide an analysis of Elgin’s presentation of coherentism. I will then argue that it fails
justification from a coherentist theory of truth. The former claims that coherence amongst
a sufficiently large set of beliefs justifies belief in them, whereas the latter holds that
coherence amongst a sufficiently large set of claims makes them true. Elgin rejects the
coherentist theory of truth as implausible, but if coherence is not the grounds for truth,
that raises the question of why it would be the grounds for justification. If coherence
amongst a sufficiently large set of propositions does not guarantee the truth of these
propositions, why would it be the case that coherence amongst them justifies belief in
inference to the best explanation. The reason for why coherence provides justification for
belief in propositions is because coherence tracks the truth; and coherence tracks the truth
due to the fact that when a large set of propositions cohere with one another by
confirming and corroborating one another, the best explanation for why this is so is
because these propositions are providing an accurate account of reality. Hence, belief in a
sufficiently large set of beliefs which cohere with one another in an appropriate manner is
justified because the truth of these propositions is the most economical means of
accounting for the manner in which they come together to form a coherent web.
Elgin provides us with a nice example to illustrate the manner in which this
works. Suppose someone steals an object and various different witnesses, who have never
communicated with one another, claimed to have witnessed the theft. Furthermore, they
all provide accounts of the theft and descriptions of the thief which share various details
with one another. Now, even though each of the witnesses on their own is unreliable for
various reasons (one has poor vision; another one is a pathological liar, etc.) the fact that
their testimonies support one another, and the fact that they have never been in contact
with one another and therefore could not have formed a conspiracy, gives us great reason
to suspect that each of them is telling the truth and that the accounts they have provided
describe events as they actually occurred. In other words, although each of these
testimonials on its own is nearly worthless, together they generate a macro-account which
provides support for each of the individual accounts that make it up, such that the
macro-level account and the individual, lower-level accounts which make it up are
supporting one another. The lower-level accounts support the macro-level account by
constituting it, and the macro-level account in turn supports its constituents.
But why does the macro-level account provide such support for its constituents?
Because, given the manner in which these different accounts corroborate one another, and
given the fact that a conspiracy has been ruled out, the best explanation to account for
this interlocking of corroboration or this 'web of coherence' is that these accounts, despite
this manner that coherence amongst beliefs in general provides justification for them.
Elgin summarizes her position as follows: "The thesis of the sort of epistemological
suitably comprehensive, coherent account, when the best explanation of coherence is that
the account is at least roughly true" (p. 246). It is important to emphasize this connection
between coherence and inference to the best explanation because it is only when the best
explanation for the coherence of an account is the truth of that account that we are
alternative explanation which accounts for the coherence of an account, such as a writer's
skill with regards to the account put forward in a well-crafted novel or a person
subconsciously ignoring various sets of data which conflict with a coherent account that
he is trying to create, then coherence no longer justifies belief in the truth of this account
via the principle of inference to the best explanation because the truth of the account is no
longer the best explanation for its coherency. Hence, by tying coherence to the principle
of best explanation, Elgin is able to guard against the danger of sets of propositions which
are coherent for reasons other than truth. The claim that coherence alone justifies belief in
propositions could require belief in elaborate and coherent accounts which have no
connection to reality at all. By adding in that belief in an account is not warranted unless
the truth of that account best explains why the propositions that make up that account
corroborate one another, then various counter-examples where coherency clearly has
nothing to do with truth and thus does not justify belief are weeded out. In summary,
those propositions. What accounts for this coherency must also be taken into account or,
as Elgin puts it, "[C]oherence conduces to epistemic acceptability only when the best
explanation of the coherence of a constellation of claims is that they are (at least roughly)
true"(p. 248).i With this basic framework laid out, Elgin then spends the rest of the paper
responding to objections and laying out the implications and benefits of her position.
One of the more fascinating issues which Elgin discusses concerns the relation of
our beliefs to our perceptions and how this relationship conflicts or accords with
coherentism. Suppose that you are reading a novel which has been written from the
second-person point of view, i.e. one which purports to describe your life. This
'biography,' although described in a very engrossing manner, fails to capture your belief
because it conflicts with how you believe you life has really unfolded. But what justifies
the belief in one account of your life over another? Is it really because the propositions
which make up the account given of you in the novel fail to cohere with another, much
larger and more coherent web of propositions which provides an alternative account of
your life that you are justified in disbelieving the first set of propositions while believing
in your other biography? Or is the real source of justification the fact that your senses fail
account of your life that is corroborated by perception would enjoy justification in belief
over the novel-account, even if it were in fact less coherent? Is it really coherence which
The problem which Elgin is trying to get at here seems to be this. We have a
belief that perceptions should carry greater weight when it comes determining our beliefs
and that, no matter how elaborate and coherent a theory or account is, that account should
have no ability to override beliefs which are corroborated by our direct perceptions.
Elgin calls them, enjoy a primacy within our system of beliefs. But is this primacy really
compatible with the claim of coherentism that belief in the truth of a proposition is
justified only by that belief's coherence with other beliefs? If a coherent system of beliefs
about the world, a theory of nature, can be rejected by means of the fact that that theory
fails to conform to observation, then it would seem to be the case that it is perception
itself which provides the justification for believing some propositions while rejecting
although it is true that perceptual deliverances carry more weight than other kinds of
beliefs, this does not mean that such deliverances are immune to revision due to
are the result of perceptual experiences cannot be ignored, even if they do contradict an
elaborate and coherent network of beliefs because "[h]owever tightly woven an empirical
(p. 250). However, it is also clear that we do not always fully trust our observations but
rather, we oftentimes discount or re-interpret them in order to make them cohere with
other beliefs. Suppose for example that you believed that you had seen something but that
a group of people who were with you and whom you trust tell you that you are mistaken.
Your response to this situation would not be to simply ignore the testimonies of your
friends. Rather you would begin to attempt to find ways to force your perceptual
deliverance into conformity with your beliefs that your friends are honest people and that
a group of observers is more reliable than a single observer. You could do this by
blaming the lighting or questioning your eyesight. Perhaps you would even begin to
doubt your sanity. Regardless, it is clear that perceptions do not enjoy a complete right of
deliverances force us to change other beliefs that we have, but sometimes the flow of
The complex manner in which empirical observation can relate to theory can be
seen in the way that science oftentimes works nowadays as well. Oftentimes it is the case
existence is posited by that theory in order to explain various other phenomena, even
though they have no actual empirical corroboration for the existence of this phenomenon
observation conflicts with a well established theory, the first action of the scientists
performing the observation will not be to re-write or trash the theory in question, but
rather to hope and pray that the equipment is merely faulty. Perceptual deliverances and
empirical testing do not always get the last word. Both can be and often are revised to
force them to cohere with a wider network of beliefs which has greater weight.
Furthermore, Elgin also argues that the reason for why we give perceptual deliverances
such a great weight is precisely because the beliefs born of such deliverances really do
cohere together. Were it the case that those beliefs did not cohere with one another, than
we would be much less willing to trust our senses and give such weight to the beliefs
directly inspired by them and would instead focus upon other sources of knowledge.
to the manner in which modern, scientific investigations are carried out, I must,
of circularity arises for Elgin because it is simultaneously the case that Elgin on the one
hand, by the very hypothesis of her theory, needs to argue that all of our beliefs,
including even belief in the principle of non-contradiction, are justified due to the manner
in which they cohere with one another, and, on the other hand, she also needs to use the
does or does not cohere with other propositions in our web of beliefs. As a result, it is not
possible for Elgin to justify belief in the principle of non-contradiction itself. But because
her whole coherentist theory relies upon this principle, without it, her system falls apart.
Let us begin by reminding ourselves that Elgin’s central claim is that “epistemic
the best explanation of coherence is that the account is at least roughly true" (p. 246).
Elgin does not lay out a full account of what it means for different propositions to be
'coherent' with one another, but clearly, at least one of the necessary conditions for two
propositions to be coherent with one another is that they do not contradict one another.
Elgin herself clearly recognizes and accepts this: "[i]nconsistencies among beliefs
conclusively demonstrate that some of the beliefs are false" (p. 252). Hence, if I have one
belief that coheres with my larger system of beliefs, that belief is justified, whereas if I
have another belief which contradicts this system, then my belief in that proposition is
not justified. Hence, as a result, my set of justified beliefs will consist only of
propositions which are not contradictory with one another because the principle of
non-contradiction is one of the criteria which I use for ruling out beliefs as unjustified.
non-contradiction? Per coherentism, it must be the case that my belief in this principle is
justified by way of the fact that it coheres with my overall system of beliefs, whose
coherence is best explained by the truth of the propositions of that system. And the reason
for why the proposition stating that the principle of non-contradiction is true coheres so
nicely with this system is because there are no counter-examples to that principle within
this system. However, it should be immediately clear that the explanation for why that is
the case is entirely trivial in nature: there are no contradictions within this system of
beliefs because the principle of non-contradiction has already been used to rule out as
false all contender propositions for entry into this system which have been excluded
because they do not cohere with it. Hence, the principle of non-contradiction coheres
with this system because it has excluded from this system the possibility of there being
present any two propositions which would contradict one another, and which would
therefore contradict the principle of non-contradiction.
argue that coherence amongst different beliefs provides justification for them. At a
minimum, coherence requires that these beliefs do not contradict one another. Hence, if I
have a variety of beliefs, and some of them contradict one another, then I am not justified
in believing those beliefs which are outside of the larger web of beliefs that I have which
these beliefs are contradictory with. As such, I can do one of two things. Either I can
discount those beliefs in question, in which case I will be left over with a web of coherent
beliefs. But then, even though the principle of non-contradiction does cohere with these
beliefs, how can I justify my belief in that principle on this basis, given the fact that, via
the principle of justification via coherence, I have already used the principle of
non-contradiction to rule out all sets of incompatible beliefs which could serve as a
of beliefs, because I have already used that principle to rule out all contradictory belief
sets which could serve as counter-examples to it, i.e. with which it could not cohere. Or,
on the other hand, I could not rule out some of my beliefs merely on the basis that they
disagree with some of my other beliefs, in which case I do have contradictory sets of
beliefs with which my belief in non-contradiction does not cohere. The problem of
circularity here is that one cannot justify a principle by using that principle to rule out all
of its counter-examples. But this is precisely what Elgin does vis-à-vis the principle of
non-contradiction. Hence, she is not able to justify this principle at all. But if she cannot
justify it, then she has no basis for the claim that coherence is what justifies belief
because coherence no longer tracks the truth because it is now possible for contradictory
propositions to be true. Or, if the principle can be justified, it is for reasons other than
i
What prevents Elgin's view from simply collapsing into a version of the argument that the principle of
best explanation itself is simply what justifies our various beliefs, such as the belief in an external, material
world? Unlike Jonathan Vogel who argues that our belief in an external world is justified when we use the
principle of best explanation to posit the existence of such a world in order to account for our perceptions
of such a world, it seems to be Elgin's belief that we do not actually have to perform this inference to arrive
at a justified belief. Rather, we can arrive at a justified belief merely by noting the coherency that a belief
of ours has with our various other beliefs. So long as the truth of the overall account which is made up of
these beliefs is the best explanation for the coherency of this account, our belief in this individual belief in
question is justified, regardless of whether or not we actually go from the perceived coherency of this belief
with our other beliefs to the additional inference that the best explanation for the coherency of this overall
system of beliefs is the truth of the propositions being believed.