PHIL 440 (FALL 2006)
INSTRUCTOR: WILBURN
SESSION 4 LECTURE NOTES
9/21/06
Last Week: We looked at Chisholm's account of foundationalist justification, which is the
main one that we will be examining (even though we’ll look at a slightly different one later, in
a couple of weeks). This week we’ll look at two papers by Bonjour, the first one criticizing all
forms of foundationalism (with regard to empirical belief), and the second endorsing an
alternative coherentist account of justification.
1.2 CAN EMPIRICAL KNOWLEDGE HAVE A FOUNDATION?
Laurence Bonjour
Again, what is the doctrine of the given???
“the claim that certain empirical beliefs possess a degree of epistemic
justification or warrant which does not depend, inferentially or otherwise, on
the justification of other empirical beliefs, but is instead somehow
immediate or intrinsic.”
Bonjour doesn’t underestimate it. Why does Bonjour compare it to a hydra?
Bonjour’s goals in this paper:
(1) Unify the problem – “distinguish and clarify the main dialectical variants of
foundationalism, by viewing them as responses to one fundamental problem which is both the
main motivation and the primary obstacle for foundationalism.”
(2) Criticism empirical foundationalism of all stripes – “offer schematic reasons for doubting
whether any version of foundationalism is finally acceptable.”
What is the fundamental problem that motivates foundationalism and makes it see
inevitable? The regress problem – What is this again???
Bonjour agrees with Chisholm as to what the four potential solutions to the regress problem would
have to be like.
(i) The regress might terminate with beliefs for which no justification of any kind is available.
1
What’s wrong with this???
(ii) The regress might proceed infinitely backwards with ever more new premise beliefs being introduced
and then themselves requiring justification.
What’s wrong with this???
(iii) the regress might circle back upon itself, so that at some point beliefs which appeared earlier in
the sequence of justifying arguments are appealed to again as premises.
What’s wrong with (iii)?
Bonjour writes,
“Although the problem of the knower having to have an infinite number of
beliefs is no longer present, the regress itself, still infinite, now seems
undeniably vicious. For the justification of each of the beliefs which figure
in the circle seems now to presuppose its own epistemically prior
justification: such a belief must, paradoxically, be justified before it can be
justified.”
(iv) The regress might terminate because beliefs are reached which are justified—unlike those in alternative
(i)—but whose justification does not depend inferentially on other empirical beliefs and thus does
not raise any further issue of justification with respect to such beliefs.
Choice (iv) is the foundationalist’s option. The foundationalist considers it the only choice left
standing after all these objections have been considered. Bonjour wants to argue that that it really
doesn’t survive these objections because it itself falls prey to something like the regress of justification
problem.
“The epistemic regress argument will prove to be a two-edged sword, as
lethal to the foundationalist as it is to his opponents.”
Two interpretations of alternative (iv):
Strong vs. weak. What is that???
Strong foundationalism: “According to strong foundationalism, the foundational beliefs
which terminate the regress of justification possess sufficient epistemic warrant, independently of
any appeal to inference from (or coherence with) other empirical beliefs, to satisfy the
justification condition of knowledge and qualify as acceptable justifying premises for further
beliefs. Since the justification of these basic beliefs, as they have come to be called is thus
allegedly not dependent on that of any other empirical belief, they are uniquely able to provide
secure starting-points for the justification of empirical knowledge and stopping-points for the
regress of justification.” (Note that this is not a certainty claim.)
Weak foundationalism: “[This] is a view which may be called weak foundationalism.
2
Weak foundationalism accepts the central idea of foundationalism—viz. that certain empirical
beliefs possess a degree of independent epistemic justification or warrant which does not derive
from inference or coherence relations. But the weak foundationalist holds that these foundational
beliefs have only a quite low degree of warrant, much lower than that attributed to them by even
modest strong foundationalism and insufficient by itself to satisfy the justification condition for
knowledge or to qualify them as acceptable justifying premises for other beliefs. Thus this
independent warrant must somehow be augmented if knowledge is to be achieved, and the usual
appeal here is to coherence with other such minimally warranted beliefs.”
Why does he make this distinction? To point out that his objection is a very broad one,
applying to both strong and weak forms of foundationalism.
The central question that Bonjour asks: What could make a belief basic or foundational???
To answer this, he asks what the difference is between epistemic justification and other kinds of
justification. . . e.g., moral justification . . .
Epistemic justification alone has an “essential or internal relationship to the cognitive goal of
truth.”
This suggests something about the essential nature and function of basic epistemic beliefs.
If basic beliefs are to provide a secure foundation for empirical knowledge, if inference from
them is to be the sole basis for the justification of other empirical beliefs, then that feature,
whatever it may be, in virtue of which a belief qualifies as basic must also constitute a good reason
for thinking that the belief is true. If we let "f" represent this feature, then for a belief B to qualify
as basic in an acceptable foundationalist account, the premises of the following justificatory
argument must themselves be at least justified:
1.
Belief B has feature (f).
2.
Beliefs having feature f are highly likely to be true.
3.
Therefore, B is highly likely to be true.
Now comes the most general form of Bonjour’s argument:
“. . . if we now assume, reasonably enough, that for B to be justified for a
particular person (at a particular time) it is necessary, not merely that a
justification for B exist in the abstract, but that the person in question be in
cognitive possession of that justification, we get the result that B is not
3
basic after all since its justification depends on that of at least one other
empirical belief. If this is correct, strong foundationalism is untenable as a
solution to the regress problem (and an analogous argument will show
weak foundationalism to be similarly untenable).”
What is Bonjour saying? So, the problem is with which principle? (1) or (2)
How might we keep (2) from posing a problem??? Why does he start talking about
externalism, this account of Armstrong’s???
First, what is Armstrong’s account of what it is to be justified in holding a basic or noninferential belief?
What is he trying to accomplish when he writes this:
“What makes a true non-inferential belief a case of knowledge is some
natural relation which holds between the belief-state...and the situation
which makes the belief true?”
in particular, where
“there is law-like connection between the state of affairs Bap [i.e. a's
believing that p] and the state of affairs that makes "p" true such that,
given Bap, it must be the case that p."
This purports to address the regress problem is very simply. On this account, the person who has a
basic belief need not be in possession of any justified reason for his belief.
Look at the form of Armstrong’s account:
1. Belief B is an instance of kind K
2. Beliefs of kind K are connected in a law-like way with the sorts of states of affairs which
would make them true, and therefore are highly likely to be true.
3. Therefore, B is highly likely to be true.
The second principle seems to simply replace the demand for justification with something
completely different?
So, what is your reaction? Is Armstrong actually dealing with the problem of justification
4
or has he simply changed the subject.
Why does Bonjour write this:
“It is clear, of course, that an external observer who knew both that P
accepted B and that there was a law-like connection between such
acceptance and the truth of B would be in a position to construct an
argument to justify his own acceptance of B. P could thus serve as a useful
epistemic instrument, a kind of cognitive thermometer, for such an external
observer. But P himself has no reason at all for thinking that B is likely to
be true . . . From his perspective, it is an accident that the belief is true. And
thus his acceptance of B is no more rational or responsible from an
epistemic standpoint than would be the acceptance of a subjectively similar
belief for which the external relation in question failed to obtain.”
So, Bonjour has this reaction to externalism (which I share): It simply changes the subject.
Think about skepticism . . .
However, considered more generally, it is a specific form of a type of strategy which needn’t
change the subject. There is another variant of this type of strategy which is, in fact, precisely what
the traditional foundationalist tends to have in mind when he asks after basic or foundational
beliefs.
This is the traditional Cartesian doctrine of cognitive giveness. How would you characterize
this notion???
“basic beliefs are justified by reference, not to further beliefs, but rather to
states of affairs in the world which are "immediately apprehended" or
"directly presented" or "intuited."
How is this like externalism similar and different in its response to principle (2) of the
original foundationalist strategy, e.g.,
1. Belief B has feature (f).
2. Beliefs having feature f are highly likely to be true.
3. Therefore, B is highly likely to be true. ???
How is it similar???
5
In both cases one’s empirical beliefs are ultimately grounded in some kind of direct appeal to
non-cognitive reality.
How’s it different???
In the case of externalism the justifying state of affairs is part of the world itself, whereas in the
case of classical givenism the justifying state of affairs is part of how the world is apprehended by
the perceiver.
Note again that we need not assume here that the classical given need not be known with
certainty or indubitability. But why would philosophers ever think that it would be
amenable to some kind of special epistemic privilege???
So, this is the idea so far:
1. Belief B has feature (f).
2. Beliefs having feature f are highly likely to be true.
3. Therefore, B is highly likely to be true.
where belief B is of the form “My immediately apprehended state of affairs is such and
such” and feature (f) is the feature that my this belief has of being somehow sourced or
prompted by immediately apprehended state of affairs actually being such and such. What is
important is that this immediately apprehended state of affairs is not itself a belief or any
sort of cognitive state since, if it were, it too would require further justification.
Now, we are finally coming to Bonjour’s most substantive point. On the above account,
there are three items involved in our having a foundationally grounded belief:
B (the belief) e.g., that I am having a yellow sensation
The state of affairs that my belief is about, e.g., the yellow sensation
The intuition or immediate apprehension of that state of affairs, e.g., my immediate
apprehension of that yellow sensation
The problem Bonjour sees is with the last of these three, e.g., the intuition or immediate
apprehension of the yellow sensation.
We’ve specified that it is not a belief or cognitive state of any kind. It has to be noncognitive. But this makes it difficult to see how it could justify B.
6
Classically, Lewis and Quinton and others suggest that my immediate apprehension of that yellow
sensation is an item of which I am "directly aware". But unless we lapse into marginally
intelligible and completely non-informative metaphors concerning the “mind’s eye”, it’s hard to
see what “direct” could mean here other than simply non-inferential.
Conclusion: the givenist is caught in a fundamental dilemma: if his intuitions or immediate
apprehensions are construed as cognitive, then they will be both capable of giving
justification and in need of it themselves; if they are non-cognitive, then they do not need
justification but are also apparently incapable of providing it. This, at bottom, is why
epistemological giveness is a myth.
We cannot simply stipulate that our direct-knowledge-yielding acts of non-cognitive
apprehension are semi-beliefs, providing justification but not requiring justification
without being ad hoc.
Consider a parallel case: The problem of free will.
A more specific analysis of the problem in this particular instance:
The basic idea, after all, is to distinguish two aspects of a cognitive state, its capacity to justify
other states and its own need for justification, and then try to find a state which possesses only the
former aspect and not the latter. But it seems clear on reflection that these two aspects cannot be
separated, that it is one and the same feature of a cognitive state, viz. its assertive content, which
both enables it to confer justification on other states and also requires that it be justified itself. If
this is right, then it does no good to introduce semi-cognitive states in an attempt to justify basic
beliefs, since to whatever extent such a state is capable of conferring justification, it will to that
very same extent require justification.
THE COHERENCE THEORY OF EMPIRICAL KNOWLEDGE Laurence
BonJour
In this paper, Bonjour offers his alternative to foundationalist accounts. Mainly, he
wants to show that coherence theory is itself coherent, that it doesn’t fall prey to
obvious and overwhelming objections.
Why go this way? How would you describe the fundamental confusion that
Bonjour thinks foundationalist accounts suffer from???
Confrontation???
CTEK (Coherence Theory of Empirical Knowledge): The most basic idea is that
the relation between the various particular beliefs is not one of linear dependence; but
one of mutual or reciprocal support. How does it answer the regress concern???
7
“There is no ultimate relation of epistemic priority among
the members of such a system and consequently no basis
for a true regress.”
“The component beliefs are so related that each can be
justified in terms of the others; the direction in which the
justifying argument actually moves depends on which
belief is under scrutiny in a particular context.”
“The apparent circle of justification is not vicious because
the justification of particular beliefs depends finally not on
other particular beliefs, as in the linear conception of
justification, but on the overall system and its coherence.”
First of all, what is coherence???
(1) Coherence is not to be equated with consistency. A coherent system must be
consistent, but a consistent system need not be very coherent. Coherence has to do
with systematic connections between the components of a system, not just with
their failure to conflict.
(2) Coherence is a matter of degree. For a system of beliefs to be justified,
according to the CTEK, it must not be merely coherent to some extent, but
more coherent than any currently available alternative.
(3) Coherence is closely connected with the concept of explanation (and prediction).
Unificationist theory of explanation.
That’s the idea. What are the objections to it? All have to do with the question
of how coherence can be equated with justification. Why is that an issue???
Here are the objections. Let’s talk about them.
(I) According to the CTEK, the system of beliefs which constitutes empirical
knowledge is justified solely by reference to coherence. But coherence will never
suffice to pick out one system of beliefs, since there will always be many other
alternative, incompatible systems of belief which are equally coherent and hence
equally justified according to the CTEK.
(II) According to the CTEK, empirical beliefs are justified only in terms of relations
to other beliefs and to the system of beliefs; at no point does any relation to the
world come in. But this means that the alleged system of empirical knowledge is
deprived of all input from the world. Surely such a self-enclosed system of beliefs
cannot constitute empirical knowledge.
8
(III) An adequate epistemological theory must establish a connection between its
account of justification and its account of truth; i.e., it must be shown that
justification, as viewed by that theory, is truth-conductive, that one who seeks
justified beliefs is at least likely to find true ones. But the only way in which the
CTEK can do this is by adopting a coherence theory of truth and the absurd
idealistic metaphysics which goes along with it.
First, deal with (II). Then use these results to deal with (I) and (III).
Again, (II): According to the CTEK, empirical beliefs are justified only in terms of
relations to other beliefs and to the system of beliefs; at no point does any relation
to the world come in. But this means that the alleged system of empirical
knowledge is deprived of all input from the world. Surely such a self-enclosed
system of beliefs cannot constitute empirical knowledge.
Distinction:
B is non-inferential iff B is not actually arrived at on the basis of inference from
other beliefs. (Psychological Inferential -- Ipsy)
B is non-inferential iff B is not warranted on the basis of inference from other
beliefs. (Logical Inferential -- Ilog)
Bonjour’s strategy: Show that the world can impact our beliefs through
observation, even though all the warrant for these observation statements is,
strictly speaking, Logical Inf
Example:
Belief P – “There is a black soda can on the desk.”
This belief is cognitively spontaneous (not Ipsy)
Bonjour needs to explain how it is still Ilog.
The form this explanation takes:
1. B is a visual belief, i.e. it is produced by my sense of sight; and I am, or at least
can be, introspectively aware of this fact.
2. The conditions of observation and attention are such and such: the lighting is
good, my eyes are functioning normally, and there are no interfering circumstances,
etc.
3. I have spontaneously have visual belief P.
9
4. If 1-3 are true, then P is likely to be true.
5. Thus, P is likely to be true.
In short, the reason that visual perceptual beliefs are epistemically justified or
warranted is that we have empirical background knowledge which tells us that beliefs
of that specific sort are epistemically reliable. This is the basic claim of the CTEK
for all varieties of observation.
“First, there must be a process of some sort which produces cognitively
spontaneous beliefs about a certain range of subject matter. The process involved
may be very complicated, involving such things as sense organs; the state of the
mind and/or brain as a result of previous training or innate capacities; perhaps also
the sorts of entities or events which philosophers have variously referred to by such
terms as "immediate experience," "raw feels," and "senses"; instruments of various
kinds; perhaps even occult abilities of some sort (such as clairvoyance); etc.’
“Second, “the beliefs thus produced must be reliable with respect to the
subject matter in question in the two distinct ways discussed above (under
specifiable conditions): on the one hand, it must be very likely that such
beliefs, when produced, are true (if the requisite conditions are satisfied); and, on the
other hand, if the person is in a situation in which a particular belief about that
range of subject matter would be true (and if the requisite conditions are
satisfied), then it must be very likely that such a belief will in fact be produced.
This second sort of reliability is crucial; on it depends, in large part at least, the
possibility of negative observational knowledge.”
“Third, and most importantly “from the standpoint of the CTEK, the person must
know all of these things, at least in a rough and ready way. He must be able to
recognize beliefs which result from the process in question (though he need not
know anything about the details of the process). He must know that such beliefs are
reliable in the two senses specified. And he must know in a given case that any
necessary conditions for reliability are satisfied. He will then be in a position, in a
particular case, to offer the following justification for such a spontaneous belief”:
(i) I have a spontaneous belief that P (about subject-matter S) which is an instance
of kind K
(ii) Spontaneous beliefs about S which are instances of K are very likely to be true,
if conditions C are satisfied.
(iii) Conditions C are satisfied. Therefore, my belief that P is (probably) true.
Is this enough to eliminate foundational or basic beliefs from Bonjour’s
scheme??? How about beliefs of type P, which are about experience and
which are introspectively accessible???
So, how do we answer (I), (II) and (III)
10
II. According to the CTEK, empirical beliefs are justified only in terms of relations to
other beliefs and to the system of beliefs; at no point does any relation to the world
come in. But this means that the alleged system of empirical knowledge is deprived
of all input from the world. Surely such a self-enclosed system of beliefs cannot
constitute empirical knowledge.
How does CTEK allow for the possibility of input from the world into the
cognitive system, a possibility which is in fact realized in our cognitive
system???
I. According to the CTEK, the system of beliefs which constitutes empirical
knowledge is justified solely by reference to coherence. But coherence will never
suffice to pick out one system of beliefs, since there will always be many other
alternative, incompatible systems of belief which are equally coherent and hence
equally justified according to the CTEK.
How does CTEK make it unlikely that there will, in the long run, end up
being an indefinitely large number of equally coherent alternative theories of
the world???
11