A Closer Look at “Internalism Defended”

The debate between externalism and internalism has been of great interest to epistemologists. This paper will give an overview of one aspect of the debate, focusing on mentalism as defined by Richard Feldman and Earl Conee in Internalism Defended. Mentalism naturally appeals to belief justification via evidence directly available to the internal mind of a subject, but there is still the problem of how one can justify for forgotten evidence and explicate the connection between a belief and its justifying evidence. I shall evaluate these proposed issues and discuss the effectiveness of mentalism in the externalism-internalism debate.

First, a few definitions are in order. Externalism is the idea that basic knowledge requires justification that is external to a person. The justification of beliefs should include external factors, such as whether the belief is caused by affairs that make the belief true, whether the belief is caused by reliable processes or whether the belief is caused by a mechanism with above average accuracy. Internalism takes the opposite view: beliefs do not need to be justified by outside sources but rather by things that are internal to a person’s mind. There are two prominent views of internalism, accessibilism and mentalism. Accessibilism affirms that beliefs are justified when it can be accessed by the person directly or through reflection. Mentalism, according to Feldman and Conee, asserts that justification of beliefs is determined by occurrent mental or dispositional factors. One distinction between the two is that mentalism allows non-reflective mental states to justify one’s beliefs; it does not depend on introspection nor does it need to access specific mental states. One is not required to be aware of the belief justifiers for the belief to be justified. Mentalism is a weaker form of internalism than accessibilism is. In this way, mentalism can be said to be superior to accessibilism because of its more flexible requirements.

Feldman and Conee strongly advocated mentalism as the version of internalism that is strong enough to defend against points posited by externalists. The authors presented examples of contrastive epistemic situations, where one agent has a justified belief while the other does not or one agent’s belief is better justified than the other’s. By using examples of pairs of people who varied slightly in mental states, they argued that epistemic differences are best explained by differences in the agents’ mental perspectives.

For example, consider the case of two people’s beliefs about the weather:

Bob and Ray are sitting in an air-conditioned hotel lobby reading yesterday’s newspaper.  Each has read that it will be very warm today and, on that basis, each believes that it is very warm  today.  Then Bob goes outside and feels the heat.  They continue to believe that it is very warm today. But at this point Bob’s belief is better justified.

At time t1, Bob and Ray are inside the hotel reading their respective newspapers. Assuming they are reading the same issues and that the newspaper is a reliable source, Bob and Ray both have justified reasons for believing today’s weather and have the same mental states that today will be very warm. At a later time t2, Bob decides to head out and so feels the heat. Now Bob has another reason, his direct experience, for believing that today is very warm. His belief is better justified than Ray’s; he has an additional and concrete reason to believe today’s forecast. Ray lacks this extra mental state and relies on nothing more than his newspaper. The idea is that getting new information makes one undergo changes in mental state and bridges a connection between the subject and the reason, “internalizing” the reason one now holds.

As another example, consider the case of the TA and the logic student:

A logic Teaching Assistant and a beginning logic student are looking over a homework assignment. One question displays a sentence that they both know to express a truth and asks whether certain other sentences are true as well. The TA can easily tell through simple reflection that some of the other sentences express logical consequences of the original sentence and thus she is justified in believing that they are true as well. The student is clueless.

At time t0, the TA and logic student starts out with different epistemic states; the TA has knowledge of logic principles and the logic student does not. At t1, the TA is able to use her existing justifying beliefs to reach a conclusion about the question. The TA, due to her expertise in logic, answers the question quickly, but the logic student does not have enough background knowledge to know what to do. To clarify, suppose that the logic TA and the logic student have the evidence (on paper) that it is not the case that (p and q). From the given statement, it logically follows that it not being the case that (p and q) entails the case that (not-p or not-q). One is able to arrive at this logical conclusion by De Morgan’s law. The TA possesses linking information (via De Morgan’s law) that connects it not being the case that (p and q) to the conclusion (not-p or not-q). The beginning logic student does not hold the same justifying information. He would not be justified in believing that (not-p or not-q) is the answer to the question unless he learns De Morgan’s law. Feldman and Conee call the TA’s background information the “cognizance of necessary truth.” This truth is the required information one needs to have in order to be justified in knowing a certain proposition p. If the logic student was to learn this law, he would then share the same epistemic status of the TA, come to understand the logical consequence relation, and be justified in concluding that (not-p or not-q) follows from not (p and q).

Based on these cases and a few similar ones, Feldman and Conee concluded “…that every variety of change that brings about or enhances justification either internalizes an external fact or makes a purely internal difference” (Feldman and Conee). These example pairs are taken to be representative of all epistemic situations. Their use of example pairs is by no means conclusive, but these examples offer the view that epistemic differences between any two individuals have an entirely mental origin. The justification of one’s beliefs is determined entirely by the reason the subject has at the moment, the epistemic state the subject is in or the non-occurent beliefs the subject has in a dispositional state.

However, one limitation of these example pairs is that they do not address the matter of initial justification. In all the examples, epistemic knowledge was already established. The agents were already justified to a certain degree and further information only enhanced those initial justifiers. The matter of how initial justifiers came about needed to be addressed – were they provided by some external source or a mental state.  Feldman and Conee did not address this point. So while their examples added more support for mentalism, this alone was not enough to say all epistemic justifiers originated from mental origins. Also, they also do not explain when one would be considered justified. The question of whether quality over quantity matters most in determining justification is an important concern. Bob is more justified than Ray because he has one more mental state, but in the circumstance where a friend informs Bob the temperature outside, one will not so quickly conclude that Bob is more justified than Ray in knowing that today is very warm. The quality of this additional information is not as valuable as experiencing the warmth in person. In a case like this, externalism would have the upper hand; process reliabilism would explain why Bob would be more or less justified than Ray.

Feldman and Conee gives more sound support for internalism in the third segment of their paper when they reply to objections. Due to the focus of this paper, I will briefly discuss only two of these objections and the authors’ responses to them.

Some authors have raised the problem of how the internalist would account for the justification of knowing proposition p for which the original evidence is now forgotten. In reply, Feldman and Conee appeal to the vivacity of one’s recollection that p and that one would have other supporting evidence for p that is deemed reliable. Based on background information and the strong belief that memory is generally reliable and accurate, one can still be justified in having knowledge about p even without knowing the exact supporting evidence of p. Yet, there is also the problem of one who learns that p from an unreliable source but that p happens to be true. This response would have meant that the reliability of a source is unimportant since acquiring evidence from reliable and unreliable sources would make one equally justified. Feldman and Conee counter-replied this argument, saying this is not a problem of justification but a problem of knowledge, which is a different matter and requires a different approach. One needs to view the case as an agent losing a defeater of justification when she forgets the unreliable source. Losing the unreliable source would lend the agent to relying on her better credentials of existing beliefs. It now seems that internalism does not have much difficulty in situations regarding forgotten evidence.

Lastly, one of the objections rose from the big question as to whether logical or probabilistic support justifiers count as epistemic justifiers. Must there be a justifying relation between a person’s belief and the evidence for that belief? If this relation is required, internalism must be able to address this. William Alston argued that there must be a higher order requirement on justification. He reasons that a believer must be able to tell which factors justify a belief. If one was aware of the relation between a person’s belief and the evidence for that belief, then one might say that one has the mental states that recognize this relation as justifier. To recognize the relation, to be aware of it, is most certainly a mental state, and mentalism would then work. Feldman and Conee disagreed with this requirement. They contended that as long as one possessed the right evidence for a belief, this by itself is enough to warrant the belief justified. One need not be concerned with knowing the exact relation. However, this counter-argument raises further arguments that this “existence of supporting evidence” may be beyond the reach of children who may have perfectly justified beliefs without any underlying support relations.

Feldman’s and Conee’s paper have well demonstrated the strong claims of internalism. Yet, it is unclear from their paper which mental states determine justificatory status or whether the number of mental states makes a difference in belief justification. They also do not explain whether knowledge in some unsophisticated persons, such as children, would count as having justified knowledge. While internalism has intuitive appeal and can explain many matters related to epistemological justification without needing to appeal to external factors, more needs to be explained to address what is the basic nature of knowledge and epistemic justification.


Feldman, Richard and Conee, Earl. “Internalism Defended”. American Philosophical Quarterly, Vol. 38, No. 1 (Jan., 2001), pp. 1-18


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s