The internalism/externalism debate in epistemology regarding justification centers on the question of what states, properties, and events can contribute the kind of justification necessary for knowledge. Internalism has been understood in a variety of ways, but roughly it is the claim that all the factors that justify beliefs are internal to the cognizer. Internalism is best construed as mentalism, the claim that the justificatory status of a person’s beliefs strongly supervenes on that person’s mental states, events, and conditions. Thus, if two people are alike mentally, then they are alike justificationally as well. Internalism has often been understood to include the claim that justifiers must be accessible to the cognizer. However, it is more straightforward and noncommittal to understand internalism as described above, with theories that include an accessibility requirement being seen as species of internalism, or one (somewhat popular) way to fill out the theory. Internalism understood as mentalism lets in all the theories that can plausibly be considered ‘internalist’. Thus, we will understand internalism as the claim that only mental factors determine justification. Foundationalism and coherentism are examples of internalist theories of justification.
Externalism is the denial of internalism. Thus, the claim is that factors that are not internal to the cognizer (factors that are not the cognizer’s mental states, properties, or conditions) make an epistemic difference in terms of justification. Extramental factors play a justificatory role. Reliabilism and proper-function theory are examples of externalist theories of justification.
Arguments in favor of internalism focus on how well internalism can handle cases of justified and unjustified belief. Imagine that Jim and George both see on the news that it is raining today. In addition to watching the weather, Jim looks outside and sees the rain falling. Internalism can explain why Jim is more justified in his belief than George is because of a mental difference – Jim has perceived the rain falling.
Imagine that Jim and George each hear a bit of testimony from Tracy. Tracy is a very reliable person, Jim knows about Tracy’s good track record, but George does not. As such, Jim is more justified in believing what Tracy says. The mental difference of memories regarding Tracy’s honesty, that Jim has and George lacks, account for the justificatory difference here.
So, internalism seems to get the cases right. Arguments for externalism largely consist in attempts to show that limiting justifiers to the mental fails to give the right result in certain cases.
Several externalists have claimed that internalists face a problem of forgotten evidence. Roughly, the idea is that one can still be justified in believing a proposition even when one has forgotten the evidence that supported that proposition; and, if so, then there must be something extramental doing some justificatory work. Goldman gives the case of Sally who reads in the NYT that broccoli is healthy. She forms that belief, but then forgets her evidential source and never comes across any further sources. Nonetheless, her belief is justified and if true is a case of knowledge.
Internalists can respond by noting that if Sally really is justified in her belief then there will be some mental factors doing the work. For instance, it is likely that Sally has a confidence or a clarity regarding the healthiness of broccoli. These phenomenal qualities are mental factors that can play a justificatory role for an internalist (contrast the support of a hazy memory). Further, Sally is likely aware of the general reliability of her memory, and that she usually does not simply believe things without having a good reason. These mental states can also provide justificatory support for her belief. If Sally lacks all such support, then it is doubtful that she truly is justified in her belief regarding the benefits of broccoli.
Externalists have responded by altering Sally’s case such that her source on the benefits of broccoli is actually an unreliable source such as the National Inquirer. Sally has forgotten the source, however, so if the internalist response above is correct, then Sally will still be justified in her belief that broccoli is healthy (provided the other mental factors mentioned above are true of her). The externalist claims that though her belief is true, it cannot be a case of knowledge, so Sally must not be justified in her broccoli belief.
This inference, however, is mistaken. It can be that Sally is justified in her true belief and yet fail to have knowledge – Gettier has shown us this. The reconstructed Sally case is indeed a Gettier case. The reasons that Sally has for thinking that broccoli is healthy provide justification, but they also contain an essential falsehood: that Sally’s reasons come from a reliable source. Thus, Sally’s case follows the recipe for Gettier cases – it is an instance of a true, justified belief that fails to be knowledge.
Some may find it strange that Sally’s belief became justified through her forgetting the source of her reasons (by forgetting it was an unreliable source). A couple things favor the internalist here, however. First, if Sally did remember the unreliable source, her belief would surely be more unreasonable. So, forgetting the unreliability of the source does seem to increase the justification. Second, if it is denied that Sally is justified in this revised case, then some distinction must be drawn to have it that she is justified in the first case, yet not in the second even though the two cases are on a par from Sally’s perspective. An account that distinguishes these cases in such a way must go contrary to stronger intuitions. There is no problem for internalism here.