1. Internalists must either say that (i) the only facts that contribute to the justification of an individual’s beliefs at this moment are reflectively accessible at this moment or (ii) they are reflective accessible or can be retrieved by memory at this moment.
2. If the internalists opt for (i), the internalist cannot account for the intuition that we justifiably believe certain propositions where we cannot identify anything in our present conscious mental life that could justify these beliefs.
3. If the internalists opt for (ii), the internalist cannot account for the intuition that we can justifiably believe many things now where we cannot identify anything in our present conscious mental life that justifies these beliefs and cannot retrieve from our memory any adequate justification for these beliefs.
4. We can justifiably believe propositions that we learned earlier where we cannot identify anything in our present conscious mental life that justifies these beliefs and cannot retrieve from our memory any adequate justification for these beliefs.
5. (Therefore) Internalism delivers the wrong verdicts by classifying justified beliefs stored in memory as unjustified.
Feldman says that internalists have the resources to account for the intuitions that underwrite Goldman’s argument. We do now justifiably believe many propositions we learned in the past where there is little in our conscious mental life now that could be identified with the evidence we had for originally forming the belief, but maybe the internalist can concede this much. If you took Latin in high school, you might now realize that you believe that Caesar once wrote that Gaul was divided into three parts. What currently justifies your belief that Caesar once wrote this? It might not be any conscious state that you are in now, but what about non-occurrent recollections of past experiences? Maybe you have some vague recollection of studying Latin. I can vaguely recall someone telling me that the one thing I will remember from my days in Latin is that Caesar once said that all of Gaul was divided into three parts.
The real force of Goldman’s argument, however, is not touched by this point. We do not have this ability to conjure up sketchy recollections of past experiences to serve as justifiers for our beliefs. Even if we can, it's hard to think we can place much weight on them. There are no sketchy recollections of past experiences that could plausibly be taken to justify any of the dozens of beliefs I have about the United States’ history or the hundreds of beliefs I have about Texas’ history. On this point, Feldman also has a response. For stored beliefs where we do not have any apparent recollection of the events that happened while we acquired these beliefs, our beliefs can currently receive justification from associated feelings of confidence as well as stored beliefs that give us justification for thinking that our memories are reliable.
Goldman’s thinks that if we were to adopt this approach, we would be led to adopt an absurdly permissive view. Suppose Sally reads about the health benefits of broccoli in the New York Times and retains the belief that broccoli is healthy while forgetting where she read this and what evidence was offered for this belief in the article. Contrast this with a case in which Sally comes to form the same belief by reading some disreputable rag like the National Inquirer. About this case, Goldman says:
[Sally’s] broccoli belief was never acquired, or corroborated, in an epistemically sound manner. Then even with the indicated current background belief, Sally cannot be credited with justifiably believing that broccoli is healthful. Her past acquisition is still relevant—and decisive. At least it is relevant as long as we are considering the “epistemizing” sense of justification, in which justification carries a true belief a good distance toward knowledge. Sally’s belief in the healthfulness of broccoli is not justified in that sense, for surely she does not know that broccoli is healthful, given that the National Inquirer was her sole source of information.
If Goldman is right, the defects a belief had earlier continue to undermine a belief’s justificatory status now even if you are currently non-culpably ignorant of these defects.
Goldman might be right on this point, but I don't think it's obvious. Suppose Cooper believes now that his head is not made of glass. His mental life now is much like yours, so if you wonder what reasons Coop has for thinking his head is not made of glass, imagine his reasons are similar to the reasons you have for thinking that your head is not made of glass. Suppose, however, that Coop’s head is made of glass. Consider two versions of the case. In the first, Coop is deceived by a Cartesian demon from the moment the lights first came on until his eventual death. The demon never gives him any clue that this is so. In the second, Coop’s mental life is identical to Coop’s mental life in the first case with only one small difference. In the second, the demon pulls back the curtain to reveal to Coop that all of humanity has been systematically deceived. Coop refuses to consider the evidence presented to him, evidence that defeats whatever justification he previously had for his beliefs. Frustrated by Coop’s refusal to see reason, the demon wipes all traces of their meeting from Coop’s mental life and Coop continues to believe that his head is not made of glass. Here's the key point. I see no difference in the justificatory standing of Coop’s beliefs in this case. Dialectically, I think the problem with the example is this. If you think that the perceptual beliefs of those who are systematically deceived are justified in spite of the fact that they are generated by a Cartesian demon, you should certainly count Coop’s beliefs as justified in the first case. You should do the same for the second. If, however, you think that the beliefs held by the systematically deceived are not justifiably held, I think you could certainly say that Goldman is right and beliefs stored in memory are not justified nor if they were unjustified when formed. The reason this is a problem is that nearly everybody thinks that it is obvious that the beliefs Coop forms when deceived by a demon on the basis of his experience are justified. To use the example Goldman uses, he either has to say that the Coop cases differ in some important way from each other or say that my case differs from his. The first option, as I said, seems implausible. The second option also seems implausible. For my case involving Coop to differ from his case involving Sally, he would have to say that Coop’s beliefs are justified in both versions of the case and that Coop is better off than Sally when it comes to the justification of his beliefs.