The content of this doctrine requires some explanation. Let us agree that a person is epistemically justified in holding a certain belief if and only if the person in virtue of holding the belief does not violate any epistemic duty or manifest any epistemic defect. The notion of a good argument I will take for granted. But what is it to have a good argument, in the sense involved in (1)? Surely a person who merely believes truly that there is a good argument for a certain conclusion on such-and-such a page of a textbook should not be said to have thereby the argument stated on that page. Furthermore the mere ability to construct an argument seems insufficient. Let us say that a person S has a good argument for a certain conclusion if and only if S knows some statement of the argument and truly believes that the argument so stated is a good argument for the conclusion. I shan't pause to say more about the nature of the knowledge and belief involved here, except to remark that they may be dispositional rather than occurrent. For yesterday I had a good argument for "394 is an even number" viz. "any number ending in an even number is an even number; 394 ends in an even number"; therefore 394 is an even number" even though before this morning I had never entertained that proposition. On the other hand S may have forgotten an argument, and so not have it, even though S can with a bit of effort reconstruct it.