When Socrates claims that knowledge is recollection, he is not only explaining what form our knowledge takes but also redefining what qualifies as knowledge at all. Clearly, the definition does not apply to everything we normally consider knowledge. When we find out in the newspaper what happened the previous day, we are not discovering things we’ve always known but forgotten. We get a hint about what counts as knowledge in the distinction Socrates draws toward the end of the dialogue between knowledge and true belief. This distinction, which plays an important role in the Republic, implies that we can be confident in knowing something only if we can give an account of, or justify, our knowledge. The slave boy may have guessed the answer to the mathematical problem at the outset, but he can be sure he knows the answer only because he went through the problem step by step, ensuring that he made no mistakes. This sort of rigorous justification applies only to subjects that consist of unchanging, abstract entities that are not subject to the errors and vagaries of everyday experience, such as mathematics. What we learn from the newspaper can never amount to more than true belief.

The argument that knowledge is recollection is bold and challenging, but it contains a number of problems. Foremost is the controversial question of whether the slave boy does in fact arrive at his own conclusions. Strictly speaking, Socrates only prompts the slave boy with questions, but he often makes statements couched in the form of questions, in which he arguably tells the boy the right answer rather than allowing him to figure it out for himself. Even if we do accept that the boy reaches the right answer on his own, it takes another leap to trust that he does so only by recollecting knowledge that he already possessed—let alone knowledge that he possessed before he was even born, as Socrates actually asserts. We could object first that the boy is not activating latent knowledge so much as latent ability. By claiming that the boy’s knowledge must be recollection, Socrates assumes that he is passively absorbing a set of facts rather than actively learning how to think mathematically. Second, we could object that the doctrine of knowledge as recollection does not explain how we first come to know things. Even if we believe that all the knowledge we possess came to us before we were born, such as in a previous life, we would still face the question of how we gained that knowledge in the first place.