UNDERSTANDING AND KNOWLEDGE-STATES
Explaining q has been defined as uttering something with the intention of rendering q understandable (in a certain way). A theory of explaining, however, that invokes an undefined notion of understanding does not take us far enough.
Much of the discussion that follows is devoted to formulating an account of understanding that will complement the conditions for explaining and allow us to develop the concept of an explanation (product) and that of a good explanation later (see chapter 7).Since explaining involves an intention to render q understandable— where q is an indirect question—I shall be concerned with cases of understanding where q is also an indirect question. Such understanding I take to be a form of knowledge.[56] One understands q only if one knows a correct answer to Q which one knows to be correct. Using quantificational notation, we can say that a necessary condition for the truth of sentences of the form “A understands q” is
(1) ($x)(A knows of x that it is a correct answer to Q).
If A satisfies this condition, I shall say that A is in a knowledge-state with respect to Q. This condition involves a de re sense of knowing. (In the de re, by contrast with the de dicto, sense, from “A knows of x that it is P,” and “x = y,” we may infer “A knows of y that it is P.”) I am construing an answer to a question as a proposition, and de re knowledge of a proposition as involving an “acquaintance” with it (a knowledge of its content). There are several points to be made about knowledge states. First, being in such a state is not simply knowing that there is some proposition that is a correct answer to Q. It is not defined simply as
A knows that ($x)(x is a correct answer to Q),
for this could be true even if A fails to know any correct answer to Q.
Second, being in a knowledge-state with respect to Q entails that Q has a correct answer; therefore, so does understanding q.
If there is no correct answer to Q, if, e.g., Q is based on a false presupposition, then understanding q is an impossibility. One cannot understand why helium is the lightest element, since it isn't.Finally, it is not sufficient that A know of (i.e., be acquainted with) an answer to Q which happens to be correct. A speaker who claims that he does not understand how a certain accident occurred might ask: Was it caused by the slippery road, did the driver fall asleep, was there a blowout, or what? Suppose that, unbeknownst to the speaker, the slippery road did cause the accident. The mere fact the he knows of an answer that happens to be correct is not sufficient to say that he understands. What he does not know of this answer is that it is correct, and this is necessary for understanding. Condition (1) is a strong one for understanding, and it might be denied. If A knows of no answers to Q at all, or only incorrect ones, then we might be willing to follow (1) and deny that A understands q. But (1) makes the stronger claim that A knows of some answer to Q that it is correct. Suppose that the police believe, but do not know, that the accident was caused by the slippery road. And suppose that their belief is correct. If (1) is a necessary condition for understanding, we cannot say that the police understand what caused the accident. This seems odd, since they can produce what they believe to be, and is, a correct answer to Q.
Nevertheless, I suggest that this refusal is justified. If the police suspect, but do not yet know, that the accident was caused by the slippery road, then, although they are well on the way toward understanding what caused the accident, they are not yet in the position of someone who does understand it. Their failure to be in this position derives not from the fact that they have yet to find a correct answer, but from the fact that they have yet to achieve a required epistemic state with respect to that answer, namely, one of knowing that it is correct.
To be sure, if the police believe correctly, but do not know, that the accident was caused by the slippery road, then it could be misleading to deny that they understand what caused the accident. This is because we would not be giving the whole truth, only part of it, regarding their epistemic situation. To avoid misleading the audience what might be said is that the police do not yet understand what caused the accident, although they believe it was the slippery road.On the other hand, our definition of a knowledge-state may be accused of being too weak for understanding. Suppose that A once heard a correct answer to Q which he can no longer remember. However, he knows that the answer he heard is correct. Therefore, he is in a knowledge-state with respect to Q, despite the fact that he cannot remember a correct answer to Q. But surely if A is to be said to understand q he cannot be in this position.
My reply is that in the case envisaged A is not in a knowledge-state with respect to Q, although he once was. At one time, but not now, he was acquainted with a correct answer to Q. He does now know that the answer he once knew is correct. His situation can be described by saying that
($x)(A knew of x that it is a correct answer to Q), and A knows that
($x)(x is a correct answer to Q).
But it is not the case that
($x)(A knows of x that it is a correct answer to Q).
3.