2- From Explanations to Why-questions
So what is the problem? Surely a theory that accounts for what is going on in these examples, a theory of the kind of scientific achievement they are instances of, deserves to be called a #theory_of_explanation
Sue, and a student Marcel; Marcel knows that electrons fly off metals, but doesn’t know the correct theoretical account of this effect. Sue imparts to Marcel the knowledge I sketched above, in a way that makes it true that (1) Sue explained the photoelectric effect to Marcel. In (1) the object of the verb “explained” is “the photoelectric effect”—so isn’t it the photoelectric effect that is being explained?
What epistemologists aim to understand is the propositional attitude knowledge—what we ascribe when we say things like “Smith knows that Barack Obama was born in Hawaii,” or “Bloggs knows where the President lives.” But not all uses of “know” attribute a propositional attitude.
One might say, for example, that Fred knows Ed, meaning not that Fred knows that P is true for some sentence P that is about Ed, but that Fred has met Ed, that they are friends, or something like that. Now suppose I present a certified epistemologist with the sentence
(2) John knows Bill’s telephone number.
and ask her for the conditions under which (2) is true. On the surface (2) looks to have the same form as “Fred knows Ed,” so the epistemologist might say that answering my question falls outside her domain of expertise. But this would be a mistake. Despite their surface similarity, “knows" in (2) is not being used in the same way as in “Fred knows Ed.” For it is evident that (2) is equivalent to
(3) John knows what Bill’s telephone number is, and in (3) “knows” is being used in its propositional-attitude sense.
Linguists say that when a noun phrase like “Bill’s telephone number” is used so that it has the same meaning as an indirect question (here, “what Bill’s telephone number is”), it is being used as a #concealed_question.
, it is evident that “the photoelectric effect” is a concealed question in (1). Make the question explicit and we get
(4) Sue explained to Marcel why the photoelectric effect happens (why electrons fly off metals when light is shone on them).
Insofar as philosophers of science are after what is going on when sentences like (1) are true, they are really after what is going on when sentences like (4) are true. And (4) contains “why” as well as “explains".
Reasons favoring this view:
- One fact that favors my thesis is that explanation and why-questions come apart, and the interests of philosophers of science seem to go with why-questions, not with explanation. One way in which explanation and why-questions come apart is that not all cases of explaining are cases of explaining why. The verb “explain” can take any indirect wh-question as a complement, not just a why-question. Someone can explain where the bread is, who is coming to the party, or how to ride a bike. But figuring out what was going on when John explained who was coming to the party does not seem like a job for the philosophy of science e. (The same goes for explaining where, or explaining what; the case of explaining how is more controversial.)
- And there are two separable parts to explaining the answer to a wh-question. The first part is to— somehow—convey to one’s audience what that answer is. One can do this first part in many ways; and it can be done without explaining anything. One can, for example, merely tell one’s audience what the answer is. The second part involves doing the first part in a way that counts as explaining. Maybe to explain an answer one must go through it slowly and clearly. - if you want to know what it takes for someone to have explained the answer to a why-question, you will have to do more than consult your best theory of explanation. You will have to combine what that theory tells you about what it takes to explain the answer to a question with what your best theory of answers to why-questions says it takes to be the answer to a why-question.
Failure to appreciate these points has fostered confusion.
Think about Carl Hempel’s DN model of explanation.The DN model says, roughly, that X explains Y iff X is a sound argument with Y as its conclusion that essentially contains a law-stating premise (a premise that expresses a law of nature). This model fits many examples in which someone explains why something is the case, but isn’t even remotely plausible as a theory of what is happening when a policeman explains where the train station is, or when a biologist explains what the theory of evolution says. (Hempel emphasized that his was a theory of scientific explanation, so maybe the policeman example isn’t fair, but in the second example we have a scientist explaining a scientific theory.)
one need not produce any valid arguments or cite any laws of nature to explain.
If I were defending the DN model I would reply: so what? Hempel may have called the DN model a model of “explanation,” but he shouldn’t have. It is best evaluated as a theory of answers to why-questions. Understood as a theory of that kind, it says nothing about answers to what-questions, or about what it takes to explain an answer to a what-question.
There are distinctions to be made—and that I will make—between partial answers and complete answers to why-questions, and between “merely” partial answers and partial answers that are also part of the complete answer to a why-question.
Hempel claimed that not all why-questions are “explanation-seeking” why questions; some are instead “epistemic.” He took explanation-seeking why-questions to be the target of his theory. Hempel’s distinction has been quite influential. Nearly forty years later Salmon reports no challenges to it, and does not challenge it himself, in his book-length history Four Decades of Scientific Explanation.
Does Hempel’s distinction even exist? “Not all why-questions call for explanations,” he wrote as he introduced it, “some of them solicit reasons in support of an assertion”. This is hard to understand. Hempel seemed to think that offering evidence (“reasons in support”), and explaining, were exclusive activities. But I can do both simultaneously: I can explain what evidence there is for, say, the proposition that the earth is heating up. And how can it be that not all why questions “call for explanations”? A natural interpretation makes this the claim that explaining why Q is not always the response called for when someone asks why Q. But since “John explained why Q” is roughly equivalent to “John explained the answer to the question why Q,” on this interpretation Hempel comes out saying that sometimes, when we ask a why-question, we are not looking for an answer to the question we are asking. That can’t be right.
Instead of asserting that every why-interrogative has an epistemic reading, Hempel may have meant to assert that why-interrogatives with certain forms are epistemic (and can only be read as epistemic), those forms being closely related to the form “Why should it be believed that P?”
Why call them the explanation-seeking why-questions? Why think that there is a theoretically interesting difference between them and the epistemic why questions?
There are evidently at least two forms that answers to why-questions can take. They can use “because,” or they can use “in order to.” “In order to” answers are connected with ends or purposes: teleology. Because that is a fraught topic I am going to set aside why-questions that demand “in order to” answers for later. I will discuss the relationship between “because” answers and “in order to” answers in section 6.1. Elizabeth Anscombe said that an #intentional_action is one “to which a certain sense of the word ‘Why?’ is given application; the sense is of course that in which the answer, if positive, gives a reason for acting” . For now I want to set aside why-questions that are requests for such reasons. This excludes some why-questions that demand “in order to” answers. It also excludes some why-questions that take “because” answers, for “because” can also be used to give someone’s reason for acting—the most natural way to hear “I am walking to the store because it sells milk” is as equivalent to “My reason for walking to the store is that it sells milk.”
Why questions of the type the author is focused on is thought to be more fundamental, in the sense that a theory of answers to them can be used to construct theories of the other kinds of answers.
2.2 More red hearring
For another example of how over-use of “explanation” can obscure what is really going on, consider Peter Lipton’s essay “Understanding without Explanation.” He begins the essay like this:
Explaining why and understanding why are closely connected. Indeed, it is tempting to identify understanding with having an explanation. Explanations are answers to why questions, and understanding, it seems, is simply having those answers.
#Lipton, however, wants to resist this identification: he holds that understanding is to be identified, not with having an explanation, but with “some of the cognitive benefits of an explanation”. He goes on to argue that “there can be understanding without explanation”.
so “there can be understanding without explanation” means that someone can understand why Q without “having” an answer to the question why Q. Lipton’s thesis becomes the claim that one can understand why Q without knowing why Q. That sounds pretty wild.
Lipton’s examples of “understanding without explanation” suggest a second interpretation of his thesis. Lipton understands why some planets engage in retrograde motion, without anyone having explained to him why they do so, and without him being able to explain to anyone else why they do so. Since the example is meant to support Lipton’s thesis, his use of it suggests that his thesis is that someone can understand why Q without anyone having explained to her why Q, and without her being able to explain why Q. But who would deny this?
The second interpretation, while it may fit the example, does not fit at all with the way Lipton introduced his thesis, as the claim that one may have understanding without “having” an explanation. Just because no one explained to him why planets engage in retrograde motion, just because he cannot explain it to anyone else, does not mean he does not “have” an explanation.