# Knowing an Answer and Knowing Who

In the thread below on knowing who, David Braun suggests that statements like the following sound contradictory.

‘Zaphod Beeblebrox kissed that guy’ is an answer to the question ‘Who did Zaphod Beeblebrox kiss?’, and B knows that Zaphod Beeblebrox kissed that guy, and so she knows an answer to the question of who Zaphod Beeblebrox kissed, but she does not know who Zaphod Beeblebrox kissed.

I think this sounds fine in some cases, or at least if it doesn’t this tells us something about what counts as an answer to a question.

Let’s get some background. As President of the galaxy, Zaphod legalised gay marriage, so ‘that guy’ denotes in the context his husband. Last night Zaphod was at a party where by all accounts he kissed celebrities from every major star system. He also gave that guy a goodnight kiss. The identities of the celebrities are a mystery. B knows that Zaphod kissed his husband, but not the mysterious celebrities.

Now what about David’s sentence? I think B doesn’t know who Zaphod kissed. She does know that he kissed that guy. Is ‘Zaphod kissed that guy’ an answer to the question of who Zaphod kissed? I’m half-tempted to say it is, so here’s a case where David’s ‘contradictory’ sounding sentence is true. But maybe I’m missing something crucial about the semantics of plurals and/or questions here.

## 4 Replies to “Knowing an Answer and Knowing Who”

1. Mike says:

Is there a contradiction? B has an answer to the question of who Zaph kissed. That sounds right to me, too. But isn’t that consistent with there being the answer to the question who Zaph kissed, where an answer needn’t be (and in this case isn’t) the answer? For instance, what’s 2+2 equal? Here’s an answer, “5”. So you have an answer to what 2+2 equals but you don’t know what 2+2 equals. But that’s too easy. You might instead have an answer that’s true and yet not know the answer. What’s 2+2 equal? An answer, “something less than 6”. So you have an answer that’s correct, but you don’t know what 2+2 equals. No contradiction there, I think.

2. David Braun says:

I claimed that the following sentence is contradictory.

‘Zaphod Beeblebrox kissed that guy’ is an answer to the question ‘Who did Zaphod Beeblebrox kiss?’, and B knows that Zaphod Beeblebrox kissed that guy, and so she knows an answer to the question of who Zaphod Beeblebrox kissed, but she does not know who Zaphod Beeblebrox kissed.

Brian says that it “sounds fine in some cases, or at least if it doesn’t this tells us something about what counts as an answer to a question.” He then goes on to describe a situation in which one might utter this sentence and it “sounds fine”. I think the example is nice, but raises issues that need separate treatment.

It might be worth mentioning why Brian brings up this sentence. I reject contextualist theories of ‘knowing who’ in the paper that Brian mentions. In one argument against contextualism, I argue that contextualists about ‘knows who’ are forced to attribute implausible context-dependence to other closely related expressions, such as ‘answer’. I do this by first claiming that a sentence similar to that above is contradictory. And so I claim that contextualists will want their theory to entail that it is false in all contexts. But if ‘knows who’ is context-dependent, then the most obvious way to assure that the sentence is false in all contexts is to do the following: (1) claim that ‘answer’ is context-dependent, and (2) claim that “what counts as an answer” in a context is yoked to what’s required for knowing who in that context. A person satisfies ‘knows who X is’ in a context iff he knows a proposition that satisfies ‘is an answer to the question of who X is’ in that context. So contextualists will want to claim that ‘answer’ is context-dependent. But I then go on to argue that ‘answer’ is not context-dependent.

Brian now argues that the sentence is not contradictory. (Or does he? All he really says is that it “sounds fine in some cases”.) If correct, this would undercut one of my premises against ‘knows who’ contextualism. But I still think that it is false in all contexts, and in that sense contradictory. The sentence still has a strong flavor of contradictoriness, at least when there is not a lot of background setup. Consider the following.

X knows an answer to the question of who likes Vegemite, but X does not know who likes Vegemite.

This seems contradictory, or impossible. One can make it sound better with a story, which is what Brian does with the earlier sentence. But Brian provides the sort of background that might allow even explicit contradictions to sound OK, such as:

Well, B knows an answer to the question, but he also doesn’t know an answer to the question.
Well, knows who Zaphod kissed, but he also doesn’t know who Zaphod kissed.

Further, I think Brian’s example exploits our mixed intuitions about what we might call “non-exhaustive answers.”

Let H be Zaphod’s husband. Zaphod kissed H. Zaphod also kissed celebrities C1, C2, … , Cn. Now consider the following interrogative and declarative.

Who did Zaphod kiss?
Zaphod kissed C2.

Is the above declarative an answer to the interrogative? I am inclined to say ‘yes’, but many have intuitions contrary to mine, or mixed intuitions. Similarly for

Who did Zaphod kiss?
Zaphod kissed H.

In any case, usually when we ask questions like ‘Who did Zaphod kiss?’, we are interested in getting some sort of exhaustive answer. We often want either an enumerative answer (an answer that names all the kissed people) or at least an answer that (in some sense) characterizes everything in the set of things that Zaphod kissed, such as ‘Zaphod kissed his husband and some celebrities’.

In Brian’s example, if B were asked ‘Who did Zaphod kiss?’, he would answer ‘Zaphod kissed H’. Would he thereby provide a (genuine) answer to the question? I say ‘yes’, but others might say ‘no’ or have mixed verdicts. But it’s certainly true that we might vary, from time to time, or context to context, in whether we find this answer interesting, and in whether we would call it ‘an answer’.

Back now to the original sentence and Brian’s case. Consider now a yoking contextualist, that is, a theorist who yokes the content of ‘answer’ in a context to the content of ‘knows who’ in that context, in the way I described above. (I reject contextualism, but it is useful to see how they might hold on to the view that the sentence is contradictory.) Such a theorist could plausibly say that the sentence is contradictory, but that Brian encourages us to shift context in mid-sentence. Brian first encourages us to think that ‘Zaphod kissed H’ is an answer to ‘Who did Zaphod kiss?’. Therefore, it is an “answer in that context”. But then ‘B knows who Zaphod kissed’ is also true in that context. Brian can easily get us to abandon that context, for usually we are not very interested in such non-exhaustive answers. When Brian says ‘B doesn’t know who Zaphod kissed’ he encourages us to enter a new context in which that negative ascription is true. But ‘Zaphod kissed H’ does not stand in the relation expressed by ‘answer’ in that new context to ‘Who did Zaphod kiss?’. The declarative is (so to speak) “no longer an answer” to the interrogative in this new context.

An invariantist (like me) can give a similar story. When Brian first encourages us to think that ‘Zaphod kissed H’ is answer to ‘Who did Zaphod kiss?’, we go along, for perhaps some people are interested in such minimal answers. (The invariantist might even say that this verdict about answers is the correct one.) When Brian later says ‘B doesn’t know who Zaphod kissed’, we take him to be implying that B does not know an answer to the question of who Zaphod kissed that is of any significant interest. We go along with Brian, changing our minds about what is an interesting answer (to him, or to us, or to others).

In short, yoking contextualists about ‘knows who’ and ‘answer’ can plausibly argue that the sentence is contradictory, but that Brian is encouraging us to shift context in mid-sentence. Invariantists about ‘knows who’ and ‘answer’ can also say that the sentence is contradictory, but that Brian’s example encourages us to (roughly speaking) change our minds about what answers are interesting.

3. Marco says:

Here’s a true instance of the allegedly contradictory schema that doesn’t involve a multiply-instantiated property: ‘Clark Kent is Superman’ is an answer to the question ‘Who is Superman?’ and Jimmy knows that Clark Kent is Superman (since he knows that Superman is Superman). So Jimmy knows an answer to the question of who Superman is. But Jimmy doesn’t know who Superman is.

Actually, one may wonder how to state this example in Braun’s terms (see his sections 1-2). Should we say that ‘Clark Kent is Superman’ is a “linguistic” answer to the interrogative ‘Who is Superman’? Or should we say that ‘Clark Kent is Superman’ expresses an answer (a “propositional” answer) to the “semantic” question expressed by ‘Who is Superman’? Braun suggests (p. 3f) that these come to the same thing. He suggests that declarative A is a linguistic answer to interrogative Q iff the proposition expressed by A is a propositional answer to the semantic question expressed by Q. This equation seems right enough to me, but it might come into tension with some other things Braun says (depending on what he means, of course).

On Braun’s IP analysis, “a proposition answers a semantic question iff it provides information about the question’s subject matter” (p. 6). He later writes of logically true propositions, and suggests (p. 9f) that they generally don’t provide information about subject matters. I take this to imply that logically true propositions (whatever they are) won’t generally count as answers.

Now, does ‘Clark Kent is Superman’ express a logically true proposition? If it does, then one way to try to defuse the Jimmy example is to deny that ‘Clark Kent is Superman’ is an answer to ‘Who is Superman?’ on the grounds that the proposition expressed by the former is not a propositional answer to the semantic question expressed by the latter. But in that case, Braun’s equation (taken as a claim about linguistic answers and not as a mere definition of them) comes apart. For it seems undeniable that ‘Clark Kent is Superman’ is a linguistic answer to the interrogative ‘Who is Superman?’. (This could also be formulated as a challenge to Braun’s claim linking semantic answers to answers of interrogative acts; see p. 4.)

I would rather keep Braun’s nice equations. What, then, should I say about Jimmy and logical truth?

4. David Braun says:

I think Marco’s entry (Oct 10) raises two separate issues.

I am a Millian, but I think non-Millians can accept nearly everything I say about questions, answers, and knowing who. The non-Millian and I would disagree about whether two sentences express the same proposition or semantic question. I think that ‘Who is Clark Kent?’ and ‘Who is Superman?’ express the same semantic question. The non-Millian would disagree. I think that if ‘Superman is Clark Kent’ is a linguistic answer to ‘Who is Superman?’, then and ‘Clark Kent is Clark Kent’ and ‘Superman is Superman’ are also (linguistic) answers to ‘Who is Superman?’. The non-Millian would disagree. (I would try to explain away the non-Millian intuitions by appeal to different ways of entertaining and believing propositions and semantic questions.)

All of this has consequences for whether Jimmy knows who Superman is. If ‘Superman is Clark Kent’ is a linguistic answer to ‘Who is Superman?’, and the former expresses the proposition that Superman is Superman, and Jimmy knows that Superman is Superman, then Jimmy knows who Superman is. All of this is correct, I think. But non-Millians could disagree with these consequences and still (I think) accept my views about questions, answers, and knowing who.

The other issue that Marco raises is about whether the proposition that Superman is Superman answers the (semantic) question of who Superman is. He thinks that the proposition that Superman is Superman is (perhaps) logically true. He (correctly) says that I suggest that if a proposition is not logically true, then it provides genuine information. So he worries that the proposition that Superman is Superman (and so the proposition that Superman is Clark) do not provide genuine information and so cannot answer the question of who Superman is.

I do suggest that being-a-non-logical-truth is sufficient for providing genuine information. But I do not say that being-a-non-logical-truth is necessary for providing genuine information. In fact, I am inclined to deny that. So (I think) what I say about answers is consistent with saying that the proposition that Superman is Superman answers the question of who Superman is.

Further, I’m not sure whether the proposition that Superman is Superman is logically true. It seems to me that it entails that Superman exists, and that is not logically true.