(2) Dionne knows the way to San Jose.

(1) Andy knows the time the bar closes, but not as such.
(2) The bar is such that Andy knows the time it closes, but not under that description.

This is still a concealed question use, and I say still propositional, not objectual.

I don’t understand why you would bother. Why not just note an ambiguity? We already know the same word “know” is used in English where other languages use different words such as “kennen” (I’ll call this “acquaintance”) and “wissen”. So it seems uses of “know” with descriptions as complements could just be ambiguous between the two.

Note the acquaintance reading I alluded to meets the criteria given for the “objectual” sense (as in “knows the mayor”): it is extensional and does allow “but not as such” with a perfectly clear sense.

I wondered at first why you starred 22 through 24, since it seemed to me they could all be made acceptable (though on an acquaintance reading only) given appropriate stories. You just have to figure out when it would be correct to say someone is acquainted with a time or the object that is a question answer.

For example: At school a test asks students to recall difficult words they’ve been taught given definitions. Bonus question 10 is a real stumper. One student could brag “My Dad has a great vocabulary, I bet he knows the answer to question 10 [though not as such]” to say simply that his father knows the difficult word itself.

Or: John is regularly awakened from his sleep by the noise of patrons leaving the bar at closing time (though he doesn’t know that’s the cause) and comes to curse that ungodly hour. To his regret, he has come to know all too well the time the bar closes (though not as such).

So it seems it is only the concealed question reading that merits the stars; in the acquaintance sense 22-24 are fine. So I think you just have to demarcate the former as the sense you are really discussing.

First, it seems to me that `but not as such’ is fine with some `knowledge the’ attributions. For example:

(2’) Donna knows the fastest way to San Jose.
(7’) Andy knows the time the bar closes on Sundays.
(8’) Sarah knows the least controversial answer to the first question on the history exam.

(22’) Donna knows the fastest way to San Jose, but not as such.
(23’) ? Andy knows the time the bar closes on Sundays, but not as such.
(24’) Sarah knows the least controversial answer to the first question on the history exam, but not as such.

(For (23’) imagine that Andy knows that the bar closes at 2 am on Mondays, Tuesdays, … , Saturdays, but thinks it might close early on Sundays.)

I still don’t quite see why the original (22), (23) and (24) were bad.

Second, isn’t it misleading to think of the phenomenon as `knowledge the’? We see it, e.g., in:

(a) Donna knows several ways to San Jose.
(b) Sarah knows every correct answer to the first question on the history exam.
(c) Andy knows a recipe for pizza dough.

Heim, Irene. 1979. “Concealed questions.” In Rainer Bäuerle, Urs Egli, and Arnim von Stechow (eds.), Semantics from Different Points of View, 51-60.

Lance Nathan, one of our graduate students, just defended a fine generals paper on concealed questions, which he should be able to supply upon request.

I don’t take the translation data quite as seriously as Mike, or for that matter a lot of philosophers. I’ve heard, for instance, that in Serbian there are different words for know-that and know-how. I don’t know whether it’s true, but I’ve heard it. But even if it’s true, I wouldn’t take that to be strong evidence against the Stanley/Williamson thesis about know-how. Even knowing the translation data, there’s still the stuff about ‘wonder’ to explain, and there’s still the pressing question of how (8) gets to express (31).

Anders, I was taking those readings to be on the table, but I still think they are propositional (or informational if you prefer). They are complicated cases, but I think basically they can be handled by familiar means. But it was just those readings I was thinking about in the paragraph after (25).

I appreciate that you are not considering this sense at all. Just wanted to note that it exists.

Now the noun objects of knowledge-the are informational too: know the way to, the answer to, the result of, the reason for, the truth about, the cause of, etc.

In (1) the information is presented clausally, in (2) nominally, but semantically they’re the same kind of thing. But is (2) then just the same as (3) where the knowledge-object is information rather than a material object? A test would be whether you can know information by acquaintance, parallel to (3) and different from (2).

Suppose you’ve previously seen an exam question come up in past papers: or you’ve previously seen a puzzle explained. If you can’t remember exactly the answer or solution, you can certainly say

I know this question/puzzle/riddle.

If knowledge is sometimes a kind of familiarity, rather than requiring certainty and confidence, could you then say

? I know this answer/solution

meaning you’ve seen it proposed before, or approached like this before — not necessarily that you know the answer in the informational sense?

interested in how to get free drinks off her
because of what she likes to drink
dubious about what/how…
*in/of/about that she likes Red Bull and vodkas

So ‘know’ with a conjunction of N-complement and that-complement is a syntactic violation, more serious than a possible semantic syllepsis.

However, it’s not a simple nominal. The punctuation is a clue:

(29) Andy knows the bar manager, and what she likes to drink
(29’) * Andy knows the bar manager and what she likes to drink

The intonation of (29’) implies a conjoint NP that’s the single object of the verb ‘know’, as in ‘Andy knows the bar manager and her sister’. This doesn’t feel right for me with (29’), which is rather better as (29), created by left node raising from

(29’‘) Andy knows the bar manager, and [Andy] knows what she likes to drink

comparable to ‘Andy bought the bar manager a tequila, and her sister a vodka’ — the omitted part in the second clause is not a constituent, but a string of words in common. This would allow the syntactic conjunction of two semantically different words, as long as the syntactic condition of a following nominal complement is met.