You can see where he'd have trouble!

[jutfrank]

What about an argument from "sluicing," Piscean?

Sluicing is a type of ellipsis that only works with questions (embedded and root). The ellipsis always follows the wh-word.

One could say: I asked where Piscean worked, and he told me where. ["he told me where" = "he told me where [Piscean worked]"]

But one couldn't say: *[strike]I have been to where Piscean works, and I like where[/strike].

Do you mean to say that for an ellipted utterance such as He told me where, we can confidently assume that a question had been asked? I see, yes.

That's interesting but where's the argument?

Without the ellipsis (He told me where he works), we cannot make the same assumption.

 

[Phaedrus]

Do you mean to say that for an ellipted utterance such as He told me where, we can confidently assume that a question had been asked?

Not exactly. I mean that a sentence like He told me where is grammatical only if where is parsed as the sole remnant of an embedded question that has undergone ellipsis. Thus, I am confining my grammatical attention to the level of the sentence, making no assumptions about what has come before or what may come after. In actual usage, of course, the elliptical sentence requires an antecedent, but that needn't be supplied by a question -- e.g.: She doesn't know where he works, but I do. He told me where.

That's interesting but where's the argument?

I had only gestured at the argument. Now I shall try to make it explicit. (I hope you enjoyed that use of "shall" as much as I did.)

1. The type of ellipsis known as sluicing only occurs with interrogative clauses, embedded or root.
2. Ipso facto, the type of ellipsis known as sluicing does not occur with free/fused relative clauses.
3. A sentence like He told me where exhibits the type of ellipsis known as sluicing.
Therefore,
4. A sentence like He told me where contains an interrogative clause, not a free/fused relative clause.

Without the ellipsis (He told me where he works), we cannot make the same assumption.

Perhaps syntactic ambiguity between embedded question and free/fused relative holds in some cases with some verbs. I'm not sure whether tell is such a verb. If, with regard to the sentence He told me where he works, we try to interpret where he works as an free/fused relative, it will have the value of a noun phrase like the place. Now consider the sentence He told me the place, which resembles the sentence He told me his name. In each case, the noun phrase seems to stand for the answer to a question. In neither case does the sentence work like He told me a story.

 

[jutfrank]

Now consider the sentence He told me the place, which resembles the sentence He told me his name. In each case, the noun phrase seems to stand for the answer to a question.

I suppose you mean that it's likely that the NP stands for the answer to a question. That doesn't mean that it does.

I am confining my grammatical attention to the level of the sentence, making no assumptions about what has come before or what may come after.

Perhaps this is the root of disagreement. If you want to understand the meaning and use of any utterance of natural language, you ought not to confine analysis to the level of the sentence, or to a purely grammatical approach.

 

[PaulMatthews]

I don't see the point in making this a subordinate interrogative clause with the whole sentence meaning "You can see the answer to the question 'Where would he have trouble?"' I would think it was less convoluted to say that it was a subordinate fused relative clause. The meaning is You can see the situation in which he would have trouble.

I disagree.

This is certainly interrogative. It is not a question of visually seeing some location but of cognitively finding the answer to the question "Where would he have trouble?".

 

[PaulMatthews]

Hmm.
You continue to assert that it's an answer to a question, but you give no reason for this. Why do you insist on the question? Why can't we analyse it as 'You can see/understand the situation in which he would have trouble'?

In the analysis that you suggest, "where he would have trouble" would be an NP functioning as Od of "see".

But this doesn't work for the reason I gave in #33.

The same applies to the earlier examples, as discussed in #21.

I'm sorry if the interrogative analysis is beyond your grasp.

 

[Phaedrus]

If you want to understand the meaning and use of any utterance of natural language, you ought not to confine analysis to the level of the sentence, or to a purely grammatical approach.

Words of wisdom, Jutfrank. Still, syntax is interesting. Our problem is essentially whether "where he'd have trouble" is an interrogative clause or an NP.

If it were a noun phrase, it would contain a clause; but the outermost parsing of it would be a noun phrase, just as the outermost parsing of a gerund is an NP.

The argument from sluicing that I gave yesterday was homegrown, and I am pleased with it. But it can't prove anything about the unsluiced version.

In their article "The Syntax of Free Relatives in English" (1978), Joan Bresnan and Jane Grimshaw observe other properties of the two constructions.

One is that Pied Piping works with embedded questions but not with free relatives. I believe we can use that as a test here favoring the interrogative parsing.

(i) You can see with what he would have trouble. [cf. "You can see what he would have trouble with."]
(ii)*[strike]Look at with what he is having trouble[/strike] / *[strike]Observe with what he is having trouble[/strike].

I do find (i) rather awkward, but I do not find it ungrammatical, as (ii) clearly is. Another property is that embedded questions accept extraposition:

(iii) It can be seen where he would have trouble. [cf. "Where he would have trouble can be seen."]
(iv) ?* It is a tough part of the exam where he had trouble. [cf. "Where he had trouble is a tough part of the exam."]

 

[jutfrank]

Still, syntax is interesting.

To be frank, I generally think it's mostly pointless and only vaguely interesting so I'm impressed at how your posts always manage to be so pleasantly engaging. :-D

Our problem is essentially whether "where he'd have trouble" is an interrogative clause or an NP.

I wouldn't count this as among my problems (I'll leave it to you) but I'm listening keenly.

(i) You can see with what he would have trouble. [cf. "You can see what he would have trouble with."]
(ii)*[strike]Look at with what he is having trouble[/strike] / *[strike]Observe with what he is having trouble[/strike].

I do find (i) rather awkward, but I do not find it ungrammatical, as (ii) clearly is.

I'm not so sure that (i) is grammatical. It doesn't sound like something a native would say. (Then again, I have just repeated it so many times that I've lost almost all sense of judgement!) This may sound like a stupid question, but by what criteria are you judging its grammaticality?

Another property is that embedded questions accept extraposition:

(iii) It can be seen where he would have trouble. [cf. "Where he would have trouble can be seen."]
(iv) ?* It is a tough part of the exam where he had trouble. [cf. "Where he had trouble is a tough part of the exam."]

I don't see your point here. Is it that the grammaticality of (iii) is evidence that it doesn't contain an NP? Or evidence that it does contain an interrogative clause? And are you suggesting that (iv) is ungrammatical because where he had trouble is not an interrogative clause?

Sorry if I'm being a bit dim but I'd really like to understand.

 

[Rover_KE]

Thread moved as the OP lost interest after post #8.;-)

 

[Phaedrus]

To be frank

By all means, be yourself.

I see that this thread is now in the Using English equivalent of Siberia, but at least it had a good run before it was exiled from the visited forum.

I'm not so sure that (i) is grammatical. It doesn't sound like something a native would say. (Then again, I have just repeated it so many times that I've lost almost all sense of judgement!) This may sound like a stupid question, but by what criteria are you judging its grammaticality?

I judge grammaticality as you do, presumably, by apperceiving whether a sentence (or an attempted sentence) sounds grammatical to my native ears.

I admit that (i) is a borderline case, probably because of interference from "see." It's as if "with" wanted to mate with "see," as in "see with glasses."

Let's try a different case. There won't be such interference if we use "under" rather than "with." I doubt you'll have any problems at all with (vi):

(v) You can see what circumstances he would have trouble under.
(vi) You can see under what circumstances he would have trouble.

(vii) He should avoid whatever circumstances he would have trouble under.
(viii) *[strike]He should avoid under whatever circumstances he would have trouble.[/strike]

If you like, we could change "whatever" to "what" in (vii) and (viii), but I think (vii) and (viii) would be even worse with "what" rather than "whatever." This shows that after "avoid" an NP complement is needed, insofar as the -ever suffix never works on the wh-word of an embedded question. Nevertheless, compare (ix) with (vi):

(ix) *[strike]He should avoid under what circumstances he would have trouble.[/strike]

 

[jutfrank]

I admit that (i) is a borderline case, probably because of interference from "see." It's as if "with" wanted to mate with "see," as in "see with glasses."

Yes, that's what I suspected too.

Let's try a different case. There won't be such interference if we use "under" rather than "with." I doubt you'll have any problems at all with (vi):

(v) You can see what circumstances he would have trouble under.
(vi) You can see under what circumstances he would have trouble.

(vii) He should avoid whatever circumstances he would have trouble under.
(viii) *[strike]He should avoid under whatever circumstances he would have trouble.[/strike]

If you like, we could change "whatever" to "what" in (vii) and (viii), but I think (vii) and (viii) would be even worse with "what" rather than "whatever." This shows that after "avoid" an NP complement is needed, insofar as the -ever suffix never works on the wh-word of an embedded question. Nevertheless, compare (ix) with (vi):

(ix) *[strike]He should avoid under what circumstances he would have trouble.[/strike]

That's much better. Got it. :up:

 

[probus]

Thread moved as the OP lost interest after post #8.;-)

All I can say is where is Raymondaliaspollyon when we need him.

 

[jutfrank]

So, what is the justification for calling it an embedded question when there is no question, direct or indirect?

Unless I've misunderstood (which is quite likely), the term 'embedded question' seems to be an unfortunate shorthand for 'embedded interrogative clause'. That is, it's a purely syntactic term that does not necessarily have the pragmatic illocutionary force of asking a question. It seems to me that an embedded question posits a 'potential' question that may or may not be realised through expression in language, either publicly or privately.

Excuse me for attempting to answer what was not addressed to me. I keenly await other responses.

 

[Phaedrus]

What I still cannot see is why both the underlined clauses below are labelled 'embedded questions:


[1] Sally asked me where John lived.
[2] Sally told me where John lived.

Paul Matthews said '[2] reports the act of giving an answer to the question "Where does John live?" 'but does not appear to have offered us a justification for this claim. I suggested that no question has been asked. Jutfrank seems to have also cast doubt on the existence of a question.

I agree with Jutfrank that "embedded question," in the sense in which I and probably also Paul Matthews are using it, is equivalent to "embedded interrogative clause," and that such clauses are not always or even mostly used to report actual questions that have been asked.

My interpretation of what Paul Matthews has in mind, in the explanation of his that you have cited, is that [2] reports the act of supplying a value for a variable, namely, the variable denoted by the embedded interrogative clause "where John lived."

Incidentally, the first sentence of yours that I have quoted in the quote box above, Piscean, is, to my lights, a pseudo-cleft sentence in which an embedded interrogative clause ("why both the underlined clauses below are 'embedded questions'") is subject complement.

I can't make up my mind about the subject. Is it a free relative or another embedded interrogative? I'm inclined to think it's another embedded interrogative. It feels much different from, e.g., "What I still cannot see is pink and blue, according to Sally." Moreover, we could have:

Piscean told me what he still could not see. He told me that he still could not see why both the underlined clauses below are 'embedded questions'.

In that sentence pair, which is based on your pseudo-cleft, "what he still could not see" (non-assertive) is a variable the value for which is supplied by the "that"-clause (assertive) of the second sentence. Both the "what"-clause and the "that"-clause are complements following "told me."