[Essay] “this unusual instrument” with plural antecedents

[Charlie Bernstein]

"The glass harp" is not mentioned in the OP. Only "glass harps" is.
The former is only implied at best.

The question is why "this thrilling book" cannot refer to the Gothic novel. . . .

Because book is particular there. The best you can do there grammatically is "these thrilling books," but that still refers to three particular book.

If you change it to "this thrilling genre," it becomes clear that you're talking about a type of book.

In the same way, saying this instrument instead of this glass harp takes it out of the particular and generalizes it.

 

[jutfrank]

So the demonstrative adjective this does not have to refer to a singular antecedent?

I think you're making a categorical mistake in thinking in terms of grammar instead of semantics. The word this refers ultimately to a singular referent, not any grammatical antecedent.

The way reference in language works is such that a reference word (e.g. this) usually refers to an antecedent word, yes. However, this antecedent word is still itself only a referring expression, which refers in turn to a thing in the world (a referent). It is quite typical of the use of ordinary language for reference to skip the intermediary (antecedent) stage.

An important point to note here is that for some philosophers and semanticists, it is not necessary to postulate actual things in the world. It is quite enough to say that the reference 'ends' at the idea in the speaker's mind (the 'signified', in Saussurean language). When the speaker says this unusual instrument, the reference is directly to a singular mental object. There is no need for an antecedent.

 

[jutfrank]

John spent the summer reading Dracula, Frankenstein, and other Gothic novels. Who would have thought he was so into this thrilling book? (I guess this one is wrong, too.)

Well, there is obviously something wrong here but the problem is not to do with grammar but with meaning. The lack of grammatical cohesion between the sentences (i.e. the mismatch of number) is evidence of a problem in the speaker's mind. In other words, there's nothing ungrammatical about these sentences—the problem is that they do not cohere, which causes problems of interpretation. It's a question of coherence—making sense (or not).

At least, what I'm saying is true for native speakers. For English language learners, it is a bit different, in that for them it is at least conceivable (although very unlikely) that an error such as above could be the result of a misunderstanding of grammaticality alone. In other words, it is possible that the thought could be quite coherent in the speaker's mind, but that a lack of competence with (or understanding of) grammatical forms has led to the wrong choice of expression.

 

[raymondaliasapollyon]

Well, there is obviously something wrong here but the problem is not to do with grammar but with meaning. The lack of grammatical cohesion between the sentences (i.e. the mismatch of number) is evidence of a problem in the speaker's mind. In other words, there's nothing ungrammatical about these sentences—the problem is that they do not cohere, which causes problems of interpretation. It's a question of coherence—making sense (or not).

At least, what I'm saying is true for native speakers. For English language learners, it is a bit different, in that for them it is at least conceivable (although very unlikely) that an error such as above could be the result of a misunderstanding of grammaticality alone. In other words, it is possible that the thought could be quite coherent in the speaker's mind, but that a lack of competence with (or understanding of) grammatical forms has led to the wrong choice of expression.

Why is the example in the original post exempt from the requirement on the matching of number? You said in post #23 that this is because an expression does not need an antecedent; it can directly refer to a singular mental object. Now you are saying the novel example is ruled out because that requirement comes into play. The question now arises why this thrilling book cannot, in your words, directly refer to a singular mental object, in this case, the category known as the Gothic novel.

A principled, non-ad hoc account is much needed.

 

[Tdol]

And the original would not work if they listed two instruments, like glass harps and steel guitars.

 

[jutfrank]

Why is the example in the original post exempt from the requirement on the matching of number?

I've explained this. Because it's not necessary. The text is coherent. The reader can follow the thought perfectly well.

You said in post #23 that this is because an expression does not need an antecedent; it can directly refer to a singular mental object.

Yes.

Now you are saying the novel example is ruled out because that requirement comes into play. The question now arises why this thrilling book cannot, in your words, directly referto a singular mental object, in this case, the category known as the Gothic novel.

If it did, the text would suffer from a coherence problem. This means the reader would be confused about what this book refers to. There's no such confusion in the instrument example. It's quite simple, really.

A principled, non-ad hoc account is much needed.

I don't know what that means. You seem to be looking for the wrong kind of answer.

 

[Tdol]

BTW, raymondaliasapollyon, English language singulars and plurals are not always able to be hammered into mathematically precise models. You will hear plenty of educated speakers say there's two. The secret is to learn to accepts the patterns. We say what feels right. We know harps come in many sizes, so we use the plural and in the next sentence we consider them to be a singular, and we do it without seeing much, or any, contradiction. Singular and plural are sometimes fungible before they are absolute in language.

 

[raymondaliasapollyon]

I've explained this. Because it's not necessary. The text is coherent. The reader can follow the thought perfectly well.

How do you define cohesion and coherence?
One definition of coherence is "The set of relationships within a text that link sentences by meaning." And one definition of cohesion is "Grammatical or lexical relationships that bind different parts of a text together." In the novel example, "Dracula," "Frankenstein," and "Gothic novels" form a hypernymy/hyponymy relationship, thereby satisfying the criterion of cohesion.

Whatever is wrong with the example cannot be related to cohesion.

If it did, the text would suffer from a coherence problem. This means the reader would be confused about what this book refers to. There's no such confusion in the instrument example. It's quite simple, really.

What is the coherence problem in this case? Attributing the problem to "mismatch of number" is merely restating the problem rather than solving it. The thread is all about why some shifts in number are allowed and some are not.

I don't know what that means. You seem to be looking for the wrong kind of answer.

I'm looking for a principled, precise, and non-circular explanation.

 

[raymondaliasapollyon]

1. Glass harps haven’t entirely disappeared, though. You can still find some street musicians playing them to entertain large crowds. If you’re lucky enough to catch such a performance, sit back, relax, and enjoy the beautiful sounds of this unusual instrument.


2. John spent the summer reading Dracula and Frankenstein. Who would have thought he was so into this thrilling book?

3. John spent the summer reading Dracula, Frankenstein, and other Gothic novels. Who would have thought he was so into this thrilling book?

In #1, the sentence in blue moves the focus from street musicians playing glass harps, musical instruments, to an example of such a performance. A single performance involves one instrument.

I think this account makes the best sense. A particular glass harp is implied by that single performance, and that glass harp is treated as representative of the class of glass harps.

 

[jutfrank]

How do you define cohesion and coherence?
One definition of coherence is "The set of relationships within a text that link sentences by meaning." And one definition of cohesion is "Grammatical or lexical relationships that bind different parts of a text together."

Yes, exactly.

In the novel example, "Dracula," "Frankenstein," and "Gothic novels" form a hypernymy/hyponymy relationship, thereby satisfying the criterion of cohesion.

Yes.

Whatever is wrong with the example cannot be related to cohesion.

I don't follow this. The hyponymy relationship is not relevant in any way.

The cohesion problem I'm talking about is what I thought you were describing as a mismatch of number—precisely the problem that we're talking about! There is an issue with grammatical cohesion between the plural books and the singular this. I thought that's what you were saying, no?

What is the coherence problem in this case?

The coherence problem is that when the reader reads this thrilling book, it doesn't make sense. He doesn't know how to understand what the phrase is referring to. That's a coherence problem. You know all this, so why are you asking me to waste time explaining this to you?

I have a feeling that you are not actually interested in my answer, so this will be the last post I offer. If you are interested, read my previous posts again, more carefully.

Attributing the problem to "mismatch of number" is merely restating the problem rather than solving it.

First, well, actually it's just a way of describing the problem grammatically. Second, no, it is no way solving any problem. Why do you think it might be? What do you mean by solving the problem? I think you must have misunderstood something here.

The thread is all about why some shifts in number are allowed and some are not.

I wouldn't use the word 'allowed', but okay, yes, sure, we know that.

The answer I'm giving you is simple: if it's clear what the speaker/writer is referring to (like in the instrument example), it's allowed. If it's really unclear and confusing (like in the novels example), it's not allowed. Another way to say that is that if it makes sense it's okay and if it doesn't make sense, it's not. Another way to say that is if it's coherent, it's fine and if it's not, it's not. That's the best 'rule' that you're going to get.

I'm looking for a principled, precise, and non-circular explanation.

Look, I don't know how to be any more clear. Nothing I've said can be described as circular. If you think so, you've misunderstood. Reread the whole thread again.

 

[raymondaliasapollyon]

If a learner asks, "Why is X wrong?" he is asking what rules of language are violated, be they syntactic, semantic, pragmatic, or a combination of these. Therefore, asking why "this thrilling book" presents a problem and cannot refer to the Gothic novel is effectively asking what syntactic, semantic, etc. rules are violated. Answers such as "it just doesn't make sense" does not address why that interpretation is semantically faulty, i.e. what semantic rules are violated, and is therefore a circular account.

To be useful, a non-circular semantic account that attributes the unacceptability to a mismatch of number needs to explain how the semantically well-formed example in the OP differs from the novel example. Answers such as " 'this unusual instrument' refers directly to a mental object' would be valid only if we could determine when an NP can directly refer to a mental object and when it cannot.

 

[raymondaliasapollyon]

To test the claim that "such a performance" activates the implied singular glass harp, consider the following example:

Glass harps haven’t entirely disappeared, though. You can still find some street musicians playing them to entertain large crowds. This unusual instrument is especially popular in areas that have lots of tourists.


Here, the sentence does not have the "bridging link," i.e., "such a performance." If the shift in number is incorrect in this case, the example would lend support to the claim.

 

[jutfrank]

I'll largely ignore post #35 because I've already given my explanation that the incoherence is the result of a problem of reference. I can't see what more you need to understand.

Regarding post #37—yes, exactly. I agree with all of that. This is exactly what I was saying way back in post #2! However, be careful of saying that something is 'incorrect'. I would guess that some people would think of your modified phrasing as incorrect and some wouldn't.

 

[jutfrank]

If the shift in number is incorrect in this case, ...

The above suggests to me that you're still thinking in terms of grammatical correctness, which is, in my opinion, a mistake. To remind you again: in the sentence about the novels, individually, both sentences are grammatical. This means that they don't violate the rules of grammar. The problem is obviously something to do with the relationship between them. That's not a grammatical concern, it's a semantic one.

One cannot speak of semantic rules with the same notion of 'correctness' as one can of grammar. What would that mean? It is quite possible that you could use incorrect grammar to express to me a thought which I understand perfectly, thus satisfying the criteria for successful and effective communication. Would you then say that the utterance was 'correct' or 'incorrect'?

When analysing the meaning of an utterance, we often have no choice but to guess what mental content the speaker is attempting to express. In the original glass harp example, instead of asking "Is the shift in number correct?", a better question to ask is "Is the decision to attempt to shift the focus of thought by shifting the grammatical number likely to have been a successful one?"

 

[Charlie Bernstein]

To test the claim that "such a performance" activates the implied singular glass harp, consider the following example:

Glass harps haven’t entirely disappeared, though. You can still find some street musicians playing them to entertain large crowds.

Right. Them agrees in number with Glass harps. Done.

This unusual instrument is especially popular in areas that have lots of tourists.

Like the first sentence, this second one is complete and correct in itself. What's popular? This instrument.

The number of this sentence doesn't have to agree with the number of the first sentence.

Here, the sentence does not have the "bridging link," i.e., "such a performance." If the shift in number is incorrect in this case, . . .

But there is no shift. The noun agrees with the verb. You can say This instrument is or These instruments are. If there were a shift, it would either say This instrument are or These instruments is. That would be wrong. But it doesn't shift. So it's correct.

The number of this sentence doesn't have to agree with the number of the first sentence.

. . . the example would lend support to the claim.

Your example supports the claim that there is number agreement in both sentences.

You seem to be arguing for number agreement from one sentence to another. I don't know what "rule" you're defending.

 

[raymondaliasapollyon]

You seem to be arguing for number agreement from one sentence to another. I don't know what "rule" you're defending.

Number agreement between the subject and the verb is different from a shift in number, which is a phenomenon on the discourse level.

 

[raymondaliasapollyon]

You can't scientifically test a claim by changing more than one element.

How would you test the claim? The stretch of sentences has been changed in one crucial respect, i.e., the sentence introducing "such a performance" was removed. Whether the bridging link licenses the shift or not is being investigated.

With the use of the singular form with a definite article to suggest a class of things in the third of the sentences I've italicised, the writer has simply switched from one way of denoting a class (the plural) to another. You/we may not find this elegant, but there is nothing 'incorrect' about it.

If you don't like the term "incorrect," fine. How about "anomalous"? Also, is this kind of shift acceptable in a school essay? Would a student writing this way risk losing points?

 

[jutfrank]

is this kind of shift acceptable in a school essay? Would a student writing this way risk losing points?

I know next to nothing about the Chinese education system, but in the English-speaking world a student cannot lose points by writing an essay. He/she doesn't gain points either, for that matter. What a teacher does is try to grade the essay according to a set of criteria. Although a lot errors certainly can create a bad impression and therefore contribute to a low grade, as a general rule we tend to look for the good things in what the student has written.

 

[raymondaliasapollyon]

The above suggests to me that you're still thinking in terms of grammatical correctness, which is, in my opinion, a mistake. To remind you again: in the sentence about the novels, individually, both sentences are grammatical. This means that they don't violate the rules of grammar. The problem is obviously something to do with the relationship between them. That's not a grammatical concern, it's a semantic one.

Or a pragmatic / discourse one.

If by "grammatical," you mean "syntactic," then it is indeed not a grammatical concern.
But linguistics has assigned a broader meaning to grammar, which is often said to include syntax, morphology, semantics, and phonology. In that sense, any semantic anomaly could be said to be ungrammatical.

One cannot speak of semantic rules with the same notion of 'correctness' as one can of grammar. What would that mean? It is quite possible that you could use incorrect grammar to express to me a thought which I understand perfectly, thus satisfying the criteria for successful and effective communication. Would you then say that the utterance was 'correct' or 'incorrect'?

It would be grammatically incorrect (in the popular, narrow sense) but semantically well-formed.
I only used "incorrect" as a shorthand label. Maybe you prefer "anomalous."
But how could one speak of "semantic rules" if one denied "incorrect" could be used to describe a semantic anomaly?

 

[jutfrank]

Or a pragmatic / discourse one.

Yes. In fact, 'pragmatic' is actually a more precise word.

If by "grammatical," you mean "syntactic," then it is indeed not a grammatical concern.

Yes, I meant 'syntactic'.

But linguistics has assigned a broader meaning to grammar, which is often said to include syntax, morphology, semantics, and phonology. In that sense, any semantic anomaly could be said to be ungrammatical.

Woah, wait a second—that's rather controversial. Personally, I don't go with that at all. Syntax and morphology, yes, but phonology is quite different. And as somebody with a background in semantics, I would certainly not include semantics as part of grammar!

I only used "incorrect" as a shorthand label. Maybe you prefer "anomalous."

Maybe very slightly but it's still the wrong word. We sometime use the words acceptable/unacceptable on this website, which I'm not crazy about but which I think are much better words. Another way to express this is to say that something 'works' or 'doesn't work', which I think is very appropriate.

But how could one speak of "semantic rules" if one denied "incorrect" could be used to describe a semantic anomaly?

If you are genuinely interested in my answer, could you rephrase the question because I'm not sure what you're asking. Are you using the word 'rule' as something that necessarily must be either correct/incorrect by definition? If so, my opinion is that you get rid of the idea of semantic rules, unless what you really mean by 'rules' is what can better be called 'principles'.


May I ask what you are trying to do here with this thread, and on this forum generally? Are you really interested in learning something or is it that you wish to try to teach us something? Or maybe you enjoy having argumentative discussions about linguistics? Are you, or have you been, a linguistics student? May I ask to what level? I only ask these last two questions to better help myself and other members improve our answers.