De Natura Deorum Chapter 1: disputed passage

Pacifica

grammaticissima
Staff member
But, Godmy, keep in mind that the verb esse doesn't necessarily require a predicate (it doesn't necessarily require a Y). It can also denote the mere existence of X, the subject — and this subject can be modified by a relative clause, like any subject.

As in, for example, quis est qui hoc putet? = Who is there who thinks/would think this? There is no Y here. Qui hoc putet here isn't a Y as in quis est [is] qui hoc putet, which would mean something different ("Who is the one who would think this?"); but it's a relative clause modifying quis. I think the quid est quod + subjunctive construction is analogous to this.

Now this was my last post here for tonight.
 

Godmy

A Monkey
1) Any sentence requires a subject and predicate in the formal syntactic analysis. In syntax, if you miss a predicate, then it is not a sentence. And if you have a predicate, you always have a subject by default, even if only implied = if the subject is not mentioned, it is implied. just as "quis est [ille/homō] quī hoc putet"

2) No matter that, since we know that X and Y in X is Y have both different functions (just as "quid est quod"), if your quod wants to supplant either of them (let's say Y), then it cannot have as the antecedent the other thing (X let's say)

3) quid alone cannot be an antecedent to a relative clause nor it causes through the agreement the number and gender in the word "quod" to be what it is. It is some whatever neuter singular implied in the main clause (let's say [id] ) //granted that we interpret it as a relative clause
 

Imber Ranae

Ranunculus Iracundus
1) Any sentence requires a subject and predicate in the formal syntactic analysis. In syntax, if you miss a predicate, then it is not a sentence. And if you have a predicate, you always have a subject by default, even if only implied = if the subject is not mentioned, it is implied. just as "quis est [ille/homō] quī hoc putet"
I think you're confusing two completely different grammatical concepts, Godmy. PP is using the word 'predicate' in the more restrictive sense of the term 'predicative complement', which only applies to linking verbs (like copulative esse) and some passive constructions. It makes no sense to say that any sentence will require a predicative complement since not all verbs or verbal constructions can even take a predicative complement in the first place. The term 'predicate' in the broader sense, which is in opposition to the term 'subject' and which every sentence requires by definition, includes the verb itself (together with its complements, objects, etc. if it has them). But I don't think PP even knows that use of the word 'predicate', nor is it necessary to know in order to have a proper understanding of Latin grammar.

It's on this faulty basis that you now appear to be arguing that the verb esse requires a predicative complement, but that simply isn't true according to any standard grammatical analysis: it can also just mean 'there is X' or 'X exists', in which case there's nothing keeping X from being the antecedent of a relative clause, just as nothing prevents the subject of any other verb being the same. Now, if you want to argue that in fact esse can't mean that and must take a predicative complement, I'm afraid the argument you've put forth is wholly insufficient to show it because it relies on a false conflation of two different grammatical concepts which just happen sometimes to share the same word 'predicate'. In fact it's nonsense on its face because in effect you're saying that non-linking verbs don't require a predicative complement to serve as the predicate of a sentence, but esse can't be anything but a linking verb because it requires a predicative complement in order to serve as sentence predicate. But that's merely circular reasoning: why, after all, should non-copulative esse be held to a different standard than every other non-linking verb? Do you see what I'm saying?
2) No matter that, since we know that X and Y in X is Y have both different functions (just as "quid est quod"), if your quod wants to supplant either of them (let's say Y), then it cannot have as the antecedent the other thing (X let's say)
But here it isn't X is Y, it's just There is/exists X, so quod can't really be said to be supplanting anything.
3) quid alone cannot be an antecedent to a relative clause nor it causes through the agreement the number and gender in the word "quod" to be what it is. It is some whatever neuter singular implied in the main clause (let's say [id] ) //granted that we interpret it as a relative clause
I'm having trouble parsing your syntax here, so I'm forced to admit that you've completely lost me on your explanation as to why quid can't be the antecedent of a relative clause. You'll need to state your case more clearly.
 

Pacifica

grammaticissima
Staff member
But I don't think PP even knows that use of the word 'predicate'
I do know it, actually; what I didn't know was the exact term "predicative complement."
 

Godmy

A Monkey
IR, I'm reacting to your three answers to my three points. I hope I can manage without quoting:

1) Isn't that a strawman? I didn't say 'a predicative complement', neither PP did. She said predicate and hence I talked about predicate only because so I could talk about the subject, that was my goal (see more about the predicative complement in my next paragraph). We talked about the subject, since I think we agreed that "what" in "what" questions is rather a predicate, not a subject, so we talked about the other thing.

But then I also made a claim about what you say is a predicative complement:

In syntactical analysis, if either the subject is missing or the predicative complement is missing, they are both implied (even if missing - even the predicative complement) - they are morphosyntactically understood as to be there, as to have a position, they are treated as if present, and because of that some other words can even agree with them through the morphological agreement (in number and gender). I talk about that more further in this reply.

If you will, they are expressed by a zero lex(eme). //I would say "zero word", but that wouldn't look terminologically kosher probably.

I never said the word can't be missing! It clearly can. But it doesn't matter in the analysis. Its position still exists and takes/has number and gender.


2) There is "X" (quid) and implied "Y", or if you want, there is "Y" (quid) and implied "X". = the implied missing word is some rudimentary neuter singular, e.g. [id] - further opened to interpretation (to, let's say "fact" or "reason")


3) Well, the statement was so simple that I don't know how to write it in any other way: quid doesn't get modified by a relative pronoun - it just doesn't, ever. That's a fact, at least where I come from.

In this sentence if "quid" is X, then "Y" is some implied neuter singular, let's say [id] - since [id] is vague enough to be opened to interpretation, which is then modified by quod, which then has also neuter singular.
 

Pacifica

grammaticissima
Staff member
1) Isn't that a strawman? I didn't say 'a predicative complement', neither PP did.
I am sorry; I didn't use the most appropriate term, but I should have thought it was clear enough what I meant when I said esse didn't necessarily mean a Y (which I now know is exactly called "predicative complement", though it can still be called more vaguely "predicate", if IR is right.)
IR:
I think we agreed that "what" in "what" questions is rather a predicate, not a subject.
Not always; it can be either (for the two expressions we were talking about, I'll link again to this post).

What you seem to still not understand is that esse can not only be used as a linking verb in the X is Y kind of sentences, but also as a verb meaning simple existence or presence of something or someone: X just is there/exists — without "y". And that X can of course be modified by a relative clause, like any noun or pronoun can. And nothing prevents that X from being quid either.

Look:

Example of quid as a predicative complement, with esse used as a linking verb: Quid est hoc? "What is this?" = Hoc/this is the subject, and you're asking what it is: quid is a predicative complement.

Example of quid as a subject, with esse used not as a linking verb, but as a verb denoting mere existence (or presence): Quid est in hac spelunca? "What is (there) in this cave?" Quid here is the subject. Then, nothing forbids this subject from being modified by a relative clause: e.g. Quid est in hac spelunca quod te tam terreat? "What is there in this cave that scares (is such as to scare) you so much?" The quid est quod + subjunctive construction is analogous to this: quid est quod hoc faciamus? = literally "What is there with regards to which we should do this?" (What reason is there for us to do this?): it is not questioning what X is, which... as in an X is Y kind of sentence, but the very existence of such an X which... as in a X is (there/exists) kind of sentence.
3) Well, the statement was so simple that I don't know how to write it in any other way: quid doesn't get modified by a relative pronoun - it just doesn't, ever. That's a fact, at least where I come from.
It does; you'll have to come to accept it.
 

Godmy

A Monkey
PP, when I spoke about the missing subject and implied words in the analysis, it was meant also for a predicative complement (the word is missing, yet it has a position in the analysis = is implied), so forgive me making a statement about a general predicate (which was the word you used).

I won't argue about "quid" - since there is nothing to argue about. It never takes a relative pronoun, period.

...
 

Pacifica

grammaticissima
Staff member
I won't argue about "quid" - since there is nothing to argue about. It never takes a relative pronoun, period.
I don't understand your persistence in believing this (which is a strange idea in the first place; I don't understand where you get it from) disregarding the obvious, but there seems to be no way I can find the words to get the message accross to you, so... I'll give up.
 

Pacifica

grammaticissima
Staff member
(which is a strange idea in the first place; I don't understand where you get it from)
To be fair, now that I think a little about it, there do seem to be some languages that don't much like modifying interrogative pronouns with relative clauses (French doesn't usually do it), so maybe that's the case in Czech and it's influencing (and misleading) you in the analysis of such Latin constructions. (Just a hypothesis as to where you got the idea from; maybe it's something else entirely.)
 

Godmy

A Monkey
You cannot find any instance of it in Latin (or, as it seems, any other IndoEuropean language), of quid taking a relative pronoun. This would be the only instance, but I argue that it is not, since that relative clause (if it is one) needs to agree with the element of the main clause whose function it either takes or whose property it specifies. If it's a subject clause or specifies the subject, then it agrees with the subject, if it is a predicate [complement] clause or agrees with it, then it must agree with the predicate complement. One of them (subject or the predicate complement) is obviously missing from the main clause, but in the analysis we take it as if it was there, expressed by a zero word/zero lexeme, but implied: having its gender and number with which other words can morphologically agree.

This is also one of the reasons why quis (pronoun, syntactically usually a noun in the clause) doesn't need a gender (apart from having to be animate) in such cases as
quis est quīc hoc putet
quis est quae hoc putet // (in such case where the context implies that you speak about and pick from a female-only group, where you don't need the ambiguous quī for both)

Because in the main clause the second element is implied: quis est [homō/vir/ille... some masculine singular] quī hoc putet
or quis est [fēmina/mulier/illa... some feminine singular] quae hoc putet

Because the relative pronoun agrees with the element whose function it supplies or specifies.


Quid est in hac spelunca?
Alright, I will correct my previous statement: quid seems to be a predicate [complement] always in the cases, where there is the verb "esse" (directly or implied) with the function of a copula (as the type of sentence we were talking about).
 

Pacifica

grammaticissima
Staff member
Why do you need to imply ille/homo, etc., in quis est qui hoc putet, Godmy? Why can't you just understand quis as the subject, and est being used not as a copula between X and Y, but as a non-linking verb meaning existence? "Who is there/exists who would think this?" This is what I'm pretty sure it means, rather than "Who is the one who would think this?" as you interpret it; because I imagine that if it were as you say and there were some is/ille or sim. implied, it would sometimes actually be used (be physically there), for purpose of emphasis or whatever; yet this is not the case. I've never come accross such a construction, and in the LL at any rate, you find tons of examples of quis est qui + subjunctive; none of quis est is/ille/homo* qui + subjunctive.

*Edit: Concerning homo, I realize you might perhaps find examples (though I didn't now) of quis homo est qui + subj.; but in that case quis would be adjectival, and the whole mean "What man is there who would..." rather than "Who is the man who would..."
 

Pacifica

grammaticissima
Staff member
You cannot find any instance of it in Latin (or, as it seems, any other IndoEuropean language), of quid taking a relative pronoun.
In English, for one, "what" can take a relative pronoun. You'll find examples here (not all results are relevant, of course, but you'll see those that are).

There are also countless examples is Latin, but since you'll keep discarding them by saying there's an implied id, it's no use for me to provide some.
 

Godmy

A Monkey
I think, in this sentence, with "quis" you're asking for identity of a hypothetical man/woman with those properties in the relative clause and I think that interpreting "est" as "exstat" there is a needless step since you just ask about identity and probably not as two separate questions, whether: 1) he exists 2) exists with these properties.

The topic of the main clause is an enquiry about identity primarily, not really about existence, that is implied automatically in the case of a definite identity. It is just one question, not two connected questions.


The general idea of a copula is that someone "is" in some state, that someone "exists" in some state. And something different's happening in speakers' brains when they use the verb this way = when they use a copula than when they truly inquire about the existence. That's what linguistics teaches us. This allows some languages differentiate between these two so much that they can omit the copula altogether, while keeping the other intact (Russian: it never uses copula, but it always uses the verb of existence in the case, when it is truly meant).
 

Imber Ranae

Ranunculus Iracundus
1) Isn't that a strawman? I didn't say 'a predicative complement', neither PP did.
It's not a strawman because PP obviously meant a predicative complement when she said that esse doesn't require a predicate, because by the other definition of predicate it makes no sense to say that a verb can't take one. This is what people usually refer to when they talk about 'the predicate' (and other more specific terms like 'subject predicate' and 'object predicate') in Latin. The other kind of predicate, the one you said 'any sentence requires', is automatically fulfilled by the inclusion of a verb in a sentence, so it makes no sense to argue that non-copular esse doesn't exist on the grounds that all verbs must have a predicate, seeing as non-copular esse is the sentence predicate, just as any other verb will be (together with its objects and complements).
She said predicate and hence I talked about predicate only because so I could talk about the subject, that was my goal (see more about the predicative complement in my next paragraph). We talked about the subject, since I think we agreed that "what" in "what" questions is rather a predicate, not a subject, so we talked about the other thing.
Sorry, but I don't understand what you're saying here. It's true that quid will always be the predicative complement, and never the subject, with copular esse. But with other kinds of verbs, such as non-copular esse, quid can be the subject as well (with non-copular esse it will necessarily be the subject because it can take neither a direct object nor a predicative complement).
But then I also made a claim about what you say is a predicative complement:
You were just talking about the predicative complement, Godmy. It doesn't seem to me that you have the terminology down because you keep confusing the two.
In syntactical analysis, if either the subject is missing or the predicative complement is missing, they are both implied (even if missing - even the predicative complement) - they are morphosyntactically understood as to be there, as to have a position, they are treated as if present, and because of that some other words can even agree with them through the morphological agreement (in number and gender). I talk about that more further in this reply.
No, no, no. That's not what a predicative complement is. You're talking about the other kind of predicate now, Godmy, the sentence predicate, not the predicative complement. Only linking verbs can have a predicative complement. You need to learn the terminology before we can can make any headway in this discussion. Please look here (this isn't negotiable): predicative complement / predicate

You're talking about the latter here, not the former. Often people use the word 'predicate' for 'predicative complement', however, which is how PP was using it earlier and how it's generally used when discussing Latin grammar. But if you're going to bring up the broader meaning of predicate as the sentence predicate it's absolutely imperative that we distinguish it from the meaning 'predicative complement'.
If you will, they are expressed by a zero lex(eme). //I would say "zero word", but that wouldn't look terminologically kosher probably.

I never said the word can't be missing! It clearly can. But it doesn't matter in the analysis. Its position still exists and takes/has number and gender.
I already know that every sentence requires a subject and a predicate (but not a predicative complement!) But that isn't germane here, because as I've said before, non-copular esse, just like any verb, already fulfills the function of the predicate in any sentence where it occurs. On the other hand, only copular esse will require a predicative complement.
2) There is "X" (quid) and implied "Y", or if you want, there is "Y" (quid) and implied "X". = the implied missing word is some rudimentary neuter singular, e.g. [id] - further opened to interpretation (to, let's say "fact" or "reason")
Not true, because esse isn't always copular (i.e. a linking a verb). If you're going to deny this I'm going to have to ask for a source, because bare assertion doesn't hold any weight in academic discussions. The only argument that you've so far put forth to support it relies on a false conflation of sentence predicates and predicative complements, so that's going nowhere.
3) Well, the statement was so simple that I don't know how to write it in any other way: quid doesn't get modified by a relative pronoun - it just doesn't, ever. That's a fact, at least where I come from.
The statement wasn't simply stated, Godmy, mainly because it doesn't parse as English (i.e. it's non-grammatical), hence my confusion. If you had reread it you'd have been able to see that. But never mind, you've clarified: as you've now stated it, I can't do anything but dispute this 'fact' because it appears to me wrong on the face of it and you've not yet supplied any supporting evidence for it. And I don't care where you're from: bare assertion doesn't carry any weight.
In this sentence if "quid" is X, then "Y" is some implied neuter singular, let's say [id] - since [id] is vague enough to be opened to interpretation, which is then modified by quod, which then has also neuter singular.
Not true. Non-copular esse exists.
 

Imber Ranae

Ranunculus Iracundus
I think, in this sentence, with "quis" you're asking for identity of a hypothetical man/woman with those properties in the relative clause and I think that interpreting "est" as "exstat" there is a needless step since you just ask about identity and probably not as two separate questions, whether: 1) he exists 2) exists with these properties.

The topic of the main clause is an enquiry about identity primarily, not really about existence, that is implied automatically in the case of a definite identity. It is just one question, not two connected questions.
But the constructions quis est qui, quid est quod, quid est cur, etc. with the subjunctive all have very much to do with existence, because the very existence of what the relative describes is put into question. In fact it's a big part of the reason why these take the subjunctive rather than the indicative in the first place. They're the interrogative equivalent of such nihil/non est qui, nihil/non est quod, nihil/non est cur, etc.
The general idea of a copula is that someone "is" in some state, that someone "exists" in some state. And something different's happening in speakers' brains when they use the verb this way = when they use a copula than when they truly inquire about the existence. That's what linguistics teaches us. This allows some languages differentiate between these two so much that they can omit the copula altogether, while keeping the other intact (Russian: it never uses copula, but it always uses the verb of existence in the case, when it is truly meant).
But esse isn't always a copula in Latin.
 

malleolus

Civis Illustris
So is it a potential subjunctive?
Nope, it's a consecutive subjunctive meaning what reason is there to.
and belongs to the group of consecutive relative clauses just like non (nihil) habeo, quod/ nihil est, quod/(non) est, quod..
You wouldn't expect an AcI with constructions like these.
 

Callaina

Feles Curiosissima
Nope, it's a consecutive subjunctive meaning what reason is there to.
and belongs to the group of consecutive relative clauses just like non (nihil) habeo, quod/ nihil est, quod/(non) est, quod..
You wouldn't expect an AcI with constructions like these.
No, but I would expect an ut...

Edit: I mean with that idea of result (coming from a reason).
 

Pacifica

grammaticissima
Staff member
No, but I would expect an ut...

Edit: I mean with that idea of result (coming from a reason).
You can most probably say that some relative clauses have a nuance of result.

In the construction quid est quod + subj., you could take it as "What is there that is such that, as a result, we should..."

Much like in what's called "relative clause of characteristic", actually (and it doesn't feel that much different at all to me, to the point that I'm wondering if this quid est quod thing couldn't be called sort of a relative clause of characteristic, even if it may have another and more exact name), which you use when you mean not someone/something that actually does X, but someone/something that is such as to do X, so it can be taken as "that is such, that, as a result, they/it do/does/could do X". (Edit: Callaina , I've added "do/does" here to reflect the fact that something so mentioned can be indeed the case, even if it isn't really stated as a fact in the same way as it would be with the indicative.)

But in both cases, it still feels basically potential to me, because the result isn't necessarily always real, and those constructions could just as easily be analysed as, respectively, "What is there with regards to which we should (potentially, if it was there)..." and "someone/something that has the potential to..."

Now, to go even further, actually even the most basic ut consecutive clauses have their origin in a potential subjunctive, but native speakers messed up here, and progressively made it shift from the meaning "so that X could happen as a result" to "so that X actually happens as a result", giving rise to an illogical use of the subjunctive.

In my humble opinion (as I've already said a few times), all those categories often overlap quite a bit in fact (even if such or such thing is habitually called X or Y by grammarians, it can very well have a bit of Z as well, you see what I mean).
 

Callaina

Feles Curiosissima
(Edit: Callaina , I've added "do/does" here to reflect the fact that something so mentioned can be indeed the case, even if it isn't really stated as a fact in the same way as it would be with the indicative.)
Right, as is the case with this very example (because presumably people in Cicero's time did offer prayers and worship, etc. to the gods).

Now, to go even further, actually even the most basic ut consecutive clauses have their origin in a potential subjunctive, but native speakers messed up here, and progressively made it shift from the meaning "so that X could happen as a result" to "so that X actually happens as a result", giving rise to an illogical use of the subjunctive.
I don't know that it's that illogical; I mean, the subjunctive is used in other places (e.g. causal clauses, clauses of characteristic, etc.) to discuss things that are real and not simply potential. I've come to see it more as a tool for indicating abstract relationships (cause, effect, purpose, etc.) than a strictly "real vs. unreal" divide.
 

Pacifica

grammaticissima
Staff member
I don't know that it's that illogical; I mean, the subjunctive is used in other places (e.g. causal clauses, clauses of characteristic, etc.) to discuss things that are real and not simply potential. I've come to see it more as a tool for indicating abstract relationships (cause, effect, purpose, etc.) than a strictly "real vs. unreal" divide.
Yes, it's used the way you say in many Latin constructions — as in circumstantial, causal and concessive cum clauses, which all indicate absolutely real facts (purpose is still logically potential, though). But that may be (perhaps, in my idea), a deviation from the original basic meaning of the subjunctive, which I imagine was potentiality.
 
Top