Sunday 23 March 2008

On all and all

A correspondent writes to say that he caught himself saying All I find is idiots, and wondered whether it shouldn't have been All I find are idiots. He was saying to a friend that, when he wanted to find an expert in a particular thing, all he found was/were idiots, and he goes on: 'I realize that if all is used with a plural noun (with or without a partitive genitive) it takes a plural verb, i.e. All (of) the men are idiots, but what happens when all is used on its own? Is it like what, which, according to certain grammarians (Fowler, Gowers), can take a singular or plural verb? Or does it always require a plural verb in sentences such as my outburst?'

Certainly all on its own can be found with either a following singular or a plural verb.

All are welcome
All know the answer.
All are idiots.
All have a role to play.

Usage usually has the all reinforced with a pronoun, such as We all know...

All has gone quiet.
All is not lost.
All's fair in love and war.
All is dreams and sexual madness for Strindberg. (corpus example)

What we have, in other words, is a distinction between countable and uncountable. In (1) all could be replaced by many people; in (2) it could be replaced by everything. As my correspondent's example is plainly countable, the normal usage would thus be All are idiots. All is idiots is ungrammatical in standard English (though I've come across it in some regional dialects).

So why did my correspondent say All I find is idiots? I think it's because the postmodification I find alters the semantics of the subject, allowing an uncountable interpretation. He is right: it is pulling the construction in the direction of what. Fowler spent over two pages in his Dictionary of Modern English Usage trying to decide which way to go. He corrects, for example, What I wish to point out are certain instances... to What I wish to point out is certain instances.... He does not actually have a separate entry on all, but I imagine, if he had, he would have treated All I wish to point out... in exactly the same way.

So my correspondent used the singular because, quite plainly, all the so-called experts out there were being perceived as varieties of a single species, 'idiot'. If there had been a singular uncountable noun to express this notion in English, he would probably have used it. (Compare: All I see are leaves vs All I see is foliage.) But there isn't one, without saying something like All I see is idiocy. The reason he (and Fowler) feels uncomfortable is that there is always a problem when subject-concord concord pulls the verb one way and complement-concord pulls the verb another way. Usage has been split on this one for two hundred years, at least, and doubtless will continue to be, depending on whether the speaker has a unified or diversified meaning in mind (cf. committee is vs committee are and many other examples).


Anonymous said...

Hi. I was thinking about this 'All...'" business when I heard my friend say, "All I can think about is her". I know that your post concerns the NUMBER of the verb after 'All', but I am wondering about that CASE in this sentence (viz. "All I can think about is her"). Something tells me that old-school grammarians would object to "her" and replace it with "she" on the same ground that they object to "it is her" and say instead "it is she". Is my suspicion correct? From a strict grammatical perspective, should it be "All I can think about is she"? If so, is it because "she" is a 'subject complement'?

DC said...

Traditional grammarians would certainly have opted for the subject form here - see my previous post on who(m)ever for some further discussion of this sort of issue. Current usage undoubtedly favours the object form. Indeed, some object forms have even achieved catch-phrase status, such as it's all me me me.

The adjective strict doesn't make any sense to me, in this context. Modern linguists are just as strict (i.e. rigorous) in their statements about grammar - indeed, more so, as they try to make their statements a realistic reflection of usage. Ths is actually much harder to do than cook up an artificial rules.

Anonymous said...

OK, but if traditional grammarians would have for "she", what reason would they have given? Is it the subject complement business I was referring to?

I can see what you're saying about the descriptive/prescriptive issue. I imagine that all most everybody would say "All I can think about is her". But if the pronoun really is the SUBJECT complement, then doesn't that give us a good reason for putting it in the SUBJECTIVE case?

I'll have a look at your earlier post on the who(m)ever issue as well. Thanks for the info!

DC said...

Yes, it's the view that the subject and its complement should be in the same case.

'Doesn't that give us a good reason...?' Only if you hold the view that there is a rule in English which says that subjects and their complements must agree in case. There is no such rule in English - only in the minds of people who think that English must work like Latin.

Anonymous said...

Thanks for that. I take your point. I think I can add to it as well. I was just browsing through "The Linguistics encyclopedia" (by Kirsten Malmkjær) and she/he (I'm afraid I don't whether it's a man or woman!) makes essentially the same point as you. Here's a quick quote (which I add in case you think it would be a fitting footnote for this blog note): "The Latin case system and the rules for using it are then imposed on English...Palmer argues that this proof suffers from two defects, one being the virtual absence in English of a case system, and the other being the unjustified assumption that Latin should be a model for English; had a case language other than Latin been chosen as a model (French, c'est moi: It's me) the rule for BE might have been different."

I'm not sure where I stand personally. I can see what you're saying (and what the above author is saying). On the other hand there does seem to be some logic in putting a SUBJECT complement in the SUBJECTIVE case. Maybe it's just a preference of mine. Certainly it's stipulated as a rule in many (prescriptive grammars). Anyway, thank you for the reply

The Ridger, FCD said...

FWIW, I believe the argument is that since "She is all I can think about" is correct, the reverse must be "All I can think about is she". But honestly, does that sound like something any English speaker would ever say? Not to me.

There's no real reason to insist that "be" take subject case forms on either side of it. Since it's only a few pronouns (I, she, he, they) for which this even matters - the others, like nouns, have no "object case form" and we do nicely, don't we? - we might as well defer to what English DOES have, which is word order. For us, complements are complements.

DC said...

Quite so. Sadly, the argument that a usage is not 'something any English speaker would ever say' has never carried any weight with prescriptivists.

Anonymous said...

I have a point that relates to what "the ridger, fcd" said. I agree that ONE of the arguments offered by the prescriptivist is the "inversion" argument. In the Oxford English Grammar, Sidney Greenbaum writes: "If the copular verb is BE and if the subject predicative identifies the subject, the subject and subject predicative can change places." Greenbaum then gives as an example:

(1) The president was Bill Clinton
(2) Bill Clinton was the president

So, the argument might go, for it to be possible for subject and subject complement to swap places, we must write "she"- as the post above notes. If we can say "X is Y", then we must be able to say "Y is X." WE can only do this if we write "she", therefore we must write "she". Not sure what I make of this argument. There seems to be some plausibility in the idea that in identity statements it should be possible to swap subject and subject complement and (doing nothing else) get a grammatical sentence. Then again, maybe the above point about word order is valid. Or, putting it differently, perhaps the fact is that it may be ok to have a COMPLEMENT in the objective case, but it is never ok to have a SUBJECT in the objective case (the "subject" being the noun phrase that comes first, the complement being the noun phrase on the right side of "is"). I wonder what you think about that. It's one thing to end a sentence with a subject predicative/complement in the objective case. But it's quite another to start writing subjects in the objective case.

Anonymous said...

Just to add to the few points above. Hemingway writes: "All you have is me." Imagine how ridiculous this would have sounded if he had written: "All you have is I." I take it that "I" is subject predicative (or subject "complement") here ("All you have" being the subject). But if you follow the rule that subject and subject complement must be in the same case (subjective) you get "All you have is I," which is, to my ear, intolerable.

As for the argument that one ought to be able to reverse subject and subject complement (and get a grammatical sentence), there is this to remember:

even if you write: "All you have is I," you still can't reverse subject and subject complement and get a grammatical sentence. You end up with: "I is all you have." "I is" is not grammatical. You have to change the part of the verb "to be". But if you're allowed to change the part of the verb "to be" , then why shouldn't you be allowed to change the case of the pronoun?

So, as was said above, word order counts.

DC said...

I think Sid was making a syntactic point, not a morphological one. It depends on how you write your grammar, but it's often the case that the realization of a syntactic generalization needs to take into account low-level morphological variation. The point of the syntactic generalization is to distinguish this kind of construction from those where the inversion is not possible, such as John ate the meat.

The only case I can think of where a subject is in the objective case is in some regional dialects - her be ready to go, etc. In such a case, of course, the distinction between subjective and objective ceases to be relevant.

Anonymous said...

I recently went to a talk about this sort of thing in Polish and English, which confirmed for me something that I've been thinking about in English for some time. Sometimes the verb agrees syntactically with the subject (or agent, if you prefer):
(1) The committee is meeting tomorrow.
but sometimes it agrees with the semantics of the subject/agent:
(2) The committee are meeting tomorrow.
Any comments?

Anonymous said...

After more thinking about the all thing, a nonlinguist and I think there's a semantic distinction between what your original correspondent said (1) and what he thought he should have said (2).
(1) All I find is idiots.
(2) All I find are idiots.
We think that (1) can be paraphrased as (3) and (2) as (4):
(3) Everywhere I go/look, I only ever see/find people who are idiots.
(4) All I can see here are people who are idiots.
Which is, I think, what you're saying.

DC said...

Not really. It's what I mentioned at the end of the original post. Several nouns in English allow the distinction between a unitary and diversified meaning, and as they're well handled in the grammar books I don't think any extra comment is called for.

Anonymous said...

Even though the prescriptivists are using Latin as their model, English is also a Germanic language; and in German, the nominative (subjective) case is used after "sein" (to be), but after no other verb.
Personally, I feel that the accusative (objective) case works best in English after any verb (including to be).

Anonymous said...

Hello. I'm a student of English who has had some trouble in understanding subject-verb number agreement. So your original post, and its discussion of countable vs. uncountable was helpful.

I was hoping you could help me out on a number question that is similar to the original "all" question posted by your correspondent.

WHen the subject is "More than one" or "more than one X" , why is it that a singular verb is required (this is what I find in the books)? Why are the (a) versions considered to be grammatical and the (b) versions not?

(1a) There are several options. More than one IS possible.
(1b) There are several options. More than one ARE possible

(2a) More than one answer IS right
(2b) More than one answer ARE right

I suppose the answer to my question hinges on the syntactic analysis of "more than one (x)" and on what the head of this noun phrase is.

Thanks for the help!

DC said...

This is a long-standing point of discussion in the usage books. Traditional grammarians have always been disturbed by constructions which are singular in form but plural in meaning, and thus by people who make the concord follow the meaning. You will hear both (1a) and (1b), the latter being an example of a concord motivated by two factors - meaning and proximity. If you're thinking in a singular way, then you will say 'more than one is'. If you're thinking plural, then 'more than one are'. It depends on which notion is foremost in your mind - just as committee, for example, can be followed by is or are depending on the meaning. In a totally singular context, such as (2), then plural concord is not likely. In a totally plural context, such as 'more than one of these books', singular concord is less likely - though, of course, traditional usage books will recommend a singular here regardless, just as they do after none.