The English-Learning and Languages Review Ć HOMEPAGE
A critical examination of the account of
some and any
Rodney Huddleston, Geoffrey K Pullum et al. (2002)
(The false trail of grammatical categorising 3)
The material discussed in this article comes mainly from chapters 5 and 9, by John Payne and Rodney Huddleston, and by Geoffrey K. Pullum and Rodney Huddleston, respectively.
Vague use of some
The authors of The Cambridge Grammar of the English Language (CGEL) claim that readers need no training in linguistics. It is true that the huge number of technical terms used are explained somewhere or other in this massive volume of 1,859 pages. But it is very doubtful whether there is anything here that would be useful for the learner of English as a foreign language, at whatever level. And a vast amount of the book is occupied by a minute analysis of sentences.
Page 423 begins like this:
9.6 Compound determinatives
Compound determinatives such as somebody occur as determiner-head with the syntactic fusion of the two functions marked by the morphological compounding of a determinative base with a nominal one.
On page 360 under the heading Count NPs the authors write:
 a. He had eaten [all of the pies]. b. He had eaten [some of the pies.]
Example [a] entails that there were none of the pies that he hadn’t eaten, while existential [b] entails that not all the pies had the property that he hadn’t eaten them. There is a complication in the existential case, however, in that some in [b] indicates not just a number greater than zero but a number no less than two ……
The CGEL devotes, partly or wholly, more than fifty pages to some and any and their compounds. It introduces the subject in Chapter 5 “Nouns and noun phrases” under the heading “5 Quantification”. In § 5.1 “Existential quantification, universal quantification, and negation” (pp.358-59) it says
 EXISTENTIAL QUANTIFICATION ………….
a. [Some of the meat] was fresh. ………….
Existential quantification indicates a quantity or number greater than zero, and has some as its most straightforward expression. The term ‘existential’ reflects the fact that elementary examples like [a] assert the existence of a quantity of meat that was fresh, i.e. a quantity having the predication property.
The examples so far have had the quantified NP located before the verb, in subject function. When we turn to postverbal NPs we find a third possible expression of existential quantification, any. …………………………….
There is probably nothing untrue in these statements, but they tell one nothing useful. One can equally call a or the expressions of existential quantification, since they too indicate a quantity or number greater than zero. And it is questionable whether any can really be described as a ‘quantifier’. With one exception, it is never absolutely necessary to use it with plural nouns, although we very often do, simply because we are often thinking in plural contexts. The exception, of course, is with the phrase any of. But the any in any of the books is no more a quantifier than best in the best of the books. Some complete sentences with any may seem to have a quantifying aspect, such as
Have we any milk?
but others, such as
If you find any solution, let me know.
do not at all.
The CGEL continues (from now on I use bold type to replace the double underlining that the original CGEL text uses to indicate the item that has “scope over” the item with single underlining):
 i a. He hadn’t eaten [all of the meat]. = b. He hadn’t eaten [some of the meat].
ii a. He hadn’t eaten [all of the meat]. = b. He had eaten [none of the meat.]
b’. He hadn’t eaten [any of the meat].
In [ia] the negative has scope over all (“It is not the case that he had eaten all of the meat”). This is equivalent to [ib], where some has scope over the negation: “There was some of the meat that he hadn’t eaten”. Version [ib] is much less likely than [ia], though the construction is certainly admissible (cf. I didn’t agree with some of the things he said). Construction [iia], with the meaning “all of the meat had the property that he hadn’t eaten it”, is somewhat marginal and contrived: He hadn’t eaten all of the meat would generally be interpreted with the negative having wide scope, as in [ia]. Much more likely than [iia], then, is a version where negation has scope over existential quantification. And this can take either of two forms, one with the negative existential quantifier none, and one with a negative having scope over any. In the context of a wide scope negative, any is normally used instead of some to express existential quantification; we will use the asterisk notation for *He hadn’t eaten [some of the meat], with the proviso that it could be used in the special context of denying a previous utterance of He had eaten some of the meat or the like.
Any can also be used in interrogative contexts, as in Had he eaten any of the meat? Here Had he eaten some of the meat? Is also possible, but it suggests the questioner is disposed to think the answer may well be positive, whereas any is neutral, giving no indication of the speaker’s attitude to the possible answers (see Ch. 10, §4.7.4). However, any cannot substitute for some in He had eaten some of the meat: in the sense we are concerned with here, any is restricted to non-affirmative contexts (negatives, interrogatives, and various others, as described in Ch. 9, §4.4). This gives us three expressions of existential quantification: basic some, non-affirmative any, and negative no.
Here we come straightaway to a fundamental weakness in the method of most grammarians. They find – quite correctly – that they cannot describe satisfactorily the meaning of words (in this case some and any). So instead they try to explain them by pushing them into grammatical categories. In the present case they are probably also making the common basic mistake of attaching the meaning of the whole to the meaning of the part; confusing the meaning of an individual word (here any) with the meaning of the sentence it forms part of.
The authors appear so engrossed with their category of ‘non-affirmative’ any, and their (perfectly sound) principle of ‘scope’, that they fail to notice that they have themselves illustrated, with their examples ib and iib’, the really important point, i.e. that some and any have completely different meanings.
He hadn’t eaten some of the meat.
He hadn’t eaten any of the meat.
They again fail to note the distinction on page 360 under , where they offer the examples
He hadn’t eaten some of the pies.
He hadn’t eaten any of the pies.
Indeed, the CGEL’s authors tell us, by implication above, and explicitly on pages 381 and 383, that any has “the same sense as some…”! This failure to distinguish between the meanings of some and any may seem strange, but it is an example of what can go fundamentally wrong when one treats all pieces of language as, first and foremost, grammatical categories. The CGEL appears to regard some and any simply as functions of grammar, rather than two separate unique meanings, meanings just about as different from each other as it is possible to be. Native English-speakers naturally understand perfectly what some and any mean. But when academics start on their grammatical analyses, they seem, in a very real sense, to stop understanding them. (At the top of page 383 the CGEL declares:
Free choice anyf is best treated as having the same sense as some in its basic proportional use, …….
I find this a very puzzling statement. What does treated mean? Best treated from what point of view? Best with what purpose in mind? The authors talk as if there were no objective truth to be conveyed. For the CGEL’s ‘free choice anyf’ see below.)
The CGEL in fact started making the mistake of ascribing to syntax what are purely semantic phenomena way back in “Chapter 2 Syntactic overview” (p.60).
A number of words or larger expressions are sensitive to polarity in that they favour negative over positive contexts or vice versa. Compare:
 i a. She doesn’t live here any longer. b. *She lives here any longer.
But b. is nonsense simply because it is nonsense to say something like
* She lives here whatever length of time you like to think of longer.
* She lives here indiscriminately longer.
One has to ask why anybody should ever dream of saying such a thing as b. It seems to me that one can only make a point like the CGEL’s here if an obsession with syntax makes one blind to the reality of meaning.
On page 829 the CGEL insists that ‘positively-oriented’ some is ‘inadmissible’ in the negative sentence
They didn’t make some mistakes
but that, in contrast, in changed ‘scope’ conditions,
I didn’t understand some of the points she was trying to make
is an ‘admissible’ sentence, which they paraphrase as “some of the points…had the property that I didn’t understand them”.
In fact the two sentences are in principle exactly the same. The ‘scope’ relationships can be interpreted as identical in the two sentences (the first sentence as “some mistakes had the property that they didn’t make them”). It is an example of how unusual or unnatural situations in real life can deceive analysts into thinking that the language expressing such situations is prohibited by some grammatical law. We can make the first sentence a bit more realistic by expanding it a little. For instance:
This time they didn’t make some of the old mistakes –
or by stressing some. But the CGEL seems to have been led astray by the rather peculiar idea expressed into thinking in advance that the sentence is ‘inadmissible’, and so believing that the ‘scope’ is such as to make it so.
The fact that any is often found in negative clauses seems to have led to a conviction among generations of grammarians of English that there is something essentially negative – or at least ‘non-affirmative’, as the CGEL puts it - about the word (see above). But affirmative declarative statements using any are perfectly normal:
He could have eaten any of the meat or
He can eat any of the meat or
He eats any meat.
The acceptability or otherwise of sentences is determined by two things: one, the particular combination of meanings in them, by whether the meanings are compatible with each other; and two, by whether it is the convention of contemporary native speakers to use a given formulation – sometimes a combination of meanings may be perfectly sensible, but is simply not used. (I like her much makes perfect sense, but breaks the current convention; while on the other hand I grew up thinking that one ‘can’t’ say It will likely rain – that it has to be It will very likely rain – but have discovered that – particularly in America? – the very is often absent in today’s English. See also www.lingua.org.uk/aintnogram.html ).
*He had eaten any of the meat
is nonsense, because it means something more or less like
*He had eaten whichever of the meat you like to think of.
He could have eaten whichever of the meat you like to think of
makes perfect sense.
The CGEL’s authors might argue that the any in He could have eaten any of the meat is a ‘free choice’ any, a different sort of any from any in *He had eaten any of the meat. I do not accept that any here is an example of their ‘free choice’ any, but let us examine their argument.
Free choice any and either
Besides their use as an existential quantifier in non-affirmative contexts, any and either can be used with what is called a free choice sense:
 a. Any of these computers will do. b. Either of these computers will do.
Again, any here implicates a set of three or more computers while either ……… The interpretation is that if you choose arbitrarily from among the set of computers – i.e. make a free choice – the one you choose will have the predication property, i.e. in this case it will do (“satisfy the requirements”). In the free choice case the quantifier is always stressed. Where relevant, we will distinguish the two senses with subscripts: anyn and eithern represent occurrences of these items with the non-affirmative existential sense, while anyf and eitherf represent occurrences of them with the free choice sense.
Anyf and eitherf are not excluded from non-affirmative contexts, so that there may be ambiguity between the two senses, as in:
In the non-affirmative existential sense, this asks if there was at least one of the students who was allowed to take part. In the free choice sense (with stressed quantifier) it asks whether permission to take part was available generally, i.e. whether a student chosen arbitrarily from the set could take part. What is at issue in this second interpretation is whether or not restrictions applied as to which or how many students were allowed to take part.
And on page 382:
ii Any policeman will be able to tell you. [count singular]
iii Any remaining dirt will have to be removed. [non-count]
Anyf occurs with all three of the main kinds of head: plural, count singular, and non-count. It indicates that there is a free choice: an arbitrary member (or subquantity) can be selected from the set (or quantity) denoted by the head and the predication property will be apply to it. In [ii], for example, there is a free choice as to which policeman can be selected, but no matter which policeman it is, that policeman will be able to tell you. Anyf implicates that the free choice will only have to be made once, but this implicature can be cancelled. In [ii], since the first policeman selected will normally provide the required information, it will not be necessary to ask another one. ……….
(Note how the above illustrates how impossible it is to describe words with other words. ‘Free choice’ is a particularly unsatisfactory way of describing any – of whatever kind. There is no choice involved in the meaning of their ‘free choice’ any. One does not wander around looking at various policemen and then select one. This is very clear in an example like Any fool could have told you that.)
And on page 383:
There are several differences between the two any’s. We noted that anyn is usually unstressed; anyf, by contrast, is always stressed. Secondly, while anyn is restricted to non-affirmative contexts, anyf is not polarity-sensitive. The anyf examples given above are all in affirmative contexts, but in a negative context we can have a contrast between the two senses:
ii [We don’t publish just anyf letters:] we reject more than half of those submitted.
Example [i] is equivalent to We publish no letters. In [ii] the free choice is negated: we ourselves are making a selection, accepting some letters, rejecting others. Thirdly, anyf, unlike anyn, permits modification, for example by the adverb almost. Compare:
ii *Jan couldn’t find [almost anyn computer magazines] in the shop.
The CGEL is convinced that any has a ‘non-affirmative’ meaning. But it has to recognise that any is often used affirmatively. So its authors are forced to postulate this second distinct meaning. This distinguishing of two senses of any may be valid up to a point. It would seem to be confirmed by examples like  above.
On page 382, though, the CGEL says:
Anyn is usually but by no means always unstressed. It can be stressed, for example, when it is the focus of negation: I don’t think ANY people left early. The negative orientation of anyn can be reinforced by the polarity-sensitive at all or whatever : We hadn’t made any progress at all / whatever.
However, I suggest that for native English-speakers any always contains the idea which the CGEL tries to describe by talking about ‘free choice’. This is the ‘whichever, whatever, it doesn’t matter which, whichever you like to think of’ idea. Whatever can also be used together with what the CGEL calls the free choice any. For instance, we could add it to the CGEL’s example  ii on page 382:
Any policeman whatever will be able to tell you.
And we can add it to the ‘ambiguous’  sentence quoted above:
Were any of the students whatever allowed to take part?
Even if we assume, as I do, that the any here is just one and the same undifferentiated any, that there are not two distinct any’s with different senses lurking here, the sentence is still ambiguous. It is not a double-meaning any that causes the ambiguity. The ambiguity is caused by the particular combination of meanings in the sentence. For example,
Can any of the students take part?
is ambiguous too; but
Do any of the students take part?
is not – only one interpretation is possible. Yet again,
Do any of the students satisfy the requirements for taking part?
 i could well be stressed. Yet in practice, even then, the any’s in i and ii would not sound the same. Most speakers would indeed add the just in a sentence like ii; but with or without just, the intonations of any in i and ii, even if both are stressed, are different.
As for almost, here again is an example of the failure to understand the decisive part played by particular combinations of meaning. It is not any supposed difference in meaning between anyf and anyn that makes  ii nonsense. It is the combination of the negative with almost any that makes it nonsense.  i would equally be nonsense (except as a rebuttal of a previous statement that Jan will etc.) if one turned Jan will into Jan won’t:
*Jan won’t read almost any computer magazines.
And in fact the distinction between two meanings of any is unclear, imprecise, and ultimately questionable. There are several examples in the CGEL that show how fuzzy and uncertain it is.
Take the case of conditional sentences. These, such as (CGEL pp.745 and 838)
If anyone has a solution to this problem, please let me know,
If I’d ever seen anything like that, I’d have reported it,
are supposedly part of the justification for categorising any as ‘non-affirmative’. But they can often be paraphrased to provide examples of what the CGEL would call presumably ‘free choice’ any:
Anyone with a solution to this problem should kindly let me know,
I’d have reported anything I saw like that.
Similarly (p.836) in
the CGEL maintains the any is the ‘non-affirmative’ sort. But the sentence can be paraphrased as
Any unnecessary risk-taking would be foolish,
where the any would clearly be called ‘free choice’ by the CGEL. And in another of their examples of supposedly ‘non-affirmative’ any’s on page 836,
Any more pudding would be quite excessive,
it is very unclear what sort the any is. To me it looks much more like one of their ‘free choice’ any’s.
And what should one make of sentences like
You have to press hard to have any effect.
Anyone with any sense would have stayed at home. ?
The any’s look more like the CGEL’s ‘non-affirmative’ any’s than their ‘free choice’ any’s. Yet according to the CGEL they must be ‘free choice’ any’s, because ‘non-affirmative’ any’s are not, they say, allowed in affirmative declarative sentences.
In any case, the CGEL fails once more to convey anything of the essential meaning of any. It is much more accurate, practical, and psychologically real – in the sense that it is the way native English-speakers normally experience any - to regard all any’s as one and the same. They all carry the ‘whatever, whichever’ idea.
The CGEL tries to fit everything, including of course the any series, into its huge, elaborate system of grammatical categories. In the process it has to resort to what sometimes seem to be rather desperate measures to get out of difficulties this approach lands it in. It claims that ‘non-affirmative’ contexts ‘admit’ ‘negatively-oriented polarity-sensitive items (NPIs)’, of which the ‘non-affirmative’ any series of words, but not ‘free choice’ any’s (pp.823,826), are an example; while ‘affirmative’ contexts ‘exclude NPIs’ (pp.822-38). But awkward facts get in the way of the CGEL’s construct. For instance, the last two CGEL sentences quoted above come from a series of examples of ‘NPI’s in what appear to be affirmative contexts. But they are not really affirmative contexts, say the CGEL’s authors (p.835):
Covertly negative lexical items with clausal or clause-like complements
Many verbs and adjectives that take clauses as complements are covertly negative in that they trigger entailments or implicatures involving the negation of the subordinate clause, and this is sufficient to sanction NPIs in those clauses. …………………..................
We group these covertly negative items into six classes: (a) failure, avoidance, and omission; (b) prevention and prohibition; (c) denial; (d) doubt; (e) counter-expectation; and (f) unfavourable evaluation.
I list below the CGEL’s examples of sentences with these supposedly ‘covertly negative items’ which contain one of the any series:
 iv We managed to avoid any further delays.
 i We kept him from telephoning anyone before the police arrived.
ii I am prohibited from so much as naming any of the principals in this case.
 i My client denies that he ever said any such thing.
ii My client denies any involvement in the matter.
iii My client completely rejects the notion that he ever said any such thing.
 iii I’m doubtful about the value of pursuing the matter any further.
iv She expressed scepticism about there being any point in continuing.
A statement like I’m surprised the car started asserts that I have experienced a reaction to the discovery that the car started because that is counter to expectation: it implies that I had a prior expectation that could be expressed as The car won’t start.
 i It astounds me that they took any notice of him.
iii We were all amazed that he had been able to write anything during that time.
 i It would be foolish to take any unnecessary risks.
ii Any more pudding would be quite excessive.
I don’t think languages work in the way the CGEL proposes: grammatical categories admitting or excluding other grammatical categories. We use any etc. in these sentences, not because they are ‘negatively-oriented polarity-sensitive items’ triggered by ‘covertly negative lexical items’, but simply because any is the word we have to use in these contexts if that is the meaning we intend. What else could we say if that is what we want to say? In all the above sentences except  i,  iii, and  iii, we could use - perfectly acceptably from the linguistic point of view - some instead of any. In  i we could say someone instead of anyone; in  iii we could say much instead of any; in  iii we could use something instead of anything. These changes would give the sentences quite different meanings. In most cases they would be odd things to say; in most cases it is very unlikely that many people – or any people – would ever say them. But they are excellent examples of the danger of confusing the rare or unlikely in life with unacceptability in language.
Any (etc.) is the natural word to use in these contexts, because a speaker of them will normally need an extreme, an emphatic word, and this is precisely what any is. See www.lingua.org.uk/saquirk.htm.
The argument under (e) seems to me an example of how desperately the CGEL has to strain in order to justify its assertions; in order to fit together pieces that don’t want to be fitted together. The fact remains that in the examples the authors offer, their supposedly “negatively-oriented polarity-sensitive items” are being used in positive contexts. Categorisation of any etc. as negatively-oriented polarity-sensitive items is, I believe, the sort of fantasy that grammarians impose on themselves.
Fundamentally, creating language is not a negative process of ‘admitting’, or ‘excluding’, or ‘sanctioning’ categories, but a positive one of combining different meanings to make the sense one wants. Actually, at this point in the CGEL the authors are not very far from saying precisely this; but unfortunately their whole approach prevents them admitting any such thing.
We do not say things like
*I have any books.
That is not because the any is a ‘non-affirmative quantifier’ in an ‘affirmative declarative’ sentence, but because it is nonsense. On the other hand,
I read any books
Returning to Chapter 5 (p.362) we find the CGEL continuing to misunderstand what determines the acceptability or otherwise of sentences:
*I had been for a long walk and was feeling hungry, so I ate any/either of the pies excludes not only the non-affirmative reading (by virtue of being a positive declarative main clause) but also the free choice one.
But one can say:
I’ll eat any of the pies or
I can eat any of the pies or
I always eat any pies.
It is the meaning of a particular occasion in the past in the CGEL’s example – once more, a particular combination of meanings - that makes any nonsensical there. Even the use of any in a past context, though, can be perfectly normal:
I was ready to eat any of the pies or
When I was a boy I ate any pies or
I could have eaten any pies.
And is it clear what sort of any the any’s are in these sentences? Are they all unquestionably the CGEL’s ‘free choice’ any’s?
On pages 364-65 the CGEL introduces a distinction between
Proportional and non-proportional quantification
The “not all” implicature is not found with all uses of some:
 i There were some children in the park.
[non-proportional use of some]
……………………………………………………………………………As for [ii], we are concerned with the reading where some is unstressed, reduced to /səm/ or /sm/. In this case there is no particular larger set of children that I have in mind of whom it would not be true that I saw them all climb over the fence. Again, then, I’m not implicating that I didn’t see all of a certain set of children climb over the fence. Some conveys “not all” only when it is interpreted proportionally, i.e. when there is a certain set involved such that the issue arises as to whether all members of that set have the predication property. …………………………………….But it isn’t only in partitives that some has a proportional interpretation. Consider:
 i Some people misunderstood the question.
[proportional use of some]
In [i] there is an implicit set of people who were asked the question (perhaps the candidates in an examination, perhaps the voters in a referendum, and so on), so some is interpreted proportionally: it contrasts with all and implicates “not all”. In [ii] I am talking about people in general, but people in general constitute a set, so again we have the “not all” implicature.
On page 380 we find:
(z) Basic non-proportional use, selecting plural and non-count heads
 i There are some letters for you. [plural]
ii We need some sugar. [non-count]
…………………………………………………..In this use some is non-proportional: we are not concerned with a subset of letters belonging to a certain larger set. There is accordingly no “not all” implicature, ………………………………………..This some is normally unstressed and pronounced /səm/.
On page 381:
(e) Basic proportional use
 i Some people left early. [plural]
ii I think some candidate expressed a view on this issue. [count singular]
Iii Some cheese is made from goat’s milk. [non-count]
Here, in contrast most directly to use (a) [Basic non-proportional use, selecting plural and non-count heads], we are concerned with quantity relative to some larger set, so that there is a clear “not all” (and indeed “not most”) implicature: “Not everyone left early, most people didn’t leave early”, and so on. As with use (a), any would generally be used in negative declarative contexts: I don’t think any people left early. In this use, however, some is stressed, and not reducible to /səm/.
(a) Non-affirmative anyn
Anyn has essentially the same sense as some in its basic non-proportional and proportional uses, but is restricted to non-affirmative contexts – prototypically either negative declaratives or else interrogatives.
The non-proportional use is seen in:
 i There aren’t any letters for you. [plural]
ii We don’t need any sugar. [count singular]
iii I haven’t got any job lined up for you today, I’m afraid. [non-count]
And on page 382:
The proportional use of anyn is illustrated in:
ii I don’t think any candidate expressed a view on this issue. [count singular]
iii I don’t think any cheese is made from goat’s milk. [non-count]
This is yet another example of how abstract analysis cannot convey the meaning of some and any and how they are used; and, indeed, in the CGEL’s authors’ insistence on differentiating between proportional and non-proportional some they reveal a basic misunderstanding. Of  ii (above) they say “there is no particular larger set of children that I have in mind …” But of  ii (above) they say “I am talking about people in general, but people in general constitute a set…”. But if this is the case, then children in  i and ii are a subset of children in general, friends in We are having some friends round ( ii b, p.365) are a subset of friends in general – or at least of the set of our friends - and letters and sugar in There are some letters for you and We need some sugar ( i and ii, p.380) constitute subsets of letters and sugar in general, and so some is ‘proportional’ in these sentences, although the CGEL claims it is non-proportional. There is in fact, in its most common use, no such thing as non-proportional some. The very essence of some is the idea of proportion, a portion, a part.
As regards any, the idea of proportionality and non-proportionality is surely irrelevant?
The CGEL combines its confusion concerning proportionality, as well as its questionable distinction of two sorts of any’s, with confusion about stress. It declares that its ‘free choice’ any is always stressed (see above on page 361 and page 383). On page 381 (see above) it asserts that ‘proportional’ some, as in Some people left early, is stressed, and not reducible to /səm/.
These statements seem to betray a basic misunderstanding of the nature of stress, caused once more, I suggest, by insisting on basing linguistic analysis on category. Virtually any word can be both stressed and unstressed, irrespective of grammatical category, depending purely on context and the speaker’s meaning.
In all three examples of the authors’ ‘free choice’ any on page 382 the any is unstressed. (Perhaps one should say, rather, not heavily stressed, because there are not just two possibilities where stress is concerned, stressed or unstressed. Stress is on a continuum, from minimum to maximum, and is judged relative to the stress on the words surrounding it. Here I would only stress any relative to computers, policeman and remaining dirt if my interlocutor had contested my assertions about these things.) On the other hand, in a sentence like
Any fool could have told you that
in most contexts the any would be stressed relative to fool; while in a sentence like
Any moisture will quickly evaporate
the any will not be stressed.
In contrast, on the same page 382 (see above) there are three examples of their ‘proportional’ any where they imply that the any’s are unstressed, when in fact they are stressed. The any’s would only be unstressed if they appeared in sentences like
I don’t think any people left early, but some pigs did or
I don’t think any butter is made from goat’s milk – cheese yes, but not butter.
In the sentence Some /sʌm/ people left early it is clear that the speaker is talking about a party, or a meeting, or perhaps a football match; in any case, it is a sentence where it is natural to stress some and not people, because people is, as it were, a default concept, so there is no reason to emphasize it. But one could add
but some /səm/ ladies stayed on to the very end,
where ladies would be stressed and some would be unstressed /səm/.
On the other hand, if people is not a ‘default’ concept, as in, for instance,
The Martians all stayed to the end of the party, but some /səm/ people left early,
people will be stressed. Similarly, one might say
No butter is made from goat’s milk, but some /səm/ cheese is.
Again, in contrast to Some people left early, consider:
Some /səm/ people came out of the house.
Here people is not a ‘default’ concept; the listener needs to be told that it was people, not dogs or cats, or anything else, that came out of the house.
As further illustrations of the possibilities in the variation of stress, consider sentences such as:
They made some /səm/ mistakes, but they weren’t important ones.
They made some /sʌm/ mistakes, but not many.
Or a conversation such as:
“Next we’ll need some /səm/ sugar.” “I don’t think we need any sugar.” “No, I’m sure we’ll need some /sʌm/ sugar - not much, but some /sʌm/.”
(Note that in all these examples, some, whether stressed or unstressed, always has the sense of “part of”, whether it is part of all the sugar or all the people in the world, or just part of the people at the party, and so on.)
There are other flaws and mistakes in the CGEL’s account of some and any.
CGEL page 380
 i It was some years before she saw him again.
ii We discussed the problem at some length.
Here some cannot be phonologically reduced to /səm/, and there is no replacement by any in negative contexts.
On the contrary, any is perfectly normal in a negative context:
We didn’t discuss the problem at any length.
Still on page 380, the CGEL is right to describe the use of some in the three sentences below as “vague”:
(c) Vague count singular use
 i When I arrived, some student was waiting outside the door.
ii Some idiot must have left the oven on!
iii Some day I will win the lottery.
But wrong to restrict it to “count singular use”. Some can be used in this vague sense with plural count and non-count nouns as well:
Some idiots have left their cars blocking the exit, so nobody can get out.
“Why do you have to collect your visa in person?” “Oh, some bureaucratic nonsense.”
Some is likewise excluded from generic NPs. Only the [a] examples in  allow generic interpretations:
 i a. [A lion] is a ferocious beast. b. [Some lion] is a ferocious beast.
ii a. [Lions] are ferocious beasts. b. [Some lions] are ferocious beasts.
iii a. [Sulphuric acid] is a dangerous b. *[Some sulphuric acid] is a dangerous
This is a revealing statement. It demonstrates the pointlessness – and misdirection - of so much of the CGEL’s analysis. It is surely completely unnecessary to point out that some is not used to make generic statements? The essential meaning of some is the opposite of the generic. To make a point of it is rather like, for instance, emphasizing that red objects cannot be described as colourless. I do not believe the CGEL could have made such a point if its authors had recognised that the essence of language is meaning, not grammar.
And the CGEL’s disqualification of sentence iii b. is another perfect example of how grammarians confuse mistakes in life with mistakes in language. Some sulphuric acid is a dangerous substance is linguistically identical with ii b., Some lions are ferocious beasts, which the CGEL allows. It is not an unacceptable sentence; it is a perfectly normal sentence, but almost certainly betrays a misunderstanding of the nature of chemistry- presumably all sulphuric acid is dangerous. On the other hand,
Some cheese is too salty for my wife
is both linguistically and factually correct.
Some and anyn cannot replace a in examples like Jill has a good knowledge of Greek .…
……………………: *Jill doesn’t have any good knowledge of Greek.
Once again the authors fail to recognize the fundamental factor of particular combinations of meanings. Surely both
Jill has some (considerable) knowledge of Greek.
Jill doesn’t have any real knowledge of Greek.
are both perfectly sensible sentences?
CGEL page 411
Explicitly partitive constructions
Here the head is followed by a complement consisting of of+a partitive oblique, an NP which can be plural, non-count, or, under restricted conditions, count singular:
iii a. some of the morning ………………………………… [count singular]
………………………………………………………………………………………………………………………………………………………Count singulars are admissible as partitive obliques only if they can be interpreted as divisible into parts in some relevant way. The clearest cases are those involving time periods, as in [iii]. We can also have, for example, Some of the loaf had mould on it, but not *Some of the car was muddy: we need a noun head, as in Parts of the car were muddy.
The authors are confused again here. Usage is the opposite of what they say. Some is used in this ‘partitive’ way with singular count nouns only if there is no clear division into separate parts. The meaning of some of, in fact, sits uneasily with single things at the best of times. It is nothing to do with ‘needing a noun head’. Indeed, some in the phrase some of is itself ‘noun’-like. Some of sounds most natural – perhaps one should say least unnatural – with single things when the thing in question is rather amorphous. For instance,
Some of their argument is sound, but much of it is irrational.
CGEL page 813
Clausal negation marked by the absolute negators is generally in alternation with verbal negation in which no is replaced by any in the case of [19i] and by ………………….. Compare:
 NON-VERBAL NEGATION VERBAL NEGATION
i a. They showed no remorse. b. They didn’t show any remorse.
There are three restrictions that apply to the alternation between the non-verbal and verbal constructions.
(a) No verbal negation counterpart with negator in clause-initial constituent
Where the negator falls within the subject or an element preceding the subject, there is no direct verbal negation counterpart:
 i a. Nobody knew where Kim was. b. *Anybody didn’t know where Kim was.
ii a. At no stage did she complain. b. *At any stage she didn’t complain.
iii a. ………………………………………………………………………
In the case of [iib/…], the ungrammaticality of the verbal negation construction can be corrected by placing the element in post-verbal position: She didn’t complain at any stage ……………………………
This is an example of trying to solve problems by resorting to assertions of arbitrary grammatical rules. In fact it is once again just a matter of compatible and incompatible meanings.  i b. is simply nonsense (whichever person you like to think of didn’t know where Kim was). As for  ii b., this is a perfectly normal case of meaning being determined by the combination of particular meanings with the meaning of a particular word order.
One day she didn’t complain
is a natural sentence.
*Every day she didn’t complain
is not; it is an incomplete sentence.
Every day she didn’t complain she was given a present
is a natural complete sentence. In the same way,
*At any stage she didn’t complain
is incomplete too. At any stage here is roughly equivalent to whenever.
*Whenever she didn’t complain
is very obviously an incomplete sentence. But
Whenever she didn’t complain they gave her a present
is a natural sentence. So is
She didn’t complain whenever they gave her a present,
but with a different meaning.
*Any days she complained
is an incomplete sentence; something like
Any days she complained, they docked her wages
is needed. But
Some days she complained.
is complete; and so on.
CGEL page 822
Items which prefer negative contexts over positive ones (such as any longer) are negatively-oriented polarity-sensitive items, or NPIs.
[According to the CGEL, the any series of words are ‘NPI’s.]
CGEL page 834
(Here I use bold type to replace the double underlining of the original CGEL text which indicates the item that according to the authors “sanctions” the underlined NPI.)
4.4 Non-affirmative contexts
All negators, whether expressing clausal or subclausal negation, sanction NPIs [negatively-oriented polarity-sensitive items]:
 i a. Kim didn’t do anything wrong.
b. No one did anything wrong. [clausal negation]
c. Hardly anyone liked it at all.
ii a. He seems not very interested in any of these activities.
b. It was a matter of little consequence for any of us. [subclausal negation]
c. It is unlikely anyone has noticed it yet.
Recall, however, that the negative context begins at the point where the negator is located. An NPI is not sanctioned by a following negator: cf. *Anyone did nothing wrong or * We had given anyone nothing.
Once more the authors seem unaware that the decisive factor is particular combinations of meaning.
*Anyone did nothing wrong [Whichever person you like to think of did nothing wrong]
does not make sense. But
Anyone with a conscience will do nothing wrong.
does, as does
We could have given anyone nothing, and they wouldn’t have complained.
CGEL page 835
Bare infinitival why interrogatives
 a. Why tell them anything about it? b. *Why not tell them anything about it?
It might initially seem surprising that the NPI is permitted in the positive question [a] but excluded from the negative [b]. But again the explanation has to do with the conveyed meaning of the construction.
Although positive, [30a] conveys the negative suggestion that there is no reason to tell them anything about it. The negative meaning of the clause that would express this negative implicature allows naturally for NPIs. With why not interrogatives, on the other hand, there is a positive implicature. The conventional meaning of Why not tell them about it? is to suggest via a rhetorical question that you should tell them about it. The positive sense of the latter makes the NPI in [b] unacceptable.
Again this sort of analysis suggests a misunderstanding of how languages actually work. The authors analyse language into rigid artificial abstract grammatical categories, and then find they have to engage in all sorts of intellectual exercises to justify the complicated edifice they have created. The words they use are significant. They say that this or that grammatical category or item “permits” or “sanctions” or “excludes” some other grammatical category or item, as if language was a system apart, a system of arbitrary immutable laws beyond the control of the people who speak it. I think the nature of language is the opposite of this. It is not a negative system of restrictions. A language is a positive system: a set of meanings out of which sense is made by combining particular meanings sensibly.
Returning to [30a] and [30b], note how selective the authors are in their examples and interpretations. Interpreting Why tell them anything about it? as a negative is purely subjective. To me (no doubt subjectively too) it is a demand for a reason, and therefore solidly ‘positive’. As for [30b], this is in fact a perfectly possible sentence. One can see this easily if one makes the question more realistic by adding a phrase:
Why not tell them anything they want about it? or
“I don’t know what to tell them.” “Why not tell them anything comforting?”
So their ‘NPIs’ are perfectly ’acceptable’ in Why not sentences; as we can see again if we paraphrase [30a].
Why not refrain from telling them anything about it?
The English-Learning and Languages Review and its individual contributors assert their Copyright © on all the material published in it. Nevertheless, the Review gives permission for unlimited reproduction of the piece above, A critical examination of the account of some and any in The Cambridge Grammar of the English Language, or parts of it, on condition that: (1) acknowledgement is made of the source (2) no changes are made except with the Review’s permission (3) no restraint is imposed on further reproduction of the material (4) copies are not sold
The English-Learning and Languages Review Ć HOMEPAGE