The BBC News Magazine, expanding on earlier BBC News coverage of the Tesco "10 items or less" flap (reported on here), passes on more misinformation from various sources about the usage of fewer and less. The piece ("When to use 'fewer' rather than 'less'?") begins inauspiciously:
Tesco is changing its checkout signs after coming under criticism from linguists for using "less" rather than "fewer". But it's not just huge, multinational supermarkets that get confused about this grammatical point.
The grammatical question of fewer versus less has been raising the hackles of plain English speakers for years.
"Plain English speakers" — where does that come from? From, I assume, the primary source that BBC News used for these stories, the Plain English Campaign, which was also the source for the replacement for "10 items or less": "Up to 10 items".
There's a lot to unpack here, including objections from some to "up to" as a replacement for "or less".
(Hat tip to Mark Eli Kalderon.)
First, I doubt very much that actual linguists (either people who study linguistics or people skilled in foreign languages) were prominent in the complaints about Tesco's signage, which surely came from the peevarazzi (a.k.a. peevologists), probably including some Plain English Campaign members.
Possibly some Plain English members consider themselves to be linguists — "prescriptive linguists", of course — and represented themselves as such to Tesco or BBC News, and that's how "linguists" got into the story's lead paragraph. (I'll put off for another occasion a fresh rant on how the opposition between "descriptive linguistics" and "prescriptive linguistics" is pernicious.)
Next question: why is the Plain English Campaign, an organization devoted to furthering clear and concise English (by campaigning "against goggledygook, jargon and misleading public information"), pronouncing on matters of "grammar" (very broadly conceived)? I think I know the answer to this one: because many of its members believe that usages that offend "the rules of grammar" are BY DEFINITION unclear. On this view, "10 items or less" is unclear because a reader (or at least a reader with good grammar) will expect less to refer to amount, while items denotes a collectivity rather than an amount, so the reader will have trouble understanding "10 items or less"; it's unclear BECAUSE it's "incorrect". (I'll get to the correctness issue in a moment.)
But I find it hard to believe that anyone actually has a problem in understanding "10 items or less"; the problem is not in extracting content, but in the reader's (un)willingness to leave aspects of form unremarked.
On to the core of the magazine piece, the question, "But when is fewer linguistically correct, and when is less more accurate?" (Note: "linguistically correct" and "more accurate".) The answer:
Both words are used as comparatives – fewer meaning "a smaller number of", less meaning "a smaller amount or quantity of" according to A&C Black's Good Word Guide.
But confusion stretches back more than 1,000 years, to the time of King Alfred the Great (9th Century), says the Plain English Campaign, and substituting less for fewer is still common in informal speech, especially in the US, it says.
The first part of this answer is a bald assertion about what fewer and less mean. When I ask people about where such bald assertions (about word meanings and the details of syntactic constructions) come from — "How do you know this?", I ask — I get blank incomprehension. They say, "That's just the way it is" or "It just stands to reason" or something similar. At root, this is a simple appeal to authority: someone once told them that this was the rule.
The twist comes when I ask them what the authorities' backing is for their assertions. My interlocutors are baffled by this question, because the authorities' claims are about matters of FACT, which the authorities must have verified by research. We have to place some trust in people who represent themselves as authorities and tell us that the capital of Portugal is (currently) Lisbon, that koalas are marsupials and not at all closely related to bears, and that the value of pi is approximately 3.14. Nobody could possibly verify all of this stuff on their own.
Claims about linguistic usage are claims of fact, too. But there's a lot of reason to be cautious about such claims when they come from people who set themselves up as usage authorities — because these people have almost never done any research on the matter. The issue is: what do educated middle-class speakers do? In particular, what do the "linguistic elite" — educated middle-class (or above) speakers who are practiced writers and speakers — do? If a significant number of the linguistic elite do something, then it counts as standard (even if many others do something else; there's a lot of variation within the standard).
The alternative, which seems to be where the Plain English people are coming from, is to assume that there is a Grammatical Truth, handed down from some even higher authority than mere usage critics. So we get such dismissive comments as this one from Ian Bruton-Simmonds of the Queen's English Society ("dedicated to preserving and improving the beauty and precision of the English language"), as reported in the BBC News Magazine coverage:
Language should not be confused because it weakens it, he says.
"It's common sense – fewer is for numbers of separate items or people, less is for quantities not thought of in numbers: there were fewer people in the shops because there was less money," he says.
He thinks language has been "rotten for a long time" and says the "efficiency" of words has been lost by "PR men such as David Cameron".
[Note 1: the BBC story refers to Bruton-Simmonds as "Bruton-Simmons".]
[Note 2: We last came across Bruton-Simmonds (author of Mend Your English (1992)) on Language Log in David Beaver's hilarious posting "Rant" from 2005.]
[Note 3: I'm hoping, for his sake, that Bruton-Simmonds is being paraphrased rather than quoted here, since "Language should not be confused because it weakens it" is a singularly flat-footed sentence. In the right circumstances, you can get away with two it's in close proximity, but this one's hard to work out. The first it is summative -- meaning 'the confusing of language' -- which many sticklers would object to on principle, as unacceptably vague. (The second it is anaphoric, with language as its antecedent.) Then there's the woolliness of "language should not be confused" -- what would it mean to confuse LANGUAGE? -- and of "weakens it" -- what would it mean to weaken LANGUAGE? (I know, I know, it's just like corrupting language, only not quite so bad).]
The assumption here seems to be that the meaning of a word is fixed for all time (though the BBC magazine site, probably echoing Plain English, does concede that "English grammar, like all other aspects of language, changes through time" — just not in this respect, apparently). It was wrong a thousand years ago, and it's still wrong. End of story.
References to CONFUSION are all over usage complaints like these. The idea seems to be that people aim for some (correct) usage, but then (inadvertently) fix on some related usage, and produce the second out of "confusion". In fact, in these contexts "S confuses X with/and Y" seems to mean no more than 'S uses X where Y is what the "rules of grammar" insist on instead'.
In the case of fewer/less, the original innovation was to extend less to expressions of number, on analogy with more, which is used with both C (count) nouns (more bushes) and M (mass) nouns (more shrubbery), thus serving in place of both C many-er and M much-er. That analogy is essentially as old as the English language. After a certain point, people who used less with C nouns were simply reproducing a piece of the linguistic systems used by (some) people around them; they weren't confusing anything with anything else.
Now to a subtlety. The BBC magazine story quotes Marie Clair from Plain English:
Phrases like "10 items or less" or "up to 50% discount" are retail speak and it would be much better to use language "from people on the streets," she says. Indeed, Tesco is not alone in committing this grammatical faux pas in public – the Good Word Guide notes that a Post Office advertisement in the Guardian stated: "Please remember, on Tuesdays and Thursdays there are less queues in the afternoon."
The reference to "retail speak" is silly; though expressions of the form "n Ns or less" (for some number n and some count noun N) do occur in retail contexts, they occur elsewhere as well. Clair is probably just assuming that usages she disapproves of come from tainted contexts, like advertising, where she notices them.
Clair follows with a conflation of different uses of less, as if they were the same thing, as if people who use less where it is proscribed are just substituting the "incorrect" less for the "correct" fewer across the board.
Now there are people who seem to do this, and invariably use less rather than fewer in examples like
(1) Less people came to the party than we expected.
But there are also plenty of people who are variable on the matter, using both (1) and
(2) Fewer people came to the party than we expected.
(perhaps tending to use fewer plus plural C noun, as in (2), in more formal contexts).
And there are many people who don't use less plus plural C noun, as in (1), but do use less, quite happily, with numerically quantified plural C nouns, as in
(3) Less than ten people came to the party.
(or, of course, in "10 items or less"). I am one of those people. In fact, I prefer less to fewer in examples like (3), though I can use either; fewer strikes me as on the stiff and formal side.
The result of all this variation is a disparity in the distribution of less and fewer in different contexts. As Mark Liberman noted in his posting, fewer strongly predominates over less with plain plural C nouns (fewer people), but less predominates over fewer with numerical quantifiers (less than ten people).
Back in the BBC magazine story, we learn that
The Plain English Campaign has a simple rule of thumb to help everyone: less means "not as much," whereas fewer means "not as many".
If the Plain English people had looked at actual educated usage, they would have seen something more complex and interesting. But having this simple "rule of thumb" available prevented them from even asking the question, "What are educated people doing with these words?"
In (partial) defense of Plain English, I should point out that they are not entirely unsubtle:
But it can be tricky when referring to quantities, says Marie Clair from the Plain English Campaign. For example, we say less than six weeks, not fewer than six weeks, because we are not referring to six individual weeks, but to a single period of time lasting six weeks.
[Well, the BBC magazine attributed this observation to Clair, but the Telegraph story (hat tip to Victor Steinbok) on Tesco signage has almost the same quote, but with a different (and vaguer) attribution:
Guidance from Oxford University Press says: "Less means 'not as much'. Fewer means 'not as many'. This can be tricky when referring to quantities. For example, we say less than six weeks, not fewer than six weeks, because we are not referring to six individual weeks, but to a single period of time lasting six weeks."
It turns out that the source is OUP's AskOxford page on less vs. fewer. On AskOxford, unidentified "experts" provide brief answers to frequently asked questions.]
This observation touches on a deeper basis (deeper than the analogy between less and more) for the long-ago spread of less into fewer's territory: the close relationship between (singular) M nouns and plural C nouns. Both can denote collectivities of individuals and both can denote expanses. Sometimes there are alternative C and M lexical items with very similar denotations; C bushes/shrubs vs. M shrubbery is my standard example. You can count either ("Count the bushes/shrubbery in front of the house") and you can treat either as an expanse ("The stretch of bushes/shrubbery in front of the house is amazingly large").
Since there's a considerable overlap between the denotations of singular M and plural C nouns, it's no surprise that they should select similar modifiers. Six weeks is easily understood as denoting an expanse, rather than simply an enumeration of weeks, so the modifier less than makes sense. But ten people can also be understood as denoting a collectivity, rather than simply an enumeration of people, so the modifier less than makes sense here also. Note that with actual collective nouns (which are C), less is the only available modifier in this domain: "Less/*Fewer than the full committee would be insufficient to pass a resolution." (If you're a British speaker with collective plural agreement, maybe "Fewer than the full committee are insufficient to pass a resolution" might do for you, but I'm not competent to judge such examples.)
A further note: the Plain English formulation — less for 'not as much', fewer for 'not as many' — comes closer to getting things right that the paraphrase they offer: fewer for "items that can be counted individually". I've made this point many times before, but it's still important: the crucial distinction is between nouns that are by default M nouns and those that are by default C nouns, and though the C/M classification is closely aligned to the semantics of nouns, it is itself a grammatical distinction, showing up in different syntax for the two types (for instance, in the contrast between much shrubbery and many bushes).
Finally, the BBC magazine invited its readers to
TELL US WHAT RULES ANNOY YOU
What rules of grammar confuse or irritate you?
Now, the magazine asked for "examples of confusing grammar rules", but readers (unsurpisingly) tended to supply pet peeves (your for you're; paninis or paninos, rather than the correct Italian panini, as the plural of English panino). These were from the first four responses (twenty more responses have now been posted here — hat tip to Randy Elzinga), which also included a citation of an anomaly in the spelling of English past tenses (lead – led, but read – read) and a complaint about up to as a replacement for or less.
This last comment questioned whether "up to 10 items" allows for 10 items (an inclusive understanding) or only for 9 or less (an exclusive understanding). The Telegraph story raised a similar question, as did several correspondents in e-mail, among them Rick Sprague, who offered a comparison:
Compare with the phrase "up to a point", where the stated point is the minimal case negating the proposition.
It now seems that some people treat up to as having only the exclusive understanding, that (many) others treat it as having only the inclusive understanding (NOAD2: 'indicating a maximum amount'), and that some treat it as ambiguous. So: not a great choice for the Tesco signs. Either "(a) maximum of 10 items" or (b) "no more than 10 items" might do, if "10 items or less" is unacceptable because some people are so passionately opposed to it and "10 items or fewer" sounds too formal to others.