Tag Archives: Prescriptivism

Future descriptive

7 Aug


To: All editorial staff

From the production editor


Dear all

Several of you have been asking for a definitive style ruling in recent weeks about the now-perennial “cannot be underestimated/cannot be overestimated” debate. I know feelings have run high on the issue, and until now we have tried to preserve the traditional distinction in meaning in our pages, even though the interchangeability between the two phrases in spoken English is now almost total.

Historically, it is true that – as recently as the early 21st century – the correct use of the phrases was highly dependent on context, and to say then that the prime minister’s intellectual capacity “cannot be underestimated”, when the opposite was meant, would have been to cause considerable offence. But the error has now become such a common one that it is time to seriously address the question of whether it is an error at all.

Of course I am aware, as some of you have kindly pointed out, that under and over “mean completely opposite things” and that the distinction is “perfectly obvious to those who are prepared to think about it”. Of course it is, but the everyday rough-and-tumble of language has a way of wearing fine distinctions – even useful ones like these – smooth. Look, for example, at how the similar (and now vanishing) terms “biennial” and “biannual” became so confused in the 1900s that the following definition once appeared in Chambers’s 20th Century Dictionary:

biannual (bi-an’-ū-əl) adj. two-yearly: also half-yearly.

And consider “head over heels” – a phrase universally understood in its metaphorical sense, but which, parsed logically, says the exact opposite of what it means.

I am reluctantly coming to the conclusion that “cannot be over/underestimated” have, through widespread usage, fallen into the same category of phrase as “head over heels”: those that can only be understood in the round, and not by parsing very word individually.

I am aware this decision will disappoint many of you, especially those of you who have pointed me to a significant strand of linguistics scholarship that disagrees with me. Writing in the early 2000s, eminent figures on the influential website Language Log contended against the acceptability of what was then called “misnegation”. Comparing “cannot be underestimated” in relation to the (now-uncontroversial) phrase “could care less”, Professor Mark Liberman wrote:

I’ve argued that “could care less”, where modality and scalar predication seem similarly to point in the wrong direction, has simply become an idiom. Shouldn’t the same be said for “cannot underestimate the importance”?

I don’t think so. As I’ve argued before, there’s a crucial difference.

Whatever is happening with “cannot underestimate” applies equally to “cannot understate”, “impossible to underestimate/understate”, “hard to underestimate/understate”, “difficult to underestimate/understate”, “cannot be underestimated/understated”, “hard to underrate”, “cannot be undervalued”, and many other common ways to re-express the same idea.

In contrast, alternative formulations of “could care less” are rare, and can only be understood as bad jokes, to the extent that they’re not simply puzzling.  Thus one semantic equivalent to “could not care less” might be “could not possibly have less concern” — and we find this in a published translation of Montaigne…

“However, if my descendants have other tastes, I shall have ample means for revenge: for they could not possibly have less concern about me than I shall have about them by that time.”

But in this case, Montaigne means to imply that his concern-meter will be pegged at zero, not at its maximum value. And more generally, we don’t see things like “I could possibly have less concern” used with the meaning idiomatically assigned to “I could care less”. This is the behavior that we expect from an idiom; and the different behavior of “cannot underestimate/understate/
underrate/undervalue” is what we expect from a psychologically probable error.

Other scholars at the time contended that “cannot be under/overestimated” was indeed an idiom; but even if they and I are wrong and it is a mistake, it seems to be a mistake that English-speakers are never going to stop making. And, as we all know to our frustration, appeals to reason over usage rarely succeed in these matters because language doesn’t listen to reason.

Therefore, henceforward,  “should not be underestimated” and “should not be overestimated” shall in all cases be deemed to be equally correct ways of saying the same thing, which is something to the effect of “should not be evaluated incorrectly”. The style guide will be updated accordingly.

Believe me, it gives me no pleasure to come to this conclusion. But our language has changed around us: and with the 22nd century just over a decade away, we have better and more significant things to do with our editorial resources than enforcing a distinction that, to our readers, is increasingly becoming inaudible.

Yours as ever



Production editor, the Tribune


Nation shall prescribe unto nation

11 Jul

‘I’d have gone for “visionnaire” myself. I’m glad we didn’t get “auralooker”:

Historian Nick Kapur’s fascinating Twitter thread about the BBC’s Advisory Committee On Spoken English and its influence on modern speech reveals just how close we came to referring to anticyclones as “halcyons”, but also offers an illuminating insight into what prescription in language really means.

Because of course, there is not one kind of linguistic prescriptivism: there are two. One opposes all language change and all neologism, and attempts to conserve current norms as an eternal standard. But the other seeks to deliberately modify language: not to reject new words, but to invent them, and to influence speech and writing to go in new directions – such as the campaigns to popularise Ms and Mx as neutral  honorifics. It is this second kind of prescriptivism, which one might call activist or progressive prescriptivism, that Kapur is tweeting about here.

The story begins, he relates, in 1926, when Lord Reith sets up a committee to help resolve one of the many problems a pioneer national broadcaster has to solve: how should you pronounce certain words on air? (This group, the Advisory Committee On Spoken English, still exists today, doing very similar work to help BBC broadcasters). Then in 1935, faced with the question of what to call users of the new media of the day – television –  a new sub-committee was set up, not just to advise on pronouncing words, but to invent some new ones. Led by the Anglo-American man of letters Logan Pearsall Smith – an eager language reformer – the Sub-Committee on Words generated the alternatives listed above to start the debate (although it eventually rejected all of them and recommended “televiewer”, subsequently shortened to “viewer”.)

After that, the sub-committee remained active, and widened its remit to mass-produce new words for broadcast far beyond the new industry’s immediate needs, eventually becoming so extravagant and implausible in its inventions that an exasperated chairman of governors closed it down in 1937. But by then it had created several terms – “roundabout” for the road junction, “serviceman” for members of all the armed forces, “art researcher/art historian” to replace the German word “kunstforscher” – that are now commonplace in modern English.

The impression descriptivist scholarship frequently gives is that language is an unknowable stew of errors, localisms, homophone confusions and misreadings, prone to unpredictable change. The emphasis, or the cultural preference, often seems to be bestowed on the unwilled variations to language, not the willed ones. But Kapur reminds us that English is also highly susceptible to the approaches of those who have a design on it, from Edwardian grammarians like Fowler to equalities campaigners to spelling reformers like McCormick at the Chicago Tribune. There are words and conventions in many registers of modern English that were created deliberately by people who wanted to see them catch on and took the opportunity to make it happen.

Sometimes, of course, prescriptivism is institutional, and benefits from that privilege. It might be justifiably argued that the BBC’s committee, as a quasi-official body proposing usage for the nation’s only broadcaster, was in a very strong position to succeed, particularly as it was inventing terms for then-unnamed phenomena. But the Academie Française, which is attempting to do for French today almost exactly what the BBC committee did for English in the 1930s – and from a similarly state-sanctioned position – is greeted with widespread indifference and derision for its efforts.

And in any case, innovative prescription does not need an official platform to succeed. This blog has discussed at length the extent to which Fowler’s suggestions have influenced modern formal and legal English, but Fowler himself was no state official, nor did his books bear any government imprimatur (although Churchill is said to have recommended Modern English Usage to his staff after it came out). His books were a success because, then as now, there is a sustained public appetite for advice on how to engage with formal English. (Indeed, given the existence of a generation of professional linguists who consider it their role to observe rather than advise, the field for such material is possibly clearer today than it was then.)

This is not to say the process is easy: frequently, big innovations just don’t catch on.  There is no doubt that some of the committee’s ideas, like some of Fowler’s, are much worse than others: for example, one member apparently felt it desirable to create a shorter term for “inferiority complex” (“inflex”), and another proposed “yulery” as a collective term for Christmas festivities. The point is not that Fowler or the committee were always “right” about what they proposed; the point is – at least sometimes – that they were successful.

Usage remains the timeless, and the only, judge of current English. But usage does not simply adjudicate on terms that have risen up unbidden from the demos; it also sits in judgment on peri-statal prescriptions and private linguistic entrepreneurialism. Due process is afforded to all new words, whether they are accidents or designs. Linguists say that language is a democracy, and it is: a democracy in which, among other things, anyone is free to prescribe and see what happens.

Invisible mending

8 Nov

“Most writers I know have tales to tell of being mangled by editors,” writes the esteemed academic John Gross,*

“… and naturally it is the flagrant instances they choose to single out – absurdities, outright distortions of meaning, glaring errors. But most of the damage done is a good deal less spectacular. It consists of small changes (usually too boring to describe to anyone else) that flatten a writer’s style, slow down his argument, neutralise his irony; that ruin the rhythm of a sentence or the balance of paragraph; that deaden the tone that makes the music.”

Here at the Tribune, we are a “writer’s paper”: that is to say, we allow our senior writers – and especially our columnists – not just their own opinions, but their own style as well.  Of course, in theory we edit everything perfectly – we intervene whenever it is required, and keep clear whenever it is not – but to the extent there is an institutional bias, it is to be hands-off: not to flatten a style or ruin an argument for the sake of enforcing “good English”. So we are, one would hope, less likely than some of Gross’s targets to “pounce mercilessly on split infinitives … and all the other supposed offenses that are often no offense at all”.

But hands-off editing comes with its own set of hazards. Specifically, it can create a culture of under-intervention: we do basic editing, correcting spellings and checking dates, but perhaps decline to step in when a columnist has mixed a metaphor, or written a sentence so long that it provokes amusement on Twitter. In the worst cases, faced with something notably angry, funny, colloquial or emotional, we can become paralysed: confronted by a confessional tour de force or celebrity stream of consciousness, we freeze, run a spellcheck and send it through without doing the whole job.

So, bearing the countervailing risks in mind, where you would you step in, and where you would you step back, here?

screen-shot-2016-11-07-at-17-40-51 screen-shot-2016-11-07-at-17-40-59 screen-shot-2016-11-07-at-17-41-59

This is Laura Craik’s “Upfront” column in the Evening Standard’s ES magazine. She is a fashion and trends commentator who writes in a  chatty, informal style typical of that genre: even if you don’t know her, that much becomes immediately apparent when you read the copy. The tone and register are easy to grasp, and so are the editing parameters: you instinctively allow “mahoosive”, “yada yada”, the sentence fragments, or “Soz” in a way that you wouldn’t if they cropped up in a Telegraph editorial.

But I’m not so sure about “pontificating”. Given the context (“I say ‘rushed’, but really I’d been pontificating since May”),  I strongly suspect what’s meant is “prevaricating”. Even if the intended sense is something closer to “I’d been talking about it to everyone for months”, “pontificating” still isn’t quite right: it carries the sense of speaking (like a pontiff) from a sense of real or imagined authority, and the whole point of the piece is that the author didn’t know what to do. In a piece where nearly everything should be allowed to stand, this is something that needs to be changed: the one reason in 600 words not to step back and wave the copy through.

Intentional malapropisms are funny. Unintentional ones on the way to making a different kind of joke are just distracting. That’s where the kind of invisible mending that broadsheet subs do comes in. Tone is exclusively the province of the writer – there is a lot of truth in the columnist’s weary complaint that “it’s my column, not yours” – but sense and cogency are the business of the newspaper as a whole, and particularly the copydesk. Making a change like that doesn’t “flatten the writer’s style” but enhances it, by removing a distraction over which a literate reader might trip. Editors shouldn’t do too much, but we usually have to do something.


* Editing and Its Discontents“, in The State of the Language, edited by Christopher Ricks and Leonard Michaels (University of California Press, 1990)

Too chill for comfort

13 Sep

If you were looking for snark, the official Twitter feed of a major American-English dictionary might not be the first place you’d look. But, oh boy.

A few days ago, Gabriel Roth of Slate unwisely allowed his inner prescriptivist out for an airing after reading the following tweet from Merriam-Webster Dictionary:

Articulating the silent twinge that many editors and writers feel at the sight of descriptivism in action, he wrote:


And then unexpectedly this reply, from the dictionary itself, appeared:


Ouch. Owned. Or – to use the correct spelling of the word in this context – “pwned“. As a rueful Roth wrote later, “I find myself wistfully remembering the days when tweeting at brands was a safe, innocuous pastime”. And other responses to M-W’s intervention have been broadly favourable: the tweet was rude, yes, commenters thought, but also uncompromisingly truthful about the ineluctable nature of language change.

However, scrolling down through M-W’s Twitter feed, it emerges that this is not the only time it’s taken a bold line in such matters. Five days earlier, in similarly lively terms, it made the following observation:

Well, hang on. Yes, “enormity” can indeed mean “great size”, and has done for centuries. But, no, it’s not “fine”: currently, as a word, it’s totally skunked. As we discussed last month, “enormity” is hovering uneasily on the brink of a permanent change in meaning, but is still tending to drag its other meaning of “moral horror” into simple discussions about size. It’s a very tricky word to be employing at the moment; a while ago, for example, we saw fit to remove it from a news story about the heated subject of the Scottish referendum because of its overtone of opprobrium. It’s far from clear that, in these circumstances, a major dictionary should be recommending it quite so breezily. Authorities are looked up to; these things get taken seriously.

As this blog has had occasion to remark before, people don’t require help with informal English. They speak it well. They do not seek the assistance of their editor friends when composing a tweet or posting on Instagram; but they do, sometimes, when updating their CV or writing to a solicitor. What they want is help with formal English: a register whose social significance they grasp, but one in which they perceive themselves not to be fluent.

This is when they turn to the dictionary: to be briefed on the meaning of a legal idiom, or the appropriate use of a word in their own reply: to find out, perhaps, whether “enormity” means what they think it means. But they are doing this at a time where one of the prime objectives of linguistics is the debunking of the prescriptive maxims about language that have been taught during last two centuries. An unsatisfactory dialogue has therefore developed between linguists and the public in which queries about the niceties of formal English are met only with assurances about the validity of informal English. For the last several decades, it seems, lexicographers have been talking about what’s changed in the language while their readers have been asking about what hasn’t.

The spirit behind this objective is democratic to a fault, and the efforts to expose the frailties of formal English are intellectually impeccable. But nonetheless, they are starting to amount to the total deconstruction of a dialect that many people still have no choice but to speak.

The ghosts of Fowler, Strunk and White still haunt the sphere of formal discourse. It is highly commendable that more modern authorities like Merriam-Webster should be getting involved in the conversation about usage. But burning a grumpy prescriptivist on Twitter? Waving off debate about a word in difficult transition?  That isn’t advice; it’s advocacy. Roth is right: counsel as blasé as this is just a little too chill for comfort.

The size and the horror

2 Aug

We held out for a long time, but it looks like even our resolve is weakening. Witness this exchange on the subs’ email list last week:

From the reader’s editor:

Can someone please tweak this: [appends link to article]
Style guide:
It might sound a bit like “enormous”, but enormity refers to something monstrous or wicked, such as a massacre, and is not just another word for “big”
 From a sub-editor:
I’ll have a look at this
From another sub-editor:

this is an odd one as our default dictionary Collins actually says it can be used informally to mean “vastness of size or extent

And then, from the website production editor, this:
I think it’s one of those words whose changed meaning is now used widely enough to possibly warrant a style guide tweak.
Have copied in the house style team for their view.
OK, so we haven’t changed anything yet. OK, so we’re just taking views at the moment. But still, compare this willingness to be descriptivist with what we were saying about enormity two years ago, when an article was summarily corrected to remove any suggestion of bias during the Scottish independence campaign:
A front-page analysis of the Scottish independence referendum said: “With only 10 days to go, the rest of Britain finally awoke yesterday to the enormity of what is happening in Scotland.” The style guide states that enormity “refers to something monstrous or wicked, not big”. The writer was, in fact, referring to the scale and importance of the vote (“Nothing else now matters in British politics”, 8 September, page 1).

I thought then, and I think now, that the word is currently best avoided in either sense. It can’t be relied upon to deliver its old meaning, but nor, as witnessed above, have the prejudicial implications of that meaning been completely extinguished. It is well and truly “skunked“, as Bryan Garner would say.

But nonetheless, the direction of travel is obvious: the “immensity” meaning is starting to appear in major dictionaries, and, in the case of Merriam-Webster, as a formal definition of equal status with the others. More than one senior and discriminating Tribune writer is using the word in relation to size without batting an eye, despite what the style guide may say. And although it is unwise to try to prove anything that relies on context with a Google Ngram, compare the usage graph for “enormity of the crime” (i.e. repugnance) with the one for “enormity of the task” (i.e. immensity):

Screen Shot 2016-08-01 at 12.03.16
Screen Shot 2016-08-01 at 12.02.49

Language changes so slowly that we perceive it to be static; we discover with bemusement that “awful” once meant “awe-inspiring” or that “egregious” once meant “eminent”, but we don’t perceive the same shifts to be happening today. Yet they are, and this is a clear example of a word conclusively changing its meaning in front of our eyes. It may still be too early to safely describe a band (as we already have) as “uptempo pop rockers destined for enormity”. But the day is getting closer.


19 Jul

My iPhone is biometrically coded, offers encryption that baffles the FBI, and can connect me to global news and networks that defy the reach of political censorship. But it does want me to capitalise ‘Polaroid’:


Digital freedom means many things, but apparently not the right to appropriate trademarks.

The extent to which registered trademarks enter common language, or “genericisation”, is one of the hard knot of editing issues, along with libel and legal reporting, where editors can really earn their money – high-stakes prescriptivism, so to speak. Lawyers representing dominant companies live in fear of their trademarks becoming nouns or verbs that define an entire market, not just  their clients’ products (see, for example, “googling”, “photoshopping”, “thermos”, “sellotape”, and once, long ago, “aspirin”). Once a word has “entered the language”, the courts are inclined to take that as a fait accompli and deny any further copyright infringement cases; so the lawyers have to act fast and early to prevent genericisation ever happening. They email, write, phone, demand corrections, suggest alternatives (Velcro likes you to say “hook and loop fasteners”). They are language change’s sharpest and best-resourced opponents.

And that’s why the Tribune’s style guide on the issue says the following:

trademarks (TM)
Take care: use a generic alternative unless there is a very good reason not to, eg ballpoint pen, not biro (unless it really is a Biro, in which case it takes a cap B); say photocopy rather than Xerox, etc; you will save our lawyers, and those of Portakabin and various other companies, a lot of time and trouble

The editor’s natural interest in enforcing distinctions chimes well with the lawyers’ determination to have them enforced, even if the legal vigilance gets a little grating at times. Portakabin, notably, used to send round a letter a week before the summer music festival season – long before any transgression had actually taken place – to “remind” editors that their clients did not provide the toilet facilities for Glastonbury, so on no account were festivalgoers to be described as using “Portaloos”. This would reliably cause the kind of grumbling, even among hard-nosed copydesk veterans, that one might almost have described as descriptivist. However, one can see the point: in court cases, one of the determining factors of a word being deemed to have entered the public domain is to what extent it appears in a general sense in media reports.

That said, it’s still not entirely clear what Apple is hoping to achieve with its suggestions on QuickType (as the predictive typing aid in iOS9 is called). As an experiment, I tried out Facebook Messenger with single-word nouns given as trademarks in the Tribune’s style guide.

Whenever it recognised the word (it didn’t in all cases), QuickType invariably suggested an initial cap, with the sole (and slightly baffling) exception of Jacuzzi (trademark of the company founded in 1915 by Giocondo Jacuzzi in Berkeley, California.)

But is that enough? As the style guide suggests, the best practice when trademarked words come up is to change them to a generic alternative – at least, better practice than scattering the copy with ™s and ®s. Is a capital letter, without more, enough to escape accusations of aiding genericisation? Is the fact that your operating system is suggesting a semi-proper noun less likely to annoy the lawyers, or more?

It might be argued that proposing some acknowledgment of copyright is enough to absolve you of blame; certainly better than offering a lowercased suggestion. But the best practice recommended to lawyers is to use the trademark as an adjective not a noun (“Xerox brand copiers”), and QuickType seems happy to suggest nouns. It may be that QuickType is following the dictionary practice of capitalising trademarked nouns when defining them; but dictionaries always quickly make clear that such words are trademarks, and QuickType does not.

In the end, the genericised fate of escalator, kerosene, laundromat and trampoline suggests that the battle – like nearly all battles against language change – may be futile. Certainly, Xerox have fought doughtily against “xeroxing” as a verb and kept their trademark protected: but there are many fewer photocopiers in today’s broadband-connected offices than there used to be, and it may be that the word will die out with the practice. But in the meantime, the letters will keep on arriving. And the lawyers aren’t the amateur grammar grumblers of the letters page: these people peeve for a living.



21 Jun

For a moment, I wasn’t entirely clear what was being corrected here:

Screen Shot 2016-05-22 at 16.18.46

Was this going to be a discussion about “refusenik”? In fact, it soon becomes clear that it isn’t: the quote is merely provided as context to explain why the non-standard noun “copyrighters” has become confused with “copywriters”. But perhaps we do need to talk about “refusenik”; because current usage is moving it rapidly away from what it used to mean.

“-nik”, (“-ник”), as imported from Russian and added to the end of English words, has become a gloriously compact morpheme for creating agent nouns (as well as a way of adding an implied hint or criticism about the subject’s politics): beatnik, peacenik, no-goodnik. But it’s so compact that it actually leaves the precise relationship between the noun and the agent unclear. A beatnik is someone enraptured by the Beat movement; a peacenik is a proud pacifist. Similarly, a “refusenik” is now someone who chooses to boycott something as a protest. But historically, the word meant something very different.

The original term, “отказник” (“otkaznik”), was applied to minorities in Cold War Russia – often Soviet Jews – who were being denied permission to leave the USSR and emigrate (in the Soviet Jews’ case, to Israel; in the case of other minorities, to join diasporas or seek asylum on religious grounds). The policy, which prompted an international humans rights dispute, came to a head for several years in the late 1960s and early 1970s before bans were lifted for a period. So the original refuseniks were not people who had refused to do something; they were people to whom something had been refused; would-be refugees who were being prevented from travelling to any other country. They were victims of refusal, not proponents of it.

In these post-Soviet days, that usage is rarely heard. Many of the major dictionaries still record it, for example, Oxford, American Heritage and Merriam-Webster –

Screen Shot 2016-06-20 at 13.25.33

Screen Shot 2016-05-22 at 14.30.19

Screen Shot 2016-06-20 at 13.28.16

– but Collins already describes the original, political meaning  as the “former” definition:

Screen Shot 2016-06-20 at 13.26.39

and it’s clear that a word that could once have been a synonym for “detainee” is now well on the way to being a synonym for “dissident”.

Usage has radically changed the word’s meaning and laundered it, willy-nilly, of its political gravity and history. One might feel that this has not been language-change’s finest hour, although the compressed ambivalence of “-nik” as a suffix probably made some confusion inevitable. Now, as is the case with “enormity”, the new meaning of  “refusenik” has overwritten the old to such an extent that it’s no longer safe to assume the original usage will be understood.

But it can still be jarring to see the new meaning employed in text. And of course, one doesn’t have to use it; one can probably be a refusenik – I mean a dissident – about that for a while longer.