The Economist this weekend has a special edition on language-learning; Lane Greene (Language columnist and Spain correspondent) writes:
Language-learning has a bit of a mystique about it. It is one of many people’s regrets never to have done. Some people conclude that they simply don’t have the brain for it, like chess or jazz improvisation.
I’m a language-learner myself—it gives me a kind of pleasure that I assume other people get from chess. And so over the years I have written about various elements of language-learning. Are some brains just too old to master another language? What technological tools are the most helpful? Why does it seem so hard to acquire another accent? Does linguistic talent overlap with the musical kind?
And, most recently, what’s the hardest language to learn? This depends on what language you already speak.[…]
Mapache gave me the heads-up, and I pass it on to you, for those who might be interested.
I don’t want to go all prescriptivist; and I understand what this is trying to say; but I’d call that sentence ungrammatical: ‘done’ needs a complement. “… never to have done so.” or “… done that.” would be OK.
I see the author is ‘Spain Correspondent’. Could there be some interference from a different language?
(Also “… one of many people’s regrets …” is rather garden-pathy: ” For many people, it is one of their regrets that …”)
Does the Economist no longer employ subeds?
That struck me as odd too.
It sounds idiomatic to me, but that may be because my wife is from the East Midlands and uses similar constructions. I speak AE (American English) and her native tongue is BE (British English) and they are mostly—not always—mutually intelligible.
That’s odd, The Spectator had an article on language learning this week, probably trying to beat The Economist. Jim Lawley, The secret to learning a language quickly. I don’t read it, just a daily email newsletter, and when I did eventually find the article, I can’t remember what it said…
It is one of many people’s regrets never to have done.
OK for me, too. I think I’d have put in “so” or “this” at the end, if I was writing for the Economist, but that actually gives the sentence a slightly literary feel for me.
It is one of many people’s regrets never to have done.
One distraction in analysing here is the compounded plurality (“people”, “regrets”). Remove one wing of the plurality and the fault is more exposed. Working just with removal of the “people” plurality (the “regrets” plurality could be removed at this point as well or instead, but let’s limit the complexity):
• It is one of his regrets never to have done.
Dreadful, but remedied with another “it” (or a “that” or a”this”, but see below):
• It is one of his regrets never to have done it.
Clumsy, but we can see the grammatical improvement more clearly by rewording slightly:
• One of his regrets is never to have done it.
• His regret is never to have done it.
I agree with DE that “so” would help (or “that” or “this”, but better to avoid a second pronoun) in the original, but also as a substitute for “it” in those revisions (in which “this” or “that” would also be good”):
• It is one of his regrets never to have done so.
• One of his regrets is never to have done so.
• His regret is never to have done so.
And so on. English is versatile, and we should be versatile in its use.
I don’t think “done so” works for me, because there’s no verb phrase for it to refer back to. It could work if “Language-learning” were changed to “Learning a new language” or sonething.
@AntC:
From a Romance language, highly unlikely. The need for a pronoun is clear there. As a result, I find myself missing it in English too, and I’m relieved I’m in the company of native speakers here. But then I don’t recall a rule stating when omitting it is grammatical or not. I’d say “many people regret they never did” but “many people regret they’ve never done it.”
No idea about potential interference from Basque!
Quite right, Keith. I (along with David E, I guess) was reading the sentence in isolation. A mistake. I therefore revise my revisions. Of the original:
• It is one of many people’s regrets never to have done that.
And of my variants:
• It is one of his regrets never to have done that.
• One of his regrets is never to have done that.
• His regret is never to have done that.
Let’s revisit the whole thoroughly awkward paragraph:
One improvement among a zillion that are possible:
I’d say “many people regret they never did” but “many people regret they’ve never done it.”
But “many people regret they never have” is possible in US English. In UK English it might have to be “many people regret they never have done.”
Keith, neither of those will fit after the nominal in the original: language-learning. Your objection, again.
@Keith I don’t think “done so” works for me, because there’s no verb phrase for it to refer back to. It could work if “Language-learning” were changed to …
I understood “Language-learning” in the immediately preceding sentence to be tantamount to ‘learning a language’ — that is, the gerund exposes it’s a verb-phrase in drag; then the “so” would naturally refer back to it.
@chuchuflete, @DE hmm interesting. My intuitions are BrE (or so I thought) and certainly not AmE. Perhaps I’ve been out of Blighty too long?
@Giacomo [interference] From a Romance language, highly unlikely.
That mixed-up quantifier (“one of” quantifies “regrets”, “many” quantifies “people”; “never” negates/banjaxes the whole quantification) seems to me very unidiomatic: either interference from a language that quantifies differently, or some careless editing in a rush to publish.
(Anyways apologies for distracting from the subject of the ‘special edition’. Since this Lane Greene purports to be some sort of Authority ‘Writing with Style: The Economist Guide’, I think they’re ‘fair game’.)
No apologies needed; distractions always welcome!
along with David E, I guess
Yes. Keith is right.
The interfering language cannot be Kusaal, natural though that might appear, what with Spain being historically part of Africa and all: ba pʋ niŋɛ “they didn’t do” is grammatical with no explicit object, but implies a definite antecedent equivalent to some specific just-mentioned “it”: otherwise you’d have to say ba pʋ niŋi alaa “they didn’t do thus.” Or something even vaguer.
In context, the second sentence of the Economist article strikes me as irredeemably clunky at best, and probably truly ungrammatical. Aliquando bonus dormitat Homerus.
It sounds idiomatic to me, but that may be because my wife is from the East Midlands and uses similar constructions.
Lane Greene was born and grew up in the US, but who knows how his style may have been affected by going to Oxford on a Marshall, writing for the Economist, living in Europe, and speaking several other languages.
@DE, did you ever notice grammatical, semantical or some other sort of influence of Kusaal on your speech?
No.
Nor on local English as spoken by others, from Kusaal specifically; though there are quite a lot of regional features which cross language boundaries and also turn up in local English.
A striking one is the absence of gender: people speaking English use “he/she” in free variation, which can be disconcerting. The use of final “or” to mark yes/no questions is another one. Pretty much every local language does that routinely.
In some respects Kusaal is not too different from English syntactically: both have basic SVO order and are more or less caseless, and otherwise strongly prefer head-final ordering. Both have just two numbers and no grammatical gender, and the verb aspect categories, though marked quite differently, resemble each other more than e.g. Russian. Lots of differences with adjectives and with clause subordination, though.
Noetica, agreed in this context, but I was thinking Giacomo was talking about the sentences in isolation.
I clicked on the link about “just too old” brains, having discovered how hard it is to learn languages over the age of 60, and found the headline “To master a language, start learning it early: New evidence suggests a drop-off in results after the age of 17”.
The article itself starts off less sensationally:
“THOSE who want to learn a foreign language, or want their children to, often feel they are racing against the clock. People seem to get worse at languages as they age. Children often learn their first without any instruction, and can easily become multilingual with the right exposure. But the older people get, the harder it seems to be.”
Then: “Scientists mostly agree that children are better language learners, but do not know why. Some posit biological factors. Is it because young brains have an extreme kind of plasticity? Or, as Steven Pinker, a Harvard psychologist, argues, an instinct for language-learning specifically, which fades as the brain ages and (in evolutionary terms) is no longer needed? Others think children have special environments and incentives, not more conducive brains.”
Unfortunately the rest is not accessible so it’s hard to judge the quality of the article.
Perhaps it is just a matter of “You can’t teach an old dog new tricks”. Has anyone ever been able to explain that?
My own view is that a lot comes down to memory. There is a lot to remember when learning a language — and not just vocabulary. If you can’t remember it you can’t use it. Mongolian makes sense to me on an intellectual level. But I can’t for the life of me access it when I need it.
The it here looks like a dummy subject, but by the end of the sentence is it revealed as an actual pronoun that refers to the preceding sentence!
I don’t think I’ve seen that before. I’m pretty sure I’ve never seen regret meaning “(in)action someone regrets” either.
I think it all boils down to the way long-term memory works: by strengthening synapses so they don’t spontaneously fall apart. That way, when one neuron is excited, the next one along that connection fires as well – the association between two topics has become automatic. Once you have a lot of knowledge, learning that some part of it is wrong and what should replace it is hard because it requires the next neuron not to fire automatically and another to fire instead.
“You can’t teach an old dog new tricks”. Has anyone ever been able to explain that?
There are as many explanations as there are owners, dogs and tricks.
Lets start with the simple fact that the claim is false. I have seen old dogs learn to propel themselves on a wheeled platform with their front legs, when their back legs have become paralyzed. Other such disabled old dogs learn to let themselves be transported in a kiddie cart, without trying to jump out. Their owners show them how it works.
Now let’s move on to motivation. Young dogs seem to learn behavior in part because they want to please their owners. But once it has been established that owner and dog are BFF (mutt buddies), why should the dog bother to learn new ways of pleasing ? It’s the owner who is stuck in his ways if he can’t learn to stop.
When your life or livelihood depends on it, you may find that it’s easier to learn a foreign language than when you expect to acquire the knowledge off the shelf, as if buying a bag of potato chips you don’t need. “Take out these earrings”, said the Queen*, “they tire me”.
*of Pisuerga
Both its are needed; one has swallowed the other. To me, the following is perfectly grammatical:*
Compare “Language-learning? I’ve never done it.”
*although three its is very awkward.
I, when I was reading the OP for the first time (originally reading the primitive post…, that is), read it like DM.
“English is cyclical, so done must refer to it.
many people’s regrets never to have done it is one of many people’s regrets never to have done it is one of many people’s regrets never to have done it is one of many people’s regrets never to have done it is one of many people’s regrets never to have done it is one of many people’s regrets never to have danced with you….
Perhaps I’ve been out of Blighty too long?
Indeed. This construction with complement-less final have done has become idiomatic in the English of England in recent decades. It’s not AFAIK to be found elsewhere, where have done it, have done that, or (as I prefer) have done so remain the ordinary constructions. This does not mean, of course that in the E of E these longer constructions have become unidiomatic, merely that a fourth construction has been added to the existing three.
having discovered how hard it is to learn languages over the age of 60
If there is in fact a cut-off, it’s probably somewhere around there if not later; my father got from “wait, are those letters?” to actually understanding some nontrivial texts in about half a year of the ulpan at age 53. I’m not sure if that actually counts as learning a language (we all know there’s a long way to go yet) but it was definitely quite a bit of progress.
(In practice I suspect it’s less of a cut-off and more of a down-slope. There does seem to be a major drop-off after the age of 5 or so, but I’d be surprised if the cut-off at 17 is real.
…I wonder if I’ll ever be as confident in any third or fourth language as I am in [my second language] English.)
Once you have a lot of knowledge, learning that some part of it is wrong and what should replace it is hard because it requires the next neuron not to fire automatically and another to fire instead.
…this (somewhat) explains why it’s so easy for misconceptions to stick despite active reminders otherwise – if the connection is already in long-term memory, then actively breaking it is quite a lot of work.
I started learning Swedish when I was 46 and Spanish when I was 58. I will of course never reach the proficiencies of people raised with the languages, or even of English which I started learning at 9, but it’s going pretty good if I say so myself.
@JC This construction with complement-less final have done has become idiomatic in the English of England in recent decades.
Do you have a citation for that claim? I’ve not noticed that construction talking to my rellies, who all still live in England. They don’t appear to adopt Anglo-Saxon attitudes when talking to me. I’ve not noticed it reading English-published media/podcasts such as the Guardian or New Statesman or Economist (usually).
Which is why the sentence immediately stood out to me.
Do you have a citation for that claim?
I have, or I do, as the case may be.
Here’s Lynne Murphy (aka Lynneguist), a linguist at the University of Sussex of American origin, who specializes in BrE-AmE differences: “Pro-predicate do and verb phrase ellipsis”. She links to “Pro-predicate do in the English of the Intermountain West”, a formal paper by Marianna di Paolo (American Speech, Vol. 68, No. 4 (Winter, 1993), pp. 339-356), which I don’t have JSTOR access to.
Apparently this construction was used in Middle English, but its modern reappearance was about 1920 in England, and it is now generally accepted there. I’m unclear if it’s generally used in the rest of the British Isles now: comments on the blog post from 2007 suggests it’s specifically English. Unfortunately, some of the comments have become anonymized, but some are by Dearieme and some by an Irish commenter.
Note that this refers only to pro-predicate do in independent clauses: in dependent clauses like That’s something (that) I shouldn’t have done, it is fine in all varieties of English.
its modern reappearance was about 1920 in England, and it is now generally accepted there.
Thank you. And [di Paolo quoting Butters 1983]
I learnt all my English considerably later than 1920. Indeed exclusively in the latter half of C20th. The di Paolo article is 1993 — before I quit Blighty. [But if it says more of relevance, I’m not getting round its paywall.]
Perhaps my reaction is no more than ‘that sentence is a train wreck’, so my pinning the blame on the floating ‘done’ is being over-specific. And I note for @Hat — not greatly noted for BrE sensitivities — it “struck me as odd too”.
Of Murphy’s (1), (2), I would probably prefer (1) (with dangling ‘done’ — the claimed BrE preference), but neither of them do I find ‘jarring’, as her correspondent claimed — certainly nothing like my immediate reaction to the Economist sentence. I don’t feel an urge to append … done so/done that — indeed I’d find that overkill.
I’m less than convinced Murphy is talking about the disfluence at issue:
Greene is originally American. Murphy claims Americans prefer just “no main verb”:
‘Jarring’ for anyone much? Preferable to the dangling ‘done’? [Reminder this is the first two sentences of the article. There’s no previous info for it to be ellipsising.]
[Murphy] The continuation of the verb phrase is just understood from context.
Well, ok: how would I have continued that sentence to supply the “missing bit”? It’s not simply a case of repeating the wording from a previous sentence (as others pointed out above). The sentence is making pro-predicate do work awful hard — that is, if that’s all there is to the construction .
I’d need several seconds to parse that; going from -learning to learned is apparently too hard.
Adding done makes it understandable, but sounds much weirder to me than adding all of done that. Of course that’s what German does: Viele Leute bedauern, das nie getan zu haben.
I learnt all my English considerably later than 1920.
Sure. But it takes time for syntactic innovations like this to spread.
@AntC: I think you are right that that sentence is practically a trainwreck, and leaving auxiliary verbs dangling at the end may actually be a relatively minor part of that. In, “It is one of many people’s regrets…,” the intended meaning seemingly requires interpretating the initial it a dummy subject, but still also taking it somehow to have language-learning as its antecedent. The mental parser for reading English (or mine, at least, but I suspect I am not alone) really resists reading the sentence that way. By the time I get to the end, it’s such a garble that the final dangling verb is almost impossible to slot into a allowed grammatical role. So
is ungrammatical to me, even though as a speaker of American English, it’s normal for me to say
In parallel, it would make sense for a British English speaker to reject
even if they would normally say
“It is a truth universally acknowledged, that a single man in possession of a good fortune, must be in want of a wife.”
1. Balanced high register
2. Emphasis on “truth” and “universally acknowledged”
3. Secondary sexist emphasis on “man” (women do not need “wives”) or irony (more frequently applied to women, who are reckoned incomplete if unmarried)
4. Could be considered unnecessarily long (“Everyone knows/thinks that a rich bachelor needs to marry” would probably do instead).
So I would say part of the unease with the “trainwreck” sentence is the lack of balance, both in style/register and the “dropping of pronoun already doing double duty”
“Everyone knows/thinks that a rich bachelor needs to marry” would probably do instead
If you’re writing a school essay rather than a novel, sure.
Many years ago perusing Burke’s “Reflections on the Revolution in France” this sentence struck me:”Is it then a truth so universally acknowledged that a pure democracy is the only tolerable form into which human society can be thrown..”.
That must predate Pride and Prejudice. Perhaps “a truth universally acknowledged” was a common trope at the time, perhaps Austen who had family connections with sufferers from the French Revolution, had read and internalised it.
Very interesting — thanks for that comparison!
Is it then a truth so universally acknowledged, that a pure democracy is the only tolerable form into which human society can be thrown, …
… that a man is not permitted to hesitate about its merits, without the suspicion of being a friend to tyranny, that is, of being a foe to mankind ? [download from here]
“a truth universally acknowledged”: 1.01 Mghits, entire first page Jane Austen
“a truth so universally acknowledged”: 3.49 kghits, more than half of first page not Burke
kghits
some combination of assimilation and dissimilation would certainly make this cluster easier to pronounce…
კღიწ
I like that.
Everything looks better in mkhedruli.
As AntC and Brett said, the “pro-predicate do” is not the issue with Lane Greene’s “regrets” sentence. (I’m one of the few who didn’t have a problem with it; I guess I subconsciously unpacked “done” to “learned a language”, so I didn’t need a second “it”.)
And the pro-predicate do wasn’t a syntactic innovation of the 20th century; even Butters 1983 didn’t claim that it was — he cites at least one example from Shakespeare — only that it had increased in popularity *in main clauses* to “virtually 100% in cultivated conversational British English”, while it had been established much earlier in subordinate clauses. (This is also paywalled so I can’t see all of it, but the first page is visible out of only six, and I’m looking at a PhD thesis at the Sorbonne (in French) on “DO britannique” that re-examines his claims carefully, and finds some of them wrong; also, short version in English. 1983 was a long time ago, and Butters didn’t have today’s corpora and search tools.)
Already in 1897, the OED’s entry on do included, under “as a substitute for a verb just used”:
with no indication that this was considered unusual by that time. (The revision of 2014 added a note: “In British, Australian, New Zealand, and South African English, do can be used after an auxiliary verb, where North American English generally has the auxiliary alone.”) Butters thought the usage in subordinate clauses had taken off in the early 19th century, and this scolding peeve in an 1819 grammar book may indicate that it was only recently noticeable, since that’s what people tend to peeve at:
(I didn’t bother to mark the excited italics on about half the words there.) There are also examples in Milton and Jonathan Swift, so it’s been at least occasionally used for a very long time; the question is about the relative frequency of using the “done” vs. leaving it off.
If competition advances‥as it has done for several years.
That is example is no good for two reasons: it is in a subordinate clause, and it is pro-verb do rather than pro-predicate do.