Friday 29 August 2008

On a top ten of endangered languages

The Guardian does some interesting 'top tens', but none more so than the one which has just appeared.
Peter Austin, who directs the endangered languages documentation project at the School of Oriental and African Studies in London, has just compiled a personal selection. An impossible task, in some ways, given that he had to select from some 3000 candidates. But it makes fascinating reading.

Wednesday 27 August 2008

On nominalisations

A correspont writes to say he is having trouble with nominalisations. He cites a style guide which advises its readers to shun them, turn them into verbs, and find an appropriate subject for the sentence. He comments: 'In my view, however, this is easier said than done. And to make matters worse, there seems to be instances where nominalisations are useful, particularly in academic writing.'

Nominalisation is the result of forming a noun from a word belonging to another word-class, e.g. writing from write. It's been a feature of English from its very beginning, in Anglo-Saxon times, so any general rule about 'shunning' nominalizations has to be absurd. What the style guides are usually getting at is the overuse of two processes: (a) long words formed with a suffix such as -ation - as in nominalisation, indeed, from nominalise; and (b) sentences where a noun phrase derives from a finite clause, as in the rejection of the proposal, instead of X rejected the proposal.

Nominalisations allow us the option of being more abstract and impersonal, which is why they are useful in academic writing. Note the problem in (b) above: we have to choose a subject for the clause, and it isn't obvious which subject to go for. Who actually rejected the proposal? And, in any case, is it relevant to know who rejected it? The important point is that it was rejected. The nominalisation allows this focus on the result without distraction.

The antipathy to abstract words is a feature of 20th-century style pundits. George Orwell inveighed against them (despite using them all over the place). So did Ernest Gowers. In a section (in Plain Words) called 'the lure of the abstract word' he comments that avoiding nominalisations 'is more important than any other single thing if you would convert a flabby style into a crisp one'. And certainly, the overuse of such forms can be turgid, as his examples show: 'The actualisation of the emotivation of the forces...', 'a mutuality of capability...', and so on.

But overuse is not the same as use. And no-one can avoid using nominalisations. A few lines before the above, Gowers himself writes about 'an excessive reliance on the noun at the expense of the verb', and there are dozens of nominalisations in his pages. The crucial word is 'excessive'. Excessive use of anything is always stylistically dangerous.

Style guides always simplify, often to the point of pastiche, and that is what has happened here. An originally sensible point - the need to avoid unnecessary abstraction, which often hides unclear thinking - has been generalised into an outright ban. I can't give a guide about when to use or not use nominalisations in a blog (as my correspondent also asks) - that would be a huge task. But I can draw attention to the gradience that exists between nouns and verbs - or, more precisely, between deverbal nouns via verbal nouns to participles - where it's fascinating to see the range of nuances of expression which English provides. It is one of the hidden gems in the big Quirk grammar (§17.54), and it goes like this, with glosses given underneath each sentence:

(1) some paintings of Brown's
(a) 'some paintings that Brown owns'
(b) 'some paintings painted by Brown'
(2) Brown's paintings of his daughter
(a) 'paintings depicting his daughter and painted by him'
(b) 'paintings depicting his daughter and painted by someone else but owned by him'
(3) The painting of Brown is as skilful as that of Gainsborough
(a) 'Brown's mode of painting'
(b) 'Brown's action of painting'
(4) Brown's deft painting of his daughter is a delight to watch
'It is a delight to watch while Brown deftly paints his daughter'
(5) Brown's deftly painting his daughter is a delight to watch
= 3a or 4
(6) I dislike Brown's painting his daughter
'I dislike the fact that Brown does it'
'I dislike the way that Brown does it'
(7) I dislike Brown painting his daughter (when she ought to be at school)
= 6a
(8) I watched Brown painting his daughter
'I watched Brown as he painted his daughter'
'I watched the process of Brown painting his daughter'
(9) Brown deftly painting his daughter is a delight to watch
= 3b or 4
(10) Painting his daughter, Brown noticed that his hand was shaking
'While he was painting his daughter...'
(11) Brown painting his daughter that day, I decided to go for a walk
'Since Brown was painting ...'
(12) The man paintng the girl is Brown
'The man who is painting...'
(13) The silently painting man is Brown
'The man who is silently painting'
(14) Brown is painting his daughter

Style guides should be explaining to people what English allows us to say and write, and pointing out the strengths and weaknesses of different usages in different contexts. Blanket bans are a nonsense.

On avoiding Oz

A correspondent writes to ask whether the conjunctions because, since, as, and for are interchangeable - as in She was late for class because/since/as/for her alarm clock was broken. And he adds: 'The reason I ask is that I would like to be able to use any one of them to connect clauses rather than being limited to because, which might create wizard-of-oz prose.' (He's recalling the words in the song - 'because, because, because...'.)

They can all be used, indeed, but there are frequency, stylistic, and contextual differences. Because ('cos) is by far the commonest. In one corpus study, the percentage ratio between the four conjunctions was: 77 (because), 12 (for), 5 (as) and 6 (since). For is used in more formal style, and is very rare in speech. As and since are much more common in writing, but are often avoided because of the occasional ambiguity with their temporal meaning.

As they were near the window, they heard noises
Since they left, there have been many arguments.

Corpus studies also show positional variation. All clauses are more commonly found in final position (i.e. after the main clause) - because clauses overwhelmingly so, especially in speech. As clauses tend to be more evenly balanced between initial and final position.

Note also that, if stylistic variation is the aim, English provides several other alternatives for expressing reason, such as:

With the bad weather coming, I've decided to close the store. (= Because the bad weather is coming...)
What with the bad weather coming...
Seeing (that) the bad weather is coming...

And there are more complex (and often more formal) constructions such as:

Inasmuch as the bad weather is coming...
By virtue of the bad weather...
In view of the bad weather...
On account of the bad weather...
Owing to the bad weather...
Due to the bad weather...

Plenty of choice, then, for the aspiring stylist.

Tuesday 26 August 2008

On Team GB

A correspondent wonders about the 'unusual' grammar of the phrase Team GB, as used for the Olympics team. It reminds him of a North London pub in the 1970s called Pub Lotus (where it seems all the chairs were replica bucket seats from Lotus sports cars) and Canadian usages such as Health Canada. He adds: 'Here in the UK we seem to be in a transitional stage, with English Heritage coexisting with Sport England. It's not unlike the pattern seen in stock phrases taken from Norman French, e.g. court martial, battle royal - the difference being that in the modern pattern nouns, rather than adjectives, are used in apposition.'

But that difference is important, making the construction very different from the case of postposed adjectives. How unusual is it, in fact? If we construe it appositionally as 'the team which is, more specifically, GB', then the construction is not very different from the many cases of restrictive apposition which have been in English for a long time:

Mount Everest (= the mountain that is Everest)
Lake Windemere (= the lake that is Windermere)
River Thames (= the river that is Thames)
County Cork (= the county that is Cork)
Queen Elizabeth (= the Queen who is Elizabeth)
Dr Brown (= the doctor who is Brown)
architect Jim Smith (= the architect who is Jim Smith)
the number six (= the number that is six)
the year 2009 (= the year that is 2009)
my brother Fred (= my brother who is Fred)
Platform 3 (= the platform that is 3)
and so on.

But Team GB does have a certain rhetorical punch, which comes, I think, not only from the reversed word order but also from the omission of any determiner: compare the GB Team. This makes it like Operation Desert Storm, Hurricane Katrina, Eggs Benedict, and so on, as well as such dramatic names as Mission Impossible. Constructions such as Health Canada are interesting because of their use of an uncountable noun as the first element. And part of the effect of Team GB may derive from the fact that its first element is a collective noun. A few other collectives work in the same way, e.g. Club Med, Department 2, Generation X

Is the usage likely to extend beyond the present top-level institutional senses? Will we get Team England, Team Chelsea, and so on? It would be good to collect a few more examples of the current fashionable trend.

Saturday 9 August 2008

On being linguistically defeated

Someone (presumably a daily blogger) with a sharp eye has noticed the absence of posts on this site during most of July, and wondered why. I've talked about this before (see an earlier post on bloglessness). This is a reactive blog, and July seems not to be a good month for raising questions about language - presumably the beach beckons. Also, when I'm travelling, blogging goes into abeyance.

But I have another reason, this time, which actually brought to light a point of (to me) fresh linguistic interest. A 3-week visit from Mateo, an energetic 3-year-old grandson, destroyed any chance of doing anything by way of serious writing or thinking, but it didn't give me a linguistic holiday, because this young man is trilingual. His mother is English, father Venezuelan, and they live in the Netherlands, where he attends a daily creche. So he is learning all three languages at once. I have, incidentally, as a result discovered the depressing side of child multilingualism: this is the fact that the child assumes you know the same languages as he does! Now Spanish I can handle, but the amount of Dutch I know would fill a thimble. And it is mildly embarrassing for a linguist (of all people) to be at such a loss when this scrap informs you, in 3-yr-old fluent Dutch, about the state of the world, and you have to beg him to translate. Which he does! While looking at you in a pitying way.

This has been my first close encounter with child trilingualism. Trilingual children are by no means unusual, of course. Some people estimate that maybe a third of the kids in the world grow up trilingually. And certainly it's the normal human condition to be bilingual. But it's one thing reading about trilingualism in books and articles, and quite another to hear it around you in daily practice. As a result, I heard something I'd not noticed before.

Three is an age where monolingual children first display serious 'normal non-fluency' (as the speech pathology world calls it). This is a phenomenon which sometimes causes parents anxiety, because with its pauses and repeated attempts at words it sounds like stammering, but in fact it's nothing like stammering at all. In particular, it lacks the tension one associates with that condition. What the child is doing is processing more complex language (notably, coordinate and subordinate clauses), and needing extra time to do it. So we hear such narratives as 'Daddy went in the garden and he - and he - and he - and he did kick the big ball'. There might be a dozen or more repetitions before the child sorts out what is needed to make a successful coordinate clause.

I've talked about all this before, in several clinical linguistic books and articles, but one thing I'd never thought of was the way normal non-fluency would be a sign of code-switching at this age. Mateo is at the stage now where he is realizing he speaks different languages. He has learnt the names 'English', 'Spanish', and 'Dutch', and is using them appropriately. Evidence? When watching Handy Manny on Play Disney - a repairman who switches between Spanish and English - Mateo shouts out 'Spanish' whenever he hears some Spanish words. And on the way back from the beach one day, as we passed a boy with a big bike, he looked at it, then at me, and said 'bicycle'. I didn't know he knew that word, so I must have appeared to be taken aback, because he then said - as if I hadn't understood - 'bicicleta' (the vocative 'prat' was in his intonation). He then added, for my benefit, 'Spanish'. 'I don't suppose you know it in Dutch as well?', I said, in a sceptical tone. He made a noise which sounded like a rude dismissal, so I queried it, and he said 'bike'. His mother told me later that he had probably said fiets, and that bike was common as a loan word in everyday Dutch. All this in a kid who's been on this earth for only just over a thousand days.

For single-word sentences, there was no non-fluency. But when attempting longer sentences, there was. A typical situation was where he would start a sentence in English, then half-way through realize that the word he needed was only available to him in Dutch. Previously he would have simply said the Dutch word without pause, in a mixing phenomenon well recognized in child bilingualism. But now, aware that this was a different language, his developing pragmatic sense of appropriateness made him pause - but (as is typical of normal non-fluency) he kept on jerkily talking. Occasionally he found the English word himself, but often he fell back on the Dutch one, or simply looked at me appealingly for help - which I was usually able to provide, as the context made it clear which word he was looking for.

Normal non-fluency doesn't last for very long - usually, less than six months. So you have to be in the right place at the right time to experience it - which is presumably why it has had so little study in the language acquisition literature. What I don't know is whether, in a multilingual context, it takes longer to resolve. We'll see.

Friday 8 August 2008

On commas, again

A correspondent writes to ask if there is a rule for the use of commas before end-placed optional adverbials.

The 'optional' is important. If an adverbial is obligatorily required to complete the sense of the verb, then there can't be a comma. An adverbial is obligatory after put, for instance, so it has to be:

I put the book on the table.

and never

I put the book, on the table.

Similarly, if there is a clear semantic contrast involved, the comma is needed to make a distinction:

He walked naturally. = He walked in a natural manner.
He walked, naturally. = Of course he walked.

With optional adverbials, it would be unusual to see commas before single adverbs, adverb phrases, and short adverb clauses:

I stopped reading the book immediately.
I stopped reading the book at three o'clock.
I stopped reading the book when the clock struck three.

Inserting a comma in such cases would be to add a dramatic reading which would have to be justified by the context, such as - in this example - conveying an 'afterthought' impression:

I stopped reading the book, immediately.
= I stopped reading the book - immediately.

With clauses, because they are semantically and grammatically more self-contained, there is a greater likelihood of a comma intervening. However, this is only likely to happen if the adjacent clauses are relatively lengthy (motivating a pause and thus a comma) or if there is no close semantic link between adverbial and verb. The following pair of sentences illustrate the first issue:

Most people took up new jobs after the war was over.
Thousands of people of all ranks and ages took up new jobs in a wide range of professions, after the peace negotiations had been brought to a satisfactory conclusion.

The comma is not obligatory, but inserting one certainly helps the reading process, whether internally or reading aloud, and it is thus widely encountered in such contexts.

In that pair of sentences, there is no strong semantic link between the verb phrase took up and the adverbial. By contrast, the earlier sentence I stopped reading the book when the clock struck three does have a strong link. The temporal meanings of stopped and when form a strong semantic bond and make a comma unlikely.

The main places where there is comma uncertainty arise when there is a clash between these two criteria. A lengthy construction would motivate a comma, whereas a strong semantic link would not. This sentence illustrates:

I stopped reading the book about how to carry out an analysis of commas in a wide range of languages(,) when I realized that it wasn't going to reach any satisfactory conclusion.

Style guides vary in their advice, in such cases. Most newspapers underpunctuate.

Writers with a strong sense of auditory style are much more likely to use commas, to point the way they want their sentences to be heard. I believe this is a major feature of my own style. Indeed, it can be seen just now (in my to point clause) and a few sentences ago (where my whereas clause was separated by a comma). Grammatically the commas are unnecessary, in these cases, but they represent the way I want the sentences to be internally heard. The issue becomes a matter of aesthetics, now, and so not everyone will like it. Indeed, a few weeks ago I got a ferocious email from someone complaining about the overuse of commas in my By Hook or By Crook. He found four in one short sentence, he said. Me, overuse commas, in a short sentence? Never, never, never, never, never.

Thursday 7 August 2008

On "can be able to"

I usually begin my posts with the phrase 'A correspondent writes...'. This time it has to be 'A Correspondent writes...' For the query comes via Her Majesty the Queen.

No, it is not the case (I imagine) that Her Majesty reads this blog. What happened was this. A Japanese professor, anxious to check on a question of English grammar in the King James version of the Bible, thought that the best way of deciding the matter was to write directly to the Queen. (I think that's wonderful, that someone should think to do such a thing.) Her Majesty's correspondence officer, observing that 'this is not a matter on which The Queen would comment', forwarded the letter to the English-Speaking Union (the Duke of Edinburgh is its patron), and they in turn passed it on to me. So I guess this is the next best thing to having a royal appointment! Maybe there will be a post of Grammar Laureate one day.

It's an interesting issue. The correspondent had noted an unusual modal auxiliary use in Exodus 10.5, where it is said that locusts will cover the face of the earth so that 'one cannot be able to see the earth'. He suggests - very plausibly, to my mind - that this has to mean, 'it is not possible that one is able to see'. And he comments that the expression is 'unnatural'.

It certainly wasn't unnatural in the 16th century. There are several examples of can/cannot co-occurring with be able to. A couple occur in Shakespeare. In The Winter's Tale (5.2.25), the Second Gentleman observes: 'Such a deal of wonder is broken out within this hour that ballad-makers cannot be able to express it.' And in Antony and Cleopatra (1.4.78), Lepidus says:

Tomorrow, Caesar,
I shall be furnished to inform you rightly
Both what by sea and land I can be able
To front this present time.

It seems that can was being treated just like may, shall. and others, all of which can precede be able to - I may be able to go, They should be able to go, and so on.

As the modal system developed, the association of can with be able to fell away, for two reasons. There was an overlap of meaning between the ability sense of can and that of be able to, which made can be able to seem tautologous. And be able to came to fill the gap of expressing ability whenever a nonfinite construction required it: one says to be able to talk, not to can talk. The two expressions thus complemented each other, and this made it less likely that they would co-occur.

But the usage didn't die away completely, especially in negative expressions. I have heard can't be able to in some regional dialects. Nor did it entirely disappear from Standard English. If you do a Google search for cannot be able to there are 3000 or so examples. Now, admittedly many of these are likely to come from people for whom English is a second or foreign language, and where learning levels are unclear, but there are several cases where the text has been written by native speakers in Standard English. Here is an example, from Articles Base:

'E-mail notifications should be provided to inform users when they are required to perform a task or need to be aware of an event. This should be able to be deactivated but at the same time there must always be some alternative form of notification that cannot be able to be deactivated.'

It would be interesting to hear if the usage is known by readers of this blog.

Monday 4 August 2008

On txtng reactions

Following up my last post: I've been impressed by the media coverage given to Txtng: the Gr8 Db8 over the past three weeks. The interest seems to have reached across the usual publishing divide. One day a full-page review appeared in The Daily Mail; the next day in The Times. This week there were reviews in the New Statesman, New Scientist, and Newsweek and - I kid you not - one column inch in Take A Break. Is this the first time the word linguistics has appeared in that publication? I imagine so!

But it's going to be difficult to dispel the urban myths about texting. Here’s an example of the problem. Txtng came out on 5th July. On the 6th there was a report in Scotland on Sunday headed ‘Professor spreads the word on joy of text’. That sounds good, and the report did summarize quite well the six main points.

- Text messages aren’t full of abbreviations - typically less than ten percent of the words use them.
- These abbreviations aren’t a new language - they’ve been around for decades.
- They aren’t just used by kids - adults of all ages and institutions are the leading texters these days.
- Pupils don’t routinely put them into their school-work or examinations.
- It isn’t a cause of bad spelling: you have to know how to spell before you can text.
- Texting actually improves your literacy, as it gives you more practice in reading and writing.

At the end, the reporter asked for a reaction from the Headteachers’ Association of Scotland. This is what the spokesman said: ‘Because of the rate in which text-speak is taking hold I shudder to think what letters will look like in 10 years’ time.’

The spokesman obviously hadn’t paid any attention at all to the report. The reaction I would hope to see is something along the lines of: ‘It’s reassuring to hear that things aren’t as bad as we thought they were’. Or even: ‘Well let’s explore ways in which we can utilize the potential of texting for improving literacy in our schools’. But no.

I struggle to find an analogy. It’s a bit like someone saying: ‘an aeroplane landed on a motorway a few years ago, and everyone worried about it happening again. It’s a real problem now, and it’s going to be even worse in 10 years’ time.’

To which the answer is: it isn’t a problem, actually. You’re imagining it. Look at the facts before you comment. It’s a risk, certainly, and we need to be alert. But there are no grounds for panicking.

A few years ago, it would have been difficult to say this about texting, because there were no facts. Things have changed now. The research is building up. My book went to press just a few months ago, and already since then I’ve come across further research findings which reaffirm its conclusions. For example, a recent article (in New Scientist for 15 May 2008) reported a study by Sali Tagliamonte and Derek Denis of the University of Toronto which confirmed that abbreviations are far less frequent in electronically mediated communication than people suppose. For every one instance of u, there are nine of you, they found. That’s exactly what I would expect.

It’ll take quite a while to get rid of the myths about texting. The trouble is that they are well established on the Internet. That hoax essay from 2003, in which a pupil was supposed to have bemused her teacher by writing an essay entirely in textisms, is still doing the rounds. Someone sent me a copy just the other day as ‘evidence’ of the terrible state we’re in. If it was a regular happening, or (more to the point) if teachers were letting this happen, we might have cause to worry. But it isn’t. They aren’t. And we shouldn’t.