Tuesday 10 June 2008

On (in)complete predications

A correspondent writes to ask about the difference between 'complete' and 'incomplete' predication, which she and her colleagues have found confusing.

I'm not surprised. It's not a pair of terms I come across much these days. Terms based on 'completeness' are more associated with Victorian grammars, and they rather fell into disrepute when structural linguistics developed. But the underlying concept is still an important part of any modern grammar, even though analyses vary somewhat.

A verb which can stand alone as a predicate (i.e. without any complementation) would be a 'complete' predicate, in these terms. It's gone, I complained, and so on.

An 'incomplete' predicate would be a verb which requires some sort of complementation for the sentence to be grammatical. In the terms I like to use, this would be an object (a transitive verb, e.g. I saw a dog), an adverbial (e.g. They kept out of trouble) or a complement (in a narrow sense, e.g. I am ready).

Notice that the class a verb belongs to depends on its meaning: I kept a cat is different from I kept out of trouble. Similarly, we need to distinguish between I'm eating and I'm eating a cake. In other words, we're talking here about different uses of verbs - a more helpful way of thinking, to my mind, than to set up an absolute contrast, such as is implied by 'complete' vs 'incomplete'.

A Comprehensive Grammar of the English Language (§§16.20ff) has an excellent listing of verbs in different complementation functions.

On @ signs

A correspondent writes to ask what is the 'official' name for the @ sign, if there is one.

Well, I guess it all depends on what one means by 'official'. What is official in one context may not be so in another - and that is the case with this sign. Traditionally, it was officially referred to as 'commercial at', because of its origins in commercial accounting; and that is the way it's described in the Unicode system and in the Telecoms standardization system (what used to be called CCITT). On the other hand, the majority usage today is in computing contexts, so that term seems a bit antiquated.

Modern terms whch have been proposed include 'asperand' and 'ampersat', but neither has more than a few thousand hits on Google. 'Amphora' is another suggestion, derived from the medieval symbol used as a measure of quantity (an amphora was a kind of jar).

Personally, I think the best technical term in English for the at sign is 'at sign'.

Note that other languages have opted for different solutions, usually based on the shape of the symbol, such as Dutch apenstaartje ('little monkey-tail'), Hungarian kukac ('worm, maggot'), and so on. There are several lists available on the Web (if you type in 'at sign' into a search engine).

Sunday 8 June 2008

On /r/s and ofs

A correspondent - presumably from the UK - writes to say she is 'fascinated by the American habit (inability?) to say mirror, terror, etc as we do in two discrete syllables. mir-r, ter-r, is what we hear.' Why do they do it?, she asks. She offers four explanations:
'1. A desire to be stylish or a reluctance to be too correct/too English?
2. Their frequent desire to speed speech up, as in giving a year as Two thousand eight instead of Two thousand and eight?
3. A form of shyness, like saying duiper instead of nappy?
4. Or maybe a Deep South accent becoming unable to embrace it?'

She also asks if we know when such usage began, adding (in a totally different connection): 'the first occurrence I have come across of, for example, would of instead of would have is in Gone with the Wind.'

The answer to the first point is much, much simpler. Most accents of US English are rhotic - that is, they pronounce the /r/ after a vowel. The phonetic character of the /r/ is retroflex, i.e. the tip of the tongue is curled back towards the palate. In words like mirror we find two retroflexed /r/s occurring in quick succession. It is therefore the most natural thing in the world to run the two /r/s together. The alternative, to drop the tip of the tongue for the vowel and then to raise it again, would slow the pronunciation down so artificially that it would sound weird. In fact, nobody ever does this. A 'long /r/' is the result. No need to delve into the imagined American psyche here.

The fallacy in the explanations given by my correspondent is clear when one realizes that this isn't exclusively an American thing. Any rhotic accent will display the same effect. You can hear the same sort of 'long /r/' in West Country UK accents, for instance, or in Northern Ireland, and many other regions. Where did the American /r/ come from in the first place? Think of the people on the Mayflower, and where many of them came from.

As for the spelling would of and other such usages... this isn't a recent American thing either. The earliest instance I have come across of have appearing as of is in one of Keats's letters (5 Sept 1819). And the elliptical pronunciation has probably been around as long as the auxiliary verbs have existed in English. there are several 'ves in Shakespeare, for instance.

Friday 6 June 2008

On good times and bad

A correspondent learning English as a foreign language writes to say he is uncertain about whether he can say for four times (vs four times) and for several times (vs several times) in English. He has found the former in Othello and the latter in Frankenstein, he says.

The Othello example is a red herring. It is: 'I have looked upon the world for four times seven years' - so this is the use of times meaning 'multiplied by'. The preposition governs the noun years, and the usage is fine (although the phrase as a whole is stylistically unusual).

The Frankenstein one is found in Chapter 12 of Mary Shelley's book, and this is more interesting. It is: 'for several times they placed food before the old man when they reserved none for themselves'. That's certainly odd. The OED has 364 instances of the phrase several times in its entire database, and none of them have a preceding for.

Numerals provide occasional examples. The OED has 600 instances of three times, and there are two preceded by for - a British one in 1830 and a US one in 1919. So it may have been more common in earlier times - but two instances out of 600 isn't very convincing evidence for a norm. Today, as earlier, the standard usage for times meaning 'occasions / instances' is as follows:

I've been to France several/three times.
I knocked on the door several/four times.

Why do the for instances arise at all? I think it's the influence of the construction used for specific time periods, which readily allows an optional for:

I stayed there (for) several/four days.
I waited (for) several/three hours.

That's probably why there's some EFL uncertainty.

Tuesday 3 June 2008

On Caxton as linguist

A correspondent writes to ask, in relation to the emergence of standardization in English, whether the choices made by William Caxton in his publications were linguistically motivated.

I've actually dealt with this question at some length in my The Stories of English (Chapter 11), so I won't repeat myself here. My view, in brief, is that the spellings in Caxton's books were so varied that it isn't possible to ascribe to him any sort of conscious standardization policy. I agree with Norman Blake, in his excellent study (Caxton and his World), who concludes that Caxton was 'an opportunist in linguistic matters'. I see Caxton as a jobbing printer, who wanted to get books out as quickly and efficiently as possible. My impression is that he was a man of the moment, devising ways of solving the immediate printing task presented by a manuscript, and not especially concerned to check that the linguistic solution for book X was the same as the one he had already adopted for book Y. Even within one book, there are many variations. In the famous 'egg' story (from Eneydos we find 'asked' spelled within a few lines of each other as axed and axyd, 'wife' as wyf and wyfe, and many more such instances of variation. His books did introduce several features which would eventually help to shape Standard English. The 'egg' story shows that he was aware of the need to make choices. But there's precious little sign of a linguistic policy in his work.

On early splitting

A correspondent writes to ask about split infinitives. He has read The Fight for English, where I say that the construction has been in the language for centuries, but I don't give any examples. He is suspicious. Are there really any before the 19th century, he asks?

Oh, yes, plenty. They've been traced back to the 13th century, at first usually with a single negative word or pronoun as the splitting element, but then with adverbs. The early 15th-century cleric Reginald Pecock frequently used them. Here is an example from his The Repressor of Overmuch Blaming of the Clergy (chapter 5):

Goddis forbode y schulde be so lewde for to so seie
[God forbid I should be so foolish as to say so]

There are examples in Wyclife, Tyndale, More, Donne, Goldsmith, Burns... From Thomas More (1557) we find 'and then to not byleve them'; from Thomas Stapleton (1565), 'to flatly gainsaye'.

The practice became less common in the Early Modern English period - I don't know why. I can't recall any instances in Shakespeare, for instance. It builds up again in the 18th and 19th century - which is presumably why late 19th-century prescriptive grammarians started to condemn it - notwithstanding instances such as 'in order to fully appreciate' from Lord Macaulay (in 1843), among others.

Monday 2 June 2008

On tolerating

A correspondent writes: 'What, if any, is the difference between tolerance and toleration? As far as I know, tolerance is the commoner word. Are they synonymous?'

Tolerance is certainly the commoner word - try a Google search and you'll see it's about forty times as common. But that's not the issue. There are certainly meaning differences. Tolerance is the broader concept. It is used in a variety of settings, some quite specialized, especially in engineering (where people talk of tolerances, meaning a permitted deviation from a specified norm), biology (where it means the ability of an organism to withstand a particular environmental condition), and in medicine (where people talk of a tolerance for a certain drug, say). In its most general sense, it expresses a sympathetic willingness to accept views or behaviour other than one's own.

You also find tolerance quite often used as part of a compound expression, as in tolerance dose, tolerance level, and tolerance limit. A 2007 proposed addition to the OED notes tolerance zone - a designated area in which prostitution is tolerated by the local authority. That's a 20th-century usage.

Toleration is usually used when someone wants to talk about a specific instance of tolerance - a particular act of allowing something. In history, there are some famous cases of this kind, as reflected in such government names as Act of Toleration and Toleration Bill. It also has a much stronger suggestion of limits, and sometimes even a sense of reluctance. Tolerance has more positive connotations (a desire to accept) than toleration, which can mean 'we have to put up with this'. Compare the phrase religious tolerance with religious toleration. The country which practises the former is more likely to be enthusiastically supporting religious diversity than the latter.

During the history of these two words, there have certainly been occasions when their meanings have overlapped. But they seem to be pretty distinct today. Note that both have the same opposite: intolerance. There is no intoleration.