Tuesday 10 February 2009

On anacolutha

A correspondent writes to draw attention to this sentence:

The table was covered with objects, although once he removed the books, he was able to drag it to the centre of the room.

He asks: can a subordinating conjunction be used to connect the complex sentence beginning with once to the preceding simple sentence or can only a coordinating conjunction join compound-complex sentences? He suggests this alternative:

The table was covered with objects, but once he removed the books, he was able to drag it to the centre of the room.

The use of 'although', in cases like this, is certainly common in spontaneous speech. What we have is technically described as an anacoluthon, defined (eg by the OED) as 'a construction lacking grammatical sequence'. Such sentences work semantically, but at the expense of syntax, because they usually omit a required element. What's missing here is something like

The table was covered with objects, but this wasn't a problem because, once he removed the books, he was able to drag it to the centre of the room.

Anacolutha are able to occur because they rely on our semantic or contextual awareness to allow us to make short cuts in grammar. They're very frequent in spontaneous speech. There's no problem understanding such sentences, of course. There's no ambiguity. But they're frowned upon in formal writing, where (in the above example) the coordinate conjunction would be recommended. Many of the strictures in Fowler's Dictionary of Modern English Usage are against anacolutha of this kind.

No traditional grammar handles them - and even linguistic grammars pay scant attention to them. The realization that the grammar of speech is very different from the grammar of writing has become a big thing, over the past 50 years, but anacolutha remain one of the neglected areas of grammatical investigation. This is a shame, as they're so common in speech. I've just spent a mind-numbing few days listening over and over to a series of lectures I recorded last year, in order to write a commentary on them for a book/DVD package which Routledge are publishing in May this year, called The Future of Language. Because these were spontaneous performances, without written notes or an Obaman autocue, they contain several anacolutha. I draw attention to them in the commentary. Interestingly, the firm contracted to add the sub-titles regularized many of them, so that they conformed to written English norms. I had to change them back. WYS (in the sub-titles) IWYG.


baralbion said...

Pam Peters in ‘The Cambridge Guide to English Usage’ (which I am glad to say I own on the strength of your recommendation) says, without further comment, in the relevant article:

‘Though’ has other roles than that of concessive subordinator which it shares with ‘although’. It also serves as a contrastive subordinator (like ‘but’), as in: He was reserving judgement, though he considered it a hopeful sign.

Anonymous said...

I'm not surprised the subtitling firm regularized your anacolutha - unless the brief is explicitly to retain speech errors and disfluencies, subtitles are almost never verbatim but are edited for length and clarity for readability purposes. Subtitling and transcription are different things.

I recently had to edit subtitles where the brief was to shorten the speech, ensure readability, and reproduce (a sense of) the original disfluencies and speech errors. Since the subtitles were also translations, this was an interesting excercise.

DC said...

Quite right. The firm needs a clear brief to that effect. But actually, where an ELT audience is concerned, writing such a brief would be quite tricky. In my case, I wanted to keep in some of my 'errors', as they help learners see what actually goes on in the grammar of everyday spontaneous speech; but I wasn't bothered about phonetic nonfluencies (such as the occasional slip of the tongue). I was able to make judgments about which to keep and which to omit as I went through the material; but I don't think I'd find it easy doing this in advance.

Anonymous said...

Could it be "...however, once he..."?

Juliusz Okuniewski

DC said...

There are usual several ways of resolving an anacoluthon, and yours is another - but there'd have to be a semi-colon or something beforehand, to avoid a reading miscue.

Unknown said...

So, Mr. Crystal, basically speaking, anacoluthon is not a deviation from the norm - it's more like the omition of the bulky content of the sentence, which is understandable from the context, something like subtle implication. And if you say that traditional grammar cannot classify this phenomenon as either deviation or a norm, why then formalists argue that it's inappropriate?

DC said...

No - omission isn't the issue. You can have cases of anacoluthon which retain the entire semantic content of the target sentences because only grammatical elements have been affected. If you want to put it in terms of deviations and norms, then anacoluthon is more like the deviant conflation of two syntactic norms, or the interference of one syntactic norm by another.

Anonymous said...

Professor Crystal, You say that “such sentences work semantically, but at the expense of syntax, because they usually omit a required element.” I see your point. However, there is also a missing link in terms of semantics and pragmatics: the step: ‘when a table is covered with objects is difficult to move’ is essential to understand the meaning of the sentence. Are anacolutha triggered by careless expressions then?

DC said...

This is a presupposition, not part of the meaning of that sentence. Noone ever expresses all the presuppositions that a sentence entails. Presupposition doesn't enter into the definition of anacoluthon.

Carelessness is a misleading way of putting it. This is a normal feature of spontaneous speech production.