On repeating oneself. On repeating oneself.

An unusual question from a correspondent: 'Do I ever repeat myself?' He was referring to an earlier post ('On writing how many books?') and wondering whether repetition presented me with a problem. It's an interesting question.

In a sense, I'm always repeating myself. That's what all lecturers do, as they present their courses, year after year. But of course, the lectures are never exactly the same. The subject changes. New facts emerge. And you think of better ways of saying the same thing. Even if your lecture is totally scripted, it's never quite the same. I asked actor son Ben once how he coped with the same thing, night after night, in a play run. 'It's never the same', he said. There's not much difference between lecturing and acting, from a performance point of view.

But the question is not so much about speaking as about writing, where repetition usually gets a bad press. Do I repeat myself there?

I think you have to make a distinction between research writing and popular writing. With the former, it is the nature of the beast not to repeat. There are new findings, new methodologies, new interpretations, always. And it would be a pretty poor journal editor who failed to spot that the content of an article was a repeat of something that had been published already, either by the same author or by others. Indeed, when I used to edit journals, that was one of the commonest causes of rejection - that a proposed article didn't actually say anything new.

But when you write for general audiences, you repeat all the time. The important word to note is 'audiences'. If you look in my website list of my publications, you'll sometimes see that an article written for one readership is then adapted for another, because the interests or level of the new audience differs. That's been a basic principle in the study of written language since the 70s: writers should always bear the needs of their audience in mind. An article written for students will differ from one written for teachers, and both will differ from one written for the general public, and differ again if it is to be presented as a broadcast talk, and differ still more depending on the channel. When I did a lot of Radio 4 broadcasting in the 1980s, a regular comment made by my producer, Alan Wilding, about my scripts would be: 'This is Radio 4, not Radio 3, David'. He usually meant my sentences were too long!

Often, a magazine or journal asks for permission to republish an article without change, or to cut it slightly. When I wrote an article on endangered languages a few years ago, it was picked up by half-a-dozen different publications and adapted in several different directions. But the article wasn't rewritten from scratch each time. What would be the point of that? I had given the first version my best shot. I couldn't think of a better way to say what I wanted to say. And if an editor felt that the piece was going to be of interest to a readership, then who am I to say otherwise? Anyway, if you have something that you think is worth saying - and there is nothing more worth saying, to my mind, than to draw attention to the issue of language death - you don't mind how many times you say it. You want the point to be recognized, and the more opportunities you find to make it, the better.

Having said that, when I do read over a piece for republication, I usually find myself changing it in some way. It might need updating. I might have fallen out of love with an earlier phrasing. I might think of a better analogy. There is an excellent English word for what one does: tweaking.

Does the notion of tweaking carry over into books? Yes. You will often see a reprint of a book with a phrase such as 'reprinted with corrections' on the publication data page at the front. It usually means no more than the correction of typos, ambiguous phrasings, cross references that don't work, and the like. It might be the correction of factual errors. There might be a minor updatings. It doesn't mean the wholsesale reworking of a text, or the addition or deletion of significant chunks which would affect the pagination. If that level of revision is required, we are talking about a new edition.

But there's another sense in which repetition is relevant for books. I have written several introductions to language or linguistics, and each one is an attempt to do the task better. The analogy here is with painting. I know an artist who spends his life trying to capture the quality of a landscape he likes to paint. In his studio you see the same scene, over and over, and yet it is never the same. I don't know - he doesn't know - whether he will ever be satisfied. I feel the same when I'm writing about language and linguistics. If that is what is meant by 'repetition', then it's an intriguing, positive thing.

Other factors affect the situation. Publishing fashions change; readership interests change; the subject itself changes; the angle changes; the author changes. I sometimes find myself reworking earlier material to satisfy new circumstances. Here's an example. The kind of publishing that Penguins were doing in the 1970s led me to write a straight textual introduction: Linguistics. The kind of publishing which Cambridge were doing in the 1980s led to a book at the opposite extreme - with full colour illustrations, The Cambridge Encyclopedia of Language. But then the wheel turned full circle. When Penguin approached me in 2003 to write an introduction to language, the idea the commissioning editor had in his mind was 'the Cambridge encyclopedia without the pictures'. That would have been deadly boring, and the only way I could think of to make the project worthwhile was to think up a wholly new slant - which is why the book came to be called How Language Works. The emphasis on 'how' gave the new project some life, but I have to admit that I did not enjoy this kind of reworking of material, and I think it shows in the writing. I shan't do it again.

Otherwise, no, I don't like repeating myself. I don't.