Someone writes to ask whether The Cambridge Encyclopedia of the English Language (CEEL, 1995) will have a new edition. I suppose one never says never. But I doubt it. Which prompts some thoughts on new editions.
One writes a new edition because the old one is out of date, for some reason. The second edition of CEEL (2003) was needed for two reasons: a century had changed, so every reference to 'in the present century', 'in the last century', and so on was wrong (plus the associated verb-form changes from present perfect to past - 'in the present century we have seen' vs 'in the last century we saw'); and the Internet had arrived, in all its many linguistic forms, requiring a whole new section. There wasn't much else to change - a handful of updates and expansions of points here and there. The population figures for English as a global language needed updating, in particular. But apart from this, the two editions are very close. Whether the amount of change warrants a new edition is a decision made by the publisher, always.
You have to distinguish a 'new edition' from a 'new impression' - that is, a reprinting of the book as the previous print-run has run out. Print-runs vary enormously. For books like mine, they're usually around 5,000 or 10,000 copies, depending on such things as the publishers' estimate of likely sales, and how much warehouse space they've got. Printing books 'on demand' in small print-runs is also changing things. A new impression is a chance to change things in a small way - getting rid of typographical errors that were missed first time round, for instance. Due to a last-minute change in printing The Stories of English, for instance, when a map was re-located towards the back of the book, some page references in the footnotes ended up being two out! A new impression allows you to get rid of that sort of problem.
You also have to distinguish a hardback edition from a paperback edition. Many publishers do not publish these simultaneously, as they naturally want to recoup their investment in a book by the higher price of a hardback. With my books, a paperback generally follows a year later. That too is a chance to make a few changes, though I don't normally do so.
Somebody mentioned Language and the Internet yesterday, and that's a nice example of a need for a new edition quickly. That book came out in 2001, and it had a new edition in 2005. That's amazingly fast. Why? Because in 2001 the first edition of that book made no mention of instant messaging or of blogging. Although blogging had been around since 1997, it didn't take off until 2003-4. Now it's the fastest growing area of Internet activity. A new edition is urgent, under the circumstances.
Am I working on any new editions at the moment? I don't like new editions, actually, and only do so when I have to. I mean, if I wrote a book in say 1999, that was because I was really into that topic then. Things have moved on since. I'm interested in other things now. To have to go back to how I was thinking in 1999 and try to take on board everything that I thought then and see how it might be improved... Well, it's a bit of a bore, to be frank. So there's just one, which has got to be revised at regular intervals, and that is The Dictionary of Linguistics and Phonetics. This has had a new edition roughly every five years since 1980, and there will be another one early next year.