On writing how many books?

This morning brings a familiar query. Someone is writing a piece in which they want to mention how many books I've written. They have looked on Google and seen references to 'over 40', 'over 80', 'over 100'... Can they have a definitive answer, they ask, from the horse's mouth?

Well, not from this horse. There's a problem, you see. It all depends on what you mean by 'book' and by 'written'. This isn't just a linguistic quibble. It's a real issue, and I never know what to say when someone asks me this question.

If you want to research this, you'll find the information at www.davidcrystal.com, where there's a complete listing of all the books that have my name on the title page. As of today, there are 160 items on that list, excluding translations. But what does that figure mean?

To begin with, you have to make a distinction between books which are written and books which are edited. This is especially important for me, because since 1986 my life has been schizophrenic. On the one hand I keep trying to be a linguist. On the other hand I find myself doing a huge amount of general reference work, as an encyclopedia editor.

The distinction between writing and editing isn't clear-cut, actually, because when you're editing something you do an awful lot of rewriting of other people's stuff. Sometimes, when the entries are not of a specialized nature, I end up writing them myself. But for the most part I find myself handling material which other people have written. In the case of the unabridged encyclopedia, I use material from over 350 consultants around the world, as well as drafts from my small in-house team of assistant editors. That work informs all the editions of the Cambridge 'family' of general encyclopedias (not the two language ones, mind, which are authored works) and of the Penguin 'family' of encyclopedias and reference books. There are 33 of them altogether.

Then on the language side, how do you quantify co-authored books or books where I've edited a collection of essays? I am not by nature a collaborator, but over the years I've worked with colleagues on a number of such projects. Indeed my very first book was a collaboration, with Randolph Quirk. (He wrote well over half of it, actually, and showed me how to write my bit.) Or, to take another example, the Skylarks reading development programme - which wasn't my idea: it was Jeff Bevington's. I was brought in to give the scheme some linguistic structure, and ended up doing so much that I became a co-author. And above all, there was the Databank and Datasearch series of information books for children with reading difficulties. There are over 30 of them, and all are co-authored with English teacher John Foster. I didn't write any of these - I re-wrote the material John provided, using developmental linguistic criteria, so that the information was presented in a way which maximised ease of reading. In all there are 46 items in my list which involve other people in one way or another.

And what does one do with 'second editions'? (or third, fourth, etc.) Several of my books have gone through more than one edition. I hate doing new editions of an old book, actually. I find it a pain having to think myself back into the mindset of how I was - usually several years before. Still, some types of book do need regular updating, and so it has to be done. And with some books it can take almost as much time to do the new edition as it did to write the first one. Each new edition of the Dictionary of Linguistics and Phonetics, for instance, takes ages. And for the second edition of The Cambridge Encyclopedia of Language, the publisher allowed me to use full colour, instead of the second-colour restriction I had to work with in the first edition - which meant re-researching all the illustrations, and that took the best part of a year. So I guess new editions of this kind should count as separate items, in any inventory. On the other hand, when What is Linguistics? was in print, each new edition involved minimal updating (mainly fresh information about the universities which were offering linguistics courses at the time), so they should hardly count as separate books.

If you ignore all the problem-cases in the language domain, you will find a core of 40 items with me as sole author. That's where the figure of 'over 40' must come from. Adding the main academic collaborations, such as Investigating English Style, The Grammatical Analysis of Language Disability, and Shakespeare's Words, would take this to over 50. Including the educational collaborations, for schools or EFL, would add another 40. That's where the figure of 'over 80' must come from. And if you add the rest... well, take your pick.

As a linguist, I try to write a book a year on a fresh topic. Sometimes I don't manage it; sometimes I manage more. The demand for new editions on the general reference side often gets in the way. I read every word of every reference publication that goes out under my name, and with the larger books this can take a lot of time. Also, as I said, new editions of the language books suck up time too. So, I see that 2003 was a poor year, with no fresh book; but you can see why when you look at 2004, which is when a lot of the writing being done over the previous two years actually appeared.

The publication dates sometimes mean very little, in fact. Publishers will often delay or hasten a book's appearance depending on marketing considerations (back to school, Xmas...). On more than one occasion I've had a book held over from one year to the next to obtain (what the publisher hopes will be) a better presence. That can lead to the unfortunate clash of having two books appear in the same month. This happened to me in January this year, when How Language Works (originally scheduled for publication in March, but brought forward by Penguin) appeared within a couple of weeks of Words, Words, Words (originally aimed for pre-Xmas, but kept back by OUP to coincide with the launch of a TV series). It's not good when this happens. The two books are usually reviewed together, so that neither gets full attention. And although bookstores will try to persuade you to buy both at once, with a discount, I doubt it often happens.

An interesting variant of this is when a book gets turned down. I don't usually write a book without a contract in place. As a free-lance academic these days, it wouldn't be sensible to spend time writing something which might never see the light of day. (An author always has to keep an eye on the year-after-next's income.) But occasionally it happens, usually because I'm carried away with the mood of a moment. I get enthused about a topic, and when that happens I can't not write about it. Often I drop everything else and just write and write until I've got the topic out of my system. Then I wonder what to do with what I've written, and offer it to the publishers I know. You might think the new work would always be welcomed with open arms? Not a bit. My bottom shelves are filled with the corpses of rejected typescripts.

But never, never throw anything away. For what is dead meat to one publisher in one year might be highly attractive to a different publisher (or even the same one!) in another year. And that's a further reason why some years in my book-list appear to be more productive than others.

For instance, A Glossary of Textspeak and Netspeak appeared in 2004. That was actually written in 2001, at the dawn of the texting era. I was at a lunch and sitting next to the chairman of Bloomsbury publishing. We were talking about the Internet, as my book on that topic was just out. I was going on about the new texting linguistic developments that were emerging (abbreviations, emoticons, and so on), and how there was as yet very little information available about them. 'What about a book?', he said. 'Good idea', I said, and promptly went away and wrote it. A few weeks later I sent it in, but got no response. Too busy with Harry Potter, I supposed. After a few months it emerged that there wasn't any interest in it after all. So I asked my commisisoning editor at Penguins whether they would be interested. Nope. So it joined the other things mouldering on the bottom shelf and I forgot about it - until a year later, when I was approached by an editor at Edinburgh Univesity Press. They were thinking of a new introductory series and were hoping I would write a short introduction to linguistics. Having already done this three times, I said no. Then I thought: hang on, I wonder whether they'd be interested in a short introduction to internet jargon... And they were. I blew the dust off the typescript, found the original in my computer archive, gave it a thorough updating, and sent it off. Bingo. It appeared in 2004, and helped convey the impression that I'd been specially busy that year.