On writing a textbook, and why you shouldn't

I've published 1.5 academic books in my working life. The "half book" ended up as a special edition of the Journal of Physics (IIRC), rather than a full book, because the Research Council that was paying me to write it ran out of money. Or so they said, anyway. Still, I guess I should be grateful for being paid up-front even to write half a book: most authors rely entirely on royalties for their remuneration.

There's something satisfying about being able to claim you're a "published author", even if your work has a miniscule circulation. I've got a photo of me in Waterstones' bookshop in San Francisco, pointing at copies of my book on their shelf. I got a bunch of free copies from the publisher, ostensibly to give to interested colleagues but, in practice, to be displayed on my mother's coffee table.

Unfortunately, a warm fuzzy is about the best you can expect from academic publishing, and the effort involved in publishing a textbook is extraordinary.

First, of course, you've got to write the darned thing. My problems in that area started early, with an argument about software. I wanted to write the whole book using LaTeX, which is what I used for all my academic writing. To its credit, the publisher did eventually manage to find an editor and a type-setter who could handle LaTeX. It took a while because it seems they didn't use this format for anything except math books. In fact, the publisher was very keen for me to use Microsoft Word. I can imagine using Word to write a letter, but a four-hundred page book full of program listings and equations? The very idea...

I expected the actual writing to be the hardest part of the authorial process but, with hindsight, I see it was the easiest. After the writing, you've got the horrors of editing. Despite what authors often say, brutal editing is necessary to produce a readable book. I approve of editing in principle -- even of my own writing. Unfortunately, in the academic world, editors are rarely experts in the subjects of the books they're working on, and they labour under severe constraints imposed by the publishers.

In my case, the editor told me that the publisher's policy was to produce all books, regardless of subject matter, to suit a reading age of twelve. Yes, twelve. I wasn't allowed to use semicolons, or any sentence with more than twenty words. There were many Byzantine rules about how to use terms that might be trademarks, or subject to some other claim of intellectual property. I was told, for example, that I couldn't use any trademarked term as a noun, only as an adjective. So I couldn't write "an Oracle administrator", I had to write "an administrator of the Oracle database". That's fine if there's only one or two of these occurrences, but when there were two dozen on every page, applying the publisher's rules led to laboured prose that I really didn't want to be associated with.

Then there was the matter of language and style. My publisher was based in the US, and I am very much a British writer. Too much, perhaps. The problem wasn't just in converting all the currency units and measures to US values -- which would have been a pain in the butt on its own -- but also of avoiding anything that smelled like idiomatic British writing.

So while it took, perhaps, six months of evenings to write the first draft of my textbook, the editing process took a further six months. And, to be honest, a fair amount of friction.

Then there's the review process. I guess it's less a problem if you're publishing a novel but, for academic books, the publisher can't take the risk of releasing something that isn't authoritative. So I had to deal with not one, but two different academic reviewers, neither of whom had much knowledge of my (rather niche) subject. I had to make fairly extensive, pointless, changes to satisfy these folks, while still struggling to leave the text making some kind of sense.

By the time I'd fought my way past the reviewers, my book was bordering on something I could barely recognize as my own work. Whether the editing and review process made it a better book, or just a different one, I really don't know. Like most writers, I'm completely unable to assess the quality of my own work. Still, after another three months or so, my book was on the shelves.

And already out of date.

At first it was nice to get a regular cheque from the publisher but, of course, sales dwindled, and they were hardly rampant to start with. Books on specialist areas of software engineering don't fly off the shelves. After about six months, cashing the cheque became a chore, the payment having become so small.

I look back, and wonder whether it was worth it. Financially, probably not. My publisher offered pretty generous royalty terms, compared to the industry as a whole, but the darned book was never going to sell like a Stephen King novel. In the end, counting up all the hours I worked, and the amount of money I made, I probably just make more money than I would have, had I spent my free evenings tossing burgers or sweeping floors. But only just.

While having a textbook (and a half) to your name does look slightly impressive on your curriculum vitae, I can't imagine it's enough to swing a job application one way or the other. In the end, I think I wrote a textbook for the same reason I write everything else: because I enjoy it.

But I don't think I enjoyed it enough to do it again.

Published 2026-01-20, updated 2026-02-23

On writing a textbook, and why you shouldn't

Categories