Keep this book to yourself

Peter Atkins is quietly impressed by an insider's guide to writing textbooks

I cannot recommend any aspiring textbook author to buy this book: it is so good. Writing Successful Textbooks is a step-by-step guide to the creation of textbooks, and it will be particularly valuable to inexperienced authors. It will also be helpful to old-timers for the manner in which it organises and elucidates what we have deployed intuitively.

Anthony Haynes is a commissioning editor for educational and professional subjects, so the strength of his exposition is in these fields. I can feel relatively safe, therefore, because there are secrets of scientific textbook publishing on which he does not touch with the same revealing zeal as he does on his familiar ground. I will expose some of these secrets later.

So, what are the secrets of successful textbook publishing? How do you turn your lecture notes into a modest, and sometimes immodest, mound of cash? Haynes talks his readers through the process of organising their thoughts into a proposal and presenting the proposal to potential publishers. As Haynes is a publisher himself, potential authors should be just a little wary of all his advice, which is likely to make a publisher's life easier. He recommends that you offer your book in sequence rather than in parallel to a selection of publishers, and thereby (although he does not refer to it) avoid a bidding war from which you, the author, could benefit enormously. "Agent" is a word that does not appear in the index, although I suspect they do appear in the nightmares of commissioning editors.

That having been said, there is a difference between trade and textbook publishing. A textbook needs to be developed, and there needs to be a mutual trust between publishing house and author. As Haynes remarks, that trust is best established at the outset.

I say that textbooks need to be developed. You have to accept that although your own judgement about the organisation, exposition and presentation of a subject is without parallel in the universe, you need advice. You need someone to stand back from your efforts and advise you on what works and what does not. More important still, although you are the bee's knees in your subject, you need to take the advice your editor offers, or at least consider it with the utmost seriousness.

Here my experience differs somewhat from the author's, and differs between the left and right sides of the Atlantic. North American publishers have a number of advantages over their counterparts in the United Kingdom, all of which combine to allow them to make a greater investment in a text.

First, their captive audience is greater. As well as the intrinsically large and uniform market of North America, adoption regimes are more stringent, so a publisher can be confident that books will be bought. All students studying science, and particularly those going on to medicine, are required to study chemistry in their freshman year, so the market for that subject alone is several hundred thousand students. Second, the North American market has been trained to accept significantly higher prices, with freshman textbooks costing typically $100 (£70) or so, so the income per sale is much greater. Third, history shows that it is much more likely that books will travel east than west, and that there is a greater likelihood that a US textbook will be acceptable in Europe (and elsewhere) than vice versa.

The attractions of the North American market are so great, and the potential rewards so great, that a publisher is prepared to make a huge investment, which makes the regime outlined by Haynes look a little underpowered. Typically a US publisher will appoint a development editor, who will collaborate with the commissioning editor, and later the production editor, to shape and mould the book. That development editor may have only your book and perhaps one other to concentrate on for the two years or so needed for its preparation.

Moreover, because the audience in North America is so diverse, a textbook alone is like an aircraft carrier with no escort: it looks vulnerable and undervalued. Consequently, a fleet of supplementary material is normally launched at the same time, including an instructor's solutions manual, a students' solutions manual, a study guide, laboratory manuals, an instructors' resource book, CD-Roms and a rich website. Fortunately, most of this supplementary material is prepared by people other than the authors of the main text. But it is essential that time-consuming collaboration take place to ensure that the style of presentation is consistent.

If anything, Haynes underestimates the effort needed to produce a textbook, which is also a slight camouflage in a publisher's favour, but he does allude to the requirements of the web. In the old days (a decade or so ago), there was a sense of completion when one's typescript went off to the publisher. Certainly, one's long vacation was marred slightly by the arrival of proofs, but in due course the first pristine books arrived to be laid ostentatiously on tables, and that was that.

Now there is no end to the process. In more recent times, the need to initiate or at least be heavily involved in the preparation of a CD-Rom comes at the busiest time, when the typescript is being finalised, and adds hugely to the burden of production. Currently, CD-Roms are thought slightly passé, and the favoured medium of electronic delivery is the internet. The website for a book is an organism with a voracious appetite for your time, and can easily occupy all the interval that once lay fallow between editions.

I suspect, but have no evidence, that publishers are having a harder time commissioning authors to write big textbooks, partly on the grounds that the work involved is now so much greater than before. Another factor plays a role, at least in tertiary education: the structure of British research is now such that the income of departments is so highly geared to research papers that heads of department are compelled to stamp on the green shoots of potential textbook authorship. It may be right that primary effort should go into the discovery of new knowledge, for nothing is more precious, but one cannot help thinking that the balance between encouragement to do research and to reach out to share insight with a worldwide audience needs adjustment. To an established author, this elimination of competition is not wholly unwelcome; but from a less selfish point of view, its effect on future generations is potentially regrettable.

Haynes takes you through the processes necessary to discover your readers' needs, to organise the material, and to lay it down on the page. Where he is a little thin is in the area of illustration. For him, illustration appears to be the simple-minded diagrams that characterise books in the humanities, with arrows sprouting from this to that and giving rise to the sense of organisation. He rightly remarks that diagrams - by which he must mean line drawings - are a help to those who think visually. They are also an essential guide to readers for whom English is a second language, as they can act as a picture dictionary.

In a science textbook, they are absolutely essential. Scientific diagrams consist of more than a few arrows. They include detailed line drawings in which the author tries to bridge the gap between the perceived (the macroscopic world of observation) and the imagined (the underworld of atoms, wherein lies the root of the observation). They also include photographs to display the reality of the discourse and graphs to show the variation of properties with parameters. I like to do all the line art myself, as I think it inappropriate in a science textbook for an artist to come between the scientist and a page. With drawing software now so powerful, even the most artistically challenged can make a good effort to convey their visual insight.

Where Haynes also underestimates effort and importance is in the preparation of worked examples. Maybe they are less important in manuals of baby care than in textbooks of chemistry and physics, but there is no doubt that they are an essential teaching aid. One of the real difficulties for students of the physical sciences is to learn how to take the step from exposition to action, or even from a qualitative perception of the solution of a problem to its quantitative implementation. This activity is the most difficult to teach, partly because it is inductive and people think creatively in so many different ways. The only way is to present lots of worked examples with detailed solutions, and to display somehow the thought processes behind the solution, and to combine the worked examples with immediately available ways of testing understanding.

The concluding chapter of Writing Successful Textbooks strikes at the heart of the issue: with electronic media now so ascendant, is there a future for any printed textbook, successful or otherwise? Here, as elsewhere, Haynes treads with a sensitive step and explores the disadvantages as well as the advantages of e-books.

There is a real difference in learning experiences between the printed book and electronic media. The two disadvantages of electronic media are the lack of intellectual rigour and the fact that it is hard to know where one is and how far one has to go. (I am ignoring the current inconvenience of multimedia devices: a temporary problem that will disappear as hand-held computers mature.) A textbook is an experience in intellectual rigour, with arguments that have to be mastered and the exploration of a subject in a linear manner. You also know, when halfway through reading a text, that you are half way. In contrast, current multimedia experiences are less taxing intellectually and do not impose the same requirements of sustained argument.

The advantages of multimedia texts, though, are huge, and the writing must be on the screen for the demise of paper in this area. Once they become convenient to use, they will allow illustrations to come alive as videos; up-to-date data to be linked in from the web; graphs to be open to exploration; all information to be hyperlinked internally and externally; and courses and examinations to be networked. For publishers there will be minimal distribution and unit costs, savings that must (unless the publisher is particularly canny) be passed on to users.

Haynes has written a really useful book. Stylistically, he has a little to learn, for I think that textbooks should be naturally gender-free (like this article) rather than resorting to the irksome repetition of "he and she" several times on a page. His presentation is thin on the sciences, but not excessively so, for what he writes is largely generic. Altogether, if aspiring authors follow his advice, the textbooks they produce should have no excuse, other than the author's competence, for not being first class. As I implied at the beginning of this review, I would be fearful if this book fell into the wrong hands - my competitors'.

Peter Atkins is professor of chemistry, University of Oxford. He has written about 40 books, most of them textbooks.