How to Write a Good Science Text

Most established scientists based in universities have probably been approached by book publishers. Acquisitions editors are always searching for essential monographs, timely conference proceedings and outstanding textbooks. The quest for good authors is highly competitive. Most publishers now use subject specialists who are able to use their own judgment when they come across an interesting proposal. These editors visit campuses and attend conventions in order to drum up business. How should th

By | April 20, 1987

Most established scientists based in universities have probably been approached by book publishers. Acquisitions editors are always searching for essential monographs, timely conference proceedings and outstanding textbooks. The quest for good authors is highly competitive. Most publishers now use subject specialists who are able to use their own judgment when they come across an interesting proposal. These editors visit campuses and attend conventions in order to drum up business. How should the working scientist respond to a request to write?

The biggest rewards for authors and publishers come from textbooks. A good text will regularly be reprinted and go into revised editions. Once firmly established, a textbook provides a steady income for some years. This is in marked contrast to a symposium volume, which normally will produce almost all of its cash flow in just one year. Publishers want texts in the backlist to offset the short-duty cycle of monograph publishing.

A textbook is the hardest type of book to write well. Conference publishing demands little more than good organizing skills to bring in the papers. In a monograph it is even acceptable for authors to cull material from their research and reviews. And in both of these styles of publishing you are addressing other researchers who speak the same language. With a textbook you have to write down to the level of the average student. You will need to heed the advice of your publisher and of external advisers. Above all, you need to give the project exclusive attention during the main writing and revision phases.

The Proposal

You may have a novel idea for a textbook, or perhaps a publisher has approached you. Either way, the first step is to write a proposal for the publisher. This summarizes the content of the book you intend to write. A helpful way of looking at a synopsis is to ask yourself: "What information would I need if the publisher were asking me whether to offer a contract for this book?" The outline should have a paragraph or so to indicate the content of each chapter. How much will you write and how long will it take you? Describe how you want to illustrate the book, particularly if color printing is called for. State clearly for whom you are writing, and list some of the courses that might adopt your book as the required text.

Your publisher will certainly want to see a writing sample as well. This could be a review article or a sample chapter. The acquisitions editor will show your proposal to advisers and probably to marketing colleagues also. He or she should then work with you to revise the outline so that it best suits the needs of the market. Once you have agreed on the strategy, the publisher offers you a contract. Because textbook writing is very time-consuming, it's sensible to wait until you have a legally binding agreement to publish your work before writing too much.

For large or specialized texts, it is worth thinking about involving a co-author, both to spread the work load and to contribute knowledge that complements your own. Remember, however, that the total amount of effort might actually increase, particularly if the human chemistry doesn't work perfectly. Each author should have clearly defined writing responsibilities. It is important to choose a person whose writing style and approach to the subject match your own. Consider whether you will be able to interact properly; 10 minutes over coffee at a busy convention might be fine for putting the finishing touches to a brief research paper, but a book will involve a lot of detailed discussion, perhaps even paragraph by paragraph. Also think carefully about the financial elements, which can lead to differences of opinion later if they are not carefully discussed and agreed to at the outset.

Writing Style

If you are writing a major introductory college text, your publisher will assign a development editor to teach you the tricks of the trade. Most textbooks are not in that category, and publishers rely heavily on the good judgment of their authors.

The first thing to cast aside is the concise and formal style used in research papers, which is much too dense and uninviting to students. The best style addresses the student directly and is written in the active voice. I encourage my authors to use the personal pronoun when writing about their own contributions to a subject. That way the students can see that an active researcher is guiding them through the subject. Use feminine as well as masculine pronouns. Any little anecdotes about your own life as a scientist can do wonders to liven up the duller parts of the story.

It is very important to treat your subject from an international perspective. This will not only enhance the worldwide sales of your text in English, it will also increase the chance of selling translation rights into other languages. So, a marine biology book ought not to take all of its examples from the Chesapeake Bay, and an astronomy text must include galaxies in both hemispheres. To get interesting data and case studies, try using your own international network of colleagues to the fullest. Scientists like to be asked for advice and they will often contribute material that will be unique to your book.

You should think about the differences between the oral presentation of your lectures and written prose. For example, in a mathematical derivation it's tedious to read phrases like "we see that … more than once or twice per theorem. Lecture notes will provide an excellent basis for your book, but they will almost certainly need extensive revision and adaption.

References to the primary literature seldom have a place in books for undergraduates. I strongly discourage citations within text because they interrupt the flow of the argument. That distracts the student, who may literally think it's necessary to seek out the paper in a reference library in order to follow the point being made.

Use a bibliography to encourage students to read more widely. This should be more than a listing of competing textbooks and recent papers from your own research group. What lecturers and students appreciate is an annotated bibliography. For each item you list, give a sentence or two saying why you have selected it. Is it a classic paper, an outstanding review or an unusually clear presentation of a difficult topic? Will it take the keen student nearer to the frontier of research? Generally it is best not to list out-of-print textbooks or articles in obscure journals that students will find hard to track down.

Sometimes authors use in-text citations to acknowledge the work of others. While that is essential in a research paper, it is better in a text to gather your acknowledgements in one place. The exception to this general rule is the reproduction of original data in diagrams and tables, where proper acknowledgement should be given in the caption.

Special Features

Any special material you can add will increase the possibility of your text's being widely adopted as a standard text, and will also give students the incentive to hold onto the book after their course has finished, rather than dumping it on the used book market. Of course, it is hard to think of unique selling features, but you should try nevertheless. For example, perhaps really comprehensive lists or tables will mean that students in your subject will want to retain the text as a reference source

On the other hand, it is important to avoid gimmicks. Booksellers hate anything that can fall out of the book, so forget ideas such as including strips of polarizing film, viewers for three-dimensional pictures, microfiche and slides. Publishers will flee if you propose fly-outs—big charts or maps that unfold from the book. These are exceedingly expensive and will simply lead to your book being priced out of the range of students.

Currently, books that include a software package are selling well. If the subject lends itself to computer programs these can be listed in the book, and disks can be made available to anyone who does not want to type out the programs personally. Any text that is mainly about software should certainly have a disk. Once again though, booksellers don't usually want the disks. Either you can sell them yourself, or the publisher should make them available by direct mail, preferably by printing order blanks in your book. Naturally, you should use a widely available language in writing the programs, and choose a microcomputer version of it.

For introductory college texts, a teacher's manual is a strong selling feature that will make all the difference when it comes to getting the book adopted by schools. The manual does not need to be elaborate or elegantly designed. It should explain to the instructor how best to use the book, how to design experiments or observations to go with it, and how long the course will take. It should also contain the full solutions to quantitative problems set as exercises in the text.

Examples and Problems

Choose your examples from as wide a range as possible, and make a conscious effort to describe situations or problems that the average student is likely to encounter. In the physical sciences and mathematics, worked-through examples are very helpful indeed, particularly for students who may not have much opportunity to receive individual instruction to help resolve problems. In many areas of the biological sciences, be sure to include first-rate, clearly-labeled diagrams, and good photographs.

Instructors generally welcome the inclusion of problems for the student to solve, as well as suggestions for discussion, project work and further reading. I have found that the problems submitted by authors are frequently too difficult. Perhaps this is because they are gathered together from old examination papers! Such collections need supplementing with much easier questions, the simplest of which can be the rederivation of a worked example using slightly different data. I consider that the first few questions after a chapter should be readily solvable even by the weakest students. That way they will gain confidence. If all the questions are too hard, the students will grumble, and the book may not be adopted the next year.


I like to be able to pick my way through a book easily. For that you need to number the illustrations and equations sensibly. I think that a simple decimal system is best: the chapter number before the decimal point and the figure or equation number after it. If you number in a single ascending sequence right through the book, the copy editor, or you, will have to renumber everything if a new picture or equation is added. With the decimal system the damage is confined to one chapter. Modern methods of printing mean that you do not need to regard the photographs as separate "plates" (unless they are in a color section) and consequently they can be numbered sequentially with the diagrams. Whatever system you adopt, it is best not to use Roman numerals (which many people find hard to read at a glance), and sensible not to overorganize the text by numbering everything of paragraph length (which gets tedious for the reader).

In some subjects, particularly physics and mathematics, it can be very helpful if all the symbolic notation is gathered and summarized in an appendix. Incidentally, except in mathematics, it is best not to use the Greek alphabet for the first-year textbooks.

A good index in a must. Although it is a chore, it is best done by the author. Most publishers have a simple guide on how to write an index.

A detailed table of contents is useful too. Quite often the contents are used by the marketing department to promote a book, so the more you put in there, the greater the information they will send out to lecturers.

If you are a relatively junior faculty member you might consider getting a big name to endorse your text by writing a foreword, which will help to establish the credibility of the text in the market.

My final piece of advice is that if you intend to submit your text on a tape or disks, do please tell your publisher beforehand.

Mitton is editorial director for science publishing at Cambridge University Press, Cambridge CB2 2RU, UK. He trained as an astronomer and has written several books for children and students.

Popular Now

  1. Publishers’ Legal Action Advances Against Sci-Hub
  2. Metabolomics Data Under Scrutiny
    Daily News Metabolomics Data Under Scrutiny

    Out of 25,000 features originally detected by metabolic profiling of E. coli, fewer than 1,000 represent unique metabolites, a study finds.

  3. How Microbes May Influence Our Behavior
  4. Decoding the Tripping Brain