Clearly there are good and bad books of this type. The success of such a venture depends in part on how one approaches the question of whether or not to attempt publication of the papers presented at a meeting and on how the papers are handled.
Publication and the issues it raises need to be considered early in the planning of a scientific meeting. Quite often the sponsors of a meeting will insist on publication, and there may already be a standing arrangement with a publisher. If there are no prior commitments there are several important reasons to act decisively. For example, some keynote speakers may not want the chore of writing yet another version of their famous lectures. When you approach them, they might ask if you insist on publication, and perhaps be reluctant to participate if you do. Researchers with major new results may not want their data to appear in print before its publication in a primary journal. Equally, there are scientists who will welcome the chance to float their new ideas tentatively if a meeting volume is going to be published promptly. The point is that from the moment you announce the conference and sound out speakers, you will have respondents asking if publication is envisioned.
From the prospective publisher's standpoint, the earlier the approach the better. A responsible publisher will have good advice to offer on editorial policy and desirable length. Furthermore, if the book is to be produced quickly, it is necessary to plan the handling at the press in advance, and to start the publicity procedures before the typescript is delivered.
Is Publication Necessary?
There can be several good reasons for deciding against publication. One of the most important is that publication can inhibit free discussion of controversial issues at the meeting. Most scientists are rather more cautious about criticizing the views of others if they know their remarks will be printed. Similarly, researchers might be more reluctant to present emerging theories if they feel they will later have to retract a hastily published version. Then there are the economic considerations; libraries are under economic pressure, and a volume that merely collects together papers already published elsewhere serves little useful purpose. Think also about the work involved, not just for contributors but also for the academic editors, which will be at the expense of time spent on original research.
There are also arguments in favor of publication. The strongest must be in those cases where a conference can draw together ideas and results that are unpublished because they are so new, or that are widely scattered throughout the primary literature. The rapid publication of entirely new results through a proceedings volume is a particularly valuable service to the scientific community.
It is feasible for a publisher to have the finished book available at the meeting itself or within as little as three months afterwards, which is faster than most journals can do. Some meetings mark events and anniversaries, such as the retirement of a distinguished researcher, the return of Halley's comet, or the opening of a new particle accelerator. In such cases the provision of a permanent memento of the meeting is entirely appropriate.
If there is no obvious publisher, approaches should be made to a handful. All that is needed is a short letter stating when the meeting will take place, what its purpose is, which papers will be included, and how long the resulting volume will be. State who will be taking editorial responsibility for the book, and indicate whether the contributors will provide camera-ready copy. Include a copy of your conference announcement. You should get positive responses, and requests for further information, within one month.
In making a final choice of publisher consider some of these issues: How long will publication take from delivery of the typescript or camera copy? Has the publisher recently published other conference volumes in your subject area, and, if so, is the editorial standard reasonable? What are the arrangements for timely international distribution? Will contributors or conference participants be able to purchase the proceedings on favorable terms? Will reprints be available? Will the publisher set a price similar to those currently charged for most books of the same type? How many free copies will be given to the sponsoring body, the editorial committee, and to reviewers? You cannot expect a generous response to all of these questions, because conference publishing is not a big money spinner for publishers. The point is to get a good deal, considered as an overall package.
The financial arrangements are particularly important. If your conference is worthy of publication, then it should be entirely at the risk and expense of the chosen publisher. If an offer to publish is contingent on your finding a subsidy or guaranteeing the publisher against a loss, then you need to ask yourself if there is a market for the book. The exception to this rule would be if you want the book to have an unusually low price, in which case a subsidy would be necessary.
Most publishers will offer some remuneration to the editors, normally as a fee, but sometimes as a royalty on sales. It is not usual to pay contributors, because of the inefficiencies of making many small payments. Sometimes editors decide that their fee can instead be paid into the general conference funds. If the publisher is willing to pay the fee when the contract is signed, this can be a useful source of funds in advance of receiving registration fees from participants.
What Should Be Published?
Among the items that can be omitted to save space are banquet speeches, lists of participants (but not of contributors), the conference photograph, spontaneous discussions after each paper, and abstracts of the poster papers, which are more usefully published as a handout at the conference.
Apart from the selected papers, you could usefully commission some extra material in order to increase the value of the published version. A good foreword, prepared by a member of the editorial board, is an excellent feature. It should introduce the topic of the conference at a level that new graduate students can understand, and explain how the sections of the book relate to each other. In a conference with a number of different themes you can ask for introductory or linking material for each section of the book. The aim is to produce a cohesive summary of the meeting in which the flow of the papers is logical to the reader. It does wonders for sales if you can get a distinguished scientist to write a concluding chapter that pulls together all the new results and the problems still to be solved.
Ideally, the editorial board needs to review and edit the papers. This means that if possible papers should be submitted in draft form first; the book will be much improved if authors can be persuaded to go through this process.
For reasons of economy, most proceedings volumes are printed from camera-ready copy provided by the individual contributors. Most publishers will give you simple guidelines for contributors that you can distribute. It will probably be your responsibility to specify the length of each paper.
The publisher's instructions will include recommendations on the typeface to be used and advice on how to lay out the text to achieve a pleasing design. Grid sheets are usually provided so that all of the text will fit the chosen page size. Unfortunately, there is no standard among major science publishers; hence, the instructions will vary from one press to another. I urge contributors to use the highest technology available to them, to use grid sheets, and to select a classic seriffed typeface. In practice, this means that most authors now furnish laser printed copy of professional quality.
Diagrams and photographs should be kept separate, but adequate space must be left for them in the camera copy. The printer must screen the photos, which is difficult if they are pasted down.
If the book is going to be fully typeset, the individual papers can be prepared just as if they were to be submitted to a major journal.
Ready for Press
If you have prepared everything in good order it should be possible to produce the book in as few as eight working weeks, although three months is quite normal. Do not expect to find your book in stores that soon, however. Most publishers run their marketing operation on a six-month cycle, so it wifi be some time before all outlets are informed of the availability of the book. In the short term, you can help increase awareness of the book by taking leaflets to conferences or by mailing them out with your preprints.
How many copies will sell? Generally, conference books sell only to libraries, which are becoming increasingly selective. In the physical sciences 600-1,000 copies is a reasonable forecast. In topical areas of the biological sciences sales can be higher, and in medicine and engineering 2,000 might be achieved with good marketing. Sometimes sales can be much higher if a book catches a fashionable trend or has some unique feature. In 1979, I published a collection assembled by Stephen Hawking and Werner Israel to mark the centenary of Einstein's birth. It was a masterly survey of gravitational physics, which reprinted several times, has sold more than 10,000 copies, and is still in demand.