Reference Books: Essential--and Profitable

As a schoolboy in England in the 1950s and ‘60s, I was first introduced to reference publishing by Kaye and Loby’s Tables. Here you could find all the “right” answers to experimental demonstrations in physics and chemistry, such as the viscosity of various mineral oils and Young’s modulus for steel, which then seemed rather remote from everyday life. And we used four-figure logarithm tables all the time. What a gold mine they were for publishers: in public examinations completely fresh copies had to be issued to every pupil. Reference books are indispensable to the working scientist. They are also the most demanding publications to write, and the hardest for publishers to commission. In 1972, when I was at the Institute of Astronomy in Cambridge, I put together the Cambridge Encyclopedia of Astronomy with a dozen active researchers. This was my introduction to the hazards of reference publishing. After I had assembled the team and put together the outline for the book, I was told constantly that some aspect was under-represented, or I should have included so-and-so in my list of contributors. When the typescripts came in they covered the full spectrum from excellent to unacceptable; many had to be heavily edited. And the reviews of the final book ranged from lavish praise to extensive nitpicking. I now have been a publisher for nine years. Of all the types of books that feature in my lists, reference books are by far the scarcest. But that handful of books is among the most important contributors to turnover as well as the prestige of the publisher’s imprint. Good reference books are fantastically. demanding on the editor and contributors. In a magazine article, or even a textbook, it is perfectly acceptable to say that the speed of light is about 300,000 km per second. In a reference handbook it would be necessary to state that the value accepted by CODATA is 2.997 924 580x10^8 meters per second. This is just a simple example to illustrate that in a reference book all numerical data have to be accurate to the highest possible degree. Most reference books, particularly dictionaries, are based on works that already exist. A common technique for compiling new reference books is to imitate the best features of several similar books already in print, update the information and add some new data. For science books there are usually not copyright problems with this approach because data as such cannot be copyrighted. However, the slavish reproduction of tables or graphs without any modification is not only morally questionable, but would certainly invite a robust reaction from the copyright owner. I have found that very few scientists can take time out to work on reference books. This is a pity, and a detriment to science. If a handbook is to be reliable it should be assembled by working scientists rather than editorial committees or junior editors in a publisher’s office. The active researcher knows where the latest experimental data are to be found and is equipped professionally to make judgments about the reliability of new data. If assembly of reference material is left to freelance editors and writers, chaos ensues. Just last week, I had two tables of solar system data on my desk, one asserting that the diameter of the planet Pluto is 3,000 km, while the other insisted it was 1,140 kin; since neither table gave the (essential) citation to the source of the data, I was left with some detective work in the library. There is one feature that all reference books should have, independent of discipline: the sources of the data must be given explicitly, with a full citation. The reason for this should be obvious; namely, that some users will want to double-check, particularly if they come to consult the book a few years after publication. Also, there is sometimes a difference between accepted values and the latest experimental data. In my own field, one example is the distance from the molar system to the Galactic center. officially this is 10 kiloparsecs and that value is used in many calculations, although the most recent observations suggest the distance is probably about 8.3 kioparsecs. A reference compilation needs to distinguish between these two acceptable values. A working scientist can only use a reference book with confidence if it has a good reputation. Some have been around for a long time and have gone through many editions. The publishers of these titles have a vested interest in ensuring that the data have been verified very carefully, and that new editions are absolutely trustworthy. For new reference books there is a problem, at least in the first few months after publication. However, if they are seriously flawed the bad word seems to get around remarkably quickly. But if they fulfill a real need, the author’s reputation is established immediately. The production of a reference book does not bring the same kudos as a dramatic discovery announced in a research paper. However, a good handbook can have a life much longer than most research papers, particularly if it is brought up to date regularly. The professional and financial rewards to the author then can be very substantial. There is also real satisfaction in knowing that numerous scientists have benefited from the assembly job. Mitton is editorial director for science publishing at Cambridge University Press, Cambridge CB2 2RU, UK

Reference Books: Essential--and Profitable

Interested in reading more?

Become a Member of