A year or so after I had entered the University of Leiden as a student of physics I noticed one day an unusual excitement among my more advanced colleagues. Its cause was the appearance of a new book Atombau und Spektrallinien by Arnold Sommerfeld. This brilliant author made it possible to learn everything about the Bohr atom, including the latest refinements of the theory, without having to dig it out of numerous, often badly written original articles. The book soon became the most important physics text and was read by all who were interested in the frontier of physics at that time, atomic structure. Everyone admired its clarity, though some claimed that the book was not always fair in giving credit (for example, it contained several sections on radioactivity, but did not mention the Curies at all).I could not afford to buy the book, but studied it avidly in the library. It suddenly entered my mind that to have been mentioned in that book was a sign of great success. In my youthful enthusiasm and lack of insight I was convinced that a footnote reference in "Sommerfeld" was sufficient to acquire some sort of immortality since it would be seen by numerous future generations. I have forgotten whether I made the grade in a later edition. But I am not even interested in checking up on it. The book has long since been assigned to oblivion. Few, if any, of the contributors to Physical Review Letters know of its existence and even fewer know of the important contributions to atomic theory made by its great author, Arnold Sommerfeld.I tell this story because it illustrates an error in authors' thinking: They do not wish to admit that their articles are only ephemeral, even when their contributions to physics may be lasting. We publish in so-called "archive journals," implying that our writings will be consulted for centuries. Our shelves are filled to overflowing with years and years of The Physical Review; we dare not discard any of them.With the growth of physics, the number of articles and journals is now so large that a physicist can keep informed only about his own field of specialization. The amount of new material a scientist can learn each year is no more now than it was in the past, while the amount of new ideas and data produced on the average per active researcher has probably also changed little. The obvious result is that an expansion of physics forces each physicist towards narrower specialization. This causes a split into a number of loosely related branches and leads to the creation of many new specialized journals.The multiplication of journals is, however, not a good solution to the problem of communication among physicists. There is still much overlap between the various branches so that it is now often necessary even for a specialist to consult several journals. Similarly, the selection of the proper journal for publishing a specific article is often difficult. The existence of many journals each with a limited number of subscribers is also uneconomical since it requires a multiplication of costly editorial, printing, and distribution services.A drastic modification in the publishing habits of physicists is needed if we do not want our science to disintegrate into a number of minor disciplines, competing for attention. Many suggestions for changes have been made, most of them based on the theme, "Please send me only the articles I want." It is not realized that this would require very elaborate distribution services of prohibitive cost.We wish to suggest just the opposite. It is far more economical to let the subscriber do the selecting; in addition, he may even discover interesting articles which would not have reached him if the selections were made at the source. A less cumber some and less expensive printing method and a wider circulation should be the aim. We envisage that a number of the now existing journals could be merged into a single "pulp" magazine. It would appear five times per week and would be mailed to the combined subscribers of the present journals. The savings to both publishers and subscribers would be considerable. Such a magazine would still be divided into sections according to subject matter, but the precise classification of an article would not be as essential as when the sections reach different readers. The wide distribution may even encourage authors to compose their articles so that they can be appreciated outside their specialized circle. Since we believe these articles to have only a transient value, the subscriber can discard the magazine after a suitably short time. However, a high-grade paper should be used for authors' reprints. Libraries could subscribe to a special edition printed on durable paper, ensuring the availability of original articles in a distant future.
Read full abstract