A Brief for Brevity

Dear Editors:

Although I have been following the debate in Physics Today about replacing traditional archival journals with electronic publishing and thereby avoiding involvement of referees and editors, I did not take any of this discussion seriously. The major drawback of such an electronic scheme for the average scientist - having to wade through reams of pages of papers that have probably never been read before by anyone except the authors - is simply too scary to consider. My experience as a referee has shown that all too often the papers submitted for publication have not even had the proper proof reading that a good secretary could supply. It is well-known that the great majority of papers are now read and referenced by very few scientists besides the authors themselves. An electronic publication scheme with no refereeing and editing would I think aggrevate an already bad situation and lead very quickly to the death of useful scientific literature. Who will have the courage to scan through all the mediocre (or worse) entries that would accumulate in a vain attempt to locate the few gems?

On the other hand, I have been shaken from my own complacency by recent demands of some journals to reduce the size of all papers. As an author, I am confronted with higher charges (penalties) for papers longer than an arbitrarily set length. As a referee, I am asked to advise other authors how they could shorten their papers. I find that such demands/requests are not at all easy to fulfill either for my own papers or for those of others. It seems to me that what we are experiencing here is a critical phenomenon soon to be followed by a phase transition. The question is what the new phase should be (electronic publishing? NOT!). One possible answer I have not seen seriously discussed is the possibility of organizing a major change in the style of the papers themselves.

My recent research has once again led me into a subject new to me, so I have recently spent some time reading original references from the physics literature of the 1930s, 40s, and early 50s. The startling difference I could not help but notice between the writing style of many of these older (and often famous) papers, and virtually all the modern papers I see is that they were SHORT -- only one or two pages! Shorter than what we call Letters today! They have no introductory or concluding discussion -- going directly to the subject and the new results being presented. Of course, this style is contrary to the one I learned in school (``Tell them what you are going to tell them. Tell them. Tell you what you told them.'')

I have used the Introduction/Body/Conclusion style repeatedly myself with some success and must say that I still think highly of it; however, I am about to reach the conclusion that in general we can no longer afford to use this style for all our papers. The Introduction/Body/Conclusion style is to some extent triply redundant. Perhaps if we worked harder at saying what we have to say well once, then the other two times might not be necessary.

Why do we use this long-winded style when our forebears were able to present their results successfully in a more concise format? Part of the problem arises no doubt from the huge difference in the number of scientists working today, and the resulting diversity of their background and training. Also, much modern research is interdisciplinary. My own work is interdisciplinary, so I try to communicate new results to an audience including not only other physicists but also fluid dynamicists, acousticians, geophysicists, engineers, mathematicians, and computer scientists. The audience constrains me to avoid the use of specialist jargon as much as possible. However, since the purpose of jargon is to abbreviate discussion of commonly understood topics, the avoidance of jargon implies that the paper must be somewhat longer. Authors writing noninterdisciplinary papers for journals like Physical Review Letters are forced to deal with the same constraints by the requirement that all papers appeal to a ``general (nonspecialist) audience.'' Although difficult to prove, I submit that the goals of writing ``shorter papers'' and writing ``papers appealing to nonspecialists'' are mutually exclusive. Thus, both the subjects written about today and, in some cases, the journals themselves force papers to be longer. Writers, editors, and referees all need to reexamine their assumptions about the writing style the average paper should use.

My recommendation? Clarity without redundancy. Or at least without so much redundancy. If an author is going to write just one paper on a given topic, then he/she should be allowed/encouraged to write the clearest possible paper - that probably means using the Introduction/Body/Conclusion style. However, if the author foresees a series of papers of the same subject, she/he should dispense with the ``self-contained'' criterion in most of the papers. I think this modest proposal could easily reduce the number of journal pages by a factor of two, and possibly more if everyone played by the same rules.

I am willing to try it if you are.

Sincerely,

James G. Berryman

Danville, CA

---

E-mail to berryman@farne.stanford.edu