The recent commentary by Philip Wyatt discusses authorship problems with both ideas and data in scientific papers (Physics Today, April 2012, page 9). I have some thoughts to add.
Scientific reports and journal publications represent a huge intellectual effort, the final result of an enormous underlying intellectual structure. Among many and various contributors to the research, who should join as coauthors? Those who publish know a great deal about various types of coauthors, be they “courtesy,” “ghost,” “gift,” “honorary,” “surprise,” or “token.”
For years, major professional societies and mainline journals remained conspicuously silent on the subject of authorship guidelines. The American Physical Society, for example, offers only an insubstantial one-sentence statement: “Authorship should be limited to those who have made a significant contribution to the concept, design, execution, or interpretation of the research study.” “Significant” joins “interesting” and “important” among the most meaningless, overused descriptors in scientific writing.
In recent years the biomedical community led a large effort to develop and popularize useful authorship guidelines. About 500 biomedical journals, including the more general journal Science, adopted guidelines for submitted manuscripts. Called the Vancouver criteria, the guidelines were put together by the International Committee of Medical Journal Editors and contain three principal requirements, paraphrased below.1
‣ Coauthors should generate a significant part of the manuscript’s intellectual content (there’s that word “significant” again).
‣ Coauthors should contribute to the manuscript, either writing the first draft, reviewing it critically, or revising its intellectual content.
‣ Coauthors should review and approve the final manuscript before it goes to the publisher.
To be listed as an author, a person must meet all three of the above criteria. Fundraising, general research-group supervision, data collection, routine calculations, routine help with manuscript preparation, and other efforts are insufficient for authorship. As a biomedical researcher told me, “Holding the door open to bring the mice into the laboratory does not entitle you to coauthorship.” With few exceptions, lending samples, other materials, laboratory equipment, and funds fails to qualify the lender for coauthorship. Lending a good idea may justify coauthorship, per the first requirement above. After all, the continued progress of science depends mainly on developing or discovering new ideas or new interrelationships among familiar concepts.
From the above criteria, there follows another guideline:
‣ Coauthors should be able to present and defend the manuscript’s content to an audience of their peers.
Some parallels may be made between the modern-day science-paper authorship issue and renga, a form of communal linked verse promoted by Basho in 17th-century Japan and by other master poets. A scribe would record the renga session, noting three items: the verse sequence number (1, 2, 3 . . .), the verse (for example, “a morning full moon, no noisy crows”), and the verse’s author. During the versing, exceptional contributions received extra recognition, perhaps even from the master poet present. At the session’s end, after 36 to 100 verses (linked by strict rules) were recorded, the scribe revealed how many verses came from each participant. A successful renga session required many contributions, from the castle’s host, the ladies who played music and served refreshments, the scribe, the audience, and even cooks, groomsmen, and stable boys. But only those who stood and contributed a verse that was accepted into the final intellectual record had their names brushed in charcoal ink onto rice paper.
