Sir

Your Special Report on plagiarism, “Taking on the cheats” (Nature 435, 258–259; 2005), does not, in my opinion, appreciate the way in which scientific research evolves.

In my experience, incremental progress is often reported at one or more scientific conferences, until a comprehensive manuscript can be submitted to a reputable journal for publication. The work may be further published in research monographs, in review articles and, on occasion, in textbooks.

It seems to me a misunderstanding to insist that a piece of work must be published only once.

A series of progress reports would naturally have extended sections in common — if this were not allowed, no scientist would consider presenting work at a conference with published proceedings. Increasingly, conference papers are being published as regular books and sometimes even as special issues of standard journals. Is it not defeating the scientific purpose of conferences if final papers are expected, rather than discussion papers on work in progress?

Similar conflicts can arise between journal and book publishers. I submitted a paper to a scientific journal and later incorporated a description of it in a research monograph. However, the efficient book publisher got the monograph on the street two months after receiving the manuscript, whereas the journal turned out to have a backlog resulting in accepted papers waiting more than a year to appear in print. In this instance, could I be accused of committing ‘self-plagiarism’ on the basis of overlaps, when I have no control over the schedules of the publishers?