Sir

Your Special Report on the software that journal editors are considering to help them catch academic cheats (Nature 435, 258–259; 2005) suggests that six words used contiguously in more than one published paper now constitutes plagiarism. I design novel oligonucleotides to inhibit STAT3 activity in hormone refractory prostate cancer. I use this phrase of more than six words to describe what I do in the introduction to my manuscripts. Am I commiting self-plagiarism?

Plagiarism must absolutely be defined not by words used but by data shown. That is the serious offence, not that someone reuses a key descriptive phrase in several papers.