Sir

You report (Nature 389, 105; 1997) that Dr Friedhelm Herrmann's lawyers argue that as senior author he would not have had any responsibility for possibly fabricated data and, moreover, that a full professor would not have a motive to commit fraud. Both statements are ridiculous.

Any scientist has to publish, or reputation and grant-money will be seriously endangered. The only way to publish for scientists no longer doing bench-work themselves is to guide research and be senior authors. A senior author must be responsible for the scientific soundness and honesty of a paper, or he or she should not be in this distinguished place. They get reputation from it after all, and the quality of a paper is often inherently inferred from the name and status of the senior author.

The minimum requirement for senior authorship should be that the paper has been read, understood and worked on by intelligently and critically discussing it with the people who did the actual bench-work. In a perfect world, the senior author will also have provided ideas and mental input through the course of the experiments and have seen the problems and difficulties of the study. There are heads of laboratories who provide just that, and are therefore rightly named the senior authors. I do not think that — as is sometimes the case — being head of a department or institute, providing laboratory space, or simply giving permission to use costly equipment as such justifies senior authorship. Unfortunately, there are no real rules, but it is time that there were.