Sir

Christopher Stubbs1 claims that science is best served by alphabetical ordering of authors. Is this really the only alternative? Why not rotating alphabetical authorship and reverse alphabetical order to alleviate ‘alphabetic disorder’2?

The earlier letter on surnames by Mark Shevlin3 reminds me of a well-known British professor's first encounter with the Genetics Citation Index in 1963. He could not comprehend why his name did not appear in it. Because he usually deferred to his junior colleagues by placing his name last on bylines, his name did not turn up. Later, when we launched the Science Citation Indexin 1964, his name was included in the Source Index where all authors’ names are included and cross-referenced to the first author.

First authorship is mistakenly considered essential if one is to be fully credited in the citation game. But any informed citation analyst relies on an author's full CV to determine the first author of the papers to be included in the citation count.

The first comprehensive ‘all-author’ study did not occur until 1978 (ref. 4). Previously, ‘first-author’ studies were the norm. Nevertheless, many administrators persist in using the first-author data found in the printed Science Citation Index, disregarding its potential bias towards second authors. The Web of Science, an expanded version of the SCI, overcomes this problem because all authors’ names, thanks to hyperlinks, are displayed in the Citation Index section.

But what about the absurd ‘no-author’ policy for leading articles published in Nature and other journals? Is this fiction designed to impress readers with the journal's authority?5 Are these editorials written by robots or people named ‘anonymous’?