Sir

Recent correspondents1,2 have suggested that authors with initials earlier in the alphabet are more likely to be cited simply because they are more numerous, a fact borne out by examination of the distribution of surnames in the London telephone book.

However, comparing the citation rate of each letter with the number of telephone users of each letter is an inferior attempt to control for differences in the representation of initials to the method employed in the original study.

The correlation originally identified3 was between the number of citations for each paper published by authors of a particular initial and their surname's alphabetic position. This controls for the uneven distribution of initials among authors — if Zs are rare, they will publish fewer papers.

Perhaps the crude technique of using Londoners as a control removes the correlation because only a subset of papers are published by people with English names. Clearly there is still a case to answer, and given the growing importance of citation indices, one deserving of serious research.