Sir

F. T. Krell in Correspondence (“Impact factors aren't relevant to taxonomy”, Nature 405, 507–508, 2000) suggests that taxonomy is subject to different “regularities” from other fields. He uses the allegedly small number of entomology, biodiversity and taxonomy journals covered in the Science Citation Index (SCI) as the reason for low impact factors. But he provides only anecdotal data on the size of this literature. How many articles are published by the low-impact journals at the Natural History Museum?

Krell refers to Bradford's law of scattering, but he needs instead to provide data to show that entomology–taxonomy diversity is somehow different from other fields. Bradford's law simply suggests that if there are, say, 1,000 journals in a field, then one-third of the papers are to be found in each of three zones containing about 10, 100 and 1,000 journals, respectively.

A critical element in determining the impact factor of a field is not the number of papers it publishes, but the citation density of the average paper and the half-life of the references cited. Adding more journals to the SCI database would not increase the impact because the increased number of cited references would have to be shared by more published papers.

The data on the 65 entomology journals covered in the SCI indicate that their impact factors are not significantly different from other fields with long half-lives (see the Institute for Scientific Information's Journal Citation Reports, http://www.isinet.com/isi/products/citation/jcr/?version=1). However, the size of a field does affect the number of super-cited papers that will be published.

Krell states: “Qualified referees must evaluate the scientific work itself.” But this is equally true of any field. Why shouldn't such evaluation be supplemented by citation analysis? The question is whether the referee has any basis for comparing articles, authors or journals across a wider horizon. Taxonomy, small as it may be, is not without its 'citation classics', as the work of R. Sokal, E. O. Wilson and others demonstrates. Their work is cited by thousands of papers covered in the SCI, by taxonomists and by other scientists.