Sir

Giulio De Leo et al. draw attention to the debate in Italy about the reform of research policy (Nature 391, 12; 1998). They also show the results of a bibliometric analysis of papers in environmental research produced by the universities, CNR, ENEA and other scientific institutions between 1981 and 1996. From these statistics they conclude that the ENEA production is low in numbers and quality.

But ENEA (Italian National Agency for New Technology, Energy and the Environment) is not a typical research institution, because it exists to produce know-how as well as to provide services (under contract), advice and support to the public administration at national, regional and local level. ENEA therefore also produces technical reports and other written and electronic material that is not included in citation indexes.

The field examined by De Leo et al. is limited to environmental research. As the title of our agency indicates, environment is only one of ENEA's fields of interest. As a consequence, even within its environment department, research projects and the publications derived from them could easily escape this categorization, because they address areas such as radiobiology, toxicology, geology and numerical modelling, and often yield high-quality publications that are not always well represented in the journals examined by the authors.

The environment department of ENEA was created in 1994 — or in 1989 under a different name and with different staff. Neither date corresponds to the period covered by De Leo et al.

We feel, therefore, that a comparison cannot be made in the oversimplified way followed by the authors. Their approach is flawed by differences in role, history, size, manpower characteristics and fields of publication of the institutions they analyse.

A discussion of the role of ENEA and of the other Italian research bodies is timely and necessary and should consider the distribution of manpower and of financial resources, two items on which discussions should be started and action taken, and De Leo et al. are going in the right direction.