Evaluating publications remains a significant factor in characterising research groups. In France, the process of evaluation falls under programme P150 of the organic law relating to the finance laws (LOLF, loi organique relative aux lois de finances). The programme’s main objectives in defining how documents should be evaluated (measuring performance by publication, technology transfer, and international co-operation) have been adopted by the Observatoire des sciences et des techniques, using its own indicators. The two-year analysis model recommended by the Institute for Scientific Information (ISI) favours applied research, to the detriment of the humanities, whose modes of publication are ill-suited to the kind of scientometric methods developed for the fields of science, technology, and medicine. Information and communication professionals need to work with the scientific community to raise awareness of such issues, leading to discussion of necessary changes to the methodology and professional ethics of evaluation practice in France.