Citation Analysis as a Method of Historical Research into Science

The use of citation analysis in research on the history of science is based on a literary model of the scientific process. In this model, scientific work is represented by the papers written and published to report it, and the relationships between discrete pieces of work are represented by the references in the papers. Price, one of the Ieading contributors to the model, has taken this view of the scientific process to the point of defining scientific papers as the chief product of a scientist’s work, and a scientist as one who writes scientific papers (1). Though the literary model is certainly a gross simplification of the scientific process, it seems to provide a functional view of that process that is both accurate and useful. Price has used it in a series of studies that have produced a number of insights into how science works and the ways in which it differs from, and interacts with, technology (l-4). Small and Griffith have used it to define the specialities that make up the leading edge of scientific development (5,6), and I have used it to clarify the interactions between broad fields of research (7). The accuracy and productivity of the model for historical studies was tested and proved in 1964 in a study conducted for the Air Force Office of Scientific Research (8). That study was concerned with determining whether citation analysis could be used to develop an accurate and useful network diagram of the cumulative research that led to a given scientific breakthrough. The idea was not entirely a new one. Bernal had used the network diagram technique in 1953 to show the antecedents and consequences of Pasteur’s discovery of molecular asymmetry but had not based it on citation analysis (9). The potential usefulness of references for historical research was suggested in 1955 (10). Then, in 1960, Dr. Gordon Allen put together the two ideas of references and diagrams with the illustration shown in Figure 7.1. A picture of the chronological and citation relationships among the papers in a bibliography on the staining of nucleic acids, the