Re-examining Price's conjectures on the structure of reference networks: Results from the special relativity, spatial diffusion modeling and role analysis literatures

In his influential article 'Networks of Scientific Papers', Derek Price used data on the N-rays reference network to exemplify his argument that natural science research literatures overcite recently published papers. In subsequent work, he further argued that this tendency is weaker in social science literatures, and may be entirely absent in scholarly literatures in the humanities. We report results from a replication of Price's N-rays analysis and data for three additional reference networks: special relativity theory, spatial diffusion modeling and role algebra analysis. Our analyses indicate that the N-rays reference network provides little support for Price's conjectures, but that those for the other three areas are generally consistent with them. We find, however, that the two social science literatures exhibit structures more closely resembling the pattern that Price claimed to be characteristic of the humanities, and suggest that the variety of structures that reference networks exhibit may be greater than Price anticipated.