Mapping the Structure of Science

Conceptually, there is not much distance between diagraming the intellectual structure of an historical line of research and doing the same thing for the whole of contemporary science. If, as the DNA-history study showed (l), citation analysis can be used to define the intellectual relationships between past research events, there is no obvious reason why it cannot be used to do the same thing for current ones. In other words, what the DNA-history study validated was not just a methodology for exploring the history of science but a methodology for defining the intellectual structure of science, past or present. The structure of science is an intriguing subject to those who study science as a system. Beyond the intuitively comfortable hypothesis that science is a mosaic of small units, rather than a structural monolith, there are many more questions than answers. What is the nature of the basic units in the mosaic structure? How do they relate to each other? Are the intellectual and social structures similar? Are they made up of the same basic units? What is the relationship between them? Is there a variety of configurations at the infrastructure level? How dynamic are the configurations? Is there a relationship between configuration and research performance? Can structural analysis help us make more effective science policy decisions? A number of studies during the first half of the 1970s began examining some of these questions in a systematic manner. Most of them provide evidence that specialties are the basic intellectual and social unit of the scientific mosaic. Crawford (2) described how the specialty of sleep research is built around the communications between a small group of key individuals and research centers. Griffith and Mullins (3) have shown that small, socially coherent groups are capable of producing major changes within disciplines. Crane’s work on specialties in rural sociology and mathematics has led her to the idea of specialties forming a structure that can be mapped (4).