In recent decades, the enormous growth of publication, especially in the sciences, has created critical problems for the users of information and for agencies that provide it. Solutions have been sought in the use of electronic, mechanical, and photographic aids to information handling. The author, while recognizing their value, believes that there is a more basic need for greater selectivity in the building of research library collections, based on an understanding of subject literatures as phenomena – their creation, behavior, and use. He draws concepts from the “new” discipline of information science – including statistical discoveries of Bradford, Zipf, and Booth, as well as epidemic theory, citation tracing, and the research front – and applies them as tools in analyzing literature of information science itself. His purpose, he explains, is to begin a synthesis of these techniques and to show relationships among the elements of knowledge they elicit concerning the structure of the literature and the processes occurring in it. He holds that “an understanding of literature's formal properties (e.g., characteristics of production and use) rather than the understanding of its conceptual content offers the best means for prediction and control in the management of library resources.”
In demonstrating the use of the techniques, Donohue applies them to a main corpus of journal articles in information science, as well as to the literature cited by them, and to the writings that in turn cite the main corpus. After analyzing the results for what they show about the discipline itself, he goes on to develop a general method for the use of the techniques in library management, especially in the problems of journal selection.