There was a time not too long ago when the doctoral student was expected to read essentially all of the previously published literature pertaining to a given field and then to continue reading currently all that was published. It was assumed that this procedure was necessary in order to provide assurance that the productivity of the graduate would not be blunted by lack of knowledge of significant developments.
Accordingly, when reference to previously published literature seemed important, the graduate, now professionally active, could rely on memory to specify subjects which could be used to recall relevant material from libraries or personal collections. This type of "information retrieval" tended to be based on a recall mechanism which generally permitted the rapid identification of the desired materials.
A subtle change in the recall mechanism began to take place as the volume of published literature reached the point at which the professional could
Kent A. Computers and Biomedical Information Storage and Retrieval. JAMA. 1966;196(11):927-932. doi:10.1001/jama.1966.03100240061014