The documentation literature of the 1950s is reviewed briefly, and some early text processing endeavors are discussed. Various predictions made in 1960 by Mooers about the creative role of computers in information retrieval are then considered, and an attempt is made to explain why some of the more exciting predictions have not been fulfilled. Conclusions are drawn concerning the limits of computer power in text retrieval applications.
1. The Documentation Literature in the Nineteen Fifties
The modem digital computers were developed in the 1940s and early 195Os, and the first commercial use of computers dates back to the early part of the 1950s. The information retrieval literature of that period was largely concerned with the description of novel hardware search systems capable of automatically searching through stored document files, and with the coordinate indexing approach in which combinations of keywords attached to text items are used to replace the more conventional classification schedules and subject heading catalogs.
Among the hardware retrieval environments that received extensive discussion in the early literature were punched card systems where document identification and keyword information was recorded on paper cards. In some systems, topic cards were used. In that case each card would represent a particular subject term, or topic area, and the card information itself would specify the identifiers of documents carrying each particular term. Documents corresponding to a set of distinct terms could then be identified by placing the relevant topic cards in a stack on top of each other, and using the "peek-a-boo" method to identify card positions with coinciding punched holes on every card. In the implementation using topic cards, each card is effec-