The fifty-year span of the Journal of Documentation coincides very precisely with the lifetime of the modern computer. When the Journal began in 1945, the first recognisably modern, working computer was a year or two away (the Manchester Mark 1 first ran in June 1948), although the foundations had been laid (the von Neumann architecture was defined in 1945).

It did take the Journal a little while to discover the computer as such: it was not, in general, interested in either the machinery itself or the theory of the machinery. It was, however, interested both in practical devices and in theories that might relate to information phenomena or to the handling of information. From the point of view of the practical devices, there is of course a considerable degree of continuity between other forms of mechanization (some already well-established in 1945) and the use of computers. Thus for example early issues of the Journal carried quarterly `Documentation surveys' -- a classified, annotated bibliography of recent publications. The early heading `Punched-card techniques' became `Mechanization in Libraries' in volume 3, September 1947, and thereafter included a number of books and articles on computers in retrieval and other aspects of information work.

In this paper, I will try to pull out and comment upon some of the themes and ideas from the history of the development of computer-based methods in information retrieval, of relevant models and theories, and of the evaluation of retrieval systems, as seen through the pages of the Journal. I attempt to represent, at least in overview, the major concerns revealed in the Journal; nevertheless, the selection of material, and particularly the comments thereon, reflect my own personal biases. Although I cover, at least in general terms, events and discussions taking place outside, references will be entirely to items (papers and reviews) appearing in the Journal. Clearly these were not always the first, nor necessarily the best, examples of their type, but they reflect a particular view of the world. The bibliography is listed strictly in chronological order, rather than in the order of reference in the text; thus as I follow particular themes through their own time-sequences, the reference numbers will indicate roughly the chronological relations with other themes.