Victorian Data Processing

This sounds great. When I wrote my dissertation, I was aware that "computer" originally referred to a human agent, but didn't really know that such computers had already been organized into large scale distributed processing networks.

Victorian Data Processing - When Software Was People
Martin Campbell-Kelly

To most people the phrase Victorian Office conjures up an image such as in Dickens' A Christmas Carol - with Bob Cratchit, the solitary clerk, seated on a high stool, quill-pen in hand. Indeed, many Victorian Offices were like this, but by the 1850s a quite different type of office was emerging - the industrialized office employing several hundred clerks. These offices were the ancestors of the modern computerized bureaucracy. In these huge organizations, clerks performed tasks that would later be done by office machines, and are today performed by computers.
In this paper the historical and economic context of the development of the industrialized office will be described. The paper will include a "tour" of a number of Victorian data processing organizations: the General Register Office, the Railway Clearing House, the Central Telegraph Office, the Prudential Assurance Company, and the Post Office Savings Bank. The data processing techniques and labor processes will be explained, and the changing gender and structure of the workforce described.
The Victorian Office can be viewed from today’s perspective as a collection of human agents, obeying procedures stored in the organizational memory. Parallels are drawn between today’s software and Victorian clerical processes in terms of dependability, evolutionary change, and the boundaries of determinacy and autonomy. Some general conclusions will be made about the nature of "information revolutions."

From a workshop on the origins of computation.

Unfortunately in a truly retro move, the paper does not seem to be available online without a subscription.