Abstract:Publishing is an important cultural industry, and its products form part of the record of our social and cultural history. Yet what of the records of the publishers themselves -the voluminous correspondence, financial information, manuscripts, policies, review reports, and so on -and what is preserved of such records? In this paper, I propose to discuss the ways in which South African publishers keep -and do not keep -archives. If we are to write a full publishing history in this country, then it is imperative that there should be archives, records, sources of information. Yet publishers, while concerned with record-keeping for corporate purposes, turn out to be not particularly good at keeping records for posterity, for examination, or for independent study. A key problem is access to source material for further study, as often it has not been retained, or certain corporate archives are closed to outsiders and considered confidential. This problem is the result of a tension between why such archives may have been compiled for 'management information', precedent, or compliance and what users such as historians or other researchers may seek in them. There is a core difference in what each party regards as valuable. In other words, we may view publishers and their products as part of a cultural industry, but they largely view themselves as business entities. Another problem with sources results from the repressive legislation of the past, which banned books, suppressed authors and constrained publishers. The effect was to create gaps, omissions and even excisions in the archives that survive. This paper aims to examine which kinds of traces survive, and under what circumstances, in the archives of South African publishers, using a case study approach. As part of a larger, ongoing project, this would be of use to researchers who are seeking to write publisher histories, as well as to publishing houses with haphazard policies for sorting and retaining records for archival purposes.