1 The costume section of the Interpretation Branch of the Canadian Parks Service headquarters office in Ottawa was established in 1972 and since then has undergone a number of changes which reflect the changing focus of the branch as a whole.
2 Originally, the in-house curators and designers undertook not only site-specific costume research and planned costume programmes, but produced the costumes worn by interpreters at some national historic sites. By 1979 the number of sites using costume had increased to the point thai the resources of the costume unit were stretched to the limit. The increased demand for costumes resulted in the creation of a costume resource centre at the headquarters office. At present, there are approximately 1000 costumed interpreters at 57 historic and national parks across the country, representing periods and themes that range from eighteenth-century French military establishments to an early twentieth-century prairie farm.
3 Between 1984 and 1986 three of the regional Parks Service offices hired costume curators. The responsibility for conducting site-specific research and for monitoring the production of costumes by independent contractors was gradually decentralized to the regional level. In addition, military curators in each region are responsible for costume matters related specifically to military costume. Fortress Louisbourg, with a costumed staff of over 100 military and civilian interpreters, is the only site with its own costume curator.
4 At the headquarters level, our staff includes two costume curators, a costume designer and a costume resource coordinator. Occasionally, contract researchers and designers are hired for specific projects. Though we are no longer directly involved with site related matters we provide guidance on national concerns, including the development of costume standards and guidelines, the operation of a research resource centre, and the compilation of interpretive activity resource manuals and source lists of costume reproduction suppliers and contractors. We also respond to hundreds of inquiries each year from outside agencies, researchers and members of the public. In addition, the costume unit continues to initiate and conduct a variety of research projects.1
5 The changes in the direction of the branch are reflected in the nature and focus of these projects. Though our research is no longer site specific, we focus on broader social and historical themes and provide "deep background," which, in the hands of the regional curators, can be applied at site level. For example, our research on protective foul-weather clothing has application to sites throughout the system offering winter programmes.
6 Another example of one of our recent major projects is our ongoing work to produce basic patterns, prototypes and research reports for a broad range of nineteenth-century men's and women's garments which can be used at appropriate sites across the system or as new costume programmes are initiated. Our intention is to represent fully the range of garment styles worn during the nineteenth century with patterns adapted to modern standard size ranges. The accompanying prototype garments set the standards for the construction methods used in the production of these costumes.
7 Due to the practical limitations of clothing hundreds of interpreters, individual original garments are seldom copied stitch-for-stitch, except for display or education programmes. Instead, principles and methods of clothing construction consistent with established practices of nineteenth-century tailors, dressmakers and home sewers are applied to authentically styled and cut patterns. Rather than focusing exclusively on changes in style, our recent research has emphasized the technical aspects of garment cut and construction, which has been given surprisingly little attention in the literature. Often, research on garment construction and production is limited to descriptions of individual garments, or is focused on the labour movement within the garment industry, not on garment construction.
8 This specific goal of producing authentically cut and constructed garments necessarily directs our attention to original garments. Our costume documentation files, which currently include information on several hundred garments in collections throughout Canada, England and the United States, as well as in the Parks Service collection, describe the garments in detail, from both a stylistic and technical perspective. These details are analyzed to establish the predominance of specific styles and the use of particular construction techniques. This information has also been useful in developing dating chronologies. Unfortunately, everyday working class garments and those with known provenance are more rare than we would wish, and costume collections are consistently weighted toward the middle and upper classes. As a result, written and visual documentary evidence plays an equally important part in our research strategies, especially when original garments are lacking.
9 These sources include paintings, drawings and photographs, inventories, letters, trade and mail-order catalogues, travel literature, journals and memoirs, newspaper advertisements, business records, registered designs, periodicals, and dressmaking and tailoring construction manuals and pattern-drafting systems.
10 Though some public repositories are rich in resources most are weak in some fields, notably in the area of fashion and garment trade periodicals. Fortunately, in recent years the Parks Service has assembled a large collection of original nineteenth-century women's fashion, needlework and homemaking periodicals which enjoyed widespread Canadian readership. A network of devoted men's costume historians in England and the United States has provided us with copies of rare British and American tailor's periodicals subscribed to by Canadian tailors. In addition, we have a good collection of nineteenth-century trade catalogues and dressmaking and tailoring construction manuals.
11 The costume research conducted by the Interpretation Branch differs from that done in either a purely academic or a conventional museum setting in one very specific way: in a sense our "conclusions" are already drawn for us. No matter where our research might lead us, our research strategies and, consequently, the methods and sources we use must always conclude with the same end product: period clothing to be worn by the interpreters at the historic sites. We cannot limit our methods to prove a methodological point, nor can our sources be restricted to those which can be neatly manipulated. Source limitations must be compensated for by consulting as wide a range of sources as necessary to provide the required answers. In some respects our approach tends to be closely allied with that of historical and experimental archaeologists. We are concerned not only with conducting costume research but with applying it.
12 Within the context of a re-created historic environment, costumes are expected to contribute to the interpretation of the site in the same way as other objects. Since a costumed interpreter is often the first point of contact for the visitors to a site, the relative impact of costume may be disproportionately great. The same stringent standards that apply to the reconstruction, restoration and furnishing of historic buildings apply to the costumes worn in those buildings. Our costume research reflects the contribution period costume makes to our interpretation of the past.