5 Corporate records for manufacturing firms producing articles of costume, and for retail or wholesale outlets, are often of great value to costume historians. The category "corporate records" includes any documentation held in a company's archives. Archives is in this context a generous word, meaning anything from a professionally run archives with staff and research facilities to perhaps a closet in an underground parking lot, or a drawer in the president's desk. If a researcher is fortunate, companies have retained some information about their own history. However, one of the limitations of this source is that whether due to lack of interest in their own past or due to factors such as fires, frequent moves, insufficient space, takeovers and mergers, or the present and future oriented thinking of businesses, the records of many firms have been obliterated over time, and exist neither in public archives nor in company offices.

6 Even when records have been retained in company hands, there are a number of difficulties in accessing and utilizing them. First is their inconsistency. The "hit and miss" nature of retaining archival records has resulted in a situation where one firm may have interoffice memos, order books, dated garment labels, samples, annual reports, correspondence, etc., spanning its entire history of operation, while another firm may have an insurance appraisal from one year, and two or three photographs (Fig. 2). It is therefore impossible to attempt any sort of meaningful comparisons between the two firms on the basis of data compiled from these sources. Second, they may not have been organized in a way that is useful to researchers. Researchers interested in corporate history often would benefit from training as archivists because it may be necessary to do a considerable amount of sorting and filing before being able to seriously analyse the data.

7 Once the data has been identified and sorted it should be treated with the cautions identified further below. An annual report then as now served several functions; it recorded the major achievements of a given year and provided financial statements. However, it may also have been used to "sell" the company to shareholders, government officials, or in some cases, staff or unions. An understanding of the operation of a garment manufacturing firm or of a retail or wholesale business, which may be obtained from annual reports, may improve one's understanding of the garments themselves and how they were interpreted at the time in question. For example, the Emery Manufacturing Company began operation in Edmonton in 1912 as a manufacturing firm supplying Ramsey's department store, a local shop interested in providing Edmontonians with locally produced clothing with "New York styling" (both because of a desire to initiate and support local industry and because of the firm's ability to acquire and therefore market clothing at lower prices). In 1914, Emery's became independent of Ramsey's and expanded its production into everyday ladies wear.3 Questions of availability, pricing, styling, quantities, and the company's solvency, etc., are some of the issues which may be addressed in annual reports.

8 Inter-office memos may have been written to protect individuals from criticisms, such as that they were producing "knock-offs" or copies of garments produced by other firms. Correspondence with individuals from outside the firm, other businesses, or government agencies may address a broad range of issues, among them questions of style, cut, construction, production, etc.

9 Production records, order books, salesmen's records, and correspondence with customers are extremely useful because they indicate what was actually purchased. In contrast to advertisements and catalogues, which only indicate what was available, this sort of material can be used to determine what was sold and, sometimes, how well it was received. Unfortunately, financial records contain little information about styles, materials, etc., so the price lists and sales figures may mean very little without further illustration of what was actually being accounted for. Ideally this information should be compared to advertisements and catalogues to enable both sets of data to be more accurately interpreted. All too often only one set of data is extant.

10 Advertisements can be very useful in documenting style, construction and fit. However, they should not necessarily be interpreted literally. Researchers, in common with today's consumers, must be suspect of a firm's claims. Costume illustrations may exaggerate features; written description may omit essential details. The actual product may not have been exactly as the advertisement suggested. Some advertisements from the period have a comical appeal which may overshadow our ability to examine them objectively (Fig. 3). These points will be discussed further under mail-order catalogues, which are a specific form of advertisement.

11 Aspects of costume as social history are also revealed through close scrutiny of advertisements. For example, an advertisement for GWG's "Blue Diamond" overalls in Alberta Labor News describes a unique relationship between the GWG firm and railway-men that evolved because of requests for particular features in the design of overalls (Fig. 4).

These overalls had a number of special features which were patented. The patent records support the claims about the unique features of these overalls. Unfortunately there are few overalls extant in museum collections and it has not been possible to compare the garments to the advertisements. In terms of the costume itself, therefore, it has not yet been possible to determine whether or not the claims made in the advertisements are true.

12 Of perhaps greater interest to social historians is the fact that the advertisement describes an effective, collaborative relationship that existed between the consumer and the manufacturer. Critics of high fashion complain that designers pay little attention to their consumers, yet here, apparently, is an example of occupational clothing being designed in concert with the people who would eventually be wearing the clothes. Researchers should be able to verify this claim through inquiries among railwaymen of the period. They could then pursue the ideas suggested by this relationship, consider to what extent it was developed by others in the industry, and perhaps why it was not developed further.

13 One feature, common in advertisements in this period, was the promotion of the firm as stable and prosperous, a part of the growth of the western frontier. Company plants were frequently photographed at an angle intended to accentuate this impression. An advertisement in Farm and Ranch Review shows the GWG building from the corner and the perspective makes it look larger than it really is (Fig. 5).5 The view of a strong man in the foreground and various settlement images in the background (breaking the land, a train, a grain elevator, etc.) was intended to make one proud to be a Westerner and to encourage men to buy GWG goods. Consumers were buying the image as much as they were buying the product. It was effective but not entirely accurate.

14 Mail-order catalogues are a popular source for material historians studying the late nineteenth and twentieth centuries. Unlike some of the other sources under discussion, they are easily accessible. Catalogues are readily available on microfilm, a limited number of reproductions exist, and many museums have collections of original catalogues for reference purposes. They are useful because they contain numerous line drawings, often some colour plates, pricing information, written descriptions, etc. They give an immediate, visual image of the range of goods that was available to customers through mail order.

15 Because they were aimed at families in the lower economic brackets, they provide examples of garments that can only rarely be found in museum collections. Everyday clothing is generally not preserved either by individuals or by institutions, a situation which museums have recognized and addressed since their attention to social history was reawakened in the 1960s.6 Mail-order catalogues therefore help to provide a balance to the better quality garments in museum collections.

16 One method of examining catalogues to determine what information they can reliably provide, which has recently gained popularity, is to undertake a content analysis.7 As a case study, women's garments on the pages of the Eaton's Spring and Summer Catalogue for the years 1920, 1923, and 1926 were analysed and compared (Fig. 6). Few construction details were included in the descriptions so some features were analysed based upon visual examination of the illustrations. For example, the location of garment fastenings was usually inferred from the illustration and previous knowledge of period construction techniques. Descriptions occasionally specified that garments "fastened at the side" but the type of fastening used was not mentioned. The information provided in the catalogues is much more fragmentary than a superficial glance suggests. Researchers studying the pages of catalogues for purposes such as costume reproduction are left making of number of suppositions on the basis of their own expertise.

17 Generally the intended occasion of use for dresses was not mentioned. Occupational clothing of any form was rarely mentioned. One dress was referred to as "very suitable for maids' or nurses' wear."8The vast majority of dresses appeared to be either house dresses or afternoon dresses. Perhaps the reason that fewer distinctions were made was related to the increasing number of washable dresses as well as to a less rigid social climate which no longer dictated that a woman had to change her clothing half a dozen times a day. But it is also clear from a close examination of these pages that Eaton's was not trying to appeal to the fashionable set.

18 Dresses were available in a broad price range, although the majority of them were under $15.00. The least expensive dresses were wash and morning dresses under $5; only a couple of dresses were available in the most expensive price range ($25-$29). Eaton's sold fewer expensive dresses as the decade progressed. Although there was a correlation between occasion of use and price, there was no similar relationship between occasion of use and length. One might expect the most fashionable dresses to be either all longer or all shorter but there was quite a variation of length, particularly at the beginning of the decade.

19 A number of the less expensive models were produced by Eaton's but Eaton's did manufacture garments in all price ranges. Although Eaton's had a large garment factory, which has gained notoriety because of labour disputes, relatively few of the ladies dresses in the catalogue were identified as having been made by Eaton's. In 1920, only 15 of the 76 dresses available were identified as "Canadian Made/Eaton's," suggesting that the remaining 61 were imported, probably from the United States. In 1923, they did not specify the number of Eaton-made dresses and in 1926 only five dresses were identified as Eaton-made. However, what this means is unclear. It may reflect the marketers' concept of the value of such forms of boosterism rather than the actual number of Eaton-made dresses. If this is true, it suggests another limitation of the value of mail-order catalogues as a source.

20 The data compiled through this systematic approach amounts to dozens of pages of minute detail. Included among the findings were a number of important observations. For example, although the text in 1920 referred to round necklines as being the most fashionable,9 in fact there was little difference between the number of round necklines and the number of square necklines - 33 to 29 respectively. There were noticeably fewer V necklines - only 14. This supports the idea that the claims made in the catalogues may not be substantiated.

21 Another observation is that the questions asked of the catalogues must be carefully developed. For example, in trying to determine the most commonly available colours of the period, if one listed all of the exotic names given for basic colours (i.e., "Copenhagen," "cadet" and "sky" for blue), the sampling would be too small to identify trends clearly. By combining them under headings of common colour names, it is possible to determine the most frequently used colours.

22 When considering the usefulness of this information it is important to remember that a content analysis of data from mail-order catalogues only indicates the range of material that was available from this supplier. It does not necessarily indicate what was the most popular style, colour or fabric. Unfortunately, records of purchases from Eaton's during this period are unavailable for comparison.

23 The illustrations can be considered within the context of the history of fashion design to determine to what extent high fashion filters down to the average person. For example, the 1920s are thought of in terms of flappers and beaded dresses but the author's content analysis of Eaton's catalogues from 1920, 1923 and 1926 showed that very few dresses with any applied beaded decoration were available through the catalogues. The content analysis does allow researchers to state, with confidence, views that a more subjective or cursory examination of the catalogues might have suggested. It could also provide contradictory evidence to previously held suppositions.

24 Although historians are quick to point out what can be gleaned from the pages of catalogues, few have addressed the question of their limitations.10 The fact that most of the reproduction catalogues have been sponsored by the companies themselves inhibits objective assessment in introductions. In discussion of goods purchased by mail order, material historians have erroneously referred to them as "the arbiters of good taste."11 Although some of their users may have felt that through mail-order catalogues they had access to some of the finer goods available to city dwellers, our interpretation of what the catalogues meant to customers is largely speculation. The typical catalogue shopper of the early twentieth century has not been identified although some tentative efforts in this direction have been taken. For example, the archives of Sears, Roebuck and Company are rich in data that have been used to interpret the significance of the catalogues to its users. Their popularity in the early twentieth century was popularized recently in a documentary which aired on the Public Broadcasting System. The approach has been largely nostalgic rather than scientific.

25 Robert D. Watt, in his introduction to The Shopping Guide of the West: Woodward's Catalogues 1898-1953, includes a number of personal comments from Woodward's catalogue shoppers from the 1920s to 1940s, which provide insight into the value these catalogues held for them but, unfortunately, he does not provide any quantitative information.12 Although not limited to costume, his comments are useful particularly in placing catalogues within the context of their times, relating the catalogues to the events which fostered their rise and fall. Further work needs to be done to determine who used catalogues (gender, income level, occupation, region, ethnic background), what they purchased from the catalogues (finished garments versus materials, quantities of ladies wear, children's wear, menswear), how mail-order goods related to those available in general stores, how availability changed over the years, etc. Unfortunately, the raw data required to answer these questions does not exist in the Eaton's archives so it must be compiled by other, less reliable, means such as oral interviews with a large sample of informants.

26 Large numbers of people relied on Eaton's to bring them current fashions; the catalogues featured garments consistently and would not have done so if it had not been profitable. A better understanding of who used the catalogues would enable costume and social historians to interpret them more accurately. While a formal series of interviews specifically concerning the use of Eaton's catalogues has not been completed, the author has, in recognition of this problem, routinely asked questions about their importance during the course of interviews related to recent museum acquisitions for Alberta Culture.13 Women have discussed various means of earning their own money (selling butter and eggs, etc.) to enable them to purchase dresses or other personal items from Eaton's. They expressed excitement in occasionally being allowed to purchase a garment from Eaton's rather than having to make all of their own clothes. They have described efforts to make clothes, without a purchased pattern, based upon illustrations in the Eaton's catalogues.14 This data is entirely subjective and remains to be verified by a systematically addressed questionnaire.

27 Another source that is useful in providing a context for interpretation of this data is written records from the period (letters, journals, etc.). For example, in letters written to her mother and sister is England, between 1912 and 1914 when she was living in Windermere, British Columbia, Daisy Phillips frequently commented about the variety, quality and price of goods available by mail order through Eaton's. To her sister Freda she wrote,

A few weeks later she wrote to her sister that,

The "Stores" referred to here by Daisy is the Army and Navy Stores. The Army and Navy Stores in England, in contrast to the Army and Navy Stores in Western Canada, were considered comparable to Harrods and "although they had very comprehensive stocks, were supplying wealthy purchasers and were not used by the vast majority of people."17 The Army and Navy Stores have been addressed by British historians and should be considered in comparison with early catalogues from Woodward's, the Hudson's Bay Company and Eaton's. Researchers focus upon Canadian catalogues forgetting that, depending upon where one emigrated from, other catalogues may have been used as well.

28 Middle-class English immigrants like Daisy Phillips may have been disappointed by the quality and selection of goods available through Eaton's but in interviews, Eastern-European immigrants, unable to read the English descriptions or to afford to purchase finished garments, said that they were inspired by the illustrations; illustrations were in many cases more important than the descriptions.18

29 It is difficult to compare the descriptions and illustrations with the actual goods to know how accurate they are because there are few known examples of garments purchased through Eaton's in public collections. Many of the items sold by Eaton's were not produced in their own factories or labelled with Eaton's labels. Accession records rarely indicate where the owners purchased their clothing. The few garments with Eaton's labels that have been found in collections have not been identified in the catalogues. One man's shirt in the collection of the Provincial Museum of Alberta was selected for reproduction for use at the Ukrainian Cultural Heritage Village a number of years ago.19 However, in preparing the pattern for the shirt it was discovered that the sleeves were cut in a very unusual fashion, raising the question of whether the shirt was a "second," and whether it had survived because it was uncomfortable and therefore rarely if ever worn. Unfortunately, the data which is usually collected with museum artifacts is rarely detailed enough to answer questions such as these.

30 A systematic study of mail-order catalogues is required, which would compare the catalogues to extant artifacts and information obtained through oral interviews and written documents such as journals and letters. Until this is completed, researchers should be careful in assumptions regarding this material. The fact that a particular style, colour, or material was available in the catalogues does not mean that a given person would have worn it; there are many other factors that need to be taken into consideration. A comparison with catalogues from other firms and with popular ladies magazines provides a price scale for costume of the period. Quantitative analysis of the catalogues clearly points out that Eaton's only featured garments from the lower end of the scale, primarily house dresses and day dresses, not particularly fashionable clothing.

31 Initially, GWG produced an illustrated catalogue which was sold to merchants for distribution to their customers. It was a form of joint advertising as the merchant's name, as well as the manufacturer's name, appeared on the front cover. Some firms introduced almanacs and household handbooks, a form of enlarged catalogue which, as GWG explained, aimed

The idea was to produce a catalogue that was useful so that as well as providing information about the company's products, it provided additional information that would ensure that customers held on to it for future reference. The firm could include photographs of its premises and the costume being produced, as well as detailed information about advances in product development, thereby educating customers.

32 The almanacs were issued annually and featured a calendar which encouraged customers to retain their copies. They included household hints, first aid, gardening, laundry and cleaning information alongside horoscopes, tea cup reading and stock breeding records — truly something for everyone. The company's logo and brief claims about various product lines appear on almost every page. Inserted between recipes for Liver in Gravy and Liver Casserole was a description and illustration of a Women's Wool Plaid Sports Jacket.21 Caveats associated with the interpretation of almanacs are essentially those of mail-order catalogues. In addition, researchers should understand that almanacs, unlike other forms of catalogues, did not include the complete line of garments manufactured by a firm, simply a sampling of some of the more popular styles.

33 The Department of Consumer and Corporate Affairs in Hull, Quebec houses a valuable collection of patents, trademarks and registered industrial designs.22 Unfortunately, the material is difficult to access. All files are organized chronologically but are indexed inconsistently. For example, several attempts to develop a comprehensive list of patents issued to the Great Western Garment Company proved unsuccessful because some features were patented by principals in the firm rather than under the company's name. These patents only turned up during an exhaustive search through all patent registrations for the period. Similarly, it is impossible to develop a list of all items patented by inventors from a given province because the material is not indexed by address. The material is organized for the use of contemporary inventors interested in finding out whether or not they have come up with a new idea.

34 Although not organized or indexed in a way that is easily accessible for historians it would be fruitful to develop a complete index of costume and textile registrations by Canadians. Some of the types of features registered include: jewellery designs for items inspired by the Klondike gold rush, colour variations in knitting patterns, various types of pockets and fastenings, innovations such as pre-shrunk fabric, labels, applied decorative features, etc. If the material was properly indexed, costume historians would find it useful for a number of purposes, including the dating of garments in their collections and as a record of inventive activity in costume and textile manufacturing.

35 As with other corporate records, researchers must be cautious about the use of this data. The fact that a design was registered does not mean that it was ever actually put into production. Many ideas were registered on speculation and once the inventor tried to develop a prototype they were abandoned as impractical. Similarly, claims made by manufacturers in patent applications can not always be supported by further research into a company's products. This data must be substantiated by documentation from directories, advertisements, catalogues, and most importantly, extant garments. Many of the claims made in patent applications can best be measured by an examination of the items or features being patented. Do they in fact live up to their claims?

36 As is discussed further in Theresa Rowat's paper elsewhere in this issue, photographs are frequently misinterpreted by researchers who do not understand the context in which they were taken. One example in terms of corporate photographs, is of a series of photographs of the Great Western Garment factory in Edmonton in the 1910s. Found by researchers in the files of the Glenbow Archives and the Provincial Archives of Alberta, these photographs have twice been used to document problems in working conditions in the garment manufacturing industry.23 In fact, this was one of a series of photographs taken on behalf of the firm to document the clean and orderly working environment in the plant. Even before finding an advertisement highlighting this photograph in Alberta Labor News, a costume historian would suspect that the photographs had been staged because of the way the workers were dressed. Their white, neatly pressed dresses and perfectly placed hairdos are not what one would expect in a factory. Furthermore, there are two views of one photograph with subtle changes in composition (Figs. 7 and 8). Locating the advertisements was simply confirmation of what had been observed.

37 The Alberta Labor News states that this photograph "illustrates the progress of industry developing hand in hand with labor in the service of the great masses of the people" (Fig. 9).24 The photograph was also used in an advertisement, in Farm and Ranch Review and the Fanner's Almanac and Home Journal, contrasted with a photograph of the first factory.25 This advertisement states that "Quality and Service were alone responsible for what is said to be the greatest stride ever made by a manufacturer in Canada." GWG's standards were beyond those required by Alberta's labour legislation at the time. These photographs were used to advertise the benefits of supporting unionized occupations. Many of the workers wearing GWG clothing were unionized and would support a unionized firm over a non-unionized firm.