1 This paper examines the role of artifacts in the study of North American textile history. After first discussing my own approach, I will outline the broad parameters of the existing literature on the topic, concluding with an examination of some historical fabrics to suggest how they can be used to further illuminate our understanding of a variety of aspects of the subject.

2 Historians are increasingly recognizing the value of incorporating artifactual evidence into their research. Objects can provide information about large numbers of people who are often absent from documentary sources. As a textile historian, I have begun to develop a framework for integrating material and documentary evidence to produce more layered and complex historical analyses.

3 My involvement with the field of textile history follows a distinct and complementary path, beginning with a practical knowledge of weaving, followed by a historical research base. These two elements are becoming increasingly important for those wishing to examine the interface between artifacts and documents. I first encountered the study of historic textiles when I worked as a weaver investigating traditional methods of non-mechanized cloth production. During this period, I analyzed a large corpus of extant nineteenth-century handwoven textiles many of which I subsequently reproduced using eighteenth and nineteenth century equipment and methods. The final product was a manual, "Reproducing Nineteenth Century Handwoven Fabrics, A Weaver's Guide to Accurate Reproductions," which is housed in several major American and Canadian museums. During the course of the technical and artifactual studies, I began to seek more information about the historical contexts of the fabrics and their methods of production. Thus, I prepared a historical case study of textile production in eighteenth-century, rural Pennsylvania.1

4 These two elements, the practical and theoretical, have informed my approach to the study of textile history and the issues with which I was initially concerned. Among the questions raised were: How available were the various textile technologies in different periods and regions? What was the impact of technology on workers? What role did gender play in cloth production? What were the levels of skill of the artisans? What were their products and how did they market them? I was also interested in examining the sources for North American textiles other than local manufacture and their effect on domestic production; the role consumer behaviour played in determining the quantities and qualities of cloth consumption; and finally, how artifacts shed light on all these issues.

5 Many of the above concerns arise from contemporary social historical thought. The ability to utilize data contained in cloth and textile tools as an integral element of historical analysis derives from my training as a weaver and work as a museum curator. In order to understand how my research is influenced by the combination of these elements, it is useful to review briefly the approaches to textile history that have characterized the field until the present.

6 Broadly speaking the artifact-based literature dealing with North American textile history can be divided into three groups: those studies primarily concerned with the objects, those that rely almost exclusively on documentary evidence, and those that combine the two approaches. This is not to say that these are mutually exclusive categories, just that a single methodology tends to dominate works in each group.

7 A survey of textile literature showed that the majority of it fell into the first group, focusing primarily on objects.2 Such works attempted to describe and catalogue textiles that might have been made and/or used on this continent and their method of manufacture. The artifact base employed in this type of analysis, consisting of a disproportionate quantity of coverlets, was collected in the field, drawn from existing museum collection, or both.

8 Some of these studies concentrate primarily on one type of textile, such as John Heisey's, A Checklist of American Coverlet Weavers, which centered on jacquard-woven coverlets; or a single region, such as the catalogue of Louisiana Cajun textiles, L'amour de Maman.3 Other are broader, perhaps the best known being Harold and Dorothy Burnham's "Keep Me Warm One Night," in which the authors not only catalogued Eastern-Canadian handwoven textiles but also discussed their ethnic origins and evolution in Canada, the tools and processes with which they were made, their structures, and pertinent weaving information.4 Descriptive analyses, while necessary, are biased by the availability of existing textiles—how representative is an artifact if it survived long past its period of use?—and by the nature of collecting—how often do people collect remarkable items, while ignoring the more mundane? However, if one understands these shortcomings, catalogues and descriptive studies serve the useful and necessary function of illustrating and categorizing objects used in the past, thereby providing a foundation on which to build more broadly-based artifactual research.

9 Although the object-oriented literature provides excellent detail about the artifacts, the historical context is often unevenly researched and perpetuates myths and stereotypes. For many authors uncritically describe the generally accepted version of early American textile manufacture in which most households had all the necessary tools for cloth manufacture, and the women, assisted by their children, spun all the yarn and wove all the fabric needed for the family's use. If the women did not weave, an itinerant weaver did. Perhaps the biggest problem with such works is that frequently scholars seeking information on textiles and their production turn to them in the absence of more knowledgeable artifact-based research. As long as object analysis remains the exclusive focus, it will be difficult to revise such vague interpretations.

10 However, during the last twenty years, North American textile history entered a second phase, as social scientists entered the field lured by the richness of the indirect evidence regarding cloth contained in household inventories, tax lists and census data, business records, and so on. Studies such as Linda Baumgarten's "The Textile Trades in Boston, 1650-1700," and Susan Prendergast Schoelwer's "Form, Function, and Meaning in the Use of Fabric Furnishings: A Philadelphia Case Study, 1700-1775," utilized household inventories to demonstrate that while most of the fabrics found in these seventeenth- and eighteenth-century port cities were imported, by far the majority were functional woolens and linens, not the fancy silks and printed cottons disproportionately present in large, urban museum collections.5 These articles represented an important step in altering interpretation of textile history which has tended to centre on the upper levels of society. They further demonstrate the need to move beyond the purely artifactual evidence in order to create a more comprehensive image of the broad spectrum of social groups engaged in various aspects of North American fabric production and consumption.

11 Other work relying on documentary sources has begun to inject even more detail into the nature of local textile production in North America. My recent research has quantified data from inventories, tax lists and account books to reveal that in some regions of colonial America cloth production was divided along gender lines with a small group of professional male weavers responsible for making some of the cloth used locally.6 David-Thiery Ruddel, in "The Domestic Textile Industry in the Region and City of Quebec, 1792-1835" drew on post-mortem inventories and discovered that there was a distinct difference in rural, versus urban, patterns of manufacture and consumption of locally-produced cloth.7 And Tom Dublin, in his study of outwork in nineteenth-century New England, utilized business records to show that many rural women wove on a commercial basis. This is a very different interpretation than that which shows them weaving cloth for their self-sufficient households.8 Such research clearly demonstrates the complexity of early handwoven cloth production and suggests the need to re-evaluate temporal and regional demographic, economic, political and social realities.

12 Perhaps the biggest problem facing the field of North American textile history at the moment is how to reconcile the two approaches of direct analysis of textiles and related objects, and indirect references to cloth found in documentary sources. A great deal of time and energy has been devoted to collecting, documenting and analyzing fabric artifacts and the methods used in their manufacture, but the question must be asked, can these objects shed new light on the concerns of historians unobtainable through documentary sources alone?

13 Despite the difficulties in integrating textile artifactual data into historical studies, a third category has emerged that attempts to do this. In 1985, David-Thiery Ruddel and I examined the value of this method as it related to Quebec in our article "Artifacts and Documents in the History of Quebec Textiles."9 Combining a database of extant French-Canadian and British household textiles with documentary sources relating to nineteenth-century Quebec, we gained insight into several aspects of French-Canadian history not easily evident from written sources alone. Of particular import was the recognition of the direct and lasting impact of the British presence in Quebec on the design of locally-made handwoven textiles. Bedcoverings woven by mid-nineteenth century, French-Canadian, rural women are direct copies of imported, mass-produced, British-made coverlets. The designs have been modified over the years but have retained strong vestiges of the originals well into the twentieth century. This transference of British motifs into popular culture has been hitherto unrecognized in Quebec ethnographic literature. In addition, the comparatively low level of skill and extensive use of recycled materials exhibited in most surviving Quebec textiles suggest that the weavers did not receive the training of the professional male weavers working in other regions of North America. All of these observations need further refinement, but they would not even have been made without examining the textile artifacts.

14 Another example of the third phase of textile scholarship in which knowledge of the methods of production played an important role is in my work on rural Pennsylvania.10 Here there was a decided lack of artifactual evidence since many of the textiles made and used prior to the War for Independence were recycled and worn out in attempts to keep the American troops clothed and warm. But a knowledge of the work involved in making fabric, and of textiles from similar contexts, allowed the interpretation of both the organization and the extent of local cloth production and some rough estimates about the proportion of locally-made and imported material required to meet the rural community's needs.

15 Studies that endeavour to incorporate artifactual and documentary data are difficult to do as they require an ability to "read" and evaluate artifacts much as one does with written evidence. But they are increasingly being done as indicated by Katherine Grier's book Culture and Comfort: People, Parlors and Upholstery.11 Grier combined her knowledge of Victorian textile furnishings with historical, art historical and anthropological methodologies to analyze how domestic artifacts mirrored the tension between culture and comfort that existed in American Victorian society.

16 Much of the research discussed in the papers appearing in this issue of Material History Bulletin could benefit from the inclusion of direct or indirect artifact analysis. Using Tom Dublin's research on nineteenth-century textile outwork in New England as an example, object knowledge would assist in assessing the importance of weaving in the lives of the women workers in several ways. First, by helping to formulate the appropriate questions to be asked of the documentary data: Where were cotton warps sized—at the mill or by the weaver? Do we know the thread count of the finished cloth? What did the cloth look like—for example, checked takes longer to weave than striped. In the absence of this information from documents, analysis of a single piece of fabric identified as the product of outwork would help to provide answers. Second, what was the weaving technology used in outwork looms? An example of such a loom compared with one belonging to a professional male coverlet weaver, for example, could shed further light on the role of gender and textile production. In addition, if either the artifactual or documentary evidence suggests that any of these outwork looms was equipped with a fly shuttle, one would need to re-evaluate substantially the amount of time spent weaving a piece of cloth.

17 The survey of the artifact-based literature relating to North American textile history suggests several areas for more detailed study. How does textile usage in North America and Europe compare? Are there regional variations within the continent? What was the impact of the industrialization of the textile industry, first in Britain, then in the United States and Canada, on the supply of and demand for fabrics on this continent? Remembering that throughout the seventeenth, eighteenth and much of the nineteenth centuries, North Americans depended on foreign fabric to supply many of their cloth needs, we must not isolate textile research to this side of the Atlantic. Finally, it would be useful to continue to look at fabric production and use in its broader social, economic and political contexts, addressing such issues as rural self-sufficiency; consumer behaviour; marketing; regional, national and international trade; imperialism; gender; the process of industrialization and modernization; and how all of this evolved.

18 Combined with the above is the need to expand our artifactual data base by collecting modern materials, examining the output of North American mills, and doing more comparative studies of Canada and the United States to see the similarities and differences of textile manufactures in two countries with many elements in common and yet some very significant differences. Finally, at some point, the disparate artifactual- and documentary-based textile literature should be synthesized to provide a revised general picture of North American cloth manufacture.