Cultural Mapping and the Digital Sphere

Place and Space

Ruth Panofsky, Kathleen Kellett, Ruth Panofsky, Kathleen Kellett

• Publisher: University of Alberta Press
• Year: 2015
• Print ISBN: 9781772120493
• eBook ISBN: 9781772120561
• Topic: Education
• Subtopic: Essays in Education

"Notwithstanding their differing approaches—digital, archival, historical, iterative, critical, creative, reflective—the essays gathered here articulate new ways of seeing, investigating, and apprehending literature and culture." – From the Preface This collection of essays enriches digital humanities research by examining various Canadian cultural works and the advances in technologies that facilitate these interdisciplinary collaborations. Fourteen essays—eleven in English and three in French—survey the helix of place and space. Contributors to Part I chart new archival and storytelling methodologies, while those in Part II venture forth to explore specific cultural and literary texts. Cultural Mapping and the Digital Sphere will serve as an indispensable road map for researchers and those interested in the digital humanities, women's writing, and Canadian culture and literature. Foreword by Susan Brown and Mary-Jo Romaniuk. Contributors: Jeffery Antoniuk, Susan Brown, Constance Crompton, Ravit H. David, Patricia Demers, Shawn DeSouza-Coelho, Cecily Devereux, Teresa M. Dobson, Sandra Gabriele, Isobel Grundy, Andrea Hasenbank, Paul Hjartarson, Kathleen Kellett, Sasha Kovacs, Vanessa Lent, Margaret Mackey, Breanna Mroczek, Bethany Nowviskie, Ruth Panofsky, Mariana Paredes-Olea, Harvey Quamen, Jennifer Roberts-Smith, Omar Rodriguez-Arenas, Mary-Jo Romaniuk, Stan Ruecker, Lori Saint-Martin, Michelle Schwartz, Stéfan Sinclair, Mireille Mai Truong, Stéphanie Walsh Matthews, Heather Zwicker.

MAPPING TAGS AND TAGGING MAPS

Leveraging Spatial Markup for Literary History

Susan Brown, Isobel Grundy, Mariana Paredes-Olea, Jeffery Antoniuk, & Breanna Mroczek

Introduction

The “spatial turn” is making itself felt throughout literary studies, as it is in other areas of the humanities and social sciences. As Barney Warf and Santa Arias argue,

Recent works in the fields of literary and cultural studies, sociology, political science, anthropology, history, and art history have become increasingly spatial in their orientation. From various perspectives, they assert that space is a social construction relevant to the understanding of the different histories of human subjects and to the production of cultural phenomena.…In other ways, however, the spatial turn is much more substantive, involving a reworking of the very notion and significance of spatiality to offer a perspective in which space is every bit as important as time in the unfolding of human affairs, a view in which geography is not relegated to an afterthought of social relations, but is intimately involved in their construction. Geography matters, not for the simplistic and overly used reason that everything happens in space, but because where things happen is critical to knowing how and why they happen. (1)

There are always complex reasons for an intensifying interest in space as a topic or analytical category, but there is no question that the current interest is fuelled by the greater accessibility of spatial technologies such as plotting tools and geographic information systems (GIS), made possible by the World Wide Web, that put a dazzling array of maps at our fingertips. These maps allow us to explore digital representations of space by interactively zooming in and out, experimenting with various layers of information, and gathering granular analysis of a vast range of materials. The application of computational spatial analysis in literary studies is relatively new, but the results are already evident in a number of impressive projects. These projects include an overview of the development of printing technologies (Prickman); an analysis of the impact of railway development (Thomas, Healey, and Cottingham); and a study of the circulation of letters in early modern Europe (Coleman, Edelstein, and Findlen).

Given that the Canadian Writing Research Collaboratory (CWRC) is dedicated to creating an online environment for research on writing and seeks to offer traditional scholars a wider array of computational tools than those they typically incorporate into their research, it is hardly surprising that the literary researchers involved in CWRC’s development have consistently identified the mapping of literary data as a highly desirable feature of a shared digital tool set. The precise nature of the mapping required for most CWRC projects has yet to be established, but there is definite interest in how the spatial and temporal aspects of literary research intersect. Historical maps are also desirable for their representation of literary materials or activities from the past, as are maps that might represent slices of time or change over time. As Peter K. Bol argues, “if we fail to see that change over time unfolds differently across space, we substitute a single history for the reality of multiple histories” (297). While the study of writing has attended to the impact of place in myriad ways, newly accessible visual representations are fast becoming an effective means of tracking changing spatial and temporal relationships.

All of the research projects affiliated with CWRC have a spatial component. Some are concerned with regional writing. A number of theatre projects seek to spatially track productions and networks of professionals. The Editing Modernism in Canada project at the University of Alberta has developed a smart phone app called “WatsonWalk” based on walks taken by Sheila Watson while she and her husband Wilfred Watson lived in Paris from 1955 to 1956, as documented in her journals. Kristine Moruzi’s project “From Colonial to Modern: Transnational Girlhood in Australian, New Zealand, and Canadian Print Cultures (1840–1940)” aims to map the relationship between places of production, places of publication, and places of dissemination. The most ambitious of CWRC’s pilot projects is “Magazines, Travel, and Middlebrow Culture in Canada 1925–1960,” led by Faye Hammill of the University of Strathclyde. This project investigates geographical mobility as a form of upward mobility, and is designed to distinguish leisure travel from the enforced movement of migration and diaspora. The project uses middlebrow magazines to map shifts in travel patterns. If one takes seriously the claim that space “is not simply a passive reflection of social and cultural trends, but an active participant, i.e., geography is constitutive as well as representative” (Warf and Arias 10), then to make available a dynamic mapping environment that visualizes spatial relations is to give scholars a potentially transformative tool for investigating those constitutive relations. Yet, given the wide range of CWRC projects, the task of devising even a basic generalized system of mapping is a considerable challenge, as the project’s core development team learned in its initial experiments with spatial data contained in its largest seed project, the Orlando Project, housed at the University of Alberta.

Orlando is the founding project behind CWRC. Initiated in the mid-1990s by a small team of literary researchers, Susan Brown, Patricia Clements, and Isobel Grundy, just as graphical browser environments were exponentially expanding the usage of the World Wide Web, this research project sought to produce, from scratch, a history of women’s writing in the British Isles using computers (Brown and Clements et al.). Rather than digitizing existing texts—the focus of many digital literary projects then and now—Orlando sought to produce born-digital scholarship in a form intended from the outset to take advantage of the digital medium. That meant trying to imagine how the scholarly text that one is researching and creating might take advantage of the yet emerging possibilities of a web publication environment. Hence, on the advice of our digital humanities co-investigator Susan Hockey, the project adopted Standard Generalized Markup Language (the precursor to Extensible Markup Language or XML) as a means of encoding the text in a form that would make it amenable to reuse in applications that could not be developed during the initial stages of the Orlando Project.

Members of Orlando were keen to map its contents. From the start, they assiduously tagged locations throughout the text base using nested tags, based on those developed by the Text Encoding Initiative (TEI), to label places mentioned in many different contexts so that it would be possible to search and eventually map the materials that were being written. Thus, spatial information in Orlando was tagged with labels that specify that the text enveloped in the tag is a reference to a place, with sub-tags or nested tags demarcating the subcomponents of the spatial information. For example, the markup for the city of London is

<PLACE><SETTLEMENT>London</SETTLEMENT><REGION REG=Middlesex> </REGION><GEOG REG=England></GEOG></PLACE>

Such tagging distinguishes this particular reference to London from the more than two dozen other settlements around the world (including London, Ontario) that take their name from the English city, from building names such as London House, and from proper names such as Jack London.

The Orlando text base comprises several types of materials in which such spatial information occurs: bibliographical entries that identify locations of publishing activities; events that usually (though not always) specify the places where those events occurred; and detailed bio-critical entries on particular writers that contain information about locations related to their subjects’ lives and writing careers, as well as the places involved in the poems, plays, novels, travel writings, or other texts they produced. Orlando’s bio-critical entries on writers thus contain rich spatial information, as do the bibliographical and contextual event material. Among Orlando’s more than eight million words of born-digital scholarship, there are (at the time of writing) a total of 65,488 place tags and a total of 10,657 unique place names. On the basis of this extensive set of materials, the Orlando and CWRC teams have conducted experimental geographical visualizations as a means of exploring how best to implement mapping functionality within the Collaboratory platform. We have begun with information related to writers’ places of residence, travel, migration, and social networks, and how this biographical data may relate to facets of their writing careers, such as places of production or reception, literary settings, and topographical descriptions. Here, we report the results of this first foray into the complex interrelationship of data, tools, and graphical interfaces.

Because the range of available spatial information varies considerably, the digital mapping of place and space contained in a literary history such as Orlando demands a flexible system that combines spatial identification with contextual information. Places may be real or fictional, domiciles or travel destinations, sites of political activity, publication, or reception of texts, and all of these place types are included in the spatial information encoded in Orlando’s text base. A map that represented all such information without distinction would be difficult to read. Moreover, the project’s historical sweep soon made it evident that many places require definition in time as well as space in order to be represented meaningfully, particularly since place names and boundaries have changed over the course of British (and international) history. Location is similarly pertinent to matters of biography and literary reception. Visualizations may be space-dominant or time-dominant, and one of our challenges is to better integrate spatial and temporal dimensions in the production of geographical representations of literary history, as Ian Gregory advocates in “Exploiting Time and Space: A Challenge for GIS in the Digital Humanities.”

Orlando’s texts use tags based on TEI for entities such as place, but they also use a custom tag set designed specifically by the research team for the production of a digital literary history. This tag set demarcates spans of text according to several contexts, such as discussions of writers’ residences, travels, or migrations; fictional settings or non-fictional topographical descriptions; sites of theatrical productions; or where a writer engaged in political activism. Thus, the tagging of a sentence such as “Frances Brooke sailed from England to join her husband in Québec” under a “location” tag allows place as a destination of travel to be distinguished from a sentence such as “This novel is best known for its picture of settler or habitant life in Lower Canada, which Frances Brooke drew from her own years there” (Brown, Clements, and Grundy). Places often mentioned under the “politics” tag, for instance, include Hyde Park and Trafalgar Square in London, the nuclear research station at Aldermaston in Berkshire as the site of demonstrations, and Holloway Prison as the place where London suffragettes were incarcerated. These references to London as a place are all distinguishable from London as a place of birth, death, or publication, even within the current published interface of Orlando, which is entirely textual.

As a result of having implemented such semantic tags in Orlando, then, we can contextualize each mention of a place name and explore the axes of geographical representation that will be useful to CWRC-affiliated researchers and in the delivery of CWRC materials. The CWRC team needs to test spatial data infrastructure that can map diverse projects and varying articulations of spatial experience. For our initial explorations, we therefore selected two axes of representation that would be of broad interest to literary scholars: the travel or migration of writers or groups of writers; and trends in writers’ geographical subject matter (that is, places represented in writing).

Creating the Maps

The methodology for generating the maps discussed here involved several stages. First, Orlando data was queried according to various parameters, drawing on the semantic tags discussed above, such as “return all biographies in a given range of birth dates,” “return all excerpts containing a given keyword,” or, “return all excerpts containing ‘Scotland’ within discussions of violence or politics.” Second, once the queried data was retrieved it was fed into a pre-processing stage in which geographical data was extracted, contextualized, and stored separately. By contextualization we mean specifically the association of geographical data with a declared temporal axis, or—when no date was mentioned in the precise context—with a speculation as to temporal location. We processed all place names returned by these search parameters into the form of events with spatial and temporal co-ordinates. In so doing, we avoided the necessity of visualizing all of Orlando’s spatial information in a single, temporally conflated map. In the absence of temporal information embedded in the spatial data itself, we drew on the fact that Orlando’s documents organize information in roughly chronological order to provisionally infer the temporality associated with a place by extrapolating from adjacent dated material. Based on this process of inference, each instance of geographical information produced by the initial extraction process was associated with a specific date or range of dates, the writer’s name, and the subject discussed, as well as the complete excerpt of the prose under each location as mentioned in Orlando.

A third step consisted of matching place names to latitude and longitude information. In advance of mapping, however, the existing spatial data needed to be enhanced. The mapping systems with which we experimented cannot produce a map from a simple place name like “London,” as it is not in itself geo-referenced, that is, related to a location on the earth. This is not a trivial process, even for materials in which the different forms of “London” have already been distinguished from one another. Automated services for geo-referencing place names exist, but they produce a certain number of false or (in other respects) inadequate matches and so require considera...

Table of contents

• Cover
• Title Page
• Copyright Page
• Contents
• Foreword
• Preface
• PART I PLACE AND THE DIGITAL FRONTIER
• PART II WRITERS AND READERS: MAPPING TEXTUAL SPACE
• Contributors / Collaborateurs
• Index
• OTHER TITLES FROM THE UNIVERSITY OF ALBERTA PRESS