HGIS Growth Elsewhere

While HGIS was gaining momentum in the SSHA, efforts by individual scholars drew attention to the revisionist potential of GIS-based research. Geoff Cunfer's On the Great Plains, a product of the Great Plains History project, won the SSHA President's Book Award for best first book manuscript in 2003. Cunfer's GIS analysis of climatological and soil data across the Great Plains challenged Donald Worster's much more local argument that farmers’ excessive plowing caused the Dust Bowl (Cunfer Reference Cunfer2005). Fellow environmental historian Brian Donahue brought a microgeographical approach to agricultural history in The Great Meadow. Donahue's GIS analysis of landholdings and land use on individual farms overturned the long-standing view that New England farmers mined the soil. He found to the contrary that generations of farmers in Concord, Massachusetts, were careful stewards who deeply understood the productive potential and limitations of their land (Donahue Reference Donahue2004). A specialist conference that I organized with the Newberry Library in 2005 featured these and many other well-advanced HGIS projects.Footnote ⁴ A few years later, Idaho State University historian J. B. Owens launched America's first graduate program with an HGIS track, a masters in historical resource management (Owens and Woodworth-Ney Reference Owens and Woodworth-Ney2005). With funding from a Mellon Foundation professorship, historian Richard White established the Stanford Spatial History Lab in 2006–7. Under Erik B. Steiner's leadership as creative director, the lab has become a magnet for creative geovisualization in the geohumanities (Dear et al. Reference Dear, Ketchum, Luria and Richardson2011). A national and international community of HGIS scholars was beginning to take shape, and interest was growing in the potential for GIS to provide an analytical framework for research at any scale, from local history to world history.

Beyond the seedbed of the SSHA, a number of institutional initiatives during this period sought to promote and support the use of GIS in social science and humanities research and teaching. The Electronic Cultural Atlas Initiative (ECAI) was based at the University of California, Berkeley. Under the leadership of Lewis Lancaster, a scholar of East Asian religions, ECAI aimed to build a global community of scholars using GIS and other digital methods for spatio-temporal research, promote collaboration across and among disciplines, enable scholars to share data through an online clearinghouse, and develop open-source tools for spatial scholarship.Footnote ⁵ Although Lancaster emphasized cultural research, ECAI's early participants also included social science historians and historical geographers, notably David J. Bodenhamer, John Corrigan, and Trevor Harris. In their continuing collaboration, they espouse the “spatial humanities,” a conceptualization of digital scholarship that emphasizes GIS as a core method for spatial research. They also advocate “deep mapping,” a vision of multimedia, immersive, exploratory research that combines GIS with other visual technologies for qualitative and quantitative analysis and representation (Bodenhamer et al. Reference Bodenhamer, Corrigan and Harris2010, Reference Bodenhamer, Harris and Corrigan2013, Reference Bodenhamer, Corrigan and Harris2015).

Several pedagogical initiatives also worked to support faculty who wished to learn GIS and use it in their teaching. The National Institute for Technology in Liberal Education, founded in 2001, hired geographer Diana Stuart Sinton as GIS Program Director in 2003. Much of her outreach to liberal arts colleges across the United States involved leading basic GIS workshops and helping faculty at institutions without a geography department or GIS technical support staff develop approaches for using spatial methods in their teaching and research. Sinton later continued this work as director of Spatial Curriculum and Research at the University of Redlands in Redlands, California (pers. comm., February 20, 2016).Footnote ⁶ Redlands happens to be the home of Esri (Environmental Systems Research Institute), the world's largest manufacturer of GIS software and applications. Esri's corporate strategy has long included providing free online instruction and materials for classrooms using Esri software. While these educational efforts aimed to stimulate sales of Esri products, owners Jack and Laura Dangermond also wanted to help fill the gap in geographic education, which remains sorely lacking in many American schools,Footnote ⁷ colleges, and universities. Esri workshops and short training courses on the company's primary software, ArcGIS, have been gateways to GIS for many faculty and students over the years. The most focused and sustained institutional effort to train social scientists in GIS principles and research methods has been the Center for Spatially Integrated Social Science (CSISS) at University of California, Santa Barbara, created in 1999 by GIScientists Michael F. Goodchild and Donald G. Janelle. CSISS programs have included some designed for specific fields, such as economics, as well as interdisciplinary workshops for criminologists, archaeologists, political scientists, and others (Goodchild, pers. comm., February 18, 2016; Janelle, pers. comm., February 18, 2016).

Challenges of HGIS

Goodchild expressed surprise that scholars in the arts and humanities have sometimes been more receptive to GIS than have social scientists. He credits this partly to the strong social critique of GIS, which has made it a suspect technology for some cultural anthropologists. I have found resistance to GIS among cultural and even some historical geographers, due to the rejection of positivistic methods and strong critiques of cartography during the theoretical turn in human geography that began in the late 1980s and is only now beginning to abate. The technical difficulty and expense of the hegemonic software ArcGIS have also been significant obstacles. A less obvious explanation for varied receptivity to GIS lies in academic disciplines’ varying conceptual foundations and customary practices. Goodchild notes that scholars of dance, architecture, and art history have been particularly quick to engage with GIS at CSISS workshops (Goodchild, pers. comm.). This is not surprising, for these fields inculcate visual and spatial awareness. During my tenure at Middlebury College, I found many kindred spirits in the History of Art and Architecture Department, and many of their students took to GIS and cartography like fish to water.Footnote ⁸ Similarly, art history faculty and graduate students who participated in a digital mapping summer institute that I codirected had less difficulty grasping the geographical concepts that underpin GIS and cartographic representation than many other historians with whom I have worked (Jaskot et al. Reference Jaskot, Knowles, Wasserman, Whiteman and Zweig2015).

Geographer Helen Couclelis pointed to the significance of conceptual frameworks for understanding GIS in a report on a faculty program mounted by CSISS. Couclelis found that academic economists, despite their experience in regional science or urban studies, had difficulty seeing the utility of GIS because they lacked “a clear conceptual foundation in order to successfully apply these methods [GIS] in a spatial context” (Couclelis Reference Couclelis2004). In addition, they identified problems in dealing with missing data—a chronic issue for a spatial methodology that seems to require complete representation across the study area—and the complexity introduced by adding space to temporal analysis. Although economic historians have proven to be among the readiest adopters of GIS among social science historians, the questions that emerged in Couclelis's study remain salient today. Why should scholars who are conducting already complex research “be further burdened with geometry, topology, and spatial autocorrelation” (ibid.: 23)? Seeking answers to spatial questions, arguably the core of geographic analysis, remains unfamiliar and can seem unnecessary to those trained in other disciplines (cf. Knowles Reference Knowles, Gregory, DeBats and Lafreniere2017).

HGIS also remains challenging because of the time and expense it often takes to create historical spatio-temporal databases. Certain kinds of HGIS databases can be built relatively easily, such as compilations of scanned maps for a city on which researchers locate historical buildings, with links to historical photographs or texts. This kind of digital curation, however, despite its popularity for pedagogical projects, rarely yields scholarly breakthroughs. Analytically powerful compilations of geolocated data, whether for large digital infrastructure projects or focused scholarly inquiry, take many years to develop, often at great expense. This is partly because the research involved in geographically locating historical entities—the necessary precursor to any kind of spatial analysis—often takes as much or more time as entering all the other data into the database. Southall and Gregory encountered this problem with their Great Britain Historical GIS (Gregory and Southall Reference Gregory, Southall and Knowles2002). Mapping nineteenth- and twentieth-century demographic data within historically accurate administrative boundaries took a great deal of time and effort, but it also made the project possible. Likewise, in preparation for my analysis of spatio-temporal patterns in the developing American iron industry, locating 1,100 iron works took many months of research by two scholars and three research assistants (Knowles Reference Knowles2013). The monumental Transatlantic Slave Trade Database began as, and continues to be, largely a locational effort to map the basic geography of the trade (Eltis and Richardson Reference Eltis and Richardson1997, Reference Eltis and Richardson2010). Locating slaves’ origins in Africa remains a daunting scholarly task three decades after the project began (Lovejoy Reference Lovejoy2013).

Visual Culture in Social Science History

One more issue is specifically germane to the relatively minor presence of HGIS in Social Science History. Outside the two special issues, there have been surprisingly few articles in the journal that are based on GIS analysis or that include maps that appear to have been made with GIS or computer graphics software. It may be that HGIS also challenges the journal's visual culture and publishing conventions, or at least that prevailing modes of representation among GIS users do not fit well with the journal's constraints. Looking through 40 years of journal issues, one does not get the impression that its contributors are strongly visual thinkers. Most articles appear to make verbal arguments, not visual ones. Few issues of the journal include more than a handful of figures, and most of those are graphs. Maps are generally scarce until the late 1990s and early 2000s, and they are never common.

I would argue, however, that Social Science History has in fact always had a visual aesthetic that reflects the association's emphasis on method. The aesthetic is straightforward; typically, bare-bones visualization of numerical or statistical evidence. The yardstick for quality is legibility. Visual argument in Social Science History takes one of three basic forms:

1. (1) Tables, which reveal relative values while suggesting thorough research.

2. (2) Graphs, which more explicitly demonstrate analytical methods and reveal historical dynamics.

3. (3) Fairly simple maps, which either situate the research geographically (as in locator maps) or reveal real differences that become the basis for geographical comparison.

These familiar modes of visual representation seem to be almost without design in Social Science History. Like Shaker furniture, however, their simplicity has a purpose: the clear explication of method and evidence. This is the same purpose that has been a driving force and inspiration for social science history since its inception. I would go so far to say that Social Science History has cultivated an egalitarian representational culture that effectively invites readers to sit beside the researcher in what often seems to be a scarcely mediated examination of historical evidence (cf. Bender and Marrinan Reference Bender and Marrinan2010). This egalitarianism corresponds with the SSHA's purposely low costs of membership, conference registration, and subscription to the journal. Keeping participation costs low has also, thus far, prohibited full-color or glossy graphics reproduction in the journal.

The visual constraints of Social Science History lessen its appeal for HGIS scholarship. Given how easily one can add layers of data to a map in GIS, it is tempting to produce maps of considerable complexity. In the age of digital cartography, the restrained, elegant art of pen-and-ink map design has given way to lavish use of color. The number of tonal values that the eye can recognize in color greatly exceeds the tones of gray we can distinguish. Area “fills” of black and white patterns (remember Zip-a-Tone?) are now unacceptably old-fashioned, but digital design templates have few good alternatives. The result, too often, is that maps created in color are reproduced in grayscale in Social Science History (as in other black-and-white journals), often lessening the figures’ legibility and attractiveness. I do not know for certain that the lack of color in our journal prevents HGIS researchers from submitting articles, but it may be a factor. Authors have a growing number of alternatives as social science and GIS journals provide full-color reproduction in print or on the journal website.

Emerging Directions in HGIS Scholarship

With color or without, I see several promising new directions in HGIS scholarship at the SSHA and related professional organizations. One is a gradually emerging interest in the issue of uncertainty. GIScientists have been wrestling with the uncertainty of spatial data for many years. Means of measuring and accounting for uncertainty became a specialty within GIScience almost two decades ago (Zhang and Goodchild Reference Zhang and Goodchild2002). It can be a jarring experience for historians when the process of trying to map historical information reveals how uncertain, vague, imprecise, or temporally unstable locations can be. This presents an important opportunity for historians to contribute substantively to issues that concern researchers in many other fields. Can we develop methods for gauging the measurable or the relative certainty of historical information? Can we go beyond the very limited cartographic devices of hollow symbols and dashed lines to represent uncertainty on maps of past landscapes and historical phenomena?

A second new direction in social-scientific spatial thinking is the study of gazetteers as sources of historical geographical information (Bol Reference Bol and Knowles2008). Southall, one of the early champions of HGIS, is in the vanguard of gazetteer research, along with historians of China, notably Ruth Mostern and Merrick Lex Berman. As Mostern's work on the spatial history of the Song Dynasty shows, for more than 1,000 years Chinese gazetteers compiled a tremendous wealth of information about places, their rulers and citizens, events, topography, ways of life, religion, government, and more (Mostern Reference Mostern2011). In their latest collaboration, Mostern and colleagues argue that gazetteers have been more useful over a much longer stretch of human history than have maps. Perhaps more importantly, they see gazetteers as a form of knowledge and geographic information that reaches from ancient history to the present, for the digital place-based information systems that people carry everywhere in their laptops, smartphones, and cars are essentially digital gazetteers (Berman et al. Reference Berman, Mostern, Southall, Mostern, Southall and Berman2016; Mostern and Southall Reference Mostern, Southall, Mostern, Southall and Berman2016). The United States, Canada, and many European nations have produced gazetteers and their cousins, directories and registers, which have already proven to be highly suitable sources for translation into GIS form (Churchill and Hillier Reference Churchill, Hillier and Knowles2008; Fyfe and Holdsworth Reference Fyfe and Holdsworth2009). More focused research on the form, content, mapability, and comparative potential of historical gazetteers promises to deepen and expand social scientists’ understanding of past places and change over time.

The third new path is bringing GIS deeply into the realm of textual analysis that has, until very recently, been the domain of the digital humanities. In collaboration with experts in corpus and computational linguistics at Lancaster University in the United Kingdom, Gregory is developing new ways of combining GIS with computational methods of textual analysis. Much of the promise of the hybrid approach is evident in Gregory's first publication in this vein. Gregory and literary scholar David Cooper mapped the movements and language of English Lake District poets Samuel Taylor Coleridge and Thomas Gray, specifically connecting the writers’ poetic descriptions and emotional evocations to the locations or views to which they refer, or which the poets were known to have seen on their rambles (Cooper and Gregory Reference Cooper and Gregory2011). As Daniel Alves and Ana Isabel Queiroz's study of urban space and literary representations of Lisbon shows, marrying deep textual analysis with spatial analysis merges traditionally humanistic source material with social scientific mappings of space (Alves and Queiroz Reference Alves and Queiroz2013). Similarly, Lafreniere and Gilliland have incorporated textual analysis from diaries to recreate the social and built environments in London, Ontario, and to map the time-space patterns of daily life in the nineteenth-century city (Lafreniere and Gilliland Reference Lafreniere and Gilliland2015). Marrying GIS with textual analysis holds enormous potential to visualize the social, semantic, emotional, and perhaps even psychological dimensions of emplaced human experience (Knowles et al. Reference Knowles, Westerveld and Strom2015).

Last but not least, there is growing interest among digital humanists from many disciplines in the differences and connections between topography (the physical environment) and the topology of networks, meanings, and human perceptions of social space and place. My own particular interest in this diffuse field of interest comes from my involvement in the Holocaust Geographies Collaborative, which is trying to develop methods that will enable us to connect victims’ trajectories through the Holocaust to the territorial spaces and specific places captured, remade, and utilized by the Nazis and their allies. The latter can be quite appropriately modeled and represented in GIS (Knowles et al. Reference Knowles, Cole and Giordano2014), but GIS is almost entirely unsuited to represent individuals’ traumatic, highly subjective experiences of displacement, violence, loss, and survival. Topography and topology are rich, complex concepts that scholars in many disciplines are exploring, theoretically, empirically, and representationally. Social science historians have much to contribute to the evolving discussion.