Theorising the Spatial Humanities

The Spatial Humanities (SH) is a discipline that emerged after the spatial turn, going hand-in-hand with the democratization of geographic information technologies. The interest in locating geographical, vague and imaginary space and place in History, Archaeology, and Literature has opened a new world of possibilities in humanities investigation, providing the opportunity to revisit traditional questions from new and innovative approaches. Usually set in a highly interdisciplinary scene, research making use of spatial analysis and other disciplines such as computational linguistics and natural language processing have led to the development of tools, techniques and methods that can answer questions that were impossible to address before. In doing so, however, the focus of the SH has not only been the development of new methodologies and tools for analysis, but also the creation of a theoretical understanding of multiple constructions of space through time, culture and/or geography.
This session is concerned with the use of Spatial Technologies in Humanities research, aiming to explore and discuss modern geospatial techniques and their contributions to Humanities, as well as the underpinning theories behind them. Therefore, the session has three interconnected objectives. Firstly, it will carry out an overall examination of current geospatial technologies, methods and approaches in the Digital Humanities. This will be a valuable exploration of the use of GIS and other technologies across diverse disciplines, the associated research tools and methodologies, and the pros and cons of these. Secondly, it will aim to look at different ways of visualising and understanding data within a spatial environment, from GIS-based methods of modelling space and time, to a variety of other techniques, including graph theory, deep mapping, and more. This aspect of the session will have the purpose of showcasing novel developments and methodological innovations in the Spatial Humanities including Linked Open Data, Natural Language Processing, and Machine Learning, amongst others. Bringing the first two aims together, the final aspect of the session will be concerned with the discussion of the theoretical underpinnings of spatial technologies and their use in Humanities research.
The session aims to bring together researchers from all disciplines engaged in geospatial innovation, both within and beyond the humanities. This includes scholars from disciplines such as History, Geography, Literary Studies, Linguistics and Archaeology, as well as GISc, DH, Computational Linguistics and Computer Science. The discussion will revolve around six presentations where critical aspects of the SH will be examined, as well as a theoretical overview of the subject, and new pathways to explore in the future will be also discussed. In doing so, the session will include the debate of concepts such as chorography or biographies of landscapes, and it will reflect the ongoing interest in new ways of visualizing spatial data, in terms of not only time, but also representing constructions of multilayer and interactive space.
1.
Re-thinking Geographic Information Systems for the Digital Humanities

Carmen Brando (Ecole des hautes études en sciences sociales)
The interest for GIS approaches to support digital humanities has grown as Humanities scholars have integrated this methodology into their practices. Beyond geo-visualisation, GIS provides with a reasonably easy-to-use toolbox for performing queries and analysis, and customising cartographies according to semiological practices. To make the most out of GIS, information must be structured relying on a layered-based data model which imposes the definition beforehand of rigid categories, a limitation that can find its origin in database systems. GI science has analysed this issue and states that formalizing semantics of GI is indispensable (Kuhn, 2002).
Modelling of Humanities data relying on ontologies is challenging. Nevertheless, Linked Data (LD) approaches allows for gaining in expressivity by reusing and extending ontologies and thesaurus thereby to create graph-based knowledge bases (KB). Also, semantic historical gazetteers (Pleiades) are a simple mean to bring structured knowledge into sources during annotation. Here, gazetteers must be understood as much more than a list of names and coordinates, they are instead KB including historical facts, useful for interlinking datasets online. Such paradigm can be further generalised to become the next generation Historical GIS, graph-based and interconnected.
Furthermore, GIS assumes that geography, as stated by Andrews (2018), can be treated as a set of specific points, connected into paths or polygons where appropriate. Such assumptions lead to consider that geometries and location prevails over semantics and that place represents an isolated entity without semantic connections. On the contrary, a place evolves and interacts in time with other entities through relationships that one can semantically qualify, and all this information can always be traced back to the source, a key issue in historical research and Literary studies. CIDOC-CRM and its spatiotemporal extension, promote such paradigm for documenting cultural heritage sources. Only time will tell how scholars adopt this formalism.
Yet today, spatialising texts represents a complicated task for humanists as they must tag, disambiguate and geolocate named entities. Several NLP pipelines facilitate the transition from texts to maps and link to KB (Paris et al., 2017), but still tool adaptation is tedious and time-consuming. GIS was not in fact built for handling texts or any unstructured content, and it may better if it stays that way, however further effort is need to facilitate synergies between multiple methodologies.
We, spatial humanists, may contribute to re-think GIS by considering the aforementioned challenges. Do we wish to move on that direction? How can we move forward? In my presentation, I intend to elaborate on these questions based on ongoing project experiences.
References

Andrews, T. (2018). Wait - When, Where? Difficult Answers to Simple Questions About Historical Time and Place, Proceedings of the 2nd Workshop on Corpus-Based Research in the Humanities, 25-26 January, Vienna,

Kuhn, W. (2002). Modeling the Semantics of Geographic Categories through Conceptual Integration. Proceedings of GIScience 2002. LNCS, vol 2478. Springer, Berlin

Paris, P-H., Abadie, N. and Brando, C. (2017). Linking Spatial Named Entities to the Web of Data for Geographical Analysis of Historical Texts,
JMGL, 13:1: 82-110

2.
Historicising the Spatial Humanities: The Antiquarian Foundations of Some “New” Ways of Knowing

Christopher Donaldson (History Department, Lancaster University)
The Spatial Humanities is often characterised as a new or emergent area of scholarly activity. The reasoning underlying such characterisations is sound insofar as it identifies computational technologies and methodologies as essential to Spatial Humanities research. In this paper, however, my aim will be to historicise the Spatial Humanities by considering its relation to the pre-disciplinary approaches of early modern antiquaries. More than just drawing this connection abstractly, I shall also demonstrate it practically by drawing connections between current trends in the use of linked open data, GIS, and “deep” mapping and antiquarian conventions such as hyper-annotation and extra-illustration. I shall use these examples to adduce evidence in support of my overarching contention: namely, that the Spatial Humanities represents less a new development than a recasting and elaboration of older models of spatially informed knowledge creation.
A key point of connection that I shall develop in this paper is the link between Spatial Humanities research and the tradition of chorography. Chorography originated in classical antiquity as a category of research focussed on inventorying and collating, at length, the features, character, customs, and manners of particular regions. This practice, combining as it does elements of history, geography, ethnography, and natural philosophy, was adopted and adapted by early modern scholars as a way of characterising the intensive investigation of specific localities. A key work within this context was William Camden’s
Britannia (1586), which, with its subtitle “a chorographical description” (chorographica descriptio), exerted a powerful influence over the subsequent development of antiquarian research in Europe and consequently informed the scholarly culture out of which an interest in space as a paradigm for humanities research has emerged.

Like much current research in the Spatial Humanities, chorography constitutes a process of accounting for the trans-historical character of a region as a landscape persisting through time, and therefore as one populated by markers and memories of the past that exert an abiding influence on perceptions of the character of that region in the present. Chorography, in other words, is a practice that seeks to push beyond the shallow time-depth of conventional planar maps to explore the evolved character of places through time. Thus, not unlike much recent research Spatial Humanities research, chorography is an analytical and descriptive process guided by an awareness of the interplay of past and present on the way in which a given location, locality, or region is perceived, experienced, and understood.
Accordingly, as I mean to show, although modern Spatial Humanities scholarship is defined by an engagement with computational resources, its methods and aspirations are broadly complementary with those of much older scholarly traditions. Acknowledging this connection, as I shall argue in conclusion, offers a salient reminder of the need for those working in the Spatial Humanities to maintain a reflexive openness to the multiple – and, at times, competing – constructions of space that define the study of human civilisation and society.
3.
A Spatial Humanities approach to integrate History and Literature: insights from a decade long experience researching Lisbon’s past

Daniel Alves (IHC, NOVA-FCSH)
Geography has always been a fundamental element in historical analysis. In other disciplines of the Humanities, such as Literature, for example, studies on the representation of space have been occupying an increasing interest. While it is true that Geographic Information Systems and other tools of the Digital Humanities have facilitated what has been termed the spatial turn, a process of reconfiguration and revaluation of the use of space in the social sciences and the humanities (see, for example, the work of Ian Gregory or Franco Moretti), this dynamic is, however, prior to the introduction of digital tools in humanistic research. Perhaps the most relevant reference in the field of historical studies is Henri Lefebvre's pioneering work,
La production de l'espace, 1974. In the case of Literature, among several possible references, see Gabriel Zoran's article,
Towards a Theory of Space in Narrative, 1984. Although Literature is thought and used as a source for History, and several works point to a necessary interdisciplinary vision in the study of the past, integrating fiction with historical sources, there are not many works that add to these two dimensions a spatial perspective. The use of GIS and other tools and methods of extraction, visualization and analysis of geographic information has strengthened this perspective. In the case of Urban History and Literature about Lisbon, there is still an immense potential to be explored, despite almost a decade of work on these subjects for the 19th and 20th centuries. Studies on the use of urban space in literature, on urban musical landscapes, on political sociability in cafes or on memory and space of revolutionary events have already been carried out, allowing now to question the effective contributions of these tools and the interdisciplinarity that they enabled in the historical and literary knowledge about the city. Is it possible to integrate a fictional view on the social space of a city and the historical knowledge about the human activities that unfolded in the same space? How can knowledge about urban history contribute to the study of literature and its evolution over time? To what extent can GIS enhance or limit these two perspectives? We will discuss these issues based on a body of evidence on the work carried out in recent years, particularly those resulting from the Atlas of Literary Landscapes of Continental Portugal and the Lisbon Retail Trade History projects.

4.
Linked Data, Data Enrichment and Public Participation for Deep Mapping: the Case of de Krook Quarter (Ghent, Belgium)

Piraye Hacıgüzeller, Fien Danniau and Christophe Verbruggen (Ghent Centre for Digital Humanities, Ghent University)
Deep mapping has served as a promising concept in digital humanities in recent years building on the unattainable yet inspiring ambition of recording and representing “everything you might ever want to say about a place” (Pearson and Shank, 2001: 65; see Bodenhamer et al., 2015). In our contribution to the session, one of our aims will be to explore recent developments in deep mapping within spatial humanities. Subsequently we will aim to to lay out major technological and conceptual elements for deep mapping in spatial humanities in general and historical deep maps in particular. Specifically we will make  three major points. Firstly, we will suggest that deep mapping in spatial humanities would considerably benefit from the creation of dense
linked data ecosystems through introduction of new standard vocabularies for linking information on, for instance, main places, events and people related to history of “small places” (e.g. urban neighbourhoods). Secondly, we will provide a critical review of
automatic data enrichment pipelines that are used in digital humanities research and how they can be implemented while creating historical deep mapping. Finally, we will place
public participation as a vital part of the process of making historical deep maps and put forward some public participatory mapping (PPM) and GIS (PPGIS) strategies that can be effectively employed to involve the public in historical deep mapping projects. We will argue that focusing on these elements in deep mapping applications, will enhance the flexible architecture and open-endedness of deep maps, enrich their content and shift the focus from the end result (i.e. the deep map) to the process (i.e. deep mapping) in spatial humanities and, as such, significantly contribute to the realisation of the deep mapping-related aspirations prominent in spatial humanities today (e.g. open-endedness, cartographic fluidity).  

We will be illustrating these ideas further by presenting a new deep mapping project initiated and coordinated by Ghent Centre for Digital Humanities at de Krook Quarter of Ghent (Belgium). Since 2017, de Krook is the architectural iconic city library and digital innovation centre in the city of Ghent. It is located in a neighbourhood that holds several iconic and cultural institutes and theatres each of which offer well-documented histories and heritage. Attracting three times the expected amount of visitors, de Krook is a new public place in town, fulfilling a range of functions and rapidly becoming a new identifier of the neighbourhood. The engagement by the citizens, the city council and the university in this place-in-the-making offers an interesting testbed for participative, digital and analogue, methodology and for the concept of deep mapping as a way to captivate the historical as well as the currently shifting dynamics of the place.
References:

Bodenhamer, D., Corrigan, J. and Harris, T. M. (2015).
Deep maps and spatial narratives. Indiana: Indiana University Press.

Pearson, M. and Shanks, M. (2001).
Theatre/archaeology : disciplinary dialogues. London; New York: Routledge.

5. Spatial History between Maps and Texts: Lessons from the Eighteenth Century
Katie McDonough (The Alan Turing Institute), Ludovic Moncla (INSA Lyon), and Matje van de Camp (Independent Scholar)
Richard White described the work of the Spatial History Project at Stanford in terms of the combination of collaboratively-built, digital visualizations and a “conceptual focus on space.” Visualizations refashioned “opaque” and “unwieldy” evidence so that it could become part of the historian’s narrative (White, 2010). More recently the
Digital History and Argument white paper stressed the utility of mixing computational methods–text, network, and spatial analysis–to strengthen arguments. It also highlights the bias lurking under the cover of technologies we depend on in digital history (Arguing with Digital History working group, 2017).

Our
Spatial History of the Encyclopédie project models the combination of text and spatial analysis, building new Natural Language Processing tools to create spatial data about places named in printed reference books. It contributes to advancing historical arguments about the ways that Europeans created and shared spatial knowledge between the Renaissance and the French Revolution. As a result, it also challenges assumptions that historians have made about spatial metadata. These metadata, collected in books and now in digital gazetteers, derive from strategic, imperialist, and elite place-naming practices that have until very recently erased local toponymic traditions from the encyclopedic tradition.

We will share our most recent work to analyze European traditions of spatial information documentation. Learning from Diderot’s
Encyclopédie–the eighteenth-century bookend to the explosion of reference works in the Enlightenment–we can begin to understand why gazetteers from the nineteenth-century onwards illuminate certain geographies and ignore others. We have created a geoparser that does not prioritize geolocation as the end goal, allowing for less specific spatial definitions and maintaining links to textual content that aids in disambiguation and relative location. This contributes a key resource to the community of geohumanities scholars interested in building early modern-specific, often locationally-ambiguous place authority records. It also makes the provenance of spatial data transparent. This is the first step towards two much needed transformations in spatial history practice: 1) awareness of gazetteers’ sources and 2) access to temporally-specific spatial data, or, data about places that is chronologically similar to our evidence (unlike Geonames). Finally, we will discuss approaches to making visualizations of the
Encyclopédie data that refer to source texts reused in
Encyclopédie articles.

Spatial History of the Encyclopédie copes with unwieldy material, but does not stop at visualizing this evidence from the past. It is also not only conceived of as a means to interpreting eighteenth-century geographical discourses and knowledge collections practices. At its core, this is a project that tackles the ethics of doing spatial history, indeed, all geohumanities work. In excavating the history of the European gazetteer tradition through the lens of the
Encyclopédie’s geography content, we illuminate the distant origins of data sets like Wikipedia and Geonames.

References:

White, R. (2010). What is Spatial History?
The Spatial History Project.
https://web.stanford.edu/group/spatialhistory/cgi-bin/site/pub.php?id=29 (accessed 25 March 2019).

Arguing with Digital History working group. (2017). “Digital History and Argument,” white paper, Roy Rosenzweig Center for History and New Media.
https://rrchnm.org/argument-white-paper/ (accessed 25 March 2019).

6.
Towards a spatial linguistics? How far can recent methodological developments in spatial humanities be practically applied for the geo-visualising dialect data?

Jacques Van Keymeulen, Veronique De Tier, Roxane Vandenberghe and Lien Hellebaut (Department of Linguistics, Ghent University), Sally Chambers and Piraye Hacıgüzeller (Ghent Centre for Digital Humanities, Ghent University), Jesse de Does, Katrien Depuydt and Tanneke Schoonheim (The Dutch Language Institute (INT), Netherlands)
Geo-visualisation of data is fundamental to dialect research. Dialectologists have been creating and interpreting language maps for many years (Ghent University, Department of Dutch Linguistics, 2014). Language maps enable dialectologists to present data in a clear and understandable manner. Particular examples include: (a)
objective maps to display the raw language data prior to interpretation,
(b)
isogloss maps which use lines to mark borders between variants and (c)
symbol maps where for example, the same dialect words are used in different geographical locations. However, do the recent developments in the spatial humanities, represent new opportunities for linguistics research? Can the geo-visualisation of linguistics data move beyond flat cartographic representations towards dynamic, interactive spatial visualisations of geo-linguistic phenomena? If spatial humanities techniques can be repurposed for linguistics research, to what extent can they be practically implemented?

The aim of the
Database of Southern Dutch Dialects (DSDD) project (Ghent University, Department of Dutch Linguistics, 2017) is to aggregate and standardise three databases of the
Brabantic,
Limburgian and
Flemish Dialects into one database for the Southern Dutch Dialect area. Now that the integration of the three dictionaries is underway, the project team can turn their attention to exploring how to geo-visualise this integrated dialect dataset, using state-of-the-art web-mapping techniques.

Originally, dialect maps were created MapInfo. The symbol maps were generated on the basis of Kloeke codes, unique codes for places and hamlets in the Dutch language area, included in the dialect data. While this system works perfectly well for the offline creation of dialect maps, the current set-up does not meet the online needs of 21
st Century dialectology. The project team is therefore looking for a new solution to geo-visualise the integrated dataset of the Southern Dutch Dialect Area.

The aim of this contribution is to: a) explore recent developments in the spatial humanities and in particular their application to linguistics research, b) explore how geo-visualisation of linguistics data, and in particular dialect data, can move beyond flat cartographic representations towards dynamic, interactive spatial visualisations of geo-linguistic phenomena and c) explore in how far the currently available technologies could provide practical implementation solutions for the Database of Southern Dutch Dialects.
References
Ghent University, Department of Dutch Linguistics (2014). Dialectkaarten.
https://www.dialectloket.be/tekst/dialectologie/dialectkaarten/ (accessed 31 March 2019)

Ghent University, Department of Dutch Linguistics (2017). Database of Southern Dutch Dialects (DSDD).
https://www.ghentcdh.ugent.be/projects/database-southern-dutch-dialects-dsdd (accessed 31 March 2019)