Engaging the Archive and its Absences: Lessons From New York City’s Nineteenth-Century Spanish-Language Press
Pace University, United States of America
This presentation considers the use of geospatial visualization as a means of recovering the Spanish-language press of the late nineteenth century in New York City. Although New York City served as a thriving Spanish-language publishing center in the nineteenth-century (and many of the leading institutions of that community held offices in the neighborhood surrounding Park Row during its heyday as the center of the popular, English-language newspapers of the city), much of the history of New York’s Spanish-language press during that period remains understudied. Moreover, with most of the sources for uncovering this history scattered (often spottily) across archives, few resources exist for engaging students with the early history of Latinx writing in New York City.
"Engaging the Archive and its Absences" draws on a mapping project, C19LatinoNYC.org, that I have been conducting with students in introductory Latinx literature courses, which involves plotting addresses found in archival sources to reimagine the community of writers, editors, printers, booksellers, who once led New York’s nineteenth-century Latinx press. I will consider cultural mapping as a research and pedagogical tool for confronting absences in the archive and for making history not just knowable, but also teachable in new ways that enable students to critique and confront structural inequality and systematic oppression. This presentation speaks especially to those who research and teach courses in Latinx Studies. It is also meant to spark interdisciplinary conversation, especially among those working in adjacent fields that must confront absences and omissions in the archive, including hemispheric studies, black Atlantic studies, and indigenous studies.
OpenGazAm: Digitization and Toponym Resolution for a Historical Gazetteer Of The Early Modern Americas
1International Institute of Social History (Netherlands); 2Meertens Instituut (Netherlands), Tilburg University (Netherlands)
In 1797, Boston-based geographer Jedidiah Morse published the first edition of his momentous ‘The American Gazetteer’. It includes around 7,000 unique place name descriptions in the newly founded United States and in the European colonies in both North and South America, and the Caribbean.
The American Gazetteer provides a unique contemporary view of the Early Modern American contents. Its entries range from just a couple of words to several pages and contain basic information on the geographic location and administrative hierarchies of the localities, as well as descriptive notes. Much emphasis is placed on distances, navigability of waterways, types of traded commodities, climates, facilities, and so forth – all relevant for merchants seeking new fortunes.
The goal of the OpenGazAm-project has been to create a Linked Open Data Gazetteer that will be interoperable with the World Historical Gazetteer and Pelagios. The HGIS de las Indias, a recently finished GIS and Linked Open Data representation of an eighteenth-century gazetteer of the Spanish Americas, is already incorporated in both platforms. It will be linked to Morse’s contemporary gazetteer, thus creating a valuable combined resource. Digital historical gazetteers such as the HGIS de las Indias and The American Gazetteer are indispensable in modern humanities research. Existing non-historical digital gazetteers, such as GeoNames, have much difficulty in identifying and disambiguating historical toponyms. Spelling variations, changing place names, and discontinued localities, are omnipresent in historical sources and hamper quick and easy identification of places.
In this paper, we will present the results of the OpenGazAm project, and focus on two of the main challenges in its creation: our approach to the digitization of the printed text (1) and an evaluation of toponym resolution techniques in order to correctly identify (2).
Digitization. No edition of the text exists. Therefore, we had to resort to OCR techniques in order to create a near-golden standard version of the text. We will present a short evaluation of different techniques. By far the best results were reached with the Handwritten Text Recognition toolkit Transkribus, further enhanced by applying a novel toolkit for OCR post-correction, Text Induced Corpus Clean-up (TICCL). TICCL is part of the Dutch national CLARIAH infrastructure for the humanities, and specially adapted for this project.
Toponym resolution. A particular challenge is identifying the correct toponym, and disambiguating between similarly-named entities. The American Gazetteer contains for instance 29 Washington’s and 9 Trinidad’s. Therefore, we need to automatically extract contextual information mentioned in the toponym descriptions (county, state, province, ‘… miles S.W. from …’, ‘along the river …’) in order to successfully disambiguate places. However, such contextual information is also time-specific. For example, the system must recognize that ‘Huntsville in Georgia’ is now equal to ‘Huntsville in Alabama’. We have manually identified 200 places and linked them to existing (mainly modern) digital gazetteers. Here, we will evaluate different techniques for toponym resolution in Morse’s gazetteer and make recommendations for best practices.
Changing Places: Using Spatio-Temporal Maps to Link Literary Texts with Movement
Université de Neuchâtel, Switzerland
We can study the relationship between a literary text and place in several ways, such as by close reading the text itself and prioritizing described places or by visiting real places that are mentioned in a text and doing scholarship in situ. In digital humanities, text and place have often been paired through GIS mapping of literary locations, for example “LitMap” by Dr. Barbara Hui, which maps extracts of W.G. Sebald’s novel Rings of Saturn. As Hui describes, computation allows us to “read literature spatially.”
With computation we can also add a time element so that texts can be read spatio-temporally. While we are accustomed to data visualizations that combine maps and timelines to present data in an experiential manner, such as population change over time, adding a time element to a literary map can evoke a sense of movement between places, which adds a new layer to literary mapmaking. By highlighting change or patterns of stability/instability, spatio-temporal maps link text to both place and movement.
In this paper, I will present two digital humanities projects that map literary archives spatio-temporally. “Summer of Darkness” is an iOS app about the summer when Frankenstein was written that comprises original letters, journal entries, and literary texts. “Mapping Bishop” is a web-based spatio-temporal map of Elizabeth Bishop’s texts from the early 1950s that is under active development. Both projects combine mapping, a textual archive, and a playback element to create an experiential or mimetic approach to literary scholarship. I will also discuss preliminary observations from my mapping of these archives.
An Experience of Digital Landscape: Iconography, Interaction, and Immersion
UNC Pembroke, United States of America
The study of cultural landscapes often takes advantage of mixed methods approaches, incorporating qualitative description, quantitative measure, or both, to understand space and place. The separation between the qualitative and quantitative is often driven by the difference in the theoretical underpinnings or applied goals that guide the study’s attempt to understand this ‘real world’ place. When considering a digital landscape, studies tend to focus on the quantitative aspects of the representation, which is understandable as the landscape itself is now made up of nothing more than 1s and 0s. However, the person maneuvering through the digital landscape, as well as the writers and artists who created the landscape, likely do not think of the space in only a quantitative way, but instead focus on their experience of the place that is characterized in much the same way that someone experiencing a ‘natural’ environment would.
Through the majority of landscape studies there has been an emphasis on what can be physically seen, yielding descriptions that have revolved around aspects such as what is visually present (or absent), the potential for interaction, and the observer’s visual perception. The Japanese landscape architect, Tadahiko Higuchi offered an approach to consider the visual and spatial structure of landscape which is readily applicable to both real world landscapes and digital, or virtual, landscapes.
This presentation will show how the visual indices and composites suggested by Higuchi can be applied to both real and virtual landscapes in three examples. The first will show how these indices and composites can be used to interpret a modern natural landscape. The second will show how these indices and composites can be used to consider a virtual landscape within a video game environment. The third will show how Higuchi’s tools, which were originally used to plan landscape experiences, can be used to design landscapes for a virtual environment. These three examples will highlight ways we can approach both the interpretation and creation of our experience of landscape and how the separation of the real and virtual is not so great when looking at our environments.