Interim Report on the LatAm Gazetteer of Colonial Latin America

Ben W. Brumfield, 14 September, 2019

LatAm is a collaborative effort to produce: (i) a historical gazetteer for Latin America covering the period at the end of the eighteenth century; and (ii) a digital edition of the two primary sources for that gazetteer, Antonio de Alcedo’s ​Diccionario geográfico-histórico de las Indias Occidentales ó América​ (1786); and George Alexander Thompson’s Geographical and Historical Dictionary of America and the West Indies (1812), an expanded, English-language version of Alcedo’s Diccionario.

Shortly after our project got under way we learned of HGIS de las Indias, developed by Werner Stangl, who has generously contributed that project’s data to what we hope will be a growing initiative. HGIS de las Indias includes over 13,000 detailed records of historical lugares (places) in Hispanoamérica and their dynamic spatial relationships with over 1,400 territorios (territories) between 1701 and 1808. A small percentage of HGIS de las Indias records are derived from Alcedo’s Diccionario, which will provide a useful overlap with our own transcription and encoding of that text.

While the ingestion of HGIS de las Indias into WHG (and the testing of the Linked Places format) progressed, the LatAm team has been focusing on Alcedo’s and Thompson’s volume 1, editing, correcting and indexing places in the collaborative edition platform FromThePage.

After conducting the evaluations of the quality of available digitizations of both Alcedo and Thompson, we improved the OCR texts and developed protocols for tagging them following FromThePage standards. You can find an explanation for this work and the scripts we used at HD CAICYT Lab’s github repository.

With the digitized volumes and the environment ready in FromThePage at University of Texas servers, we started our editing work, using the categories for colonial Latin America used in HGIS de las Indias, a stellar project conceived and developed by LatAm member Werner Stangl to index places.

Figure 1 — Thompson with place names encoded in FromThePage

HD CAICYT Lab also conducted an index-a-thon at The Ravignani Institute (University of Buenos Aires) on September 6th, in Buenos Aires, as part of the Argentina’s Science Week:

In October, LLILAS Benson Latin American Studies and Collections members will carry out a second index-a-thon. This indexing work will feed the data needed for our gazetteer, completing and improving the data from HGIS de las Indias.

LatAm has as its goal the contribution of place data encoded from the Alcedo and Thompson transcriptions to both the World-Historical Gazetteer (WHG) and Pelagios projects. WHG and Pelagios now share a contribution format (Linked Places format) and, going forward, will have compatible “registry” back-ends indexing ingested data.

As an offshoot of the LatAm grant work, during these months, WHG lead developer Karl Grossner made a conversion of lugares data from HGIS de las Indias into the Linked Places format, and Pelagios lead developer Rainer Simon ingested it into the Pelagios system, where it is now available as an optional gazetteer in the Recogito transcription tool.

The same data, with the addition of territorios, will also be the first major contribution to WHG, which will have an alpha release in early 2019. This has proven to be an invaluable exemplar for the joint development work of WHG and Pelagios. Going forward, there will be a wide variety of contributions in Linked Places format to the WHG and Pelagios aggregation systems. At one end of that spectrum are relatively simple records transcribed from historical texts and maps, including as few attributes as a name, a place type, its source, a containing or “parent” administrative division or state, and a link to corresponding record in some authority (e.g. GeoNames).

Figure 2–13,214 lugares from HGIS de las Indias

At the other end of the spectrum are historical GISes (HGIS), with detailed records, often including dynamic relationships between places and numerous attributes relevant to a particular region and period, and to a particular set of research questions. HGIS de las Indias is in this category, as is the China Historical GIS at Harvard.

Because the WHG and Pelagios systems are receiving data from sources of varying complexity, both the standard formats for places and annotations and the ingest “pipelines” and practices offered must be both generic and somewhat elaborate — a difficult balance to strike.

The ongoing work of transcribing and encoding Alcedo and Thompson will produce a dataset that is complementary to what is now in place from HGIS de las Indias, substantially filling out the growing coverage of colonial-era Latin America places made available in the graphical interfaces (GUIs) and APIs of World-Historical Gazetteer and Pelagios’ Recogito and Peripleo. This has been greatly aided by the unforseen and fortuitous joining of forces with Werner and HGIS de las Indias.

Our digital edition can be accessed at: https://fromthepage.lib.utexas.edu/benwbrum/latam-digital-edition-and-gazetteer

More about the HGIS de las Indias gazetteer: https://github.com/LinkedPasts/linked-places/tree/master/examples

LatAm is

• Gimena del Rio Riande, Nidia Hernández, and Romina De León from CAICYT-CONICET Humanidades Digitales Lab (Argentina)
• Karl Grossner from the World Historical Gazetteer (USA)
• Werner Stangl from HGIS-Indias (Austria)
• Albert Palacios from LLILAS Benson Latin American Studies and Collections (USA)
• Ben and Sara Brumfield from Brumfield Labs, principal investigators (USA)