LACOL Digital Humanities 2023

Nahom Ayalew, Jesse Koblin, Kalea Ramsey, Bernard Wongibe

Tools Used 

1. Voyant Tools: Our team used Voyant Tools for text analysis on the Jasper Parrish Papers. The static, exploratory data visualizations we produced, including Cirruses, Line Trends graphs, and a Collocations chart, illuminated topics within the text which expanded our understanding of Native American relations with the U.S. Government in the 17th and 18th centuries. 
2. Recogito: We used the Named Entity Recognition function in the text annotation tool Recogito to extract locations from the Jasper Parrish Papers transcription text file and the metadata descriptions of other objects in our archives. 
3. Excel & Google Sheets: After extraction, the locations are entered as rows in Excel. In Google Sheets, the add-on “Geocode by Awesome Table” generates coordinates for the locations. However, as many Native American locations are either unrecognized or inaccurately assigned coordinates, we are manually checking every coordinate. 
4. Tableau: Tableau is a data mapping software that accepts spreadsheets and allows us to map the locations with a lot of toggling options. Tableau is a fantastic explanatory tool because it recontextualizes our data in a geographic dimension, creating a distinct vantage point from which to view our topic and evoking Native American indigeneity. 

Methods. Did you use a specific method like “topic modeling” or “network analysis?” Did you rely on specific theories or readings to help you understand and apply this method?

Using the tools we described above, our group carried out a spatial analysis of the geographical focus of our colleges on Native American objects in archives and museums. Our methods remind me of two specific methods described in Chapter 2 of digital_humanities by Burdick et al. The first one is ‘distant/close, macro/micro, surface/depth.’ This chapter argued that we should not choose between distant or close, rather, it is good to toggle between views of data, zooming in and out. I think our choice of sources demonstrate that well, since our mapping data came from both the Jasper Parrish Papers (micro) and subsets of larger museum and archive metadata (macro). The second one is ‘locative investigation and thick mapping.’ As mentioned before, geocoding runs inaccurately for Native American locations. Our data has become a locative investigation of itself because of how much attention we have to pay to the context of each individual dot on a map. 

Data Modifications. What did you do to your data? What did you add or take away? Did you consolidate fields or separate them?

For the Recogito-Excel-Tableau workflow, we decided to only use a subset of our colleges’ archive data for practical reasons: we only transcribed a portion of the Jasper Parrish Papers, and we could not directly access an export of the entire archive records, so we are building our dataset with search results on archive websites. The systems our museums’ use allows for an export of all of the objects, and we have received them. We are planning to consolidate the subset of archive data and the cleaned version of museum data, create a separate sheet containing mappable columns of accurate coordinates, accession date, and provenance, and then input everything into Tableau for mapping. Our data consolidation will allow easier manipulation, exploration, and visualization across both our archival and museum data, allowing us to extrapolate information jointly gathered from previously-disparate sources. 

Anything Else?While conducting research and extracting data for this project, our group has reconciled with the understanding that our collegiate education is founded on Native American dispossession and genocide. Amherst is built on Nonotuck land, Vassar is built on Munsee-Lenape land, and Williams is built on Stockbridge-Munsee Mohican land. It is critical our group remains conscious of our privilege to attend colleges which enable our understanding of the significance of Indigeneity.