The kepler.gl mapping tools highlight time-space relationships and trends in a dataset. Starting with a CSV file of 130 text entries for the Slave Narratives project as its data source, kepler.gl produced a point map that visualized the locations of interviews with former slaves living in Alabama in 1937. Just the basic point map alone provided insight into where most interviews took place and where former slaves settled within Alabama after their emancipation. The heat and cluster map features articulated this information even more effectively; these visualizations distilled 130 individual location data points into concentrated “masses” that conveyed trends about the volume of interviews in different regions of the state. Unsurprisingly, most centered around the major cities of Mobile, Montgomery and Birmingham. The category feature in kepler.gl took the analysis one step further – not only could I learn about the number and locations of interviews but also about the interviewees themselves. This did require me to switch back to a point map; however, at this point, I was interested in more detailed information and found myself zooming in on specific cities to learn more about the nature of the work that these former slaves performed (the type categories being house, field, both or unknown).

At the same time, there are some limitations on these tools to consider. First, they are only as good as the underlying data. In the case of the Slave Interview project, the dataset included exact latitude and longitude coordinates which made for a very crisp and accurate visualization. Alternatively, the MGG project team encountered some issues in this space which they managed to address somewhat effectively. A considerable number of locations in the Damron guides did not list clear addresses that the project team could link to coordinates. As a result, some of their data entries use generalized city/town locations that produce skewed or misleading results. The project team dealt with this by creating an option to filter out these particular results, leaving the decision to the individual user based on their preference.

Other limitations worth mentioning are the embedded assumptions contained within maps, which Presner and Shepard foreground in their discussion of “spatial humanities.” While I believe that each of the projects and activities lent themselves best to a “traditional” cartography, in other circumstance, particularly when approaching topics in ancient history, a different approach may be appropriate. Presner and Shepard correctly identify these concerns, but do not help the reader to understand alternatives to overcome the modern, Western biases inherent in mapping. Perhaps in practice this would look like a mapping tool that focuses more on one object’s location in reference to another object, landmark or natural feature rather than on GPS coordinates, as ancient peoples would have understood their world more in these terms without a fully developed sense of latitude and longitude.