Student Reflections: Skills in Transnational History Module

Konrad Lawson
Wednesday 24 May 2017

This guest posting is by Katherine Bellamy, who completed the MLitt in Transnational, Global and Spatial History in 2016 with the dissertation “Ninnimissinouk Networks: The Endurance of Identity in a Transnational Context.” Katherine showed particular aptitude and a well-rewarded curiosity in the skills component of the programme, and made use of geographic analysis, network visualisations, and also impressively mastered some particularly challenging ‘regular expression’ high wizardry to extract and clean data from historical databases. We have invited her to share her experiences.

Katherine

The Skills in Transnational History module proved to be a valuable opportunity for me to explore my interests in the digital humanities, in a way which complemented the broader themes of transnational history. We explored numerous avenues open to historians wishing to pursue digital methods, including the use of GIS (Geographic Information Systems) software as a tool for analysing historic and geographic data. Having previously used GIS myself, though in a purely geographical context, the opportunity to use this tool in a historical context was of particular interest to me. My first project for this module aimed to present John Murra’s theory of the ‘vertical archipelago’ in the Andes with a GIS map. I had initially planned to utilise both climatic and agricultural data from Alexander Humboldt and Aimé Bonpland’s 1805 map, Géographie des Plantes Equinoxales, but ultimately chose to omit the agricultural data as there was no clear way to show the significant variations across different altitudinal ranges. This meant I was unable to clearly demonstrate Murra’s link between the varying agricultural production zones, ecological zones, and settlements as I had first hoped. The first of my final maps depicted the average temperature and population figures within key intendancies, aiming to demonstrate any correlation between population size and environmental circumstance (though the result suggested large population sizes could be sustained at both ends of the spectrum); the second map depicted the altitude of the cities associated with the intendancies of the previous map, alongside the varied vegetation zones within the Andes.

 

These maps drew on various sources, including the aforementioned 1805 Géographie (providing temperature data); an 1875 map of South America authored by Louis Stanislas D’Arcy Delarochette (from which I identified the intendancies/cities depicted on the maps); and an 1822 map of Peru which I georeferenced in order to establish accurate contemporary boundaries.

Whilst the georeference was not entirely precise, it allowed me to create an additional polygon within the GIS software to represent the area of the Arequipa region which previously extended beyond current boundaries. Neither map is by any means perfect, both as a result of inherent issues with map creation (all maps lie!), as well as the broader problems associated with utilising largely qualitative historic data in a quantitative setting. The lack of detailed, accurate quantitative data created difficulties throughout the process of creating these maps.

Good data is of the utmost importance when pursuing these research methods. For the second assignment, which focused on the application of other key skills learnt in the module – namely the creation of relational databases and/or social network analysis – I was able to utilise data which was more appropriate. The source of my basic dataset was a list of 2,855 employees of the Hudson Bay Company (roughly ranging from the mid-eighteenth to mid-twentieth centuries), which included employee names, dates of birth, dates of death, and dates active with the HBC. This relatively rich dataset offered the potential for a substantial database, and even some network analysis. However, the amount of information available for each employee varied significantly, nor was the ordering consistent. As such, I utilised regular expressions in order to limit my selection to only those employees who had complete date ranges, and then further narrowed my selection to include only those active between 1821 (when the North West Company and HBC merged) and 1860. I was left with a list of 211 employees, and pdfs corresponding to each employee with further biographical details and their various postings with the HBC. Inputting all this data into a LibreOffice database, I created a form linking the basic biographical information (obtained from using regular expressions) to the more detailed position information (from the pdfs). On a larger scale, a digital humanities approach could yield valuable results, through the creation of a searchable database, displaying biographical information alongside each individual’s career history with the HBC. I had personally hoped to create some form of social network graph utilising this data, but was unable to do so due to time constraints and the nature of the data. Instead, I used the data to create a series of QGIS maps which depicted key HBC post locations, and how frequently these posts were mentioned over the course of the nineteenth century.

Whilst I was faced with obstacles which often prevented me from achieving what I had initially planned to show, the process of dealing with these challenges in this course proved immensely useful for me. Not only did it demonstrate the challenges faced more broadly by historians wishing to adopt digital humanities methods, it also improved my knowledge and understanding of these tools, and, critically, the best way to utilise them. The digital humanities have been viewed with a certain degree of scepticism, yet it is important to recognise that, when applied correctly, the results of adopting these tools can be extremely rewarding. Ongoing projects such as Stanford University’s Spatial History Project, and the University of St Andrews’ own Digital Humanities Network, continue to demonstrate the value of adopting this interdisciplinary approach, utilising innovative methods to yield results which enrich historical enquiry.

Map Sources

– Alexander Humboldt and Aimé Bonpland, ‘Géographie des Plantes Equinoxiales(1805, Langlois)
– Louis Stanislas D’Arcy Delarochette, ‘Map of South America’, (1875, James Wyld)
– C. Carey and I. Lea, ‘Map of Peru’ (1822, H. C. Carey and I. Lea)