I am working on a paper for the ESEH in France in a few weeks which focuses on the role of an industrialist, J.E. Howard, in supporting the efforts of British government officials and economic botanists to establish cinchona plantations in Asia. Cinchona was a key strategic commodity, used to produce quinine, an anti-malarial drug. It is also a key ingredient in the tonic water used to mix a Gin and Tonic. The British, along with the Dutch, took seeds and plants from western South America and used their botanical garden networks to establish plantations in their Asian colonies. This is one of the more interesting plant transfers during the nineteenth century because it played a key role in supporting imperial expansion. Lucile Brockway explored the role of British government officials in stealing cinchona seeds in South America in her Science and Colonial Expansion, but did not discuss the important role played by Howard.
I’ve done a lot of archival research on this topic, but I thought it would be interesting to see what I could find in the Trading Consequences database. Trading Consequences is a collaborative text mining research project between environmental historians, computational linguists and computer scientists, which mined the House of Commons Parliamentary papers and other collections to identify mentions of commodities and place names. The computer scientists at the University of St. Andrews created a Location Cloud Visualization to explore the results. When you search for cinchona it clearly shows the increased association with India and Ceylon in the 1860s, but I needed to dig down past these web visualizations to see what the database has to say about a particular person. To do this, I extracted every sentence that mentions the commodity cinchona in the Trading Consequences database, ordered them by year and exported a text file. This yields a file with 3762 sentences that mention cinchona.
Uploading this data into Voyant Tools makes it easy to explore some of the patterns in the text as it changes over the course of the nineteenth century.
For example, we can see the initial importance of India (which would include some mentions of the East India Company) and the growing significance of Ceylon (modern Sri Lanka) and Java (modern Indonesia) as the century went on. It is also notable that the words Peru and Peruvian appear less frequently in these British government documents. It would be possible to create these graphs with a proper x-axis along the bottom showing the exact year, but for the purpose of exploratory research, the speed and interactive nature of Voyant Tools shows the trends and makes it possible to click and figure out which years are associated with the spikes and valleys.
Using the same tool, we can see the rise and decline in popularity of an alternative spelling of cinchona, “chinchona”, during the middle of the 19th century. Spelling variations are essential to keep in mind when working with nineteenth-century digital sources. A key word search for cinchona alone would miss a lot of the documents related to this commodity.
More to the point, we can search for the last names of five of the key individuals involved in the transfer of cinchona: Clement Markham, Richard Spruce, the father and son, William and Joseph Hooker, and John Eliot Howard. Markham was a Indian Office geographer who led an expedition to Peru to steal cinchona seeds. Spruce, a botanist, collected further seeds from New Granada. The Hookers were both directors of Kew Gardens, with Joseph taking over from his father in 1865 (we’ve also visualized decades of correspondence with the Kew Directors). Howard was one of the sons in the Howard & Sons company, which produced much of the quinine manufactured in Britain. In addition to his expertise as a manufacturer, Howard was a leading expert on the botany of cinchona. The visualization below shows that while Markham, Spruce and William Hooker were key figures in the initial planning and transfers of the early 1860s, Howard gains significance in the corpus in the years that follow.
When you dig a little deeper into the results, you find that Howard was only found in 25 times in sentences containing extracted by the Trading Consequences pipeline with the words “cinchona”, “chinchona” or “Peruvian bark”. He shows up more frequently in documents discussing cinchona, but often in sentences that don’t mention the commodity explicitly. These results are not particularly large, but we can use the tool to find interesting sentences. For example, the first sentence in our results that mentions Howard comes from a 1857 report from Dr. Royle and confirms John Eliot was involved with the project from the beginning: “Howard, who is the largest manufacturer of quinine in this country, and the person who is probably better acquainted than any one else with all the different kinds of bark known in European commerce, has entered warmly into tho project, and has written to his managing agent in South America to obtain seeds of the best kinds in his neighbourhood.” (East India (Cinchona Plant))
Even with the limited results, it is possible to observe the pattern above, where Howard gains significance in segment 4, and explore the data with the other visualizations, to develop a hypothesis: Howard’s role in establishing cinchona plantations in British India increased during the late 1860s. We can then return to the Trading Consequences commodity search and identify the most important document relating to cinchona in this time period and feed it into Voyant Tools for further analysis (“Copy of All Correspondence between the Secretary of State for India … and the Governors of Madras and Bombay Relating to the Cultivation of Chinchona Plants, from April 1866 to April 1870”). After some experimenting and skimming through the sentences including the word Howard, I found a close correlation between “Howard” and “analysis”, suggesting he played an important role in testing the quality of cinchona bark from British India during these years.
Text mining and visualization techniques provide an additional tool for historians to explore the vast quantities of digitized historical documents. In most cases, these distant reading techniques do not replace traditional historical methods, but they allow us to find the most relevant documents and see interesting macro-trends that merit further exploration. Text mining ten million pages of historical documents requires a collaborative approach, but it is a lot easier to work with Voyant Tools to explore a smaller corpus of documents downloaded from the Internet Archive or Jstor.
This research is funded by the Social Science and Humanities Research Council.
Latest posts by Jim Clifford (see all)
- Call for Participants and Proposals: Canadian History and Environment Summer Symposium - December 12, 2017
- Film Review: Guardians of Eternity - January 25, 2017
- “Two chemical works behind him, and a soap factory in front”: Living and Working in London’s Industrial Marshlands - November 25, 2015
- Tracking Cinchona with Digital Methods - June 15, 2015
- Announcing Quelques arpents de neige Environmental History Workshop XXV - October 4, 2014
- Canadians Heading to Guimarães - June 30, 2014
- GIS and Time - December 17, 2012
- The Environment & Society Portal - July 8, 2012
- Aménagement du territoire et normativité de l’espace. Quelle place pour la forêt? - May 7, 2012
- Downstream: reimagining water - March 7, 2012