See your words: Using Voyant on my MRE script

In this post, I’ll be talking a bit about my graduate research, and about the tool Voyant.

As part of my graduate research  at Carleton University I wrote a script based on a shooting outside of Ottawa, in the township of Goulbourn. One evening in August 1882, Robert McCaffrey was confronted by his lover, Maria Spearman, and her brother, on the side of the road. Maria was reportedly in the “family way”, and the two had sought out McCaffrey in order to arrange a marriage. When McCaffrey refused, a struggle occurred, which resulted in a gun being fired, and McCaffrey’s death. The headline “Murder! Shot Through the Heart” was splashed in newspapers as far away as Washington. Maria and her brother were arrested for murder, and taken to the Carleton County Gaol (now the HI-Ottawa Jail Hostel) to await the upcoming fall assizes. The story was taken up by anonymous writers to discuss the current issues of the day, including women’s rights, and the inequality of the justice system. Public outrage over the death of Robert McCaffrey soon turned to sympathy, and Maria quickly became characterized as a helpless victim, who had no other course but to take matters into her own hands. In the end, although Maria admitted she had accidentally fired the gun, the jury found her not guilty, and she returned to Goulbourn following her release.

The script I’ve written tells this story, using historical records as the skeleton of the piece. I created dialogue by combining verbatim excerpts from primary sources and then using my imagination to fill in the remaining gaps. The script also features characters based off of individuals who were involved in the creation of the script, including myself, that work to highlight the complexity of creating and performing the past. They also are a reflection on the evolution of script, and my journey throughout my research.

This script was performed by The Cellar Door Project at the end of February 2016. Now, I’m in the phase of my research that involves writing a reflection on this project.

A few weeks ago I was introduced to a tool in my Digital History course called Voyant. Voyant is a web-based tool that searches through a text you’ve uploaded, and provides information on words that frequently appear in it. For it can provide you with a graph that shows the trend of a particular word through multiple texts.

Someone in my class had mentioned they had tried Voyant on their thesis/MRE paper, to see what kinds of trends in the words they could see. I decided to give this a try on my own work, specifically with the script.


So, there are a total of 5,315 words. The most frequently used words are the (182), you (162), I (148) etc.

If you click on the small cog wheel, you have the ability to edit out these kinds of words. Select English (Taporware) from the dropdown list, and then click “OK.”


Pretty neat!


Looking at the new cirrus (the word bubble), there are a couple of words that I expected to see “Emily” “Maria” “Chester” “Robert.” These are all major characters throughout the script. I can even generate graphs that show trends in specific words. (Sorry, I went a little bananas!)

By clicking on a specific word in the cirrus, I can see how many times it appears. “Maria” appears just 5 more times than “Emily.”

Seeing “Maria” just barely scrape ahead of “Emily” prompts a wave of guilt. Seeing them side by side hit the issue of authorial presence I’ve be struggling with in my research.

“Emily” (surprise surprise) is based on me. The idea to insert myself into the script came in a roundabout way; I had been speaking with friend of mine, and while discussing my research I remember that he had been the one who had taken me out to the site of McCaffrey’s death when I was doing my early research.  I reminded him of the trip, and to my surprise he remembered it right away, even referencing the music we had been listening to on our drive out.  I was struck by this.  While I have shared my research (and the story of the shooting) with many, I hadn’t realized until that moment that my audience was listening, or that they would might take part of the story away with them.

I decided to write a scene for the script based on this conversation, which eventually became the last scene in the script. I felt that this would help me navigate some of my feelings on the subject, and be cathartic.  When I discussed the conversation I had had with members of The Cellar Door Project production team I was encouraged to cultivate this more. They urged me to consider adding myself as a primary character to the script. Immediately my guard was up.  I didn’t want to include myself in the script, after all, this was supposed to be a play about Maria Spearman.

Greg Dening in book Performances, articulates my reluctance.  He explains that most historians find authorial presence disturbing. Furthermore, he explains that the use of the subjective “I” is seen as “complicated and untrustworthy.”[1]  Historians find authorial presence disturbing because we have been instructed that our writing should be objective.  Bruno Ramirez in his work explains that the application of a structured rationality is inherent to the discipline of history, and is perceived as necessary for the attainment of historical truth.[2]  Logically, historians know objectivity is impractical, as well as unobtainable.  However, a small part of us still clings to the illusion.

Seeing “Maria” and “Emily like this a tangible manifestation of all this. Whether I like it or not, the script I’ve written is just as much about me as it is about Maria. And that, like everything I create, it comes from me.

It’s also pretty moving to see all my research filtered down into a colourful blob of words. Since after all, all the script is is a stringing together or words. It makes me think more about the words I’ve chosen to use, and what they reflect about me.

Here are some of the word trends.




[1] Greg Dening, Performances (Chicago: The University of Chicago Press, 1996), 111.

[2] Bruno Ramirez, “Clio in Words and in Motion: Practices of Narrating the Past,” The Journal of American History 86, no.  3 (1999): 998.