For my midterm project, I did a text analysis of the play “Votes for Women,” using Voyant Tools. I was curious as to what I could find in analyzing a play that I was unfamiliar with. I based my analysis on the three acts of the play, and compared word frequencies.
“Votes for Women” is 27,705 total words with top words: miss (350); jean (298); stonor (267); john (201); lady (198); women (149); mrs (125); know (109); oh (106); men (98); it’s (93); man (88); think (87); yes (82); woman (74); great (72); lord (72); farn (66); little (65); levering (63).
Below is a visualization comparing the frequency of the top five words in the “Votes for Women” across the three acts. You can interact by clicking on the drop down menu below the chart and choosing which top words to compare.
Act I has 11,265 words with top words: jean (166); miss (159); john (135); mrs (124); lady (123); lord (65); stonor (61); great (54); oh (53); farn (48).
Act II has 7,948 words with top words: women (77); man (56); miss (52); men (51); voice (48); jean (37); know (34); woman (34); stonor (32); it’s (29).
Act III has 8,492 words with top words: stonor (174); miss (139); jean (95); lady (53); john (52); think (30); know (29); men (26); oh (26); women (25). Something particularly interesting about the chart below is that ‘lady’ and ‘john’ are almost always used together.
Sources
I used the Project Gutenberg EBook of Votes for Women by Elizabeth Robins.
Processes
I knew that I wanted to perform a text analysis of this play and that Voyant Tools allows you to compare word frequency over time. However, I couldn’t upload the entire play into Voyant and expect it to recognize the three different acts. So, the first thing I did was create three separate documents for Act I, Act II, and Act III. Additionally I deleted the extra EBook information at the beginning and end of the document, as it was not part of the play.
Then I uploaded all three documents into Voyant simultaneously and it allowed me to analyze the play as a whole, and across acts. I liked the data I gathered from this method, but wanted to know what the top words for Act II were because they seemed to be very different from Act I and III. So I uploaded each act into Voyant one at a time, to perform a text analysis of each individual act. This gave me new information, but not that much more than what I gained from analyzing all three acts simultaneously.
Presentation
Voyant tools gives you a lot of different information and a few different visualizations. I included the trends graph, as it shows top words over time in each act and across the three acts. This allows you to see the presence of different characters across the play.
I also included a word bubble, which draws attention to words used most frequently, but still includes many different common words from the play.
Finally, I thought it was important for context to include some basic statistics about each act in my initial description of the DH project.
Significance
Text analysis can be really insightful when it comes to long works, such as “Votes for Women.” It also allowed me to learn some key information about the structure of the play, without actually reading it (or watching it). From analyzing “Votes for Women” I learned which characters spoke most frequently, some differences between acts, and the type of language used.
What I would be curious about going forward, is what a text analysis of this play would look like if you took the names of speakers out of the document. I wanted to do this as part of my text analysis, but it would have taken too much time and I felt like I still gained valuable information by including the speaker names in my analysis. In Voyant, you can also unselect top words, such as character names, to just analyze other common words. Through this I found that this play uses a lot of gendered language, and frequently talks about men and women.
This tool is really useful for getting a sense of a long piece of text and it outputs numerical information that is not easily obtained through solely reading the play. While text analysis did not give me a plot summary of “Votes for Women,” it allowed me to dig deeper into the text in new ways and it prompted me to ask different questions about analyzing a play than I would have if I didn’t have access to this digital tool.
