I created a very simple network graph from the characters in the short story “Sisters” by Anjum Hasan. In the story Jaan hires a maid, Jamini, who she becomes very close to. Jamini then dies and Jaan goes to the slums to try to figure out how and why. The nodes where the five characters in the stories that had names and dialogue. Their edges where if they talked to one another during the course of the story.
My network graph illuminates that Jaan is the central character in the story because she is the only one who conversed with everyone else. She has the highest degree of 4. Shakti and Jamini seem to be the second most connected character with the degree of 2. Without creating a network graph and analyzing the text without digital tools, I would agree that Jaan is the main character, but I would not agree that Shakti and Jamini are tied for second most important. Because this network is so simple, there are many limitations.
For example, I only included dialogue that existed in the story. Shankar is the son of Jamini so it can be assumed that they spoke to each other, although a scene did not exist within the confines of this story. The same can be said with Jamini and her boss’s (Jaan) husband, Javed. These characters knew each other and probably spoke, but the network graph only shows dialogue between characters in the story. If it showed relation perhaps, Jamini would seem as central of a character as Jaan.
I also could of added in secondary edges, like friendly interactions and unfriendly interactions. I could compare this with which characters who spoke to each other and see if characters that had friendly interactions also spoke to more people.
I could of also coded for more characters that were not named and did not have dialogue, such as the baby or the group of slum women. This could of shown more of the division of classes based off of the smaller compounds within a larger network.
Overall it is obvious to me that network graphs can act as a close reading for text. Even with a simple list of characters that spoke to each other acted as a piece of evidence for something that I thought was true just through reading. With more elaborate graphs I’d expect not only confirmation to what I already thought was true but also revealing new information that I had not considered before.