Beginning network analysis

My first introduction to network theory was in a sociology course I took as a junior in high school. We read Granovetter’s landmark paper on weak ties. While I found Granovetter’s writing to me far more accessible than that of other sociologists of his day, I found the subject matter quite dull (a sentiment that has significantly changed over the years). Using television and film as an introduction to network analysis would have been much preferable, and I appreciate that this is how the topic is being covered in this course.

Prior to class last week, I had never watched a single episode of Once Upon a Time, so I really had no prior conceptions of the social networks within the show. It was a bit overwhelming to be thrown into a world with so many characters, right from the beginning, and have to attempt to build a web or network of sorts that encapsulates their affiliations (proximity) as well as (perhaps in a separate but related web) their dialogue (who speaks with whom). An initial thought is that it is, perhaps, useful to have watched a movie or film prior to attempting to catalogue these associations. At the very least, this would have helped with character names - I found myself referring to characters by physical features, on occasion, as often characters’ names weren’t used until several minutes after their introductions.

I don’t really have any other experience with which to compare my data collected from Once Upon a Time, but I suspect that features of the show’s plot and setting will lead to network features that are rather unique to the show. For example, it appears that some of the characters may actually have two distinct roles - a fairy-tale character may have a different actor/name/and role within the “real” world (I could be wrong about this, but this is how I understood the population of the town in which Henry lives). Not only does this complicate things (do both roles count as a single person?), it would also lead to interesting communities or clusters between characters.

I noticed one potential shortcoming of the methods we used in class to collect network data. Emma speaks with and is near Henry’s mother in the first episode more often than Henry is. This would give the relationship between Emma and Henry’s mother more weight (in a network diagram) than that between Henry and his mother. However, this would seem to inaccurately capture the relationship between Henry and his mother. Despite hints of a not-so-great relationship between these two, their affiliation is obviously going to be stronger than that of Emma and Henry’s mother.

Written on October 3, 2016 by Josh Guberman