News/Research

Revisited: "Network Analysis of Media Studies"

04 Oct, 2016

Revisited: "Network Analysis of Media Studies"

Miyoko Conley, a BCNM Designated Emphasis candidate in Theater, Dance, and Performance Studies, recaps Miriam Posner's History and Theory of New Media Lecture Series workshop on network analysis for film & media, which took place on September 29, 2016.

The 2016-17 History and Theory of New Media lecture series kicked off on September 29th, with a combination lecture and workshop by Miriam Posner, entitled, “Network Analysis for Media Studies.” Posner, a Digital Humanities core faculty member at UCLA, introduced network analysis as a methodology, and showed how this valuable tool can open new channels for knowledge formation in the humanities.

The workshop began with the simple question, “What is network analysis?” Though a method and a tool used to track and analyze connections between data, Posner emphasized it as a way of understanding the world as a set of relationships. Network analysis can help scholars understand seemingly basic questions like, who knew whom, or how might ideas or formal motifs have traveled? However, network analysis also gives visual form to sets of relationships, which brings to light questions such as, have we [scholars] overlooked important figures in communities of practice? Does a certain community have unusual or unexpected properties? Throughout her demonstration, Posner referred back to not only this theme of relationships, but also to the challenge network analysis brings to dominant modes of thinking.

One way this challenge is issued is through the way network analysis programs lay out data. When looking at a set of nodes (entities) and edges (lines), the brain wants to assign significance to the proximity of particular nodes; however, network analysis graphs do not assign a particular value to proximity. Posner demonstrated this through her own analysis of filmmaker Oscar Micheaux’s cast network. While the example provided a step-by-step guide on how one might start data analysis, it also showed how one might begin to interpret the visual data, including avoiding the pitfall of assuming proximity indicates value.

After showing multiple ways to construct graphs and explaining terminology, Posner continued the theme of connectedness quite literally with a live demonstration. She gathered volunteers and connected them together with string, to illustrate the different ways connections can be read for importance. It is not always the same type of information that is given weight; for example, one might assume that the node with the most direct connections has the most importance, but that does not take into account the nodes that are connected to other clusters, and would therefore control the flow of information.

The question and answer session focused on the question of potential in network analysis. Many of the projects Posner cited illustrated how network analysis in the digital humanities can be used for exploration and experimentation, and to expose normalized structures of thought. This latter idea was also shown through another project in which Posner is involved, Maori Subject Headings, which strives to improve the accessibility of Maori materials in libraries. Since the Library of Congress’ subject headings do not match up with how Maori culture groups objects, Maori materials have become harder to search, particularly for Maori users. Through data analysis, Maori Subject Headings is one way to re-catalogue the information and decenter dominant knowledge structures. By looking at how data analysis offers a representational form for ways of knowing, Posner closed her workshop with not only excitement for its possibilities, but also with an emphasis on critically thinking about the data structures used.

2016 HTNM: Posner