A social network analysis approach to examining gendered character positions in popular film narratives

Abstract

This thesis explores what the relational perspectives and tools of social network analysis can add to our understanding of the marginalisation of women within popular Hollywood film narratives. Despite empirical and theoretical suggestions that the female voice is marginalised in Hollywood cinema, we know very little about how dialogue is distributed in popular cinema, or how these patterns of character interaction structure film narratives. This thesis proposes that a network-based approach can illuminate these questions and operationalise ideas about gendered character positions in popular film narratives. There have been a number of papers proposing network models for narrative analysis in the last decade or so, primarily aimed at literature and plays. The research in this thesis aims to be the first to engage seriously with literature from both film studies and narratology to explore what these methods can add to our toolkit for understanding the dynamics of film texts. Thus, as an interdisciplinary project, the research aims both to illuminate the substantive research questions regarding the gendered positions of characters in popular cinema, as well as to develop and refine network-based approaches to the analysis of narrative texts in general. The thesis develops a method which represents film narratives as character interaction networks by recording each line of dialogue between named speaking characters. Using this method, I construct a corpus of 27 blockbuster films. The thesis develops a number of central arguments through examination of this corpus. Firstly, using character interaction networks to explore popular film texts reveals that the narrative marginalisation of the female voice in mainstream Hollywood cinema is more complex than the current empirical benchmarks based on the distribution of speaking characters. Secondly, through a comparative character network-based analysis of Wonder Woman (2017), Thor (2011) and The Hunger Games (2012), I illustrate that our understanding of “female-led” in popular cinema requires more than simply identifying the presence of “strong female characters”, as Hollywood narratives can work to individualise and contain the empowerment of their female protagonists within relational frames. Finally, by taking a closer look at the question of how we relate network centrality measures to the concept of centrality within a narrative, the thesis argues that the character network approach can be a useful tool for revealing underlying patterns in how a story is told through its characters, so long as we think of character networks in terms of dynamic network representations. To this end, the thesis proposes a novel dynamic centrality measure designed to capture character positions within narratives in light of recent developments in the analysis of dynamic networks.

Date of Award	1 Aug 2020
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Eithne Quinn (Supervisor) & Martin Everett (Supervisor)