Female_Objectification_Quantifying_in_Novels

This Repository is code for our paper: Reflecting the Male Gaze: Quantifying Female Objectification in 19th and 20th Century Novels.

Data Introduction

The novel list used in our study comes from the 100 most popular downloaded books from Project Gutenberg as of August 25, 2023. We selected the novels which were published after 1800 and were written in or translated into English. Our final dataset consists of 79 novels, with novel_79_list.txt showing the titles of them.

Measuring methods

Our treatment of female objectification is by defining two bias metrics:

1. Agency Bias: A text exhibits agency bias if male entities are more likely than female entities to appear in the text as grammatical agents.
2. Appearance Bias: A text exhibits appearance bias if “female” words are distributionally closer to “appearance” words than “male” words.

More explaination of this two concepts can be found in our paper.

Following is the flowchart of our research:

Code part

Research and Analysis