Words that hurt : a legal semantic analysis of sexual violence and rape in international judicial bodies of criminal character with the help of natural language processing

Abstract

A lot of attention has been given to issues concerning Sexual and Gender Based Violence, which by many is synonymous to violence targeting only women, while neglecting how sexual violence affects male victims. Yet, there has been even less analysis on the language used to describe these crimes and in particular how International Criminal Law and the decisions of international judicial bodies of criminal character portrait this situation. The legal semantics used to try and to describe sexual crimes in Court documents have been an uncharted territory, creating various questions in regards to how sexual violence and rape against women and men is being portrayed by judicial bodies or even what ideas and messages are being conveyed in the cases where the perpetrators of such crimes have been convicted. With the help of the methodology of Natural Language Processing and the tools of Word Frequency and Word Embeddings, I analyze the Cases of Akayesu, Simić, Ntaganda and Ongwen to identify semantic differences in the portrayal of sexual violence and rape in regards to the gender and the role of the victim in the armed conflict. Key Words: legal semantics, rape, sexual violence, natural language processing, word embeddings, word frequency