The Language of the Maze: Analyzing thelanguage of the Greek and Spanish legal genderrecognition frameworks

Legal gender recognition (LGR) laws mark a key point in how states define and regulate genderidentity in Southern Europe. This thesis investigates the political dynamics and specific actors thatinfluenced the legislative processes in Greece and Spain, aiming to identify the divergent pathwaystaken by these two nations. Unlike traditional qualitative assessments, this study uses a quantitativemethodological framework using Natural Language Processing (NLP) to analyze the discourse ofpolitical parties, transgender advocacy groups, and opposition organizations.

The research utilizes a dual-metric approach: a fine-tuned sentiment analysis model to eval-uate the political stance of the actors, and a semantic similarity index to quantify the proximityof their discourse to the finalized legal text. The comparison shows clear differences between theGreek and Spanish cases; specifically, the discourse surrounding the Spanish legislation exhibitedsignificantly higher positive sentiment and semantic alignment with the final law compared to theGreek context. Furthermore, the data indicates that leftist political factions served as the primaryparliamentary supporters, and that both transgender advocacy groups and anti-trans political actorsdemonstrated high semantic proximity to the enacted legislation in both countries. While this anal-ysis cannot establish causal influence, the strong semantic alignment suggests that both pro- andanti-minority groups can influence minority rights legislation; therefore, the direct cooperation ofaffected communities and political parties is a crucial predictor of legislative alignment regardingthe passage of minority rights legislation.