Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Examining Gender and Power on Wikipedia Through Face and Politeness

Adil Soubki, Shyne Choi, Owen Rambow

TL;DR

This work tackles how face acts and politeness interact in discourse, focusing on Wikipedia talk pages. It introduces a theory-grounded framework, creates a corpus annotated for face acts, and trains a FA tagger using Llama-3-8B with LoRA to scale analysis to roughly $1.3$ million sentences. The analysis reveals that female editors are generally more polite and show self-humbling face acts, while admin power narrows gender differences; experience interacts with gender in nuanced ways that modulate politeness and face acts. By delivering a scalable tagging tool and large-scale sociolinguistic analysis, the paper advances understanding of gender, power, and politeness in online communities and provides resources for future research.

Abstract

We propose a framework for analyzing discourse by combining two interdependent concepts from sociolinguistic theory: face acts and politeness. While politeness has robust existing tools and data, face acts are less resourced. We introduce a new corpus created by annotating Wikipedia talk pages with face acts and we use this to train a face act tagger. We then employ our framework to study how face and politeness interact with gender and power in discussions between Wikipedia editors. Among other findings, we observe that female Wikipedians are not only more polite, which is consistent with prior studies, but that this difference corresponds with significantly more language directed at humbling aspects of their own face. Interestingly, the distinction nearly vanishes once limiting to editors with administrative power.

Examining Gender and Power on Wikipedia Through Face and Politeness

TL;DR

This work tackles how face acts and politeness interact in discourse, focusing on Wikipedia talk pages. It introduces a theory-grounded framework, creates a corpus annotated for face acts, and trains a FA tagger using Llama-3-8B with LoRA to scale analysis to roughly million sentences. The analysis reveals that female editors are generally more polite and show self-humbling face acts, while admin power narrows gender differences; experience interacts with gender in nuanced ways that modulate politeness and face acts. By delivering a scalable tagging tool and large-scale sociolinguistic analysis, the paper advances understanding of gender, power, and politeness in online communities and provides resources for future research.

Abstract

We propose a framework for analyzing discourse by combining two interdependent concepts from sociolinguistic theory: face acts and politeness. While politeness has robust existing tools and data, face acts are less resourced. We introduce a new corpus created by annotating Wikipedia talk pages with face acts and we use this to train a face act tagger. We then employ our framework to study how face and politeness interact with gender and power in discussions between Wikipedia editors. Among other findings, we observe that female Wikipedians are not only more polite, which is consistent with prior studies, but that this difference corresponds with significantly more language directed at humbling aspects of their own face. Interestingly, the distinction nearly vanishes once limiting to editors with administrative power.
Paper Structure (18 sections, 5 figures, 6 tables)

This paper contains 18 sections, 5 figures, 6 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Theoretical Framework
  4. Face Act Tagging
  5. Dataset
  6. Annotation
  7. Modeling
  8. Data Representation
  9. Experimental Setup And Evaluation
  10. Results
  11. Application and Analysis
  12. Admin Differences
  13. Experience Differences
  14. Gender Differences
  15. Intersectional Differences
  16. ...and 3 more sections

Figures (5)

  • Figure 1: Frequency of face acts for admins and non-admins.
  • Figure 2: Frequency of face acts by editor experience
  • Figure 3: Frequency of face acts by gender.
  • Figure 4: Differences between relative usage of face acts by gender, broken down by non-admins (blue) and admins (red); lines to the right (left) indicate that women (men) perform the face act more often
  • Figure 5: Differences between relative usage of face acts by gender, broken down by inexperienced (green) and experienced (orange) users; lines to the right (left) indicate that women (men) perform the face act more often