This repository contains the experiments, data, analyses, and write-ups having to do with my work on the role of individuals' subjective social beliefs in linguistic processes. The project was advised by my Qualifying Paper Committee, which consisted of Rob Podesva (Chair), Judith Degen, and Meghan Sumner.
Broadly, this project examines the role that individuals' subjective social beliefs about gender (e.g. whether or not it is biologically immutable) influence the way they process and produce gendered and gender-neutral terms. The items under investigation in this particular project are role nouns (Misersky et al 2014), such as 'congressperson' which mark specific social and professional positions in the world. We are particularly interested in those role nouns which show morphological marking for social gender. This can be via compounding, resulting in a ternary split, as in 'congressperson/woman/man', or it can be via affixation, resulting in a binary split, as in 'actor/actress'.
Using a self-paced reading task and a forced-choice task, we tested the role of individuals' beliefs on the reading times and selection rates of terms like ';congressperson'. We found that, in on-line processing, there was no significant effect of gender ideology, such that neither individuals' scores on Baber and Tucker's (2006) Social Roles Questionnaire nor their political affiliations significantly predicted reading times. In the forced-choice, offline task, however, there was a significant effect, such that Democrats used significantly more gender neutral forms than did Republicans. Moreover, within the Democratic party-aligned population, those with more progressive gender ideologies produced more gender-neutral forms than did their more conservative Democratic counterparts. The same pattern was not observed in the Republicans.
A version of this work will be presented at the 2022 annual meeting of the Cognitive Science Society in Toronto, co-authored with Judith Degen and Robert J Podesva. You can read the manuscript here.
You can find in this folder a series of subfolders for each of the four experiments conducted in the course of this project:
- norming_study_one: Norming study for 39 potential critical items (not included in CogSci paper) | Study Link
- 1_Reading_Neutrality: Self-paced reading study (Experiment 1 in CogSci paper) | Study Link
- 2_Maze_Neutrality: MAZE-task reading study (not included in CogSci paper) | Study Link
- 3_Production: Forced-choice production study (Experiment 2 in CogSci paper) | Study Link
Each experimental folder includes the experiments themselves (experiment.html), as well as support folders and a folder for automated stimuli creation(except in the norming study). Thanks to Max Zvyagin for helping with this stimuli creation process.
You are welcome to click through any and all of the experiments; the html pages here do not collect responses or user data.
You can find in this folder my analysis scripts for all of the relevant studies. The 'final' folder includes those scripts whose contents made it into the final paper(s). The 'exploratory' folder includes other analysis scripts, including those in which I was learning how to do some of the analyses!
Admittedly, this part of the repo is a bit messy at the moment, and will be cleaned up in future weeks.
This folder contains my write-ups from this research project. This includes a write-up of my Qualifying Paper (Qualifying_Paper/Papineau-QP1.pdf), as well as our CogSci 2022 paper (CogSci_2022/papineau-et-al-2022.pdf). It also includes all output images used in these presentations.
Again, a bit messy at the moment. Will be cleaned up shortly.
This folder contains the frequency data I pulled from the Corpus of Contemporary American English in the process of adding frequency predictors into various models.