[Paper Review] Using Sociolinguistic Variables to Reveal Changing Attitudes Towards Sexuality and Gender

Sky CH-Wang, David Jurgens|arXiv (Cornell University)|Sep 22, 2021

Gender Studies in Language4 citations

TL;DR

This paper uses computational sociolinguistics to analyze longitudinal changes in U.S. attitudes toward sexuality and gender by examining lexical variation in gendered vs. gender-neutral language on Twitter and Reddit. It finds increasing use of gender-neutral terms like 'partner' reflects growing acceptance of non-heterosexual relationships, while reduced gender marking in indefinite references signals rising gender equality—driven by specific demographics and catalyzed by same-sex marriage legislation.

ABSTRACT

Individuals signal aspects of their identity and beliefs through linguistic choices. Studying these choices in aggregate allows us to examine large-scale attitude shifts within a population. Here, we develop computational methods to study word choice within a sociolinguistic lexical variable -- alternate words used to express the same concept -- in order to test for change in the United States towards sexuality and gender. We examine two variables: i) referents to significant others, such as the word "partner" and ii) referents to an indefinite person, both of which could optionally be marked with gender. The linguistic choices in each variable allow us to study increased rates of acceptances of gay marriage and gender equality, respectively. In longitudinal analyses across Twitter and Reddit over 87M messages, we demonstrate that attitudes are changing but that these changes are driven by specific demographics within the United States. Further, in a quasi-causal analysis, we show that passages of Marriage Equality Acts in different states are drivers of linguistic change.

Motivation & Objective

• To investigate how linguistic variation in gendered and gender-neutral language reflects changing societal attitudes toward sexuality and gender in the U.S.
• To examine whether shifts in word choice—particularly around romantic partners and indefinite persons—signal broader cultural changes in acceptance of LGBTQ+ identities and gender equality.
• To determine whether policy changes, such as the passage of Marriage Equality Acts, causally influence linguistic behavior in online communities.
• To identify demographic patterns in language use that correlate with evolving social attitudes toward sexuality and gender.

Proposed method

• Identified two sociolinguistic lexical variables: (1) referents to romantic partners (e.g., 'partner' vs. 'girlfriend'), and (2) indefinite references to people (e.g., 'some folks' vs. 'some guys').
• Collected and analyzed 73 million Twitter posts and 14 million Reddit comments over nearly a decade to track longitudinal changes in word choice.
• Used demographic inference to associate linguistic choices with identity groups (e.g., gender, political leaning, sexuality) based on subreddit and user-level data.
• Applied quasi-causal analysis to assess the impact of state-level Marriage Equality Act (MEA) passages on linguistic shifts, particularly in LGBT+ communities.
• Employed statistical modeling to measure changes in relative frequency of gendered vs. gender-neutral forms over time, controlling for demographic and temporal confounders.
• Validated findings through manual review of subreddits and community categorization to ensure alignment with target social groups and reduce noise from content overlap.

Experimental results

Research questions

• RQ1 How has the use of gender-neutral terms for romantic partners changed over time in U.S. social media, and what does this reveal about shifting attitudes toward non-heterosexual relationships?
• RQ2 To what extent does the reduction in unnecessary gender marking for indefinite references reflect increasing acceptance of gender equality?
• RQ3 Which demographic groups are driving changes in linguistic choices related to sexuality and gender?
• RQ4 Does the passage of Marriage Equality Acts in U.S. states lead to measurable shifts in language use, particularly among LGBT+ communities?
• RQ5 How do linguistic patterns in online discourse correlate with broader societal trends in sexual and gender identity acceptance?

Key findings

• The use of gender-neutral terms like 'partner' to refer to romantic partners has increased significantly across heterosexual-identifying communities, signaling growing social acceptance of non-heterosexual relationships.
• There is a statistically significant increase in the use of gendered terms (e.g., 'husband') among LGBT+ communities following the passage of state-level Marriage Equality Acts, indicating reduced social cost of openly identifying sexual orientation.
• The decline in gendered marking for indefinite references (e.g., 'some folks' over 'some guys') is observed across multiple demographic groups, reflecting broader progress toward gender equality.
• Demographic analysis reveals that changes in language use are not uniform—specific groups, including heterosexual and non-binary individuals, show distinct but converging trends toward gender-neutral expression.
• Quasi-causal analysis confirms that the timing of Marriage Equality Act passages correlates with measurable increases in explicit same-sex relationship labeling in LGBT+ communities.
• The study demonstrates that linguistic variation in lexical variables can serve as a reliable, large-scale indicator of evolving cultural attitudes toward sexuality and gender.