No Longer Conforming to Stereotypes? Gender, Political Style and Parliamentary Debate in the UK

Wendy Chavez Tacuri & Nico Solidoro

Introduction

In this presentation, we will summarize and discuss the article titled:
“No Longer Conforming to Stereotypes? Gender, Political Style and Parliamentary Debate in the UK”

This study investigates how gender influences the linguistic style of political speech. The authors use computational methods to analyze a large dataset of campaign speeches.

Research Question

The central research question is: Do male and female candidates adopt different linguistic styles, and are these styles influenced by traditional gender norms?

This question is motivated by common assumptions that women speak in more emotional and inclusive ways, while men are more assertive and analytical.

Why does this matter?

Understanding gendered speech patterns in politics can help us:

  • Challenge stereotypes of political behavior
  • Understand broader shifts in representation
  • Track changes in democratic communication norms

Traditional Stereotypes

Gender and Political Style

  • Women: expected to be cooperative, collective, emotional, and less confrontational
  • Men: expected to be individualistic, rational, and aggressive

These norms shape public expectations and influence how female MPs may behave or be perceived in parliamentary settings.

Text Analysis Method

The authors used quantitative content analysis on UK parliamentary speech transcripts. They built a dictionary of terms aligned with 5 stylistic dimensions:

  • Rationality
  • Collectivity
  • Individualism
  • Aggressiveness
  • Emotionality

Each MP’s speech was analyzed using this lexicon to classify the style.

Method details

Steps in the Analysis

  1. Gathered debates from 1997-2017
  2. Cleaned and segmented speech text by MP
  3. Applied dictionary to label style
  4. Aggregated results per MP
  5. Ran aggression models to identify predictors

This approach allowed them to quantify style and link it to gender, party, role, and time.

Dataset Description

What Data Was Used?

The authors relied on official records from the UK Parliament, specifically, the Hansard transcripts, which document every speech made in Parliament. The dataset covers a 20 year period, from 1997 to 2017.

In total they analyzed around 1.5 million individual speech segments from over 500 different Members of Parliament.

So what kind of information did they collect?

  • the gender of the MP,
  • their party affiliation,
  • whether they were in the government or opposition at the time,
  • the year the speech was given,
  • indicators that capture aspects of speech style, like how emotional or aggressive the language was, based on the text analysis.

Ensuring Reliability

How the authors made sure their analysis was reliable and trustworthy:

  • they carried out manual checks on a subset of the speeches.
  • they cross-referenced their results with findings form earlier LIWC (Linguistic Inquiry and Word Count) based studies. LIWC is a tool used to measure things like emotional or rational language.
  • they used multilevel regression models to analyze the data in order to avoid statistical issues like clustering or repeated observations form the same person.

LIWC tool

The LIWC tool measures: * Use of pronouns (e.g., I, we) * Articles and prepositions * Emotional tone * Cognitive complexity * Social and inclusive language

These are important because they reflect unconscious stylistic habits.

Findings: Persistent Gender Differences

Some small but consistent gender patterns exist:

  • Female MPs more likely to use:
    • emotional, collective, and positive language
  • Male MPs more likely to use:
    • rational, aggressive language and more articles and prepositions

These differences remain consistent across parties and time but are slowly narrowing.

Findings: Individual Differences

The strongest finding is that individual differences in style are more important than gender differences.

Each candidate has a distinctive way of speaking. This variation overshadows any general trends related to gender or political party.

Findings:

The context matters a lot.

  • For example, female MPs who are in the opposition tend to speak just as confrontationally as their male counterparts.

  • Also, the study found that party affiliation and whether an MP is in government or opposition are actually stronger predictors of speech style than gender alone.

Hence, gender still plays a role, but it’s not the most powerful factor shaping how MPs speak.

Theoretical Contribution

This paper challenges gender stereotypes in political communication. The authors show that style is not rigidly tied to gender norms.

Instead, the evidence suggests that style is personal, and that both men and women adopt a range of stylistic strategies.

Broader Implications

This paper also suggests that the increasing presence of women in politics does not automatically lead to more “feminine” styles of speech.

Instead, both men and women are adapting and converging in their use of language, often defying traditional expectations.

Critical Analysis

There are limitations:

  • the dictionary method is a bit limited, it can’t pick up things like irony or rhetorical nuance.
  • there’s also no intersectional lens, so it does not explore how race, class, or age might affect speech.
  • it focuses only on parliamentary speech, nothing from media or interviews.

Suggestions

  • They could use advanced NLP tools like BERT, for more nuanced analysis.
  • They could also include interviews or surveys to understand how MPs view their own speech.
  • they could look at how speech is framed in the media

Recap

  • Gender matters, but not as much as expected; in fact,context matters more
  • Personal style is more significant than group identity
  • Computational methods offer powerful tools for analyzing political communication

Questions and Discussion

Thank you for your attention.

Here are two questions to consider:

  1. To what extent do media and voter expectations still reinforce gendered stereotypes, even if candidates don’t?
  2. Could this kind of stylistic analysis be applied to debates or interviews, and would the findings differ?