The rise of "gut feelings" in US political rhetoric

(Press-News.org) Discussion of "alternative facts" has gained sad notoriety in US politics. Yet the question has been around much longer: How do people conduct political debates – is the focus more on facts or personal opinions? A team of international researchers led by the Cluster of Excellence "The Politics of Inequality" looked into the matter. The researchers examined political rhetoric in eight million speeches by members of the US Congress between 1879 and 2022 to see if the focus of their language was more on data and facts or personal convictions and subjective interpretations.

The team noticed a significant decline in the use of evidence-based political rhetoric since the 1970s, with a historic low in the present. Over the same period, the researchers observed a decline in legislative productivity, an increase in the political polarization of both political parties as well as growing economic inequality in the US.

Between facts and personal convictions
"In many democracies, there is currently much concern about 'truth decay': the blurring of the boundary between fact and fiction, not only fuelling polarization but also undermining public trust in institutions", explains lead author David Garcia, a professor of social and behavioural data science at the University of Konstanz. Together with his research colleagues in the UK, Israel, Austria and Germany, Garcia examined the political rhetoric used in congressional speeches over a period of more than 140 years. Their focus was on the question of how politicians express their conceptions of truth in their language: are they more likely to use objective facts or personal convictions?

"Productive democratic discourse balances evidence-based and intuition-based conceptions of truth", Garcia says. While evidence-based discourse provides a foundation for 'reasoned' debate, intuition contributes emotional and experiential dimensions that can be critical for exploring and resolving societal issues. However, if facts are given less weight and the balance is off, it jeopardizes political discourse. This is exactly the development the researchers have observed in the US' congressional speeches.

Between 1879 and the middle of the 20th century, the ratio of facts to intuition used in congressional speeches was relatively stable and balanced. After 1940, the balance even tipped towards facts and peaked in the mid-1970s. From 1976 to 2022, however, there was a significant, continual decline in the use of facts in congressional speeches, with a historic low in the present. Both US parties are affected by this downward trend, although the drop has been even steeper for Republicans since 2021. This negative trend not only applies to congressional speeches: the results are similar for analyses of Twitter/X posts by members of Congress from 2011 to 2022, the researchers say.

“One remarkable aspect of our results is the strong association between evidence-based language and performance,” co-author Professor Stephan Lewandowsky, Chair in Cognitive Psychology at the University of Bristol, says: “The more speeches in Congress reflect a reliance on evidence and facts rather than intuition, the better the performance of Congress and the less polarization between parties. Conversely, the increasing reliance on intuition-based language since the 1970s has been associated with a decline in performance and increasing polarization. Clearly it matters how politicians use language in the legislature.”

The method behind the study
How can you analyse eight million congressional speeches? The team led by David Garcia relied on computational data analysis methods. "We undertook a massive effort to track long-term trends in how the language of the U.S. Congress has evolved, by analyzing Congressional records spanning nearly fifteen decades," first author Segun Aroyehun explains. "We used advanced text analysis to assess the meaning of words in speeches and compared them to the meaning of words in dictionaries capturing conceptions of truth. This allowed us to observe the focus of speeches over time."

They started by identifying representative, distinctive key words linked to either evidence-based or intuition-based rhetoric. The list included 49 key words for fact-based language (e.g. terms such as "analyze", "data", "findings" and "investigation") and 35 key words for intuition-based language (e.g. "point of view", "common sense", "guess" and "believe").

The team then calculated the ratios of the respective categories of key words used in eight million texts. The resulting figure, EMI or "Evidence-Minus-Intuition", describes the relationship between evidence-based and intuition-based rhetoric. A positive EMI indicates a higher proportion of facts, while a negative value indicates a higher proportion of personal opinions.

The analysis method was developed in a previous project. Jana Lasser, who has been Professor of Data Analysis at the University of Graz since 2024, examined the linguistic patterns of Twitter posts by members of the US Congress between 2011 and 2022. “Even during this period, there was a change in the argumentation style”, explains Lasser. “Personal beliefs gradually gained in importance and were increasingly presented separately from scientific facts.”

 

 

Key facts:

Embargoed until Thursday, 10 April 2025, 11:00 CEST
(10:00 London time, 05:00 US Eastern Time) Original publication: Segun Aroyehun, Almog Simchon, Fabio Carrella, Jana Lasser, Stephan Lewandowsky and David Garcia, Computational analysis of US Congressional speeches reveals a shift from evidence to intuition, Nature Human Behavior
DOI: 10.1038/s41562-025-02136-2
Link: https://www.nature.com/articles/s41562-025-02136-2 Computational analysis of the political rhetoric used in eight million speeches by members of the US Congress from 1879 to 2022. Study by the Cluster of Excellence "The Politics of Inequality" at the University of Konstanz in cooperation with the Ben-Gurion University of the Negev, the University of Bristol, the University of Potsdam, the University of Graz and the Complexity Science Hub (Austria). David Garcia is a professor of social and behavioural data science at the University of Konstanz. He is a member of the Cluster of Excellence "The Politics of Inequality" and the Centre for Human | Data | Society at the University of Konstanz. The computer scientist is an expert in the field of computational social science. The Cluster of Excellence The Politics of Inequality at the University of Konstanz studies the political causes and effects of inequality from an interdisciplinary perspective. The research focuses on some of the most pressing issues of our time: access to and distribution of (economic) resources, the global rise of populists, climate change and unfairly distributed educational opportunities.   

 

Note to editors:

A photo is available for download here:

 

https://www.uni-konstanz.de/fileadmin/pi/fileserver/2025_extra/der_siegeszug_des_bauchgefuehls_david_garcia.jpg

Caption: David Garcia is a professor of social and behavioural data science at the University of Konstanz. He is a member of the Cluster of Excellence "The Politics of Inequality" and the Centre for Human | Data | Society at the University of Konstanz.

Copyright: Ines Janas

 

https://www.uni-konstanz.de/fileadmin/pi/fileserver/2025_extra/der_siegeszug_des_bauchgefuehls_segun_port.jpg

Caption: Segun Aroyehun is PostDoc in David Garcia's Social Data Science Group at the Univesity of Konstanz.

Copyright: Segun Aroyehun

 

https://www.uni-konstanz.de/fileadmin/pi/fileserver/2025_extra/der_siegeszug_des_bauchgefuehls_jana_lasser.jpg

Bildunterschrift: Jana Lasser is a professor of Data Analysis at the University of Graz

Copyright: Uni Graz/Kanizaj

 

https://www.uni-konstanz.de/fileadmin/pi/fileserver/2025_extra/der_siegeszug_des_bauchgefuehls_stephan_lewandowsky.jpg

Caption: Stephan Lewandowsky is a professor of Cognitive Psychology at the University of Bristol

Copyright: Stephan Lewandowsky

 

END