Epistemic orientation in parliamentary discourse is associated with deliberative democracy

2026-04-21 · Source: Computation and Language · Field: Government & Public Sector — Public Policy & Governance, Artificial Intelligence & Machine Learning · Depth: Expert, quick

Summary

Researchers developed the Evidence--Minus--Intuition (EMI) score, a scalable metric to quantify epistemic orientation in political discourse, distinguishing between evidence-based and intuition-based reasoning. This score is derived using large language model (LLM) ratings and embedding-based semantic similarity. Applying the EMI approach to 15 million parliamentary speech segments from 1946 to 2025 across seven countries, the study found a positive association between higher EMI scores and deliberative democracy within countries over time. This relationship was consistent in both contemporaneous and lagged analyses. Furthermore, EMI was positively associated with the transparency and predictable implementation of laws, indicating its relevance for governance quality.

Key takeaway

For policy analysts evaluating democratic health, this research suggests that monitoring the epistemic orientation of parliamentary discourse, specifically the EMI score, can serve as a valuable indicator. A higher EMI score correlates with stronger deliberative democracy and improved governance transparency. You should consider integrating such quantitative discourse analysis into assessments of political system quality and stability.

Key insights

Epistemic orientation in political discourse, measured by EMI, positively correlates with deliberative democracy and governance quality.

Principles

• Truth-seeking is vital for democratic deliberation.
• Epistemic orientation can be quantitatively measured.

Method

The Evidence--Minus--Intuition (EMI) score is calculated using LLM ratings and embedding-based semantic similarity to differentiate evidence-based from intuition-based reasoning in text.

In practice

• Apply EMI to analyze political speech.
• Monitor discourse for epistemic shifts.

Topics

• Epistemic Orientation
• Deliberative Democracy
• Parliamentary Discourse
• EMI Score
• Large Language Models

Best for: AI Scientist, Research Scientist, Policy Maker

Related on AIssential

• Analyzing Debate Dynamics in the Portuguese Parliament with Dialogue Action Flows — Dialogue Action Flows (DAFs) model multi-party interactions by integrating verbal and non-verbal cues probabilistically.
• Epistemic Constitutionalism Or: how to avoid coherence bias — Large language models require explicit "epistemic constitutions" to regulate belief formation and address biases like source attribution.
• The Illusion of Certainty: Why PR’s Obsession with “Cold Data” Is Costing Us True Trust — PR's over-reliance on "cold data" misses real-time sentiment; informal intelligence is crucial for true trust.
• Polar: A Benchmark for Evaluating Political Bias in LLMs — LLM political bias is context-dependent, varying by language, ideology, and geography.
• A Context-Aware Dataset for Stance Detection in Bioethical Controversies on Reddit — BioStance offers a large, context-aware dataset for stance detection in complex bioethical social media discourse.
• Quantifying Media Representation Dynamics Across 25 Years of News Reporting on Policing-related Deaths — News narratives on policing-related deaths disproportionately amplify state perspectives, though civilian voices gained prominence post-2020.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Computation and Language.