Political Discourse Analysis

Overview

CVX applies temporal trajectory analysis to political speech, tracking how rhetorical strategies evolve over time. This has been validated on real parliamentary data with quantitative results.

ParlaMint-ES Analysis (B8)

Data

ParlaMint-ES v5.0: 32,541 speeches from the Spanish Parliament (2015-2023)
841 MPs (355 female, 486 male), CC-BY 4.0 license
TEI XML + plain text + TSV metadata with speaker gender, party, date

Rhetorical Anchors

8 rhetorical dimensions defined as Spanish-language reference phrases:

Anchor	Description
`ataque_personal`	Ad hominem attacks, personal accusations
`politica_social`	Social policy, welfare, public services
`economia`	Economic policy, budgets, fiscal matters
`emocional`	Emotional appeals, dramatic rhetoric
`institucional`	Institutional references, rule of law
`territorial`	Territorial debates, regional autonomy
`genero`	Gender equality, feminist policy
`seguridad`	Security, defense, public order

Key Results

Gender prediction from rhetoric: F1=0.94, AUC=1.00 — rhetorical profiles strongly differ by gender
Party > Gender: Within-party distance (1.084) < within-gender distance (1.231) — party affiliation drives rhetorical similarity more than gender
COVID impact on female MPs: Counterfactual analysis shows female MPs’ rhetorical trajectory diverged significantly from pre-COVID trends, with increased focus on politica_social and emocional anchors

CVX Features Used

Feature	Purpose
`project_to_anchors()`	Rhetorical proximity profiles per speaker
`anchor_summary()`	Mean position + trend per anchor dimension
`counterfactual_trajectory()`	What-if analysis (COVID divergence)
`cohort_drift()`	Gender/party group-level drift
`temporal_join()`	Convergence windows between political groups
`granger_causality()`	Cross-group rhetorical influence
`discover_motifs()`	Recurring rhetorical patterns

Notebooks & Reports

Notebook	Focus	Status
B8_parlamint_polarization	Real ParlaMint-ES data, gender/party analysis, COVID counterfactual	Complete
B7_political_polarization	Synthetic political polarization framework	Superseded by B8

Interactive Results

Political Rhetoric Analysis (B3) — Full interactive tutorial with temporal analytics on political speech: rhetorical drift, partisan clustering, counterfactual analysis

RFC-007: Advanced Temporal Primitives (temporal join, Granger, motifs)
Research: Political Rhetoric — Methodology and theoretical background