Abstract This study examines how Spanish political parties, Partido Socialista Obrero Español (PSOE), Partido Popular (PP), VOX, and Unidas Podemos (UP), construct and contest representations of Ukrainian refugees on X using Natural Language Processing (NLP), Critical Discourse Analysis (CDA), and statistical validation. PSOE exhibits higher positive sentiment and lower hate speech; PP shifts toward greater positivity; VOX displays negative sentiment, fear-related emotions, and elevated hate speech; UP supports refugees while criticizing systemic inequalities. Linguistic analyses of deixis, modality, and verb voice complement these patterns: inclusive pronouns and active constructions align with humanitarian discourse, whereas passive voice, deontic modality, and distancing deixis index securitization. The Spanish case is examined due to the implementation of the EU Temporary Protection Directive and relatively favourable public framing, enabling analysis of how inclusionary and exclusionary logics manifest in digital discourse. Integrating affective/structural cues, the study shows how political actors shape perception and legitimize inclusion or exclusion.
Dehdarirad et al. (2026) studied this question.