"Undocumented Immigrants" != "Illegal Aliens": Decomposing the Conceptual and Narrative Landscapes of Partisan Immigration Terms

2026-06-25 · Source: Paper Index on ACL Anthology · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Social Sciences & Behavioral Studies, Research Methodology & Innovation · Depth: Expert, medium

Summary

A study presented at the 15th Joint Conference on Lexical and Computational Semantics (*SEM 2026) in July 2026, spanning pages 348–365, investigates whether politically charged terms like "undocumented immigrants" (UI) and "illegal aliens" (IA) differ in meaning beyond speaker identity. Researchers Yejin Cho, Gabriella Chronis, Nitin Sudarsanam, Kevin Barcenas-Martinez, and Katrin Erk employed a dual-track methodology, projecting contextual embeddings into psycholinguistic feature space and extracting narrative scenes using LLMs. They found that in partisan news, "IA" contexts emphasize causation, fear, criminality, and threat, while "UI" contexts highlight consequences experienced, shared humanity, vulnerability, and governance. This research demonstrates that these terms impart distinct construals on migrants, framing them as either agents of harm or patients of circumstance.

Key takeaway

For computational linguists or research scientists analyzing political discourse, this work highlights the profound impact of specific terminology. You should recognize that terms like "undocumented immigrants" and "illegal aliens" are not interchangeable; they actively shape public perception by embedding different narratives. Consider applying this dual-track methodology, combining contextual embeddings and LLM-based scene extraction, to uncover subtle yet powerful framing biases in other politically charged topics within your research.

Key insights

Partisan immigration terms carry distinct conceptual and narrative framings beyond merely indexing speaker identity.

Principles

• Terms shape perception: "IA" frames migrants as threats, "UI" as vulnerable.
• Contextual embeddings reveal psycholinguistic features of language.
• LLMs can abstract narrative scenes from text for discourse analysis.

Method

The methodology involves projecting contextual embeddings into interpretable psycholinguistic feature space and extracting narrative scenes with LLMs to analyze conceptual framing.

In practice

• Analyze partisan language framing in news media.
• Deconstruct conceptual biases in political discourse.
• Apply dual-track methodology to other charged topics.

Topics

• Computational Semantics
• Political Discourse Analysis
• Large Language Models
• Contextual Embeddings
• Narrative Framing
• Psycholinguistics

Best for: NLP Engineer, AI Scientist, Research Scientist

Related on AIssential

• Political Backflow From Europe — European political narratives often misinform American discourse due to differing social and economic realities.
• From 50K to 8.2 Million in 24 Hours: Vozinha's Algorithmic Consecration and the Multilingual Making of World Cup Visibility — Algorithmic visibility for public figures is shaped by distinct cross-lingual narrative frames and platform metrics.
• 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.
• Trump’s Obama and Bad Bunny posts crystallize his political philosophy | Sidney Blumenthal — Trump's political philosophy is rooted in historical racist and nativist ideologies, amplified by modern digital manipulation.
• Polar: A Benchmark for Evaluating Political Bias in LLMs — LLM political bias is context-dependent, varying by language, ideology, and geography.
• The Cosine Similarity Trap: Why Embeddings Can’t Distinguish “War” from “Union” — Language models often cannot distinguish complementarity from conflict due to how embeddings encode relatedness, not relationship.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Paper Index on ACL Anthology.