Metaphor, multimodality and generative AI: Synthetic images prompted by figurative expressions in anti-immigration discourse
- 1Department of Linguistics and English Language, Lancaster University, Lancaster, Lanacshire, United Kingdom of Great Britain and Northern Ireland
Abstract
Complex social processes like migration are frequently understood via metaphors. In right-wing media and far-right political discourses, two metaphors that abound draw on flood and invasion frames to problematise immigration and legitimise hostile attitudes and actions towards it. In so far as metaphor involves the construction of mental imagery, it is intimately bound with visual experience, including semiotically mediated visual experience. In a semiotic landscape increasingly populated by synthetic images, I argue that the meaning of metaphors may be modulated by AI-generated visual content and that this is especially important in contexts of political communication. The study investigates images generated by prompts instantiating an immigration is flood versus an immigration is invasion metaphor to inquire if and how they differ by source-frame and what are the ideological implications of the images returned. Following an initial pilot study, several specific semiotic features are analysed, including colour, properties of the figure (size, shape, orientation) and ground elements (environment, securitising features and frame-consistent features like fire and water). Results show significant differences between metaphorical frames and highlight the potential that AI-generated images have to reflect and reinforce ideological positions.
Related
- Tangled in natural and synthetic microfibers: prevalence, grooming, and spread across honeybees
- Multi-frequency seismo-acoustic measurements of synthetic ice with fluid inclusions: implications for seismic and through-ice acoustic estimation of sea-ice thickness
- On-Fault Parameters and Fault Connectivity Control Synthetic Earthquake Catalogs and 30-Year Rupture Probabilities in RSQSim Simulations of the Eastern San Francisco Bay Area
- A review on the use of Synthetic Aperture Radar (SAR) for monitoring shorelines and intertidal areas
- Unsupervised probing critical transitions in complex systems using generative diffusion models
BibTeX
@article{hart2026metaphor,
title = {{Metaphor, multimodality and generative AI: Synthetic images prompted by figurative expressions in anti-immigration discourse}},
author = {Christopher Hart},
journal = {PLOS One},
year = {2026},
month = oct,
doi = {10.1371/journal.pone.0359980},
url = {https://doi.org/10.1371/journal.pone.0359980},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Metaphor, multimodality and generative AI: Synthetic images prompted by figurative expressions in anti-immigration discourse
AU - Christopher Hart
PY - 2026
DA - 2026/10/16
JO - PLOS One
DO - 10.1371/journal.pone.0359980
UR - https://doi.org/10.1371/journal.pone.0359980
AB - Complex social processes like migration are frequently understood via metaphors. In right-wing media and far-right political discourses, two metaphors that abound draw on flood and invasion frames to problematise immigration and legitimise hostile attitudes and actions towards it. In so far as metaphor involves the construction of mental imagery, it is intimately bound with visual experience, including semiotically mediated visual experience. In a semiotic landscape increasingly populated by synthetic images, I argue that the meaning of metaphors may be modulated by AI-generated visual content and that this is especially important in contexts of political communication. The study investigates images generated by prompts instantiating an immigration is flood versus an immigration is invasion metaphor to inquire if and how they differ by source-frame and what are the ideological implications of the images returned. Following an initial pilot study, several specific semiotic features are analysed, including colour, properties of the figure (size, shape, orientation) and ground elements (environment, securitising features and frame-consistent features like fire and water). Results show significant differences between metaphorical frames and highlight the potential that AI-generated images have to reflect and reinforce ideological positions.
SN - 1932-6203
ER -