Frischlich, L., Olsson, H., Roy, A., Schulze, H., Rhodes, S., & Mansheim, A. (2025). The complexity of misinformation extends beyond virus and warfare analogies. npj Complexity, 2, 1–8. https://doi.org/10.1038/s44260-025-00053-z
Summary
This perspective paper argues that the dominant analogies used to describe misinformation—“infodemic” (virus), “information warfare” (weapon), and “information pollution” (toxin)—each capture only a slice of the phenomenon while systematically obscuring the rest. Drawing on complexity science and belief-dynamics modeling, the authors propose a four-level framework (micro cognition, meso-I social groups, meso-II digital platforms, macro societal systems) for treating misinformation as a complex adaptive system. They map each analogy onto this framework to expose its “conceptual mileage” and “conceptual baggage,” concluding that single-analogy thinking produces incomplete and sometimes counterproductive policy. Instead, they call for portfolio-style, complexity-aware interventions and evaluation methods modeled on climate change research.
Key Contributions
- Introduces a four-level (micro / meso-I social groups / meso-II platforms / macro systems) framework tailored to misinformation, extending integrative belief-dynamics models.
- Provides a structured codebook and methodology for evaluating analogies against complexity criteria.
- Offers a comparative diagnostic making the blind spots of the three dominant analogies explicit.
- Translates the framework into concrete policy guidance favoring multilevel intervention portfolios and iterative evaluation.
- Sets a future research agenda: meta-dynamics, cross-platform and cross-regime comparison, longitudinal feedback effects, and qualitative work to counter ecological fallacy.
Methods
The work is a conceptual/theoretical synthesis adapting integrative belief-dynamics models to misinformation. The authors constructed a four-level framework with prototypical actors and ties, then ran a systematic mapping exercise: a co-developed codebook was applied by a multidisciplinary team to evaluate the three analogies (plus “fake news”) against the framework, with iterative consensus revision. This was supplemented by bibliometric scoping (Web of Science and Google Scholar hit counts) and a Google Trends analysis of normalized search interest for each analogy from 2005 to May 2025.
Findings
- Bibliometric scoping (May 2025) found 1,275 Web of Science results for “infodemic,” 1,258 for “information war*,” and 103 for “information pollution” (Google Scholar: ~41,900, ~69,300, and ~6,280 respectively).
- The virus/infodemic analogy captures micro-level transmission and supports inoculation approaches but ignores social embedding, platforms, macro political systems, and strategic motives.
- The information-warfare analogy highlights macro-level strategic state action but treats society as monolithic and downplays individual agency, non-political motives, and feedback loops.
- The information-pollution analogy best represents systemic dynamics and tipping points but underplays recipient agency, intergroup dynamics, and platform affordances.
- No single analogy captures the recursive coupling of psychological, social, technical, and political factors—and interventions built on one analogy can backfire (e.g., crude media literacy training raising generalized distrust).
Connections
This paper’s advocacy for inoculation as one (partial) micro-level tool connects it to debates on prebunking and psychological resistance in van-der-Linden2026-jt and to critiques of intervention efficacy such as Spampatti2026-kx. Its complexity-informed reframing of the “overstated vs. underappreciated threat” dispute speaks to broader assessments of misinformation prevalence and impact in Budak2024-ef and DeVerna2025-dl, while its emphasis on platform-level (meso-II) dynamics relates to structural analyses of information ecosystems like Gonzalez-Bailon2024-rq.