Polarization is the divergence of opinions, attitudes, or positions toward ideological extremes, within a social group or between groups. It reduces in-group common ground, and fosters growing gaps between opposing sides.[i] Such fragmentation impedes collective decision-making, and therefore coordinated responses to external threats. Since cohesive social systems can resist manipulation, adversaries seek to generate or amplify polarization.
Hybrid warfare—the integration of military (kinetic) and non-military tools—weaponizes polarization strategically to undermine societal cohesion. It exploits existing grievances regarding economic inequality, cultural divides, or political identity, amplifying them through information operations and through social media.
Social systems consist of numerous individual decision makers, organizations and governments acting non-linearly and adapting in time to each other’s actions, and generating chaotic dynamics. These complex systems’ sensitive dependence on initial conditions impedes predictability of outcomes with traditional “all-else-being-equal” approaches.
Hybrid warfare (HW) takes advantage of social systems’ complexity. A small-scale cyberattack or disinformation campaign can lead to large societal disruptions; rumors quickly amplified online can destabilize trust in institutions; targeted cyber intrusions can ripple through supply chains. Moreover, HW operates across military, cyber, economic and informational domains. These domains are tightly coupled. A disruption in one system spreads to others, leading to cascading failures: a cyberattack can cause entire power grids’ disruption, leading to economic slowdown and even public unrest. HW actors may trigger dynamics they don’t fully control. Attribution and response become much harder, because complex interactions obscure causality. Polarization undermines resistance to HW attacks.
Socio-physics uses physics tools to model social systems dynamics, overcoming analytic difficulties posed by complexity. For example, Kaufman, Kaufman & Diep (KKD)[ii] used statistical physics to model polarizing political opinion dynamics. In their model, political groups’ members are agents more or less connected in a network, with attitudes ranging from one ideological extreme to the other. They interact in time within their group and between political groups, with various intensities. The stability or volatility of attitudes is captured by a “social temperature.” High social temperature corresponds to fluid, persuadable populations with lower polarization; low temperature reflects rigid ideological commitments and strong internal group cohesion. The system of groups exhibits “phase transitions:” under certain conditions, societies can rapidly shift from moderate to highly polarized states difficult to reverse (these abrupt transitions resemble magnetization effects in physics).
KKD applied the model to several 2- and 3-group contexts: Brexit in 2016, Bosnia-Hercegovina elections in 2018, and US elections in 2015 and 2023. In the US elections, for example, they explored opinion dynamics between 3 groups: Democrats, Republicans, and Independents. As in reality, independents provided a reservoir of voters the other two camps could recruit, influencing overall polarization dynamics in time. In all these examples, the model, calibrated with survey data, correctly anticipated outcomes, some of which ran counter contemporary analyses. Besides outcome anticipation, the model can also produce scenarios under various assumptions. KKD showed that relatively small interventions fail to reduce polarization; only efforts exceeding a critical threshold can shift the system to a less polarized state. Below this threshold, interventions are effectively absorbed without impact. This challenges conventional policy approaches that rely on incremental or small-scale programs. Focusing events (e.g., a crisis) and leaders’ actions can temporarily unify adversarial groups.
In general, democracies face a structural dilemma. Their openness—free speech, pluralism, and broad information access—is a strength, but it creates vulnerabilities to HW. In the terms of the KKD model, democracies tend to have a relatively high social temperature, which supports adaptability but also allows external actors to manipulate attitudes relatively easily. When polarization deepens, societies can become simultaneously rigid within groups and volatile between them. Polarized democracies become resistant to depolarizing interventions, not least because of lack of communication among groups (homophily). They can miss threats, due to the emergence of non-intersecting sources of information. These lead to sharply divergent information and interpretations of reality, and to divided or misdirected attention. One result is reduced collective ability to recognize divisive actions, opening avenues for HW action. The same dynamics operate both inside and among countries, where HW can promote rifts among historical allies, whether inside the European Union, among the NATO countries, or between the United States and Canada.
Some policy implications follow from the KKD scenarios. Effective depolarization requires interventions at sufficient scale, coordination across sectors, and strategic timing. All these are vulnerable to, and undermined by HW. According to the model, even successful depolarizing efforts tend to be short-lived, which empirical studies have also found around the world.
Traditional defenses tend to be siloed, despite struggling against threats which don’t stay in one domain and cascade unpredictably. Social systems must become robust, by recovering quickly, and adapting in the long run. Socio-physics dynamic modeling is one approach for generating anticipatory scenarios of emerging HW threats against which response strategies can be built and tested.
—Miron Kaufman, Professor Emeritus, Department of Physics, Cleveland State University
—Sanda Kaufman, Professor Emerita, College of Urban Affairs & Education, Cleveland State University
[i] DiMaggio, J., Evans, J., & Bryson, B. (1996). Have Americans’ social attitudes become more polarized? American Journal of Sociology, 102(3), 690–755.
[ii] E.g., Kaufman, M., Kaufman, S., & Diep, H. T. (2022). Statistical Mechanics of Political Polarization. Entropy, 24(9), 1262. https://www.mdpi.com/1099-4300/24/9/1262; and, Kaufman M, Kaufman S & Diep HT 2025. Application of the Three-Group Model to the 2024 US Elections. Entropy Special Issue Computational & Statistical Physics Approaches for Complex Systems & Social Phenomena, 3rd Edition. https://www.mdpi.com/1099-4300/27/9/935