Intellectuals see a world rotating around ideas. They love the idea, for example, that America is an idea, and that Israel went to war with an idea on October 7.
I don’t see things this way. I think America is primarily a nation and that the Israel nation went to war with parts of the Gazan nation on October 7.
Per David Pinsof’s Alliance Theory, ideas are coalition technologies.
When Haviv Rettig Gur says Israel went to war with an idea on October 7, he is pointing at something real. Hamas is not merely a military organization. It is structured around a moral vocabulary, armed resistance, religious-national framing, rejection of Israel’s legitimacy, that shapes tactics, recruitment, and endurance in ways that purely material analysis misses. But that vocabulary does not float free. It is embedded in a tribe, a coalition, a set of institutions, and material interests. The tension between these two observations is not a contradiction. It is the entry point into the deepest analysis.
David Pinsof, David Sears, and Martie Haselton provide the precise mechanism in their Alliance Theory paper. When coalitions support their allies, they apply perpetrator biases, minimizing their allies’ transgressions, downplaying responsibility, emphasizing mitigating circumstances, and attributing bad outcomes to external forces rather than to the ally’s choices. They apply victim biases, amplifying their allies’ grievances, attributing opponents’ motives to irrational malevolence, and embellishing the severity and duration of harm. They apply attributional biases, crediting allies’ successes to internal virtues and allies’ failures to circumstances beyond their control. These are not cynical performances. They are genuine psychological mechanisms that produce the moral vocabulary of any conflict, and they operate symmetrically across coalitions. The Israeli coalition applies perpetrator biases to Israeli military actions and victim biases to Israeli casualties. The Palestinian coalition does the reverse. Neither side is applying abstract principles about justice, proportionality, or resistance. Both are running the same psychological machinery on different inputs, generating different moral vocabularies that feel, to the people running them, like straightforward descriptions of reality.
This is what Rettig Gur misses and what the Alliance Theory paper clarifies. When he says Israel went to war with an idea, he is pointing at the output of this machinery. The moral vocabulary of martyrdom, resistance, and national liberation is what perpetrator and victim biases generate when applied over decades to a coalition under existential pressure. The ideas are real. They are not epiphenomenal decoration on top of a tribal reality that is somehow more true underneath. Pinsof and colleagues are careful about this: the contents of belief systems are determined by alliance structures and propagandistic biases, but those biases are genuine psychological mechanisms, not conscious deceptions that participants could abandon if they chose to be honest. The person who applies victim bias to their ally’s casualties is not lying. They are doing what the human alliance psychology produces when activated by coalition membership. The self-deception layer that Trivers identifies explains why this is stable: the person who genuinely believes their side’s casualties are more important is a more effective advocate than someone who privately suspects the symmetry.
Ideas do not act. Groups do. What ideas do is coordinate groups under pressure, justify sacrifice, signal loyalty, define enemies, and sustain cooperation when material incentives alone would produce defection. The martyrdom ideology Hamas deploys is not a philosophical choice made by individuals reasoning from first principles. It is what the propagandistic bias machinery produces when applied to a coalition under conditions of extreme asymmetry, poverty, blockade, and repeated military defeat. It tracks social debts within the group, promises heroic status that outlasts the body, and enforces internal discipline by making defection from the resistance narrative equivalent to moral betrayal of the community’s dead. Ernest Becker supplies the immortality motive. Robert Trivers supplies the accounting mechanism. Pinsof, Sears, and Haselton supply the specific bias machinery that generates the moral vocabulary. The idea turns a tribal calculator into a soul, and it makes the tradeoff feel necessary rather than strategic, which is precisely what the coalition needs its members to believe in order to function.
Intellectuals prefer the language of ideas for reasons that are themselves explainable within this framework. Intellectuals belong to a professional caste whose primary tradeable currency is interpretation. If a conflict is a tribal struggle over land, resources, and security, the intellectual is a secondary actor, a commentator on events driven by forces that do not require his expertise. If the conflict is a war between ideas, the intellectual becomes the high priest. He defines what the idea is, explains what it means, narrates its victories and defeats, and declares when it has been conquered. Saying Israel went to war with an idea, or that America is an idea, or that Nazism was crushed in 1945, converts the intellectual from a bystander into an essential participant. The interpretive class uses ideas as coalition technologies to claim jurisdictional authority over the meaning of violence. This is not cynical. It is what the propagandistic bias machinery produces when applied to an intellectual’s own coalition membership and professional position. Intellectuals cannot see the tribal machinery generating their moral vocabulary any more than anyone else can, because the self-deception mechanism makes the output feel like principled analysis rather than coalition maintenance.
The paper’s symmetry finding matters enormously here. Pinsof et al. document that liberals and conservatives apply identical propagandistic biases toward their respective allies, finding no evidence that either side is uniquely principled or uniquely hypocritical. The same applies internationally. The American intellectual who frames the Hormuz crisis as a violation of the rules-based order is not more principled than the Iranian analyst who frames it as legitimate asymmetric retaliation. Both are applying perpetrator biases to their allies’ actions, victim biases to their allies’ grievances, and attributional biases to their allies’ outcomes. The American says Iranian forces violated freedom of navigation. The Iranian says American sanctions violated Iranian sovereignty. Each is minimizing their coalition’s transgressions and amplifying their coalition’s grievances, and each experiences this as accurate description rather than advocacy.
The biological framework clarifies what happens to ideas inside closed tribal loops. The martyrdom ideology of Gaza has been cloned in a high-pressure, isolated environment for decades. It has suffered the Müller’s ratchet of inbred ideas: accumulating distortions, extremist mutations, and epistemic closure because it has lacked contact with rival moral frameworks that would force adaptation and recombination. Intellectuals who call it an idea miss the inbreeding depression of the logic. They treat what is partly a mutated propagandistic bias system as a philosophical position subject to rational revision. The comparison to the Jerusalem and Babylonian Talmuds is instructive: the Jerusalem version stayed in its origin environment and became brittle; the Babylonian version survived contact with Persian legal culture and gained the hybrid vigor that made it the dominant text. A coalition’s moral vocabulary that cannot encounter rival coalition’s moral vocabularies without triggering immune rejection accumulates the equivalent of deleterious mutations, not metaphorically but structurally, in the sense that its internal logic progressively detaches from the conditions it must navigate to survive.
The crypsis layer completes the picture. In elite Western institutions, the idea is the camouflage. A biosocial realist, a tribal strategist, or simply a person with interests that the dominant coalition prefers not to name must adopt the coloration of the environment to survive. DEI statements, land acknowledgments, rules-based-order affirmations: these are surface proteins that allow the organism to pass through detection systems calibrated to flag visible coalition aggression while remaining blind to the same aggression when it presents in the approved vocabulary. The summons reproduces this crypsis at the individual level. The intellectual is called into being as a defender of the idea, hailed repeatedly by institutions that reward interpretation and punish naked interest-acknowledgment, until he genuinely believes that what he is defending is a universal principle rather than a coalition’s preferred classification system. The self-deception is not incidental. A person who knows he is performing coalition maintenance cannot perform it convincingly. A person who believes he is defending civilization can.
The phrase America is an idea illustrates the full machinery. The package contains constitutionalism, individual rights, equality before the law, democratic self-government, and the promise of mobility. Its key feature is portability: in principle, anyone can subscribe, which means the nation’s expansion can be framed as universalism rather than empire. Political elites across the spectrum benefit: the right claims to defend founding principles, the left claims to realize them, and both factions invoke the same idea to authorize competing policy programs. The professional-managerial class gains interpretive jurisdiction. If America is an idea, then lawyers define its boundaries, academics interpret its meaning, journalists narrate its successes and failures, and policy experts specify its implementation. The attributional bias machinery operates throughout: American successes are credited to the idea’s internal virtue, American failures are attributed to insufficient adherence to the idea, and competing nations are described as opponents of the idea rather than as actors with their own coalition interests. What the framing hides is what the framework predicts it would hide: the coalition power through which the idea is reproduced and policed.
The rules-based international order performs the same function at global scale. The package combines post-1945 institutions, dollar-denominated trade architecture, alliance structures, and legal vocabulary of sovereignty and human rights. Its defining feature is not what the rules are but who writes, interprets, and enforces them. Perpetrator biases operate continuously: when allied forces cause civilian casualties, the framing emphasizes mitigating circumstances, proportionality assessments, and procedural compliance. When adversary forces cause civilian casualties, the framing emphasizes moral culpability, intentionality, and systemic patterns. Victim biases operate with equal consistency: the casualties of allied nations receive full moral weight while the casualties produced by allied operations receive humanitarian framing that partially redistributes responsibility. These are not double standards in the sense of conscious hypocrisy. They are the perpetrator and victim bias machinery running as designed on the inputs provided by coalition membership.
Every major power runs this same machinery with different content. China’s package, common future for mankind, win-win cooperation, non-interference, sovereignty and development first, generates victim biases toward Chinese claims of historical grievance and perpetrator biases toward foreign criticism, attributing China’s development successes to internal virtue and its challenges to external interference. Russia’s package, multipolar world, civilizational sovereignty, traditional values, generates victim biases toward Russian security concerns and perpetrator biases toward NATO expansion, attributing Russian actions to defensive necessity and adversary actions to aggressive malevolence. Iran’s package, resistance, anti-imperialism, axis of resistance, generates victim biases toward Iranian populations under sanctions and perpetrator biases toward Iranian proxy operations, framing sacrifice as sacred obligation and retaliation as justice. Each system packages interests as principles, defines who is legitimate, coordinates allies, and disciplines defectors. Each generates its specific moral vocabulary through the same propagandistic bias machinery that Pinsof et al. document in American partisan politics. The mechanism is universal. The content is coalition-specific.
The Iran war makes this machinery visible with unusual clarity because all sides are describing the same physical events in completely different moral vocabularies. The Strait of Hormuz closure is simultaneously a rule violation, a legitimate retaliation, a destabilizing disruption, a symptom of system breakdown, and a pricing event, depending entirely on which coalition’s propagandistic bias machinery you inhabit. The attribution is the key move. The American coalition attributes the closure to Iranian aggression, emphasizing Iran’s internal disposition toward violence and expansionism. The Iranian coalition attributes it to American provocation, emphasizing sanctions and military threats as the external circumstances that necessitated the response. Neither is fabricating. Both are running the attributional bias mechanism on the same event and producing different causal stories that feel like accurate descriptions from inside each coalition.
The closing of the strait is also a competitive victimhood event in the paper’s precise sense. Pinsof et al. document that victim biases on both sides of a conflict lead coalitions to strive to establish that their coalition was subjected to more injustice than the other. The American coalition emphasizes the victims of Iranian regional aggression, Yemeni casualties, Lebanese instability, and now global energy disruption. The Iranian coalition emphasizes Iranian victims of sanctions, assassinations, and military encirclement. Neither side denies the other’s casualties entirely. Each insists its own are more severe, more systematic, and more morally significant, which is precisely the competitive victimhood pattern the paper describes.
Rettig Gur is not wrong that ideas shape behavior. He is describing the output of the propagandistic bias machinery accurately: the martyrdom ideology really does constrain what Hamas can offer its fighters, really does shape recruitment and endurance, really does make the coalition harder to dissolve than a purely instrumental organization would be. What he misses is the machinery that generates that output, and what the intellectual preference for idea language does is systematically direct attention toward the output and away from the mechanism. Across every case from Hamas to NATO to BlackRock’s stewardship vocabulary to the CCP’s serving the people language, the pattern is identical: a coalition’s propagandistic biases generate a moral vocabulary, that vocabulary is experienced by its users as principled description, and intellectuals who are themselves inside coalitions provide the interpretive labor that makes the output look like philosophy rather than alliance maintenance.
The synthesis Alliance Theory enables is not that ideas are less real than tribes but that ideas are what tribes generate when they run their propagandistic bias machinery on their alliance structures. The ideas are genuine. The biases that produce them are genuine. The coalitions that shape the bias inputs are genuine. None of these layers is more or less real than the others. The intellectual who says this conflict is about ideas and the analyst who says it is about tribal interests are both pointing at real features of the same system. The paper points at the machinery that connects them.
I love Alliance Theory in particular and David Pinsof’s work in general because they explain and predict without partisan bias. This is not just a personal virtue. It is a structural achievement that requires specific intellectual choices, and understanding what produces it clarifies why it is so rare.
The paper’s primary mechanism for avoiding partisan bias is the symmetry commitment. Pinsof et al. do not merely assert that both sides use propagandistic biases. They construct the entire theoretical apparatus around the prediction that the biases operate identically across coalitions, and then test that prediction empirically. The paper is not balanced in the journalistic sense of finding something critical to say about each side. It is symmetric in the deeper sense of applying the same causal model to both sides and letting the evidence determine whether the model fits. When the data show that liberals and conservatives are equally intolerant of their respective rivals, equally likely to apply perpetrator biases to their allies, and equally likely to use attributional biases in coalition-serving directions, that finding is not a diplomatic gesture toward balance. It is a theoretical prediction that could have been falsified and was not. That is a different thing from performed neutrality.
The explanatory mechanism also resists partisan capture in a specific way. By locating the explanation in the universal structure of alliance psychology rather than in the particular content of any coalition’s beliefs, the paper makes it structurally difficult to weaponize against one side. If the argument were that conservatives are uniquely hypocritical, or that liberals are uniquely principled, it would function as coalition technology for the opposing side. Because the argument is that everyone runs the same machinery on different inputs, it cannot be appropriated by any coalition without that coalition simultaneously indicting itself. That is a rare property in political psychology, where most findings, whatever the author’s intentions, tend to be adopted as ammunition by one side.
The distinction the paper draws between psychological equivalence and moral equivalence is also worth noting. Pinsof et al. explicitly reject the inference that because both sides apply identical mechanisms, both sides are equally right or equally harmful. Different moral consequences can stem from the same underlying psychology, and categorizing more vulnerable groups as enemies might produce more harm than categorizing less vulnerable groups as enemies. That distinction allows the paper to be scientifically symmetric without being morally relativistic, which is a difficult line to hold and the paper holds it.
What makes this rare in practice is that most researchers in political psychology have themselves been summoned into coalitions, in Tavory’s sense, by the institutional environments that trained and rewarded them. The academic social science environment has strong directional incentives, and the propagandistic bias machinery operates on researchers just as it operates on everyone else. Producing work that genuinely resists those incentives requires either unusual self-awareness about one’s own coalition membership, or a theoretical framework powerful enough that its implications constrain the analysis even when those implications are uncomfortable, or both. The Alliance Theory paper achieves the second condition clearly: the framework predicts symmetry, the data confirm symmetry, and the conclusion follows regardless of what the authors might prefer to find.
That is why the predictive and explanatory power and the partisan balance are not separate virtues. They are the same virtue. A framework powerful enough to generate precise, falsifiable predictions about both sides of a political divide cannot easily be bent to serve one side without breaking the framework. The rarity of Pinsof’s work in this regard reflects how seldom researchers build frameworks with that property, and how much easier it is to build frameworks that generate predictions about one side while treating the other as the implicit baseline of rationality.
