Introduction
During every crisis, the first question in the room is almost always the same: what are the other side’s intentions? It is a natural question, but it is also the one most likely to mislead us. Intentions are hidden; they can be changed overnight; and in adversarial settings they are routinely masked behind denial, deception, and cheap talk designed to feed whatever story we already prefer. Under uncertainty, analysts and decision makers reach for coherent narratives because narratives reduce anxiety, organize scattered facts, and create an illusion of control. Psychologists have described for decades how easily intuition and attribution drift into overconfidence, even among experienced professionals.[1]
We propose a doctrine that is simple to state and difficult to practice: in adversarial decision-making, begin with behavior, not mind. The Behavior-First Postulate gives evidentiary primacy to observable actions that are costly, difficult to reverse, and reveal constraints —deployments that take time and money; procurement and training cycles that close options; institutional moves that bind leaders; and durable patterns that converge across independent channels. Intent remains relevant, but only as a hypothesis that must be derived from, and continuously tested against, the opponent’s behavioral record. This ordering is not a semantic preference; it reflects a hard strategic fact. Words—including propaganda, diplomatic theater, and deliberately calibrated ‘signals’—are often cheap. In contrast, actions that consume resources and reduce freedom of maneuver are harder to fake at scale and over time. Signaling theory formalizes why costly commitments tend to carry more information than rhetoric in settings where interests diverge.[2]
What we Learn from History
Israel has lived through the consequences of mind-primacy. In 1973, the prevailing concept discounted the probability of a coordinated Egyptian Syrian attack unless specific preconditions were met, and it pushed observable preparations into the category of bluff, bargaining, or routine. The Agranat Commission later made clear that rigid assumptions—not merely lack of data—contributed to strategic surprise.[3]
A generation later, the Iraq WMD episode showed a different variant of the same pattern: a strong prior assumption about Saddam’s intentions interacted with ambiguous indicators and produced conclusions that later reviews judged to be overconfident.[4]
Behavior-first does not claim it would have removed uncertainty; it claims it would have forced a different weighting regime. The burden would have shifted toward concrete, constraint-producing evidence and away from narratives that were persuasive but difficult to falsify.
The behavior-first lens also sharpens warnings as events unfold. Russia’s posture toward Ukraine in late 2021 and early 2022 illustrates why: Moscow coupled denials and diplomacy with a steady accumulation of enabling behavior—units repositioned and stayed in place; forward stocks of fuel, ammunition, and medical support were moved closer; command-and-control nodes and logistics corridors were put on a war footing; engineering and bridging assets appeared; aircraft redeployed; and commercial imagery documented the continuing build-up even as official messages spoke of restraint.[5]
The analytic question therefore should not have been whether the Kremlin ‘wanted war’ in the abstract, but whether it had chosen to close the gap between rhetoric and feasibility. When an actor pays the costs that make large-scale action executable on short notice, capability itself becomes a strategic commitment— it creates readiness momentum, narrows political options, and compresses warning time for everyone else. In such conditions, propaganda and denial are weak evidence: they are cheap, reversible, and often designed to make sure the target continues to debate intent while preparation is completed.
The Hamas attack of October 7, 2023, is a painful reminder of how intent narratives can become anchors for mistaken analysis. Public reporting and post-event reviews describe an assessment environment in which Hamas was widely treated as deterred and governance-focused, and in which warning signs were repeatedly interpreted through that lens. Investigations and inquiry summaries reported in 2025 acknowledged significant failures, including underestimating capabilities and misjudging intentions.[6] At the same time, reporting indicates that Israeli intelligence possessed, well before October 2023, a detailed Hamas offensive operational plan that was assessed as aspirational or not actionable, and therefore did not become a reference scenario that drove collection priorities and weighting.[7]
An Analytic Approach
A behavior-first discipline treats such material as a forcing function. A plan document alone is not proof, but it is a hypothesis generator that should raise the weight assigned to compatible, costly behavior: sustained training and rehearsals, procurement and engineering effort, communications patterns consistent with operational security, and multi-channel preparations that converge over time. Where propaganda, selective leaks, and day-to-day governance signals project moderation, the doctrine demands a concrete test— are the adversary’s constraints and investments moving toward escalation or away from it? If preparations continue to accumulate while rhetoric remains calming, the analyst’s job is not to defend the story; it is to explain, in auditable terms, why the behavioral record does not raise the probability of a worst case scenario.[8]
For policymakers and intelligence leaders, the implications are immediate. A behavior-first assessment begins by stating plainly what kinds of behavior count most. The highest-weight evidence is that which ties the adversary’s hands— irreversible deployments, long-lead procurement, training cycles, binding directives, sustained resourcing, and durable posture changes that create opportunity costs. Next comes convergence: independent indicators that point in the same direction across different channels, as consistency across domains is harder to fake. Only then should low-cost signals—speeches, propaganda, selective leaks, and diplomatic theater—enter the analysis, and even then, primarily as cues to test against behavior. This ordering is not about cynicism; it is about protecting decision makers from being seduced by plausible stories that adversaries want us to tell ourselves.
This doctrine also changes how assessments are written and how they should be challenged. Instead of asking analysts to explain intent, decision makers should demand an auditable behavioral chain: which observable commitments drive the judgment, what constraints they reveal, and what would have to be observed for the judgment to change. The last element is decisive. A professional assessment should be able to say, plainly, what would disconfirm it. Without that clause, analysis slides into belief. With it, forecasting becomes a discipline: teams are forced to update when behavior diverges from the prior story, and leaders can see whether a judgment is anchored in evidence or in preference. Scoring forecasts and measuring calibration are not a luxury; they are how organizations learn whether their tradecraft reduces surprise.[9]
Conclusion
Objections should be welcomed, not dismissed: behavior can be deceptive. Feints, exercises, diversionary moves, and staged signals are as old as war itself. The answer is not to retreat to mind-reading; it is to refine what we mean by behavior. The doctrine privileges sustained costly, multi-channel behavior precisely because it is harder to fabricate over time without revealing inconsistencies or incurring real opportunity costs. When deception is plausible, analysts should look for the enabling details that make action feasible, for persistence rather than headlines, and for the places where the adversary has paid a cost that is difficult to recover. Intent inferences will never disappear, and nor should they. But they should be treated as the end of the analytic process, not the beginning. In a region where surprise is paid for in lives and strategic setbacks, the safer starting point is the pattern of actions the adversary is already willing to undertake.
[1] On bias, narrative attraction, and overconfidence in intelligence judgment see Richards J. Heuer Jr., Psychology of Intelligence Analysis (Center for the Study of Intelligence, CIA, 1999); Amos Tversky and Daniel Kahneman, “Judgment under Uncertainty: Heuristics and Biases,” Science 185, no. 4157 (1974): 1124-1131; Lee Ross, “The Intuitive Psychologist and His Shortcomings: Distortions in the Attribution Process,” in Advances in Experimental Social Psychology, Vol. 10 (Academic Press, 1977), 173-220; and Gerd Gigerenzer, Gut Feelings: The Intelligence of the Unconscious (Viking, 2008).
[2] On costly signals and why commitments can be more informative than rhetoric, see Michael Spence, “Job Market Signaling,” The Quarterly Journal of Economics 87, no. 3 (1973): 355-374; James D. Fearon, “Signaling Foreign Policy Interests: Tying Hands versus Sinking Costs,” Journal of Conflict Resolution 41, no. 1 (1997): 68-90; and Vincent P. Crawford and Joel Sobel, “Strategic Information Transmission,” Econometrica 50, no. 6 (1982): 1431-1451.
[3] Report of the Commission of Inquiry into the Events of the Yom Kippur War (Agranat Commission), 1974.
[4] On the Iraq WMD analytic failures and post-event reviews, see U.S. Senate Select Committee on Intelligence, Report on the U.S. Intelligence Community’s Prewar Intelligence Assessments on Iraq (2004); and Commission on the Intelligence Capabilities of the United States Regarding Weapons of Mass Destruction, Report to the President of the United States (2005).
[5] On the pre-invasion Russian build-up and the visibility of enabling logistics and medical preparations, see Reuters, “Exclusive: Russia moves blood supplies near Ukraine, adding U.S. concern,” Jan. 28, 2022; Reuters, “Satellite images show new Russian military deployments near Ukraine,” Feb. 11, 2022; and Reuters, “U.S. says Russia inching troops closer to Ukraine, stocking up blood supplies,” Feb. 17, 2022.
[6] On post-event acknowledgments of misjudged intent and warning failures regarding October 7, see Reuters reporting on the Israeli military inquiry summary (February 27, 2025) and Reuters reporting on the Shin Bet self-investigation (March 4, 2025).
[7] On reporting that Israeli intelligence possessed a detailed Hamas operational concept before October 2023 and the debate around how it was assessed, see Associated Press (December 1, 2023) and related press reporting cited therein.
[8] Additional analytic perspectives on the October 7 intelligence failings that emphasize weighting, indicators, and deception include Mark Wyss, “The October 7 Attack: An Assessment of the Intelligence Failings,” CTC Sentinel 17, no. 9 (2024); and Institute for Defense Analyses (IDA), Israeli Intelligence Failures Prior to Hamas’s October 7 Attack (2024).
[9] On forecast scoring, calibration, and institutional learning in adversarial prediction, see Philip E. Tetlock and Dan Gardner, Superforecasting: The Art and Science of Prediction (Crown, 2015).
JISS Policy Papers are published through the generosity of the Greg Rosshandler Family.
