By now, the cadence is familiar: deny the existence of a “grand plan,” recast incendiary rhetoric as metaphor, and insist that any violence was directed not at civilians but at criminals — a distinct “genus,” as defense counsel put it.

In his recent defense speech, Atty. Kaufmann advanced a theory built on a single premise: that the public speeches of former President Rodrigo Roa Duterte are insufficient to prove the existence of a coordinated plan to kill civilians, whether during his tenure as mayor of Davao City or as president of the republic. The former president’s repeated threats to “kill” drug suspects, the defense argues, were hyperbolic tools of deterrence, designed to instill fear among criminals, not evidence of a state policy to murder.

The problem with that argument is that international criminal law does not treat words as weightless entities.

At the International Criminal Court, speeches by powerful officials are not evaluated in isolation. They are assessed in context — against patterns of violence, chains of command, and institutional behavior. The Rome Statute does not require a written master plan or a signed death warrant to establish crimes against humanity. It requires proof of a “widespread or systematic attack directed against any civilian population,” carried out pursuant to a state or organizational policy, with knowledge of that attack.

Three elements matter here: attack, policy, and knowledge.

First, the “attack.” Under Article 7 of the Rome Statute, an attack is not limited to armed conflict; it can be a course of conduct involving multiple acts of violence against civilians. The Pre-Trial Chamber in the Kenya situation clarified that “attack” refers to a campaign or operation carried out against a civilian population, not a single isolated act. When thousands die in police operations or vigilante-style killings under an announced national campaign, the legal question is not whether each victim was innocent, but whether the violence formed part of a broader pattern targeting a civilian population.

The defense’s attempt to carve out a “specific genus” of civilians — drug suspects — does not necessarily resolve the issue. International jurisprudence has repeatedly held that suspected criminals remain civilians unless they directly participate in hostilities. The Rome Statute’s protections do not evaporate because a head of state labels a category of people as “criminals.”

Second, the “policy.” Kaufmann’s core assertion is that fiery speeches alone are insufficient to prove a grand plan. On its face, that is correct: rhetoric alone does not establish policy. Yet ICC jurisprudence has never required a formal, written blueprint. In Prosecutor v. Bemba, the Court held that policy may be inferred from repeated acts, the mobilization of resources, coordination among state actors, and the toleration — even encouragement — of violence.

Similarly, in Prosecutor v. Katanga, the Court emphasized that policy can be inferred from the manner in which acts occur, their repetition, and the involvement of authorities. The Appeals Chamber has noted that the policy element exists to distinguish random crimes from organized campaigns — not to impose an unrealistic evidentiary burden.

In that light, public speeches matter not as isolated sound bites but as potential indicators of intent and knowledge. When a chief executive repeatedly promises immunity, praises lethal force, and frames killings as necessary purification, those statements can serve as contextual evidence that violence was not rogue behavior but aligned with leadership messaging.

Third, knowledge and intent. Under Article 30 of the Rome Statute, criminal responsibility attaches when a person acts with intent and knowledge. Leaders need not pull the trigger. Articles 25(3)(b) and 25(3)(d) establish liability for ordering, soliciting, inducing, or contributing to crimes committed by a group acting with a common purpose.

Article 28 further establishes command responsibility when superiors knew or should have known that subordinates were committing crimes and failed to prevent or repress them.

The defense’s deterrence argument — that threats were intended to frighten criminals — inadvertently underscores knowledge. If the messaging was designed to instill fear in a defined class of people, and if killings predictably followed such messaging, the question becomes whether those outcomes were foreseeable and tolerated.

The ICC has rejected the notion that leaders may hide behind rhetorical ambiguity. In Prosecutor v. Gbagbo and Blé Goudé, inflammatory speeches were examined as part of the contextual matrix for assessing whether violence against civilians formed part of a common plan. The Court scrutinized not only the words themselves but also their effects and the structures that translated them into action.

Moreover, international law does not permit a government to redefine “civilian” to suit its objectives. The jurisprudence of international tribunals, including the ICTY, has consistently held that even individuals suspected of crimes remain part of the civilian population unless actively engaged in hostilities. Under international humanitarian law, a domestic “war on drugs” does not transform urban neighborhoods into battlefields.

There is also a revealing undertone in Kaufmann’s address: an apparent acceptance that the Court may authorize a full trial. Procedurally, this suggests a pivot from an outright jurisdictional denial toward a merits-based defense. Under Article 61 of the Rome Statute, the Pre-Trial Chamber must determine whether there are “substantial grounds to believe” the accused committed the crimes charged — a lower threshold than proof beyond a reasonable doubt. If the defense is already shaping its narrative around the insufficiency of proof, it tacitly acknowledges that the evidentiary bar at this stage is not insurmountable.

That is significant. The ICC does not require the prosecution to prove the entire case at the confirmation stage. It must show that evidence — including patterns of killings, official statements, operational directives, and command structures — forms a coherent picture consistent with crimes against humanity. If that threshold is met, the matter proceeds to trial, where a higher standard applies.

Kaufmann’s argument rests on narrowing the frame: isolate the speeches, detach them from operational reality, and insist on a written “grand plan” that likely never existed on paper. But international criminal law was designed precisely for situations in which mass violence is carried out without formal decrees. Modern atrocities are rarely accompanied by signed manifestos. They are executed through atmospherics — signals, incentives, and tolerated excesses.

The ICC’s founders understood this. The Rome Statute’s architecture reflects the lessons of the 20th century: that leaders can cultivate climates of violence without issuing explicit kill lists; that policy may be informal yet systematic; and that fear can be weaponized as effectively as orders.

None of this establishes guilt. The former president is entitled to the presumption of innocence, and the prosecution bears the burden of proof. But to argue that repeated public exhortations to kill, coupled with a sustained campaign that resulted in thousands of deaths, are legally irrelevant unless accompanied by a signed master plan is to misread the evolution of international criminal jurisprudence.

Words are not stray sparks in the wind when spoken by a head of state. They are signals — amplified through institutions, interpreted by subordinates, and often translated into action.

The defense would have the Court believe that those signals were merely rhetorical thunder, intended to frighten wrongdoers into submission. The Rome Statute asks a more difficult question: when thunder is followed by lightning, and the lightning strikes the same neighborhoods again and again, can the storm truly be called accidental?

That question — not the flourish of any single speech — is what the judges at The Hague must decide.