What is an Armed Conflict?

Saturday 21 February 2026

War is easy to recognise and surprisingly difficult to define.

A missile strike on a capital city — a rebel column entering a provincial town — a militia massacre in a remote district — these events are plainly violence. But when does violence become an armed conflict? When does an armed conflict become a civil war? And why do scholars disagree?

Three of the most influential frameworks — the Heidelberg Institute for International Conflict Research (HIIK), the Uppsala Conflict Data Program (UCDP), and ACLED — answer that question differently. The differences are not academic trivialities. They shape headlines, humanitarian funding, diplomatic urgency and the historical record.

The first fault line is numerical. UCDP traditionally defines an armed conflict as a “contested incompatibility” involving organised armed actors resulting in at least 25 battle-related deaths in a calendar year. For classification as a war, the threshold rises to 1,000 battle-related deaths. It is precise, quantitative and relatively conservative. Violence below the death threshold does not count — even if it terrorises a population.

HIIK adopts a broader, qualitative scale. Rather than relying solely on death counts, it classifies conflicts along five intensity levels, from “dispute” to “war”. A violent crisis (Level 3) may involve sustained armed force even if the annual fatality count does not cross 1,000. HIIK’s approach recognises that organised armed violence can destabilise a state long before the morgues are overwhelmed. It is more sensitive to low-intensity but persistent warfare — the type that afflicts much of the Sahel.

ACLED, by contrast, is event-driven. It records individual political violence incidents — protests, battles, explosions, riots — in granular detail. It does not impose a rigid threshold before acknowledging a conflict. Instead, analysts derive patterns from accumulated events. ACLED therefore often signals deterioration earlier than threshold-based systems. But its breadth can blur the line between insurgency, communal violence and organised crime.

The second fault line is conceptual. Must an armed conflict be political? UCDP insists upon a “contested incompatibility” over government or territory. HIIK similarly emphasises incompatibility concerning values of national relevance. ACLED includes organised criminal violence where it interacts with political authority. This is why Mexico’s cartel wars may appear prominently in ACLED and HIIK analyses but remain more ambiguous in stricter civil war taxonomies.

The third fault line is structural. Does the violence have to involve the state? Intrastate conflict classifications often require state participation. Yet many contemporary wars — especially in Africa — involve hybrid configurations: militias fighting each other, state forces fighting one militia while covertly backing another, foreign private military companies reinforcing fragile regimes. The neat binary of state versus rebel rarely survives contact with reality.

Why does this matter?

Because labels carry consequences. If a situation is called a “civil war”, insurance contracts shift, refugee law engages particular protections, humanitarian funding appeals intensify, and diplomatic mediation frameworks activate. If it is coded merely as “criminal violence”, it may receive less international urgency — even if its lethality is comparable.

Consider Sudan. Under any framework, the conflict between the Sudanese Armed Forces and the Rapid Support Forces qualifies as war. But what of Nigeria’s Middle Belt violence? UCDP may categorise specific dyads; HIIK may code multiple parallel violent crises; ACLED will map a dense web of communal and insurgent events. Each portrayal shapes how outsiders perceive the state’s stability.

The modern battlefield is increasingly fragmented — decentralised militias, armed gangs exercising quasi-sovereign authority, insurgents operating beneath conventional war thresholds, states deploying proxy actors to preserve deniability. In such an environment, rigid fatality thresholds may understate instability, while overly inclusive event recording may overstate systemic collapse.

No definition is neutral. Every threshold embodies an implicit theory of what matters: death tolls, political incompatibility, territorial control or state involvement.

The uncomfortable truth is that “armed conflict” is less a fixed category than a spectrum. At one end lies political dispute. At the other, industrialised war. In between exists the contemporary world’s prevailing condition — chronic, fragmented, ambiguous violence.

And ambiguity is not accidental. Many regimes prefer violence that remains just below international alarm thresholds. Many armed groups calibrate attacks to avoid triggering overwhelming retaliation. Warfare today is often conducted in the grey zone — intense enough to dominate a population’s daily life, restrained enough to evade formal classification as war.

Thus the question “What is an armed conflict?” is not merely definitional. It is political. It determines who counts as a victim of war, which crises command global attention, and which remain peripheral footnotes.

If the twentieth century was defined by declared wars between states, the twenty-first is defined by contested sovereignties within them. Our statistical tools struggle to keep pace with that transformation.

And so we return to the beginning. War is easy to recognise — but difficult to measure. The answer depends not only on how many die, but on how we choose to count them.