A map can tell you where a conflict is happening today, but it is blind to where that war has been – and where it is stubbornly determined to go. Analysts and planners assess geopolitical fault lines, patterns of violence, human factors, and threat networks to understand the dynamics of irregular warfare inherent in crisis, conflict, cooperation, and competition. However, these environments remain ambiguous, complicated, and difficult to predict.
In an attempt to understand local conflict dynamics, policymakers often use platforms like the United Nations Development Programme’s Crisis Risk Dashboard or Armed Conflict Location and Event Data’s Conflict Index, where incident maps are the foundation of efforts to inform crisis response decision-making. That information matters, but it is incomplete. Two localities showing similar levels of violence today could be moving toward divergent futures. Without understanding the conflict’s potential trajectory, policymakers can overreact in some contexts and underreact in others.
This is the core insight of new research that uses a trajectories approach to irregular conflict: local violence is not just a sequence of incidents. It is a sequence of place-specific conditions. In plain language, conflict has memory. Analysts and planners should remember this dynamic.
A Tale of Three Violent Cities
Politically motivated violence in Africa is unevenly distributed across space and time. It tends to cluster in some regions, such as the Great Lakes or the Horn, leaving large areas unaffected by conflict for decades. In North and West Africa, more than 90% of violent events occurred in just Nigeria—by far the main epicenter of violence—as well as Burkina Faso, Cameroon, Mali, and Niger. Armed conflict also tends to affect localities differently, as they pass through different stages during their existence.
A conflict life cycle is essentially a series of sequential years with violence interrupted by at least one year of calm. These cycles can be identified by monitoring the interplay between where violent incidents happen along with the number of conflict actors and rivalrous relationships among them as conflict develops over time.
Consider three West African cities that are embedded within the same overall regional dynamic, but show three entirely different trajectories. Maiduguri in northern Nigeria experienced twenty consecutive years of intense and concentrated violence, a long uninterrupted cycle that began in 2006 with no clear endpoint. Over the same period, Agadez in central Niger cycled through six distinct episodes of fighting, each lasting between one to four years with periods of relative calm in between. Bobo Dioulasso in southeastern Burkina Faso endures sporadic, brief flare-ups that dissipate quickly every year.
Figure 1. Same regional dynamics, different trajectories. Three cities affected by West African insurgencies experienced fundamentally different conflict pathways. Source: Authors derived from ACLED data; adapted from Walther, Radil and Russell (2025).
What Trajectories Can Reveal
Most conflict analyses are snapshots in time and space. They describe what happened this week or this quarter, then rank places or districts by risk. Snapshot tools are necessary for situational awareness, but they flatten time and atomize places. They obscure whether a locality is improving, relapsing, or locking into a persistent violent system.
A trajectories lens restores time to the center of the analysis of a place. Instead of asking only how much violence exists now, it asks how local conflict behavior evolves from one period to the next. Does violence remain geographically concentrated or spread across wider areas? Does it sustain high frequency or drop to sporadic incidents? Does a locality return to calm after an episode of violence, or does it cycle through patterns of recurrent instability?
Research tracking contemporary conflict patterns across Africa since the late 1990s reveals a striking empirical regularity: most local conflicts are brief and ephemeral, some are recurrent, and a smaller but highly consequential set becomes entrenched. In our analysis of over 3,700 conflict-affected localities in Africa, 77 percent experienced cycles of violence lasting just one year. These ephemeral outbreaks typically resolve within an average of fourteen months once violence begins.
The remaining localities tell a different story. Recurrent zones sequence through repeated episodes averaging three to four years each with intervening periods of calm. Entrenched zones—places like Maiduguri—settle into patterns of concentrated, high-intensity violence with mean durations of six to eight years. Once a locality enters this entrenched state, the probability of remaining there year-to-year exceeds 70 percent. Quick de-escalation becomes structurally unlikely. The practical importance of that distinction is difficult to overstate.
What ‘Memory’ Means in Practice
Calling conflict ‘path dependent’ sounds abstract. In concrete terms, it simply means that yesterday’s situation constrains today’s options. When local violence persists, armed actors adapt. Civilian movement patterns shift around threat corridors. Public authorities lose influence in practical terms while informal power brokers gain influence by mediating fear and access. Economic life turns to smuggling and crime.
These adaptations often make short-term stabilization harder over time. In the West African Sahel, armed groups affiliated with Al Qaeda or the Islamic State have implemented an alternative governance and service provision model that has supplanted the state for many years.
At that point, conflict is no longer just a contest among armed organizations. It becomes embedded within local systems of security, mobility, and survival. This is why an entrenched trajectory can endure even when insurgent leaders are captured or killed or if counterinsurgency operations succeed tactically. The underlying local system has already adapted to continue to reproduce violence. The phrase ‘conflict has memory’ captures this cumulative process. It reminds us that each period is not a reset but an inheritance.
The Cost of Category Errors
Many policy failures in irregular warfare can be read as category errors. A recurrent conflict is treated as a one-off disturbance. An entrenched conflict is treated as if a single short-term operational cycle can produce resolution. In northern Mali, for example, the recapture of Kidal in 2023 was presented as a major victory for the government and its Russian mercenaries, even though the historical root causes of the Tuareg rebellion remained unaddressed. In the time since, the region remains mired in fighting, with Kidal, nearby villages, and the Tinzaouaten border crossing with Algeria all experiencing ongoing attacks from insurgents.
These errors are common because incident-heavy reporting rewards immediacy. Trajectory analysis instead asks for patience. It does not reject urgency, but it separates urgent action from strategic over- or underreaction.
This distinction is especially important where policymakers face pressure to demonstrate rapid and visible results. In such circumstances, the temptation is to optimize responses for short-term incident suppression. That can be necessary, but if not paired with trajectory-informed design, it leaves the underlying conflict system intact.
The result is familiar: temporary reductions in violence followed by predictable relapse. NATO’s 2011 intervention in Libya is a classic example. While the bombing campaign against Colonel Gaddafi temporarily reduced violence, it did not produce a clear end to the conflict. Violence remerged two years later and lasted until the Government of National Accord and the Libyan National Army signed a ceasefire agreement in 2020. A memory-based framework makes relapses like this less surprising. It does not make it acceptable, but it does make it legible.
Three Practical Implications
If conflict has memory, then analysis, intervention, and stabilization planning should be trajectory-aware by default. That requires three practical shifts for policymakers and practitioners.
Classify localities by pathway, not only incident load.
Incident counts tell you where violence is high today, not whether a locality is moving toward calm or systemic lock-in. Analysis should routinely classify whether local areas are on ephemeral, recurrent, or entrenched pathways. The strategic question becomes “What transition is plausible here in the next twelve to twenty-four months?” not simply “How high is violence now?” This means reserving resource-intensive response efforts for localities showing signs of conflict entrenchment, while relying on limited, yet rapid responses for areas experiencing ephemeral spikes in violence.
Track movement between, not just within, pathways.
A district that moves from a series of short single year episodes of violence to a multiyear cycle can signal meaningful change in conflict dynamics even if total incidents are stable. Similarly, an annual decline in incidents in an entrenched district is not enough to conclude that the cycle is ending. Decision-makers need trajectory metrics that are attuned to how cycles transition, especially about how they end. This requires continuous commitment to tracking not just incidents, but also their locations and the relationships among the conflict actors.
Match intervention intensity to conflict stage.
A one-size-fits-all response can be wasteful at best and destabilizing at worst. Ephemeral local flare-ups often need rapid containment and calibrated support, not maximal militarization. Recurrent zones require repeated stabilization cycles plus governance continuity. Entrenched zones require long-horizon strategies that combine security pressure with institutional and economic redesign, since short engagements rarely alter structural conditions.
A Diagnostic Frame for Planners
Decision-makers do not need advanced modeling language to apply the core logic. They need a disciplined habit of asking temporal questions at locality scale, such as “Is this place showing signs of short shock and recovery, repeated cycling, or deep persistence?”, “Are we seeing transition toward reversibility, or transition toward lock-in?”, or “Are our interventions designed for the trajectory we have, or the trajectory we wish we had?”
Those questions can structure planning discussions at all levels, from policymakers to the tactical level. They can also sharpen communication with policymakers by replacing generic risk levels with interpretable conflict pathways.
The most consequential shift in this framework is both conceptual and practical. Analysts and planners should not consider violence as a sequence of isolated incidents, but rather a sequence of local pathways with momentum.
When response strategy ignores memory, policy drifts toward repetition. The same instruments are applied to different problems, and mixed results are misread as bad luck. But when strategy accounts for memory, interventions can be calibrated to trajectory type, timing can be aligned to transition points, and assessment can focus on whether pathways are actually changing.
Irregular warfare is often described as a contest of adaptation. That description should include our own analytic and planning habits. If conflict systems learn over time, intervention approaches have to learn over time too. Conflict has memory. Responses should remember it.
Dr. Steven M. Radil is a political geographer, geospatial scientist, and founder of GeoPublic Analytics. He works with international organizations to examine how geography shapes armed conflict, with a focus on policy-relevant conflict analysis in Africa and beyond. Dr. Olivier J. Walther is an Associate Professor of Geography at the University of Florida. His work has pioneered the introduction of social network analysis to the study of borderlands in Africa. Together, they have introduced the concept of the spatial conflict life cycle and trajectory analysis in conflict studies.
Main image: Jihadists in northern Mali pose a threat to countries in the broader Sahel. Source: Magharebia.com.
The views expressed are those of the author(s) and do not reflect the official position of the Irregular Warfare Initiative, Princeton University’s Empirical Studies of Conflict Project, the Modern War Institute at West Point, or the United States Government.
If you value reading the Irregular Warfare Initiative, please consider supporting our work. And for the best gear, check out the IWI store for mugs, coasters, apparel, and other items.
Leave a Reply