Recent reporting from Ukraine, Gaza, Israel, and other conflicts around the world provide disturbing anecdotal accounts of torture, including physical abuse, sexual violence, and psychological torment. Yet, despite decades of research, we still know surprisingly little about what torture actually looks like in practice. Why do different armed groups torture in different ways? Do they develop their own recognisable torture “repertoires”? And does greater public scrutiny deter perpetrators and push them to hide what they are doing?
To answer these questions, my co-author and I examined more than 1,200 documented torture incidents from Peru’s internal armed conflict (1980–2000), combining data collected by Peru’s Truth and Reconciliation Commission with witness and survivor testimonies. Peru provides a valuable case because the conflict involved multiple types of armed actors—the military, the police, Shining Path militants, and civil defence forces—who exhibited clear variation in torture practices between the groups and across time.
The Idea of a Torture Repertoire
A common argument in the literature is that organisations develop distinct “repertoires” or “regimens” of torture. According to this view, armed groups learn and reproduce recognisable combinations of techniques that become part of their organisational culture, training, and operational practice.
There was some evidence for this in Peru—but much less than we expected.
Different organisations did exhibit a preference for particular methods. Police were more likely than other actors to use electric shocks, asphyxiation, and psychological torture. The military used these methods too, but with different frequencies. The Shining Path relied heavily on beating, cutting, and immobilisation, with the civil defence forces often replicating those practices.
In fact, one of our most striking findings was the degree of overlap.
Rather than adhering to rigid and distinctive torture repertoires, most actors employed a broad range of methods. Military personnel, police officers, insurgents, and civil defence forces all used multiple techniques, often in changing combinations. Only the Shining Path displayed something approaching a distinctive repertoire, and even there the variation was substantial.
This matters because much of what we think we know about torture comes from broad historical or cross-national accounts. Looking closely at individual incidents in a single case reveals a messier reality. Torture appears to be more adaptable and situational than theories of consistent organisational repertoires would suggest.
Not All Torture Leaves Visible Marks
Another question we examined concerns what scholars call “stealth” torture.
Some methods leave lasting physical evidence. Others inflict severe suffering while leaving few visible traces. This observation has led some to suggest that governments facing legal scrutiny, media attention, and human rights monitoring may come to favour methods that do not leave visible marks.
The Peruvian case offers mixed support for this claim.
The police were the actor least likely to use torture methods that left permanent marks. At the other end of the spectrum, Shining Path militants were the most likely to employ visible scarring techniques. The military and civilian self-defence forces fell somewhere in between. At first glance, this pattern aligns with expectations: actors subject to greater oversight appeared somewhat more inclined toward less visible forms of abuse.
But other evidence complicated this observation.
As human rights monitoring increased over the course of the conflict, government actors did not systematically reduce their reliance on scarring methods. Victim characteristics associated with their ability to access torture reporting mechanisms—such as education, literacy, language, or whether they lived in urban or rural areas—similarly had little effect on the likelihood of being subjected to scarring torture.
In short, concerns about monitoring seem to explain some organisational differences in torture practices, but they do not provide a complete explanation.
Violence as Communication
The statistical patterns become clearer when paired with witness and survivor testimonies.
For the military and police, torture generally occurred while in detention facilities, military bases, and interrogation settings. The evidence suggests that perpetrators used a variety of methods and frequently combined them, but without a single recognisable formula.
The Shining Path was different.
Especially as the conflict progressed, torture often became part of a broader effort to intimidate civilians, demonstrate resolve, and impose high costs for resisting. In the early years of the insurgency, the movement used violence more selectively while attempting to build wider public support. As opposition to the movement grew and its prospects for success declined, highly visible forms of violence became more common. Torture served not only to punish but also to communicate. It signalled the costs of collaboration with the state and the consequences for opposing the Shining Path’s revolutionary goals.
The civilian self-defence forces present an intriguing case. Although they fought alongside the state against the Shining Path, some of their torture practices more closely resembled those of the insurgents than those of the military. This suggests that local conditions, imitation, and shared environments may sometimes shape torture practices as much as formal organisational expectations or training.
Why These Findings Matter
The broader lesson from Peru is that torture practices are more flexible than we often assume. Different actors use different methods, in different combinations, for different purposes.
Our findings provide only limited support for the idea that organisations follow a stable torture playbook. Instead, torture practices appear to evolve with changing circumstances, local dynamics, strategic objectives, and the decisions of individual perpetrators.
The study also raises questions about how torture spreads. If organisations with very different political goals and structures end up using many of the same techniques, then internal expectations and formal training may be only part of the story. Perpetrators can also learn through imitation, experimentation, and informal socialisation. Torture may not require doctrine to reproduce itself. The logics and underlying motives for torture are diverse and complex.
This has implications for efforts to prevent torture. Human rights monitoring remains important, and some actors clearly appear more sensitive to scrutiny than others. But reducing torture is unlikely to be as simple as exposing abuses or eliminating a particular set of techniques. The problem is deeper. Torture is not a single practice with a single logic. It is a flexible tool that different actors adapt under different circumstances for different purposes.
If we want to understand why torture persists—and how it can be stopped—we need to move beyond assumptions about repertoires and pay closer attention to how violence actually unfolds on the ground. Peru’s conflict offers an important reminder that the reality is often far more complex than existing theories suggest.
Dr Steven T. Zech is a Senior Lecturer in the Department of Politics and International Relations at Monash University where he teaches courses on contemporary security topics, political violence, and comparative politics. His general research interests include political violence, human security, militias, contentious politics, civility, and countering threats to liberal democracy.
This article is published under a Creative Commons License and may be republished with attribution.
