Several years ago, my colleague Adam Nadel and I took confessions from a group of convicted war criminals. They were veterans of the Imperial Japanese Army who, during the Second Sino-Japanese war, committed the worst crimes imaginable: torture, rape, murder of children, and diabolical medical experiments upon kidnapped, unsedated civilians. They did not commit their crimes in moments of berserk breakdown or temporary insanity. They committed them over and over again, for years—cunningly, creatively, and with a joyful sense of competition over who could do the most.

In my book about these war criminals, who call themselves the Chukiren, I ask two primary questions. Here is the first: What turns an ordinary man into a monster? The Chukiren came from all walks of life, but they all ended up doing the same things. Insofar as there was a pattern, it was simple and disturbingly familiar. To start, they were separated from their families and social networks, isolated into all-male groups that were walled off from outside judgment and moral referencing. Locked in their ethical echo chamber, they were then trained to see the world in restrictive binary terms -- safe vs. unsafe, good vs. evil, pure vs. impure, us vs. them. Either/or thinking promised simple, total solutions to complex problems, and in that way prepared them for the irreversibility of extreme and violent action. Meanwhile, they were beaten down physically and mentally. Stress, sleep deprivation, and an arbitrary system of punishments and rewards made them feel like they had little or no control over what would happen to them at any given point. They began to crave a sense of control as an almost physical need: violent domination of others was the quickest and most satisfying method of satisfaction. Finally, and most importantly, they were trained in atrocity through a deliberate, step-by-step process of incremental escalation. Whenever they acclimated to a particular level of violence (for instance, practicing their bayoneting on dead bodies), they would be initiated into a more extreme version (practicing their bayoneting on frightened prisoners tied to trees). Eventually, the most appalling cruelties became background noise.

When Nadel approached me to set up the interviews with the Chukiren, I hesitated. I worried that the effort to understand perpetrators might be construed as an effort to exonerate them. I worried that telling the perpetrators’ stories would be an injustice to those they had permanently silenced so many years ago. But in the end I agreed. Because in the many years I had spent interviewing survivors of atrocity, I had been haunted continuously by a single question: What kind of person could have done those things to you? Meeting these war criminals, hearing their stories, was my chance to find answers.

At least that is what I told myself. But over the course of the project, I found it increasingly difficult to disentangle rigorous inquiry from morbid curiosity. And I found it especially hard to write about atrocity from the perspective of the perpetrator without sinking into grisly sensationalism or licensing the voyeuristic gaze. Confusion about my own motivations forced me to question some of my basic assumptions about human rights work. And this led to my book’s second primary question: What makes us want to hear stories of atrocity?

The simplest and most widely believed answer is this: we are drawn to stories of suffering because we are ruled by empathy. We look because we care; we bear witness to atrocity because it is in our nature to grieve for each other. But others argue that this is a mistake. They argue that what we feel when witnessing the suffering of others is not empathy at all. It is a variant of narcissism and self-care. Jean-Jacques Rousseau characterized this suspicion eloquently when discussing the seemingly extraordinary capacity of humans to care deeply even about the suffering of invented characters in stories and plays. Perhaps we are drawn to displays of suffering in fiction not because we are so capacious in empathy, but because we are so capacious in our self-regard. Rousseau writes:

In giving our tears to these fictions, we have satisfied all the rights of humanity without having to give anything more of ourselves; whereas unfortunate people in person would require attention from us, relief, consolation, and work, which would involve us in their pains and would require at least the sacrifice of our indolence, from all of which we are quite content to be exempt. It could be said that our heart closes itself for fear of being touched at our expense. In the final accounting, when a man has gone to admire fine actions in stories and to cry for imaginary miseries, what more can be asked of him? Is he not satisfied with himself? Does he not applaud his fine soul? Has he not acquitted himself of all that he owes to virtue by the homage which has just rendered it? What more could one want of him? That he practice it himself? He has no role to play; he is no actor. (Letter to D'Alembert and Writings for the Theater, 1758)

Rousseau’s case against empathy has serious implications for contemporary human rights work. As rights workers, we tell awful stories to generate empathy in distant spectators because we believe empathy promotes individual helping behaviors and spurs institutional interventions. But what if human empathy is as confused and compromised as my own motivations were in undertaking this project? Indeed, what if empathy is not a solution but a problem? Some, for instance, argue that empathy is ineffective. It inspires spasmodic urges to help that fade quickly in the face of complex problems that require long-term commitment. Indeed, the uncomfortable experience of sharp empathy may very well promote empathy-avoidance in the future: that is, a refusal to notice the suffering of others.

Others add that empathy is, as a matter of simple resource distribution, counterproductive. Empathy exacerbates social suffering by inspiring a desire for simple, personalizing solutions to large structural problems. We are more likely to care about a single named baby trapped in a well than about hundreds of thousands of nameless babies born into lethal poverty, as Paul Bloom has argued. Worse, we are willing to let ourselves off the hook for responsibility to those hundreds of thousands precisely because we have performed the emotional labor of worrying over the one.

In the end, whatever we make of empathy, I do not believe it can suffice as an answer to the question, “What makes us want to hear stories of atrocity?” Indeed, if meeting the Chukiren taught me anything, it was that the two questions I had been pursuing as if they were independent of each other – “What turns an ordinary man into a monster?” and “What makes us want to hear stories of atrocity?” – are in fact only comprehensible as subcomponents of a larger and more difficult question: “What draws us to evil?”

There is no simple answer, but when I have come closest to thinking I understand, it is because I have begun to see the connections between our everyday desires and our historical monstrosities.