Carnegie Mellon University
April 12, 2011

Press Release: Crossing the Line: What Constitutes Torture?

New Study Suggests Inability To Empathize With Those Suffering Pain May Prohibit One’s Ability To Identify Torturous Acts

Contact: Shilo Raube / 412-268-6094 /

george lowensteinPITTSBURGH—Although nearly all nations of the world condemn torture, and most have signed on to treaties that prohibit it, there is much less agreement about what specific acts constitute torture.  Almost all statutes dealing with torture define what is torture based on severity of pain. For example, the United Nations Convention against Torture — ratified by more than 150 countries — defines torture as the “infliction of severe physical or mental pain or suffering.” 
In a forthcoming issue of the journal Psychological Science, however, researchers Loran Nordgren of the Kellogg School of Management at Northwestern University, George Loewenstein of Carnegie Mellon University and Mary-Hunter Morris McDonnell of Harvard University demonstrate just how subjective and biased such judgments can be.
In four studies, the research team asked participants to make two types of judgments about interrogation tactics: the degree of pain produced by the tactic, and whether that tactic should be classified as torture (and hence prohibited). The four studies focused on three common interrogation techniques: exposure to cold temperatures, sleep deprivation and solitary confinement. Drawing on their own prior research on “empathy gaps” — the idea that people in one emotional state are unable to imagine how they would feel or behave in another — the authors predicted that subjects who experienced even a mild version of the same physical or psychological pain produced by a tactic would rate the pain produced by the tactic as worse, and would be likely to classify that tactic as torture. 
In all studies, some of the subjects made these judgments without actually experiencing the distress of the interrogation tactic (as is generally the case). Other subjects made the same evaluation while experiencing a mild version of the specific form of pain produced by the tactic, by submerging their arm in ice cold water, standing outside in the Chicago cold, participating while in a state of mental fatigue at the end of a nighttime class after a full day of work, or playing a social exclusion game. In all four experiments, subjects who got a small taste of the pain judged the tactic to be more painful, and were more likely to define it as torture, than those who made their judgments pain-free.
“This research comes at a critical moment in the debate surrounding the limits of interrogation procedures,” said Nordgren, a professor of management and organizations at the Kellogg School of Management. “By demonstrating inconsistencies and biases in people’s judgments of what constitutes torture, our findings suggest that those who set torture policies are likely to be fundamentally misinformed. Our studies provide a systematic investigation of what talk show host Erich ‘Mancow’ Muller learned first hand when, to prove that waterboarding was not torture, he voluntarily subjected himself to the procedure. Within seconds, he had changed his perspective.” 
Moreover, an interesting finding from the studies presented in the paper was that having experienced the pain only minutes earlier was no different from never having experienced it; judgments of severity and of whether a method constituted torture were only affected by the judge’s immediate state of pain. Loewenstein, the Herbert A. Simon Professor of Economics and Psychology at CMU, has observed similar patterns in earlier research.
“While people may feel they can judge a tactic because they have experienced the pain it produces at some point in the past, this research suggests that people can only empathize with pain when they are directly experiencing it,” Loewenstein said. “Those who make policies that guide the use of torture, of course, are unlikely to be in a state of pain when determining what interrogation tactics can be used in what circumstances.”
In a companion piece, published in the Vanderbilt Journal of Transnational Law, the authors explore the legal implications of their findings. Legal scholars draw a distinction between illegal actions such as rape that are malum in se — i.e., inherently bad — and those, such as illegal tax accounting practices that are malum prohibitum — wrong only because they are prohibited. Acts that are malum in se should be avoided, even if they are not expressly prohibited; in contrast, it is perfectly acceptable to hew close to the line on acts that are malum prohibitum — e.g., if a zoning regulation prohibits buildings from being more than 90-feet tall, a building that is 89-feet-tall is perfectly acceptable. McDonnell, a student at Harvard Law School, says that by showing that judgments of torture are highly subjective, the research shows that a malum prohibitum approach is not practical. 
“Since torture is, to a great extent, in the eye of the beholder, any attempt to hew close to the legal standard risks crossing the line,” McDonnell said.

Pictured above is George Loewenstein, the Herbert A. Simon Professor of Economics and Psychology at CMU.