Read our COVID-19 research and news.

Torture Can't Provide Good Information, Argues Neuroscientist

Much ink has been spilled about the right and wrong of torturing terrorists, particularly since 16 April when the U.S. Department of Justice released memos detailing the “enhanced interrogation” of terrorism suspects during the Bush Administration.

But what is less often discussed is whether or not torture works at all. Generally the debate has been whether torture causes detainees to say anything, possibly lying, to end the torture. But in an article published today in Trends in Cognitive Sciences, a neuroscientist compares torture’s “folk psychology” assumptions to current models of brain function during stress and trauma. His conclusion: Torture causes brain damage that can wipe out memories of the desired information, or even create false ones.

Author Shane O’Mara, a neuroscientist at Trinity College in Dublin, Ireland, says the CIA torture strategy seems:

based on the idea that repeatedly inducing shock, stress, anxiety, disorientation and lack of control is more effective than standard interrogatory techniques in making suspects reveal information. Information retrieved from memory in this way is assumed to be reliable and veridical, as suspects will be motivated to end the interrogation by revealing this information. No supporting data for this model are provided; in fact, the model is utterly unsupported by scientific evidence.

But, O’Mara writes, research on stressed and traumatized brains have found changes that compromise memory retrieval:

Brain imaging in persons previously subjected to severe torture suggests that abnormal patterns of activation are present in the frontal and temporal lobes, leading to deficits in verbal memory for the recall of traumatic events. A recent meta-analysis of the relationship between pharmacologically-induced cortisol elevations (in the upper physiological range) concludes that it impairs memory retrieval in humans, as do psychosocial stress-induced cortisol elevations. On the other hand, mildly stressful events generally facilitate recall. The experience of capture, transport and subsequent challenging questioning would seem to be more than enough in making suspects reveal information.

In particular, O’Mara cites waterboarding as being counterproductive for obtaining reliable information:

Waterboarding is cited in the legal memoranda as causing elevations in blood carbon dioxide levels. Data on the effects of hypercapnia (increased blood carbon dioxide) or hypoxia (decreased blood oxygen) on brain function are not cited; nor are data on carbon dioxide narcosis (deep stupor or unconsciousness), which may be expected as a result of acute and repeated waterboarding. Brain imaging data suggest that hypercapnia and associated feelings of breathlessness (dyspnea) cause widespread increases in brain activity, including brain regions associated with stress and anxiety (amygdala, prefrontal cortex) and pain (periacquiductal gray). These data suggest that waterboarding in particular acts as a very severe and extreme stressor, with the potential to cause widespread stress-induced changes in the brain, especially when these are repeated frequently and intensively.

The top suspect for the 9/11 terrorist attack, Khalid Sheikh Mohammed, was waterboarded 183 times by the CIA. According to O’Mara, no one can provide reliable intelligence after this treatment.

In an email with ScienceInsider, Metin Basoglu, head of the Trauma Studies section at the Institute of Psychiatry of King’s College London criticized the paper.

’Torture does not work’ type of arguments ignore an important moral question. What if it did work? Would it then have justified its use in the name of ‘national security’ for example? By engaging in these discussions scientists implicitly acknowledge the moral validity of the latter position. Indeed, it is this position that has made possible arguments such as ‘light torture’ being acceptable under some circumstances...International law categorically prohibits torture under all circumstances and that's it! Torture is first and foremost a moral issue and this should always be remembered.

O’Mara replied in an email to ScienceInsider:

The job of the scientist (excuse my pomposity here) is to describe the world as it is. One could generate an infinite number of counterfactuals for the sake of the discussion. However, the moral arguments stand anyway, as does the legal position. I do note (I was allowed only 1500 words) that there are strong ethical and legal objections. But what I want to do in the paper is examine the lay psychology and lay neurobiology which holds that torture is efficacious, and show that the evidence is against the lay position: that Psych 101 classes teach already - extreme and sustained stressors have a deleterious effect on the brain systems supporting memory.