Torture, Culture, War Zone Exposure, and Posttraumatic Stress Disorder Criterion A’s Bracket Creep—Reply

In reply

Bracha and Hayashi appear eager to challenge the conclusions of our March 2007 article¹ because of “major methodological and conceptual limitations that severely limit its generalizability.” Linkage sampling might have its problems but these do not necessarily invalidate the results. The study conclusion that “physical” and “nonphysical” stressors show substantial overlap in their immediate impact needs no statistical analysis; the data presented in Table 1 of our article¹ speak for themselves. In addition, linkage sampling does not characterize the whole sample; the 102 Belgrade site participants were mostly consecutively recruited from 2 associations for war veterans and prisoners of war. The results based on this subgroup do not present a different picture.

A good way of testing the generalizability of findings is their replication in different samples. The recent study is the last in a series of 4 studies that examined this issue using the same methods. The previous studies, altogether involving 203 torture survivors from a Middle Eastern country, included highly resilient political activists,² nonactivist ordinary civilians,³ and a mixed group of activists and nonactivists (M.B., unpublished data, 1996). We have therefore validated the results in 4 different samples showing considerable variance with respect to culture, ethnicity, socioeducational status, political belief systems, resilience levels, and trauma severity and context (eg, war, a military regime, torture for nonpolitical reasons). In a 1996 article,⁴ we reported a finding of no association between physical torture and posttorture psychological status. The recent study findings are thus fairly robust and replicable across different samples.

The argument that our study of “European civilians exposed to a civil war” is not relevant for international law “regarding interrogation procedures of terrorist-camp–trained detainees in Iraq and Afghanistan” is seriously flawed for many reasons. Implicit in this statement is that international law applies to ordinary people “vulnerable” to traumatic stress but not to “enemy combatants” with “diminished vulnerability” to PTSD. First of all, this is a morally unacceptable position. Furthermore, it is a serious logical fallacy to think that torture is not torture if someone can effectively endure it. Diminished vulnerability is a misleading term, playing down the likely effects of torture on “enemy combatants.” Our work shows that highly resilient militant activists are much more severely tortured than ordinary people³ and about a third subsequently develop PTSD.²

The comparison by Bracha and Hayashi between our sample of “European civilians” and “terrorist-camp–trained detainees” is based on unjustified assumptions about the nature of these groups. How do they know how many of the interrogated detainees are “terrorist-camp–trained” people? In a methodological critique of a research study, such statements need to be based on reliable evidence, not just on personal opinion. Opinions differ on this issue, as illustrated by a recent New York Times editorial,⁵ which noted that “ . . . a large portion, if not a majority, of the inmates at Guantanamo Bay” are “innocent men.”

The misconceptions of Bracha and Hayashi about our study sample are not supported by our data. As indicated in our article, 68.5% of the participants were combat veterans. Of these, 88% had volunteered for military service and 72% volunteered for active combat duty. Using a validated Psychological Preparedness for Torture scale,³ we found that 58% were moderately to very well trained in endurance and 56% were fairly to very committed to a war cause. Such a profile hardly fits a characterization of our sample as helpless “European civilians exposed to a civil war.” And this should not come as a surprise to Bracha and Hayashi. How else do they think such a bloody war with all sorts of violence on all sides would have been possible without a strong commitment to a war cause?

We also have data on the participants' beliefs about various issues measured by a 48-item Emotions and Beliefs After Trauma questionnaire.⁶ Eighty percent of the survivors reported stronger faith in God, 80% reported no effect of torture on their religious beliefs, 89% felt their faith in God was of great help during difficult times, 71% attributed their suffering to God's will, 77% believed God is on the side of the poor and oppressed, 74% believed their torturers would be punished in the afterlife, and 70% believed their suffering would be rewarded in the afterlife. Such beliefs hardly reflect the image of Bracha and Hayashi of a “formerly communist European culture” under “45 years of official state atheism with firm belief in the finality of death.” Moreover, if “the belief in lavish rewards in the afterlife for fighting and especially dying for their cause” is an indication of “fundamentalist religiosity,” as Bracha and Hayashi appear to contend, then our results suggest that this is not particular to Islamic faith, given that only 18% of the study participants were Muslim, while 77% belonged to the Catholic or Orthodox faith (5% stated no allegiance to any faith). Let us also stress here the important point that religious beliefs had no association with PTSD in our study.⁶

We completely agree with the argument that the Cartesian “physical vs psychological” dichotomy is a flawed approach. We are intrigued by the fact that Bracha and Hayashi raise this issue as a criticism of our study, considering that the purpose of the whole article was to demonstrate precisely this point. We categorized the stressors in ways that would allow testing of particular arguments that certain forms of “ill treatment” do not amount to torture. In any event, this issue has no bearing on our conclusion regarding our first study hypothesis (ie, stressors that do not involve physical pain are not associated with less distress and uncontrollability). Table 1 in our article¹ shows how individual stressors compare with each other in this respect. There is not even a need to group the stressors in any way to see the message that comes across.

Our second study hypothesis (ie, physical torture is more likely to be associated with PTSD and depression than nonphysical stressors), however, did require a categorization of stressors as physical and nonphysical. The results show that nonphysical stressors have a cumulative impact comparable with that of physical torture. In our article,¹ we noted that stressors in a captivity setting occur concurrently in clusters and interact in complex ways to potentiate the effects of each other. A priori judgments about the relative impact of individual stressors independent of their context are thus likely to be very misleading. Nevertheless, Bracha and Hayashi do not hesitate to express personal opinions about the nature of the stressors in total disregard of the evidence presented in the article. Common sense dictates that if sound empirical evidence does not support an a priori conceptualization of trauma, then it is the latter that needs to be questioned, not the former.

We are quite amazed by the view that defining falaqa as physical torture “trivializes” the meaning of the term. The notion that falaqa is a “widespread form of parental discipline in the Middle East” is far from the truth, reflecting nothing more than a stereotypical image of the Middle Eastern culture. Moreover, beating of all kinds is a prevalent act of violence in all cultures and this certainly does not make it a less severe form of trauma. Indeed, (unpublished) evidence from our first study² showed that beating in any form was the second most important predictor of PTSD (after asphyxiation) in highly resilient political activists. In our recent study,¹ it was rated as one of the most distressing forms of torture (a mean rating of 3.6 on a 0-4 scale), on par with electrical torture, burning parts of the body, and forced extraction of teeth. Bracha and Hayashi conveniently ignore such evidence.

In conclusion, none of the issues raised by Bracha and Hayashi have any serious implications for the validity of our study conclusions. We do, however, agree with them on one important point. The “physical vs psychological” dichotomy is not only a seriously flawed methodological approach but also a dangerous one that should be avoided in all considerations of what constitutes torture. What also needs to be avoided are a priori judgments about which stressor is worse than the other, which unfortunately characterize much of the current debate on what constitutes torture.

Correspondence: Dr Başoğlu, Section of Trauma Studies, Institute of Psychiatry, King's College London, Box PO91, DeCrespigny Park, Denmark Hill, London SE5 8AF, England (spjumeb@iop.kcl.ac.uk).

Financial Disclosure: None reported.