The Logic Behind why Torture is an Ineffective Interrogation Tactic

One question over the last decade since the beginning of the war on terror (the undercase is intentional) is whether or not torture should be used against captured terrorists. Due to the lack of direct nationality association for most terrorist a number of individuals believe that the provisions of the Geneva Convention do not apply to these individuals when captured. Any lingering morality concerns could be explained away through utilitarian arguments (the provided information will help many more individuals at the expense of physically and/or mentally injuring one person) or nationalistic arguments (these individuals are have declared war against us thus they are our enemy and deserve no quarter). Thus with no international legalities or moralities to prevent the act of torture itself, to these individuals the issue entirely involves the usefulness of torture in an interrogation environment.

The goal of interrogation is to extract accurate and useful information. With this goal in mind it is important to understand that there are two types of prisoners that can be captured: one who is utterly loyal to the particular cause whatever it may be (some could regard these individuals as fundamentalists) and one who is loyal to the particular cause due to some conditional element (money, kinship, etc.). For the purpose of an analysis with respects to the goal of interrogation it makes sense to define a best case and worse case scenario to determine the usefulness of torture on both prisoner types.

For the utterly loyal prisoner it stands to reason that success of the cause is the only important result, whether or not the individual is alive to welcome that success to reality is of secondary concern, (being alive to see it may be preferred, but not required). This attitude typically makes up the mindset of martyrs. With this attitude in mind it is rational to expect that from the interrogator’s perspective an utterly loyal captive will not give up any information that would prove detrimental to the cause. Therefore, the best case scenario emerging from torturing this type of individual is that the individual does not divulge any information and the worse case scenario is that all information given is either false or irrelevant.

A torture based interrogation of a non-utterly loyal individual is more interesting because there is a chance that the prisoner may talk, however the reliability of that information is certainly questionable. For non-utterly loyal individuals it is appropriate to move from the interrogator’s perspective to the prisoner’s perspective relative to the nature of the interrogation because as previously mentioned utterly loyal individuals have only one perspective. For the prisoner he/she is being tortured to give up information, so saying something should stop the torture session for a given time period. For the moment suppose that the prisoner gives up accurate information that is detrimental to the cause he/she is supposed to be supporting. After giving up this information the torture session is ceased and the information is checked and/or acted upon.

The best case scenario for the prisoner in the ‘accurate information’ scenario is that when the information is confirmed as accurate and is used to damage the opposing side the prisoner will no longer be tortured for the duration of his/her captivity because he/she has provided a honest exchange with his/her captors. Any expectation of release is not rational based on two reasons; first it does not make sense for an opposing party to release a captured enemy during an ongoing conflict unless required for some form of negotiation or prisoner exchange. Second if the individual is willing to give up accurate information under torture it makes that captive a valuable commodity, especially if more information is thought to be required in the future, which it almost always will be. Related to that second reason for not releasing the prisoner, the worse case scenario for giving accurate information is a higher probability to be tortured again because accurate information was given during the first torture session. If the point of the torture is to acquire accurate and actionable information it is very probable that the worst case scenario will win out much more often than the best case scenario for the prisoner.

Suppose the prisoner gives inaccurate information that will either be detrimental to the torturer’s cause or simply be of no consequence. The best case scenario for the prisoner is that because the information is inaccurate the prisoner is considered an unreliable source of information for the remainder of his/her imprisonment and is not tortured further. The worst case scenario is the prisoner is tortured again, not to extract information but as punishment for providing false information.

For summary purposes the best and worst case scenarios for each situation are shown below.

Truthful Best Case =
No further torture and damage to the captive’s cause;

Truthful Worst Case =
Greater torture than average and damage to the captive’s cause;

Non-Truthful Best Case =
No further torture and no damage to the captive’s cause with potential damage to captor’s cause;

Non-Truthful Worst Case =
Greater torture than average and damage to the captive’s cause with potential damage to captor’s cause;


After outlining the best and worse case scenarios it is important to rationally determine which of the scenarios is more probable. As previously mentioned there is little point in determining the probability for the utterly loyal captive because from the captive’s perspective the only outcome that matters is not revealing any information that would be detrimental to the cause, therefore the interrogator’s actions are irrelevant.

In the case of the non-utterly loyal captive giving accurate information the worse case scenario seems more probable because if the goal is to acquire accurate and relevant information why would captors deny themselves a proven avenue to continue to achieve that goal. Now it can be argued the frequency of the torture may change in order to ensure the psychological stability of the prisoner, but arguing a general cessation of torture in such a situation does not appear to be valid.

In the case of the non-utterly loyal captive giving inaccurate information determining the more likely scenario is dependent on two factors: the temperament of the captive and the temperament of the captor. If the captor is inherently savage or views the group represented by the captive as such then the worse case scenario of torture for punishment sake is more probable using the guise of ‘information extraction’. If the captor is not savage or is concerned about the treatment of any of their potential prisoners then the best case scenario of no future torture is more probable.

Therefore, from the prisoner’s perspective with the only major difference between the best case scenario for accurate information vs. inaccurate information being whether or not damage is done to the cause the prisoner support combined with a higher probability for the best case scenario when giving inaccurate information vs. accurate information it stands to reason that a prisoner should either give inaccurate information over accurate information when tortured. If this logical deduction holds then from the captor’s perspective engaging in torture for the purpose of extracting accurate and usable information is a futile effort and may even damage the cause.

Overall based on the characterization and basic personality traits of individuals that would be eligible for torture, the act of torture itself in interrogation does not appear to be beneficial on a logical perspective towards achieving the overall goal of extracting accurate and useful information.