What We Fight About When We Fight About Torture

Right now we are locked in a complex dispute over the claims in the SSCI Torture Report that the CIA’s torture program was ineffective (as well as illegal). Part of the dispute can be frustrating because I think we are conflating a number of more distinct questions when we ask whether the torture was effective or not. Consider the following article from John Yoo who says that the torture report should be confined to the “dustbin” of history because it is inaccurate. He claims that torturing the detainees helped the CIA find Osama Bin Laden plain and simple.

We need to be more precise in order to have this conversation. Effectiveness or ineffectiveness are actually cluster concepts composed of more specific elements. I will try to tease apart the components here (there are at least five).

First, there is the issue of whether the CIA received intelligence from torture. In evaluating the dueling claims from the SSCI report and the CIA, it seems clear to me that the CIA did receive some intelligence about Bin Laden’s courier from the detainees who were tortured. The real question is evaluating the significance of that intelligence as compared to the other data points in the overall intelligence assessment of the CIA. That’s a complicated question and simply asking whether or not the CIA received intelligence about the courier from the tortured detainees does not tell you anything about the significance of the information. For that you need to ask some different questions which I now elaborate.

Second, there is the question of whether that intelligence was also received from other sources. Even if the tortured detainees provided intelligence about the courier, the more relevant question is whether that information was also received from our sources. That makes a huge difference. The critics of the report (including the CIA) make it sound as if the information from the courier came exclusively from the tortured detainees, but in fact this might have been a situation of overdetermination. The CIA already knew about the courier. If they received the information again from a tortured detainee, it is literally true that they received intelligence from the tortured detainee but the significance of that information is substantially reduced (perhaps to zero).

Third, there is the question of whether the intelligence could have been received from a non-torture source, either by non-coercive interrogation or some non-interrogation method. This question is important because it is relevant to the issue of “unavoidability” that plats a part in the legal analysis of necessity. At least some of the information came from multiple sources including detainees who provided the information before they were tortured. This suggests that torturing the detainees was avoidable because there were other non-torture avenues available for the CIA to get the information. It is also important to ask — and not enough people are asking and discussing this — whether the CIA could have used methods other than interrogation to get information about the courier. Of course, this discussion is stymied by the fact that the public does not have access to CIA methods and practices, which are classified. But how can we determine that the torture was indeed “necessary” without making explicit reference to the lack of other avenues? Unfortunately the CIA does not discuss these other avenues, but they really need to if they want anyone to accept their conclusion that the torture was truly necessary.

Fourth, there is the question of whether the torture saved lives. This is a counterfactual question because it requires imagining a world without torture and asking which terrorist attacks would — and would not — have occurred. This is guess work. When the CIA and their surrogates argue that the torture saved lives, they are asking everyone to engage in a mighty big thought experiment and what the world would have looked liked if they had followed the legal prohibition against torture. This is closely connected to the issue of unavoidability but it really is a separate question. Did it save lives? I have no idea. But at the very least the SSCI report shows that the CIA has failed to make the case that it saved lives.

Fifth, there are the first-order normative questions that are allegedly separate from effectiveness entirely, i.e. regardless of the answers to (1)-(4), was tortured legally or morally appropriate? Obama says that torture is wrong but he refuses to say whether it was effective or not. But these questions are linked in an interesting way. If we are debating whether the necessity defense should apply to torture (which I’ve written about extensively), at least part of the analysis is whether the torture is unavoidable. If torture is ineffective and useless, then it is clearly avoidable and the necessity argument does not apply — regardless of the rest of the legal argument. Of course, there might be other moral and legal reasons to reject torture, but the application of the necessity argument seems central to me.

The present discussion in Washington, DC, is conflating all of these questions into one incoherent mess.