Friday, November 11, 2016

The Scharff Technique: On How to Effectively Elicit Intelligence from Human Sources


Much of the time intelligence agencies and law enforcement capabilities do not have the whole picture about a particular problem which is being analyzed. To that end, the need for individuals who are trained in human intelligence (HUMINT) is immensely important to gathering the right information to fill the gaps in knowledge. Part of this is the process of elicitation.

During World War II an individual by the name of Hanns Joachim Scharff (1907-1992) worked at the German Luftwaffe’s Intelligence and Evaluation Center (Auswertestelle West). During Scharff’s tenure with the Luftwaffe he had the opportunity to refine his interpersonal skills and hone in on his engaging personality. Additionally, during this posting he was able to through trial-and-error interrogate over 500 American and British fighter pilots.

At Dulag Luft where he began his interrogations he operated contrary to that of his predecessor. “Rather than compelling his prisoners to reveal classified data through the employment of coercive methods, his success was the result of carefully orchestrated, outwardly friendly exchanges with his prisoners.” Contributing to his success in eliciting information from his sources was the fact that U.S. Army Air Corps airmen were taught to expect violence and even torture when information was being elicited from them.

To be effective as an interrogator and collector of human intelligence three key pillars for success of gathering quality HUMINT are the collector’s ability to “create a non-adversarial relationship with the prisoners,” and “create an environment that might make the [Prisoner of War] POW momentarily forget he was being interrogated.”  Finally, to avoid hostility, the collector needs to “display a friendly demeanor.”

Scharff also knew the value of being able to put himself in the shoes of the adversary and develop strategies to counter the prisoner’s ability to conceal information. The following three principles are paramount about HUMINT contexts:

a       a.) A source typically forms a hypothesis regarding how much and what information the interviewer already holds;
b    b.) The source’s perception (of the information already held by the interviewer) will, in turn, affect his or her counter-interrogation strategies; and
c    c.) The counter-interrogation strategies employed will affect how much and what information the source reveals.

The conceptualization of Scharff’s method requires five distinct parameters to operate effectively:

1        1.)  To employ a friendly approach.
2        2.)  Do not press for information.
3        3.)  The illusion of knowing-it-all. (Interviewer pretends to know more than they do).
4        4.)  The use of confirmations/disconfirmations.
5        5.)   Ignore new information.

To measure the effectiveness of the technique Granhag et. al. (2016) developed five prime measures to assess the aforementioned parameters effectiveness in eliciting information:

1      1.) The amount of new information elicited.
2      2.) The extant understanding of the interviewer’s information objectives.
3      3.) The source’s perceptions of the interviewer.
4    4.) The more information the source perceives the interviewer to have the more favorable for        information will be elicited.
5     5.) The source does not realize they have divulged any additional information on their own volition.

After running an experiment which allowed a variety of interviewers to elicit information from human sources and corroborate findings with the above mentioned parameters while meshing the data gathered with extent literature the authors found the technique effective. “In brief, [the authors] have shown that the Scharff technique consistently outperforms the *Direct Approach on the most critical efficacy measures. The combined evidence marks that Scharff technique as an effective tool for eliciting human intelligence.”

*Direct Approach: U.S. Army Field Manuel: An initial approach of a source through the use of “open-ended and explicit questions posed in a business-like manner.


This article provided the academic community with research in an otherwise under-researched area shy of the classified community. Given this reality, it is hard to find issues with the research. The analysis would have been stronger if more insight into the methodology of the experimentation took place, however, the overall strength of the study is bound to the qualitative nature of assessing the Scharff technique. Also, the study addressed a limitation which revolved extensively around the studies inability to successfully replicate itself due to the fact, “typical HUMINT interaction[s] are very difficult or even impossible to reproduce in a laboratory setting.”


Par Anders Granhag, Steven M. Kleinman & Simon Oleszkiewicz. (2016). The Scharff Technique: On     How to Effectively Elicit Intelligence from Human Sources, International Journal of Intelligence and Counterintelligence, 29:1, 132-150.


  1. You could kind of guess that finding academic works on HUMINT methods was going to be a little too tough to find. However, this was still a good find. Scharff's more personable methods did yield great results. I feel that he did have a slight advantage in that the people he was getting information from were a bit more of a "captive audience" than someone else. But the methods are still great for use with gaining information through normal conversation. Good read, Tom.

    1. Thanks, Ruark! I found it interesting that DoD brought Scharff to the Pentagon after the war to be debriefed on his methods. Definitely revolutionary with some respects.

  2. Tom, nice find with this one. I was expecting articles like this when I think of elicitation practices. I was going to ask you if you knew how the authors set up their experimental interviews so they would mirror a captor/captive relationship but it seems from your critique that they were vague on that. I did also wonder why the fifth parameter of operating effectively was to forget new information. Can you tell me what that means in your opinion?

    1. Thanks, Eric!

      To answer your question about the fifth parameter. I read it to mean that if the source divulges information that is new to the interrogator/handler they should ignore the presence of that information so as to alert the source of that information being important. In doing so, they could continue to give up information that they think you still have. It's more of an OPSEC precaution for the interview or conversation.

  3. It was quite a coincidence that the packet of materials I sent the group to look over before the exercise contained this very article! Especially considering I hadn't seen your summary yet. Nice write up, hope you enjoyed the exercise. It was basically an extension of the role-playing setup your team made the week prior.