During our training seminars, participants have asked about the value of identifying a suspect's truthful behavior by asking questions to which it is known that the suspect is telling the truth. There are a number of valid reasons for starting an interview with several minutes of nonthreatening background questions. These include:
- Establish rapport with the suspect;
- Establish a non-accusatory tone to the interview;
- Evaluate the suspect's intelligence, comprehension of English and mental state;
- Evaluate the suspect's "normal behaviors" e.g., eye contact, response timing, typical speech patterns, etc.;
- Establish a pattern of taking a written note following each of the suspect's response.
However, not included among these goals is the attempt to establish a particular suspect's behavior when telling the truth.
Since the act of lying or telling the truth represents a specific, identifiable behavior, there is an expectation that there will be some corresponding specific, identifiable behavioral manifestation of lying or truth-telling. This was the underlying presumption of many early lie detection instruments. The Fordham Pathometer, for example, had a needle that pointed to either "truth" or "lie" based on changes in the subject's electrodermal resistance when responding to the investigator's questions. How accurate was the Pathometer in detecting truth or deception? It had no accuracy. To understand why, consider the following research finding:
Sixty randomly selected video-taped interviews of criminal suspects were selected as part of a major study. Thirty of these suspects were innocent (truthful) and thirty were guilty of the crime under investigation (deceptive). Each interview started with background questions to which the suspects were telling the truth (spelling of their name, providing marital status, current address, employment information, etc.) Six evaluators, who had training in detection of deception but no knowledge of the cases selected for analysis, were asked to render an opinion of guilt or innocence based on evaluating these suspect's behavior when answering these background questions. Even though the 30 innocent and 30 guilty suspects were all telling the truth when answering the questions, the evaluators were able to accurately identify which suspects were guilty of the crime above chance levels.
Or consider the "Silent Answer Test" which is administered during a polygraph examination. During this test, the subject is instructed to think the truthful answer to the test questions, but not to answer the questions out loud. So when the examiner asks relevant questions addressing the crime under investigation, the suspect remains silent and says nothing. Research shows that guilty suspects still exhibit significant physiological responses to relevant questions during a Silent Answer Test - even though no lie was told!
These, and other examples, clearly indicate that lie detection methods do not detect lies or truth. Rather, they identify a person's underlying expectations (based on that person's knowledge of guilt or innocence) within the environment. These expectations, in turn, affect the person's behaviors, thoughts, emotions and perceptions -- it is the analysis of these secondary psychological processes that are used to infer whether or not a person is telling the truth or lying.
Consider two individuals, Bob and Bill, driving through a border crossing. Bob knows that he is a US citizen with a proper passport and is not trying to smuggle anything illegal across the border. Bill enters the border crossing with quite a different expectation. He knows that he has a falsified passport and that there is cocaine hidden in a compartment within the dashboard.
Bob approaches the crossing point and appears casual, but interested. He has his passport ready to show the agent and responds to initial questions directly and appropriately. His behavior and demeanor is similar to 99.9% of other people the agent has seen that day. Bill approaches the crossing point looking straight ahead through the windshield with his hands fixed at the 9 and 3 o'clock position. He exhibits little eye contact and offers short, hesitant responses to the agent's initial questions. After being pulled aside for further questioning and inspection, the hidden cocaine is found and Bill confirms that his passport is falsified.
Did the agent detect Bob's truthfulness and Bill's deception? No. The agent identified manifestations of Bob and Bill's different expectations going through the crossing point. Many of the observations that lead to inferences of guilt or innocence have nothing whatsoever to do with lying or telling the truth.
An observation that may challenge this concept is that there do seem to be certain behaviors that occur when someone tells the truth (use of illustrators, direct eye contact, use of descriptive terminology, on time responses, broad denials, etc. ), and other behaviors that occur when someone lies (poor eye contact, delayed responses, grooming behaviors, erasure etc.). While this is true, upon further examination, an investigator would find that these behaviors are not unique to truth-telling or lying. In other words, people who are telling the truth could also exhibit poor eye contact, delay their response to questions, engage in grooming behaviors and erasure at some point during the interview. Consider the following question/response block from a suspect who is telling the truth:
Q: Have you ever been questioned before by a police officer?
A: (Eyes, drop, delay) Yeah. Back when I was in high school I got caught with some marijuana. (Cover mouth with hand) I ended up paying a fine and going through a drug education program. It was stupid of me I know. (Laugh, dusting sleeve).
The listed behavior symptoms are not the result of lying. They indicate that the suspect is experiencing embarrassment, shame and indecision as to whether to reveal the prior offense to the investigator.
Despite hundreds of years of attempts to identify a behavioral, physiological or cognitive response that only occurs when a person lies or tells the truth, current research and technology have not identified any unique response associated only with truth or deception. Consequently, assessing another person's credibility requires making inferences that involve a number of underlying assessments. Among these are that behavior symptoms must be evaluated in context considering such factors as the environment in which it occurred, the nature of the question asked of the suspect, and the content within the suspect's verbal response.