|LSI Laboratory for Scientific Interrogation, Inc. | home
SCAN: Deception Detection by Scientific Content Analysis
reprinted with permission from "Law and Order"
Vol 38 No. 8, August 1990
by Tony Lesce
Police investigators repeatedly question people who may be untruthful, or who may have an interest in concealing information. Therefore, investigators need a method of separating fact from fiction. "SCAN" can help them do this.
Various systems of deception detection have been devised. Some depend on analyzing the subject's body language; one consists of watching the movements of the subject's eyes. Gestures vary with nationality and culture, and can be misleading to someone unfamiliar with the culture and who doesn't understand the code. In the United States, with inhabitants from many cultures, this can be confusing.
Avinoam Sapir, a former Israeli police lieutenant, teaches a four- day course in SCAN in which he instructs police officers in how to obtain and analyze subject's statements. He began his career as a polygraph examiner for the Israeli police while working on his master's degree in criminology at Tel-Aviv University. In 1981, he began to conduct classes in SCAN for American agencies, such as U.S. Customs, Social Security Administrations, and others. He has led seminars for a variety of American and foreign police and investigative agencies. Among them are: U.S. Immigration & Naturalization Service, U.S. Army Criminal Investigative Division, U.S. Army Military Intelligence, Saskatchewan Police College, Canada, Ohio Association of Polygraph Examiners, Royal Australian Police College, the American Society for Industrial Security, and the FBI Academy.
How SCAN Works
SCAN depends heavily upon analyzing the statement, not upon observing "kinesics" or "body language". The basic premise is that the structure and contents of a subject's statement reveal whether there's an attempt at deception. SCAN is cross-cultural, in the sense that an interrogator can conduct an analysis in any language he understands.
According to Sapir, at least 90% of statements made are truthful, and most people do not attempt to lie directly. Instead, they hedge, omit crucial facts, feign forgetfulness, and pretend ignorance. The reason for this is that liars are reluctant to commit themselves to their deceptions, instead preferring to use conversational tricks to avoid damaging admissions.
Some use tricks, such as answering a question with a question, or glossing over critical points in the narrative. Another trick is bridging gaps with uninformative statements such as; "We talked" (about what?), or "afterwards" (after what?).
SCAN studies speech patterns, seeking tell-tale signs of deception by analyzing both structure and contents of statements. To do so, it is first necessary to obtain a suitable "pure" statement from the subject.
Obtaining A Pure Statement
Experienced investigators know that the most difficult way to obtain beneficial information is the simple question-and-answer routine, with the subject answering "yes" or "no". The point in eliciting a statement is to ask open-ended questions to avoid prompting the subject and encourage spontaneity, with the subject providing new information in response to the questions.
Investigators should ask as few specific questions as possible to avoid introducing their own information into the subject's statements. This is the well-known principle: "Get, don't give." At all costs, the interrogator avoids letting the subject know what he already knows or doesn't know. By giving the subject as little information and coaching as possible, the investigator seeks to obtain an uncontaminated version of the events. The preferred form is first-person, past tense.
This is essential because, as Sapir makes the point in his seminar, the interviewer cannot be objective. He has before him a file on the case, and probably has performed some investigation himself. If he has interviewed other subjects, or examined physical evidence, he's already developed some facts. This makes it very hard not to arrive at some conclusions, and an effort must be made to keep his ideas from contaminating the testimony. Most subjects will talk with relatively little encouragement or coercion. Even if the subject shows reluctance, it's often possible to cajole him into making a statement by using authority or by persuading him that he needs to unburden himself. In other cases, the investigator can condition the subject into being responsive by asking him a series of innocuous questions. Once the subject begins responding to simple questions, such as his name and address, it is difficult for him to stop, and the skillful interrogator can lead him gradually into a discussion of sensitive topics. Sapir advises that the investigator should admit defeat only if the subject digs in his heels and states:"I don't want to talk" in response to several efforts.
Sapir points out that eliciting a statement is not a 50/50 proposition. He considers an interrogation successful if the subject speaks about 95% of the time, and the investigator only about 5%. This minimizes the investigator's contribution to the final statement. He states it even more dramatically: "The person is dead. The statement is alive," to show that the person's words are more important than the person himself.
The need to avoid contamination is paramount. Sapir counsels interrogators not to introduce any topic into the conversation unless the subject first mentions it. This ensures that every fact or allegation the interrogator elicits comes from the subject, and not from another source.
It's not necessary to see the subject to analyze his statement. The interview is necessary only to obtain a statement, and the tool for SCAN is really the statement itself. The investigator goes over the statement word by word, line by line, teasing out important and subtle details with Scientific Content Analysis.
Speech patterns help distinguish between statements originating in memory or imagination. Often, pronoun use shows deception, such as an alleged kidnap victim using "we" in describing where the kidnaper and victim went and what they did. The "we" is inappropriate in this context, because the relationship is the "he took me," or "he forced me." Sapir demonstrates that there cannot be confusion about pronouns by pointing out that, even while recalling memories from early childhood, we can remember whether we were alone or with someone else. Therefore, a shift in pronoun use in a narrative denotes a shift in the relationship.
Gaps in the narrative also betray deception. Minor contradictions often conceal major omission. A person may skip over a critical point in his statement because he doesn't want to reveal what happened just then. The statement; "I don't remember" often conceals a critical detail which the speaker would rather forget, or at least avoid mentioning.
Some subjects adroitly pose their own questions to avoid revealing information. Asking "Is this important?" at some point in the narrative suggests that there is information with which the subject does not feel comfortable.
A change in tense indicates a strong emotional response. For example, relating the story of a rape can begin with the subject telling it in the past tense. When the rape begins, some subjects change to the present tense because recalling the memory is painful.
Content analysis is also important. An important rule that Sapir stresses is: "Nothing happens in a vacuum." This highlights the need to examine the circumstances of the alleged incident, as well as the incident itself. If the subject tries to be deceptive, he not only has to conceal or change the crucial detail he's safeguarding, but he has to "massage" the other facts of the case to fit his version. This often leaves traces, either inconsistencies between allegations, or gaps in the narrative.
The reason is that the deceptive person is working from imagination, according to Sapir, and the truthful person is working from memory. In reality, there are many unimportant details that coexist with critical facts, and these often find their way into truthful narratives. The result is that a deceptive statement will be a "stripped-down" version, while the truthful statement also contains tangential or irrelevant information.
According to Sapir, most deceptive stories (80%-90%) push the main issue to the end of the statement, and do not continue the narrative after that. Another difference between truthful and deceptive statements is that most deceptive stories do not mention emotion, and those that do locate the emotions logically, near the most threatening point. In reality, often the most severe emotional reaction is an after-shock, which takes place after the most threatening moment has passed. Yet another feature of most deceptive stories is that they are governed by logic, while true stories are not necessarily logical.
How Reliable is SCAN
According to Sapir, his method is as reliable as a polygraph examination. In this country, we must remember, the polygraph is an investigative tool, and not evidence. During the years since its introduction during the early 1920s, the polygraph has not succeeded in gaining admission in courts, except for rare exceptions. Some states, such as New Mexico, allow introduction of polygraph examinations as evidence, but also allow the jury the discretion of judging it valid or not.
Because SCAN depends only upon the subject's statement, it's a "cold" technique. Scoring is based only upon structure and content, not upon gestures, perspiration, eye movements, and other factors that are open to individual interpretation. Thus, it avoids the weakness of techniques that depend upon "global scoring." Global scoring considers all possible factors that may affect the situation, and much depends upon the interpreter's judgment. The bottom line is that SCAN is potentially more capable of development into a precise technique.
SCAN is not a universal solution. To apply SCAN, it's essential to obtain the subject's cooperation. It's also essential to work with an uncontaminated statement. An inept interrogator who discloses information to a subject can impair the quality of the results. Likewise, it's difficult to use SCAN if another person has contributed to, or edited the statement, as often happens with press releases, published books and articles. although the subject may acknowledge or claim authorship, it's often difficult to tell whether a part of the statement originated with the subject, a ghost-writer or editor.
SCAN is a versatile method because it's free from constraints that limit other techniques. Unlike the polygraph, it does not require attaching the subject to a machine, and obtaining a signed release, which is sometimes required. The subject is therefore unaware that his statement will be the object of specialized treatment and analysis.
Unlike both the polygraph and the voice stress analysis method, SCAN does not require high-tech equipment, only a pencil and paper. It also does not require the subject's presence during analysis, because once the pure version of the statement is on paper, the analysis can proceed when convenient. Indeed, an investigator can obtain a statement, and have it analyzed later by a specialist.
Like other psychological and behavioral techniques, such as profiling, SCAN serves to focus an investigation. It a reasonably reliable method of probing for leads. This method is adaptable to many law enforcement investigative situations.