Before reading this post, we highly suggest that you read Deceptive Dimensions: Intro to Deception and The Ekman Nursing Student Study.
Lying, in essence, is just a game: the deceiver faces off against a detector, each using offensive and defensive strategies, both trying to succeed at singular opposing goals. It’s zero-sum, one-on-one competition with an equal, the third type of social situation “where unpredictability can be applied in Robert Greene’s strategic sense.”
So far, we’ve learned that the odds of success for high-stakes lying are random at best. We think you’ll agree that when your job, your marriage, or your criminal record is on the line, you need more control. To fix this, we need to know what your opponent, the detector, is looking for and what he’s basing his decisions on.
To form an effective unpredictable strategy in this game, you must first know what’s expected of you; only then can you avoid being thwarted by a counter-strategy. And to know what’s expected, you must steal the detector’s playbook.
The playbook we’re imagining is nothing more than a list of common behavioral cues used by average men and women when they’re forced into the detector role of a deceptive scenario. Some behavior-based detection strategies are probably air-tight, and aren’t likely to be circumvented without special training. Some will be specific to the type of social backdrop against which the deception is taking place. Some fluctuate in frequency of use depending on what’s at stake. And some are so universal, so predictable, that we can’t, in good conscience, let you be taken down by them. These are what we’re looking for. Once we know what behaviors detector’s use to make their judgments, we can identify which you should try to control. Not all of them can be controlled, of course, but those that can should definitely not be ignored.
Polygraphs (lie-detector machines) do not detect lies effectively, contrary to popular belief. This scene from Lie to Me demonstrates why. Take a second to watch it. You’ll see that when the polygraph demonstrator answers control questions (questions the answers to which are obviously and undeniably true), the polygraph only interprets them as true when the demonstrator is calm. When he’s sexually aroused, the device malfunctions. Polygraph machines only register arousal, not actual lies. Introducing any stressor stimulus into the environment – for instance, a sexy latina chick in a skin-tight, v-neck dress – will cause the machine to interpret true statements as lies.
Popular methods for “catching” liars are just as rudimentary as those used by polygraphs. Surveys show that the average untrained detector looks exclusively for obvious signs of nervousness. That’s all. Now, liars in general could very well be more nervous than truth-tellers, but nervousness about lying is indistinguishable from nervousness about being disbelieved, nervousness about whatever consequences may result from failure, and nervousness related to the imposing presence of the detector. When persistent anxiety extends over the entire duration of a deceptive interaction, it garbles the signals. This helps explains why successful deception rates are random; not only are detectors looking for deception leakage in all the wrong places, but they’re looking for the wrong things to start with.
Nonetheless, in symmetrical/high-salience situations where the stakes are high, the detector makes the rules, and you, the deceiver, are playing his game. All the more reason to steal that playbook.
The Hocking Study
In 1980, John Hocking and Dale Leathers, speech communication professors from the University of Georgia replicated the Ekman nursing student study we examined last time, but using a different theoretical perspective.
Prior to his experiment, Hocking analyzed survey data to outline the popular cultural stereotype of a liar. In the survey, respondents overwhelmingly described liars as nervous, defensive, and fidgety. Liars, they said, will display a wide range of anxious behaviors, such as facial manipulators (i.e., touching the face), restless lower body movements, and lipwetting. About 65 percent of them also expected liars to avoid eye-contact. (That last belief will come up repeatedly after this study; research suggests that eye-contact avoidance is a myth; liars actually make more eye-contact.)
Since the stereotype of a liar is so pervasive, Hocking hypothesizes that avoiding stereotypical typecasting is paramount for successful deception; a liar, he says, must monitor and control the behaviors that are under scrutiny, suppressing stereotypical (read, “nervous”) lying behaviors and maximizing stereotypical (“calm”) honest behaviors.
As a final pre-experiment preparatory step, Hocking categorizes all the behaviors from the survey into three classes: 1. gestures (Class I); 2. facial expressions (Class II); and, 3. vocal changes (Class III). Class I behaviors are the easiest for a liar to monitor and control. Class II behaviors are easy to control, but hard to monitor; after all, a liar can only guess what his face looks like. Class III behaviors are easy to monitor, but practically impossible to control.
So, according to Hocking’s hypothesis, liars, by default, will exercise more control over their bodily and ocular gestures than anything else. This contradicts Ekman, who argues that liars control Class II behaviors over and above all others.
Deceivers: Criminal justice students. First, they view a neutral video clip. Then, they view a positive video of a landscape scene, followed by the negative medical training videos used previously by Ekman, with scenes of limbs being cut off and burnt flesh being peeled away and debrided. In the interviews that follow, the students will selectively lie about facts pertaining to the first video and will lie completely when they see the medical videos.
Detector: An interviewer. Hocking recruits a local detective to interview the students. Hocking provides no details about what the detector says or does; all we know is that he asks questions and gets lied to.
Stakes: Job success. The students are recruited with a letter bearing the signature and seal of the Director of the School of Criminal Justice. Hocking not only tells the students that their job success is directly related to their performance, but also convinces them that their results will be reported to the faculty.
Salience: High, symmetrical. As expected, the detective is fully aware that lies are coming. The students understand, too, that he’s trying to catch them. The belief that their careers depend on success means they are also closely watching all sources of feedback, external and internal.
Leakage: Class II and Class III behaviors. After training observers to identify the survey behaviors, Hocking showed them three edited versions of the students’ interviews. The first version was a silent, face-only recording, which the observers used to identify facial expressions; the second was a silent, full-body recording, used to identify eye and body gestures; and the third was an audio-only recording, used for vocal changes. The observers counted as many behaviors as they could, and the results were compared against which interviews were truthful and which were deceptive. No judgments were rendered by the observers.
The results only partially support Hocking’s hypothesis. In Class I, several nervous behaviors decreased in frequency during deception (foot movements, head movements, and facial manipulators), but the decreases were not dramatic. Overall, nervous gestures decreased by about 10 percent during deception. Class II behaviors neither increased nor decreased, disproving part of Hocking’s hypothesis and again raising the question: Are liars exceptionally good at controlling their facial expressions, or are detectors just really bad at reading them? Class III’s results do suggest Hocking was right about one thing: the liars couldn’t control their vocal changes. They spoke faster, paused more, and interrupted themselves much more often. Finally, Hocking’s eye-contact hypothesis was wrong, too: the liars looked away more often and held eye-contact less. This is the last study where you’ll see that result; in all subsequent studies we’ve seen, eye-contact frequency and duration both increase during deception.
Later on, we’ll see that Hocking’s experimental design was significantly improved upon by subsequent researchers to fix issues like sample size, subjects’ anxiety fluctuations, deceiver motivation levels, and individual subject’s baselines. But Hocking’s theory makes a major contribution: the accuracy (or lack thereof) of the average detector is stereotype accuracy. Once you know how detectors (in general) expect you to act, acting in the opposite way will increase your odds of successful deception considerably.
For now, our advice is, Don’t appear nervous. Impractical, yes. Vague, of course. But as of right now, we can’t say more. We’re not done, though. As we progress, we will identify which behaviors receive the most attention and, of those, which are the most easily controlled. Once we’ve isolated these, we believe a little easy practice will make you appear much more honest when the appearance of honesty counts the most.
Hocking, John E. & Leathers, Dale G., (1980). Nonverbal Indicators of Deception: A New Theoretical Perspective. Communication Monographs, 29, 119-131.