Nov 21, 2021 7:00 AM

Body Language Pseudoscience Is Flourishing on YouTube

In celebrity interviews and homicide cases, video sleuths are searching for the truth—but what if the signals are all wrong?

Some say it was significant when she touched her nose. Others, when she showed the tip of her tongue. One person noted that a particular blink looked especially resigned. But in March 2021, the body language analysts of YouTube all agreed on one thing: Meghan Markle’s interview with Oprah Winfrey was content gold. They were right: Combined, the top five videos analyzing the Duchess of Sussex’s nonverbal communication during the tell-all have a total of 5.9 million views. The claims made within them range wildly, from “I feel like she might be having some sort of itchiness to her nose” to “This is lying … She does a different kind of lip press.”

Analyzing the body language of the British royal family is hardly new—it’s been a cottage industry in the UK tabloids for decades, and big red circles and blocky arrows signaled major clues in print long before they became staples of screens. But over the course of the past few years, the idea that a twitch or an itch reveals a person’s innermost secrets has gained traction on YouTube—where videos with millions of views analyze everything from celebrity apologies to serial killer interrogation tapes. More than 38 million people have watched a 37-minute video analyzing the nonverbal communication of family murderer Chris Watts; just under 2.8 million tuned in to see a body language breakdown of Amber Heard’s domestic abuse deposition against Johnny Depp.

The story sold in the vast majority of these videos is simple: X gesture can be directly translated into Y meaning, case closed. Viewers who seek out this content rarely seem to question it—the second most popular comment on the Heard analysis reads, “They deadass need to have body language readers in court rooms” (1,700 likes). Logan Portenier, the 27-year-old Seattle resident behind the video, says that down in his comment section he has seen a recent rise in “people starting to use nonverbal communication channels to help them make their own decisions.” That, he says, is “not something that I really wanted to have be my responsibility.” His channel, Observe, has 650,000 subscribers, 80 percent of whom are women, mostly aged 21 to 35.

Wrinkle your nose. Scratch your head. There’s a huge problem at the heart of this, which is that studies have repeatedly shown that body language cannot accurately be “read” like a book, particularly when trying to detect deception. While it’s one thing to use hindsight to claim that a convicted killer gave away his guilt, it’s another thing entirely to use unfounded science to attempt to prove a public figure is lying, or to wade into a recent tragedy to claim to “finally reveal” what happened to a murdered young woman. How do the body language gurus of YouTube clash with the academic community? How true are their claims, and what are the consequences of spreading them across the internet?

It’s fairly likely that you’ve crossed your arms and tensed your shoulders when feeling uncomfortable—you’ve probably balled up your fists when getting confrontational, and most of us have likely made snap judgments about other people’s body language from time to time. Vincent Denault, a postdoctoral fellow in psychology at McGill University who previously cofounded the Center for Studies in Nonverbal Communication Sciences at the Research Center of the Montreal Mental Health University Institute, says that researchers from a number of disciplines—“psychology, communication, biology, ethology”—legitimately study nonverbal behavior and the role it plays in our daily lives. But, he stresses, there is no innate, universal “language” of the body.

“When specific gestures are associated with specific meanings, and when this is implicitly or explicitly presented as scientific, then it begins to fall under the umbrella of pseudoscience,” Denault says. While scientists codify certain behaviors to better understand communication in various contexts, Denault says these systems cannot, in turn, be used to “decode.”

“The public thinks that nonverbal behavior is only good for one thing: detecting who is lying and who is telling the truth. It is not the case,” Denault says. One 2020 study from the University of Portsmouth tasked people with identifying smugglers in videotaped ferry crossings; while the observers claimed to look for signs of nervousness, only 39.2 percent accurately identified smugglers, “significantly below chance level.”

In his September 2020 video about Amber Heard, Portenier films himself reacting to the actress’ testimony, and he laughs, smirks, and rubs his face in disbelief before claiming that her snacking on food and seeming unenthused is “not a good indicator for Amber being the victim. It’s a very good indicator for her being an abuser.” In hindsight, Portenier stands by the statements made in the video but says he “probably spoke a little strongly” and would be “a little bit more mellow” if he was to make such a video now. Perhaps surprisingly, he agrees with Denault about the dangers of pseudoscientific analysis.

“On the internet, it’s so easy right now to just claim that you know things, and there’s nobody to really counteract it … It is something that concerns me for sure,” he says. Portenier’s knowledge of body language is largely self-taught, though he also took some psychology classes at university. He says he has been studying the topic for a decade, consuming the work of former FBI agent Joe Navarro (who has also made multiple videos with WIRED). Portenier also studies psychologist Paul Ekman’s work on microexpressions, which are facial expressions that last for a fraction of a second and are difficult to conceal. (By Ekman’s own admission, microexpressions that reveal concealed emotions aren’t all that common, and academics note he has not published data empirically proving that microexpressions can be used to detect lies.)

Bruce Durham, a 41-year-old from Newcastle, England, who made a video showing the “Exact Moment” Meghan Markle “Lies” to Oprah, is also self-taught. Durham says he has been working in performance coaching for more than 20 years. “I’ve had thousands of hours just sitting in front of people and letting them speak,” Durham says. “When you’ve spent that much time looking at people and you practice your observation skills, you can quickly develop trends and analysis, you sort of join the dots.” His channel, Believing Bruce, has just under 200,000 subscribers.

Both Portenier and Durham stress that they’re not leading experts in their field, and both say they try to communicate the limitations of what they do to the audience. “A lot of people look for who’s lying and who’s not, but you can’t ever really tell that. What you can do is, they fall into two categories of looking comfortable and looking uncomfortable,” Durham claims (his analysis of Markle is interspersed with clips of Pinocchio’s nose growing in Disney’s 1940 film). Durham says that identifying when someone looks uncomfortable provides a jumping-off point to ask further questions and is not a conclusion in itself, but he confesses that he makes his video thumbnails and titles more “evocative” in order to gain clicks. Still, he argues: “I always start or end my videos with, ‘You need to be fair and balanced.’ And I always say that multiple times as well.”

At the start of his Heard video, Portenier issues a disclaimer: “There’s a few things I want to be able to say, and I’ll say at the beginning of all of my videos that are on nonverbal communication. It is at best 70 percent accurate; there is definitely a talent to reading it; it is not a complete science as something like psychology is; it is not quite a pseudoscience; it is a mixture in-between.” In general, he comes across as more cautious than many body language analysts on YouTube. (It was he who argued that Markle could simply have had an itchy nose on the day of the Oprah interview.) But by his own admission, “audiences can and will 100 percent just blow right by any disclaimer that you give.”

This is likely because body language analysis on YouTube feeds confirmation bias: If you hate a celebrity, what could be better than having “scientific proof” that they are lying and conniving? Numerous videos analyze apologies from influencers such as James Charles, Shane Dawson, and Jeffree Star; after talk show host Ellen DeGeneres was accused of treating staff poorly in 2020, Portenier gained over a million views on a video promising to show her “INSINCERE apology faces.” Portenier says he was “surprised” that his video on Heard’s “CRINGE” deposition became his most popular ever, but there is a large appetite for content that discredits the actress; the same is true of Markle. Portenier, who has 1,684 regular donors on Patreon, takes suggestions from his audience on who he should analyze next.

“Unfounded and discredited claims about nonverbal behavior are popular. They have been around for thousands of years,” Denault says. “And even if I try, from time to time, to set the record straight, unfounded and discredited claims will continue to thrive because there are people who want to believe.”

What are the consequences of believing? Denault thinks that most who tout their body language expertise do so in good faith, believing what they say is correct. “But even if they are acting in good faith, good faith does not equate good practice,” he says. “The kind of claims made by body language ‘experts’ sometimes end up in the hands of people in positions of power, such as judges and jurors, where people’s liberties and even lives may be at stake.”

In August 2020, The Intercept published an investigation into “the junk science cops use to decide you’re lying”; that same year, Denault and 50 other researchers signed an article on the dangers of using body language pseudoscience within the justice system.

Aldert Vrij, a psychology professor at the University of Portsmouth who specializes in deception, explains that law enforcement agencies still use body language analysis because a number of organizations offer this training, and it “sells well.” “It sounds fascinating: The liar ‘leaks’ nonverbal cues without being aware of it … but the highly skilled investigator can spot these cues,” he says. Arguably, it plays into the vanity of investigators and offers a shortcut to success, which is particularly appealing when faced with an interviewee giving short answers. But, Vrij notes, training programs claim they work without providing empirical evidence.

Denault believes YouTube videos contribute to “an ecosystem of misinformation about nonverbal behavior”—it’s easy to see how a viewer could end up on a jury or in a HR department and draw conclusions that could damage another person’s life. Portenier argues that a better understanding of body language has many uses: It has helped him in past sales jobs and also helped him give the impression of confidence when nervous. He also believes body language training can help people understand each other’s emotional states.

“I’m trying hard to have an educational voice in it,” he says of the body language scene on YouTube. “I do have a personal goal on my channel to make sure that, while I’m talking about these things, I bring up the realistic side of things, because I fully agree with the skeptics when they’re like, ‘OK, just because politician number one scratched his nose for five seconds, I don’t think that that’s grounds to say that he’s considering any sort of mischievous things.'”

For Durham, body language has its uses for managers looking to make their teams more comfortable and to “provide the right sort of environment” for their employees. He describes his channel as a hobby; his main income comes from coaching and keynote speaking. He is concerned that creators who live off their YouTube earnings end up “chasing the views” by creating videos with foregone conclusions. “They’ll say stuff because it’s popular, because they believe that people want that answer,” he says. Both Portenier and Durham recognize there is a problem with body language analysis on YouTube, but neither seem to believe they’re part of it.

Shortly after the crowd surge at the Astroworld music festival on November 5 in Houston, Texas, which killed 10 people, the event’s headliner, Travis Scott, posted an Instagram Story responding to the tragedy. Many inside and outside the body language community were quick to note the way Scott repeatedly rubbed his forehead during the video. It is compelling to think that his gestures could be directly translated to reveal his emotional state—many people will quietly believe that they can tell whether his video was heartfelt, despite having no body language training at all. It’s unlikely that an appetite for easy answers will diminish; it’s equally unlikely that celebrities will stop scratching their noses.

“Extraordinary claims require extraordinary evidence,” says Denault. “If someone talks about body language as a way of ‘decoding’ the mental states of others, that is a pretty extraordinary claim about nonverbal behavior. This claim would require extraordinary evidence. And there is no such evidence.”