If you haven’t read the introduction first, check that out here.
Below I cover the building blocks of nonverbal communication. I give a comprehensive overview of the history of research into emotions, expressions, and mimicry.
It’s currently not complete, but you can see what I’ve written so far about the evidence for universal emotions/expressions.
Are Emotions and Expressions Hardwired? Is Their Meaning Universal?
If you never saw the show Lie to Me, it stars Tim Roth as Dr. Lightman.
Lightman is basically a professional human lie detector who helps investigators with his near magical feats of people-reading genius.
That character and show was actually inspired by psychologist Paul Ekman and his research on deception. But way before he became known as ‘the’ expert on lying, Ekman was researching facial expressions.
In one of his most famous studies Ekman traveled to the isolated Fore tribe in Papua New Guinea to help answer a controversial question about humanity:
Are the associations between some facial expressions and emotional states universal for all humans?
In other words, are we all hardwired to smile when happy, frown when sad, etc.?
Or did we just learn to make all the facial expressions that we do because of growing up in a specific culture?
Scientists, psychologists, and philosophers had been asking this questions for well over a hundred years.
Now to be Perfectly Clear: Some Expressions are Purely Cultural, and aren’t “Hardwired”
For instance, there are many gestures like the thumbs up and peace sign (making a V with the pointer and middle finger). These are called emblems and can be used widespread by a culture, but they don’t have a universal meaning.
In fact, some cultures actually find these hand signs offensive.
Nodding the head to say yes or shaking it to say no are interesting gestures. Many diverse cultures use them, but they’re not used that way universally.
It’s been argued that nodding is a shortened form of bowing. And that bowing, or physically lowering yourself, or even prostrating yourself on the floor (as an extreme), is a universally submissive or deferential posture (more on that later).
Others have claimed that turning the head side to side is practiced by the baby refusing the breast. They argue that this early refusal behavior is the origin of associating refusal with side to side head movements.
These claims behind the origin of the head nod and head shake are interesting, but they don’t lend themselves well to scientific study.
We’re going to focus on just discussing the expressions and postures which we have scientific evidence of being universally wired together with our emotional/physiological states.
Just remember not every expression is universal—there are many that we acquire from our culture, or even adopt from an individual.
Additionally, there is controversy over the existence of universal emotional categories, I’ll touch on that in just a bit though.
Early Evidence of Universal Facial Expressions
Darwin had written about the universality of some facial expressions in his book The Expression of Emotions in Man and Animals way back in 1872.
In addition to his own observations, he corresponded with friends traveling throughout the world and asked them a series of very specific questions about the natives and the physical expressions and faces they made.
In the sixties psychologist Silvan Tomkins expanded on Darwin’s ideas, and along with Ekman and Carroll Izard they created a set of photos depicting what they theorized to be the prototypical facial expressions for:
After testing out their photos in the US, Ekman and Izard separately traveled the world showing photos of the universal expressions theorized by Darwin and Tomkins.
Their experiment was simple, they just wanted to see if people would match the facial expressions to the same emotional descriptions that we do.
Or would all the natives somewhere have a culturally different use of these facial expressions?
The two visited over 20 countries all over the world including: Japan, Chile, Brazil, Ethiopia, China, Turkey, Estonia, Scotland, and Indonesia.
Their experience confirmed that different cultures had different social norms for displaying expressions (something already explored by anthropologists).
These differences in acceptable emotional expressions are called display rules.
Acceptable expressions to make could vary based on whether the interaction is public or private, the genders of the involved, the situation, etc.
For some cultures there are strong taboos against expressing certain emotions like anger. (The result is emotional suppression, which we’ll talk more about later.)
Amazingly despite these differences, across various cultures people seemed to match the photos of the facial expressions to the same emotions that we do with very high consistency.
However, the possibility remained that mass media had somehow influenced the people in these studies. They might have learned the expressions from images they saw printed or televised.
This is what brought Ekman to Papua New Guinea, to meet the preliterate Fore tribe who lived isolated from the modern world. They had no media and few encounters with outsiders.
So, if the emotional expressions that Darwin theorized about decades earlier were truly universal than even this isolated group of tribesmen should smile when happy.
And, again when shown the photos these isolated tribesmen made the same connections between expressions and emotional context that we did.
Since then studies like these have been repeated over a hundred times.
They’ve even added additional prototypical expressions to the original six.
It’s also been observed that even people born blind (and thus unable to ever witness facial expressions), will make these prototypical facial expressions to some extent as well, although their capacity in this regard is diminished.
Now, you might think the universality of facial expressions may seem a little obvious.
Especially today where we’ve all grown accustomed to seeing pictures and video of people emoting from all over the world.
It’s important to note that besides display rules, there are other differences in emotional expressions across cultures. But there still seems to be evidence of some universality to the basic expressions.
But have you ever wondered why?
Why do we make faces?
The truth about the relationship between our emotions and expressions gets weirder…
The Magic of Making Faces
What if instead of our emotions causing our expressions, our expressions actually caused our emotions?
In the 1880’s, William James (philosopher, psychologist, and physician) and Carl Lange (physician, psychologist, and neurologist) both independently proposed similar theories that our bodily reactions were reflexes that inform our emotional reactions.
James’s theory was that the sensations of the bodily state is the emotion itself. “Without the bodily states following on the perception, the latter would be purely cognitive in form, pale, colorless, destitute of emotional warmth.”
In what’s come to be known as the James-Lange Theory the sensations of changes to our physiology forms the basis of our emotional experiences.
Or in other words, reflex bodily reactions relay information back to the brain which creates the emotional state. This makes changes to heart rate, breathing pattern, muscular tension, and visceral changes (like “gut feelings”) the essential part of an emotion.
In the century since their theories, yet more theories have come out (surprise!), as well as many studies focusing on whether facial expressions affect our emotions.
As you can imagine, it’s hard to investigate whether or not a facial expression affects an emotion.
If the subjects realize that the experiment is studying the relationship between facial expressions and emotions it could influence the results.
Not to mention how do you measure what a person is feeling?
Most people aren’t great at discussing their feelings, and a trend you’ll notice across many studies is that people will often not have much of a difference in their mood when it comes to self-report, however,
Take depression as an example, we commonly think of depression as a state of extreme chronic sadness.
However, many depressed people wouldn’t describe themselves as sad.
Instead, it’s very often the other signs of depression that get noticed.
They might not be taking pleasure in things that they used to enjoy. They might feel unmotivated, maybe to the point that it’s hard to get out of bed and go to work or to school. They may have a pessimistic attitude.
These signs along with others persisting over a period of time might indicate that someone is depressed.
That being said, if you were to go and tell someone who seemed like they were depressed, “Hey you seem depressed,” depending on who it is there is a good chance that they will disagree with you.
They might not understand the symptoms of depression, or think that it applies to them. So there might be differences in meaning. On top of that psychological problems are often misunderstood, and even stigmatized in some communities.
As a result, some one might not want to see themselves as being depressed, because depression is not something they want to identify with. And if they feel that way they definitely don’t want others to think that of them that way as well.
So in many of these earlier studies researchers would essentially trick subjects into unknowingly smile or frown while reading a comic (or similar entertainment), and then rate how entertained they were.
Asking subjects to rate the comic is a simple solution to get a measure of their mood, instead of asking them outright how they feel (which might get a distorted answer).
Results show that a happier person will enjoy the experience of the comic more while a less happy person will be more negative.
Which just leaves us with the last problem: How do you make people smile, without them becoming aware of it?
A number of studies have been designed using a cover story to trick subjects into believing that something else was being studied.
Early on in 1974, JD Laird told subjects that he was measuring muscle activity in the face. He placed surface electrodes on subject’s faces and had them contract their faces based on which set of electrodes he touched.
This would cause them to engage the muscles used for smiling or frowning.
It wasn’t a perfect cover story. In one of his studies 16% and in another 19% of subjects figured out that the faces they were making were emotional ones and their data wasn’t considered.
The analysis of the remaining subjects showed that the facial expressions seemed to influence people’s responses, as well as self-reported emotional state.
In the years after, other studies using slightly different designs got similar results showing that facial expressions were having an impact.
In 1988, Fritz Strack used one of the cleverest cover stories.
Strack told subjects that they were doing research on motor coordination and on people’s abilities to perform various tasks without the use of their dominant hand (such as in the case of the disabled).
They were led to believe that they were just a pilot study which they were using to test some of the tasks that might be included in a larger more complex study they would be doing later.
He’d have subjects hold the pen either: just in the lips (in a way that would suppress the smiling muscles), just in the teeth (which would engage smiling muscles), or hold it in their non-dominant hand.
Then he’d also have them do a variety of other tasks involving using the pen in a coordinated fashion and making decisions. This primarily served to further convince them of the cover story and make the true study harder for subjects to figure out.
This cover story and experiment was much more convincing. In a small pretest they found that no one made the connection that mood and facial expressions were what was actually being studied.
There is a pretty strong consensus across these studies and about a dozen others like them, that subjects who were simulating a smile would enjoy the comic more, and if they were suppressing a smile or simulating a frown they’d tend to enjoy things less.
So if there is such a strong connection between physical expression and emotion, then you might expect that people who have mood disorders like depression have more activity in their muscles related to negative expressions.
Mood Disorders and Facial Expressions
Now of course, it’s important to note that our thoughts influence muscle activity in the face. You know that whole emotions causing expressions thing (which I’m assuming we all agree is a given).
In particular, it’s been found that negative thoughts tend to increase activity in the corrugator muscles (used in frowning the brow) and decrease activity in the zygomaticus muscles (used in smiling).
It was commonly believed that depressed people are less expressive, particularly that they are less likely to smile.
There have been many studies that have used EMG (electromyography, which measures muscle activity) on depressed people during positive versus negative thoughts or imagery.
Studies have conflicted as to whether or not facial activity of depressed people differ from nondepressed people during this task (Oliveau 1979, Greden 1986).
On the other hand, several studies have shown that when a depressed patient improves their corrugator activity tends to decrease to roughly the same extent as well (when compared to activity measured when they were more depressed) (Schwartz 1978, Carney 1981, Greden 1984).
So there is some evidence that facial expressions are impacted by mood disorders. Of course many people are likely to be putting on a face.
But what’s really interesting is how influencing the ability to make facial expressions seems to affect our psychology. Which is further evidence of the facial feedback hypothesis.
Evidence That Because Botox Alters Our Facial Expressions it Also Alters Our Emotions
Botox partially denervates facial muscles (like paralysis). This decreases muscle activity which, “weakens” the ability to make certain facial expressions.
So if weakening the expression weakens the emotion, then we would expect that Botox treatments would weaken the emotions associated with treated muscles.
Thanks to Botox we have quite a bit of research looking into the psychological effects that turning off a specific muscle has.
In one study, scientists collected before and after photos of people who had Botox on the corrugator muscles. These muscles are used in frowning, and associated with negative emotions like sadness, fear, and anger.
The researchers then taught a group of “viewers” how to discriminate between different intensities of the facial expressions. They used a reference set of pictures for varying levels of the emotions for: anger, fear, sadness, and happiness. Of the 40 viewers, 39 were able to successfully discriminate between different intensities of the emotions.
These viewers were then given the set of before and after Botox photos in random order. However, they were not told that people in the pictures had the Botox procedure or that they were before and after photos.
Despite not knowing any of that, the viewers rated the people in the after Botox photos as seeming 49% less afraid, 40% less angry, 10% less sad, and 71% happier than their before photos.
So having Botox tune down this frowning muscle involved in the negative emotions weakened people’s expression of those emotions. This made them appear happier.
Which means Botox at least makes us appear emotionally different to others. But are appearances only skin deep?
In another study, Michael Lewis and Patrick Bowler (2009) wanted to see if using Botox to denervate the corrugator muscles, which is used for expressing negative emotions, would make people feel those emotions less.
They found that people who had reduced frowning ability because of Botox scored significantly lower on a psychological questionnaire measuring anxiety and depression, when compared to people who had used other cosmetic treatments (control).
In 2006 Finzi and Wasserman found that 8 out of 10 patients had their depression go into remission after one treatment of Botox. This early study didn’t use a placebo control or blinding.
In 2012 Wollmer took 30 depressed patients who had both noticeable frown lines and hadn’t improved significantly on medication. Using a double-blind placebo-controlled study, patients randomly received either Botox or a saline injection (placebo) to the corrugator muscles.
After a single injection they followed up with these patients for 16 weeks, and the Botox group had a significant improvement in mood compared to the placebo group over several different rating scales for depression.
In another double blind placebo controlled study of corrugator muscle Botox injection done by Finzi and Rosenthal (2014) they analyzed data from 74 depressed subjects, and found similar levels of improvement to Wollmer’s study in the Botox group compared to the control.
Maybe one of the most compelling of the Botox studies was a brain imaging study by Hennenlotter in 2008.
Earlier studies had shown that having subjects imitate facial expressions affected brain activity in limbic regions of the brain which are associated with emotions and behavior.
In Hennenlotter’s study they had half of their 38 subjects undergo fMRI and either just observe or to imitate pictures of people making angry or sad facial expressions. The other half went through this process 2 weeks after a Botox injection to the corrugator muscles, which impaired their frowning ability.
What they found was that when people with the Botox treatment imitated an angry expression they had less activity in their left amygdala (which is associated with anger) compared to the control group who hadn’t received Botox.
Meanwhile the activity in motor and premotor areas of the brain (involved in the making of expressions) didn’t seem to be impacted by Botox. This means that the amygdala activity didn’t decrease because of a change in motor function, it decreased because of lack of feedback from the facial muscles themselves!
The amygdala is also linked with parts of the brain stem involved in controlling the autonomic nervous system (which we’ll talk more about later), and the Botox treatment seemed to affect this link.
Of course, you have to remember two important points:
1. Botox itself isn’t an anti-depression or anti-anxiety treatment. Botox just interferes with facial muscle contraction. It’s because facial expressions influence our psychology that Botox treatments to facial muscles has a psychological impact.
2. Treating facial muscles with Botox, isn’t necessarily going to have a positive psychological impact (And I don’t just mean that you might be unhappy with your new look). The psychotherapeutic benefits in these studies came from Botox muting muscles involved in negative emotions (specifically the corrugator muscles).
Weakening the ability to frown is what makes the people in these Botox studies less sad.
On the other hand, if Botox is used on muscles involved in positive expressions it seems to actually make you less happy.
In 2013, Michael Lewis presented a study with results supporting this possibility. Instead of studying Botox injection to muscles involved in negative emotions, he looked at Botox on muscles involved in positive ones.
Crow’s feet are wrinkles formed by the muscles around the eyes used in smiling. As you might expect by now… getting Botox to these muscles weakened subjects’ ability to use these muscles when smiling, and this seemed to make people less happy.
So there seems to be psychotherapeutic benefits of Botox, there is emerging evidence that Botox may be helpful in treating people with anxiety, depression, and potentially OCD and other mood disorders… just remember that it depends on where you get it.
On the other hand, there is another negative psychological side effect of Botox, which I’ll get to in a little bit…
So far we’ve seen evidence supporting the facial feedback theory; that facial expressions affect people’s emotional state.
-suppressing an expression weakens the emotion
-exaggerating an expression strengthens the emotion.
Interestingly, this also seems to be true with pain.
A Shocking Discovery About the Expression of Pain
John Lanzetta shocked his subjects with electricity and would have them either exaggerate their pain face, suppress the expression of pain, or react naturally.
He would vary the intensity of the shocks, but wouldn’t tell subjects how powerful each shock would be. He recorded both how they perceived the intensity of the shock and their skin conductance (a measure of their physiological response).
Compared to reacting naturally, subjects when exaggerating the expression reported a more powerful shock. And skin conductance responses were higher, too.
Conversely, suppressing the pain expression also seemed to reduce their experience of pain and lower skin conductance.
So your facial expressions can actually modify your experience of pain.
Facial Expressions are Actually Shaping Your Physiology
Although these experiments have shown that making an expression seems to affect mood, the effect isn’t huge, but it’s pretty reliable across a variety of studies.
Now I understand that a major weakness of many of these studies, is that they tend to rely on someone reporting their experience or taking a psychological survey.
However, there have also been studies on the effects of facial expressions on physiological measures like autonomous nervous system (ANS) activity.
Studies by Ekman, Robert Levenson, Wallace Friesen, and others have shown that the body functions slightly differently in different emotional states.
In one study they showed that either “reliving” an emotionally charged event by trying to vividly recall it for 30 seconds, or making one of the emotion-prototypic facial expression for 10 seconds was enough to affect the body in a variety of ways.
They found that there were some distinct changes for different emotions, and not just distinction between positive and negative emotion, but also some differences between negative emotions.
For example, making the anger and fear expressions would increase heart rate from baseline by about 8 beats per minute, while happiness would only increase it by 2, and disgust would slightly decrease heart rate. Anger increased finger temperature the most, while fear, surprise, and disgust decreased it slightly.
What might be the most interesting thing, is that voluntarily making the facial expression caused more significant changes than did mentally “reliving” an emotionally charged event.
However, classical conditioning (like Pavlov’s dog trained to salivate when a bell rang) could also affect ANS function.
So in order to rule out the possibility that these effects are conditioned instead of hardwired Levenson, Ekman, Friesen, and Heider travelled to West Sumatra to study the Minangkabau.
The Minangkabau are a very different culture from ours, they are matrilineal Muslim culture with cultural taboos against displaying negative emotions (particularly anger).
Additionally, “In comparison with Americans, for whom the internal experience of emotion is very important, Minangkabau more commonly emphasize the external aspects of emotion, focusing primarily on the implications of emotion for interpersonal interactions and relationship.”
Meaning they are more concerned with how their expressions are influencing others.
Despite cultural differences with regards to emotions, they found that their distinct patterns of ANS activity specific to different emotions were fairly consistent with the patterns they found in the west.
So interestingly enough, our expressions also seem to influence how our bodies function.
This would also seems to be true of bodily expressions (and even vocal expression, but we’re just focusing on body language in this article).
Most of the early research was face centric.
The Universal Bodily Expressions and Their Influence on our Emotions
There have been several studies that have explored the effect of posture on our psychology. Particularly how postures influence physiology and behavior.
Riskind and Gotay in 1982 performed a series of 4 studies, investigating posture’s influence on emotion as well as motivation.
In their first study, they tested whether subjects would give up sooner on a frustrating task if their posture was slumped.
Like many of these studies, researchers misled subjects about what was being studied in order not to taint results.
First they told subjects (who were university students), that they would be participating in two unrelated studies.
In the fake first study, the first experimenter claimed that they would be studying muscle activity and galvanic skin response. For about 8 minutes, the subjects were hooked up to equipment with electrodes. Half of the subjects were posed into a slumped posture, and the other half were posed into an upright one.
During this time, the experimenter gave subjects a questionnaire to assess their mood, which they explained was necessary to control for mood fluctuations which might affect the measurements being taken.
Then subjects were sent to another room, where they believed the second unrelated study would be taking place. Here they met with another experimenter who did not know whether the subject was slumped or upright in the first half of the experiment (“blinding” the experimenter like this was a precaution to prevent him from having any influence on the subjects).
This second experimenter gave subjects geometric puzzles, and told them that it was to measure their spatial reasoning.
In reality, 2 of the 4 puzzles were actually impossible to solve (Tricks within tricks in this study).
What the researchers were really trying to see was whether posture would affect motivation during a frustrating activity.
After screening the subjects to remove any who might have figured out what was actually being studied, the results confirmed what the researchers had suspected: being recently slumped caused subjects to give up more quickly on the puzzles than did being in an upright posture.
Even though the difference was significant, the student’s moods in the self-report questionnaire weren’t very different. As we went over earlier, when asked on the spot people tend not to be aware of much of a difference in their moods during these studies. Despite the fact that there can be a very obvious difference in behavior when comparing groups.
In their second study of four, the researchers replicated the first, but with a different group of experimenters, and a different group of students from a different part of the country.
They also slightly modified the procedure by having experimenters describe to the subjects how they should position themselves (versus the experimenter physically posing them like in the first study).
The results of this second replication study were very similar to the first, with the previously slumped students giving up sooner than the upright students to a similar extent.
In their third study they looked at how a person’s posture affects our perception of them.
They showed study participants (students again) pictures of people in slumped or upright postures. They’d see these pictures and read stories about problems common to college students (like stress over academic achievement, feeling socially isolated, etc). The faces in these pictures were obscured so that the influence of posture could be examined in isolation without the influence of facial expressions.
They were told to imagine themselves as being the person in the picture while going through the problem in the story. Then they were given four possible reactions, and told to pick the reaction that the person pictured would most likely be having in that situation. Two out of the four options were how a non-depressed person might react in the scenario. While the other two options reflected reactions of helplessness and sadness characteristic of depressed people.
Subjects had to do this for 6 scenarios both for pictures of characters that were both slumped as well as upright.
The results were that subjects perceived the characters who were slumped as reacting more helplessly depressed in situations, than characters who were pictured as upright.
Study 4 subjects were again posed by experimenters and were lead to believe that they were really participating in two unrelated studies (like in the first two studies).
This time instead of a slumped posture, they posed half the subjects into either a:
-tense posture with hunched shoulders shrugged up and legs held close together. This posture has been associated with anxiety and feeling threatened.
-relaxed posture where the other half of subjects were leaning back in the chair and kept their legs open.
In addition, for the second experiment half of subjects were told that they would be taking a test that had nothing to do with general intelligence (your IQ could be high and you could do worse on the test, then someone with a lower IQ).
And the other half were told that the test they were taking was going to be adopted by their university and that it was an accurate measure of IQ.
In reality, both groups were given the same test either way. But the students who were told that it was an IQ test now had something to lose.
The researchers called this a higher threat situation.
With this experimental design there were four groups:
Then they were surveyed in two ways.
First, their moods were rated by picking 1-10 on a scale for adjectives related to stress (eg apprehensive, calm, nervous, etc.).
Second, they also had subjects rate their physical sensations by picking 1-10 for physical stress descriptions (eg, “There is a knotted feeling in my stomach.”).
Subjects who were hunched (tense) were the most stressed out for both measures of stress, compared to relaxed postured subjects. As you can imagine, the students who thought it was an IQ test (tense and high threat) were more stressed out than those who were hunched, but didn’t think the test was related to IQ (tense and low threat).
What’s interesting is that being hunched made students more stressed out regardless of whether they thought it was an IQ test. Which means that, in this study at least, posture had more of an impact on stress levels than did the situation.