Skip to main content

Verified by Psychology Today


How Universal Is Body Language?

New research suggests emotional body language may transcend culture

For all the importance we place on words, whether spoken or written, much of the communicating we do on a regular basis comes through body language.

According to pioneering research by Dr. Albert Mehrabian, only seven percent of the meaning we derive from human communication comes from the actual spoken words used. An additional 38 percent comes from tone of voice while a whopping 55 percent comes from body language alone.Though these findings remain controversial, there is no disputing that facial expressions, physical gestures, body posturing,and even our patterns of breathing can provide an amazing array of information for other people to interpret.

Researchers have long identified that certain kinds of body movements and facial expressions can convey information about the emotions we happen to be experiencing at the time. Even when physical movements are broken down into point-light displays that convey minimal information about how we move, research subjects are still able to interpret emotional states based solely on body language.

But are these emotional signals shaped by different cultures or are they universal to all humans? A new research article published in the journal Emotion attempts to answer this question through an ambitious cross-cultural study. Thalia Wheatley of Dartmouth College and a team of co-researchers travelled to Ratanakiri, Cambodia to study members of a remote Kreung hill tribe. One of the indigenous groups living in Cambodia's highlands, the Kreung are still largely isolated from the outside world except for occasional visitors.

With the assistance of representatives from the Cambodian government and local authorities who acted as translators, the researchers collected a series of videos featuring a Kreung male who was asked to display different emotions (anger, disgust, fear, happiness, sadness). The participant used was an experienced performer of traditional dance and music in the Kreung community and had considerable experience in performing before an audience.

With each emotional display, the participant was presented with different scenarios and was asked to perform each scenario as if he were the character described. Scenarios included: " I am very mad that I lost the stuff in my home" (anger), "I want to vomit. This soup is spoiled" (disgust), "I am so scared. Why are there so many tigers in this forest?" (fear), "I am very happy to be sharing these stories with other people, (happiness), and "I feel so miserable when my child has gone far away", (sadness).

These videos were later used in a study involving twenty-eight Dartmouth students or employees (thirteen were female and the average age was 21.9) who were asked to judge which emotions were being displayed. The videos were displayed in a random order with no sound or other verbal cues) and replayed on a continuous loop. In the first study, All participants were given a choice of five emotional labels to endorse and asked to view each video carefully before making a choice.

Results showed an eighty-five percent success rate which was far greater than what would be expected by chance alone. Of the emotions studied, participants were most accurate in rating fear followed by anger, disgust, and sadness. Happiness was the emotion least likely to be rated accurately though participants still scored better than chance.

In another study, a set of videos were prepared featuring an American woman displaying three positive emotions (happiness, love, pride) and three negative emotions (anger, fear, sadness) using body language alone. The effectiveness of these videos was tested using thirty-four participants recruited through Amazon's Mechanical Turk. Also, to minimize the visual cues that would be received, the videos were converted into point-light displays featuring fourteen light points corresponding to the the major joints in the body as well as the torso and head.

The videos were then presented to twenty-six Kreung individuals (eleven of whom were female). Since Kreung don't formally document age, there was no way to distinguish between adults and adolescents who participated. All participants were presented the videos and a translator helped explain the experiment and what they would be required to do. Instead of being given specific emotional labels such as in the first experiment, the Kreung participants were asked to describe the emotions being displayed in their own words.

Results showed that the Kreung participants tended to be quite accurate in guessing which emotions were being presented. The overall accuracy rate was sixty-two percent though their accuracy in detecting specific emotions such as anger and happiness was far higher (virtually everyone guessed anger correctly). They were also reasonably accurate in detecting sadness and, to a lesser extent, fear.

For emotions such as love and pride however, the Kreung participants did much worse and often misidentified these videos as examples of happiness. Overall, there was no significant difference between Kreung and American raters in detecting emotions such as anger, happiness, sadness, or fear though American participants did much better in detecting pride and love.

In a third study, sixteen Kreung participants were given only five words to choose from in identifying emotions (the Kreung words for: anger, disgust, fear, happiness, sadness). This was intended to make this study as similar to the first study as possible. As with the previous study, Kreung participants detected anger, disgust, and happiness at rates far above chance though their performance on sadness and fear was much lower.

So, what do these results suggest? While Kreung and American participants showed no significant difference in detecting emotions, there were still limitations with this kind of research considering differences in how the research was conducted. For example, Kreung were interviewed directly while American participants did their rating online and without any direct interaction with researchers.

Still, the results of these studies do seem to suggest that body movements can convey emotions such as anger, fear, sadness, and love even for individuals belonging to different cultures. By using remote tribal groups such as the Kreung who have yet to be assimilated as many other preliterate societies have been, Thalia Wheatley and her colleagues were able to show that emotional signals may well be universal since they reflect basic human needs and desires that all humans share.

As the world becomes more assimilated, studies such as this will likely become rarer with time. That may also mean that culture clashes will become more common, something we are already seeing firsthand in many countries. Learning more about how basic biology and social factors shape the way we communicate may well be vital in helping to understand ourselves better.


Parkinson, C., Walker, T. T., Memmi, S., & Wheatley, T. (2017). Emotions are understood from biological motion across remote cultures. Emotion, 17(3), 459-477.