Emotions may be universal, but they aren’t easy to translate

In the English language, “love” is a pretty important word. It describes one of our most powerful feelings, so it’s no surprise that plenty of other languages have words for this emotion too, from the French amour to the Turkish word sevgi. At the very least these words all seem to capture the same emotion. But when it comes to language, feelings are a tricky business. A massive new study of thousands of languages around the world has revealed that the words we use to describe our feelings often don’t sync up.

“We might have a word like ‘love’ in English and look it up in a translation dictionary in any language and find a comparable word,” says Joshua Conrad Jackson, a doctoral student in psychology at the University of North Carolina at Chapel Hill and lead author of the new paper, which was published online December 19 in the journal Science. “But whether those two words actually refer to the same underlying concept, that’s an open question and that’s what we were testing.”

Jackson and his colleagues analyzed a database of 2,474 languages from 20 different language families, which are languages that share a common ancestor or parent tongue. To investigate whether words that represent an emotion have different shades of meaning around the globe, they used a phenomenon called colexification that happens when the same word is used to describe multiple ideas. In English, for example, the word “funny” can mean “humorous” or “odd.” Meanwhile, the Russian word ruka refers to both the hand and the arm. Often, words that become colexified describe concepts that the speaker views as related; in many languages, Jackson says, the same word can be used to describe leather, skin, and bark.

Across the languages Jackson and his colleagues studied, they identified around 66,000 cases of colexification. They used these instances to draw up maps of the concepts that people link to emotions and how they vary between languages.

It turned out that people understood emotion words very differently around the world. In Indo-European languages, anxiety was closely tied to anger. But among Austroasiatic languages (a group of languages spoken in areas of Southeast Asia and India), anxiety was more related to grief and regret. Anger, meanwhile, was connected to envy in Nakh-Daghestanian languages (which are spoken in the Caucasus Mountains), but was more closely linked to words like hate, bad, and proud in Austronesian languages (which can be found across Southeast Asia and the Pacific Islands and cover one-fifth of the world’s languages).

In fact, emotion words had about three times more variability in their meanings than words linked to different colors. How universal our concepts of color actually are remains an open question, though, and more research is needed. “What we can show is that however universal color is, emotion is significantly less universal,” Jackson says. “The way we interpret these feelings and the way we communicate them are potentially more sensitive to social processes.”

The differences Jackson and his team observed between languages weren’t random. The meanings ascribed to emotions shared more similarities amongst language families that were spoken in areas near to one another than those in lands separated by great distances.

And however differently we think about emotion across cultures, there were a few patterns that held true around the world. Emotions that feel positive—like love—rarely shared meanings with unpleasant feelings like anger (although there was one exception; while love was often connected to positive concepts such as happiness in Indo-European languages, in Austronesian languages love was more related to pity).

People also drew a distinction between highly-charged emotions like anger—which come with a racing heart and rising blood pressure—and less heady feelings such as contentment or sadness.

The research implies neither that emotion is universal nor that it is a social construct, Jackson says. “If anything our findings suggest that both are true—that there are some universal building blocks of emotion, but that the way we build on those depends on where we were raised, who we’re learning from, and the culture that we identify with,” he says. “When you’re speaking about 2,500 languages, you’re going to have a lot of nuance and a lot of richness.”

Hyejin Youn, an assistant professor of management and organizations at Northwestern University, who has studied similarities in the meanings of words used to describe nature across languages, praised the new work for tackling the complex morass that is emotion at an unprecedented scale. One strength of the investigation, Youn says, is that the researchers probed whether physiological reactions like a racing heart affected how emotions were defined. “If we look at not only just the name or the label of the concepts, but also the physiological dimension, then we find some kind of universal structure,” she says.

However, she adds, it’s important to keep in mind that emotions are very difficult to define and measure. In her view, a possible next step would be using brain imaging to identify whether concepts like anger or love prompt the same physiological signals in different languages.

The differences in how we conceptualize emotion across languages may impact our conversations with people who have another mother tongue, says Kristen Lindquist, an associate professor of psychology and neuroscience at the University of North Carolina at Chapel Hill and senior author of the new paper.

“We take for granted that other people understand our experiences…in cross-cultural diplomacy we assume that our expression of anger is going to be understood by another foreign diplomat, or our expression of joy or love,” Lindquist says. “It may actually be more difficult to understand emotions in people around the world than we have previously thought.”