Emotion has always seemed like one of the true human universals. All humans appear to cry when we are sad and smile when we are happy. We sweat from fear, our hearts beat in anger and swell with love. Hundreds of languages from around the world have words for concepts such as “sadness,” “happiness,” “fear,” “anger,” and “love.” Intuition tells us that these concepts are so common because they refer to basic human experiences that are hardwired into our DNA.
For many decades, behavioral scientists shared these intuitions, writing that certain emotions such as sadness, anger, and fear were universal to all humans. Psychologists focused on the cross-cultural recognition of facial expressions, whereas ethnographers and linguists examined whether different languages possessed translations of English words such as love.
But should we expect people around the world to conceptualize emotions, like “love” or “hate,” the same way, simply because we can point to a translation in a dictionary?
Should we expect people around the world to conceptualize emotions, like “love” or “hate,” the same way, simply because we can point to a translation in a dictionary?
Our research, published recently in Science, suggests that emotions might be less universal than previously believed. In an analysis of the meaning of emotion words across 2,474 languages, we found that words such as “anger” and “fear” varied widely in their meaning across cultures, far more than color concepts such as “red” and “blue”.
We were able to analyze such a large sample of languages by harnessing big data tools and focusing on a phenomenon known as “colexification.” Colexification refers to cases where a single word names two or more concepts. More often than not, when concepts get colexified by speakers of a language, it’s because those speakers see those concepts as having similar meaning. For example, the Russian word ruka is used to name the concepts of both “hand” and “arm,” and the English word “funny” colexifies the concepts of “odd” and “humorous.” Studying colexifications can thus illustrate how people understand concepts across different cultures. If languages around the world colexified “pity” with “regret,” this would suggest that these concepts have a similar meaning for all humans.
Drawing on a newly updated linguistic database, we computed large networks of colexifications across a global sample of languages, ranging from English to Turkish to Chinese, and even including small languages such as Yagua (from present-day Peru) and Sakha (from present-day Russia). Rather than universality, we found wide variation in colexification patterns. For Indo-European languages, such as English and other Western European languages, “love” was associated with words such as “want” and “like,” but for Austronesian languages in the Pacific Islands (e.g. Java or Fiji) and Tacanan languages of South America (e.g. Tacana, Yaminahua), it was associated with words such as “pity” and “grief.”
Emotions did not vary randomly in their meaning across cultures. Instead, languages in close geographic proximity appeared to share the most similar understanding of emotions. Since neighboring languages have more contact through historical patterns of migration, trade, and conquest, this suggests that the meaning of emotions can be shared between language groups.
There were also some universal features of emotion. Most notably, all language families differentiated emotions primarily based on whether they were pleasant or unpleasant (e.g. “joy” vs. “fear”), and whether they involved low or high levels of physiological arousal (e.g. “sadness” vs. “anger”). This suggests that feelings of positivity and arousal might be universal experiences, even if the meaning we endow these basic feelings is sensitive to our culture.
Our results suggest that while there appear to be some universal building blocks for emotion, the way that we construct specific emotions from these building blocks depends on where we live and who we learn from.
Our results suggest that while there appear to be some universal building blocks for emotion, like the feelings of positive and negativity just mentioned, the way that we construct specific emotions from these building blocks depends on where we live and who we learn from. “Surprise,” “fear,” and even “love” might mean very different things for people living in the United States, Turkey, and Fiji.
This study shows how language offers a window into the human mind. The average person speaks about 16,000 words a day, and many of these words are used to label and communicate our experiences. We now have the tools to systematically analyze this language, and compare the way that humans use language in different cultures.
These insights suggest that we should be cautious of translation dictionaries that equate words such as the English love and the Turkish sevgi.
These analyses can reveal startling truths about our inner lives, such as the universal and culturally specific ways that we experience and express emotion. One potentially fascinating area of analysis would be to test whether the different ways that people express emotion through language correlate with different physiological expressions of emotion. For example, do languages in the Pacific Islands show more aversive physiological reactions to scenarios involving love, where it’s associated with “pity” and “grief,” than people in the United States, where its associated with “like” and “want”?
At the very least, these insights suggest that we should be cautious of translation dictionaries that equate words such as the English love and the Turkish sevgi or beware of assumptions about shared experiences when interacting with individuals from other cultures. On a deeper level, we may also need to completely revise the way we think about human emotion. All humans may tremble sometimes and smile others, but this does not necessarily mean that smiling and trembling mean the same thing to all humans.