Human Language Has a Bias for Happiness

An Orthodox priest and parishioners laugh at while talking together on a sidewalk
Researchers tested the Pollyanna Hypothesis, which theorizes that because we enjoy communicating, our language will reflect it.

J Stimp / Flickr / CC BY 2.0

The most frequently used words in human language are more likely to have positive connotations than negative ones, revealing that we may have a bias toward happiness.

The research — the largest study ever of language and its emotional capacity — is the first to use "big data" to study the Pollyanna Hypothesis.

The hypothesis posits that because humans take pleasure in communicating with each other, their language will reflect this positivity.

Psychologists at the University of Illinois put forth the idea in 1969 when their research revealed a "universal human tendency" to use positive words more often than negative ones.

In 1978, another set of researchers from that university furthered this theory when their work revealed that people take longer to recognize unpleasant items and that we have a tendency to recall past events as more positive than they actually were.

In other words, we tend to see the glass as half full, and our language reflects that.

However, the Pollyanna Hypothesis was never investigated on a large scale until recently when the University of Vermont's Peter Sheridan Dodds led an international group of researchers to test it.

Linguists, mathematicians and other scientists combed through The New York Times, the Google Books Project, Google's Web Crawl, Twitter, song lyrics, and a library of movie and TV subtitles to create lists of the 10,000 most commonly used words in each of these 10 languages: English, Spanish, French, German, Korean, Chinese, Russian, Brazilian Portuguese, Indonesian and Egyptian Arabic.

Then, native speakers of these languages rated these words on a nine-point scale, where one is the most negative and nine is the most positive.

For example, in English, the word "laughter" received a score of 8.50 while "terrorist" got a score of 1.30.

The researchers ran the scores through a computer and found that each of the 10 languages uses positive words more often and in a wider range of forms than negative ones.

In terms of linguistic happiness, Spanish scored highest, followed by Brazilian Portuguese, English and Indonesian. Chinese was ranked as the least happy language in the study.

This research seems to confirm the Pollyanna Hypothesis; however, the authors point out that it only does so if our words truly convey our emotions.

It’s a big "if," but past research by the authors suggests that language could be an adequate measure of happiness.

In their previous study, they analyzed commonly used English words on Twitter and found that emotions on the social network rose and fell similarly to Gallup's polls on emotional well-being.