Human language prioritizes safety over happiness, study finds

A new study suggests human language may prioritize safety over happiness, with researchers finding that words related to safety are used more frequently than those expressing positive emotions. The findings challenge traditional models of linguistic meaning. Researchers at the University of Vermont analyzed billions of word usage examples across books, news, social media, and broadcasts. They found that a model based on power, danger, and order explained over 90% of word meaning variation, compared to about 72% for the traditional VAD model, which focuses on positivity, excitement, and dominance. The study, published in Science Advances, offers a new interpretation of the Pollyanna principle, suggesting the preference for positive words may stem from a focus on safe and predictable states rather than happiness. The researchers note this could reflect survival instincts, though further study is needed to confirm evolutionary origins, and the findings may improve AI's understanding of language.