(Phys.org) -- Scientists at ETH Zurich have studied the use of language, finding that words with a positive emotional content are more frequently used in written communication. This result supports the theory that social relations are enhanced by a positive bias in human communication. The study by David Garcia and his colleagues from the Chair of Systems Design is published in the first issue of the new SpringerOpen journal EPJ Data Science, and is freely available to the general public as an Open Access article.
Previous studies focused on word lengths and frequency. They demonstrated that frequency depends on the length of words used, as a result of the principle of least effort influencing the use of shorter words. In contrast, this study focused on how the emotions expressed in words relate to the word frequency and its information content. The authors focused on words used in written emotional expression in the three most popular European languages online: English, German and Spanish.
They exploited a dataset on human behavior on the Internet, which includes texts from blogs, chat rooms and forums, among other sources. After performing a quantitative analysis on this dataset, the authors found that positive words appeared more frequently than words associated with a negative emotion. This suggests that the emotional content affects the words' frequency, even though the overall emotional content of the studied words is neutral on average. These findings support existing theories that there is a positive bias in human expression to facilitate social interaction.
Going one step further, the authors also focused on words within their context and realised that positive words carried less information than negative ones. Therefore, because of the positive bias observed in human communication, positive words are more likely to be used whereas negative expressions could be reserved to transmit information about urgent threats and dangerous events.
Explore further: New poll reveals what Americans fear most
More information: Garcia D., Garas, A., Schweitzer F., Positive words carry less information than negative words, European Physical Journal Data Science (EPJ Data Science); www.springerlink.com/content/487p8w5635603351