Scientists have found that when reading a message with a smiling emoji, the brain begins to pick up its emotional tone even before the gaze falls directly on it. The results of an experiment with eye-movement recording were published in the Quarterly Journal of Experimental Psychology.
Emoji are small pictures that are inserted into digital messages to convey an emotion or intonation. In ordinary speech, a similar role is played by facial expression. Previous work has shown that people recognize emotions from emoji 73 milliseconds faster and more accurately than from photographs of real human faces, and joyful expressions are recognized fastest of all.
The new study, led by Demian Stoyanov, then a doctoral student at Macquarie University and the University of Groningen, involved 50 psychology students (36 women and 14 men aged 18 to 34, mean age 19.5 years). They read silently two-part sentences separated by a dash, for example: "You changed your hairstyle - something has changed." In the middle of each sentence stood an emoji: either positive or neutral. The speed and direction of gaze were recorded by an infrared camera at a rate of 1000 times per second.
The researchers were interested in the effect in which a word or symbol located in the field of so-called peripheral vision (the area slightly to the side of the point where the gaze is directed) influences how long a person lingers on the current word. This phenomenon is called the parafoveal effect. It helps to understand whether a person processes words one at a time or several neighboring ones at once.
When participants looked directly at the emoji, positive faces were read faster than neutral ones: on average 298 versus 362 milliseconds. However, on the word immediately before the emoji, the picture was the opposite: if a positive emoji fell into the field of view, readers lingered on that word slightly longer than before a neutral one. For example, the duration of the single fixation (a short stop of the gaze) averaged 196 milliseconds before a positive face and 192 milliseconds before a neutral one. The difference is small but statistically reliable - provided that the scientists adjusted for familiarity with the emoji, its emotional intensity, visual complexity and ambiguity. Without such adjustment, the effect was not significant.
On the word immediately after the emoji, no significant differences were found between the positive and neutral face. However, participants finished the whole sentence with a positive emoji faster: on average in 2523 milliseconds versus 2585 milliseconds with a neutral one.
The authors emphasize that this is a statistical association, not a proven cause. They also note limitations - the experiment used only emotionally neutral two-part sentences read in the laboratory. In addition, the comparison with versions without emoji was conducted on a different group of readers. The scientists suggest that an emoji between parts of a sentence may work as a punctuation mark, giving a pause before the next thought.
"You may begin to pick up the emotional tone of an emoji even before you look directly at it," Stoyanov explained. "In addition, even a face classified as neutral can give a businesslike message a slightly negative shade."