On September 25, 2026, another meeting of the “Co-Author Reading” Readers’ Club, part of the “Science in Publications” series, was held with the support of Boiling Point. The meeting focused on the results of a series of studies examining how people recognize the emotional states of others.
The speaker, Olga A. Korolkova, Candidate of Science (Psychology), Associate Professor, Leading Researcher at the Institute of Experimental Psychology, Moscow State University of Psychology and Education, explained that most classical studies of emotions have relied on static images of faces, often displaying pronounced basic emotions. In natural or near-natural settings, however, several channels contribute simultaneously to person perception, including facial expressions, body movements, posture, gestures, voice, and the context of the situation. Moreover, the face and body are perceived not in isolation but as an integrated whole. This information is integrated automatically and at early stages of perception. Emotion recognition depends both on the emotion itself and on the observer’s individual characteristics.
Olga Korolkova described six series of studies on emotion perception involving more than 500 participants aged 22–60. The methodology included, among other procedures, viewing video clips depicting emotional expressions. Participants were presented with seven emotional states: joy, fear, disgust, excitement, boredom, disappointment, and calmness.
The researchers systematically varied the amount of visual information available. In some series, participants saw the whole person — the face, body, and situational context. In others, the face was blurred, only the face was presented, or a point-light figure conveying the structure and dynamics of body movement was used. In the most challenging condition, information about movement was also reduced: the stimuli were presented as a sequence of individual frames. After viewing each clip, participants answered three questions: What emotional state is the person in? What cues were used to identify this state? What might have caused this state?
Overall, the results supported the hypothesis that reducing the amount of available information makes emotional states more difficult to recognize. Emotions were identified most accurately when full information was available, as well as when the face was blurred but the dynamics of body movement remained visible.
Joy was found to be recognized primarily from static cues, such as the shape and configuration of the face and body. As long as at least a partial image of the person remained visible, participants generally identified this state quite accurately. A substantial decrease in accuracy was observed when only moving points were presented. Fear was recognized more accurately when both static and dynamic information was available, with body movements playing a particularly important role.
The perception of disgust and disappointment depended on both form and movement dynamics. Recognition accuracy progressively declined as the amount of visual information decreased. Boredom was also identified largely from body movements: the less movement information remained available, the more difficult it was for participants to recognize the state. Excitement was particularly strongly associated with kinematics — the characteristics and intensity of movement.
The study of calmness produced the most unexpected result. In some cases, participants recognized a calm state more accurately when they could see the body but not the face. One possible explanation is that when the face was visible, observers began looking for signs of other emotions and interpreted a calm facial expression as sadness, distress, or dissatisfaction.
The discussion of calmness generated particular interest among the audience. Participants shared the observation that a calm face is sometimes perceived by others as sad or upset. Olga Korolkova emphasized that in such situations it is important to consider not only facial expression but also the overall situation, the dynamics of behavior, and other available cues.
A separate part of the presentation was devoted to a comparative analysis of verbal descriptions. The results showed that, in both correct and incorrect recognition, participants may focus on the same cues — for example, eyebrow position, an open mouth, facial expression, or the nature of movement. However, these cues may be interpreted differently: one participant may perceive fear, while another sees surprise or disappointment.
Differences were particularly pronounced in participants’ assumptions about the causes of an emotional state. When the emotion was recognized correctly, descriptions of its possible causes were more likely to correspond to the situation presented. When recognition was incorrect, interpretations became more diverse and often included alternative scenarios of what might be happening. Interestingly, when visual information was minimal, participants began to reconstruct the context more actively. Even when only a moving point-light figure remained on the screen, their responses included descriptions of actions and possible situations.
Participants of different ages were found to describe the same emotional expressions differently. At higher levels of lexithymia — difficulties in recognizing and verbalizing emotions — responses tended to be shorter and less detailed.
Gender differences were identified in some clusters: on average, women provided more detailed descriptions, whereas men more often gave concise, sometimes one-word responses. Olga Korolkova noted, however, that the sample was predominantly female, and this finding therefore requires further investigation.
It was emphasized that the research is ongoing and that the findings have yet to be synthesized and subjected to further analysis.
Veronika G. Anikina, Candidate of Science (Psychology), Deputy Director for Academic Affairs at the Institute of Experimental Psychology, Moscow State University of Psychology and Education, also attended the meeting and, in particular, emphasized the importance of context in emotion perception.
During the discussion, participants asked questions about the role of context and cultural differences in emotional expression. Particular attention was given to the extent to which the findings could be generalized across different nationalities. The labor-intensive nature of processing the resulting textual data with the help of AI was also discussed.
Video recording of the meeting
A video recording has also been published on VK, Dzen, and Rutube.
2026 Calendar of Webinars and Meetings of the “Science in Publications” Project.