Tears, tears, tears: Current topics in the field of emotional crying

July 12th A1.02
Janis Zickfeld (University of Oslo)

Many scholars ever since Darwin have been fascinated with the human capacity to shed emotional tears. Reflecting recent interest in the intra- and interpersonal aspects of emotional crying, the present symposium provides an interdisciplinary overview of current topics focusing on theoretical and methodological advances. The symposium starts by applying automated analyses techniques such as machine learning to the context of spontaneous tears. The first talk elucidates how such techniques could be applied to dynamic stimuli and discusses the possibilities and limitations by examining videos of crying individuals. The following two presentations focus on the interpersonal effects of emotional crying from different angles. The second talk explores the importance of a specific context in light of emotional attributions to crying individuals and discusses how individuals perceive the emotions of criers. The third talk highlights the influence of gender roles on judgments of crying and presents systematic evidence from different cultures. Focusing on intrapersonal aspects of emotional tears the fourth talk poses the question whether crying can be beneficial by investigating its effect on pain perception and mood. Results from two studies found that shedding emotional tears did not reduce pain perception or alter general mood but might target specific aspects with regard to tension. Finally, the fifth talk proposes a model of positive tears based on qualitative and quantitative cross-cultural evidence. In conjunction, the five talks shed a fascinating light on different aspects of emotional crying and contribute to the field by providing different perspectives and methodologies.

Are there any beneficial effects of crying? The case of pain perception and mood

Asmir Gračanin, Michelle C. P. Hendriks & Ad J. J. M. Vingerhoets

Previous research on the effects of crying on different aspects of well-being yielded mixed results. For example, retrospective self-report studies showed that mood improvement following crying was reported by a significantly larger percentage of participants in comparison to those reporting mood deterioration. However, diary and quasi-experimental studies showed no effects or even significant detrimental effects of crying on mood. Finally, only one quasi-experimental study showed decreases in overall negative mood, but only after a somehow longer time period following crying. Nevertheless, the observed positive effects of crying on well-being found in some of the studies deserve additional attention. To determine whether crying influences pain perception, as a potential mechanism at the basis of its putative effects on well-being, and whether it affects tension-related aspect of mood rather than negative mood in general, we conducted two laboratory studies, in which we exposed participants to pain induction procedures (electric shock in Study 1 and cold pressor in Study 2) after they had watched a sad movie. In study 1 crying was elicited in 28 out of 57 participants and in study 2 it was elicited in 49 out of 69 participants. In addition to baseline and one immediate post-crying mood evaluation, in study 2 we repeated the pain induction procedure and mood measurements two more times. Crying failed to predict changes on all measures of pain perception. It also did not influence overall mood improvement over longer time period. However, crying specifically facilitated the improvement of the tension-related aspects of mood.

A model of positive tears

Janis Zickfeld & Beate Seibt

Although several scholars acknowledge the existence of tears of joy there is little systematic theoretical or empirical evidence on how positive tears are experienced, what elicits them, what actions or impulses they motivate in the crier, how they differ from tears of sadness or distress and whether there are different types. We investigated these issues and drafted a taxonomic model of positive tears. Drawing on more than one thousand reports of positive tears and including more than 2000 participants from 9 diverse countries and 7 languages the studies employed a strong mixture of quantitative and qualitative techniques. The final results showed evidence of the occurrence of positive tears and found four qualitatively different types and profiles that we termed achievement, beauty, affection and amusement tears. Achievement tears are often shed in contexts of extraordinary performance or when someone overcomes an obstacle and often include feelings of pride. Beauty tears occur commonly in situations of overwhelming elegancy including nature, music or visual arts and feature feelings of awe or experiencing chills. Affectionate tears are often experienced in situations including unexpected kindness or exceptional love such as wedding ceremonies or reunions and often feature feelings of warmth, increased communality and feeling touched or compassionate. Finally, amusement tears are shed when something especially funny occurs and include feelings of amusement or lightness and the inclination to laugh or giggle. We also investigated inter-individual differences with regard to these categories and discuss possibilities and implications of our taxonomy of positive tears.

Crying in context: The interaction between weeping and situational cues in attributions of emotion

Marc Baker

Context plays a crucial role in understanding facial expressions. For instance, changes in body language of the expresser or the surrounding environment as well as our own beliefs, experiences, and expectations modify what emotion we attribute to a facial expression. Using context to manipulate an audience’s perception of a targets facial expression has been called the ‘Kuleshov effect’. The effect is robust and has been consistently replicated. What is less well understood is how the Kuleshov effect interacts with other strong emotional signals such as tears. Tears are found to increase attributions of sadness regardless of the facial expression being displayed. This is called the ‘tearing effect’ and is also a hugely robust finding. In our experiment, judges (N= 150) rated the emotions of people crying whilst watching sad videos. Emotional context was manipulated by showing the judge a clip from a film associated with either joy, anger, fear or sadness prior to seeing the criers video. Preliminary results will be discussed showing how context affects how people perceive the emotions of criers and how tears may signal emotional intensity as opposed to strictly sadness.

Endorsement of Gender Roles Across Cultures and their Influence on Evaluations of Crying

Leah Sharman, Genevieve Dingle, Ad Vingerhoets, Harrison Manley, Marc Baker, Asmir Gračanin, Agneta Fischer, Kunalan Manokora, Igor Kardum & Eric Vanman

This study is an attempt to further understand how conformity to gender roles, rather than physical sex alone, and beliefs about crying may interact to affect evaluations of crying. This research uses a cross-cultural survey design across 5 countries (Australia, Croatia, Netherlands, Thailand, and the UK), hypothesising a mediational role of crying beliefs between gender roles and both crying intensity and mood following crying. Results will be analysed on a sample of over 750 people with gender balanced within countries. All participants, aged 18-40, were asked questions about their last crying experience related to a negative event (e.g., sadness, anger/frustration) and whether they received help from others, their general crying behaviours and beliefs, and their gender role prescription. Preliminary results from two countries suggest that more feminine gender endorsement is associated with stronger beliefs that crying is helpful, and greater likelihood of crying more intensely when crying because of a negative experience. Beliefs that crying is helpful for emotional recovery were not only related to more intense crying, but were also associated with mood improvement following crying. Early gender differences suggest that men reported crying with significantly less intensity and believing that crying was less helpful overall compared to women. However, women felt that crying was also the most unhelpful, with beliefs that crying was more unhelpful in both social and individual situations compared to men. Final results will be discussed including mediation models and the impact of social presence in crying evaluations.

Crying and the machine: What automatic analyses may (not) reveal about spontaneous emotional tears

Dennis Küster, Marc Baker & Tanja Schultz

Empirical crying research is often based on either vignettes or on still images featuring posed facial expressions. Unfortunately, however, we still know rather little about dynamic spontaneous crying. In addition, vignettes and static images are vulnerable to design and selection biases. Automatic analyses might shed new light on spontaneous crying. Using a combination of state of art commercial and open-source facial image analyses tools (AFFDEX, FACET, OpenFace), we re-examined dynamic spontaneous responses of a recent dataset of 35 female ‘fluent criers’ to a self-chosen sad movie. We extracted per-frame evidence for more than 20 Action Units, head pose, eye gaze, and blinking rates during and prior to crying and sad moments, as well as during a neutral baseline period. Preliminary results suggest the presence of facial responses typically associated with sadness, but also that of other expressions, including smiling both during and prior to tearful crying. Further changes in eye blink rates, head pose, and gaze raise the possibility that machine learning approaches might be able to detect both onset and attempted regulation of tears, without the need to examine complex dynamic changes in skin glossiness changes observed during tearful crying. We discuss how automatic facial image analyses may help to generate new hypotheses about human spontaneous crying, and how it could help inform systematic generation of suitable research materials for perception studies. Finally, using tears as an example, we aim for a critical discussion about the potential, and potential pitfalls of using automatic analyses to classify human emotional states.