The research, conducted by Ramona Siebert, Nick Taubert, Martin A. Giese, and Peter Thier at the University of Tübingen, trained two rhesus macaques to categorize videos of four distinct facial expression types: neutral faces, lip-smacking, silent bared-teeth displays, and open-mouth threat displays. These categories span the social spectrum of macaque communication, from affiliative gestures that signal benign intent to aggressive displays that warn of possible attack. The animals watched dynamic, naturalistic video clips rather than static photographs, an important design choice because real facial expressions unfold in time and involve coordinated movements of many facial structures at once.

Once the monkeys had learned the task, the researchers tested whether the animals could generalize their classifications to videos they had never seen before, featuring novel individuals. The macaques succeeded, demonstrating that they had not merely memorized specific clips or individual faces but had extracted something general about each expression category. Performance, however, was not perfect, and the pattern of errors proved to be one of the most informative aspects of the study. Misclassifications did not simply track how visually similar two expression categories looked, which suggested that something more interesting than raw visual resemblance was shaping the monkeys’ judgments.

To understand what drove those errors, the team developed a novel automated approach that combined deep learning-based motion tracking with the Macaque Facial Action Coding System, a standardized framework for decomposing macaque faces into discrete action units such as lip movements, brow raises, and ear positions. This pipeline allowed the researchers to objectively quantify the facial movements in every video clip, replacing subjective human coding with measurable, reproducible descriptors. The analysis revealed that the monkeys’ category boundaries could not be reduced to visual similarity metrics alone, a result that challenges simple feature-matching accounts of expression perception.

Some expressions, it turned out, were far easier for the monkeys than others. Open-mouth threat displays were readily categorized correctly and reliably distinguished from lip-smacking, and they also produced the strongest arousal responses, as measured by pupil dilation. Pupillometry has become a valuable tool in animal cognition research because pupil size is modulated by autonomic arousal, providing a physiological readout of how salient or emotionally charged a stimulus is. The combination of accurate categorization and heightened arousal suggests that threat displays carry particular functional significance for macaques, commanding both perceptual precision and an emotional response.

Silent bared-teeth displays told a very different story. Categorization of these expressions varied substantially across stimulus identities, meaning that the same expression type produced different judgments depending on who was making it. The researchers found that this variability was influenced by the signaler’s body weight, gaze direction, and coordinated movements of the eyebrows and ears. In macaque societies, the bared-teeth display can serve multiple functions, from signaling submission to facilitating affiliation, and the new findings hint that monkeys may interpret such displays in a context-dependent way, factoring in who is signaling and what the rest of the face and body are doing.

To probe these mechanisms further, the team turned to synthetic stimuli. Using morphed avatar faces, they generated expressions whose component intensities could be precisely controlled. The monkeys categorized these avatar expressions according to the intensity of the expression components, demonstrating graded perception rather than all-or-nothing category boundaries. This is a significant result because it shows that macaque expression perception is not a rigid sorting process but a graded one, in which the strength of facial actions maps continuously onto perceptual judgments, much as human observers perceive emotional intensity along a continuum.

The avatar experiments also allowed the researchers to dissect which ingredients of a facial expression are actually necessary for recognition. Manipulations revealed that categorization was robust against the lack of coherent motion and against the lack of facial realism, so long as a basic level of texture was present. In other words, the monkeys did not need a fully lifelike, smoothly moving face to classify an expression; a sufficiently textured but simplified avatar sufficed. This finding has practical implications for laboratory research, since it validates the use of synthetic faces in primate studies and suggests that the perceptual system extracts expression-relevant structure even from impoverished stimuli.

One boundary of the monkeys’ abilities emerged clearly: human facial expressions elicited no systematic categorizations and no differential arousal. Despite the superficial resemblance between some human and macaque expressions, the macaques did not treat human faces as meaningful within their classification framework. This cross-species result underscores that expression perception in these animals is tuned to conspecific signals, shaped by the specific morphology, dynamics, and social functions of macaque facial behavior rather than by a generic facial expression detector.

Taken together, the study paints a picture of facial expression perception in rhesus monkeys as functionally meaningful, context-dependent, and holistic. Expressions are not fixed categories defined by mouth movements alone; they are social signals whose interpretation is shaped by all components of facial behavior and by characteristics of the signaler, including body weight and gaze. The methodological contribution may prove as influential as the behavioral one: by fusing deep learning motion tracking with an established action coding system, the researchers have created a toolset for quantifying facial behavior objectively, opening the door to comparable studies across laboratories and species. For a field that has long relied on human judgment to define what a monkey’s face means, the message is clear that the monkeys themselves may be reading far more than we assumed.

Subject of Research: Dynamic facial expression perception and categorization in rhesus macaques using naturalistic videos and synthetic avatars

Article Title: Behavioral characterization of dynamic facial expression perception in rhesus monkeys (Macaca mulatta) using naturalistic and synthetic stimuli

Article References: Siebert, R., Taubert, N., Giese, M. A., & Thier, P. (2026). Behavioral characterization of dynamic facial expression perception in rhesus monkeys (Macaca mulatta) using naturalistic and synthetic stimuli. Animal Cognition. https://doi.org/10.1007/s10071-026-02109-6

Image Credits: AI Generated

DOI: 10.1007/s10071-026-02109-6

Keywords: rhesus macaque, facial expression, animal cognition, social signaling, pupillometry, deep learning, facial action coding system, avatars, graded perception, threat display, lip-smacking, Tübingen

The study, led by Claudia Chiarolanza of Sapienza University of Rome together with Stefano Isolani of Guglielmo Marconi University and Danilo Garcia of the University of Stavanger, drew its inspiration from Greek mythology. Nemesis, the goddess of retribution, embodied the outrage that follows perceived violations—a fitting metaphor, the authors argue, for the emotional and social forces that swirl around revenge in romantic relationships. Betrayal within a couple is not merely a private wound; it is a violation of a relational contract, and the response it elicits carries both private and public dimensions. The researchers set out to test whether the perceived legitimacy of a vengeful act depends jointly on the sex of the avenger and the visibility of the act itself.

To do so, the team designed a controlled 2 × 2 factorial experiment. Each of the 908 participants, all Italian adults recruited through social networks and community circles, was randomly assigned to read one of four scenarios depicting a marital betrayal and a subsequent act of revenge. The scenarios varied the sex of the avenger—male or female—and the context of the retaliation—public or private. In the public conditions, the vengeful act unfolded before an audience: one male avenger posted about his wife’s affair in a town’s Facebook group, while one female avenger publicly humiliated her husband and his lover at a company dinner. In the private conditions, the retaliation was directed at the partner alone, such as a message to a small WhatsApp group or throwing a spouse’s belongings out the window. Participants rated how plausible they found each scenario and how likely they would be to act as the avenger did, on five-point scales.

The design was carefully controlled. An a priori power analysis had indicated that at least 299 participants would be needed to detect a medium effect with 90 percent power; the final sample far exceeded that threshold. Randomization checks confirmed that the four groups did not differ in age, education, relationship status, or any of the psychological variables measured. The researchers also collected data on participants’ own sex, mental well-being, impulsivity, dispositional forgiveness, and dispositional vengeance, using well-validated instruments including the Barratt Impulsiveness Scale, the Bolton Forgiveness Scale, the Vengeance Scale, and the Warwick-Edinburgh Mental Wellbeing Scale. Internal consistency was acceptable to excellent across all measures, with the Vengeance Scale reaching a McDonald’s omega of 0.91. All analyses employed robust standard errors and bootstrap estimation with 5,000 samples to guard against violations of normality.

The results on perceived plausibility were subtle but telling. Neither the avenger’s sex nor the context alone significantly shaped how believable participants found the scenario. Instead, the two factors interacted: in public settings, scenarios featuring a female avenger were rated as more plausible than those featuring a male avenger, with a statistically significant difference corresponding to a small effect size (partial eta-squared of 0.008). In private settings, no such difference emerged. The Female/Public condition received the highest mean plausibility rating of 3.48, while the Female/Private condition received the lowest at 3.18. The authors caution that this interaction, though significant, represents a modest variation in social perception rather than a dramatic practical effect.

The picture changed dramatically when participants were asked whether they themselves would act as the avenger did. Here the overall model explained a substantial 20 percent of the variance, and a large main effect of the avenger’s sex emerged: female-avenger scenarios received markedly higher endorsement than male-avenger scenarios in both contexts. Crucially, the gap widened when the act was public. In the public condition, the difference between female and male avengers corresponded to a large effect (d = 0.91), nearly double the moderate effect observed in the private condition (d = 0.44). Descriptively, the Female/Public condition again topped the ratings at a mean of 2.61, while the Male/Public condition sank to the bottom at 1.42. In other words, participants were most willing to imagine themselves taking public revenge as women, and least willing to do so as men.

Perhaps the most important finding is that this pattern survived statistical adjustment. When the researchers re-ran the model controlling for perceived plausibility, participants’ own sex, well-being, impulsivity, forgiveness, and vengeance, the avenger’s sex by context interaction remained significant, with effect magnitudes essentially unchanged. Among the covariates, dispositional vengeance, perceived plausibility, and impulsivity each predicted endorsement, whereas participants’ own sex, well-being, and forgiveness did not. This persistence indicates that the contextual and gendered contribution to vengeance endorsement is not reducible to stable individual differences. Something about the situation itself—who acts and who watches—shapes how acceptable revenge appears, above and beyond the psychological profiles of the people doing the judging.

The authors interpret these results through a communicative-signaling framework. Drawing on work by Boon and Yoshimura, they argue that a vengeful act simultaneously delivers three messages: that an injury has occurred, that the avenger is willing to defend their standing, and that future violations will carry a cost. When the act is enacted before others, this signal becomes more legible and more consequential, because the audience to whom it is implicitly addressed expands. Social media platforms amplify this dynamic further, extending the reach of a retaliatory act far beyond the immediate circles of the avenger and the target. On this view, the public versus private context of revenge is not a peripheral detail but part of what the act means—and the finding that visibility amplifies the difference between female and male avengers suggests that gendered expectations about who may legitimately perform retribution in front of others are woven into those signals.

The study is not without limitations, which the authors acknowledge candidly. The convenience sample was recruited in Italy, limiting generalizability to other cultures; the study lacked a formal manipulation check confirming that participants interpreted the public and private conditions as intended; it was not preregistered; and written scenarios may not capture the emotional intensity of real betrayal. The researchers also note an ethical concern: the scenarios may inadvertently reinforce stereotypical portrayals of the sexes, and future work should incorporate counter-stereotypical narratives and direct measures of gender identity and gender norm endorsement. Still, the practical implications are tangible. Because impulsivity predicted vengeance endorsement, clinical conversations about betrayal may benefit from attention to emotion-regulation strategies of the kind used in cognitive-behavioral interventions. And because the visibility of a vengeful act shaped its acceptability, therapists may want to explore how clients construe the audience and reputational dimensions of their own potential responses to infidelity.

What Project Nemesis ultimately offers is a reframing of one of humanity’s oldest impulses. Revenge, the study suggests, is not simply retaliation—a raw discharge of hurt. It is a socially embedded, communicative process that depends on who enacts it, who witnesses it, and the gendered expectations that surround both. The perceived plausibility and endorsement of vengeful acts arise from the joint contribution of situational cues and dispositional tendencies, consistent with a biopsychosocial model in which fast, emotionally driven reactions are tempered by slower, intentional character processes. As the authors conclude, perceptions of vengeance are not purely reactive but reflect the dynamic interplay of personal characteristics, social context, and the audience to whom the act is, at least implicitly, addressed. In an era when a single post can broadcast a betrayal to thousands, understanding how visibility and gender shape the social calculus of revenge has never felt more urgent.

Subject of Research: Gender and context effects on the perceived plausibility and endorsement of vengeful behavior after relational transgressions

Article Title: Project Nemesis: vindictive behavior in public and private contexts

Article References: Chiarolanza, C., Isolani, S., & Garcia, D. (2026). Project Nemesis: vindictive behavior in public and private contexts. Current Psychology, 45(18), Article 1545. https://doi.org/10.1007/s12144-026-10082-z

Image Credits: AI Generated

DOI: 10.1007/s12144-026-10082-z

Keywords: vengeance, revenge, relational transgressions, infidelity, forgiveness, gender differences, public and private contexts, social signaling, impulsivity, social psychology, Project Nemesis, Current Psychology