(Credit: GaudiLab on Shutterstock)
Males unconsciously advertise their fighting ability through two simple movements, study shows.
In A Nutshell
- People judged dominance from ~3–4-second walking clips of size-standardized avatars, so only gait cues were visible.
- Two motion cues stood out: side-to-side torso sway and holding the shoulders out from the chest. Both added information beyond body size.
- Different mixes of size, sway, and shoulder position can lead to similar “tough” ratings. Smaller men can look just as intimidating as larger ones.
- The effect was modest on its own. In the world, people read many cues at once, so gait likely combines with size, face, voice, and context.
NEWCASTLE, England — In a crowded bar, on a dark street, or across a parking lot, the human brain makes snap judgments about who poses a threat. Scientists have long known that height and muscle mass drive these assessments. But a new study from Northumbria University reveals something unexpected: how men walk adds to body size in shaping threat impressions, and the visual system can read these movement cues quickly, even with limited detail.
Researchers used 3D motion capture technology to record 52 men walking naturally, then asked 137 people to rate how physically dominant each walker appeared based solely on their gait. What emerged was a pattern where different combinations of body size and walking style can lead to similar dominance ratings. A smaller, weaker man who walks with exaggerated movement patterns can register as just as dominant as a substantially larger, stronger man who moves with minimal flair.
Two specific walking patterns stood out: torso sway (how much the upper body rocks side-to-side with each step) and shoulder abduction (holding the shoulders away from the chest). Men who displayed more of these movements were rated as significantly more dominant, even when viewers couldn’t see their actual body size, face, or anything else besides the walking pattern.
“Physical dominance” in this context refers to the ability to win physical fights, a trait that has shaped human evolution through male competition. Throughout history, correctly judging who could inflict harm helped people avoid dangerous conflicts. But previous research focused almost exclusively on static cues like facial structure, chest width and height. This study shows that dynamic movement provides its own independent channel of information about fighting ability.
How the Brain Detects Threats Through Motion
Male competition has driven the evolution of sexually dimorphic traits in humans, from thicker skulls to greater upper body strength. These features either help inflict damage or defend against it. Upper body muscle mass, for instance, correlates with punching power.
Research suggests the sudden stretching of active jaw muscles following a blow may stabilize the jaw and help reduce head acceleration, though this remains an area of ongoing investigation. But traits that directly facilitate violence are only part of the equation. Equally important are the neurocognitive mechanisms that assess risk. Deciding whether to engage in a violent altercation carries high stakes, as even the victor can sustain life-threatening injuries. As a deeply social species, humans have developed sophisticated abilities to infer fighting ability. Even children show an understanding of conflict dynamics.
Previous research found that both men and women from culturally diverse samples made similar judgments when identifying physically dominant men from photographs. Body size emerged as a key determinant of knockout power across species. Visual cues appear to track upper-body strength and, by extension, formidability.

But here’s the catch: participants in those studies judged formidability from static images. But photographs freeze people in place. Real-world encounters involve movement, and the way joints shift and bodies flow through space might reveal different information about someone’s capacity for individual fighting ability.
Research on biological motion shows that humans extract remarkable amounts of social information from movement alone. Scientists often use “point-light displays” for these studies, which are basically just dots placed on shoulders, elbows, wrists and other key spots. Even with these stripped-down animations, observers can figure out gender, personality traits, mood and intentions.
Studies show that even scrambled versions of these dot displays trigger specialized brain regions that recognize living things. Research has found that people born blind who later gain sight can immediately recognize human walking patterns, and newborns show signs of this ability too.
Walking gait tends to happen without conscious thought, which means it might reveal honest information about someone. Many animal courtship displays work this way, with females judging males based on performances that are hard to fake.
Capturing Natural Gait Patterns
The Northumbria research team captured walking patterns from 52 men aged 18 to 41, recording their gaits with a 14-camera Vicon motion capture system. Each participant walked naturally along an eight-meter walkway while 38 reflective markers tracked their movements at 200 frames per second.
To help participants relax into their natural stride, researchers framed the initial walks as “dynamic calibration.” After about five minutes, six one-way recordings were made without alerting the walkers. One recording from the latter three walks was selected for each participant.
Motion data were then used to create standardized, featureless humanoid avatars of uniform height and build, eliminating all visual information except the walking pattern itself. The clips showed approximately two full gait cycles and lasted between three and four seconds.
Beyond the motion capture, researchers collected hand grip strength (as a proxy for overall strength), circumference measurements of biceps, shoulders, chest, waist and hips, plus scores from the Buss Perry Aggression Questionnaire, which measures tendencies toward physical aggression, verbal confrontation, anger and hostility.
How Walking Style Can Make You Look Dominant
Before getting into the technical details, researchers showed the five most dominant-looking walks and the five least dominant-looking walks to ten people and asked what made them different. Everyone agreed on two things: swagger and sway.
Sway is straightforward: how much someone’s torso rocks left and right with each step. Researchers measured this by tracking the distance between a marker on the middle back and markers on the hip bones. Since people walk at different speeds, they adjusted these measurements so everyone could be compared fairly.
Swagger involved three movements: holding the shoulders away from the chest, bending the elbows, and rotating the upper arms inward. Together, these created the look of an inflated upper body, with arms held away from the torso instead of swinging straight alongside. When researchers ran their final analysis, shoulder position mattered most.
The data, published in Scientific Reports, showed strong connections between how dominant someone looked and both sway and shoulder position. But here’s what makes it interesting: these movement patterns also connected to actual physical measurements. Men with more sway and wider shoulders tended to have larger bodies, stronger grips and higher scores on aggression questionnaires.
A statistical model that pitted all these factors against each other found that sway, shoulder abduction, and body size/strength each made independent contributions to dominance ratings. Movement patterns weren’t just proxies for size. They added unique information that the brain appears to extract and process automatically.


When Size Doesn’t Tell the Whole Story
The statistical analysis revealed something particularly interesting: different mixes of these three factors could create the same impression. A big, strong guy who walks normally might look just as dominant as a smaller, weaker guy who amps up his sway and shoulder position.
This raises the possibility that smaller men might unconsciously (or consciously) adjust how they walk to project more dominance. Meanwhile, very large men might not need to advertise through movement.
The model explained about 6.6% of the differences in how people rated dominance. That might sound small, but remember: viewers saw nothing but movement. No face, no actual body size, no context. In real life, where all those other cues are visible, movement likely plays a bigger role.
Interestingly, men who scored higher on aggression questionnaires actually showed less sway. This might make sense if someone truly plans to fight rather than avoid one. A more stable, centered walk provides better balance for both throwing punches and staying upright when hit.
Another surprise: older raters gave higher dominance scores across the board, though the study doesn’t dig into why.
Gait Matters When Size Is Hidden
From an evolutionary perspective, quickly spotting threats would have been crucial for survival. Body size can be estimated at a glance. Sway takes a few seconds of watching. So why bother with movement cues if size alone works?
The researchers suggest that movement might provide reliable information when size gets masked. In a crowd, at dusk, or when someone wears bulky clothes, body shape can be hard to gauge. But movement stays visible. Studies show these motion signals hold up well even with visual noise, potentially making them a backup channel for threat assessment.
Many species size each other up before fighting, often called ritualized aggression. Red deer, for instance, will go through escalating displays before actual combat. In humans, visual assessment of the upper body matters for judging fighting ability, and big enough size differences can prevent fights entirely.
Animals sometimes fake size too. Pufferfish inflate dramatically to scare off predators. Similarly, holding the shoulders wide might make the upper body look bigger than it really is.
Can Men Fake Dominance Through Their Walking Style?
This raises questions about honest versus dishonest signals. In animal communication, signals generally stay honest on average. If dishonest signals become too common, receivers would learn to ignore them and the whole system would collapse. But some dishonesty can persist, especially when fighting risks are extremely high and judging ability is difficult.
Could smaller, weaker men bluff their way out of confrontations with exaggerated walking? The data leave room for this possibility. Weaker men might use extra sway and shoulder width to appear more dominant. In theory, this might work to slightly boost perceived threat level and avoid fights with stronger men, particularly because walking patterns may be hard to fake convincingly.
But if bluffing becomes too obvious or too common, stronger men would catch on and start ignoring these cues. The researchers suggest that walking dynamics might be genuinely difficult to fake well, which could keep the signal relatively honest over evolutionary time.
The relationship between aggression and sway adds another wrinkle. Someone who actually wants to fight (rather than avoid one) would benefit from a stable, centered walk. Better for throwing punches and maintaining balance. In that case, the swagger and sway that project dominance might actually interfere with the goal.
Real-World Limits
This study looked only at normal walking. Other research has examined different dominance cues like how straight someone stands or the expansive gestures athletes make after winning.
The research can’t tell whether men consciously control these walking patterns or whether they naturally reflect each person’s strength, size and temperament. Some studies report that people across cultures can judge male strength from walking patterns, and that women in some cultures find stronger men’s walks more attractive.
One important caveat: this study used only animated figures that revealed nothing but movement. Real life, of course, includes faces, actual body size, voices, social context and more. How all these channels work together remains unclear.
The sample included only men, focusing on male-on-male competition. Whether similar walking patterns affect how people judge women’s dominance, or judgments across genders, hasn’t been studied.
People almost certainly aren’t consciously tracking torso angles when deciding if someone looks threatening. But the research shows that specialized brain systems are pulling out this information automatically. Male competition has shaped human evolution for thousands of years, leaving its mark on both physical traits and how we read them. The brain can spot a potential threat across a room based on nothing more than walking style. And in that split-second assessment, size doesn’t tell the whole story.
Paper Summary
Methodology
Fifty-two men aged 18-41 had their walking patterns recorded using a 14-camera Vicon motion capture system running at 200 frames per second. Thirty-eight reflective markers were attached to each participant following the Plug-In-Gait kinematic model. Participants walked naturally along an eight-meter walkway, with one recording selected from the latter three of six one-way walks to ensure natural gait. The recordings showed approximately two full gait cycles lasting 3-4 seconds. Motion data were used to create standardized, featureless humanoid avatars of uniform height and build, eliminating all visual information except walking dynamics. Researchers also collected anthropometric measures (biceps, shoulder, chest, waist and hip circumferences), hand grip strength, and Buss Perry Aggression Questionnaire scores from each walker. A separate group of 137 raters viewed all 52 walking videos in random order and rated each on a 1-7 scale for “physical dominance,” defined as the regularity of winning physical fights. Biomechanical analysis focused on two features identified through initial observation: sway (mediolateral torso oscillation) and swagger (shoulder abduction, elbow flexion and shoulder internal rotation).
Results
Dominance ratings showed excellent reliability across raters (Cronbach’s α=0.97). Sway, shoulder abduction and shoulder internal rotation (in simple correlations) all correlated significantly with dominance ratings. These biomechanical measures also correlated with a principal component (ANTH) capturing body size and strength, and trait aggression scores. A linear mixed-effects model predicting dominance ratings found independent contributions from sway (β=0.037, p=.02), shoulder abduction (β=0.077, p=.006), and ANTH (β=0.25, p=.009). Rater age also influenced ratings, with older raters assigning higher dominance scores (β=0.013, p=.04). The model explained 6.6% of variance in dominance scores (f²=0.071), characterized as a small to medium effect. Shoulder internal rotation, elbow flexion, trait aggression and rater sex did not make independent contributions in the final model.
Limitations
The study focused exclusively on walking gait, so findings may not generalize to other movement patterns or postural displays. The sample included only men, limiting conclusions about gender differences or cross-gender perceptions of dominance. The controlled laboratory setting and use of standardized avatars enhanced internal validity but reduced ecological validity compared to real-world encounters where multiple cues (size, face, voice, context) are simultaneously available. The model explained only 6.6% of variance in dominance scores, indicating other unmeasured factors play substantial roles. The study couldn’t determine whether gait patterns are consciously controlled or automatic, nor whether they function as honest signals or can be strategically manipulated. Cross-cultural generalizability remains unclear, as all participants and most raters were from similar cultural backgrounds.
Funding and Disclosures
The paper does not explicitly mention funding sources or declare competing interests beyond stating “The authors declare no competing interests.” The research was conducted at the Department of Psychology, Northumbria University, Newcastle Upon Tyne, UK, with ethics approval from the department’s ethics committee (reference SUB05_KM, approved June 16, 2014).
Publication Information
McCarty, K., Leslie, C., Anguera de la Rosa, A., Saxton, T. K., & Cornelissen, P. L. “Perceiving intimidation through kinematic cues in men’s gait,” was published September 29, 2025 in Scientific Reports. doi:10.1038/s41598-025-16400-y





