Study finds the “bowing effect” on attractiveness is cultural, not universal

People in Japan tend to view a face as more attractive when the person performs a traditional bow, and this boost occurs regardless of the individual’s racial appearance. Observers in the United States, on the other hand, show little or no reaction to bowing, suggesting that cultural habits rather than universal body language explain the effect. The findings, published in Japanese Psychological Research, provide evidence that social etiquette can shape how aesthetic qualities are judged.

Bowing is a physical gesture that involves lowering the head and bending the upper body forward. In Japan, it serves as a customary greeting used to show respect, express gratitude, or offer an apology. The gesture can range from a slight fifteen-degree tilt when greeting someone in passing to a thirty-degree bow of respect for a senior colleague, or a deep forty-five-degree bend reserved for heartfelt apologies or profound thanks.

The research, led by Takayuki Osugi of Yamagata University, builds on earlier findings showing that tilting an image forward to simulate a bow increases perceived facial attractiveness among Japanese raters. Psychologists suggest that this reaction occurs because bowing activates cognitive schemas, which are mental frameworks that organize knowledge and expectations. In this context, the gesture triggers cultural associations with politeness and submissiveness, or non-threatening social deference. In Japanese society, these traits encourage a favorable overall impression that spills over into aesthetic judgments, a pattern known as a halo effect.

In related work, an experiment exploring head orientation found that downward head tilts increased perceived femininity and attractiveness for female faces, whereas upward tilts increased perceived masculinity. That earlier work examined vertical head motion as an evolutionary cue related to physical height and biological sex differences. In contrast, Osugi and his team approached bowing as an intentional, culturally learned greeting intended to communicate social respect.

One competing explanation holds that lowering the head is a universal signal of submission, observed across non-human primates, which would predict that bowing boosts attractiveness in any culture. Testing American observers offered a way to distinguish that account from a culturally specific one. In addition, past studies on the bowing effect relied almost entirely on East Asian faces shown to Japanese observers. Because the raters and the faces shared the same ethnic background, it was difficult to tell whether the boost occurred because the gesture was culturally familiar or because people were observing members of their own group.

Free daily newsletter

To investigate, the researchers developed three-dimensional computer-generated models representing four racial appearances: East Asian, European, South Asian, and African. Both male and female models in their twenties were created with varied hairstyles and everyday clothing. In an initial setup phase, independent groups of Japanese volunteers rated the static appearance of these faces on a scale from 0 (unattractive) to 100 (attractive) so the team could select models with average baseline attractiveness scores of roughly 42 points.

Experiment 1 focused on Japanese participants recruited through an online crowdsourcing platform. Across six sub-experiments, each testing different combinations of model gender and race, between 45 and 49 participants evaluated faces under three movement conditions: a 1.5-second long bow, an immediate short bow, and a static upright control condition. About half of the Japanese participants took part in more than one sub-experiment, though the sub-experiments ran intermittently over roughly three years. A follow-up analysis accounting for repeated participation produced the same pattern of results.

Japanese observers consistently awarded higher facial attractiveness ratings to bowing models than to standing-still models across all sub-experiments. For instance, when rating female European models, the average score was 43.2 for the static condition, 51.1 for the short bow, and 59.0 for the long bow. Across the different model types, scores rose by roughly 6 to 12 points for short bows and 15 to 28 points for long bows compared to static faces. Across all tests, holding the bow longer produced higher ratings than a brief bow, and the interaction between the motion condition and the model’s racial appearance was not statistically significant.

In Experiment 2, the researchers presented the identical video sequences and rating tasks to American participants recruited through Amazon Mechanical Turk. Between 30 and 40 American participants completed each sub-experiment. Participant overlap was minimal in this group, ranging from just one to seven participants across sub-experiments.

American participants showed minimal change in their attractiveness ratings when models bowed. Some differences did reach statistical significance, but the gaps were small, typically under three points on the 100-point scale, compared with gains of 15 points or more among Japanese raters. In female European model tests, for example, American ratings averaged 56.5 for the static pose, 57.4 for the short bow, and 56.9 for the long bow. A direct statistical comparison between the two national samples confirmed that the attractiveness boost from bowing was robust among Japanese participants but essentially absent among American participants.

The authors then conducted Experiments 3 and 4 to test whether the bowing effect generalizes across different levels of baseline physical attractiveness. These follow-up experiments also addressed a potential confound: all initial attractiveness ratings had been completed by Japanese university students, leaving it uncertain whether American raters would judge the facial aesthetics of the models similarly. In Experiment 3, 49 Japanese participants evaluated female East Asian models and 48 evaluated male East Asian models, with faces categorized as either distinctly attractive (averaging around 55 to 58 points) or less attractive (averaging around 28 to 30 points).

Bowing increased attractiveness scores for both attractive and less attractive faces among Japanese raters. For less attractive female models, scores rose from 33.3 when standing still to 42.0 for a short bow and 49.0 for a long bow. For highly attractive female models, scores rose from 48.0 when standing still to 54.3 for a short bow and 62.7 for a long bow. Holding the bow longer produced a steady enhancement regardless of baseline looks, though a brief bow provided a slightly smaller relative lift to already attractive faces.

To evaluate American aesthetic perceptions and task engagement, Experiment 4 presented the same high- and low-attractiveness models to 35 American participants for female models and 32 for male models. The participants easily distinguished between the two tiers, giving high-attractiveness female models an average score around 65 points and low-attractiveness models around 44 to 45 points. Within each tier, bowing motion produced almost no change in ratings, and though a small overall effect of motion appeared for male models, no individual comparison held up after correction. These results indicate that American viewers were paying close attention and evaluating facial aesthetics normally, yet simply did not factor the bowing motion into their judgments.

There are a few things to keep in mind when interpreting these findings. The researchers did not record how much exposure individual American participants had previously had to Japanese culture, anime, or travel. While some participants might have had passing familiarity with bowing etiquette, measuring this background would likely have highlighted the cultural divide even further.

The authors also did not collect race-related demographic information from the American sample, meaning future studies will need to examine whether an observer’s own racial background interacts with the racial appearance of the bowing person. Additionally, participants completed the tasks on their personal computers, so screen size, resolution, and viewing distance were not standardized. Live interpersonal encounters also involve vocal tone, body posture nuances, and spontaneous eye contact that digital avatars on computer monitors cannot capture.

The study, “Effect of Bowing on Facial Attractiveness of Other Races,” was authored by Takayuki Osugi and Jun I. Kawahara.