Red Backgrounds Enhance Dominance in Human Faces and Shapes

Abstract

Red color signals dominance in both animals and humans. This study investigated whether a red background color influences the perception of dominance in human faces and geometric shapes. The facial stimuli consisted of computer-generated faces, quantitatively morphed into nine levels of dominance, ranging from less dominant to more dominant. This included East-Asian female faces in Experiment 1 and male faces in Experiment 2. The face stimuli were presented against three background colors: red, green, and gray. Participants were instructed to categorize the faces as either obedient or dominant by pressing the corresponding labeled keys. The results showed that faces were more likely to be perceived as dominant when presented against a red background than against green or gray backgrounds, for both female and male faces. Additionally, two questionnaire surveys showed that the perception of dominance also increased for shapes presented against a red background. However the effect of red diminished in the absence of the actual perception of the color red. These results suggest that the perception of dominance in both human faces and objects is enhanced by the presence of red, possibly due to evolutionary factors related to the perception of red.

Introduction

People live in a colorful world, where colors are perceived on almost every object in daily life. Research suggests that colors carry significant information that can influence emotions, cognition, and behavior (Elliot & Maier, 2014). Among all colors, red has been associated with dominance, aggression, threat, anger, danger, arousal, and sexual attractiveness, with its effects having notable behavioral consequences (Elliot & Maier, 2012, 2014; Elliot & Niesta, 2008; Fetterman et al., 2012; Mentzel et al., 2017; Pravossoudovitch et al., 2014; Wiedemann et al., 2015).

Comparative data suggest that the psychological effects of red have deep evolutionary roots in dominance signaling. Red is associated with dominance and aggression across a wide range of animal species, including anthropoid primates (Khan et al., 2011; Rhodes et al., 1997; Setchell & Jean Wickings, 2005), fishes (Dijkstra et al., 2005), birds (Pryke, 2009), reptiles (Healey et al., 2007), and humans (Hill & Barton, 2005). In primates, trichromatic color vision (red, green, and blue), in addition to dichromatic vision (green and blue), is believed to be optimal for identifying changes in skin redness, thereby facilitating the assessment of social and emotional states in conspecifics (Changizi et al., 2006). This provides evidence for the role of red in social interactions. In humans, a red face in an opponent often signals anger, dominance, aggressiveness, while pallor signals fear (Archer, 2006; Changizi et al., 2006; Drummond, 1997; Hackney, 2006; Hill & Barton, 2005; Montoya et al., 2005; Young et al., 2013). Facial redness is associated with higher testosterone levels and/or increased blood flow due to emotional arousal, such as anger or excitement, which is often associated with dominance and aggressive behavior (Setchell et al., 2008). Displaying red skin may signal dominance and a readiness to fight through a bottom-up processing (Changizi et al., 2006; Stephen et al., 2012). On the other hand, a red contextual color (such as wearing red or having a red background) can have a similar effect to facial redness but might bias the processing of emotional expressions in a top-down manner (Peromaa & Olkkonen, 2019; Wiedemann et al., 2015; Young et al., 2013). Rooted in evolutionary biology and reinforced by social interactions, the color red is commonly associated with dominance (Mentzel et al., 2017). Individuals wearing red are perceived by themselves and others as more dominant and aggressive (Feltman & Elliot, 2011; Wiedemann et al., 2015). Hill and Barton (2005) showed that athletes randomly assigned red sportswear (rather than blue) were more likely to win in combat sports at the Olympic Games. Even abstract shape colored red is rated as more dominant, aggressive, and more likely “to win a fight” compared to those colored blue (Little & Hill, 2007). However, some studies have failed to demonstrate the effect of red clothing on dominance/aggressive performance in competitive contests (García-Rubio et al., 2011; Kocher & Sutter, 2008; Kramer, 2016; Pollet & Peperkoorn, 2013). So far, few studies have directly examined the effect of a red background on the perception of dominance. Young et al. (2013) demonstrated that a red background color to presented faces enhances the perception and identification of anger. Research has shown that there is a positive association between anger and human dominance behavior (Cabral et al., 2016). It is possible that a red background could bias the perception of dominance.

Faces play a crucial role in social interactions, and humans can rapidly and automatically evaluate faces based on various trait dimensions, including dominance and trustworthiness (Bar et al., 2006; Bonassi et al., 2021; Oosterhof & Todorov, 2008). Dominance has long been considered a central trait in interpersonal relations, perceptions, and emotions (Wiggins, 1979). It involves the use of intimidation and coercion to achieve a social status based on the induction of fear (Cheng et al., 2010; Henrich & Gil-White, 2001). Displays of dominance help establish social hierarchies and assign roles within a group (Rule et al., 2012; Mast, 2001). Dominant individuals are generally perceived as socially influential, aggressive, fearlessness, highly competitive, and often have primary access to resources, including materials and mates, while obedient or submissive individuals may expect protection from those of higher rank (Fiske, 1992). Thus, judging the dominant individuals is critical to survival and tends to be consistent among perceivers (Keating & Bai, 1986; Keating et al., 1981; Mazur, 2005). Certain facial features and postures, as well as aggression-related emotional expressions (e.g., eye gaze, facial shape cues), contribute to the perception of facial dominance (Chiao et al., 2008; Hess et al., 2000; Mileva et al., 2014). For instance, males with a higher facial width-to-height ratio are generally perceived as more dominant (Mileva et al., 2014).

The current study aims to address this research gap by examining how the interaction between red background color and facial features affects the perception of dominance in human faces. The study uses computer-generated faces with varying degrees of dominance cues, morphed from obedient to dominant, based on East-Asian female faces (Experiment 1) and male faces (Experiment 2). These faces are presented against red, green, and gray background colors. Green was selected because it is the color opposite to red in many well-established color models, and gray was chosen as an achromatic contrasting color that can be matched for saturation and lightness (Elliot et al., 2010). We hypothesized that a red background color would interact with facial dominance features to enhance the perception of dominance in both female and male faces. Furthermore, to gain a deeper understanding of the impact of the red effect on the perception of dominance, we conducted two additional online questionnaire surveys. The first survey aimed to investigate whether the observed effect of red background in enhancing dominance judgments extends to evaluations of simple geometric shapes or if it is restricted to faces. The second questionnaire delved deeper, exploring whether the effect of red on dominance was limited to perceptual color or if it also applied to the semantic representations of color in the background. We hypothesized that the perceptual red background color significantly influences perceptions of dominance in both faces and shapes.

Methods

Experiment 1: Effect of Background Color on Female Face Dominance

Participants

Twenty-eight Japanese undergraduate students (14 males and 14 females, mean age = 20.1 years, SD = 1.9) from Waseda University participated in the experiment. All participants had normal or corrected-to-normal visual acuity and normal color vision, and were unaware of the purpose of the experiment. The sample size was set a priori at 23 participants, based on a target of 0.8 power with a medium effect size by power analysis (Cohen's d = 0.6). More data were collected in case of poor performance (e.g., failed to fit the psychometric function). The experiment, as well as subsequent experiments, was approved by the institutional review board (IRB) of Waseda University (2015-033), and conducted in accordance with the ethical standards of the 1964 Declaration of Helsinki. Written informed consent was obtained from all participants in advance.

Apparatus and Stimuli

E-Prime 2.0 (Psychology Software Tools, Inc.) was used to present the stimuli and collect the data. Stimuli were displayed on a 24-inch LCD monitor (EIZO FG2421, EIZO corp, Hakusan, Japan), with a 1920 × 1080-pixel resolution and a refresh rate of 100 Hz. Participants viewed the monitor binocularly at a distance of approximately 60 cm.

In this study, an average-looking East-Asian female face was manipulated in a facial dominance dimension using the average dominance rating scores that 20 Japanese participants (10 males and 10 females, mean age = 21.40, SD = 1.43) gave to 200 East-Asian female faces. Morphs were created using FaceGen Modeller (Singular Inversions, Toronto, Canada). The facial dominance transformation was carried out in a data-driven manner the same as that Nakamura and Watanabe (2019) transformed facial attractiveness (see Supplemental materials for detailed information). Nine faces were chosen (SD = −5, −3, −1.5, −0.5, 0, 0.5, 1.5, 3, 5; see Figure 1) from less dominant (obedient) to more dominant. The face stimulus was fitted in a frame of 512 × 512 pixels, presented with three different background colors created by Photoshop CS2 (Adobe Systems Inc., San Jose, CA, USA). The three background colors were measured by PR-655 (Photo Research, Chatsworth, CA, USA) and each color was measured 10 times and calculated the average color values. The color information was as follows: Red: L* = 56.66, a* = 80.59, b* = 57.60; Green: L* = 60.03, a* = −64.56, b* = 41.64; Gray: L* = 57.59, a* = −2.46, b* = −30.15.

Figure 1.

The morphed female faces in nine levels of dominance SDs with three background colors.

Procedure

The experiment was conducted in a laboratory under dim lighting conditions. At the beginning of the experiment, participants were instructed to denote the dominance of a centrally presented face stimulus by pressing a key on the computer keyboard (e.g., pressing “z” if the face stimulus is dominant using the left index finger, or pressing “m” if the face stimulus is obedient using the right index finger). Each trial began with a fixation for 500 ms, followed by a face stimulus with a background color (i.e., red, blue, or gray) for 500 ms, after which a black screen was presented until participants responded (Figure 2). Participants were requested to identify the dominance of the face stimulus (dominance or obedience) as quickly as possible. Trials were separated by an inter-stimulus interval (ISI) of 500 ms. Each face stimulus was presented 12 times in a random order, resulting in 324 trials (9 faces × 3 background colors × 12 times = 324 trials). The experiment was preceded by 20 practice trials. The entire experiment took approximately 20 min to finish.

Figure 2.

Example of the trial sequence. Participants were required to categorize the dominance of the face stimulus (dominance or obedience) by pressing one of two keys.

Analysis

All statistical analyses were performed using R (R Core Team, 2020). The dominance categorization rate was computed for each face stimulus with different background colors. The dominance categorization rate from each participant was fitted with a psychometric function using a generalized linear model with a binomial distribution. A psychometric function describes the probability of a correct answer as a function of stimulus intensity. Fitting a psychometric function can reveal the level of stimulus variable at which perceptual performance transitions from one perceptual category to the other (i.e., the point of subjective equality [PSE]) and the precision with which the participant is able to perceptually differentiate stimuli along the dimension (i.e., the just noticeable difference [JND]; Klein, 2001; Kingdom & Prins, 2016). To further verify the effect of background color on dominance perception, the PSE and JND obtained from the fitted curves were used to show the perceived differences in dominance judgments and discrimination sensitivity, respectively. The PSE was determined from the x-value at which the fitted curve had a y-value of 0.5, indicating the probability of the face stimuli being judged as equally dominant or obedient. It represents the perceived midpoint of dominance perception, showing where the participant perceives the two faces as having the same perception of dominance. The shifts of PSE show how the subjective perception of dominance changes under different background color conditions. The JND was calculated from one-half of the difference between the x-values at which the fitted curve had y-values of 0.25 and 0.75. The JND represents the participant's sensitivity to changes in the face stimulus and measures the minimum difference needed for the participant to detect that one face is more dominant than the other. The PSEs and JNDs of the three background colors from the fitted curves of each participant were compared using one-way ANOVAs. Paired-sample t-tests using Bonferroni's correction for multiple comparisons (α = 0.05/3 = 0.017) were used to further reveal the differences between the three background colors for the PSEs and JNDs, separately.

As the dependent variable was binary (i.e., dominance vs. obedience responses), we used the generalized linear mixed-effect models (GLMMs; the glmer function) to examine the effect of background color on the dominance of face stimuli (Bates et al., 2015; Jaeger, 2008). We anticipated that the perception of face dominance would depend on those two factors with the background color and the face morphing levels. Thus, we included the background color, face dominance level, and their interactions as fixed factors in the model, and then eliminated factors that reduced the model's overall goodness of fit. The Akaike information criterion (AIC) was also used to compare the goodness of fit of the alternative models (a lower AIC value indicates a higher quality model). After the model selection, analysis of variance (Type II Wald x² test) was performed to identify significant effects and interactions, and Tukey post hoc pairwise comparisons were performed (using car and emmeans packages; Fox et al., 2013; Lenth et al., 2018). The data analysis can be found on OSF https://osf.io/xpwek/.

Results

Five participants whose responses failed to fit the psychometric function were removed from the data analysis (s1, s2, s5, s14, s18; mean SD > 2, see plots for each participant in Supplemental materials, https://osf.io/xpwek/). Thus, 23 participants’ data were used for further analysis. We first compared the overall responses of obedience and dominance across the conditions of background colors and facial dominance levels, and found that participants made obedience responses more frequently than dominance responses, t(22) = 2.97, p = .007, Cohen's d = 0.63. It might be related to the fact that face stimuli were morphed based on female faces. The proportion of dominance response to each face stimulus and their psychometric curves with the three background colors are shown in Figure 3.

Figure 3.

The proportion of dominance responses for each female face stimulus by psychometric functions. Colored curves show the logistic fits to the psychometric results for the three background colors. The violin plot showing the distribution of individual PSEs and JNDs in the three background colors are presented. The horizontal line represents the mean. Error bars represent the standard mean error. Asterisks indicate significant differences (** p < .01, after Bonferroni correction).

The data points represent the mean proportion of trials in which participants responded with dominant judgement for each morphed face stimulus presented with three background colors. All statistical analyses were conducted on those fitted to each individual's data. A one-way ANOVA analysis of the PSE showed a significant main effect of background color (Figure 3), F(2, 44) = 33.31, p < .001, η_p²= 0.07. The PSEs with red background color (Mean = 0.19, SD = 1.09) were significantly smaller than with green background color (Mean = 0.82, SD = 1.02; t(22) = 6.78, Bonferroni corrected p < .001, Cohen's d = 1.41), and gray background color (Mean = 0.80, SD = 1.10; t(22) = 6.99, Bonferroni corrected p < .001, Cohen's d = 1.46). No difference was observed in PSEs between the gray and green background colors (t(22) = 0.22, Bonferroni corrected p = 1, Cohen's d = 0.05). Thus, an ambiguous face presented with a red background color was more likely to be perceived as dominant than one presented with gray or green background colors. A one-way ANOVA analysis of JNDs showed that there was a marginally significant difference between the three background colors (Figure 3), F(2, 44) = 2.48, p = .096, η_p²= 0.06. It may suggest that the sensitivity of face dominance perception is little influenced by the background colors.

The generalized linear mixed-effect model analysis showed that there was no significant interaction effect between the background color and face dominance morphing levels (i.e., 9 levels of SDs), x(16)²= 18.15, p = .32. After removing the interaction effect from model analysis, a final model including the background color and face morphing levels as fixed factors, and intercepts of participants as random factors showed a main effect of background color, x(2)²= 119.07, p < .001, and face dominance morphing levels, x(8)²= 1492.79, p < .001. Participants made more dominance responses when the face stimulus was presented in red background colors than in gray (p < .001), and green background colors (p < .001). No difference was observed between the gray and green background colors (p = .95). These analyses performed on the categorical responses led to the same conclusion: the red background color had a greater likelihood of enhancing the perception of dominance in female faces when compared to green and gray background colors.

In summary, those results showed that a red background color could bias a morphed female face to be perceived as dominant, compared with gray and green background colors. It may suggest a robust congruency effect of red-dominance association on face dominance perception. In Experiment 2, we replicated those findings with examining whether the effect of red background color could bias the dominance perception of morphed male faces.

Experiment 2: Effect of Background Color on Male Face Dominance

Participants

Twenty-seven newly recruited Japanese undergraduate students (13 males and 14 females, mean age = 20.22 years, SD = 1.5) from Waseda University participated in the experiment. All participants had normal or corrected-to-normal visual acuity and normal color vision, and were unaware of the purpose of the experiment. The experiment was approved by the institutional review board (IRB) of Waseda University (2015-033), and conducted in accordance with the ethical standards of the 1964 Declaration of Helsinki. Written informed consent was obtained from all participants in advance.

Stimuli and Procedure

Nine computer-generated East-Asian male faces that varied in a dominance dimorphic cue (SD = −5, −3, −1.5, −0.5, 0, 0.5, 1.5, 3, 5; Figure 4) from obedient to dominant dimensions were created. Facial dominance of males was generated in the same way as those of female faces in Experiment 1. An average-looking East-Asian male face was manipulated in a facial dominance dimension using the average dominance rating scores that 20 Japanese participants (10 male, mean age = 21.40, SD = 1.43) gave to 200 male faces. The face stimulus was presented with the same three background colors as in Experiment 1, and the experimental setting and procedure were identical to those of Experiment 1.

Figure 4.

The morphed male faces in nine levels of dominance SDs with three background colors.

Analysis

The data analysis was identical to that in Experiment 1.

Results

Five participants whose responses failed to fit the psychometric function were removed from the data analysis (s1, s6, s8, s11, s14; mean SD > 2.5, see plots for each participant in Supplemental materials, https://osf.io/xpwek/). Twenty-two participants’ data were used for further analysis. There was no significant difference between the obedience responses and dominance responses, irrespective of the extent of facial dominance manipulation, t(21) = 0.15, p = .88, Cohen's d = 0.03. The proportion of dominance response to each male face stimulus and their psychometric curves with the three background colors are shown in Figure 5.

Figure 5.

The proportion of dominance responses for each male face stimulus by psychometric functions. Colored curves show the logistic fits to the psychometric results for the three background colors. The violin plot showing the distribution of individual PSEs and JNDs in the three background colors are presented. The horizontal line represents the mean. Error bars represent the standard mean error. Asterisks indicate significant differences (*p < .05; **p < .01; after Bonferroni correction).

A one-way ANOVA analysis of the PSE showed a significant main effect of background color (Figure 5), F(2, 42) = 9.39, p = .0004, η_p²= 0.10. The PSEs with red background color (Mean =−0.28, SD = 0.62) were significantly smaller than with green background color (Mean = 0.20, SD = 0.71; t(21) = 3.71, Bonferroni corrected p = .003, Cohen's d = 0.79), and gray background color (Mean = 0.10, SD = 0.61; t(21) = 3.18, Bonferroni corrected p = .01, Cohen's d = 0.68). No difference was observed in PSEs between the gray and green background colors (t(21) = 0.97, Bonferroni corrected p = 1, Cohen's d = 0.21). Thus, a male face presented with a red background color was more likely to be perceived as dominant than one presented with gray or green background colors. A one-way ANOVA analysis of JNDs showed that there was no significant difference between the three background colors (Figure 5), F(2, 42) = 2.01, p = .15, η_p²= 0.04. Thus, the sensitivity of face dominance perception is little influenced by the background colors.

As a further test of population categorical responses, these were fitted using a GLMM. Results showed no significant interaction effect between the background color and dominance levels (i.e., 9 levels of SDs) of male faces, x(16)²= 5.81, p = .99. After removing the interaction effect from the model analysis, a final model including the background color and face dominance levels as fixed factors, and intercepts of participants as random factors showed a main effect of background color, x(2)²= 46.03, p < .001, and face dominance levels, x(8)²= 1674.51, p < .001. Participants made more dominance responses when the face stimulus was presented in red background colors than in gray (p < .001), and green background colors (p < .001). No difference was observed between the gray and green background colors (p = .42).

Those results replicated the findings in Experiment 1 that red background color increases the possibility of faces to be perceived as dominance, compared with gray and green background colors. Thus, red background color biases towards a dominant face perception for both female and male faces.

The effect of a red background on enhancing the perception of dominance in human faces may stem from a strong red-dominance association rooted in evolutionary biology and sociocultural learning. To determine whether this red effect arises more from perceptual evolutionary roots or semantic learning factors, we designed two questionnaire surveys to further explore the reasons underlying this red-dominance association. In the first survey, we tested the red-dominance association with simple geometric shapes to examine whether the observed effect of a red background on enhancing the perception of dominance is specific to human faces or whether it could also influence the perception of dominance in geometric shapes. We hypothesized that a red background could also enhance the perception of dominance in geometric shapes. In the second survey, we replaced the background colors with the semantic words representing those colors, to explore whether the observed effect of red on dominance was specific to the perceptual experience of the color red or could also be influenced by the semantic associations of color words. We hypothesized that the effect of red might be restricted to the perceptual experience of color and would not bias the perception of dominance in semantic contexts.

Questionnaire Survey 1—Effect of Background Color on Shape Dominance

Ten basic geometric shapes (Figure 6a) were presented against red, green, and gray background colors, generating 30 visual stimuli (see an example in Figure 6b). A total of 61 participants (mean age = 19.38 years, SD = 1.19) recruited mainly from Waseda University were instructed to rate the dominance of the shape stimuli using a seven-point Likert scale. Data was collected using Qualtrics Survey via online platforms.

Figure 6.

(a) Visual stimuli used in the questionnaire; (b) An example of visual stimulus with an oval in red background in survey 1; (c) An example of visual stimulus with an oval presented with the word of red (あか) in survey 2.

The mixed-effect model analysis showed that there was no significant interaction between background colors and geometric shapes, x(18)²= 9.77, p = .94. Model analysis showed significant main effects of background colors, x(2)²= 447.55, p < .001, and shapes, x(9)²= 247.72, p < .001. Multiple comparisons showed that shapes presented in red (mean = 4.47, SD = 1.61) background colors were rated as more dominant than gray (mean = 3.26, SD = 1.55) and green (mean = 3.05, SD = 1.47) background colors (ps < .01; Figure 7, color).

Figure 7.

Mean rating of dominance for geometric shapes presented in colored background and background with words representing colors. Error bar represents error of the mean. Asterisks indicate significant differences (**p < .01).

The results from Questionnaire Survey 1 indicated that the effect of red background color on the perception of dominance was not limited to human faces, as geometric shapes were also rated as more dominant when presented on a red background compared to green and gray backgrounds. This suggests that the association between red and dominance can influence the perception of dominance for both faces and objects. To further investigate whether the observed effect of red on dominance is specific to the perceptual color itself or influenced by the meaning of color, another online questionnaire survey was conducted. In this survey, instead of using the actual perceptual colors, semantic words representing colors were used to examine the effect of color meaning on shape dominance.

Questionnaire Survey 2—Effect of Words of Color on Shape Dominance

The visual stimuli consisted of the same geometric shapes accompanied by Japanese words representing the three colors: あか (red), みどり (green), and はいいろ (gray). These color words were presented in black ink, both above and below the shapes (see an example in Figure 6c). 67 newly recruited participants (mean age = 26.39 years, SD = 6.68) were instructed to rate the dominance of the shape stimuli using a seven-point Likert scale.

Model analysis showed that there was a significant interaction effect between the word of colors and shapes, x(18)²= 77.61, p < .01. A significant effect of shape, x(9)²= 27.51, p = .001, and no significant effect of word of colors, x(2)²= 2.81, p = .25 were observed. Thus, there was no significant difference between the three words of colors (red: mean = 3.88, SD = 2.03; gray: mean = 3.84, SD = 1.84; green: mean = 3.74, SD = 1.81) on the rating of shape dominance (ps > .05; see Figure 7, word).

When the background color was replaced with words representing the meaning of colors, the effect of red on the perception of dominance in shape was diminished. This suggests that the effect of red on the perception of dominance is specific to the perceptual color itself.

Discussion

The present study investigated the effect of red background color on the perception of dominance in both human faces (male and female faces) and geometric shapes. The results showed that when a morphed face stimulus was presented with a red background, it was more likely to be perceived as dominant compared to when it was presented with green or gray backgrounds. Additionally, geometric shapes presented on a red background were also rated as more dominant than those on green and gray backgrounds. Therefore, it can be concluded that a red background enhances the perception of dominance for both human faces and geometric shapes. Furthermore, the effect of a red background on the perception of dominance in shapes is specifically linked to the perceptual experience of the color, as the use of semantic words representing colors failed to induce a similar bias. Our findings provide further evidence for the enhancing effect of red on dominance in both animals and humans (Elliot et al., 2009; Hill & Barton, 2005; Setchell & Dixson, 2001; Stephen et al., 2012; Wiedemann et al., 2015; Young et al., 2013), suggesting evolutionary roots of dominance associated with the perceptual experience of the color red.

When used as a background color for presenting ambiguous face stimuli with varying levels of dominance, red interacted with facial dominance features, leading to a higher likelihood of perceiving the faces as dominant compared to green and gray backgrounds. The psychometric function analysis, which involved fitting dominance responses for both female and male faces (Figures 3 and 5), demonstrated that the presence of a red background altered the category boundary along the facial dominance continuum, shifting it away from less dominance perception and increasing the probability of perceiving dominance. Thus, the red background interrupted and modulated the processing of facial dominance, leading to a bias toward perceiving a more dominant facial impression. This interference effect aligns with previous research that has demonstrated the influence of red color on face perceptions (Minami et al., 2018; Sivananthan et al., 2021; Stephen et al., 2012; Young et al., 2013). Studies have proposed that facial expressions, including emotions, are inherently ambiguous, and contextual information can easily influence the perception of faces, shifting the perception in one direction or another (Hassin et al., 2013). Specifically, contextual colors such as a red background, can influence the perception of emotional facial expressions (Gil & Le Bigot, 2015; Young et al., 2013). Thus, the present findings provide further evidence of the influence of red on facial perceptions, indicating a strong congruent effect of the red-dominance association that can be automatically activated to bias dominance perception for both female and male faces.

Moreover, previous research indicates that red facilitates the processing and categorization of angry faces, and faces that are redder or presented with a red background are more likely to be perceived as expressing anger (Drummond & Quah, 2001; Ruba et al., 2021; Peromaa & Olkkonen, 2019; Young et al., 2013, 2018). Dominant faces share similarities with angry faces in terms of their basic shape features (Said et al., 2009). It might be possible that the red-dominance association may stem from and be mediated by the perception of an angry expression, leading to the red effect on facial dominance impressions.

Furthermore, the results from the questionnaire survey demonstrated that the effect of a red background extends beyond facial dominance perception and also enhances the perception of dominance for geometric shapes. This indicates that the effect of red on dominance is not limited to facial stimuli but can also generalize to non-face objects. However, when the red background color was replaced with semantic words representing colors, the effect of red on dominance perception diminished. This implies that the specific perceptual experience of the color red plays a crucial role in augmenting dominance perception, rather than the mere semantic association with the word “red” itself.

According to the color-in-context theory proposed by Elliot and Maier (2012), the strong association between red and dominance may have both evolutionary roots and socially relevant properties. The effect of red-dominance has been observed across human cultures and in many animal species. For example, in primates like mandrills and rhesus macaques, males with more intense red coloration often signal higher testosterone levels and physical dominance, making them more intimidating to rivals and more attractive to potential mates (Hill & Barton, 2005; Khan et al., 2011; Setchell & Jean Wickings, 2005). In humans, red is also considered a testosterone-based biological cue that elicits dominance, aggression, and enhances competitive behavioral performance (Elliot et al., 2007; Hill & Barton, 2005). When people are angry or aroused, their faces often flush red due to increased levels of oxygenated hemoglobin in the blood. This visual cue may have evolved as a way to signal heightened emotional states or readiness for confrontation (Changizi et al., 2006; Pryke, 2009; Montoya et al., 2005). Through evolutionary context, people have learned to associate the occurrence of the signal of red with dominance behaviors, establishing the red-dominance association. Red visual cues, such as red clothing and red backgrounds, are believed to enhance perceptions of dominance, aggression, anger, and competitive ability (Attrill et al., 2008; Dreiskaemper et al., 2013; Elliot et al., 2010; Fetterman et al., 2011, 2012; Hill & Barton, 2005; Peromaa & Olkkonen, 2019; Sorokowski et al., 2014; Wiedemann et al., 2015; Young et al., 2013). Studies in sports competitions have shown that individuals wearing red have a higher probability of winning contests compared to those wearing blue (Attrill et al., 2008; Dreiskaemper et al., 2013; Hagemann et al., 2008; Hill & Barton, 2005; Piatti et al., 2012). In human social contexts, red is commonly associated with power, strength, competitive, dominance, and social status (Elliot et al., 2010; Feltman & Elliot, 2011; Williams et al., 1970; Wen et al., 2022; Wu et al., 2018). Historical and cultural examples illustrate the symbolic significance of red in various societies. For instance, in ancient China, only officials in high social positions were permitted to wear red attire (Jiang et al., 2016). In Western culture, red has historically been associated with nobility and rank, often worn by kings, cardinals, and judges (e.g., the coat of arms of a cardinal is indicated by a red galero; Elliot et al., 2010; Wu et al., 2018). Wu et al. (2018) found that both Chinese and British participants associated red with power and social status. Moreover, studies on color connotation show that red is perceived as dominant, aggressive, and dangerous, while green and blue are associated with safety and peacefulness (Chen et al., 2020; Harashima, 2016; Williams et al., 1970). These findings suggest that the observed effect of red-dominance association may stem from biological heritage, further enhanced through social interactions and language learning.

The red-dominance association may also be rooted in the anatomy of the human retina and our color processing mechanisms. Approximately 35 million years ago, catarrhine primates (Old World monkeys and apes, including humans) developed trichromatic vision, enabling them to detect ripe fruits and nutrient-rich foliage with heightened sensitivity to red and orange hues, which provided a significant evolutionary advantage in their frugivorous and arboreal lifestyle. Notably, nearly two-thirds of the cones in the human retina process longer light wavelengths, including reds, oranges, and some yellows (Regan et al., 2001). This makes the retina particularly sensitive to red light compared to other colors, rendering red more vivid and attention-grabbing. The heightened perception of red extends beyond foraging, influencing emotional and social responses, such as recognizing emotional states through flushed skin and perceptions of dominance, aggression, and attraction. Future studies could explore the effects of orange and yellow hues on psychological functions, as these colors also played significant roles in foraging and may theoretically elicit similar attention and dominance effects. It could help clarify the extent to which different wavelengths contribute to dominance perception and determine whether the evolutionary advantages associated with these colors persist in modern contexts.

In summary, the current study provides evidence that a red background enhances the perception of dominance in both human faces and geometric shapes. Importantly, the study underscores the significance of the specific perceptual experience of the color red, suggesting a biological basis for the red-dominance association. Future research is needed to delve deeper into the mechanisms underlying the formation of these associations. For example, investigating the developmental aspects among infants and children could shed light on the origins and development of red-dominance associations. Additionally, examining cultural differences can offer valuable insights into how cultural factors influence the formation of these associations. Further exploration of different perceptual dimensions of the color red, such as hue, lightness, and saturation, may help disentangle the specific color characteristics that contribute to perceptions of dominance, anger, health, and attractiveness. By addressing these research avenues, we can enhance our understanding of the red-dominance association and its implications across various domains of human perception, cognition, and behavior.

Supplemental Material

sj-docx-1-evp-10.1177_14747049241284602 - Supplemental material for Red Backgrounds Enhance Dominance in Human Faces and Shapes

Supplemental material, sj-docx-1-evp-10.1177_14747049241284602 for Red Backgrounds Enhance Dominance in Human Faces and Shapes by Na Chen, Yidie Yang, Maiko Kobayashi, Koyo Nakamura and Katsumi Watanabe in Evolutionary Psychology