The Morphometrics of “Masculinity” in Human Faces (2024)

Journal List
PLoS One
PMC4324773

As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsem*nt of, or agreement with, the contents by NLM or the National Institutes of Health.
Learn more: PMC Disclaimer | PMC Copyright Notice

PLoS One. 2015; 10(2): e0118374.

Published online 2015 Feb 11. doi:10.1371/journal.pone.0118374

PMCID: PMC4324773

PMID: 25671667

Philipp Mitteroecker,^1,^* Sonja Windhager,² Gerd B. Müller,¹ and Katrin Schaefer²

Pasquale Raia, Academic Editor

Author information Article notes Copyright and License information PMC Disclaimer

Associated Data

Data Availability Statement

Abstract

In studies of social inference and human mate preference, a wide but inconsistent array of tools for computing facial masculinity has been devised. Several of these approaches implicitly assumed that the individual expression of sexually dimorphic shape features, which we refer to as maleness, resembles facial shape features perceived as masculine. We outline a morphometric strategy for estimating separately the face shape patterns that underlie perceived masculinity and maleness, and for computing individual scores for these shape patterns. We further show how faces with different degrees of masculinity or maleness can be constructed in a geometric morphometric framework. In an application of these methods to a set of human facial photographs, we found that shape features typically perceived as masculine are wide faces with a wide inter-orbital distance, a wide nose, thin lips, and a large and massive lower face. The individual expressions of this combination of shape features—the masculinity shape scores—were the best predictor of rated masculinity among the compared methods (r = 0.5). The shape features perceived as masculine only partly resembled the average face shape difference between males and females (sexual dimorphism). Discriminant functions and Procrustes distances to the female mean shape were poor predictors of perceived masculinity.

Introduction

Many studies have investigated whether the phenotypic “masculinity” of men, especially in facial appearance, plays a role in human mate preference and social perception (e.g., [1–3]). In these contexts, masculinity is assumed to be determined by the expression of steroid hormones, such as testosterone, in development. Since these hormones are immunosuppressive, a masculine phenotype is considered an honest signal of immunocompetence and mate quality (e.g., [4]). Yet, empirical evidence for the association between phenotypic facial masculinity, perceived masculinity, and attractiveness in humans remains scarce and inconsistent [5–10]. Body height has been found to play an important role in studies of social inference and mate choice [11,12]. Hence, facial allometry—aspects of face shape reflecting body size—is likely to contribute to the perception of facial masculinity [13].

Early attempts to construct “masculine faces” for rating studies were mainly based on manual image manipulation, without the use of empirical morphometric data (e.g., [14–16]). Other studies, based on simple morphometric approaches (using linear distances and distance ratios), intended to provide a quantitative measure of masculinity (a “masculinity score”) for actually measured faces (e.g., [17,18]). Modern morphometrics and image analysis have opened an array of new possibilities for multivariate studies of the perception of human faces and bodies (e.g., [19,20–22]). Among these, geometric morphometrics allows for a powerful visualization of shape differences and for the construction of artificial faces based on statistical estimates. The wide range of methods that have been used to generate continuous masculinity scores from morphometric variables differ considerably in their statistical and biometric properties. In many of these approaches, the underlying biological and psychological concepts of masculinity were not properly distinguished and the statistical techniques remained unjustified (see below).

In this paper we outline a strategy for estimating biologically and statistically meaningful notions of masculinity. We then critically review other morphometric approaches to masculinity in the current literature and discuss possible biological and psychological interpretations (see Table 1 for a summary). Finally, we demonstrate these methods by an application to a dataset of human facial photographs—still the most common data source in face research [23].

Table 1

A selection of different masculinity concepts and their statistical properties.

Masculinity concept	Reference data	Computation	Interpretation
Perceived masculinity(e.g., [22,29])	Rating by naïve subjects	Multivariate regression of morphometric variables on the masculinity rating	Morphological pattern driving the masculinity rating
Hormone-mediated masculinity(e.g., [22])	Measurement of sex steroid levels (postnatally: salivary, blood; prenatally: approximated by 2D:4D)	Multivariate regression of morphometric variables on hormone level or 2D:4D	Morphological effect of the measured hormone
Average morphological sexual dimorphism (e.g., [3,19,53,54,])	Average male and average female morphology	Difference between male and female mean shapes	Average morphological effect of sex chromosomes (XY, XX)
Allometric versus non-allometric sexual dimorphism (e.g., [32, 38])	Average male and average female morphology; a measure of size of the investigated structures	Regression of morphology on both size and sex	Sexual dimorphism in shape resulting from dimorphism in size versus size-independent dimorphism
Sum of standardized dimorphic traits(e.g., [17,18,42])	Prior selection of dimorphic traits; standard deviation of each variable	Sum of standardized measurements	Variables with low sexual dimorphism have high weights; no obvious multivariate biometric interpretation
Linear discriminant function(e.g., [7,20])	Average male and average female morphology; within-sex covariance matrix	Mean difference vector multiplied by the inverse within-sex covariance matrix	Classification technique based on dimorphic variables with low variance within the sexes
Deviation from female mean shape(e.g., [43])	Average female morphology	Masculinity scores are given by the Procrustes distance between each shape and the female mean shape; no corresponding axis in shape space	Deviation from female mean shape in all directions of shape space, including non-dimorphic features

Open in a separate window

The morphometrics of masculinity

Chromosomal sex is a binary property. Except for chromosomal aberrations, a person is either male (XY) or female (XX). “Masculinity”, by contrast, is a continuous property that refers to two different concepts. First, it can designate the individual variation of shape features that differ on average between the sexes. Such sexually dimorphic features are commonly referred to as secondary sex characteristics (such as body height, skeletal structure, muscle mass, facial and body hair, mandibular prominence), which typically are mediated by sex steroids. In this paper we use the term maleness (or femaleness) in the sense of a continuous score that reflects the individual expression of sexually dimorphic features. Second, the term masculinity can designate perceived masculinity—a graded property ascribed by one or several observers.

Whereas maleness is a purely morphological concept, perceived masculinity is a psychological concept—a mental construct—that is typically assessed via rating studies. It is an empirical question to which degree perceived masculinity resembles maleness. Multiple studies reported that variation in human facial secondary sex characteristics correlates with ratings such as masculinity/femininity, dominance, and attractiveness, as well as with estimates of hormone status (e.g., [19,24–26]). Cross-cultural agreement for trait attributions to faces is high and even generalizes to face-like inanimate objects (e.g., [27]).

Morphometric studies of facial masculinity thus need to distinguish between (a) average sexual dimorphism in face shape, (b) individual variation of sexually dimorphic shape features within the sexes, and (c) shape features affecting perceived masculinity.

The morphological pattern that drives, on average, a psychological rating such as perceived masculinity, can be estimated by a multivariate regression of morphological traits on the rating scores (e.g., [22,28,29]). Suppose we have p morphometric measurements x₁, x₂, …, x_p of n human faces. Regressing the morphometric variables (e.g., Procrustes shape coordinates) on a masculinity rating results in a vector of regression slopes that comprises the p univariate regression slopes b₁, b₂, …, b_p. One can compute faces with different perceived masculinity by adding corresponding multiples of the vector to a reference configuration. In geometric morphometrics and several image analysis approaches, these estimated morphological configurations can be displayed as actual shapes or images [21,28,30]. The vector of regression slopes also serves as an axis in the p-dimensional data space, and the n coordinates or scores of the assessed individuals along this axis can be computed as the linear combination (weighted sum) b₁x₁ + b₂x₂ + … + b_px_p. When standardizing the vector of regression coefficients to unit length, the scores can be interpreted as the orthogonal projections of the data points onto the vector (regression scores). We refer to these scores as masculinity shape scores.

The vector of multivariate regression slopes can be interpreted as a linear gradient in shape space that contrasts shapes with low versus or high masculinity rating. The scores along this gradient are a linear combination of the shape variables, weighted by their covariance with the rating. This linear combination of morphometric variables thus has the maximum possible covariance with the rating (it can be considered a two-block partial least squares analysis where one block consists of a single variable only). If the rating lacks a proper scale, the vector of multivariate regression slopes may be substituted by the difference vector between average shapes of high-rated and low-rated individuals. In contrast to a multivariate regression, partial regression coefficients from a multiple regression of the rating on morphology are more difficult to interpret and often computationally unstable (see below).

Multivariate regression can also be used to estimate the morphological pattern induced by the expression of androgenic hormones, such as testosterone (e.g., [5,26]), but a meaningful quantification of hormone levels is difficult. Testosterone concentration in adults may not be a reliable indicator of hormone activity during development. Other authors instead used the 2D:4D ratio (the length of the index finger relative to the length of the ring finger) as a proxy of prenatal testosterone exposure and based shape regressions on this variable (e.g., [19,26,31]).

The average difference between male and female shapes, which can be considered a regression of morphology on chromosomal sex (Fig. 1), is a vector comprising the p mean differences for all variables. It is an axis in the p-dimensional data space (Fig. 2a) along which scores for the assessed individuals can be computed. These scores are a linear combination of the original variables with weightings equal to the mean difference for each variable (scaled to unit sum of squares). Thus we obtain the linear combination with maximum sexual dimorphism. We refer to these continuous scores as maleness shape scores in order to distinguish them from masculinity shape scores. The maleness shape scores are morphometric estimates of maleness (individual expression of dimorphic features), whereas masculinity shape scores are morphometric estimates of perceived masculinity. Komori et al. called the difference vector between male and female mean shapes the “sex-relevant vector” and the subspace perpendicular to this vector the “sex-irrelevant vectors” [32].

Open in a separate window

Further published approaches to the measurement of masculinity

Brown et al. [20] and Scott et al. [7] proposed a linear discriminant function between males and females to represent “an objective measure of masculinity” ([7], page 8). Under the (often unrealistic) assumption of hom*ogenous variance-covariance matrices, a discriminant function is a tool for maximum likelihood classification. The linear discriminant function between two sexes is the vector for which the squared difference between male and female average scores is a maximum relative to the variation of the scores within the sexes (e.g., [40,41]). Hence, sexually dimorphic traits with large variance within the sexes contribute less to the discriminant function than variables with low within-sex variance. It is not a priori clear that such a score resembles the human perception of masculinity. Computationally, the discriminant function is equal to a regression of sex (as a binary variable) on morphology, which conveys no obvious biological meaning (Fig. 1e; [41]). A considerable problem can be the dependence of the discriminant vector on the exact list of variables. Adding or skipping a variable may modify the coefficients of all other variables. In many morphometric applications, the data are first reduced to a small number of principal components (PCs) before computing a discriminant function, because a stable computation requires many more individuals than variables [41]. But the discriminant function typically depends on the number of selected PCs. Brown et al. [20] and Scott et al. [7] performed a step-wise elimination of principal components in the course of their discriminant function analysis. But this does not circumvent the problem, and, unlike other variables, principal components cannot be arbitrarily selected because of their hierarchical way of computation.

In other studies, the sum of a small selection of supposedly androgen-responsive traits was used for a measure of masculinity. Scheib et al., for example, computed a masculinity index by standardizing (z-transforming) and summing relative lower face length and relative cheekbone width [17]. Penton-Voak et al. [18] and Burriss et al. [42] derived a masculinity index by the summing of several standardized traits (distance measurements) that had significantly higher values in men and subtracting from that sum the traits that had higher values in women. Such indices are linear combinations of the original variables with weights equal to plus or minus the standard deviation of the respective variable. Since the standard deviation comprises both variability within the sexes and between the sexes (i.e., sexual dimorphism), variables with small sexual dimorphism contribute more to this kind of masculinity score than highly dimorphic variables. This is probably not what most researchers intend. The summing of untransformed variables would give an unweighted linear combination to which all selected variables contribute equally, regardless of their actual contribution to sexual dimorphism or some measure of masculinity. Both variants of this approach cannot be used in geometric morphometrics and other multivariate contexts.

Sanchez-Pages and Turiegano suggested using Procrustes distance between the shape of a male individual and the average female shape as a measure of masculinity [43]. Procrustes distance is a measure of overall shape difference, usually approximated by the Euclidean distance between the two superimposed configurations of landmarks (measurement points). An interpretation of this measure as a masculinity score is problematic because it summarizes deviations from average female shape in all directions of shape space, including deviations along the average male-female axis, but also along the directions perpendicular to it (corresponding to shape features with identical averages in both sexes). Furthermore, because Procrustes distance always is positive, males with a more feminine shape than the female average would still have a positive masculinity score.

For a comparison of these methods, we applied them to a small sample of human facial photographs, the most common data source in face research despite the limitations imposed by the two-dimensional representation. We estimated perceived masculinity and sexual dimorphism (maleness), which we decomposed into an allometric and a non-allometric part. Furthermore, we computed a discriminant function and the Procrustes distances between the male shapes and the female average in order to demonstrate the problematic behavior of these two approaches.

Empirical Analysis

Material & Methods

Our sample comprises frontal photographs of 21 Caucasian women (age 20–34 years) and 24 men (age 20–33 years) from the Viennese student population, who were recruited on campus. A camera with 200 mm lens was positioned at the eye height, 3.5 m away from the face. The heads were adjusted according to the Frankfort Horizontal Plane, and a scale bar was placed next to each head. In addition, body height and body weight were measured for each individual. Participation in the study was voluntary and based on written consent. Each participant was informed about the project and the measurement procedure.

For the rating, the male facial photographs were standardized with regard to white balance, contrast, and brightness. The images where then transformed into grey scale and superimposed by a blurred ellipse to disguise contextual information such as hairstyle and clothing. The final stimulus, including the face, the ellipse, and the uniform grey background, was the same size for all faces in order not to provide any direct size cues. Ninety-one age-matched (24.36 ± 3.49 years) female, self-reported heterosexual Caucasian subjects participated in the rating of the male faces. The raters were students and approached at the university of Vienna. No rater reported to have seen any photographed man before. They were asked to judge masculinity using a slider, ranging from “feminine” to “masculine”. The scale was hidden for the participants and consisted of a continuous range from 0 to 80. Only one face was presented at a time in a pseudo-randomized order. There was a minimum of 17 raters for each face.

For the morphometric analysis we digitized 33 anatomical landmarks and 37 semilandmarks on each original image to describe overall facial form (Fig. 3A; see [29] for details). Semilandmarks are points on curves, for which the exact location along the curve cannot be identified and hence is statistically estimated. We used the sliding landmark algorithm for this purpose, which minimizes the bending energy (a measure of local shape difference) between each individual and the sample average [44,45]. After sliding the semilandmarks, the 45 landmark configurations were symmetrized by averaging each configuration with its relabeled reflection [21,46]. Subsequently the landmarks were superimposed by Generalized Procrustes Analysis [13,47].

Open in a separate window

Fig 3

Landmark configuration used for studying face shape and perceived masculinity.

(a) Face with the 33 landmarks (open circles) and 37 semilandmarks (filled circles) used in the morphometric analysis. (b) The shape features determining perceived masculinity are visualized by deformation grids from the mean shape to shapes predicted for deviations of ±20 rating scores from the average.

The face shape patterns relating to perceived masculinity and allometry were calculated by multivariate regressions of the male face shapes on the masculinity rating and body height, respectively. Sexual dimorphism was computed as the difference between average male and average female shape. The non-allometric component of sexual dimorphism was computed as the difference between average male and average female shape after projecting out the allometry vector (projection of the sexual dimorphism vector in the subspace perpendicular to the allometry vector; [37,48]). All shape patterns were visualized using thin plate spline deformation grids [28,49].

Results

The regression of facial shape on rated masculinity indicated that male faces with a higher masculinity attribution tended to have wider faces with a wider inter-orbital distance, a wider nose, thinner lips, and a larger, more rounded lower facial outline (Fig. 3B). This shape pattern, as a vector in shape space, accounted for 19.4% of total facial shape variation in males, and the corresponding masculinity shape scores had a Pearson product-moment correlation with the masculinity rating of 0.50.

Fig. 4 contrasts female and male mean shapes. Men, on average, had thicker and lower positioned eyebrows, relatively smaller eyes, thinner lips, and a more massive and angulated lower jaw than women. These dimorphic shape features, as a vector in shape space, accounted for 15.4% of total variation across all male face shapes, and the individual maleness shape scores along this vector had a correlation with the masculinity rating of 0.26.

Open in a separate window

Fig 4

Sexual dimorphism is visualized by deformation grids between average female and average male facial shape, together with two-fold extrapolations of these shape differences.

Discussion

In the investigation of social inference and human mate preference, a wide but inconsistent array of tools for computing scores of facial masculinity has been devised (see also [50]). Several of these approaches implicitly assumed that the individual expression of dimorphic shape features, which we refer to as maleness, resembles shape features perceived as masculine. We outlined a morphometric strategy for estimating separately the face shape patterns that underlie perceived masculinity and maleness by regressions of shape on rated masculinity and sex, respectively. Geometric morphometrics allows for the computation of shape scores for perceived masculinity and for maleness, as well as for the visualization of these shape patterns and the construction of faces with different degrees of masculinity or maleness.

When we applied these methods to a set of facial photographs, we found that shape features typically perceived as masculine are wide faces with a wide inter-orbital distance, a wide nose, thin lips, and a large and massive mandible. The individual scores for this combination of shape features—the masculinity shape scores—had a correlation with the actual masculinity ratings of 0.5, which clearly exceeds the correlations reported in Sanchez-Pages et al. [50] for a range of univariate and multivariate morphometric estimates of masculinity. In our data, the estimated pattern of face shape thus accounts for 25% of the variation in masculinity ratings. The remaining 75% of variation must be attributable to other factors, including variation in facial texture and hair color. Despite great efforts to photograph all individuals in a standardized way, variation in head posture probably contributes to the unexplained variation (however, it does not seem to have a systematic effect in our data because the shape patterns depicted by the deformation grids are more local and not related to head posture). Furthermore, it is known that the phase of menstrual cycle and other aspects affecting the psychological status influence the perception and preference of male face stimuli ([51], but see [8]). It has also been reported that facial asymmetry is related to measures of masculinity [52], and we find a correlation of 0.42 between rated masculinity and facial fluctuating asymmetry [46] in our data. The effect of fluctuating facial asymmetry on the rating is not accounted for by the regression approach and, hence, contributes to the unexplained variance.

The pattern of sexual dimorphism only partly resembled the shape features of perceived masculinity. The maleness shape scores, i.e., the individual scores along the sexual dimorphism vector, accounted for 7% of variation of the masculinity rating. This is slightly below the correlation of 0.33 (R² = 0.11) reported by Komori et al. [32]. When decomposing sexual dimorphism into an allometric and a non-allometric part, the relation to the masculinity rating was largely driven by the allometric part, despite similar variation of both parts in male faces. The pattern of allometry in face shape, mainly involving the lower face, was similar to the one found in Mitteroecker et al. [13]. Even though the faces were all scaled to the same size for the rating, face shape appears to contain cues to facial size, which we found to be correlated with rated masculinity.

The discriminant function was not successful in predicting rated masculinity, providing for less than 2% of explained variance. Furthermore, the discriminant function considerably depended on the number of selected principal components. The shape features combined by the discriminant function resembled neither the pattern of sexual dimorphism nor the pattern of perceived masculinity. Likewise, the Procrustes distance between male shapes and the female average, as suggested by Sanchez-Pages and Turiegano [43], accounted for only 3.6% of variation in the rating.

To conclude, proper quantification of the influence of biological factors such as size, hormones, immunocompetence, and body composition, on perceived masculinity and attractiveness via facial shape is a promising direction of research for understanding social behavior and the evolution of human mate choice. The distinction between the perceived masculinity of a face and its expression of sexually dimorphic shape characteristics (which we termed maleness) is crucial in this research agenda. Discriminant functions and Procrustes distances to the female mean shape are poor predictors of perceived masculinity. Still, our findings show that it is possible to estimate the shape pattern of sexual dimorphism in human faces, and we can compute scores that are maximally dimorphic (maleness shape scores) in order to assess individual variation of secondary facial sex characteristics. Furthermore, it is possible to compute the shape pattern that corresponds, on average, to high or low perceived masculinity ratings. The individual expressions of this average pattern (masculinity shape scores) provide the best predictor of rated masculinity. These are two principal approaches to the study of sexual dimorphism and perceived masculinity in the human face. It remains to be evaluated to which degree the relationship between face shape and perceived masculinity is indeed linear and which threshold levels exist for this relationship at extreme ranges of variation.

Acknowledgments

We are grateful to the study participants, and appreciated the thoughtful comments of two reviewers and the associate editor.

Funding Statement

This article was supported by the Open Access Publishing Fund of the University of Vienna, which is covering the publication costs. This research was supported by the Focus of Excellence grant “Biometrics of EvoDevo” and the Emerging Field grant “Comparative Human Life History: A Multilevel Approach” from the University of Vienna. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Data Availability

All data files are available from the DRYAD database. The DOI is doi: 10.5061/dryad.cp0p5.

References

1. Cunningham MR, Barbee AP, Pike CL. What do women want? Facialmetric assessment of multiple motives in the perception of male facial physical attractiveness. J Pers Soc Psychol. 1990; 59: 61–72. [PubMed] [Google Scholar]

2. DeBruine LM, Jones BC, Little AC, Smith FG. Are attractive men’s faces masculine or feminine? The importance of controlling confounds in face stimuli. J Exp Psychol Hum Percept Perform. 2010; 36: 751–758. 10.1037/a0016457 [PubMed] [CrossRef] [Google Scholar]

3. Perrett DI, Lee KJ, Penton-Voak IS, Rowland D, Yoshikawa S, Burt DM, et al. Effects of sexual dimorphism on facial attractiveness. Nature. 1998; 394: 884–887. [PubMed] [Google Scholar]

4. Thornhill R, Gangestad SW. Facial sexual dimorphism, developmental stability, and susceptibility to disease in men and women. Evol Hum Behav. 2006; 27: 131–144. [Google Scholar]

5. Penton-Voak IS, Chen JY. High salivary testosterone is linked to masculine male facial appearance in humans. Evol Hum Behav. 2004; 25: 229–241. [Google Scholar]

6. Puts DA. Beauty and the beast: Mechanisms of sexual selection in humans. Evol Hum Behav. 2010; 31: 157–175. [Google Scholar]

7. Scott IML, Pound N, Stephen ID, Clark AP, Penton-Voak IS. Does masculinity matter? The contribution of masculine face shape to male attractiveness in humans. PLoS ONE. 2010; 5: e13585 10.1371/journal.pone.0013585 [PMC free article] [PubMed] [CrossRef] [Google Scholar]

8. Peters M, Simmons LW, Rhodes G. Preferences across the menstrual cycle for masculinity and symmetry in photographs of male faces and bodies. PLoS ONE. 2009; 4: e4138 10.1371/journal.pone.0004138 [PMC free article] [PubMed] [CrossRef] [Google Scholar]

9. Jones B, DeBruine L, Perrett D, Little A, Feinberg D, Law Smith MJ. Effects of menstrual cycle phase on face preferences. Arch Sex Behav. 2008; 37: 78–84. 10.1007/s10508-007-9268-y [PubMed] [CrossRef] [Google Scholar]

10. Little AC, Jones BC, DeBruine LM. Preferences for variation in masculinity in real male faces change across the menstrual cycle: Women prefer more masculine faces when they are more fertile. Pers Indiv Differ. 2008; 45: 478–482. [Google Scholar]

11. Courtiol A, Raymond M, Godelle B, Ferdy J-B. Mate choice and human stature: hom*ogamy as a unified framework for understanding mating preferences. Evolution. 2010; 64: 2189–2203. 10.1111/j.1558-5646.2010.00985.x [PubMed] [CrossRef] [Google Scholar]

12. Pawlowski B, Dunbar RIM, Lipowicz A. Evolutionary fitness: Tall men have more reproductive success. Nature. 2000; 403: 156 [PubMed] [Google Scholar]

13. Mitteroecker P, Gunz P, Windhager S, Schaefer K. A brief review of shape, form, and allometry in geometric morphometrics, with applications to human facial morphology. Hystrix. 2013; 24: 59–66. [Google Scholar]

14. Sternglanz SH, Gray JL, Murakami M. Adult preferences for infantile facial features: An ethological approach. Anim Behav. 1977; 25: 108–115. [PubMed] [Google Scholar]

15. Wade TJ, Dyckman KA, Cooper M. Invisible men: Evolutionary theory and attractiveness and personality evaluations of 10 African American male facial shapes. J Black Psychol. 2004; 30: 477–488. [Google Scholar]

16. Swaddle JP, Reierson GW. Testosterone increases perceived dominance but not attractiveness in human males. Proc R Soc Lond B Biol Sci. 2002; 269: 2285–2289. [PMC free article] [PubMed] [Google Scholar]

17. Scheib JE, Gangestad SW, Thornhill R. Facial attractiveness, symmetry and cues of good genes. Proc R Soc Lond B Biol Sci. 1999; 266: 1913–1917. [PMC free article] [PubMed] [Google Scholar]

18. Penton-Voak IS, Jones BC, Little AC, Baker S, Tiddeman B, Burt DM, et al. Symmetry, sexual dimorphism in facial proportions and male facial attractiveness. Proc R Soc Lond B Biol Sci. 2001; 268: 1617–1623. [PMC free article] [PubMed] [Google Scholar]

19. Fink B, Grammer K, Mitteroecker P, Gunz P, Schaefer K, Bookstein FL, et al. Second to fourth digit ratio and face shape. Proc R Soc Lond B Biol Sci. 2005; 272: 1995–2001. [PMC free article] [PubMed] [Google Scholar]

20. Brown WM, Price ME, Kang J, Pound N, Zhao Y, Yu H. Fluctuating asymmetry and preferences for sex-typical bodily characteristics. Proc Natl Acad Sci U S A. 2008; 105: 12938–12943. 10.1073/pnas.0710420105 [PMC free article] [PubMed] [CrossRef] [Google Scholar]

21. Mitteroecker P, Gunz P. Advances in geometric morphometrics. Evol Biol. 2009; 36: 235–247. [Google Scholar]

22. Schaefer K, Mitteroecker P, Fink B, Bookstein FL. Psychom*orphospace—from biology to perception, and back: Towards an integrated quantification of facial form variation. Biological Theory. 2009; 4: 98–106. [Google Scholar]

23. Sforza C, de Menezes M, Ferrario V. Soft- and hard-tissue facial anthropometry in three dimensions: what’s new. J Anthropol Sci. 2013; 91: 159–184. 10.4436/jass.91007 [PubMed] [CrossRef] [Google Scholar]

24. Keating CF. Gender and the physiognomy of dominance and attractiveness. Soc Psychol Quart. 1985; 48: 61–70. [Google Scholar]

25. Neave N, Laing S, Fink B, Manning JT. Second to fourth digit ratio, testosterone and perceived male dominance. Proc R Soc Lond B Biol Sci. 2003; 270: 2167–2172. [PMC free article] [PubMed] [Google Scholar]

26. Schaefer K, Fink B, Mitteroecker P, Neave N, Bookstein FL. Visualizing facial shape regression upon 2nd to 4th digit ratio and testosterone. Coll Antropol. 2005; 29: 415–419. [PubMed] [Google Scholar]

27. Windhager S, Bookstein FL, Grammer K, Oberzaucher E, Said H, Slice DE, et al. "Cars have their own faces": cross-cultural ratings of car shapes in biological (stereotypical) terms. Evol Hum Behav. 2012; 33: 109–120. [Google Scholar]

28. Bookstein FL. Morphometric tools for landmark data: Geometry and biology. New York: Cambridge University Press; 1991. [Google Scholar]

29. Windhager S, Schaefer K, Fink B. Geometric morphometrics of male facial shape in relation to physical strength and perceived attractiveness, dominance, and masculinity. Am J Hum Biol. 2011; 23: 805–814. 10.1002/ajhb.21219 [PubMed] [CrossRef] [Google Scholar]

30. Zollikofer CPE, Ponce de Leon MS. Visualizing patterns of craniofacial shape variation in hom*o sapiens. Proc R Soc Lond B Biol Sci. 2002; 269: 801–807. [PMC free article] [PubMed] [Google Scholar]

31. Burriss RP, Little AC, Nelson EC. 2D:4D and sexually dimorphic facial characteristics. Arch Sex Behav. 2007; 36: 377–384. [PubMed] [Google Scholar]

32. Komori M, Kawamura S, Ishihara S. Multiple mechanisms in the perception of face gender: Effect of sex-irrelevant features. J Exp Psychol Human. 2011; 37: 626–633. 10.1037/a0020369 [PubMed] [CrossRef] [Google Scholar]

33. Sylvester AD, Kramer PA, Jungers WL. Modern humans are not (quite) isometric. Am J Phys Anthropol. 2008; 137: 371–383. 10.1002/ajpa.20880 [PubMed] [CrossRef] [Google Scholar]

34. Gould SJ. Allometry and size in ontogeny and phylogeny. Biol Rev. 1966; 41: 587–638. [PubMed] [Google Scholar]

35. Klingenberg C. Heterochrony and allometry: the analysis of evolutionary change in ontogeny. Biol Rev. 1998; 73: 79–123. [PubMed] [Google Scholar]

36. Rosas A, Bastir M. Thin-plate spline analysis of allometry and sexual dimorphism in the human craniofacial complex. Am J Phys Anthropol. 2002; 117: 236–245. [PubMed] [Google Scholar]

37. Mitteroecker P, Gunz P, Bernhard M, Schaefer K, Bookstein FL. Comparison of cranial ontogenetic trajectories among great apes and humans. J Hum Evol. 2004; 46: 679–698. [PubMed] [Google Scholar]

38. Schaefer K, Mitteroecker P, Gunz P, Bernhard M, Bookstein FL. Craniofacial sexual dimorphism patterns and allometry among extant hominids. Ann Anat. 2004; 186: 471–478. [PubMed] [Google Scholar]

39. Rohlf FJ, Bookstein FL.A comment on shearing as a method for “size correction”. Syst Biol. 1987; 36: 356–367. [Google Scholar]

40. Mardia K, Kent J, Bibby J. Multivariate Analysis. London: Academic Press; 1979. [Google Scholar]

41. Mitteroecker P, Bookstein FL. Classification, linear discrimination, and the visualization of selection gradients in modern morphometrics. Evo Biol. 2011; 38: 100–114. [Google Scholar]

42. Burriss RP, Roberts SC, Welling LLM, Puts DA, Little AC. Heterosexual romantic couples mate assortatively for facial symmetry, but not masculinity. Pers Soc Psychol Bull. 2011; 37: 601–613. 10.1177/0146167211399584 [PubMed] [CrossRef] [Google Scholar]

43. Sanchez-Pages S, Turiegano E. Testosterone, facial symmetry and cooperation in the prisoners’ dilemma. Physiol Behav. 2010; 99: 355–361. 10.1016/j.physbeh.2009.11.013 [PubMed] [CrossRef] [Google Scholar]

44. Bookstein FL. Landmark methods for forms without landmarks: Morphometrics of group differences in outline shape. Med Image Anal. 1997; 1: 225–243. [PubMed] [Google Scholar]

45. Gunz P, Mitteroecker P. Semilandmarks: a method for quantifying curves and surfaces. Hystrix. 2013; 24: 103–109. [Google Scholar]

46. Mardia K, Bookstein F, Moreton I. Statistical assessment of bilateral symmetry of shapes. Biometrika. 2000; 87: 285–300. [Google Scholar]

47. Rohlf FJ, Slice DE. Extensions of the Procrustes method for the optimal superimposition of landmarks. Syst Zool. 1990; 39: 40–59. [Google Scholar]

48. Burnaby TP. Growth-invariant discrimination functions and generalized distances. Biometrics. 1966; 22: 96–110. [Google Scholar]

49. Klingenberg CP. Visualizations in geometric morphometrics: how to read and how to make graphs showing shape changes. Hystrix. 2013; 24: 15–24. [Google Scholar]

50. Sanchez-Pages S, Rodriguez-Ruiz C, Turiegano E. Facial masculinity: how the choice of measurement method enables to detect its influence on behaviour. PLoS ONE. 2014; 9: e112157 10.1371/journal.pone.0112157 [PMC free article] [PubMed] [CrossRef] [Google Scholar]

51. Penton-Voak IS, Perrett DI, Castles DL, Kobayashi T, Burt DM, Murray LK, et al. Menstrual cycle alters face preference. Nature. 1999; 399: 741–742. [PubMed] [Google Scholar]

52. Gangestad SW, Thornhill R. Facial masculinity and fluctuating asymmetry. Evol Hum Behav. 2003; 24: 231–241. [Google Scholar]

53. DeBruine LM, Jones BC, Little AC, Boothroyd LG, Perrett DI, Penton-Voak IS, et al. Correlated preferences for facial masculinity and ideal or actual partner’s masculinity. Proc R Soc Lond B Biol Sci. 2006; 273: 1355–1360. [PMC free article] [PubMed] [Google Scholar]

54. Feinberg DR. Are human faces and voices ornaments signaling common underlying cues to mate value?Evol Anthropol. 2008; 17: 112–118. [Google Scholar]

Articles from PLOS ONE are provided here courtesy of PLOS

The Morphometrics of “Masculinity” in Human Faces (2024)

FAQs

What are the morphometrics of masculinity in human faces? ›

In an application of these methods to a set of human facial photographs, we found that shape features typically perceived as masculine are wide faces with a wide inter-orbital distance, a wide nose, thin lips, and a large and massive lower face.

Show Me More ›

Does masculinity matter the contribution of masculine face shape to male attractiveness in humans? ›

These findings suggest that facial morphological masculinity may contribute less to men's attractiveness than previously assumed.

Explore More ›

Who are masculinized versions of male faces more likely to be preferred by? ›

One-sample t-tests (compared to 50%) demonstrated that women preferred the masculinized face more often than the feminized face when the faces were transformed using the sexual dimorphism method (mean = 67%, t₁₂₃ = 6.85, p < 0.001), the perceived masculinity method (mean = 74%, t₁₂₃ = 10.03, p < 0.001) and the pubertal ...

See Details ›

What are the 4 types of masculinity? ›

Connell developed a theory of masculinity that was first published in 1995. In her renowned book, she lays out the way in which contemporary, Western masculinity is constructed through the following categories: hegemonic masculinity, complicit masculinity, subordinate masculinity, and marginalized masculinity.

Learn More ›

What are examples of male masculinity? ›

Traditional Standards Of Masculinity and Being a Man

Being perceived as “weak” due to crying or showing fear.
Emphasis on physical strength.
Professional Success.
Being the “breadwinner” of the family.
Exerting dominance in relationships.
Being assertive.

More items...

Apr 8, 2021

Read On ›

Is a masculine face more attractive? ›

This can mean that attractiveness and facial masculinity tend to have a threshold-like correlation, rather than a linear or curvilinear one. In other words, women show a preference for faces with higher masculinity, if the faces have low masculinity levels.

What is the most masculine face shape for men? ›

Square If your forehead, cheekbone, jawline, and face length measurements are similar, you have a square face shape. This face shape is considered the masculine ideal because it features a sharp, angular jawline with defined cheekbones and a wide forehead.

Get More Info Here ›

Do men prefer feminine faces? ›

Although young men in their 20s typically judge facial femininity as more attractive than facial masculinity, at all ages, men with higher sexual desire and testosterone levels tend to show a marked preference for feminine faces.

Get More Info ›

Which facial type is more trustworthy masculine or feminine? ›

Interestingly, females with more feminine faces elicited less trust in both male and female partners, while males with more masculine facial shape were more trusted by females, but less trusted by males.

Discover More ›

What is a masculine face trait? ›

The width of a masculine face is the same across the cheekbones as it is across the angle of the jaw (with the jawline sometimes slightly larger). The masseter muscle (the muscles around the jaw used for chewing) are bulkier and the chin is wider. Facial features are crisp and angular; cheeks are 'chiselled'.

Find Out More ›

Which gender is better at recognizing faces? ›

Past studies have shown that women typically perform better than men in facial-recognition tests.

What are portrayals of masculinity? ›

Masculinities in the Media

In media targeting adults, male characters reflect societal expec- tations regarding masculinities and are more often portrayed as unemotional, violent, tough, dominant, and aggressive compared to women (Collins, 2011; Scharrer & Blackburn, 2018; Zeglin, 2016).

Read The Full Story ›

What are the hallmarks of masculinity? ›

Standards of manliness or masculinity vary across different cultures, subcultures, ethnic groups and historical periods. Traits traditionally viewed as masculine in Western society include strength, courage, independence, leadership, and assertiveness.

Show Me More ›

What are the conceptualizations of masculinity? ›

In sum, masculinities can be conceptualized as what people believe about themselves (Domain 1) and others (Domain 2), situational cues (Domain 3), what people do in social interactions (Domain 4), as well as ecological influences on groups and societies (Domain 5).

What is the perception of masculinity? ›

Masculine identities

Masculinity involves displaying attitudes and behaviours that signify and validate maleness, and involves being recognised in particular ways by other men and women.

Discover More Details ›