- Орангутаны научились использовать орудия с выгодой для себя [2019-03-22]
- Мозг собак отличил настоящие слова от тарабарщины [2019-03-22]
- Какаду Гоффина выклевали палочки из картонки и использовали их в быту [2019-03-22]
- Орангутанов уличили в создании крючков [2019-03-22]
- Любовь самок гуппи к ярким самцам объяснили генами и освещением [2019-03-22]
- Орангутаны рассказывают друг другу о прошлом [2019-03-22]
- Блеск и нищета этологии [2018-05-03]
- Почему хозяева похожи на своих собак? [2018-03-15]
- Интервью с Анатолием Протопоповым [2019-02-13]
- Цирки: развлечение vs мучение [2018-12-15]
• 1.1. The ratio of second-to-fourth digit (2D:4D)
• 1.2. Mental Rotation ability
• 2.2.2. Measures and questionnaires
• 2.3.3. High and low probability of conception
• 3.1. Male facial preferences
• 3.3. Measures of individual differences
• 3.3.2. Relationships between femininity/masculinity, 2D:4D, and MR ability
• 3.4. Individual differences in facial preferences
• 3.4.1. LTM and STM difference and 2D:4D
• 3.4.2. Attractive male face as a function of menstrual phase and 2D:4D
• 3.4.3. Attractive male face as a function of menstrual phase and psychological femininity (BSRI)
• 3.5. Menstrual regularity and 2D:4D
• 3.6. Parental Bonding Instrument and 2D:4D
• 3.7. Longest intimate relationship (LIR) and 2D:4D
• 3.8. ATM, STM, LTM, and average male faces
• 4.1. Attributes of preferred faces
• 4.1.1. Attractive versus average faces
• 4.1.2. Attractive versus healthy male
• 4.2. Attributes of the perceiver
1. Introduction
The standard model of sexual differentiation of the mammalian body and brain maintains that the “default” sex is female, and maleness is a consequence of early androgen exposure that both defeminizes and masculinizes the developing embryo (Jost, 1972, Woodson & Gorski, 2000). The human literature examining the effects of androgen insensitivity (Masica et al., 1971, Wisniewski et al., 2000), congenital adrenal hyperplasia (Hampson et al., 1998, Hines, 2000, Resnick et al., 1986), idiopathic hypogonadotropic hypogonadism (Hier & Crowley, 1982), and Turner's syndrome (Collaer et al., 2002, Hines, 2000) are all consistent with this model. In nonhuman mammals, this androgen-dependent sex reversal is accompanied by a change in sexual preferences (Gorski et al., 1980, Phoenix et al., 1959). That is, the “default” sex (female) is normally attracted to males in adulthood but following early exposure to androgens, the vast majority of the resulting sex (male) is now attracted to females. This reversal in preference may be a consequence of a change in sensitivity to the cues to which each sex is attracted. If early androgens influence human sex preferences, then the degree of early androgen exposure within women and within men may influence both their psychological femininity/masculinity scores and the degree to which they are attracted to sexually selected characteristics displayed on the faces and bodies of the opposite sex. That is, individual differences in preferences for the opposite sex may be attributable, at least in part, to the degree to which the different individuals' brains were defeminized/masculinized by the organizational effects of early androgens. The current study evaluates this proposition by examining how the ratio of second-to-fourth digit (2D:4D) and mental rotation (MR) ability, potential indices of early androgen exposure, are related to a woman's preferences for male faces and her femininity/masculinity scores on the Bem Sex-Role Inventory (BSRI; Bem, 1981). Differences in facial preferences as a function of menstrual phase and parental bonding questionnaire scores are also examined.
1.1. The ratio of second-to-fourth digit (2D:4D)
The 2D:4D is a sexually dimorphic trait that has been known for over 100 years (Baker, 1888), with men, on average, having lower 2D:4Ds than women do. Unlike sexually dimorphic characteristics that appear at puberty, 2D:4D is established by the 13th week after conception (Garn, Burdi, Babler, & Stinson, 1975). Based on a large number of studies, Manning (2002) has concluded that 2D:4D is positively associated with prenatal estrogen and negatively associated with prenatal testosterone. There is also evidence for population differences in 2D:4D that are significantly larger than the between-sex differences within any population (Manning et al., 2000). To avoid the possible confounding effects of population differences, the current study examines the relationship between 2D:4D and mate choices in a group of women of European (Anglo) descent.
1.2. Mental Rotation ability
MR ability is a second sexually dimorphic trait that may also serve as an index of early androgen exposure. Although early studies indicated that the male advantage in this task was not apparent until the onset of puberty (Linn & Petersen, 1985, Petersen, 1976), more recent experiments have demonstrated that male–female differences can be detected as early as 4 years of age (Levine, Huttenlocher, Taylor, & Langrock, 1999) and that MR ability is correlated with prenatal testosterone levels (Grimshaw, Sitarenios, & Finegan, 1995). The current study examines the relationship between MR ability and 2D:4D and explores how each is related to participants' mate choice preferences.
1.3. Menstrual phase
Preference for male faces has been shown to change as a function of the perceiver's menstrual phase (Johnston et al., 2001, Penton-Voak & Perrett, 2000, Penton-Voak et al., 1999). Women tested during the high probability of conception phase of their menstrual cycle have been found to prefer a more masculinized male face than their stated preference outside this fertile window. This observation suggests that the activational effects of menstrual hormones may influence facial preferences (Penton-Voak & Perrett, 2000). However, Johnston et al. (2001) found that the magnitude of this menstrual shift in facial preferences was also a function of women's femininity/masculinity scores on the BSRI. If femininity/masculinity scores reflect the defeminizing/masculinizing effects of early androgens, then women's preferences for masculinized male faces may be a function of both the organizational and activational effects of steroid hormones. To evaluate these hypotheses, the current study examines how women's facial preferences for male faces vary as a function of both their menstrual phase (a potential activational action of steroids) and 2D:4D (a proposed organizational effect of steroids).
1.4. Mating strategies
Cashdan (1996) has argued that women are confronted with a mate choice dilemma in the form of choosing between a “good genes” man, who offers direct benefits to her reproductive success, and a “good dad” man, who may indirectly enhance her reproductive success by providing resources that facilitate offspring survival. These considerations should be reflected in women making a different choice for a short-term mate (STM; brief affair) than a long-term mate (LTM; committed relationship). Several studies have supported this model of female mate choice (Gangestad et al., 2004, Johnston et al., 2001, Penton-Voak & Perrett, 2000). With respect to facial preferences, the model implies that a woman should prefer a more masculine male face when selecting an STM than an LTM. This prediction is based on the finding of Gangestad and Thornhill (2003) that the degree of masculinization of male faces is linearly related to lower fluctuating asymmetries in body measurements, a reliable index of “good genes” across species (Møller & Thornhill, 1997). Facial masculinity appears to be a signal of “good genes”, the dominant consideration in selecting an STM, but not an LTM. The current study evaluates this hypothesis by having female participants select their preferred STM and LTM from male faces that differ in masculinity.
Bowlby, 1969, Bowlby, 1973, Bowlby, 1980, Draper & Harpending, 1982 have proposed that mate choice strategies are closely related to the security of attachment to parents during childhood. For example, girls growing up without fathers have been found to mature earlier, be sexually precocious, have low self-esteem, and have difficulty forming long-term relationships (Chisholm, 1993, Draper & Harpending, 1982, Jones et al., 1972, Moffitt et al., 1992, Surbey, 1990). If this is the case, then low paternal bonding in females may be associated with a “good genes” strategy when selecting both STMs and LTMs. The current study evaluates this mate selection hypothesis by examining participants' facial preferences (STM and LTM) as a function of their scores on the Parental Bonding Instrument (PBI; Parker, Tupling, & Brown, 1979), which measures perceived maternal and paternal bonding.
2. Method
2.1. Participants
Participation was limited to self-identified heterosexual women of European (Anglo) descent (N=41) from 18 to 30 years of age (M=19.6 years) who were enrolled in introductory psychology classes at New Mexico State University. All participants received experimental credit for their voluntary participation.
2.2. Materials
2.2.1. Stimulus
Facial preferences were measured using a 1200 frame QuickTime movie that morphs an extremely masculine male face (Frame 0) into an extremely feminine female face (Frame 1199) passing through the average male and average female faces towards the middle of the movie. This movie was developed to systematically change the major facial features and proportions that characterize human maleness and femaleness following puberty, while maintaining a high degree of facial symmetry. Because such secondary sexual characteristics are mainly a consequence of different levels of pubertal hormones (Grumbach, 2000, Tanner, 1990), this methodology provides a basis for interpreting mate choices in terms of the degree to which hormonal markers are displayed on the faces of men and women. This experimental tool is more fully described in Johnston et al. (2001).
2.2.2. Measures and questionnaires
Two independent measurers used vernier calipers, accurate to 0.01 mm, to measure the index and ring fingers from photocopies made of the palm side of both the right and left hands. Hands were rejected if either of the measurers stipulated that a photocopy was unclear. Using this criterion, right-hand 2D:4Ds were obtained for all 41 participants and left-hand 2D:4Ds were obtained for 37 of those participants. 2D:4Ds were calculated by dividing the length of the index finger (2D) by the length of the same-hand ring finger (4D). The 2D:4D correlations between the two measurers for the right (N=41) and left (n=37) hands were each r=.97. For each hand, the two 2D:4Ds were averaged across the two measurers. The averaged 2D:4Ds for the right (M=0.98) and left (M=0.97) hands were significantly correlated (r=.52). Because the right hand is the hand most often associated with sex steroids (Lutchmaya et al., 2004, Manning et al., 2003, Manning et al., 1998), the averaged 2D:4D ratio for the right hand is used in all analyses unless otherwise stipulated.
MR ability was measured using a computerized version of the 2D MR task developed by Cooper and Shepard (1973). This task used 20 pairs of F-like figures, with one member of each pair rotated to varying degrees from the other member and both members rotated with respect to vertical. One member of each pair of Fs was either the mirror image of the other (different) or nonreflected (same). During the task, participants were exposed to 10 “same” and 10 “different” trials presented in a random order. They were instructed to indicate whether the two Fs presented on each trial were the “same” or “different” (i.e., mirror images) by striking either the “s” or the “d” key on a computer keyboard, respectively. Participants were told to be as fast and as accurate as possible. The computer recorded the number of errors and the average response time in milliseconds for all correct responses. To validate sex differences in this MR task, response times for the participants in the current study were compared with those from a parallel study examining mate choice in 46 Anglo men [Vonnahme, 2003; [2D:4D ratios and MR reaction time for Hispanic and Anglo men]. Unpublished raw data). Men had faster response times (M=1867 ms) than the women did [M=2137 ms; t(85)=2.50, p=.01].
The 60-item BSRI (Bem, 1981) consisting of 20 masculine, 20 feminine, and 20 filler traits was used as a measure of the participants' feminine/masculine traits. Self-ratings on the 20 masculine and 20 feminine items have been used to classify individuals as psychologically masculine, feminine, androgynous, or undifferentiated. Bem (1981) reports the test–retest reliabilities ranging from .76 to .94, and internal consistencies ranging from .75 to .90.
Participants' parental bonding was measured using the PBI developed by Parker et al. (1979). This 25-item questionnaire measures two important dimensions of the perceived parent–child relationship, care/rejection, and control/autonomy, which can be combined to produce individual bonding scores for each parent. Parker et al. report test–retest and split half reliabilities of .76 to .88 for the care/rejection dimension, and .63 to .74 for the control/autonomy dimension. Construct and predictive validity measures of .51 to .78 have been demonstrated in several studies (Klimidis et al., 1992, Parker, 1983, Parker et al., 1979), and, by comparing parental interviews to their children's PBI scores, Parker (1981) established the PBI as a valid measure of both perceived and actual parental bonding.
Participants also completed a personal history questionnaire, which included items pertaining to their sexual orientation, age, duration of longest intimate relationship (LIR), and whether the participant was using hormonal contraception or other steroids at the time of testing. A final questionnaire requested information about the participant's menstrual cycle, including length, regularity, age at menarche, and date of onset of their most recent menses. Participants were also required to report the onset of their next menses following the final experimental session.
2.3. Procedure
To examine facial preferences over the menstrual cycle, each woman participated in two experimental sessions, with exactly 2 weeks between sessions.
2.3.1. Session I
Participants read and signed a consent form immediately before the first session. After completing the personal history questionnaire, they watched the QuickTime movie and received instructions on the use of the program. Participants were then required to select a set of defined target faces, one face at a time. The targets were the most healthy male face, most healthy female face, most dominant male, most dominant female, most attractive male (ATM), most attractive female, best male provider, best female provider, most baby-faced male, most baby-faced female, best male for LTM, best male for STM, and the androgynous face. In each case, a descriptive phrase was used to clarify the desired target face (Johnston et al., 2001). The selected faces (i.e., healthy male, attractive female, ATM, etc.) were presented in a random order that varied among participants and between sessions. The experimenter recorded the frame number selected by the participant for each target face until she had made selections for all targets. Then, for each participant, the male faces and the androgynous face she selected were displayed again, one face at a time, in a random order. While each selected face was so displayed, the participant was asked to rate that face on 14 attributes, using seven-point Likert scales. These attributes were coercive, emotionally supportive, sexually exciting, intelligent, sensitive, trustworthy, selfish, healthy, threatening, cooperative, physically supportive, controlling, physically protective, and helpful. Next, participants performed the MR task and completed the BSRI and PBI questionnaires.
2.3.2. Session II
Photocopies of each participant's hands were made, followed by the facial preference and MR tasks using the same procedure as in Session I. Then, participants completed the menstrual cycle questionnaire and were debriefed. Experimental credit was given when the first day of the first menses following this experimental session was reported to the experimenter by e-mail.
2.3.3. High and low probability of conception
The date of last ovulation was computed for participants based on their reports of menses. Variation in the duration of the menstrual cycle is almost exclusively confined to the follicular (preovulatory) phase, with ovulation being almost exactly 14 days prior to the onset of menses, irrespective of cycle length (Fluhmann, 1957, Lein, 1979, Matsumoto et al., 1962). This corresponds with the 14th day of a 28-day cycle and the 20th day of a 34-day cycle (Katchadourian, 1980). Without direct hormone measures, this procedure allowed each participant's menstrual state on test days to be defined in terms of the number of days prior to or after ovulation.
For the purposes of the present research, a high probability of conception is defined by a window of 9 days, which include the 7 days prior to ovulation, the day of ovulation, and the day following ovulation (Barrett & Marshall, 1969). A woman was classified as being in a low probability of conception state when tested outside of this window.
3. Results
3.1. Male facial preferences
Frame numbers for each male face, including the androgynous face, were averaged over sessions. The mean frame number for each face over all participants was the following: most dominant male (96), STM (233), most healthy male (250), ATM (265), best male provider (266), LTM (295), most baby-faced male (606), and the androgynous face (659). Using a significance criterion of p<.05, a Tukey–Kramer HSD comparison for all pairs [omnibus F(7,320)=163.80, p<.01] revealed that the DoM was significantly more masculinized than the other seven faces, whereas the androgynous and baby-faced male faces were not statistically different from each other, but were significantly less masculinized than all other faces. Fig. 1 shows the average frame of the movie selected by participants as representing the target faces of interest in the current study. As predicted, the preferred STM was significantly more masculinized than the preferred LTM [mean difference=62 frames, S.E.M.=21.89; two-tailed paired t(40)=2.82, p<.01] and the difference between the ATM and healthy male faces was not significant [two-tailed paired t(40)=1.00, p>.32].
;)
|
|
Fig. 1. Mean movie frames selected by participants as best representing their STM, LTM, and ATM. The dominant (DOM), healthy (HTM), and average male (AVM) faces are included for comparison. |
3.2. Attribute ratings
Each of the 41 participants had selected eight male target faces (including the androgynous face) during each experimental session and subsequently rated each face on 14 different attributes. Averaged across sessions, this procedure yielded 328 faces, with mean ratings on all 14 attributes. A principal components analysis with Varimax rotation was performed on the correlation matrix between the attribute ratings that the participants had assigned these faces. Three factors (Table 1) accounted for 69.2% of the variance: 38.5%, 24.4%, and 6.3%, respectively. In this order, these three factors and their associated attributes are very similar to the “Friend,” “Lover,” and “Enemy” factors reported by Johnston et al. (2001).
| Table 1.
Rotated components matrix for attribute ratings |
 |
Attributes | Loadings | | ||
|---|---|---|---|---|---|
|
F1 (Friend) | F2 (Lover) | F3 (Enemy) | | |
|
Sensitive | .838 | −.174 | −.214 | |
|
Helpful | .798 | .049 | −.194 | |
|
Trustworthy | .772 | −.007 | −.320 | |
|
Emotionally supportive | .772 | .095 | −.237 | |
|
Cooperative | .755 | −.271 | −.261 | |
|
Intelligent | .574 | .377 | .208 | |
|
Physically supportive | .127 | .824 | .054 | |
|
Sexually exciting | −.012 | .805 | .109 | |
|
Protective | −.076 | .796 | .354 | |
|
Healthy | .114 | .749 | .100 | |
|
Selfish | −.232 | .015 | .810 | |
|
Controlling | −.325 | .356 | .773 | |
|
Threatening | −.322 | <.001 | .764 | |
|
Coercive | −.212 | .542 | .572 | |
|
Variance accounted for by each factor (%) | 38.5 | 24.4 | 6.3 | |
|
Significant loadings for each of the three factors are indicated in bold. |
The bivariate scatterplot and fitting platform of JMP-IN 4.0.4, a SAS Institute statistical software package, was used to examine how the mean factor scores of all three factors varied over the 328 rated frames located on the male side of the QuickTime movie (Fig. 2). This analysis revealed that the Friend factor decreased linearly with increasing facial masculinity (b=0.0014, p<.01, adj. R2=.08). A third-order polynomial curve revealed that the Lover factor reached its maximum for a degree of masculinity (Frame 121) that was higher than the average but less than the extreme masculine face (b=−.0044, p<.01; b2=−0.000005, p<0.01; b3=0.000000021, p<.01; adj. R2=.42). Finally, the extreme masculine face (Frame 0) was rated highest on the socially undesirable Enemy traits. This factor declined rapidly with decreasing masculinity and was statistically best described by a third-order polynomial curve (b=−.0011, p<.01; b2=0.0000065, p<.01; b3=−0.000000013, p<.01; adj. R2=.17).
;)
|
|
Fig. 2. Best fit polynomial curves showing how the loadings of the friend, lover, and enemy factors vary over the male portion of the facial movie, beginning with the most extremely masculinized male face (0). The relative position, on average, of the STM, LTM, and ATM is indicated in the figure. |
3.3. Measures of individual differences
3.3.1. BSRI
To derive factors based on the variance within this group of experimental participants, a principal components analysis with Varimax rotation was performed on the 60 BSRI items. Four factors accounted for 42.6% of the variance, with F1, F2, F3, and F4 accounting for 14.1%, 12.4%, 8.9%, and 7.2% of the variance, respectively. In order of importance, items loading on F1 (the “feminine” factor) were tender, sympathetic, affectionate, sensitive, compassionate, sincere, warm, and eager to soothe hurt feelings. All of these items correspond to the psychological characteristics of females according to Bem (1981). F2 (the “masculine” factor) is represented by aggressive, forceful, soft-spoken, dominant, shy, willing to take a stand, act as a leader, assertive, and leadership abilities. Other than shy and soft-spoken, which are feminine traits that loaded negatively on F2, all of the other attributes loading on the second factor were identified by Bem as psychologically masculine traits. The remaining factors (F3 and F4) did not vary with any of the independent variables examined in this study.
3.3.2. Relationships between femininity/masculinity, 2D:4D, and MR ability
Simple linear regressions revealed that an increase in right-hand 2D:4D was associated with increasing femininity (b=9.45, p<.03, adj. R2=.10) but not masculinity, the F2 factor (b=5.31, p=.22). In contrast, MR time became faster as masculinity increased (b=−195.28, p<.01, adj. R2=.16), but there was no association between MR time and femininity (b=0.86, p=.99). The relationship between MR ability and 2D:4D was insignificant for either the right or left hand (b=1966.66, p=.92; b=699.13, p=.72, respectively). Finally, MR ability was not related to any of the other dependent variables discussed below.
3.4. Individual differences in facial preferences
3.4.1. LTM and STM difference and 2D:4D
As noted earlier, participants selected more masculine male faces for STMs than for LTMs [t(36)=2.82, p<.01]. However, this difference varied as a function of a participant's left-hand 2D:4D. Women with left 2D:4Ds below the mean (M=.97) selected almost identical STMs and LTMs (mean difference=1 frame), whereas those with high 2D:4Ds considered very different males (mean difference=123 frames) as appropriate for long-term compared with short-term relationships [t(35)=8.44, p<.01]. Because the right-hand 2D:4D was also examined, the p value was Bonferroni adjusted to p<.02.
3.4.2. Attractive male face as a function of menstrual phase and 2D:4D
For the 33 participants tested at both high and low probability of conception, preference shifts over the menstrual cycle were computed for all target faces by subtracting the movie frame number selected by each participant during the high probability of conception phase from the frame number that she selected for the same target face during her low probability phase. A series of multiple regressions with right-hand 2D:4D, MR ability, and normal cycle/oral contraception as predictor variables examined the magnitude of shift for each selected target face. These analyses revealed that only one predictor variable, 2D:4D, was significantly related to a shift in facial preference over the menstrual cycle, and this preference shift was only significant for the ATM face (b=−728.45, p<.01, adj. R2=.20). Both the size and direction of the ATM menstrual shift varied with a woman's 2D:4D. Low 2D:4D women shifted their preference toward a more masculinized ATM during high probability of conception, whereas high 2D:4D women shifted toward a less masculinized ATM during the high probability phase.
3.4.3. Attractive male face as a function of menstrual phase and psychological femininity (BSRI)
The size and direction of the ATM shift also changed as a function of a woman's “femininity” score (b=−30.68, p<.01, adj. R2=.19). During a high probability of conception, women who scored low in psychological femininity shifted their preference to a more masculinized ATM, whereas women high in psychological femininity shifted preference to a less masculinized ATM. Once again, only the ATM face showed this relationship.
3.5. Menstrual regularity and 2D:4D
Based on the menstrual regularity question on the menstrual questionnaire, participants were categorized into (a) regular, within 1 day, (b) fairly regular, within 4 days, or (c) irregular, greater than 4 days difference in the duration of their menstrual cycles. A one-way ANOVA (Digit Ratio×Menstrual Regularity) with Tukey–Kramer comparisons for all pairs revealed that women with regular menstrual cycles had higher 2D:4Ds than did those reporting fairly regular cycles, and the latter had higher 2D:4Ds than did those with irregular cycle lengths [F(2,38)=5.14, p=.01; Fig. 3].
;)
|
|
Fig. 3. Right-hand 2D:4D as a function of reported menstrual regularity (regular, fairly regular, or irregular). |
3.6. Parental Bonding Instrument and 2D:4D
An examination of STM and LTM facial preferences as a function of maternal and paternal bonding scores revealed no significant relationships. A simple linear regression revealed that paternal but not maternal bonding scores were positively correlated with 2D:4D (b=140.71, p<.01, adj. R2=.14).
3.7. Longest intimate relationship (LIR) and 2D:4D
A t test for LIR (scaled by age) revealed that women with left-hand 2D:4Ds below the mean had significantly shorter intimate relationships (M=0.69, S.E.M.=0.23) than did women with high 2D:4Ds [M=1.55, S.E.M.=0.25; t(35)=2.57, p<.02; Bonferroni adjusted p=.04]. A t test for LIR scaled by years since menarche also found that low 2D:4D women had shorter intimate relationships (M=2.04, S.E.M.=0.57) than did those with high 2D:4D [M=4.06, S.E.M.=0.25; t(35)=2.57, p=.01; Bonferroni adjusted p=.02]. Indeed, low 2D:4D women had intimate relationships that were, on average, half as long (M=14 months) as that of the high 2D:4D women (M=28 months).
3.8. ATM, STM, LTM, and average male faces
An earlier study by Johnston et al. (2001), using the same QuickTime movie, established that the ATM (M=Frame 284) was significantly more masculinized than the average male face was (M=Frame 370). A one-sample t test comparing the means of the ATM from the former and present studies (mean difference=19 frames) indicated no significant difference between the two studies [two-tailed t(40)=−1.23, ns]. Further one-sample t tests comparing the average male face from the first study to the ATM, LTM, and STM of the current study revealed that all of these mate choices were more masculinized than the average male face [two-tailed t(40)=6.80, p<.01; two-tailed t(40)=4.73, p<.01; two-tailed t(40)=7.31, p<.01, respectively].
4. Discussion
4.1. Attributes of preferred faces
4.1.1. Attractive versus average faces
In the current study, the ATM face was found to be significantly more masculinized than the average male face. This result replicates the findings of Keating (1985), Grammer and Thornhill (1994), and Johnston et al. (2001) and refutes the “average is attractive” hypothesis posited by Langlois and colleagues (Langlois & Roggman, 1990, Langlois et al., 1994, Langlois et al., 1991) and the “feminine male” preference reported by Penton-Voak et al. (1999) and Penton-Voak and Perrett (2000). These contradictory findings may be attributable to methodological differences (Alley & Cunningham, 1991, Johnston, 2000, Johnston et al., 2001, Pittenger, 1991) or cross-cultural differences (Penton-Voak, Jacobson, & Trivers, 2004).
4.1.2. Attractive versus healthy male
Replicating the findings of Johnston et al. (2001), the ATM and the healthy male face were not significantly different, and both faces were more masculinized than the average male face. These results suggest that facial androgen markers are perceived as signals of good health, and such faces are aesthetically pleasing to a female observer. Further support for the relationship between androgen markers and good health comes from a recent study by Gangestad and Thornhill (2003). These authors demonstrated a systematic linear relationship between the degree of facial masculinization and low fluctuating asymmetries of the body, an established measure of immunocompetence across species (Møller & Swaddle, 1997, Møller & Thornhill, 1997), including humans (Waynforth, 1998). From this perspective, androgen markers are honest signals of immunocompetence, and females attracted to masculine faces are expressing a preference, albeit unconscious, for a male with “good genes.” By itself, however, this “good genes” interpretation does not account for why females do not select an even more extreme masculine face for their attractive mate choice. An examination of participants' STM and LTM preferences, and their associated personality attributes, offers a possible answer to this question.
Theories on mating strategies in females suggest that two important considerations underlie mate choice and thus influence gene survival (Cashdan, 1996, Dawkins, 1989, Trivers, 1972). Cashdan (1996) defined the female's dilemma as choosing between a mate with “good genes” and a “good dad.” A choice for “good genes” may directly benefit offspring through the inheritance of immunocompetence genes, whereas a choice in favor of a “good dad” indirectly benefits offspring by affording paternal care and/or physical and psychological resources. This analysis suggests that females select a very masculine male when there is little or no expectation of resource investment (an STM) but prefer a less masculine male when other resources (kindness, cooperation, etc.) are important (an LTM).
In the current study, the STM was not statistically different from the ATM, but both of these choices were significantly more masculinized than the LTM and less masculinized than the dominant male face. If the ATM and STM are “good genes” males, then there are other factors involved in the selection of LTMs. Indeed, participants attributed a higher degree of friendliness to their LTM but placed more emphasis on the lover attributes when evaluating their STM. That is, the STM/ATM may be a “good genes” choice, whereas the LTM has qualities more appropriate for a “good dad.” A similar argument can explain why women do not select STM or ATM faces that are even more masculinized. As the degree of masculinity increases beyond the preferred level, there is a rapid increase in the socially undesirable traits that characterize the Enemy Factor. From this perspective, the STM and ATM represent the optimal “good genes” selections that avoid the negative attributes associated with higher degrees of masculinity. These observed mate preferences are consistent with the theoretical viewpoint of Cashdan (1996), and they further indicate that females are capable of making such choices based on the androgen markers displayed on male faces.
4.2. Attributes of the perceiver
In the current study, masculinity scores (BSRI) increased systematically with faster MR, and femininity scores (BSRI) decreased systematically with decreasing 2D:4D (see Csathó et al., 2003). These findings are consistent with early androgen exposure that both defeminizes and masculinizes the brain and is reflected in psychological measures of masculinity and femininity in adult humans. That is, within this group of women, there is evidence for different degrees of defeminizing and masculinizing that parallel the physical (2D:4D measures), cognitive (MR times), and psychological differences (BSRI scores) that exist between the sexes. However, if defeminizing and masculinizing were consequences of a single concurrent androgen exposure, there should also be a consistent relationship between MR times and 2D:4Ds. This was not the case. Both Clark (2004) and the current study found no correlation between these variables.
Testosterone levels in the amniotic fluid of normal children (male and female) in the second trimester have been significantly correlated with performance on MR tasks administered when the children were 7 years old (Grimshaw et al., 1995). This suggests that an organizational action of testosterone may underlie MR ability. However, there is also substantial evidence for an activational effect of testosterone on MR ability. For example, Van Goozen, Cohen-Ketenis, Gooren, Frijda, and van de Poll (1994) found that women receiving testosterone injections prior to woman-to-man sex change surgery showed markedly enhanced MR performance. Unlike 2D:4D that is fixed early during embryological life and remains constant thereafter, MR ability is influenced by circulating androgens. It is possible, therefore, that 2D:4D and MR may both be a function of androgen exposure but (a) the hormonal influences occur at different developmental times, (b) activational effects influence MR but not 2D:4D, and/or (c) different testosterone metabolites are involved in each case.
In the current study, femininity scores systematically decreased with decreasing 2D:4D. The most parsimonious explanation would be that early androgen exposure influences both the defeminization of the brain (Jost, 1972, Woodson & Gorski, 2000) and 2D:4D (Manning, 2002). In other mammals, the direct administration of androgens during early development has been shown to reduce the size of the anteroventral periventricular nucleus of the hypothalamus that controls the estrous cycle. The periventricular nucleus, which is larger in female than in male rats, can be systematically defeminized by increased perinatal exposure to androgens (Matsumoto, Sekine, Murakami, & Arai, 2000). In the current study, the menstrual cycle irregularity observed in women with low 2D:4Ds parallels these findings. That is, women with low 2D:4Ds show evidence of physiological defeminization in addition to the psychological defeminization noted earlier. Taken together, these results offer strong support for using 2D:4D as an index of early androgen exposure in human females.
In many respects, low 2D:4D women appear to be the antithesis of high 2D:4D women. Women with low 2D:4Ds are less stereotypically feminine than high 2D:4D women (i.e., they score low on the BSRI “femininity” factor that includes traits such as sympathy, tenderness, and sensitivity) and also have irregular menstrual cycles. A decrease in these characteristics may hinder emotional bonding, leading to the observed shorter relationships of low 2D:4D compared with high 2D:4D women. In contrast, the latter women score higher on femininity, consider themselves as sympathetic, tender, sensitive, affectionate, compassionate, and sincere, and have intimate relationships that are more than twice as long as their low 2D:4D counterparts do.
A recent report by Clark (2004) contributes a significant additional element to the list of differences between high and low 2D:4D women. In a group of Canadian women, Clark found a significant, negative correlation between a woman's right hand 2D:4D and her score on the Sociosexual Orientation Inventory (SOI), a measure of a woman's willingness to engage in casual (uncommitted) sex (Simpson & Gangestad, 1991). That is, low 2D:4D women reported an unrestricted sexuality, whereas high 2D:4D women preferred committed relationships. The result of Clark dovetails with the LIR findings of the current study, hence, in the interest of completeness, it has been added to the behavioral profiles of low and high 2D:4D women shown in Table 2.
| Table 2.
Summary of the differences associated with the 2D:4D digit ratio of Anglo women |
|
Low digit ratio | High digit ratio | |
|---|---|---|---|
|
Defeminized | Not defeminized | |
|
Irregular menstrual cycles | Regular menstrual cycles | |
|
Prefer masculinized STM and equally masculinized LTM | Prefer masculinized STM and less masculinized LTM | |
|
Shift toward a more masculinized ATM when probability of conception is greatest | Shift toward a less masculinized when probability of greatest | |
|
Low paternal bonding scores | High paternal bonding scores | |
|
Short-duration intimate (sexual) relationships | Long-duration intimate (sexual) relationships | |
|
(High SOI scores) | (Low SOI scores) | |
|
| ||
|
Conclusion | | |
|
“Good genes” mate choice strategy | “Good dad” mate choice strategy | |
|
Differences in parentheses are from Self-Perceived Attractiveness and Masculinization Predict Women's Sociosexuality, by Clark (2004). |
Mating strategies are commonly believed to result from the degree of attachment to primary caregivers during childhood (Bowlby, 1969, Bowlby, 1973, Bowlby, 1980, Draper & Harpending, 1982). The current study, however, found no significant relationships between a woman's perceived bonding to her parents (maternal or paternal) and her mate choices (STM or LTM). However, women who scored low in paternal bonding also had low 2D:4Ds, the trait that is closely associated with a “good genes” mate choice strategy (Table 2). Given these relationships, it is possible that the reported associations between low paternal bonding and unrestricted sexuality may reflect the actions of an underlying variable, early androgen exposure, that has a widespread influence on both bonding and mate choice.
Table 2 summarizes the observed differences between low and high 2D:4D women. They appear to differ in emotional makeup, which, in turn, may influence their social relationships. High 2D:4D women bond to their parents and their mates, and seek men with the traits of a good father when they contemplate enduring intimate relationships or the possibility of pregnancy. In contrast, low 2D:4D women appear to be emotionally independent and report less paternal bonding and shorter intimate relationships. In addition, unlike high 2D:4D women, they are attracted to very masculine “good genes” men for both short- and long-term relationships and seek such males when their fertility is highest. However, because most of the variables listed on Table 2 vary systematically with digit ratio, this high 2D:4D/low 2D:4D classification should not be interpreted as a simple dichotomy, but rather, the end-points of a digit–ratio continuum with most women falling somewhere between these two extremes. From this perspective, the current results suggest that these individual differences and their associated mate choice strategies depend, at least in part, on the degree of androgen exposure during early development.
The results of this study suggest that physical allure depends upon an interaction between the perceiver and perceived. The chemistry behind this attraction may involve prenatal hormonal influences that organize the brain of the perceiver and pubertal hormonal influences that are expressed on the face (and body) of the perceived. Further studies using male participants and/or different ethnic groups would be invaluable for establishing the generality of this hormonal model of physical attraction.
References
Alley & Cunningham, 1991 1.. Averaged faces are attractive, but very attractive faces are not average. Psychological Science. 1991;2:123–125. CrossRef
Baker, 1888 2.. Anthropological notes on the human hand. American Anthropologist. 1888;1:51–76.
Barrett & Marshall, 1969 3.. The risk of conception on different days of the menstrual cycle. Population Studies. 1969;23:455–461. CrossRef
Bem, 1981 4.. Bem Sex-Role Inventory: Professional manual. Palo Alto: Consulting Psychologists Press; 1981;.
Bowlby, 1969 5.. Attachment and loss. Attachment. vol. 1. New York: Basic Books; 1969;.
Bowlby, 1973 6.. Attachment and loss. Separation: Anxiety and anger. vol. 2. New York: Basic Books; 1973;.
Bowlby, 1980 7.. Attachment and loss. Loss. vol. 3. New York: Basic Books; 1980;.
Cashdan, 1996 8.. Women's mating strategies. Evolutionary Anthropology. 1996;5:134–143.
Chisholm, 1993 9.. Death, hope, and sex. Current Anthropology. 1993;43:1–24. CrossRef
Clark, 2004 10.. Self-perceived attractiveness and masculinization predict women's sociosexuality. Evolution and Human Behavior. 2004;25:113–124.
Collaer et al., 2002 11.. Cognitive and behavioral characteristics of Turner syndrome: Exploring a role for ovarian hormones in female sexual differentiation. Hormones and Behavior. 2002;41:139–155. CrossRef
Cooper & Shepard, 1973 12.. Chronometric studies of the rotation of mental images. In: Chase WG editors. Visual information processing. Oxford: Academic Press; 1973;p. 75–176.
Csathó et al., 2003 13.. Sex role identity related to the ratio of second to fourth digit length in women. Biological Psychology. 2003;62:147–156. MEDLINE | CrossRef
Dawkins, 1989 14.. The selfish gene. 2nd ed.. Oxford: Oxford University Press; 1989;.
Draper & Harpending, 1982 15.. Father absence and reproductive strategy: An evolutionary perspective. Journal of Anthropological Research. 1982;38:255–273.
Fluhmann, 1957 16.. Irregularities of the stages of the menstrual cycles. Western Journal of Surgery, Obstetrics and Gynecology. 1957;65:25–277.
Gangestad et al., 2004 17.. Women's preferences for male behavioral displays change across the menstrual cycle. Psychological Science. 2004;3:203–207. CrossRef
Gangestad & Thornhill, 2003 18.. Facial masculinity and fluctuating asymmetry. Evolution and Human Behavior. 2003;24:231–241.
Garn et al., 1975 19.. Early prenatal attainment of adult meta-carpal–phalangeal rankings and proportions. American Journal of Physical Anthropology. 1975;43:327–332. MEDLINE | CrossRef
Gorski et al., 1980 20.. Evidence for the existence of a sexually dimorphic nucleus in the preoptic area of the rat. Journal of Comparative Neurology. 1980;193:529–539. MEDLINE | CrossRef
Grammer & Thornhill, 1994 21.. Human (homo sapiens) facial attractiveness and sexual selection: The role of symmetry and averageness. Journal of Comparative Psychology. 1994;108:233–242. MEDLINE | CrossRef
Grimshaw et al., 1995 22.. Mental rotation at 7 years: relations with prenatal testosterone levels and spatial play experiences. Brain Cognition. 1995;29:85–100.
Grumbach, 2000 23.. Pubertal maturation in aromatase deficiency and resistance to estrogen. In: Bourguignon JP, Plant TM editor. The onset of puberty in perspective. Amsterdam: Elsevier; 2000;p. 247–267.
Hampson et al., 1998 24.. Spatial reasoning in children with congenital adrenal hyperplasia due to 21-hydroxylase deficiency. Developmental Neuropsychology. 1998;14:299–320.
Hier & Crowley, 1982 25.. Spatial ability in androgen-deficient men. New England Journal of Medicine. 1982;306:1202–1205. MEDLINE
Hines, 2000 26.. Gonadal hormones and sexual differentiation of human behavior: Effects on psychosexual and cognitive development. In: Matsumoto A editors. Sexual differentiation of the brain. Boca Raton: CRC Press; 2000;p. 257–278.
Johnston, 2000 27.. Female facial beauty: the fertility hypothesis. Pragmatics and Cognition. 2000;8:107–122.
Johnston et al., 2001 28.. Male facial attractiveness: Evidence for hormone mediated adaptive design. Evolution and Human Behavior. 2001;22:251–267.
Jones et al., 1972 29.. Factors influencing the age of menarche in a lower socio-economic group in Melbourne. Medical Journal of Australia. 1972;21:533–535.
Jost, 1972 30.. A new look at the mechanisms controlling sex differentiation in mammals. Johns Hopkins Medical Journal. 1972;130:38–53. MEDLINE
Katchadourian, 1980 31.. Fundamentals of human sexuality. 3rd ed.. New York: Holt, Rinehart, and Winston; 1980;.
Keating, 1985 32.. Gender and the physiognomy of dominance and attractiveness. Social Psychology Quarterly. 1985;48:61–70. CrossRef
Klimidis et al., 1992 33.. The PBI-BC: A brief current form of the Parental Bonding Instrument for adolescent research. Comprehensive Psychiatry. 1992;33:374–377. CrossRef
Langlois & Roggman, 1990 34.. Attractive faces are only average. Psychological Science. 1990;1:115–121. CrossRef
Langlois et al., 1994 35.. What is average and what is not average about attractive faces?. Psychological Science. 1994;5:214–220. CrossRef
Langlois et al., 1991 36.. A picture is worth a thousand words: Reply to “On the difficulty of averaging faces”. Psychological Science. 1991;2:354–357. CrossRef
Lein, 1979 37.. The cycling female: Her menstrual rhythm. San Francisco: Freeman; 1979;.
Levine et al., 1999 38.. Early sex differences in spatial skill. Developmental Psychology. 1999;35:940–949. MEDLINE | CrossRef
Linn & Petersen, 1985 39.. Emergence and characteristics of sex differences in spatial ability: A meta-analysis. Child Development. 1985;56:1479–1498. MEDLINE | CrossRef
Lutchmaya et al., 2004 40.. 2nd to 4th digit ratios, fetal testosterone and estradiol. Early Human Development. 2004;77:23–28. Abstract | Full Text | Full-Text PDF (121 KB) | MEDLINE | CrossRef
Manning, 2002 41.. Digit ratio: A pointer to fertility, behavior and health. New Brunswick: Rutgers University Press; 2002;.
Manning et al., 2000 42.. The 2nd to 4th digit ratio, sexual dimorphism, population differences and reproductive success: Evidence for sexually antagonistic genes. Evolution and Human Behavior. 2000;21:163–183.
Manning et al., 2003 43.. The second to fourth digit ratio and variation in the androgen receptor gene. Evolution and Human Behavior. 2003;24:399–405.
Manning et al., 1998 44.. The ratio of 2nd to 4th digit length: A predictor of sperm numbers and level of testosterone, LH and oestrogen. Human Reproduction. 1998;13:3000–3004. MEDLINE | CrossRef
Masica et al., 1971 45.. Fetal feminization and female gender identity in the testicular feminizing syndrome of androgen insensitivity. Archives of Sexual Behavior. 1971;1:131–142. CrossRef
Matsumoto et al., 2000 46.. Sexual differentiation of neuronal circuitry in the hypothalamus. In: Matsumoto A editors. Sexual differentiation of the brain. Boca Raton: CRC Press; 2000;p. 203–228.
Matsumoto et al., 1962 47.. Statistical studies on menstruation: A criticism on the definition of normal menstruation. Gunma Journal of Medicine. 1962;11:297–309.
Moffitt et al., 1992 48.. Childhood experience and the onset of menarche: A test of a sociobiological model. Childhood Development. 1992;63:47–58.
Møller & Swaddle, 1997 49.. Asymmetry, developmental stability, and evolution. Oxford: Oxford University Press; 1997;.
Møller & Thornhill, 1997 50.. Bilateral symmetry and sexual selection: a meta-analysis. American Naturalist. 1997;151:174–192. CrossRef
Parker, 1981 51.. Parental responses of depressives: An investigation of several explanations. Journal of Affective Disorders. 1981;3:131–140. MEDLINE | CrossRef
Parker, 1983 52.. Parental overprotection: A risk factor in psychosocial development. New York: Grune & Stratton; 1983;.
Parker et al., 1979 53.. A Parental Bonding Instrument. British Journal of Medical Psychology. 1979;52:1–10.
Penton-Voak et al., 2004 54.. Population differences in attractiveness judgments of male and female faces: Comparing British and Jamaican samples. Evolution and Human Behavior. 2004;25:355–370.
Penton-Voak & Perrett, 2000 55.. Female preference for male faces changes cyclically: Further evidence. Evolution and Human Behavior. 2000;21:39–48.
Penton-Voak et al., 1999 56.. Menstrual cycle alters face preference. Nature. 1999;399:741–742. MEDLINE | CrossRef
Petersen, 1976 57.. Physical androgyny and cognitive functioning in adolescence. Developmental Psychology. 1976;12:524–533. CrossRef
Phoenix et al., 1959 58.. Organizing action of prenatally administered testosterone propionate on the tissues mediating mating behavior in the female guinea pig. Endocrinology. 1959;65:369–382. MEDLINE
Pittenger, 1991 59.. On the difficulty of averaging faces: Comments on Langlois and Roggman. Psychological Science. 1991;2:351–353. CrossRef
Resnick et al., 1986 60.. Early hormonal influences on cognitive functioning in congenital adrenal hyperplasia. Developmental Psychology. 1986;22:191–198. CrossRef
Simpson & Gangestad, 1991 61.. Individual differences in sociosexuality: Evidence for convergent and discriminant validity. Journal of Personality and Social Psychology. 1991;60:870–883. MEDLINE | CrossRef
Surbey, 1990 62.. Family composition, stress, and human menarche. In: Bercovitch F, Zeigler T editor. The socioendocrinology of primate reproduction. New York: Liss; 1990;p. 11–32.
Tanner, 1990 63.. Fetus into man: Physical growth from conception to maturity. Cambridge: Harvard University Press; 1990;.
Trivers, 1972 64.. Parental investment and sexual selection. In: Campbell B editors. Sexual selection and the descent of man 1871–1971. Chicago: Aldine Publishing Company; 1972;p. 136–179.
Van Goozen et al., 1994 65.. Activating effects of androgens on cognitive performance: Causal evidence in a group of female-to-male transsexuals. Neuropsychologia. 1994;32:1153–1157. MEDLINE | CrossRef
Waynforth, 1998 66.. Fluctuating asymmetry and human male life-history traits in rural Belize. Proceedings of the Royal Society of London. 1998;265:1497–1501.
Wisniewski et al., 2000 67.. Complete androgen insensitivity syndrome: Long-term medical, surgical, and psychosexual outcome. Journal of Clinical Endocrinology and Metabolism. 2000;85:2664–2669. CrossRef
Woodson & Gorski, 2000 68.. Structural sex differences in the mammalian brain: Reconsidering the male/female dichotomy. In: Matsumoto A editors. Sexual differentiation of the brain. Boca Raton: CRC Press; 2000;p. 229–255.
New Mexico State University, NM 88003, USA
Corresponding author. Department of Psychology, MSC 3452, New Mexico State University, P.O. Box 30001, Las Cruces, NM 88003, USA. Tel.: +1 505 646 2502; fax: +1 505 646 6212.
PII: S1090-5138(05)00018-8
doi:10.1016/j.evolhumbehav.2005.03.002
© 2005 Elsevier Inc. All rights reserved.