Received 20March 2003 Accepted 30 June 2003 Publishedonline 21August 2003

Second to fourth digitratio, andperceived maledominance Nick Neave1* ,Sarah Laing 1,BernhardFink 2 and JohnT. Manning 3 1Human CognitiveNeuroscience Unit, Schoolof Psychology and SportSciences, NorthumberlandBuilding, NorthumbriaUniversity, Newcastleupon Tyne NE1 8ST, UK 2Ludwig-Boltzmann-Institutefor Urban Ethology, Althanstrasse 14, Vienna, Austria 3Departmentof Psychology, University ofCentral Lancashire, Preston PR1 2HE, UK Previous studieshave shownthat male faceswith extreme featuresassociated with testosteroneare per- ceivedas dominant and masculine. Women have beenreported to prefer more masculinizedmale faces asthey may considertestosterone markers tobe an ‘ honest’indication ofgood health, andsuch consider- ationsmay underlietheir aestheticpreferences. However, pronounced testosterone facial markers are also associatedwith dominance,and several negative traits. This suggeststhat female aesthetic preferencesmay bean adaptive compromise betweenpositive attributes associatedwith higher than aver- age testosterone,and negative attributes associatedwith more extreme masculinization. This currentstudy attempts toclarify therole ofhormone markers in female perceptionsof dominance, masculinity and attractiveness,in male facial images. Recentevidence suggests that therelative length ofthe 2nd to 4th (2D :4D ratio) is apointer toprenatal testosteronelevels andmay thusserve as a windowto the prenatal hormonal environment.We measured 2D :4D in asample ofmale college studentsand took salivary samplesto analyse circulating levels oftestosterone. Women rated facial images ofthese males for dominance,masculinity andattractiveness. Our resultsshow that male 2D :4D wassignificantly negatively related toperceived dominance and masculinity butnot attractiveness. Circulating testosteronelevels were notrelated todominance, masculinity or attractiveness.These findings suggest that: (i) high prenatal levels oftestosterone serve to ‘ organize’male facial featuresto subsequently reflect dominance and masculine characteristics presumably activated during ;and (ii) attractivenessis not directly related totestos- teronelevels. We conclude that facial dominanceand masculinity reflecta male’s perceivedstatus rather than his physical attraction towomen. Keywords: finger length ratio; testosterone;dominance; masculinity

1. INTRODUCTION 1996), further suggesting alink todominance behaviour (Mazur &Booth 1998). The faceplays acrucial role in animal andhuman social AsMazur &Booth (1998) pointedout, early exposure cognition andbehaviour. Evidencefrom primates shows tohigher levels oftestosterone are likely toproduce more that certain brain structuresspecialize in perceiving facial male-like characteristics (masculinization) andfewer expressionsof emotionsand intentions, and in regulating female characteristics (defeminization) whereasless emotional andbehavioural responsesto these expressions exposureto testosterone causes the reverse. These pre- (Morris et al. 1996). Humanfaces signal qualities that are andperinatal hormoneeffects are regarded asorganizing stable over time, suchas attractiveness as a potential mate thearchitecture ofthebody and brain. Whenmale testos- (Grammer &Thornhill 1994; Barber 1995) or dominance teroneincreases later in life during puberty,it is thought (Zebrowitz& Montepare1992; Collins& Zebrowitz toactivate pre-existing structures;for example, in males 1995; Mueller& Mazur 1997). Although weassume that higher androgenserum levels at puberty together with a facial beauty conveysa wealth ofinformation concerning higher androgenreceptor expression at certain skeletal anindividual’ s mate value, theactual role ofperceived sites,may contributeto sex differences in facial mor- facial dominancestill remains equivocal. phology (Kasperk et al.1997). In males,a high testoster- Facial dominancemay signal subjectiveintentions one-to-oestrogenratio facilitates thelateral growth ofthe (Maynard-Smith &Harper 1988; Harper 1991) aswell as cheekbones,mandibles andchin, the forward growth of anobjective potential for action (Mazur &Booth 1998). thebones of the eyebrow ridges, and the lengthening of Featuresthat contributeto perceived facial dominance, thelower facial bone,all ofwhich are consideredmascu- suchas strong jawsor broad cheekbones (see, for line facial features(Thornhill &Gangestad1999; Fink & example, Cunningham et al. 1990), may indicatesuperior Penton-Voak2002; Grammer et al. 2003). Oestrogen physical strength.Several authorshave suggestedthat inhibits this growth, leading toa feminizedfacial shape suchfeatures may relate tocirculating testosteronelevels with high eyebrows,more gracile jawand fuller lips. A (Grammer &Thornhill 1994; Thornhill &Gangestad preferencefor sex-typical traits may operate in females’ judgementsof male facial attractiveness,and males’ pref- erencesfor female faces.Whereas some studies support thehypothesis that womenprefer masculinizedmale faces, *Authorfor correspondence ([email protected]). other studiesindicate that womendo not have clear

Proc.R. Soc.Lond. B (2003) 270, 2167–2172 2167 Ó 2003 TheRoyal Society DOI10.1098/ rspb.2003.2502 2168N. Neaveand others Perceptions of male dominance preferencesfor suchtraits in males.Perrett et al. (1998) age andis correlated with testosteroneconcentrations showedthat females ’ preferencesfor male faceswere (Manning et al. 1998; Ronalds et al. 2002); (ii) alow apparently driven by stereotypical personality attributions: 2D :4D ratio isassociated with male-typical attributes highly masculinizedmale faceswere perceived as less suchas better spatial ability (Manning& Taylor 2001), warm, lesshonest and more dominantthan feminized left-handedness(Manning et al. 2000) anda predis- male faces.Such attributions may have akernel oftruth, positiontowards autism (Manning et al. 2001); (iii) the ashigh testosteronehas beenlinked with anti-social waist: hip ratio ofmothers, a positive correlate oftestos- behaviour in men(Dabbs et al. 1991). terone,is negatively associatedwith the2D :4D ratio of In mammals, thegrowth ofsecondary sexual traits is their children (Manning et al. 1999); and(iv) children linked tolevels ofandrogens (Owens & Short 1995), with congenital adrenal hyperplasia, ageneticdisorder whichdepress immune systemfunction (Folstad &Karter associatedwith high prenatal ,have lower 1992). Evolutionary theory suggeststhat only males in 2D :4D ratios than normal controls(Brown et al. 2002). goodcondition can bear the ‘handicap’ oflarge secondary Astestosteroneis strongly implicated in establishing the sexual traits that representan honest advertisement of sexually dimorphic 2D :4D ratio, andas this hormonehas male viability. Assignalling facial dominanceis therefore also beenlinked with male dominancebehaviours, Man- costly,it is assumedthat dominantlooks may signal high ning (2002) proposedthat alow2D :4D ratio may be status.Across a widevariety ofspecies, behaviours associatedwith assertivenessand dominance but as yet, intendedto achieve, maintain andenhance status are this possiblelink remains speculative.In thecurrent paper observedprimarily among high-testosteroneindividuals weaimed toestablish whetherorganizational levels oftes- (Kraus et al. 1999; Josephs et al. 2003). If perceivedfacial tosterone(as measured by 2D :4D ratio) are related to dominancedoes indeed relate toapotential for high status activational levels oftestosterone, and perceived facial in male dominancehierarchies, thenin thehuman ‘dominance’, and ‘masculinity ’.Wepredicted that males resource-basedmating system(Buss 1989; Kenrick& with low2D :4D (i.e.high prenatal levels oftestosterone) Keefe1992), it may signal afitnessrelevant quality wouldbe rated by femalesas being higher in dominance (Dewsbury1982; Mueller1993; Ellis 1995). andmasculinity. It isparticularly important for humansto distinguish In addition,although someevidence demonstrates a betweendominant behaviour, whichaims at achieving and female preferencefor exaggerated male facial character- maintaining high statusand greater controlof resources istics(e.g. Scheib et al. 1999) therelationship between over aconspecific,and aggressive behaviour, whichaims facial dominanceand attractiveness remains unclear. at inflicting physical injury onaconspecific.In this paper, Swaddle& Reierson(2002) recentlyshowed that high tes- weonly discussthe former. While therelationship tosteronefaces reveal dominance.However, they didnot betweentestosterone levels andhuman remains findevidence of directional selectionfor increased(or equivocal (Archer 1991), Mazur& Booth (1998) con- decreased)testosterone in termsof male facial attractive- cludedthat high levels oftestosterone were linked with ness.Consequently, these authors argued that this reflects dominanceand competitiveness in human males;specifi- stabilizing selectionacting ontestosterone through mate cally, they suggestedthat testosteronerises in theface of preferences.We thereforeaimed toinvestigate thepossible achallenge andactivates behaviours intendedto dominate relationship betweenmale 2D :4D ratio andfacial attract- andenhance status. ivenessjudgements made by females.We predictedthat if In support,several studieshave reportedlinks between womenconsider dominant and masculine faces as attract- testosteroneand human dominanceencounters (Booth et ive, then—in addition todominance and masculinity — al. 1989; Schaal et al. 1996; Salvador et al. 1999; Dabbs 2D :4D shouldalso benegatively related tomale et al. 2001; Neave &Wolfson2003). However,some perceivedattractiveness. researchershave criticized thesimplicity ofthis theory (see Finally, toidentify possiblelinks betweenactivational author commentsfollowing Mazur &Booth 1998), and levels oftestosterone and female perceptionsof male have pointedout the problems in attempting tolink asin- dominance,masculinity andattractiveness, we also took gle hormonal measureto certain behaviours (Brain 1998). salivary samplesof free testosterone from themale volun- Furthermore, Campbell et al. (1998) notedthat as teers. dominant/aggressive behaviour emergesearly in child- hood,the major impact oftestosterone may beacting at an organizational rather than anactivational level. 2. MATERIALAND METHODS Until recently,ascertaining foetal exposureto testoster- onehas beendifficult. However, Manning (2002) has (a) Participants recentlysummarized theevidence that testosteronestimu- Werecruited 48 male participants fromundergraduate lates prenatal growth ofthe 4th finger while oestrogenpro- coursesat Northumbria University.The meanage of the sample motesthe growth ofthe 2nd finger. Alow2D :4D ratio was 21.3years (rangeof 18 –33years, s.d. = 3.4).All claimed to (4th finger longer than thesecond) may thusact asa beheterosexual and right-handed. marker for auterineenvironment high in testosteroneand lowin oestrogen,and such a ratio is mostoften seen in (b) Procedure males.Conversely, a high 2D :4D ratio may serveas a (i) Datarecording marker for auterineenvironment low in testosteroneand Participantsgave theirinformed written consentin accord- high in oestrogen,and is mostoften found in females.In ancewith Northumbria UniversitySchool of Psychology and support,(i) 2D :4D ratio demonstratesa sexually dimor- Sport SciencesEthics Committeeguidelines. Age, body height phic pattern that appears tobeestablished at avery early and weight wererecorded.

Proc.R. Soc.Lond. B (2003) Perceptions of male dominance N.Neaveand others 2169

Wemeasuredthe lengths ofthe 2ndand 4th digits ofthe left Table1. Correlation of hormonal variableswith perceived and right fromthe tip of the fingerto the ventralproximal measures. creasefrom photocopies. Wherethere was aband ofcreasesat (LH, left ; RH,right hand; T,testosterone.) the base of the digit,we measuredfrom the most proximalof these.For 16 right and lefthands, the 2ndand 4th digitswere LH2D: 4DRH 2D:4D T alsomeasured directly from the hand and fromthe photocopies to establish repeatabilities.All measurements were made with age 0.111 20.071 20.298¤ height 20.085 0.040 20.198 digitalVernier callipers measuring to 0.01mm. dominance 20.306¤ 20.278¤ 0.086 Colourdigital images of eachparticipant ’sfacewere taken at masculinity 20.309¤ 20.305¤ 0.076 high resolutionunder standardized light conditionsin frontal attractiveness 20.231 20.064 20.017 view.Participants were advised to removeany facialadornment, testosterone 0.034 20.052 lookdirectly into the cameraand presenta neutralfacial expression. ¤ p , 0.05. Finally,male participants provideda salivary samplefrom which circulatinglevels of freetestosterone couldbe measured. 3. RESULTS They weregiven labelled, lidded cups and sugarless chewing gum,and wereasked to chew the gum and deposit enough saliva In theright handthe first andsecond measures of to fillthe bottom ofthe cup(5 ml)the sampleswere then frozen 2D :4D from photocopieshad ahigh intraclass corre- to 220 °Cbeforeanalysis. lation coefficient( r1 = 0.96), asdid the mean of these two measurescompared toright 2D :4D calculated from

(ii) Faceratings measurementsdirectly onthe ( r1 =0.91). Repeated To obtain independentratings ofperceivedmale facial domi- measuresANOVA analysesshowed that measurement nance,masculinity and attractiveness,36 female raters froma error waslow in relation toreal differencesin 2D :4D differentlocal university, mean age of 22.1years (rangeof 19 – ratio betweenindividuals (2D :4D from photocopies 30years,s.d. = 2.7)rated eachface. Faces were randomly F = 49.37, p = 0.0001; 2D :4D from photocopiesand presentedby computer,and remainedon screenuntil raters pro- directfrom fingers F = 21.70, p = 0.0001). In theleft videda measureof perceiveddominance, masculinity, and handthe intraclass correlation betweenmeasures of attractiveness using a7-point scale(1 = extremelysubordinate/ 2D :4D from photocopieswas high ( r1 = 0.94), butfrom feminine/unattractive, 7 = extremelydominant/ masculine/ acomparison ofphotocopied and direct-measured attractive).At the endof the session,female raters wereasked 2D :4D it wasquite low ( r1 = 0.67). However,repeated ifthey recognizedany of the facesthey had seen(none did). measuresANOVA analysesshowed that both intraclass correlation coefficientswere significant (2D :4D from (c) analyses photocopies F = 31.52, p = 0.0001; 2D :4D from pho- Salivary testosterone was usedas itislessstressful and invas- tocopiesand direct from fingers F = 5.10, p = 0.009). We ivethan serumsampling, and causesminimal disruption to nor- concludedthat ourmeasures of 2D :4D from photocopies malroutines (Ellison 1988). A strong correlationhas been ofthehands reflected real differencesbetween individuals reportedbetween salivary and serumtestosterone levels(Vittek in thestudy. et al. 1985).As circadianand circannualchanges intestosterone Pearsoncorrelations werecalculated analysing the have beenreported (Nieschlag 1974; Dabbs 1990)samples were relationships betweendigit ratios, perceiveddominance, collectedat the sametime of day (between10.00 and 15.00) masculinity, andattractiveness ratings, andsalivary testos- within aperiodof two weeks(in November). terone.Values of p equal toor lessthan 0.05, in one-tailed Testosterone levelsin salivasamples were measured by using testswere considered significant. Correlations are shown amodifiedserum radioimmunoassay. Before assaying, the pre- in table 1. viouslyfrozen samples were thawed and centrifuged.The assay Asage andbody height may influencecertain physio- usedthe Coat-A-Count total testosterone kit (Euro/DPC,UK), logical variables aswell asanatomical traits wefirstly cor- which isasolidphase radioimmunoassay,based ontestosterone- related thesefactors with the2D :4D ratio. Neither were specificantibodies immobilized to the wallof apolypropylene associatedwith the2D :4D ratio, butage wasassociated tube. 125I-labelledtestosterone competesfor afixedtime with with attractivenessratings ( r = 20.265, p = 0.034)—the testosterone inthe salivasample for antibody sites.The tube is older theparticipant thelower therating, andsalivary tes- then decanted,to separate antibody-bound testosterone (on tosteronelevel ( r = 20.298, p = 0.022)—theolder thepar- wallsof the tube)from free testosterone (inthe decanted ticipant thelower the level (table 1). Left-handand right- solution).The tubes arecounted using agamma counterand hand2D :4D ratios werestrongly correlated with one the amount of testosterone presentin the salivasample is another (r = 0.642, p = 0.000). determinedfrom a calibrationgraph ofknown standards, where Both left-handand right-hand 2D :4D ratios weresig- higher gamma counts equate to lowertestosterone levels.This nificantly negatively correlated with female ratings ofper- procedurehas to bemodified to allowfor measurementof the ceivedmale dominance( r = 20.306, p = 0.017; r = lowerhormone concentrations in salivaby dilutingthe supplied 20.278, p = 0.028, respectively) indicating thelower the Coat-A-Count testosterone standards 20-foldand extendingthe ratio, thehigher therating ofmale facial dominance(see incubationtime from 3 hat 37 °Cto 24hat roomtemperature. figure 1). Each samplewas assayed induplicate and the meanvalue for Similarly, both left-handand right-hand 2D :4D ratios the two duplicatestaken as the valuefor eachsample. The tubes weresignificantly negatively correlated with female ratings wereeach counted for oneminute using aPackardCobra II ofperceived masculinity ( r = 20.309, p = 0.016; r = gamma counter. 20.305, p = 0.018, respectively) indicating thelower the

Proc.R. Soc.Lond. B (2003) 2170N. Neaveand others Perceptions of male dominance

(a) (a) 7 7 6 g

n 6 i t

a 5 r

e

c 5 g n 4 n a i n t i a

r 4

3 m y o t d i n

2 i l 3 u c

1 s a

m 2 0 0.85 0.90 0.95 1.00 1.05 1 left-hand 2D : 4D 0 0.85 0.90 0.95 1.00 1.05 (b) left-hand 2D : 4D 7

6 (b) 7

g 5 n i

t 6 a r

e 4 c

n 5 g a n n i i t

3 a m r 4

o y d t i n

2 i l 3 u c s 1 a

m 2

0 1 0.85 0.90 0.95 1.00 1.05 right-hand 2D : 4D 0 0.85 0.90 0.95 1.00 1.05 Figure 1. ( )Scatterplot of left-hand 2D: 4Dratio and a right-hand 2D : 4D femaleperceptions of male dominance. ( b)Scatterplot of right-hand 2D:4Dratio and femaleperceptions of male Figure 2. (a)Scatterplot of left-hand 2D:4Dratio and dominance. femaleperceptions of masculinity. ( b)Scatterplot of right- hand 2D:4Dratio and femaleperceptions of masculinity. ratio, thehigher therating ofmale facial dominance(see figure 2). In addition,attractiveness ratings didnot significantly correlate with 2D :4D (left hand: r = photographs (figures 1and2). To clarify this,measure- 20.231, p = 0.057; right hand: r = 20.064, p = 0.333). mentsmade of male digit ratios werestrongly related to Neither left-handand right-hand 2D :4D ratios, nor female perceptionsof these same participants from head- faceratings of ‘dominance’, ‘masculinity ’ and ‘attractive- only colourphotographs. This finding implies that high ness’ displayed significant correlations with salivary testos- prenatal levels oftestosterone (resulting in alow2D :4D terone(table 1). Finally, attractivenessratings were ratio) serveto ‘organize’ male facial features(presumably significantly positively correlated with both ratings of activated during puberty) tosubsequently reflect facial dominance( r = 0.462, p = 0.000) andmasculinity dominance/masculinecharacteristics. This is in accord (r = 0.428, p = 0.001). with theassertion of Campbell et al. (1998) that human dominancebehaviours may reflectorganizational rather than activational processes. 4. DISCUSSION The developmentof certain facial featuresthat contrib- The aim ofthis studywas to ascertain possiblerelation- uteto perceptions of dominance may betestosterone shipsbetween organizational effectsof testosterone (by induced,and several authorshave proposedthat such 2D :4D finger length ratio), activational levels oftestos- ‘honest’ markers affectjudgements of certain facial terone(via saliva) andperceived male facial character- characteristics. For example, Swaddle& Reierson(2002) istics.Manning (2002) suggestedthat alow2D :4D ratio digitally manipulated male facesto mimic shapevariations may beassociated with male assertivenessand dominance, causedby varying levels oftestosterone. Female raters andwe findsome support for this prediction.Our results selectedthe most dominant and attractive faces,with showedthat both left-and right-hand 2D :4D ratios were ‘high-testosterone ’ facesproducing higher dominancerat- significantly negatively correlated with female ratings of ings (though lower attractivenessratings). Dominant- perceivedmale dominanceand masculinity from colour looking males may bepreferred as mates because their

Proc.R. Soc.Lond. B (2003) Perceptions of male dominance N.Neaveand others 2171 looks are ‘honest’ indicators oftheir potential toachieve Brain, P.F.1998 Androgens and human behaviour: acomplex andmaintain ahigh status(Mueller & Mazur 1997). relationship. Behav. BrainSci. 21, 363–364. However,we found no evidence that circulating levels Brown, W.M.,Hines, M., Fane, B.A.&Breedlove, S.M. oftestosterone in ourmale volunteerswere associated with 2002 Masculinized finger length patterns in human males female ratings oftheir perceiveddominance/ masculinity, and femaleswith congenital adrenal hyperplasia. indeed,testosterone levels within themales werenot Behav. 42, 380–386. Buss, D.M.1989 Sexdifferences in human matepreferences: related totheir 2D :4D ratios. This is perhaps notso sur- evolutionary hypothesestested in 37 cultures. Behav. Brain prising, if male facial characteristics are organized before Sci. 12, 1–49. birth andthen sculpted during puberty,then we should Campbell, A., Muncer, S.&Odber, J.1998 Primacy of organ- notexpect fluctuating levels tobe related toperceived ising effectsof testosterone. Behav. BrainSci. 21, 365. facial characteristics. Collins, M.&Zebrowitz, L.1995 Thecontributions of appear- Our data revealed that perceiveddominance and mas- ance to occupational outcomes in civilian and military set- culinity are both significantly correlated with attractive- tings. J.Appl. Social Psychol. 25, 29–163. nessratings, butthis may notbe simply takenas support Cunningham, M.R., Barbee, A.P.&Pike, C.L.1990 What for adirectlink betweenthese features. In theliterature, do women want? Facialmetric assessmentof multiple several reports ofan association between attractiveness motives in theperception of male facialphysical attractive- anddominance exist (e.g.Keating 1985; Barber 1995), ness J.Person.Social Psychol. 59, 61–72. butwe know that this also entails someneonate features DabbsJr, J.M.1990 Age and seasonal variation in serum tes- suchas large eyes,a small noseand a high forehead. tosterone concentrations in men. Chronobiol.Int. 7, 245– Consequently,as Mueller &Mazur (1997) point out,a 249. dominant-looking man may notbe rated asattractive, and DabbsJr, J.M., Jurkovic, G.&Frady, R.L.1991 Salivarytes- anattractive man neednot be perceived as being domi- tosterone and cortisol among late adolescent male offenders. J.Abnormal Child Psychol. 19, 469–478. nant.Mueller & Mazur(1997) thereforepropose that the DabbsJr, J.M.,Bernieri, F.J., Strong, R.K., Campo, R.& effectsof facial dominanceand attractiveness should be Milun, R.2001 Going on stage: testosterone in greetings consideredseparately by statistical analysis. Our data sup- and meetings. J.Res.Person. 35, 27–40. port this view ofthe 2D :4D ratio asa pointer toprenatal Dewsbury, D.A.1982 Domiance rank, copulatory behavior, (andactual) levels oftestosterone correlated with per- and differential reproduction. Q.Rev. Biol. 57, 135–159. ceiveddominance and masculinity, butnot with attract- Ellis, L.1995 Dominance and reproductive success among iveness.In addition,as we did not find a relationship nonhuman animals: across-species comparison. Ethol. betweenattractiveness and circulating levels oftestoster- Sociobiol. 16, 257–333. one,we may speculatethat any associationbetween these Ellison, P.T.1988 Human salivarysteroids: methodological featuresalso operatesat anearly stage in life. considerations and applications in physicalanthropology. During early development,very high levels oftestoster- Yearbook Phys.Anthropol. 31, 115–142. oneare supposedto negatively affectdevelopment, Fink, B.&Penton-Voak, I.S.2002 Evolutionary psychology of resulting in theexpression of bodily asymmetries. The facialattractiveness. Curr.Direct. Psychol. Sci. 11, 154–158. 2D :4D ratio has beensuggested to serve as a pointer to Folstad, I.&Karter, A.J.1992 Parasites, bright malesand the theprenatal hormonal environmentof a foetus.For our immunocompetence handicap. Am. Nat. 139, 603–622. Grammer, K.&Thornhill, R.1994 Human facial attractive- study,we may speculatethat theassociation between ness and sexual selection: theroles of averageness and sym- attractiveness,dominance and masculinity onthebasis of metry. J.Comp.Psychol. 108, 233–242. underlying hormones,namely testosterone,is basedon the Grammer, K., Fink, B., Moller, A.P.&Thornhill, R.2003 factthat: (i) theattractiveness –hormonelink is apointer Darwinian aesthetics: sexual selection and thebiology of todevelopmental stability andimmunocompetence; and beauty. Biol. Rev. 78, 385–407. (ii) featuresdeveloped under the influence of testosterone Harper, D.G.C.1991 Communication. In Behavioural ecology donot directly accountfor attractivenessbut rather for (ed. J.R.Krebs &N.B.Davies).Oxford: Blackwell Scien- male dominanceand masculinity, both ofwhich are fea- tific. turesof perceived behavioural social statusrather than Josephs, R.A., Newman, M.L., Brown, R.P.&Beer, J.M. mate value. The presentfindings are in accordwith pre- 2003 Status, testosterone, and human intellectual perform- viousstudies (Mueller & Mazur 1997; Mazur &Booth ance: stereotype threat asstatus concern. Psychol. Sci. 14, 1998) proposing that facial dominanceis anhonest signal 158–163. ofdominant behaviour. Asit suggests,a male ’s mate value Kasperk, C.,Helmboldt, A., Borcsok, I., Heuthe, S., Cloos, dependson his potential toachieve high statusin social O.,Niethard, F.&Ziegler, R.1997 Skeletalsite-dependent contexts. expression of theandrogen receptor in human osteblastic cell populations. Calc. Tissue Int. 61, 464–473. Keating, C.1985 Gender and thephysiognomy of dominance and attractiveness. Social Psychol. Q. 48, 61–70. REFERENCES Kenrick, D.T.&Keefe, R.C.1992 Age preferences in mates Archer, J.1991 Theinfluence of testosterone on human reflect sexdifferences in human reproductive strategies. aggression. Br.J. Psychol. 82, 1–28. Behav. BrainSci. 15, 75–133. Barber, N.1995 Theevolutionary psychology of physical Kraus, C.,Heistermann, M.&Kappeler, P.M.1999 Physio- attractiveness: sexual selection and human morphology. logical suppression of sexual function of subordinate males: Ethol. Sociobiol. 16, 395–424. asubtle form of intrasexual competition among male sifakas Booth, A., Shelley, G., Mazur, A., Tharp, G.&Kittok, R. (Propithecus verreauxi ). Physiol. Behav. 66, 855–861. 1989 Testosterone and winning and losing in human compe- Manning, J.T.2002 Digit ratio: apointer to fertility, behaviour, tition. Hormones Behav. 23, 556–571. and health.NewBrunswick, NJ:Rutgers University Press.

Proc.R. Soc.Lond. B (2003) 2172N. Neaveand others Perceptions of male dominance

Manning, J.T.&Taylor, R.P.2001 2nd to 4thdigit ratio Owens, I.P.F.&Short, R.V.1995 Hormonal basisof sexual and male ability in sport: implications for sexual selection in dimorphism in birds: implications for new theories of sexual humans. Evol. Hum. Behav. 22, 61–69. selection. TrendsEcol. Evol. 10, 44–47. Manning, J.T.,Scutt, D., Wilson, J.&Lewis-Jones, D.I.1998 Perrett, D.I., Lee, K.J., Penton-Voak, I., Rowland, D., Yoshi- Theratio of 2nd to 4thdigit length: apredictor of sperm kawa,S., Burt, D.M.,Henzi, S.P., Castles, D.L.&Akam- numbers and concentrations of testosterone, luteinizing hor- atsu, S.1998 Effectsof on facial mone and oestrogen. Hum. Reprod. 13, 3000–3004. attractiveness. Nature 394, 884–887. Manning, J.T.,Trivers, R.L., Singh, D.&Thornhill, R.1999 Ronalds, G., Phillips, D.I., Godfrey, K.M.&Manning, J.T. Themystery of femalebeauty. Nature 399, 214–215. 2002 Theratio of second to fourth digit lengths: amarker Manning, J.T.,Trivers, R.L., Thornhill, R.&Singh, D.2000 of impaired fetalgrowth? Early Hum. Dev. 68, 21–26. The2nd :4thdigit ratio and asymmetryof hand perform- Salvador, A., Suay, F., Martinez-Sanchis, Simon, V.M.& ance in Jamaican children. Laterality 5, 121–132. Brain, P.F.1999 Correlating testosterone and fighting in Manning, J.T.,Baron-Cohen, S., Wheelwright, S.&Sanders, male participants. Physiol. Behav. 68, 205–209. G.2001 The2nd to 4thdigit ratio and autism. Devl Med. Schaal,B., Tremblay, R.E., Soussignan, R.&Susman, E.J. 1996 Maletestosterone linked tohighsocial dominance but Child Neurol. 43, 160–164. low physicalaggression in early adolescence. Maynard-Smith, J.&Harper, D.G.C.1988 Theevolution of J. Am. Acad. Child Adolescent Psychiat. 34, 1322–1330. aggression: can selection create variability? Phil. Trans.R. Scheib, J., Gangestad, S.W.&Thornhill, R.1999 Facial attract- Soc. Lond. B 319, 557–570. iveness, symmetryand cues of good genes. Proc.R. Soc. Lond. Mazur, A.&Booth, A.1998 Testosterone and dominance in B 266, 1913–1917. (DOI10.1098.rspb.1999 .0866.) men. Behav. BrainSci. 21, 353–397. Swaddle, J.P.&Reierson, G.W.2002 Testosterone increases Morris, J.S., Frith, C.D.,Perrett, D.I., Rowland, D.,Young, perceived dominance but not attractiveness in human males. A.W., Calder, A.J.&Dolan, R.J.1996 Adifferent neural Proc.R. Soc. Lond. B 269, 2285–2289. (DOI10.1098/rspb. response in thehuman amygdalain fearful and happyfacial 2002.2165.) expressions. Nature 383, 812–815. Thornhill, R.&Gangestad, S.W.1996 Theevolution of Mueller, U.1993 Social statusand sex. Nature 363, 490. . TrendsEcol. Evol. 11, 98–102. Mueller, U.&Mazur, A.1997 Facial dominance in Homo Thornhill, R.&Gangestad, S.W.1999 Facial attractiveness. sapiens ashonest signalling of male quality. Behav. Ecol. 8, TrendsCogn. Sci. 3, 452–460. 569–579. Vittek, J., L ’Hommedieu, D.G., Gordon, G.G., Rappaport, Neave,N. &Wolfson, S.2003 Testosterone, territorilaity, and S.C.&Southren, A.L.1985 Direct radioimmunoassay the ‘home advantage ’. Physiol. Behav. 78, 269–275. (RIA) of salivarytestosterone: correlation withfree and total Nieschlag, E.1974 Circadian rhythmsof plasmatestosterone. serum testosterone. Life Sci. 37, 711–716. In Chronobiological aspects ofendocrinology (ed. J.Aschoff, F. Zebrowitz, L.A.&Montepare, J.1992 Impressions of baby- Ceresa &F.Halberg), pp. 117 –128. NewYork: Schat- facenessand attractiveness across thelife span. Devl Psychol. tauer Verlag. 28, 1143–1152.

Proc.R. Soc.Lond. B (2003)