The human anger face evolved to enhance cues of strength

    The human anger face evolved to enhance cues of strength Aaron Sell, Leda Cosmides, John Tooby PII: DOI: Reference:

S1090-5138(14)00067-1 doi: 10.1016/j.evolhumbehav.2014.05.008 ENS 5911

To appear in:

Evolution and Human Behavior

Received date: Accepted date:

9 April 2014 21 May 2014

Please cite this article as: Sell, A., Cosmides, L. & Tooby, J., The human anger face evolved to enhance cues of strength, Evolution and Human Behavior (2014), doi: 10.1016/j.evolhumbehav.2014.05.008

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Submitted Manuscript: Confidential 27 December 2011 ACCEPTED MANUSCRIPT

Title: The human anger face evolved to enhance cues of strength

RI P

T

Authors: Aaron Sellǂ12, Leda Cosmides1, and John Tooby1

Affiliations:

Center for Evolutionary Psychology; University of California, Santa Barbara; Santa

SC

1

2

NU

Barbara, CA 93106-3210; USA.

School of Criminology and Criminal Justice; Griffith University, Mount Gravatt; Mount

MA

Gravatt, QLD 4121; Australia.

ED

ǂ Corresponding Author: [email protected]

AC

CE

PT

Word Count: 1904

ACCEPTED MANUSCRIPT Main Text The human anger face is an early (Stenberg, Campos & Emde 1983), reliably-developing (Galati

RI P

T

et al. 2003), species-typical (Ekman 1973) expression consisting of a stereotyped array of coordinated muscle contractions (Ekman & Friesen 1978) (see figure 1). But why did evolution

SC

give the human facial expression of anger the particular form that it has? The most common position is that the anger face is a universal but arbitrary signal of aggressive intent (Blair 2003;

NU

Matsumot, Hee Yoo & Chung 2010; Schmidt & Cohn 2001). According to this widespread view,

MA

the expression would have evolved to be salient and distinguishable from other emotional expressions (Darwin 1872), but there need be nothing functional about the particular pattern of

ED

muscle movements per se. We agree that the anger expression functions as a signal, but we propose that the specific array of muscle contractions that constitute the anger face was in fact

PT

tailored by selection to be functional rather than arbitrary. Specifically, during conflicts of

CE

interest, natural selection favored displaying those configurations of muscle activation that amplified others’ assessments of the sender’s fighting ability—in the human case, those

AC

configurations that amplified cues of strength. This hypothesis is made plausible by recent work showing the existence of cues of strength in the face – cues that are rapidly and spontaneously assessed when estimating another’s fighting ability (Sell et al. 2009a; Trebicky et al. 2013; Zilioli et al. 2014).

1.1 Theoretical background An animal’s fitness is crucially dependent on the outcomes of conflicts of interest. Accordingly, selection should have organized behavioral systems in animals that bargain for improved

ACCEPTED MANUSCRIPT outcomes in these conflicts. In particular, aggression is a type of bargaining behavior that deploys the threat or actuality of cost-infliction as a tool to incentivize others to reduce their

T

resistance to the aggressor’s realization of its interests (Huntingford & Turner 1987). In humans,

RI P

these tactics are largely regulated by anger - a neural system that evolved to orchestrate bargaining behavior to cost-effectively resolve conflicts of interest in favor of the angry

SC

individual (Sell, Tooby & Cosmides 2009; Sell 2011; Tooby et al. 2008). On this view, the

NU

triggering of anger signals the expressor’s assessment that the existing situation does not sufficiently reflect the angry individual’s interests, given its bargaining power. The greater the

MA

fighting ability of the individual, the more costs it can inflict, and the better its bargaining position is. This is why selection favored the evolution of aggression, and phenotypic traits that

ED

enhance the ability to inflict costs on competitors.

PT

As has been documented in hundreds of species, relative fighting ability in animals is a crucial

CE

determinant of winning conflicts (Arnott & Elwood 2009). In consequence, selection generally favors the evolution of neural designs that (a) assess fighting ability in conspecifics and (b) use

AC

those assessments to make decisions about whether to cede or contend for a resource (Arnott & Elwood 2009; Enquist & Leimar 1983). Mutant designs that enhance the cues of fighting ability used by these assessment mechanisms will spread (e.g., piloerection, dewlap inflation) because such shifts in assessment by rivals will lead to better outcomes for the animal displaying enhanced cues. Such enhancement of cues will persist provided that the predictive relationship between the cue and the animal’s actual fighting ability is maintained to some degree (Maynard Smith & Harper 2003). Cues that are constrained by the physical structure of the feature being assessed (i.e. “indices”) generally meet this requirement, and thus many species have been selected to configure their morphology in ways that send enhanced signals of size and strength

ACCEPTED MANUSCRIPT during agonistic bargaining (Darwin 1872; Huntingford & Turner 1987; Maynard Smith & Harper 2003).

RI P

T

This logic applies equally to humans: Selection is expected to have favored assessment mechanisms among our ancestors that exploited available indices of fighting ability. Recent

SC

evidence demonstrates that humans can and do assess cues of physical strength when evaluating fighting ability, presumably because strength was a substantial component of fighting ability.

NU

Indeed, these assessment systems appear well-designed to estimate fighting ability rather than

MA

strength per se, e.g., assessments of fighting ability privilege cues of muscularity over simple height or weight, upper body strength over lower body strength, and more accurately assess men

ED

than women1 (Sell et al. 2009a, 2010; Sell 2012). Moreover, as predicted by the model that anger is the expression of a neural bargaining system, physically stronger men get angry more

PT

easily, experience more favorable resolutions of conflicts of interest, and consider aggression a

CE

more legitimate way to settle conflicts (Sell, Tooby & Cosmides 2009). Crucially, accurate cues of fighting ability have been found to exist in the face alone, and are spontaneously and rapidly

AC

assessed when estimating fighting ability (Sell et al. 2009a; Trebicky et al. 2013; Zilioli et al. 2014). If anger functions as a bargaining system in humans, then it follows that humans should have evolved to deploy facial morphology in a way that enhances these cues during aggressive bargaining. We hypothesize that the universal anger expression evolved for that function. Over evolutionary time, selection would build neural circuits that, at the onset of anger, co-activated those discrete muscle movements in the face that increased the appearance of strength. Hence, 1

Numerous data sets have established that that upper body strength in males is about 90% greater than in females (Abe, Kearns & Fukunaga 2003; Lassek & Gaulin 2009), with the distributions of the two sexes only minimally overlapping. Hence, ancestral humans typically lived in social groups in which women would rarely have been strong enough to settle conflicts in their favor through violence. For this reason and others, aggression will be a more important tool of social negotiation for men rather than women (Archer 2004; Campbell 2002; Daly & Wilson 1998; Wrangham & Peterson 1996), and both men and women should be (and are) better calibrated to assess male strength (Sell 2012).

ACCEPTED MANUSCRIPT the signal of the onset of a bout of power-based bargaining (i.e. anger) would be simultaneous to and constituted by facial enhancements of cues of strength. In short, the universal human anger

T

facial expression is an adaptation for enhancing cues of strength in the face during aggressive

RI P

bargaining.

SC

The accepted method for establishing the evolved function of an adaptation was best articulated by George Williams (1966): Researchers must show that features constituting the

NU

candidate adaptation are sufficiently well-organized to bring about the solution to an identified

MA

adaptive problem that the coordination between adaptive problem and phenotypic solution cannot be plausibly explained by chance. Therefore, if the anger face is an evolved adaptation

ED

designed for enhancing cues of strength and fighting ability in the face, then each of the seven distinct major modifications that comprise the anger face should, independently, make the face

PT

appear physically stronger. The null (and widely accepted) hypothesis is that the features of the

CE

anger face are arbitrary, and so should show no particular relationship to judgments of strength. To test between these hypotheses, we used a program calibrated with a large number of statistical

AC

composites of real faces to generate seven pairs of faces. Each pair contrasted a single feature of the face as modified by anger (e.g. lowered brow) with the opposite modification (e.g. raised brow). Raters then chose the stronger of the pair to determine the effect of each component of the anger face.

2.1 Methods The actual facial components that are modified by human anger have been extensively documented by Ekman and colleagues in their Facial Action Coding System (FACS) (Ekman & Friesen 2002a). Ekman identifies facial movements as various combinations of “action units”

ACCEPTED MANUSCRIPT (hereafter AU), which consist of stereotyped contractions of specific facial muscles. In the case of anger, Ekman and his colleagues identified the classic closed-mouth anger expression (as

T

modeled in figure 1) as the result of the following action units: AU: 4, 5, 7, 10, 17, 22, 23, 24

RI P

(see Ekman & Friesen 2002a, table 10-1). Quotes from the FACS manual (Ekman & Friesen 2002b) documenting the effects of action units 4, 10, 17, 22, 23 and 24 on facial structure are

NU

SC

listed in Table 1.2

Table 1: Ekman’s FACS identifies seven key manipulations in the anger face Lowers brow-ridge

Raises infraorbital triangle

Widens nose

FACS code

AU 4: Brow Lowerer

AU 10: Upper Lip Raiser

AU 10: Upper Lip Raiser

“Lowers the eyebrow” pg.17

“Pushes the infraorbital triangle up” pg.95

“Widens and raises the nostril wings.”

2.1.1 Subjects

ED

PT

CE

Inner Brow Ridge – high/low

Cheek bones – low/high

Nose – Nostrils wide/thin

Raises mouth

AU 10: Upper Lip Raiser AU 17: Chin Raiser “Raises the upper lip” pg.95 “Pushes the lower lip upward” pg.98 Mouth – up/down

Enlarges chin and chin bun

Lips pushed forward

Lips thinned

AU 17: Chin Raiser

AU 22: Lip Funneler

AU 23: Lip Tightener AU 24: Lip Pressor

“Pushes the chin boss upward” pg. 98

“Lips funnel outward” pg. 235

“Tightens and narrows the lips” pg. 241

Chin – wide/thin

Lips puckered /retracted

Lips thin/thick

AC

FACS manual description of changes (Ekman, et al. 2002b) FaceGen feature manipulated

MA

Effect of anger face

The raters in our experiment were taken from the psychology and criminology subject pools at the University of California, Santa Barbara (35 subjects [25 female], mean age = 18.9, SD = 1.2) and Griffith University, Mount Gravatt (106 subjects [80 female], mean age = 24.9, SD = 8.3) respectively. Students received course credit for participation.

Two action units— the glare (“upper lid-raiser” – AU 5) and the squint (“the lid-tightener” – AU 7) were excluded because they are mutually exclusive and have opposite effects on the face, naturally occurring at different times or in different facial displays (see Supplemental Material for additional data and discussion on this topic). 2

ACCEPTED MANUSCRIPT

2.1.2 Stimuli

T

The stimuli were created with FaceGen 3.1, a face modeling software package made by Singular

RI P

Inversions. FaceGen creates three dimensional models of faces which can be morphed by age, race, sex and through direct manipulations of facial features. FaceGen works with a formula that

SC

was calibrated with a large number of statistical composites of real faces, enabling average faces

NU

to be generated for various age groups and across sexes. We used as our starting point the average European 20 year old male face. Then, for each of the seven constituent features of the

MA

anger face this average face was modified in the direction of anger and in the opposite direction to create two faces: an anger-modified face and a control face. Manipulations were done to

ED

approximately the midpoint of the scale (setpoint 4 on a scale from 0 to 10) (see table 1). Our

PT

stimuli consisted of male faces because – in humans – fighting ability is more accurately assessed on male targets than female targets (Sell et al. 2009a; 2010). Indeed, both males and

CE

females are more accurate when assessing male targets’ fighting ability; a pattern that holds in

AC

general and when assessing fighting ability from the face specifically (Sell 2012).

2.1.3 Procedure

Subjects were presented with each pair of faces side by side (each 9cm2, resolution: 150 dpi) in random orders. For each pair of faces (anger modified or its control), subjects were asked to judge which appeared physically stronger. The anger modified face was randomly placed either on the left or right side.

3.1 Results and Discussion

ACCEPTED MANUSCRIPT Results are summarized in figure 2. As predicted, chi-squared tests show that each modification that the anger system makes to the face significantly and independently increases perceived

T

strength (see supplemental material for a replication with different stimuli and additional tests).

RI P

If the anger face was not designed to enhance cues of strength, there would be no reason for each constituent of the anger face to increase others’ assessments of the strength of the individual

SC

displaying that facial modification.3

NU

These results are perfectly consistent with previous literature showing that several of the

MA

components of the anger face are more prominent in the sex that shows evidence of combat design (Becker et al. 2007; see Sell, Hone & Pound 2012 for evidence of combat design). Male-

ED

biased sexually dimorphic features include prominent cheekbones, wider noses, lower browridges, and larger chins (Lefevre et al. 2012; Penton-Voak et al. 2001; Trebicky et al. 2013).

PT

Similar research shows components of the anger expression positively correlate with judgments

CE

of dominance – e.g. lower browridge, wider nose (Toscano, Schubert & Sell 2014); browridge, thin lips, wide nose (Windhager, Schaefer & Fink 2011). However, the current study is the first

AC

systematic test of the individual components of the anger expression, and in doing so confirm that they are improbably well-designed to solve the specified adaptive problem.

4.1 Conclusion

3

It is important to recognize that not all conflicts of interest are resolved by threats of violence, especially in a highly cooperative species such as humans. When individuals have a history of cooperation, they can also bargain by threatening to reduce the weight they place on each other’s welfare, as reflected in their actions (Sell, Tooby & Cosmides 2009). Nevertheless, aggression is a far more ancient and zoologically widespread form of bargaining. Once the anger face system was forged in aggressive contexts, it then was available to be co-opted as a wider signal that the sender contests a course of action by another (is “angry”), even when the bargaining threat is not violence but the withdrawal of cooperation.

ACCEPTED MANUSCRIPT Taken together, these data show that the constellation of facial muscle contractions comprising the anger expression is not arbitrary; instead, the anger face is well-designed to increase the

T

perceived fighting ability of the angry individual during bargaining. These results support the

RI P

view that, over evolutionary time, the addition of each modification would have functionally enhanced perceived formidability in the eyes of competitors, advantageously increasing the

SC

likelihood that they would defer in conflict. These data explain why the enhancement of facial

MA

NU

cues of strength would be timed with, and even constitute the signal of anger.

References

ED

Abe T, Kearns CF, Fukunaga T. (2003). Sex differences in whole body skeletal muscle mass measured by magnetic resonance imaging and its distribution in young Japanese adults.

PT

Brit. J of Sports Med., 37(5), 436–440.

CE

Archer J (2004). Sex differences in aggression in real-world settings: A meta-analytic review.

AC

Rev Gen Psychol., 8(4), 291-322. Arnott G & Elwood R (2009). Assessment of fighting ability in animal contests. Anim. Behav., 77(5), 991-1004.

Becker DV, Kenrick DT, Neuberg SL, Blackwell KC, Smith DM. The confounded nature of angry men and happy women. Journal of Personality and Social Psychology 92(2), 179190. Blair RJR (2003). Facial expressions, their communicatory functions and neuro-cognitive substrates. Philos Trans R Soc Lond B, 358(1431), 561-572.

ACCEPTED MANUSCRIPT Campbell A (2002). A mind of her own: The evolutionary psychology of women. Oxford: Oxford

RI P

Daly M & Wilson M (1988). Homicide. Transaction Books.

T

University Press.

Darwin C (1872). The expression of the emotions in man and animals. London: John Murray.

SC

Ekman P. (1973). Cross-cultural studies of facial expression. In P. Ekman (Ed.), Darwin and

NU

facial expression (pp. 169-222). New York: Academic Press. Ekman P & Friesen WV (1978). Facial Action Coding System. Palo Alto, CA: Consulting

MA

Psychologists.

Ekman P, Friesen WV & JC Hager (2002a). Facial Action Coding System: Investigator’s Guide.

ED

Research Nexus: Salt Lake City.

PT

Ekman P, Friesen WV & JC Hager (2002b). Facial Action Coding System: The Manual.

CE

Research Nexus: Salt Lake City. Enquist, M. & Leimar, O. (1983). Evolution of fighting behavior: decision rules and assessment

AC

of relative strength. Journal of Theoretical Biology, 102, 387-410. Galati D, Sini B, Schmidt S & Tinti C. (2003). Spontaneous facial expressions in congenitally blind and sighted children aged 8-11. Journal of Visual Impairment & Blindness, 97(7), 418-428. Huntingford FA & Turner AK (1987). Animal conflict. Chapman & Hall, New York. Lassek WD & Gaulin S (2009). Costs and benefits of fat-free muscle mass in men: Relationship to mating success, dietary requirements, and natural immunity. Evol. & Hum. Beh., 30, 322–328.

ACCEPTED MANUSCRIPT Lefevre CE, Lewis GJ, Bates TC, Dzhelyova M, Coetzee V, Deary IJ & D Perrett (2012). No evidence for sexual dimorphism of facial width-to-height ratio in four large adult samples.

RI P

T

Evolution and Human Behavior, 33, 623-627. Marsh AA, Adams RB, & RE Kleck (2005). Why do fear and anger look the way they do? Form

SC

and social function in facial expressions. PSPB, 31(1), 73-86.

Matsumoto D, Hee Yoo S & Chung J (2010). The expression of anger across cultures. In (eds.)

NU

Potegal, Stemmler & Speilberger’s International Handbook of Anger. Springer: New

MA

York.

Maynard Smith J & Harper D (2003). Animal Signals. Oxford University Press.

ED

Penton-Voak IS, Jones BC, Little AC, Baker S, Tiddeman B, Burt DM & Perrett DI (2001).

PT

Symmetry, sexual dimorphism in facial proportions and male facial attractiveness. Proc. R. Soc. B, 268, 1617–1623.

CE

Schmidt K & Cohn J (2001). Human facial expressions as adaptations: evolutionary questions in

AC

facial expression research. Am J Phys Anthropol., 33, 3-24. Sell A, Hone LSE & N Pound (2012). The importance of physical strength to human males. Human Nature, 23, 30-44. Sell A (2011). The recalibrational theory and violent anger. Aggression and Violent Behavior, 16, 381-389. Sell A (2012). Evolved cognitive adaptations. In J. Vonk & T. Shackelford (eds.), The Oxford Handbook of Comparative Evolutionary Psychology. Oxford University Press.

ACCEPTED MANUSCRIPT Sell A, Bryant G, Cosmides L, Tooby J, Sznycer D, von Rueden C, Krauss A & M Gurven (2010). Adaptations in humans for assessing physical strength and fighting ability from

RI P

T

the voice. Proc. R. Soc. B, 277, 3509-18. Sell A, Cosmides L, Tooby J, Sznycer D, von Rueden C & Gurven M (2009a). Human

SC

adaptations for the visual assessment of strength and fighting ability from the body and face. Proc. R. Soc. B, 276, 575-584.

NU

Sell A, Tooby J & Cosmides L (2009b). Formidability and the logic of human anger. Proc. Nat.

MA

Acad. of Sci., 106(35), 15073-78.

Stenberg CR, Campos JJ & Emde RN (1983). The facial expression of anger in seven-month-

ED

old infants. Child Dev., 54(1), 178-184.

PT

Tooby J, Cosmides L, Sell A, Lieberman D & Sznycer D. (2008). Internal regulatory variables and the design of human motivation: A computational and evolutionary approach. In

CE

Elliot, Andrew J. (2008). Handbook of approach and avoidance motivation. (pp. 251-271).

AC

New York, NY, US: Psychology Press. Trebicky V, Havlicek J, Roberts CS, Little AC & K Kleisner (2013). Perceived aggressiveness predicts fighting performance in mixed-martial-arts fighters. Psych. Sci., 24(9), 1664-72. von Rueden C, Gurven M. & Kaplan H. (2008). The multiple dimensions of male social status in an Amazonian society. Evol. & Hum. Beh., 29: 402-415. Walker R, Hill K, Kaplan H & G McMillan (2002). Age dependency in hunting ability among the Ache of eastern Paraguay. J of Hum. Evol., 42, 639–657. Williams G (1966). Adaptation and Natural Selection. Princeton University Press.

ACCEPTED MANUSCRIPT Windhager S, Schaefer K, Fink B (2011). Geometric morphometrics of male facial shape in relation to physical strength and perceived atractiveness, dominance, and masculinity.

RI P

T

American Journal of Human Biology, 23, 805–814. Wrangham R & DE Peterson (1996). Demonic males: Apes and the origins of human violence.

SC

Houghton Mifflin Harcourt.

Zilioli S, Sell A, Stirrat M, Jagore J, Vickerman W & NV Watson (2014). Face of a fighter:

NU

bizygomatic width as a cue of formidability. Aggressive Behavior, revised and

AC

CE

PT

ED

MA

resubmitted.

ACCEPTED MANUSCRIPT Figure Captions

RI P

T

Figure 1. Effects of the anger face (i.e. action units 4, 5, 7 10, 17, 22, 23, 24)

SC

Note for Figure 1: The right side of the model’s face is seen on both the left (as it appeared in the photo) and right (flipped along the Y-axis) to control for any fluctuating asymmetry in the face.

AC

CE

PT

ED

MA

NU

Figure 2. The seven components of the universal anger face increase perceived strength

AC

CE

PT

ED

MA

NU

SC

RI P

T

ACCEPTED MANUSCRIPT

Figure 1

ACCEPTED MANUSCRIPT

Figure 2

AU 4: Brow Lowerer

Raises infraorbital triangle

AU 10: Upper Lip Raiser

Widens nose

AU 10: Upper Lip Raiser

Raises mouth

AU 10: Upper Lip Raiser AU 17: Chin Raiser

Enlarges chin and chin bun

AU 17: Chin Raiser

Lips pushed forward

AU 22: Lip Funneler

Lips thinned

AU 23: Lip Tightener AU 24: Lip Pressor

% picking angered face as stronger 94% Χ2 = 111.8 p = 10-25

MA

Levator Labii Superioris; Caput Infraorbitalis

PT

ED

Levator Labii Superioris; Caput Infraorbitalis, Mentalis

Mentalis

CE

AC

Control Face

T

Lowers brow-ridge

Depressor Glabellae; Depressor Supercilli; Corrugator

Anger Modified Face

RI P

Muscles used

SC

FACS code

NU

Effect of anger

86% Χ2 = 73.3 p = 10-16

85% Χ2 = 70.4 p = 10-16

75% Χ2 = 36.5 p = 10-8

78% Χ2 = 45.1 p = 10-10

80% Χ2 = 52.1 p = 10-12

Orbicularis Oris 71% Χ2 = 25.4 p = 10-6