The Adaptive Significance of Human Scleral Brightness. An Experimental Study

doi:10.21203/rs.3.rs-1994935/v1

Download PDF

Article

The Adaptive Significance of Human Scleral Brightness. An Experimental Study

https://doi.org/10.21203/rs.3.rs-1994935/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Homogeneously depigmented sclerae have long been proposed to be uniquely human - an adaptation to enable cooperative behaviour by facilitating interpersonal coordination through gaze following. However, recent evidence has shown that deeply pigmented sclerae also afford gaze following if surrounding a bright iris. Furthermore, while current scleral depigmentation is clearly adaptive in modern humans, it is less clear how the evolutionarily intermediate stages of scleral pigmentation may have been adaptive. In sum, it is unclear why scleral depigmentation became the norm in humans, while not so in sister species like chimpanzees, or why some extant species (presumably as our ancestors did at some point) display intermediate degrees of pigmentation. We created realistic facial images of 20 individually distinct hominins with diverse facial morphologies, each face in the (i) humanlike bright sclera and (ii) generalised apelike dark sclera version. Participants in two online studies rated the bright-sclera hominins as younger, healthier, more attractive and trustworthy, but less aggressive than the dark-sclera hominins. Our results support the idea that the perceptual affordances of more depigmented sclerae increased perceived traits that fostered trust, increasing fitness for those individuals and resulting in depigmentation as a fixed trait in extant humans.

Human eyes are exceptional: in addition to their unusual horizontal elongation, they have homogeneously pale sclerae, which makes them highly conspicuous and transforms them from sensory organs into signalling devices. However, the understanding of the exact role of this morphology has remained highly speculative, and it has not yet been systematically and comprehensively studied with a view to evaluating the variety of its possible underlying adaptive functions. Interest in this topic was first sparked with the early comparative studies on primate external eye morphology studies by Kobayashi and Kohshima (1997, also 2001)^1,2, which singled humans out as having a uniquely conspicuous coloration, and proposed that this coloration evolved to facilitate gaze-following and triadic communication (cf. Tomasello et al. 2007; Perea Garcia et al. 2017)^3,4. Several influential accounts of human socio-cognitive evolution (e.g. Tomasello et al. 2007, Knight and Lewis 2017, cf. also e.g. Hare 2017)^3,5,6 have followed this idea, with claims that dating the emergence of this conspicuous morphology “would suggest a possible date for the origins of uniquely human forms of cooperation and communication” (Tomasello et al. 2007: 319)³.

However, more recent research points to a broad variety of social and ecological factors that influence the diversity in external eye appearance in primates; for example, such diversity may be driven by the behavioral ecology of specific lineages, such as the spectral quality and quantity of light from the sun in the species’ range (Perea-García et al., 2022)⁷. Furthermore, functional studies⁸ as well as simulations^9,10 show that the eye-gaze of other primate species previously considered to have cryptic eye gaze is actually conspicuous, supporting the proposal by Perea-García et al (2019)¹¹ based on comparative morphological evidence. This is especially interesting in the case of the chimpanzee, which are characterized by deeply pigmented conjunctiva, but whose typically bright amber irises create a stark contrast between these two tissues.

In sum, pigmented sclerae may also afford gaze following by onlookers, and we assume that scleral pigmentation is the ancestral state in primates - since most studied primate species do display some degree of scleral pigmentation (Perea-García et al., 2022)⁷. This begs the question: if pigmented sclerae afford gaze following, why did depigmented sclerae become fixed in our lineage? The transition from pigmented to depigmented sclerae in humans may have been mediated by several mutually non-exclusive proximate functions (see esp. Danel et al. 2020)¹², among which the literature suggests the signaling of attractiveness and health (e.g. through symmetrical depigmentation that affords the perception of changes in vasculature or pigmentation), signalling of reliability and trustworthiness (e.g. by facilitating the perception of one’s gaze and emotions and, thus, intentions), and signalling of reduced emotional reactivity and reactive aggression (possibly sharing proximal mechanisms with self-domestication). In addition to affording signaling functions, the loss of populations of epithelial stem cells in the nasal and temporal limbus (Grieve et al., 2015)¹³ may have been a prerequisite for the extensive and homogeneous loss of pigmentation that characterizes our species. This is because the superior and inferior limbus are typically covered by the eyelids, protecting these cell populations from increased risk of UV damage that are essential to keeping a transparent cornea (Perea-García et al., 2021)¹⁴ and, therefore, vision. By contrast, the presence of lateral stem cell populations may necessitate increased pigmentation to reduce the chances of keratopathy (or corneal abnormality) (Hingorami et al., 2012)¹⁵.

We believe that a detailed account of the natural history of our extant eye appearance necessitates a holistic approach combining sources of evolutionary, environmental, developmental, behavioral and psychological evidence. In this study, however, we restrain our focus to the factors pertaining to the psychology of the differential perception of psychological traits on the basis of changes in scleral brightness. To start evaluating the relative contribution of these factors, we developed a novel rating design that allowed us to collect evaluations of humanlike faces with a bright versus dark sclera while keeping the participants unaware of the experimental modification. To this end, we morphed and photo-edited images of the faces of a range of reconstructed species of ancient hominins (see 2.1 below), to create realistic facial images of 20 individually distinct hominins with diverse facial morphologies; each face in the (i) “humanlike” bright sclera and (ii) “generalised apelike” dark sclera version. This stimulus was presented online to 250 (Study 1) and 100 (Study 2: self-replication) demographically diverse participants, who rated the individuals in the images on their perceived age, health, attractiveness, trustworthiness and aggressiveness.

1.1. Predictions

We predicted that the bright-sclera hominin faces would be rated as younger, healthier, more attractive and trustworthy, and less aggressive than the dark-sclera faces.

1.1.1. Health

The external appearance of the sclera is known to change with a wide range of health conditions, making it a useful tool in medical diagnosis, and, by extension, a reliable cue to several aspects of an individual’s health status. The depigmented sclera in humans provides a blank canvas that showcases changes in vasodilation and blood flow through vessels of the conjunctiva (Leibiwitz 2000, Roche & Kobos 2004)^16,17. The resulting differences in eye redness may indicate a range of underlying health issues, including inflammation, irritation and trauma of ocular structures (Mueller & McStay, 2008; Leibowitz, 2000; Donshik, 1988)^16,18,19, subconjunctival haemorrhage (Doshi & Noohani, 2022)²⁰, glaucoma (Donshik, 1988)¹⁹, and more general and systemic conditions such as hypertension, diabetes (Owen 2005)²¹, sickle cell and autoimmune diseases (Paton, 1961, Patel & Lundy, 2002)^{22, 23} as well as lack of sleep (Sundelin et al. 2013)²⁴ and intoxication (McLane & Caroll, 1986, Li et al. 2008, Dhingra et a. 2019)^{25, 26, 27}. A yellowish coloration of the sclera is a typical symptom of jaundice and high blood level of bilirubin, and may indicate multiple underlying diseases causing impaired bilirubin metabolism, liver dysfunction or biliary-tract obstruction (Al-Qaisi et al., 2021)²⁸

Several experimental studies showed that digitally manipulated redness and yellowness of the sclera affected the health perception of observed eye photographs. Compared to unmanipulated stimuli, individuals with redder and yellower eyes were perceived as less healthy (Provine et al. 2011, 2013)^29,30. Similarly, Caucasian female facial pictures with decreased eye redness and yellowness were perceived as healthier than identical portraits with eye colours manipulated in the opposite direction (Russell et al. 2014)³¹.

1.1.2. Age

The range of cues that scleral coloration provides to an individual’s health partly overlaps with those indicative of the individual’s age. For instance, yellowing of the sclera, caused by fat deposition between collagen fibres over the years, is a visible sign of ageing (Watson & Young 2004; Broekhuyse 1975)^{32, 33}. In addition, age-related lipids accumulation is visible in other eye structures, such as arcus senilis (Salvi et al. 2006)³⁴, which can further enhance the impression of the eye having a yellowish coloration. Also, continuous exposure to sunlight and ultraviolet radiation, which causes colour changes in the elastic fibres in the conjunctiva, may be another factor causing the yellowing of the sclera (Grundl 2012)³⁵. Similarly, it has been suggested that cumulative lifetime experience of health conditions affecting blood vessel dilation in the conjunctiva may result in a higher baseline redness in older eyes (cf. Russell et al. 2014)³¹.

All the above factors may cause the eyes to turn less white with age (Grundl et al. 2012, Russell et al. 2014, Krammer & Russell, 2022)^31,35,36. Experimental studies show that people indeed correctly associate changes in eye colouration with increased age. Experimentally increasing the redness and yellowness of the sclera in digital photographs leads to the modified stimuli being perceived as older (Russell et al. 2014, Provine et al. 2013)^{30, 31}, and individuals with lighter sclerae are perceived as younger than those with darker sclerae (Russell et al. 2014; Grundl et al. 2012; Provine 2013)^{30, 31, 35}.

1.1.3. Attractiveness

From the perspective of human mate choice, both good health and youthfulness indicate a higher reproductive fitness of an individual (Etcoff 1999, Symons 1979, 1995)^{37, 38, 39}. Therefore, both those attributes are important determinants of human mate value and are preferred in sexual partners (e.g. Buss, 2019).⁴⁰ As a result, sexual selection will favor visual markers of these qualities; that is, individuals displaying signals and cues of health and youthfulness – such as a white sclera – should be perceived as attractive and preferentially selected as sexual partners.

This prediction has indeed been confirmed in several experiments with digital manipulations of scleral coloration. For example, individuals with redder and yellower eyes were assessed as less attractive compared to control portraits with the sclera left unmodified (Provine et al., 2011, 2013; Russel et al. 2014)^{29, 30, 31}. By analogy, individuals with darkened sclerae were also perceived as less attractive than those with untinted eyes (Russell et al. 2014)³¹. Importantly, scleral whiteness directly measured on individual and unmodified pictures of the eye region in women was also positively correlated with attractiveness ratings (Grundl et al. 2012)³⁵. However, the positive effect of sclera whiteness on attractiveness may be limited. This is because enhancing sclera whiteness above original values (i.e. having "super-white" eyes) did not increase ratings of attractiveness of the photographed eye regions (Provine et al. 2013)³⁰. In the context of sexual preferences, it is also worth noting that the attractiveness rating may be further modified by the sexually dimorphic nature of the ocular morphology. As recent findings show, both sclera shape (Danel et al. 2018, 2020) ^{12, 41} and colour (Krammer & Russel, 2022) ³⁶ are sexually dimorphic at least in Caucasians. However, the link between attractiveness and differences in ocular morphology in men and women should be inspected in more detail, as it may be mediated by age (brighter sclerae are associated with youth, and younger people are perceived as more attractive).

1.1.4. Aggressiveness

A preference for eye-like stimuli in newborns (Dupierrix et al. 2013)⁴² and its subsequent development in neurotypicals, and more broadly, for specifically human patterns of complex sociality may have built on selection against reactive aggression in our lineage. Reduced emotional reactivity to eye contact may have enabled the other signalling functions of the eye, and in particular the co-option of eye-gazing signals for purposes other than negotiating dominance in agonistic displays. The idea that scleral depigmentation could be a by-product of selection against aggression was proposed by Perea-Garcia et al. (2019)¹¹. The authors showed that humans and bonobos (Pan paniscus) have relatively light sclerae compared to common chimpanzees (Pan troglodytes). Such observation, later confirmed by other authors (Mearing & Koops, 2021)⁴³, led to the question of why closely related species evolved different patterns of eye colouration. Selection against aggression in humans (Hare, 2017)⁶ and bonobos (Hare et al., 2012)⁴⁴ are one candidate answer. The self-domestication hypothesis (Hare et al., 2012)⁴⁴ proposes that many of the traits that characterize humans (increased fecundity from an earlier age, reduced reactive aggression, increased period of socialization, extended weaning periods, reduced facial prognatism) are by-products of selection against aggression, similarly to a suite of seemingly unrelated traits that appear in domesticates compared to their wild counterparts (Wilkins et al., 2014)⁴⁵. When compared to the wild-type, depigmentation is one of the most commonly found features in domesticated animals (Wilkins et al. 2014, Balayev et al. 198922)^{45, 46.} This is due to the shared embryogenic origin of melanoblasts and the adrenal medulla (Wilkins et al., 2014)⁴⁵. Thus, by selecting reduced emotional reactivity, reduced pigmentation is also indirectly selected. While scleral depigmentation may be originally a functionless byproduct of selection against aggression, it is possible that bright sclerae became a marker of reduced emotional reactivity, which could then itself become target of selection, as it would reliably announce less emotionally reactive individuals (Perea-Garcia, 2019)¹¹.

1.1.5. Trustworthiness

Finally, a body of evidence supports the idea that the whiteness of the human sclera functions to advertise gaze direction as well as a range of information regarding the internal state of the owner of the eyes. As best illustrated by the wide utility of eye-tracking technology in behavioral research, gaze is a rich, reliable and relatively easily observable source of information on an individual’s cognitive processing, including mindfulness and cognitive load (e.g. Bixler and D’Mello 2021⁴⁷, Perhofer and Refner 2019⁴⁸), but in particular one’s attentional states. Naturally, all this valuable information can be accessed by conspecifics and other animals, and potentially exploited in competitive contexts. Indeed, apes have been shown not only to make use of gaze cues in competitive foraging (Bräuer et al. 2005⁴⁹, Kano and Call 2014⁵⁰), but also to be aware of the information carried by their own gaze, and occasionally to use own gaze direction strategically to mislead competitors (Hall et al. 2017⁵¹). Similarly, humans are aware of the social information that is leaked through their gaze direction, and make strategic decisions on what to look at, depending on whether or not they believe their gaze is visible to third parties (Dudarev et al. 2022⁵², see also Myllyneva and Hietanen 2015⁵³; Foulsham and Lock 2015⁵⁴; Hausfeld, von Hesler, and Goldlücke 2020⁵⁵. In short, gaze is one of the basic predictors of immediate behavioural intent that humans use in interaction, including emotionally based intent (such as approach vs avoid, cf. Adams & Kleck 2005⁵⁶) and aggressive action (such as hand-to-hand combat, cf. e.g. Hausegger et al. 2019⁵⁷).

The depigmented sclera characteristic of humans functions to advertise gaze direction, helping conspecifics infer one’s attentional states, which has been proposed to facilitate cooperation. Kobayashi and Kohshima (2001: 433)² proposed that “gaze-signal enhancement might aid the conspecific communication required for increased co-operative and mutualistic behaviours to allow group hunting and scavenging”, an insight later developed into the “cooperative eye” hypothesis (Tomasello et al. 2007³, see also Wolf et al. 2023⁵⁸), which claims that enhanced gaze-following would have been adaptive by facilitating coordination and thus increasing the effectiveness of collaborative subsistence tasks (see also, Perea-García et al. 2017⁴). There is however a complementary account of the relation between a depigmented sclera and cooperativeness, not focusing on the efficiency of cooperative action but instead on evaluating the reliability of the cooperative partner. As is well-known, cooperation is typically not an evolutionarily stable strategy (Axelrod 1984⁵⁹), because despite the benefits it might bring in the long term, it is typically outcompeted by defection in the short term. Partner choice is therefore key, since cooperation is only profitable on condition one chooses reliable partners that one can trust to act as cooperators rather than defectors (e.g. Roberts 2015)^60. A depigmented sclera makes one’s gaze patterns easier to observe, thus facilitating inferences about that individual’s knowledge, attention and intentions that are crucial in predicting behavior. By making the individual more readable and easier to monitor, a depigmented sclera could serve as an advertisement of their honesty, trustworthiness, and reliability. Humans indeed routinely use gaze information in assessing the trustworthiness of interactants, and in particular strongly associate gaze aversion with dishonesty; despite a lack of evidence in its support, the belief that “liars avert gaze” is in fact the single most dominant “pan-cultural stereotype” about liars (Global Deception Research Team 2006⁶¹, see also Slessor et al. 2011)⁶² Interestingly, a recent bottom-up study into facial correlates of trustworthiness (Mo et al. 2022)⁶³ identified a darkened pupil-iris area (thus, increased perceptual contrast with the surrounding white sclera) as the only facial feature stably cross-culturally correlated with higher trustworthiness judgments.

We developed a rating study with the faces of artificially generated ‘generalised hominins’, which allowed us to collect evaluations of humanlike faces with a bright versus dark sclera while keeping the participants unaware of the experimental modification. The hominin facial morphologies used in the study made the embedding of both sclera variants (i.e. bright, humanlike vs dark, apelike) look sufficiently natural that it did not attract conscious attention from the participants. Equally, the facial morphologies were sufficiently humanlike for the participants to feel natural rating this stimulus on human attributes. Our study was conducted in accordance with the principles of the Declaration of Helsinki. Ethical approval for this study was obtained from the Scientific Research Ethics Committee at the University of Warmia and Mazury in Olsztyn (decision 16/2021). Our study was pre-registered with the AsPredicted service at https://aspredicted.org/TM4_PSK

2.1. Stimulus

Stimulus in our study was a set of 20 artificially created realistic facial images of generalised ancient hominins, each in a bright sclera and dark sclera version. As source images we used 31 medium- to high-quality facial images of reconstructions of a range of extinct hominin species (see Additional online information, https://osf.io/7cr9e/?view_only=49f985066f77441fb29e015aa077c6a3), acquired through targeted internet searches. The "Facemixer" utility in the Abrosoft FantaMorph Deluxe for Windows software was applied to mix the features and shapes of the source faces; typically three to four input faces were used to compose an individual output face, which made it possible to retain considerable individual detail of the input faces. Each of the resulting output images was corrected by an artist to preserve the visual uniformity of the set. Photoshop CC was used to eliminate artefacts, introduce specific corrections such as slight opening of the mouth, and add a black mask around the face; each image was edited individually. The Lightroom CC software was used to improve the overall visual quality of the images. This resulted in 38 realistic facial images of individually distinctive ‘hominins’ portrayed en face, looking straight and with a neutral facial expression; 20 were selected to form the final set to maximise individual differences between the images. Finally, the brightness of the scleral areas of each facial image was adjusted by +50 Photoshop brightness points to create a “humanlike” bright sclera and, depending on the image, by either -150 or -200 points to create a “generalised apelike” dark sclera version (Fig. 1). Extensive piloting that preceded the study confirmed that in an overwhelming majority of cases, the experimental manipulation is not detected by the participants.

2.2. Procedure

250 demographically diverse participants informedly consented to taking part in the experiment through the online crowdsourcing platform Prolific. Eligibility criteria were set to having completed at least two previous studies, 100% approval rate on previous studies, and declaring normal or corrected to normal vision as well as fluency in English. Participants could use desktop or laptop computers, but not tablets or smartphones. Participation was remunerated (1 GBP). The study was implemented with the Labvanced online application. Each participant was shown 10 bright-sclera and 10 dark-sclera faces, presented in random order. The stimuli were described to the participants as “the faces of prehistoric ancestors of modern humans (so-called hominins, i.e., "apemen" or "cavemen")”. Each participant saw the faces of all the 20 hominins in the set, but only one condition per face; that is, participants saw each face only once, with either bright or dark sclerae. Participants rated the faces on five dimensions – age, health, aggressiveness, attractiveness (“to others like her/him”), and trustworthiness – by moving sliders between the left (0) and right (100) extremes for each trait. The extremes of the scale had verbal labels (e.g. “very young” and “very old” for Age), but the underlying numerical values were not made visible to the participants.

2.3. Analysis

First, we excluded participants who did not complete the task, or whose completion time was shorter than Q1 - 1.5x of the interquartile range (IQR) in our sample (the other planned exclusion criterion, “not moving the slider on any 4 or more consecutive ratings”, was not applied due to this being common in our sample). We excluded 31% of the original data set (n=82), after which the rest of respondents (N=168) were included in analyses. We fit separate generalized mixed models (GLMM) for each of our dependent variables (except Age, see below) with our manipulation (dark/bright sclerae) as an independent variable. For age, we fit a linear regression model (LM). We included stimuli pairs and participants as random effects in all our models (through a software glitch, one of the faces in the bright sclera version was never displayed to participants; this was corrected in Study 2). We predicted that bright-sclera hominin faces would be rated as less aggressive, more attractive, more trustworthy, healthier, and younger than the same faces with a dark sclera.

2.4 Study 1: results

All models found statistically significant differences between pairs of stimuli in the predicted direction. These models are summarized in Table 1 below.

Table 1.- Summary of the models for each of our dependent variables. The model for Age is a linear regression with error structure. All other models are generalized linear models with error structure.

	Intercept	Dark-Bright difference	t/z value	p
Age	63.13	4.03	5.30	<0.01
Aggression	-0.44	0.28	38.369	<0.01
Attractiveness	-0.19	-0.37	-50.76	<0.01
Health	0.51	-0.39	-53.17	<0.01
Trustworthiness	0.24	-0.31	-43.25	<0.01

Study 2 was conducted as a self-replication study to address minor limitations detected in the design of Study 1.

3.1. Study 2: design and procedure

The stimulus and procedure were the same as in Study 1, with the following revisions.

All the stimulus images were equalised for overall brightness level; that is, the overall image brightness were made constant within each bright sclera - dark sclera pair of stimuli.

Participants were asked to complete a screen calibration procedure to ensure a proper rendering of the color scale on the display devices. Small changes were introduced to the phrasing of the instructions for participants, to further ensure compliance with the instructions. The only exclusion criterion for responses was time-based: we excluded all trials completed below 5 seconds. Finally, the Labvanced application settings were corrected so as to ensure that all stimuli items were displayed and that the demographic data of the participants were associated with their responses. We further aimed to increase the demographic diversity of the sample by recruiting four batches of 25 participants (total 100 participants), at times 4:00, 10:00, 16:00, 22:50 CET, so as to make the study equally available to Prolific users across the different time zones. This self-replication study was pre-registered with the AsPredicted service at https://aspredicted.org/blind.php?x=GCY_FXV

3.3. Analysis

We excluded participants that did not complete the task, and we excluded all trials completed in less than 5 seconds. Lastly, we solicited general feedback on the task through two post-study questions: "Did you notice anything particular about any area in the faces that influenced your decisions?", and "Do you have any other comments or other helpful information?". We conducted separate analyses for groups of participants categorised by their mentions of the eyes in the feedback comments: a) eyes not mentioned at all or in an irrelevant way (e.g. “eyebrows”, “scar under the eye”); b) eyes mentioned in general (e.g. “I think the eyes influenced my response”), or; c) specific mention of our IV (e.g. “the darkness of the eye/sclera influenced my response”). These additional results are included in the Additional online information (https://osf.io/7cr9e/?view_only=49f985066f77441fb29e015aa077c6a3) - the results are qualitatively comparable to the main results we present here. We fit separate generalized linear mixed effects models (GLMM, glmer function in the package lme4 of R) for each of our dependent variables (except Age, see below) with our manipulation (dark/light sclerae) as an independent variable. We included participants as random effects in all our models to allow the intercept to vary for each of them. For Age, we first fit a simple linear regression model (lm in base R) to inspect the residuals. Upon confirmation that the residuals conformed to assumptions, we fit a linear mixed effects model (LMM, lmer function in lme4, R) with Age as dependent variable and our manipulation as the independent variable.

3.4. Study 2: results

All models found statistically significant differences between pairs of stimuli in the predicted direction. Boxplots with a comparison in ratings between pairs of stimuli for each variable are shown in Figure 2. These models are summarized in Table 2 below.

Table 2.- Summary of the models for each of our dependent variables. The model for Age is a linear regression with error structure. All other models are generalized linear models with error structure.

	Intercept	Dark-Bright difference	z/t value	p
Age	62.41	4.88	4.78	<0.01
Aggression	-0.12	0.26	26.89	<0.01
Attractiveness	-0.49	-0.23	-22.96	<0.01
Health	-0.01	-0.33	-33.92	<0.01
Trustworthiness	-0.07	-0.23	-23.33	<0.01

The results of our experiments provide first direct experimental evidence that a depigmented sclera typical of H. sapiens is associated with the perception of youth, health, attractiveness, trustworthiness and reduced aggressiveness relative to a pigmented scleral phenotype typical of apes. This extends existing findings that subtle changes in scleral coloration are interpreted as cues to several parameters such as age, health or emotional state (Russell et al. 2014; Gründl et al. 2012, Provine et al. 2013)^{31, 35, 30}, or that human infants have a generalised preference for a humanlike, bright scleral morphology (Dupierrix et al. 2013)⁴², and five-year-olds prefer the human morphology when choosing cooperative partners (Wolf et al. 2023)⁵⁸. The effects we observed are robust: in particular, the self-replication and post-study feedback eliminate alternative interpretations related to factors such as participant pool composition, overall stimulus brightness, lack of direct control over participants display settings, and conscious biases related to the IV.

Our results lay the grounds for an interpretation that the transition from pigmented to depigmented sclerae in the ancestors of H. sapiens could be mediated by several convergent selection pressures related to the signalling of multiple parameters. Of course, drawing evolutionary conclusions from the results of laboratory studies with modern humans is invariably problematic, because inferences about past selection pressures rely on several additional assumptions. In particular, such interpretations necessitate a degree of acceptance of the uniformitarian assumption, whereby the preferences (and with them, selection pressures) detected now can be extrapolated into the past: here, extending the evaluations provided by modern sapiens onto the biologically as well as culturally different populations of extinct hominins. However, some degree of reliance on the uniformitarian assumption underlies a large body of human-evolutionary research, for example, experimental research in the cultural evolution of language applies uniformitarianism as a methodological heuristic (Walkden 2019; Trudgill 2020)^{64, 65}.

More generally, the main limitation in making evolutionary interpretations lies in establishing the chronology of adaptive functions, i.e. distinguishing between the original adaptive function of the complete or partial depigmentation of the sclera and the functions that later co-opt this original structure. For example, the signalling of reduced aggressiveness and greater trustworthiness, and the signalling of age, health, attractiveness may have been shaped, respectively, by selection pressures for cooperative partner choice and sexual selection, which were proposed to be prominent at different points in the evolutionary history of our species (Hare et al., 2012)^44. As we have no way of reaching into the past to interview our extinct ancestors, answers to such questions can only be approximated through comparative research with other animals, in particular other primates (cf. esp. Caspar et al. 2021, Kano et al. 2021a, 2021b; Mearing et al. 2021, Mearing & Koops 2021; Perea García et al. 2019, 2021)^{66, 67, 68, 69, 43, 11, 14}. The recent spike in such studies highlights the importance of comparative research in assessing the adaptive significance of human ocular morphology. While far from conclusively supporting that our ancestors perceived differences in scleral brightness the way our participants did, our results support said notion.

Finally, we see our study as a methodological contribution to this fast-developing field. Interest in the adaptive significance of human ocular morphology was sparked a quarter-century ago with the influential study by Kobayashi and Kohshima (1997)¹, and has now been revived with several very recent publications (Danel et al. 2020)¹². This particular wealth of studies within the last few years shows not only that this is a very timely topic, but most importantly testifies to the urgency of discussion on developing methodological standards. Our approach of using artificial humanlike facial images that represent no extant human population complements existing approaches that rely on humanlike stimuli which however fall outside of naturally occurring phenotypes (e.g. Yorzinski et al. 2021, Wolf et al. 2023)^{70, 58}. It also has the advantage of tapping into our perceptual systems without triggering culturally acquired associations relative to extant human groups.

Author contributions

SW conceptualised the study, coordinated stimulus acquisition and data collection, and co-led the writing of the manuscript.

JOPG co-led the writing of the manuscript, analysed the data, and contributed to conceptualising the study design and preparing the stimulus.

DD contributed to writing the manuscript, analysing the data, conceptualising the study design and preparing the stimulus.

ZL prepared the stimulus for the studies.

Competing interests
The authors declare no competing interests.

Data availability statement

The datasets generated and analysed during the current study are available in the OSF repository, https://osf.io/7cr9e/?view_only=49f985066f77441fb29e015aa077c6a3

Ethical approval

Ethical approval for this study was obtained from the Scientific Research Ethics Committee at the University of Warmia and Mazury in Olsztyn (decision 16/2021).

Acknowledgements

We are grateful to the members of the Center for Language Evolution Studies, NCU Toruń, for their feedback and help in developing the study. We are also grateful to dr Daria Różycka for an intellectual property rights analysis. All errors and omissions are exclusively ours.

Work on this article was supported by the cooperation agreement between Nicolaus Copernicus University, Toruń and Hirszfeld Institute of Immunology and Experimental Therapy Polish Academy of Sciences.

Image credits

The following images were used to create the composite faces in Figure 1:

https://www.frontiersin.org/articles/10.3389/fevo.2021.639048/full

https://www.flickr.com/photos/sortingoutscience/5071085511/sizes/o/

https://commons.wikimedia.org/wiki/File:Homo_habilis_-_forensic_facial_reconstruction.png

Additional information

https://osf.io/7cr9e/?view_only=49f985066f77441fb29e015aa077c6a3

Kobayashi, H. & Kohshima, S. Unique morphology of the human eye. Nature 387, 767–768 (1997).
Kobayashi, H. & Kohshima, S. Unique morphology of the human eye and its adaptive meaning: comparative studies on external morphology of the primate eye. J. Hum. Evol. 40, 419–435 (2001).
Tomasello, M., Hare, B., Lehmann, H. & Call, J. Reliance on head versus eyes in the gaze following of great apes and human infants: the cooperative eye hypothesis. J. Hum. Evol. 52, 314–320 (2007).
Perea-García, J.O., Ehlers, K.R. & Tylén K. Bodily constraints contributing to multimodal referentiality in humans: the contribution of a de-pigmented sclera to proto-declaratives. Lang. Commun. 54, 73–81 (2017).
Knight, C. & Lewis, J. Wild voices: mimicry, reversal, metaphor, and the emergence of language. Curr. Anthropol. 58, 4; https://doi.org/10.1086/692905 (2017).
Hare, B. Survival of the friendliest: Homo sapiens evolved via selection for prosociality. Annu. Rev. Psychol. 68(1), 155–186 (2017).
Perea-García et al., 2022 (under review)
Kano, F., Kawaguchi, Y. & Yeow H. Experimental evidence that uniformly white sclera enhances the visibility of eye-gaze direction in humans and chimpanzees. eLife 11, e74086; https://doi.org/10.7554/eLife.74086 (2022).
Whitham, W., Schapiro, S. J., Troscianko, J., & Yorzinski, J. L. Chimpanzee (Pan troglodytes) gaze is conspicuous at ecologically-relevant distances. Sci. Rep. 12, 9249; https://doi.org/10.1038/s41598-022-13273-3 (2022).
Whitham, W., Schapiro, S. J., Troscianko, J., & Yorzinski, J. L. The gaze of a social monkey is perceptible to conspecifics and predators but not prey. Proc. R. Soc. B. 289, 20220194; https://doi.org/10.1098/rspb.2022.0194 (2022).
Perea-García, J. O., Kret, M. E., Monteiro, A., & Hobaiter, C. Scleral pigmentation leads to conspicuous, not cryptic, eye morphology in chimpanzees. Proc. Natl. Acad. Sci. 116(39), 19248–19250 (2019).
Danel, D.P. et al. Sex differences in ocular morphology in Caucasian people: a dubious role of sexual selection in the evolution of sexual dimorphism of the human eye. Behav. Ecol. Sociobiol. 74, 115; https://doi.org/10.1007/s00265-020-02894-1 (2020).
Grieve, K. et al. Three-dimensional structure of the mammalian limbal stem cell niche. Exp. Eye Res. 140, 75–84 (2015).
Perea-García, J. O., Danel, D. P. & Monteiro, A. Diversity in primate external eye morphology: previously undescribed traits and their potential adaptive value. Symmetry 13, 1270; https://doi.org/10.3390/sym13071270 (2021).
Hingorani, M., Hanson, I., & Van Heyningen, V. Aniridia. Eur. J. Hum. Genet. 20, 1011–1017 (2012).
Leibowitz, H. M. The red eye. N. Engl. J. Med. 343, 345–351 (2000).
Roche, S. P. & Kobos, R. Jaundice in the adult patient. Am. Fam. Physician 69, 299–304 (2004).
Mueller, J. B. & McStay, C. M. Ocular infection and inflammation. Emerg. Med. Clin. North Am. 26, 57–72 (2008).
Donshik, P. C. Allergic conjunctivitis. Int. Ophthalmol. Clin. 28, 294–301 (1988).
Doshi, R. & Noohani, T. Subconjunctival hemorrhage. 2022 Feb 23. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing; 2022 Jan–. https://www.ncbi.nlm.nih.gov/books/NBK551666/ (2022).
Owen, C. G., Newsom, R. S. B., Rudnicka, A. R., Ellis, T. J. & Woodward, E. G. Vascular response of the bulbar conjunctiva to diabetes and elevated blood pressure. Ophthalmology 112, 1801–1808 (2005).
Paton, D. The conjunctival sign of sickle cell disease. Arch. Ophthalmol. 66, 90–94 (1961).
Patel, S.J. & Lundy, D.C. Ocular manifestations of autoimmune disease. Am. Fam. Physician. 66(6), 991–998 (2002).
Sundelin, T. et al. Cues of fatigue: effects of sleep deprivation on facial appearance. Sleep. 36(9), 1355–1360 (2013).
McLane, N. J. & Carroll, D. M. Ocular manifestations of drug abuse. Surv. Ophthalmol. 30, 298–313 (1986).
Li, J., Tripathi, R.C. & Tripathi, B.J. Drug-induced ocular disorders. Drug Saf. 31(2), 127–141 (2008).
Dhingra, D., Kaur, S. & Ram, J. Illicit drugs: effects on eye. Indian J. Med. Res. 150(3), 228–238 (2019).
Al-Qaisi, M. T., Batheja, M., & Fallon, M. B. Jaundice in Cecil Essentials of Medicine (eds. Wing, E. & Schiffman, F.) 420–425 (Elsevier, 2021).
Provine, R.R., Cabrera, M.O., Brocato, N.W. & Krosnowski, K.A. When the whites of the eyes are red: a uniquely human cue. Ethology 117, 395–399 (2011).
Provine, R.R., Cabrera, M.O. & Nave-Blodgett, J. Red, yellow, and super-white sclera. Hum. Nat. 24, 126–136 (2013).
Russell, R., Sweda, J.R., Porcheron, A. & Mauger, E. Sclera color changes with age and is a cue for perceiving age, health, and beauty. Psychol. Aging. 29(3), 626–635 (2014).
Watson, P.G. & Young, R.D. Scleral structure, organisation and disease: a review. Exp. Eye Res. 78(3), 609–623 (2004).
Broekhuyse, R.M. The lipid composition of aging sclera and cornea. Ophthalmologica 171(1), 82–85 (1975).
Salvi, S.M., Akhtar, S. & Currie, Z. Ageing changes in the eye. Postgrad. Med. J. 82(971), 581–587 (2006).
Gründl, M., Knoll, S., Eisenmann-Klein, M. & Prantl, L. The blue-eyes stereotype: do eye color, pupil diameter, and scleral color affect attractiveness?. Aesth. Plast. Surg. 36, 234–240 (2012).
Kramer, S.S. & Russell, R. A novel human sex difference: male sclera are redder and yellower than female sclera. Arch. Sex Behav. 51, 2733–2740 (2022).
Etcoff, N. Survival of the prettiest (Doubleday, 1999).
Symons, D. The evolution of human sexuality (Oxford University Press, 1979).
Symons, D. Beauty is in the adaptations of the beholder in Sexual nature, sexual culture (eds. Abramson, P. R. & Pinkerson S. D.) 80–118 (University of Chicago Press, 1995).
Buss, D.M. Evolutionary psychology: the new science of the mind (Routledge, 2019).
Danel, D.P., Wacewicz, S., Lewandowski, Z., Żywiczyński, P. & Perea-Garcia, J.O. Humans do not perceive conspecifics with a greater exposed sclera as more trustworthy: a preliminary cross-ethnic study of the function of the overexposed human sclera. acta ethol. 21, 203–208 (2018).
Dupierrix E, de Boisferon AH, Méary D, et al. Preference for human eyes in human infants. J Exp Child Psychol. 123, 138–146 (2014).
Mearing, A. S. & Koops, K. Quantifying gaze conspicuousness: are humans distinct from chimpanzees and bonobos? J. Hum. Evol. 157, 103043; https://doi.org/10.1016/j.jhevol.2021.103043 (2021).
Hare, B., Wobber, V., & Wrangham, R. The self-domestication hypothesis: evolution of bonobo psychology is due to selection against aggression. Anim. Behav. 83(3), 573–585 (2012).
Wilkins, A. S., Wrangham, R. W. & Fitch, W. T. The "domestication syndrome" in mammals: a unified explanation based on neural crest cell behavior and genetics. Genetics 197(3), 795–808 (2014).
Belyaev, D. K. & Trut, L. N. The convergent nature of incipient forms and the concept of destabilizing selection in Vavilov’s Heritage in Modern Biology (eds. Ovchinnikov, Y. A. & Rapoport, I. A.) 155–169 (Nauka, 1989).
Bixler, R. E. & D'Mello S. K. Crossed Eyes: Domain Adaptation for Gaze-Based Mind Wandering Models. In: ACM Symposium on Eye Tracking Research and Applications (ETRA '21 Full Papers). Association for Computing Machinery 8, 1–12. https://doi.org/10.1145/3448017.3457386 (2021).
Perkhofer, L. & Lehner, O. Using Gaze Behavior to Measure Cognitive Load. In: Davis, F., Riedl, R., vom Brocke, J., Léger, P. M., Randolph, A. (eds) Information Systems and Neuroscience. Lecture Notes in Information Systems and Organisation, vol 29. Springer, Cham. https://doi.org/10.1007/978-3-030-01087-4_9 (2019).
Bräuer, J., Call, J. & Tomasello, M. All great ape species follow gaze to distant locations and around barriers. J Comp Psychol. 119(2), 145–154 (2005).
Kano, F. & Call, J. Great Apes Generate Goal-Based Action Predictions: An Eye-Tracking Study. Psychological Science 25(9), 1691–1698 (2014).
Hall, K. et al. Chimpanzee uses manipulative gaze cues to conceal and reveal information to foraging competitor. Am J Primatol. 79, e22622; https://doi.org/10.1002/ajp.22622 (2017).
Dudarev, V., Kamatani, M., Miyazaki, Y., Enns, J.T. & Kawahara J.I. The Attractiveness of Masked Faces Is Influenced by Race and Mask Attitudes. Front. Psychol. 13, 864936; https://doi.org/10.3389/fpsyg.2022.864936 (2022).
Myllyneva, A. & Hietanen, J.K. There is more to eye contact than meets the eye. Cognition 134:100–109. https://doi.org/10.1016/j.cognition.2014.09.011 (2015).
Foulsham, T. & Lock, M. How the eyes tell lies: Social gaze during a preference task. Cogn. Sci. 39, 1704–1726 (2015).
Hausfeld, J., Hesler, K. & Goldlücke, S. Strategic gaze: an interactive eye-tracking study. Exp. Econ. 24(1), 177–205 (2020).
Adams, R. B. Jr., & Kleck, R. E.. Effects of Direct and Averted Gaze on the Perception of Facially Communicated Emotion. Emotion 5(1), 3–11 (2005).
Hausegger, T., Vater, C. & Hossner, E.J. Peripheral Vision in Martial Arts Experts: The Cost-Dependent Anchoring of Gaze. J. Sport Exerc. Psychol. 41, 1–9 (2019).
Wolf, W., Thielhelm, J. & Tomasello, M. Five-year-old children show cooperative preferences for faces with white sclera. J. Exp. Child Psychol. 225, 105532; https://doi.org/10.1016/j.jecp.2022.105532 (2023).
R.M., Axelrod & Hamilton., W. D. The Evolution of Cooperation. (New York: Basic Books, 1984).
Roberts, G. Partner Choice Drives the Evolution of Cooperation via Indirect Reciprocity. PLOS ONE 10, e0129442; https://doi.org/10.1371/journal.pone.0129442 (2015).
Voracek, M. et al. A World of Lies. J. Cross-Cult. Psychol. 37, 60–74 (2006).
Slessor, G. et al. Investigating the “Deceiver Stereotype”: Do Older Adults Associate Averted Gaze With Deception? J. Gerontol. Ser. B. 67(2), 178–183 (2012).
Mo, C. et al. Culture-free perceptual invariant for trustworthiness. PLOS ONE 17, e0263348; https://doi.org/10.1371/journal.pone.0263348 (2022).
Walkden, G. The many faces of uniformitarianism in linguistics. Glossa J. Gen. Linguist. 4, 1–17 (2019).
Trudgill, P. Prehistoric Sociolinguistics and the Uniformitarian Hypothesis: What Were Stone-Age Languages Like?. A: Millennia of Language Change: Sociolinguistic Studies in Deep Historical Linguistics. Cambridge: Cambridge University Press, pàg, 7–16 (2020).
Caspar, K. R., Biggemann, M., Geissmann, T. & Begall, S. Ocular pigmentation in humans, great apes, and gibbons is not suggestive of communicative functions. Sci Rep 11, 12994; https://doi.org/10.1038/s41598-021-92348-z (2021).
Kano, F., Sato, Y., & Yamanashi, Y. How chimpanzees look at movies: The" Art and Science" project in Kyoto City Zoo. Japanese Journal of Animal Psychology, 1–71 (2021).
Kano, F., Kawaguchi, Y., & Hanling, Y. Experimental evidence that uniformly white sclera enhances the visibility of eye-gaze direction in humans and chimpanzees. Elife, 11, e74086; https://doi.org/10.7554/eLife.74086 (2022).
Mearing, A. S., Burkart, J. M., Dunn, J., Street, S. E., & Koops, K. The evolutionary origins of primate scleral coloration. bioRxiv 07.25.453695; doi: https://doi.org/10.1101/2021.07.25.453695 (2021).
Yorzinski, J. L., Thorstenson, C. A. & Nguyen, T. P. Sclera and Iris Color Interact to Influence Gaze Perception. Front. Psychol. 12, 632616; (2021).

No competing interests reported.

Download PDF

Editorial decision: Major revision
17 Oct, 2022
Reviewers agreed at journal
04 Oct, 2022
Reviews received at journal
19 Sep, 2022
Reviewers agreed at journal
02 Sep, 2022
Reviewers invited by journal
29 Aug, 2022
Editor assigned by journal
29 Aug, 2022
Editor invited by journal
29 Aug, 2022
Submission checks completed at journal
29 Aug, 2022
First submitted to journal
24 Aug, 2022

You are reading this latest preprint version

The Adaptive Significance of Human Scleral Brightness. An Experimental Study

Status:

Version 1

Abstract

Figures

1. Introduction

2. Study 1

3. Study 2

4. Discussion

Declarations

References

Additional Declarations

Status:

Version 1