Unit 2 Lesson: Portraits as Faces

Abstract

Portraits have been a part of human civilization for thousands of years. According to the famous Roman historian and naturalist, Pliny the Elder (AD 23/24–79), the first painting ever made, was a portrait. It was presumably invented by the daughter of a man called Butades, a potter from Sicyon in Greece. To preserve the memory of her beloved who was going away for a while, she decided to trace the profile of his head from the shadow cast on a wall.

Portraits have been a part of human civilization for thousands of years. According to the famous Roman historian and naturalist, Pliny the Elder (AD 23/24–79), the first painting ever made, was a portrait. It was presumably invented by the daughter of a man called Butades, a potter from Sicyon in Greece. To preserve the memory of her beloved who was going away for a while, she decided to trace the profile of his head from the shadow cast on a wall.

With ‘portrait’ we think of a representation of a person in drawing, painting, sculpture, or photography. It can be of the whole body or only the head, but generally we take for granted that it fixates the physiognomy of the face in such a way that we can recognize the identity of the person. Since we want to know who the painting represents, the image must be the likeness of the model. In addition, we also sometimes require that the image conveys some of the model’s personality and inner feelings. Our tendency to endow images with a presence is a phenomenon that can be investigated scientifically under the heading ‘theory of mind’—a psychological term that refers to our capacity to understand other people’s emotions (Folgerø et al., 2016; Martínez, 2020).

Many of the portraits that we know from history’s greatest artists were painted long before the invention of photography, and had, perhaps precisely for this reason, much of the same social function that photographs today have. But can we from this draw the conclusion that the audience before, say, 1800 perceived a real person in the painting and that the spectator’s encounter with a portrait had some of the atmosphere of a meeting between real people? A painting is a work of art, and a portrait seen in a painting is not the same as the face of a physical person. Studying paintings is not the same as studying faces. Yet, from art history, we have a myriad of examples that show that people actually saw (or wanted to see) real people in the pictures. If we look attentively at images of God the Pantokrator in mosaics typical of the Orthodox East in the Middle Ages, or icons with the same subject from that period and region, we will often find that the eyes, although the face is turned outwards towards the spectator, are slightly averted. It is possible that this tendency can be explained with reference to the belief in the ‘evil eye’ which was widespread, especially in the Eastern Mediterranean area. Superstitious people believed that a malevolent person could cast a curse on another person by means of the gaze, causing injury or misfortune. The gaze of God was, of course, the opposite of malevolent, but the tradition testifies the power that people attributed to the gaze, whether good or bad—even when cast by a painting! This is probably also the reason why in Byzantine depictions of the Last Supper Judas never had his face turned towards the spectator (Gombrich, 1961).

Although, as mentioned, it is the likeness that makes us accept a picture of a person as a portrait, the likeness is not necessarily as faithful as a photograph. We have an innate desire to see a thing ‘as something’. You have probably been lying on your back in the grass on a summer day, looking at the clouds that drift in the sky. The clouds have different shapes, and often they resemble something, perhaps an animal or some other object. Our capacity to discover hidden figures behind inanimate objects is spurred by a desire for meaning. It is probably this tendency—our desire to look for meaning where there is none—that made people in different cultures and historical epochs ascribe magical effects to certain images. From the medieval Christian tradition we have the so-called acheiropoieta—pictures ‘made without hands’. Having come into being miraculously, without the participation of a painter, such images were believed to be true images of the Saviour. In a culture without photographic reproductions, this would almost be a direct confrontation between the faithful and God. Acheiropoieta were venerated as relics and pilgrims travelled long distances to see them.

A similar status was ascribed to the so-called vera icona (which means ‘true image’). This refers to the story of a pious woman called Veronica (her name is an anagram for vera icona), who was a witness to Christ when he carried his cross to Golgotha. Moved to pity, she gave him her veil so that he might wipe the sweat of his face. When the veil was returned to her, she discovered that the Lord’s facial traits were mysteriously impressed upon it (The veil is sometimes referred to as a sweat cloth or sudarium [Hodne, 2013a]). A number of churches claim to be in possession of Veronica’s cloth, which is venerated as a relic like the acheiropoieta.

There is an interesting link between the medieval Veronica tradition and the more exact and historically reliable example narrated by Pliny. Both stories explain how the face of a person is traced or imprinted mechanically on a surface, to convince us that this is the true image of the model, not a product of phantasy or artistic creativity. The difference between the two examples lies in the head direction, for while in Pliny’s example the image is an exact profile, the face on Veronica’s clothes is seen frontally. It is, as we shall see, the frontal image of Veronica, not Pliny’s profile, that is ascribed to supernatural powers. In fact, the frontal image exploits a well-known optical effect that we will briefly explain. Some of you may know the famous 1914 poster by Alfred Leete that was used in a campaign to recruit soldiers for the British Army during World War I (Fig. 1). The poster shows the Secretary of State for War, Lord Kitchener, with his face turned outwards and a finger pointing towards an imagined spectator. The accompanying text ‘… wants you’ is not a continuation of the word ‘Britons’ at the top but must refer to the face of Lord Kitchener himself, whose eyes seem to follow the spectator in front of the image. In fact, in designing this poster, Leete exploited a well-known optical trick that was documented already in Antiquity. The first to talk about such cases of ‘omnivoyance’ was probably, again, the Roman historian Pliny. In his Naturalis Historia, he described a painting of the goddess Minerva, made by the artist Famulus, which had the appearance of always looking at the spectators, from whatever point it was viewed. The same optical effect was reported in the following century by the Assyrian satirist Lucian (AD 125–180), who, during a visit to the Temple of Hera in Hierapolis saw an image of a goddess that seemed to look you ‘…in the face, and as you pass it the gaze still follows you, and if another approaching from a different quarter looks at it, he is similarly affected’ (Lucian, 1913).

Fig. 1

Alfred Leete, Lord Kitchener Wants You. Published in London Opinion 1914

Although ancient Greek scientists like Ptolemy (AD 90–168) tried to give the phenomenon a scientific explanation, using terms normally associated with Renaissance perspective, like ‘visual cone’ and ‘visual axis’ (Ptolemy, 1996), medieval authors persistently saw it as mystic and enigmatic. In fact, the divine image that the thirteenth-century theologian Nicholas Mesarites described in the Church of the Holy Apostles, was that of the enthroned Christ, the divine judge, whose eyes are ‘wholly directed toward all at once and at the same time toward each individually’ (Hodne, 2013a; Pliny, n.d.). The eyes that fall upon the righteous are mild and gentle, he said, whereas to ‘those … who are condemned by their own judgment they are scornful and hostile’ (Mesarites, 1957).

The Renaissance theologian Nicholas of Cusa used exactly this effect in a treatise called De visione Dei that addressed certain aspects that he believed to be characteristic of God as an Almighty being. Writing to fellow monks in a monastery, he described God as an omnipotent, endowed with the capacity to see everyone, everywhere, all the time. Interestingly, to illustrate the effect, along with the letter he also sent a drawing of the vera icona type. The term that Cusa used to denote the effect was ‘omnivoyance’, which means that our feeling as spectators of being watched must be attributed to the image itself, if not even the one (God) who is represented in the image. The word ‘omnivoyance’ is fitting, for it denotes the effect as well as all those images that possess the quality to induce it. However, if you try to look up the word on the Internet, you might find a different term instead: the ‘Mona Lisa effect’. This is commonly used in research articles on this topic. The name ‘Mona Lisa-effect’ was probably introduced for the first time in an article by Kinya et al. (1985), which published the responses from 12 test persons in 4 experiments, using (1) photographs with a direct gaze and slightly averted face (like in Leonardo’s Mona Lisa), (2) a schematic face, (3) a full face (en face) photograph with direct gaze, and (4) images with averted gaze. The persistence of perceived gaze contact through slanted stimuli (face direction) was by this group explained by a hypothesized gaze anchoring effect (Kinya et al., 1985).

From this, we learn that the optical effect of ‘being observed’ is not limited to pictures like that of Christ in Majesty, where the face of God is turned directly towards us, but also works like Leonardo da Vinci’s famous Mona Lisa in the Louvre. In fact, the gaze of Leonardo’s model (presumably a portrait of the Florentine noblewoman Lisa Gherardini) follows us in exactly the same way as the aforementioned examples, even if in this case only the eyes are directed towards the spectator, the face is seen in three-quarter profile (Fig. 2). This should remind us that there is nothing magic with holy faces. It is an optical effect that can be studied scientifically. One aspect of it was addressed in a famous study by W. H. Wollaston as early as 1824 (Wollaston, 1824). Wollaston’s study focused on pictures of the Mona Lisa type, where the eyes of the sitter are directed towards the spectator while the face is seen in three-quarter profile. Realizing that the optical illusion only works when the eyes of the model are turned out from the picture at an angle perpendicular to the picture surface, Wollaston’s question was not how the omnivoyance effect functions, but how the artist must paint the eyes to induce the feeling that the portrayed person is looking at us. The answer to this question is not at all easy, as the image that accompanied his text demonstrates (Fig. 3). In fact, the eyes of the two heads that we see here, the one to the left and the one to the right, are exactly the same, yet only the one on the left appears to look at us. This shows that gaze direction is not only a question of eye direction—also the direction of the face must be considered.

Fig. 2

Leonardo da Vinci, Mona Lisa (La Gioconda), Louvre, Paris

Fig. 3

Wollaston, W.H. Page from ‘On the Apparent Direction of Eyes in a Portrait’, 1824

How can this effect be explained optically and psychologically? According to some the human visual system has a tendency to recreate the original viewpoint of a scene by compensating for the retinal distortions that arise from an observer’s atypical vantage points (Kubovy, 1986). A number of studies have found empirical evidence for the observer’s capacity to compensate for the slant of the projection surface (Boyarskaya & Hecht, 2012). Others have maintained that the reason why the portrait’s eyes appear to follow the observer is that the shape of the nose and eyes, as well as the outline of the face, remain relatively stable during the observer’s horizontal displacement in front of the picture (Halloran, 1989). With the exception of extremely oblique angles, a face looks more or less the same from all positions. This means that all images in a certain sense move with the spectator, even in portraits with their eyes turned towards the side (Boyarskaya & Hecht, 2012).

It is almost certain that the images described by Mesarites and Cusa, differently from Leonardo’s Mona Lisa, were full frontal. Although it must have been very useful to exploit the omnivoyance effect for images of God, the fact that it can be obtained in painted and photographic images regardless of whether the face is turned directly towards the spectator or is slightly averted, means that it cannot by itself be the hallmark of the representations of ‘the divine’ in art. It is therefore worthwhile to study face direction independently of gaze direction.

Based on a hypothesis that full face view in Western art was more common in depictions of holy persons (Christ, saints) than secular persons, a survey was performed based on reproductions of artworks in books and catalogues of Italian, German, and Flemish art from the fifteenth and the sixteenth centuries. The catalogues used were Max Friedländer’s Die altniederländische Malerei; Ernst Buchner’s Das deutsche Bildnis der Spätgotik und der frühen Dürerzeit; Werner Richard Deusch’s German Painting of the Sixteenth Century: Dürer and His Contemporaries; and Raimond van Marle’s comprehensive The Development of the Italian Schools of Painting.

The masterpieces by European artists contained in these catalogues comprised all types of subjects and paintings, not only portraits. A definition of ‘portrait’ was needed, which limited the search to images that included the face of a person, and, in some cases, the shoulders and breast, but not the entire body. Moreover, only portraits of single persons (not group portraits), and only depictions of secular persons were included. In addition, a definition of pose was needed that distinguished between the frontal view of the head and different versions of the profile. Basically, in art one usually differentiates between en face (full-frontal view of the face, like in a passport photo), profile (like the example from Pliny, above), and three-quarter view (the head is turned obliquely to the picture plane). But in this survey only two categories were used—full face and not full face.

Of more than a thousand paintings reproduced in Max Friedländer’s catalogue there were no less than 172 portraits, but, interestingly, none of these was a full face; while of the 188 portraits in Ernst Buchner’s there were two. Also, when surveying Raimond van Marle’s books two en face portraits were found among the total of 196 portraits included in his volumes. Of the 32 portraits in Werner Richard Deusch’s there were three; one of these was also reproduced by Buchner: the famous Self-portrait as Christ by Albrecht Dürer.

True, a few full-frontal portraits of secular persons exist. In addition to Dürer, the catalogues mention an example by Hans Holbein. There is also a quite well-known self-portrait by Tintoretto which is not mentioned in the catalogues, but the overall impression is that the en face pose was almost never used when depicting secular persons. The study found no vital difference between regions in Europe; North and South in Europe either a pure profile or a variation of three-quarter view was preferred when representing ordinary people. Of the total of all 580 paintings analysed in this study, only seven are shown frontally. Taking into account works that are reproduced more than once (a very limited number), these seven comprise about 1.2% of the total!

This study by Hodne addressed the distinction between the frontal (and often almost perfectly symmetrical) view of the face and averted faces. It did not however analyse variations of profile and their meanings. A study that specifically addressed the three quarter view was done by Chris McManus, who used portraits from polyptychs (panel paintings showing Christ or the Virgin flanked by saints on either side) selected from among the more than 5000 images in Berenson’s eight volumes. Examining the ways in which left–right asymmetries were used by Italian Renaissance artists, McManus discovered that occasional asymmetries in polyptychs were more likely to involve the substitution of a left cheek for a right cheek, than vice-versa (McManus, 2005). This is especially true of women, where, among the 551 paintings of this type, 68% turned the left cheek towards the spectator. This number was significantly lower in the male portraits, where 56% of the 932 paintings analysed showed the left cheek. Several later studies have confirmed that the number in each case is significantly different from 50%.

‘Left cheek’ was predominant in both cases (male and female), but the choice of pose (head turn) also depended on the sitter’s relationship to the painter. A similar study by the same author on portraits by Rembrandt showed that portraits of kin (where the model has some kind of relation to the painter), be they male or female, were more likely to show the right cheek than were portraits of non-kin. From this he developed a hypothesis that the right and the left half of the face are perceived differently by the spectator, with the right cheek signalling ‘kin’, i.e. someone who is like myself, while the left is ‘not kin’ (unlike myself).

Comparisons with photographs, where the same bias is found, refute the hypothesis that left cheek bias is in some way connected with the fact that most artists use their right hand when they paint; yet, the phenomenon may very well be related to handedness in a symbolic way. Despite the numeric predominance of ‘left’ in portraits, there is reason to conclude that our preferences lean towards ‘right’, since this view is associated with ‘self’. The most natural explanation of this is a tendency across cultural and geographical barriers to associate ‘right’ with positive values and ‘left’ with negative. This conclusion is based on a 1909 study by the French sociologist Robert Hertz, who, studying funerary and mortuary practices, pointed to the fact that right is always associated with positive values (honours, flattering designations, prerogatives), while left is negative (despised and reduced to an auxiliary role). It is the structure of the human organism and the asymmetry of our brain, McManus concludes, that explains why ‘right-left symbolisms are not arbitrary, but are constrained by biology, …’ (McManus, 2005).

We saw that a portrait signals different things, depending on whether it turns its left or right cheek towards the spectator. But in both cases (left or right) we are confronted with the picture of a face that is turned in a three-quarter pose. In the Renaissance, secular persons were either portrayed in pure profile or a three-quarter view. Although their eyes were sometimes turned outwards, like the famous Mona Lisa, the face was almost never seen frontally. Was the full-frontal pose used in Renaissance art at all, and if so, in what connection? One way that we can shed light on this question is to turn to the works by one of the period’s most prolific painters of portraits, the German-born Dutch artist Hans Memling. From his hands there are at least 36 surviving portraits of secular persons, and four small-scale paintings of Christ. Interestingly, while none of his secular portraits are seen in full face, three of the Christ portraits that he made are shown this way. The three portraits of Jesus that are seen frontally are very similar. They show the Lord with a calm, almost insensitive, facial expression; his right hand is raised in a blessing gesture, while the left is placed on the lower frame as if it was a window opening (Fig. 4). His dark, long hair is parted in the middle, and the same goes for the beard, which is cut the same way as the paintings from the Van Eyck school that we will discuss below. If we drew a vertical axis right through the centre of the portrait it would intersect the hair and beard exactly where it is parted, dividing the painting into two perfectly symmetrical halves. The fourth Christ portrait that Memling made, was totally different. In this, Jesus has a sad and tormented facial expression, as we know him from Crucifixion scenes or scenes from the Calvary (Fig. 5). What’s more, on his head he wears the Crown of Thorns and from his wounds issue drops of blood. His right hand is raised, not to bless, but so that we can see that he has stigma marks.

Fig. 4

Hans Memling, Christ giving blessing (Holy Face), Museum of Fine Arts, Boston

Fig. 5

Hans Memling, Man of Sorrows, Musei di Strada Nuova, Genoa

On the basis of these observations, we can make two testable hypotheses. First: The full-frontal and symmetric view of the face is rare in portrayals of secular persons because it signals something divine or holy that we seldom associate with ordinary people. Second: The fact that the frontal view was common, but by no means compulsory, in representations of Christ, could be an indication that pose (frontal or averted) signalled different views of Christ as a person with both divine and human aspects. These two aspects may very well correspond to two well-established art historical types of Christ’s image. The first is also found in depictions of Christ as universal ruler (Pantokrator) or judge in scenes of the Last Judgement or the Apocalypse; small panel paintings of this will here be referred to as the Holy Face. The second type is in German often denoted as the Schmerzensmann; in the present context we will use the familiar English term Man of Sorrows.

In a 2013 study images of Christ from catalogues were surveyed according to the same principles as the already mentioned study of Renaissance portraits of secular persons (Hodne, 2013b). This time images were retrieved from Raimond van Marle’s catalogue, The Development of the Italian Schools of Painting. Van Marle’s volumes cover Italian art from more or less the period of Giotto (slightly before the year 1300) until the verge of the High Renaissance around 1500. The number of Christ portraits in his catalogue is not very high. Among the thousands of paintings reproduced in his 18 volumes, there are only 24 portraits of Christ. Of these, 10 are of the Holy Face type, whereas 14 can be categorized as Man of Sorrows. When these two types were further divided into sub-groups, we found that of the Holy Face type, there was a clear preference for full face (7 full face portraits and only 3 half profiles). As for the Man of Sorrows, the tendency was even clearer in the opposite direction: of the 14 panels, only one was full face. The rest were all in half profile.

It must be emphasized that since the aim of this study was to find out whether there was a connection between pose (frontal or averted) and type (Holy Face or Man of Sorrows), the two types had to be defined initially by other means. Hence, the Man of Sorrows was defined as an image where Jesus shows signs associated with Passion (stains of blood, Crown of Thorns, stigma marks) and has a sad expression; in contrast, the Holy Face shows Christ with a calm facial expression and without signs of Passion.

When confronted with other catalogues, the survey of Christ portraits from Van Marle also confirmed our suspicion that frontality was much more used in portrayals of Christ than ordinary persons. While, as we remember, only 1.2% of the secular portraits showed the sitter en face, this number was significantly higher in depictions of Christ. Of the 24 Christ portraits in Van Marle, no less than 8 are shown frontally. In Friedländer’s Die altniederländische Malerei, the percentage is even higher: 7 of 10 Christ portraits reproduced here are frontal.

How a photo or a painting relates to the spectator depends both on face direction and gaze contact. The studies discussed in the previous chapter mainly focus on the role of face and head orientation. In research, head orientation and gaze direction are often considered as different and separable aspects of vision, and there is also a discussion about what—gaze or face—is the most important when we make judgements about others. Among scholars there is a tendency to consider gaze as the more important. In a gender-categorization task Baron-Cohen demonstrated that test persons judged the sex of people faster when the targets were looking straight ahead (Baron-Cohen, 1997). This effect was independent of the orientation of the face (full face versus half profile), thereby confirming that gaze direction overrides face direction when judging the sex of other people. These results have been confirmed by not gender-related experiments, leading researchers to conclude that gaze contact has precedence over contextual information such as head orientation.

A view that stresses the joint interaction of gaze and face direction was presented by Pageler et al. (2003)**. An investigation of brain activation in the fusiform gyrus and the posterior superior temporal sulcus during face tasks revealed a significant effect of both head and gaze on the speed of gaze processing. Reaction time was faster and activation greater in both brain areas when both face and gaze were directed towards the spectator (the participant). In addition, activation in the fusiform gyrus was greater for the forward gaze compared to the angled gaze, but only when the gaze was supported by head orientation (face forward). Similarly, Langton (2000), in a study where participants were asked to make keypress responses to a photograph of a male face, revealed that incongruent cues (head orientation different from gaze orientation) slowed the participants’ responses, suggesting, like Pageler et al., that head and gaze are mutually influential (Langton, 2000). Indeed, a number of recent behavioural studies suggest that contextual information like head orientation plays a greater role in the processing of social attention than was formerly believed.

In a study from 2016, Folgerø et al. used images from photographs as well as paintings to test people’s reactions to varying gaze and face directions. Portraits were selected from a database with images of the heads of persons photographed from different angles. For each person was chosen a frontal view and a view that was what we here call three quarters (approximately midway between frontal and profile). Then the eyes on the photos were manipulated in such a way that for each pose there was one image with direct gaze and one with averted gaze. Test persons thus had four alternatives to choose from: direct face/direct gaze, direct face/averted gaze, averted face/direct gaze, and averted face/averted gaze.

Test persons (36 females and 16 males, mostly students) were asked to respond to the faces according to a choice of pre-selected adjectives. These were common words from everyday speech that are descriptive of personality traits, either positive (‘harmonious’, ‘trustworthy’, ‘caring’, ‘inclusive’, and ‘respectable’) or negative (‘authoritarian’, ‘monitoring’, ‘evasive’, ‘intimidating’, and ‘dominant’). The results showed that people gave frontal faces with a direct gaze as well as three-quarter view with averted gaze high scores on attributes from the group of positive adjectives. The test was then repeated using paintings instead of photographs, manipulated in the same way as above, with four alternative views of the face (this time the test persons were 39 females and 10 males). The results of this test were similar to the first, but even more pronounced in favour of direct face/direct gaze. The test persons saw this type of portrait as representing a person who was more caring, trustworthy, harmonic, inclusive, and respectable than the corresponding images with averted gaze and face.

The findings in the study of Folgerø et al. are interesting when we turn to our analysis of the two categories of Christ portraits. The positive emotions elicited by frontal gaze and frontal face direction may shed light on why Renaissance artists preferred this pose when representing the divine aspect of Christ, while generally avoiding it when representing the Man of Sorrows. What’s more, the fact that the results obtained from the test that used paintings are similar to those based on photographs, might be an indication that responses to images resemble, in some way, reactions we have in real-life encounters.

We have seen that in art face direction was used systematically to distinguish between representations of ordinary persons and the holy, as well as between different aspects of Christ. It is interesting to ask, as regards the different types of Christ portrait, whether there are early Biblical or Apocryphal sources or documents from, for instance, the Church Fathers, that can confirm the existence of different (and opposing) visions of Christ?

Interestingly, Tertullian’s discussion of the Messiah in Adversus Marcionem seems to confirm the idea of a twofold nature of Christ. Discussing the Old Testament prophecies, Tertullian concluded that the Messiah must have two quite different aspects. First he drew an image of him ‘in humility’ with ‘no appearance nor glory, … or beauty, but his appearance was unhonoured, defective more than the sons of men, a man in sorrow, …’ (Tertullianus & Evans, 1972). Tertullian’s description was part of an argument against heretics who denied that God could have a body like human beings. Christ’s suffering was to Tertullian proof that he had a body, and his ugliness is related to his suffering. This idea was based on the prophet Isaiah, who described Messiah as ‘disfigured beyond that of any man and his form marred beyond human likeness’ (Is 52, 14). What’s more, ‘he had no beauty or majesty to attract us to him, nothing in his appearance that we should desire him. He was despised and rejected by men, a man of sorrows, and familiar with suffering’ (Is 53:2-3, my italics).

Christ suffered during his life on earth. He could be seen by everyone who lived at that time, but the risen Christ showed himself only to the faithful. The calm expression that we know from paintings of the Holy Face refers to this second aspect. Emphasizing that ‘the tokens of sublimity’ only apply to the Lord’s Second Advent, Tertullian said that Christ then ‘will have an honourable appearance, and beauty unfading, more than the sons of men’ (Tertullianus & Evans, 1972). This must be based on the Psalms, where the King is described as ‘beautiful above the sons of men’ (Psalm 44 (45):2).

There can be little doubt that the necessity felt by painters and their employers to distinguish between two different and quite distinct types of Christ had a foundation in texts by religious authorities. We notice that the title usually applied to the suffering type—Man of Sorrows—was also used in the Bible and other religious texts. It is more difficult to find an equal for Holy Face in the Bible, but that this kind of picture can be seen as expression of the ‘honourable appearance’ that Tertullian described, should be clear from some of the very oldest paintings of this type that still exist today. In fact, these paintings often had inscriptions referring to the mentioned passages from the Bible on the frame. An example of this is a painting that formerly was attributed to Jan van Eyck, now in the Groeninge Museum in Bruges, which, in addition to the painter’s false signature and quotations from the Gospels, bears the inscription: Speciosa forma pfiliis hominum (Koerner, 1993; Fig. 6). This is a misspelling of ‘Speciosus forma præ filiis hominum’ (‘you are beautiful above the sons of men’) from Psalm 44 (45):3, that was also quoted by Tertullian in relation to the sublime aspect of Messiah.

Fig. 6

Jan van Eyck (follower of), Vera Icon, Groeningemuseum, Brugge

From this it is quite clear that the artists’ distinction between two contrasting aspects of Christ was theologically founded, but we still lack a final answer to why the Man of Sorrows should be represented with averted face and the Holy Face with direct. To explore why the direct pose was chosen to represent the holy, we must turn to another aspect of frontality, namely symmetry. We know that the human body is almost symmetric; we have a right hand and a left hand, a right foot and a left foot. Ears and eyes are symmetrically placed on either side of our nose, and so on. However, this symmetry is only approximate, for nobody is perfectly symmetric. Studies demonstrate that almost perfectly symmetric faces are generally perceived as more beautiful than less symmetric faces. Such preference for symmetry is also found among animals. Swallows and peacocks with symmetric tail feathers, it seems, are preferred to individuals with irregular patterns as the most attractive mate choices, most likely because the opposite, asymmetry, in some cases will be a result of impairment (Jones et al., 2001).

One reason why we prefer the frontal view of the body may be that it is from this position that we can study its symmetry and, consequently, judge whether it is healthy and attractive. For the same reason, the en face portraits must also be analysed in terms of their symmetry. To continue our discussion of the painting from the Van Eyck school above, we may first recall the inscription on the frame that clearly stated it was intended as an image of Christ in his beautiful aspect, ‘above the sons of men’. In the painting the face is shown frontally to reveal its symmetry, and symmetry, we know now, is something that we associate with beauty.

Everything in the painting seems to be designed to enhance its symmetry. As in Memling’s works, Christ’s hair is carefully done and parted in the middle with locks falling to either side in equal numbers (three) on both shoulders. Even the beard is parted in the middle. The ensemble in Fig. 7 juxtaposes Van Eyck’s Early Renaissance painting with 5 late medieval precursors from Catholic Western Europe as well as the Orthodox East. Except the background (which in some cases shows the cloth of the vera icon, see above), these images are all very similar. Christ turns his face frontally towards the spectator, his long hair falls in an equal number of locks on both shoulders, hair and beard are parted in the middle, and so on. The perfect frontality is broken in two cases, where the gaze is slightly averted (Byzantine artists usually avoided direct gaze from fear of ‘the evil eye’), but there can be little doubt that the style of the hair and beard was carefully designed to enhance the strict frontal symmetry of the face.

Fig. 7

Comparison of variants of Vera Icon