RIHA Journal 0229 | 30 September 2019
In Search of a Divine Face
Physiognomy and the Representation of Sanctity in Christian Art
The article aspires to show how physiognomy was used in late antique art in order to give substance to the theophanic dimension of a person, especially of a saint. Drawing on monumental art, sculpture, daily life objects and catacomb paintings or mosaics, it is possible to discern that the physiognomic features were used as iconographical attributes, which were adjusted depending on the context and of the degree of holiness of the person depicted (saints, deceased, patrons). Therefore, the article allows to follow the transition from portrait to 'icon' characterizing the late antique period.
Representing saints: ascribing physiognomy
Recognizability and substantiation of the saint
Popes and bishops: which physiognomy for the homo spiritualis?
Towards a physiognomy of holiness
Portraits and physiognomical manipulation
 In keeping with the topic of the volume, dedicated to the iconic dimension of bodies, this article discusses the ways in which physiognomy was employed in late antique portraits. Drawing on a selection of famous cases, I analyse how physiognomy translated status and character, and, with them, the revelatory dimension of holy and ecclesiastical figures.
 The manipulation of one’s physiognomy in his or her portrait was legitimated already in Aristotle’s thinking, with the Greek philosopher praising the artisan who was able to rework the model’s outlook in order to better reproduce its virtues.1 In Antiquity, the portrait developed between a double quest, that for a mimetic reproduction of the model’s physiognomic features, and that for elements that indicated its status.2 Studies of Graeco-Roman portraiture thus have focused on the dichotomy between 'physiognomic' and 'typological' images, although the two can only be separated at a conceptual level.3 When portraits rather than theories are discussed, the two are often found to overlap (Fig. 1).4 It is precisely the tension between the physiognomic/ individual and the typological/ idealised features that catalysed, in the second half of the twentieth century, some of the main theses regarding the representation of individual character. In the context of his research on the semiotics of authority in the Middle Ages, Ernst Kantorowicz postulated the existence of two bodies of the king, the mortal and the mystical/ political.5 Paul Zanker stressed the use of "Zeitgesicht" in Roman imperial portraits; faces that became en vogue thanks to certain typological features, and which came to influence also private commissions of the time.6 The same scholar showed that, when it came to representing intellectuals, certain physiognomical elements such as the beard, the upward gaze, and baldness were chosen and adapted in order to transmit their aspirations, ideals, and privileged relation with the divine.7
 The dichotomy between the physiognomic and the idealised dimension of a portrait, and the importance ascribed to physiognomy on account of its capacity to indicate the person’s interiority were further stimulated in the Christian milieu. The desire to represent the individual’s spiritual life—that is, his or her vicinity to the divine—and even the degree of holiness often took precedence over that of producing a realistic image of the person.
 Synthesising the phenomenon is a famous passage by Paulinus of Nola (354–431), a Roman nobleman who adopted an ascetic life and eventually became bishop of the community at Nola. Faced with his friend Severus’ desire for a portrait, he restated the embarrassment felt and expressed a century before by the Neoplatonic philosopher Plotinus (203–270) in face of his embodied state.8 Paulinus underlined the impossibility of rendering both the carnal and the spiritual man (homo spiritualis) in a portrait.9 This incapacity to account for the model’s spiritual state essentially annuls the logic of the portrait. The point Paulinus makes can be linked to the diffusion, from the dawn of the Christian era, of a new type of image; an image able to account for the model’s sanctity, produced for mnemonic and venerational purposes.10 In this type of devotional image, physiognomic accuracy was not meant to reproduce mimetically the features of the model, but to render the holy individual present to the devotee, who sought contact with him or her through the image. As argued by Katherine Marsengill in her recent study, the margin between portrait and icon was blurred, with the birth of the latter being related to the gradual abandonment of the physiognomical portrait.11 In Late Antiquity, the portrait thus goes through a transition, whose interaction with cult images has been studied in detail in the now classic studies by Hans Belting and Maria Andaloro.12
Representing saints: ascribing physiognomy
 As the more educated part of Roman society converted to Christianity, the elements that in ancient portraits stood for erudition, spiritual concerns, and self-control—those features that represented the mark of the philosophical life, like thinness, baldness or the beard—came to be used to indicate sanctity; that is the theophanic dimension of saints and ecclesiastic figures.13 The process through which physiognomy was used to construct an image of a holy Christian individual can be traced by analysing how the apostles Peter and Paul received their typical portraits in the course of the fourth century. Since, in all likelihood, these were no mimetic representations of the two apostles, an artificial physiognomy was invented.14 Personal features thus make way to an attributive physiognomy as Christian try to recover the faces of the apostles.15
 Paul is described in a second-century text as a man whose face is simultaneously human and angelic, characterised by a tension between the fascination he stimulated and his unattractive visage:
a man small of stature, bald-headed, bandy-legged, healthy, a brow meeting in the middle, a somewhat longish nose, a gracious presence; for some times he appeared as a man, but at other times he had the face of an angel.16
In order to put a face on this grave, inspired, and angelic character, artists of the fourth century adopted the image of the philosopher: a middle-aged man with pointed beard, dark hair, and advanced baldness.17 Peter’s portrait, which had to capture the diverse model of sanctity that the apostle embodied, was imagined as opposite to that of Paul. In the same period in which the face of Paul was gaining its features, that of Peter was also being established as that of a man with decisive and prominent features, rich hair, oval head and short beard. Peter emerged as a strong and mature fisherman able to balance, and complement, the erudite Paul (Figs. 2, 3).
 In imagining the appearance of the two apostles, the semantic possibilities offered by the physiognomic tradition were exploited in order to create two types of individuals able to translate a notion of embodied authority, if in diverse manners. Together, the two testified to the unity of a universal Church, reunited around the proselytising philosopher and the elect fisherman. Through the former God spoke the language of the elite, while in the latter the power of God was revealed as accessible to all. The role physiognomy played in the definition of a notion of incarnated sanctity, embodied by the two apostles, is confirmed, paradoxically, by the use of the opposing concept. However, on a number of gold glass pieces from the same period, physiognomic differences between the two are completely cancelled. Represented from the side and facing each other, Peter and Paul are shown with identical features, with their identities attested only by inscriptions (Fig. 4).18
In this case, the identity implications of physiognomy are used to advertise the notion of concordia apostolorum, using a strategy attested in the depiction of tetrarchic rulers, whose unity and common source of power were indicated through the use of identical faces.19
Recognizability and substantiation of the saint
 Peter and Paul with their specific features stand out from a mass of early Christian saints characterised by physiognomic anonymity. An exception in this sense is represented also by the Roman martyr Agnes. The recognizability of her portrait, which spreads in the course of the fourth century, probably linked to the effigy on her tomb, allows us to distinguish physiognomy’s use to render the saint present and stress her theophanic dimension. In a context populated by generic figures such as those on gold glass, the portrait of Agnes is broadly diffused and characterised by a great degree of iconographic coherence. Even though we cannot speak of a mimetic representation, and her identification is ascertained by inscriptions, the saint is consistently depicted as a plump young woman in orant position, with earrings, a bowl haircut and a type of chignon; all features that we later find on the mosaic 'icon' which adorns the apse of the sanctuary raised next to her tomb in the second quarter of the seventh century.20 Although she was martyred during the reign of Diocletian (284–305), thus more recently than the apostles, there is no certainty that portraits of her made while alive ever circulated. Thus, the first attempt to give her a face, or at least an official image, could be associated with the organisation of her tomb during the episcopate of Liberius (352–366). Then, an engraved marble pluteum dominated by the saint’s figure bore the features that would become hers (Fig. 5).21
The selected physiognomy underlines the youth and dedication of the girl, serene and smiling. It was certainly not by chance that her figure appears similar to that of the Virgin, as depicted on gold glass pieces of the same period. In between a funerary portrait and an icon, the image with a plausible physiognomy that was marked on the pluteum seems to have gained the status of an 'official' portrait of the saint, worthy of being reproduced.
 It is in a specific context, such as the one of the saint’s tomb, that his or her image is often attested as having a particular, individualised physiognomy. That is the case with the chapel of St. Victor in Milan (490–512), part of the Basilica Ambrosiana. There, the martyr Victor is depicted at the centre of a wreath, on the apex of the dome (Fig. 6). Identified by the name "Victor", written on a book he holds in his left hand, the figure is characterised by a marked individualisation that distinguishes him from Sts. Nabor, Felix, Gervasius and Protasius, who are depicted in conventional fashion on the walls below (Fig. 7).22
Victor is rendered in the manner of a philosopher, as a mature individual with grey hair and beard, prominent nose, plump mouth and protruding ears. As proposed by Ivan Foletti, the choice of a highly individualised physiognomy was meant to manifest the saint’s presence in situ in the eyes of who entered the chapel. Without engaging in the complex matter of the various stylistic modes identifiable in the mosaics of the chapel, discussed by Foletti in his study, it is, nevertheless, necessary to stress that the saint stands out through the almost 'impressionistic', 'colouristic' depiction. This element differentiates him from other figures depicted in the chapel, including St. Ambrose’s portrait, which can be defined as mimetic on account of its individualised features. The difference can be explained by the use of diverse kinds of models: In the case of Victor, a mimetic portrait on panel—a type of image that sources attest in tombs of venerated individuals—seems a likely model.23
 Representation of individuals in mimetic fashion finds its strongest expression in funerary contexts, as attested by the famous portraits from Fayoum (Fig. 8). In catacomb tombs too, carved spaces attest to the practice of inserting panels and/ or canvases with the defunct person’s portrait.24 In the latter instance, the images were potentially made while the person was alive, which assured the transmission of his (or her) 'true image', preserving his memory and providing, in the case of saints, a focus for his cult. The physiognomic individualisation of a holy figure such as Victor thus appears linked to the developments of the funerary portrait, while potentially also reflecting the desire to render visible the saint’s presence. His remains, buried beneath the chapel, assured his physical presence, in a context where the theophanic dimension of the martyr, and the possibility for the viewer to interact with him, were provided by the depiction of the effigy in the apex of the golden dome.
Popes and bishops: which physiognomy for the homo spiritualis?
 If in funerary context the choice of an individualised physiognomy can be traced back to earlier traditions, for other contexts one needs to explore the potential meaning of the choice. The series of pontifical portraits depicted in the nave of San Paolo fuori le mura in the time of Leo the Great (440–461) present us with such an instance.25 There, one had to represent individuals for whom no previous models existed, with the exception of Peter and, possible, the most recent two bishops, the patron Leo and Sixtus III (432–440). Adopting the common motif of the imago clipeata, widespread in imperial context, these images are not 'intentional' portraits, but show nevertheless their subjects with cleverly individualised features. The forty-two bishops, identifiable through inscriptions that mention their names and number in the episcopal succession, are set on the underlying model of the philosopher. Based on this type of portrait of mature man with incipient baldness, each bishop received certain individualising features. Freed from the need to render the individuals recognisable, physiognomy was here employed as an iconographic tool, used to bring to life a number of different, yet complementary individuals. The resulting figures embodied both the prestige of the episcopal office and appeared as suitable intermediaries between the earthly and the heavenly Church.
 At the intersection between the funerary and the ecclesiastical milieux emerges a new manner of displaying individual physiognomies, through a synthesis able to flesh out that ideal homo spiritualis indicated by Paolinus of Nola; a figure permeated by the divine.26 This process is discernible in the so-called Cripta dei Vescovi in the catacomb of San Gennaro in Naples. There, a number of arcosolia decorated with mosaics bear the portraits of bishops and other high ecclesiastical figures (Figs. 9, 10).27 The creation of this space is closely related to the translation of the remains of the martyr Gennaro to the catacomb. As indicated by written sources, this was the work of bishop John I (413–432), whom we find buried ad sanctos, in the immediate vicinity of the martyr. His identification is confirmed by the inscription (of his titular saint’s name) SCS IOHANNES on the arch of the arcosolium (Fig. 9).28 The second arcosolium to be excavated was most likely the one in the wall to John’s right, in which is generally recognised the image of the North African bishop Quodvultdeus (Fig. 10). Exiled by the church in Carthage, he found refuge in Naples, where he died in 454.29