A Cultural History of the Emotions in Antiquity
A Cultural History of the Emotions in Antiquity

Douglas Cairns

Douglas Cairns is Professor of Classics at the University of Edinburgh, UK. He has published widely on Greek lyric and tragedy, including Bacchylides: Five Epinician Odes (2010) and Aidôs: The Psychology and Ethics of Honour and Shame in Ancient Greek Literature (1993). Author affiliation details are correct at time of print publication.

Search for publications

Bloomsbury Academic, 2019





Greece, Italy

Related Content

The Visual Arts

DOI: 10.5040/9781474207027.ch-005
Page Range: 83–108

Ancient emotions are en vogue: within the last decade, classical studies have brought forth a large corpus of research on the representation and elicitation of emotions in ancient literature, philosophy and politics. In classical archaeology and the history of ancient art, however, the subject is still underrepresented: most studies concentrate on single emotions (especially grief and mourning), or treat them within the wider field of body language and gestural communication.[1] Another field concerns the representation of violence and its (emotional) effect on the recipient.[2] But a comprehensive survey on the manifold “languages of emotion” in ancient art remains a desideratum. Also, while classicists have eagerly applied theoretical and methodological approaches developed in other disciplines and contributed substantially to the science of emotions, most studies on emotions in ancient imagery have not yet drawn upon the full potential of interdisciplinary research.[3]

This is all the more surprising in that, in modern Western culture, visual media are commonly regarded as particularly suited to convey emotional content and, more importantly, to enable an emotional reaction in the beholder (cf. Frevert and Schmidt 2011). Visual language is always highly suggestive; it creates the illusion that it is intuitively comprehensible and does not require any previous knowledge to be read and understood. This is especially the case for ancient Greek and Roman art whose naturalistic style “encouraged (and still does encourage) the imagination to believe that the visual world of a painting or a sculpture is just like our world” (Elsner 2007: 1). This twofold susceptibility to subjectivity might be precisely the reason why classical archaeologists did not jump on the bandwagon but remained skeptical towards the study of emotions (cf. Mylonopoulos 2017: 73).

Naturally, the present chapter can tackle only a small selection of the manifold aspects of the representation and elicitation of emotions in visual media. The first section opens the debate and presents the dominant strategies for conveying emotions in ancient Greek and Roman art: besides gestures and facial expressions, we shall look at personifications of emotional states and the role of context in representing and eliciting emotions. The second section deals with gender- and status-typical emotion expressions, based on a close reading of the iconography of grief and mourning. Thirdly, I elaborate on the differentiation between acute emotional episodes and long-lasting emotional dispositions, taking lust and love as a starting point. In the final section, we turn towards the emotive power of visual media concentrating on the visual representation of violence as a means of eliciting emotions in the recipient.

Embodied Emotions: Making Inner States Visible

In interpersonal communication, we rely on verbal and non-verbal expressions (i.e. words, tone of voice, gestures, postures, facial features) to convey and read internal emotional states. These “emotion codes” are not universal but shaped by display rules, that is, culture-specific prescriptions about who can show which emotions to whom, when, and, most importantly, how. In the representation of emotions in the arts, these “emotion codes” are again modified in accordance with the possibilities and limits of the respective artistic genre and its iconographic or literary conventions. Thus, when we deal with the representation of emotions in literature and art, we deal with a “double coding of emotions.”[4]

Maria Luisa Catoni has systematically developed the ancient term schêma (lit. “form,” “shape,” but also “appearance,” “fashion,” “manner,” etc.) as a tool for the analysis of images. Starting from the Platonic dialog Cratylus, she describes how all mimetic techniques—terms, images, music, dance, and so forth—use schêmata to represent the nature of things in the most exact way. Since every technique has its specific representational limits, it is always a selective imitation (Catoni 2008: 73). In the case of visual media, schêmata operate as external properties that transmit information about inner states and characteristics. They imitate not only a figure’s outer appearance but also their êthos (“character”) and pathos (“emotion”).

Ancient Greek art, especially of the Archaic and Classical era, relies primarily on gestures and postures to display feelings and other inner states, while the faces are quite often relatively calm. Outright manifestations of emotions through features, such as broad smiles, open mouths, angry scowls, disgusted grimaces, or sad frowns, are extremely rare and tend to be restricted to non-Greek figures or members of the lower classes. The western pediment of the Zeus temple in Olympia, which shows the battle between the centaurs and the Lapiths, a legendary people from Thessaly, is an (often-cited) prime example for this phenomenon (see Fig. 5.1).[5] While the coarse, distorted features of the rampant centaurs reflect rage, strain, and agony, their human opponents maintain a relatively calm appearance: although the Lapith is being bitten by his beastly enemy, he is only slightly frowning his forehead instead of displaying his physical pain outwardly.

Starting in the fourth century and catching on in the Hellenistic era, a new style in the visual arts seeks to exploit the full potential of emotional expression. Animated facial features and agitated gestures enter the realm of large-scale sculpture that was hitherto dominated by the vocabulary of idealism and restraint. Even heroes can now be characterized in their physical and psychological struggles as the famous Farnese Herakles impressively shows: “Gone are the days in which Herakles fought monsters, wild animals, and antiheroes with an emotionless face revealing no effort . . . Lysippos’ Herakles is a tired hero who challenges and questions his quest” (Mylonopoulos 2017: 82). The “Pasquino Group” aptly portrays Menelaus frowning in agony and pathetically gazing into the distance as he retrieves the lifeless body of Patroclus, draped over his arms in a mannered, hyperextended pose (see Fig. 5.2).[6] Nevertheless, the most excessive forms of affect continue to characterize figures who fall outside the “good order” of Greek culture (Maderna 2009: esp. 31–2). In the Roman imperial period, artists are in the comfortable position of being able to cherry-pick from the century-old pool of different styles and strategies: they prefer to depict their gods and heroes in the sublime style of the fifth century but rely on the effervescent corporality and pathos of the Hellenistic era to represent giants, satyrs, and centaurs (cf. Prioux 2011: 136–7).

Figure 5.1. Centaur biting Lapith, western pediment of the temple of Zeus at Olympia, c. 460 bce.
Figure 5.2. “Pasquino Group,” Roman copy from Hellenistic original, c. 150–120 bce.

The general reluctance towards the depiction of facial expressions can be explained in terms of the ideal of sôphrosunê (lit. “soundness of mind,” but also “self-control,” “moderation,” etc.) that from the fifth century onwards forms an essential element in the conception of good character.[7] Throughout antiquity, the ability of (emotional) self-control is a key concern in the upbringing and education of both male and female children and therefore also constitutes a decisive means of status distinction. The calm features in the visual arts hence have to be understood as idealized manifestations of actual behavioral norms and display rules—a textbook example of the “double coding of emotion.”

So, except in the Hellenistic period, which tests the boundaries of artistic expression in all directions, distorted features are, if at all, applied primarily to denote the “other.” Instead of manifesting affect through the face, as emphasized in Darwin’s Expression of the Emotions in Man and Animals, ancient images operate mainly with gestures and postures to convey a figure’s internal characteristics. In his extensive study on body language in ancient Greek art, Gerhard Neumann differentiates between two types of gestures and postures: sudden and extensive movements are subsumed under the term “Momentangebärden” and associated with affective reactions to concrete situations. “Zustandsgebärden” are characterized by quiet, contemplative movements and, in his reading, illustrate a figure’s mental state, a permanent internal condition.[8] With regard to their different functions and meanings, he also refers to them as “pathetic” versus “noetic” gestures (Neumann 1965: 107–8). Claudio Franzoni offers a slightly different model: starting from Aby Warburg’s famous classification of certain “emotionally charged visual tropes” as “pathos formulas,” he proposes a second category called “êthos formulas” that translate a figure’s character and attitude into bodily schêmata (Franzoni 2006: 245).

While it seems justified to interpret sudden gestures and agitated postures as visual codes for acute emotions, it is in most cases impossible to discern the exact emotion without taking into account the context. The notorious ambiguity of bodily schêmata can be demonstrated by the gesture of raised arms. A famous, albeit only fragmentary, dinos by Sophilos features the funerary games for Patroclus, identified as such by an inscription, with spectators sitting on pyramidal terraces and watching a chariot race (see Fig. 8.1). Two of them are stretching their arms out forwards as if they were commentating on the events taking place before them (modern football fans may come to mind). In prothesis scenes and other images of the funerary context, however, the raised arms can be interpreted as part of the traditional mourning ritual that also included tearing one’s hair and beating one’s chest (see Figs 3.1, 5.45). And finally, on a wall painting in Pompeii portraying the sacrifice of Iphigenia, the protagonist raises her arms to the sky in fear and supplication (see Fig. 5.6).

There is no clear-cut formula to link a certain way of raising the arms with a certain emotion. If it were not for composition and context, it would be difficult to differentiate between the excitement of the chariot race enthusiasts, the ritualized grief of the mourners, or the desperate plea of Iphigenia. As Évelyne Prioux notes, one and the same schêma “can be used to render different emotions and its interpretation will depend on the context in which it appears” (Prioux 2011: 142). Most of the “loud gestures,” then, do not necessarily denote a distinct emotion but rather a very basic and in itself neutral bodily reaction, a purely physiological “arousal”—be it negative or positive (see Scherer 2005: e.g. 718). They denote a vague state of commotion, an undefined affect that needs further information to be fully understood.

Another problem concerns the distinction between pathetic and noetic gestures or, respectively, between pathos formulas and êthos formulas. A red-figure lekythos in New York shows Poseidon pursuing the heroine Amymone with his trident at the spring of Lerna (see Fig. 5.3).[9] The subject of “erotic pursuit” was very popular in early fifth-century Athenian vase painting and usually shows a male figure (divine or mortal) running after a young woman or boy. In most instances, the “sexual prey” raises both arms in fear and alarm (another meaning of this gesture). But instead of expressing panic, Poseidon’s love interest elegantly raises her veil or mantle—a gesture you wouldn’t expect in this moment of threat. In archaeological scholarship, this gesture is traditionally associated with the unveiling of the bride during a stage of the Greek wedding ceremony called anakaluptêria and interpreted as an iconographic marker to denote brides in particular and married women in general (see Llewellyn-Jones 2003: 85–120, esp. 99–104). In this reading, Amymone would hence be characterized as the, albeit only short-term, “bride” of the sea god. In our case, as in most other “erotic pursuits,” however, the woman’s hand position suggests that she is rather covering than uncovering herself, which points to another interpretation.

In ancient Greece and Rome, women were generally expected to cover themselves in the presence of men other than their kin in order to demonstrate their modesty: aidôs in Greek, pudicitia/pudor in Latin (cf. Cairns 2002: 75–6). Written sources even suggest a connection between aidôs and the veil that goes far beyond mere behavioral norms: “The veil . . . is a typical symbol of aidôs, used both as a concrete physical expression of the emotion and as a metonymy for the emotion itself” (Cairns 2016a: 36). Transferred to ancient iconography, the gesture of pulling one’s veil must be understood primarily as a sign of modesty; only in certain contexts may it also refer to the bridal or married status of the figure. But there is more to it: the Greek term aidôs oscillates between pathos and êthos, as it can refer to both the acute emotion of shame and the character trait of modesty—and so does the act of veiling (cf. Cairns 2002: 75). This ambiguity is also inherent in the visual accounts. Amymone might be pulling her veil in this exact moment because she feels ashamed as a result of a man (or god) appearing on the stage and advancing upon her in a most compromising way. Nevertheless, she is not succumbing to the acute affect of panic but maintains her general disposition of modesty. We shall return to the differentiation between emotional and dispositional phenomena later in this chapter, when dealing with visual representations of desire and love; at this point, it is sufficient to say that most gestures are (deliberately) ambiguous and can have various levels of meaning.

Figure 5.3. Poseidon chasing Amymone. Athenian lekythos, c. 420 bce.

In addition to the translation of pathê into bodily schêmata, emotional states can be rendered by means of personifications. In a first concrete sense, personifications of emotions have to be taken seriously as real agents that may be invoked, channeled, or controlled by worship (cf. Naiden, Chapter 2 in this volume). In a more abstract sense, however, they also represent the idea of pathos as an external force or, as Franzoni puts it (2006: 59), “la provenienza ‘da fuori’ del pathos,” a concept already inherent to the ancient emotion terms pathos (“suffering” or “experience”) and affectus (“being affected”), but also formulated in various philosophical treatises and literary accounts (cf. Cairns 2016b). So, without denying the real existence of personifications for (at least some) ancient minds, they can also be understood (and studied) as a “metaphor of emotion as an antagonist or external force—as an entity that ‘seizes’ or ‘holds’ a person, or as something that ‘comes over’ or ‘comes upon’ us from outside.”[10]

Visual representations of personified emotions are sometimes quite unspecific in their iconography and can only be identified by context and/or inscriptions. In many cases, however, there is a visible effort to translate the pivotal aspects of their mode of action into bodily schêmata and attributes (cf. Korte 1874). Nemesis, the goddess of indignation and retribution, for example, is often depicted as a winged figure with a wheel that will punish the arrogant and “roll out” wickedness (see Fig. 2.1; cf. Faraone 1999: 64, n. 105). The wings are a common feature of personifications and hint at their status as daimones “mid-way between god and men” (Seltman 1923–5: 90). Eros/Amor too appears as a young boy or youth with wings, which in his case may also refer to the fleeting character of erotic desire. When Eros enters Greek art in the late sixth century bce, he is represented in the iconography of the ideal erômenos, the younger party in pederastic relationships, and thus as the object of his own power (see Stafford 2013: 179–90). In a number of vase paintings of this early period, he is chasing a young boy and a woman and thus paradoxically assumes the active role usually ascribed to adult men—in these images, Eros manifests “a pure form of embodied desire” (Lewis 2002: 143). In the course of the fifth century, the adolescent erômenos type is by and large replaced by a childlike figure that often serves “as a caption for the desire of others” (Lewis 2002: 144; cf. Figs 5.89). This iconography of Eros as a young boy with wings will persist in the centuries to come (cf. Figs 5.10ab). Childhood and adolescence are traditionally associated with playfulness, foolishness, and self-indulgence—characteristics that one would easily ascribe to passionate love as well. A poem by Alcman articulates this notion in a parable of êros as a negligent child:

Aphrodite it is not, but wild Eros playing like the boy he is,
coming down over the flower-tips—do not touch them, I beg
you!—of the galingale.[11]

As we shall see, the image of the “wild boy” (margos) bringing ruin to his victims (i.e. lovers) is a persistent topos in Greek and Roman art and literature that can be enriched ad libitum with various additional elements and anecdotes. At this point, it is important to note that the personification of emotions conveys a core aspect of emotional perception. While the ancient concept of pathê as external powers differs from certain modern scientific definitions of emotions as internal processes, we can trace striking similarities from a phenomenological perspective: then and now, the image of romantic love as an autonomous entity (apparent in English expressions such as “hit by love” or “love-struck”) emulates its affective, ambush-like character. Then and now, the externalization of emotions in language and art has a metaphorical meaning that reflects the subjective experience of being at the mercy of one’s own feelings.

With facial features, body language, and personifications we have discussed the most important strategies of making emotions visible within an image. As emerges from the previous considerations, however, the context of a visual code also plays a decisive role in deciphering its emotional content. On an initial level, the image itself provides a good deal of information by means of composition, figure types, and further pictorial elements. Beyond this, the function of the medium, the context of reception, and the general cultural background can also be crucial indicators in the assessment of the emotional content of an image.

These broader contexts are particularly important when it comes to the elicitation of emotions through visual media: the emotive quality of an image is not necessarily dependent upon the representation of emotions (and vice versa), but can also be generated through the context of use and reception. A funerary monument within its original setting, for instance, can trigger a feeling of sorrow or melancholy in the viewer without making use of an overtly emotional iconography. In the famous fourth-century bce Attic grave relief of Ampharete, the deceased woman elegantly reclines on a chair and gazes at a small infant on her lap, who is reaching for the little bird she is holding in her raised hand (see Fig. 7.4). The scene is not characterized by vigorous pathos but portrays a rather restrained form of mother–child interaction. At first sight, Ampharete’s moderate composure is in line with the ideal of emotional self-control that was so crucial in classical society. Yet the emotions are there, subtly conveyed by the dyadic composition of the figures, the longing gesture of the child, and the maternal gaze, filtered through the “double coding of emotion,” that is, the display rules and artistic conventions of the time.

The ancient viewer must have perceived this image as the epitome of ideal motherhood—and must have been surprised by the epigram that identifies Ampharete as the grandmother of the infant: “It is my daughter’s child that I hold here with love, the one whom I held on my lap while in life we looked on the light of the sun and now (still) hold, dead, like me.”[12] Far from mitigating the affective power of the relief, the epigram serves as an emotion catalyst: not only does the text explicitly refer to the affection of Ampharete towards her grandchild (τέκνον φίλον); it also characterizes her as an attentive mother supporting her daughter in early childcare. And thirdly, by deliberately disappointing the viewers’ expectations, the epigram commands their attention and consequently heightens the emotive effect of the monument. Within the framework of classical Athens, the calm and restrained habitus of Ampharete is more than enough to elicit an emotional reaction in the beholder. Of course, the desired emotive effect follows the same feeling-rules we see in the image itself: the beholder is expected to develop a feeling of melancholic sympathy rather than excessive grief.

Gender and Status: The Case of Grief and Mourning

As pointed out above, both the expression of emotion in interpersonal communication and the artistic representation thereof are culture specific. Some emotion codes apply to all members of a given culture while others pertain only to subgroups or social communities and thus help to identify someone as belonging to a certain age group, gender, or status. In the visual arts, too, the particular modes of depicting emotions are a crucial means of differentiating between members of different social groups. Images not only reflect the various “emotional communities” of their time but also contribute to their formation and consolidation (on the concept of “emotional communities,” see Rosenwein 2006).

The function of emotion codes as an element of social distinction can be well demonstrated with the “emotion family” of sorrow, grief, and mourning. Both in Greek and in Roman culture, grief and mourning were governed by numerous restrictions and regulations and thus had a strong ritualized aspect that not only affected the display of emotion, but also its subjective dimension. In other words, “the imposition of limitations regarding the period of mourning, attire, and other facets reflects an understanding of grief as a social function, not merely a private sentiment” (on the ancient Greek concept of “grief,” see Konstan 2006b: 244–58, here 252).

From Geometric to Classical times, the most significant image subject for the study of grief and mourning is the prothesis, the laying out of the corpse amidst a group of male and female mourners. Already in the earliest prothesis scenes we find gender-typical mourning habits: female figures are shown raising both arms and beating their heads, while male figures stretch out only one hand, thus paying their respects in a more controlled way. This basic iconographic pattern reflects the different social roles ascribed to men and women within the funerary ritual, which in turn enshrines in visual discourse the culturally constructed dichotomy between male rationality and female emotionality. It will prove extremely effective and is maintained in later periods, sometimes refined by additional pictorial elements which allow for further differentiation regarding the role and status of the depicted figures and their affiliation to the deceased.

A black-figure pinax (votive or funerary tablet) in New York shows a typical archaic prothesis scene for a young, beardless man mourned by female and male relatives (see Fig. 3.1). Two men approach the bier from the right in the traditional greeting gesture; one of them is characterized as being old by his white hair and beard. Although their heads are thrown back to indicate that they are singing the thrênos (ritual lament), their behavior is relatively calm. Next to the bier, three women gesticulate wildly, tearing their disheveled hair, while a fourth woman is singled out as she gently touches the head of the deceased. Textual evidence suggests that the last physical contact with the deceased person—albeit automatically accompanied by ritual contamination (miasma) that necessitated subsequent purification rites—was perceived as both an emotional need and a privilege of the closest relatives and dependants. In Euripides’ Suppliants, for instance, the Chorus of Argive mothers demand to touch their sons’ “blood-dripping corpses” (σώμαθ᾽ αἱματοσταγῆ, 812): “Let me embrace and hold my children to my bosom in my enfolding arms.”[13] As Angelos Chaniotis has shown on the basis of Greek epigrams of the imperial period, for instance, the touching of the dead body was not just a literary topos but formed an essential element both in ancient Greek and Roman funerary ritual and was considered an essential means of attesting and perpetuating an affective bond beyond the grave (Chaniotis 2006a: esp. 206–7): In the visual record, too, the physical contact between the dead and the living clearly signals emotional intimacy and seems to be, as far as we can identify the figures and their relationship to one another, restricted to the parents, spouses, and the old nurse of the deceased.[14]

Returning once again to the pinax in New York, there is a fifth female figure squeezed into the right edge of the scene: significantly smaller in size and thus clearly identified as a servant, she is squatting on the floor, lowering her head in despair. Unlike the highly ritualized mourning gestures of the standing figures, her body language indicates that, as a slave, she is neither able nor expected to maintain her composure. The woman touching the deceased and the servant crouching on the floor display emotional reactions that are not included in the standard scheme and only apply to certain members of the community.

A white lekythos, used in the funerary ritual, pushes the stereotype of the hyper-emotional slave woman to the extreme (see Fig. 5.4):[15] an old nurse, characterized by wrinkled, distorted features, sparse hair, a simple dark garment, and tattoos covering her arms, crouches on the ground in front of a tumulus and wails excessively for her deceased foster daughter, who is visible on the left side of the grave as a “shadow image,” quietly contemplating her own untimely death (the hydria-loutrophoros crowning the grave mound denotes the deceased as an unmarried, childless woman).[16] With her ideal features, her elegant dress and her calm composure, the young woman forms a sharp contrast to the nurse who is marked as her social inferior not only by her non-ideal appearance but also by her frantic behavior.

Content unavailable

This content is not available for display due to rights reasons.

The “outsourcing” of extreme emotion expressions to members of the lower classes can be observed in Roman art too. In the so-called conclamatio scenes on sarcophagi, for instance, the family members, that is, free Romans, are usually shown in rather introverted forms of sorrow while the (female) servants play out the entire gamut of mourning gestures, from raising their arms to beating their bare breasts (cf. Dimas 1998: 21). The central image of a child sarcophagus in Agrigento (see Fig. 5.5) shows the deceased boy on his funeral bed flanked by his parents, who bow down with sorrow, fully wrapped in their cloaks, hiding their desolation under their veils.[17] The domestics behind the bier compensate for this rather reticent emotionality with intense gestures and loud lamentation. An aged nurse, identified by her characteristic bonnet, caresses the corpse and thus demonstrates her strong affection and close bond to the little boy, while an agitated pedagogue and another servant raise their arms in distress. Their more excessive forms of mourning help further to heighten the emotive power of the scene while maintaining the norm of passionate restraint for individuals of higher status.

These gender- and status-distinctive mourning habits are not exclusively pictorial phenomena but are based on real display rules well attested in various literary and visual genres (cf. Toynbee 1971: 45). In Plutarch’s famous Consolation to his Wife, the author tries to calm and comfort her on the death of their youngest daughter by stressing the importance of renouncing the “never-sated passion for lamentation” (θρήνων ἄπληστος ἐπιθυμία, 609b):

For not only “in Bacchic riot” must the virtuous woman remain uncorrupted; but she must hold that the tempest and tumult of her emotion in grief requires continence no less, a continence that does not resist maternal affection, as the multitude believe, but the licentiousness of the mind.[18]

Figure 5.5. Roman sarcophagus, 120–130 ce.

Whether in Dionysian ecstasy or in mourning, a virtuous woman ought never to let herself be carried away by her feelings, but should rather perform a “controlled loss of control” that helps to channel individual emotions in socially acceptable directions. The mourning gestures in the images are not necessarily intended to convey the “real” sentiments of the depicted figures, but rather the appropriate forms of conduct in the face of loss. Or as Beate Wagner-Hasel puts it, “[Es ist] verfehlt, exzessive Trauer mit unkontrollierter Emotionalität gleichzusetzen” – “[It is a] mistake to equate excessive grief with uncontrolled emotionality” (Wagner-Hasel 2006: 84).

It is essential to emphasize that the intensity of emotion expression cannot be equated with the intensity of the emotion (either in interpersonal communication or in its visual representation). Some pictorial elements may appear at first sight to conceal the depicted figure’s inner state but turn out to be powerful emotion codes when read against their cultural background. This can be well demonstrated by the act of veiling, which in the Graeco-Roman world was “both a spontaneous expression of grief and an element in mourning ritual” (Cairns 2016a: 35). On a first, phenomenological level, the veil can be understood as a metaphor for the subjective emotional experience of grief and sorrow, as something that “covers” a person like darkness.[19] As demonstrated above, covering with a veil or mantle is also a sign of shame or modesty (aidôs in Greek and pudicitia/pudor in Roman culture); it was also regarded as the appropriate reaction in various “emotional scenarios in which honour and propriety are in question, for example to express grief or anger” (Cairns 2002: 75; on the veil as an expression of shame, see also above). Returning to the Agrigento sarcophagus, the veiling of the parents thus symbolizes both their physical experience of being “shrouded with grief” and their need to maintain, at least on the surface, their composure (which in turn establishes a hierarchy between them and their servants in terms of emotional restraint and appropriate mourning). As an emotion code in the arts, the veil can be used as a visual symbol for grief as such and thus reveals more than it hides without risking infringing the cultural rules of propriety.

The “aesthetic of decorum” also resonates in a number of ancient descriptions of a now lost painting by the Greek painter Timanthes that showed the sacrifice of Iphigenia (cf. Perry 2002: 154–6). The artist displayed varying degrees of grief and sorrow among the Greek spectators (commensurate with their relationship to the victim), but concealed the despair of her father Agamemnon by covering his face with a veil. This pictorial solution can be traced to Euripides, who has a messenger describe Agamemnon’s immediate reaction to the horrifying spectacle: “When king Agamemnon saw the girl entering the grove to be sacrificed, he groaned aloud, and bending his head backward he wept, holding his garment before his face.”[20] Notwithstanding this literary model, a narrative description of offstage (i.e. non-visible) action, all ancient commentators praise the painter for his artistic ingenuity in rendering the scene visually, yet they differ slightly in their explanation as to why exactly Timanthes chose to shroud the father’s grief: Pliny seems to imply that the veil reflected the sheer inability of the painter (or the visual arts in general) to represent Agamemnon’s suffering (“cum . . . tristitiae omnem imaginem consumpsisset, patris ipsius voltum velavit, quem digne non poterat ostendere”; Natural History 35.73); Cicero discusses the painting in the context of rhetoric and praises its decorum (Cicero, Orator 74); and Valerius Maximus stresses the emotive power of omission when he brings up the viewer’s compassion to imagine the father’s weeping (“patris fletum spectantis adfectu aestumandum reliquit”; Valerius Maximus 8.11.6). The most comprehensive account is given by Quintilian, who combines the earlier interpretations:

Are not certain things likewise to be covered up in a speech, either because they ought not to be disclosed or because they cannot be expressed adequately? This is what Timanthes of Cythnus (I think it was he) did in the picture with which he won the prize over Colotes of Teos. Having depicted, in his Sacrifice of Iphigenia, Calchas sad, Ulysses even sadder, and given Menelaus the most complete expression of grief that his art could produce, he found he had used up all his means of representing emotion and could discover no way of adequately portraying her father’s face; so he covered his head in a veil, and left it to the imagination of the spectators.[21]

A visual reference to the painting, or maybe rather to the rhetorical tradition that developed from it, can be found in the “House of the Tragic Poet” in Pompeii (see Fig. 5.6):[22] covered in a cloak and concealing his face with his right hand, Agamemnon stands to the far left and turns away from the scene in the center, where his daughter is being dragged to the altar.[23] But unlike in the work by Timanthes, the expressions of the other Greek men do not resemble anything like grief: their faces are filled with astonishment and surprise as they turn their eyes on Artemis and her companion approaching from above with the redeeming sacrifice. In combination with Artemis’ intervention, Agamemnon’s veiling not only serves as a powerful tool to express his insuperable suffering, but also creates an interesting twist for connoisseurs (a “Pompeian pun,” as it were) as he ipso facto misses the miracle of Iphigenia’s salvation.

Figure 5.6. Wall painting from the “House of the Tragic Poet” in Pompeii, 62–79 ce.

Illuminating as these aesthetic discourses may be with regard to ancient (Roman) art production and reception, we must keep in mind that the act of veiling as an expression of grief was not just a stylistic device but had its roots in real experience: it drew upon collectively shared emotional concepts, reflected actual display rules in mourning, and could thus be intuitively understood by members of Graeco-Roman culture.[24] Some things never change: in the perfectly orchestrated funeral of John F. Kennedy, his wife Jackie became an icon of grief (see Fig. 5.7). Covering her face with a lace veil, she “did not compromise the ritual with outbursts of personal emotion [but] presented a calm demeanour of silent dignity” (Mulvaney and De Angelis 2010: 19) that affected the American public far more than excessive wailing and unrestrained mourning would have.

As we have seen, ancient Greek and Roman imagery developed numerous codes to express grief and mourning, from spontaneous manifestations of sorrow to highly ritualized gestures of lamentation. While the idiosyncratic possibilities and limits of different visual media certainly contributed to the concrete formation of emotion codes, it must be reiterated that most of them were grounded in everyday experience. The case studies presented also show that emotion codes not only are culturally specific but also vary between different social groups within one and the same culture (e.g. defined by status or gender). Naturally, the visual strategies we identified do not apply solely to the representation of grief, sorrow, and mourning, nor are they implemented only to differentiate between individuals of different status and gender: the same logic of distinction by means of emotional expression comes into effect in the representation of fearful barbarians, pain-stricken monsters, laughing hetaeras, or salacious satyrs.

Figure 5.7. John F. Kennedy’s family at his funeral in Washington, DC, November 1963.

Affects and Attitudes: The Many Faces of Love

The English term “love” (like most of its equivalents in Western languages) comprises a huge spectrum of different emotional phenomena: while some forms of love are experienced as the sudden and uncontrollable onset of intense psycho-physiological changes, others appear to be long-term affective dispositions that are relatively stable and can be formed and modified by deliberate choice (cf. Scherer 2005). The following section will elaborate on the pictorial possibilities of distinguishing between these different affective experiences, taking representations of erotic desire and conjugal love as case studies.

The ancient Greek term erôs is multilayered and ambiguous (one need only think of the tributes to Eros in Plato’s Symposium), but at this point a basic definition as “sexual longing” or “erotic desire” will suffice. The episode of the Trojan saga where Menelaus encounters Helen for the first time after her momentous adventure with Paris may serve as our first example.[25] In Athenian vase paintings of the first half of the fifth century bce, the story is often rendered using the scheme of “erotic pursuit” that we have already discussed. However, Menelaus’ initial intentions are diametrically opposed to those of Poseidon and other philanderers chasing young girls or boys: driven by the bitterness and the fury of the cuckold, he is pursuing his adulterous wife with a drawn sword, ready to take vengeance (see Fig. 5.8).[26] The minute he sees her, however, he is struck by her beauty and falls in love again. The Athenian vase painters found a truly striking formula to emphasize both the instantaneous and emotion-driven nature of Menelaus’ change of mind: the suddenness and vehemence of his reaction to the sight of Helen finds its pictorial expression in the sword slipping out of his hand—he has literally dropped his deadly plan—while the initiating sensation is made visible by a little Eros, the personification of erotic desire, fluttering between the two. In a couple of vase paintings, Eros confronts the avenger with a phiale in his outstretched arm; a red-figure lekythos in St Petersburg even depicts a mysterious fluid streaming from the bowl directly into Menelaus’ eyes.[27] The closest textual parallel to this visual code can be found in Euripides’ Hippolytus, where the Chorus, consisting of young married women of Troezen, call on “Eros, god of love” and describe him as “distilling liquid desire down upon the eyes, bringing sweet pleasure to the souls of those against whom you make war . . .”[28] Numerous written sources convey the notion of erôs as an experience mainly induced by visual stimuli. In Plato’s Phaedrus, for instance, we find the idea that “the lover’s desire is the result of the effluence [aporrhoê] of particles from the beautiful body that enter the lover’s soul via his eyes” (Cairns 2011a: 43). Both in literature and art, the image of Eros instilling passion through the eyes not only underlines the importance of vision in the ancient concept of erotic desire but also stresses the passive nature of the sensation: as erôs invades the body, the lover becomes a helpless victim of passion.

Figure 5.8. Athenian krater, c. 450 bce.

There are other instances, however, where Eros conveys a different notion of erotic love, which is more in line with the ideal of mutual attraction as a long-lasting disposition arising from marriage. In many wedding images on Athenian vases, one or more Erotes are present to symbolize the (desired) desire between the bridal couple. On a loutrophoros fragment in Oxford, the bride is being led to her new home by the groom in the “hand on wrist” gesture, a code closely associated with legitimate marriage (see Fig. 5.9).[29] A winged Eros accompanies the two and carries along two vessels for the nuptial rites, thus supporting the union in the most tangible way by engaging in the wedding ceremony. In these scenes, the concept of erôs as an external power does not necessarily signify its uncontrollable nature; on the contrary, by deifying erôs, it is possible to address him directly, bribe him with worship, and thus, at least to some extent, also activate and control him (cf. Naiden, Chapter 2 in this volume). And so, immediately following the invocation of Eros “distilling liquid desire down upon the eyes,” the Chorus in Hippolytus continue: “never to me may you show yourself to my hurt nor ever come but in due measure and harmony.”[30] While the married women of Troezen are begging to be spared from the burden of uncontrolled desire, the newly weds have every reason to hope for the support of Eros “in due measure and harmony.” In the context of bridal imagery, Eros is not perceived as a warlike spirit but as a benevolent power that reinforces the matrimonial bond and thus leads to a fulfilled life according to the norms and ideals of society—the two poles of erôs, condensed in the image of a little boy.

Figure 5.9. Wedding ceremony. Athenian loutrophoros, 420 bce.

Roman art, too, deploys the ambiguous figure of little Eros/Amor to create complex and sometimes witty statements on the capricious nature of erotic love. The “House of Punished Love” features Amor twice on opposite walls of the tablinum, both times in a family environment (see Fig. 5.10ab).[31] The southern wall shows Mars and Venus in open nature, accompanied by a winged cupid and a female servant stooping over a small box on the floor. The fully-armed god of war approaches his sitting lover from behind and ardently touches her breast, to which she responds with a gesture of “affirmative consent.” Mars and Venus embody the paradigmatic couple, the man showing an active and acute interest in the woman who in turn reacts to his intentions with mild control (cf. Lorenz 2008: 155). Amor floats by from the right, charming the divine couple with a iunx in his hands, a magical device used in love spells.

While this image sets Amor in the context of mutual attraction and thus stresses his benevolent power (only Hephaestus/Vulcan may object), the northern wall of the tablinum shows his unfortunate tendency to cross the line: Venus sits on a rock surrounded by an idyllic landscape and watches her son being “brought to justice” by another female figure (perhaps Nemesis). Presented in back view and tied in bonds, little Cupid is wiping his bitter tears with his right hand, standing before his mother, who has taken his quiver into custody. The scene has been convincingly interpreted as Amor’s punishment for torturing lovers, a popular subject in Graeco-Roman art.[32] The Hellenistic Anthologia Graeca contains a number of epigrams on a statue of Eros in bonds (alas, we cannot tell for certain whether the statue ever existed or was only imagined by the poets); the poet Crinagoras, for example, directly addresses the sinner, leaving no doubt that the punishment is well deserved:

Weep and groan, schemer, the sinews of your arms bound fast; such are your deserts. There is no one to untie you. Let us have no more piteous glances up. You, Eros, were the one to squeeze tears from others’ eyes; you fixed your bitter arrows in the heart, and instilled the poison of passion incapable. The agonies of mortals are your mirth. What is done to you is what you did; justice did an excellent thing.[33]

Figure 5.10a. Fresco from the “House of Punished Love” in Pompeii, c. 20 ce.
Figure 5.10b. Fresco from the “House of Punished Love” in Pompeii, c. 20 ce.

Another epigram from the Anthologia Graeca (16.195) draws an analogy between the specific nature of Eros’ penalty and his typical mode of action, wondering whether, “perhaps, this prisoner himself did once enchain the mind of the artist” (μή ποτ’ ἐκείνου οὗτος ὁ δεσμώτης αὐτὸς ἔδησε φρένα), thereby referring to another topos well known in the Graeco-Roman culture, that of the “chained lover,” incapacitated by desire.

On an initial level, the image of little Amor being punished for “squeezing tears from others’ eyes” can be enjoyed as a delightful story about the notoriously naughty child who “crushes the tops of the flowers” (Alcman fr. 58) and now finally gets his comeuppance. But against the background of erotic metaphor, it unfolds its true potential and opens an entire field of poetic images and ideas revolving around psychological bonding and constraint—they are all grounded in the physical experiences of desire. So again, the visual arts draw upon the same emotion concepts as the written sources: in this case, that erotic desire has to be restrained in order to not restrain you. Returning to our tablinum in the “House of Punished Love,” the two opposite walls present quite different concepts of erotic love that we have already encountered in earlier descriptions: a desirable state when enjoyed in “due measure and harmony,” it tends to exceed the acceptable boundaries and therefore must be chastened.

Against this background, then, it is no wonder that fickle erôs was not perceived as the ideal emotional condition in married life. Sexual attraction and amorous love between spouses did not take high priority in Greek and Roman societies and were even eyed with suspicion due to their destructive potential. Ordinary mortals, as opposed to gods, were well advised to establish a long-term relationship that was not susceptible to the turmoils of violent passion, “an emotional tie stronger than pure sexual fascination” (Salvo 2016: 266). In ancient Greece, the ideal relationship between spouses is covered by the umbrella term philia, a multifaceted concept that applies to a variety of non-hierarchical relations and encompasses notions of mutual obligation, respect, and affection (cf. Hartmann 2002: 126–30; on the manifold notions of philia, see also Konstan 1997; Konstan 2006b: 169–84). There is no Roman concept of emotion that covers the whole range of Greek philia, but with regard to the relationship between spouses, the closest equivalent might be concordia (Larsson Lovén 2010: 204). Unlike the violent passions erôs and amor, these emotion concepts are characterized by longevity and moderation as they involve a strong rational element that is informed by cultural ideals and social necessities.

On Greek grave reliefs of the classical era, the ideal of a well-tempered and long-lasting emotional disposition towards one another finds its visual expression in the so-called dexiôsis (handshake). On classical grave reliefs, it is by far the most common motif to depict the bond between the dead and the living. In most cases, it is performed between husband and wife, but also parents and children or other relatives may be joined by the dexiôsis. These images are not necessarily intended to portray concrete situations (e.g. farewell or reunion), but rather they work on an abstract level, as witnesses of a connection beyond death (cf. Meyer 1999: 116–17). The fine grave stele of “Korallion, wife of Agathon” shows the deceased sitting on a stool and shaking hands with her husband, a bearded man slightly past his prime (see Fig. 5.11).[34] The sincerity of their handshake is strengthened by their intense eye contact and Korallion touching her husband’s forearm, a gesture mainly performed by female figures to further emphasize the attachment to their partner.[35]

Many scholars have questioned the emotional significance of the dexiôsis because of its range of functions within Greek imagery: while it certainly conveys the ideal of family cohesion in funerary art, it also appears on contemporaneous document reliefs where it denotes the contractual obligation between the parties.[36] By interpreting the dexiôsis as a visual code for the equally versatile concept of philia, the putative contradiction disappears: both philia and the handshake are based on mutual agreement and reciprocity and thus include, to varying degrees depending on the context, both formal and emotional elements (cf. Räuchle 2017: 235–7).

Figure 5.11. Athenian grave stele of Korallion (drawing), c. 350 bce.

The Roman equivalent of the dexiôsis is the so-called dextrarum iunctio, which often features in Roman imperial art to display the concordia between two parties (e.g. the emperor and a representative of the army). On private funerary monuments, it refers to the matrimonium iustum; thus it responds to legal concerns, but also symbolizes “a union in harmony.”[37] So again, the gesture does not allow for a clear decision whether it is supposed to express emotional or formal aspects of marriage—rather, these two dimensions seem to be inextricably linked.

While the handshake on Greek grave stelae is almost always combined with eye-contact and often reinforced by explicitly emotional gestures, the Roman dextrarum iunctio appears rather formal and rigid: it “involved a certain degree of touch, but it was usually only the right hands of husband and wife that made contact . . . The couple were often placed side by side, both looking out of the relief at the viewer, rather than at each other” (Davies 2017: 172).[38] Only in rare cases is the impression of aloofness sometimes softened by the woman turning towards her husband and gently touching his shoulder, as can be seen on a funerary altar from Rome (see Fig. 5.12).[39] As in Greek funerary art, these more direct displays of affection are usually restricted to women, while men tend to maintain their representative habitus (Davies 2017: 172–3).

Figure 5.12. Roman funerary altar, first century ce.

The gender differences we can observe in the display of erotic desire and marital love in visual media are in accordance with the emotion concepts formulated in written sources: Aristotle (Nicomachean Ethics 1158b20–5) holds that affection should be proportionate, that is, that men would generally receive more love than they return since they are superior (cf. Hartmann 2002: 126). In Roman culture, too, “spousal devotion and loyalty were virtues more commonly attributed to women” (McCullough 2011: 175). Needless to say, these feeling norms played a decisive role in shaping the emotional experiences of the members of the respective community: Greek curse tablets dealing with love matters, for example, show that women were primarily interested in affirming and increasing affection (philia), while men primarily wished to arouse passionate love (erôs) in their love interests (Salvo 2016: 264).

As could be demonstrated, Greek and Roman art provides manifold strategies of distinguishing between different forms of love—from acute passion to lifelong affection. The first explicit classification of these different emotional phenomena is provided in the Nicomachean Ethics, where Aristotle elaborates on the nature of philia and defines it as a “fixed disposition,” comparable to êthos (character)—and opposed to pathos: “Liking [philêsis] seems to be an emotion [pathos], friendship [philia] a fixed disposition [hexis], for liking can be felt even for inanimate things, but reciprocal liking involves deliberate choice, and this springs from a fixed disposition [hexis].”[40] Of course, there is no sharp distinction between these different emotional phenomena: an emotional disposition can encourage the experience of acute emotions, while the repeated experience of certain emotions can lead to the formation of a fixed disposition.[41] The same logic applies to the visual expressions of emotions: many of the codes that appear to be spontaneous emotional reactions can at the same time point to a character trait. By skillfully playing with the ambiguous and multivalent nature of their visual codes, Greek and Roman art has managed to convey the fascinating world of emotions in all its complexity.

The Emotive Power of Images: The Suffering “Other”

Having tackled in previous sections various aspects of the representation of emotions in Greek and Roman iconography, we shall now address the emotive qualities of visual media, that is, their ability to elicit emotional reactions in the recipient. There are countless visual strategies for affecting the beholder and innumerable possibilities to emotionally engage with a work of art. The emotive power of an image is not coextensive with the use of explicit pathos formulas, but to a large extent also depends upon other factors such as cultural background and context of reception. As the grave relief of Ampharete with her grandchild has shown, even extremely restrained representations can have a strong emotional effect on the viewer.

In what follows, we shall look at the other side of the spectrum and turn to a work of art that combines the subject of violent death and total defeat with the excessive use of emotion expression and pathos formulas: the “Greater Attalid Dedication.” Also known as the “Greater Barbarians,” this famous victory monument was commissioned by King Attalos I on the occasion of his battles against the Gauls between 238 and 223 bce, and (most likely) erected in the sanctuary of Athena in Pergamum.[42] The original bronze monument is gone, but two figures are preserved in Roman marble copies, namely the “Dying Gaul” and the “Ludovisi Gaul.” Both statues can be regarded as epitomes of Hellenistic pathos as they are characterized by expressive gestures and postures, exaggerated facial features, and a sophisticated composition that actively involves the onlooker (Kunze 2002: 40–3, 47–51).

The “Dying Gaul,” identified by the torques around the neck, the shaggy hair and moustache, has sunk to the ground, severely injured on the back by an absent enemy. The “closed” composition of the figure ingeniously captures the solitude of death: he has withdrawn into himself, suffering in silence and waiting for the “black cloud of death” to enfold him (θανάτου δὲ μέλαν νέφος ἀμφεκάλυψεν, Homer, Iliad 16.350; cf. Cairns 2016a: 31, n. 29). The more dynamic “Ludovisi Gaul” represents a Galatian chief who, finding himself in the hopeless situation of being besieged by invisible opponents, has killed his wife and is now turning to suicide as the last resort (see Fig. 5.13).[43] Presenting his muscular body in a dramatic vertical torsion that culminates in a vehement head rotation, he is vigorously driving the sword into his breast while holding the lifeless body of his wife, who has just collapsed under his fatal blow. It is generally accepted that the original dedication did not represent the Greek victors, but focused on the defeated parties (cf. Kunze 2002: 40, n. 148; Winkler-Horaçek 2011: 141). This deliberate omission can be interpreted as a means of actively involving the viewers as they are invited to fill the gap in their mind’s eye or even to assume the role of the vanquisher (Pirson 2000: 72–5; Kunze 2002: 51). The radical, unabashed focus on the Gauls’ suffering yields an extreme emotive power; yet it is strongly contested what kind of emotions were elicited in the ancient onlooker.

Figure 5.13. “Ludovisi Gaul,” Roman marble copy from Hellenistic original, c. 220 bce.

Earlier scholarship held that these images of total defeat portrayed the Gauls in their heroic willingness to self-sacrifice and thus served as a means of glorifying them as worthy opponents. In this rationale, the suicide of the Celtic chief was placed in the tradition of Stoic philosophy and hence interpreted as an admirable act to escape dishonor at the hands of the enemy (e.g. Wenning 1978: 50). Others argued that the close view on the barbarians’ misery aimed at the sympathy (eleos) of the beholder (Schober 1936: 123). Critics object that both the glorification of the defeated and the elicitation of pity for their lot would be rather counterproductive for a victory monument (Hölscher 1985: 129–30; Winkler-Horaçek 2011: 141). Tonio Hölscher takes a slightly different view and suggests that the strategy of portraying the defeated with “pathetic empathy” (Hölscher 1985: 134) serves as a means of reflecting the vulnerability of human existence, the harsh and merciless nature of fortune—a perspective that is also adopted by the “tragic historians” of the Hellenistic era (cf. Cairns, introduction to this volume). At the same time, Hölscher emphasizes that this “emotional absolutization of defeat” by no means reduces the effect of the monument, but even contributes to the glory and splendor of the victors: their triumph is just as absolute as their opponent’s ruin (Hölscher 1985: 128–30).

Luca Giuliani argues against these interpretations that a sympathetic, other-regarding stance towards the enemy was not compatible with the ancient concept of pity (eleos), but rather influenced by the Christian ideal of compassion as an unconditional and altruistic sympathy for all mankind (Giuliani 2004: 18). He bases his claim on Aristotle, who at some point in the Rhetoric defines eleos as “a kind of pain in the case of a destructive or painful harm in one not deserving to encounter it.”[44] In this reading of eleos as depending “on the judgement of whether the other’s suffering [is] deserved or not” (Konstan 2006b: 201; cf. Konstan 2001: 34–43), the wretched barbarians of the “Greater Attalid Dedication” would indeed not qualify for the pity of the beholder. However, numerous sources suggest that even the woe of the villain could elicit sympathy (Cairns 2004: esp. 65, n. 3).[45] Especially in the Hellenistic era, both literature and visual art did not so much aim at a moral valuation of their accounts but rather at their vividness (enargeia) that stimulated the imagination (phantasia) and thus created an emotional effect of its own. So, on the ground of the extant sources, we cannot dismiss the possibility that ancient beholders felt sympathy at the sight of the suffering Celts.

A diametrically opposed interpretation is offered by Andrew Stewart, who claims that any Greek would “enjoy a put-down of barbarians, especially the hated Gauls” (Stewart 2004: 229), and characterizes the monument in highly suggestive prose:

So in the statues, the Celts’ heads are indeed satyrlike and their physiques are absolutely un-Greek, extremes of either defect or excess. For their skin is either so thick that it obscures the bodies’ rational construction, or so thin that the muscles pop out like tumors all over the surface. Their nakedness too is insane, a nightmarish example of Greek artistic convention come to life, but now in the person of the barbarian berserker . . . Even the Suicidal Celt’s “heroically” poised head is a perversion and parody . . . his questing glance is not a sign of strength but an index for weakness and the ultimate, irremediable lack.

 --Stewart 1997: 220

This vivid account says much more about Stewart’s powers of imagination than about the actual compositional and stylistic properties of the statues in question. While the expressive facial features are indeed rooted in the long tradition of the “barbarian berserker,” other pathos formulas such as the contortion of the bodies or the exaggeratedly protruding muscles have to be understood as general stylistic devices of the time that could be used for representations of Greek heroes as well (cf. Fig. 5.2). Certainly, it cannot be ruled out that some ancient beholders reacted to the view of the dying barbarians with sadistic satisfaction, but the monument’s composition and the rendition of the figures by no means predetermine this particular reaction.

The interpretations just cited differ radically with regard to conjectured emotive effect but nonetheless assume a highly affective reaction in ancient viewers. Luca Giuliani rejects such an emotionalized reading entirely and reconstructs a “curious but cool” recipient who appreciates the precise and concrete description of violence from a purely aesthetical perspective.[46] His concept of an attentive and critical audience is supported by ancient authors such as Plutarch, who states that “a successful imitation manifests a cunning and authority of its own, so that we take a natural delight in the performance but are distressed by the reality” (Plutarch, Moralia 673 F–674 A, trans. Clement and Hoffleit 1969: 381). However, as a professed Platonist and a member of a highly educated elite, Plutarch’s statement does not necessarily reflect the perspective of common folk.

In short, the ancient sources do not allow for a single interpretation of the “Greater Barbarians” (and other works of art), but rather suggest a diversity of possible perspectives: the commander of a victorious campaign against the Gauls would recognize their relentless fighting spirit in these statues and reactivate the pride of his triumph over such worthy opponents. A veteran who had seen his comrade being brutally slaughtered by a ferocious Celt perhaps experienced flaring hatred at the sight of the barbarians but also a sense of satisfaction at their demise. A woman beholding the statue of the Gaul who just murdered his wife, however, may have reacted with a certain sympathy for the predicament of the female as the perpetual victim of male power games. And finally, a rhetorically trained intellectual may have contemplated the monument primarily on the basis of aesthetic criteria such as successful imitation, artistic ingenuity, and craftsmanship. Of course, the possible reactions are not completely arbitrary but develop within the range of collectively shared values and build upon a common understanding of justice and desert—the “Greater Attalid Dedication” is a victory monument after all. Yet our fictitious recipients open up the possibility that the precise cognitive-affective reaction to an artwork not only depends on the cultural background but also on status and gender, as well as personal experiences and dispositions.

And yet, there seems to be a common core of aesthetic reception, an instinctive response to the sight of human suffering, or to the “vividness” of its representation in visual art, that is shared among all these different reactions. For even the cool eye of the distanced observer needs to be affected in some way in order to appreciate “successful imitation.” We may be able to distance ourselves from the misery represented in art—but the full aesthetic experience nevertheless requires an emotional engagement, a basic physiological response to the artwork. We have arrived at the fundamental mystery of aesthetic emotions.

The ancient term phrikê may contribute to our understanding of the complex processes at work here (see Cairns 2016b: 3–15; Cairns 2017a). Literally translated as “shivering” or “shuddering,” it can also be the name of an (aesthetic) emotion “that responds to the misfortunes of others” (Cairns 2017a: 71). It is this peculiar shiver of excitement at the view of another human being’s suffering, a shiver “that springs from a fascination with the spectacle . . . and yet also entails an instinctive revulsion towards that spectacle” (cf. Cairns 2017a: 55). The nature of phrikê as an “involuntary, instinctive response especially to immediate sensory stimuli” (Cairns 2016b: 14) does not predetermine the appraisal of an artwork, but allows for various emotional reactions. Phrikê, then, might be interpreted as an instinctive somatic experience, that enables us “to put some phenomenological flesh on the bare bones” of aesthetic reception (Cairns 2017a: 73).


There are various ways in which social and cultural structures stipulate the representation and elicitation of emotions through visual media: the reluctance of ancient artists to manifest emotional states through facial features is in line with the ideal of self-control and respective behavioral norms in everyday interactions. The status- and gender-typical emotion codes in images of grief and mourning are connected to real display rules and different roles in funerary ritual, but at the same time hint at the culturally constructed dichotomies between male rationality and female emotionality, between the self-restraint of the free and the uncontrollable nature of slaves; likewise, the visual distinction between female and male desire and affection in the images can be linked to cultural ideals about gender-conforming behavior and experience. These distinctive patterns in emotion expressions reach beyond the emotional communities of Greece and Rome when barbarians and mythical figures enter the stage: even in the vigorous pathos of Hellenistic sculptures one can detect a fine differentiation between the ordered passions of the Greeks and the dangerous passions of barbarians. In short, visual codes of emotion establish a relatively strict hierarchy in emotion management that reproduces and at the same time substantiates the social and cultural structures of the community in which they originated.

In classical antiquity, as in other periods, the spheres of emotion and cognition are inextricably linked and cannot be considered separately. In the ancient Greek (and Roman) thinking, “emotions necessarily involve judgements and beliefs, and without these, one can hardly speak of emotions as [sic] all. In a word, emotions are profoundly rational” (Konstan 2017a: 39). It is no surprise, then, that the distinction between acute emotions and long-lasting dispositions or character traits in the images is anything but clear-cut. As could be demonstrated, this fuzziness is not due to the limits and restrictions of visual media alone, but deeply rooted in ancient perceptions of emotion and cognition: the ambivalence of the visual code of veiling is related to the ambiguity of aidôs/pudor as both an acute feeling of shame and a long-lasting, moral disposition towards modesty. The versatile use of the handshake (dexiôsis and dextrarum iunctio) in Greek and Roman art aptly conveys the equally multifaceted concepts of philia and concordia that comprise rational and affective aspects. For the ancient mind, the boundaries between emotion and character are fluent.

Thus, the visual languages of emotion are first and foremost cultural phenomena that have to be analyzed in their context(s): the pictorial (and literary) traditions of single codes, figure-types, and compositions; the function and intended message of the image; the conditions of use and reception; the underlying beliefs, norms, and values of the respective community; and finally, the broader context of ancient “mentality.”

And then, sometimes, something happens: when we look at the image of Agamemnon covering his face at the sacrifice of his daughter, when we think of the peculiar display rules in grief and mourning at that time, when we then recollect the manifold metaphorical layers of the veil in ancient culture, we can relate to his need to withdraw from the outer world and identify with his experience of being shrouded with sorrow. Ancient literature and art are replete with metaphors that draw on the experiential, embodied nature of emotions; they enable us to glimpse behind the curtain of contextual and cultural specificity and understand, in Dilthey’s sense, the emotional and cognitive experiences of the past. There is no cultural history without the emotions.

[1] Single emotions: McNiven 2000b (fear in ancient Greece); Huber 2001 (Greek grief and mourning); Oakley 2004 (grief and sadness on lekythoi); Sojc 2004 (sadness on Classical grave reliefs); Räuchle 2017 (maternal emotions in classical Athens). On gestures and body language: Kenner 1960 (laughing and weeping in Greek art); Brilliant 1963 (gesture and rank in Roman art); Neumann 1965 (gestures in Greek art); Franzoni 2006 (pathos and êthos in Greek art); Catoni 2008 (schêmata in Greek art); Maderna 2009 (dangerous passions in Greek and Roman art); Masséglia 2015 (body language in Hellenistic art); Davies 2017 (proxemics in Roman art).

[2] On extreme violence in the fifth century bce, see Fischer and Moraw 2005; Seidensticker and Vöhler 2006; Muth 2008; Zimmermann 2009.

[4] Cf. Schnell 2007, esp. 177–80; see also Chaniotis 2012c: 14: “In direct communication, we have other media to increase the accuracy of the expression of feelings, such as facial expressions, raising or lowering the voice, and body-language . . . mimic and gestures are also represented in art. This information is, however, filtered and sometimes subject to conventions of representation.”

[5] Group of centaur biting Lapith, c. 460 bce, Archaeological Museum, Olympia: Ridgway 1965: 49; Maderna 2009: 10, Fig. 2.

[6] Roman marble copy from Hellenistic original around 150–120 bce, Loggia dei Lanzi, Florence: Mylonopoulos 2017: 81, Fig. 7.

[7] Prioux 2011: 137; see also Franzoni 2006: 63–8. On sôphrosunê in general, see North 1966. On sôphrosunê and emotion control, see Harris 2001: 80–7. On female sôphrosunê, see North 1977; Räuchle 2017: 28–30, 139–40.

[8] Neumann 1965: 106: “Während die ‘Momentangebärden’ affekthaft und durch ausfahrende Gesten gekennzeichnet sind, sind die ‘Zustandsgebärden’ überwiegend von kontemplativ-nachdenklichem Charakter. Sie geben Kunde von der spezifischen Verfassung der Gestalt, von der mehr oder weniger dauerhaften inneren Zuständlichkeit.” (“While the Momentangebärden are emotional and marked by outward gestures, the Zustandsgebärden are predominantly contemplative and thoughtful in character. They give information about the specific consitution of the figure, of the more or less permanent inner state.”)

[9] Red-figure lekythos, c. 420 bce, Metropolitan Museum, New York 17.230.35: BAPD 214280; Reeder 1995: 356–7, no. 113.

[10] Cairns 2016b: 7. On personifications between allegory and belief, see Stafford 2000: 1–44.

[11] Alcman fr. 58 (trans. Campbell 1988: 435): Ἀφροδίτα μὲν οὐκ ἔστι, μάργος δ᾿ Ἔρως οἷα <παῖς> παίσδει, / ἄκρ᾿ ἐπ᾿ ἄνθη καβαίνων, ἃ μή μοι θίγῃς, τῶ κυπαιρίσκω.

[12] IG I3 1290: Ἀμφαρέτη. | τέκνον ἐμῆς θυγατρὸς τόδ’ ἔχω φίλον, ὅμπερ ὅτε αὐγάς: / ὄμμασιν ἠ|ελίο ζῶντες ἐδερκόμεθα, / ἐ̑χον ἐμοῖς γόνασιν καὶ νῦν φθίμενον φθιμένη ’χω.

[13] Euripides, Suppliants 815–17 (trans. E. P. Coleridge): δόθ᾽, ὡς περιπτυχαῖσι δὴ / χέρας προσαρμόσασ᾽ ἐμοῖς / ἐν γκῶσι τέκνα θῶμαι.

[14] See Räuchle 2017: 210–18. Funerary plate with inscription labeling the woman embracing the head of the deceased as the mother (mêtêr): black figure pinax, c. 500 bce, Louvre, Paris MNB 1905: BAPD 463; Neils and Oakley 2003: 165, Fig. 3. Older man and woman touching the deceased youth’s head, likely to be identified as father and mother: white lekythos, 410–390 bce, Antikensammlung Berlin F 2684: BAPD 217904; Oakley 2004: 84, Fig. 54. Old nurse standing at the top of the bier and bidding farewell to the untimely deceased woman: red figure loutrophoros, 475–450 bce, National Museum, Athens 1170: BAPD 205750; Huber 2001: 127.

[15] White lekythos, c. 430 bce, National Museum, Athens 19355: BAPD 214321; Oakley 2004: 163.

[16] Originally used in the nuptial rites, the loutrophoros became a grave marker for those who died unmarried and childless: see Sabetai 2008.

[17] Sarcophagus, c. 120–130 ce, Museo Nazionale, Agrigento: Amedick 1991: 121 cat. no. 2.

[18] Plutarch, Consolation to his Wife 609a: (trans. De Lacy and Einarson 1959: 587) οὐ γάρ ‘ἐν βακχεύμασι’ δεῖ μόνον τὴν σώφρονα μένειν ἀδιάφθορον, ἀλλὰ μηδὲν οἴεσθαι ἧττον τὸν ἐν πένθεσι σάλον καὶ τὸ κίνημα τοῦ πάθους ἐγκρατείας δεῖσθαι διαμαχομένης οὐ πρὸς τὸ φιλόστοργον, ὡς οἱ πολλοὶ νομίζουσιν, ἀλλὰ πρὸς τὸ ἀκόλαστον τῆς ψυχῆς.

[19] E.g. Homer, Iliad 18.22: “The black cloud of sorrow enwrapped Achilles” (τὸν δ’ ἄχεος νεφέλη ἐκάλυψε μέλαινα); 8.124: “Then was the soul of Hector clouded with dread sorrow” (Ἕκτορα δ᾽ αἰνὸν ἄχος πύκασε φρένας); Euripides, Heracles 1140: “Alas! a cloud of lamentation envelops me” (αἰαῖֹׄ στεναγμῶν γάρ με περιβάλλει νέφος). For further examples, see Cairns 2016a: esp. 34–6; cf. Cairns 2009, esp. 51: “The grief itself is the cloud or the concealing garment of which the literal garment is the external visual symbol.”

[20] Euripides, Iphigenia at Aulis 1547–50 (trans. Kovacs 2002: 335): ὡς δ᾽ ἐσεῖδεν Ἀγαμέμνων ἄναξ | ἐπὶ σφαγὰς στείχουσαν εἰς ἄλσος κόρην, | ἀνεστέναξε, κἄμπαλιν στρέψας κάρα | δάκρυε, πρόσθεν ὀμμάτων πέπλον προθείς. Cf. Cairns 2011b: 19.

[21] Quintilian 2.13.12–13 (trans. Russell 2001: 345): “Quid? non in oratione operienda sunt quaedam, sive ostendi non debent sive exprimi pro dignitate non possunt? Ut fecit Timanthes, opinor, Cythnius in ea tabula qua Coloten Teium vicit. Nam cum in Iphigeniae immolatione pinxisset tristem Calchantem, tristiorem Ulixem, addidisset Menelao quem summum poterat ars efficere maerorem: consumptis adfectibus non reperiens quo digne modo patris vultum posset exprimere, velavit eius caput et suo cuique animo dedit aestimandum.”

[22] Wall painting from the “House of the Tragic Poet” in Pompei, 62–79 ce, Museo Archeologico Nazionale di Napoli 9112: Lorenz 2008: 392, Fig. 200; Chaniotis et al. 2017: 115, no. 31.

[23] Iphigenia does not “stand at the altar awaiting her doom,” as Pliny describes (“qua stante ad aras peritura,” Naturalis Historia 35.73), but is carried by two male figures and raises her arms in a gesture of heightened emotionality. As noted above, this gesture is ambiguous and could in the present context either denote panic or supplication.

[24] Cf. Cairns 2009: 47–8: “The gesture of veiling, then, is a characteristic and powerful marker of grief, one that will have been familiar to Greek audiences not only from epic, tragedy and the visual arts, but from their own experience.”

[25] On visual representations of the episode, see Mangold 2000: 80–102, 184–96; Ritter 2005. On the various versions of the episode in literature, see Mangold 2000: 81–2. On the conflicting emotions of Menelaus in this episode, see Kaltsas 2017: 101–2.

[26] Athenian red-figure krater, c. 450 bce, Louvre Paris G 424: BAPD 214486; Ritter 2005: 276, Fig. 6.

[27] Athenian red-figure lekythos, c. 460–450 bce, State Hermitage Museum St Petersburg b 4524: BAPD 215792; Ritter 2005: 272, Fig. 4a–b. For other examples of Eros with phialê, see Ritter 2005: 270–6.

[28] Euripides, Hippolytus 525–7 (trans. Kovacs 1995: 175): Ἔρως Ἔρως, ὁ κατ᾽ ὀμμάτων στάζων πόθον, εἰσάγων γλυκεῖαν ψυχᾷ χάριν οὓς ἐπιστρατεύσῃ . . .

[29] Athenian red-figure loutrophoros, c. 420 bce, Ashmolean Museum Oxford 1966.888: BAPD 34; Reeder 1995: 169, no. 25.

[30] Euripides, Hippolytus 528–9 (trans. Kovacs 1995: 175): μή μοί ποτε σὺν κακῷ φανείης μηδ᾽ ἄρρυθμος ἔλθοις.

[31] Wall paintings from the “House of Punished Love” in Pompei, first century ce, Museo Archeologico Nazionale di Napoli 9249. 9257: Lorenz 2008: 153, Fig. 48, 266, Fig. 125.

[32] For visual representations of the punishment of Eros, see Curtius 1930; Delivorrias 1984: no. 1252–4; Hermary and Cassimatis 1986: no. 417–26; Blanc and Gury 1986: no. 77–84. For the poetic tradition regarding a statue of “Eros in chains,” see Anthologia Graeca 16.195–9.

[33] Anthologia Graeca 16.199 (trans. Paton 1943: 257–7): καὶ κλαῖε καὶ στέναζε, σὺσφίγχθεὶς χεροῖν / τένοντας, ὦ ’πίβουλε∙ τοῖά τοι πρέπει. / οὐκ ἔσθ’ ὁ λύσων∙ μὴ ’λεείν’ ὑπόβλεπε. / αὐτὸς γὰρ ἄλλων ἐκ μὲν ὀμμάτων δάκρυ / ἔθλιψας, ἐν δὲ πικρὰ καρδὶᾳ βέλη / πήξας ἀφύκτων ἰòv ἔσταξας πόθων, / Ἔρως∙ τὰ θνητῶν δ’ ἐστί σοι γέλως ἄχη. / πέπονθας, οἷ’ ἔρεξας. ἐσθλὸν ἡ δίκη.

[34] Grave stele, c. 350 bce, Athens Kerameikos Museum P688: Clairmont 1993: 4, no. 4.415. Inscription (IG II311891): Κοράλλιον Ἀγάθωνος γυνή.

[35] Gestures that can be interpreted as explicitly emotional, e.g. the touching of another person’s shoulder or chin, are mostly performed by women; see Meyer 1999: 122–9; on the gender-specific display of emotion, see below.

[36] On the handshake in general, see Neumann 1965: 49–59. On Greek funerary art, see Davies 1985; Pemberton 1989; Meyer 1999; Räuchle 2017: 235–7.

[37] Cf. Alexandridis 2000: 17–18; Larsson Lovén 2010: 214; Davies 2012: 29.

[38] The differences in composition and style create a curious and paradoxical effect: the Greek dexiôsis scenes purport to reproduce reality, though they do not, whereas representations of the dextrarum iunctio often have a certain unrealistic touch even though the gesture formed an integral part in the actual wedding ritual.

[39] Funerary altar, first century ce, Museo Nazionale Romano 124514: Larsson Lovén 2010: 212, Fig. 22.

[40] Aristotle, Nicomachean Ethics 1157b27–32 (trans. Rackham 1926): ἔοικε δ᾽ ἡ μὲν φίλησις πάθει, ἡ δὲ φιλία ἕξει. ἡ γὰρ φίλησις οὐχ ἧττον πρὸς τὰ30 ἄψυχά ἐστιν, ἀντιφιλοῦσι δὲ μετὰ προαιρέσεως, ἡ δὲ προαίρεσις ἀφ᾿ ἕξεως. Considered as a hexis, then, (conjugal) philia depends upon a specific emotional attitude towards the other person (i.e. spouse) and a certain choice to act accordingly, thus comprising both rational and emotional components; cf. Ward 1996: 160.

[41] This interrelation betwen pathos and hexis is also apparent in ancient physiognomic treatises: while they are generally more interested in hexis than in pathos, they derive many of their judgements on permanent character traits from facial expressions of acute emotions; e.g. Aristotle, Physiognomics 805a29–b9; cf. D. L. Cairns 2005: 127, n. 18 with further references.

[42] On the “Greater Dedication,” see Schober 1936; Hölscher 1985; Pirson 2002; Marszal 2000; Kunze 2002, 40–51; Winkler-Horaçek 2011. The nicknames were coined to distinguish the votive from the “Lesser Attalid Dedication,” which stood on the Athenian Acropolis and was significantly smaller in scale. The highly controversial discussion regarding the original context, dating, and reconstruction of the Lesser Dedication cannot be considered here. For a thorough discussion, see Engel 2015.

[43] “Ludovisi Gaul,” Roman marble copy from Hellenistic original c. 220 bce, Museo Nazionale Romano 144: Kunze 2002: tab. 5, Figs 9–10.

[44] Aristotle, Rhetoric 2.8, 1385b13–16 (trans. Konstan 2006b: 204): ἔστω δὴ ἔλεος λύπη τις ἐπὶ φαινομένῳ κακῷ φθαρτικῷ ἢ λυπηρῷ τοῦ ἀναξίου τυγχάνειν, ὃ κἂν αὐτὸς προσδοκήσειενἂν παθεῖν ἢ τῶν αὑτοῦ τινα, καὶ τοῦτο ὅταν πλησίον φαίνηται.

[45] Moreover, even Aristotle allows for a basic emotional reaction to another person’s suffering. In his Poetics (13, 1453a2–6) he states that the suffering of a profoundly bad person may not be able to trigger eleos or phobos but to philanthrôpon, an emotional response to another’s misfortune that is indifferent to desert, “an instinctive sensitivity to the suffering of others” (Konstan 2006b: 217).

[46] “Gerade in der Antike haben wir vielfach mit einem neugierigen aber kühlen, genießenden Blick zu rechnen, dem die möglichst präzise und konkrete Schilderung von Gewalt in erster Linie Anlass zu ästhetischem Vergnügen gewesen zu sein scheint” (“In antiquity, we often have to reckon with a curious but cool, appreciative look, to which the most precise and concrete portrayal of violence seems to have been primarily a cause of aesthetic pleasure,” Giuliani 2004: 20).