Jump to ContentJump to Main Navigation
Word And Image In Ancient Greece$

N. Keith Rutter and Brian Sparkes

Print publication date: 2000

Print ISBN-13: 9780748614066

Published to Edinburgh Scholarship Online: March 2012

DOI: 10.3366/edinburgh/9780748614066.001.0001

Show Summary Details

Eidôla in Epic, Tragedy and Vase-Painting

Eidôla in Epic, Tragedy and Vase-Painting

(p.140) 8 Eidôla in Epic, Tragedy and Vase-Painting
Word And Image In Ancient Greece

Ruth Bardel

Edinburgh University Press

Abstract and Keywords

Greece was not only the birthplace of the dramatic arts, it was also the birthplace of the stage-ghost, a much-neglected but utterly fascinating dramatic character. Ghostly etiquette demands that a ghost appear at night to one unaccompanied person, a code of behaviour which renders all ancient Greek stage-ghosts rather impertinent, appearing as they do in broad daylight. In order to avoid such complicated categorisation and its attendant problems, this chapter focuses on the word ‘eidolon’, the term that is used to designate stage-ghosts in the dramatis personae of ancient Greek tragedy. By focusing on this word, however, it soon becomes apparent that this figure not only seemed to refuse any one categorisation, but was also a provocative amalgamation of many areas normally held to be distinct; in particular, iconography.

Keywords:   Greece, stage, ghost, eidolon, tragedy, categorisation, iconography


A 1999 ADVERTISEMENT1 cast Greece as the ‘longest running theatrical event’, a status paradoxically endorsed by ‘Friends, Romans, countrymen’. The caption underneath the full page advert stated, ‘Ask anyone who has been to Greece about the spectacular open air amphitheatres. Witness the grandeur of these astonishing monuments either as a visitor or a spectator in the musical shows, theatrical plays and other cultural events featured each summer in the land that is the birthplace of the dramatic arts’ (my italics). The advertisement’s background picture shows a drama in mid-flow, staged in the Odeion of Herodes Atticus (built in the second century AD) at the foot of the Acropolis. Assuming for a moment that the drama in question is a re-enactment of an ancient Greek tragedy, the disjuncture between image and text (and even context) becomes apparent. It is precisely this often uneasy alliance between image and text that this chapter examines in relation to the figure of the ghost or, more specifically, the eidôlon.

Greece was not only the birthplace of the dramatic arts, it was also the birthplace of the stage-ghost, a much neglected but utterly fascinating dramatic character. Since Ruby Hickman’s 1938 monograph, Ghostly Etiquette on the Classical Stage, there has been no comprehensive study of ghosts in ancient Greek drama.2 ‘Ghostly etiquette’ demands that a ghost appear at night to one unaccompanied person, a code of behaviour which renders all ancient Greek stage-ghosts rather impertinent, appearing (p.141) as they do in broad daylight, in the open-air theatre of Dionysus, often in the presence of the chorus – not to mention the spectators. Ghosts are slippery customers and classification does cause problems. Hickman (1938: 16, 62–3, 124–5, 160), for example, categorises ghosts in Greek and Roman tragedy as dream-ghosts, stage-ghosts, off-stage-ghosts, doubtful ghosts, anonymous ghostly shapes, hallucination, pseudoghosts, borderline-ghosts, fictitious dream-ghosts and minor pseudoghosts. In order to avoid such complicated categorisation and its attendant problems I have focused on the word eidôlon, the term which is used to designate stage-ghosts in the dramatis personae of ancient Greek tragedy and which, although it does in many ways support the kinds of distinction that Hickman makes, certainly facilitates the attempt to categorise such elusive figures as ghosts.

By focusing on the word eidôlon, however, it soon became apparent that this figure not only seemed to refuse any one categorisation but was also a provocative amalgamation of many areas normally held to be distinct, such as religion, philosophy, epistemology, ontology, representation and, in particular, iconography. As the lengthy title of Peifer’s 1989 monograph, Eidola und andere mit dem Sterben verbundene Flügelwesen in der attischen Vasenmalerei in spätarchaischer und klassischer Zeit, demonstrates, iconography generally designates the small winged figures found in vase-paintings as eidôla, in contrast to the literary and dramatic deployment of the word eidôlon, which designates an unwinged, life-sized and life-like figure. The attempt to secure an effective definition of the word eidôlon and its connotations for a fifth-century spectator, whether of the dramatic or visual arts, is perhaps analogous with Odysseus’ futile attempt to embrace the ghost of his dead mother in the Odyssey. However, an attempt to clasp the ghost, as it were, may shed some light on the nature and meaning of the eidôlon.

Beginning, somewhat anachronistically but nevertheless in deference to one of the eidôlon's defining features, a discussion of the well-known ‘Medea vase’ (Figures 8.18.2) will graphically highlight the disjuncture between the dramatic and/or literary and the iconographic use of the word eidôlon. The only figure in Greek vase-painting which can with absolute certainty be identified as a tragic ghost is the figure labelled ‘EIΔΩΛON AHTOY’ in this fourth-century volute-krater by the Underworld Painter, inspired by an unknown Medea tragedy.3 According to Siebert, Aeetes’ clothes mark him as barbarian, and the column of smoke beneath his feet marks him as a tragic ghost (Siebert 1981b: 67); apart from these two signifiers there is little to denote ‘spectre’. Trendall and Webster (1971: (p.142)

Eidôla in Epic, Tragedy and Vase-Painting

Figure 8.1 Apulian red-figure volute-krater by the Underworld Painter, c.330–320 BC (Munich 3296; photo: Blow Up)

110) thought the ‘column of smoke’ beneath the eidôlon's feet was a rock: but, whatever the case, it does seem significant that the ghost of Aeetes is the only figure in this representation who stands, for want of a technical expression, on a wobbly, smoke-like rock. All the other figures are, as it were, firmly grounded. Aeetes’ ghost looks as substantial as the other figures in the vase-painting, played as he must be by a very live actor. The eidôlon may have delivered the prologue, as Taplin suggests (1997: 80), or he may have returned, as Shapiro states (1994: 181), ‘in spirit to remind his daughter of the betrayal of her hearth and home that has brought her ultimately to this sorry state’. Whether at the beginning or at the end of the narrative, the eidôlon often stands outside the temporal sequence of the (literary, dramatic or pictorial) narrative proper.

Whatever the rôle of the ghost, it is clear that the artist has compressed time and space: the painter has combined, as Shapiro (1994: 181) notes, ‘at least three discrete scenes into one multi-level composition, with (p.143)

Eidôla in Epic, Tragedy and Vase-Painting

Figure 8.2 Detail of Figure 8.1

subsidiary figures who may allude to several more’. The eidôlon itself, in many significant ways, acts within the representational – or even theatrical – space as a marker of this compression of time and space, a figure from the past whose gaze seems to be directed downwards, towards the final sequence in the present narrative. Aeetes’ eidôlon stands at the fringes of the pictorial space, spanning two of the three distinct scenes.

(p.144) Unlike the other figures in the vase, who all seem to be interacting with, or in response to, at least one other character, the eidôlon's isolated, marginal position is, therefore, all the more marked and one suspects that he has little or no relation to the dramatic action unfolding around him – very much the ideal candidate for a prologue speaker. But, most significantly, within this pictorial representation, the image and the text/inscription cohere: the eidôlon is a life-size, life-like and fully recognisable figure, and Homeric epic seems to endorse this late, that is fourth-century, use of the tag eidôlon.


The ghost, like the word eidôlon, always has a past and its history and/or story helps to explain its appearance in the world of the living. Since ‘Pictures and descriptions of ghosts are not easy to come by’ (Winkler 1980: 160), it will be worthwhile to examine the deployment of the word eidôlon in Homeric epic, which, undoubtedly, influenced the nexus of ideas surrounding this dramatic figure. The first extant use of the word eidôlon appears in the fifth, and arguably the most chaotic, book of the Iliad. The already anomalous conditions of warfare are heightened, in the fifth book, by the direct intervention of the gods in the conflict between the Trojans and the Achaeans. Ontological categories which are normally kept distinct here interact and the correlation between the human and the divine problematises cognitive certitude and is set against a backdrop of uncanny experiences. One such instance of ‘supernatural events’ (Fenik 1968: 39) is the creation, by Apollo, of Aeneas’ eidôlon.

Wounded by Diomedes, Aeneas is speedily removed from the battlefield by Apollo, who thus completes the rescue mission initiated by Aphrodite but frustrated by Diomedes. Apollo then devises an eidôlon (v.449) resembling Aeneas’ self and wearing similar armour (v.450), and around this eidôlon, the Trojans and Achaeans continue fighting, unaware that any substitution has taken place (v.451–3). The very short episode of the creation of Aeneas’ eidôlon and its insertion into the heart of battle provides discernible evidence, a token, of the gods’ intervention in the affairs of mortals. Most significantly, however, Aeneas’ eidôlon marks Diomedes’ breach of Athena’s injunction not to do battle against the gods and thus acts as a sign of this transgression. It is clear from this episode that the word eidôlon need not exclusively denote a ‘wraith of the dead’ (Kirk 1990: on Il. v.449–50): Aeneas is alive but wounded, and it may be that this precarious state qualifies the making of an eidôlon, whose essential qualities seem to be that of representation, substitution for an absent person and fidelity to the original.

Homer does not tell us precisely what Apollo used in the construction (p.145) of Aeneas’ eidôlon: in fact, it matters little since the emphasis is placed on the eidolon's external appearance. Apollo fashions (Eidôla in Epic, Tragedy and Vase-Painting) an eidôlon of Aeneas just as Athena makes (Eidôla in Epic, Tragedy and Vase-Painting) an eidôlon of Iphthime, Penelope’s sister, in the Odyssey (iv.795–800).4 The eidôlon is thus ostensibly a divinely, but ultimately a poetically, generated image of a living, wounded or, as we shall see later, dead but unburied individual. In all of these instances, however, the aspect of the eidôlon is crucial to its recognition: just as the clothes (or costume) of the barbarian king on the Medea vase would facilitate its recognition in the absence of the explanatory label, so Aeneas’ armour, as sported (or rather, imitated) by the eidôlon, acts as a signifier for the man himself, and in the same way, Iphthime is recognised by Penelope because the eidôlon adopts her corporeal aspect (Eidôla in Epic, Tragedy and Vase-Painting, iv.796).

But the apparent corporeal reality of Iphthime’s eidôlon contrasts strikingly with its mode of entry and exit: this eidôlon passes through the keyhole into Penelope’s room (iv.802), delivers its message of comfort and finally dissolves into the winds (iv.838–9). The distinction between what the narrator describes and what Penelope sees is explicit: in response to Penelope’s distress, the narrator describes what the ‘dim phantom’ (Eidôla in Epic, Tragedy and Vase-Painting, iv.824, 835) says in response. An eidôlon may be dim and shadowy (unreal) to its creator, but it is vivid and (real) to the percipient (cf. iv.841).

Similarly, the departure of Patroclus’ ghost coincides with Achilles’ attempt to embrace it: it is also at this point that Achilles, ‘seized with amazement … sprang up’ (Il. xxiii.101), and his response to what happened is significant. Striking his hands together, Achilles ‘spoke a word of lament’ (xxiii.102):

Eidôla in Epic, Tragedy and Vase-Painting

– even in death there is something, a psuchê and an eidôlon. As with the encounter between the sleeping Penelope and the eidôlon of Iphthime, there is a clear discrepancy between what the poet describes – the dim eidôlon or psuchê (respectively) – and what Achilles describes – both a psuchê and an eidôlon. And this is not mere textual redundancy, the conjunction of two synonymous terms. Etymologically, psuchê and eidôlon could not be more distinct: breath and image do not seem to cohere; the invisibility of the former and the visibility of the latter appear contradictory. The phenomenon described by the poet (text) and that of the character (image) are crucially different.

(p.146) Verbal descriptions, unhampered by concrete visual portrayal, as in Homeric epic, can portray eidôla slipping through keyholes or disappearing like a puff of smoke – all actions which would be demanding on the ancient, or even modern, stage just as they would be in static pictorial representations. As Aristotle notes (Poetics 1460al0–15), epic affords greater scope for the alogon, the inexplicable, because we do not actually see the persons of the story – rather as we do not actually see the ghosts of a ghost story. And yet, during the process of visualisation (enargeiâ), the images are conjured up in the mind’s eye, as the listener listens and gives shape and form to the poets’ words, imaginatively fleshing out, as it were, the bare bones of the narrative.

Like the eidôlon of Aeneas, that of Patroclus is ‘in all things like his very self, in stature and lovely eyes and in voice and in like clothes was he clad’ (II. xxiii.66–7). It is these attributes, the ‘as-in-life’ appearance of the eidôlon, that seems to anchor it more firmly within the world of the living than in that of the dead. The difference, however, between the two descriptions is significant: the focus on the armour of Aeneas’ eidôlon is frigid in comparison with the intimate and tender description of Patroclus’ eidôlon, which is consistent with his characterisation as kind and gentle.5 The distinction between these two eidola – Aeneas (and indeed, Iphthime) and Patroclus – could not be more apparent; the one mechanical, deliberately created and used as a tool, the other spontaneous, sensitive and, as it were, full bodied. Both eidola are, however, reflections, so to speak, of the ontologically unstable status of their originals, the one wounded and the other dead but unburied. The ghost of the unburied Patroclus wanders in vain (II. xxiii.74): like Elpenor in the Odyssey (xi.54), Patroclus is an ‘exile’ in the other world, isolated from the living by death and from the ‘home’ of the dead until the proper rites of death have been performed.

It is also significant that Patroclus’ eidôlon appears without armour: that is, as he was before he was clad – and killed – in the armour he borrowed from Achilles. The lack of armour in the description of Patroclus’ eidôlon divests Patroclus of his warrior, and social, status and stresses the fraternity between the two men. Significantly, as Edwards notes (1991: on Il. xviii.338–42), it is the times that Achilles and Patroclus talked alone together, rather than their fighting exploits, that the eidôlon recalls. The eidôlon's evocation, in a ‘more leisurely narrative development’, of their childhood together in the house of Peleus and the ‘recollection of their closeness in life’ (Richardson 1993: on Il. xxiii.69–92) accentuates the memory of the living, dear comrade Patroclus rather than the wounded corpse of the dead warrior. The eidôlon is thus, in two senses, intimately (p.147) connected with memory, both of the recent and the more distant past, a memory which is articulated by a figure, itself dependent largely upon recollection.

It seems paradoxical that vase-painters, unlike Homer, focus on the martial qualities of the dead Patroclus; his diminutive ‘eidôlon’ (in iconographic nomenclature) is often armed and winged. The correlation between Hector’s and Patroclus’ deaths and the discrepancy between the treatment of their two corpses, a crucial theme in the Iliad, are brought together on a black-figure lekythos (c.510–500) of the Leagros Group6 which shows the tomb of Patroclus, Achilles in his chariot with Hector’s corpse tied to it, and two small, armoured and winged warriors hovering in the two comers of the scene, one on the left and one on the right. The wings attached to such ‘eidôla’ in vase-paintings suggest (rapid) movement rather than any intrinsic attribute of the eidôlon, a mobility that contrasts well with the inert corpse of Hector. Iris is also present, dispatched by Zeus, and here she makes a ‘warning gesture’ (Schefold 1992: 260) towards Achilles. This depiction is, as far as I am aware, unique in its presentation of two small, winged warriors: are they the so-called ‘eidôla’ of Patroclus and Hector? The presence of the two small winged figures illustrates, I suggest, in pictorial terms, the essence of the appearance of Patroclus’ ghost in Homer’s epic narrative. Patroclus’ eidôlon appears in Homer at the height of Achilles’ violation of Hector’s corpse and of the ethics of burial and of normal warfare: revenge, as Kerrigan (1996: 299) rightly argues, is ‘an overdetermining factor in combat’, where simply to have an opponent justifies killing. Achilles has, in effect, two restless souls on his hands.

The eidôlon's speech, by uniting the themes of proper burial rites, exile and supplication, marks the disruption of normal and acceptable behaviour in such circumstances. Furthermore, the appearance of Patroclus’ eidôlon is itself a symptom of this dismption: unwept and unburied, the restless ‘psuchê of hapless Patroclus’ (Il. xxiii.65) is an inadvertent casualty of Achilles’ relentless pursuit of vengeance. The vase-painting from Delos appears to unite the themes of the unwept and unburied corpses – both of Patroclus and of Hector – and the two ‘eidôla’ I propose, mark this abnormal state of affairs, just as the eidôlon of Aeneas marked Diomedes’ transgressive behaviour. If revenge was the focus of Homer’s text, he could have painted a very different picture of the ghost of an unwept and unburied man, one who had, furthermore, been violently slain and threatened with mutilation. The small winged warrior figures at the fringes of the representational space of the vase discussed above, as in (p.148) others depicting the same theme, are a small, but pertinent, reminder of Achilles’ violation of the cosmic order.

Patroclus’ eidôlon requests a speedy burial since the spirits (Eidôla in Epic, Tragedy and Vase-Painting), the phantoms of men that have done their work (Eidôla in Epic, Tragedy and Vase-Painting, xxiii.73), will not let him cross the river to ‘mingle’ with them.7 In the Odyssey, the same phrase recurs with a stress on the distinction between mortals and immortals – Eidôla in Epic, Tragedy and Vase-Painting (xi.476) – which highlights the laborious life of mortals in contrast to the ease of the immortals. This generic and formulaic phrase is a generalisation that dehumanises the dead, and also articulates their lamentable and depleted post-mortem state, mere images of worn-out mortals. Reduced, metonymically, to heads (Od. x.521, 536; xi.29, 49), the dead are barely recognisable, an undistinguished mass of vague, flitting shadows (Od. x.495). Such phrases seem to be reserved for the anonymous dead who have no attachment to the living percipient (in this case, Odysseus) and who have been dead for a long time. Against the backdrop of anonymous flitting shadows, the eidôlon stands out in conspicuous contrast.

The dead in general are conceived of in terms of superfluity: they are the hoi polloi (xi.42; cf., for example, x.521, 529–30, 536; xi.29). The sheer number, coupled with the notions of powerlessness, frailty and transience, heightens their insubstantiality and their distance from the world of the living. Most telling is the phrase, Eidôla in Epic, Tragedy and Vase-Painting (xi.632; cf. xi.25), a spectral multiplicity that overwhelms and frightens Odysseus both at the very beginning (xi.42–3) and at the very end of his encounter with the ‘glorious tribes of the dead’ (x.526; xi.29). It is the anonymous legions of the dead tribes which frighten Odysseus (xi.37–43), not the individual eidôla of familiar people such as Elpenor or his mother. But to characterise Homer’s ghosts as ‘whining, impotent things of little use … as imposing as puffs of smoke … drab creatures trapped in an utterly dull space’, as does Finucane (1996: 5), is to miss the point. In Homeric epic, it is the psuchê that is consistently described as insubstantial by analogy to a wisp of smoke, a dream or a shadow. These similes and other figurative language used to describe the post-mortem state suggest that metaphysical notions originate at the very level of the image, the eidôlon.

It is the images of the anonymous dead that, I suggest, influence the small winged ‘eidôla’ schema in iconography, in particular on white-ground lekythoi. The Charonian scenes with their legions of small winged figures serve, as Sourvinou-Inwood argues (1995: 336–7), to render death as ‘Other’. The small winged ‘eidôla’ in scenes depicting the reception of (p.149) the deceased by Charon are ‘part of the landscape’, which helps to denote the ‘localisation of the scene in the underworld’; they also ‘participate in the shade’s reception, and help articulate her introduction into the community of the dead’ (Sourvinou-Inwood 1995: 337). One such scene on a black-figure lekythos by the Sabouroff Painter8 shows (from left to right) Charon in his boat with Hermes, who reaches out his hand to a life-sized, veiled woman: many ‘eidôla’ are also represented, ‘the majority shown coming from the direction of Hades on the other side of the water, like a reception committee’ (ibid.).

Here, the central figure of Hermes bridges the gap between two literary and iconographic articulations of the post-mortem state: on the one side (the left), the small winged ‘eidôla’ predominate and on the other (the right), the full-scale image (eidôlon) of a veiled woman. I suggest that, rather than acting as a reception committee, the small-scale ‘eidôla’ are, like the anonymous dead of Homeric epic, the frightening image of what one becomes upon death (psuchai). One wonders if the vase-painter sought to convey how the movement from the world of the living to the world of the dead corresponded to a change in size and status. Does the visual, textual or dramatic representation of the dead depend upon which side of the fence – or boat – one is on? Just as Patroclus’ eidôlon stands out from the rest of the dead, so the eidôla of the Odyssean Nekuia stand out in sharp relief from the hoi polloi, and so the eidôlon of the veiled woman stands out amid the psuchai that flit about. The throngs, the anonymous hoi polloi, thus provide the background – as in the lekythos discussed above – against which the recognisable individual ghosts (eidôla) appear and serve to emphasise the distinction between these two categories of the dead in both literary and pictorial representations.

Elpenor bids Odysseus ‘remember me’ (Od. xi.71), a spectral injunction that highlights the association of the eidôlon with memory, and one which resonates throughout ghostly appearances within drama. The motif of the unwept and unburied corpse is a prominent one in tragedy: improper burial disturbs the social and cosmic order, throwing up plaintive ghosts (Polydorus in Euripides’ Hecabe, for example), whereas proper burial mediates between the desire to remember and the necessity of forgetting the dead. It may well be that Elpenor’s privileged status as an eidôlon derives from his close association with Odysseus, a reflection of the fact that, as in the cases of Aeneas, Patroclus and Iphthime, the eidôlon draws upon the framework of previous relationships. The eidôlon points as well to the instability of Elpenor’s ontological status; he is dead but not yet fully integrated into the realm of the dead. Elpenor’s eidôlon is also (p.150) significant in that it acts as a prologue to the ‘drama’ of the Nekuia, an episode that prompts Page (1955: 44–6) to propose that the figure of Elpenor was a device used to introduce the Nekuia to a poem in which it does not belong. Whether the Nekuia was originally part of the Odyssey or not, it is clear that the Elpenor episode sets a precedent wonderfully manipulated by Euripides in his Hecabe, a drama introduced by the eidôlon of Polydorus.

Significantly, it is the figure of Elpenor, like that of Tiresias, who captures the vase-painter’s imagination: it is Elpenor’s or Tiresias’ eidôlon, rather than the mass of flitting shadows of the indistinct dead, that is portrayed in visual representations of the Nekuia. Odysseus’ consultation with the ghost of Tiresias features on a Lucanian calyx-krater by the Dolon Painter of the early fourth century9 – the so-called ‘Tiresias Vase’. Odysseus is seated, sword in hand, and between his feet lies the head of the ram killed as a sacrifice. In the bottom left-hand comer, at the feet of Odysseus and the figure on the viewer’s left (Perimedes; Eurylochus may well be the figure on the viewer’s right), is the head of Tiresias’ ghost, looking up at them, rising from the depths of Hades.10 The dead seer Tiresias is here, reduced to a head, emerging from the decorative border defining the representational space and very much reminiscent of the oracular head of Orpheus.11 The strengthlessness of the dead is graphically illustrated on an Etruscan mirror in the Vatican (Beazley 1949: 69, pl. 5b) showing Odysseus and the ‘floppy’ eidôlon (Etr. hinthial, which is inscribed) of Tiresias, who not only supports himself on his walking stick but also hangs limply on the arm of Hermes – a slack posture that will presumably be rectified once the ghost has quaffed some blood. Hermes in this image seems to be introducing the eidôlon of Tiresias to Odysseus, who is seated, sword in hand. Both Odysseus and Hermes are unshod, unlike the eidôlon of Tiresias,12 whose cloaked body contrasts with the naked and athletic figures of Hermes and Odysseus.

This contrast between the clothed dead and the (almost) naked living is reversed in one well-known representation of Elpenor’s eidôlon. An Attic (p.151) red-figure pelike in Boston13 by the Lykaon Painter, c.440 BC, shows Odysseus, Hermes and the naked and rather athletic eidôlon of Elpenor, who emerges at knee-level from the reedy marshes, his two hands leaning on the banks of the Styx as if to support or steady himself. Elpenor’s ghost is one of the privileged few to be specifically referred to as eidôlon, in Odyssey xi, and yet Peifer confidently asserts that ‘Hier ist der Tote nicht als Eidolon dargestellt … Es ist kein Eidolon’ (1989: 131): this is one discrepancy between text and image, an inconsistency between iconographical distinctions and textual description. Apart from the suggestive familiarity with certain details of the Homeric text, the illustrators of Odysseus’ meeting with his dead comrade Elpenor, or his consultation of the dead seer Tiresias, are engaging in a creative process whereby private mental images are translated into public, concrete and recognisable visual portrayals. Drawing on both the individual and socio-mythic imagination, the vase-painter, like the tragedian, gives concrete shape and form to the poets’ words.

The beginning of the speech by the ghost of Anticlea draws on one of the most revealing aspects of Odysseus’ encounter with the dead: Anticlea asks her son how he came to be there, for it is ‘hard for those that live to behold these realms’ (xi. 156). Tiresias also asks why Odysseus has left the light of the sun and come to ‘behold the dead’ (xi.93–4), a question that casts Odysseus in the role of spectator and establishes at the outset a relationship between viewer and thing(s) viewed. Odysseus is unable to embrace his mother’s ghost (206–8), a literal empty gesture that signifies the recognition of its non-existence in the empirical world. Once again, as with Achilles’ attempt to embrace the ghost, the discrepancy between (visual) appearance and (tactile) reality is made explicit, prompting Odysseus to ask whether his mother’s ghost was some eidôlon (213) stimulated, or ‘roused’ (214), by Persephone to make him grieve all the more. Already in the sphere of the anonymous images of mortal men who have done with their earthly labours, Odysseus’ grief is heightened by the life-like appearance of his mother’s familiar ghost, which can be seen but not touched. In his Helen, Euripides was later to elaborate upon the problems relating to sensory perception, epitomised in the eidôlon of Helen, to create a thoroughgoing epistemological enquiry. Far from writing a play devoid of metaphysical depth (Dale 1967: xvi), Euripides employs the ruse of Helen’s eidôlon to address perplexing ontological issues.

Odysseus sees (xi.601) only Heracles’ eidôlon (602),14 for Heracles himself (Eidôla in Epic, Tragedy and Vase-Painting, 602) is in Olympus: like Iphthime and Aeneas, he, by (p.152) virtue of his phantom, can be in two places at once. As Heracles strides towards Odysseus, the other ghosts raise a clamour as of birds flying everywhere in terror (605–6), a striking example of the conspicuousness of the eidôlon. The phantom of Heracles looks like dark night and glares about him terribly like one always about to shoot (606–8).15 Heracles’ eidôlon is the only spectre who is physically described in terms of what he is wearing – a ‘terrible belt’ (609). The brief ekphrasis of the belt is suitably prefaced by an image of frozen action, as though the intention of Heracles’ eidôlon – to shoot – is perpetually suspended in time, a sharp contrast with his energetic life-time activities. Heracles’ belt serves as an iconographic curriculum vitae: it features bears, wild boars, lions with flashing eyes, conflicts, murders and Eidôla in Epic, Tragedy and Vase-Painting (‘slayings of men’, 611–12). The construct of Heracles’ eidôlon is articulated and accentuated by the belt upon which ‘wondrous deeds were fashioned’ (Eidôla in Epic, Tragedy and Vase-Painting Eidôla in Epic, Tragedy and Vase-Painting, 610). The link between art, representation, the rôle of the artist, artistic creation of any sort, and the manufacturing or presentation of an eidôlon is provocative.

The ekphrasis of the belt of Heracles’ eidôlon also acts as a meeting point for strategies of visual and verbal persuasion: Odysseus is not a passive but an ideal spectator who converts seeing into story-telling, by which the past becomes present through (re)presentation, evoking an immediacy that excites emotional involvement. Heracles’ eidôlon, differentiated from the other dead both linguisticially and iconographically, thus articulates the fundamental association between the eidôlon and representation, whether in narrative, dramatic or visual strategies. The very fact of telling a (ghost) story gives a kind of existence, presence, form and voice to what is ultimately an imaginary (mythical) being. And, like the consummate story-teller that Odysseus proves to be, he draws on the socio-mythic imagination to describe his encounter not only with the eidôlon of Heracles but also with the seer Tiresias, the noble (but notorious) women, Agamemnon, Achilles, Ajax, Minos, Orion, Tityus, Tantalus, Sisyphus and the Gorgon as well as the anonymous tribes of the dead. What the Nekuia as a whole admirably demonstrates is that the ghost is intimately connected to narrative strategies which place the ghost within a given context, space and time. The relationship between the living and the dead is, as Schmitt notes (1998: 185), ‘formed on the spatiotemporal line of the tale’.

Each ghost story is, in fact, two stories: the story of how the ghost came to be a ghost and the story of one’s encounter with the ghost (the Nekuia). Odysseus’ encounter with the eidôlon of Heracles highlights the way in which images (eidôla) generate more images (ekphraseis) and more (p.153) spectral narratives. The talkative ghost, of which there are many in Homeric epic, in stark contrast to the silent corpse, is a fundamental characteristic of many spectral episodes, and yet the sublime silence in the Nekuia of the ghost of Ajax, who carries his anger to the grave, conforms rather more readily to the characterisation of the dead as silent and speechless. The poets and the dramatists give voice and action to these dead mythical characters time and again: for example, the once reticent ghost of Ajax in the Odyssey is released from the underworld, revivified as a character and given one of the most poignant speeches in Greek tragedy in Sophocles’ Ajax. Far from being ‘guilty of too much explanation and descriptive detail in their speeches’ (Braginton 1933: 45), the ghosts, their narratives and the images conjured up as a result are inseparable, a factor greatly influencing the dramatic personae of Darius in Aeschylus’ Persians and Polydorus in Euripides’ Hecabe.


The Homeric portrayal of eidôla is crucial to our understanding of the stage-ghost in ancient Greek drama. The Homeric characterisation of eidôla would have worked on the collective imagination of the spectators of any given drama that either presented ghosts on stage or recounted sightings of ghosts. Elpenor’s ghost acts as a prologue to the narrative of Odysseus’ encounter with the dead, thereby setting a precedent for Polydorus, the prologue-speaker of Euripides’ Hecabe. In a grand finale to the Nekuia, the eidôlon of Heracles appears, a wonderfully apt conclusion to the spectacle of the dead which adumbrates Heracles’ dramatic appearance ex machina. Most significantly, Odysseus is cast throughout the Nekuia as a spectator, an eye-witness: he essentially presents an autopsy to the enthralled Phaeacian audience and it is, arguably, just such a context that furnishes a precedent for the reported sightings of ghosts in tragedy, such as that of the ghost of Achilles in Euripides’ Hecabe.

Like forensic orators, the dramatic poets practise eidôlopoiia, animating a known but dead person, one who has ceased from speaking. Forensic oratory, according to Aristotle (Rhet. 1357a36–b29), looks to the past and urges justice, and it has been recently argued (Hall 1995: 45–6) that one connection between tragedy and forensic rhetoric lies precisely in their mutual concern with the past and with justice. Characterisation, êthopoiia, was crucial to speech-writers and tragedians alike and, in the case of a dead relative of a client, the orator or the tragedian imagined what the person would have said if alive – eidôlopoiia, the imitation of a character of a dead person, created in order to animate the past. This connection between the forensic orator and the tragedian is perhaps most explicit in the figure of Clytemnestra’s eidôlon, who appears at the beginning of a (p.154) drama which treats of the establishment of the Areopagus. Aphthonius in his Progymnasmata 1116 links eidôlopoiia specifically with Eupolis’ Demoi, but the same could be argued for most, if not all, tragedies and their characters. One such is Aeschylus’ fragmentary Psychagogoi, whose title evokes another correspondence between rhetoric and drama, both of which were held to beguile (psuchagôgein) the minds of their audience. Much later, both eidôlopoiia and psuchagôgein are used in a derogatory sense by Plato (Rep. x.599d3, 600e5, 605c3), who accuses Homer and the dramatic poets of deliberate and deceptive ‘image-making’.

Extant tragedy furnishes us with three stage eidôla; fragmentary tragedies and the summaries of lost epics suggest that this type of literary and dramatic character was rather popular. In a recent article, Taplin (1997: 69–70) suggests that only two fifth-century vase-paintings can plausibly be claimed to show a play in performance, and both are early, from the era of Aeschylus.17 One of these is the Basle column-krater,18 an Attic vase dated to around 490 BC, which shows six youths dancing in unison before a bearded and shrouded figure who rises behind, or from, a structure which has been variously interpreted as a tomb, an altar or a monument. Indecipherable lettering, interpreted as the chorus’ song (Schmidt 1967: 71 with n. 4), issues from the open mouths of these ‘Basle Dancers’. It seems clear that these six youths are a masked chorus;19 as Taplin (1997: 70) notes, ‘Their identical hair, head-dresses and features are suggestive of masks, though there is no decisive indicator. And their military costumes, with some indications of ornate decoration, appear to be a signal of their mimetic role as soldiers (bare feet seem to be standard for choruses).’20 Taplin suggests that the structure they are dancing in front of (or around) seems to be a tomb rather than an altar and that the facing figure may be rising from the tomb rather than standing behind it.

(p.155) If this is so, then ‘we would have a ghost-raising scene, as, for example, in Aeschylus’ (lost) Psychopompoi, where Odysseus’ men summoned the dead prophet Tiresias’ (Taplin 1997: 70, n. 2).

The chief interest of the Basle column-krater has focused on choral formation, costume and choreography, but it is also an important piece of fifth-century Athenian evidence for the ghost-raising motif in tragedy,21 and the two facets of the vase can be harmoniously combined if one bears in mind the active role of the chorus in the necromancy scene of Aeschylus’ Persians. This chorus is ‘surely raising the ghost of a dead hero’ (Green 1991: 35). Whom does this ghost represent? Is it, as Taplin suggests (see above), the dead prophet Tiresias summoned by Odysseus’ men? The figure rising from the structure appears to be veiled – perhaps one of Aeschylus’ infamous veiled and muffled figures, derided in Aristophanes’ Frogs (911–13). The analogy between veiling and death is strong in literature and serves as iconographic short-hand for a dead figure (as in the lekythos discussed earlier, p. 149). Whatever the specific performance this vase-painting refers to, it does demonstrate the existence of the motif in ancient Greek drama before Aeschylus’ Persians of 472 BC.

Aeschylus’ Persians, a play which, according to Hickman (1938), would be ‘dull’ without its ghost, features an eidôlon who, under the rubric of ‘ghostly etiquette’, is a very gauche ghost appearing, as he does, in broad daylight to the chorus and the queen – not to mention the theatai, raised both dramatically and literally in the centre of the polis. There is a strange injunction in [Paulus] Sententiae (iii.4b2), admittedly a late source, against publicly appearing in the costume of a savage or a ghost: Darius is not only a barbarian ghost (ethnic and ontological states which mark his character as decidedly other), he is a kingly barbarian ghost – rather like the eidôlon of Aeetes mentioned at the beginning of this chapter, with whom Darius has certain affinities. During the evocation of the ghost of Darius, the chorus implore him to rise to the peak of his funeral mound, ‘revealing your yellow-dyed slippers on your feet and revealing the tip of your kingly tiara’ (660–2). This is often regarded as an indication of what the spectators are beginning to see as Darius rises from his tomb: a literal dramatic interpretation of these lines would have Darius rising feet-first, which is clearly not the case.

The ghost of Darius is asked to appear in the ‘distinguishing marks of royalty’: why, asks Gow (1928: 151), does Aeschylus mention only Darius’ hat and shoes? Gow suggests that the language about clothes is deliberately vague: Persian dress is dissimilar to Greek, but its essential (p.156) garment, the jacket or chitôn, differs in cut, colour and material, though not in name, from what Aeschylus and his audience wore themselves. To name it, in the description of Darius, would be to weaken what Aeschylus sought to strengthen – the illusion of a strange foreign fashion. So, Gow concludes, ‘betwixt head and heel, what else was there for him to mention? The trousers? But the Greeks thought them ridiculous (Eur. Cycl. 182, Ar. Wasps 1087); and, after all, in no age have trousers been a fitting theme for tragedy.’ But the costume and aspect of the Persian eidôlon of Darius clearly draws not only on the Homeric ‘as-in-life’ motif but also on the collective imagination. Darius is a regal ghost from foot to head, a being whose magnificence is measured – and recognised – by or at its extremities, and whose social status is confirmed by its attire.

Furthermore, the inability to describe anything ‘betwixt head and heel’ appears to be symptomatic of the play as a whole. The world of the Persians is a cosmos of extremes in which simple binary oppositions are spanned by artificial (and unstable) constructs. The physical collapse of the Persian empire is dramatically portrayed by the figure of Xerxes: what the empire once was is articulated by the impressive ghost of Darius, which also signifies the collapse of ontological and natural boundaries – men behaving as if they were gods, bridges erected where there should be none, and the dead appearing among the living.22 The corpses of the Persian youth are tossed to and fro by the sea on the shores of Salamis (272–7), exposed to the light of day just as that which should remain hidden – the dead King Darius – comes into the light of day (630). Is the raising of Darius’ ghost another defiance of natural laws, both spatial and temporal, which compounds the Persian disaster?

Smyth (1924: 81) describes the raising of Darius’ ghost as the ‘greatest ghost scene in all literature’, surpassed, if at all, only by the ghost of Clytemnestra in Aeschylus’ Eumenides. Clytemnestra’s eidôlon is portrayed in the process of waking the sleeping Erinyes on an Apulian red-figure bell-krater by the Eumenides Painter, c.380–360.23 Clytemnestra’s eidôlon is veiled, a clear indicator of her ontological status. An Erinys emerges at waist level from the decorative border of the vase, an indication, as Bérard notes (1974: 24), of vertical movement, and one which also stresses the chthonic origins of the Erinyes, who intrude into the world of the living and the representational space of the vase and/or theatre. Both Clytemnestra’s eidôlon and the Erinyes are placed at the fringes of the main focus of the image – Orestes, at the omphalos, is in the (p.157) centre of the image. Unlike the eidôlon of Patroclus, who makes a subtle plea against the ethics of revenge, Clytemnestra’s eidôlon in the Eumenides, without a shadow of doubt, is a precursor of the classic revenge ghost of the sort parodied in 1599 in an anonymous play24 by the character of Comedy: ‘Then of a filthy whining ghost / lapped in some foul sheet of leather pilch / comes screaming, like a pig half-sticked / and cries Vindicta! Revenge! Revenge!’. However, in this particular vase-painting, Clytemnestra’s eidôlon looks rather benign, and its inoffensive ghostly behaviour – in contrast to the eidolon’s portrayal in the Eumenides – seems to be characteristic of ancient Greek ghosts in general. It was left to the Hellenistic period, perhaps, and even more so to the Roman era, to amplify the horrific and frightening aspects of ghosts.

The eidôlon of Polydorus who delivers the prologue to Euripides’ Hecabe does not, as far as I am aware, feature in the vase-painters’ repertoire. Was he too much of a Euripidean rags-and-pathos figure to inspire the vase-painters? How did the ghost of Polydorus enter? Although a late source, the Bobbio scholiast on Cicero’s pro Sestio 126 (Hildebrandt 1971: 102) comments that the ghost of Deiphylus in Pacuvius’ Ilione appears creeping low upon the stage, dirty, and wearing mournful clothes as do those who are brought on stage as dead. The contrast between the (reported) resplendent ghost of Achilles and the dirty, ragged ghost of Polydorus is perhaps one which Euripides would not have missed. It certainly seems a preferable entrance to Flickinger’s (1939: 359–60) suggestion that a ‘wraith-like puppet was suspended from a pole above the scene building while an actor spoke the prologue from off-stage’. The visual impact of a phantom prologue-speaker must have been considerable, especially if it was unprecedented. It also complicates the audience’s response to the figure of Polydorus: is the primary focus of his identity that of a prologue-speaker or that of a ghost? Our response must be twofold: Polydorus is, literally, split into two distinct entities: his corpse lies on the sea shore, unwept and unburied (30), while his itinerant ghost (eidôlon) hovers near the temporary Achaean camp. What the theatai actually see on stage is Polydorus’ eidôlon, what they are invited to imagine, through Polydorus’ evocative prologue, is his murdered body tossing to and fro on the sea-shore.25

Polydorus’ eidôlon also reiterates far more explicitly than any of the Homeric eidôla that the existence of the eidôlon is dependent upon its being seen. To whom, if not to the audience, does the eidôlon of Polydorus appear? The absence of any other dramatic characters on stage at the time (p.158) of the eidôlon's prologue may help to explain the absence of this particular figure from vase-paintings, which, if they derive from tragic or other literary episodes, place the character(s) within a definite context. Polydorus’ eidôlon appears in a performative (and therefore representational?) vacuum and, were it not for the spectators who are vital to dramatic action, he would not exist at all. Polydorus’ first stage presence, as prologue (1–59), is complemented by his silent stage presence as a corpse for more than half the play (658–1295). By casting Polydorus’ ghost as the prologue-speaker, Euripides establishes, right at the beginning of the play, the crucial, and synchronous, components of a dramatic performance and experience – the spectator’s imaginative powers and the palpably real figures appearing on the stage.

The task of the dramatist – or vase-painter – was to convert the private mental images evoked by epic into public representations on stage, or in the visual arts. As Rohde (1925: ii. 225) notes, ‘what had hitherto seemed a dream-vision (Traumbild) of the imagination (Phantasie) now visibly presented itself to the eyes of the beholder … Thus rewakened to a palpable (greifbarer) and fully realized life (voll lebendiger Gestalt erweckt), the myth was seen in a new light.’ Text and image may not always cohere, but it is abundantly clear that the ‘dead have no existence other than that which the living imagine for them’ (Schmitt 1998: 1), and palpable and fully realised characters are, I propose, precisely what the word(s) eidôlon and eidôla meant to a fifth-century spectator of the dramatic and visual arts.


Bibliography references:

Beazley, J. D. (1949), ‘The world of the Etruscan mirror’, JHS, 69, pp. 1-17.

Beazley, J. D. (1955), ‘Some hydria fragments in Corinth’, Hesperia, 24, pp. 305-19.

Bérard, C. (1974), Anodoi: essai sur l'imagerie des passages chthoniens, Rome: Institut Suisse de Rome.

Bieber, M. (1961), The History of the Greek and Roman Theater (2nd edn), Princeton NJ: Princeton University Press.

Braginton, M. V. (1933), The Supernatural in Seneca’s Tragedies, Menasha WI: George Banta.

Bridjer, H. A. G. (ed.) (1984), Ancient Greek and Related Pottery, Amsterdam: Allard Pierson Museum.

Dale, A. M. (1967), Euripides: Helen, Oxford: Clarendon Press.

Easterling, P. (ed.) (1997), The Cambridge Companion to Greek Tragedy, Cambridge: Cambridge University Press.

Edwards, M. W. (1991), The Iliad: A Commentary, vol. V (Books 17-20), Cambridge: Cambridge University Press.

Fenik, B. (1968), Typical Battle Scenes in the Iliad, Stuttgart: Franz Steiner.

(p.159) Finucane, R. C. (1996), Ghosts: Appearances of the Dead and Cultural Transformation, New York: Prometheus Books.

Flickinger, R. C. (1939), ‘Offstage speech in Greek tragedy’, CJ, 34, pp. 355-60.

Garezou, M.-X. (1994), ‘Orpheus’, LIMC, VII, pp. 81-105.

Gasparri, C. (1986), ‘Dionysos’, LIMC, III, pp. 420-514.

Gow, A. (1928), ‘Notes on the Persae of Aeschylus’, JHS, 48, pp. 133-58.

Green, J. R. (1991), ‘On seeing and depicting the theatre in classical Athens’, GRBS, 32, pp. 15-50.

Green, J. R. (1994), Theatre in Ancient Greek Society, London and New York: Routledge.

Hall, E. (1995), ‘Lawcourt dramas: the power of performance in Greek forensic oratory’, BICS, 40, pp. 39-58.

Hammond, N. G. L. and Moon, W. G. (1978), ‘Illustrations of early tragedy at Athens’, AJA, 82, pp. 371-81.

Hickman, R. M. (1938), Ghostly Etiquette on the Classical Stage (Iowa Studies in Classical Philology 7), Cedar Rapids IA: Torch Press.

Hildebrandt, P. (ed.) (1971), Scholia in Ciceronis Orationes Bobiensia, Stutgardiae: Teubner.

Kahil, L. and Icard, N. (1984), ‘Artemis’, LIMC, II, pp. 618-753.

Kerrigan, J. (1996), Revenge Tragedy, Oxford: Clarendon Press.

Kirk, G. S. (1990), The Iliad: A Commentary, Vol. II (Books 5-8), Cambridge: Cambridge University Press.

Kossatz-Deissmann, A. (1981), ‘Achilleus’, LIMC, I, pp. 37-200.

Page, D. L. (1955), The Homeric Odyssey (The Mary Flexner Lectures), Oxford: Clarendon Press.

Peifer, E. (1989), Eidola und andere mit dem Sterben verbundene Flügelwesen in der attischen Vasenmalerei in spätarchaischer und klassicher Zeit, Frankfurt: Lang.

Prier, E. (1989), Thauma Idesthai: The Phenomenology of Sight and Appearance in Archaic Greek, Tallahassee: Florida State University Press.

Richardson, N. (1993), The Iliad: A Commentary, Vol. VI (Books 21-24), Cambridge: Cambridge University Press.

Riess, E. (1897), ‘Superstitions and popular beliefs in Greek comedy’, AJPh, 18, pp. 189-205.

Robertson, M. (1977), ‘Jumpers’, Burlington Magazine, 119, pp. 78-88.

Rohde, E. (1903), Psyche: Seelencult und Unsterblichkeitsglaube der Griechen, 2 vols, Tübingen: J. C. B. Mohr (Paul Siebeck).

Rohde, E. (1925), Psyche: The Cult of Souls and Belief in Immortality among the Greeks, London and New York: Routledge and Kegan Paul, Harcourt and Brace.

Roller, L. E. (1984), ‘The Greek view of Anatolia’, in Brijder 1984: 260-3.

Schefold. K. (1992), Gods and Heroes in Late Archaic Greek Art, Cambridge: Cambridge University Press.

Schmidt, M. (1967), ‘Dionysien’, AK, 10, pp. 70-80.

Schmidt, M. (1972), ‘Ein neues Zeugnis zum Mythos vom Orpheushaupt’, AK, 15, pp. 128-37.

(p.160) Schmitt, J.-C. (1998), Ghosts in the Middle Ages, Chicago: Chicago University Press.

Shapiro, H. A. (1994), Myth into Art, London and New York: Routledge.

Siebert, G. (ed.) (1981a), Méthodologie iconographique: actes du colloque de Strasbourg, 27-28 avril 1979, Strasbourg: AECR.

Siebert, G. (1981b), ‘Eidôla. Le problème de la figurabilité dans l'art grec’, in Siebert 1981a: 63-74.

Smyth, H. W. (1924), Aeschylean Tragedy, Berkeley CA: University of California Press.

Sourvinou-Inwood, C. (1995), ‘Reading’ Greek Death: To the End of the Classical Period, Oxford: Clarendon Press.

Stanford, W. (1940), ‘Ghosts and apparitions in Homer, Aeschylus and Shakespeare’, Hermathena, 56, pp. 84-92.

Taplin, O. (1997), ‘The pictorial record’, in Easterling 1997: 69-90.

Trendall, A. D. (1967), The Red-Figured Vases ofLucania, Campania, and Sicily, Oxford: Clarendon Press.

Trendall, A. D. (1989), Red-Figure Vases of South Italy and Sicily, London: Thames and Hudson.

Trendall, A. D. and Cambitoglou, A. (1978 and 1982), The Red-Figured Vases of Apulia, 2 vols, Oxford: Clarendon Press.

Trendall, A. D. and Webster, T. B. L. (1971), Illustrations of Greek Drama, London: Phaidon.

Vermeule, E. (1979), Aspects of Death in Early Greek Art and Poetry, Berkeley CA, Los Angeles, London: University of California Press.

Winkler, J. J. (1980), ‘Lollianos and the desperadoes’, JHS, 100, pp. 155-81.


(1) This advertisement was run by the Hellenic National Tourism Organisation for about three successive weeks in the Sunday Times Magazine. I would be grateful to any reader who could furnish details of the performance pictured. If the performance in question is a re-enactment of an ancient Greek drama, the scene from Aeschylus’ Eumenides, where Clytemnestra’s ghost rouses the Erinyes, presents itself as a likely candidate.

(2) All of the material discussed in this chapter is examined in greater detail in my doctoral thesis, which focuses on the stage-ghost, submitted in Michaelmas Term 1999, under the title ‘Casting shadows on the Greek stage’.

(3) Munich 3296 (J. 810): Bieber 1961: figs 121,122 and 289; Trendall and Webster 1971: 110, no. III.5.4; Trendall and Cambitoglou 1982: 18.283, pl. 195.

(4) Prier (1989: 102) is wrong when he states that, ‘Athena assumes an appearance of Penelope’s sister … Homer describes Athena as a “dim likeness” (eidôlon amauronOd. iv.824)’. Athena makes an eidôlon (Eidôla in Epic, Tragedy and Vase-Painting, iv.796) of Penelope’s sister, she does not assume a ‘dim likeness’ of the sister’s appearance – a crucial difference.

(5) See further Edwards 1991: on Il. xvii.669–73 and xix.300, with references.

(6) Delos 546: Beazley 1956: 378.257; Carpenter 1989: 100; Schefold 1992: fig. 313; Kossatz-Deissmann 1981: 588, pl. 117.

(7) Although Eidôla in Epic, Tragedy and Vase-Painting (to mix, to mingle) can denote social intercourse, it is properly used of liquids and thus evokes the mercurial nature of the psuchai and the eidôla kamontôn who refuse Patroclus this contact.

(8) Athens NM 1926: Beazley 1963: 846.193; Carpenter 1989: 297; Boardman 1989: fig. 255.

(9) Paris, Cab. Méd. 422: Trendall 1967: 102.532; 1989: fig. 79.

(10) This vase-painting, in particular the ghost of Tiresias, is often linked to a fragment from Crates’ Heroes (fr. 12 KA): Eidôla in Epic, Tragedy and Vase-Painting (‘turning his head towards them from the ground’). See further Riess 1897: 193.

(11) For example, the Attic red-figure hydria of the mid-fifth century in Basle Museum BS 481: Schmidt 1972: pls. 39–41.1; Vermeule 1979: 197, fig. 21; Garezou 1994: no. 18.

(12) Shoes were important to the dead in medieval society and ‘were supposed to help in their passage into the hereafter’ (Schmitt 1998: 204). Precisely what the signification of the shod Tiresias designates in this image is open to debate. Two ancient ghosts demonstrate a concern with clothing: Melissa’s ghost (Hdt v.92) and the dead wife who returns to claim a gilt sandal (Lucian, Philopseudes 27).

(13) Boston MFA 34.79: Beazley 1963: 1045.2; Carpenter 1989: 320; Boardman 1989: fig. 150.

(14) This passage (xi.602–4) is often regarded as an interpolation.

(15) For the colour of ghosts see the discussion in Winkler 1980.

(16) Rhet. Gr. i.101.12W; ii.44.28S.

(17) The other fifth-century vase-painting that (according to Taplin) can plausibly be claimed to show a play in performance are the five hydria fragments (Corinth T1144: Beazley 1963: 571.74; Carpenter 1989: 261; Taplin 1997: 71, fig. 5), c.460s, from Corinth, published by Beazley in 1955. The hydria fragments have attracted two main interpretations: Beazley suggested that they might be evidence for a ‘Croesus’ tragedy during the first quarter of the fifth century. Considerably later, Hammond and Moon 1978 proposed that the fragments showed Darius rising from his tomb, as in Aeschylus’ Persians. Since these fragments were published by Beazley, another fragment has been found, see Roller 1984: 262–3 with fig. 3, who interprets these fragments as depicting Croesus on his pyre.

(18) Basle BS 415: Schmidt 1967: pls. 19.1 and 21.1; Boardman 1975: fig. 333; CVA 3 (7) pl. 6.3; Gasparri 1986: no. 845, pl. 401; Taplin 1997: 70, fig. 4.

(19) Robertson 1977: 81, with n. 5, and Green 1991: 34–5; 1994: 17–18, with n. 5 on 177.

(20) Green 1991: 35 interprets the line on the left ankles of the nearer figures as indicating footwear: this would have been a strange item of footwear indeed, as all the chorus members’ toes are very clearly delineated.

(21) As Schmidt 1967: 74 notes, ‘Solche Beschworung der Toten war in der Tat in äschyleischer Zeit ein beliebtes Tragodienmotiv.’ See also Green 1991: 37 and 1994: 17–18.

(22) In Antigone (1067–71), Tiresias rebukes Creon for having hurled below one of those above (Antigone), blasphemously lodging a living person in a tomb and keeping here (above ground) something belonging to the gods below, the corpse of Polyneices.

(23) Paris, Louvre K 710: Bieber 1961: 27, fig. 96; Trendall and Cambitoglou 1978: 97.229; Kahil and Icard 1984: no. 1382, pl. 560; Shapiro 1994: 147, fig. 104.

(24) The author of this play is anonymous: the verses (47–50) come from the Induction to ‘A Warning for Fair Women’. See Stanford 1940: 91–2.

(25) So Hickman 1938: 53, who does not, however, develop the dramatic, epistemological or ontological implications.