I

The de septem orbis spectaculis attributed to Philo of Byzantium (though almost certainly a text of the late Empire) provides a telling introduction to my argument. In the prologue, he constructs an extraordinary defence of the power of ecphrasis over and against mere tourism. This is a patently self-serving claim to introduce his description of the Seven Wonders of the World – no need to travel with this book in your hand – but it shows vividly how forms of attention become a significant element of ecphrastic thinking:

This paragraph puts together an intricate mix of familiar expressions of the Greek intellectual tradition of the later Roman era. It places paideia ‘culture’, ‘education’ as its central category, as so many Greek authors from the Roman Empire do.Footnote ⁴ It elegantly applies the dynamics of educated paradoxology to reverse the usual value of an eye-witness account over the second-hand experience of reading (recalling the traveller's self-assertions from Herodotus onwards, with his celebrated calculation of which senses are the most persuasive).Footnote ⁵ It manipulates the philosophical, rhetorical and physiological discourse of phantasia with its vocabulary of enargeia, mirroring and imprinting in the soul: the turn towards the inwardness of the experience of wonder and paradox – ‘seen in the soul’ – is also typical of later Greek intellectual arguments.Footnote ⁶ But what I find most striking is the contrast between two forms of attention. The person who visits sites sees them only once, and passes by and into forgetfulness. ‘Precision’, to akribes, a standard criterion of art criticism, is missed, and the memory is inadequate for keeping an organised sense of the different parts of a seen work.Footnote ⁷ The tourist has a momentary experience, and the reliance on memory guarantees an insufficient grasp of detail or relationships of parts. But the reader of an ecphrasis (such as Philo's book) engages in ‘research in/through discourse’, logô historêsas, and experiences ‘the working out of vividness’ – that is, he undergoes the rhetorical power of the description to bring sites to life (enargeia). The reader not only can perceive the whole work as if in a mirror, but also retains (phulassei) ineradicable imprints in the soul. Textuality appropriates the language of sight – the written work is ‘like a mirror’, and its ‘images’ imprint the soul – and the book, unlike the gaze, promises a totalising experience: the whole site is captured by a text in a way that vision cannot grasp. The contrast is clear: the visitor's momentary gaze misses details and cannot provide the attention which leads to lasting impressions; the reader in his armchair [oikoi], however, has the attention of a researcher and thanks to the vividness of the rhetorician's ecphrastic mastery, receives lasting impressions. Narrative is better than seeing; it produces images that last in time; the traveller's flighty gaze contrasts with the reader's profound attention to detail and form. The ecphrasis, as opposed to the mere gaze, offers a temporal lastingness rather than fragmentary and fading perceptions.

Most ancient ecphrases, indeed, as Philo suggests, circulate as pieces of verse (or prose) separate from the object they describe, and which indeed often paradigmatically describe objects which do not have a separate existence in the world, such as Achilles' Shield or the Gates of Carthage. Even where the artwork is known to have existed – Myron's Cow, say – most scholars have assumed, correctly to my mind, that the epigrams which respond to it circulate as poems, and were read and appreciated without any necessary visual contact with the Cow itself.Footnote ⁸ Ancient ecphrases do not appear to have been used as guide-books with which to cruise the gallery, even when, as with Philostratus or Callistratus, they have precisely that form (though they never simply describe an art object). Counter examples of ecphrases actually attached to an artwork are few and often marginal – and have begun recently to be discussed by classicists at length. There is a unique bilingual stele, probably from the third century BCE, found in Attica with an inscription in Phoenician and in Greek, which offers a dramatised decoding of the strange and potentially baffling figures represented on the stele itself.Footnote ⁹ There is the tomb of Menophila from Sardis, which has the form of an epitaph that decodes the riddling signs on the tomb, a form found frequently in the Anthology.Footnote ¹⁰ Two Greco-Roman sites in Italy have artwork with ecphrastic verses attached – the cave at Sperlonga, where a plaque has been added as a response to the statues, and the so-called House of Propertius at Assisi where the first-century Roman wall-paintings have Greek poems added at a later stage which comment on and re-frame the images. Along with the House of the Epigrams at Pompeii, these examples have led modern scholars to explore how pictures and poems can frame each other in a complex dynamic of expressiveness.Footnote ¹¹

This dynamic established by texts attached to pictures becomes particularly interesting with the development of Christian art, with its particular relation to Scripture. Prudentius' little poems for a series of church paintings from the Bible are called either Tituli historiarum – ‘Titles/placards from histories/narratives’ – or the Dittochaion, which is usually translated as ‘Double nourishment’ from oche, understood thus as ‘Double testament’ (though I have sometimes wondered whether it couldn't be heard as ditto-chaion, ‘double shepherd's staff’, in the sense of a double shepherding of the viewing flock).Footnote ¹² These 49 quatrains, each written to be under a painting in a church, are marked, then, in the manuscript tradition as ‘titles’ or as ‘double’ – that is, as poems doubling the visual comprehension with a verbal, imaginative direction. The titulus creates a double perception, a dynamic between word and image. Each is (to be) affixed to a well-known biblical scene; but each also expands into narrative, and is often directive towards a Christian interpretive gaze. So – to take a simple but paradigmatic example – quatrain 14 on the grove of Elim in the wilderness reads as follows:

The picture of the arrival of the Israelites at Elim is to be seen as part of the story of the Exodus (devenere, ‘arrived’ implies the journey passed). But the twelve springs are glossed as a ‘mystic’ grove (that is, to be read within a wider theological narrative of transformation), and taken as a symbolic indication of the twelve apostles. Prudentius makes the grove the subject of the verb pinxit, ‘depicts’, ‘paints’, as if grove is both subject and object of representation. That is precisely the logic of the figural reading it demands: the twelve trees do indeed represent the twelve Apostles; the story of the Hebrew Bible is to find fulfilment in the Christian Bible: hence the addition of scriptis, by which we are encouraged to see the picture as part of a scriptural reading. The water which nourishes the palms slips easily thus into the symbolic expressivity of Jesus' encounter with the Samaritan woman at the well, and his promise of living water in John 4.10. The stay of the Israelites at Elim had already been invested with significance by Jewish writers: Philo (Moses 1.188), and Ezekiel's Exagoge describe it in lavish, Utopian terms; Josephus (AJ 3.9) imagines the feeding of the people from such sandy soil and brackish water as a miracle; the midrashim are happy to see the numbers of fountains as an indication of the tribes of Israel, and the water as the water of the Torah.Footnote ¹³ Prudentius redrafts such reading into a Christian vision. The double testament of the Dittochaion is to be found in this doubled figural vision: what is to be seen depends on a Christian temporality where the Christian order is always already anticipated, embodied and fulfilled – as discussed with stunning sophistication by Augustine in Confessions xi, where in God's language of creation everything is spoken simul ac sempiterne, ‘at the same time and eternally’, ‘in the simultaneity of eternity’ [Chadwick], though not everything which God causes to exist by speaking is made simul et sempiterna, ‘in simultaneity and eternity’ [Conf. xi 7. 9].

The importance to Christian theology and practice of this dynamic between pictures in churches and directive ecphrastic writing is picked up in my first key example, taken from the poems of Paulinus of Nola, an example which has surprisingly not yet entered the canon for scholars of ecphrasis. Paulinus' 27th poem contains a fascinating ecphrastic moment. Paulinus not only describes the paintings in a church, but also tells us that they have tituli, so that the words can expound what the hand has proclaimed, and so that the ignorant rural Christians, recently and not entirely deprived of pagan pleasures, can feast their eyes on art and gain through this sensory delight some scriptural soul-food. That is: first of all, we have explicit recognition of a specific Christian theology of ecphrasis, as a directive and moulding sermon on pictures which might otherwise involve mere pleasure;Footnote ¹⁴ but second, and most extraordinarily, we have an ecphrasis of an ecphrasis, a poem describing a poetic description attached to an artwork.

Paulinus began his lengthy ecphrasis of the church by offering it as a tour (veni, pater, et socio mihi iungere passu 360), a tour which will be an extended trip around every detail (dum te circumagens operum per singula duco). With classic ecphrastic gestures, he has repeatedly encouraged the viewer to see what he wants to show – ecce vides 362; ecce vides 387; aspice 480; volo…videas 511–12 – and encourages wonder and joy at the sights. But for a Christian, in the light of the biblical prohibition of graven images, such decorations can never be simply aligned with the history of Classical art and ecphrasis (for all the echoes of Virgil in Paulinus' verse).Footnote ¹⁵ It needs an explanation, as Paulinus declares (542–4):

The answer is that such images lead the rural crowd, who are neither strong in faith nor learned in their reading (neque docta legendi 548), to take pleasure in the surrounds of the church, and, as error can creep unnoticed into the simple soul (mentibus error irrepsit rudibus 564–5), so through reading the holy stories virtue too can creep up (subrepit honestas 593), brought on by examples of piety. The astonishment (adtonitas/stupefactis), and pleasure (grata oculis), which is typical of the ecphrastic moment, becomes ‘useful’ (utile Footnote ¹⁶ ) – an appropriation of the standard historiographical oppositionFootnote ¹⁷ – because by reading to one another (relegunt sibi) the stories of the pictures, love of food may become a love of fasting, and their gaping mouths can be filled with virtue rather than excessive wine (an image that also hints at the proper wine of the eucharist). These new spectacula are to be a display of Christian value (as Tertullian demanded with such aggressive rhetoric in de spectaculis). This ecphrasis is thus an ecphrasis of the paintings, the poems attached to them, and the worshippers reading the images and poems. In this way, the ecphrasis becomes an apologetics fully inscribed within a Christian theology of the role of didactic exemplars. The reader of the poem, led like the addressee around the Church of Felix, is encouraged to see the church and its decoration within a particular model of worship, and we are taught about teaching.Footnote ¹⁸

An extraordinary letter of Paulinus vividly depicts this exemplary power of artistic and verbal decoration in church – and, equally significantly, the active religious re-framing of pictures by attaching poems to them. This is further precious testimony of the saint's fascination with the work of art and text, but it is surprisingly a text that has been barely mentioned in the literature on ecphrasis. In a letter to Severus, perhaps from 403, Paulinus reacts strongly to the fact that Severus has had a portrait of him painted on the wall of the baptistery of the church at Primuliacum, facing a painting of Saint Martin.Footnote ¹⁹ Paulinus' objection is a test-case for the awkward mixture of humility and literary pride in his writing.Footnote ²⁰ Paulinus worries that he – his image – has no right to be there at all (serpentibus et columbis, hoc est nobis et Martino? ‘What have serpents to do with doves – a true comparison of myself with Martin?’ Ep. 32.2); and finds glory in it for Martin, and recognises that it was only Severus' deep love of Paulinus that prevents a charge of malice, since the contrast between them exposes him ‘to merited contempt once Martin's countenance is sighted…demonstrating the heinousness of this absurd comparison’, nos vero potius deformares, quos iure conspecta Martini facie despuendos ad probrum ridiculae comparationis exponeres. The only possible justification Paulinus can find for the display of the two paintings together is in their contrasting exemplary force of the saintly life of Martin and his own humble insufficiencies. Newly baptised Christians could thus ‘simultaneously see the exemplar to follow and the model to avoid’, vitandum et sequendum pariter conspicarentur. To make this point clear, Paulinus writes a poem to be inscribed on the wall and sends it to Severus:

Paulinus' poem constructs the two pictures as a strongly marked opposition (which is unlikely to have been the intention of Severus), and demands that the viewers not only see the pictures thus, but be drawn up as such, with the saintly and blessed instructed to stare at Martin, the sinners and guilty to look at Paulinus.Footnote ²¹ The ecphrastic poem, with the familiar injunction to look/see, re-frames the pictures – and asks us to see the difference between them, and, what's more, in an act of self-recognition to see oneself within the matrix of sin and blessedness.

But Paulinus then adds a second, longer poem as an alternative inscription. This too establishes the pictures in opposition to each other, but this time while Martin represents a man who ‘armed his faith with strong examples and words’ (ille fidem exemplis et dictis fortibus armat), Paulinus shows how you can ‘redeem sins by pouring away one's wealth’: iste docet fusis redimens sua crimina nummis. Paulinus here represents himself as redeemed by his charity – a model for wealthy proselytes to follow.Footnote ²² Donor portraits in churches were not uncommon; although there is no suggestion that Paulinus funded the church, he imagines his portrait within such a framework of expectation. The didactic force of the poem's descriptive rhetoric is equally marked, but now Paulinus allows himself a slightly less humble self-positioning. Now he is a model for the redeemed sinner, and for the benefactors of the church.Footnote ²³ Nonetheless, he concludes that the sole purpose of such a pair of paintings must be the contrast: ut in Martino forma iustitiae et summa virtutum, in nobis conscientiae iniquitatis obtritio et confessio cerneretur, ‘so that one can see in Martin the model of justice and the summation of virtues, and in me contrition for conscious sin and consciousness of it’. See justice, see sinfulness!

Paulinus' letter enacts his self-representation, as all letters do, but also is about his self-representation: in all humility – though the situation and his response put great pressure on such a term – he offers two different poems to change the way viewers might view his portrait on the wall of the church (another double testament, about the two pictures now viewed as a diptych). The letter, firmly set within the highly contested world of Christian (self-)representation, where self-scrutiny is such a charged term, sets out to perform an act of ecphrastic re-framing, a determined and determinative re-reading of his own portrait. This is a unique and fascinating case. Looking at one's self is a fundamental act of Christian askesis; here it has been reformulated within a long literary tradition as an analysis of a portrait of oneself, a painterly version of the gaze into the mirror of the soul, which doubles its own self-understanding with an increasing sense of his own paraded unworthiness.

Nor does it stop with these two poems. Paulinus notes that Severus has already put up a poem of his own for the portraits – a poem ‘of light and honeycomb’, luminis et fauorum – and suggests his verses should be just a supplement to it, like a setting for Severus' jewels (again the gesture of politesse does not wholly conceal the re-reading involved in the framing). And then this hugely long letter continues with a further collection of more than twenty poems to be added to various spots in the church, as the letter starts its ecphrastic trip round the church's site – the sort of poems he describes as attached to the paintings in his description of Felix's church in Poem 27, and that we saw with Prudentius' Dittochaion. The whole building is to be perceived through Paulinus' ecphrastic writing. Just as the opening suggestion of two possible epigrams to be attached to the two paintings demonstrates with stark clarity how adding a poem to a picture in a church is designed to direct the reader's comprehension in a thoroughgoing didactic manner, so covering the building with ecphrastic verse is designed to create an all-embracing understanding of what it is to move through this church, to worship in it, to inhabit its spaces.Footnote ²⁴ Ecphrasis as a humble display of power…the creation of a Christian visionary gaze.

Christian ecphrasis is firmly located within a new intellectual context of theological anxiety about the pleasures and seductions of the gaze (among other appetites). Ecphrastic aesthetics, that is, need to be placed within such a historically determined cultural context.Footnote ²⁵ Christian ecphrasis can echo Hellenistic language and the gestures of the Hellenistic viewer, but cannot be the same, because the act of viewing has developed a new moral and intellectual positioning. Paulinus demands a particular form of attention. On the one hand, he wants to cover the church with poems, to direct the Christian worshipper with a constant ecphrastic sermon, to create a pilgrimage of the gaze through the material world of the church towards a religious comprehension of the architecture, an almost liturgical passage through space and time. Each poem stops the pilgrim in his tracks and offers a narrative to create a religious experience of the moment, an experience framed within a Christian teleological temporality. On the other hand, with his own self-portrait – and within the paradoxical logic of competitive self-effacement – he requires us to stop and re-read the diptych of his own and Martin's portraits, first by re-framing the poem that is already there attached to the painting – this is an ecphrasis of an ecphrasis – in order to see his portrait as a negative figural reading of Martin's excellence; then by composing two further poems as alternative framings. The three poems create a horizon of expectation of what a Christian comprehension of these paired portraits could look like, a matrix of sin, humility, charity and baptism that finds little parallel in the Classical tradition of ecphrasis. But the three poems also construct three readings of the portraits, and in the gaps and differences between them a space of interpretation: a form of attention. Look, look and look again. Take time: at stake is eternal life. The picture must be viewed and re-viewed to produce a proper Christian reading: in a remarkable dynamic of self-representation, Paulinus wants us to look at himself very carefully, and performs that self-scrutiny directively and repeatedly for us. The Christian viewer is asked to look beyond the particularities to an eternal truth, to view figurally, a specific sort of doubled religious vision, for which the double portraits, read and re-read, framed and re-framed, become a model for this form of attention.

II

In the second part of my argument, I want to develop the notion of the forms of attention provoked by ecphrasis, to explore further the issue of temporality, raised especially by Christian figural reading, and how this might contribute both to an understanding of the history of ecphrasis and to its aesthetics. It has been observed that the ecphrastic epigram is an invention of Hellenistic culture.Footnote ²⁶ Although there are also some necessary qualifications to this bald statement – archaic inscriptions, for example, which require a passer-by to stop and observe a monument are evidently forerunners of the form, and the familiar tropes of epigram, such as the opposition of the material and the verbal, have long histories – nonetheless it is the case that the institution of the production and competitive circulation of epigrams describing a work of art seems to start in the Hellenistic era. It should be no surprise, then, that the ecphrastic epigram can be seen as self-consciously poised against epic's ecphrasis. The awareness of genre, and the dominance of the burden of the past over the Hellenistic imagination, make such a dynamic between privileged forms of poetry inevitable.

Now, obviously epic ecphrases tend to be rather long and epigrammatic ecphrases rather short. But the implications of this have not been followed through adequately, particularly with regard to forms of attention and the issue of temporality. First, it should be noted that, in Latin at least, this question of length was reflected on explicitly by Pliny. In Epistles 5.6, a centrepiece of current discussion of ecphrasis and identity in Roman culture,Footnote ²⁷ after his description of his villa, Pliny breaks off to reflect on his account (5.6.41):

The elegant irony of the letter writer's self-promoting mock hesitation – his ‘habit of “evasive display”’Footnote ²⁸ – and the work of self-presentation in the ecphrasis of the villa – ‘the autobiograph…as a “figure of style”’Footnote ²⁹ – has been brilliantly explored by John Henderson in particular. What is especially relevant about Pliny's rhetorical gesture to my argument here, however, is, first of all, his recognition that the potentially dismaying or tedious length of the description could be punctuated by taking a rest – where sitting down on a tour and putting down the letter are made parallel, in the same way as the journey round the house and the reading of the letter have been carefully intertwined throughout. That is, the reader's reaction to the length of the description is treated as a question of attention, where, in the world of the exchange of letters, there is always a tension between the pithy instantaneousness of the epistolary present and the leisurely otium of a gentleman's discourse. Looking after time – parsimonia temporis – is a theme of Pliny's self-fashioning.Footnote ³⁰ So too he places himself within the tradition of writing letters by forging a contrast between Cicero's abundance and his own restrictedness – of intelligence, material, rhetoric (9.23):

For Pliny, size matters. It is a question of matching rhetoric to subject and talent. Umbraticas (with ut ita dicam to make sure you read it carefully) implies letters sent out not just from the private world, a world of shadow to contrast with the bright lights of public business, where fame is hidden or darkened, but also from the shady retreats of an Epicurean garden or scholar's villa, where long discursive letters fit the lack of negotium. Pliny's self-fashioning enjoys displaying his expansively articulate ‘narrow confines’.

So the lengthy description of his villa in 5.6 needs more than an ironic hesitation to justify itself. Thus, Pliny continues with a lengthy paragraph which links his own ecphrasis to the ecphrases of the Shield of Achilles in the Iliad and the Shield of Aeneas in the Aeneid. It is worth quoting in full (5.6.42–4):

The question of length is insistently raised not least by the repetition of words of size (longum, longissimum, quot, quot, brevis, minutissima, magna – and, with a smart irony, parva magnis, as his own collocation with Homer, Virgil and Aratus is drawn into the same world of appropriate measurement). Description should be fitted to the scale of the subject and determined by relevance. Aratus can collect all the tiniest stars in his project of describing the vastness of the heavens, because astronomy is his grand subject. So, the argument's logic would have us believe, it is the grandeur of Pliny's villa that demands the length of the description, and, since such a description is his announced subject (titulus), it is a suitable procedure. Typically, Pliny exits from this dodgy argument (vel iudicium meum vel errorem) with an elegant and self-conscious gesture: verum illuc, unde coepi, ne secundum legem meam iure reprehendar, si longior fuero in hoc, in quod excessi, ‘But back to where I began, in case I am rightly condemned by my own principle, if I were to stay any longer in this digression’. By his own lights, he cannot delay any longer in the discussion of length.

The slow walk/description round the house of Pliny and the local countryside is a form of extended engagement, and the reader is invited to admire the altius ibi otium et pinguius, a ‘more profound and comfortable leisure there’, and to take pleasure in it. A slow and engaged gaze, with a focus on every detail (omnes angulos).Footnote ³¹ Such a long, detailed look is also the mark of Aeneas' response to the shield and other weapons his mother gives him.Footnote ³² Aeneas ‘could not get his fill and turned his eyes over each detail, and was amazed’, expleri nequit atque oculos per singula volvit,/ miraturque (Aen 8.618–19). So too he had wondered, in silent awe, transfixed by the image of the temple doors in Carthage (Aen 1.494–5): Haec dum Dardanio Aeneae miranda videntur,/ dum stupet obtutuque haeret defixus in uno…, ‘While these images seemed amazing to Dardanian Aeneas, while he was silent, and, transfixed, was lost in an unbroken gaze…’ – as he ‘viewed the details’, lustrat singula (453), and ‘fed his soul on the empty image with many a groan’, animum pictura pascit inani multa gemens (464–5). While he is gazing in this way, he is interrupted by the arrival of the queen (dum stupet…regina ad templum…incessit, ‘while he was silent…the queen…came into the temple’ [494–6]). In a similar way, when Aeneas stares at the Temple of Apollo in Aeneid 6, the spell of the engrossed gaze, and thus the length of the description, is broken by the arrival of Achates and the priestess (6.33–4):

Spectacula, the delights of the visual, demand in epic a long gaze and a long description – unless or until they are broken off. Time, the present moment, here demands instead that the action move on. One foundation of ancient epic narrative, we have come to see, is the tension between narrative teleology and narrative deferral – Achilles' refusal to fight, Odysseus' endlessly deferred ceasing of travel, Aeneas' unwillingness to leave Dido, Lucan's passionate desire to delay the arrival of the principate. One standard (if difficult) rhetorical understanding of ecphrasis depends on the opposition of description and narrative, where description is said to pause narrative, to hold still narrative's progress.Footnote ³³ With Aeneas standing transfixed before an image, only to be interrupted by his companions or the queen, the ecphrastic moment becomes synecdochic for the epic narrative's dynamic of delay and progress.Footnote ³⁴

Aeneas' emotional responses to the images he sees are marked. He cries, he rejoices, he is amazed. The Iliad provides a foundational model, as ever, for this. When Thetis brings the new weapons to Achilles, and they clash out loud (ἀνέβραχε 19.13), as she carries them, first the Myrmidons are terrified and cannot look (19.14–15):

The Myrmidons have to turn their eyes away from the shield's images in fear. But Achilles looks at it, and his eyes take on a burning flame of rage – as if the light of the gleaming shield has entered his soul – and he rejoices as he gazes at it (19.15–19):

The sight provokes different, powerful emotions from its different spectators. That Achilles can react with both anger and pleasure, and, at the same time, that the imagery of peace and war on the shield can provoke fear in the Myrmidons, have both played an important role in modern interpretations of the significance of Achilles' new armour in the epic.Footnote ³⁵ Strong emotions, then – but there is no explicit work of interpretation, commentary or understanding of the imagery within the poem. Although ‘wonder’ (θαῡμα) is repeatedly built into the language of the ecphrasis (miratur, miranda), the pattern of focalisation excludes the heroic characters of the epic (unlike in Virgil). Neither Achilles, nor any other hero, looks at the imagery and seeks for or finds any meaning there. In Homer, ecphrasis is not a scene of recognition.

Virgil, reading, as ever, through Homer, makes the eschewal of recognition a failure of recognition. Aeneas' response to his shield famously juxtaposes his pleasure in looking at the images to his ignorance of the images' import: miratur rerumque ignarus imagine gaudet, ‘he was amazed [by the shield], and, ignorant of the events, rejoiced in the image’ (imbricated with ‘ignorant, rejoiced in the image of events’). Rerum here may recall Aeneas' response to the gates at Carthage – sunt lacrimae rerum – a statement whose prophetic force for the story of Carthage and its queen, and his role in that story, Aeneas, the speaker, cannot (yet) understand. So, too, as Aeneas looks at the story of Daedalus on Temple of Apollo in Book 6, which acts as a multilayered prelude to his father's tale of future inheritance in the underworld, the image of the labyrinth has been taken as a model for the inextricabilis error (6.27) of (his) reading, much as the opening description of the shield as enarrabile textum (8.625), ‘an indescribable fabric’ not only plays with the rhetorical hyperbole of ‘beyond expression’, ‘even if I had a hundred mouths…’, but also suggests that there will always be something that goes beyond, that escapes the interpretation of the shield's imagery. The self-conscious tradition of epic ecphrasis may – should – encourage the reader to explore the relationship between the ecphrasis and the framing narrative, but the heroes themselves are debarred from any such knowledge (or from getting such knowledge right). Ecphrasis in epic is not a stage for the display or revelation of knowledge (let alone wit or interpretative flair). Indeed, critics over the centuries may have turned so obsessively to analyse the ecphrases in Homer, Apollonius or Virgil, precisely because this refusal or failure within the text to interpret the imagery provokes the reader to make good the silence or the misrecognition.

We are now in a position to see more precisely how Hellenistic ecphrastic epigram is set against epic. First, the epigram demands a different form of attention. The epigram paradigmatically does not imagine the viewer lost in awe, silent and transfixed. Rather, it insists on the interpretative moment, the flash of recognition. It does not depict the activity of looking, and thus description, as a potentially ever-expandable engagement: even when the epigram offers a series of details, as, for example, it does in the riddling funerary monuments, it is a discrete list where each element is precisely identified and interpreted. Where epic revels in the enarrabile textum, the viewer lost in thauma, unbroken contemplation (obtutu uno), epigram insists on viewing as pointed recognition, response as hyper-articulacy, and, above all, writing as staging the performance of knowingness. The two genres offer contrasting modes of attention. Secondly, where epic temporality sets in tension narrative progression and the delay of the ecphrastic process, in epigram the inherited trope of ‘stop, passer-by, and look’ becomes a bounded event, which is decontextualised; decontextualised, that is, not just in the sense that there is no framing narrative to the ecphrastic epigram, but rather in the sense that in epigram we find the depersonalisation of the viewing subject (never named in epigram), and the removal of any precise spatial or temporal frame of viewing. We do not find ecphrastic epigrams which begin ‘I was walking in the agora when…’, or even ‘the light at dawn falls on…’, language we do find in Pliny as well as in epic: to read the epigrams on Myron's Cow reveals nothing of where it is displayed or how any viewer came into contact with it.Footnote ³⁶ The ecphrases of Achilles' shield, Jason's cloak or Aeneas' shield (or the images on the doors at Carthage or the Temple of Apollo) are all deeply concerned with an ever broadening temporal framework: Achilles' death, beyond the end of the epic; Jason's afterlife in Euripides' Medea; Aeneas' famamque et fata nepotum (8.731), the history of Rome in the future. Each acts as a form of foretelling, a directive frame for things to come. An epigram very rarely indeed steps outside the dramatic present. The two genres offer different versions of how the understanding of an image relates to time.

But, thirdly, the difference between epic and epigram is also a question of a style of knowing. While epic describes images at length but restricts or removes the activity of interpretation from the voices of the poem, the epigram focuses precisely on the act of viewing and its cognitive delights, but restricts or removes actual description of the object itself (as we have seen in an exemplary way with the Myron's Cow epigrams). The viewer in – and writer of – epigram is a knowing subject, whose epigram is the display and performance of his or her knowingness. A hero of epigram cannot be someone who rerumque ignarus imagine gaudet. Any paraded doubt or ignorance in epigram is a foil to a climactic recognition. Indeed, even the emotions of epigram are almost invariably subordinate to the gestures of knowing. These different forms of knowing suggest different models of reader engagement: the openness of signification that we find in epic ecphrasis can be contrasted with the normative, the directive and the restricted expectation of epigrammatic ecphrasis. These contrasting forms of attention, contrasting forms of temporality, and contrasting forms of knowingness, are interrelated, then, and articulate contrasting discourses of epigrammatic and epic ecphrasis.

III

In the final section of my argument, I want to explore this sense of contrasting forms of ecphrastic attention in three further ways. First, I will take two paradigmatic examples of how epigrammatic knowingness plays with temporality, and its self-aware contrast of style with epic. Second, I will argue that a prose ecphrasis (of the same statue as one of the epigrams) not only offers a strikingly non-epigrammatic form of attention but also seems to re-frame an epic styling of knowledge. Third, we will see that the novel, with its characteristic polyphony and self-conscious wit, combines both epigrammatic and epic forms of attention with precise dramatic force.

My first epigram is an anonymous poem on a statue of Heracles defeating the Ceryneian hind (AP 16.96):

The narrator is depicted as awe struck (θαῡμα again), mentally and physically – in mind (φρεσί) and eye (ὄσσοις) – and rejoices in the verisimilitude and emotional power of the artwork. In the final couplet, Heracles' joy in victory is assimilated to the pleasure taken in art, where, it is suggested, the ‘all gold’ of art trumps the golden horns of myth. Heracles' violence is typically sexualised – here, as he presses on the deer's loin and grasps her beautiful horns, she gasps and pants – and the erotic language helps the slide between the hunt and the aesthetic pleasure. In this way, this poem is exemplary of the strategies of ecphrastic epigram.Footnote ³⁷ The first line, however, clearly echoes a famous line of Odysseus in Homer, as he begins to tell his long history to the Phaeacians (Od 9.14: τί πρῶτόν τοι ἔπειτα, τί δ' ὑστάτιον καταλέξω; ‘What first, then, what last shall I say?’). What is in Homer a choice between ‘first’ and ‘last’ has become in the epigram a choice between beginning, middle and end, ‘first, then, last’, where the Homeric connective epeita, ‘then’ has become in the epigram a full temporal marker of a third stage, a different use of ‘then’. This redrafting of Homeric language (no surprise in Hellenistic poeticsFootnote ³⁸ ) may suggest not only that a long, self-representing epic narrative is to come, but also that it will be even longer, or in more articulate form, even than the apologoi of Homer's Odysseus. In fact, however, the three time-scales introduce three couplets, one on Heracles, one on the deer, and one on the narrator's witty conclusion ([first] he grasps…[then] she gasps…[last] yo, art!). The Homeric echoing thus acts as a foil as much as an introduction for the epigrammatic briefness. The promise of epic length before the suitably epic subject of a Heraclean agon is a game of scale, pointing up the poem's pointed techne. The appropriation of epic language becomes a joke on epigram's temporality, its form of attention. We will not sit long in silence, as if at the end of a night's bardic performance, after this poem.

My second epigram is a poem which has been analysed with a good deal of sophistication and flair in recent years,Footnote ³⁹ APlan 275, Posidippus xix G–P, and which deals with temporality in a more complex way than the example I have just discussed:

The statue speaks, in answer to the passer-by, as in so many funerary epigrams, and the poem embodies the process of interpretation: ‘the pressing series of questions and answers [may]…be read as a direct enactment of the viewer's/reader's craving for interpretation while performing it’.Footnote ⁴⁰ The poem itemises each detail of viewing as a precise question with a precise answer, and ends with the didactic aim of the sculptor as a summary conclusion to the activity of reading: a bounded scene of recognition, of knowingness. To this degree, then, the poem fits straightforwardly into the generic expectations of Hellenistic ecphrastic epigram, as we have been discussing them. But as Kathryn Gutzwiller has sharply expressed it, the interpretation of the allegorical statue of Right Moment [Kairos], adds a further level to the reading process:

(As Augustine will discuss with the greatest profundity, there is an integral connection between the time it takes to say words and the comprehension of time.Footnote ⁴² ) Lucia Prauscello has intelligently noted also that prothurois in the final line has a broader metaphorical meaning than some more literal minded scholars have suggested. ‘In the portico’ may seem to offer a form of contextualisation for the statue (of a sort that I indicated above was extremely rare indeed in ecphrastic epigrams). While this physical meaning is, of course, a possible reading, which may have indicated a familiar location to some ancient readers, it has proved quite impossible for modern scholars to agree even on a city for the display of the statue, let alone a building. Prauscello is surely right to see a broader understanding of the term, indicating, as part of the teaching of the poem/statue, that time is to be understood here within the contemporary philosophical, and especially Epicurean, understanding of ‘time’ and ‘seizing the right moment’: understanding how the ‘right moment’ affects your life is ‘among the prolegomena (ἐν προθύροις) to leading a successful life’.Footnote ⁴³ Indeed, kairos itself has been read by Prioux as a political and, above all, an aesthetic term, which gives the poem a self-reflexive turn particularly suited to epigram: ‘“L'occasion” est un bel emblème pour définir l'air d'un epigrammatopoios’.Footnote ⁴⁴

There is one complementary twist I would want to add to Gutzwiller's, Prauscello's and Prioux's fine readings, which will tie the poem closely into the argument of this article. The poem starts with the front of the statue, and moves round to the back, as if the viewer is walking in the portico past the statue. Indeed, the specificity of en prothurois as a location, metaphorical or literal, – rather than a gallery or a temple – emphasises that this is a statue to be seen while passing through – as the dim echoes of the old trope of funerary statuary speaking to a passer-by may also suggest. Once you get to the back of Kairos, you have passed it by. The statue and its placement encourage everyone to miss the moment, and the poem takes you through the process. The lesson (last word) is not to miss the moment (passed by). The poem plays a lightly ironic game with its own sense of Kairos – with, that is, the temporality of the epigram as the bounded moment.

The sharpness of Posidippus' poem is worth contrasting with Callistratus' prose ecphrasis of the same statue (Callistratus 6), which is surprisingly not discussed in the modern critical literature on this poem.Footnote ⁴⁵ There is, first of all, in Callistratus' account a set of physical descriptions lacking in the epigram: the statue is in bronze; Kairos was represented as a beautiful boy at the height of perfection; the flesh had a particular bloom, and the forehead glistened with grace; the cheeks appeared to blush; he stood on a sphere; and so forth (6.1–2). Second, the governing trope of this prose ecphrasis is the familiar stunned reaction of the observer to the amazing realism of the art (6.3):

In contrast to the epigram sequence on Myron's Cow, Callistratus provides a rather prosaic wonder that this art ‘deceives the eye’, and that the brute bronze by artistic skill seemed to have sensation and colour. (Callistratus largely avoids the paradoxes and playfulness integral to the epigrams.) There is a mild eroticism in the image of the bronze melting and yielding to art, which is picked up later in the rather laboured suggestion that ‘beauty’ and the ‘right moment’ are integrally connected: ‘beauty is always a thing of the right moment, and the right moment is the only maker of beauty’. Perhaps most telling, however, is the silence of the spectators. They are awed by the sight of the statue, and indeed they have to turn to expert advice to understand it (6.4):

The narrator gets an exegete (ἐξηγούμενος) someone who knows, whose perception is more skilled and more trained in artistic matters, and it is this expert who goes on to explain the allegorical meaning of the statue, providing, indeed, a very similar reading to Posidippus. (This is a classic example of the standard scene of a visitor and exegete.)Footnote ⁴⁶ Callistratus can rejoice in the beauty and skill of the image, but remains ignarus rerum. The epigrammatist is the knowing subject, who parades the brilliance of his perception, who directs our gaze and understanding before an artwork. Callistratus, like a baffled and prosy Aeneas, needs to find someone else to provide this knowingness. Where Posidippus dramatises the statue itself answering the questions of its origin and meaning, as a display of his status among the sophoi and a mark of his own aesthesis technikotera, Callistratus stages his own failure to understand without help from another better trained sophos. The contrast could hardly be more pointed. It is from a later period and in prose rather than verse, but most strikingly it betokens a different form of attention, a different form of knowing.

My final case is from Achilles Tatius, ever the ‘scopophiliac's paradise’, and a wonderful scene which can be said to dramatise forms of attention as a theme of the novel.Footnote ⁴⁷ In Book V, the lovers, Leucippe and Clitophon, are preparing an expedition from Alexandria – in fact, being lured into an ambush by the bandit Chaereas. As they leave the house, a hawk chasing a swallow hits Leucippe on the head with its wing – an evil omen – at which point Clitophon turns and sees a picture in a gallery. The picture is of the rape of Philomela, which, notes Clitophon, is a ‘subject which also had a secret significance’ (5.3.4). This introduction leads into a lengthy ecphrasis of the painting. Menelaus, Clitophon's friend and confidante, follows the ecphrasis by expounding the point of the picture, with a generalising sententia typical of Achilles Tatius's novelFootnote ⁴⁸ (5.4.2):

And he concludes from the painting's depiction of erotic misdeeds that the expedition should be put off. Menelaus' language here, as Shadi Bartsch notes, is the technical language of art interpretation, familiar both from other prose of the time and from epigrams' appropriation of it.Footnote ⁴⁹ In no other place in Achilles Tatius (or indeed in the other extant Greek novels) do we get such an explicit interpretation of an ecphrastic image, complete with a buttress of the theory of interpretation. Indeed, Leucippe immediately asks (5.5.1): ‘what signifies the myth in the image? What are these birds? Who the women? Who that shameless man?’, and Clitophon answers by offering a long exposition of the myth of Philomela and Tereus, the back-story, as it were, to the ecphrasis of the painting. Scholars have explored at length the multiform ironies of how this ecphrasis and this myth relate to the framing narrative: as Bartsch puts it, ‘the author is not only inviting the reader to take an “inferential walk”, he is rolling out the red carpet!’.Footnote ⁵⁰

But what I want to emphasise is how Achilles Tatius' novel with its customary generic and discursive licentiousness seems here to combine the epic and epigrammatic ecphrastic styles, to good effect. Normally, following its epic pretensions, the novel presents its prophetic and over-determined ecphrastic descriptions without any directive commentary. Here, uniquely, it adds both an epigrammatic reading of the image (ἔρωτος παρανόμου, μοιχείας ἀναισχύντου, γυναικείων ἀτυχημάτων ‘illegal desire, shameless adultery, women's misfortunes’ – a zippy summary of a plot) and also a further, more extended reading of the myth in the image. Achilles' playfulness, however, is not merely to be seen in his generic promiscuousness. Clitophon and Leucippe, for all the work of interpretation, put off the expedition for one day only, and then sail straight into the ambush anyway. As so often in this narrative, the cleverness of the first-person narrator, a sophos about town, rebounds against him. Even as hero of his own story he cannot interpret his way out of the disaster to come. His smartness, his skill at offering a clever reading of the image, is all to no avail. The joke is on the sophisticated interpreters in the story – and, inevitably, on the readers of the story, as they too go on inferential walkabout.Footnote ⁵¹ Introducing the epigram's gestures of snappy, knowing interpretation into an extended, epicising narrative, does not reduce the semantic productiveness of the ecphrasis, but wilfully plays with it. He may think he is a sophos at large, but Clitophon remains resolutely and amusingly rerum ignarus. In Achilles Tatius, the sophisticated knowingness that the epigram encapsulates, becomes part of the novel's plot and humour.

This example in Achilles Tatius, I would suggest, should encourage us to look forward to the ecphrastic strategies of Heliodorus, or sideways at Lucian's construction of the civilised viewer, as further explorations of the relation between ecphrastic description and knowing interpretation. Greek prose in the Roman Empire, in short, repeatedly and amusedly engages with the extreme versions of the styles of interpretation and forms of attention on display in epic and epigram.

IV

Paulinus of Nola is well aware of the tradition of Classical ecphrasis, as his sophisticated reworkings of Virgil and other Latin poets indicate. His composition of multiple poems to adorn a church and to re-frame his own self-portrait require the focused, sublime form of attention similar to that we have seen in epic, but also demand a knowledge that is specifically Christian: the Christian is not ignarus rerum, however humble and insufficient he may wish to present himself to be, but has to know that his redeemer liveth, and has to be able to see a figure of eternal truth in the particularities of architecture, a portrait, or an image of the Hebrew Bible – and to view from within a specifically Christian matrix of moral values. The gaze of epic and the gaze of the Christian worshipper are both in sharp contrast with the form of attention embodied in the stance of the epigrammatic ecphrasis, where revelation is so often of the viewer's own rhetorical brilliance, a moment of wit or paradox. The ecphrastic viewer in epic, epigram and Christian writing each engages differently with time, and finds expression in different narrative voices. The novel, in its turn, plays with these different voices of ecphrasis and dramatises different forms of ecphrastic attention.

Historians of ecphrasis since Friedländer have often talked of ecphrasis as a recognised tradition of poetry and prose, a form of literary expression with tropes and expectations, even, indeed, following ancient theorists, as a rhetorical genre – with the result that examples of ecphrases are repeatedly taken from different genres and across different historical eras without due regard for the generic and cultural frames within which they have been composed. This article has demonstrated, I hope, that looking at forms of attention and modes of temporality allows us to develop a more nuanced comprehension of ecphrasis as a historical and aesthetic expression, and consequently to re-think this tradition as a far more varied and dynamic interaction than the critical horizon has so far allowed.