The Temple of Hadrian

This is a slightly modified version of an article that I published in 2009, entitled ‘Personifications and the Ancient Viewer: The Case of the Hadrianeum ‘Nations’’. The original piece, which was based on a chapter of my Courtauld PhD thesis, appeared in the journal Art History (vol. 32.1, pp. 1-20).



Some of the most exciting work on classical art in recent decades has focused on the figure of the viewer. Scholars working to uncover the meanings of ancient art have shifted their focus away from its makers and patrons, in a bid to reconstruct the experiences and responses of those who looked at it. This approach has shed new light on monuments as familiar as the Parthenon and the Ara Pacis, as well as drawing attention to previously neglected aspects of ancient material culture [1]. However, the viewer has remained conspicuously absent from the study of ‘allegorical’ art. This is surprising, given the fact that allegorical images by their very nature represent one thing in the form of something else, and therefore implicate their viewers in a particularly complex and self-conscious act of interpretation. This article offers an initial attempt at the reappraisal of allegorical figures from classical antiquity and places particular emphasis on the dynamics of their construction and reception.

The literal meaning of allegory, as Marina Warner has summarized it, is ‘‘‘other speech’’ (alia oratio), from allos, other, and agoreuein, to speak openly, to harangue in the agora; it signifies an open declamatory speech which contains another layer of meaning’ [2]. While the etymology of the term concerns the spoken word, much of the theoretical discussion of allegory has centred upon written texts [3]. Nevertheless, visual allegories were also widespread in antiquity, and were adopted with enthusiasm by artists and audiences in later periods. The most common type of allegorical image is the personification, which represents an abstract or non-human concept in the form of a human body. The inherent ambiguity of these figures (how do we differentiate the personification of Virtue, for instance, from a ‘real’ virtuous woman?) is mitigated by the provision of visual or textual clues: some personifications are identified by written labels; others are differentiated from the figures that surround them by artistic style, or size; others carry ‘unreal’ attributes, like globes, or cornucopiae.

The personifications discussed in this article have been interpreted as representing the provinces or nations of the Roman Empire. They are depicted as idealized females carrying an array of large weapons and wearing the varied costumes of their distant homelands (plates 1–8). I refer to these figures as ‘nations’, but it is important to stress at the outset that my use of this term is simply a convenient shorthand: the modern term ‘nation’ fails to reflect the real character and diversity of ethnic groupings within the Roman Empire, or the relationships of these groups to the territory they occupied. On the other hand, the nation personifications themselves work quite hard to disavow any sense of the real diversity of the Roman world. The setting of the formally similar human bodies within relief frames of identical shape and size imposes an unnatural uniformity onto the Empire, (mis)representing it as an easily quantifiable collection of discrete and equal units.

1 ‘Nation’ relief from the Hadrianeum showing figure in military costume, c. 145 CE. Marble, height: relief 2.13 m; figure 1.51 m. Rome: Capitoline Museums. Photo: K. Anger, D-DAI-Rom 1999.1438.  2 ‘Nation’ relief from the Hadrianeum showing figure with crossed arms, c. 145 CE. Marble, height: relief 2.13 m; figure 1.51 m. Rome: Capitoline Museums. Photo: K. Anger, D-DAI-Rom 1999.1437.  3 ‘Nation’ relief from the Hadrianeum showing figure holding curved sword and flowers, c. 145 CE. Marble, height: relief 1.68 m; figure 1.51 m. Rome: National Museum. Photo: K. Anger, D-DAI-Rom 1999.1444.  4 ‘Nation’ relief from the Hadrianeum showing figure holding pomegranate and flowers, c. 145 CE. Marble, height: relief 1.70 m; figure 1.51 m. Rome: National Museum. Photo: K. Anger, D-DAI-Rom, 1999.1445.

1 ‘Nation’ relief from the Hadrianeum showing figure in military costume, c. 145 CE. Marble, height: relief 2.13 m; figure 1.51 m. Rome: Capitoline Museums. Photo: K. Anger, D-DAI-Rom 1999.1438.

2 ‘Nation’ relief from the Hadrianeum showing figure with crossed arms, c. 145 CE. Marble, height: relief 2.13 m; figure 1.51 m. Rome: Capitoline Museums. Photo: K. Anger, D-DAI-Rom 1999.1437.

3 ‘Nation’ relief from the Hadrianeum showing figure holding curved sword and flowers, c. 145 CE. Marble, height: relief 1.68 m; figure 1.51 m. Rome: National Museum. Photo: K. Anger, D-DAI-Rom 1999.1444.

4 ‘Nation’ relief from the Hadrianeum showing figure holding pomegranate and flowers, c. 145 CE. Marble, height: relief 1.70 m; figure 1.51 m. Rome: National Museum. Photo: K. Anger, D-DAI-Rom, 1999.1445.

5 ‘Nation’ relief from the Hadrianeum showing figure with knotted costume, c. 145 CE. Marble, height: relief 2.14 m; figure 1.51 m. Rome: Capitoline Museums. Photo: K. Anger, D-DAI-Rom 1999.1434.  6 ‘Nation’ relief from the Hadrianeum showing figure with vexillum (standard), c. 145 CE. Marble, height: relief 2.17 m; figure 1.48 m. Rome: Capitoline Museums. Photo: K. Anger, D-DAI-Rom 1999.1433.  7 ‘Nation’ relief from the Hadrianeum showing figure with double axe, c. 145 CE. Marble, height: relief 2.16 m; figure 1.54 m. Rome: Capitoline Museums. Photo: K. Anger, D-DAI-Rom 1999.1435.  8 ‘Nation’ relief from the Hadrianeum showing figure with Phrygian-style cap, c. 145 CE. Marble, height: relief 2.16 m; figure 1.50 m. Rome: Capitoline Museums. Photo: K. Anger, D-DAI-Rom 1999.1436.

5 ‘Nation’ relief from the Hadrianeum showing figure with knotted costume, c. 145 CE. Marble, height: relief 2.14 m; figure 1.51 m. Rome: Capitoline Museums. Photo: K. Anger, D-DAI-Rom 1999.1434.

6 ‘Nation’ relief from the Hadrianeum showing figure with vexillum (standard), c. 145 CE. Marble, height: relief 2.17 m; figure 1.48 m. Rome: Capitoline Museums. Photo: K. Anger, D-DAI-Rom 1999.1433.

7 ‘Nation’ relief from the Hadrianeum showing figure with double axe, c. 145 CE. Marble, height: relief 2.16 m; figure 1.54 m. Rome: Capitoline Museums. Photo: K. Anger, D-DAI-Rom 1999.1435.

8 ‘Nation’ relief from the Hadrianeum showing figure with Phrygian-style cap, c. 145 CE. Marble, height: relief 2.16 m; figure 1.50 m. Rome: Capitoline Museums. Photo: K. Anger, D-DAI-Rom 1999.1436.

Which nations do these figures represent? As they were found without any identifying inscriptions, previous studies of the relief series have all attempted to identify the nations from an interrogation of their individual costumes and attributes. Any difficulties encountered in this task are perceived as simply marking our remoteness from the classical past, where viewers are thought to have interpreted these same images with fluency and ease. In this article, I argue that these nation personifications would have eluded the ancient viewer, too, regardless of whether or not they originally bore identifying labels. I base this argument on an analysis of the process by which the figures were made, which reveals that the unique, identifiable characteristics of the individual nation have been sacrificed in the construction of an holistic and balanced ‘group aesthetic’. This phenomenon is observable more widely amongst figural groups in classical and classicizing art. In the case of the Hadrianic nations, however, it has particularly profound ideological consequences, both for the nation whose identity was distorted by its inclusion in the group, and for the Roman viewer who contemplated these exotic, monumental figures.

The Hadrianic Nations in Context

This first section outlines the historical and archaeological context of the Hadrianic nations, and addresses some common misconceptions about the nature of the Empire that these sculptures represent. The Proconnesian marble reliefs were excavated between the sixteenth and twentieth centuries from the area around the Piazza di Pietra, which lies off the modern Via del Corso just south of the Antonine Column, within the boundary of the ancient Campus Martius [4]. Nineteen reliefs survive from what was potentially a much larger series [5]. Each relief originally measured over two metres in height, and each shows an almost life-sized figure standing on the lower moulding of the relief frame with her head receding into the upper moulding. Although the reliefs were found without any accompanying inscriptions, they were immediately interpreted as representing personifications of provinces or nations of the Roman Empire. This genre of image had a long ancestry in Roman imperial art, the most well-known and well-preserved precedent being the first-century CE series of figures from the Sebasteion at Aphrodisias, which were discovered in 1979. At the Sebasteion, inscribed bases identify the personifications as ethne of the Roman Empire collected during the reign of Augustus [6]. Other monumental series of nation personifications are mentioned in ancient texts, like the series of fourteen nationes commissioned by Pompey after his conquests in the East, and the elusive Augustan ‘porticus ad nationes’ [7].

9. The Hadrianeum building in the Piazza di Pietra, Rome. Photo J. Hughes

9. The Hadrianeum building in the Piazza di Pietra, Rome. Photo J. Hughes

Like their predecessors, the Piazza di Pietra figures are female – not the scruffy, empty-handed women who were supposed to represent ‘real’ barbarians in Roman art, but idealized icons of femininity, whose blank, classicizing faces, unwieldy weapons and masculine costumes (see, for example, plate 1) helped scholars to decide that they represented allegorical figures rather than realistic ‘ethnic portraits’ [8]. The figures originally decorated the attic of a portico struc- ture, which flanked (and may even have surrounded) the building whose remaining columns are visible today in the wall of the municipal borsa (plate 9) [9]. For a long time this building was thought to be the basilica of Neptune, a structure which the Historia Augusta tells us was restored by the Emperor Hadrian [10]. However, it was re-identified at the start of the twentieth century as the temple of the Deified Hadrian, or ‘Hadrianeum’, dedicated in 145 CE by that emperor’s adopted son and heir, Antoninus Pius [11].

Since the re-identification of the building there has been a strong desire to write it – and its personification reliefs – into modern stories about Hadrian’s foreign policy, and Antoninus’s filial piety. In Hadrian’s lifetime he made two long journeys around the Roman world, now, in Jocelyn Toynbee’s words, ‘the most striking, significant and familiar feature of his reign’ [12]. These journeys were memorialized by a series of coins minted between 134 and 138 CE, discussed at length in Toynbee’s study of The Hadrianic School in 1934 [13]. Twenty-five provinces or provincial cities appear on three different reverse types (the obverse always bears a portrait of Hadrian). The first reverse type shows a single personification, inscribed with the name of the province it represents (plate 10). As on the other types, this personification bears stereotypical attributes, chosen to evoke (and construct) provincial identity: Achaea bears an amphora, Aegyptos an Isiac sistrum; Arabia a camel, and so on. The second reverse type shows the personification sacrificing with the Emperor to commemorate his arrival in the province: the legend of this type reads Adventus plus the name of the province in the locative case (plate 11). A third reverse type has the legend Restitutori with the name of the province in the genitive case: here the Emperor grips the hand of the kneeling province, as if to raise her up (plate 12). Studies of the Hadrianic province coins unanimously agree that they embody a philosophy of unity and respect towards the empire they portray. Toynbee, whose book remains the standard study on the coins, writes that:

They reflect the great idea for which Hadrian stood, the idea of the Empire as a vast unity, a brotherhood of fellow-citizens of the world living together on an equality in prosperity and peace, under the aegis of a beneficial central government, to which the well-being of each member was a matter of vital interest [14].

This perception of the coins as commemorating the shared benefits of empire was, in turn, mapped by scholars onto the province reliefs from the Hadrianeum [15]. Toynbee, for instance, stated that:

The underlying idea and spirit of the Hadrianeum Provinces and of the ‘province’ coin-series are essentially the same. In the reliefs, the Provinces appear, not as unwilling, conquered subjects, but as peaceful, contented, prosperous units of the Roman world. They show no signs of ‘resignation’, they adopt no ‘attitude of grief’ [16].

10 Hadrianic ‘Province’ coin, minted c. 134–138 CE. Silver denarius, diameter 17 mm. Private collection. Photo: courtesy of the Classical Numismatic Group, Inc.,  11 Hadrianic ‘Adventus’ coin, minted c. 134–138 CE. Silver denarius, diameter 17 mm. Private collection. Photo: courtesy of the Classical Numismatic Group, Inc.,  12 Hadrianic ‘Restitutor’ coin, minted c. 134–138 CE. Bronze sestertius, diameter 31 mm. Private collection. Photo: courtesy of the Classical Numismatic Group, Inc.,

10 Hadrianic ‘Province’ coin, minted c. 134–138 CE. Silver denarius, diameter 17 mm. Private collection. Photo: courtesy of the Classical Numismatic Group, Inc.,

11 Hadrianic ‘Adventus’ coin, minted c. 134–138 CE. Silver denarius, diameter 17 mm. Private collection. Photo: courtesy of the Classical Numismatic Group, Inc.,

12 Hadrianic ‘Restitutor’ coin, minted c. 134–138 CE. Bronze sestertius, diameter 31 mm. Private collection. Photo: courtesy of the Classical Numismatic Group, Inc.,

Other historians gave similar readings. Eugenie Strong contended that the Hadrianeum personifications were made from an entirely different mould to the more aggressive ethnographic ‘portraits’ of earlier periods:

The conception, it has been well pointed out, is not so much of the conquered country; it is no longer the Germania, the Gallia or the Judaea capta, who sit desolate on the reverse of so many Imperial coins, as of the friendly allied province, tenderly regretful, perhaps, of past inde- pendence, yet proud to be raised to equality with Rome [17].

Roger Hinks agreed, commenting that none of all the figures ‘represents a province as capta or devicta’, and noting that the figures seem to represent outlying provinces of the empire, the clearest beneficiaries of ‘Hadrian’s foreign policy of pacification and unification’ [18].

13. ‘Achaea’ and ‘Trophy’ reliefs from the Hadrianeum, c. 145 CE. Rome: Capitoline Museums. Photo J. Hughes.

13. ‘Achaea’ and ‘Trophy’ reliefs from the Hadrianeum, c. 145 CE. Rome: Capitoline Museums. Photo J. Hughes.

In the decades after their discovery, then, the province personifications were incorporated into broader eulogies of the Roman Empire at the height of its power. But a closer look at the iconography of the personifications indicates that there is far more at stake than these accounts suggest. In the first place, several of the Hadrianeum nations do, in fact, bear strong iconographic similarities to those captured barbarians of earlier art. The personification in plate 2, for instance, crosses her arms over her stomach: such a gesture was frequently used in triumphal representations from earlier periods of imperial Art, as in the sculpted barbarians which decorated the Forum of Trajan (some of which are now visible in their re-used context on the Arch of Constantine) [19]. Secondly, the nation reliefs were discovered along with a number of ‘trophy’ reliefs, representing conquered enemy armour (plate 13) [20]. The trophies originally alternated with the nation personifications at the top of the portico, facing the temple [21]. Their place on the architecture of the building would have reflected the old Roman practice of fixing enemy spoils to porticoes and temple buildings [22]. Meanwhile, the placing of the personifications between the trophy reliefs would have suggested the nations’ own status as conquered boot [23].

The trophy reliefs were so far incompatible with their expectations of Hadrianic imperialism that some scholars excluded them entirely from their reconstructions of the building. Toynbee argued that ‘the Trophies themselves are scarcely in harmony with the spirit of the Province reliefs, which do not commemorate the triumphs of Roman arms, [and] it is perfectly possible that the Provinces and Trophies did not belong together’ [24]. Toynbee’s optimistic vision appears to have been shared by the post-antique restorers of the figures. For while the restorations were limited for the most part to repairing damage to the outer frames and missing corners, two of the figures had lost hands, which were replaced with new attributes. The personification shown in plate 3 was given a new right hand holding a sheaf of corn and flowers; while both the hands of the personification shown in plate 4 were also restored with their attributes – a pomegranate in the right, and flowers and corn (again) in the left. These restorations are fundamentally misleading, for while the personifications in the province coin series do indeed carry flora and fauna, none of the other Hadrianeum figures are shown with agricultural produce; rather, they carry weapons. The Hadrianeum personifications were not envisioned as peaceful representatives of distant lands bringing the varied resources of empire into its centre, but rather as warriors. And while for some they may have called to mind provincial auxiliaries serving in the Roman army, epiphanized here as obliging guardians of the imperial temple, the potential for them to revolt would have been perennially present [25].

Both scholars and restorers of the reliefs, then, have tried to iron out differences between the Hadrianeum reliefs and the Hadrianic and Antonine coin series, partly because of their expectation, or desire, to see unity within the corpus of images commissioned by a single imperial dynasty. However, as I have underlined here, there are some marked differences between the vision of empire presented by the coins and those presented by the Hadrianeum reliefs. One difference perhaps outweighs all the others in its importance. While each of the provinces represented on the coin reverses has her name written in large letters, curling around her body, those on the reliefs were found without any inscriptions. No inscribed bases like those accompanying the ethne at Aphrodisias have been found at the Hadrianeum; neither is there any conceivable space for such blocks in the architectural structure of the portico. Neils Hannestad has suggested that ‘When [the personifications] were set up, the names were probably painted on the flat relief background, corresponding to the coin inscriptions’ [26]. But there would be no obvious place for these inscriptions to fit on the background of the reliefs. Moreover, according to the current restoration, the attic of the portico stood fifty Roman feet (14.8 m) above the ground [27]. At this distance, any inscriptions that might have been painted onto the reliefs would scarcely have been legible, even to literate viewers.

Interpreting Personifications

Reading an inscription is not, however, the only way to get information out of an allegorical figure. Personifications in ancient art and literature wore and carried distinctive costumes and attributes, which, in theory, bore a meaningful relationship to the concept they represented. For those allegorical figures which also bore identifying written labels, these costumes and attributes served to confirm existing perceptions or create new stereotypes about the personified entity. For those figures without identifying inscriptions, the costumes and attributes ostensibly provided the means to recognize the personified concept, this time by drawing on preconceived ideas about what it ought to look like. Literary texts from classical antiquity preserve numerous accounts of viewers interacting with personifications in this way. One of the best-known examples is the epigram describing an anonymous viewer’s vision of Lysippos’s Kairos in the Anthologia Graeca (275):

A. Why do you have a pair of wings on your feet?

B. I fly like the wind.

A. And why do you carry a razor in your right hand.

B. As a sign to men that my appearance is more abrupt than any blade [is sharp].

A. And your hair, why does it hang down over your face?

B. So that he who encounters me may grab it.

A. By Zeus, and why is the back of your head bald?

B. Because nobody, once I have run past him with my winged feet, can ever catch me from behind, even though he yearns to [28].

The question-and-answer form of the epigram provides the framework for a normative representation of how to read allegorical imagery, whereby each symbolic attribute is to be interrogated, unfolded and inscribed into a meaningful relationship with the abstract concept represented. In this case, the personification has already revealed its own identity as ‘Kairos’ (‘Time’ or ‘Opportunity’), but in other texts, the reader is left to identify an unnamed allegorical figure from a list of their physical and behavioural characteristics. For example, in Xenophon’s account of the Prodikean Choice of Herakles, that hero is confronted by two personifications at a crossroads. Only ‘Kakia’ (‘Vice’) identifies herself by name, while her companion remains anonymous, simply being described by the author as unadorned, sober and modest [29]. Recognizing her as ‘Arete’ (‘Virtue’) is not difficult, especially as she is represented as the polar opposite of her companion [30]. But the fact that the reader is left to name her independently is significant, because it suggests that the mechanics of this sort of allegorical interpretation had been internalized by people in antiquity. It also bears witness to the tacit agreement between author and readers as to what (female) ‘Virtue’ actually is.

Scholarship on the Hadrianeum figures is based on this same idea, that the correct way to interact with a personified figure is by unravelling the meaning of its symbolic attributes. The conventional approach to the reliefs involves taking each figure individually, listing its attributes, and attempting to find parallels for them in ancient material and literary culture, with the ultimate goal of identifying the figure that holds or wears them. Modern scholars, then, approach the Hadrianeum personifications in much the same way that many of the ancient viewers would have done. However, while the ancient viewer looking up at the reliefs on the top of the portico would have had to fall back on their existing knowledge about the ethnography of the Empire, the modern scholar is able to call on a whole range of literary and visual sources – often obscure, and rarely contemporary–in their priviliged, eye-level scrutiny of the reliefs [31]. Given all the advantages of this retrospective viewpoint, it seems surprising and potentially uncomfortable that the identities of the figures are all heavily contested. The personification in plate 5, for instance, was identified by Hans Lucas as Egypt, on account of her semi-nudity and her knotted costume, which is similar to that worn by priests of Isis; Toynbee later re-identified the figure more precisely as the city of Alexandria, after comparisons with coins; then Erika Simon offered a new identification of the figure as Achaea, firstly on account of her headband, which evokes the iconography of statues of Olympian athletes, and secondly because of the fragmentary attribute which has been identified tentatively as an anchor, referring, perhaps, to Athenian control of the sea [32].

For these scholars, our failure to decipher the iconography of the personifications is simply evidence of our own remoteness from the ancient world, whose inhabitants interpreted the same images unproblematically [33]. In his study in 1939 of Myth and Allegory in Ancient Art, Hinks complained of the Hadrianeum figures, ‘It is true that they have not all been successfully identified, but this difficulty is due rather to our ignorance than to any lack of care on the artist’s part’ [34]. However, the notion that personifications could be problematic for the ancient viewer finds strong support in ancient literary texts. In the ‘Kairos’ epigram cited above, the attributes are ambiguous and esoteric, and the viewer has to ask the statue itself for help in understanding their significance. Another particularly vivid example is found in the first book of the Ars Amatoria, where Ovid advises his reader/protégé on how to impress a girl at a triumphal procession:

And when some girl among them asks the names of the monarchs, or what places, what mountains, what rivers are borne along, do you answer everything, nor only if she ask you; ay, even if you know not, tell her as if you knew it well. That is Euphrates, his forehead fringed with reeds; he with the dark blue locks hanging down must be Tigris. These, say, are Armenians, here is Persia, sprung from Danae; that was a city in the Achaemenian valleys. That one, or that, are chieftains; and you will have names to give them, correct, if you can, but if not, yet names that are fitting [35].

The personifications mentioned here are closely related to the Hadrianeum personifications, representing units of imperial geography in human form, this time in an explicitly triumphal context. The two rivers, the Euphrates and the Tigris, are represented in the form of personifications well known from mosaics and the plastic arts. ‘Persia’ and the un-named Achaemenid city may also have been personified in images like the Hadrianeum reliefs, and the reference to Persia’s ‘birth’ from Danae might be particularly evocative if she were shown in anthropomorphic form [36]. However, the interest of this passage does not lie in its recording of possible iconographic antecedents for the Hadrianeum personifications. Rather, it offers proof that such images could be problematic for the ancient viewer attempting to identify them: the boy, Ovid admits, probably has no idea who or what these personifications are supposed to represent – not that this really matters, as the girl to whom he is talking certainly will not know any better. Moreover, the passage also indicates the extent to which the interpretation of images functioned in antiquity to configure networks of social relations. Identifying the personifications correctly meant that the viewer had the cultural resources to recognize their attributes, and the mental agility to put them together to tell a coherent story. That such displays of cultural capital could contribute directly to the construction of social status is implied in the boy’s (anticipated) conquest of his girlfriend, and the simultaneous demonstration of Ovid’s own intellectual superiority over the pair of woefully unsophisticated adolescents.

These texts confirm that the interpretation of personifications was sometimes problematic for the ancient viewer. However, simply attributing the ancient viewers with ignorance is as unsatisfactory and as methodologically unsound as attributing them with positive knowledge of the same facts. The kind of top-down analysis which sees ‘meaning’ as ‘meaning at the original moment of conception’ (rather than at the different points of an image’s reception) has always been the dominant approach in art history for a very good reason: any other approach generally depends on speculation. In fact, in the introduction to his recent book on Art in the Lives of Ordinary Romans, John Clarke states ‘My method is purposely speculative. At the end of many of the case studies I construct scenarios – ‘what if’ viewing scripts’ [37]. This approach is admirable in as much as it draws attention to the plurality of an image’s meanings, and alternative viewers other than the upper-class male, but once this aim is achieved, it is inadequate as a way of telling us what images ‘really meant’ to their ancient audiences. However, I would argue that the Hadrianeum figures give us a valuable opportunity to prove, rather than simply to assume, that certain images were problematic for the ancient viewer. This opportunity lies in re-inserting the individual personifications into their original context of the group.

The ‘Group Aesthetic’

The modification of individual figures within a group to complement and interact with one another was a formal development linked to the increasing ‘naturalism’ of classical Greek art. In Geometric and Archaic art, groups were normally constructed through the identical multiplication of individual figures, as in the famous scene of mourners around a bier portrayed on the belly of the Dipylon amphora [38]. In such images, the conceptual similarity between figures was reflected by the physical similarity of their bodies and gestures: the definitive characteristic of the individuals portrayed was that they were (all) mourners or tribute-bearers, and their lack of visual differentiation only served to emphasize their social connection. However, in the Classical period, groups of figures began to embody the sense of balance between difference and similarity that has been a basic element in the construction of figural groups in Western art ever since. The group of bronze ‘dancers’ from the Villa of the Papyri at Herculaneum (plate 14) provides one famous, large-scale example [39]. The dancers still draw our attention to the fact that they belong to a group, being of equal size, made from the same material and dressed in the same kind of columnar garment. In the National Archaeological Museum in Naples the sculptures are also placed at equal distance from one another, as they may have been in their original display context [40]. At the same time, any sense of repetition or hieratic stillness is alleviated, because the artist made each dancer hold her arms in a position that complemented and contrasted with the poses of the rest of her company. Even this decorative group of figures without attributes shows us how taking any one of these figures singly, in isolation, would leave our understanding of that figure somewhat impoverished.

14 Group of ‘dancers’ or ‘peplophoroi’ from the Villa of the Papyri, Herculaneum, first century BCE – first century CE. Bronze, height 1.50 m. Naples: National Museum. Photo: J. Hughes.

14 Group of ‘dancers’ or ‘peplophoroi’ from the Villa of the Papyri, Herculaneum, first century BCE – first century CE. Bronze, height 1.50 m. Naples: National Museum. Photo: J. Hughes.

This subtle modification of a figure to respond to its neighbours applies not only to poses and gestures, but also to the costumes and attributes that it wears or carries. Images of the four Seasons provide us with a particularly cogent example. In Egypt and early Classical Greece, the three Horai were nature goddesses and were visually identical; in Hellenistic and Roman times their number rose to four, and they were each given different attributes, to symbolize the four seasons of the year [41]. In these later images, it is clearly membership of the group that defines both the iconography of the individual personification, and the conceptual boundaries of the abstraction it represents. For example, in many representations, Winter is portrayed as a markedly older figure than the other Seasons; often she looks ‘back’ at the other figures, signalling, with that gesture, that she marks the end of the year. In other representations, the change in the meteorological climate is also registered in the changing of costumes, from Spring’s light tunic to Winter’s heavy mantle.

15. Details from Seasons Mosaic from the imperial palace at Ostia, showing (left to right, top to bottom) Spring, Summer, Autumn, Winter, mid-second century CE. Rome: San Paolo alle Tre Fontane.

15. Details from Seasons Mosaic from the imperial palace at Ostia, showing (left to right, top to bottom) Spring, Summer, Autumn, Winter, mid-second century CE. Rome: San Paolo alle Tre Fontane.

Significantly, in the majority of cases each Season holds or wears attributes deemed suitable for showing both continuity and contrast amongst all four figures – generally fruit and/or foliage. And while each attribute makes logical sense in its relation to its wearer (flowers bloom in Spring; wheat grows in Summer; and so forth), the connection between attribute and personified concept would not necessarily be the obvious choice, were each Season considered in isolation. For example, the figure of Winter in a Season mosaic from Ostia (plate 15) is represented wearing a ‘cloak’ of aquatic foliage.42 This attribute successfully illustrates Winter’s simultaneous continuity with, and difference from, the other Seasons, who were also shown wearing foliage: the aquatic foliage was clearly chosen for this purpose rather than because it was itself intrinsically ‘wintery’, and there is a sense in which the amount we can learn from our personification about the character of Winter as a concept is actually very limited. This subordination of the individual element to the group frequently problematizes the identification of figures when they are removed from their original context. For instance, the British Museum has in its collections a statue dating from the second century CE of a heavily draped reclining female holding a hare in a basket, whose identity has fluctuated between the personification of Winter and the personification of the province of Hispania [43].

If we now turn back to the Hadrianeum personifications, it is clear that here, too, the iconography of the individual personification has been modified to complement and interact with the other members of the group. This is immediately visible at the level of the pose. Nine of the figures turn their heads varying degrees to the left, while six turn varying degrees to the right. In 1979, when the nations were still thought to belong to the interior of the Hadrianeum temple, Anna-Maria Pais took these variously oriented gazes to suggest that the nations on the walls of the cella originally all turned to look towards the statue of the Emperor [44]. In their new place at the top of the external portico, the figures could still have been oriented to face the temple, or, equally, they might have been arranged according to alternating left- and right-hand gazes. Whichever of these reconstructions is correct, most viewers would agree that the head-turn of the figures serves as a formal device to relax the potentially stiff frontality of the long row of figures, which bears little or no meaningful relationship to the character of the individual nation. The same would go for the position of the figures’ hands, which are variously shown as outstretched, crossed, touching a shoulder, or fiddling with a cloak. Gestures and body language sometimes do impinge on the identity of the personification – the case of Virtue and Vice at the crossroads is a good example. However, in this long series of nations, as with the Herculaneum dancers, most viewers would simply view the bodily stances of the individual figures as functioning, collectively, to create an internal aesthetic of regulated, balanced diversity. In other words, the turning of the heads and the gesturing hands, together with other bodily features like the opening and closing of the figures’ mouths and their shifting contrappostos, would be read, on the whole, as redundant signifiers, lacking corresponding signifieds.

Few viewers would approach a group of figures like the Hadrianeum personifications expecting to identify or understand them fully from their gestures. Nevertheless, the same principle of studied diversity is also visible in the head-dresses, hairstyles, costumes and weapons of the Hadrianeum figures. Of all the head-coverings worn by the figures, every one is different. A number of the figures have bare heads, but even amongst these, no single hairstyle is replicated. Every single one of the surviving weapons is also unique, as are the costumes, with their complicated drapery folds, and their tassels, fringing, knots, ties, bands and pins. The crucial question is whether these unique outfits and attributes are really to be read as accurate reflections of the ethnography of empire, or if they belong to the same class of purely decorative variation that marked the bronze statues from Herculaneum? Did the ‘real’ women of any particular province, on seeing that their neighbours in the next province tied up their hair, really (all) decide to leave theirs flowing down on their shoulders? Did the inhabitants of Moesia pick up a short dagger so that they would not be confused with the inhabitants of Phrygia, even if a long axe would be a lot more effective at mowing down their enemies? Perhaps they did – material culture was certainly used in antiquity to create distinctive national identities. But by raising the possibility that the figures’ costumes and attributes, like their gestures, are the result of studied artistic improvisation, I aim to problematize the idea that we can neatly divide the formal and iconographic characteristics of the individual personifications into ‘those which exist purely for aesthetic/decorative variation’ (as the contra- pposto, head-turns and other ‘embodied’ features would conventionally be seen) and ‘those which bear a concrete, semantic meaning’ (i.e., the visibly cultural costumes and attributes).

The notion that the costumes and attributes may be decorative improvisations rather than accurate inventories of ethnic difference would help to explain the problems which modern scholars have encountered in identifying the individual nations (not to mention the suspiciously Roman feel of some of the weapons and hairstyles) [45]. However, even if we suppose that the artists of the Hadrianeum reliefs did faithfully reproduce clothes and weapons as observed, the process of identifying the nations would still be problematized by the subordination of the individual element to the context of the group. Unlike the Hadrianic province coins, whose personifications held a wide range of evocative attributes (Panathenaic amphorae, camels, etc.), every one of the Hadrianeum figures held a weapon, regardless, presumably, of whether this was their most definitive and easily recognizable characteristic in the mind of the Roman viewer. By contrast, it is possible to compare the literary descriptions of the provinces by ‘geographical’ writers like Pliny and Strabo. While these texts clearly do not give an objective representation of the Empire, they nevertheless admit far more heterogeneity than does the Hadrianeum group [46]. We learn from Pliny, for instance, that ‘the cheese most highly prized at Rome, where the goods of all nations are judged side by side, is from the provinces of Nemausus’ [47]. At the Hadrianeum, a personification of Nemausus would have to be recognized by its military equipment, regardless of the fact that the most evocative symbol of Nemausus for a Roman viewer was its cheese.

So far I have shown how far the appearance of the individual personification is determined by its environment – that is, by its inclusion in a larger series. The construction of a balanced group, where each single element complemented all the others, necessitated at least a biased selection of costumes and attributes, if not a free improvisation of these features. In the case of the Hadrianic nations, this mode of representation had profound consequences for the way in which the Empire was presented to the viewers back in Rome. As far as the Hadrianeum is concerned, each nation of the Empire only makes visual sense when inserted back into the group: in this way, the whole (Empire) becomes a holistic, quasi-natural entity, in being shown to equal more than the sum of its parts. This visual Gestalt may have been underscored by the way in which the reliefs were arranged on the building. It is interesting to note that, while the original display of the reliefs remains a controversial issue, all the suggested reconstructions envisage them as surrounding or flanking the temple, whether on the podium, the portico, the attic, or around the walls on the temple interior [48]. Rome, embodied by the imperial temple building, may thus have literally been surrounded by her Empire; in this case, the ancient viewers may have been prompted to make analogies with those Roman season mosaics in which personifications of Spring, Summer, Autumn and Winter revolve around a central emblema of ‘Annus’, the personification of the year [49]. Like the seasons in these mosaics, the nations of the Empire can be understood as constitutive components of the centre; at the same time, neither Annus nor Roma is reducible to the sum of the parts that surround it.

Most important here, however, is the fact that the subordination of the individual nation to the group as a whole renders the interrogation of individual costumes and attributes inappropriate: any viewer – ancient or modern – who stops to consider the single figure in the hope of finding out which nation it represents, and what that nation is like, is bound to be disappointed. This reading of the Hadrianeum personifications as figures that simultaneously solicit and problematize interpretation has clear implications for our understanding of the temple building in its sociopolitical context. As Pierre Bourdieu has stated (writing about the reception of high culture more generally):

When the message [of a work of art] exceeds the possibilities of apprehension or, to be more precise, when the code of a work exceeds in subtlety and complexity the code of the beholder, the latter lose interest in what appears to him to be a medley without rhyme or reason, or a completely unnecessary set of sounds or colours. In other words, when placed before a message which is too rich for him, or ‘overwhelming’ as the theory of information expresses it, he feels completely ‘out of his depth’ [50].

Once the ‘subtlety and complexity’ of its imagery has been exposed, the Hadrianeum building ceases to represent (just) a shared celebration of Roman power over the foreign enemy: instead, the viewing of this imperially commissioned building becomes a way of reinforcing existing power hierarchies within the ancient metropolis. The Hadrianeum personifications constructed an uncomprehending viewing subject, whose failure to interact successfully with the personifications implicitly rationalized their position as the ruled subjects of the Empire. A similar process is visible in the passage of Ovid’s Ars Amatoria discussed above. Here, the lazy incomprehension of the protagonist raises a smile: however, if this alienation of the viewer is seen not as an isolated, fictional episode, but as a potentially accurate reflection of how Roman citizens experienced monumental representations of ‘their’ Empire, then the implications are rather more sinister.


In this article I have challenged two closely related assumptions about the production and reception of allegorical images in antiquity. Firstly, I have used ancient texts and images to dispute the pervasive notion that personifications were easily interpreted by their ancient viewers. By extension, I have problematized the widely held belief that allegorical personifications were used by patrons to facilitate the straightforward communication of information to diverse and largely illiterate audiences [51]. In formulating these ideas I have focused my attention on a series of personifications which lack any identifying inscriptions. However, it should not be too hard to see how this reading might also be applied to groups of personifications which did have their names supplied. If excavations ever produced a series of identifying inscriptions for the Hadrianeum figures, the reading of them given here would have to be only slightly modified. The biased selection and/or free improvisation of the figures’ costumes and attributes would still mean that the appearance of the individual personifications failed to match up to the preconceived notions of the nation represented: in this way, the figures would continue to underline the viewer’s dependence on the imperial patron as a revealer of previously unknown facts.

I hope to have indicated that introducing the figure of the viewer into the study of ancient allegorical art can provide positive and useful information about the function and meaning of these images in ancient society. In the specific case studied here, considering the dynamics of reception has indicated the double-sided nature of such imperially motivated representations of empire. On the one hand, monuments like the Hadrianeum ostensibly served to illustrate and celebrate the power of a (unified) Rome over her foreign nations; on the other hand, these exotic images of empire served as a mechanism to stratify and subordinate the viewing population of the city. The choice of the classicizing, allegorical mode for the representation of the Empire at the temple of Hadrian was a loaded one – far from allowing the easy and immediate communication of facts to its audience, it actually enacted the disenfranchisement of the Roman viewer from any real knowledge or control of the empire to which they belonged. Modern scholars who lament our inability to reconstruct the intentions and knowledge of those who commissioned and designed the Hadrianeum sculptures may, therefore, take cold comfort from the fact that this sense of alienation was probably shared by the majority of viewers who looked at these images in antiquity.

Screenshot 2018-12-26 at 16.34.34.png


This article develops a theme from my 2005 Courtauld PhD thesis, ‘Embodiments of Empire – Roman Imperial Geography in Human Form’. I thank Peter Stewart, Robin Osborne, Mary Beard, and the two anonymous readers for Art History for improving earlier drafts of these ideas – any mistakes of course remain my own. I also thank the AHRC and the Courtauld Institute for funding my doctoral research, and the British School at Rome for the Rome Award that I held from October 2003 to January 2004.

1  See, for instance Robin Osborne, ‘The Viewing and Obscuring of the Parthenon Frieze’, Journal of Hellenic Studies, 107, 1987, 98–105; Jas Elsner, ‘Cult and Sculpture: Sacrifice in the Ara Pacis Augustae’, Journal of Roman Studies, 81, 1991, 50– 61; John Clarke, Art in the Lives of Ordinary Romans: Visual Representation and Non-Elite Viewers in Italy, 100 BC–AD 315, Berkeley, Los Angeles and London, 2003.

2  Marina Warner, Monuments and Maidens: The Alle- gory of the Female Form, Berkeley and Los Angeles, 1985, xix.

3  The term ‘allegorical’ has been debated since antiquity, but in this article my use of the word is general and conservative. See Jon Whitman, Allegory: The Dynamics of an Ancient and Medieval Technique, Oxford, 1987, especially appendices 1 and 2 on the histories of the terms ‘Allegory’ and ‘Personification’.

4  On the limits of the Campus Martius, see T. P. Wiseman, ‘Campus Martius’, in Eva Margaret Steinby, ed., Lexicon Topographicum Urbis Romae, Rome, 1993, vol. 1, 220–4.

5  For reasons of space, only eight figures are illustrated here. All the surviving figures can be seen in the most recent publication of the building and reliefs: Marina Sapelli, ed., Provinciae fideles: il fregio del tempio di Adriano in Campo Marzio, Milan, 1999. Drawings of four other now-lost figures (preserved in manuscripts) are shown at Sapelli, Provinciae, 78–80. For the reconstructions of the temple portico where the reliefs were displayed, see Amanda Claridge ‘L’Hadrianeum in Campo Marzio: storia dei rinvenimenti e topografia antica nell’area di Piazza di Pietra’, in Sapelli, Provinciae, 117–27. If this portico extended not only along one side of the temple but all around its perimeter (as Luigi Canina argued at the end of the nineteenth century, and as the latest reconstruction continues to allow), this could mean that in excess of eighty figures were to be seen on the building. Claridge, ‘L’Hadrianeum’, 125.

6  For discussion of the type of monument, see R.R.R. Smith, ‘Simulacra Gentium: the Ethne from the Sebasteion at Aphrodisias’, Journal of Roman Studies, 78, 1988, 50–77; Paolo Liverani, ‘‘‘Nationes’’ e ‘‘Civitates’’ nella propaganda imperiale’, Mitteilungen des Deutschen Arch- ologischen Instituts. R ̈omische Abteilung, 102, 1995, 219–49; and Ann Kuttner, Dynasty and Empire in the Age of Augustus: The Case of the Boscoreale Cups, Berkeley and Los Angeles, 1995, 69–93.

7  For Pompey’s fourteen nationes, made by Coponius, see Pliny, NH 36.41, Suetonius, Nero, 46. The Augustan Porticus ad Nationes, whose location is unknown, appears at Pliny, NH 36.39 and Servius, Ad Aen. 8.722: see Filippo Coarelli, ‘Porticus ad Nationes’, in Steinby, ed., Lexicon Topographicum, vol. 4, 138–9.

8 . For the typological division of province and nation personifications into realistic ‘provinciae captae’ and idealized ‘provinciae pia fideles’, see Piotr Bienkowski, De Simulacris Barbararum Gentium Apud Romanos, Krakow, 1900.

9 . In 1831 Canina and Fea found a wall made from peperino underneath Palazzo Cini, and claimed that it formed part of the external wall of a portico. Luigi Canina, Indicazione topografica di Roma antica, Rome, 1850, 182. Canina subse- quently reconstructed this portico as forming a large rectangular enclosure with the temple at the centre. Canina, Gli edifizi di Roma antica e contorni cogniti per alcune reliquie descritti e dimos- trati nell’intera loro architettura, Rome, 1848, 312, pl. 144. Cf. Canina, Pianta topografica di Roma antica, Rome, 1850, pl. 2.

10 . SHA, Hadr., 19.

11 . Hans Lucas, Zur Geschichte der Neptunsbasilika in Rom, Berlin, 1904, 3, 24–7. Cf. SHA, Verus, 3.

12 . Jocelyn Toynbee, The Hadrianic School: A Chapter in the History of Greek Art, Cambridge, 1934, 2. On the imperial journeys, see Helmut Halfmann, Itinera principum: Geschichte und Typologie der Kaiserreisen im R ̈omischen Reich, Stuttgart, 1986. For general information about Hadrian’s relationship with his Empire, see Mary T. Boatwright, Hadrian and the Cities of the Roman Empire, Princeton, 2000.

13 . Toynbee, Hadrianic. On the coins, see also Paul Strack, Untersuchungen zur r ̈omischen Reichspra.gung des zweiten Jahrhunderts. Teil II: die Reichspra.gung zur Zeit des Hadrian, Stuttgart, 1933. Strack showed that the figures on the coins do not correspond precisely to the imperial administrative divisions of the provinces, and are there- fore better understood as representing gentes or nationes.

14 . Toynbee, Hadrianic, 3.

15 . The similarities between the Hadrianic and Antonine series – in both iconography and ‘spirit’ – are discussed by Toynbee, who sees the two series as produced by the same school of artists, if not the very same workshop. Toynbee, Hadrianic, 147, cf. 156. In a review of Bienkowski’s study (1900, see note 7 above) Percy Gardner endorses Bienkowski’s use of the Hadrianic coins as comparative material for the reliefs: ‘[The reliefs] were fully discussed by Dr. Lucas in a paper in the Jahrbuch of the Institute for 1900: but Dr. Lucas was not altogether successful in understanding or attributing the statues, probably because he missed the true key to them, which is furnished by that remarkable series of coins issued in the reign of Hadrian, and commemorating his journeys into the various provinces of the Roman Empire. On these coins we have not only figures of the provinces visited, but the name of the province in each case. They thus enable us to attribute the uninscribed statues of the Basilica of Neptune.’ Percy Gardner, ‘Bienkowski’s iconography of barbar- ians’, Classical Review, 16: 2, March 1902, 137–8. However, attempts to identify the reliefs from the coins only highlight the incongruity between these series. See, for instance, Toynbee, Hadrianic, 156–9.

16  Toynbee, Hadrianic, 156.

17  Mrs Arthur Strong, Roman Sculpture from Augustus to Constantine, 1907, London and New York, 245.

18  Roger Hinks, Myth and Allegory in Ancient Art, London, 1939, 74.

19  See Catharine Edwards, ‘The Art of Conquest’, in Catharine Edwards and Greg Woolf, eds, Rome the Cosmopolis, Cambridge, 2003, 67–8, fig. 3. One of the figures from the Sebasteion at Aphrodisias holds her arms in the same position: see Smith, Ethne, pl. 2. For the iconography of captured barbarians and their relation to the Roman nation personifications, see Bienkowski, De Simulacris.

20  For the trophy reliefs, see Sapelli, Provinciae, nos. 2, 4, 11, 14, 16, 18 and 30–2. On trophies in Greece and Rome generally, see Tonio Ho ̈lscher, ‘The Transformation of Victory into Power’, in Sheila Dillon and Katherine E. Welch, eds, Representations of war in ancient Rome, Cambridge, 2006, 29–34.

21  Claridge, ‘L’Hadrianeum’, 125.

22  The topography of the immediate area also connects the nations to war. The temple was situated in the Campus Martius, the military zone of the city; moreover, entry to the precinct where the nations were displayed was gained through a triumphal arch. See F. Castagnoli, ‘Due archi trionfali della via Flaminia presso piazza Sciarra’, BCom, 70, 1942, 74–82.

23  Cf. Livy 22.57.10: ‘They gave orders that armour, weapons and other equipment should be made ready, and took down from the temples and porticoes the ancient spoils of enemies.’ See Ida Ostenberg, Staging the World: Rome and the Other in the Triumphal Procession, Lund, 2003, 20; and Katherine E. Welch, ‘Domi militiaeque: Roman domestic aesthetics and war booty in the Republic’, in Dillon and Welch, eds, Representa- tions, 110–12.

24  Toynbee, Hadrianic, 155.

25  The ambiguity of such representations is dramatized in the eerie episode recorded by Suetonius, when the fourteen personifications of Pompey’s long-since conquered and incorporated nationes turn against the Emperor Nero in a nightmare, moving towards him from all sides. Suetonius, Nero, 46: ‘modo a simulacris gentium ad Pompei theatrum dedicaturum circumiri acerique progressu.’

26 Niels Hannestad, Roman Art and Imperial Policy, Aarhus, 1986, 198. See also Anna-Maria Pais, Il ‘podium’ del tempio del divo Adriano a Piazza di Pietra in Roma, Rome, 1979, 122.

27 Claridge, ‘L’Hadrianeum’, 125.

28 Translation from J. J. Pollitt, The Art of Ancient Greece: Sources and Documents, Cambridge, 1990 [1965], 103. For this statue, see Andrew Stewart, ‘Lyssipan Studies I: The Only Creator of Beauty’, AJA, 82, 1978, 163–71.

29 Xenophon, Memorabilia, 2.1.20–34.

30 The simplest examples of this phenomenon are pairs of personifications which illustrate opposite and mutually exclusive concepts. In this story of the choice of Herakles, for instance, the polarization of the concepts of Virtue and Vice is achieved by measuring both figures against exactly the same parameters: cosmetic adornment, facial expression, speech and movement. Vice is made up to look whiter and rosier than reality, while Virtue wears only ‘purity’ for adornment; Virtue wears white clothing while Vice wears clothes ‘in which her bloom would be most conspicuous.’ Virtue’s eyes are ‘modest’, Vice’s eyes are ‘wide-open’; Vice speaks loudly, and immediately; Virtue, correspondingly, speaks quietly, after a modest pause. The road Virtue offers is ‘long and hard’, whereas the road offered by Vice is ‘short and easy’. The characterization of Virtue and Vice using the opposite ends of the same spectrum effectively represents the total opposition of the lifestyles that each figure represents. Herakles cannot be both virtuous and ‘vicious’ – the choice is a clear one between long and short, difficult and easy.

31 John Clarke’s criticism of the unempathetic slant of much modern scholarship seems tailor-made for many previous studies of the Hadrianeum figures: ‘The only viewer that modern scholarly literature has given us [is] an upper- class male who knows everything because he has read all of Greek and Latin literature and has the advantage of photo archives and history books . . . No ancient viewer had the advantages of the modern scholar; to see Roman art exclusively from the scholar’s point of view is to distort its purposes and meanings for the Roman viewer.’ Clarke, Art, 12.

32 See Lucas, ‘Neptunsbasilika’, 10; Toynbee, Hadrianic,156–8;SimoninHelbig,Fu.hrerdurchdie ̈offentlichen Sammlungen klassischer Altertu.mer in Rome, 1963–72, vol. II, 245.

33 One exception is found in Paolo Liverani’s discussion of a typologically similar group of city personifications from an imperial monument in Ephesos. Here, Liverani raises in passing the idea that the ancient viewers may not have been able to identify the individual personifications: ‘Dubito, infatti che lo spettatore efesino di media cultura potesse riconoscere a colpo d’occhio tutte le personificazioni presenti su questi rilievi . . . Non sembra, dunque che tali rappresenta- zioni fossero cos ́ı caratteristiche e di impiego tanto frequente da essere immediatemente rico- noscibili.’ However, he continues to assert that the figures would have been universally recog- nized as generic city personifications: ‘Le figure erano globalmente riconoscibile come personifi- cazioni di citta`.’ Liverani, ‘Nationes’, 242. The reading of the Hadrianeum personifications given here can be transferred to the Ephesian city personifications to support Liverani’s intui- tion that not all viewers could successfully identify the cities portrayed. I would, however, disagree with the categorical statement that all ancient viewers necessarily accessed the ‘secondary’ meaning of such figures as allego- rical portrayals of cities or nations, as opposed to simply accepting them as representations of the ‘real’ human bodies of foreign women.

34  Hinks, Myth, 74.

35  Ovid, Ars Amatoria, 1.219–27. Trans. J. H. Mozey for the Loeb Classical Library. On this passage, see Clarke, Art, 10, and Mary Beard, ‘The Triumph of the Absurd: Roman Street Theatre’, in Edwards and Woolf, Cosmopolis, 35–7.

36  For the appearance of nation and province personifications in triumphal processions, see Ida Ostenberg, Staging the World, 216–24. Ovid’s Tristia, 4.2 provides a vivid description of perso- nifications carried in a triumph of Tiberius: ‘This thing with broken horns and sorry covering of sedge was the Rhine himself, discoloured with his own blood. See! Even Germany is borne along with streaming locks, seated in grief at the feet of the unconquered leader. Offering her proud neck to the Roman axe she wears chains on that hand in which she carried arms.’

37  Clarke, Art, 12.

38  John Boardman, The History of Greek Vases, London, 2001, fig. 13.

39  Carol Mattusch, The Villa of the Papyri at Herculaneum: Life and Afterlife of a Sculpture Collection, Los Angeles, 2005, 195–215.

40  This similarity between statues was also a practical measure in bronze-working: see Mattusch, Villa, 212. Our expectations of visual symmetry in such sets of figures have led to a sixth bronze ‘dancer’ being excluded from the group, because of her slightly smaller size and the fact she is the only one wearing sandals. Mattusch, Villa, 214–15.

41 See Ellen Rice, The Grand Procession of Ptolemy Philadephus, Oxford, 1983, 51. Examples of these images can be found in David Parrish, Season Mosaics of Roman North Africa, Rome, 1986.

42 Mid-second century CE, from the imperial palace at Ostia. See the Lexicon Iconographicum Mytholo- giae Classicae (henceforth LIMC) vol 5.1, s.v. ‘Horai/ Horae’, 528, no. 167.

43 For the identification as Winter, see LIMC, 5.1, s.v. ‘Horai/Horae’, 521, no. 95; she is identified as Hispania by Toynbee, Hadrianic, 106.

44 Pais, Podium, 101, 118–19.

45 Ostenberg, Staging the World, 22, note 39, comments on depictions of trophies in Roman art generally: ‘Even if many depictions are extremely thorough in detail, each shield, cuirass or greave is often depicted ‘‘incorrectly’’ when compared to the actual arms used by the enemy forces at that specific time. On the reliefs, there is often a mixture of Greek, barbarian and at times perhaps even Roman pieces, contem- porary and historical.’ See E. Polito: Fulgentibus Armis: Introduzione allo studio dei fregi d’armi antichi, Rome, 1998.

46 Although Trevor Murphy sees similar ‘centra- lizing’ tactics at work in Pliny’s Natural History: ‘the role of the Encyclopedia . . . was to assimilate the diverse kinds of information that Pliny’s research had turned up and to express them in terms digestible by Roman culture.’ Trevor Murphy, Pliny the Elder’s Natural History: The Empire in the Encyclopedia, Oxford, 2004, 15.

47 Pliny, NH 11.97.

48 Claridge, ‘L’Hadrianeum’, 121–5.

49 See Parrish, Season Mosaics, 47, esp. note 176.

50 Pierre Bourdieu, ‘Outline of a sociological theory of art perception’, International Social Science Journal, 20: 4, 1968, 589–612 (2.3.3).

51 One of the most important works on personification to come out in recent years describes the personifications on Attic document reliefs as ‘ephemeral devices making the substance of the document intelligible to the illiterate, or merely lazy, viewer’. Emma Stafford, Worshipping Virtues: Personification and the Divine in Ancient Greece, London, 2000, 15.