In fourth-century Roman Britain, mosaic production was at an all-time high. The majority of the island’s c. 2000 Roman mosaics date to between AD 300 and 375. During this same period, Britain experienced substantial upheaval. Seats of power and administration shifted from established towns and cities to the countryside, specifically the villas of local aristocrats who became representatives of Roman rule within their regions (Scott 1997: 54; 2012; Bowes 2010: 20). The Late Antique villas of these wealthy and well-positioned individuals became social and political hubs for the surrounding community. This is reflected in fourth-century villa architecture, when social spaces like apsidal dining rooms and private bath complexes became more common in the archaeological landscape (Scott 2000: 77–112; Cosh 2001). Scholars have varied opinions on the appearance of these social spaces. Ellis (1991; 1995) argues for an increasingly hierarchical society in Late Antique society that is reflected by the appearance of apsidal rooms, dining rooms, and more complex villa plans. Bowes (2010: 28–30) disagrees with this argument and, instead, suggests that public architecture’s influence upon private domestic space can explain the appearance of such rooms. Such social spaces, in addition to being fairly new in the architectural vernacular, were often decorated with expensive and intricate mosaics.

The fourth century AD brought about renewed interest in the literary classics of Rome. Therefore, it is not surprising that two fourth-century Romano-British villa owners commissioned mosaics that draw upon Vergil’s first-century BC epic, the Aeneid. Vergil, among other authors, functioned as part of the ‘language’ of the elite (Barrett 1978; Scott 1997: 58). Literature-influenced mosaics like those at Lullingstone and Low Ham have been interpreted as evidence of erudite villa owners demonstrating their knowledge of the Roman classics, thereby legitimizing their claim as the bearers of Roman culture on the island (Dunbabin 1999: 99; Scott 2000: 127; Witts 2005; Mackenzie 2019: 24). The corpus of literary mosaics in England is still expanding. For example, in 2020 archaeologists uncovered the Rutland Roman villa mosaic, which depicts the Siege of Troy as described in the Iliad (Historic England 2021; 2022). However, an iconographical analysis of these mosaics tells only one part of the story. Unlike a painting hanging on a wall, viewers moved across and around a mosaic as they absorbed the imagery and composition.

In this article, I analyze the fourth-century mosaics at Lullingstone and Low Ham through a phenomenological lens, centering the embodied, Roman-era viewer in the context of a province undergoing political and social transformations. In the Roman villa, there existed a dynamic, ontological relationship between physical, inhabited spaces, and the social interactions that occurred within. Phenomenology, therefore, allows us to understand how mosaics influenced, and were influenced by, the viewer walking across them. Such approaches to archaeology have taken hold within the last decades with the work of scholars such as Christopher Tilley (1994) and Yannis Hamilakis (Hamilakis et al. 2002). They, and others working in the same vein, emphasize that we must take into account the experiential encounters that occur within historical landscapes or monuments, when seeking to understand archaeological material. Hamilakis et al. (2002: 1–21) call this methodological approach an ‘archaeology’ of the body and such studies have found their theoretical roots in philosophers such as Martin Heidegger and Maurice Merleau-Ponty (Tilley 1994: 11–14).

Heidegger’s conception of being-in-the-world—that a ‘world’ does not exist without those who dwell or act within it—has been particularly influential. In his metaphor of the Greek temple in The Origin of the Work of Art, Heidegger (2002: 23) stresses that ‘World is that always-non-objectual to which we are subject as long as the paths of birth and death, blessing and curse, keep us transported into being’. We cannot separate space, including the material remains of civilizations past, from those who dwelled within it. For Merleau-Ponty (1964: 162), vision and movement are inextricably tied together and to one’s perception of the world. His embodied perspective is especially fruitful for Roman mosaics given that one’s movement is directly linked to the viewing process.

Recent critiques of phenomenology, especially the approach of Tilley, have noted that one-to-one comparisons of ancient and contemporary experiences of a landscape are problematic given how they may have changed over millennia (Brück 2005: 54). Roman mosaics, especially when contextualized within their original architecture, allow for a much closer approximation of the ancient viewing experience. Although particular experiential elements may be missing, such as lighting and portable furniture, there is a much narrower margin of ‘error’ than with the analysis of prehistoric landscapes. Scholars like John Clarke, Rebecca Molholt, and Ellen Swift have revealed how Roman mosaics benefit from a phenomenological lens. Clarke (1979: 20) coined the term ‘spectator address’, or ‘kinesthetic address’, to describe how mosaics confront and affect (or direct) a viewer. He argues that figural mosaics are integral to understanding one’s navigation of space, because they mimic a viewer’s body with their own placement and action (Clarke 1979: 21). Building on the work of Clarke, Molholt (2011: 287) emphasizes that mosaics must be analyzed within their archaeological context and not ‘as if they were paintings or were created in emulation of painting’. The viewer of a mosaic must, by the very nature of the medium, set foot upon and experience it. Mosaics, therefore, allow us to more fully understand how a viewer may have experienced a historical space.

Swift (2009) uses a phenomenological lens to examine geometric Roman mosaics and analyzes how they influence a viewer’s movement. Swift argues that the confusion and disorientation caused by specific geometric patterns may have resulted in a viewer feeling powerless. Geometric mosaic floors therefore can possess agency, allowing the homeowner to assert their authority and dominance over a viewer. When considered within the context of the Roman house and its related social dynamics, the complex and illusionistic patterns of the later Roman Empire may have functioned not only as aesthetically pleasing pavements but also to communicate the dominance and status of the owner.

Much like the geometric patterns discussed by Swift, the story of Vergil’s Aeneid plays an integral role in understanding how the mosaics at Lullingstone and Low Ham could influence or guide their viewers and their movement within each space. At Lullingstone, an inscribed allusion to the Aeneid in the villa’s main reception hall appeals to the seated, static viewer and makes a social distinction between the owner’s invited guests, who alone have the proper vantage to read it, and anyone else in the room. Meanwhile, scenes from the Aeneid in the frigidarium mosaic at Low Ham suggest how a viewer might move through the space, ultimately circumnavigating the room. By applying a phenomenological framework to these two case studies, I demonstrate the role of narrative in mosaics and the ways it structured a viewer’s experience of Roman social spaces.

The Villa at Lullingstone

The Roman villa at Lullingstone lies on the River Darent, about 20 miles south-east of London. Built in the late first century AD, the villa was continuously occupied until the fifth century AD (Meates 1979; Meates and Allen 1987; Neal 1991). The layout consists of a central reception space flanked by living quarters and smaller reception rooms, a configuration known regionally as the winged-corridor villa (Perring 2002: 74–79). As indicated by a coin in the foundation, the villa’s owner undertook a major renovation between AD 330 and 360 (Meates 1979: 84; Mackenzie 2019: 25). Builders removed the western portico and added an apsidal triclinium (dining room) to the western end of the central reception hall, creating two interconnected spaces newly decorated with mosaics (Neal and Cosh 2009: 379–385).

Upon arriving at the villa, the Roman-era visitor climbed a set of stairs to reach the main entrance. From there, three more steps led into the audience chamber. The villa’s main entrance gave the visitor a direct line of sight through the audience chamber and into the recently added dining room (Figure 1).

Figure 1
Figure 1

Reconstruction of Lullingstone Audience Hall (© Historic England Archive PLB_N000046. Reproduced with permission, further reuse not permitted).

A square mosaic surrounded by a border of coarse red tesserae lies at the center of the audience chamber (Figure 2). A thick band of white and red guilloche encloses a figural panel of the winged horse Pegasus carrying Bellerophon, who is about to strike the Chimera. Four dolphins surround the hero with an open shell between each pair. A swastika-meander band frames the entire figural scene, and at each corner of the cushion-shaped composition, there are depictions of the four Seasons. A geometric carpet mosaic encompassing hearts, swastikas, and crosses separates the Bellerophon scene in the audience chamber from the dining room. A visitor invited to enter the dining room proceeded to the rear of the audience chamber and up a single step, decorated with a triangular motif. Apsidal triclinia, like the one at Lullingstone, came to define Late Antique domestic spaces across the Roman Empire beginning in the third century AD. Diners reclined upon a stibadium (semi-circular dining couch) that held about seven people. The central panel of this mosaic features Europa riding Jupiter disguised as a bull, while the rest of the apse is covered in coarse red tesserae (Figure 2; Meates 1979: 76; Henig 1997). The coarse red tesserae were regularly placed near the mosaic, but became more irregular as they radiated out; likely because there would have been permanent couches to cover this area.

Figure 2
Figure 2

Lullingstone Villa Mosaic (© Historic England Archive PLB_J910065. Reproduced with permission, further reuse not permitted).

Europa is nude but for a thin veil that covers her lower half, which she holds as a canopy above her head. In front of the bull, a winged cupid moves to the right and gestures back at the couple, while a second cupid tugs on Jupiter’s tail. All figures appear in red outline on white ground. The blue-gray tesserae at the bottom of the scene provide the only indication of a background or groundline. At the top of this panel, set just inside the dining room, a Latin inscription on two registers reads: INVIDA SI TA[URI] VIDISSET IUNO NATATUS, IUSTIUS AEOLIAS ISSET ADUSQUE DOMOS. The English translation is: ‘If jealous Juno had seen the swimming of the bull she would have with greater justice on her side gone to the halls of Aeolus’ (Neal 1991: 15). The text is oriented to be read from the stibadium and refers to a passage in Vergil’s Aeneid (1.65–75) where Juno asks Aeolus, the god of the winds, to conjure up a storm that would overturn Aeneas’ fleet. A geometric border of alternating red and white circles encloses the scene. This is one of a number of inscribed mosaics in England. Leader-Newby (2007: 189–192) references one from East Yorkshire Aith animals in roundels including a lion and a bull with inscriptions that say ‘Lion Spear-Bearer’ and ‘Bull Manslayer’ respectively as well as a mosaic from Hampshire depicting Dionysus with busts of the Seasons and a partially preserved inscription that reads Quintus Natalius Natalinus et Bodoni. A more recently excavated example is the Boxford mosaic, depicting scenes from Greek mythology and including an inscription (Beeson et al. 2019; excavated in 2017).

Neal and Cosh (2009: 385) note that the workmanship of the Lullingstone mosaic is ‘variable’. The figural scenes are sophisticated and, as they suggest, may represent the hand of the workshop’s lead mosaicist. While scenes of Bellerophon and the Chimera appear in two other British mosaics at Hinton St. Mary and Frampton, the geometric patterns are unique. There are no other fourth century villa mosaics in Lullingstone’s county of Kent, reinforcing the somewhat unique nature of these pavements within the corpus of Roman Britain.

Mosaics and Hierarchy at Lullingstone

The dining room and audience chamber at Lullingstone were interconnected spaces with direct lines of sight between them, yet viewers would have experienced each room’s mosaics in a different way. Let us begin by considering the placement and content of the Vergil-inspired mosaic in the dining room. The dining room is elevated above the audience chamber, meaning viewers in the latter had an oblique and obscured view of the mosaic depicting Jupiter and Europa. In other words, one literally cannot stand in the audience chamber and see the Vergilian mosaic. The composition is also oriented towards the niche, so only those reclining on the stibadium had a proper view of the scene and its inscription. Late Antique depictions of banquets, such as the third-century AD mosaic from the House of the Buffet at Antioch, feature small, portable tables set alongside the couches, which would allow diners at Lullingstone to easily view the composition from their reclined positions (Dunbabin 2003: 156–164). Within this group of diners, those at the center of the stibadium had the best vantage point, reinforcing their rarified status.

The different bodies of knowledge required to properly ‘read’ the Vergilian mosaic reinforce its limited accessibility. First, the viewer would need to identify the figures as Jupiter and Europa from Greco-Roman mythology. Second, one would need to be familiar with the plot of the Aeneid to understand the inscription’s reference. The viewer would have to make connections beyond simple identification of the figures, however, since the mythological couple appear nowhere in the Aeneid. Finally, the inscription is an elegiac couplet, a poetry form popularized by Ovid in the first century AD (Cosh 2016). The mosaic, therefore, demands knowledge on many levels from the viewer; being able to read the Latin language, Classical mythology, the epic of Vergil, and Ovidian poetry. As Michael Squire (2009: 170) notes, ‘the combination of text and image at Lullingstone widens the referential scope of both media’. The inscription references an episode from the Aeneid, but also makes clear that the image of the bull is, in fact, Jupiter in disguise. Upon witnessing the adultery of her husband as he abducted Europa, Juno would have had even more justification for going to Aeolus to produce the windstorm with which she tried to drown Aeneas. Because of the layout and referential scope of Lullingstone’s Vergilian mosaic, with its subtle interplay between text and image, the pavement was only fully accessible to those with the necessary position (spatially and socially, figuratively and literally).

The mosaic ensemble in the large open audience chamber also had a mythological theme, depicting Bellerophon and Pegasus, which is oriented to face those in the dining apse. The Seasons, in roundels at the corner of the Bellerophon scene, are arranged in a temporally accurate progression; beginning with hooded Winter in the top left, Spring with a swallow on her right shoulder in the top right, and clockwise from there (Witts 2005: 82). Winter and Spring face towards the apse like Bellerophon whereas Summer, with her ears of corn, and Autumn, now lost, face the opposite direction towards the main doorway. While the main figures in the composition face towards the dining room and favor those reclining in the apse, the orientation of peripheral figures like the Seasons meant this mosaic was accessible to a wider range of viewers, including those moving through the audience chamber. The clockwise progression of the Seasons, for example, appealed to a mobile visitor who may not have had the privilege of a long stay. As this comparison reveals, the mosaics at Lullingstone reflect and reinforce the hierarchy of the two spaces and the viewers who occupied them.

Dining in the Roman world varied considerably but, as we see at Lullingstone, often reinforced hierarchies. Sensory experiences made clear who belonged and who did not. In her article ‘The Sensory Experience of Food Consumption’, Erica Rowan (2019) compares eating in a ‘fast food’ taberna, where the diner sat upright on a bench or seat and was surrounded by loud conversation, to reclining in a sumptuous elaborately decorated triclinium with a small group of chosen companions. Sitting upright in a hurried and chaotic environment produced a vastly different, and less refined, dining experience than a leisurely meal at an elite house. Lullingstone’s two-part dining and reception space sorted individuals based on their position—reclining in the apse versus moving through the audience chamber—and resultant ability to read or experience each mosaic pavement. The small, enclosed apsidal dining room at Lullingstone was an exclusive space, in direct contrast to the open, spacious audience hall that served as the villa’s central artery and main gathering place. Those invited into the dining room took their place as static viewers, as they sat upon a stibadium to eat and drink. Viewing in this space would have been reciprocal, as the apse formed both a ‘stage’ upon which to view the diners but also a ‘viewing box’ for the feast participants to witness the comings and goings through the villa (Witts 2000).

The large open space in front of the apse served as a place for the Roman villa owner to receive clients and visitors from local tenant farmers to fellow landowners. We can imagine that the atmosphere in the audience chamber would have been hectic and filled with people coming and going, especially servants going about their daily routine (Mackenzie 2019: 25–26). As Dunbabin (2003: 151) notes, Late Antique depictions of banquets often conspicuously feature not only the diners, but the attendants as well. Dinner parties like those at Lullingstone would have included attendants to carry a wash basin, serve wine and food, fan guests, and more. The mosaic in the audience chamber is, therefore, necessarily accessible to a wider range of viewers. Its placement and more legible content, a straightforward mythological story, did not encourage the viewer to linger as they walked across the mosaic floor to greet the patron or to move throughout the house completing duties.

The Vergil-inspired mosaic within the apse at Lullingstone had only one optimal viewing point—reclining upon the stibadium. In fact, given the elevation of the apse, one was not able to see the pavement from the audience chamber. The mosaic also hinged upon at least three different bodies of knowledge to recognize the subtle allegorical relationship between Vergil’s text and the image of Jupiter and Europa. On the other hand, the mosaic in the large, open audience chamber catered to a wide range of viewers through the positioning of the Seasons and the geometric ‘carpet’. It also did not require the same literary capital to understand the mythological composition of Bellerophon and the Chimera. The mosaics at Lullingstone, therefore, served as an immediate and tangible reinforcement of social status within these spaces.

The Villa at Low Ham

The Roman villa at Low Ham is in south-west England and was first discovered in 1938. Excavations between 1945 and 1955 revealed a reception hall, courtyard, and bath complex (Radford 1947a; 1947b; 1948; Linford et al. 2019). In contrast to the Lullingstone villa, we have a somewhat piecemeal understanding of the villa’s layout, which can be divided into roughly two areas: a public reception space to the north and a bath suite directly to the south. The bath complex was connected to the main living quarters via a long and narrow corridor. Excavators discovered the villa’s Vergilian mosaic in the frigidarium (bathing room with a cold-water pool) (Cosh and Neal 2005: 253–257).

As at Lullingstone, Low Ham’s mosaic was situated in a part of the villa used for entertainment. In the Roman world, private bath complexes were nearly on par with dining and reception rooms as spaces of social interaction. The concept of daily bathing arrived in Britain with the Romans (Rook 1997). Tacitus (Agricola 21.3), writing in AD 98, describes the introduction of such Roman practices in Britain saying, ‘Step by step they were led to things which dispose to vice, the lounge, the bath, the elegant banquet’. From the time of Hadrian, lavish bath complexes could be found throughout private homes in Britain and the western provinces more generally (Yegül 2010: 181). Bathing was a social ritual, during which the villa’s owner and his guests would have leisurely washed, soaked, and perhaps even received a massage. During the early Roman Empire, written records attest to poetry readings, music, and singing in the private bath complex (Martial, Epigrams 11.52; Yegül 2010: 18). Bathing remained a central part of daily life well into the fourth century AD. For example, the fifth-century AD poet and politician Sidonius Apollinaris (Epistulae 2.2.5) wrote about his own private bath in a letter to his friend Consentius, bragging that his frigidarium can stand comparison to any public complex. As Maréchal (2020: 52) suggests, large, private baths like the ones in Apollinaris’ letter were large enough for several people and remained a setting for social encounters. A flourish of newly built bath complexes in Roman Britain suggests that bathing went hand in hand with dining as part of an elaborate social ritual (Savani 2017: 10).

Most bath complexes, from grand public ones to those found in private houses, had at least two rooms—the caldarium or hot room and the frigidarium or cold room. The caldarium was warmed through a hypocaust or under-floor heating system, whereas the frigidarium remained unheated and usually contained a plunge bath of cold water (Yegül 2010: 128). In domestic spaces, where the complex was generally smaller, the bather would apply oil after undressing, move to the hot room to induce sweating, and then proceed to the frigidarium for a plunge in the cold bath (Rook 1992: 13). The frigidarium, being the last part of the bathing ritual, was often the most luxuriously appointed room within the complex, as this is where people tended to gather and socialize (Yegül 2010: 17).

In the case of the Low Ham villa, a visitor moving south through its narrow corridor came to the antechamber that preceded the frigidarium, which is divided into two parts. A geometric mosaic in tones of blue-gray, red, and off-white is centered at the entrance to the frigidarium. The room then widens to reveal the Vergilian mosaic and, beyond it, steps leading into the cold plunge pool. The mosaic dates to the final reconstruction of the villa around AD 330 and was in use until at least AD 370 (Radford and Dewar 1954: 2). It depicts various episodes from Books One and Four of the Aeneid, during which Aeneas meets and falls in love with Dido, the illustrious and beautiful queen of Carthage (Figure 3). The five-part mosaic consists of three square panels in the middle with two longer, rectangular scenes on the sides.

Figure 3
Figure 3

Low Ham Villa Frigidarium Mosaic (By permission of the Somerset Archaeological and Natural History Society and South West Heritage Trust).

Upon crossing the geometric threshold, a visitor encountered the first square panel, oriented to face the frigidarium’s entrance. It shows the first meeting of Aeneas and Dido. A nude Venus is flanked by Dido on one side and by Aeneas and Cupid in the guise of his son Ascanius on the other. The two other square panels are oriented to face the plunge bath. At the center, we find another depiction of Venus, nude but for a blue cloak, and flanked by winged cupids bearing torches, while the second panel depicts Aeneas and Dido embracing between two trees. Long, narrow panels with scenes of swift movement occupy the two elongated side panels of the mosaic. Three galley ships in profile, to the left of the entrance, represent the arrival of Aeneas to Carthage in Book One of the Aeneid. On the right, Dido, Aeneas, and Ascanius ride on horseback in single file, an excerpt from Book Four. The figures and narrative procession in the Low Ham villa’s mosaic offer evidence for recreating prescribed routes through the frigidarium.

Cosh and Neal (2005: 257) note that the Low Ham mosaic has been attributed to the Durnovarian mosaic workshop, based in and around Dorset. And yet, other than the winged cupids, which resemble those from the nearby Pitney villa, the figures from Low Ham are not directly comparable stylistically to any surviving Durnovarian examples. In terms of representations of Aeneas, the only possible comparison is from the Frampton mosaic in Dorset, which likely depicts the hero with the Golden Bough in one of a series of heroic scenes. Overall, then, the Low Ham mosaic has stylistic connections to a number of nearby surviving mosaic pavements, but no direct comparisons.

The Pathways of Aeneas

The Low Ham mosaic is a continuous narrative, telling a story through multiple scenes or frames within one artwork. A viewer must shift into the embodied role of storyteller to reconstruct narrative scenes in a process Eleanor Leach (1988: 323) calls ‘suggestive address’. At Low Ham, the viewer physically moves through the narrative as storyteller, while walking across the pavement. This experiential ‘reading’ of a narrative has been more fully analyzed by scholars of Roman wall painting. Timothy O’Sullivan (2011: 141), examining the Odyssey Landscapes from Rome, draws a connection between the movement of the story and the kinetic experience of the viewer. He uses the term ‘narrative walk’ to describe a viewer’s experience of looking at painted scenes from a story and reconstructing its plot, while circulating in a room. The narrative has agency in providing a route that the viewer can use to ‘read’ the painted images and navigate the space. Nathaniel Jones (2018: 10) conceptualizes such wall paintings as ‘active objects’ that ‘model and spur intellectual processes’. Roman wall painting, along with sculpture, have been extensively analyzed in terms of narrative and viewership, but Roman mosaics also benefit from such a phenomenological approach. The Low Ham mosaic, for example, uses narrative to suggest how one might circumnavigate the frigidarium through successive scenes.

Unlike the apsidal mosaic at Lullingstone, there is no single point within the frigidarium that allows the viewer to see each panel from a straight-on angle, necessitating a full circulation of the room to take in each scene in turn. It suggests that the Low Ham villa’s mosaic was arranged to encourage movement through the room, where there were only two places to pause for a static vantage point, based on the activities taking place within it—the doorway and the plunge bath. The arrangement of the mosaic corresponds to the room’s functional uses, moving the visitor through the space between these two fixed positions. Moreover, the depiction of events from Virgil’s Aeneid reinforced a prescribed counterclockwise circulation pattern in the frigidarium based on the way the narrative unfolds. The figural composition of the mosaic and its use of the well-known Aeneid narrative as subject matter work together to suggest how a viewer might move through the space. A comparable example comes from Room C, a frigidarium, of the Baths of Neptune at Ostia (Clarke 1979: 27). The mosaic in this room, as at Low Ham, addresses viewers from the doorways and plunge baths through black and white figured mosaics.

An inherent feature of many figural mosaics is that the viewer notices and feels the presence of other bodies in the space (i.e. those depicted in the mosaic). As noted earlier, Clarke (1979: 21) coined the term ‘kinesthetic address’ to describe this phenomenon. He uses this term to describe the ability of a mosaic to confront and direct the viewer or, in other words, the human reaction to design directions. He differentiates between three specific types of such mosaics; 1) framed static panels with one ideal viewing point; 2) geometric motifs; and 3) unframed figural mosaics that must be seen from multiple angles and have the greatest impact on the viewer in terms of movement.

In the Low Ham frigidarium, the architect or artist solved the problem of multiple vantage points by combining different types of kinesthetic address—using both static, framed scenes and longer panels of moving figures to create a circulation pattern in the frigidarium. Framed, figural mosaic panels differ from geometric patterns because they have an optimal viewing point. At the Low Ham villa, the three panels at the center of the room fit this description. Based on the layout of the room, the prioritized vantage point was from the plunge bath; two of the three static scenes in the center of the room faced in that direction, the Venus panel at the center and Aeneas embracing Dido. The scene showing Aeneas and Dido meeting for the first time, meanwhile, faces the entry to the room. From either vantage point, the plunge bath or doorway, the viewer would be looking upon a key part of the narrative.

On the other hand, the two rectangular scenes on the perimeter of the room depict the dynamic movement of boats and horses respectively, which move in opposite directions. From the plunge bath, there was an implied counterclockwise or left-to-right movement. The boats, in processing from left to right, suggest how the viewer might move along the perimeter of the room from the pool to the entrance. The suggested movement towards the doorway is emphasized by the figure of Achates, who is aligned with the figures in the static mosaic that faces the entrance. The second long, rectangular panel showing a hunting scene with Dido, Aeneas, and Ascanius on horseback mimics that of the boats, but in the opposite direction. The billowing capes of all three figures reinforce their direction of movement. Upon circling back to the plunge bath, we come full circle and are confronted with another static, framed panel. The alternation of static and dynamic figural panels in the Low Ham villa’s frigidarium moved the visitor through the space via a set circulation pattern, which was reinforced by the narrative quality of the scenes.

The use of a specific story, Vergil’s Aeneid, serves to reinforce this path of movement. It also suggests that the ideal viewer was educated enough to follow the plot. The viewer became their own narrator and storyteller since the mosaic lacked any inscriptions: one would need to know, even cursorily, the plot of the Aeneid. In a sense, the viewer was encouraged as a sight-reader to take a narrative ‘walk’ around the panels in the correct order. Aeneas, Venus, Dido, and Ascanius retain the same formal qualities, such as costume and coloring, throughout all five panels. This repetition of the figures implies a continuous narrative and conveys an ongoing plot. The guilloche motif becomes an illusionistic frame, separating one scene from the other.

The plot begins and ends at the entrance to the plunge bath, which was the ideal vantage point in the room given the placement of the central static panels. Additionally, the two most important scenes from Books One and Four, the meeting of Dido and Aeneas and their first embrace, are strategically placed in front of the plunge bath and entrance. It is an unlikely coincidence that two of the most important scenes from Books One and Four, and the story in general, occur in static panels in front of the doorway and bath. The viewer is asked to linger and think about these two key moments in the narrative at the room’s stationary points. The storyline, like the figures themselves, encouraged the viewer to move in a counterclockwise circulation pattern from the plunge bath. At the Low Ham villa, the Vergilian mosaic served to show how one might navigate the bath complex, while simultaneously addressing the elite, well-educated viewer via narrative allusions to the most popular literary work of the time.


In this article, I have argued that a phenomenological embodied framework allows for a closer approximation of lived experience in archaeological spaces like the Roman villa, where the décor expressed the aspirations and status of the patron. The villas of Late Antique Britain are especially fruitful for such studies as numerous examples survive with their mosaic floors intact. The two case studies presented here, the Lullingstone and Low Ham villas, have in common Vergilian mosaics in prime reception spaces. Only elite guests would have been invited to recline and dine in the apse with the villa’s owner, whereas a much wider part of the population would have entered, or moved through, the audience chamber. The inscribed panel of Europa and Jupiter appealed to the elite, reclining viewer, whereas the figural panel of Pegasus and Bellerophon was accessible to a much wider audience. At Low Ham, the mosaic in the frigidarium suggested a path of circulation between the plunge bath and entrance, which was reinforced by the figural scenes and the narrative of the Aeneid. Overall, the mosaics of Lullingstone and Low Ham, when analyzed through the lens of phenomenology, reveal how narrative is used to frame social interactions within the Roman villa. This study has wider implications for the role of literature in both wall painting and mosaics and presents one way of approaching such decorative ensembles. The discovery of related mosaics, such as that from the Rutland villa with scenes from the Iliad (Historic England 2021; 2022), will allow us to evolve our understanding of the dynamic interplay between Roman-era viewers, popular literature, and villa décor.

Competing Interests

The author has no competing interests to declare.


Ancient Sources

Martial (Translated by D.R. Shackleton Bailey 1993). Epigrams, Volume III: Books 11–14. Loeb Classical Library 480. Cambridge, MA: Harvard University Press. DOI:

Sidonius Apollinaris (Translated by W.B. Anderson 1936). Letters: Books 1–2. Loeb Classical Library 296. Cambridge, MA: Harvard University Press. DOI:

Tacitus (Translated by M. Hutton and W. Peterson, Revised by R. M. Ogilvie, E. H. Warmington, and Michael Winterbottom 1914). Agricola. Germania. Dialogue on Oratory. Loeb Classical Library 35. Cambridge, MA: Harvard University Press. DOI:

Vergil (Translated by H. Rushton Fairclough 1916). Eclogues. Georgics. Aeneid: Books 1–6. Loeb Classical Library 63. Cambridge, MA: Harvard University Press). DOI:

Modern Sources

Barrett, A.A. 1978. Knowledge of the literary classics in Roman Britain. Britannia 9: 307–313. DOI:

Beeson, Anthony, Matt Nichol, and Joy Appleton. 2019. The Boxford Mosaic: A Unique Survivor from the Roman Age. Newbury: Countryside Books.

Bowes, Kimberly Diane. 2010. Houses and Society in the Later Roman Empire. London: Duckworth.

Brück, Joanna. 2005. Experiencing the past? The development of a phenomenological archaeology in British prehistory. Archaeological Dialogues 12(1): 45–72. DOI:

Clarke, John R. 1979. Roman Black-and-White Figural Mosaics. New York: New York University Press/College Art Association of America.

Cosh, Stephen R. 2001. Seasonal dining-rooms in Romano-British houses. Britannia 32: 219–242. DOI:

Cosh, Stephen R. 2016. The Lullingstone mosaic inscription – a parody of Martial? Britannia 47: 262–266. DOI:

Cosh, Stephen R. and David Stanley Neal. 2005. Roman Mosaics of Britain Vol. II. London: Society of Antiquaries of London. DOI:

Dunbabin, Katherine M.D. 1999. Mosaics of the Greek and Roman World. Cambridge: Cambridge University Press.

Dunbabin, Katherine M.D. 2003. The Roman Banquet: Images of Conviviality. Cambridge: Cambridge University Press.

Ellis, Simon P. 1991. Power, architecture and décor: how the Late Roman aristocrat appeared to his guests. In: Elaine K. Gazda (ed.). Roman Art in the Private Sphere: New Perspectives on the Architecture and Decor of the Domus, Villa, and Insula. Ann Arbor, MI: University of Michigan Press: 117–135.

Ellis, Simon P. 1995. Classical reception rooms in Romano-British houses. Britannia 26: 163–178. DOI:

Hamilakis, Yannis, Mark Pluciennik, and Sarah Tarlow. 2002. Thinking Through the Body: Archaeologies of Corporeality. New York: Kluwer Academic/Plenum. DOI:

Heidegger, Martin. 2002. Off the Beaten Track. Translated and edited by Julian Young and Kenneth Haynes. Cambridge: Cambridge University Press.

Henig, Martin. 1997. The Lullingstone Mosaic: Art, Text and Literature in a Fourth-Century Villa. Mosaic 24: 4–7.

Historic England. 2021. Roman Villa East of Foster’s Bridge. Available at [Last accessed 18 July 2023].

Historic England. 2022. Investigations at Rutland Roman Villa, 28 November 2022. Available at [Last accessed 30 October 2023].

Jones, Nathaniel. 2018. Starting from places: continuous narration and discontinuous perspectives in Roman art. The Art Bulletin 100(1): 7–35. DOI:

Leach, Eleanor Winsor. 1988. The Rhetoric of Space: Literary and Artistic Representations of Landscape in Republican and Augustan Rome. Princeton, NJ: Princeton University Press.

Leader-Newby, Ruth. 2007. Inscribed mosaics in the Late Roman Empire: perspectives from east and west. In: Zahra Newby and Ruth E. Leader-Newby (eds). Art and Inscriptions in the Ancient World. Cambridge: Cambridge University Press: 179–202.

Linford, Neil, Paul Linford, Andrew Payne, and Lawrence Rees. 2019. Low Ham Roman Villa, High Ham, Somerset Report on Geophysical Surveys, July 2018. Available at [Last accessed 30 October 2023].

Mackenzie, Caroline K. 2019. Culture and Society at Lullingstone Roman Villa. Oxford: Archaeopress. DOI:

Maréchal, Sadi. 2020. Public Baths and Bathing Habits in Late Antiquity: A Study of the Evidence from Italy, North Africa and Palestine A.D. 285–700. Leiden: Brill. DOI:

Meates, Geoffrey Wells. 1979. The Roman Villa at Lullingstone, Kent, Volume I: The Site. Maidstone: Kent Archaeological Society.

Meates, Geoffrey Wells and D.F. Allen. 1987. The Roman Villa at Lullingstone, Kent Vol. II. Maidstone: Kent Archaeological Society.

Merleau-Ponty, Maurice. 1964. Primacy of Perception: And Other Essays on Phenomenological Psychology, the Philosophy of Art, History and Politics. Evanston, IL: Northwestern University Press.

Molholt, Rebecca. 2011. Roman labyrinth mosaics and the experience of motion. Art Bulletin 93(3): 287–303. DOI:

Neal, David S. 1991. Lullingstone Roman Villa. English Heritage.

Neal, Davis S. and Stephen I. Cosh. 2009. Roman Mosaics of Britain Vol. III. London: Society of Antiquaries.

O’Sullivan, Timothy M. 2011. Walking in Roman Culture. Cambridge: Cambridge University Press.

Perring, Dominic. 2002. The Roman House in Britain. London: Routledge. DOI:

Radford, Courtenay Arthur Ralegh. 1947a. The Roman villa at Low Ham. Somerset and Dorset Notes and Queries 25(132): 1–6.

Radford, Courtenay Arthur Ralegh. 1947b. The Roman villa at Low Ham. Somerset and Dorset Notes and Queries 25(135): 61–64.

Radford, Courtenay Arthur Ralegh. 1948. The Roman villa at Low Ham. Somerset and Dorset Notes and Queries 25(139): 141–144.

Radford, Courtenay Arthur Ralegh, and H.S.L. Dewar. 1954. The Roman Mosaics from Low Ham & East Coker. Somerset: Somerset Archaeological and Natural History Society.

Rook, Tony. 1992. Roman Baths in Britain. Princes Risborough: Shire Publications.

Rook, Tony. 1997. Thoughts on the Lullingstone europa floor. Kent Archaeological Review 128: 177–180.

Rowan, Erica. 2019. The sensory experience of food consumption. In: Robin Skeates and Jo Day (eds). The Routledge Handbook of Sensory Archaeology. London: Routledge: 293–313. DOI:

Savani, Giacomo. 2017. Rural Baths and Bathing: Socio-Cultural Interactions in the Romano-British Countryside. Unpublished thesis (PhD), University of Leicester.

Scott, Sarah. 1997. The Power of images in the Late Roman house. In: Ray Laurence and Andrew Wallace-Hadrill (eds). Domestic Space in the Roman World: Pompeii and Beyond. Portsmouth, RI: The Journal of Roman Archaeology, Supplementary Series Number 22: 53–68.

Scott, Sarah. 2000. Art and Society in Fourth-Century Britain: Villa Mosaics in Context. Oxford: Oxford University School of Archaeology.

Scott, Sarah. 2012. Fourth-century villas in the Coln Valley, Gloucestershire: identifying patrons and viewers. In: Stine Birk and Birte Poulsen (eds). Patrons and Viewers in Late Antiquity. Aarhus: Aarhus University Press: 189–216. DOI:

Squire, Michael. 2009. Image and Text in Graeco-Roman Antiquity. Cambridge: Cambridge University Press.

Swift, Ellen. 2009. Style and Function in Roman Decoration: Living with Objects and Interiors. Farnham: Ashgate.

Tilley, Christopher Y. 1994. A Phenomenology of Landscape: Places, Paths, and Monuments. Oxford: Berg.

Witts, Patricia. 2000. Mosaics and room function: the evidence from some fourth-century Romano-British villas. Britannia 31: 291–324. DOI:

Witts, Patricia. 2005. Mosaics in Roman Britain: Stories in Stone. Stroud: Tempus.

Yegül, Fikret K. 2010. Bathing in the Roman World. Cambridge: Cambridge University Press.