The End and Afterlife of Roman Villas in Sicily During the Late Antique, Byzantine, and Islamic Periods

The fourth century CE was the “golden age” of late antique Roman villas in many of the western provinces.¹ There are several themes related to Roman villas in the historiographical debate on Late Antiquity: The first regards the very existence, or not, of a particularly late antique type of villa within a very long tradition that had its roots in the Republican age.² Moreover, one of the most debated topics at the European level for over twenty years has been that of the crisis and the “end” of the villas as elite residences and their subsequent destiny, use, and function between Late Antiquity and the Early Middle Ages.³

New research has shown that these villas often had a longer life than previously believed, and residential and prestigious building activities also occurred in many cases during the fifth century.⁴ Nevertheless, between the end of the fifth and the first half of the sixth century, or even earlier, the villas generally came to an “end” as residential-type buildings. This did not, however, exclude their survival in other forms, through a series of reuses and transformations. In fact, similar dynamics occurred in villas across the Roman and post-Roman Mediterranean, in a historical context characterized by many transitions.

The investigation of the afterlife of Roman villas has emerged as a seminal pursuit in the scholarly effort to transcend simplistic binaries of continuity versus discontinuity, aiming instead to unravel the complex layers of postclassical landscape transformations throughout the Mediterranean. Central to this inquiry is the recognition that Roman villas were not merely elite residences but pivotal “central places” within Roman landscapes serving as tangible manifestations of the complex economic (ranging from slave-based economies to the colonate system) and cultural (encompassing the dialectics of otium and negotium) systems of the Roman Empire. These entities, intricately woven into the socioeconomic and political shifts within the entire Roman Empire, from West to East, offer a profound window into the adaptive and transformative capacities of rural landscapes. More than architectural or artistic (self) representations of Roman elites, these villas represent territorial palimpsests whose lengthy biographies of birth, life, death, and afterlife serve as fertile ground for dissecting the multifarious material and socioeconomic metamorphoses of changes in rural landscape trajectories. In the biological metaphor, the notion of the villas’ “afterlife” becomes especially salient, positioning them not merely as relics of a bygone and vanishing Roman world but as dynamic entities influencing and integrating into the new sociocultural environments of successive epochs, such as the Germanic kingdoms and territories under Byzantine or Islamic rule.

This extensive and multicultural diachronic analysis of the villas’ demise positions Sicily as an ideal setting for adopting this perspective, especially in light of growing archaeological research. The island’s unique historical and cultural trajectory offers unparalleled insights into the enduring legacy and transformation of Roman villas, situating Sicily at the heart of contemporary debates on the postclassical Mediterranean landscape.⁵

In addition to the presence of one of the most important examples of the entire category of the Roman villa, the extraordinary Villa del Casale at Piazza Armerina (Enna), as well as other structures of great interest such as the Villa of Patti Marina (Messina) and that of Caddeddi on the Tellaro River (Syracuse), a renewed season of archaeological research on postclassical Sicily has brought new data on material culture and Byzantine and Islamic rural settlements.⁶ The aim of this research has been to collect and analyze the principal data available on the last phases of the Sicilian villas as residential buildings, together with evidence related to abandonments, reuses, transformations, and recoveries over the long centuries of Late Antiquity. These efforts extend from Late Antiquity to the Byzantine age and the Islamic and Norman periods, taking into account not only the main examples but also lesser-known contexts that are useful for defining these issues (Figure 1).⁷

Our investigation delves into the intricate interplay between historical events, socioeconomic factors, and religious developments that shaped the trajectory of villa lifeways during this transformative period. In addition to providing an archaeological overview in a diachronic sense, our work focuses on several crucial issues for the history of Sicily between Late Antiquity and the Islamic-Norman era. Central to our examination are three interwoven themes that illuminate the multifaceted nature of Sicilian villa culture: the impact of seismic activity, the dynamics of economic development and landownership, and the process of Christianization. Through a meticulous synthesis of archaeological evidence and historical records, we elucidate how these threads intertwine to depict a nuanced portrait of rural life in Sicily. Starting from the last stages of the use of villas as aristocratic residences, we analyze the relationship of the constructional and residential events of the buildings with the natural disasters attested by literary sources and archaeological evidence to verify their actual impact in individual contexts. A transversal and fundamental theme for all epochs is the examination of the economic aspects related to the ownership of agricultural estates that justify the development or crisis of individual settlements; a specific focus is dedicated to the relationship between villas and Christianization, which, as we will see, seems to assume peculiar characteristics here compared to other areas of Italy and the empire. After Sicilian villas cease to be aristocratic residences, their role as central places is examined in a different socioeconomic context, from the Byzantine to the Norman period. By offering a detailed material synthesis of Sicilian villa evolution and a focused discussion on some key aspects of villa lifeways, this essay seeks to contribute to a deeper understanding of the socioeconomic dynamics that characterized rural Sicily from Late Antiquity to the Early Middle Ages.

A key context for addressing this problem is offered by the Villa del Casale in Piazza Armerina. For some time, the complex architectural plan and, above all, the extraordinary mosaic apparatus of the villa have been a central focus of scholars regarding the initial phases of its construction and the connected problem of its ownership. In the absence of rigorous and reliable archaeological data from the excavations that brought the monument to light in the 1950s, carried out by Gino Vinicio Gentili,⁸ the stylistic elements of the mosaic floors, variously evaluated over time, have influenced the chronologies proposed by scholars. The excavations by Andrea Carandini and his colleagues in 1970 provided the first reliable stratigraphic data, which places the construction of the villa in the Constantinian period.⁹ The most recent research, conducted by the Sapienza University of Rome between 2004 and 2014 under the direction of Patrizio Pensabene, has confirmed the Constantinian chronology proposed by Carandini, at least for the baths, peristyle, great corridor, and basilica. A second construction phase has also been recognized, probably subsequent to the earthquake of 365 and during the Theodosian age: It includes the hall with three apses-oval court complex, the triple-arched entrance, the buttresses of the basilica, and the remaking of mosaics and wall decorations in some rooms of the villa (Figure 2).¹⁰ An analogous sequence was also found in the construction of the so-called southern baths, brought to light by the Sapienza excavations (Figure 3): The baths, in fact, would have been built at the same time as the first nucleus of the villa but would have received a substantial renovation in the second phase, mainly attested by a new flooring of the frigidarium with a rainbow-style mosaic, similar to that of the oval courtyard, and by a panel with the name Bonifatius inserted in the mosaic of the east side of the peristyle in front of the central staircase connecting to the corridor of the Great Hunt.¹¹

The dating proposed by Patrizio Pensabene is disputed by Roger J. A. Wilson, who finds the style of the mosaics in the rooms arranged around the oval courtyard to be very similar to that of the rest of the rooms in the villa; the same can be said for the harvesting cupids in the Ambrosia mosaic in the southern apse of the triclinium, who have the characteristic V and X signs on the forehead that are found in other cupids in the villa.¹² Furthermore, according to Wilson, no decisive archaeological elements support a delayed construction of the triclinium with the oval court after the earthquake of 365 or in the Theodosian age.¹³ Thus, in his opinion, the three-apsed dining hall “was made in the course of construction, perhaps around 330, and not decades later.”¹⁴

The question of the precise construction of the various parts of the villa required a separate study.¹⁵ Nevertheless, still during the fourth century there were interventions of a firmly residential character at the villa, such as marble slab coverings put over previous frescoes, as in the hall of the main apartment; mosaic replacements, as in the case of the mosaic known as the “girls in bikinis,” which was superimposed on a geometric floor; and new fresco decorations on the south side of the peristyle, to be connected to those in the courtyard entrance and generally outside the villa (e.g., on the buttresses of the basilica)—all signs that the villa was maintained at a high level throughout the century.¹⁶ Pensabene observes different repairs and new structures in the full masonry aqueduct line, to the east of the villa, in the buttresses of the apses of the basilica and the northern halls, in the first phase of the infilling of the arches of the northern aqueduct, after the main construction of the fourth century.¹⁷ During the fifth century, on the contrary, there does not appear to be any evidence of major building interventions or additions. In Pensabene’s view, this means that the villa, or the usable part of it, was no longer inhabited by high-ranking people but rather by the actor or conductor of the estate.¹⁸ Gentili observed that the restorations of the mosaic floors, evidence of which can be found in various rooms of the villa, are evidence of its continued use as a place of representation over the centuries. In his view, the monument remained at its maximum splendor until at least the first half of the fifth century and continued to be inhabited, with some adaptation, throughout the Byzantine period and the subsequent Arab era, until the early Norman age. The pottery that was found in the lower layers attests to continuous use for at least eight centuries.¹⁹ Nevertheless, this “persistence of life” did not have the same modalities over time, and it is not yet clear when the villa lost its function as an aristocratic residence and was transformed into a different settlement form.

In the southern baths, extensive burnt layers, the collapse of the columns of the gymnasium (Figure 4), reinforcing walls with reused material leaning against the external walls, and the introduction of a vault retaining wall in the center of the frigidarium highlight structural problems that could be attributed to a catastrophic event dated to the middle or the third quarter of the fifth century.²⁰ After this event, the original function of the southern baths ceased.²¹ The same event could have caused the collapse of the columns of the main peristyle, but it cannot be ruled out that this collapse occurred later. Unfortunately, there is no excavation data: It is known, however, that the columns of the corridor of the Great Hunt collapsed later, as evidenced by the fact that they were found on a layer of abandonment on the east side of the peristyle (Figure 5).²²

Nevertheless, the villa would continue to be used for a long time. The restorations of the mosaics in the central part of the frigidarium of the western baths, although difficult to date precisely, together with data that can be obtained from the materials found in the room, demonstrate the continuity of its use up to the Byzantine period.²³ A more in-depth study of the sixth- to seventh-century phases is needed to understand how long the villa was used with representative functions, even if they were different compared to the fourth century. From the mid-sixth to the seventh century, we witness a progressive defunctionalization of the rooms, the insertion of artisanal/manufacturing structures, and the spoliation of furnishing elements.²⁴ Subsequent floods caused the progressive burial of other parts of the villa and, in the end, of the entire building after the phases of the Arab-Norman age in which some rooms continued to be used, sometimes even at the level of the mosaic floors.²⁵

Interesting data is also available for other Sicilian villas. The large peristyle villa of Patti Marina, near Tindari in the Messina area, was built in the first half of the fourth century (Figure 6); excavations in the 1970s documented the collapse of the masonry pillars, with the remains of slightly curved lintel arches in the east, south, and west sides of the porticos of the peristyle, which can be attributed to a sudden and violent seismic event, dated between the end of the fourth and the beginning of the fifth century.²⁶ The structures rested on a layer of tiles, found almost everywhere in the portico area in contact with the mosaics, which means that the villa was already in a state of abandonment at the time of its destruction. According to Giuseppe Voza, the materials from the collapse of the structures were removed and reused in later periods, above all in a Byzantine phase when a village developed in the area.²⁷

In the triconch dining room, however, there were some walls built directly on the mosaic, a sign of a division and repurposing of the space (Figure 7); in other rooms, we can observe an infilling of the entrances and other interventions preceding the collapse of the structures. The earthquake must therefore have occurred when the villa had already been abandoned, reused, and transformed. New research, launched by the University of Messina in 2015, although concentrated mainly in the baths, has made it possible to clarify the phases of the life of the villa. First of all, renovations and uses of the structures with residential characteristics were present throughout the fourth century and continued into the early to mid-fifth century.²⁸ Room 54, located southwest of the room with three apses, is characterized by a trapezoidal shape and by the fact that it follows a different orientation compared to the other rooms of the villa. This room, accessible from the hall with three apses, featured valuable geometric mosaics, demonstrating that the building maintained high housing standards during the fourth century and, perhaps, still into the first half of the fifth.²⁹ The construction of new walls and the closure of entrances would instead take place in the second half of the fifth century, part of a phenomenon of splitting the spaces into smaller units, which is well-attested in other contemporary building complexes. We can also note the practice of burying large containers even beneath the mosaic floors, a sign of the change in the use of the rooms.³⁰ Then, around the end of the sixth century, following a destructive event, some walking surfaces were raised, and the area was subsequently inhabited at least until the ninth century.

The villa of Caddeddi sul Tellaro was probably built in the third quarter of the fourth century, perhaps around 370 CE (Figure 8).³¹ According to Voza, who excavated the site in the 1970s, the villa was destroyed by fire at the end of the fourth century, with no evidence of later uses.³² Wilson, however, states that the fire should be dated to the second half of the fifth century.³³ Recent research also reveals a situation not dissimilar to that of other contexts: new structures made with reused materials, the closure of entrances to some rooms, the subdivision of rooms into smaller spaces.³⁴

In the villa of Gerace in the Enna area, not far from Piazza Armerina, six excavation campaigns were conducted under the direction of Wilson between 2013 and 2019 (Figure 9), after initial research conducted by the superintendence of Enna.³⁵ According to the results of the latest investigations, the villa was built in the second half of the fourth century, after the destruction of a preexisting horreum built around 300. A small nucleus of rooms (Area A), some of which have mosaic floors, was in fact built next to the collapsed warehouse (Area B). It is a small villa with a compact plan, perhaps built around 370–75 and remained in use for at least seventy-five years. The name of the owner is known thanks to stamps on the bricks produced in a kiln on site, located and partly excavated to the south (Area E), and to an inscription in the mosaic of the frigidarium of the balneum excavated about 50 meters to the northeast (Area D): It is Philippianus, to whom horse breeding activity can be attributed, exalted by his own name (Figure 10).³⁶ In the second half of the fifth century, a devastating fire, perhaps associated with an earthquake, seems to have destroyed the villa. Between the end of the fifth and the beginning of the sixth century, a village developed in the area to which some structures can be attributed, in particular in Area C, which continued to be frequented in the Byzantine age, until the seventh century and beyond.³⁷

The research of the Messina superintendence between 2003 and 2004 at the villa of Castroreale San Biagio, on the Tyrrhenian coast between Milazzo and Patti, allowed the specification of the phases of use of the building, whose eleven periods with subphases reach up to the modern and contemporary age.³⁸ The villa was built in the late Republican age, with development up to the middle of the first century and with a radical restructuring of the thermal complex that can be placed in the second century.³⁹ In the fourth to fifth centuries, the central part of the villa and the thermal rooms were affected by structural interventions and changes in the decorative apparatus: In a first phase (Period IX A), rooms for housing were built on the collapsed servant quarters, and three buildings were arranged on the sides of a trapezoidal courtyard; in phase IX B, there was some restoration work done on the decorative apparatus and the use of prestigious elements, such as a fountain on a foot placed in the basin of the peristyle and covered with marble slabs and a column on a plinth inserted in a room, perhaps even having structural functions; we also note the reconstruction of the frigidarium swimming pool with an apsidal basin whose walls were covered with marble slabs, while the bottom had a covering in opus sectile similar to that used for the entrance to Room 2. In the last phase of this period (IX C), a semicircular exedra was constructed in the triclinium, perhaps used as a fountain or niche for statues; other interventions were carried out on the baths and the external areas. While not dealing with an overall renovation of the building, as in the case of other villas examined so far, the late antique interventions with architectural and decorative characteristics typical of the period (presence of apses and use of opus sectile and, in general, marble coverings) show the maintenance of the residential function of the building, which was then abandoned between the end of the fifth and the beginning of the sixth century.⁴⁰ In the areas investigated in 2004, it seems that the villa suffered a violent destruction, represented by collapses upon which layers of abandonment have been deposited. In the outbuildings, the layers of destruction were obliterated by an alluvial deposit, on which some poorly preserved structures were later built.

Another interesting context is the villa in Contrada Castellito, on the edge of the Catania Plain, where archaeological investigations have recently been resumed.⁴¹ The building was built on a low hillock near the Dittaino River and has rooms with geometric mosaic decoration arranged around a peristyle dating back to the third century.⁴² The villa has an older phase that, on the basis of planimetric comparisons, can be related to the first phases of the villas of Patti Marina and Castroreale San Biagio and dates to the early imperial age.⁴³ Some renovations to the main phase with mosaics are also attested, while a traumatic event seems to mark the end of the villa as an aristocratic residence in the second half of the fourth century.⁴⁴ Later structures in opus incertum were built, and this reoccupation seems to date between the fifth and sixth centuries and was connected to the production facility constructed in the southeastern sector of the building.⁴⁵ Furthermore, in the area surrounding the site of the villa, topographic surveys have revealed traces of occupation from the beginning of the second century until the late seventh century.⁴⁶ Rosa Maria Albanese Procelli and Enrico Procelli advance the hypothesis that the villa of Castellito constituted the management center of the Massa Capitoniana, as demonstrated by its proximity to the connecting road between Catina and Agrigentum. This circumstance would underline the villa’s continuing importance during the fourth century.⁴⁷

As emerges from this brief review, the history of Roman villas in Sicily is interlaced with that of natural events to which the last stages of destruction or abandonment have often been linked. In the historical and archaeological literature of both urban and rural settlements on the island, the earthquake of 21 July 365 CE is often mentioned. It had catastrophic consequences throughout the Mediterranean, as Ammianus Marcellinus describes in detail: “On the twenty-first of July in the first consulship of Valentinian with his brother [AD 365], horrible phenomena suddenly spread through the entire extent of the world, such as are related to us neither in fable nor in truthful history. For a little after daybreak, preceded by heavy and repeated thunder and lightning, the whole of the firm and solid earth was shaken and trembled, the sea with its rolling waves was driven back and withdrew from the land.”⁴⁸ Ammianus is here clearly describing an earthquake with a tidal wave. Antonino Di Vita has found effects of this earthquake in Crete, Cyprus, Libya, and Tunisia, while in Sicily traces of destruction in archaeological levels of the second half of the fourth century have often been linked to this event.⁴⁹ According to Wilson, however, there are many reasons to doubt that the earthquake and its tsunami affected Sicily so severely. More likely, only some coastal areas were affected by this event, which in any case is difficult to trace archaeologically.⁵⁰ Rather, another earthquake—which according to Libanius brought the largest Sicilian cities to ruin, referring to the death of Julian (363 CE)—took place at a slightly earlier moment. The earthquake could have caused very serious damage, even if we might rightly question whether this event led to the destruction of all the largest Sicilian cities.⁵¹ Libanius groups earthquakes that occurred after the death of Julian and presents them as a reaction of the earth to the death of a good emperor.⁵²

There is archaeological evidence for earthquake damage in many sites in central Sicily. In Sofiana, 6 kilometers from the Villa del Casale, a thermal building connected to the activities of the mansio identified as that of Philosophiana, mentioned in the Itinerarium Antonini along the Catina–Agrigentum road, ceased to function around 360 CE, when it was being used for productive purposes.⁵³ This event was linked, like the construction of the buttresses of the Piazza Armerina villa, to the earthquake of 365; however, it is possible to think instead of another seismic event that occurred at a slightly earlier date, such as the one mentioned by Libanius, which could also have affected the warehouse of the villa of Gerace, where a collapsed roof was found that had never been rebuilt.⁵⁴ Furthermore, it cannot be excluded that the buttresses of the Villa del Casale were built because of a suspected instability of the structures and not following a seismic event.⁵⁵ In the villa/farm of Saraceno di Favara, a layer of destruction, accompanied by fire, is well dated by the presence of pottery to no earlier than 360 CE.⁵⁶ This site is a short distance from Agrigento, however, where no irrefutable signs of an earthquake have been observed; thus, according to Wilson, the destruction at Saraceno is more likely to have been caused by a fire or localized event.⁵⁷ To these testimonies, we can also add the case of the villa in Contrada Castellito in Ramacca, in the Catania area. It has been hypothesized that a traumatic event determined the end of this villa in the second half of the fourth century, and indeed, we can note traces of fire and signs of structural failure in various sectors of the building, attributable to the earthquakes of 361–65.⁵⁸

The northeastern corner of the island, in particular the sites of Tindari, Bagnoli S. Gregorio, and Patti Marina, also shows evidence of seismic activity in Late Antiquity, but Wilson argues that the event should be placed later, between the end of the fourth and the beginning of the fifth century.⁵⁹ An earthquake in the Messina and Reggio area was, in fact, hypothesized in the 360s,⁶⁰ but it is not clear whether this event can be related to the damage found in sites on the Tyrrhenian coast. If the so-called basilica of Tindari can be dated to the first half of the fifth century, its construction may have been an attempt to restore city life after a violent destruction that Luigi Bernabò Brea and Madeleine Cavalier have attributed to an earthquake that occurred, at the latest, at the beginning of the fifth century.⁶¹ Given the proximity of the Patti Marina villa, where the traces of destruction due to a seismic event are particularly evident, one could think that it was the same earthquake in the beginning of the fifth century.⁶²

The catastrophic event that caused the collapse and damage in the southern baths of the villa in Piazza Armerina has been linked to what was found in the excavation of the villa in Gerace, where an earthquake, not documented by ancient sources, caused serious damage to the bath building (Figure 11) and, to a lesser extent, to the villa. This event dates to around the middle of the fifth century for the Villa del Casale, but, according to what is known about Gerace, it could have occurred even later in the fifth century, but there is still no decisive evidence.⁶³ According to Wilson, this was the moment in which “the elite villa life as it had been known in Sicily finally came to an end.”⁶⁴ All the villas that we have examined, in fact, were destroyed or partially abandoned by the end of the fifth century. Yet in most cases life on the sites continued, though with different forms and functions.

In any case, during the fifth century, significant building interventions redesigned the layout of the Sicilian villas, as it had in other contexts, such as that of the villa of Faragola in Apulia.⁶⁵ Ancient sources attest to the large senatorial properties of the Simmachi, Nicomachi, and Valeri on the island at the end of the fourth and beginning of the fifth century,⁶⁶ and it is probable that, as the archaeological evidence reveals, the life of the main villas lasted for a few decades with limited building interventions. According to literary sources, the great owners of the senatorial class owned patrimonia sparsa per orbem,⁶⁷ and therefore they had to entrust the care of their estates to vilici, actores, procuratores, and conductores. Symmachus, for example, refers to a conductorem rei meae Siciliensis.⁶⁸ The Nicomachi also had interests on the island, as evidenced by the governorships of Volusius Venustus and Flavianus senior.⁶⁹ It is also well known that Virius Nicomachus Flavianus revised Livy’s text near Enna around 407 CE, a circumstance used previously by Biagio Pace and recently by Brigitte Steger to attribute the construction of the Piazza Armerina villa to his father.⁷⁰ It is discussed whether the possessio nimis praeclara of the Valeri mentioned in the Life of Melania could be placed in the area of the Strait of Messina,⁷¹ linking it with the one from which Rufinus allegedly witnessed the burning of Reggio by the Goths.⁷² The importance of the senatorial presence, with the flourishing of large villas, can also be understood from the fact that the road system was remodeled according to land ownership, as demonstrated by the Itinerarium Antonini, which passes through predial centers between Catina and Agrigentum (Calvisiana, Capitoniana, Philosophiana, etc.), indicated as mansiones nunc istitutae.⁷³

This is not surprising, considering that the fourth century represented the golden age of the villas due to the importance that the island had in the political dynamics of the empire, the connection to the Italian diocese in the Diocletian reorganization, and, above all, its economic value: Sicilian wheat, in fact, supplied the capital when there were problems with the supply from Africa, but it was mainly destined for free trade.⁷⁴ Even the emperors had many properties in Sicily, and, starting from the Constantinian age, a rationalis trium provinciarum is attested that supervised the patrimonium of Sicily, Sardinia, and Corsica; later, a rationalis rei privatae per Siciliam was created.⁷⁵

In terms of the Ostrogothic and Byzantine periods, it is necessary to investigate the presence of an aristocratic class in Sicily that could have still lived in, or in any case managed, the villas and the adjoining land properties: In the fifth century, changes and the presence of new owners can be noticed in the patrimonial geography of the island.⁷⁶ Lauricius, praepositus sacri cubiculi of Honorius, received some property on the island and used conductores to administer it,⁷⁷ and Pierius, comes domesticorum of Odoacer, too had been entrusted with properties in Sicily.⁷⁸ According to Elena Caliri, these are two atypical owners compared to traditional aristocratic standards, but their behavior was absolutely in line with that of classic late antique aristocratic owners.⁷⁹

In the fifth century, Sicily was at the center of political and military dynamics that created insecurity connected to Vandal raids, as demonstrated by the “special laws” issued in 440 CE: The emperor Valentinian III allowed contingents to be armed for autonomous defense and promoted tax relief for the possessores damaged by the barbaricae vastitates, due to the first incursion of Genseric.⁸⁰ However, the episode was probably circumscribed, and only in 455, when the political scenery changed, did Genseric repeatedly attack the largest islands in the Mediterranean. These were looting campaigns that took place in the spring and, while not leading to real territorial conquests, determined control of the Mediterranean and Rome’s food supply.

In Sicily, there was no permanent occupation, but sudden and violent raids hit some areas while sparing others, even nearby.⁸¹ However, not all the archaeological traces of destruction and fires in the second half of the fifth century should be attributed to Vandal actions. It is always risky to try to directly connect historical events and archaeological documentation. As Caliri notes, there is not a single source that certifies that the Vandals expropriated lands, carried out confiscations, or forced Roman aristocrats to abandon their possessions.⁸² Still, in 526 Cassiodorus underlined a certain stability in the Sicilian agricultural economy,⁸³ and, according to Wilson, “the abundance of 5th and 6th century pottery on rural sites tends to confirm the impression of Cassiodorus … that this was a peaceful time for Sicily, with agriculture booming and the population expanding.”⁸⁴ Furthermore, until the early Byzantine period, there were no substantial changes in the structure of land ownership.⁸⁵ The destructive events that marked the end of the villas at different times in the fifth century could, therefore, have had different causes and cemented processes that had been going on for some time. This did not mean the end of large property holdings but rather a crisis for the lifestyle that made luxury villas not only centers for the management of agricultural production and the accumulation of products but also stages for competition between elites.⁸⁶ In this perspective, for example, the Villa del Casale would have been transformed into the home of the staff who were in charge of monitoring agricultural activities; gradually, in Byzantine times, some production activities moved inside the villa.⁸⁷ Even in lower-ranking estates, such as the one in Contrada Castellito in Ramacca, residential spaces lost their original function and were repurposed for productive installations between the fifth and sixth centuries.⁸⁸

One of the most common ways Roman villas were reused, already from the end of the fourth and beginning of the fifth centuries, is indicated by the presence of buildings for Christian worship.⁸⁹ The difficulty lies in determining whether these buildings were constructed when the villas were still in use as residential buildings, and thus repurposed with the intervention of the owners themselves, or only once the buildings had been abandoned as elite residences.⁹⁰ The theme fits into the broader picture of the Christianization of the countryside and the development of ecclesiastical property. In Sicily, there are very few traces of Christian presence in residential buildings, and there are no confirmed churches in the area of villa sites, with very rare and controversial exceptions.

In the Piazza Armerina villa, there are only a few archaeological materials relating to a Christian presence. The early Christian and Byzantine oil lamps that Gentili recovered in large numbers during the excavations and, in particular, in the northern and southern pools of the frigidarium of the baths (Figure 12) could support the hypothesis that, at some time during the Byzantine period after the cessation of secular use, the octagonal space, which echoes the central plan design of a circular basilica or baptistery, was repurposed as a worship site.⁹¹ In the catalogue of Gentili’s oil lamps, drawn up by Daniela Patti, the specimens found in the baths are dated to a period between the second half of the fourth and sixth centuries.⁹² There are no other elements to indicate a religious use of the room, even though the remaking of the mosaics demonstrates an inhabitation that lasted over time, more than in other parts of the villa. The oil lamps with Christian symbols cannot be considered in themselves a certain indication of the transformation of this space into a place of Christian worship. The presence of many lamps in the frigidarium in the Piazza Armerina villa could therefore suggest that the hall, having ceased its primary function as a thermal room, could have become a meeting place, perhaps because this part of the villa was better preserved than others, which had been destroyed or already covered by alluvial layers, or that it had been used as a storage place for reusable materials.

The existence of a Christian building near the villa has been hypothesized due to the out-of-context discovery of a perforated brick with crosses and Christian symbols in the layers of the southern baths, dating back to the sixth century, but no other elements have been found to support the hypothesis.⁹³

Wilson hypothesized that a large room with an apse located behind the hall with three apses of the Patti Marina villa could have been used as a hall for Christian worship, but this hypothesis is not supported by other evidence;⁹⁴ it could instead be a room built by the owner in a second phase with representative functions.⁹⁵ Unfortunately, however, the room has not been excavated down to floor level and cannot currently be explored because it is located on private property.

A building for Christian worship has been recognized on the site of Contrada Saraceno in Favara, near Agrigento, in a villa of the imperial age articulated on a courtyard peristyle, with a thermal structure and a rustic sector. A fourth-century building phase did not modify the previous structure.⁹⁶ After its abandonment following a destruction event in the mid-fourth century, some phases followed with productive purposes. An apsidal room in which fragments of glass chalices and an iron cross were found has been recognized as a chapel of Christian worship from the first half of the sixth century. Finally, in the villa of Castronovo di Sicilia, contrada San Luca, abandoned as early as the fourth century, an apsidal building was built in the sixth century; it can be interpreted either as a place of metallurgical processing for the recovery of ferrous slag or, due to its planimetric characteristics, as a church.⁹⁷

This evidence forms part of a very lively and complex picture relating to the diffusion of Christianity in Sicily, supported by numerous historical and archaeological documents.⁹⁸ However, as observed by Mariarita Sgarlata, it is difficult to find an adequate correspondence between literary and epigraphic sources and archaeological evidence. For example, the epigraph of Aithales, found in Catacomb A of the necropolis of the Trippiedi district in Modica and dated to 396 CE, recalls the construction of a church and a cemetery in Hortesiana by the deceased. Aithales could perhaps be connected to the construction of the church of S. Foca, which, according to the brick documents found in the area, should fall within the massa Hortesiana.⁹⁹

Other documents reveal the activity of private individuals for the construction of ecclesiastical buildings.¹⁰⁰ Even in Sicily, therefore, private owners would join ecclesiastical ones in the construction of ecclesiastical buildings until the end of the sixth century. For example, Gregory the Great asked the bishop of Tindari Benenatus to consecrate the oratory built for his devotion by the religious femina Ianuaria.¹⁰¹ However, the role of the bishops would have been preponderant, and this could perhaps explain the characteristics of the archaeological documentation and the absence of churches in the villas or above the villas. The Sicily of the large private property holdings of the fourth and fifth centuries appears to be essentially pagan while, as Caliri states, ecclesiastical property is the true protagonist of the Sicilian landscape of the sixth century.¹⁰² In addition to the rare and controversial examples of religious buildings that would be connected to villa sites in the early Byzantine period, it should finally be noted that the most ancient rural churches in Sicily¹⁰³—dated to the late fourth to fifth centuries—are found in villages, such as Kaukana,¹⁰⁴ Sofiana/Philosophiana,¹⁰⁵ Eraclea Minoa,¹⁰⁶ or the site of San Miceli near Salemi in the province of Trapani.¹⁰⁷ In this period, in fact, the villages became important landmarks in the countryside, with warehouses, manufacturing facilities, houses, and churches.¹⁰⁸

Settlements that can be defined as villages, agro-towns or agglomérations secondaires, or sites that maintained a quasi-urban layout (as in the cases of Sofiana/Philosophiana,¹⁰⁹ Casale San Pietro in Castronovo di Sicilia,¹¹⁰ Carini/Hyccara,¹¹¹ and maybe Contrada Muratore in Castellana Sicula),¹¹² often in continuity with the stopping points of the late Roman road network, replaced the villas as central places in the settlement network from the fifth century onward,¹¹³ especially during the early Byzantine period in Sicily (sixth and seventh centuries).¹¹⁴ In this context, the case of villas from the high imperial era is significant, having already, from the mid-fifth century, seen the development of rural villages such as Contrada Cignana and Contrada Saraceno in the territory of Agrigento, which were occupied at least until the seventh century (Figure 13).¹¹⁵

In this change in the hierarchical role between villas and villages, one can read the reflection of the new agrarian property structure in Sicily, which saw the centrality of massae fundorum, an agrarian configuration well documented from the age of Constantine, but with certainly older origins and significant development between Late Antiquity and the early Middle Ages, as demonstrated in the foundational studies of Domenico Vera.¹¹⁶ Thanks primarily to the correspondence of Gregory the Great, it is known that the massa was the “place of registration for fiscal purposes, determining the origo of its residents and their connection to the land,”¹¹⁷ and that the conduma was the physical place within it where the production tools and labor force were concentrated.¹¹⁸ Therefore, between the sixth and seventh centuries, the conduma took on an active role in the civil and fiscal classification of the peasant population residing in areas far from inhabited centers, often connected to the large ecclesiastical properties belonging to the patrimonium of the churches of Rome and Ravenna.¹¹⁹ In a letter dated July 592, Pope Gregory the Great addressed Peter,¹²⁰ the subdeacon and administrator of the Syracuse estate, mentioning the presence of four hundred conductores overseeing an equal number of stud farms. These farms were responsible for redistributing mares for breeding purposes, except in cases involving a large sale of the church’s horse herds on the island.¹²¹

In light of these changes in agrarian property structures, it is possible to reinterpret the phenomenon of the “end of the villas” in Sicily,¹²² with the loss of their role as residences of wealthy owners in favor of a different, production vocation with indirect management by lower-ranking actores or conductores.¹²³ The discovery of Byzantine glass weights and lead seals in the Villa del Casale in Piazza Armerina, one of which is related to a “Gregoras ex eparco” and the other to a “Dionysos,” could confirm the existence of new tenants responsible for property management and in contact with Byzantine administrative officials.¹²⁴ Materially, the basilica was occupied by burials, while to the north of the peristyle, a furnace was installed, possibly connected to some Byzantine grooved tiles from the sixth or seventh century (Figure 14). At Patti Marina, between the late sixth and early decades of the seventh centuries, the floor levels were raised in a significant part of the villa. In the thermal sector, a large room with two central pillars was built, around which about ten pit-type burials were arranged, with dry-stone perimeters or perimeter walls and flat stone roofing (Figure 15).¹²⁵

After a dramatic event (an earthquake around 475 CE), a new phase of occupation also occurred in the villa of Gerace, starting in the sixth century and lasting until at least the first half of the seventh century.¹²⁶ In various phases of this Byzantine period, new structures were built with masonry in different construction techniques compared to the Roman period and characterized by the presence of tiles with combed or finger-impressed decorations, including a building with an external paved courtyard (Figure 15).¹²⁷

Another interesting element in the formation of new landscapes “after the villas” is the phenomenon of the transfer of land rights from the late antique senatorial aristocracy to ecclesiastical entities. It is necessary, therefore, to investigate the transition from villa to monastery in Sicily, as in the cases of Contrada Favarotta and Contrada Grammena in the province of Catania, for which the presence of a Byzantine monastery has been hypothesized in superimposition on late Roman rural structures.¹²⁸ An occupation from the Byzantine era in the area of a probable villa also appears to be documented in the site of “Conventazzo” in San Pietro di Deca, Torrenova (Messina). It is not clear whether an apsed wall associated with burials, in which a pitcher dating back to the sixth or seventh century was found, represents evidence of an early Christian reuse of the site. The site also yielded a coin from the time of Michael II. In the Norman era, it was occupied by a large new church associated with a monastic complex established during the twelfth century.¹²⁹

The establishment of the Byzantine thema Sikelias between the late seventh and early eighth centuries represented the point of no return for the decline of the landscapes of the “long Late Antiquity” and the formation of new early medieval settlements.¹³⁰ A decisive factor was the militarization of the island, which witnessed the planned construction of a network of fortifications in both urban and rural areas. The most striking example of this is the site of Monte Kassar in Castronovo di Sicilia (Palermo),¹³¹ where the sheer scale of the work leaves no doubt about the intervention of the Byzantine state.¹³²

In recent years, thanks to knowledge about ceramic indicators from the eighth and ninth centuries (especially for eastern Sicily),¹³³ it has been possible to identify a continuity of occupation of villa sites during the so-called thematic era. In the southern part of the Villa del Casale in Piazza Armerina, a large fortification wall, possibly equipped with an outer wall, built with irregular blocks embedded with tiles, both combed and vacuolati (with vegetal inclusions in their material), has been dated to the eighth century.¹³⁴ It is possible that the sealing of the aqueduct arches of the villa (Figure 16) also belongs to this phase of general “fortification” of the site, which seems to have been continuously occupied until the ninth century, as evidenced by specific ceramic indicators.¹³⁵

In the villa of Patti Marina (Messina), recent excavations have partially investigated four rooms belonging to a large-sized building (3.1 × 5.0 m; 6.5 × 4.5 m; 1 × 5 m; 7.0 × 5.8 m) south of the thermal complex, and the external open areas were paved with stone slabs.¹³⁶ Their collapse was characterized by the presence of semilunar tiles impressed with a wavy pattern, while in the occupation layers, fragments of ceramic forms typical of the eighth and ninth centuries were found (lamps a ciabatta, vetrina pesante pottery of the Petal ware types, Byzantine-type amphorae, others produced in the Phlegraean area, and cooking wares including ollae, pots, and casseroles). In other villas in the province of Trapani, such as that of Contrada Mirabile (Mazara del Vallo), some transformations were already occurring in the middle of the fifth century, but with a continuity in occupation (only indicated by pottery findings) until the end of the sixth to the early seventh centuries.¹³⁷ Another villa in the Genna area in the Mazara hinterland was occupied until the sixth century, but starting from the seventh century and perhaps up to the ninth century, the area of occupation contracted.¹³⁸

The Islamic period marks the final stage in the long history of Sicilian “villas after villas.” Although there is still no significant evidence for the initial phase of the Islamic occupation of the territory during the conquest period in much of the ninth century, phases of continued use of the area of the ancient villas are attested during the tenth and eleventh centuries, and sometimes even during the Norman era.¹³⁹ We have already discussed the importance of the major villages along the road network, and it is worth noting the evidence for their long-term continuity, even as new settlements emerged, especially from the mid-tenth century.

In this context, we want to focus on the phenomenon of the persistence, even between the tenth and eleventh centuries, of clustered, open settlements on valley floors, in continuity with the topographical basins—that is, spaces suited for long-term human occupation—of the villas. With the Villa del Casale of Piazza Armerina, for instance, a settlement began to develop in the mid-tenth century that occupied the entire area of the residential part of the villa,¹⁴⁰ and it expanded to the recently investigated southern and northern areas.¹⁴¹ A complex settlement with rectangular rooms arranged around central courtyards emerged, seeming to have a functional vocation related to agriculture and animal husbandry, as evidenced by the discovery of agricultural tools as well as the archaeobotanical and zooarchaeological records.¹⁴² Important elements regarding the long continuity of exploitation of the agricultural production basin around the villa are also provided by various grain storage structures in large underground pits.¹⁴³

After the first Islamic phase, the life of the settlement continued with two other sequences of stone buildings in the twelfth and late twelfth-thirteenth centuries (Figure 17). From a cultural and religious perspective, traces have been found that confirm the presence of a Muslim population in the settlement. These include the discovery of amulets/plaques with apotropaic functions bearing Kufic inscriptions,¹⁴⁴ as well as the identification of Islamic burials with a right lateral decubitus position and the head facing southeast, found in the area to the north of the villa and dating back to the second half of the eleventh century (Figure 18).¹⁴⁵ Another important point that also reveals changes in agricultural techniques related to the increased use of irrigation during the Islamic age, probably related to the spread of the so-called Islamic Green Revolution,¹⁴⁶ is given by the finding in an early medieval village of Piazza Armerina of senia pots (ceramic buckets) used in water-wheel machines formed by a gear wheel moved by an animal (often a donkey) for the irrigation of fields by drawing water from rectangular wells and accumulating it in tanks.¹⁴⁷

A reevaluation of the limited excavation documentation of the Tellaro villa in Contrada Caddeddi has allowed the discovery of about thirty Islamic ritual burials found throughout the eastern area of the complex,¹⁴⁸ without following a specific arrangement, although some appear to form small groups, possibly linked to particular families. The individuals, buried in simple earthen graves, were laid to rest in a right lateral decubitus position with their upper and lower limbs slightly flexed, and their skulls oriented with their faces to the southeast. The absolute chronology of these burials is uncertain in the absence of stratigraphic contexts and radiocarbon dating, but the discovery of materials dating to the late tenth and early eleventh centuries inside a silo located approximately 100 meters from the cemetery area could provide a rough dating. These findings, therefore, indicate a continuity in the topographical and territorial basin of the villas even during the full Islamic period until the arrival of the Normans, suggesting the persistence of population centers in those territories and, more importantly, the maintenance of a certain connection between the settled areas and the agricultural and silvopastoral potential of the landscapes. A recent excavation in Contrada Cifali (Chiaramonte Gulfi, Ragusa) uncovered Islamic burial sites within the area of a site that had been occupied for a long time and where a villa (or a statio) developed.¹⁴⁹ Part of the thermal sector has been uncovered, indicating a very long period of occupation extending up to the thirteenth century.¹⁵⁰

The topographical continuity between Roman / late antique rural settlements classified as farms and rustic villas is also documented in other contexts on the island. For example, at a distance of about 500 meters from the San Luca villa (Lercara Friddi), a concentration of medieval materials (from the tenth and twelfth or thirteenth centuries) has been identified through surface surveys, suggesting the development of an open settlement in the same “topographical basin” as the Roman settlement.¹⁵¹ A similar situation appears to be the case in the area of Islamic-era fragments recently identified not far from the Roman site of Contrada Zuccarone (Corleone) (Figure 19).¹⁵² Occupational continuity during one phase dated to the ninth and tenth centuries and a final one between the eleventh and the first half of thirteenth centuries have been detected in the last phases of a villa site (already transformed during the Byzantine period) in Contrada Saraceno (Favara),¹⁵³ and the development of an agricultural settlement (a casale) between the eleventh and early thirteenth centuries, associated with an Islamic-rite cemetery, may confirm the long occupation of a late antique rural site—only indicated by the presence of pottery recovered during the excavation—in Contrada Caliata (Montevago).¹⁵⁴ From survey data and a small survey, the development of open settlements like the casale has been documented in the area of a late Roman villa, occupied until the first half of the seventh century at the site of Casale Nuovo (Mazara del Vallo).¹⁵⁵ This is where a half-solidus gold coin from the Syracuse mint, dated to 866–67 CE, was found. Based on the discovery of ceramics from the second half of the tenth to eleventh centuries, it could also be related to the Casal Bizir, mentioned in a document from 1093 regarding the foundation of the bishopric of Mazara.¹⁵⁶ Also, in the territory of Contessa Entellina, a medieval casale (maybe the site of al-Ḥammām, mentioned in a Norman jarīda of 1182) was developed on the site of a villa occupied until the end of the fifth century at Case Vaccara.¹⁵⁷ In the villa of Santa Marina (first and second centuries), located in the Pellizzara district in the area of Petralia Soprana, there are reports of medieval burials (not Islamic rite) radiocarbon dated to around 1000, while sporadic sherds of glazed pottery (from the twelfth and thirteenth centuries) and walls that overlap with the Roman age layers indicate occupational continuity during the Middle Ages.¹⁵⁸

Our investigation of post-Roman settlement in the Sicilian countryside demonstrates that the villas on the island persisted as crucial palimpsests that shed light on various facets of late antique and early medieval settlement dynamics. Perhaps most notably, we have observed from current documentation the lack of investment by elites in decorative elements or new construction works during the fifth century. Nonetheless, despite the departure from the conventional villa model, the initiation of various transformative processes, diverging from the luxuries of the fourth century and showcasing resilience to earthquakes, led certain sites to maintain their centrality. This centrality persisted within a transformed social and economic framework. Concurrently, there was a notable rise in the role of secondary settlements along the road network. The metamorphosis of the villas reflects the culmination of a shift in management practices and the culture of self-representation of the elites during the fourth century.

Changes in ownership are pivotal for comprehending certain transformations, even if archaeology cannot fully resolve the many questions regarding this highly complex topic. Written sources highlight the Church’s emergence as a key player from the sixth century onward, managing centers (condumae) that, in some instances, may precisely correspond to postvilla sites where new conductores resided. These sites likely served as hubs for coordinating livestock, tools, and manpower.

Regarding the topic of the Christianization of the villas, there is at present limited evidence supporting a transition from villa to church. Based on the overall picture of the archaeological evidence known to date, it appears that the activity of building churches in villa complexes was not a widespread phenomenon in Sicily. Instead, the construction of an ecclesiastical network at key road junctions seems more prevalent, following a deliberate strategy most probably driven by ecclesiastical powers. This strategy witnessed the establishment of rural churches in strategic mansiones/stationes along the road network. We have argued for shifting the focus from the villa-in-itself, characterized by a building complex representing a homogeneous construction culture and the influence of the late Roman elite, to the villa-estate, or a topographical basin involving relationships with the exploitation of the territory and exchange network. This shift allows for a more detailed understanding of the prolonged occupation of many contexts.

The villa sites retained their centrality throughout the Byzantine age, often serving as the “topographic basin” for subsequent developments. These developments occurred either in direct continuity or with minor spatial shifts, leading to the establishment of new open settlements. Remarkably, some of these settlements remained active during the Islamic period and, in some instances, persisted up to the Swabian age.

Due to its complex and stratified history, and the continuous contributions of new populations and cultures over the centuries, Sicily confirms itself as a particularly important region for the study of the archaeology of villas in the Mediterranean. The study of villas over a long-term perspective in contexts like Sicily, where significant political regime changes occurred (Goths-Byzantines-Arabs) along with cultural-religious shifts (from paganism to Christianity, from Christianity to Islam), can introduce new elements to the debate on the extended period of Late Antiquity and the interactions between different chronological and disciplinary fields, which often remain separate. The proposed general framework, therefore, positions the end of the villas in Sicily as still very fertile ground for future comparative research between the western and eastern parts of the Roman Empire, especially in evaluating cultural-religious phenomena such as Christianization and Islamization, political topics such as adaptation to the new spheres of power of Byzantium and Islam, or environmental issues such as the ability to adapt and react to traumatic events like earthquakes. Further microregional and interdisciplinary approaches, combined with comparative perspectives, will be fundamental for rethinking and reframing the phenomenon of the end and afterlife of Roman villas in the Mediterranean in a wider holistic perspective.