The villa lies in a rural district called Caddeddi or Vaddeddi on the south bank of the 45km-long Tellaro river, about 3km from its mouth on the Mediterranean. Locally it is called the ‘Villa del Tellaro’, but I prefer to use the name of the locality, since no doubt other villas await discovery along this lush and pleasant river valley. It is not a new find. A substantial 18th- and 19th-century farm stands on the spot, and it seems that its owners were already indulging in a little surreptitious digging themselves as early as the 1950s. A local guidebook, presumably not much circulated, refers in 1962 to Roman walls and ‘traces of mosaic’, revised in its 1970 edition to read ‘ample traces of mosaic’. That winter, the archaeological authorities in Syracuse intervened, and started expropriation proceedings. Three figured mosaics, already in part uncovered by the owners inside the east wing of their farm, were raised in 1971 and taken to Syracuse for restoration. Part of the farm was demolished, and a temporary shed was put over the Roman walling of the rooms where the figured mosaics had been found. Also within the shed was the mosaic in the corridor (7 on the villa plan, see p.18) that formed the north side of the villa’s central peristyle – an ornate courtyard – but that pavement has never been lifted and remains in situ. For decades it was hidden by sand. It has a wonderfully intricate geometric pattern of interlaced laurel-wreath medallions, featuring two different types of elaborate rosette, each alternating between one circular medallion and the next. The octagons between the medallions contain an awning pattern (looking a little like a modern umbrella, seen from below). The figured mosaics were not seen in public until 2003, over 30 years after their removal from the site, when they were put on temporary show in a church in Noto. They were then returned to the site when a new cover building was finished, and the villa finally opened to the public in 2008.
A grand design
So what do we know now about the Caddeddi villa? Excavation during the 1970s and 1980s has revealed most of its ground plan, but since part of it lies beneath the farm, and the farm buildings themselves are worthy of preservation as an example of Sicilian vernacular architecture (and could potentially serve as a convenient site museum), there are inevitably still some gaps. The main block forms a compact square about 60m by 60m, with rooms arranged around a central courtyard garden bordered by corridors on all four sides (7, 23-26, and another on the east). The main show rooms, as we shall see, are likely to lie on the north side, overlooking the river. There was a small apsed room in the middle of the south side (6), perhaps a reception room or a small dining room. Only part of its geometric mosaic survived, but, as it lay inches below the main entrance into the later farm, it was a miracle that any of it was preserved at all. The southern part of the west wing, ending in an apse (27), is lost below the farm. So we do not know, for example, if there was a bath-suite here (the apse could have accommodated a hot-water pool, perhaps), or whether the baths were in a totally separate building nearby, still awaiting discovery. Our ignorance of the whereabouts of this important facility (a sine qua non for any grand rural mansion in the Roman world) is one of many unanswered questions about villa Caddeddi.
The plan is slightly misleading, in that the Roman walls marked here are not all at the same level. The peristyle and the central parts of the west and south wings are on the highest, flat part of the site, also occupied by the later farm buildings. But the ground then falls away sharply on the north and the east, and what has been excavated here are the walls of rooms that formed the substructures of the villa, used for stores and service quarters. The upper rooms in this part, belonging to what we might call the piano nobile, are totally lost. So the splendid mosaic pavements of the north corridor (7), and in rooms 9 and 10, on flat ground at the higher level, survive in a good state of preservation, but many further floors that must have once paved chambers at the same level to the north and east of them have completely disappeared. A few fragments of mosaic were found collapsed into the basement rooms, but these have not so far been made available for systematic study. One day it might be possible to reconstruct from such pieces some of the floor designs of these lost upper chambers.
So rooms 1-5, 8A, and 11-19, as marked on the plan (p.18), belong to the substructures of the villa, and while they hint at the likely position and size of some of the chambers above, we cannot be certain that the ground plans of the two floors matched one another precisely. For example, substructures 8, 8A and 11 are likely to reflect a single apsed hall (aula) at the upper level, the largest room in the villa (11m by 8m), and one wonders whether there was also just a single corridor or loggia, with splendid views over the River Tellaro, running the full width of the building over substructures 12-18. What we do know is that the entrance to the villa was at the north-east corner, convenient for access from the major Roman road along the east coast of Sicily, which passed a short distance to the east of the villa. The entrance was rather grand: a flight of seven broad steps in white marble led the visitor up to a small vestibule (1) with a simple lozenge-patterned mosaic floor; but how that vestibule connected with the rest of the rooms at the upper level is unknown.
The villa belongs to the second half of the 4th century. This seems clear from a small hoard of coins found under a floor level in a room off the south-east corner of the peristyle, the latest of AD 348, and from another of the same date reported as coming from under the mosaic in room 8. So the villa and its mosaics are likely to be after c.AD 350. The question is how long after. In July 365, a cataclysmic earthquake off Crete set off a vast tsunami: a contemporary writer, Jerome (the later saint), speaks of Sicily (presumably its east coast) and its offshore islands (the Aeolian Isles?) being overwhelmed, and of countless people losing their lives. One is reminded of the tragic, devastating effects over a vast area of the Boxing Day tsunami in 2004. We do know that there was an earlier villa on the site: some walls belonging to it on a different alignment are known under rooms 10 and 19 (see plan on p.18). Caddeddi lies on a low shelf above the wide river valley close to the sea – so could the earlier villa have been overwhelmed by the tsunami of 365, on its still-destructive course penetrating westwards inland? This remains at present no more than an intriguing hypothesis. If supported by future work at the site, the late Roman villa that we see today may be shown to have been a post-tsunami replacement for the earlier one, and constructed sometime around AD 370/375. Establishing a more precise date for the villa must be high on the agenda for any future research programme.
‘Ample traces of mosaic’
Let us turn now to the three figured mosaics that are the great glory of the Caddeddi villa today. The first (in room 8) is only a fragment, but that any of it survives at all is another miracle. Farmhouse walls, later demolished, were resting right on top of it: the oblique gash in the panel that we see now marks the line of one of those walls. The scene shows an episode from the Trojan Wars (in a variant version, later than Homer), when Achilles agrees to release the body of Hector on payment by King Priam of the equivalent in gold of the weight of Hector’s corpse. A large weighing scale dominates the scene, with the ransomed gold vessels in the pan at lower left; but at lower right only the legs of Hector survive, lying on what would have been another pan. To the left of the scales’ tripod stand Odysseus, Achilles (with peacock feather helmet), and Diomedes, while to the right of it are one of Priam’s sons (probably) and Priam himself, of whom only a single arm survives. This is gesturing at the gold, balancing Achilles’ arm, which points to Hector’s body.
It seems likely that one further figure is missing from the scene, and that this was originally a square panel with just six figures, three Greeks and three Trojans, with the scales’ tripod at dead centre. A Greek inscription at the top aids identification. Careful measurement of the panel and the width of the two surrounding borders (the inner one surrounding each panel, and the outer one going round the walls of the whole room), as well as of the dimensions of room 8 itself, shows that there must have been 12 square figured panels adorning this magnificent room, perhaps all containing scenes from the Trojan War, or alternatively different episodes from the varied and eventful life of Achilles. The room must have been the grand showpiece of the villa, probably a banquet hall and the main reception room. The subject matter was no doubt chosen to highlight the literary erudition (or at least pretentions) of the villa’s owner.
The small room to the west (9), 4.80m by 4.70m, has a complex design consisting of a central square (with concave sides) and four festoons of fruit and flowers within a matrix of laurel leaves; these spring from large bowls (craters) at each corner. Overflowing from these are four different kinds of fruit – peaches, pears, pomegranates, and medlars. The central tableau carried a bust of Bacchus, god of wine, but only a shoulder and part of his accompanying thyrsus survives. In each of the four rectangular panels along the sides, a satyr wearing a fawn skin (nebris) attempts, not very successfully, to win the affections of a maenad. The craters are depicted in a rich brown suggestive of gold, and the whole floor celebrates the joys of wine and other fruits of the earth, perhaps in tribute to the local estate’s amazing fertility and productivity. The unusual X-pattern of the floor with the central panel reflects the powerful lines of a cross-vaulted roof, here transferred to make a decorative pattern underfoot.
The tour de force at Caddeddi is the stunning mosaic featuring a hunt scene, 6.40m by 6.25m, which adorns the adjacent room 10. There is a little damage near the middle and a strip along the north side is missing, but otherwise the pavement is well preserved. The flowing action-packed scenes are knitted together in a tight overall scheme with minimal use of register lines. The composition is dominated at the centre by a large seated female figure, clearly a personification. At top right is a scene showing the capture of leopards, lured into a baited box. Armed soldiers are on hand in case anything goes wrong, while one unlucky recruit does trapdoor duty on the box, ready to slam it shut once the leopard is inside. At top left chaos has broken out: in a now fragmentary scene, a lion and two leopards are attacking a lioness. A concertina fence lies flattened in the foreground, while, below it, two soldiers (one an archer and the other armed with a club) are about to put a violent end to the feline mêlée. To the right of this pair, a lion – feasting on an addax (a type of antelope) – is being disturbed by a stealthily advancing armed soldier. Below and to the left, alongside the personification, another soldier – who has fallen to the ground, terror in his eyes – is about to be savaged by an enraged tigress, while a now-headless companion arrives in the nick of time to thrust a spear into the feline’s back. A frightened horse, the mount of the fallen soldier, bolts from the scene.
In the middle of the right-hand part of the mosaic is a charming scene showing a cart that carries a box containing one of the captured animals across a shallow stream. One man alongside is concerned about the wheels becoming stuck in the mud, and another with a whip is spurring the oxen on to greater efforts. Two dogs are also in attendance. Above, a horseman urges on his mount while another, who has already reached dry land, has dismounted and is dragging his recalcitrant horse out of the stream. Three officers in charge of the hunt watch over the proceedings to the left of the ox-driver. All wear gold crossbow brooches as a sign of their rank, but the man at the centre with the mushroom-topped staff (baculus) is clearly the senior one.
Along the bottom of the mosaic, under a generous awning spread between the trees, a picnic is taking place, with six diners reclining on a semicircular bolster. Or rather there were six, but only the ghostly arm of the third diner from the left survives. Clearly damage to the mosaic at some later period (in the 5th century?) needed repair, but the mosaicist then on hand was incapable of composing figured work, and so filled the gap crudely with plain tesserae (apart from the bolster net). A roast bird (a duck?) lies on an oval silver platter before them, along with pieces of bread. A servant, who is not having a good day, pours water over the hands of the diner at the far right, while at far left a glass cup is being filled with red wine and then water (a dilution in accordance with normal Roman custom). The differing shapes of the silver jugs used for the wine and the water are both matched in contemporary surviving silverware; the Caddeddi mosaic helps determine their respective functions.
At bottom left one attendant is gutting a stag attached to a tree (the dogs, both with tongues out, think it is for them), and another man attends to an authepsa (a portable water heater), with a powerful flame being emitted from it, alarmingly close to a tree (in fact, the mosaicist has not understood how these objects function). At bottom right six frisky horses pull at their tethers, anxious for the meal to be over and to return to action. The elaborate border has 3D swastika-meanders alternating with oblong panels containing birds. Two hen and two cock pheasants face the entrance threshold, and along each side are a wild pigeon, a duck, a Barbary partridge, and a rock dove, with the male of each species depicted on the left (west) side of the floor, the female on the right. A white goose stood in each corner panel.
The action is envisaged as taking place in Mauretania (Morocco and the western part of modern Algeria), an important source in antiquity of leopards, and the tight curly ringlets of the seated figure at the centre denote her as a personification of Mauretania. Many of the details of this hunt are matched by similar compositions at the Piazza Armerina villa: even the little ‘v’ on Mauretania’s forehead (a trademark signature of a particular mosaic workshop?) also occurs on many of the cupids depicted at the more famous villa. A comparison of one scene (the lion devouring an addax) represented at both villas shows clear stylistic differences, as one would expect if Caddeddi is a generation or two later than Piazza Armerina. Certainly, details like the silver jugs and the oval platter with beaded rim in the picnic scene, and the introduction of the bow-fronted (‘steppe-type’) saddle (replacing the four corner-horns of earlier Roman type) are all in keeping with a date for Caddeddi in the second half of the 4th century. But for all its restless, frenetic action, packed with detail (areas of white tesserae background are kept to a minimum), it would be wrong to treat the composition as a factually accurate and realistic portrayal of the hunting and rounding up of wild animals for the games. Lots of ‘mistakes’ occur: to take the most glaringly obvious one, tigers are not and were not to be found in Mauretania or indeed anywhere else in Africa, although some ancient sources thought they existed in deepest Ethiopia. But the presence of fantastical elements and inaccurate details does not matter: the aim of the master mosaicist was surely to create a vivid, colourful, overall impression of all the thrills and spills of the hunt, and in this, I think, he has succeeded brilliantly.
The comparison with Piazza Armerina is important for another reason. So close are the parallels for many of the compositions there, both figured and geometric, to floors in North Africa, especially at Carthage, that it has long been accepted that the pavements at Piazza Armerina were laid by mosaicists invited to Sicily from overseas, from what is now Tunisia. For Caddeddi, too, elements in all three of the figured mosaics have close parallels in Carthage: to take just one example, a mosaic (c.AD 400) from there depicting the Months and the Seasons, on display in the British Museum, has the same type of craters with S-shaped handles and pelta-shaped feet that occur in room 9 at Caddeddi, and the same type of banded altars or bases too (as in Caddeddi’s satyr and nymph panels, see p.16). It seems therefore highly likely that the owners of both the Piazza Armerina and Caddeddi villas, at two different moments in the course of the 4th century, found the standards of local Sicilian mosaic craftsmanship wanting, and so summoned mosaicists from Carthage to fulfil their ambitious plans for the interior décor of their respective country mansions. At Caddeddi there must surely have once been far more floors for the African team to lay than those that survive now (or have been discovered to date), since three floors and a corridor would not have been enough to make an overseas commission worthwhile.
We know nothing about the owner, and not a single small find has so far been published. A tile kiln (28) and one outbuilding (29) have been identified, but none of the agricultural structures assumed to lie nearby. The villa perished in a fire (burn marks on the mosaics from falling roof timbers are very obvious), probably sometime c.AD 450/500. An early medieval settlement of c.950/1120 close by is implied by over 30 burials in the Muslim rite (inhumations on their sides, facing Mecca) uncovered to the east and south of the villa. For the Roman period, a programme of geophysical survey currently under way, directed by Davide Tanasi of the University of South Florida, can be expected to yield fresh results. Villa Caddeddi, with its spectacular mosaics, still has, surely, more secrets to reveal.
R J A Wilson, Caddeddi on the Tellaro: a late Roman villa in Sicily and its mosaics (Bulletin Antieke Beschaving Supplement 28), Leuven, Paris, and Bristol, CT: Peeters, 2016.
The author would like to thank Lorenzo Guzzardi for his help and friendly collaboration. All photographs appear by kind permission of the Regione Siciliana, Assessorato dei Beni Culturali e Identità Siciliana.
ALL IMAGES: R J A Wilson, unless otherwise stated.