In 2006, the Institute of Fine Arts, in collaboration first with the Superintendency of Trapani and later the newly established Archaeological Park of Selinunte, started a new project of topographical, architectural, and archaeological investigation of the main urban sanctuary on the Akropolis. This was one of the largest public areas of the city, best known for its two large peripteral temples of the second half of the sixth century: Temple C, dedicated to Apollo, well-known for its large pedimental gorgoneion and relief metopes featuring divine figures and heroic feats; and Temple D, dedicated to Athena. These large temples, their altars, and the stoas along the eastern and southern edges, mark this sanctuary as one of the main spaces for public cultic and ritual activities before the Carthaginian capture in 409 BCE. In the third quarter of the fourth century, with a new development of Selinus within a general recovery of Sicily, the concentration of civic life on the south hill led to the occupation of much of the sanctuary with shops and residential housing, until the abandonment of Selinus in the First Punic War. Within this large and complex area, our attention has focused upon the southern portion of Temple C, Temple B (ca. 300 BCE), Temple R (ca. 580 BCE), and the so-called “South Building,” the latter placed against the peribolos wall to the south of Temple B. As regards our topographical and architectural investigations, the new block-by-block study of the buildings, based on a detailed graphic and digital documentation of the remains both at the site and in the Archaeological Museum in Palermo, has provided a new, complete reconstruction of Temple B (restored by us as a prostyle tetrastyle of the Doric order, resting on a podium) and Temple R. These temples represent respectively one of the latest and one of the earliest monumental buildings at the site, and are among the most significant representatives of temple architecture on the island for their respective periods. No less important is the new analysis of the South Building, which suggests its identification as an impressive theatral viewing area, simple in design yet monumental in scale, with a capacity of c. 500 seated people. This building may have served for viewing a variety of ritual performances in the open space between Temple R, probably of Demeter Thesmophoros, and the entrance to the larger sanctuary: one thinks, in particular, of processions for sacrifices, contests and festivities. The discovery of an almost intact stratigraphic sequence ranging from the Early Hellenistic period down to the Bronze Age contributes significant new evidence concerning the history of occupation at the site, particularly for its earliest phases, including the time of the colonial foundation. The excavation of the intact Archaic, Classical, and Late Classical levels, sealed by a thick Hellenistic fill, inside the cella of Temple R, provides important new evidence concerning cult and ritual practice in this area of the sanctuary between ca. 630 and 350 BCE. Furthermore, the new analysis of Temple B and the excavations around its altar contribute significantly to our understanding of the cultural dynamics during the later phase of occupation at Selinus, under the control of Carthage yet with a mixed Greek and Punic population.

The mission’s first excavation season in the main urban sanctuary of Selinunte, conducted between May and June, saw the opening of three trenches in the southwestern area of the sanctuary. The largest trench was opened along the western front of Temple B (Trench A, 1.20 x 2.70 m). Within the trench, the entire stratigraphic sequence was documented, with bedrock reached in the southern half. This excavation was particularly significant, in that it immediately revealed the presence, in this area of the sanctuary, of an intact stratigraphic sequence, from the Early Hellenistic period all the way down to the Bronze Age. The reason why these lower levels were left unexcavated may depend on the fact that the excavations carried out in this area in the nineteenth century (1823, 1832, and 1876) reached only the level of Temple B’s lower foundations, likely on the assumption that these foundations would lie on bedrock. Instead, the foundations of Temple B lie upon layers of earth and sand, quite above the bedrock. Excavation revealed that a stratum of red earth mixed with sand covered the bedrock, which included a fragment of an obsidian blade. The presence of Bronze Age pottery in the upper strata further attests to a frequentation of this area in the prehistoric period. Upon this was a second stratum of red earth mixed with sand, this time sterile. A brown layer of earth mixed with sand lay above, containing pottery dated to the Orientalizing and Early Archaic period, including Protocorinthian: this layer is to be connected with a first phase of use of this area as a sanctuary. The next stratum (partly disturbed by the construction of Temple B, as was the stratum above), was a yellow layer of earth mixed with limestone chips, with a significant amount of pottery of the Early Archaic period: this layer is best understood as a phase of leveling of the area in connection with the construction of Temple R (ca. 580 BCE). Upon this rested a second brown layer of earth mixed with sand, with pottery dating from the Early Archaic period down to the early fifth century. This stratum represents the period of ritual function in front of Temple R. The large quantity of votive materials points to the cult of a goddess, evidenced by the large number of pyxides and the dedication of several jewels. Immediately on top of this layer were three strata of earth all datable to ca. 300 BCE based on ceramic analysis. These layers, which raised the ground level by at least 0.60 m, are part of the leveling of this area prior to the construction of Temple B. The second trench was opened inside the cella of Temple B, between the courses of blocks serving as foundations for the floor (Trench B, 0.40 x 2.00 m). In this trench, we could reach to a depth of only 0.80 m, uncovering five unique layers of fill, each consisting of earth mixed with sand, chips of stone from the dressing of the ashlar blocks of the temple, and fragments of tiles and pottery, particularly transport amphorae. The pottery suggests a date of ca. 300 BCE for the construction of Temple B. The third trench was opened against the northwestern corner of Temple B (Trench C, a square of 0.80 m l). We reached only to a depth of 0.85 m, yet found the same stratigraphic sequence as in Trench A. Particularly notable was the presence, in the Early Hellenistic fill, of fragments of the architectural terracottas from Temple C to the north.

The project’s second excavation season was carried out between May and June, with four trenches opened in the southern area of Selinunte’s urban sanctuary. The first trench was opened against the east front of Temple B, within the original boundaries of the building’s staircase, in correspondence to its southern third (Trench D, 1.60 x 2.50 m). The entire stratigraphic sequence was examined from topsoil to bedrock. At the lowest level, a sterile red stratum of earth mixed with sand covered the bedrock. This stratum was covered by a thick (max. h. 1.30 m) layer of nearly sterile yellow sand. Upon this layer was a relatively thin layer of grey earth, which included a considerable number of tile fragments, along with chips of stone and pottery. It was in this layer, datable to ca. 300 BCE based on the pottery, that were laid the foundations of the staircase, consisting of only one course of ashlar blocks. The bottom foundation course of Temple B, instead, rests upon the layer of sand below, and features a larger block at its south corner. The second trench was opened between the south side of Temple B’s altar and the north front of the South Building (Trench E, 2.42 x 2.9 m). Here too the entire stratigraphy was excavated down to bedrock across the trench. As in Trench D, a layer of red soil was covered in turn by a thick (max. h. 1.20 m) layer of yellow sand. Resting upon this layer of sand were three layers of grey earth, which included a considerable amount pottery. These layers of grey earth are all datable to ca. 300 BCE, based on the pottery. Together, they contributed to raising the ground level by about one meter. The altar was built taking this new level into account, and thus the fill should be considered as an intentional reshaping of this area on the occasion of the construction of Temple B. In the same operation, a course of ashlar blocks was added to the north front of the South Building. Our third trench was opened between the north front of the South Building and the staircase of Temple B (Trench F, 1.62 x 2.18 m). The entire stratigraphic sequence was excavated. A sterile, red layer of earth mixed with sand covered the bedrock. Upon this stratum was a thick (max. h 1.30 m) layer of yellow sand, with a significant amount of stone chips at the top and fragments of Archaic pottery. Two layers of grey earth mixed with pottery covered this layer of sand. The pottery from the upper layer suggests a dating to ca. 300 BCE. The two layers are probably part of the same action, which raised the ground level by at least a half a meter. The foundation for Temple B’s staircase cuts to the bottom of the upper of the two layers. Our excavation has revealed the presence of a lower foundation course on this side of the South Building, which rests upon the yellow sand stratum, rather than bedrock. The fourth trench was opened along the north side of Temple B, in correspondence to its eastern end and including the beginning of the foundations of the staircase (Trench G, 1.26 x 2.20 m). The entire stratigraphic sequence was excavated down to bedrock. A layer of red sand with relatively few pottery fragments covered the bedrock. Upon this stratum was a relatively thick (max. h 0.80 m) layer of red sand, with few pottery fragments. In this layer a large pit was cut, with pottery in its fill datable to the end of the fourth century BCE, and including a number of pieces belonging to the Late Classical period. A layer of grey earth rested above, with numerous tile and pottery fragments, suggestive of a dating to ca. 300 BCE. It was on this stratum that Temple B’s lower foundation course and the foundations of its staircase were laid.

Between May and July we carried out our third campaign of excavation in the main urban sanctuary of Selinunte, opening four trenches. The largest trench was opened along the western side of the altar of Temple B (Trench H, 2.00 x 3.85 m). In this trench, the entire stratigraphic sequence was excavated, reaching bedrock in the southern half. The bedrock was covered by a red layer of earth mixed with sand, sterile, upon which was a thick (ca. 1.10 m) layer of yellow sand of aeolic formation, in which we found only a few pottery fragments of the Archaic period: it remains unclear whether this upper layer represents an artificial or, more likely, a natural formation. The top of this layer of yellow sand was marked by a series of cuts and refuse pits dating to the Early Hellenistic period. To the same period date two layers consisting of earth mixed with pottery and tile fragments, which were deposited right on top of the layer of yellow sand. The two layers are part of the same fill (dated to ca. 300 BCE by both coins and pottery), which raised the level of this area of the sanctuary of about 0.70 m. It was on this fill, and the layer of aeolic sand underneath, that the altar of Temple B was built. Our excavation indicates that the altar consists of three courses of ashlar blocks on its western face. The first course from the bottom (ca. 0.40 m h) is not neatly aligned with the two courses above, and includes a series of reused blocks. A second, small trench (Trench J, a square 1.50 m l) was opened against the western side of Trench H, in order to fully explore a large refuse pit found against that limit. Another trench was opened against the north side of Temple B (Trench I, 0.95 x 2.25 m), near to its center. In this excavation, we reached to a maximum depth of only 0.20 m, uncovering two layers: a thick stratum of almost sterile sand at the bottom, and a thin stratum of earth mixed with pottery and tile fragments at the top, the latter presumably part of the same Early Hellenistic leveling found in the rest of the area. The foundation trench for the north wall of Temple B cut through both levels: pottery found in the fill of the trench, datable to ca. 300 BCE, provides further evidence for the chronology of Temple B. Our last trench this summer (Trench K, 0.86 x 3.10 m) was opened in correspondence to the southwestern corner of the South Building, which abuts the peribolos wall. In this trench, the entire stratigraphic sequence was excavated to bedrock. Covering the bedrock was an Archaic stratum, which included pottery datable to before the middle of the sixth century BCE. A second Archaic layer was immediately above, with pottery dated to the end of the sixth-early fifth century BCE. The South Building’s lower foundations are above this layer, which contributes to the dating of this building to the Late Archaic period. The last phase identified in our trench dates to the third century BCE, in which the South Building was systematically dismantled reusing some of its stones.

In May and June we carried out our fourth campaign of excavation in the main urban sanctuary. A trench was opened (Trench L, 1.65 x 4.40 m) along the course of ashlar blocks placed against the front of Temple R. The eastern limit of this trench consisted of the west facade of Temple B; whereas the trench’s western limit was the Hellenistic wall nearby the east facade of Temple R. By the end of the season, the entire stratigraphic sequence was uncovered, down to bedrock in the eastern end of the trench. Our excavation has produced significant evidence concerning the stratigraphy in this area of the sanctuary, particularly in relation to the earlier phases of occupation on the Akropolis. The bedrock was in fact covered by two thick layers, both sterile: the stratum beneath of a red hue, consisting of earth mixed with sand, and the upper stratum brown at the time of the excavation and grey immediately afterwards, consisting of sand. The red layer should be regarded as the result of the erosion of the bedrock, whereas the brown/grey layer should be considered instead as an aeolian formation: not, as it has been thought for similar layers of sand found in other areas of the main urban sanctuary, an Archaic fill. This is best indicated by the fact that on top of this brown/grey layer stood a structure of the Orientializing period: a rectangular building with an approximate EW orientation, occupying the NW sector of the trench, including a portion of the south wall (2.60 m long) and a tiny section of the east wall (0.75 m long). The elevation of these walls was in mudbrick, found in situ disintegrated on top of the foundations, which were made of rubble (preserved for a height of 0.30 m and 0.40-0.50 m thick). This rubble consisted of limestone chips and small blocks held together with clay. The first Temple of Malophoros offers a good parallel for the size and structure of the foundation of this wall. Adjoining the two walls were the remains of a clay floor, particularly hard and consisting of two layers. The rim of a Samian amphora comes from underneath this floor, helping to date the building to the seventh century BCE. In the upper layer of the floor fragments of Phoenician bichrome pottery were found, of particular significance given the sacred context and the early date. In the early sixth century, the Orientalizing building was dismantled, and the area was subject to a leveling most likely connected with the construction of Temple R and the reshaping of the area in front of this building. Coinciding with this leveling, right above the foundations of the Orientalizing building, and immediately to the south, were found the tips of two large iron spears. The following strata, identified as phases of use of this area of the sanctuary during the Archaic and Classical periods, include a large quantity of votive material, including pottery and some terracotta sculpture, such as the right foot and lower leg of the statue of a youth. These Archaic and Classical layers were ultimately disturbed in the Hellenistic period by a pit and by the leveling associated with the construction of the wall which represents the western limit of this season’s trench.

In May and June we carried out our fifth campaign of excavation in the main urban sanctuary. Excavation this season consisted of the opening of a small trench (Trench M, 1.80 x 2.40 m) against the east front of Temple R, in correspondence of the southern half of the door and part of the adjoining wall. By the end of the season, the entire stratigraphy was excavated down to bedrock across the trench. Our excavation has proved particularly significant regarding the identification of the stratigraphic phases associated with the life of Temple R and its environs between the Orientalizing and Early Hellenistic periods. Our first finding consisted of the bottom layer (ca. 50 cm thick) of a Hellenistic fill (datable to ca. 300 BCE based on bronze coins and pottery) that substantially raised the ground level in front of the temple. This action corresponded to a raised floor level within the temple, whose Archaic threshold (originally accessed from the outside by a step consisting of a single block placed against the center) was elevated by ca. 1.35 m, up to the level of the present threshold. The fill rested on top of a yellow clay floor, datable to the late fifth-early fourth century BCE (based on pottery analysis), which included a number of fifth-century terracottas: the polychrome bust of a goddess wearing a polos, comparable in style to the female heads of the western metopes of Temple E (Hera) and thus datable to ca. 470-460 BCE; the lower part of a female protome; the upper part of a small peplophoros, datable to ca. 460-450 BCE (with jewelry originally painted in Egyptian Blue); and finally a small female head wearing a polos, with traces of color on the lips, datable to ca. 440 BCE. Most likely, these terracottas served as votives in this area, and were redeposited on the occasion of the restructuring of the building after 409 BCE. From this layer also comes the fragment of a geison revetment of Roof A Scichilone, which shows traces of burning. A compact and burnt red clay floor was found under the preparation for the yellow clay floor. Inserted in this floor was a large cover tile placed upside down against the front of the building. This floor may be dated to the early fifth century, based on the pottery from its preparation, including underneath the tile. Immediately under this floor, was an earlier floor of yellow clay, which is datable to the original phase of construction of the building (early sixth century BCE), based on the pottery from its preparation. This preparation included a thick packing of stones, which has substantially preserved the strata underneath. These layers include the foundation trench of the eastern wall of Temple R, with its fill of chips of stone from the dressing of the ashlar blocks for this building, and the remains of an earlier structure obliterated by the construction of Temple R and almost at the same altitude as the Orientalizing building excavated the previous year in Trench L. The remains of this earlier structure include a clay floor, featuring two cuttings, and a conspicuous number of burnt fragments from a thatch roof. The pottery associated with this early structure dates to the seventh century BCE. This early structure rested on top of a red layer of earth mixed with sand, which was placed immediately above the bedrock.

In May and June we carried out our sixth campaign of excavation in the main urban sanctuary. Excavation this season consisted of the opening of a trench (Trench O, 2.62 x 3.43 m) in correspondence to the SE corner Temple R’s cella, including part of the door that gave access to this building. By the end of the season, the entire stratigraphic sequence was excavated down to bedrock in the northern half of the trench. Our first finding this season was the identification of a massive fill inside the building, which has raised the level of the floor by more than one meter, but was not documented before. This fill is datable to ca. 300 BCE and has sealed the Late Classical, Classical, and Archaic levels of Temple R. The fill consists of three levels, and it contains a significant amount of roof tiles fragments at the top, and amphorae fragments towards the bottom. In the fill was also found fragments of Archaic and Classical votive objects, including part of an Attic white-ground lekythos stylistically close to Douris and depicting Eros offering a wreath to a youth. The first level identified under the Hellenistic fill, consisted of a floor datable to roughly the end of the fifth century, and in phase with the pilasters extant in the cella. This floor sealed a layer of burnt material, bearing clear signs of violent destruction, including a bronze sheet smashed on the ground and over twenty arrowheads. This destruction is best associated with the Carthaginian sack of Selinus in 409. Beneath the Classical level and the preparation for the fifth century floor, which was significantly raised above its original level, was the well-preserved Archaic floor. Embedded in this floor, a terracotta figurine of a veiled goddess was found, wearing a mantle and a low polos, of the Daedalic style: this figurine is more likely a representation of the divinity of Temple R. This is best identified with Demeter, based on the discovery in front of the temple last year of a polychrome terracotta bust of a goddess wearing a polos, a type often associated with Demeter and Kore; one may also add the considerable number of piglet bones among the remains from animal sacrifice found in the area of our building. At the level of the Classical and Archaic floors, against the east and south interior walls, we found a significant number of votive depositions, which included vessels for libations, weapons, terracotta figurines, a faience amulet of a bull, and two fragments of a bone aulos. This aulos, datable to ca. 580-570 based on its association with a Middle Corinthian amphoriskos, offers significant evidence for the performance of music and ritual dancing in association with cult in this part of the sanctuary. Particularly significant was the discovery, under the fill of stone chips that served as foundation for the floor, of two large postholes with a precise EW orientation. These two postholes may tentatively be identified with the central supports of an early building, whose floor and walls were uncovered in 2010 in the excavation of Trench L. At the level of the floor of this early structure, smashed in the foundation trenches for the eastern wall of Temple R, we found a large conical oinochoe close in style to Late Protocorinthian-Early Transitional vases, depicting an animal frieze with two grazing deer. This vase undoubtedly represents a votive dedication of some significance, datable to the first generation of life of the Greek colony.

In May and June we carried out our seventh campaign of excavation in the main urban sanctuary. Excavation this season consisted of the opening of a large trench (Trench P, 3.90 x 4.66 m) in correspondence of the western two-thirds of the adyton, the innermost chamber of Temple R. By the end of the season, the Hellenistic and Late Classical levels in the trench were excavated, down to a layer of burnt material corresponding to the Carthaginian destruction of Selinus in 409 BCE. This excavation, along with the reassessment of the evidence from past seasons, has been critical for establishing the various phases of our building: I) Original construction in ca. 580 BCE; II) Renovation of roof and floor ca. 500 BCE; III) Damage by fire within the context of the Carthaginian conquest of 409; IV) Substantial rebuilding a few years later, presumably by the Selinuntine refugees returning home from Acragas; V) Creation of a large fill raising the floor of ca. 1 m and major restructuring of the building, ca. 300 BCE; VI) Final use as an arsenal within the context of the First Punic War, before the abandonment of Selinus in 250 BCE. Likely, Temple R was also used in later periods: however, the relevant evidence was removed during the 19th century excavations, which reached to the level of the Hellenistic fill across the interior of the building. The 2013 excavation has provided important evidence concerning, in particular, phases IV-V. After damage caused to the temple in 409, the building underwent a major restructuring, which included the addition of the west chamber, entered from the south, and a new roof. Furthermore, three pilasters along the main axis, with two preserved in the cella (recently mistakenly identified as bases for baetyls) and one, later dismantled, in the adyton, supported this new roof. The dating of this phase is provided by the discovery of a bronze hemilitron from Acragas (ca. 425-406) in the new floor associated with this renovation. Particularly notable is the redeposition, in this floor, of Archaic votive materials. Towards the end of the fourth century, the building may have ceased to be used as a cult place. This is inferred from the thick fill dated ca. 300 BCE, which substantially raised the floor level within the building. In the adyton, this fill consisted of two layers: the top of tile fragments and a dense concentration of stones (not documented in 2012); the bottom of several fragments of transport amphoras, like the corresponding level of the fill in the area of the door. In addition, in the area of the adyton, as in part of the cella, the fill included a series of buttressing blocks—consisting of reused architectural elements—laid against the interior walls. In this fill, refuse pits were found filled with charcoal, butchered and burnt bones, and amphora fragments, the latter datable to ca. 300 BCE and thus associated with the making of the fill. During this action, the pilaster in the adyton was discarded; it remains unclear whether the two pilasters in the cella were still used for structural reasons. To this same phase, may belong the opening of a small door in the north wall of the cella.

In May and June we carried out our eighth campaign of excavation in the main urban sanctuary. Work this season consisted of the continued excavation of Trench P, opened in the previous year in the adyton of Temple R. By the end of the season, the Classical and Archaic levels in the trench were excavated, down to the floor of an earlier structure datable to the seventh century BCE. Our first finding this season was a wide looting pit dug into the center of the adyton, which reached all the way down to the bedrock, cutting through the stratigraphy associated with the earlier phases. This pit was sealed by the late fifth century floor excavated in 2013, and it is thus best associated with the Carthaginian conquest of Selinunte in 409 BCE. Fragments of precious votive offerings (including gold, silver and ivory) from the cut and fill of the pit suggest that it may have served for the looting of a foundation deposit placed in the center of the adyton. In this operation, the floor against the north and south interior walls was left untouched. The excavation of this floor revealed the complete absence of votive offerings underneath, showing that while in the cella votive depositions were generally lined against the interior walls, the practice in the adyton was quite the opposite. Of particular significance was the discovery of dozens of loomweights in the fill of the looting pit. Under the burnt layer corresponding to the Carthaginian destruction of 409, we excavated the thick layer (ca. 1 m high) of stone chips (from the dressing of the ashlar blocks of our building) that served as the foundation for the Archaic floor of Temple R. The materials from this layer confirm the dating of the construction of our building to ca. 580 BCE. Particularly notable has been the discovery, towards the bottom of this layer, in the SE corner of the trench, of the remains of an animal sacrifice, consisting of a thin layer of ash and bones, deposited directly upon the limestone chips. The latter bore clear signs of burning themselves, thus attesting to a primary sacrifice. The bone fragments (burnt at a temperature exceeding 600°) belong to a ram (including a large horn fragment), two sheep, and one goat. This was likely a sacrifice associated with the construction of Temple R, which has particular significance in consideration of the attribution of this building to Demeter and Kore, and the strong connection of the ram with the cult of Persephone. Finally, under the layer of stone chips, we found the remains of a large structure taking most of the area of the trench. Evidence for this structure consists mainly of a clay floor (pierced by five pits of varying width and depth) similar to the one excavated in 2010 in trench L, but thicker: this, along with the distance between the two floors, may point to two separate structures, both dated to the first generation of life of the Greek colony. Finally, it may be noted that in the excavation of the fill of the looting pit, particularly towards the bottom, we have identified a significant number of fragments of Late Bronze Age pottery.