1.1 The International Institute of Ligurian Studies and the University of Genoa’s Department of Medieval Archaeology recommenced investigations on the cathedral of S. Maria in 2007. The new phase of research – fifty years on from those undertaken by Nino Lamboglia – involved over fifty people from the university and the Civic Museum of History and Archaeology. The excavation, directed by Carlo Varaldo and Rita Lavagna, examined an entire room of the House of the Colonels and the pre-existing habitatione of the Citadel Commander, over an area of circa 100 m2. 1.2 The excavations began at a floor of stone basoli and brick datable to the final phase of use of the rooms of the House of the Colonels, that is the phase which concluded with the demolition in the 1940s. The various phases of flooring in the room were documented and removed, bearing in mind the successive phases of maintenance and repairs which between the 17th and 20th centuries had led to the formation of a composite floor level. The original and earliest phase, datable to between the end of the 16th-beginning of the 17th century, was constituted by an extremely worn brick pavement with much evidence of relaying and repairs. Below this original floor level there was an extensive layer of very compact detritus with a mainly lime matrix. The finds recovered indicated that this deposit had formed in the 16th century. The demolition dated by the written sources to the end of the 16th century was extreme. Only a small part of the original deposit relating to the medieval cathedral was preserved below the building site phase for the construction of the rooms belonging to the buildings inside the Genoan citadel. In the southern part of the excavation this sequence covered the razing of a thick wall on an east-west alignment. The wall was only preserved at foundation level and may be identified as a perimeter wall of the nave of the cathedral of S. Maria. There were no traces of the cathedral’s floor, but only part of the beaten mortar make up. Both inside and outside the nave – below the layers relating to the construction site of the end of the 16th century – part of the bedrock appeared which had been artificially levelled. In the north part of the excavation, in correspondence with the central nave, a beaten floor surface and a masonry-built burial loculus were identified, but above all two steps emerged which led down to the crypt. The traces of earlier phases and structures were very limited. A wall bonded with rough mortar rich in charcoal inclusions may be interpreted as a perimeter wall from one of the cathedral’s phases. A number of burials in earth graves showing clear signs of reuse, can be placed within this phase. 1.3 The archaeological analysis of the phases relating to the Habitazione of the citadel’s commander and the House of the Colonels has produced important elements for the reconstruction of the phases which date from the demolition of the cathedral of S. Maria to the present day. The elements acquired for the reconstruction of the building phases of the cathedral to a large degree integrate the data from the excavations and the previous investigations. Based on this body of data, which is totally complementary to that from the written sources, reconstructions of the cathedral’s plan and occupation phases between the 12th and 16th century have been elaborated. A perimeter wall and a cemetery phase belonging to a building given a preliminary date in the late Middle Ages was uncovered. This late medieval phase may be linked to the presence of architectural sculptural elements of Lombard and Carolingian date, recovered during the course of the investigations. The layers that can certainly be dated to the late antique period – in the excavated area – were quantitively and qualitatively very limited. These elements document the occupation of this part of the Priamàr hill during the Byzantine period, even though they cannot be directly associated with the presence of a church. Only the continuation of the excavations in 2010 and the widening of the excavation area will clarify whether: - the Byzantine occupation phase (6th-mid 7th century A.D.) was associated with a primitive church whose structures have yet to be identified. - the construction of the cathedral of Santa Maria should be dated to between the late Lombard period and the Carolingian period, as suggested by the chronology of the marble architectural sculptures recovered during the excavations.

As part of the archaeological investigations aimed to render more attractive the remains of the medieval cathedral of Santa Maria Assunta in the Priamàr architectural complex, excavations were undertaken by the International Institute of Ligurian Studies and Genoa University. This campaign concentrated on the area situated south-west of the nave (brought to light during the 2008-2009 campaigns). In fact, according to the sources this was the site of the cloister and structures belonging to the cathedral. The excavation area was delimited to the east by a wall (US 11) belonging to the Palazzo dei Colonelli, to the west by the wall of the fortress’s bastion (US 25), and to the north-east by the remains of the supporting wall of a great brick-built vault that was also part of the Palazzo dei Colonelli.The stratigraphy in this point was disturbed in the modern period by the construction of a well covered by a slab of reinforced concrete (US 40). Following the removal of the dense vegetation and humus, about 10 cm thick, three archaeological features, relating to occupation of the Genoese fortress between the 16th and 18th century, came to light. The first, constituted by a row of squared stone slabs bonded with mortar (US 02) situated exactly at the centre of the excavation area, was linked to the east to the second feature, formed by a patch of cobbled floor (US 03) in a mortar make up (US 46). The slabs forming US 02 were characterised not only by their regularity and large dimensions (90 x 40 cm) but also by the thickness, on average 15cm. The third feature, even more macroscopic, was an ample paving situated in the north-western sector of the excavation area, also formed by very large stone slabs. However, although the dimensions and thickness were similar, the layout was very different from US 02 given the irregularity both of the working of the slabs and their positioning. All three of the above stratigraphic units were on the same level, and thus it is possible to interpret them as contemporary and part of the same occupation level. Both US 03 and US 16 presented clear traces of robbing, cut US -22 for the cobbles and US -26 for the paving. Presumably the paving slabs were removed in order to be reused. The earthy fill of both cuts contained pottery fragments of diverse dates which presumably came from outside of the site. The removal of the fill from cut US -26, a compact layer of greyish-brown soil containing brick and stone fragments (US 15), revealed the floor’s construction. A mortar make up (US 17) emerged which overlay a substantial layer of broken brick/tile mixed with lime (US 14). Both layers can be interpreted as a specific intervention to reorganise the site and reconstitute a floor level. The most important contribution to the identification of the remains of the earliest building, presumably late medieval, came with the removal of US 23. Characterised by an earthy layer similar to US 15 but darker in colour and situated between US 25 and 16, it was the fill of another cut (US -27) in US 16 itself. Its cleaning revealed a brick wall (US 30) on the same alignment as the perimeter wall of the fortress and separated from it by a cut (US -58), probably a drainage channel. The removal of US 10, that is the fill of cut US -22, uncovered a further two dumps of material (US 45 and 48). This led to the identification of a second brick wall bonded with mortar (US 55) forming a 45° angle with the south edge of the excavation area, and which appeared to continue below US 02. At the end of the campaign evidence of a least one structure preceding the occupation level constituted by US 02, 03 and 16 emerged. The removal of the stone slabs constituting US 02 and 16, and of the related make ups, together with the excavation of US 45 and 48, could therefore make it possible to go down to earlier levels belonging to the same occupation phase as walls US 30 and 55. It is hoped that, once uncovered, this phase will be prove to be coeval with the evidence already brought to light in previous campaigns in the cathedral and thus provided a more complete picture of the entire complex of Santa Maria Assunta.

All’inizio dei lavori, dopo un operazione di messa in sicurezza dell’intera area di indagine, si è operato un sensibile ampliamento dell’area di scavo. Il settore compreso tra il limite nord della pavimentazione/basamento individuata nel 2010 e il muro perimetrale del Palazzo dei Colonnelli è risultato interessato da opere di canalizzazione delle acque, realizzate in una fitta sequenza di costruzione e dismissione di pozzetti, condutture che hanno profondamente intaccato il deposito stratigrafico e, sulla base dei reperti rinvenuti, sono state attive fino al 1943/45. L’andamento e la pendenza delle condutture in laterizi e ardesia indica un complesso sistema di raccolta delle acque dagli elevati del Palazzo dei Colonnelli e un successivo deflusso verso ovest, in direzione di una cisterna o di scarichi probabilmente esterni alla Cittadella. Il muro perimetrale del Palazzo dei colonnelli è risultato rifasciato da una struttura ad esso esterna e parallela, avente il medesimo andamento. Le due strutture erano separate da un’intercapedine. L’intera sequenza delle attività di realizzazione e dismissione delle strutture idrauliche è stata realizzata tagliando un potente strato di materiale detritico, presente sull’intera area, anche al di sotto del pavimento/basamento. Si è quindi deciso di rimuovere il basamento e la sua spessa preparazione. Risulta indubbio che la funzione di questa struttura era collegata alla esigenza di movimentare, ancorare e mettere in posizione pezzi di artiglieria, nell’ambito della vita della fortezza del XVII-XVIII secolo. Lo strato sottostante era formato da materiale detritico sciolto, un pozzetto/cisterna presente nell’angolo nord est dell’area di scavo e il muro perimetrale della cittadella. Tale strato copriva un suolo compatto, da interpretare come piano di cantiere, una superficie (ossia un’interfaccia di trasformazione) costituita da strati a matrice leggermente differenziata, formatasi su tutta l’area dopo la costruzione del muro perimetrale della Cittadella, a seguito di una prolungata azione di calpestio. Da tale piano di cantiere emergevano le rasature di due strutture murarie. I reperti recuperati confermano una datazione al XVI-XVII secolo. Tale superficie è stata tagliata per la costruzione del muro 122, come si è detto, ad andamento parallelo al muro perimetrale meridionale del Palazzo dei Colonnelli. La prosecuzione dello scavo ha comportato la rimozione di strati, fino a mettere in luce una sequenza costituita da strutture murarie rasate e strati di crollo e demolizione in posto. L’insieme delle evidenze può essere interpretato come l’esito della demolizione di strutture murarie medievali, preesistenti alla fortezza e pertinenti - nel loro complesso - al chiostro della cattedrale. Le strutture murarie sono pertinenti ad almeno quattro ambienti, con le aperture di passaggio e di comunicazione e con i crolli interni dei muri perimetrali. L’insieme delle unità stratigrafiche è ben leggibile nella documentazione grafica e fotografica di fine scavo che documenta l’articolazione degli ambienti demoliti al momento della costruzione della fortezza. L’analisi archeologica ha fornito preziosi elementi per la ricostruzione delle fasi che vanno dalla demolizione degli ambienti esterni alla Cattedrale di Santa Maria (area del chiostro) fino all’età contemporanea. Gli elementi acquisiti per la definizione degli interventi di demolizione delle strutture medievali e costruzione della fortezza e degli ambienti residenziali della Cittadella sono complementari a quelli desunti dalle fonti scritte, consentono di elaborare valide ipotesi ricostruttive della planimetria e delle fasi di vita del complesso tra XVI secolo e XX secolo. La prosecuzione dello scavo avrà il compito di delineare la sequenza cronologica e l’assetto planimetrico delle fasi di vita medievali ( e probabilmente precedenti) del complesso episcopale.

The excavation is run as a field school and archaeological laboratory for students from the University of Genova. Sector 1000 This sector corresponds with one of the rooms in the medieval cloister, identified in 2012/2013. In 2014, excavation began of three masonry-built loculi situated at the centre of the sector, completing the excavation of the central one. Inside the loculi, in the upper part of the fill the skeletal remains were mixed with detritus and did not appear to be in primary deposition. The excavation of the central loculus confirmed this. The bones were mixed with soil containing demolition debris (brick/tile, stones, lime, and sand). The floor of the loculi was formed by the bed-rock. In 2015, a small triangular area was excavated between the three masonry-built loculi and the wall separating sector 1000 from sector 2000. This part of the archaeological deposit, cut by the foundations of the late medieval walls and the construction of the funerary loculi (late 14th-late 15th century) preserved a small part of the stratigraphy pre-dating the medieval construction phases. This consisted of a few patches, in direct contact with the bed-rock, in which fragments of late antique amphorae and a few pieces of slag were present. Sector 2000 At the end of the 2013 campaign, a structure built with moderately regular blocks of “Finale” stone was identified in this area. It was interpreted as a funerary loculus, cut on the south side by the construction of a cistern. In 2014, the internal stratigraphy of the loculus was excavated. The original cover of the loculus was missing (it was perhaps a marble or stone slab) and – in the upper part – it was covered by a layer made up of marble fragments that appeared to be burnt and semi-calcified. Below this layer were skeletal remains in secondary deposition, mixed with detritus. The excavation revealed the presence of a least three individuals, disturbed by being moved although some elements were still articulated. In 2015, a sequence was identified consisting of burials in primary deposition, alternating with episodes of the reduction of the skeletons and fills of sandy soil. In practice, each individual deposition was preceded by a phase of levelling of the surface inside the loculus in order to create a suitable new burial space. This way of managing the funerary space means that almost all the primary depositions were disturbed by subsequent activity, consequently the skeletal remains were only partially preserved. Sector 4000 This sector corresponds with the entrance area, situated to the east of room 1000. In 2014, the excavations reached final medieval floor surface of stone slabs. This season, a limited trench was excavated in order to complete understanding of the stratigraphic sequence down to the Priamàr bed-rock. The rock surface was covered by a beaten mortar make-up for a brick floor, originally connected to the step leading into room 1000. Following the deterioration of the brick floor, probably in the 15th century, layers of soil and detritus were put down and a new stone floor was laid. The latter remained in use until the rooms were demolished and the cloister structures abandoned in the second half of the 16th century. Conclusion Research will continue in 2016 with the extension of the excavation area in view of the general reorganisation and enhancement of the area of the cathedral of Santa Maria and associated structures.

Over 20 students from Genoa University and 25 local school students took part in this season’s excavations and finds study. Area 7000 The excavations in this area aimed to identify the link between the left aisle of the cathedral and the adjacent octagonal baptistery, originally an independent structure outside the church, and later joined to it. In this area, the Lamboglia-Restagno excavations of 1956-58 had uncovered traces of a burial cut into the bedrock, of the same type as those in the 4th-7th century cemetery identified between 1969 and 1989 underneath Palazzo della Loggia, inside the Priamàr fortress. The burial was presumed to be part of the vast cemetery that extended, according to late medieval sources, along the left flank and facade of the church. Work took place in two adjacent sectors (7100 and 7200) situated respectively, to the west and south of an imposing bunker built during the Second World War that had completely obliterated the baptistery. However, it was possible to identify some of the walls joining the church to the baptistery and a long stretch of the left perimeter wall of the church itself. The latter was formed by two distinct phases: an early medieval one (presumably datable to the 8th century) and its successive elongation during the reconstruction of the Romanesque period. Thus, it is possible to suggest that the original single-naved early medieval church as well as being enlarged on the right side and transformed into a church with three aisles, also had the facade enlarged by the addition of two bays. In addition to the burial found during the 1956-58 excavations, two more tombs were found. These were also rock-cut and part of the early medieval cemetery adjacent to the cathedral. A system for collecting rainwater was identified that was part of the Genoese fortress. It had two small channels, which carried water to a large cistern that was built in 1730 and designed by Pietro Morettini. Area 8000 The 6 x 6 m trench opened at the centre of the citadel’s courtyard aimed to find the cathedral’s facade wall in one of the few spaces not affected by the construction of the 18th century cistern. In addition to the 16th century curtain wall on the western front of the courtyard, the excavations exposed a stretch of the cathedral’s right perimeter wall on top of which the facade was probably constructed. Only through continuing this research will it possible to identify its exact position and therefore the precise dimensions of the church, obliterated during the course of the 16th century and of which only a schematic plan exists.

The 2017 campaign was run as a field school, as were the post-excavation laboratory work, with the participation of over 30 students from Genova University, in addition to some school pupils. Logistical and scientific support was provided by the Civic archaeological Museum and the International Institute of Ligurian Studies. The excavations concentrated on area 8000, opened in 2016 in order to find the facade wall of the medieval church, destroyed during the 16th century for the construction of the Genoese fort, and thus gain a more precise idea of the cathedral’s overall plan. The area was extended to wards the NE and this revealed a substantial wall attached to the right side of the building, which can probably be identified as the facade wall. This would make the church larger than previously thought (c. 61 m long instead of 55m), but excavation needs to continue to the east in order to verify this. In the meantime, a complete survey was made of all the emerging monumental remains in order to precisely document the layout and make a 3D reconstruction of the entire archaeological site, using the appropriate equipment, including drones thanks to the collaboration of the Istituto Geometri Boselli-Alberti. Only the continuation of the excavations will verify the 2017 results and provide further knowledge of the most important religious building of medieval Savona.