THE CHURCH OF THE SAVIOUR AT THE SOURCE OF THE CETINA RIVER AND THE WESTWORK IN THE CROATIAN PRE-ROMANESQUE

Analizom podataka koje pruža crkva Sv. Spasa na vrelu Cetine i grupiranjem nekoliko drugih spomenika oko nje, vodeći pritom računa o zapadnom masivu, daju se svi raspoloživi argumenti za tumačenjefunkcije westwerka u Hrvatskoj, dospijeće westwerka u hrvatsku kneževinu, parageneze oblika (aksijalni zvonik), te za održanje i transformaciju westwerka u, funkcijom drugačije odredene, arhitektonske sklopove u Xl. st. (toranj-portik). Naglašava se programatsko unošenje dvorskog ceremonijala i liturgije karolinškog doba u Hrvatsku izpolitičkih razloga.The Church ofthe Saviour at the source ofthe Cetina River is an unevitable issue when discussingpre-Romanesque architecture in Croatia, since it is nearly the only wellpreserved large pre-Romanesque church built that did not undergo any greater interventions subsequently. The Church of St. Saviour has not been, however, completely studied Quite the contrary we may say, since it is about to face largescale research. Only after initial amateur research did S. Gunjača manage to approach the monument more systematically. His research determined a more preciseplan ofthe church, without the subsequently added Gothic presbytery lt yielded the altar screen bearing the inscription ofthe district-prefeet Gastika, determining that the beifry that lied in the axis of the church was its integral part. lt is important to assess that, although the results ofthe research were not published as an analySiS, some strong!y determined crucial facts, make the church of St Saviour a prime-category pre-Romanesque monument in Croatia. The more, because it is the only church that was preserved up to its roofing, and, analogically, offers facts about other, similar buildings of the pre-Romanesque in Croatia. Not only is the Church of St. Saviour by its shape one of the representative churches, those that were by J. Petricioli defined as Early Croatian in the true meaning, but by its specific western structure, it is crucial for the comprehension ofthe westwork in the Croatian pre-Romanesque. One of the main problems ofpre-Romanesque architecture in Croatia is the Carolingian influence on its making, which has only recently become the issue ofcomplex research, and that ofthe westwork is the most important among them. The term westwork was for the first time used and noticed on some pre-Romanesque churches in Croatia by V. Gvozdanović, discussing it from the point ofview ofform, while as for its functional aspect he considered the "Kaiserkirche" as the dominant one. The papers written by Gvozdanović undoubtedly gave more space for interpreting various impacts on Croatian pre-Romanesque architecture. However, there were still many problems that remained open such as dating, genesis, paragenesis, the way in which the westwork was introduced in Croatia as an architectural motif and the insujJiciently studied issue of how it survived. Regarding thefunction and terminology ofthe westwork in preRomanesque architecture in Croatia, there are some new arguments that should be added to Gvozdanović\u27s -burial of the dignitary in the crypt of the westwerk, the hypothesis of dual patronage of the church at Solin that would prove that the westwork was usually dedicated to a patron other than the church. Determining the westworks in our country l referred to them as being atrophied or reduced which as such, were introduced to our regions at the time when they were already atrophied in the Carolingian world. When speaking of the first preserved westwerk in Croatia, l set forth the hypothesis about monk Gottschaik as the possible conveyer ofthe first westwork known to us, anto the Croatian court in the time ofprince Trpimir. This hypothesis derived from the previous greater discussion about the shapes ofchurches in the first halj of the9th century and how their forms were introduced into our regions. The time span between the construction of such churches (e.g. Crkvina at Biskupija near Knin) and the group of churches with round counterforts, thus seems logica/. In order to determine whether the western complexes of preRomanesque churches in Croatia bear the tnte functions ofthe westwork, the main issues of the functional determination of the western structure are their dating, paragenesis of shapes (its genesis can in no way be searched for in our regions), the diffusion, and finally the subsequent transformation in architectural parts that are similar by shape but different by function. Only after such analysis can we consider the reasons which brought the westwork into Croatia. The church ofSt Saviour is the only preserved example of a preRomanesque church with westwerk in early mediaeval Croatia, and therefore the crucial example in comparative analyses. By its ground-plan, the Church ofSt Saviour (Pig. 1) is an aisleless three apsed church with a trefoil presbytery and westwork on its western side. As the church is rather very well preserved, elements ofthe westwork construction can be seen. The elevation shows part of a unique block ofthe church ofwhich only the beljry at the facade rises high. The groundfloor ofthe westwerk is divided into three segments by vaults and arches in such a way that the latteral narrow areas have a barrel vault, while the central, square shaped area has a cross-vaulting Thegroundfloor of the beljry has a groin vaulting, reconstructed to a great extent. On the first floor, the central room is widely open towards the church through an arch. The first floor ofthe beljry is in a better way connected with the main room ofthe westwok, open at its side with its entire heigth. The first floor of the westwork was accessed by an exterior stairway with direct entrance to the first floor of the beljry. The altar screen divides the three-apse trefoil presbytery from the nave. Its architrave and part of the pediment were preserved and once reconstructed, they show the exact width of the church. Although there are no proofs about burials at St. Saviour (all graves date from the later period, even though they might have been placed above the ground -sarcophagi), the westwork wich such function was most accurately recorded at St. Stjepan at Otok at Solin that houses the sarcophagus of queen jelena. Yet there was no capsa reliquiarum that held the relics of St. Saviour, but the patron of the church, very rare in Croatia, indicates a relation with the Saviour\u27s cu/t. The southern wall of the first floor of the beljry have a small semicircular niche that appears only at 80 cm above the floor and rises high nearly to the vault ofthe first floor ofthe beljry (Fig 1). Its shape definitely indicates that its purpose was to serve as a small portable altarreliquary, which would confirm the liturgicalfunction of the first floor of the westwork in our region as well, certainly in a somewhat reduced form. When speaking of another possible function, the westwerk ofSt. Saviour is the only to provide any facts whatsoever. on the first floor towards the presbytery of the church, it opens with a large arch and two smaller lattera/ anes. This was the place where the nobleman, either lay or sacerdotal attended the liturgy on the main a/tar (Fig 3). It is this partiCular element that eams it the title oj"Kaiserkirche", an area intended for the ruler, but also an "Eigenkirche", a private feudal church. The district-prefect Gastika, a high ranking nobleman of the 9th century Croatian court, had the church furnished. Comparing the decoration on the altar screen to those bearing the name of Branimir, this one at St Saviour is dated to the time ofprince Branimir, as a product of a stone mason \u27s workshop at the Croatian court. Comparative analyses ofthe church ofSt. Saviour at the source of the Cetina River, may determine a smaller group ofpre-Romanesque churches all of which have a western structure. Their similarities can only be drawn from the ground-plan as none ofthem has apreserved e/evation. The nucleus of this group are a few churches with round buttresses that are by their ground-plan similar to the church ofSt Saviour, the church at Lopuška glavica at Biskupija and St. Cecilia at Biskupija near Knin (Fig 5). The church at Lopuška glavica is identical by its ground-p/an to St Saviour, only in a smaller scale and without the beljry that lies in its axis within the westwork. The Church of St. Cecilia is a three apsed church whose vau/ted ais/es are separated by cruci/orm pillars. At its west it has a westwork and an axially prominent beljry. The strips that divide the westwork into three parts as seen from the groundplan, indicate that the space in its elevation was equally divided as in St Saviour, only more monumentally. On the basis of round buttresses and vau/ting, I. Petriciali assigned the Biograd cathedral to this group (that only has an articulated beljry at the facade) as well as the so-called Fourth church at Biskupija (whose western end remains unknown) and characterized this group as a Significant contribution of Early Croatian builders to the European pre-Romanesque Equally, analyses of the scutpture indicated that the Early Christian church at Žažvić had its altar screen furnished in the mid 9th century, cut by the same workshop that made the furnishing of the Church at Lopuška glavica and at Rižinice with Trpimir\u27s inscription. The church at Žažvić is an Early Christian church adapted in the early Middle Ages with the construction of the westwork. Among the pre-Romanesque westwerk.s two of them stand out by their complexness and dimensions -the one ofSt. Mary at Crkvina in Biskupija and St. Stjepan at Otok in Solin (Fig 6) Only the foundations of these churches were excavated, but the ground-plan of St. Stjepan in Otok is however more precise. This is a three aisled basilica whose naves were divided by three pairs afpillars. :rhe span between the first and the secondpair which is greater than between otherpairs, shows that the eastren part of the church must have had a dome or something similar to a low lantem The westwork at the western facade is the most elaborate example in Croatia, whose entrance was divided into three parts that in the elevation indicate two taller latleral towers (the southern one was the staircase) that flanked the vestibule that, following the (;erman facades, must have been lower than the lalteral towers. Looking at it vertical/y, its division into three parts is not only indicated by the presumed height of certain segments, but also the lesenes along thefacade in such a way that they clearly correspond to each of the three parts The central space that is accessed on the first floor by stairs in the southern tower, in the groundfloor is transversally divided by two elongated pillars and the corresponding lesenes along the side walls (it is in fact a wall with an arches). The elevation thus shows tripple, probably sem~circular arches, that, so typical ofCarolingian architecture, make the space even more monumental. Unlike such articulation of space within the westwork itself, access to the church in the groundfloor is possible only by one central passage, but we are unaware ofthe way the central westwork hall on the first floor opened towards the church. The groundfloor, the crypt of the westwerk was vaulted as proved by the disposition uf supporters and the stairway for entering the first floor. The function of the crypt was determined not only by vauits, but also by the area where noblemen were buried since the sarcophagus ofqueen Jelena was placed here in 976 Studying these westworks chronologically it might be said that churches with westwerk did exist in Croatia at least from the mid 9th century. They were built in the follOwing order: Lopuška glavica and Žažvić in the mid century, St. Saviour at Cetina and St. Cecilia during the reign ofprince Branimir, followed by St. Stjepan at Otok built before the year of 976. Chronologically speaking Crkvina should also he dated to the mid 9th century, after the construction of the church, simultaneously with Lopuška glavica and Žažvić. The westwork at Crkvina could have been of a somewhat earlier date if the master craftsman of the \u27Kojjane pluteus" was to be considered the parameter for the first liturgical church furnishing The genesis of the westworks cannot be studied in our regions, since it came as a final product with specific, clearlyprogrammed intentions, in my opinion, nearly as a political act. Since it had to conform to specific functions, builders of our westworks were not allowed any greater variations in shape. Inspite of thiS, they proved their skill, within given circumstances at least on churches with round buttresses that, by vaulting greater spans in monumental churches, undoubtedly represent a contribution to the European architecture of the pre-Romanesque. lt seems, however that, follOWing the paragenesis offorms we can notice that the 9th century builders Wt:re vt:ry active in introducing new shapes in such a way that the belfries that lied in the axis ofpreRomanesque churches can be explained not only as a Carolingian influence on Croatian architecture, but quite the contrary, as a final reduction of the westwork as a form (it has nothing to do with the function) from examples already existing in Croatia. There are two facts that support this interpretation. Basically the westwork too, cannot be compared within a somewhat greater area (Istria, northern Italy). This may be interpreted, as already said, by direct Carolingian influence, but for which reasons? W\u27by would some forms be introduced from far away passing over great areas in order to settle in these particular region:; if, unlike for the westwork, there was no special reason for it The preseroation and transformation of westwork forms in Croatia may be followed in several directions. The axial belfry survived in the Croatian Romanesque. By forms -emporae on the first floor, and by functions -they are mostly private churches, they support the hypothesis that the pre-Romanesque axial belfry in Croatia developed by reducing westwerkforms. In the 11th century pre-Romanesque there were however several western structures, in front of the Church of ss Peter and Moses and the one ofSt. Laurent in Zadar. The western stnlcture in front ofss Peter and Moses did indeed result from the pre-Romanesque, but it no longer bears the shape ofearlier westworks and is isolated by its structure, shape and spatially as a separate structure beyond the rectangle of the church, By its shape it is similar to Romanesque tower-porticos that by theirform and partly by function evolved from the westwork, W\u27bile we are able to follow the continuity of the axial belfry and the transformation of westwerk into a tower-portico, the Church of St Laurent in Zadar indicates that the well-known shape was preseroed for a longer time, From the chronalogical picture ofpre-Romanesque westwerk in Croatia it is evident that nearly all of them, like most of the axial belfries, belong to the 9th c, The first known exapmles may be dated to the mid 9th century, the rule ofprince Trpimir, After determining their function, we have to determine the time they appear in Croatia and why. To answer to this question we should refer to the first half of the 9th century, In the beginning of the 9th century, the Caro!inigian army came to the borders of the then Roman province Dalmatia, establishing previously the boundaries of the empire, Baltles against the Avars, the conflict with Byzantium, reqUired the southeastern boundaries of the empire to be kept permamently safe. The peace in Aachen established solid boundaries for fields of interest within which the recently established Croatian state belonged to the Franks, In the follOwing decades the Franks keep on prOving their presence and influence in Croatia, Crafstmanship, swords and spurs are only outside indicators of more permanently established contacts, but meant nothing else but trade, The Carolingian control over Croatia has to be considered through two enicial aspects -the church and state that were in the Carolingian time, inseparable forces. This had to be provided as to contribute to the establishing oj theyoung state and its church authorities Priests were in charge oj both -C;umper1us at Bijaći, abbot \u27Jbeodebenus at Nin, and many others before them and after, Liturgic objects brought along such as reliquaries, thuribles, and the cult of saints they worshipped, became part oj the church inventories along with liturgic books, Christianization of the Croats definite!y got its greatest impulse Jrom the west. This is, among other; proved by churches bui!t in the Jirst half of the 9th century, and the later anes as well, The Jirst churches known to usJollowed the pattern ofregiOns close to northern Italy, together with the liturgy that required several altars in the church. However, nowhere was the Frankish presence so evident as in Carolingian westwork, mostly because it united and symbolized the sacral andprofane authorities. Carolingian westwork reveals elements of administrative structure, feudal relations that did not exist in Byzantine Dalmatia, liturgic practice typical for the west, and most ofall, the power ofthe feudalists. This unification of the church and the lay power is already sybolized in considering King as Christ\u27s legate on earth. This explains the liturgic scenography of the Aachen Chapel. Royal monasteries were built, royal churches, the royalty was buried in their monasteries. The situation was very much the same in our region. The Monasetrium regale ofSt. Bartholomew near the royal site ofKnin, the Crkvina at Biskupija as the mausoleum, so significant that it became the cathedral of the Croatian court bishop, St. Stephen at Otok where the queen s sarcophagus was guarded by a group of monks, etc. are all prove a well-established system of ruling and church administration. An entire archaeological stratum -the upmost one -materializes the Carolingian programme presence in Croatia in the 9th c. It was implemented by establishing institutions of the system into the Croatian society. And these very institutions reflect the westwork itself. We have already seen what the westwerk can possibly mean in architecture and what its versatile and different functions might have been. The first item in question for our region could be regarding the person who built the westworks. We have already seen that they were rulers, disrict-prefects. In such a way large churches, monasteries, churches ofpower-holders were crucial points of the sate policy, and the purpose ofthe westwork was to gather the authorities ofthe province. When it comes to private churches, it may be said that our churches are, if not royal churches, then private anes of noblemen, from district-prefects onwards. Ofcourse, there are no written proofs about this from the 9th c, but knowing that private churches existed later, and that they were introduced from the Frankish world, and that westworks were often private churches, clearly supports our case as well The cult of St Saviour played an important part in the promotion of the state as well. Considering Christ equal to the emperor transformed the Saviours cult in a way ofstrengthening the central administration. Therefore in the case ofSt. Saviour we do not have to search for a full liturgic sense other than the titular of the church. Otherwise, the lauds sang to the emperor, after the first Saviour\u27s name, invoke Mary, then the archangels and St Stephen. Is it necessary to remind that apart from St Saviour our churches that had westwork, were also dedicated to the virgin Mary (Crkvina) and St Stephen (at Otok), and both of them perhaps to both, as previously said. Among the complex functions of the westwork are liturgical anes with those granted to the authorities and the westwerk, as an architecturalframework for implementing such policy symbolizes the power of such state, By these functions it is no longer only a glorification of the Saviour, and therefore the emperor, but also a means of the feudalization policy. Those are the reasons for which the westwork should be searched for in Croatia as early as the first half of the 9th c, within the process ofestablishing civil and also church authorities