The monastery of Eusebius in Khirbet er-Ras (Kefar Truman)

,


INTRODUCTION
Khirbet er-Ras is located in the northern Judean shephelah, south of the road to Niblat, on the alluvial land of Kefar Truman (map ref. 1930/6540; 67 m asl).The ancient site of Khirbet er-Ras (Fig. 1) was identified following the discovery of the remains of a church (including mosaic pavements), foundations of fieldstones, and pottery from the Byzantine and Early Islamic periods. 1 The site was first excavated in 1958 by Varda Sussman on behalf of the Department of Antiquities.Beside the documentation of a winepress and a cistern filled with soil as well as of the ruins of ancient buildings visible on the ground, the most important discovery was that of the remains of an early Christian monastery church paved with mosaics (Fig. 2). 2 Unfortunately, except for the subsurface foundation walls, almost nothing of the built structures of the church was left intact.Most of the walls appear to have been destroyed already in ancient times and by later agricultural activity.
Further wall and floor remains of the Byzantine period were documented in the course of a later excavation conducted on the same site, but at some distance from the location of Sussman's original dig. 3 The preserved remains were assigned to two different chronological phases.Walls built of fieldstones and floor sections composed of flat stones were dated to the later (upper) phase.Another fieldstone wall represented the earlier (lower) phase alongside another wall covered with stone on the outside and filled up with soil.A sunken oval hearth with a step in its northern section likewise dated from this phase.Unfortunately, it was impossible to identify a larger coherent structure based on the above-described elements.
Even though the Kefar Truman mosaics have been mentioned in short reports and have been very briefly described in several catalogue entries during the past half century, 4 no detailed publication of the site has appeared to date.The present paper thus seeks to make V. Sussman's excavation results available to the public.

THE MONASTERY
The excavated structure The three-aisled basilica was oriented towards the east (Fig. 2).Even though the building's eastern section was not completely excavated, its main elements were uncovered 4 BIRAN 1959, 32; YEIVIN 1960, 46; ZVILICHOVSKY 1960; OVADIAH  1970, 101-102, No. 94; COHEN 1975, 309;  OVADIAH-OVADIAH 1987,  90, No. 146; SCHICK 1995, 364; BAGATTI 2002, 211; MADDEN 2014, 95,   No. 129.  during the dig.The basilica had a central nave measuring 4.32 3 9.50 m and two identical side aisles (1.73 3 9.50 m each).According to the original field record, both pastophoria rooms (diaconicon: 1.4 3 1.7 m, prothesis: 1.73 3 2.10 m) were uncovered during the excavation.Although the original ground plan appended to the field report (on which Fig. 2 is based) indicates a possible internal division both within the prothesis and the diaconicon, no further information is available on this.Between the two pastophoria, the one-time existence of a paved surface, the location of the apse and the steps leading to it from the nave can be reconstructed, of which, however, hardly anything has survived.Thus, it is difficult to determine the original shape of the apse.Given that no clear evidence pointing to the one-Fig.2. Plan of the church with its annexes (drawn by Magda Eber (Institute of Archaeology, Research Centre for the Humanities) after V. Sussman's original plan, kept in the Israel Antiquities Authority Archives, Folder of excavation files: Kefar Truman, Varda Sussman, &-21/1958) time existence of an internal apse was documented, it is possible that the church originally had a rectangular instead of a regular round apse. 5However, there is also the possibility that no architectonically distinct apse was built, as in the case of the H _ orvath H _ ermeshit church. 6The area where the apse is assumed to have been was paved and there must once have been a screen in the front of the space with the mosaic inscription (see below).
As customary, the narthex is located on the western side of the basilica (3.2 3 9.1 m).Since the above-ground walls of the church are missing, it is impossible to identify the location and number of the entrance(s).It may be surmised that there was a single main entrance.Neither have any traces of one or possibly more entrances leading from the narthex to the basilica been found, nor of column bases, nor of walls that separated the nave from the aisles.The single indication of the separation of the central and side spaces is the marked discontinuity of the excavated mosaic pavements.As is clearly visible on the archival photos taken at the time of the excavation (Figs 3 and 4), the central rectangular paved area is unmistakably separated from the mosaics of the aisles by a ca.0.65 m wide unpaved strip, covered with earth on the photos, on its southern side, and another ca.0.7 m wide strip, filled with earth and fairly irregular stone blocks, on its northern side.The majority of the stones from this area were probably removed for recycling as building material in later periods.Since the aisles are quite narrow, perhaps one main entrance served the entire basilica.
Additionally, two wings, a northern and a southern one, were identified during the excavation.They are neither identical in terms of their plan, nor were they precisely adapted to the dimensions of the basilica.
The internal dimensions of the northern wing are 3.65 m wide by 11.5 m long (if the walls are included, the external dimensions are 5.4 m by 15 m).Traces of an internal dividing wall were also uncovered, which divided the building into two larger rooms.Similarly to the main church, neither the entrance of the northern wing, nor the door of the internal dividing wall could be identified.No substantial traces of the original wall were found on the Fig. 3.The church with its mosaic floors during the excavation, seen from the north-east (photo: Israel Antiquities Authority Photo Archive; No. 23301) northern side, although the presence of several stones seems to indicate the wall's original line.Likewise, additional stones appear to mark the one-time existence of two supporting columns attached to the wall from the outside.There is a courtyard paved with large stones along the western façade (Fig. 5).Here, the width of the building's western wall was enlarged to 2.6 m, possibly reinforced after an earthquake, or to provide support at a time when the northern courtyard was covered with a roof.
The southern wing was 5.6 m wide and extended along the length of the basilica, i.e. it had a length of 18.5 m.As in the case of the northern wing, a courtyard paved with carefully laid large stones adjoined its western wall.The walls, the entrance to the main courtyard (2 m in width) and the threshold of the door (0.2 m above the floor level) were preserved here.According to the original field record, traces of an entrance (2.8 m in width) paved with stones similar to those of the courtyard leading from the courtyard to the nave were also identified (although not recorded on the ground plan).The eastern hall had three rows of columns, each with two columns attached to the longer walls and a single one in the hall's central longitudinal axis.Only the foundations of the central pillars remained of the three columns.The columns were built of large stones, in all probability as necessary supports of the roof.The eastern wall of the hall was not discovered, and the walls continued eastward.It seems likely that there were entrances from the wings to the main basilica, but their secure traces, if any, were impossible to document in the course of the excavation.The only wall that was well preserved was at the western end of the northern wing.It was excavated to a depth of 2 m below the surface (Fig. 6).A pair of identical columns supported the outer wall of the southern wing, too.
The stones used for the construction of the building were most probably quarried from the nearby hills.Remains of plaster were discovered on the walls, indicating that they had been plastered on the inner sides.A small number of roof tiles were also discovered, and thus it can be assumed that the structure was covered with a tiled roof.Based on the method and style in which the walls were built, the excavator concluded that the northern and southern wings were probably added to the basilica later, some time after its initial construction.According to the original field report, traces of a later structure erected in the narthex were likewise discovered.There was another structure near the entrance to the southern wing that was connected to the southern wall of the basilica.Neither the date, nor the exact function of these later structures can be established with any degree of certainty.

The mosaic floors
The narthex was paved with white tesserae of ca. 2 3 2 cm in size.The excavator noted that the presence of carelessly The white mosaic strips framing the central panel on the northern and southern sides are fairly narrow in width and do not bear any decorative motifs (Figs 3-8).In contrast, the wider strips at the western and the eastern ends of the central carpet, near the entrance of the church and in front of the sanctuary area (Fig. 7), are further decorated with lozenges in two alternating sizes, made of brick red and black tesserae on a white background according to the following pattern (Class E in Avi-Yonah's classification7 ).The centres of the lozenges are made up of either one or four white mosaic stones.Brick red squares, each made up of one or four mosaic stones, were set on the four sides of these white squares, while the outer three sides of the brick red squares were flanked by black squares, each made up of either one or four tesserae as shown on Fig. 9.2.The sizes of the larger lozenges flanking a tabula ansata (see below) vary between ca. 14 3 14 cm and ca.The only major difference between the mosaic strips at the western and the eastern ends of the central carpet is that the latter, located in front of the sanctuary area, contains a five-line Greek inscription enclosed by a tabula ansata (ca.93.5 3 143.5 cm in size, Figs 7 and 10, see below).The letters vary from ca. 10 cm-15 cm in height, their majority falling between ca. 12 cm and 13 cm, and are set in black tesserae against a white ground.The bands enclosing the letters are ca.12-13 cm to 16-17 cm in height (decreasing in height from top to bottom) and made up of white tesserae.The lines framing the entire inscription as well as the ones separating the single bands are of brick red stones, set in one row.The latter lines are separated by a double line of white tesserae (ca.3.0-3.5 cm) from the black outline of the tabula ansata framing the composition.The bases of the triangular "hands" of the tabula ansata are 38 cm-40 cm in length, while the two other sides are 33.5 cm-35.5 cm in length.The triangles are outlined with black stones in one row on a white ground, and both include three black tesserae attached to the line of the base and a small rosette made up of five brick red stones (5.5-6.5 cm in height) in the third of the triangle closer to the inscriptional field.
The central panel (Fig. 8) is framed by a tripartite border (Figs 7 and 11), whose outer framing lines are made of black tesserae in one row.9Identical black lines frame the border's central band, too, which is filled with a finely made simple guilloche ornament (Class B2 in Avi-Yonah's classification, 10 a type of interlace or Flechband ornament) crafted from white, brick red, mustard, and black mosaic stones (ca.18 cm-21 cm in width).Both the outer and the inner bands enclosing the central band are identical (the inner being 11 cm-14 cm, while the outer 13 cm-15 cm in width): rhythmically placed triangles follow each other along the entire composition (Classes A5 and A6 in Avi-Yonah's classification 11 ).Their bases are constituted by the border's outer black lines.The triangles' outlines are likewise made up of identical black tesserae.The inner spaces are filled with brick red mosaic stones (up to four tesserae, i.e. 5.5 cm-7 cm in height).Strictly speaking, the border is tripartite, to which a fourth and final element is attached, which separates the border from the central field of the central panel and which is made up of two lines of plain white tesserae flanked by the inner black outline of the border and the outer black outline enclosing the central field.
The central field (ca.281 3 573 cm) framed by the mentioned single line of black tesserae and enclosed by the above-described border is evenly filled with floral semis, made up of the repetition of small vegetal ornaments against a scalepattern background (Fig. 9.1; Class J3 in Avi-Yonah's classification 12 ).The entire composition is set against a white ground.The scales are outlined, the outer lines are marked with a line of black, while the inner lines with another line of brick red tesserae.The small floral ornaments in the centres of the scales are crafted according to the following pattern (Fig. 9.3): their short stems are marked by single black mosaic stones, each calyx is created out of five whole and two halved black tesserae arranged in V-shapes, while the petals are formed of five whole and two halved brick red stones.The heights and the widths of the floral motifs are 8 cm-10 cm and 9 cm-10 cm, respectively, while the heights of the scales vary between ca.33 cm and 38 cm.When viewed from a certain distance (Fig. 12), the unevenness of the scale pattern's distribution is not particularly noticeable; however, a closer look quickly reveals its oddities.Taken together with the medium size and rough workmanship of the tesserae, however, it amply illustrates the pavement's mediocre quality.
Besides the eastern and the western ends of the nave mosaic, the simple lozenges recur in the eastern section of the southern aisle's pavement, too (Fig. 13).(Neither of the aisles' mosaics are preserved.)According to the excavator's notes, the latter mosaic was divided into three different sections, of which the available archival photos document only the eastern part (Fig. 13), as well as the westernmost extremity of the pavement, where a ca.40 cm wide and ca.20-25 cm deep depression for collecting water, paved with white mosaic tesserae, was preserved (Figs 14 and 15). 13The pavement in the northern aisle was divided into two sections according to its decorative pattern.
All decorative elements of the above-described pavements enjoyed extremely wide popularity in the mosaic art of the late antique eastern Mediterranean in general and particularly in the provinces of Palaestina Prima, Palaestina Secunda and Arabia.The semis of the nave's central panel embedded into the scale imbrication pattern background was one of the highly popular mosaic carpet designs from the fifth century onwards. 14It was employed both as a self-contained decorative element, as in our church, and as a background to animal figures and animal combat scenes integrated into the carpet design, as on the famous fifth-century Phoenix mosaic at Antioch, 15 a fifth-century mosaic from the environs of Hama, 16 the pavement of the later fifth-century Michaelion at H _ uarta, 17 the late fifth-earlier sixth-century pavement in a private house in Androna, 18 and the narthex mosaic of a sixth-century church at Hanita, 19 to name merely a few examples. 20Scale patterns were created using both plain and outlined versions of the scales, but one pavement always used solely one of the two types.For understandable reasons, the scales' plain variant was preferred when creating a background for an animal imagery, while outlined scales predominate in panels without additional figures.
Focussing now on the pattern's independent usage, several close counterparts of our mosaic can be mentioned both from Palaestine Prima, the Byzantine province in which our site is located, and from more distant regions of the eastern Mediterranean.To begin our overview with examples closer to our site in geographical terms, mention may be made of the province's most important political and cultural centres, Caesarea and Jerusalem.In the provincial capital, both plain and outlined scales with florets appear on at least three pavements in the city's northern area: on a mosaic of the Samaritan synagogue erected east of the Byzantine dux' palace, 21 as well as on two pavements of a large luxurious private mansion, including the finely crafted mosaic of a long corridor. 22  TALGAM 2014, 116, Fig. 162.20   For further examples, see LAVIN 1963, 195; DONCEEL-VOÛTE 1988, 69-77, Fig. 43

200
Acta Archaeologica Academiae Scientiarum Hungaricae 72 (2021) 1, 193-213     the mosaics of the Church of Eleona, not far from the former, 24 on the mosaics in the nave and the northern and the southern aisles of the fifth-century Church of the Siloam Pool, 25 and in the northern apse of a church on Mount Zion. 26he same pattern was also employed by the mosaicists working in the wider area of Jerusalem.For example, it can be found on the sixth-century mosaics of 'Room 1' of a monastic complex at Kh an Saliba, east of the Holy City. 27It also appears in the narthex and the southern aisle of a church in 'Ayn al-H _ ann ıya, ca.7 km south-west of Jerusalem, 28 in the southern aisle of the mid-to later fifth-century Northern Church of the Herodion, 29 on the narthex mosaic of the fifthcentury Cave Church and on one of the bema panels in the sixth-century basilica at the Shepherds' Field 30 as well as among the mosaics of the monastery at Khirbat Siyar al-Ghanam 31 and of a church (?) at Khirbet Luqa. 32Outlined Fig. 12.The mosaic carpet of the nave seen from a certain distance (photo: Ad am Boll ok, © Eretz Israel Museum, Tel Aviv)   CORBO 1955, 147, Tav.49, Fot.159-160; OVADIAH-OVADIAH 1987,   97.scales filled with florets also figure prominently both in the sixth-century and the Umayyad-period phases of the Kathisma Church, where the apse mosaic of the south-west chapel date to the former, while the apse mosaic of the northwestern chapel and the central panel of the north-west corner can be assigned to the latter period. 33The pavements in the inner southern aisle of the Basilica of the Nativity in Bethlehem 34 and in the narthex of the Khirbet 'Asida church 35 have small squares within the scales instead of the usual florets, while on the ca.fifth-century, roughly contemporaneous intercolumnar mosaic in the Khirbet Jufra church 36 scales appear without florets.Farther to the south, outlined scales with florets adorn the eastern half of the central mosaic panel in the monastery's chapel at Khirbat al-Qas _ r37 and the same pattern covers the sixth-century northern aisle mosaic of the Central Church at Bayt 'Ayn un, 38 while along the shores of the Dead Sea, outlined scales with very schematic florets can be found among the decorative elements of the late fourth-to fifth-century mosaic pavements in the western aisle of the 'Ein Gedi synagogue. 39Moving in the opposite geographical direction, north of Jerusalem, and at the opposite end of the chronological range, a mosaic pavement with a floret pattern set into outlined scales also appears in the north-western aisle of the earlier eighth-century Jericho synagogue, which clearly attests to the long-standing acceptability of this decorative scheme in synagogue art. 40Yet, even in the latter city, this decoration is not restricted to synagogues, as indicated by a sixth-century pavement from the northern aisle of a church. 41arther to the north-west, the north aisle of the Northern Church at Shiloh was paved with a mosaic embellished with outlined scales enclosing small leaves in the sixth century. 42n Church A at Magen in the western Negev, the western panels of the northern and the southern aisles, which were most probably created in two different phases, with one copying the other, also bear outlined scales with florets. 43lose to Magen, at Horvat Be'er Shema', the apse pavement of the church, believed to have been set during the last decades of the sixth century, is decorated with florets embedded in scales. 44Another church edifice in the Negev, the Western Church at Mampsis (in Palaestina Tertia), provides examples of the scale pattern, this time without florets, on the fifth-century intercolumnar mosaics. 45In the coastal area, a variant of our pattern decorated a portion of the narthex mosaic of a church in Ashkelon-Barne'a. 46loser to our site, in Hazor-Ashdod, the early sixthcentury pavement in the southern aisle of a church building displays a scale pattern, 47 while at H _ orbat Sokho, located approximately halfway between Khirbet el-Ras and Jerusalem, another ca.fifth-century pavement was discovered with outlined scales and florets in the northern room of what was tentatively identified as a church building. 48Outlined scales with florets likewise appear on the late fourth-to early fifth-century panel in the southwestern intercolumnation of the Samaritan synagogue discovered in Ramat Aviv. 49However, in geographical terms, the closest site with florets embedded in a scale pattern is Mazor.It is thus quite unfortunate that neither its date, nor the function of the site can be established with the necessary degree of certainty, 50 not least because its rather slipshod quality also seems to match that of the Kefar Truman pavement.
Turning to the north, outlined scales with florets embellish the southern aisle of a church in Bahan. 51In Samaria, our motif appears in the eastern room of the annex building erected along the northern wall of a Byzantineperiod church in Abud. 52Two mosaic fragments, one with plain scales, the other with outlined scales, both with florets, associated with the original late antique church building were uncovered in the porch and the southern aisle of the medieval Church of Saint John the Baptist in Sebaste, ca.40 km north-east of Khirbet el-Ras, 53 while a further fragment came to light from the adjacent monastery. 54arther to the north, scale mosaics with florets were discovered in Scythopolis/Bet She'an, the capital of Palaestina Secunda, both in Christian churches and in public edifices.To begin with the latter, a large earlier fifth-century (?) mosaic pavement embellished with a pattern of endlessly repeating plain scales with florets was discovered on the western covered portico of the Palladius street. 55The same pattern appears in Rooms 5 and 14 of the slightly earlier (ca.400) Nile Festival Building at Sepphoris. 56In contrast, Christian contexts from Bet She'an, such as the small niche south of the west door of the late fifth-to early sixth-century Round Church, 57 the sixth-century aisle mosaics of a church at Tel Estaba and the pavement of its north-east chapel, 58 and the central section of the mosaic panel of Room E as well as the apse mosaic of the monastery chapel of the mid-sixthcentury Monastery of Lady Mary 59 provide evidence for the presence of florets enclosed by outlined scales.Immediately west-northwest of the city, scales with florets also embellish the sixth-century mosaic discovered in the south room of a chapel in Sede Nahum as well as the mosaic of 'Room 2' in the monastery excavated at Tell Bazul. 60 . 63Still in Palaestina Secunda, but farther to the east, the nave and the aisles of the great fiveaisle basilica of al-Suwayd a' were paved in the late fifth or in the sixth century with florets enclosed in plain scales. 64he appearance of this pattern is also recorded in the opposite geographical direction, among the earlier, most probably fifth-century mosaics in the southern aisle of the Basilica of the Annunciation in Nazareth. 65It was likewise reported to have been used on the later sixth-century upper pavement in the apse of the South Chapel at Kafr Kama. 66he same pattern recurs in the bema and the nave of a church building excavated at Khirbet Samra on the eastern shore of the Sea of Galilee. 67The late fifth-to early sixthcentury mosaics in the nave and later sixth-century pavement in the baptistery set up in the south-eastern room of the monastery church at Kursi/Gergesa bear plain scales with florets. 68Not far from Kursi, florets set in outlined scales decorate the large central panel of the sixth-century pavement of the northern aisle of the North-West Church at Hippos/Sussita, while in the southern aisle and the southern sacristy, floral semis embedded in plain scales cover the entire surfaces. 69The apse of the South-West Church is likewise embellished with florets enclosed in outlined scales, 70 and the same motifs appear on the late sixth-century pavement in the baptistery of the South-East Church (the Cathedral). 71n the opposite side of the Sea of Galilee, plain scales appear in the eastern room of the northern wing of the church on Mount Berenike in Tiberias, 72 while outlined scales can be found among the mosaics adorning the eastern aisle in the Lower Synagogue at Hammath Tiberias. 73As we have seen in the above, the latter pavement is hardly the single occurrence of this decorative system in synagogue art.Yet, churches remain the main contexts of our pattern.To the north of Tiberias, at Migdal/Magdala, 'Room 17' of a Byzantine-period monastery was paved with plain scales enclosing florets, 74 while halfway between Tiberias and Khirbet Samra, at Khirbat al-Karak (Bet Yerah), the main sanctuary area and both pastophoria were paved with mosaics of this type in both sixth-century phases of the church. 75At the opposite, northern side of the lake, outlined scales fill the central space of the fifthcentury Octagon at Capernaum 76 and they reappear in the mosaic panels of the bema and the northern nave in the fifthcentury mosaics of the Church of the Multiplication at nearby Tabgha. 77In the latter village, plain scales with florets decorated a mosaic panel in the hall of the chapel of the monastery erected on the 'Mount of Beatitudes'. 78ites in western Galilee can also be mentioned, which brings us to the eastern border region and south-eastern end of the province of Phoenicia Maritima.To name but a few, let us refer to the plain scales and their florets covering the nave of the church at Horbat Medav, 79 the ones in the intercolumnar spaces of the earlier sixth-century church at Horbat Hesheq, 80 and the same design in the south-eastern annex room of the somewhat later church at Kirbet el-Ghureiyib. 81Beside church edifices, the same design was also employed in domestic contexts in the region, as the recently published seventh-century mosaic floor discovered at Pi Maz _ uva demonstrates. 82Florets set into outlined scales are displayed in the mid-sixth-century southern aisle of the Khirbet Bata church, 83 on the mosaics set adjacent to the narthex pavement of the sixth-century church at Horbat Kenes 84 and on the later eighth-century pavement of the northern aisle in a church at Khirbet el-Shubeika. 85n the coast, in the province of Phoenicia Maritima, in Shavei Zion, the entire early fifth-century pavement of the nave of a church is covered with florets integrated into plain scales.Another mosaic panel decorated with florets, this time enclosed in outlined scales, appears among the decorative elements of the later fifth-century pavement of the north-eastern chapel of the same church. 86A comparable design can be found in Nahariya, immediately north of Shavei Zion, where plain scales with florets grace both the nave and the two aisles of the probably earlier sixth-century pavements in the church, and the same pattern adorns the fragment of a pavement set in 'Room 1' of the building attached to the southern wall of the church edifice. 87In the neighbouring settlement of Evron, the late fifth-century mosaic of the narthex bears plain scales with florets, too. 88lorets embedded plain scales dominate among the Tel Shiqmona mosaics, where pavements decorated with this pattern can be counted among the very popular ones. 89lose to Tel Shiqmona, at Kiryat Ata, outlined scales with florets covered the sixth-century mosaic of the nave. 90o the north, in Dayr al-Zahr an ı, south of Sidon, florets inscribed in plain scales appear on the earlier sixth-century narthex mosaic, 91 and in all likelihood the same pattern adorned the later sixth-century pavement covering the entire interior of a church building at Nab ı Y unus, north of Sidon. 92The same pattern was also documented on the midsixth-century nave mosaic of the Upper Church at Kh an Khalda, south of Beirut, 93 while in Beirut itself, florets inscribed in plain scales decorate a mosaic pavement discovered in a villa building dated roughly to the fifth-sixth centuries. 94s we have seen in the above, the predominance of florets inscribed in plain scales characterizing the churches of Phoenicia Maritima is not universal elsewhere.The same holds true for the Syrian provinces, too, where a more even distribution of the plain and the outlined variants can be noted.In the northern regions of Syria Prima, the later fourth-or perhaps fifth-century mosaic panel in the southern annex room of the eastern nave of the suburban martyrion of Saint Babylas at Antioch 95 displays the outlined versions, while a predilection for the plain variant can be noted among the possibly earlier sixth-century pavements of the nave and both aisles of the Machouka church, located outside the walls of Antioch. 96Additionally, mention can be made of a pattern made up of outlined scale motifs without florets on the mid-fifth-century mosaics of the martyrion of the Church of Julianos at Br ad. 97Yet, a cautionary remark is also in order as to the frequency of florets with plain scales, because in household contexts at Antioch, which constitute the major source of our knowledge of Roman to late antique mosaic production in Syria Prima, the predominance of plain scales is more than evident.(See, e.g., the earlier-to mid-fifthcentury small apse in the House of the Buffet-Supper, 98 the famous fifth-century phoenix mosaic, 99 the later fifthor early sixth-century pavement set in an apse of the later phase of the House of Aion, 100 and the earlier sixth-century mosaic in the centre of a room in the House of the Bird-Rinceau. 101n contrast, the exact opposite can be said of the mosaic floors of late antique churches on Cyprus, where outlined scales were the preferred type, 102 which also appear on Crete. 103In more western provinces of the Later Roman Empire, the use of scale patterns filled with the floral semis was rather limited. 104Instead, as an unbroken continuation of earlier Roman tradition, 105 polychromatic scales were preferred both in ecclesiastic and profane contexts. 106The above-described tendencies are copiously attested in the archaeological record of late antique Anatolia, too.Here, in the regions closer to the Syrian provinces, plain scales constituting the background to animal imagery are frequently documented, as at Edessa/S ¸anliurfa in Oshroene, 107 Germanicia Caesarea/Kahramanmaraş 108 and Korucak K€ oy 109 in Euphratensis, near Alimpinar in Armenia Secunda, 110 and in Cappadocian Parnassus. 111Outlined scales enclosing florets 112 and triangles 113 are also documented both in these regions and slightly farther to the west, e.g. at Eleaiussa Sebaste in Cilicia Prima. 114In the more western provinces, however, the use of polychromatic scales is hardly unusual, as at Tlos in south-western Anatolian Lycia. 115n the province of Arabia, outlined scales enclosing florets seem to appear only slightly later in the currently known mosaic record than in the two Palaestinae and they play a more restricted role in mosaic decoration.This pattern occurs relatively rarely on its own covering larger surfaces.On a late sixth-century pavement in the Church of Saint Basil at 91 DONCEL-VOÛTE 1988, 424-425, Fig. 422.92 DONCEL-VOÛTE 1988, 407-409, Fig. 400.93 DONCEL-VOÛTE 1988, 374, 380, 383, Figs 357, 369.94 TURQUETY-PARISET 1982, 3, 6, 14, 20, Figs 15-16.95   LASSUS 1938b, 25, Fig. 24; DONCEL-VOÛTE 1988, 28-29, Fig. 8.   96   LEVI 1947, 368-369, Pls CXLd, CXLIa-c; DONCEL-VOÛTE 1988, 175-176, Fig. 145.97 TCHALENKO 1979-1990, Pls 16-17; DONCEL-VOÛTE 1988, 39-43,   Fig. 17 R ıh _ ab, the apsidal area this pattern, 116 while in the Memorial of Moses on Mount Nebo, the earlier sixth-century U-shaped pavement around the baptismal font is covered by it. 117The same can be seen in the western ends of both the northern and the southern aisles of the Church of Saint Peter at Khirbat al-Samr a', paved in the earlier sixth century. 118An even smaller surface was allotted to the outlined scales, in this particular case without enclosed florets, decorating a narrow mosaic strip set along the north-western wall of the northern aisle of the eighth-century Church of Saint Stephen in Umm al-Ras _ as _ . 119 An only slightly larger area was decorated with outlined scales in one of the intercolumnar spaces separating the nave from the northern aisle in the later sixthcentury Church of Bishop Sergius, adjacent to the former edifice. 120Similarly, one of the intercolumnar spaces, this time between the columns separating the nave and the southern aisle, was embellished with outlined scales in the sixth-century Church of Procopios in Geresa. 121In the later sixth-century chapel at Khirbat al-Muniyya, some 6 km north of Geresa, only the pavement of the entrance leading from the narthex to the nave bears this ornament. 122n the foregoing, the widespread popularity of the floral semis ornament embedded in plain or outlined scales was demonstrated, and the same holds particularly true for the decorative elements used in other parts of the Kefar Truman pavements.A quick look at the extensive mosaic corpus of the Roman and late antique Mediterranean demonstrates the extreme popularity of simple lozenges, simple guilloche ornaments, and the rhythmically set triangles both as selfcontained ornaments and, in the case of the latter two, their frequent combination, as in the central panel's tripartite frame on the Khirbet el-Ras pavement. 123Unfortunately, given the long use and widespread popularity of these ornaments, they cannot provide a good chronological anchor for dating our mosaic.Yet, it is perhaps not a futile exercise to name a few examples where a simple guilloche (Caesarea Maritima, 124 Kiryat Ata, 125 Hippos/Susita, 126 Bethlehem, the Basilica of the Nativity, scales filled with squares 127 ), a row of triangles (Tiberias, Mount Berenike 128 ) or their combination (Jerusalem, Mount Zion, 129 Kh an Khalda, scales filled with florets and squares 130 ) frame a scale pattern.
Two distinctive traits of our pavements among the period's monuments are the lack of figural scenes, even though this was hardly unknown in the period's mosaic art, 131 and the internal arrangement of the otherwise oft-recurring decorative schemes within the church space.As to the former, the need for due caution must be emphasized in view of the complete loss of the sanctuary mosaic.As other examples clearly demonstrate, a minimal figural decoration was sometimes introduced even into overwhelmingly aniconic decorative schemes.To mention but a few telling examples, let us refer here to the pavements discovered at Khirbet Samra 132 and the one known from Khirbet Beit Sila. 133While fully aware of the above constraint, the use of an exclusively aniconic decoration both in the central and the side spaces does not seem to be a particularly common choice in the period's church art in the southern Levant.Yet, completely preserved sets of pavements characterized by a rich array of decorative patterns and a high quality of workmanship, like the ones excavated in the Northern Church of the Herodion 134 and in the Western Church at al-Yas _ ıla, 135 to name but a few sites, underline that neither the appearance of church pavements made up exclusively of aniconic decorative elements is entirely surprising, nor can they be ascribed to limited funds of the pavements' patrons or the lack of the appropriate artistic skills of their craftsmen.
The latter observation takes us to our second point.As noted in the above, the mediocre quality of the Kefar Truman mosaics is indicated by several features.Although it is hardly unusual for geometric pavements to be normally set with larger tesserae and to employ a more limited array of colours than in the case of figural ones, 136 the use of fairly large and irregularly-cut mosaic stones for the Kefar Truman pavements cannot merely be explained by this practice.In the spots where the density of tesserae is the highest, namely the inscribed panel and the tripartite frame of the central panel (Figs 7, 10 and 11), it varies between 55 and 65 stones per square decimetre, while in the central field of the central panel it rarely exceeds the 40 and 50 stones per sq.dm.The slight, but well-discernible variance in the sizes of the unevenly spaced scales points as much in the same direction as does the use of a limited number of colours (four) 137 and the placement of the lozenges against a plain white background without the slightest to employ an ornamental element that would merge them into a unified pattern.Taken together, one can hardly escape the impression that in this particular case, a monastic community with rather limited financial means hired a workshop of secondary importance.Whether this state of affairs played any role in that they opted for an entirely geometric design, or whether this choice was rather influenced by the views of the community commissioning the mosaic as to what can be considered as appropriate and acceptable for decorating sacred spaces, is more difficult to tell.What is quite certain, however, is that their choice of adorning the entire nave with florets enclosed by outlined scales was a fairly uncommon solution.A quick look at the above list of the use of florets inscribed in scales reveals that in the majority of the known instances, this pattern was employed in side aisles, intercolumnar spaces, narthexes, baptisteries, and other less prominent spaces rather than in naves.In a sense, this is clearly a continuation of the age-old Roman habit of employing the scale pattern in marginal and auxiliary spaces. 138When we do encounter this pattern in naves, some regional factors seem to be in play.The appearance of the plain variant in the fifth-to mid-sixth-century churches in Phoenicia Maritima (Horbat Medav, Kh an Khalda, Nahariya, Nab ı Y unus, Shavei Zion, Tel Shiqmona, alongside the outlined variant at Kiryat Ata) is one obvious regional trend, while the preference for the outlined variant in the sixth century east of the Sea of Galilee (H _ awf a al-Wast _ iyya, Khirbet Samra, al-Suwayd a') might indicate another.The main difference between the plain and the outlined versions may be sought in their possible symbolic associations.In the case of the florets set against plain scales, it is not at all unimaginable that for the late antique beholder, the pattern in question evoked an association of a natural landscape, which, in view of the paradisiacal connotations of certain church spaces, would make it an ideal choice for covering central liturgical spaces.On the other hand, the outlined version with its strongly articulated geometric design would be less likely to conjure up such a direct association and was thus better suited as a space-filler rather than a bearer of some symbolic meaning.However, even in these cases it cannot be excluded that outlined scale patterns played a certain function beyond mere decoration.As simple scale patterns were often employed in Roman mosaic design to direct the beholders' view and attention toward certain directions, 139 the same role could have been fulfilled by mosaic floors covered by outlined scales filled with florets.It is thus perhaps no coincidence that in the Kefar Truman church's nave the scale pattern was set to direct the visitors' gaze from the entrance area towards the sanctuary, and not in the opposite direction.The lack of an explicit and unequivocal symbolism is perhaps one of the reasons why patterns made up of outline scales were considered as being appropriate for decorating the central spaces of both churches, synagogues, and private buildings in the Umayyad period, as shown by the examples of the Khirbet el-Shubeika church, the Jericho synagogue, and Walid II's bathhouse at Khirbet el-Mafjar. 140In any case, given the polysemantic nature of ornaments, their interpretation is largely open-ended and thus their occurrences in diverse contexts offer different potential readings.Thus, the appearance of outlined scales filled with florets in a fresco discovered on the narthex wall of an early Christian basilica at Eleutherna (Crete) 141 can be read in at least two different ways, either as "mere ornament" applied with the aim of beautification, or as a visual allusion to a natural landscape symbolized by the florets, which the beholder can admire through an openwork parapet wall denoted by the scale imbrication pattern (a reading hardly conceivable in the case of floors).

The inscription
Five-line Greek inscription written in round letters, set in a tabula ansata (0.93 3 1.42 m; Fig. 10).It was found almost intact, only the last fourth of the lower three lines are lost and two shorter sections of lines 1 and 2 are damaged.Despite these losses, the inscription is wholly legible.It was written in black tesserae against a white ground: † EPITOYQEO " † In the time of the most god-loving priest and the abbot Eusebius the nave was renewed." Based on palaeographic considerations, Leah di Segni suggested a date in the third quarter of the fifth century for the inscription. 142e finds Pottery.According to the original field report, a number of pottery sherds, oil lamps, and glass finds were collected during the excavation.Unfortunately, none of these objects were available for study at the time we attempted to locate them.
Metal finds.Two of the three metal small currently available for study from among the ones discovered in 1958 clearly postdate both the church's construction phase and the building's use for Christian religious purposes.The copperalloy fragment decorated with a continuous tendril ornament (Fig. 16.3) seems to be of medieval or post-medieval date, while the iron horse-shoe (Fig. 16.4) is a modern artefact. 143he small copper-alloy ring bent from a simple wire (Fig. 16.5) represents a very common form and therefore cannot be precisely dated merely on typological grounds.

Coins by Gabriela Bijovsky
Four coins were discovered during the excavations at the church in Kfar Truman. 144Unfortunately, the identification cards lack locus and basket information.All coins are folles made of copper.The earliest coin belongs most likely to Justinian I's undated series (527À538 CE) (No. 1).This is followed by two folles of Maurice Tiberius minted in Nicomedia.One is clearly dated to 599/600 CE (No. 2), while the other follis bears an illegible date (No. 3).The latest coin is an Arab-Byzantine overstruck imitation (No. 4).This series is roughly dated to the years 647-670 CE and was most probably minted in jund Filastin.

CONCLUSIONS Date
Based on the above data, it is not particularly easy either to ascribe a precise date to the church's construction phase or to reconstruct the site's later occupation history, not least because purely geometric mosaic pavements are notoriously difficult to date on their own. 145Given the closer and more distant parallels of the central pavement reviewed briefly in the above, a mid-fifth-to mid-sixth-century date could be proposed for the mosaic floor as a fair approximation.Yet, a cautionary note seems in order.Although it has been widely assumed for a long time that strongly stylized florets set into outlined scales, as seen on the Kefar Truman mosaic, should rather be dated to the sixth century, based mainly on the stylistic and chronological development of the mosaic pavements of Antioch, 146 a slightly more cautious approach is necessary.Even if the prevalence of plain scales among the Levantine pavements of the fifth century is still detectable (as rightly observed by F. Vitto 147 ), outlined scales also appear on mosaics dated to the latter part of that century (see, for example, the church on Mount Zion and the Northern Church of the Herodion 148 ).Without recapitulating here the diverging views on the origins and eastern connections of the structured semis decoration, 149 we would highlight a single point.The appearance of the elaborately-crafted florets embedded into outlined scales on the apsidal mosaic in the domina's apartment at Piazza Armerina, 150 set in the 320s-330s most probably by a North African workshop, duly explains how florets reflecting widely different levels of schematization appear at different sites already at the end of the fourth and the beginning of the fifth centuries. 151While the pavements of the Nile Festival Building at Sepphoris, dated to ca. 400, display more schematized florets than the somewhat later Phoenix mosaic at Antioch, the stylization of the florets of the 'Ein Gedi synagogue mosaic, dated to the fourth to fifth centuries, shows very advanced stage.The differences between the stages of stylization of the florets on the mosaic of the late fourth-/early fifth-century synagogue in Ramat Aviv, the early fifth-century pavement of Shavei Zion, the mid-to later fifth-century mosaic panel of the Herodion thus seem to depend as much on the quality of their workmanship as on their respective dates.It is therefore hardly surprising if florets representing more or less the same level of stylization do appear on a large number of pavements between the fifth and eighth centuries. 152 is also true, though, that more elegantly crafted semis ornaments rarely occur after the late fifth century, while the truly elaborately-crafted floret depictions rich in details, as the ones displayed on the church floor at the Cappadocian Parnassos, dated by its inscription to 469-470 CE, are currently among the exceptionally rare occurrences of that time. 153ince based on palaeographic considerations the inscription should be ascribed to the third quarter of the fifth century or possibly earlier, a later fifth-century date can be proposed for the Kefar Truman pavement.Neither does this date Fig. 16.Coins and metal finds (B-781167-781172) 152 Cf.DONCEL-VOÛTE 1988, 457. 153 ARSLAN et al. 2011, 196, 202-203, 205-207, Drawing 2, Figs 3, 6-11.contradict the (largely?)aniconic nature of the mosaic and the use of tesserae.
On the testimony of the nave's mosaic inscription, it is also clear that the church was rebuilt or renovated at least once during its existence.It likewise seems reasonable to assume that the northern and southern wings are later additions to the main church building, which might have occurred at the same time, but in the lack of any obvious stratigraphic indications, it could have happened earlier or later, too.In view of the inscription's reference to renovation work undertaken in the church, there is good reason to assume that only the mosaic floor(s) was/were laid at that time, while the church building itself was built earlier, perhaps at the beginning of the fifth century, although a fourth century date cannot be entirely excluded in the lack of the necessary evidence.The stylistic date of the existing mosaic floors would not contradict this hypothesis.
Regrettably, neither the excavated coins, nor the pottery and other small finds are of any help in ascribing a more precise date to the phases of the monastery's foundation and its subsequent renovation mentioned in the mosaic inscription.What they do indicate is that the site was in use, perhaps continuously, through the Byzantine and at the beginning of the early Islamic periods.
Function of the excavated remains and the place of the monastic church in the network of monastic institutions in the wider region In her original brief report, V. Sussman suggested that the excavated archaeological remains can be identified as those of a monastery. 155Her suggestion was accepted by later scholars, albeit with reservations for not being provided with the necessary details as regards the rationale underlying this interpretation. 156In view of the abovereviewed evidence, the identification as a monastic site seems possible, especially in view of Eusebius's title (hygoumenos) given by the inscription.The annex built against the main church edifice as well as the presence of a winepress and a cistern in the excavated area next to the church complex may also be seen as pointing toward a monastic complex. 157he church at Kefar Truman is located near a main junction of roads leading from the coastal plain to Jerusalem.Thus, it could have been part of the network of monasteries that were established along the main roads to serve pilgrims during the Byzantine period. 158Due to its strategic location, it may have continued to serve as a residential building or a farmhouse in later times, sometime after the beginning of the Islamic period.

Fig. 9 .
Fig. 10.Mosaic panel of the nave with the Greek inscription (photo: Ad am Boll ok, © IAA)

Fig. 13 .
Fig. 13.The eastern section of the mosaic panel of the southern aisle (photo: Israel Antiquities Authority Photo Archive; No. 23300)

Fig. 14 .
Fig. 14.The church during excavation, seen from the north-east, with the paved depression at the south-western end of the mosaic pavement of the southern aisle (marked with black arrow) (photo: Israel Antiquities Authority Photo Archive; No. 23293) Fig. 17.
† Epὶ τoy ~qeo-4iλeστάτoy preσb(yτέroy) kὲ ἡgo(y)mένoy Eὐσebίoy [ἀνe-] νeώqh ὁ νa[ός] Several church buildings are likewise known to have been decorated with this pattern in Jerusalem.It covered the entire northern aisle of a church erected on the Mount of Olives, 23 and it appears among For the forms and development of carpet designs in the mosaic art of the region and the period, see the recent overview in TALGAM 2014.