Preliminary Report on the 1996 Swiss-Liechtenstein excavations at ez Zantur
by Bernhard Kolb (with contributions by Daniel Keller and Regine Fellmann Brogli)
IV. Ez Zantur IV: Wall decoration in room 1
The wall paintings in room 1 were certainly the most spectacular finds of the 1996 campaign (figs. 3. 7–8). Aware that the preservation of these walls had absolute priority we left a protective layer of earth in front of the paintings – apart from the few patches already exposed – while continuing work until the restorers U. and T. Bellwald arrived at Petra. They only had 20 days to expose the paintings and to consolidate the walls resp. the decoration. This work carried on almost to the end of the season, so that there was very little time left to study the walls in a more detailed manner. The following comments are thus to be taken as provisional.
The polychrome frescoes extend across the entire length of the north face of wall C to a length of 2.9 m between the blocked doorway in the east and the corner with wall A in the west (fig. 8).
The paintings carry on along the east elevation of wall A to a so-far uncovered length of 1.8 m. A narrow epistyle and supporting pilasters frame the architectural paintings on three sides (figs. 7–8). The upper edge of the epistyle runs at only about 1.5 m above the actual floor level in room 1. Considering the proportions of the composition a dado of about 0.5 m is to be expected. Above the painted main zone are the remains of stuccoed architecture in the upper zone. The pilaster-motif – painted as framing elements in the main zone – is repeated in stucco in the upper zone.
Fig. 7 shows that the pilaster’s base and the supporting cornice have sagged a few cm from their original position, but are still sticking to the wall despite the broken metal dowels. The narrow area between the main zone and the moulded cornice in the upper zone was decorated with profiled stucco-panels. On both walls illusionistic architectural facades are depicted on a closed background imitating alabaster The forerunners and most probably the models for painted imitation of alabaster are known from hellenistic Alexandria. Cf. the Hellenistic Alabaster Grave in Alexandria, which is built of huge monolithic alabaster slabs, in Adriani1963/6: nr. 89, 140ff. pls. 61–63. Hypogeum 2 in the Anfushi-Necropole dated to the 2nd. or 1st century BC shows early examples of painted alabaster imitation; cf. Adriani 1963/6: nr. 142, 192–194, Pl. 109. .
The basic design of the frescoe on wall C is composed symmetrically and evolves as trompe l’oeil from the closed background into the open room of the spectator. On the background of the broad main building a four columned central pavilion with a segmental pediment is flanked by two sidewings with entablature, each of which is decorated with two disc shaped acroteres depicting gorgoneia The rendering of the acroteres are reminiscent of hellenistic shieldbusts resp. of imagines clipeatae, as depicted on the walls of atrium 5 in the Villa of Oplontis. See R. Winkes, Clipeata Imago (Bonn, 1969) 10ff. resp. De Franciscis 1975: 19 fig. 4; 25 fig. 11. .
The perspective, slightly from below, shows the coffered ceilings of the pavilions. The decoration of the main building is rather unusual with its twodimensional but colourful geometric patterns. As a consequence of these patterns the illusionistic architecture is set off from the main building in a strangly untectonic manner. The changing patterns on the other hand accentuate to a certain degree the illusion of depth. This is possible because the perspectively depicted column-architecture sets the – again very untectonic – limits of the different decoration-areas. The threedimensional effect is increased by the uniformity of colour of the columns at the front and the connecting entablature of the facade. These elements are painted in a red marble-imitation on white background. The rear columns show a strongly contrasting blue-green. The same blue-green is used in the epistyle, supported by the side-pilasters, which frame the main zone.
For decades the Kazneh and the other monumental facades in Petra have been quoted as parallels in the discussion of the architectural paintings of the Second Pompeian Style. The question as to whether the architecture depicted on the frescoes actually existed or not and where it originated from cannot be discussed here For a short summary of the said discussion see A. Barbet, La peinture murale romaine (Paris, 1985) 49ff. , but it should be noted, that the architecture on the paintings discovered on EZ IV is actually very close to the architecture of the monumental facades just mentioned. The closest stylistic parallels to the frescoes of EZ IV seem to originate from the early Augustean period – for example the rear wall of the so-called Room of the Masks in the House of Augustus (ca. 30 BC) in Rome Carettoni 1983: 23ff. pl. E. . The general impression of reduced illusionistic effects, the monochrome screen-walls, the brittleness of the architecture and the relatively flat rendering of the architectural framework consisting of pilasters and a narrow epistyle are quite similar to the mural paintings of EZ IV. But two major differences remain: The first concerns the illusion of depth. We have seen that the architectural paintings from EZ IV develop from a closed wall surface in the direction of the observer. The Pompeian Second Style, on the other hand, works very much with a perspective opening of the front plane of the picture: The illusion of depth is striven after. This is particularly clear in the vista through the central aedicula in the Room of the Masks. The second difference concerns the size of the painting in relation to the available surface. The painting in the Room of the Masks covers, as usual in the Second Style, the whole height of the wall, whereas the paintings of EZ IV comprises a relatively narrow band in the lower half of the wall.
We know a number of architectural wall paintings of the Second Style in Israel originating from the time of Herod Fittschen 1996: 139ff. , for instance in the Northern Palace at Masada, which E. Netzer dates in the years 30–20 BC Netzer 1991: 617f. . The walls in tepidarium 106 for example are decorated in a relatively simple scheme of dado, orthostats, string-course and main zone, reflecting the early Second Style Foerster 1995: pl. VIb. . Somewhat closer to the frescoes on EZ IV are the paintings in the Wadi Siyyagh cave near the inner city of Petra Zayadine 1987: 140f. . A row of painted blind doorways – completely lacking in perspective – is set within an architectural framework of slim columns and an epistyle stretching over them. Comparable to the paintings on EZ IV is the placement of the paintings low down on the wall. To sum up we may say, that somewhat simple versions of the architectural paintings related to the Second Pompeian Style were quite common in Nabataea and especially in the neighbouring kingdom of Judaea. Also evident is the fact, that the stylistical rendering of the frescoes of EZ IV is unique among the examples of architectural painting in the Near East as yet known.
There is a great deal of speculation about the origin of the architects and artists who created and decorated the buildings of Petra. We will examine on the basis of a few details what the paintings of EZ IV allow us to contribute to this discussion: The painted pilaster in the corner of walls A and C (figs. 7–8) is divided into square fields decorated with red lozenges on a white background alternating with white lozenges on a red background. The less well preserved eastern pilaster near the blocked doorway shows the same basic decorative system (fig. 8).
The black and white painted motifs are also arranged in a system of squares decorated with a series of circles and quadrilobes of four diagonally placed peltae around a central concave square. The basic design of the pilasters – also known from hellenistic Alexandria Adriani 1963/6: 192ff. nr. 142. – has direct forerunners in the architecture of Petra – both that built and that carved out of the cliffs as well as in the stucco decoration: The antae of Qasr el-Bint (late 1st century BC) are decorated with stuccoed square panels, themselves ornamented with octagonals and circles McKenzie 1990: 135ff., pls. 67. 71–72. . The stuccoed architecture on the south facade of Qasr el-Bint gives another example Zayadine 1987: 138, fig.14. McKenzie 1990: 137. pls. 73–74; cf. the pilaster decoration of the aedicula in room 468 oposite ed Der, in McKenzie 1990: 150f. pls. 111–112; cf. the pilaster decoration of the Temenos Gate, in McKenzie 1990: 132ff. pls. 56–57. .
The central stuccoed facade shows a segmental pediment in the middle and broken pediments at the sides. Under the entablature are the familiar pilasters with panelled decoration. The same scheme of decoration show the stuccoed pilasters on the east facade of the temple Zayadine 1987: 138, fig. 13. .
The general similarity of the stuccoed facade on the south facade to the painted one on EZ IV is striking: Both depict a tripartite facade with a hexastyle front. A final example for the specific decoration of the pilasters should be mentioned – the architectural framework of an eye idol Moutsopoulos 1990: 71, Pl. VI. found in the Temple of the Winged Lions, which was built in the 20s of the 1st century AD Hammond 1986: 29. . The pediment’s supports show framed lozenges and panel decorations again very similar to the painted ones on EZ IV.
On the basis of the small selection of parallels, it seems reasonable to assume that the painter employed was a local from Petra, who knew the monuments at first hand and allowed certain elements of Nabataean architecture decoration to flow into his work.
The unusual patterns covering the background-architecture imitate opus sectile, as proven by the red-on white marbling prominent in all but one of the patterns. Identical or very similar patterns are known since the 1st century BC from opus sectile and mosaic floors: The orthogonal pattern consisting of black intersecting circles on red marbled background with two inner squares (see figs. 7–8) is known from Pompeii and from the Northern Palace at Masada Pattern without inner squares: Black and white mosaic in the oecus of the House of M. Caesius Blandus (P. VII 1, 40), around the middle of the 1st century BC; cf. Pompei. Pitture e mosaici. Vol. VI (Rome, 1996) 392, fig. 26. Black and white mosaic in cubiculum 78 of the Northern Palace at Masada; cf. Foerster 1995: 151ff., figs. 260. 261a. Pattern with inner squares: Casa del Sacello Iliaco (P. I 6, 4), cubiculum q. See Pompei. Pitture e mosaici. Vol. I. Parte prima (Milan, 1990) 326, fig. 80; Casa del marinaio (P. VII 15, 2); cf. E. Pernice, Die hellenistische Kunst in Pompeji, VI. Pavimente und figürliche Mosaiken (Berlin, 1938) 64, pl. 27.5. . Another pattern shows the combination of red lozenges and black rimmed hexagons on a white background. Contemporary examples of hexagonal flagstone-pavements are known in Petra Compare for instance the pavement in room XVII of the Nabataean house on EZ I, in Petra-Ez Zantur I: 77, fig. 112; Pavement of the „Lower Temenos“ of the Southern Tempel at Petra: M. Sharp Joukowsky, 1994 Archaeological Excavations and Survey of the Southern Tempel at Petra, Jordan. ADAJ 39, 1995, 246, fig. 4. . We are familiar with the pattern, executed as a black and white mosaic, in the Northern Palace at Masada Petra-Ez Zantur I: figs. 57. 59. Compare G. Foerster 1995: 151ff., figs. 262–263. .
We may conclude, that the artist created a most original but in the same time provincial interpretation of architectural paintings of the early Augustean period which were already „old-fashioned“ at the time of execution in the 1st century AD. Comparisons with the architectural paintings of the Second Style in Judaea and Petra have also shown, that the twodimensional interpretation of the Second Style wall systems was wide-spread in this geographical zone. I suppose, that the extraordinary graphic decoration of the main building on wall C has to be seen within this context.
Above the main zone follows the upper zone already mentioned decorated with stuccoed architecture – a decoration-system known as the Masonry Style A. Andreou, Griechische Wanddekorationen (Michelstadt, 1988) 109f., kat. nr. 136. 138: Tombs I and III in Lefkadia (end of 4. resp. 3rd century BC); idem, 106, kat. nr. 132: Hellenistic House in Knidos. Cf. R. Gordon, Late Hellenistic Wall Decoration of Tel Anafa (Ann Arbor, 1977) 230 fig. 24. in the Hellenisitic East and as 1. Pompeian Style in the Roman West A. Laidlaw, The First Style in Pompeii: Painting and Architecture (Rome, 1985) 89ff., pl. 97b–c: Casa della Nave Europa (Pompeji I 15, 1/3), cubiculum IV. .
The design of the upper zone had two phases: In the first phase the wall was divided up by the stuccoed pilasters and finely profiled ledges, which seperated panels. In the second phase the recessed spaces were filled in with plaster and painted with geometrical patterns already familiar from the main zone. One plaster slab from this second phase was found just a few cm below its original position (fig. 9).
It is lined on both long sides by the ledges of decoration-phase 1 and shows an orthogonal pattern of triangles in black and white forming hourglasses. Further proof of the two decoration phases is also given by a series of plaster fragments from soundings 1 and 2.
Still unresolved is the chronological relation of the decoration of the main zone with those of the two phases in the upper zone. The broken lower edge of the painted main zone shows no trace of two phases. On the other hand the similarity of the geometric motives of the main zone with those of phase 2 of the upper zone may suggest a simultaneous production. Another indication for a re-decoration of the main zone in phase 2 could be the painting’s unusually modest height of just 1.5 m. If we consider that the stuccoed and painted orthostats of the Masonry Style decoration in room 3 reach an approx. height of 1.7 m (fig. 4), the narrow zone may reflect the height of the orthostats, i. e. the decoration of a hypothetical first phase, which had been removed completely up to the stuccoed panels and re-done in a different style.
A plaster human head affix of 30 cm height, which was most certainly part of the lost upper wall decoration in room 1 came to light in the debris in front of wall C (fig. 10). The thin finishing coat of stucco with the detailed modelling of the physiognomy is unfortunately mostly broken away. What one can still see is that the head is slightly tilted to the left in a typical pathos formula. We know from Philip Hammond’s excavation reports that the Temple of the Winged Lions was also decorated with plaster head affixes Hammond 1977–8: 246, pl. LX. . Since the sizes of all but one of the heads remain unpublished it is difficult to make comparisons. The published fragmentary head is only 10 cm high and thus much smaller than ours and its relief seems to be lower Hammond 1986: 27, fig. 18. .
An interesting point is Hammond’s distinction of two phases of decoration: The original richly painted stucco-panels of the 20s AD had been replaced with bare plaster panels in a phase of re-decoration provisionally dated to the reign of Malichus II (40–70 AD). In regard to the mural decoration on EZ IV with a provisional terminus post quem of 20 AD, it would be very helpful to know in detail how the two phases of the temple-decoration were related to one another. Until these findings are available to us we must content ourselves with the fact that the dating of the stucco and paintings on EZ IV is apparently very close to that of the related examples from the Temple of the Winged Lions, an assumption corroborated by a fragmentary Corinthian capital from sounding 2 (figs. 11–12).
The almost completely preserved lower part has a collar of acanthus leaves with plastered details. The state of preservation of the upper part is good enough to identify a floral capital of McKenzie’s Type 1 McKenzie 1990: 95. 190, fig. f. . The tendril decoration of the abacus is to my knowledge an, up till now, unknown feature of this type. As to type and style the capital resembles strongly the specimens from the temple and does not seem to contradict an approximate contemporary dating of the two buildings.