The Monumental Entrance to the Palace Complex

So as to distinguish it clearly from the seven city gates in the 7km of defences, the large and visually impressive public monument, with its two massive towers and a passageway leading towards the Audience Hall, has been labelled the 'Monumental Entrance' to the Palace Complex (Figs 19 and 20). The monumentality is obvious enough. It is undoubtedly an entrance, and thus a type of gate, so that by definition it demarcated space and controlled passage between the outside and the inside.

The report that follows represents a culmination of several seasons of survey, clearance and excavation (Fig. 19). Work on final detailed plans and architectural elevations is not yet complete. Many architectural details need to be resolved by further work on the smashed blocks now in the excavation depot. Nevertheless, the general scheme and the plan are already clear (Fig. 20). The entrance passageway, overlooked by two massive towers, is orientated east-west and the symmetry of the architecture, both in plan and in elevation, is striking. The large expanses of stone paving, leading to and beyond the entrance as well as in the passageway itself, surely indicate that it also served a variety of public functions, formal and informal.

The size, intended to impress and doubtless intimidate, might also reflect the need, on this high southern ridge at Kerkenes, for cover and shelter from the natural elements during much of the year. Large and impressive though it was, it does not seem to have been designed with defence (against hostile enemies) as the foremost concern. Indeed it can hardly have been defensible at all whilst the combustible materials from which it was constructed almost invited burning.

Aims of the 2005 Excavation at the Palace Complex

Excavation in 2005 was aimed at revealing the entire entrance passage from the front of the Towers to the central partition (Fig. 21 trenches TR15 to TR19) as well as uncovering part of the inner passageway, from the front (eastern) façade to the rear (western) façade and, if time allowed, beyond towards the Audience Hall (Fig. 21 trenches TR14, TR20 and TR21). There were three purposes; firstly to recover any further fragments of sculpture and inscription that might be in the disturbed fill of the passage leading to the front façade, secondly to gain better understanding of the architectural scheme and, thirdly to enhance the visual appearance of the monument while at the same time making it less hazardous to visitors. All three of these aims were achieved and thus the present program of excavation in this Monumental Entrance has been completed.

Progress and Methods of Work

The loose fill of the entrance comprised, for the most part, very loose stone rubble and debris with, immediately on the pavement, ash and patches of charred reed thatch (Fig. 22). Extensive and irregular robbing, down to and occasionally through the stone paving, was the cause of the random scattering of carved, sculpted and inscribed fragments throughout the pit fills. The fill was carefully sorted and removed by hand. Granite rubble, which made up most of the fill, was examined and discarded. Burnt and vitrified debris and large, broken and shattered sandstone blocks were stacked on the site. All other sandstone was taken to the Excavation Laboratory where it was washed and sorted. Where necessary sieves were used to extract the sandstone, but it was not possible to shake a sieve without further breaking the friable stone, so that material was sorted on perforated plastic screens. Broken and shattered architectural blocks (Fig. 23), some with bolsters or bolster-like reliefs, were recorded in situ before being taken to the laboratory for cleaning and reassembly. These processes made excavation slow and tedious. The extent of damage caused by large blocks tumbling from the tower wall onto the pavement combined with the loss of many architectural fragments as a result of vitrification during the fire as well as dispersal in later robber pits, hampered recognition, excavation and conservation.

While it may be said with confidence that nothing of importance was missed, it is sad to report that the intensity of the fire had so broken the masonry of the passage walls that very little remains in situ. The horizontal gap left between courses after the horizontal beam completely burnt down caused the wall of the Towers to become structurally unstable. Many of the large stones shattered as the result of the intensity of the fire had to be removed for safety. As a temporary measure wood was used to prop up what remains of wall facing, but this has been done for safety pending implementation of a conservation and restoration program.

Excavation Records

Excavations were documented as in previous seasons with plans, sections and photographs. An east-west longitudinal section through the centre of the Monumental Entrance was completed and recorded by making mosaics of rectified digital photographs at a scale of 1:20, printing the images and annotating in the field. The faces of the passage walls were dealt with by the same method whilst other sections, less likely to collapse, were drawn in the traditional way. Rectified digital photographs were also used to record the stone pavement and associated features. Digitising these photographs so as to produce detailed plans and sections is in progress. The careful recording of features by digital photography will contribute to the creation of a virtual environment and monument.

The Architecture of the Monumental Entrance

An architectural scheme of remarkable symmetry boasts a monumental entrance between two massive towers (Fig. 25 and 26). Nothing can be seen of the internal arrangement and floors of the towers, the core of each having been extensively dug into by robbers. A large column base, adjacent to the front end of the tower walls, stands proud of the sloping pavement on either side of the passage.

The towers are separated by a 10.50m wide stretch of stone pavement leading to a façade that crosses the passage close to the centre (Figs 27 and 28). The passageway enters the front façade through a large double doorway and exits, most probably through a similar doorway, in the rear facade. The inner passageway was reduced in width by a small room which is presumed to have been mirrored on the other side. Beyond the rear façade the entrance opened up onto another stone paved area where a third column base sat, before it was disturbed by robbers, just outside the backline of the tower against the end of the rear façade. A small drain set in the paving runs parallel to the façade preventing water from running down through the doors into the passageway. A plinth on one side of the front façade and an anaconic stele standing at the door of the rear façade have, in anticipation, both been mirrored in the plan (Fig. 27).

The Approach to the Monumental Entrance

The front half of the broad passage, which is inclined at a gradient of approximately 1 in 5, is paved with granite worn to a polish (Fig. 30). Setting lines in the portion of pavement in front of the towers are not parallel to the orientation of the entrance but lead to the doorway of the Audience Hall (Figs 20 and 27), demonstrating that the original pavement and the Audience Hall pre-date the construction of the Monumental Entrance. At the inner front corner of each tower the top of a foundation stone can be seen slightly above the level of the top of the pavement and projecting in both directions for a few centimetres beyond the line of the upper walling. This slightly wider footing appears, however, to be restricted to the corner stones and is not observed along the tower walls rising up the passage slope (Fig. 31). The square sandstone column bases on either side, set back c.2.05m from the front of the towers, have shallow circular recesses c.0.85m in diameter. In all probability these were for tall, freestanding wooden columns provided with sandstone capitals (although an alternative possibility is discussed below). These bases were shaped in position, as was the rectangular conglomerate plinth preserved at the southwest corner of this front portion of the entrance. The plinth, measures 2.10m by 0.70m and was presumably mirrored by another one on the opposite side of the door. From the front of the towers inwards the paving was re-laid up to the walling, column bases and the, presumably two, plinths. This may be demonstrated by the row of particularly large stones along the line of the tower fronts.

It is now thought that this front section of the Monumental Entrance was unroofed for several reasons. Firstly the span, while not impossible, seems very wide, particularly as there was only a single pair of columns. Secondly, the steep slope of the pavement would have made roof construction difficult. Thirdly, if it is correct to reconstruct a large timber façade founded in the wide slot at the top of the slope any roof would have obscured it from view. Lastly, it would seem that the distribution of burnt debris, in as far as it could be ascertained between the robbing pits, is perhaps more consistent with burning of the façade and double doors than that of a passage roof. The fragments of burnt roofing, including pieces of clay with reed and wood impressions as well as deposits of charred reeds most probably fell into the entrance passage from the towers on either side.

The Position of Sculpted Monuments

Pieces of sculpture and inscribed monument, described later in this report, were recovered from the debris from the front portion of the entrance. These include a statue of a human figure and an inscribed block bearing small-scale relief sculpture. Other elements, perhaps associated with one or both of these pieces, are still being reassembled from the fragments that escaped both fire and later looting. A horn of sheet gold formed around a wooden core points to the presence of embellished wooden pieces.

It can be noted that there is no indication as to what stood on the stone plinth (and its presumed twin) in front of the partition. Measurements and proportions make it rather unlikely that any of the small scale fragments recovered stood on them. It has however been suggested that the sandstone column bases did not support tall freestanding (votive) columns of wood with stone capitals, but that they carried short, columnar, wooden statue bases.

In any event, there are no reasons to doubt that everything set up within the entrance was torn down and smashed before the fire.

The Front Façade between the Towers

Approximately half way along the side of the towers, a 2.20m wide slot appears to be a foundation trench to a structure which has been described as the front façade (Fig.32). The burnt and vitrified debris and perhaps some traces of vertical posts indicate a timber frame structure with infill rubble and mudbrick.

Along the edge of the pavement, in the centre, five large stone pavers indicated the position of large double doors. Traces of the door posts could be discerned on either side. It would seem that these doors were set in a large wooden frame partition. The total width of the five pavers is c. 4.20m whilst the width of the slot for a façade between the different elements of the entrance is c. 2.20m, i.e. wide enough to take the width of each of the inward opening doors. The preserved southern wall of the passage changes abruptly from impressively large cut blocks to rubble construction with timber elements, clear evidence that it was hidden from view by the partition.

An absence of ash and charcoal in the wide foundation slot, which was filled with masses of fused stone rubble and mud that most probably came from the upper walling of the towers as well as the core of the façade itself, suggest that the doors and possibly much of the partition itself may have been pulled down before the fire. Additional evidence that the doors were pulled out prior to the burning came in the form of fragments from a small sandstone bolster recovered from the hole for the southern doorpost.

In 2004 it was plausibly suggested that this partition took the form of a monumental, decorated, wooden façade, and that such a façade might have resembled rock-cut architectural façades in the Phrygian Highlands. The number and range of sandstone bolsters recovered in 2005 raise the possibility that similar elements in wood formed part of the decorative scheme and that valuable embellishments, perhaps hinted at by the discovery of bronze cut-outs of ibex in 2002, might have been affixed to the doors and the screen as well as the pediment. It may also be suggested that the door sockets were valuable, inciting robbers to tear them out.

One question which remains unresolved for the moment is how to reconstruct the 2.20m wide façade. It seems possible that the structure carried a walkway between the two towers, presumably behind a wooden pediment. It is further possible that there were wooden stairs or ladders on either side of the doorway which would have been concealed when the doors were open.

The Inner Portion of the Entrance between the Towers

Passing through the double doors of the front façade led to a second paved and levelled area where the paving stones are smaller and exhibit less signs of wear (Fig. 33). A burnt clay floor indicated the location of a small room in front of which was a bin like feature. As suggested on the plan (Fig. 27), it is reasonable to assume that there was a symmetrical arrangement and a room on the unexcavated southern side mirrored that on the north. No alterations were made to either this portion or to the rear section of the entrance before the destruction.

A slightly raised step was found in front of the doorway of the room on the north side (Fig. 34). This featureless room was provided with an earthen floor. The awkwardly narrow space between the east end of the room and the partition was lined with a single course of stone above the pavement forming a shallow bin-like feature.

The room itself measures approximately 2.10 by 2.50m with a single entrance, indicated by raised stones in front of the doorway, towards the eastern end of the south wall. The walls stood on stone footings which stood a single course above the level of the paving. The upper walling was of square mud-bricks each c. 32 by 32cm. The room was badly disturbed by robbers and no trace of the north and east walls survived. It is not known whether the walls were laced with timber.

The room and, and its presumed twin, were obviously roofed. Since no roof debris was recovered from the extant floor of the room it can be assumed that the roof comprised reed thatch which had been reduced to ash. In keeping with common practice at Kerkenes it is probable that the central paved area was open, a suggestion reinforced by the raised stones in front of the door and the absence of burnt roofing remains on the pavement.

The Rear Façade of the Entrance

The rectangular inner passage was enclosed at the rear by another façade (Fig. 35) similar in many respects to the front façade described above. It can be supposed, although not certainly demonstrated, that double doors stood in the centre of the rear façade. The presence of a fallen granite stele of truly aniconic form (Fig. 36) set up behind a sunken square granite stone with a finely worked upper surface supports this suggestion. It is assumed that the west wall of the inner room stood close to the back of the façade structure. Robbing had, however, removed all evidence.

The Western Paved Area

An area of stone pavement (Fig. 37), partly robbed, extended towards the Audience Hall. A drain set into the pavement along the rear façade took surface runoff away from the entrance in a northerly direction. The portion of this drain that runs across the front of the entrance was more neatly and regularly built than its northern end (Fig. 38).

On the excavated north side, set into the paving beyond the second partition, was another square column base with a circular recess for a wooden column (Fig. 39). The diameter of the recess was, at c. 0.80m, slightly less than in the bases at the front of the entrance. Looters had, at some time in antiquity, dug around this base and then beneath it, with the result that it was discovered pitched into a robber pit. It is highly probable that before the commencement of the extensive robbing in the entrance the top of this base stood proud of the surface and was thus visible.

It is likely that the two columns at the rear of the entrance were set a little beyond (i.e. to the west) of the back of the towers. Such an arrangement would have formed a roofed porch that reflected the open porch at the front of the entrance. The rear walls of both towers are very ruinous and partially disturbed by robbing, with the result that their precise position cannot be determined without considerable excavation.

The Towers

In plan, the towers are rectangular and each measure approximately 15.60m by 13.00m. The position of the rear face has not been exactly ascertained, precluding precise determination of their length. The corner of the South Tower was exposed after the collapse (Fig. 40) was carefully recorded and removed. Little was done by the North Tower apart from cutting back further the dangerously loose edge of the excavated area (Fig. 41).

The preserved lower parts of the towers were solid and provided a platform where the sloping ground rises a vertical distance of more than three metres over the front half of the entrance. Doubtless they surround and conceal outcrops of bedrock and the lower, preserved, walls are retaining walls for the stone rubble fill. Of the internal arrangements and floors nothing has been seen because of very extensive robber pits dug into the core of both towers. From what is known of similar terracing in Structure A (behind the glacis to the north and east of the entrance) and from test trenches excavated at the north end of the city in 1996 and 98, it might be expected that the foundations of internal walls were constructed within the rubble fill of the terrace, but most if not all, will have been destroyed.

Turning now to the walling of the towers, although a better understanding has been gained through revelation of the south tower there are outstanding problems that restoration of architectural blocks may partially resolve in 2006. What follows is, therefore, both preliminary and somewhat tentative.

It is reasonable to assume that all of the four exposed outer faces of each tower were faced with cut granite to the same level. This level was in fact one course, about 0.80m, above the top of the stone pavement at the higher western end of the entrance. As the height of the top course of cut granite facing stones increased down the sides, the number of courses increased to four courses on the eastern wall of each tower.

In the passage, above these cut and tightly fitted granite blocks (described in detail below) was a course of yellow sandstone above which was a course of soft white (chalky) limestone. It is now thought that the sandstone and limestone courses continued along the tower fronts. The courses were interspersed with large horizontal timber beams at least 0.30m high. It can be expected that these beams were squared to ensure structural stability. The top of the sandstone course does appear to have been tied by swallow-tailed wooden clamps which most probably fixed the floor joists. If this is correct the course of white chalky limestone would have stood higher than the internal floor. Some confirmation of this idea may be the evidence that vertical timbers of wooden framing for the upper walling, infilled with stones and some clay, began at the level of the sandstone. It must be admitted however that the evidence for these ideas might also be interpreted differently.

From the centre of the front façade across the passage back to, presumably, the rear of the tower, the wall faces were composed of angular fieldstones. In other words, cut facing stones were not used for those portions of the passage wall that were hidden by the inner portion of the passageway and parts of the wooden façades. This, incidentally, is further indication that the small rooms within the inner portion of the passageway were part of the original construction.

Additional questions that currently evade satisfactory answers relate to the original height of the towers and the treatment of their corners. Here the difficulties are compounded by the existence of numerous fallen blocks of sandstone with engaged bolsters and bolster elements, and other forms of architectural embellishment. The majority of these smashed blocks are awaiting further study and restoration in 2006 after which, it is anticipated, their number and architectural arrangement will become clearer.

At a more detailed level, it was observed that the granite blocks were snugly fitted with seating for inserts to level courses between beams. Behind each of the two column bases at the front, the granite stone making up the first course running up the slope of the side walls were made of very much smaller, but nevertheless neatly squared and tightly fitted blocks (Fig. 42). Jointing between adjacent granite blocks was by no means always vertical, with triangular or even curved granite plugs being shaped to fit where necessary. Such deviance from vertical jointing has not been seen in the softer sandstone masonry, although large blocks could be L-shaped and plugs were often used.

As to the upper walling, burnt debris filling the passage points to the use of angular granite rubble and some mud as filling between timber elements. Height is difficult to estimate. Because of the extent of the robber pits in the passageway estimation of the height of the upper walling of the towers is not straight forward. However, the fallen mass of infill from the south tower would not be inconsistent with the concept that at the front the height of the upper timber frame portion of the towers was equal to the lower stone facing. There is no clear evidence as to the width of the upper walling which, to judge from shapeless fused and vitrified blocks, may have been no more than 0.80m. In the absence of other indications it can be assumed that the towers were roofed (otherwise they might have been solid to the top). It now seems highly probable that the fragments of burnt clay with reed and timber impressions came from roofing of the towers rather than as had been thought at one time, a passage roof.

Clamps

An important detail, not least for the chronology, is the use of wooden swallow-tail clamps, the charred remains of one having been recovered in the cutting itself. Small neatly carved clamp cuttings were recorded on the two large, faced, granite blocks, that stood on the northeastern corner of the South Tower and the adjacent block to the west. These smoothed clamp cuttings were for mending granite blocks that had broken during construction. In the sandstone, by contrast, there were numerous swallow-tailed clamp cuttings. There were deeper, not particularly regular and rather rough clamp cuttings, exhibiting considerable variation in size, in the sandstone (but not apparently in the soft limestone). They appear to have been cut with the same tools as were used for shaping and finishing the sandstone and limestone blocks (but not the much harder granite), and essentially these appear to be woodworking adzes and chisels. Clamps were used to join adjacent blocks in wall faces and, at 45º angles, to brace blocks on tower corners (Fig. 43). In addition there were, as reported last year, clamp cuttings that perhaps secured facing blocks to floor joists. It can be argued that the beams were floor joists rather than timber lacing within the core of the tower since in the latter case some evidence of charred timbers would have been recovered.

Proposed Architectural Reconstruction

It behoves us to make some attempt at reconstructing this monumental entrance. As he approached the entrance, the visitor would have seen a broad stone paved entrance between two massive towers. At the top of a steep incline, between the towers, stood a decorated wooden front façade containing broad double doors at its centre. When both these doors and the similar double doors through the rear façade were thrown open, the entrance to the Audience Hall would have been visible. Walking up the worn and slippery granite pavement towards the doors, the visitor would have passed a pair of wooden columns on stone bases, carrying stone capitals or perhaps statuary, and seen sculpture on large stone plinths either side of the doors. Reaching the top of the incline would reveal the pavement continuing through the entire passageway up to the Audience Hall. The front of the towers on either side rose perhaps as much as a total of eight metres. While the bottom half of their front wall was constructed of silver-grey granite, yellow sandstone and white limestone, the top half probably comprised a mud plastered timber framed structure. Somewhere, perhaps in rows along the front, were sandstone bolsters with additional embellishments on the corners. The façade, following rock-cut examples from the highlands, would have had a triangular pediment crowned by an akreterion which might have hidden from view a raised walkway between the towers. The rear façade would have looked similar but here the freestanding columns (or columnar bases) stood forward of the tower walls and the doors were flanked by a pair aniconic stele with rectangular sunken offering places in front of each.