Tel Dor Excavation Project

Google Custom Search
Staff login:

Dor 2009 Preliminary Report


The Search for the Hellenistic Palatial Edifice

The original motivation for opening area D4 in 2004 had been the realization that the central part of a massive structure which had first been located in the 1980’s in area D1 (the [then] so-called ‘Persian’ Palace of phase D1/4), and then connected to a system of thick ashlar walls cutting right through the width of area D2 (phase D2/4) – must lie to the north of D2, northeast of D1. It was also hypothesized that said structure might be the source of the many carved architectural elements found in area D2 and D1 – mostly (but not only) reused as building stones in Roman foundations, as well as the exquisite Hellenistic mosaic found in 2000 broken up in a Roman pit.

In 2008– after four seasons of excavating (mainly) a Roman industrial complex – we thought we had finally located parts of the same structure in area D4 (in phase D4/3). The early phase of W05D4-060 was uncovered that season and proved to be constructed in a similar fashion to some of the other walls of the ‘Persian Palace’. W05D4-060 is dovetailed at its base to W04D1-050, and so presumably that wall, too, belongs to the same structure (Figure 9). A third feature associated with this structure was W08D4-363 – a massive square pier made of large ashlars. This association, too, was based on both stratigraphic situation (the pier is partly under a Roman wall) and partly on construction technique.

Hellenistic Palatial structure

Figure 9 Area D, Aerial photo (looking east) with the lines of the Hellenistic Palatial structure highlighted.

The 2009 season complicates this picture. A first clue was given already in 2008. As soon as the Hellenistic phase of W05D4-060 showed up, we also noted a foundation trench along it. If that observation were true, then all we have of the walls of this structure are foundations. Soon into the current season, moreover, an additional building, consisting of two long ashlar walls – W09D4-554 (N-S) and W09D4-533 (E-W) – under the phase 3 structure were uncovered, labeled Phase D4/4. This structure, too, is robbed down to its foundations. Only one short section of its superstructure exists – built in header-stretcher technique. Under this section, the construction is headers-only. The headers present a flat face towards the outside (north and east) and serrated on the inside (south and west) – indicating that on that side, at least, these are foundations.

Aerial view (looking east) of area D4

Figure 10 Aerial view (looking east) of area D4 at the end of the 2009 season, with features discussed in text highlighted

In contrast to the two sets of walls, there is only one set of floors in the relevant elevations (F09D4-512, F09D4-514, F09D4-517, F09D4-549 # c. 13.90-14.00). Paradoxically, they seem to reach both the phase 3 and phase 4 walls (Figure 11). In the attached detailed report, they are attributed to phase 3 – but they may well actually belong with 4.

Figure 11 (p09D4-9007 - 09) (a) F09D4-517 apparently reaching both W09D4-518(=W09D4-533a) and W04D1-050 (top right). (b) F09D4-549 apparently reaching W09D4-533 and W05D4-060 and W04D1-050 (erroneously labeled W04D1-067). (c) F09D4-512 reaching W09D4-533

Hellenistic period – area D2 vs. D4

Figure 12 Comparison of structural phases of the Hellenistic period – area D2 vs. D4.

Our working hypothesis has been that all the thick walls belong to the same structure; i.e. D4/3 = D2/4. Note, however, that the thin header walls of D4/4 actually match best with those of D2/3; i.e. D4/4 = D2/3. Contrast Figure 10 - where the thin header walls in D4 are earlier than the thick ‘interlocked ashlar boxes’ wall’ with Figure 13, where the thin header wall stand on top the thick ‘interlocked ashlar boxes’ wall.

Figure 13 Unit AN/14 of Area D2 at the end of the 1999 season. Note W17527 – a thin ashlar wall of phase 3 (Header-stretcher on header-only foundation) – on top of Phase 4 W17595. W17595 is dovetailed into W17562 (the so-called “Big Mother Wall”). Both the latter walls are wide “interlocked header-stretcher” walls of the Hellenistic “palatial edifice”

The conundrum might be solved in three ways:

Although phase D4/4 is constructed of thin header walls, it too is hardly a domestic habitation phase. The walls enclose a space of at least 10 x 10 m. – and possibly as much as 20 x 20, if the equation D4/4 = D2/3 holds – with no apparent internal divisions. Either this is a very large hall (unlikely – in view of the thinness of these walls) or a large courtyard (but there is no evidence thus far for rooms around it). The odd feature noted above – that the walls are straight on their east and north faces, but serrated on the other side – brings to mind some sort of podium (filled on the inside, free-standing on the outside); but what stood on that podium is moot.

A further discovery that may add further snags is a group of column drums designated L09D4-584 (Figure 14). These raise several new questions. First, is this a built feature? I.e., are the drums being reused in some fashion (e.g. as the foundation of a pier similar to 08D4-363); or are they a more-or-less random configuration (e.g. – they were felled from wherever they originally stood, but then left lying)? Contingent on that is the question of their phasing. Both 09D4-584 and 08D4-363 are sealed under phase 2 floors and so belong to Phase 3 at the very latest. If thought-of as a foundation for a built feature, 09D4-584 might conceivably be regarded as belonging to phase 3 – e.g. some sort of ante attached to W04D1-050 – although the superstructure of that wall bears no mark of such. Also – no trace of any robber trench of such a postulated feature was noted. Judging by elevations alone – the column drums fit the elevations of phase 4, or even – if we judge the phase 4 structure to be a filled podium – pre phase 4. This would most definitely be the case if it were found out that the columns are lying on some surface.

fluted columns

Figure 14 (p09D4-9308) The fluted columns (L09D4-584) reaching (?)W04D1-050 (=W07D4-061E)

The appearance of the columns themselves adds to the riddle. First, they are fluted – which is not the norm in the Levant (and never hitherto found at Dor). Second, the sections are about 1m. long – too long to be considered ‘drums’ in the usual sense – but both ends are cut flat i.e. they do not appear to be broken monolithic columns.

Minimally, then, these columns are re-used in Phase 3 construction and thus originate in Phase 4. They may be even older. Hitherto, the earliest phase to which we could definitely assign Greek-style architectural elements is phase 3 (or the [provisionally] contemporary Phase D1/4 = D2/4, and see reservations above). While a thorough survey of when this type of architecture appears in the Levant is outside the limits of this report, we cautiously suggest that these examples may be among the earliest.

The “Alexander Gem”

A preliminary PR press-release which was given to the University of Haifa press office at the end of the season, was picked up by the wire-services and internet blogs and had given this single engraved gemstone (Figure 15) a brief period of notoriety around the globe. We will not add any information here on the object itself, but record its find-spot and clarify somewhat the circumstances of its finding.

The Alexander Gem

Figure 15 The “Alexander Gem”

The gemstone was found while cleaning the bottom of the southeast unit of area D4, preparatory to re-opening it for excavation (Figure 10). Since it was found in a cleaning operation, it was not assigned to any specific locus (“L” 05D4-099 is a generic “surface find” / “cleaning” designation for area D4). It was only a few days later, while being cleaned in the museum, that the significance of the find was realized.

Although technically found out-of-context, a debriefing held with all concerned postfactum established that it was found at the bottom of the pit under the protective plastic put there at the end of last season. We can thus say with reasonable confidence that it is not intrusive. Both the loci immediately beneath it as well as the last loci excavated above it in previous seasons are Hellenistic in date. The recorded elevation puts it about 10 cm. above the top of preservation of Phase 4 W09D4-544 (which was not yet visible at the time). Since no floor separated phase 3 from phase 4 deposits at this point, we shall probably never be able to say, with any confidence, to which of the two it belongs (for the problematics of assigning floors to these phases, see above).

A quick survey of [the less bizarre] commentary offered by the media and on the web following the preliminary report reinforces – by and large – Jessica Nitschke’s initial identification of the object. To wit – it is a portrait of Alexander the Great, or (less likely) a later Hellenistic ruler trying to appropriate Alexander’s attributes.

The Roman Period
Roman streets

Figure 16 Aerial photo (looking east) of all of the D’s, showing lines of Roman streets.

Not much remained of the Roman industrial complex excavated here in 2004 – 2008, and much of what remained was dismantled. The only addition was one unit – AP/18 – which was opened from topsoil to the north of the existing area. The purpose of this unit was to check for the existence of a Roman street, and to check whether the Hellenistic monumental architecture continues north of W05D4-060 or if that wall forms the northern façade of the monumental Hellenistic building.

After much digging through topsoil and sand pits the excavation of this unit did hit the typical paving-stones of a Roman street. This street continues the line of a paved street excavated in 2006 through units AT-AV/17-18 (Figure 16) and presumably continued just south of area D3 to join the ‘Cardo’ approximately in AJ-AK/18-19. This allows the reconstruction of a c. 30 m. wide insula (about twice as wide as the ones previously found in area C (Stern, et al., 1995: Fig. 4.3). The southern street and façade of this insula were exposed over the years to a distance of some 80m. (from the western edge of area D5 to the eastern edge of D2. Except for the intersection with the ‘Cardo’ in AI-AJ/11-12, no other cross-street was noted across this 70 m. expense, though small blind alleys permit access to the inner structures in this insula. One such alley, splitting off the southern street, was found long ago in area D1. It is possible that a similar one – splitting off the northern street – runs between ‘upper D1’ and D4.

Connecting the known sections of streets, it is clear that – as in other cases at Dor – the street lines are not strictly orthogonal and ruler-straight, but tend to follow the topographical contours of the tell while still keeping more-or-less parallel lines to allow for regular insulae.