Tel Dor Excavation Project

Google Custom Search
Staff login:

The 2004 Season at Dor

Table of Content

A Personal Assessment and Preliminary Report (Ilan Sharon)


The depth of winter here in Israel is a good time to look back at summer season (July 1 - 30) and assess, with some backwards perspective, what has been achieved. This text is meant to be an informal report to those who participated - or had some other investment in - our work this season.

We had, throughout the season, about 50 volunteers - of whom about 40 were actually digging at any one time, plus a staff of almost 30. Note the difference - 80 people / 40 diggers about which more below. These numbers are small in contrast to 120 and even 150 in the "roaring 90's". However, they need to be seen in the context of the general situation. After three years of virtually no academic volunteer-and-student excavations in Israel at all, 2004 marked a modest recovery -- but no excavation rebounded to its '90's strength. For all that we had 'a small season' we were probably the largest excavation in the field this year. In view of the fact that we have been out of the field and 'out of touch' for three years, and that we are a rather expensive excavation, such numbers are nothing short of miraculous. They are due mainly to the efforts of Andrew Stewart and Sarah Stroup, whose recruitment achievement is all the more striking in view of the fact that the return of students / volunteers to other projects (and the return of tourism to Israel in general) has mainly been European. We are the only project that had a substantial American involvement.

Following the small staff-seasons of 2002 and 2003, we lodged at the Nahsholim Seaside Resort. The group was much bigger this time, but Ami Elram and his staff coped with all of the problems, even to the extent of allowing our R&B coordinator - Talia Goldman - some time-off to actually work at her real job. While several of the old-timers lamented the loss of Pardess-Hanna's distinctive ambience (especially the kiosk...) most of those who experienced both accommodations prefer Nahsholim despite its being more expensive. The closeness to the tell (no bussing to and fro..), the beach, more privacy in the accommodations, the beach, air-conditioning in all rooms, the beach, the quality of the food and the beach were the reasons most often mentioned in favor of Nahsholim.


We worked this year in one major area (D1) where 9 surface units were excavated and taken down to Roman levels; a limited 5 unit exposure on the margins of (previous excavations') D2 - where the Iron Age - Persian transition was sought; and a very limited 1-unit probe into early Iron Age I in area G. I'll describe these from early to late.

Early Iron Age - Area G

In 2002 the Weizmann Institute of Science's team had started a micro-stratigraphic study of the remains in the balks of area G. In 2003 this was expanded by a limited probe in unit AJ33 (Sloan's Saloon) - the only place in area G where 'steps' of Iron Age remains were left above the Bronze Age pit. The idea was mainly to provide some horizontal exposure as a control to the main sampling strategy which was perforce limited to the vertical balks. On the last day of 2003 (as these things usually go...) we hit a white layer - judged to be either the floor of phase 9 ('the destruction layer') or the roof collapse above that floor. We therefore decided to continue to investigate this area in 2004. As a further experiment in inducting themselves into tell archaeology the Weizmann team decided that they will excavate area G on their own - so that all operations - actual dirt removal, registration, pottery reading (as well as, of course, sediment-sampling) would be done by them. Our own 'contribution' to the operation would be an area supervisor. As it happens, this supervisor was Avshalom Karasik - so that area G became a Weizmann in-house operation.

As it turns out, the tracing of the 'white floor' became the starting point of a fascinating search, which lasted throughout the season. The 'floor' turned out to be in actuality a thin layer of phytoliths (the mineral 'skeleton' of grasses). A few centimeters under it, a similar surface was found - but microscopic analysis (by Rosa Alberts) showed that the two were not the same. Whereas the upper phytolith layer was almost all one species (probably barley) - the lower one contained a variety of grasses - some of them flowering. What does all of this mean? That the room was used for storing animal feed? That the different phytolith composition represents winter storage (hay) vs. spring (fresh flowering grasses)? Or that we have the collapse of a thatch roof over the grasses which were stored in the room? Or perhaps that grasses were stored under the eaves of the thatched roof, providing for both food and extra insulation? And perhaps we are not dealing with herbs and grasses at all, but rather with dung? (indicating either that animals were quartered inside the house, or that dung was used in either construction or as fuel) - all of these possibilities need to be investigated further.

The 'phytolith floors' date to the early part of the Iron Age I in area G. The stratigraphic evidence is somewhat ambivalent as to whether they should be attributed to local phase G/9 ('Ir1a(l)') or G/10 ('Ir1a(e)'). Artifacts on the floor included a nearly-complete krater, and a necklace from which several dozens tiny beads, of as yet unidentified metal and probably some siliceous materials (about 1mm. long and half as wide) were recovered by careful sifting. The scant pottery in this room may provide further support for the hypothesis that it was not of 'normal' domestic use. But! Among this scant material, was a fragment of strainer jug, similar to the unique one found in the southern part of D2 (on display in the glass house). This pot was taken by some as the representative of Sikila ceramic industry (a.k.a The Sikil pot, or Zorn ware. So there were two (actually, we have a third such fragment. The generous quantities of phytoliths in this layer will be used by Elisabetta Boaretto in her attempts to date phitoliths by 14C (see below).

While cleaning and preparing the area for excavation small areas belonging to later phases of the Iron Age were also exposed - namely parts of the original "John Sloan's floor" (phase G/6), where a jar was found in situ.

Late Iron Age - Area D2

Operations in Area D2 were limited this year to the high units west and north of the core-area (Heidi's area of 1995 - 2000). These units were left more-or-less at top-of-Iron-Age at the end of 2000. The reason for digging these units is partly conservational - to lower and step the high baulks around central D2, see below - and partly to get another glimpse at several interesting phenomena we had noticed when digging late[r] Iron Age and Persian strata in D2. It was particularly helpful that the stratigraphy of the upper phases of these very units had been recently [re]analyzed by John Berg, Effie Shalev, Talia Goldman and myself in Seattle in the spring. Excavation could thus both recheck our conclusions and start out from a (relatively) secure stratigraphic basis.

Work was divided between two sub-areas: the three western units were supervised by Liz Bloch-Smith, and the two northern ones by Aaron Brody. With them worked the stalwart South-African team. Willem Boshoff and his students could only come for half the season this year, but two of his graduate students, Neels Kruger and Erika Cruywagen, stayed on throughout the season as assistants to Liz and Aaron. The Haifa University study excavation had also worked in Area D2. In addition to both these groups, most of the 'individual volunteers' worked in this area, as well as several of the UC-UW students who were interested in excavating earlier periods.

The main late IrA features (but see below) we encountered during the previous excavation were the corner of a massive ashlar building just peeking in [what used to be] the NW corner of the area; a very extensive white floor (a.k.a. 'Liz's floor [Liz Honniset - not Liz Bloch-Smith...]) - portions of which were found in various units throughout the area; and many pits - including a complex of pits along [what used to be] the northern balk of the area. The exact stratigraphical relationship, and the dating, of these elements was not clear, however - and this years' excavation may have helped in sorting it out. The pits apparently date, by their pottery, to the 7th century. We entertained the possibility that a 6th century horizon may be identified here, to account for that missing century (at Dor and elsewhere). Some loci indeed contained a mixture of late-Iron-Age-looking and Persian-looking pottery, but being (mostly unsealed) contexts in secondary deposition, this is probably what they are - mixed loci. Yet again we have failed to define the Iron Age/Persian period transition!

The southern wall of the above-mentioned building, consisting of massive ashlar headers of over .5 x .5 x 1 m. in size and cornering onto the already known section of the eastern wall, was indeed located in Liz's area. This building thus marks an important change in the topography of D2. It is the first building in the history of the area which is built over the long-lived Bronze Age or early Iron Age which for many centuries separated the "original" area of D2 from the area to its west (the presumed Bronze Age town). A complex of thick kurkar floors (possibly alternating-with or sitting-on a stone pavement) was found throughout the area (both within- and outside-of- the area of the ashlar building). One of these "floors" was actually an installation of sorts, consisting of a sunken plastered basin, found embedded with charred seeds. Whether these are contemporary with 'Liz' floor' has not yet been determined. An actual interface between the kurkar floors and the ashlar walls was only observed at one or two points, where it appeared that the floor reached up-to-and-over the top corner of the ashlars. The working hypothesis is that the floor is later than the wall - though a possibility that the two are coeval (e.g. that the stones over which the floor is laid are a wide foundation to a narrower superstructure) can still be entertained.

Ayelet Gilboa was somewhat surprised to note that the (scant - and none of it in situ) pottery on the floor appears rather early - possibly as early as Ir2a, or Ir2a|b. Our impression from the (scant) material on Liz's floor was that it was late Ir2b - i.e. possibly dating to the Assyrian destruction. All of this will have to be verified further. If the floor is indeed early, and it indeed covers the ashlar building, than that building might be Ir2a.

The thick kurkar floor extends into Aaron's area, where it was clearly cut by the afore-mentioned industrial pit (or, rather, pit-complex - because sub-features were identified within it). The pit was filled with Ir2c pottery (almost exclusively containers - Phoenician commercial jars and small jugs, but also a few Assyrian bowls and other types), slag, and nodules of vitrified quartz, as well as other, more exotic and unidentified materials, at least one antler, and a fragment of worked ivory. As the pit extends into previously-excavated units and thus is sectioned at the balk, we developed a strategy whereby, as Aaron was excavating the pit from the top, Steve Weiner and Naama Yahalom were sampling each tipline from the side.

Due to the high iron-oxide contents of the slag, as well as iron 'hammer scales' previously noted in this pit, we had always assumed that we are dealing with some sort of Assyrian iron industry. Sana Shilstein, however, points out that one cannot automatically assume that red slag = iron industry. Assuming a high efficiency in copper extraction - the leftovers of copper industry will contain mostly iron. Indeed, XRF analysis of the slag indicates high concentration of copper as well. If so, than why is the late-IrA industry in D2 so different (in the type of deposits it leaves) from the early-IrA copper industry in area G? Naama Yahalom will investigate this question further as part of her Ph.D. program.

Persian and Hellenistic periods - Area D2

Despite our intense research interest in these two periods, they were hardly excavated this season. The units which started at topsoil hardly reached them, and the ones in which we continued older excavations were by-and-large already below them. Findings of these periods were thus limited to the dismantling of walls, digging of robber- and foundation-trenches, and to some elements which were found in reuse in later features.

Phase 5 in D2 is a heavily robbed Persian period phase. Despite the fact that almost no walls were left intact, we had managed to delineate at least three sub-phases (5a - 5c) of it during our stratigraphy-session in Seattle in the spring, by careful delineation of the complex of robber-trenches and floors. Summer excavation confirmed most of these observations and yielded some new information. Some wall fragments which were left in the robber trenches indicate that the robbed building was constructed in the typical Phoenician ashlar-pier-and-rubble-fill technique (with the ashlar piers almost invariably robbed). The orientation of the walls confirms that this phase is the earliest one in which orthogonal ('Hippodamic') town plan was employed.

The most important find of the period was an Athenian silver tetradrachm, probably dated to the middle of the fifth century BCE, found sealed under one of the floors of the structure. It is not yet clear whether this is indeed the lowest floor of phase 5 - in which case the coin would antedate the structure, and - by implication - the introduction of orthogonal planning at Dor. At any rate, the coin is early-within the phase 5 sequence. This underscores once again the fact that the urban fluorit at Dor in the Persian period is late (end of fifth - fourth centuries BCE); and raises again the perennial problem of what the character of the settlement was like in the early Persian period. We have some artifacts of this period. Indeed Andy Stewart's and Becky Martin's work on Greek pottery located several fine pieces dating to the sixth and early fifth centuries - in addition to ones we knew about already. But this far we have not been able to associate any architecture with these finds.

Recent seasons' (namely 1998) work had led us to suspect that the 'Persian Palace' in area D1 may not be Persian in date (nor is it necessarily a palace...). Rather, it consists of the deep foundations of a later structure that cut into Persian period deposits. This was confirmed by the work of Allen Estes et. al. on the stratigraphy of D1 during the 2003 'staff season' and the spring 2004 workshop in Seattle. One discovery made at the very end of the 2000 season was that W17562 (a.k.a. 'Big Momma') - the defining feature of phase D2/4 - continues all the way to area D1, and is abutted by the north-east corner of the 'Persian Palace'. This, at the very least, provides a valuable peg for correlating the stratigraphy of D1 and D2. This crucial connection was further cleaned and investigated this season (in Becky Martin's sector of D1 - see below) and it was established that the 'PP' is constructionally later than 'BM'. Aaron Brody dismantled sections of the 'Big Momma' wall which were floating (and indeed in danger of collapsing into the deep D2 pit) in his area, and retrieved potsherds from the makeup of the wall as well as its foundation trenches. Some of these were Hellenistic - and not necessarily early Hellenistic at that. This may necessitate re-thinking as to the date of the entire 'Persian Palace' - 'Big Momma' complex - namely that it is even later than its 'revised' date.

Lastly - the foundations of the Roman complex (see below) contain numerous architectural elements in reuse. Most or all seem to belong to the Doric order - and are consistent in size with elements from D1 already published by Andy Stewart and Becky Martin. Evidently, some major Hellenistic edifice was dismantled in order to build the Roman one. It does not seem likely that this dismantled building was the 'Big Momma' - 'Persian Palace'. For one thing, the size is wrong - the walls bearing the Doric columns which were found in reuse should be thinner than 'Big Momma'. What or where this building might have been is still unknown, though.

The Roman Era - Area D1

Nine new units were opened by Andy Stewart and Allen Estes this season between Area D1 and D2, with the bulk of the students and volunteers of the University of California at Berkeley and the University of Washington, Seattle. Their objectives were to connect the two areas and to cross-correlate them, and in keeping with our long-term excavation-conservation-presentation strategy of concentrating on a large exposure in the SW part of the tell. All of the 10m. wide strip between D1 and D2 was opened up (except for a wide balk along the west edge of D2, for dirt removal out of the two areas); as well as three additional units - two to the east of this strip, north of D2, and one west of it, north of D1. We arbitrarily decided to call this intermediate area 'D1' - instead of 'D½' ? - although as it turns out the architecture uncovered is more closely connected to elements formerly found in D2 than in D1. For reasons mainly having to do with the staff makeup of the UC-UW group this area was divided to three sectors, each run by a crew of three - an area supervisor (who also directly supervises two units), an assistant (who also supervised a unit of his/her own) and a recorder. This crew makeup makes for a reduction of manpower, relative to the former UC staffing-strategy in which each unit has its own supervisor and area supervisors do not do any direct unit-supervision and recording. It is still not as thrifty as the 'Israeli' system used in D2, where three units and 9-12 diggers would be supervised by a crew of two. The three crews were: Sector 1 (NE) - Yiftah Shalev, Erin Dintino and Bruce Redwine. Sector 2 (S) - Becky Martin, Rebecca Karberg, and Barbara Del Rio. Section 3 (NW) - Jessica Nager, Melissa Thornton and John Chesley.

As it turns out, most of the new area (all of sector 1, a large part of sector 2) is dominated by a single Roman building. This building had very thick walls - most of them unfortunately robbed completely or to foundation level - and thick cement-plaster floors. The plan consists of a series of tiny cubicles or basins (about 1m. x 1m. each) with thick floors, arranged in a N-S line along the building's western wall; and three elongated E-W rooms/spaces east of these. The southern room was tiled with large tesserae and a portion of the floor makeup consisted of a very large stone drum - either a column in reuse or a round stone cut specifically to serve the function of the room. The middle room had a cement-plaster floor, which was connected to the lowest room by a small plaster channel that had been constructed through the wall. The north space was a corridor with a drain running through it, and a doorway to another space (courtyard?) to the east, which has not been excavated. North of this corridor was another series of small rooms or basins.

There was clear evidence in several places that this structure had at least two construction-phases. Two superimposed floors were found in several rooms, and some of the walls had been rebuilt. It is probable that the southwestern extension of this building (into sector 2) is a later addition. A pile of water pipes and hypocaust tiles was found on the floor of the southernmost room of that extension. At least one hypocaust tile was in situ and impressions of others could still be seen on the surface of the floor. John Berg made the observation that the earlier phase of the building was ashlar-built (on rubble foundations); and that wherever cement-mortar and/or 'blobby' concrete foundations appear we are probably in the later rebuild stage.

What can be the function of this building? Two very different hypotheses were put forward. The hypocaust tiles and pipes (found mainly in one room of the southern extension) suggest a bathhouse. The row of thick-floored cubicles along the western wall might be furnaces or water tanks. The rather rough mosaic in [what would have been] the main room of that bathhouse might argue against this interpretation. The other possibility is that we are dealing with some sort of an industrial complex - possibly a large winepress. This would explain the rough tessellation in the main room and the intricate drainage system between the three E-W halls. But what about the hypocaust elements? For now, the working hypothesis is that we are excavating a (public?) bathhouse. But all bets are still on.

The biggest find of the 2000 season (some would argue the biggest find ever…) was the Hellenistic 'mask and garland' mosaic, found in a dump in the southeastern unit of D1. Sector 2 of D1 was opened in 2004 directly north of this unit, in the hope of recovering more of the mosaic and narrowing down its stratigraphic provenience. The northern part of the same dump was excavated - filled with many architectural fragments (ashlars of various sizes, column drums, thresholds, voussoirs, and rafter-supports) as well as fresco and mosaic fragments. Unfortunately, no pieces of any significant artistic merits were recovered. Most of the recovered mosaic was white limestone with haphazard tessellation; presumably these areas of the floor that would have been obscured by the klinai, or dining couches. However, it was established that the dump is earlier than the 'bath-house' complex. The robber trench of the southern wall of the building clearly cut the dump, and several small mosaic pieces - which clearly belong, size-wise, to the same mosaic - were sealed beneath the thick floors of that building.

A probe in area F

A single unit which remained unexcavated at the top of the staircase leading to the southern entrance to temple F was now opened. In their analysis of the plan of the temple, John Berg and Andrew Stewart had noted that the (robbed) temenos wall north of this unit was at a different orientation from the part adjoining the southern entrance - this raising some doubt as to whether or not these are in fact parts of the same wall.

John Chesley conducted a 'quick and dirty' excavation to verify this. The temenos wall was not found in this unit, but its robber trench was located - as well as several super-imposed crushed kurkar surfaces of the street east of the temenos. John apparently liked the experience enough to contemplate coming back to Dor with a group from Rhodes College - where he assumed a new teaching position upon returning from the excavation.

On-site Scientific Analysis

2004 was the third season of experimenting with integrating a 'classical' excavation approach with an on-site lab doing on-line analyses of sediments and finds. We started small -in 2002, Ayelet Gilboa and some of the Israeli field-staff spent a week with Steve Weiner and his staff conducting intensive sampling of the baulks of the 'old' excavation. In 2003 the Weizmann contingent joined the study/publication season, and in addition to continued sampling of sediments from the vertical balks we opened a small area (in G) specifically in order to obtain horizontal sampling surfaces and to try out strategies of integrating scientific sampling and analyses into the normal work-flow of the excavation. This season was the first attempt to use the on-site lab in a full excavation environment. The difficulty is one of scale and pace. To obtain a macro picture of an urban site of Dor's size and complexity a strategy of large-scale exposure by big crews working fairly fast is deemed necessary. For the micro scale the ideal situation is one where a static field can be studied in great detail - being sampled and resampled at leisure. The two strategies are seemingly at odds. However, the potential dividend of integrating them is enormous. It is vital for the excavator to have as much information as possible on line while she is excavating - exactly because archaeological excavation is not a process of 'observation' but of intervention - where decisions made in the field affect the overall interpretation. The challenge is to develop sampling strategies and work-flow tactics which will enable us to excavate 'as fast as possible and as slowly as necessary' so as to integrate the new tools into the normal work procedures. Much has been learned by both parties this season, and Steve Weiner, John Berg and I will continue to work on this problem in the next one.

In the context of this work Ruth Shachak-Gross continued her program of micro-morphological examination of 'floor' surfaces - complementing the time-consuming preparation of sediment blocks for thin-section analysis with ad-hoc examination of raw soil samples under a petrographic microscope. Such examinations had enabled, inter alia the identification of phytolith-layer 'floors' - as in area G - already mentioned above.

Francesco Berna completed his research on high temperature heat-effects on sediments, concentrating on the 'burnt layer' and the 'metallurgical installations' - phases 9 and 10 in area G, respectively. Francesco finished his post-doc research period at the Weizmann institute and has moved on to accept a position at the Universita degli Studi di Bologna. We all wish him the best of luck in his new work. The research on heat-effects will be continued, however, by Adi Behar, who will use the techniques developed by Francesco in her study of pyrotechnic installations - tabuns, kilns, etc. She started this research this season - sampling the kilns left in the balk in area H - and will continue it as part of her Ph.D. program.

Besides undertaking (with Genia Minz) the collection of samples in areas not covered by the other members, and acting as deputy-director of the Weizmann team - Elisabetta Boaretto continued her quest to locate ecofact categories suitable for radiocarbon analysis. The most exciting prospect seems to be the dating of phytoliths. Though primarily silicate, phytoliths do contain minute amounts of carbon. Ruth and Rosa's discovery that [at least some of] the thin white 'floors' often encountered in our Iron Age levels are actually phytolith layers, and that each of these layers, if followed and stripped carefully enough, can be considered a single-year 'event' -means that we may have stumbled onto a plentiful source of short-lived radiocarbon samples suitable for high-precision dating. Elisabetta and Genia have demonstrated that the phytoliths from a single 'white floor' can contain enough carbon for AMS dating and that the dates are 'in the same ballpark' as the ones obtained on other materials from the same typo-stratigraphic horizon. To check the generality of this phenomenon Elisabetta also sampled phytolith floors from Megiddo. Much more careful sampling and comparison of seeds and phytoliths from the same contexts will have to be conducted to assess the accuracy and precision of phytoliths as raw-material for high-resolution dating, but the prospects are certainly exciting.

Glasshouse work

I have often repeated that, my personal main objective at this stage of the project is to work out a system whereby the published output of the excavation exceeds the future debt incurred by each additional season. (to hope for more than that is unrealistic…) This means a) to dig less; b) to find ways to speed the processing-time for new data coming in; c) to direct some of the effort at working on past materials. We see this not only as a compulsory task, but as a scientific goal: the delay (and often lack of) publication is the number one predicament of contemporary field archaeology, and innovative ways to use modern technology to overcome this must be found. In this quest, as in other topics, the new Dor team seeks to take a leading position. This season has seen the beginning of the realization of this goal.

A large part of the staff was employed in the glasshouse - either throughout the season or on a part-time basis, as well as a greater proportion of the volunteers than has been the case in the past. The [often-vexing] task of running work-details and of coordinating between the glasshouse and the tell fell onto a captive director, a baby and a graduate student: Sarah Stroup (a.k.a. the captive), her captor - three-months old Maxwell Stroup, and the indefatigable Amir Haim directed the excavation's least popular feats with much good nature and aplomb. They were aided by a string of students and volunteers cajoled or coerced to leave the sun-drenched topside of the tell for the dark (but air-conditioned…) recesses of the museum.

Computerized Registration

After contemplating and planning it for the better part of three years, we took the big plunge this season - switching to a new, all digital, registration system. The idea is to do away with cards, tracing paper, and film altogether (except as drafts and backup) - at the end of the season we walk away with all of the season's data encapsulated on a CD. This is the first necessary step towards a much more ambitious goal - of computerizing the entire workflow from tell to final publication, and putting out the latter as a 'live', hyper-linked, searchable e-publication.

The design and implementation of the new system were accomplished by the twin-magicians Sveta Matskevich and Talia Goldman, and the main burden of beta-testing it fell on the area recorders - Bruce Redwine, Barbara Del Rio, John Chesley, Neels Kruger and Erika Cruywagen. Two elements of the former computer data-base - the 'basket list' and the 'locus cards' remain essentially unchanged (except for the basket and locus designations, which have changed, and the graphics on the locus card which are now digitized, about which see below). The three major additions are:
A 'graphics' table has been added to the database. It is controlled, like the other tables, by a 'bank' of unique image-designators assigned to every person who produces images on the excavation, whether they are top-plans, sketches, field-photos, field drawings, artifact drawings or computer-produced graphics.
The photographs - taken by Theresa Ortballs - are now all digital. Field-images are sorted, inserted to the graphics database, and linked to the appropriate locus cards by the area supervisors.
Field-drawings (the famous 'tzetalach') were produced by Sveta Matskevich. Talia Goldman then scanned them, rectified them onto the site-grid, vectorized them, attributed each drawing-object (e.g. 'outside line of wall', 'elevation point' etc.) and defined 'features' (groups of drawing elements e.g. Wxxx, Fyyy). These features were then added to the active top-plan for the area.
Each recorder received a [rasterized copy of] the active top-plan every day. On top of this they added (using Adobe Photoshop®) locus numbers, basket numbers, daily elevations, and new features which had not yet been drawn. Copies of the completed daily top-plans are saved in the graphics database.
Unit supervisors either use the rasterized top-plan as a basis for the sketches in the locus cards - or drew their own manually and then scan them. These sketches too are inserted as independent images to the graphics database and linked to the appropriate locus cards.
Artifact drawings or computer-renderings (see below) are similarly entered into the graphics database.

Digital artifact rendering and analyses

The mathematical artifact shape-analysis project, headed by Uzy Smilanski, continues - and this season its objectives were to integrate the already-developed 'profilograph' applications into the 'normal workflow' of the excavation, experiment with a new 3-D scanning system, and compare the results of the three drawing-systems - 3D laser scanning, profilograph, manual drawing and scanning.

The newest gadget we experimented with is a 3-D laser-scanning camera, loaned to us, complete with its operator / developer, by the Technical University of Vienna. Hubert Mara, who developed the archaeological software for the camera, demonstrated its use by teaching a volunteer who had no prior experience with it - how to scan potsherds. By the end of the season she managed an average production speed of 3 minutes per sherd. By far faster than any manual method (or even semi-automatic, as with a profilograph). Comparison of the products of the three systems (manual - by artist Vered Rosen, profilograph - operated by Igor Mintz, 3-D scanner) shows that both computerized systems achieve very similar results - sometimes different than those of the manual drawing. However the positioning of the sherds (which is done manually both in manual drawing and the profilograph) is often less accurate with the automatic 3-D procedure. In addition, the fact that the 3-D scan produces far more information opens up new avenues for research - for instance estimating the regularity in the construction of the vessel.

Catching up on publication

As already stated, completing the publication of Ephraim Stern's excavation is one of our main goals, and work on that continued during the season. Jeffrey Zorn came down from Cornell and stayed with us for two weeks to finish working on the stratigraphy of area G. Jeff finished analyzing the last of the grid-units which were not yet done, and completed the locus-index. Based on that index Ayelet Gilboa the Area G Iron Age loci into assemblages - by which the finds will be presented in the publication, and, aided by Naomi Smilansky, Francine Landau and Tim Doran, began working on pottery plates for area G. All the drawn potsherds were scanned, inserted to the graphics data-base, sorted according to phase and assemblage, the drawing of incomplete assemblages and missing potsherds was completed and the assembly of drawings into plates was begun. Orna Hillman and Nati Kranot had meanwhile continued working on the stratigraphy of area D2 - though but the happy occasion of the birth of Ronny cut Nati's season short. Noa Raban and Yisrael Shur continued meanwhile to work on the faunal remains and pottery of area D2, respectively.


Following the state-of-conservation survey conducted during the 2003 season, we set out to rectify the most pressing of the recommendations of this survey. Several conservation projects were undertaken this season, directed by Ravit Lin.

A dire safety warning of the survey was that two high cement walls, standing on the balk above area D2, were in imminent danger of collapse, and needed to be removed. This entailed an engineering problem. The walls could not be manually dismantled without either having people on them or under them - which could not be contemplated for reasons of safety. Simply knocking them down would have made irreparable damage to conservation-worthy Iron age remains below them. After consultation with IAA conservation engineers the course taken was to partially backfill a portion of area D2 on one side of the walls, to sandbag the features on the other side of the walls against possibly mishap, and then to pull the walls onto the backfilled area using a long-armed heavy excavator. Preparation for the tractor-work and cleaning up afterwards necessitated several mornings of concentrated work of the entire excavations' workforce - and the tractor work itself was no small expense.

Another safety issue pointed out by the survey was the tall vertical balks around Area D2, which are deteriorating. This problem was partly addressed by locating our main excavation areas around the core of D2. Thus excavation in the Late Iron Age portions of D2 had lowered the balks between these areas and the Early Iron Age pit. Opening new units west and north of these, in turn, lowers the vertical balks around these. Of course the new units too, will, in the long run, develop high vertical balks which will need to be gradually stepped over the seasons.

The same safety problems apply for area G. Here the deterioration of the balks, as well as high-standing walls was at a very advanced stage. Furthermore, the location of the area at the center of the tell makes it virtually un-drainable, and hence the deterioration process is practically unstoppable. This has led to a heart-rending decision to backfill the area - an operation undertaken at the end of the season.

A fourth project undertaken this year was the construction of an access-stairway from the southern beach to the tell. This was done partly as a token to the community which uses the site and cares about it, and on whose support we continually draw, and partly as a straightforward conservation project - to give the public an opportunity to access the site without clambering over antiquities.

The above only specify urgent and extreme safety-measures undertaken this season to combat the cumulative destructive processes of twenty-five years of conservation-less excavation. The highest priority for actual conservation as indicated by the survey was the western areas - which are more exposed to damage by the elements and by humans. Accordingly, we decided to concentrate our efforts on [part of] area H. The area was cleaned, documented, walls were consolidated, and floors were leveled to provide drainage. The main workforce for this project were the students of the conservation course (see below). Preparation for work (weeding etc.) and the completion of work left undone (or half-done) by these student was then conducted by the 'regular' students/volunteers. We intend to continue the same modes of operation in the next seasons, and tackle small conservation projects on piecemeal basis - until such time as a big investment in conservation can be made by the relevant authorities.

Lastly, a public meeting was held on-site, during the season, with the purpose of starting a dialog between all parties with vested interest in the future of the site. The relevant authorities (Israel Antiquities Authority and the Nature and Parks Authority) laid out their plans for the park, we presented our relevant plans, and members of the community, as well as prospective concessionaires, heard and were heard. This symposium was the brainchild of Ann Killebrew, and is the first step in her heritage program, designed to draw community involvement in the shaping of the archaeological park and to make us - the excavators - and the authorities, take each others' wishes and plans into consideration as we make our own.

Academic Program

Three field-schools were run at Dor during the season. These included the UCB-UW field-school, coordinated by Sarah Stroup and Becky Martin. It consisted of a series of lectures, a series of field-exercises, tasks which had to be completed by the students, and quizzes. The field-school accorded the participants 4 UCB undergraduate credits.

Haifa University used Dor as one of its study excavations. Two ten-excavation-day sessions were held and 14 students participated. Unfortunately, for financial reasons, we could not keep the Haifa U. students in our base at Nahsholim, and they had to commute every day. This was awkward, and is clearly something we need to improve-on in the future. The students usually arrived about 7:00 AM, worked in the field till 1:00, and had lectures and/or field exercises between 2:00 and 3:30, after which they were driven back home.

A completely new experiment was adding a conservation-course to the curriculum of field-schools at Dor. This was (to our knowledge) the first-ever on-site practical course of archaeological conservation held in Israel. It was initiated by Ravit Lin and Ann Killebrew, and was a week-long workshop for graduate students. Archaeology students from Haifa University and Hebrew University enrolled, as well as architecture students from Tel Aviv's historical conservation program. The course consisted of fieldwork (7:30 AM - 1:00 PM) - in actual conservation projects in area H (see above) and in the museum, and lectures (2:00 - 6:00). Next season it is planned to enlarge this course by adding an English-speaking session on conservation and heritage, sponsored by Haifa University's Department of Overseas Studies.

Impending Dor events

Two Dor-workshops are in the works: In January, Liz Bloch-Smith will come to Israel to begin work on a new final-publication - the Iron Age of area B - funded by the Shelby White-Leon Levy Program for archaeological publications. The workshop will be held in part at the Hebrew University in Jerusalem and in part at Dor.

In February, we intend to hold the second Seattle workshop on the publication of the Persian-Hellenistic strata in D, supported by the Getty foundation's collaborative research grant. The workshop will be integrated with a research seminar for students in the University of Washington.

A third, one-weekend long session is planned in Israel, in order to finalize the basic stratigraphy of the early Iron Age of Area D2.


The excavation, as usual, is supported by the Berman Foundation for Biblical Archaeology in the Hebrew University. We wish to thank Israel Hirshberg, the director of the Glass House Museum, for allowing us to take over a substantial part of the building, to Eyal Goren from Kibbutz Nahsholim, for the possibility to place our two new containers near the cow sheds (but beware! The cow sheds are no more and the area is intended for development), and to Aharon Refter, head of the faculty of Humanities administration at Haifa U., who financed the sealing of our leaking containers. The Israel Exploration Society handled, as always, the finances and logistics. Specific donations were made by the Guttman Foundation and by individuals who wish to remain unnamed.


As I write these words I realize that perhaps the best testimonial to what has been accomplished at Dor this summer is that a mere listing of the different projects engaged-in by different members of the team has become a dozen-page article! I sincerely thank each participant for whatever part - large or small - they have played in each of these projects, and invite you all, as well as anyone who had not yet been at Dor and happens to read this, to join us in July 2005 in the next season at Dor.