Between June 15th and September 7th, 1994, INA completed its 11th and final excavation campaign on the Late Bronze Age shipwreck at Uluburun, near Kas in southern Turkey. During the campaign we made 1,857 dives totaling some 607 hours of excavation time under water. This brings the number of dives to 22,413, for a total of 6,613 hours of excavation time on the wreck, and makes the project the longest and deepest excavation yet conducted by INA.
During the 1993 excavation campaign, it was expected that the removal of the remaining bulky artifacts from the site, comprising two stone anchors, 15 copper oxhide ingots, and a large storage jar or pithos, as well as the completion of the few unexcavated areas could be accomplished within a short period of only two months. When the last layer of copper ingots was raised, however, additional hull remains were exposed. Due to the poor preservation of these remains, coupled with the realization that what was deemed to be a shallow deposit directly upslope of the large boulder-like rock outcrop located centrally on the site, was in reality a deep gully densely packed with artifacts and organic materials, it became apparent that work on the site could not be completed during the allotted time. In 1994, it was thought that a campaign of two months' duration would be sufficient for a full study and excavation of these remains. But the depth of the deposit foiled our estimates again, and the season had to be extended by three weeks in order for the site to be completely cleared. Our toil and frustration did not go unrewarded, as many of the 1994 campaign's important discoveries were made in this very deposit.
Fieldwork at Uluburun in 1994 may be summarized in the following four categories: the complete excavation of a deep deposit in the central part of the site; probing and excavation in a large area just upslope of the designated wreck boundaries; excavation, mapping, and recovery of all hull remains; and the production of a three-dimensional, computerized topographic map of the entire site.
As during the previous season, much of the excavation in 1994 was centered in the area just northwest of the large rock outcrop (grid squares K-P 15, L-P 16, N-P 17-19, fig. 4). Work here was devoted not only to the completion of previously unexcavated areas, but also to the study, documentation, and recovery of the portions of the ship's hull discovered in 1993. After the hull remains were removed, the areas under them were excavated down to bedrock so as not to miss artifacts that might have slid under the ship's planking.
The area roughly within grid squares L-N 15-16 again yielded a nearly complete assemblage of the general artifact types found on the wreck during earlier campaigns. Excavation progress in these areas was painfully slow due to the profusion of beads, other small finds, and a dense layer of organic deposits, all of which required great care to excavate. In view of the ship's east-west orientation and that the hull came to rest on and listing against the northern side of the large rock outcrop, it is apparent that the area in question acted as a natural catchment basin for objects sliding down the slope. This may be especially likely if part of the hull, including the turn of the bilge, broke away and came to rest downslope, against the western perimeter of the large rock outcrop. The starboard turn of the bilge might have acted as a primary catchment that, later in wreck formation, dropped down to the base of the rock outcrop. The dense concentration of organic materials found in the lower levels of the area do not contradict this idea. The presence of two near-certain hull planking fragments found here also makes it quite reasonable to imagine that some portion of hull planking ultimately came to rest against or near the large rock outcrop's western perimeter. Both pieces, oriented parallel to the ship's keel- plank axis, clearly display tenons and tenon pegs consistent with those found earlier on the portion of the hull raised in 1993.
Because of the richness of deposits in this area, both in organic and artifactual material, much of the sand was bagged and tagged for sieving on the surface. This facilitated the recovery of vast quantities of seeds (pomegranate, fig, grape, olive, coriander, sumac, various grass and weed seeds, and pulses), nuts (almond and pine), and other organic materials, and many faience and agate beads (figs. 2, 3). The sievings also yielded a profusion of fish bones, but it is not known whether these bones represent the remains of shipboard food or of fish that made the wreck their home before their death.
The area just to the northwest of the large rock outcrop (grid square M-16), where the southern end of the second row of copper ingots originally lay, was riddled with wood fragments. These pieces, seemingly deposited in two distinct layers, were not oriented in any specific direction. They comprised a variety of shapes and profiles, none of which resembled hull planking. Although too poorly preserved to be positively identified, it is possible that they were parts of some superstructure or wooden object carried on the ship. Under this jumble of wood were more Cypriot export ceramics, almost certainly spilled from one of the pithoi found in this general area. Cypriot ceramics recovered here included milk bowls (White Slip II), oil lamps, and several White Shaved juglets, making the last type by far the most common example of Cypriot ware to be carried aboard the ship. The largest of the ten pithoi on the wreck, which was removed from this precise location during our first campaign in 1984, contained 18 pieces of Cypriot ware, neatly nestled inside the jar.
Also in this area was a pithos unlike any found on the site (KW 5833). Not only is this jar the shortest of the ten recovered, but it also has a pair of ribbed handles attached to its rim, terminating on its upper shoulder zone. As with the other pithoi, the newly discovered jar appears to be Cypriot in origin. It was with great anticipation, therefore, that we awaited its full excavation to learn what the jar contained. Had it originally held some liquid, such as olive oil or water, or did it serve as a china barrel, as did at least three of its counterparts found earlier on the wreck?
The sediment inside the jar was carefully decanted and placed in marked polyethylene bags for sieving ashore to recover small artifacts and organic material, which otherwise may have been missed. The jar was virtually packed with nearly every organic material recovered elsewhere on the site. We were surprised to find two short lengths of rope, our first from the site, and bits of twine. The first significant artifact to appear from the jar was a beautifully crafted and preserved wooden (probably boxwood) lid of an oval container. That some luxury objects carried on the ship were of perishable nature is clearly demonstrated by the chance survival of this lid, which must have fallen in the jar accidentally, and another wooden piece in the form of a circular base for a cylindrical container found next to the pithos. The latter piece is similar in shape to the base of ridded ivory containers (pyxides), well known from this period. At the base of the pithos were two Cypriot milk bowls nestled inside a carinated bowl (KW 5734,5735,5736: fig. 1). In addition to the complete bowls, several milk-bowl and White Shaved juglet sherds surfaced. It seems that this pithos also was used for transporting Cypriot export wares, and that a specific storage jar type or size was not reserved exclusively for this function. As the shape of the pithos does not appear to specify the nature of its contents, it is more likely that the jars themselves comprised a part of the ship's ceramic cargo, and were not intended as permanent shipboard storage containers. The practice of storing more delicate ceramic vessels in larger jars is conducted even in modern times during the porterage of these wares from production centers to markets.
In addition to hull related items such as planking, dunnage and ballast stones, finds here include shipboard materials comprising fish hooks, fish-net sinkers, sewing needles possibly for repairing sails, drill bits, and a bow drill. Artifacts from the ship's cargo include raw materials in the form of quarter oxhide tin ingots, a complete copper bun ingot and several fragments thereof, glass ingots of two colors, a hippopotamus tusk, and ostrich eggshell fragments. Manufactured goods comprise Syro-Palestinian (pilgrim flasks and amphora sherds) and Cypriot export wares (oil lamps, milk bowls, White Shaved juglets), fragments of metal vessels, seashell rings (KW 4852, 4877, 5036: fig 5), tortoise carapace fragments (almost certainly sound boxes of musical instruments), and some beads. Personal effects of the ship's crew and that of the merchants aboard, on the other hand, include pan-balance weights (complementing the two pans in a wooden case found in 1993), a spear head (KW4885) and projectile point, a whet stone pottery (two Mycenaean transport stirrup jars), an astragal, an ivory finial in the shape of an opium capsule (KW 5156), beads (faience, glass, agate, quartz, steatite, and amber), many fossil sea shells, and beach polished lips from helmet (cassidae) shells for use as necklace pendants or some other form of adornment. Highly regular and polished beach pebbles found in this area probably were collected for their attractiveness rather than simply representing ballast accidentally shoveled into the ship's hold along with the larger cobbles. One such pebble was modified partially at one end by grinding into a V-shaped section.
To the list of personal possessions, we may add a steatite lentiod seal blank or bead, and two cylinder seals. The smaller of the seals is of a soft, black stone, probably steatite. Because the seal is worn extensively, it is nearly impossible to make out the nature of the carved designs on its surface, although one may depict an outline of a human form. The second seal (KW 5084), recovered in several fragments, is of faience. Its design of three parading stags with a plant (one resembles an opium poppy) separating each register, is clearly preserved. According to Dominique Collon of the Department of Western Asiatic Antiquities of the British Museum, the representations of stags are in the traditional Mittannian style, with numerous examples found at Alalakh (on the Syrian border of southern Turkey) and Beth Shan in Israel, although the antlers on the Uluburun seal are executed somewhat differently. She notes also that the type is most common in Palestine and probably spread to Syria from there.
What may be the season's most important find also was made here. At the level of the wood, and resting on the shoulder of the pithos described above, was a leaf from a diptych (KW 4863: fig. 6), the second such item to have come from the site. Recovered intact, the leaf is narrower and taller, but otherwise identical in appearance to the one found in 1986. Unfortunately, its three-piece hinge, probably of ivory as in the first example, was detached and missing. One of the two ivory hinge sections found during earlier campaigns may belong to this diptych.
Beneath the pieces of scattered hull planking in this area lay several sections of wood that at first were puzzling. They measured only 8-11 cm wide and 2-3 cm at their thickest point, but one was preserved for about 90 cm of its length, making them too light to serve as structural members of the hull. Because of their proximity to the latticework fencing discovered during the 1993 season, at first we wondered if they might be related, or whether the former pieces represented a separate feature also situated at deck level. That they were found at a deeper level than the planking suggests that they may have originated high up on the ship structure. Presumably, they would have fallen down before the heavier hull planking was deposited on top of them. Although most pieces were poorly preserved, our examination of the largest piece suggests that they may be blades from the ship's sweeps or the oars of the ship's tender, because they are too light to have served as the ship's quarter rudders. The rounded end of the best preserved piece, which tapers to an edge in section view, and its hydrodynamic foil shape, appear to strengthen our interpretation.
We spent much of the 1994 field time recording and removing hull wood preserved in areas corresponding to the starboard section of the hold laden with the third (grid squares N-O 17) and fourth rows (N-O 18) of copper ingots (figs 7, 8, 9). Each section of hull is somewhat disarticulated, crushed, and distorted, making the identification and association of some of the poorly preserved planking in these areas difficult and arbitrary. Furthermore, the parts of the keel-plank preserved in the deepest section of the site are slightly askew (to the south) of direct alignment with the 1.7 m long section of keel-plank preserved higher up on the slope, and removed in 1993.
Unlike the hull remains at the deepest end of the site, much of the wood under the third row of copper ingots was badly eroded and had shifted a little after settling on the sea bed. Therefore, the identification of strake numbers and their specific association with the remaining hull timbers was more problematic. The force of dispersal appears to have been more violent here than at the upper or lower hull areas. This probably was due to the steepness of the seabed gradient in this area, which, compounded by the weight of the copper ingots, caused the hull to snap off at this point, leaving the forwardmost section of hull to settle on the seabed with less distortion. The planks appear to have slid a little to the south, becoming slightly out of alignment with the preserved hull section just downslope of them. Strake number three (the third strake from the keel) overlaps strake four, while strakes five and six meet at an angle. Fragments of planking were wedged or forced under the upper end of strake six. Organic sediment, artifacts, and wood fragments settled underneath the upper edges of this section of hull wood.
Although only partially preserved and in poor condition, this section of hull includes an important construction feature not found elsewhere on the wreck. As the hull remains have yet to be studied in detail, the purpose of this feature is not fully understood. It is possible that we have here a flat scarf (a pining of two planks whose diagonal ends were cut off perpendicular to their lengths), or a drop strake (a strake of planking discontinued near the bow or stern because of decreasing hull surface area), indicating that we are approaching the ship's bow. Support for these options comes from the slight tapering of the garboard towards the bow, which is especially pronounced at the deepest or easternmost section of hull re- mains. There is one well-preserved mortise-and-tenon joint at the scarf, and vestiges of a second. All mortise-and-tenon joints on the planking are pegged with a c. 2 cm diameter peg driven into the tenon on each strake. Not all the peg pairs could be located due to the poor state of plank preservation in some areas, but spacing between adjacent pegs on the same plank averages 23 cm. A large preserved patch of thorny burnet, a kind of brushwood placed under the ingots to provide a cushioning effect on the planking, indicated the exact position of ingot placement on the wood. Additional dunnage under the layer of thorny burnet is represented by only a single large branch at the lower (east) edge of the hull remains. In some places the planks were partially impregnated with copper, causing a slight "petrification" of the wood, a feature particularly noticeable on the sixth strake.
Many artifacts had found their way under the wood, including pan -balance weights, ivory finials, parts of ivory scepters (KW 4751, 5169: fig. 10) and a finial for a scepter in the shape of a pomegranate (KW 4806: fig.10), an ivory figurine in the shape of an acrobat that may have served as a handle for some unknown object (KW 5754: fig. 11), crescentic pendants visually identified as a tin-lead alloy, a blank amethyst scarab (KW 4851: fig. 13), a possible scarab blank of a gray stone (KW 4860: fig.13), amphora stoppers or bungs, thousands of faience beads, and ceramic vessels. The entire area was thick with organic materials; under the patch of thorny burnet were heavy concentrations of murex opercula, probably thicker than anywhere else on the site. The last item may be an ingredient used in making incense, a practice still alive today in many Arab countries.
The remaining section of hull at the deepest end of the site, or the area corresponding to the fourth row of copper ingots (N-O 18, fig.9), is approximately one square meter and comprises a section of the keel-plank and portions of the first six starboard strokes. Because part of the keel- plank was preserved here, the starboard garboard and most of the subsequent starboard strakes could be identified and numbered. The planks followed the same width-pattern observed in the upper hull area removed in 1993. Staffing with the garboard (c. 15-16 cm wide), a narrow strake alternates with a wide (c. 25 cm) strake. This pattern is apparent in all three preserved sections of hull wood, and appears to have been a regular feature of the hull. The orientation of the keel-plank in this section does not directly conform to that of the hull section removed in 1993. The slight difference is probably due to the shifting of this section to the south, against the large boulder-like rock outcrop, as it settled onto the seabed. Strakes three through six roughly correspond to four of the remaining five strakes in the hull section preserved under the third row of copper ingots, but this relationship is not very clear due to the poor state of preservation of the hull wood in that area.
Overlying the hull wood was considerable dunnage comprising heavy branches lying perpendicular to the strakes and keel-plank. Partially overlying and compressed into the branches from the weight of the ingots above was a thick layer of thorny burner. Most of the dunnage was found on the upper half of this section of hull, and was presumably meant to cushion the heavy row of ingots placed directly on top of it. The position of the ingots, however, corresponds to the entire section of hull preserved, not just its upper half. it is likely, therefore, that the ingots were originally centered on a swath of dunnage, and that the entire row slipped slightly downslope when the ship settled on the seabed.
Once mapped, the dunnage was removed to expose the hull remains for further study and detailed recording. The planking revealed a total of 27 pegs, but no construction features such as scarfs. There is only one peg pre- served securing a joint between the garboard and the keel- plank, the latter extensively eroded on the underside. The garboard itself is also poorly preserved and lies somewhat askew to the main section of planking. Strakes two and three are the best preserved with strakes four, five, and six progressively less well preserved. The seabed slopes toward the large rock outcrop in this area, and apparently as the periphery of this section of hull eroded away, the preserved section listed to the south and slipped toward the rock outcrop. Additionally, the planks display a pronounced concave curvature, like a slide, from top to bottom, almost certainly caused by the weight of the ingots pressing on the wood as it softened over time. During deformation in both directions, the upper (or eastern) plank ends curved up and over to the south as they were pressed into the sea bed. Strakes one, two, and three have tight seam lines preserved. At strake three, they begin to pull away from each other and tenons could be seen between planks. In strake six, the tenons are broken. Certainly, the curvature does not represent the ship's original shape, nor does it conform fully to the contour of the seabed beneath it. There is no visible evidence for frames in any of the preserved sections.
The season's greatest concentration of organic material and artifacts was found under planks four, five, and six, caught in the deep "V" formed between the southerly slope of the seabed and the large rock outcrop. The deep, densely packed layer of organic material (12 cm thick in some places) seems to be a compressed matrix of the eroded outer side of the planking and the usual murex opercula, olive pits, fish bones, etc., which percolated under the planks. Many glass and faience beads that had slid down the slope also were found here, along with a steatite lentoid seal of a Mycenaean merchant (KW 4855: fig. 16), the second such seal found on the wreck.
With the hope of discovering several obviously missing artifacts from the site, such as the pommels of the two Mycenaean swords found earlier, and the base half for a Mycenaean flask, we incorporated into our excavation area some 50 m: of seabed located directly upslope of the site proper. Because it became apparent after several years of excavation that the Uluburun ship originally had struck the seabed at some higher point on the slope before sliding a little and eventually settling in its present location, we thought that some of these missing items might be found there. Although a thorough excavation of this area did not yield any artifacts, we left the site with the knowledge that the only previously unexcavated area of seabed suspected of containing these artifacts was sterile.
To learn what may have remained of the original hull profile, cross-sectional views were made at three stations, one on the hull section under the third ingot row, and two on the section under the fourth ingot row. While these views help us to understand better the hull and the process of wreck formation, for the most part they revealed that the surviving portions of the hull had deformed to conform to the shape of the seabed at those locations. An additional mapping scheme included a three-dimensional contour map of the entire seabed under excavation. The purpose of such a map is to illustrate the irregularity and pronounced slope of the sea bed at Uluburun, and to enable us to comprehend the natural forces that affected the dispersion of the ship and its cargo. While this project was high on our agenda, it defied our attempts until this year for a variety of reasons. In 1994, it was realized only through the diligence of three expedition members who spent their entire season taking some 7,000 measurements under water. Mark Smith processed this data on a computer to obtain the map reproduced here (fig. 12).
After 11 seasons of excavation, we now can proceed at full pace with a meaningful study of the Uluburun ship's hull remains, cargo, and the personal effects of those who sailed on her either as crew or passengers. Already, preliminary examination of the finds has suggested to us that the ship had set sail from either the Syrian coast or Cyprus, or both, and was sailing for a destination somewhere west of Uluburun before its demise on the rocky promontory. Much of the material from the site provides important new information about the history of interregional contacts in the eastern Mediterranean and beyond. Before these objects may be taken at face value and integrated into our expanding knowledge of Late Bronze Age trade, the establishment of a precise date for the wreck is a primary concern.
Stylistic considerations of the pottery from the wreck, especially that of Mycenaean origin, suggest a broad date in the last three quarters of the 14th century B.C. In 1986, this date was somewhat refined by the discovery of a gold scarab bearing Nefertiti's name. Jim Weinstein's study offers maximum dates for this unique scarab as high as about 1376-1358 B.C., or as low as approximately 1339-1317 B.C., depending on which chronology is used. Therefore, the Uluburun ship must have sunk sometime during the range of dates suggested for the scarab or, considering the signs of extensive wear on the scarab, somewhat later. Recent tree-ring counts of cedar logs from the ship by Peter Kuniholm of Cornell University have given us a date of 1356+37 B.C. for the last preserved ring in the logs. The ship must have sunk sometime after that date as the logs do not include bark. Due to the floating nature of the Bronze Age Mediterranean conifer tree ring sequence, this date appears somewhat vague. However, Peter Kuniholm informs us that some refinement may now be possible, allowing for the placement of our date at the lower end of this range, or sometime in the last quarter of the 14th century B.C. We hope to fine-tune this date with additional samples that include the ship's keel-plank. As it stands, this absolute date has considerable implications for crossdating pottery and other artifacts from the Aegean, Syria- Palestine, and Cyprus.
Again, our preliminary work has shown that the ship itself was probably Levantine or Cypriot in origin, as evidenced by the type of stone anchors carried aboard (petrological work on the anchors is currently being undertaken), Syrian oil lamps with wick-charring at their nozzles (Cypriot lamps, quite distinct in shape and fabric from the Syrian lamps, carried as cargo are in pristine condition), the Syro-Palestinian origin of a few bowls and other vessels probably comprising a part of the ship's galley wares and the Near Eastern weight standards represented by the more than 130 pan-balance weights found on the site (fig 15, p.16). To these considerations we also should add the predominantly Syro-Palestinian origin of most shipboard tools, weapons, and personal effects of those aboard.
That the ship included among its passengers at least one, and more likely two, wealthy Mycenaean merchants, envoys, or individuals of some rank is clearly shown by the several knives, spears, chisels, jewelry (primarily in the form of quartz, faience, amber, and glass pendant beads), a cloak pin, and, more importantly, two Mycenaean swords, and a pair of lentoid seals, one of which was found only this season. Moreover, Jeremy Rutter of Dartmouth College, who is studying the Mycenaean pottery from the wreck, informs us that most of the two dozen or so pots from the wreck are types not common among the repertoire of trade goods in the Levant. Probably these represent personal wares of the merchants aboard. Half a dozen socketed-spear heads, including one from 1994 (KW 4885: fig. 14), of a type presumed to be introduced to Greece from neighboring regions to its north, along with a stone ceremonial mace-axe or scepter of a type found only in the northwestern and western Black Sea regions, with the best parallel coming from Rumania, suggest connections that reach far beyond the boundaries of Greece. The nature of this connection, its development, and its socioeconomic implications for the Late Bronze Age cosmopolitan world will be the focus of future studies. We now will embark on a new avenue of investigation: a quest for knowledge through the systematic and careful interpretation of the material we have so painstakingly excavated over the past 11 years.
With the fieldwork behind us, we already have begun directing our energy and resources to the conservation, restoration and documentation of the Uluburun finds. A positive move in that direction has been launched through the efforts of Jack Kelley, chair of INA's Board of Directors. As a direct result of this effort, Claire Peachey, INA's second full-time conservator, already has begun work exclusively on the conservation of Uluburun material in the Bodrum Museum of Underwater Archaeology.
Acknowledgements: The final excavation campaign at Uluburun was generously funded by the INA Board of Directors, Texas A&M University, the National Endowment for the Humanities, and the Institute for Aegean Prehistory without whose unwavering help over the years, the project could not have been brought to such a successful conclusion. As in previous years, Cressi-sub of Italy continued to give us a significant concession on the purchase of their diving equipment. The National Geographic Society recently awarded us a conservation grant to help fund six conservation assistants for seven months and to buy much-needed equipment and chemicals.
Under the overall directorship of George F. Bass, the 1994 team comprised Cemal Pulak, co-director; INA staff Dr. Donald A. Frey, Robin C.M. Piercy, Tufan Turanli; INA staff archaeologist Sheila Matthews; INA counsel James Goold; hyperbaric specialists David Perlman, M.D., and Tom Sutton, P.A. The excavation again was made possible by the participation of veterans, who embraced the final campaign enthusiastically, and put on hold other commitments and responsibilities to show up at Uluburun for one last time. The field participants included archaeologists Dr. Faith Hentschel, Dr. Patricia Sibella, Michael Fitzgerald, Nicolle Hirschfeld, Christopher Monroe, Lillian Ray Martin, Osman Unsal; Texas A&M Nautical Archaeology Program graduate students Brendan McDermott, Stephen Paris, Edward Rogers, and Mark Smith. Harun Ozdas of the Bodrum Museum of Underwater Archaeology represented the General Directorate for Monuments and Museums of the Turkish Ministry of Culture. In Bodrum, Uluburun Ends continued to be conserved under the guidance of INA staff conservator Jane Pannell by Gunes Ozbay and Gulser Sinaci, and volunteers Wadeea Boutros, Franca Cole, Margot Healey, Mohammed Moselhy, and Emily Williams. The Uluburun project also benefitted from a summer school for advanced conservation organized by INA and funded by the National Endowment for the Humanities. The training program, headed by Dr. Claire Dean and assisted by Claire Peachey, included Paula Artal-Isbrand, Norine Carroll, Catherine Magee, Won Ng, and Asaf Oron as participants. INA staff illustrator Sema Pulak continued to prepare object drawings for final publication.