by kind permission of the ETANA team, search the ABZU database:
 |
Home
Periods
Sites
People
papers
projects
Resources
for Study
RESOURCES BY COUNTRY
TECHNIQUES
site map
by kind permission of the ETANA team, search the ABZU database:
|
THE LATE BRONZE AGE SHIPWRECK AT ULUBURUN
The Institute of Nautical Archaeology’s (INA) shipwreck excavation between 1984 and 1994 at Uluburun, near Kas on the southwestern coast of Turkey, brought to light one of the wealthiest and largest known assemblages of Late Bronze Age trade goods yet discovered in the Mediterranean (#1). Dendrochronological dating of a piece of presumably fresh-cut dunnage or firewood suggests that the ship sank sometime around 1300 B. C. (#2)
The ship’s primary cargo comprised approximately ten tons of Cypriot copper ingots and nearly a ton of tin ingots of unknown origin, both cast predominantly in the typical oxhide shape (Stos-Gale, et al., 1998). Other raw materials included glass ingots, ebony logs, hippopotamus and elephant ivory, ostrich eggshells, terebinth resin carried in Canaanite jars, and various foodstuffs. Manufactured goods included ten large storage jars (pithoi), some of which contained Cypriot wares for export. Also found were metal and faience vessels, ivory containers, a jeweler’s hoard of gold and silver, and tools and weapons of bronze. The study of the cargo and personal effects of those on board suggests that the ship was a Canaanite or a Cypriot vessel sailing from the Syro-Palestinian coast or Cyprus to the Aegean. The finds further suggest that there were at least two Mycenaeans of rank traveling on the ship, probably as envoys.
The distribution of the copper oxhide ingots and other objects on the site showed that the shipwreck was oriented approximately east to west, its western extremity being uppermost on the slope, with the rows of ingots running athwartships. The dispersion of some of the cargo to the south indicated that the ship had settled on the slope with a list to starboard, which, on excavation of the ship’s keel, was revealed to be approximately 15 degrees. The stern of the ship rested at the shallower end of the site at approximately 44 meters and its bow at about 52 meters, with additional cargo scattered down the steep slope to a depth of 60 meters or more. The distribution of the artifacts on the seabed suggests a length of about 15 meters for the ship and a total capacity of at least 20 tons, which is an approximation based on the total weight of the cargo, anchors, and ballast stones recovered during excavation. It is not known, however, how much of the ship’s original cargo perished. Any estimates regarding the hull’s basic dimensions and capacity, therefore, must remain speculative.
The ship’s cargo of copper oxhide ingots was originally stowed in four neat rows, with each row running completely from one side of the hold to the other. Many ingots either slipped down the slope after the ship sank or were displaced as the hull settled under the tremendous weight of the cargo, but the basic arrangements of the rows survived. This displacement of ingots, evident in all four rows, but particularly pronounced in the first row (highest on the slope, near the stern) and the fourth (lowest on the slope), makes it impossible to ascertain any ingot’s original position. Those still in place, however, revealed that the ingots overlapped one another in each row like roof shingles, with the direction of overlap alternating in each layer (Pulak 1998, 197, fig. 12).
On the ship were also 24 stone anchors, all of the single-hole type. Extensive surveys carried out by the excavation team around the site and in the area between the shipwreck and the terminus of the cape, from shore to a depth of 50-55 meters, did not reveal additional anchors that could have been deployed as a last measure to prevent the ship from being driven against Uluburun’s jagged rocks. It is possible, however, that such anchors were simply missed by the surveyors or that they lay in inaccessible depths. Two of the anchors are much smaller (21.9 and 25. 9 kg) than the rest and may have been for use on the ship’s boat, or served another purpose entirely. Although only about a quarter have thus far been cleaned and weighed, it does seem that the remaining twenty-two anchors can be loosely grouped into three basic weight categories: the heaviest at about 210 kg, the mid range at 164-182 kg, and the lightest at about 120 kg. Eight were found between the first and second rows of copper oxhide ingots and were probably stored amidships in the hold, on their sides as pairs, one on top of the other. Another 16 anchors were uncovered at the deep end of the site in the area corresponding to the ship’s bow. Of these, 12 appeared to have been stowed downslope of the fourth row of copper ingots, in an area probably corresponding to the forward extremity of the ship’s hold. These anchors, and the eight stored amidships discussed above, must have been spares for replacing anchors that were lost during the voyage. Judging by their sheer number, losses must have been frequent. As such, these anchors would have been kept low in the hold to maximize the ship’s stability, but readily accessible when needed. On the other hand, four anchors, two per side, may have been kept on the fore deck for use in routine anchoring of the ship. How the ancient mariners managed these heavy anchors is not known, but some lever-type device may have been used in their recovery. What this device looked like and how it functioned is now lost to us, but it may have been in the form of a simple derrick fixed to the mast (Frost 1995, 168, fig. 3), or the boom or yard arm of the square sail may have been rigged in a special way to facilitate their handling.
The Uluburun anchors are of a type and size virtually unknown in the Aegean, (#3) but are often found in the sea off the coast of Israel (Wachsmann 1998, 265-70, 272-73, 285-86; Galili, et al. 1994); in the walls of temples and tombs at Ugarit and its port at Minet-el Beida (Schaeffer 1978; Frost 1969a, 1991) and Byblos (Frost 1969b, Wachsmann 1998, 271-73) on the Syro-Palestinian coast, and at Kition on Cyprus (Karageorghis 1976a, 875-78; 1976b, 60, 69, 72, 78, 169; Frost 1985; Wachsmann 1998, 273-74). Such anchors seem to have been manufactured at Tell Abu Hawam and Tell Nami (Bass 1991, 74). A similar type of anchor is also known from the Late Bronze Age shipwreck at Cape Gelidonya (Pulak and Rogers 1994, 20-21; Wachsmann 1998, 283, 285).
In addition to the cargo and anchors, a unique aspect of this Late Bronze Age shipwreck is the preserved parts of its hull. Due to the steep, rocky nature of the seabed at Uluburun, somewhat reminiscent of that at Point Iria, what little of the hull that was preserved is extremely scanty and fragmented. Nonetheless, these remains may be easily grouped into three distinct sections, with a fourth consisting of only a few completely disarticulated and damaged bits of planking. The largest section of hull was located in the only relatively flat and sandy area of the site near the eight stone anchors high up on the slope between the first and second rows of copper ingots, mentioned above. This section owed its preservation, in part, to the weight of the stone anchors, which forced the keel and planking into the sand that had accumulated on a relatively flat ledge. The second and third hull sections, on the other hand, were preserved under the second and third rows of copper ingots, respectively. Although much of the planking from these last two sections is stained green and severely distorted from the immense weight of the copper oxhide ingots that had been placed on them, they nevertheless were preserved by these ingots. The toxic environment created by the copper ingots evidently discouraged marine borers and other organisms detrimental for the preservation of hull timbers. Results of wood species analyses revealed that the hull was built of cedar (Cedrus sp.), (#4) instead of fir (Abies sp.), as previously identified and published (Bass 1989, 25). This new identification is not at all surprising when one considers that Bronze Age references often mention cedar as the timber most preferred for building ships. The physical and mechanical attributes of cedar is well suited for shipbuilding: shrinkage is minimal and seasons without significant distortion, easily worked and has little dulling effect on tools, and is more resistant to decay in saltwater than most other woods.
In
addition to the hull itself, several fragmented oar or sweep blades and remains
of bulwark fencing have also been tentatively identified. A row of five well-rounded
stakes, the only nearly fully preserved example of which is 1.7 meters long
and nearly 7 centimeters in diameter, was excavated several meters to the
north (starboard side) of the hull remains. One end of all five stakes had
been sharpened to a point with four or five strokes of an axe or adze; the
sharpened points of all stakes faced the keel. Lying on and somewhat perpendicular
to the stakes were closely-spaced parallel withies. There is no surviving
evidence to suggest that the withies were in fact fastened to the stakes,
but this assemblage almost certainly represents the type of wickerwork weather
fencing visible in all Syrian ships depicted in nearly contemporary Egyptian
tomb paintings (Davies and Faulkner 1947, pl. 8; Davies 1963, pl. 15; Säve-Söderbergh
1957, pl. 23; Basch 1978, 102, figs. 4-6; 1987, 63-65, with figs.), and it
reminds us also of the wicker fencing assembled by Odysseus to keep the waves
out of the boat he built to leave Calypso’s island (Odyssey 5.256-57).
Since 1984, when the first hull section was exposed during the initial excavation campaign, we had known that the ship’s planking was assembled in the “shell-based” mortise-and-tenon joinery similar to that found on Greek and Roman ships of more than a millennium later. In contrast to the present-day “skeleton-based” construction technique, whereby a ship’s planking is shaped around and fastened to a pre-erected skeleton of the vessel, the ancients used the “shell-based” method that employed edge-joining the planks with mortise-and-tenon joints, which were then locked in place with wooden pegs driven through the tenons. Mortise-and-tenon joints had, of course, been used previously in Bronze Age ships in Egypt, as in the construction of the Khufu’s boat at Giza (ca. 2600 B. C.) and Senwosret III’s boats (ca. 1850 B. C.) at Dashur (Lipke 1984, 64; Steffy 1994, 25-27, 32-36, Patch and Haldane 1990). These early Egyptian examples of mortise-and-tenons, however, were freestanding and not pegged to lock adjacent strakes to one another. Rather, their primary function was to align the planks during construction, which were then fastened to each other with ligatures. This tradition of shipbuilding appears to have persisted at least as late as the 5th century B. C. when Herodotus observed nearly identical construction methods still in use in Egypt. In his oft-cited quotation, Herodotus noted that short planks were joined to each other with long, close-set tenons, which were then bound in the seams from within with papyrus fibers (Haldane & Shelmerdine 1990). There is no mention of locking the close-set tenons with pegs. The Egyptians were, however, fully aware of pegged mortise-and-tenon joints at last since the Old Kingdom (Dynasty III: ca. 2700-2600 B. C.) and used them in woodwork requiring this type of fastening (Lucas & Harris 1962, 451), but, as far as we can determine, they did not resort to their use in shipbuilding, unless they restricted their use to seagoing ships only, for which we have surviving examples.
When and where exactly pegged mortise-and-tenon joints were first
used in building seagoing ships is not known. Earliest documented use of
pegged mortise-and-tenon joints in the Near East, on the other hand, occurs
much later than its use in Egypt, and is seen in a wooden table from a Middle
Bronze IIB (ca. 1800/1750-1650 B. C.) tomb at Jericho (Ricketts 1960, 530,
fig 229.1). Wachsmann (1998, 239-41) suggests that pegging mortise-and-tenon
joints in shipbuilding developed on the Levantine littoral and spread westward
from there, and that this type joint was termed “Phoenician joints” by Romans
may have been more than mere coincidence (Sleeswyk 1980; Basch 1981). It
is of interest, therefore, that the earliest archaeologically documented use
in shipbuilding occurs on the Uluburun ship, which almost certainly was built
somewhere along the Syro-Palestinian coast or on Cyprus. Moreover, re-examination
of published hull wood drawings (Bass 1967, 50-51) and unpublished photographs
from the Cape Gelidonya shipwreck (ca. 1200 B. C.) indicates that pegged mortise-and-tenon
joints were also used in its construction. More will be said on this below.
Before the Uluburun ship, the earliest conclusive use of the shell-based method of shipbuilding using locked mortise-and-tenon joinery was revealed through the excavation and detailed study of a late-fourth century B. C. merchantman found near Kyrenia, Cyprus (Steffy 1985; 1994, 42-59). While the Kyrenia ship still remains the only well-documented and published example of a ship built by this ancient method, recent discoveries have shown that the method goes back by at least several centuries. At present, the earliest archaeologically attested use of locked mortise-and-tenon joinery in post-Bronze Age ships is seen in the 7th-century B. C. shipwreck at Mazarron, Spain (Negueruela et al. 1995, 195-96), but this site is still under investigation and preliminary reports do not allow for full evaluation of the hull’s constructional details. More complete documentation exists for the late 6th-century B. C. Jules Verne 7 ship at Marseilles (Pomey 1995, 475-80), and the late 5th-century B. C. Ma’agan Michael shipwreck near Haifa (Steffy 1994, 40-42). Although archaeological evidence for it is extremely limited, it seems highly likely that the late 5th- or early 4th-century Porticello shipwreck was also built by the same method (Lawall 1998; Eismann and Ridgway 1987, 10-13). Although nothing of the ship’s hull has survived, that pegged mortice-and-tenon joints were used in its construction is strongly suggested by the discovery of several loose tenons and a sliver of wood, presumably representing a plank fragment. Of these five shipwrecks, three also employ lashing or lacing (lashing) in the fastening of some timbers, specifically, in frames to strakes in the case of the Mazarron ship, and the ends of the lower strakes to stems and sternposts on the Jules Verne 7 and Ma’agan Michael ships.
| Please
send any comments or suggestions for the project or website to
Co-ordinator@ancientcyprus.ac.uk |
 |
The Ancient Cyprus Web Project is affiliated to the Council for British Research in the Levant |
Technical note: this site is best viewed using Internet Explorer. Visitors
using other browsers, for example Netscape, may experience some difficulties
on certain pages of this site. We are working to address these difficulties.
|
|
|||
