Shipwrecks are found in the strangest places.  In the Netherlands they are found under cabbage leaves like proverbial children, in farmland reclaimed from the sea.  In Kansas, paddle-wheeled steamboats appear in farmers' fields where rivers once ran, and in New York Dutch and British colonial-period ships lie under skyscrapers built on landfill.  So why not a shipwreck in an Indian coconut grove?  In 1990, farmers planting trees at Kadakkarapally, near Chertala, Kerala, found such a vessel (figure 1).  After a brief investigation, they reburied the craft and left it to sit.  In 2002 the villagers notified the archaeologists of the Kerala State Institute of Archaeology, Art History, Conservation, and Museology (SIAACM), and thus began the unveiling of a boat type no one had suspected existed.

In May 2003, at the invitation of Dr. M.V. Nair, then head of SIAACM, and funded through a grant from RPM Nautical Foundation, I flew from Texas to India across twelve time zones, which is about as far as one can go on this planet.  At that distance day became night, night became day, and dazed and jet-lagged I found myself peering into a muddy pit surrounded by coconut trees.  As the sun dappled through the palm fronds overhead, I sipped milk from a freshly-picked coconut and stared at a boat dated to almost one thousand years ago.  “Well,” someone asked, “what do you think?”

Over the next several days I walked on the boat’s timbers still solid after centuries in the mud, waded through black trench water that stained my toenails dark, scraped from my feet mud so thick and sticky it seemed impervious to water, and I pondered the nature of India’s first reported ancient hull. Curiosity brought out honeymooning Britons, a daily parade of nattily-dressed men and sari-clad women, and the press showed up too festooned with microphones and cameras.  All wanted to know, who built her?  What was she used for, and what was the boat doing in a coconut grove?

 Archaeological investigations of shipwrecks in the Indian Ocean and its tributary seas of the Persian Gulf and Red Sea are relatively rare.  INA's own projects involved a 17^th century wreck at Sadana Island in the Red Sea, a 5^th or 6^th century wreck at Black Assarca Island, Eritrea, and a survey in Bahrain where a shipwreck of undetermined date was found.  Other groups have excavated a 9^th century wreck in Indonesia, a Dutch East Indiaman in Sri Lanka, and of course, a number of wrecks in Australia.   Many of the areas around this vast ocean remain unexplored by nautical archaeologists.  Indian waters are themselves virtually virgin territory. 

A country with a coastline as large as India's must have had a long and busy seafaring tradition.  Little is known about it, however, as the archaeology of India has largely concentrated on terrestrial sites, and Indian seafaring lies outside of the main concerns of nautical archaeology, which tends to focus on the Mediterranean and Atlantic roots of modern global maritime endeavors.  For the scholar of Indian Ocean ships and seafaring, the best information is found in ethnographic studies.  These mostly concern the sewn boat, a type once ubiquitous in the Indian Ocean before the arrival of the Portuguese at the end of the fifteenth century (figure 2). 

European merchants and colonizers supplanted traditional boat construction methods, and by the beginning of the current century, the indigenous types such as the sewn boat survive only in a few places.  One of these areas is the coast of southwestern India.  It would be expected then that a vessel predating the European colonization would be of the traditional sewn type.  That expectation is wrong.  The vessel lying in the pit at the village of Kadakkarapally is unique, unlike anything we expected.  It is built in a method lost and forgotten for centuries and even contains a feature not seen outside the time and land of the Pharaohs.  

THE HULL

Investigating the shipwreck was a race against time.  The yearly monsoon threatened to start any day- - when it came, all work would yield to the torrential rains.  Pre-monsoon heat battered India, but our site, just a mile inland, was kept cool by the sea breezes.  Inland, hundreds of people died from the unrelenting temperatures.  It was hard to imagine such events in our peaceful coconut grove where exotic birds flew overhead and children played among the palms. 

Archaeologist Dr. V. Selvakumar and geoarchaeologist Dr. Paul Shajan of the Centre for Heritage Studies (CHS) directed the excavation of the wreck.  As they handled the nitty-gritty details of the hull recording, I concentrated on the vessel's shape and how it was put together.  This gave me the time to pause over curious features and think about how and why things were done.  With over 18 meters of remaining hull length, there was plenty to ponder.

Radiocarbon dating done in the United States yielded a date of AD 1020 to 1270.  This corroborates the C-14 date derived by a lab in India that indicated a date of AD 920 to 1160.  The dates only point to when the tree was cut down, not when the craft was built or when she was abandoned.  Typically, timber is used within a few years of its cutting, but on occasion may lie unused for longer.  As there were no associated artifacts on the wreck that might have been useful for comparative dating, the overlap in the radiocarbon dates, that is 1020 to 1160, should be regarded as the likely date of the vessel’s construction.

Bottom Planking

Thick planks comprise the bottom, which was originally flat forward and aft and side to side.  Lying in two layers, the inner one is readily observable while only a small section of the outer layer can be seen in the port stern quarter.  Here, the curious villagers stripped the inner layer away and broke through the outer layer upon the wreck’s discovery. 

The planks are carved.  A number of cleats are carved out or cut into them in rows reaching from side to side.  The planking is smooth and exhibits only a few tool marks around some cleats.  There is no edge joining-- the only fastenings holding the planking together are cut iron nails hammered through the outer layer into the inner.  The nails are randomly placed, without regard to the location of the inner planking seams as some nails were hammered into them. 

A gray substance fills the planking seams, which are tight and expertly made, and it is smeared between the two layers of planks.    The substance was used as a sealant, or perhaps an adhesive, and may be a mixture of oil, lime, and possibly tree sap- a traditional sealant on Indian boats known as "chunnam."

The inner planking is a mixture of lengths and widths.  Some planks are simply “stealers”- narrow, tapering ones inserted into spaces between larger planks to fill a gap.  There is a central plank, but as it is no thicker than the other bottom planks, it does not serve as a backbone.  This plank has a channel gouged down the centerline to facilitate drainage through limber holes bored through the floors (figure 3 at arrow).  There are at least two repair patches in the inner layer, both expertly fitted.  The repairs indicate the vessel was in use for a while before its abandonment or wrecking.

Each plank varies in width over its length.  These varying widths give the strakes an “interlocking” appearance.  Widths vary suddenly and odd ends and corners protrude into abutting or adjoining planks.  This pattern continues into the bow and stern areas where shorter planks are used (figure 4). The interlocking pattern adds longitudinal strength to the hull and keeps the planks from slipping against each other, which is particularly important in the absence of edge joining. 

On the sides of the boat there is a third layer of planking.  A small number of square-sectioned treenails, combined with iron nails, fasten the outer layers to the inner. The third layer of planking may be a “rubbing strake” protecting the planking from the wear and tear of rubbing against wharves and banks.  As such, the third layer would have been considered somewhat temporary and easily replaceable.

Chine Strakes

Perhaps the most striking feature of the boat are the chine strakes.  These two massive timbers once both reached over 14.5 meters long, although now only the starboard side remains to that length.

Figure 5.

The strakes are carved in a slightly open “L”-shape, at least ten centimeters at its thickest in cross-section (figure 5).  Spaced at regular intervals along the inside of the strake are carved blocks to receive the ends of the floors, which fit tightly and expertly.  The blocks reach to the upper edge of the chine strake essentially forming a carved framing member (figure 6).

Halfway between each frame-block are cleats and a crossbeam-block.  The lower cleat is in the chine itself, and the second is carved several centimeters above it. Each cleat has a hole carved through it.  Higher yet is a carved block for the support of a crossbeam. The crossbeam-block is notched dovetail fashion on its upper surface to accept the beam end (figure 7).  The notch does not penetrate the outer surface of the chine strake.

Between the frame block and crossbeam-block/cleat array are two, but sometimes one, iron fastenings set one above the other.  These are now mostly corroded away leaving only holes and impressions, including that of a square rove on the inner surface, indicating that these were rivets.

Carved into the top of each frame-block is a mortise for the tenon of an upper futtock.  This timber, of which only two fragmentary examples survive, is pegged through the mortise-and-tenon joint with a square treenail.  This is its sole fastening to the lower hull member.  The upper strake is nailed to the futtock and it lies carvel fashion with the chine strake.  Only one small section this strake survived.

Floors

The floors are large, single-piece timbers that stretch from chine to chine (see figure 6).  Originally there were ten floors, as seen by the recesses cut into the planking in which they sat, spaced approximately 1.75 meters apart.  Of the ten, five remain intact, and three are completely missing.  Another transverse timber in the forward section of the craft lies between two floors and appears to exist solely for the purpose supporting an additional bulkhead.  This timber is smaller in cross-section than the floors and it does not sit in carved blocks on the chine strake as the floors do.

 On the upper surface of each surviving floor a groove runs from end to end for bulkhead panels.  Standing vertically, surviving panels sit in the groove with no additional fastening.  It is unclear how the panels were secured at top, and they do not seem to be fastened to each other.   The bulkheads served to divide the interior of the vessel into sections.

 Each floor is fastened to the bottom planking with two large rectangular treenails, one to either side of the centerline, with one hammered in on the forward face and the other one on the after face (figure 8).  Each floor is additionally fastened on either end to the chine strake with two large iron spikes, one forward and one aft of the bulkhead groove.  These are hammered down into the chine strakes.  An exception to this pattern is found in a floor in the forward part of the boat that is fastened with pairs of iron spikes across its width.  This may have reinforced the transition of the planking from the bottom to the bow.

In each recess, between bottom planking and the floors a gray substance was applied either as a sealant or an adhesive.  As each floor has a centrally located limber hole, as previously mentioned, making the compartments created by the floors and bulkheads not watertight, the latter function of the substance is more probable.     

Cleats & Lashing Channels

The craft is notable for the cleats and channels found throughout the interior of the hull (figure 9).  These are the most curious trait of its construction.  Such features are usually associated with ropes and lashings, and remains of cordage in some cleats indicate that this is the case with the Kadakkarapally Boat.  I believe the cleats are evidence of transverse lashing, a system of hull fastening not seen outside of ancient Egypt (third through mid-first millennia BC). 

Figure 9.  (Left) Lines of cut channels and cleats for ropes cross the bottom between each floor (at arrow).  (Right) Closeup of two cleats with the underlying channel cut through.  Remains of roping were found in some of these.

In each space between floors cleats, or more often recessed channels, are carved or cut into the planks in a line from side to side.  Some of these protrude above the surface of the planking, while others lie flush to the surface with the holes recessed into the planks.  The cleats and channels align with those carved on the chine strakes.  With transverse lashing, ropes would run down through the cleats on one chine strake, across the bottom weaving in and out of the cleats and channels, then up the cleats on the other chine strake.  Then the ropes would run parallel to and below the crossbeams, and the two ends would be tied together near the midline, creating tension.  This tightening would pull the chine strakes and planking together much like pulling the string of a drawstring bag, closing tightly all the seams and strengthening a hull that has little internal reinforcement and no edge-joining of the planking. 

This lashing pattern not only occurs between each floor, but also in the bow as seen in the rows of aligned cleats, and in the stern as indicated by the remains of cleats there.  Clearly, the lashing was a significant contributor to the integrity of the hull. 

The other possible explanation for the presence of the multiple cleats is a lashed-lug system such as is found on boats of the western Pacific.  This system used flexible ribs lashed to the cleats, or lugs, with ropework linking the ribs to upper crossbeams thereby compressing the hull.  This system, however, relies on a convex hull shape section.  On a flat-bottomed vessel such as that at Kadakkarapally, the lashed-lug system would distort the bottom planks by pulling them up out of alignment thus undermining its watertightness.  Also, the lack of any indication of flexible ribs, no rope wear on the cross-beams, and a wear pattern on the cleats inconsistent with the lashed-lug method leads me to rule this out. 

Conclusions

The construction of the Kadakkarapally boat is clearly not a European style.  While cleats do appear in the European archaeological record, such as on Britain’s Bronze-Age Brigg Raft and Ferriby Boats, as well as on Viking Ships, these also are found on watercraft of the western Pacific and thus cannot be used as an indicator of cultural origin.  The C-14 dates reinforce the non-European origin or design of the hull as, except for an intrepid few like Marco Polo, Europeans did not arrive in the region until the late fifteenth century, and only then did their own shipbuilding practices begin to displace indigenous Indian ones.

The construction is not Arabian.  Arabian boats traditionally were sewn in the same manner as is found in Kerala today on the Keralan fishing canoes.  This method has been used in western Asia for at least three thousand years and is well documented as being the primarily, if not sole, construction method of Arabian ships and boats.  Likewise, the construction is probably not East African.  Our knowledge of the watercraft of East Africa extends back two thousand years and by all accounts, these were also sewn.

I believe the boat is most likely local, or of southern Indian origin as:

·        The predominant wood type, Anjili, is native to the area of Kerala.  This tends to indicate a local origin of the vessel, as opposed to having found a wood type that was clearly foreign.  Yet coincidence cannot be discounted, and research is needed on the range of the Anjili tree in Asia. 

·        Although now inland, Kadakkarapally was once seafront and toredo damage to some upper timbers attest to an exposure to seawater for a period, probably after abandonment.  There did not appear to be any wormholes in lower timbers, however, indicating that the boat was not a sea-going vessel.

·        The flat bottom, the absence of a keel or other backbone, and the hard chine argue for a local origin.  The box-like shape makes for poor, even dangerous, sailing in all but calm seas.  The flat bottom offers virtually no lateral resistance to the wind, making sailing in anything but a wind from the stern quarters laborious, as the boat would tend to slide sideways.  Rolling waves going under the flat bottom would also put strong stresses on the hull threatening its integrity.  The sharp chine, the transition from bottom to side, is similarly detrimental, as the seas would tend to tear at such a corner.  This craft was best sailed on bays, large rivers, and perhaps on coastal runs, but only in fair weather and calm seas.  Thus, the vessel is well suited to sailing on Lake Vembanad and the large estuary separating the Alappuzha peninsula from the mainland and whose opening to the sea is at Cochin.

Afterword

In the end, we beat the monsoon.  The rains began the week after the team finished the season's recording.  It would rain steadily for the next eight weeks.  The pit in the coconut grove remained open, filled with black trench water.  The Kadakkarapally Boat awaits further study, and possible raising for conservation and display.  It is not a historic craft-- it did not take part in any great battle.  Nor is it a grandiose vessel for important personages— it was a simple cargo-carrying sailing craft.  The boat is important nevertheless.  Aside from being India's first known ancient hull, the boat represents a technology of woodworking and seafaring probably once commonplace to southern India, but replaced either by the sewn method or by the European plank-on-frame system brought by European colonizers.  As such, the boat at Kadakkarapally represents a part of Indian history submerged by the tidal wave of cultural change.

Acknowledgements

I thank Dr. M.V. Nair and the State Institute of Archaeology, Art History, Conservation, and Museology and the personnel of Centre for Heritage Studies for the opportunity to participate on this project.   Special thanks to Dr. P.K. Gopi, Dr. V. Selvakumar, Dr. Paul Shajan, and the staff of CHS for their hospitality and friendship; Mr. Rajagopal Kamath, who was instrumental in bringing the Kadakkarapally Boat to the attention of the archaeological community outside India; my field assistants Mr. V. Valsan and Mr. Provane; Father V. P. Joseph of the local History and Heritage Protection Council; and the many others who have taken an interest in the Kadakkarapally Boat.  I am particularly grateful to the RPM Nautical Foundation for funding my participation in this project., and to the  Institute of Nautical Archaeology.

Citation information:

Ralph K. Pedersen.  "The Kadakkarapally Boat: A Thousand-Year-Old Shipwreck in Kerala, India, 2007."  www.wedigboats.org

The text on this page is adapted from the author's publication: "The Shipwreck in the Coconut Grove: The Kadakarapally Boat," in The INA Quarterly 31.2 (2004): 3-9.

Page edited: December 2007

Author retains all copyrights to text and images.  For reproductions, see www.wedigboats.org