Return to Long Serpent
"Right away the mast was rigged with its sea-shawl; sail-ropes were tightened, timbers drummed and stiff winds kept the wave-crosser skimming ahead; as she heaved forward, her foamy neck was fleet and buoyant, a lapped-prow loping over currents, until finally the Geats caught sight of coastline and familiar cliffs. The keel reared up, wind lifted it home, it hit on the land."
Beowulf (lines 1905-1910), describing the hero's return from Denmark to southern Sweden
The finest and best preserved Viking longship is the Gokstad, which was built about AD 900 and excavated almost a thousand years later from the eponymous farm on the Sandefjord south of Oslo. Like all such ships, the keel was laid first and secured to the stem posts fore and aft by way of curved transition pieces (Old Norse lot). The first plank (the garboard strake) was bored with an auger and joined to the keel (which was shaped to fit) by iron rivets clinched to a rove (washer) in what is called clinker construction. The next strake or plank was secured in the same way, each overlapping one another and fastened by rivets. Marking the transition from the bottom of the ship to the waterline, a heavier strake (ON meginhufr) ran along each side. Notched floor timbers then were lashed to cleats cut out of the strakes and fitted to the sides by treenails (wooden pegs). Above these ribs, crossbeams (ON bitis) braced by curved knees provided lateral support for the upper strakes. A deck of loose pine boards, which could be removed to access cargo or bale out the bilge, rested on top. Without additional beams above the deck to serve as thwarts, the crew would have sat on their sea chests when rowing. A heavy keelson secured to the ribs supported a pine mast and yard from which hung a woolen sail. The ship was made waterproof by a caulking of twisted lamb wool or cow hair saturated with pine tar and fitted in a groove between the clinched boards. Such construction permitted an exceptionally strong and flexible ship, one that was seaworthy enough to be propelled by a large sail on the open ocean but with a draft shallow enough to allow it to be shelved on a beach or navigated by oar far upstream.
Measuring seventeen feet at the beam, the Gokstad was a wide and stable ship, made more so by riding low in the water. It measures only about six feet from the keel to the gunwale (the uppermost edge of the ship), with a freeboard (from the waterline to the gunwale) only half that distance—and the oarports half that distance again. The deck, in other words, was virtually at the waterline and the oarports themselves only a foot and a half above it. And yet, with sixteen strakes on a side, the Gokstad was twice as high as the Ladby ship, which was buried about the same time. Indeed, that ship has such a shallow freeboard that an extra eighth strake is presumed to have existed to allow sufficient room for the rowers, who also lacked thwarts on which to sit.
The Gokstad ship has sixteen oars to a side, with the oarports (which could be covered in bad weather by a disk of wood that pivoted into place) cut at the third strake (on the Oseberg ship, they are positioned at the first). This permitted a higher freeboard and so offered a height advantage over the enemy, certainly more so than if the oars had been secured by oar locks on the gunwale, as they would be on a smaller boat. The remnants of thirty-two overlapping shields, alternately painted yellow and black, were fixed to each side (which implies that the crew was doubled, one resting while the others rowed). Securely tied, they hung from a batten on the uppermost or sheer strake (and not slotted behind a rail on the outside of the ship, as with Skuldelev 5).
The Gokstad ship had served as a burial chamber for a local chieftain. Among the grave goods, there were oars and spars, tubs and kegs for food and water, kitchen utensils, wooden furniture (including six beds and a sledge decorated with brass nails), a gaming board and horn pieces, intricately wrought bronze fittings for a belt, and even remnants of a woolen sail cloth sewn with red stripes. (Sails often was interwoven to give a checkered or striped pattern, e.g., The Saga of Olaf Tryggvason (CI), St. Olaf's Saga (CXXIII), or the striped sails depicted in the Bayeux Tapestry). Scattered around the hull were the skeletons of a dozen horses and eight dogs, as well as two goshawks and two peacocks, which must have seemed exotic indeed. On board, there also were three small boats of different sizes, all of which had been deliberately broken up. The absence of any jewelry, gold or silver, or weapons argue that these more precious artifacts had been robbed fifty to a hundred years later during the rule, and possible at the instigation, of Harold Bluetooth, who as the Danish king may have sought to delegitimize the authority and legacy of earlier ruling dynasties.
Certainly, the few skeletal remains (only four leg bones, an upper arm bone and shoulder blade, and fragments of the skull) suggest that the barrow was deliberately desecrated. When it was reopened in 2007 and the skeleton re-examined, it was found to be that of a tall, powerfully-built warrior forty to fifty years old who, from the cutting strikes to both his legs (unprotected by shield or chain mail), likely died violently in battle. Originally, they were identified as belonging to Olaf Geirstad-Alf (Elf of Geirstad), a chieftain of the Ynglings and the older half-brother of Halfdan the Black. But Olaf died half a century before the burial mound was erected and the man's identity is, in fact, not known.
Preserved under a barrow, only the tall prow and stern posts, a portion of the mast, and upper two strakes that had protruded above the impermeable layer of anaerobic blue clay were lost to rot. The Gokstad ship, which was excavated from its mound in two large pieces, was so well preserved in fact that initially it was exhibited without any restoration. This picture was taken in 1880, one of only two that record the excavation (the other from the stern). Behind the mast partner that supported the mast, the triangular frame of the burial chamber is readily visible.
This photograph is from the frontispiece of A Short Guide for the use of visitors to the Viking-ship from Gokstad (1898) when the ship still was kept (as it had been since its discovery almost twenty years earlier) in the garden of the University of Christiana. (Oslo then still was named after the Danish King Christian IV, who had rebuilt the city after a disastrous fire in the seventeenth century, and would not regain its former name until 1925). The following year, the new Viking Ship Museum opened and the Oseberg ship was put on display. In 1932, the hall for the Gokstad ship was completed and, after nearly fifty years in its temporary shed, it, too, was exhibited, having been extensively restored over the previous three years, its parts disassemble, steamed, and bent back into shape. What could not be restored or was missing (such as the upper two strakes) was replaced.
In this picture, taken from the port side of the ship, the steering oar is almost hidden by the hull. But four overlapping shields can be seen hanging from the gunwale, effectively covering the oarports. In combat, the mounted shields would have provided extra protection and, in a heavy sea, some shelter from the waves. But there also was a risk of their being washed overboard, and Brøgger dismisses the notion that they ever did hang from the shield rack except when the ship was in port.
In April, 1893, the Viking, an exact replica of the Gokstad ship (save for the decorations on the bow and stern posts and a protective tent amidships, for which there is no archaeological evidence), sailed with a crew of only twelve men across the Atlantic from Bergen, Norway to Newfoundland. Averaging ten or eleven knots, it made landfall after a stormy crossing in only twenty-eight days (one period so rough that the sail had to be reefed) and then went on to New York City. After a journey along the Hudson river, through a very narrow Erie Canal, and onto the Great Lakes, the ship arrived at the World's Columbian Exhibition in Chicago, which was commemorating the 400th anniversary of Columbus's voyage to America—a reminder (perhaps insensitive, given the theme) that Leif Erikson had reached North American half a millennium earlier. (He had been sent to Greenland by Olaf Tryggvason to preach Christianity there and "on the same journey he discovered Wineland the Good," The Saga of Olaf Tryggvason, XCVI). The ship then went down the Mississippi to New Orleans and returned to Chicago the next year, where it remained moored in a lagoon at Jackson Park until 1920, when it was moved to Lincoln Park. There, the deteriorating ship continued to languish until 1994, when it was transported to West Chicago and stored. Two years later, the newly christened Raven finally was moved to Good Templar Park in Geneva, just outside of Chicago, where it still is in need of better accommodation.
The Gokstad is about seventy-six feet long, its keel appreciatively longer than either the Oseberg ship or Skuldelev 5. Crumlin-Pedersen has calculated the cubic feet of oak that would have been required to build such a ship. Taking the average of his figures, one seventy-four feet long, the approximate size of the Gokstad, would have required 1,907 cubic feet of wood or 22,884 board feet, i.e., a board one foot square and one inch thick. A mature oak several hundred years old with a straight, knot-free trunk three feet in diameter and eighteen feet long yields about 1,526 board feet. Fifteen such trees, therefore, would have been required just for the planking—which, given the length of the hull, had to be constructed from more than one piece of wood and joined by scarf joints riveted together. Allowing an average of four planks for each strake, approximately 128 would have been required to construct the ship or, if only smaller trees were available and a single log was split lengthwise and each half hewed to a plank, half that number of trees. The keel, itself, is fifty-six feet long and shaped from a single piece of oak. It would have required a straight tree even longer than that, of which there cannot have been many in the oak forests of southern Norway—or even that many forests, it would seem, when one reads in the Anglo-Saxon Chronicle that 350 ships were rowed up the Thames and London attacked in AD 851. The replica launched in 1893 required, in fact, that the wood for the keel be imported from Canada. By the nineteenth century, there simply were no longer any mature trees of that size in Norway.
Scans conducted by the Museum of Cultural History in 2019 revealed that, during the previous five years, the ship has been slowly collapsing under its own weight. The bow and stern were shifting toward the ship's center, the strakes at midship bowing outward and the underside sagging on its supports. The beam on which the keel rests and vertical iron supports were not distributing the weight of the ship evenly which, together with fluctuations in temperature and humidity, as well as floor vibrations, threaten its structural stability. Indeed, cracks were beginning to appear in the ancient wood. Eventually, the Gokstad and Oseberg ships will be relocated to a new Museum of the Viking Age, which is scheduled to open in 2026, in celebration of the centenary of the old museum. In preparation for the move, the Gokstad ship was raised one millimeter off its supports and found to weigh 9,301.3 pounds.
Although the Gokstad has been characterized as a longship (Danish langskip), properly it is not. Rather, ships from the tenth and eleventh centuries that have a length of fifty feet or more, and a length at least six times their width were true longships. Aside from the dragon ships of the sagas, those from the Viking Age include the recovered Ladby ship, Skuldelev 2 (a smallest type, with twenty-six oars) and Skuldelev 5, Rosekilde 6 (the longest), and Hedeby 1, among a few others. Although the Gokstad ship does have thirty-two oars and certainly is long enough, its beam is much wider than that of a true longship (a ratio of length to breath of 4.7).
In 2018, using ground-penetrating radar, archaeologists from the Norwegian Institute for Cultural Heritage Research discovered a Viking ship buried in a field at Gjellestad, about fifty miles southeast of Oslo. The burial mound itself had been plowed away over the centuries, leaving the lower part of the ship less than two feet below the surface. Initially, there were no plans for excavation, but a nearby drainage ditch had made the ground soggy and damp, prompting the growth of fungus that threatened to further rot the oak hull. Mold also was introduced when initial work was done to determine the condition and age of the ship (c. AD 733). Recovery began in 2020 and the keel and lower-most ribs uncovered a year later. Although in poor condition, the Gjellestad Ship is the first modern excavation in more than a century, since the Oseberg ship was discovered in 1904. Prior to that there had been only two others, the Tune ship in 1867 and the Gokstad ship in 1880.
The massive rudder or steering oar affixed to the right side of a ship's stern provides the origin of the word "starboard," from the Old English stéor (steer) and bord (board). Inexplicably, in the popular Vikings television series, some of the ships recreated for the show have their steering oar on the port (left) side. On Leif's and Cnut's ships, for example, it is placed correctly—but not on Olaf's. (It was Olaf Haraldsson, in fact, who proposed that London Bridge be pulled down—and not Leif, as portrayed in the series.)
The left side of the ship was the larboard, from the Middle English lade ("load"), as that was the side from which the ship was loaded (so as to protect the protruding steering oar). Given the potential confusion in the sound of the two words, larboard was replaced by "port" in the nineteenth century.
In the series, too, the king's name is spelled "Canute." In Old English texts (such as the Anglo-Saxon Chronicle) and lives written in Latin (the Encomium Emma Reginae and Vita Ædwardi Regis), the spelling is monosyllabic. In Latin texts by French or Norman authors (such as William of Jumièges and William of Poitiers), it is disyllabic. In a phonological process called epenthesis, the spelling shifted from Cnut to Canute when these texts were translated into English. An intervening vowel was inserted to separate the initial cluster of consonants in the name and so make it easier to pronounce.
References: "Dendrochronological Dating of the Viking Age Ship Burials at Oseberg, Gokstad and Tune, Norway" (1993) by Niels Bonde and Arne Emil Christensen, Antiquity, 67, 575-583; "Skjelettet fra Gokstadskipet ny vurdering av et gammelt funn" (2008) by Per Holck, Michael Quarterly, 5(4), 292-304 (in Norwegian, English abstract); "The Skeleton from the Gokstad Ship: New Evaluation of an Old Find" (2009) by Per Holck, Norwegian Archaeological Review, 42(1), 40-49; "Revisiting the Gokstad" (2014) by Jason Urbanus, Archaeology, 67(4), 34-38; "Viking Collection Deteriorating" (September 2, 2019), Museum of Cultural History (Oslo); "Gokstadskipet slår sprekker og må støttes opp for å unngå kollaps" (May 6, 2019) by Arnfinn Mauren. Aftenposten; Vikingeskibs Museet: The Gokstad Boat website; "Prehistoric Naval Architecture of the North of Europe: The Gokstad Ship" (1892) by George H. Boehmer, Annual Report of the Board of Regents of the Smithsonian Institution, pp. 618-628; Museum of Cultural History (Oslo) website; The Viking Ships: Their Ancestry and Evolution (1971) by A. W. Brøgger and Haakon Shetelig; Beowulf (2000) translated by Seamus Heaney; Snorri Sturluson, Heimskringla: History of the Kings of Norway (The Saga of Ólaf Tryggvason) (1964) translated by Lee M. Hollander; "'Viking', a Gokstad Ship Replica from 1893" (1986) by Arne Emil Christensen, in Sailing into the Past, edited by Ole Crumlin-Pedersen and Max Vinner; "The Gjellestad Ship" (2018/2020), Norwegian Institute for Cultural Heritage Research press releases; "How Cnut became Canute (and how Harthacnut became Airdeconut)" (2014) by Thijs Porck and Jodie Mann, NOWELE: North-Western European Language Evolution, 67, 237-243.
Return to Top of Page